train.py

"""Modified from https://github.com/huggingface/diffusers/blob/main/examples/text_to_image/train_text_to_image.py
"""
#!/usr/bin/env python
# coding=utf-8
# Copyright 2024 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 gc
import logging
import math
import os
import pickle
import shutil
import sys
from einops import rearrange
from omegaconf import OmegaConf
from packaging import version
from tqdm.auto import tqdm
import datasets

import numpy as np
import smplx
import torch
import torch.nn.functional as F
import torch.utils.checkpoint
from torch.utils.data import RandomSampler
from torch.utils.tensorboard import SummaryWriter
from torchvision import transforms
from pytorch3d.transforms import matrix_to_rotation_6d, rotation_6d_to_matrix

import accelerate
import diffusers
import transformers
from accelerate import Accelerator
from accelerate.logging import get_logger
from accelerate.state import AcceleratorState
from accelerate.utils import ProjectConfiguration, set_seed
from diffusers import FlowMatchEulerDiscreteScheduler
from diffusers.optimization import get_scheduler
from diffusers.training_utils import (
    EMAModel,
    compute_density_for_timestep_sampling,
    compute_loss_weighting_for_sd3
)
from diffusers.utils import check_min_version, deprecate, is_wandb_available
from diffusers.utils.torch_utils import is_compiled_module
from transformers import AutoTokenizer
from transformers.utils import ContextManagers

current_file_path = os.path.abspath(__file__)
project_roots = [os.path.dirname(current_file_path), os.path.dirname(os.path.dirname(current_file_path)), os.path.dirname(os.path.dirname(os.path.dirname(current_file_path)))]
for project_root in project_roots:
    sys.path.insert(0, project_root) if project_root not in sys.path else None

from comovi.data.bucket_sampler import ComoviSampler, RandomSampler
from comovi.data.dataset_image_video import ComoviDataset, ImageVideoSampler, get_random_mask
from comovi.models import (
    AutoencoderKLWan,
    AutoencoderKLWan3_8,
    WanT5EncoderModel,
    Wan2_2Transformer3DModel,
    ComoviTransformer3DModel
)
from comovi.pipeline import WanPipeline, WanI2VPipeline
from comovi.utils.discrete_sampler import DiscreteSampling
from comovi.utils.utils import get_image_to_video_latent, save_videos_grid

if is_wandb_available():
    import wandb

def filter_kwargs(cls, kwargs):
    import inspect
    sig = inspect.signature(cls.__init__)
    valid_params = set(sig.parameters.keys()) - {'self', 'cls'}
    filtered_kwargs = {k: v for k, v in kwargs.items() if k in valid_params}
    return filtered_kwargs

def get_random_downsample_ratio(sample_size, image_ratio=[],
                                all_choices=False, rng=None):
    def _create_special_list(length):
        if length == 1:
            return [1.0]
        if length >= 2:
            first_element = 0.75
            remaining_sum = 1.0 - first_element
            other_elements_value = remaining_sum / (length - 1)
            special_list = [first_element] + [other_elements_value] * (length - 1)
            return special_list
            
    if sample_size >= 1536:
        number_list = [1, 1.25, 1.5, 2, 2.5, 3] + image_ratio 
    elif sample_size >= 1024:
        number_list = [1, 1.25, 1.5, 2] + image_ratio
    elif sample_size >= 768:
        number_list = [1, 1.25, 1.5] + image_ratio
    elif sample_size >= 512:
        number_list = [1] + image_ratio
    else:
        number_list = [1]

    if all_choices:
        return number_list

    number_list_prob = np.array(_create_special_list(len(number_list)))
    if rng is None:
        return np.random.choice(number_list, p = number_list_prob)
    else:
        return rng.choice(number_list, p = number_list_prob)

def resize_mask(mask, latent, process_first_frame_only=True):
    latent_size = latent.size()
    batch_size, channels, num_frames, height, width = mask.shape

    if process_first_frame_only:
        target_size = list(latent_size[2:])
        target_size[0] = 1
        first_frame_resized = F.interpolate(
            mask[:, :, 0:1, :, :],
            size=target_size,
            mode='trilinear',
            align_corners=False
        )
        
        target_size = list(latent_size[2:])
        target_size[0] = target_size[0] - 1
        if target_size[0] != 0:
            remaining_frames_resized = F.interpolate(
                mask[:, :, 1:, :, :],
                size=target_size,
                mode='trilinear',
                align_corners=False
            )
            resized_mask = torch.cat([first_frame_resized, remaining_frames_resized], dim=2)
        else:
            resized_mask = first_frame_resized
    else:
        target_size = list(latent_size[2:])
        resized_mask = F.interpolate(
            mask,
            size=target_size,
            mode='trilinear',
            align_corners=False
        )
    return resized_mask

# 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__, log_level="INFO")

def log_validation(vae, text_encoder, tokenizer, transformer3d, args, config, accelerator, weight_dtype, global_step):
    try:
        logger.info("Running validation... ")
            
        transformer3d_val = Wan2_2Transformer3DModel.from_pretrained(
            os.path.join(args.pretrained_model_name_or_path, config['transformer_additional_kwargs'].get('transformer_subpath', 'transformer')),
            transformer_additional_kwargs=OmegaConf.to_container(config['transformer_additional_kwargs']),
        ).to(weight_dtype)
        transformer3d_val.load_state_dict(accelerator.unwrap_model(transformer3d).state_dict())
        scheduler = FlowMatchEulerDiscreteScheduler(
            **filter_kwargs(FlowMatchEulerDiscreteScheduler, OmegaConf.to_container(config['scheduler_kwargs']))
        )

        if args.train_mode != "normal":
            pipeline = WanI2VPipeline(
                vae=accelerator.unwrap_model(vae).to(weight_dtype), 
                text_encoder=accelerator.unwrap_model(text_encoder),
                tokenizer=tokenizer,
                transformer=transformer3d_val,
                scheduler=scheduler,
            )
        else:
            pipeline = WanPipeline(
                vae=accelerator.unwrap_model(vae).to(weight_dtype), 
                text_encoder=accelerator.unwrap_model(text_encoder),
                tokenizer=tokenizer,
                transformer=transformer3d_val,
                scheduler=scheduler,
            )
        pipeline = pipeline.to(accelerator.device)

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

        images = []
        for i in range(len(args.validation_prompts)):
            with torch.no_grad():
                if args.train_mode != "normal":
                    with torch.autocast("cuda", dtype=weight_dtype):
                        video_length = int((args.video_sample_n_frames - 1) // vae.config.temporal_compression_ratio * vae.config.temporal_compression_ratio) + 1 if args.video_sample_n_frames != 1 else 1
                        input_video, input_video_mask, _ = get_image_to_video_latent(None, None, video_length=video_length, sample_size=[args.video_sample_size, args.video_sample_size])
                        sample = pipeline(
                            args.validation_prompts[i],
                            num_frames = video_length,
                            negative_prompt = "bad detailed",
                            height      = args.video_sample_size,
                            width       = args.video_sample_size,
                            guidance_scale = 6.0,
                            generator   = generator,

                            video        = input_video,
                            mask_video   = input_video_mask,
                        ).videos
                        os.makedirs(os.path.join(args.output_dir, "sample"), exist_ok=True)
                        save_videos_grid(sample, os.path.join(args.output_dir, f"sample/sample-{global_step}-{i}.gif"))

                        video_length = 1
                        input_video, input_video_mask, _ = get_image_to_video_latent(None, None, video_length=video_length, sample_size=[args.video_sample_size, args.video_sample_size])
                        sample = pipeline(
                            args.validation_prompts[i],
                            num_frames = video_length,
                            negative_prompt = "bad detailed",
                            height      = args.video_sample_size,
                            width       = args.video_sample_size,
                            guidance_scale = 6.0,
                            generator   = generator, 

                            video        = input_video,
                            mask_video   = input_video_mask,
                        ).videos
                        os.makedirs(os.path.join(args.output_dir, "sample"), exist_ok=True)
                        save_videos_grid(sample, os.path.join(args.output_dir, f"sample/sample-{global_step}-image-{i}.gif"))
                else:
                    with torch.autocast("cuda", dtype=weight_dtype):
                        sample = pipeline(
                            args.validation_prompts[i],
                            num_frames = args.video_sample_n_frames,
                            negative_prompt = "bad detailed",
                            height      = args.video_sample_size,
                            width       = args.video_sample_size,
                            generator   = generator
                        ).videos
                        os.makedirs(os.path.join(args.output_dir, "sample"), exist_ok=True)
                        save_videos_grid(sample, os.path.join(args.output_dir, f"sample/sample-{global_step}-{i}.gif"))

                        sample = pipeline(
                            args.validation_prompts[i], 
                            num_frames = 1,
                            negative_prompt = "bad detailed",
                            height      = args.video_sample_size,
                            width       = args.video_sample_size,
                            generator   = generator
                        ).videos
                        os.makedirs(os.path.join(args.output_dir, "sample"), exist_ok=True)
                        save_videos_grid(sample, os.path.join(args.output_dir, f"sample/sample-{global_step}-image-{i}.gif"))

        del pipeline
        del transformer3d_val
        gc.collect()
        torch.cuda.empty_cache()
        torch.cuda.ipc_collect()

        return images
    except Exception as e:
        gc.collect()
        torch.cuda.empty_cache()
        torch.cuda.ipc_collect()
        print(f"Eval error with info {e}")
        return None

def linear_decay(initial_value, final_value, total_steps, current_step):
    if current_step >= total_steps:
        return final_value
    current_step = max(0, current_step)
    step_size = (final_value - initial_value) / total_steps
    current_value = initial_value + step_size * current_step
    return current_value

def generate_timestep_with_lognorm(low, high, shape, device="cpu", generator=None):
    u = torch.normal(mean=0.0, std=1.0, size=shape, device=device, generator=generator)
    t = 1 / (1 + torch.exp(-u)) * (high - low) + low
    return torch.clip(t.to(torch.int32), low, high - 1)

def parse_args():
    parser = argparse.ArgumentParser(description="Simple example of a training script.")
    parser.add_argument(
        "--input_perturbation",
        type=float,
        default=0,
        help="The scale of input perturbation. Recommended 0.1."
    )
    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(
        "--revision",
        type=str,
        default=None,
        required=False,
        help="Revision of pretrained model identifier from huggingface.co/models.",
    )
    parser.add_argument(
        "--variant",
        type=str,
        default=None,
        help="Variant of the model files of the pretrained model identifier from huggingface.co/models, 'e.g.' fp16",
    )
    parser.add_argument(
        "--train_data_dir",
        type=str,
        default=None,
        help=(
            "A folder containing the training data. "
        ),
    )
    parser.add_argument(
        "--train_data_meta",
        type=str,
        default=None,
        help=(
            "A csv containing the training data. "
        ),
    )
    parser.add_argument(
        "--max_train_samples",
        type=int,
        default=None,
        help=(
            "For debugging purposes or quicker training, truncate the number of training examples to this "
            "value if set."
        ),
    )
    parser.add_argument(
        "--validation_prompts",
        type=str,
        default=None,
        nargs="+",
        help=("A set of prompts evaluated every `--validation_epochs` and logged to `--report_to`."),
    )
    parser.add_argument(
        "--output_dir",
        type=str,
        default="sd-model-finetuned",
        help="The output directory where the model predictions and checkpoints will be written.",
    )
    parser.add_argument(
        "--cache_dir",
        type=str,
        default=None,
        help="The directory where the downloaded models and datasets will be stored.",
    )
    parser.add_argument(
        "--seed",
        type=int,
        default=None,
        help="A seed for reproducible training."
    )
    parser.add_argument(
        "--random_flip",
        action="store_true",
        help="whether to randomly flip images horizontally",
    )
    parser.add_argument(
        "--use_came",
        action="store_true",
        help="whether to use came",
    )
    parser.add_argument(
        "--multi_stream",
        action="store_true",
        help="whether to use cuda multi-stream",
    )
    parser.add_argument(
        "--train_batch_size",
        type=int,
        default=16,
        help="Batch size (per device) for the training dataloader."
    )
    parser.add_argument(
        "--vae_mini_batch",
        type=int,
        default=32,
        help="mini batch size for vae."
    )
    parser.add_argument(
        "--num_train_epochs",
        type=int,
        default=100
    )
    parser.add_argument(
        "--max_train_steps",
        type=int,
        default=None,
        help="Total number of training steps to perform.  If provided, overrides num_train_epochs.",
    )
    parser.add_argument(
        "--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=1e-4,
        help="Initial learning rate (after the potential warmup period) to use.",
    )
    parser.add_argument(
        "--scale_lr",
        action="store_true",
        default=False,
        help="Scale the learning rate by the number of GPUs, gradient accumulation steps, and batch size.",
    )
    parser.add_argument(
        "--lr_scheduler",
        type=str,
        default="constant",
        help=(
            'The scheduler type to use. Choose between ["linear", "cosine", "cosine_with_restarts", "polynomial",'
            ' "constant", "constant_with_warmup"]'
        ),
    )
    parser.add_argument(
        "--lr_warmup_steps",
        type=int,
        default=500,
        help="Number of steps for the warmup in the lr scheduler."
    )
    parser.add_argument(
        "--use_8bit_adam",
        action="store_true",
        help="Whether or not to use 8-bit Adam from bitsandbytes."
    )
    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(
        "--use_ema",
        action="store_true",
        help="Whether to use EMA model."
    )
    parser.add_argument(
        "--non_ema_revision",
        type=str,
        default=None,
        required=False,
        help=(
            "Revision of pretrained non-ema model identifier. Must be a branch, tag or git identifier of the local or"
            " remote repository specified with --pretrained_model_name_or_path."
        ),
    )
    parser.add_argument(
        "--dataloader_num_workers",
        type=int,
        default=0,
        help=(
            "Number of subprocesses to use for data loading. 0 means that the data will be loaded in the main process."
        ),
    )
    parser.add_argument(
        "--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(
        "--prediction_type",
        type=str,
        default=None,
        help="The prediction_type that shall be used for training. Choose between 'epsilon' or 'v_prediction' or leave `None`. If left to `None` the default prediction type of the scheduler: `noise_scheduler.config.prediciton_type` is chosen.",
    )
    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(
        "--report_model_info",
        action="store_true",
        help="Whether or not to report more info about model (such as norm, grad)."
    )
    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(
        "--report_to",
        type=str,
        default="tensorboard",
        help=(
            'The integration to report the results and logs to. Supported platforms are `"tensorboard"`'
            ' (default), `"wandb"` and `"comet_ml"`. Use `"all"` to report to all integrations.'
        ),
    )
    parser.add_argument(
        "--local_rank",
        type=int,
        default=-1,
        help="For distributed training: local_rank"
    )
    parser.add_argument(
        "--checkpointing_steps",
        type=int,
        default=500,
        help=(
            "Save a checkpoint of the training state every X updates. These checkpoints are only suitable for resuming"
            " training using `--resume_from_checkpoint`."
        ),
    )
    parser.add_argument(
        "--checkpoints_total_limit",
        type=int,
        default=None,
        help=("Max number of checkpoints to store."),
    )
    parser.add_argument(
        "--resume_from_checkpoint",
        type=str,
        default=None,
        help=(
            "Whether training should be resumed from a previous checkpoint. Use a path saved by"
            ' `--checkpointing_steps`, or `"latest"` to automatically select the last available checkpoint.'
        ),
    )
    parser.add_argument(
        "--noise_offset",
        type=float,
        default=0,
        help="The scale of noise offset."
    )
    parser.add_argument(
        "--validation_epochs",
        type=int,
        default=5,
        help="Run validation every X epochs.",
    )
    parser.add_argument(
        "--validation_steps",
        type=int,
        default=2000,
        help="Run validation every X steps.",
    )
    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"
        ),
    )
    
    parser.add_argument(
        "--snr_loss",
        action="store_true",
        help="Whether or not to use snr_loss."
    )
    parser.add_argument(
        "--uniform_sampling",
        action="store_true",
        help="Whether or not to use uniform_sampling."
    )
    parser.add_argument(
        "--enable_text_encoder_in_dataloader",
        action="store_true",
        help="Whether or not to use text encoder in dataloader."
    )
    parser.add_argument(
        "--enable_bucket",
        action="store_true",
        help="Whether enable bucket sample in datasets."
    )
    parser.add_argument(
        "--random_ratio_crop",
        action="store_true",
        help="Whether enable random ratio crop sample in datasets."
    )
    parser.add_argument(
        "--random_frame_crop",
        action="store_true",
        help="Whether enable random frame crop sample in datasets."
    )
    parser.add_argument(
        "--random_hw_adapt",
        action="store_true",
        help="Whether enable random adapt height and width in datasets."
    )
    parser.add_argument(
        "--training_with_video_token_length",
        action="store_true",
        help="The training stage of the model in training.",
    )
    parser.add_argument(
        "--auto_tile_batch_size",
        action="store_true",
        help="Whether to auto tile batch size.",
    )
    parser.add_argument(
        "--motion_sub_loss",
        action="store_true",
        help="Whether enable motion sub loss."
    )
    parser.add_argument(
        "--motion_sub_loss_ratio",
        type=float,
        default=0.25,
        help="The ratio of motion sub loss."
    )
    parser.add_argument(
        "--train_sampling_steps",
        type=int,
        default=1000,
        help="Run train_sampling_steps.",
    )
    parser.add_argument(
        "--keep_all_node_same_token_length",
        action="store_true", 
        help="Reference of the length token.",
    )
    parser.add_argument(
        "--token_sample_size",
        type=int,
        default=512,
        help="Sample size of the token.",
    )
    parser.add_argument(
        "--video_sample_size",
        type=int,
        default=512,
        help="Sample size of the video.",
    )
    parser.add_argument(
        "--image_sample_size",
        type=int,
        default=512,
        help="Sample size of the image.",
    )
    parser.add_argument(
        "--fix_sample_size", 
        nargs=2, type=int, default=None,
        help="Fix Sample size [height, width] when using bucket and collate_fn."
    )
    parser.add_argument(
        "--video_sample_stride",
        type=int,
        default=4,
        help="Sample stride of the video.",
    )
    parser.add_argument(
        "--video_sample_n_frames",
        type=int,
        default=17,
        help="Num frame of video.",
    )
    parser.add_argument(
        "--video_repeat",
        type=int,
        default=0,
        help="Num of repeat video.",
    )
    parser.add_argument(
        "--config_path",
        type=str,
        default=None,
        help=(
            "The config of the model in training."
        ),
    )
    parser.add_argument(
        "--transformer_path",
        type=str,
        default="",
        help=("If you want to load the weight from other transformers, input its path."),
    )
    parser.add_argument(
        "--vae_path",
        type=str,
        default=None,
        help=("If you want to load the weight from other vaes, input its path."),
    )
    parser.add_argument(
        '--require_grad_modules', 
        nargs='+', 
        help='Enter a list of trainable modules'
    )
    parser.add_argument(
        '--trainable_modules', 
        nargs='+', 
        help='Enter a list of trainable modules'
    )
    parser.add_argument(
        '--trainable_modules_low_learning_rate', 
        nargs='+', 
        default=[],
        help='Enter a list of trainable modules with lower learning rate'
    )
    parser.add_argument(
        '--allow_missisng_modules',
        nargs='+', 
        default=[],
        help='Enter a list of modules can be missed by given checkpoint'
    )
    parser.add_argument(
        '--rgb_loss_weight', 
        type=float,
        default=1.0
    )
    parser.add_argument(
        '--motion_loss_weight', 
        type=float,
        default=1.0
    )
    parser.add_argument(
        '--pose_loss_weight', 
        type=float,
        default=1.0
    )
    parser.add_argument(
        '--vertices_loss_weight', 
        type=float,
        default=1.0
    )
    parser.add_argument(
        '--joint3d_loss_weight', 
        type=float,
        default=1.0
    )
    parser.add_argument(
        '--tokenizer_max_length', 
        type=int,
        default=512,
        help='Max length of tokenizer'
    )
    parser.add_argument(
        "--use_deepspeed",
        action="store_true",
        help="Whether or not to use deepspeed."
    )
    parser.add_argument(
        "--use_fsdp",
        action="store_true",
        help="Whether or not to use fsdp."
    )
    parser.add_argument(
        "--low_vram",
        action="store_true",
        help="Whether enable low_vram mode."
    )
    parser.add_argument(
        "--boundary_type",
        type=str,
        default="low",
        help=(
            'The format of training data. Support `"low"` and `"high"`'
        ),
    )
    parser.add_argument(
        "--train_mode",
        type=str,
        default="normal",
        help=(
            'The format of training data. Support `"normal"`'
            ' (default), `"i2v"`.'
        ),
    )
    parser.add_argument(
        "--motion_type",
        type=str,
        default="smpl",
        choices=["smpl", "dwpose"]
    )
    parser.add_argument(
        "--motion_only",
        action="store_true"
    )
    parser.add_argument(
        "--interaction",
        type=str,
        default="dual",
        choices=["dual", "single_m2v", "single_v2m", "none"]
    )
    parser.add_argument(
        "--interleave",
        type=int,
        default=2
    )
    parser.add_argument(
        "--predict_smpl",
        action="store_true"
    )
    parser.add_argument(
        "--predict_smpl_num",
        type=int,
        default=1
    )
    parser.add_argument(
        "--abnormal_norm_clip_start",
        type=int,
        default=1000,
        help=(
            'When do we start doing additional processing on abnormal gradients. '
        ),
    )
    parser.add_argument(
        "--initial_grad_norm_ratio",
        type=int,
        default=5,
        help=(
            'The initial gradient is relative to the multiple of the max_grad_norm. '
        ),
    )
    parser.add_argument(
        "--weighting_scheme",
        type=str,
        default="none",
        choices=["sigma_sqrt", "logit_normal", "mode", "cosmap", "none"],
        help=('We default to the "none" weighting scheme for uniform sampling and uniform loss'),
    )
    parser.add_argument(
        "--logit_mean",
        type=float,
        default=0.0,
        help="mean to use when using the `'logit_normal'` weighting scheme."
    )
    parser.add_argument(
        "--logit_std",
        type=float,
        default=1.0,
        help="std to use when using the `'logit_normal'` weighting scheme."
    )
    parser.add_argument(
        "--mode_scale",
        type=float,
        default=1.29,
        help="Scale of mode weighting scheme. Only effective when using the `'mode'` as the `weighting_scheme`.",
    )

    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

    # default to using the same revision for the non-ema model if not specified
    if args.non_ema_revision is None:
        args.non_ema_revision = args.revision

    return args


def main():
    args = parse_args()

    if args.report_to == "wandb" and args.hub_token is not None:
        raise ValueError(
            "You cannot use both --report_to=wandb and --hub_token due to a security risk of exposing your token."
            " Please use `huggingface-cli login` to authenticate with the Hub."
        )

    if args.non_ema_revision is not None:
        deprecate(
            "non_ema_revision!=None",
            "0.15.0",
            message=(
                "Downloading 'non_ema' weights from revision branches of the Hub is deprecated. Please make sure to"
                " use `--variant=non_ema` instead."
            ),
        )
    logging_dir = os.path.join(args.output_dir, args.logging_dir)

    config = OmegaConf.load(args.config_path)
    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,
    )

    deepspeed_plugin = accelerator.state.deepspeed_plugin if hasattr(accelerator.state, "deepspeed_plugin") else None
    fsdp_plugin = accelerator.state.fsdp_plugin if hasattr(accelerator.state, "fsdp_plugin") else None
    if deepspeed_plugin is not None:
        zero_stage = int(deepspeed_plugin.zero_stage)
        fsdp_stage = 0
        print(f"Using DeepSpeed Zero stage: {zero_stage}")

        args.use_deepspeed = True
        if zero_stage == 3:
            print(f"Auto set save_state to True because zero_stage == 3")
            args.save_state = True
    elif fsdp_plugin is not None:
        from torch.distributed.fsdp import ShardingStrategy
        zero_stage = 0
        if fsdp_plugin.sharding_strategy is ShardingStrategy.FULL_SHARD:
            fsdp_stage = 3
        elif fsdp_plugin.sharding_strategy is None: # The fsdp_plugin.sharding_strategy is None in FSDP 2.
            fsdp_stage = 3
        elif fsdp_plugin.sharding_strategy is ShardingStrategy.SHARD_GRAD_OP:
            fsdp_stage = 2
        else:
            fsdp_stage = 0
        print(f"Using FSDP stage: {fsdp_stage}")

        args.use_fsdp = True
        if fsdp_stage == 3:
            print(f"Auto set save_state to True because fsdp_stage == 3")
            args.save_state = True
    else:
        zero_stage = 0
        fsdp_stage = 0
        print("DeepSpeed is not enabled.")

    if accelerator.is_main_process:
        writer = SummaryWriter(log_dir=os.path.join(logging_dir, args.tracker_project_name))

    # 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:
        datasets.utils.logging.set_verbosity_warning()
        transformers.utils.logging.set_verbosity_warning()
        diffusers.utils.logging.set_verbosity_info()
    else:
        datasets.utils.logging.set_verbosity_error()
        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)
        rng = np.random.default_rng(np.random.PCG64(args.seed + accelerator.process_index))
        torch_rng = torch.Generator(accelerator.device).manual_seed(args.seed + accelerator.process_index)
    else:
        rng = None
        torch_rng = None
    index_rng = np.random.default_rng(np.random.PCG64(43))

    # Handle the repository creation
    if accelerator.is_main_process:
        if args.output_dir is not None:
            os.makedirs(os.path.join(args.output_dir, args.tracker_project_name), exist_ok=True)

    # For mixed precision training we cast all non-trainable weigths (vae, non-lora text_encoder and non-lora transformer3d) 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
        args.mixed_precision = accelerator.mixed_precision
    elif accelerator.mixed_precision == "bf16":
        weight_dtype = torch.bfloat16
        args.mixed_precision = accelerator.mixed_precision

    # Load scheduler, tokenizer and models.
    noise_scheduler = FlowMatchEulerDiscreteScheduler(
        **filter_kwargs(FlowMatchEulerDiscreteScheduler, OmegaConf.to_container(config['scheduler_kwargs']))
    )

    # Get Tokenizer
    tokenizer = AutoTokenizer.from_pretrained(
        os.path.join(args.pretrained_model_name_or_path, config['text_encoder_kwargs'].get('tokenizer_subpath', 'tokenizer')),
    )

    def deepspeed_zero_init_disabled_context_manager():
        """
        returns either a context list that includes one that will disable zero.Init or an empty context list
        """
        deepspeed_plugin = AcceleratorState().deepspeed_plugin if accelerate.state.is_initialized() else None
        if deepspeed_plugin is None:
            return []

        return [deepspeed_plugin.zero3_init_context_manager(enable=False)]

    # Currently Accelerate doesn't know how to handle multiple models under Deepspeed ZeRO stage 3.
    # For this to work properly all models must be run through `accelerate.prepare`. But accelerate
    # will try to assign the same optimizer with the same weights to all models during
    # `deepspeed.initialize`, which of course doesn't work.
    #
    # For now the following workaround will partially support Deepspeed ZeRO-3, by excluding the 2
    # frozen models from being partitioned during `zero.Init` which gets called during
    # `from_pretrained` So CLIPTextModel and AutoencoderKL will not enjoy the parameter sharding
    # across multiple gpus and only UNet2DConditionModel will get ZeRO sharded.
    with ContextManagers(deepspeed_zero_init_disabled_context_manager()):
        # Get Text encoder
        text_encoder = WanT5EncoderModel.from_pretrained(
            os.path.join(args.pretrained_model_name_or_path, config['text_encoder_kwargs'].get('text_encoder_subpath', 'text_encoder')),
            additional_kwargs=OmegaConf.to_container(config['text_encoder_kwargs']),
            low_cpu_mem_usage=True,
            torch_dtype=weight_dtype,
        )
        text_encoder = text_encoder.eval()
        # Get Vae
        Chosen_AutoencoderKL = {
            "AutoencoderKLWan": AutoencoderKLWan,
            "AutoencoderKLWan3_8": AutoencoderKLWan3_8
        }[config['vae_kwargs'].get('vae_type', 'AutoencoderKLWan')]
        vae = Chosen_AutoencoderKL.from_pretrained(
            os.path.join(args.pretrained_model_name_or_path, config['vae_kwargs'].get('vae_subpath', 'vae')),
            additional_kwargs=OmegaConf.to_container(config['vae_kwargs']),
        )
        vae.eval()
            
    # Get Transformer
    if args.boundary_type == "low" or args.boundary_type == "full":
        sub_path = config['transformer_additional_kwargs'].get('transformer_low_noise_model_subpath', 'transformer')
    else:
        sub_path = config['transformer_additional_kwargs'].get('transformer_high_noise_model_subpath', 'transformer')
    transformer3d = ComoviTransformer3DModel.from_pretrained(
        os.path.join(args.pretrained_model_name_or_path, sub_path),
        transformer_additional_kwargs=OmegaConf.to_container(config['transformer_additional_kwargs']),
        interaction=args.interaction,
        interleave=args.interleave,
        predict_smpl=args.predict_smpl
    ).to(weight_dtype)

    # Get smpl body model
    body_model = smplx.SMPLLayer(model_path="./body_model/SMPL/SMPL_NEUTRAL.pkl", num_betas=10).to(accelerator.device).to(weight_dtype)

    # Freeze vae and text_encoder and set transformer3d to trainable
    vae.requires_grad_(False)
    text_encoder.requires_grad_(False)
    transformer3d.requires_grad_(False)

    if args.transformer_path:
        print(f"starting from customized checkpoint: {args.transformer_path}")
        if args.transformer_path.endswith("safetensors"):
            from safetensors.torch import load_file, safe_open
            state_dict = load_file(args.transformer_path)
        else:
            state_dict = torch.load(args.transformer_path, map_location="cpu")
        state_dict = state_dict["state_dict"] if "state_dict" in state_dict else state_dict

        m, u = transformer3d.load_state_dict(state_dict, strict=False)
        assert len(u) == 0, f"no unexpected module is allowed for given customized checkpoint {args.transformer_path}"
        if len(m) > 0:
            for _m in m:
                allowed = False
                for am in args.allow_missisng_modules:
                    if am in _m:
                        allowed = True
                        break
                assert allowed, f"module {_m} is missed by the given checkpoint, which is not allowed"
        print(f"customized checkpoint: {args.transformer_path} loded successfully")

    if args.vae_path is not None:
        print(f"From checkpoint: {args.vae_path}")
        if args.vae_path.endswith("safetensors"):
            from safetensors.torch import load_file, safe_open
            state_dict = load_file(args.vae_path)
        else:
            state_dict = torch.load(args.vae_path, map_location="cpu")
        state_dict = state_dict["state_dict"] if "state_dict" in state_dict else state_dict

        m, u = vae.load_state_dict(state_dict, strict=False)
        print(f"missing keys: {len(m)}, unexpected keys: {len(u)}")
        assert len(u) == 0
    
    # A good trainable modules is showed below now.
    # For 3D Patch: trainable_modules = ['ff.net', 'pos_embed', 'attn2', 'proj_out', 'timepositionalencoding', 'h_position', 'w_position']
    # For 2D Patch: trainable_modules = ['ff.net', 'attn2', 'timepositionalencoding', 'h_position', 'w_position']
    transformer3d.train()
    for name, param in transformer3d.named_parameters():
        for trainable_module_name in args.require_grad_modules:
            if trainable_module_name in name:
                param.requires_grad = True
                break

    # Create EMA for the transformer3d.
    if args.use_ema:
        if zero_stage == 3:
            raise NotImplementedError("FSDP does not support EMA.")

        ema_transformer3d = Wan2_2Transformer3DModel.from_pretrained(
            os.path.join(args.pretrained_model_name_or_path, config['transformer_additional_kwargs'].get('transformer_subpath', 'transformer')),
            transformer_additional_kwargs=OmegaConf.to_container(config['transformer_additional_kwargs']),
        ).to(weight_dtype)

        ema_transformer3d = EMAModel(ema_transformer3d.parameters(), model_cls=Wan2_2Transformer3DModel, model_config=ema_transformer3d.config)

    # `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
        if fsdp_stage != 0:
            def save_model_hook(models, weights, output_dir):
                accelerate_state_dict = accelerator.get_state_dict(models[-1], unwrap=True)
                if accelerator.is_main_process:
                    from safetensors.torch import save_file

                    safetensor_save_path = os.path.join(output_dir, f"diffusion_pytorch_model.safetensors")
                    accelerate_state_dict = {k: v.to(dtype=weight_dtype) for k, v in accelerate_state_dict.items()}
                    save_file(accelerate_state_dict, safetensor_save_path, metadata={"format": "pt"})

                    with open(os.path.join(output_dir, "sampler_pos_start.pkl"), 'wb') as file:
                        pickle.dump([batch_sampler.sampler._pos_start, first_epoch], file)

            def load_model_hook(models, input_dir):
                pkl_path = os.path.join(input_dir, "sampler_pos_start.pkl")
                if os.path.exists(pkl_path):
                    with open(pkl_path, 'rb') as file:
                        loaded_number, _ = pickle.load(file)
                        batch_sampler.sampler._pos_start = max(loaded_number - args.dataloader_num_workers * accelerator.num_processes * 2, 0)
                    print(f"Load pkl from {pkl_path}. Get loaded_number = {loaded_number}.")

        elif zero_stage == 3:
            # 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:
                    with open(os.path.join(output_dir, "sampler_pos_start.pkl"), 'wb') as file:
                        pickle.dump([batch_sampler.sampler._pos_start, first_epoch], file)

            def load_model_hook(models, input_dir):
                pkl_path = os.path.join(input_dir, "sampler_pos_start.pkl")
                if os.path.exists(pkl_path):
                    with open(pkl_path, 'rb') as file:
                        loaded_number, _ = pickle.load(file)
                        batch_sampler.sampler._pos_start = max(loaded_number - args.dataloader_num_workers * accelerator.num_processes * 2, 0)
                    print(f"Load pkl from {pkl_path}. Get loaded_number = {loaded_number}.")
        else:
            # 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:
                    if args.use_ema:
                        ema_transformer3d.save_pretrained(os.path.join(output_dir, "transformer_ema"))

                    models[0].save_pretrained(os.path.join(output_dir, "transformer"))
                    if not args.use_deepspeed:
                        weights.pop()

                    with open(os.path.join(output_dir, "sampler_pos_start.pkl"), 'wb') as file:
                        pickle.dump([batch_sampler.sampler._pos_start, first_epoch], file)

            def load_model_hook(models, input_dir):
                if args.use_ema:
                    ema_path = os.path.join(input_dir, "transformer_ema")
                    _, ema_kwargs = Wan2_2Transformer3DModel.load_config(ema_path, return_unused_kwargs=True)
                    load_model = Wan2_2Transformer3DModel.from_pretrained(
                        input_dir, subfolder="transformer_ema",
                        transformer_additional_kwargs=OmegaConf.to_container(config['transformer_additional_kwargs'])
                    )
                    load_model = EMAModel(load_model.parameters(), model_cls=Wan2_2Transformer3DModel, model_config=load_model.config)
                    load_model.load_state_dict(ema_kwargs)

                    ema_transformer3d.load_state_dict(load_model.state_dict())
                    ema_transformer3d.to(accelerator.device)
                    del load_model

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

                    # load diffusers style into model
                    load_model = Wan2_2Transformer3DModel.from_pretrained(
                        input_dir, subfolder="transformer"
                    )
                    model.register_to_config(**load_model.config)

                    model.load_state_dict(load_model.state_dict())
                    del load_model

                pkl_path = os.path.join(input_dir, "sampler_pos_start.pkl")
                if os.path.exists(pkl_path):
                    with open(pkl_path, 'rb') as file:
                        loaded_number, _ = pickle.load(file)
                        batch_sampler.sampler._pos_start = max(loaded_number - args.dataloader_num_workers * accelerator.num_processes * 2, 0)
                    print(f"Load pkl from {pkl_path}. Get loaded_number = {loaded_number}.")

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

    if args.gradient_checkpointing:
        transformer3d.enable_gradient_checkpointing()

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

    # Initialize the optimizer
    if args.use_8bit_adam:
        try:
            import bitsandbytes as bnb
        except ImportError:
            raise ImportError(
                "Please install bitsandbytes to use 8-bit Adam. You can do so by running `pip install bitsandbytes`"
            )

        optimizer_cls = bnb.optim.AdamW8bit
    elif args.use_came:
        try:
            from came_pytorch import CAME
        except:
            raise ImportError(
                "Please install came_pytorch to use CAME. You can do so by running `pip install came_pytorch`"
            )

        optimizer_cls = CAME
    else:
        optimizer_cls = torch.optim.AdamW

    if accelerator.is_main_process:
        accelerator.print(
            f"Trainable modules '{args.trainable_modules}'."
        )
    trainable_params = list(filter(lambda p: p.requires_grad, transformer3d.parameters()))
    trainable_params_optim = [
        {'params': [], 'lr': args.learning_rate, 'name': "_".join(args.trainable_modules)},
        {'params': [], 'lr': 0, 'name': "null" if len(args.trainable_modules_low_learning_rate) == 0 else "_".join(args.trainable_modules_low_learning_rate)}
    ]
    in_already = []
    for name, param in transformer3d.named_parameters():
        high_lr_flag = False
        if name in in_already:
            continue
        for trainable_module_name in args.trainable_modules:
            if trainable_module_name in name:
                in_already.append(name)
                high_lr_flag = True
                trainable_params_optim[0]['params'].append(param)
                if accelerator.is_main_process:
                    print(f"Set {name} to lr : {args.learning_rate}")
                break
        if high_lr_flag:
            continue
        for trainable_module_name in args.trainable_modules_low_learning_rate:
            if trainable_module_name in name:
                in_already.append(name)
                trainable_params_optim[1]['params'].append(param)
                if accelerator.is_main_process:
                    print(f"Set {name} to lr : {0}")
                break

    if args.use_came:
        optimizer = optimizer_cls(
            trainable_params_optim,
            lr=args.learning_rate,
            # weight_decay=args.adam_weight_decay,
            betas=(0.9, 0.999, 0.9999), 
            eps=(1e-30, 1e-16)
        )
    else:
        optimizer = optimizer_cls(
            trainable_params_optim,
            lr=args.learning_rate,
            betas=(args.adam_beta1, args.adam_beta2),
            weight_decay=args.adam_weight_decay,
            eps=args.adam_epsilon,
        )

    # Get the training dataset
    sample_n_frames_bucket_interval = vae.config.temporal_compression_ratio # 4
    spatial_compression_ratio = vae.config.spatial_compression_ratio        # 16

    if args.fix_sample_size is not None and args.enable_bucket:
        args.video_sample_size = max(max(args.fix_sample_size), args.video_sample_size)
        args.image_sample_size = max(max(args.fix_sample_size), args.image_sample_size)
        args.training_with_video_token_length = False
        args.random_hw_adapt = False

    train_dataset = ComoviDataset(
        args.train_data_meta, args.train_data_dir,
        video_sample_size=args.video_sample_size, video_sample_stride=args.video_sample_stride, video_sample_n_frames=args.video_sample_n_frames, 
        video_repeat=args.video_repeat, 
        image_sample_size=args.image_sample_size,
        enable_bucket=args.enable_bucket, enable_inpaint=True if args.train_mode != "normal" else False,
        motion_type=args.motion_type
    )
    
    if args.enable_bucket:
        batch_sampler_generator = torch.Generator().manual_seed(args.seed)
        batch_sampler = ComoviSampler(
            sampler=RandomSampler(train_dataset, generator=batch_sampler_generator),
            batch_size=args.train_batch_size,
            drop_last=True
        )
        
        def collate_fn(examples):
            """
            organize all the training data in fixed resolution, fps, and length
            """
            target_token_length = args.video_sample_n_frames * args.token_sample_size * args.token_sample_size

            # Create new output
            new_examples                                = {}
            new_examples["target_token_length"]         = target_token_length
            new_examples["pixel_values"]                = []
            new_examples["motion_values"]               = []
            new_examples["firstframe_values"]           = []
            new_examples["motion_firstframe_values"]    = []
            new_examples["smpl_poses"]                  = []
            new_examples["smpl_shapes"]                 = []
            new_examples["text"]                        = []
            # Used in Inpaint mode 
            if args.train_mode != "normal":
                new_examples["mask_pixel_values"]       = []
                new_examples["mask_motion_values"]      = []
                new_examples["mask"]                    = []

            for example in examples:
                # To 0~1
                pixel_values = torch.from_numpy(example["pixel_values"]).permute(0, 3, 1, 2).contiguous()
                pixel_values = pixel_values / 255.

                motion_values = torch.from_numpy(example["motion_values"]).permute(0, 3, 1, 2).contiguous()
                motion_values = motion_values / 255.

                firstframe_values = torch.from_numpy(example["first_frame"]).permute(0, 3, 1, 2).contiguous()
                firstframe_values = firstframe_values / 255.

                motion_firstframe_values = torch.from_numpy(example["motion_first_frame"]).permute(0, 3, 1, 2).contiguous()
                motion_firstframe_values = motion_firstframe_values / 255.

                smpl_poses = torch.from_numpy(np.concatenate([example["smpl_values"]["global_orient"], example["smpl_values"]["body_pose"]], axis=1))
                smpl_poses = matrix_to_rotation_6d(smpl_poses)
                smpl_shapes = torch.from_numpy(example["smpl_values"]["betas"])

                transform = transforms.Compose([
                    transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
                ])
                new_examples["pixel_values"].append(transform(pixel_values))
                new_examples["motion_values"].append(transform(motion_values))
                new_examples["firstframe_values"].append(transform(firstframe_values))
                new_examples["motion_firstframe_values"].append(transform(motion_firstframe_values))
                new_examples["smpl_poses"].append(smpl_poses)
                new_examples["smpl_shapes"].append(smpl_shapes)
                new_examples["text"].append(example["text"])

                batch_video_length = len(pixel_values)

                if args.train_mode != "normal":
                    mask = get_random_mask(new_examples["pixel_values"][-1].size(), image_start_only=True)
                    mask_pixel_values = new_examples["firstframe_values"][-1].repeat(mask.shape[0], 1, 1, 1) * (1 - mask)
                    mask_motion_values = new_examples["motion_firstframe_values"][-1].repeat(mask.shape[0], 1, 1, 1) * (1 - mask)
                    # Wan 2.1 use 0 for masked pixels
                    # + torch.ones_like(new_examples["pixel_values"][-1]) * -1 * mask
                    new_examples["mask_pixel_values"].append(mask_pixel_values)
                    new_examples["mask_motion_values"].append(mask_motion_values)
                    new_examples["mask"].append(mask)

            # Limit the number of frames to the same
            new_examples["pixel_values"] = torch.stack([example[:batch_video_length] for example in new_examples["pixel_values"]])
            new_examples["motion_values"] = torch.stack([example[:batch_video_length] for example in new_examples["motion_values"]])
            new_examples["smpl_poses"] = torch.stack([example[:batch_video_length] for example in new_examples["smpl_poses"]])
            new_examples["smpl_shapes"] = torch.stack([example[:batch_video_length] for example in new_examples["smpl_shapes"]])
            if args.train_mode != "normal":
                new_examples["mask_pixel_values"] = torch.stack([example[:batch_video_length] for example in new_examples["mask_pixel_values"]])
                new_examples["mask_motion_values"] = torch.stack([example[:batch_video_length] for example in new_examples["mask_motion_values"]])
                new_examples["mask"] = torch.stack([example[:batch_video_length] for example in new_examples["mask"]])

            return new_examples

        # DataLoaders creation:
        train_dataloader = torch.utils.data.DataLoader(
            train_dataset,
            batch_sampler=batch_sampler,
            collate_fn=collate_fn,
            persistent_workers=True if args.dataloader_num_workers != 0 else False,
            num_workers=args.dataloader_num_workers,
        )
    else:
        # DataLoaders creation:
        batch_sampler_generator = torch.Generator().manual_seed(args.seed)
        batch_sampler = ImageVideoSampler(RandomSampler(train_dataset, generator=batch_sampler_generator), train_dataset, args.train_batch_size)
        train_dataloader = torch.utils.data.DataLoader(
            train_dataset,
            batch_sampler=batch_sampler, 
            persistent_workers=True if args.dataloader_num_workers != 0 else False,
            num_workers=args.dataloader_num_workers,
        )

    # 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

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

    # Prepare everything with our `accelerator`.
    transformer3d, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
        transformer3d, optimizer, train_dataloader, lr_scheduler
    )

    if fsdp_stage != 0:
        from functools import partial
        from videox_fun.dist import set_multi_gpus_devices, shard_model
        shard_fn = partial(shard_model, device_id=accelerator.device, param_dtype=weight_dtype)
        text_encoder = shard_fn(text_encoder)

    if args.use_ema:
        ema_transformer3d.to(accelerator.device)

    # Move text_encode and vae to gpu and cast to weight_dtype
    vae.to(accelerator.device if not args.low_vram else "cpu", dtype=weight_dtype)
    if not args.enable_text_encoder_in_dataloader:
        text_encoder.to(accelerator.device if not args.low_vram else "cpu")

    # 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))
        tracker_config.pop("validation_prompts")
        tracker_config.pop("require_grad_modules")
        tracker_config.pop("trainable_modules")
        tracker_config.pop("trainable_modules_low_learning_rate")
        tracker_config.pop("allow_missisng_modules")
        tracker_config.pop("fix_sample_size")
        accelerator.init_trackers(args.tracker_project_name, tracker_config)

    # Function for unwrapping if model was compiled with `torch.compile`.
    def unwrap_model(model):
        model = accelerator.unwrap_model(model)
        model = model._orig_mod if is_compiled_module(model) else model
        return model

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

    logger.info("***** Running training *****")
    logger.info(f"  Num examples = {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:
        os.makedirs(os.path.join(args.output_dir, args.tracker_project_name), exist_ok=True)
        if args.resume_from_checkpoint != "latest":
            path = os.path.basename(args.resume_from_checkpoint)
        else:
            # Get the most recent checkpoint
            dirs = os.listdir(os.path.join(args.output_dir, args.tracker_project_name))
            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:
            global_step = int(path.split("-")[1])

            initial_global_step = global_step

            pkl_path = os.path.join(os.path.join(args.output_dir, args.tracker_project_name, path), "sampler_pos_start.pkl")
            if os.path.exists(pkl_path):
                with open(pkl_path, 'rb') as file:
                    _, first_epoch = pickle.load(file)
            else:
                first_epoch = global_step // num_update_steps_per_epoch
            print(f"Load pkl from {pkl_path}. Get first_epoch = {first_epoch}.")

            accelerator.print(f"Resuming from checkpoint {os.path.join(args.output_dir, args.tracker_project_name, path)}")
            accelerator.load_state(os.path.join(args.output_dir, args.tracker_project_name, path))
    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,
    )

    if args.multi_stream and args.train_mode != "normal":
        # create extra cuda streams to speedup inpaint vae computation
        vae_stream_1 = torch.cuda.Stream()
        vae_stream_2 = torch.cuda.Stream()
    else:
        vae_stream_1 = None
        vae_stream_2 = None

    # Calculate the index we need
    boundary        = config['transformer_additional_kwargs'].get('boundary', 0.900)
    split_timesteps = args.train_sampling_steps * boundary
    differences     = torch.abs(noise_scheduler.timesteps - split_timesteps)
    closest_index   = torch.argmin(differences).item()
    if args.boundary_type == "high" or args.boundary_type == "low":
        print(f"The boundary is {boundary} and the boundary_type is {args.boundary_type}. The closest_index we calculate is {closest_index}")
    if args.boundary_type == "high":
        start_num_idx = 0
        train_sampling_steps = closest_index
    elif args.boundary_type == "low":
        start_num_idx = closest_index
        train_sampling_steps = args.train_sampling_steps - closest_index
    else:
        start_num_idx = 0
        train_sampling_steps = args.train_sampling_steps

    idx_sampling = DiscreteSampling(train_sampling_steps, start_num_idx=start_num_idx, uniform_sampling=args.uniform_sampling)

    for epoch in range(first_epoch, args.num_train_epochs):
        train_loss_log = 0.0; rgb_loss_log = 0.0; motion_loss_log = 0.0
        pose_loss_log = 0.0; vertices_loss_log = 0.0; joint3d_loss_log = 0.0
        batch_sampler.sampler.generator = torch.Generator().manual_seed(args.seed + epoch)
        for step, batch in enumerate(train_dataloader):
            # Data batch sanity check
            if epoch == first_epoch and step == 0 and accelerator.is_main_process:
                pixel_values, motion_values, texts = batch["pixel_values"].cpu(), batch["motion_values"].cpu(), batch["text"]
                pixel_values = rearrange(pixel_values, "b f c h w -> b c f h w")
                motion_values = rearrange(motion_values, "b f c h w -> b c f h w")
                os.makedirs(os.path.join(args.output_dir, "sanity_check"), exist_ok=True)
                for idx, (pixel_value, motion_value, text) in enumerate(zip(pixel_values, motion_values, texts)):
                    pixel_value = pixel_value[None, ...]
                    motion_value = motion_value[None, ...]
                    gif_name = '-'.join(text.replace('/', '').split()[:10]) if not text == '' else f'{global_step}-{idx}'
                    save_videos_grid([pixel_value, motion_value], f"{args.output_dir}/sanity_check/{gif_name[:10]}.gif", rescale=True)
                if args.train_mode != "normal":
                    mask_pixel_values, mask_motion_values, texts = batch['mask_pixel_values'].cpu(), batch['mask_motion_values'].cpu(), batch['text']
                    mask_pixel_values = rearrange(mask_pixel_values, "b f c h w -> b c f h w")
                    mask_motion_values = rearrange(mask_motion_values, "b f c h w -> b c f h w")
                    for idx, (pixel_value, motion_value, text) in enumerate(zip(mask_pixel_values, mask_motion_values, texts)):
                        pixel_value = pixel_value[None, ...]
                        motion_value = motion_value[None, ...]
                        save_videos_grid([pixel_value, motion_value], f"{args.output_dir}/sanity_check/mask_{gif_name[:10] if not text == '' else f'{global_step}-{idx}'}.gif", rescale=True)
            
            with accelerator.accumulate(transformer3d):
                # Convert images to latent space
                pixel_values = batch["pixel_values"].to(weight_dtype)
                motion_values = batch["motion_values"].to(weight_dtype)

                # sample which layer to predict 3D smpl
                layer_smpl_predict_idx = [len(transformer3d.motion_blocks)-1] + rng.choice([layer_i for layer_i in range(len(transformer3d.motion_blocks)-1)], size=args.predict_smpl_num-1, replace=False).tolist()
                smplpose_values = batch["smpl_poses"].to(weight_dtype)
                smplshape_values = batch["smpl_shapes"].to(weight_dtype)

                # Increase the batch size when the length of the latent sequence of the current sample is small
                if args.auto_tile_batch_size and args.training_with_video_token_length and zero_stage != 3:
                    if args.video_sample_n_frames * args.token_sample_size * args.token_sample_size // 16 >= pixel_values.size()[1] * pixel_values.size()[3] * pixel_values.size()[4]:
                        pixel_values = torch.tile(pixel_values, (4, 1, 1, 1, 1))
                        if args.enable_text_encoder_in_dataloader:
                            batch['encoder_hidden_states'] = torch.tile(batch['encoder_hidden_states'], (4, 1, 1))
                            batch['encoder_attention_mask'] = torch.tile(batch['encoder_attention_mask'], (4, 1))
                        else:
                            batch['text'] = batch['text'] * 4
                    elif args.video_sample_n_frames * args.token_sample_size * args.token_sample_size // 4 >= pixel_values.size()[1] * pixel_values.size()[3] * pixel_values.size()[4]:
                        pixel_values = torch.tile(pixel_values, (2, 1, 1, 1, 1))
                        if args.enable_text_encoder_in_dataloader:
                            batch['encoder_hidden_states'] = torch.tile(batch['encoder_hidden_states'], (2, 1, 1))
                            batch['encoder_attention_mask'] = torch.tile(batch['encoder_attention_mask'], (2, 1))
                        else:
                            batch['text'] = batch['text'] * 2
                
                if args.train_mode != "normal":
                    mask_pixel_values = batch["mask_pixel_values"].to(weight_dtype)
                    mask_motion_values = batch["mask_motion_values"].to(weight_dtype)
                    mask = batch["mask"].to(weight_dtype)

                    # Increase the batch size when the length of the latent sequence of the current sample is small
                    if args.auto_tile_batch_size and args.training_with_video_token_length and zero_stage != 3:
                        if args.video_sample_n_frames * args.token_sample_size * args.token_sample_size // 16 >= pixel_values.size()[1] * pixel_values.size()[3] * pixel_values.size()[4]:
                            mask_pixel_values = torch.tile(mask_pixel_values, (4, 1, 1, 1, 1))
                            mask_motion_values = torch.tile(mask_motion_values, (4, 1, 1, 1, 1))
                            mask = torch.tile(mask, (4, 1, 1, 1, 1))
                        elif args.video_sample_n_frames * args.token_sample_size * args.token_sample_size // 4 >= pixel_values.size()[1] * pixel_values.size()[3] * pixel_values.size()[4]:
                            mask_pixel_values = torch.tile(mask_pixel_values, (2, 1, 1, 1, 1))
                            mask_motion_values = torch.tile(mask_motion_values, (2, 1, 1, 1, 1))
                            mask = torch.tile(mask, (2, 1, 1, 1, 1))

                if args.random_frame_crop:
                    def _create_special_list(length):
                        if length == 1:
                            return [1.0]
                        if length >= 2:
                            last_element = 0.90
                            remaining_sum = 1.0 - last_element
                            other_elements_value = remaining_sum / (length - 1)
                            special_list = [other_elements_value] * (length - 1) + [last_element]
                            return special_list
                    select_frames = [_tmp for _tmp in list(range(sample_n_frames_bucket_interval + 1, args.video_sample_n_frames + sample_n_frames_bucket_interval, sample_n_frames_bucket_interval))]
                    select_frames_prob = np.array(_create_special_list(len(select_frames)))
                    
                    if len(select_frames) != 0:
                        if rng is None:
                            temp_n_frames = np.random.choice(select_frames, p = select_frames_prob)
                        else:
                            temp_n_frames = rng.choice(select_frames, p = select_frames_prob)
                    else:
                        temp_n_frames = 1

                    # Magvae needs the number of frames to be 4n + 1.
                    temp_n_frames = (temp_n_frames - 1) // sample_n_frames_bucket_interval + 1

                    pixel_values = pixel_values[:, :temp_n_frames, :, :]

                    if args.train_mode != "normal":
                        mask_pixel_values = mask_pixel_values[:, :temp_n_frames, :, :]
                        mask = mask[:, :temp_n_frames, :, :]
                    
                # Keep all node same token length to accelerate the traning when resolution grows.
                if args.keep_all_node_same_token_length:
                    if args.token_sample_size > 256:
                        numbers_list = list(range(256, args.token_sample_size + 1, 128))

                        if numbers_list[-1] != args.token_sample_size:
                            numbers_list.append(args.token_sample_size)
                    else:
                        numbers_list = [256]
                    numbers_list = [_number * _number * args.video_sample_n_frames for _number in  numbers_list]
            
                    actual_token_length = index_rng.choice(numbers_list)
                    actual_video_length = (min(
                            actual_token_length / pixel_values.size()[-1] / pixel_values.size()[-2], args.video_sample_n_frames
                    ) - 1) // sample_n_frames_bucket_interval * sample_n_frames_bucket_interval + 1
                    actual_video_length = int(max(actual_video_length, 1))

                    # Magvae needs the number of frames to be 4n + 1.
                    actual_video_length = (actual_video_length - 1) // sample_n_frames_bucket_interval + 1

                    pixel_values = pixel_values[:, :actual_video_length, :, :]
                    if args.train_mode != "normal":
                        mask_pixel_values = mask_pixel_values[:, :actual_video_length, :, :]
                        mask_motion_values = mask_motion_values[:, :actual_video_length, :, :]
                        mask = mask[:, :actual_video_length, :, :]

                # Make the inpaint latents to be zeros.
                if args.train_mode != "normal":
                    t2v_flag = [(_mask == 1).all() for _mask in mask]

                    new_t2v_flag = []
                    for _mask in t2v_flag:
                        if _mask and np.random.rand() < 0.90:
                            new_t2v_flag.append(0)
                        else:
                            new_t2v_flag.append(1)
                    t2v_flag = torch.from_numpy(np.array(new_t2v_flag)).to(accelerator.device, dtype=weight_dtype)

                if args.low_vram:
                    torch.cuda.empty_cache()
                    vae.to(accelerator.device)
                    if not args.enable_text_encoder_in_dataloader:
                        text_encoder.to("cpu")

                with torch.no_grad():
                    # This way is quicker when batch grows up
                    def _batch_encode_vae(pixel_values):
                        pixel_values = rearrange(pixel_values, "b f c h w -> b c f h w")
                        bs = args.vae_mini_batch
                        new_pixel_values = []
                        for i in range(0, pixel_values.shape[0], bs):
                            pixel_values_bs = pixel_values[i : i + bs]
                            pixel_values_bs = vae.encode(pixel_values_bs)[0]
                            pixel_values_bs = pixel_values_bs.sample()
                            new_pixel_values.append(pixel_values_bs)
                        return torch.cat(new_pixel_values, dim = 0)
                    if vae_stream_1 is not None:
                        vae_stream_1.wait_stream(torch.cuda.current_stream())
                        with torch.cuda.stream(vae_stream_1):
                            latents = _batch_encode_vae(pixel_values)
                            motion_latents = _batch_encode_vae(motion_values)
                    else:
                        latents = _batch_encode_vae(pixel_values)           # (b, 48, 21, 30, 52)
                        motion_latents = _batch_encode_vae(motion_values)   # (b, 48, 21, 30, 52)

                    if args.train_mode != "normal":
                        mask = rearrange(mask, "b f c h w -> b c f h w")                                    # (1, 1, 81, 480, 832)
                        mask = torch.concat(                                                                # (1, 1, 84, 480, 832)
                            [
                                torch.repeat_interleave(mask[:, :, 0:1], repeats=4, dim=2), 
                                mask[:, :, 1:]
                            ], dim=2
                        )
                        mask = mask.view(mask.shape[0], mask.shape[2] // 4, 4, mask.shape[3], mask.shape[4])# (1, 21, 4, 480, 832)
                        mask = mask.transpose(1, 2)                                                         # (1, 4, 21, 480, 832)

                        if args.train_mode != "ti2v":
                            mask = resize_mask(1 - mask, latents)
                        else:
                            mask = F.interpolate(mask[:, :1], size=latents.size()[-3:], mode='trilinear', align_corners=True).to(accelerator.device, weight_dtype)
                            # (1, 1, 21, 30, 52)
                            # mask[:, :, :1] is 0; mask[:, :, 1:] is 1
                        
                        # Encode inpaint latents.
                        rgb_mask_latents = _batch_encode_vae(mask_pixel_values)
                        motion_mask_latents = _batch_encode_vae(mask_motion_values)
                        if vae_stream_2 is not None:
                            torch.cuda.current_stream().wait_stream(vae_stream_2) 

                        if args.train_mode != "ti2v":
                            rgb_inpaint_latents = torch.concat([mask, rgb_mask_latents], dim=1)
                            rgb_inpaint_latents = t2v_flag[:, None, None, None, None] * rgb_inpaint_latents

                            motion_inpaint_latents = torch.concat([mask, motion_mask_latents], dim=1)
                            motion_inpaint_latents = t2v_flag[:, None, None, None, None] * motion_inpaint_latents
                        else:
                            rgb_inpaint_latents = rgb_mask_latents
                            motion_inpaint_latents = motion_mask_latents
                                                    
                # wait for latents = vae.encode(pixel_values) to complete
                if vae_stream_1 is not None:
                    torch.cuda.current_stream().wait_stream(vae_stream_1)

                if args.low_vram:
                    vae.to('cpu')
                    torch.cuda.empty_cache()
                    if not args.enable_text_encoder_in_dataloader:
                        text_encoder.to(accelerator.device)

                if args.enable_text_encoder_in_dataloader:
                    prompt_embeds = batch['encoder_hidden_states'].to(device=latents.device)
                else:
                    with torch.no_grad():
                        prompt_ids = tokenizer(
                            batch['text'], 
                            padding="max_length", 
                            max_length=args.tokenizer_max_length, 
                            truncation=True, 
                            add_special_tokens=True, 
                            return_tensors="pt"
                        )
                        text_input_ids = prompt_ids.input_ids
                        prompt_attention_mask = prompt_ids.attention_mask

                        seq_lens = prompt_attention_mask.gt(0).sum(dim=1).long()
                        prompt_embeds = text_encoder(text_input_ids.to(latents.device), attention_mask=prompt_attention_mask.to(latents.device))[0]
                        prompt_embeds = [u[:v] for u, v in zip(prompt_embeds, seq_lens)]

                if args.low_vram and not args.enable_text_encoder_in_dataloader:
                    text_encoder.to('cpu')
                    torch.cuda.empty_cache()

                bsz, channel, num_frames, height, width = latents.size()
                noise_shape = (bsz, channel*2, num_frames, height, width)
                noise = torch.randn(noise_shape, device=latents.device, generator=torch_rng, dtype=weight_dtype)    # (b, 96, 21, 30, 52)

                if not args.uniform_sampling:
                    u = compute_density_for_timestep_sampling(
                        weighting_scheme=args.weighting_scheme,
                        batch_size=bsz,
                        logit_mean=args.logit_mean,
                        logit_std=args.logit_std,
                        mode_scale=args.mode_scale,
                    )
                    indices = (u * noise_scheduler.config.num_train_timesteps).long()
                else:
                    # Sample a random timestep for each image
                    # timesteps = generate_timestep_with_lognorm(0, args.train_sampling_steps, (bsz,), device=latents.device, generator=torch_rng)
                    # timesteps = torch.randint(0, args.train_sampling_steps, (bsz,), device=latents.device, generator=torch_rng)
                    indices = idx_sampling(bsz, generator=torch_rng, device=latents.device)
                    indices = indices.long().cpu()
                timesteps = noise_scheduler.timesteps[indices].to(device=latents.device)

                def get_sigmas(timesteps, n_dim=4, dtype=torch.float32):
                    sigmas = noise_scheduler.sigmas.to(device=accelerator.device, dtype=dtype)
                    schedule_timesteps = noise_scheduler.timesteps.to(accelerator.device)
                    timesteps = timesteps.to(accelerator.device)
                    step_indices = [(schedule_timesteps == t).nonzero().item() for t in timesteps]

                    sigma = sigmas[step_indices].flatten()
                    while len(sigma.shape) < n_dim:
                        sigma = sigma.unsqueeze(-1)
                    return sigma

                # Add noise according to flow matching.
                # zt = (1 - texp) * x + texp * z1
                sigmas = get_sigmas(timesteps, n_dim=latents.ndim, dtype=latents.dtype)
                noisy_latents = (1.0 - sigmas) * latents + sigmas * noise[:, :channel, :, :, :]                    
                noisy_motion_latents = (1.0 - sigmas) * motion_latents + sigmas * noise[:, channel:, :, :, :]

                # Add noise
                target_rgb = noise[:, :channel, :, :, :] - latents
                target_motion = noise[:, channel:, :, :, :] - motion_latents
                target = torch.cat([target_rgb, target_motion], dim=1)
                
                target_shape = (channel*2, num_frames, width, height)
                seq_len = math.ceil(
                    (target_shape[2] * target_shape[3]) /
                    (accelerator.unwrap_model(transformer3d).config.patch_size[1] * accelerator.unwrap_model(transformer3d).config.patch_size[2]) *
                    target_shape[1]
                )

                if args.train_mode == "ti2v":
                    t2v_in_ti2v = 1 # first image condition

                    mask_bs = mask.size()[0]
                    if t2v_in_ti2v:
                        noisy_latents = (1 - mask) * rgb_inpaint_latents + mask * noisy_latents
                        noisy_motion_latents = (1 - mask) * motion_inpaint_latents + mask * noisy_motion_latents
                        
                        temp_ts = (mask[:, 0, :, ::2, ::2] * timesteps[:, None, None, None]).flatten(1)
                        timesteps = torch.cat([temp_ts, temp_ts.new_ones(mask_bs, seq_len - temp_ts.size(1)) * timesteps[:, None,]], dim = 1)
                    else:
                        timesteps = mask.new_ones(mask_bs, seq_len) * timesteps[:, None,]

                # Predict the noise residual
                with torch.cuda.amp.autocast(dtype=weight_dtype), torch.cuda.device(device=accelerator.device):
                    if args.motion_only:
                        noise_pred = transformer3d.forward_motion_only(
                            x=[noisy_latents, noisy_motion_latents],
                            context=prompt_embeds,
                            t=timesteps,
                            seq_len=seq_len,
                            y=None
                        )
                    else:
                        noise_pred, smpl_pred = transformer3d(
                            x=[noisy_latents, noisy_motion_latents],
                            context=prompt_embeds,
                            t=timesteps,
                            seq_len=seq_len,
                            y=None,
                            init_smpl_pose=smplpose_values[:, :1].repeat(1, 80, 1, 1).flatten(2) if args.predict_smpl else None,
                            layer_predict_smpl_idx=layer_smpl_predict_idx if args.predict_smpl else None
                        )
                        # noise_pred: (1, 96, 21, 44, 80)
                        # smpl_pred: (1, args.predict_smpl_num, 80, 24*6)
                
                def custom_mse_loss(noise_pred, target, weighting=None, threshold=50):
                    noise_pred = noise_pred.float()
                    target = target.float()
                    diff = noise_pred - target
                    mse_loss = F.mse_loss(noise_pred, target, reduction='none')
                    mask = (diff.abs() <= threshold).float()
                    masked_loss = mse_loss * mask
                    if weighting is not None:
                        masked_loss = masked_loss * weighting
                    final_loss = masked_loss.mean()
                    return final_loss
                
                weighting = compute_loss_weighting_for_sd3(weighting_scheme=args.weighting_scheme, sigmas=sigmas)
                rgb_loss = custom_mse_loss(noise_pred[:, :channel, 1:, :, :].float(), target[:, :channel, 1:, :, :].float(), weighting.float())
                motion_loss = custom_mse_loss(noise_pred[:, channel:, 1:, :, :].float(), target[:, channel:, 1:, :, :].float(), weighting.float())
                loss = args.rgb_loss_weight * rgb_loss.mean() + args.motion_loss_weight * motion_loss.mean()
                
                if args.predict_smpl:
                    pose_loss = custom_mse_loss(smpl_pred.float(), smplpose_values.unsqueeze(1)[:, :, 1:].flatten(3).float(), weighting.float())
                    loss += args.pose_loss_weight * pose_loss.mean()
                    
                    smpl_bs, smpl_f, _ = smplshape_values.shape
                    smpl_ln = smpl_pred.shape[1]
                    smplshape_values = smplshape_values[:, 1:].view(smpl_bs*(smpl_f-1), 10)                             # (80, 10)
                    smplpose_values = rotation_6d_to_matrix(smplpose_values[:, 1:].view(smpl_bs*(smpl_f-1), 24, 6))     # (80, 24, 3, 3)
                    smpl_gt_output = body_model(
                        global_orient=smplpose_values[:, :1, :, :],
                        betas=smplshape_values,
                        body_pose=smplpose_values[:, 1:, :, :]
                    )
                    smpl_gt_vertices = smpl_gt_output.vertices.view(smpl_bs, smpl_f-1, 6890, 3).unsqueeze(1)            # (1, 1, 80, 6890, 3)
                    smpl_gt_joints = smpl_gt_output.joints[:, :24, :] - smpl_gt_output.joints[:, :1, :]      
                    smpl_gt_joints = smpl_gt_joints.view(smpl_bs, smpl_f-1, 24, 3).unsqueeze(1)                         # (1, 1, 80, 24, 3)

                    smpl_pred = rotation_6d_to_matrix(smpl_pred.view(smpl_bs*smpl_ln*(smpl_f-1), 24, 6))                # (1*30*80, 24, 3, 3)
                    smpl_pred_output = body_model(
                        global_orient=smpl_pred[:, :1, :, :],
                        betas = smplshape_values.repeat(smpl_ln, 1),
                        body_pose = smpl_pred[:, 1:, :, :]
                    )
                    smpl_pred_vertices = smpl_pred_output.vertices.view(smpl_bs, smpl_ln, smpl_f-1, 6890, 3)            # (1, 30, 80, 6890, 3)
                    smpl_pred_joints = smpl_pred_output.joints[:, :24, :] - smpl_pred_output.joints[:, :1, :]
                    smpl_pred_joints = smpl_pred_joints.view(smpl_bs, smpl_ln, smpl_f-1, 24, 3)
                    
                    vertices_loss = custom_mse_loss(smpl_pred_vertices.float(), smpl_gt_vertices.float(), weighting.float())
                    loss += args.vertices_loss_weight * vertices_loss.mean()
                    
                    joint3d_loss = custom_mse_loss(smpl_pred_joints.float(), smpl_gt_joints.float(), weighting.float())
                    loss += args.joint3d_loss_weight * joint3d_loss.mean()
    
                if args.motion_sub_loss and noise_pred.size()[2] > 2:
                    gt_sub_noise = noise_pred[:, :, 1:].float() - noise_pred[:, :, :-1].float()
                    pre_sub_noise = target[:, :, 1:].float() - target[:, :, :-1].float()
                    sub_loss = F.mse_loss(gt_sub_noise, pre_sub_noise, reduction="mean")
                    loss = loss * (1 - args.motion_sub_loss_ratio) + sub_loss * args.motion_sub_loss_ratio

                # Gather the losses across all processes for logging (if we use distributed training).
                avg_loss = accelerator.gather(loss.repeat(args.train_batch_size)).mean()
                avg_rgb_loss = accelerator.gather(rgb_loss.repeat(args.train_batch_size)).mean()
                avg_motion_loss = accelerator.gather(motion_loss.repeat(args.train_batch_size)).mean()
                train_loss_log += avg_loss.item() / args.gradient_accumulation_steps
                rgb_loss_log += avg_rgb_loss.item() / args.gradient_accumulation_steps
                motion_loss_log += avg_motion_loss.item() / args.gradient_accumulation_steps
                if args.predict_smpl:
                    avg_pose_loss = accelerator.gather(pose_loss.repeat(args.train_batch_size)).mean()
                    avg_vertices_loss = accelerator.gather(vertices_loss.repeat(args.train_batch_size)).mean()
                    avg_joint3d_loss = accelerator.gather(joint3d_loss.repeat(args.train_batch_size)).mean()
                    pose_loss_log += avg_pose_loss.item() / args.gradient_accumulation_steps
                    vertices_loss_log += avg_vertices_loss.item() / args.gradient_accumulation_steps
                    joint3d_loss_log += avg_joint3d_loss.item() / args.gradient_accumulation_steps

                # Backpropagate
                accelerator.backward(loss)
                if accelerator.sync_gradients:
                    if not args.use_deepspeed and not args.use_fsdp:
                        trainable_params_grads = [p.grad for p in trainable_params if p.grad is not None]
                        trainable_params_total_norm = torch.norm(torch.stack([torch.norm(g.detach(), 2) for g in trainable_params_grads]), 2)
                        max_grad_norm = linear_decay(args.max_grad_norm * args.initial_grad_norm_ratio, args.max_grad_norm, args.abnormal_norm_clip_start, global_step)
                        if trainable_params_total_norm / max_grad_norm > 5 and global_step > args.abnormal_norm_clip_start:
                            actual_max_grad_norm = max_grad_norm / min((trainable_params_total_norm / max_grad_norm), 10)
                        else:
                            actual_max_grad_norm = max_grad_norm
                    else:
                        actual_max_grad_norm = args.max_grad_norm

                    if not args.use_deepspeed and not args.use_fsdp and args.report_model_info and accelerator.is_main_process:
                        if trainable_params_total_norm > 1 and global_step > args.abnormal_norm_clip_start:
                            for name, param in transformer3d.named_parameters():
                                if param.requires_grad:
                                    writer.add_scalar(f'gradients/before_clip_norm/{name}', param.grad.norm(), global_step=global_step)

                    norm_sum = accelerator.clip_grad_norm_(trainable_params, actual_max_grad_norm)
                    if not args.use_deepspeed and not args.use_fsdp and args.report_model_info and accelerator.is_main_process:
                        writer.add_scalar(f'gradients/norm_sum', norm_sum, global_step=global_step)
                        writer.add_scalar(f'gradients/actual_max_grad_norm', actual_max_grad_norm, global_step=global_step)
                optimizer.step()
                lr_scheduler.step()
                optimizer.zero_grad()

            # Checks if the accelerator has performed an optimization step behind the scenes
            if accelerator.sync_gradients:

                if args.use_ema:
                    ema_transformer3d.step(transformer3d.parameters())
                
                progress_bar.update(1)
                global_step += 1
                log_dict = {
                    "loss/total_loss": train_loss_log,
                    "loss/rgb_loss": rgb_loss_log,
                    "loss/motion_loss": motion_loss_log,
                    "loss/pose_loss": pose_loss_log if args.predict_smpl else 0,
                    "loss/vertices_loss": vertices_loss_log if args.predict_smpl else 0,
                    "loss/joint3d_loss": joint3d_loss_log if args.predict_smpl else 0,
                    "metric/gradient_norm": norm_sum,
                    "metric/rgb_loss_weight": args.rgb_loss_weight,
                    "metric/motion_loss_weight": args.motion_loss_weight,
                    "metric/pose_loss_weight": args.pose_loss_weight,
                    "metric/vertices_loss_weight": args.vertices_loss_weight,
                    "metric/joint3d_loss_weight": args.joint3d_loss_weight
                }
                for pg in optimizer.param_groups:
                    log_dict.update({f"learning_rate/{pg['name']}": pg["lr"]})
                accelerator.log(log_dict, step=global_step)
                train_loss_log = 0.0; rgb_loss_log = 0; motion_loss_log = 0
                pose_loss_log = 0.0; vertices_loss_log = 0.0; joint3d_loss_log = 0

                if global_step % args.checkpointing_steps == 0:
                    if args.use_deepspeed or args.use_fsdp or accelerator.is_main_process:
                        # _before_ saving state, check if this save would set us over the `checkpoints_total_limit`
                        if args.checkpoints_total_limit is not None and accelerator.is_main_process:
                            checkpoints = os.listdir(os.path.join(args.output_dir, args.tracker_project_name))
                            checkpoints = [d for d in checkpoints if (d.startswith("checkpoint") and (not d.endswith("outputs")))]
                            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, args.tracker_project_name, removing_checkpoint)
                                    shutil.rmtree(removing_checkpoint)

                        gc.collect()
                        torch.cuda.empty_cache()
                        torch.cuda.ipc_collect()
                        save_path = os.path.join(args.output_dir, args.tracker_project_name, f"checkpoint-{global_step}")
                        accelerator.save_state(save_path)
                        logger.info(f"Saved state to {save_path}")

                if accelerator.is_main_process:
                    if args.validation_prompts is not None and global_step % args.validation_steps == 0:
                        if args.use_ema:
                            # Store the UNet parameters temporarily and load the EMA parameters to perform inference.
                            ema_transformer3d.store(transformer3d.parameters())
                            ema_transformer3d.copy_to(transformer3d.parameters())
                        log_validation(
                            vae,
                            text_encoder,
                            tokenizer,
                            transformer3d,
                            args,
                            config,
                            accelerator,
                            weight_dtype,
                            global_step,
                        )
                        if args.use_ema:
                            # Switch back to the original transformer3d parameters.
                            ema_transformer3d.restore(transformer3d.parameters())

            logs = {
                "step_loss": loss.detach().item(),
                "rgb_loss": rgb_loss.detach().item(),
                "motion_loss": motion_loss.detach().item(),
                "pose_loss": pose_loss.detach().item() if args.predict_smpl else 0,
                "vertices_loss": vertices_loss.detach().item() if args.predict_smpl else 0,
                "joint3d_loss": joint3d_loss.detach().item() if args.predict_smpl else 0,
                "lr": lr_scheduler.get_last_lr()[0]
            }
            progress_bar.set_postfix(**logs)

            if global_step >= args.max_train_steps:
                break

        if accelerator.is_main_process:
            if args.validation_prompts is not None and epoch % args.validation_epochs == 0:
                if args.use_ema:
                    # Store the UNet parameters temporarily and load the EMA parameters to perform inference.
                    ema_transformer3d.store(transformer3d.parameters())
                    ema_transformer3d.copy_to(transformer3d.parameters())
                log_validation(
                    vae,
                    text_encoder,
                    tokenizer,
                    transformer3d,
                    args,
                    config,
                    accelerator,
                    weight_dtype,
                    global_step,
                )
                if args.use_ema:
                    # Switch back to the original transformer3d parameters.
                    ema_transformer3d.restore(transformer3d.parameters())

    # Create the pipeline using the trained modules and save it.
    accelerator.wait_for_everyone()
    if accelerator.is_main_process:
        transformer3d = unwrap_model(transformer3d)
        if args.use_ema:
            ema_transformer3d.copy_to(transformer3d.parameters())

    if args.use_deepspeed or args.use_fsdp or accelerator.is_main_process:
        gc.collect()
        torch.cuda.empty_cache()
        torch.cuda.ipc_collect()
        save_path = os.path.join(args.output_dir, f"checkpoint-{global_step}")
        accelerator.save_state(save_path)
        logger.info(f"Saved state to {save_path}")

    accelerator.end_training()


if __name__ == "__main__":
    main()