1534 lines
64 KiB
Python
1534 lines
64 KiB
Python
|
|
import argparse
|
|
import gc
|
|
import logging
|
|
import math
|
|
import os
|
|
import pickle
|
|
import shutil
|
|
import sys
|
|
import random
|
|
import accelerate
|
|
import diffusers
|
|
import numpy as np
|
|
import torch
|
|
import torch.nn.functional as F
|
|
import torch.utils.checkpoint
|
|
import torchvision.transforms.functional as TF
|
|
import transformers
|
|
from accelerate import Accelerator, FullyShardedDataParallelPlugin
|
|
from accelerate.logging import get_logger
|
|
from accelerate.state import AcceleratorState
|
|
from accelerate.utils import ProjectConfiguration, set_seed
|
|
from diffusers import DDIMScheduler, 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 einops import rearrange
|
|
from omegaconf import OmegaConf
|
|
from packaging import version
|
|
from PIL import Image
|
|
from torch import nn
|
|
from torch.distributed.fsdp.fully_sharded_data_parallel import (
|
|
FullOptimStateDictConfig, FullStateDictConfig, ShardedStateDictConfig, ShardedOptimStateDictConfig)
|
|
from torch.utils.data import RandomSampler, ConcatDataset
|
|
from tqdm.auto import tqdm
|
|
from transformers import AutoTokenizer
|
|
from transformers.utils import ContextManagers
|
|
|
|
import datasets
|
|
|
|
from wan.data.portrait_data import LargeScalePortraitVideos
|
|
from wan.models.face_align import FaceAlignment
|
|
from wan.models.face_model import FaceModel
|
|
from wan.models.pdf import FanEncoder
|
|
|
|
from wan.models.portrait_encoder import PortraitEncoder
|
|
|
|
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 wan.models import (AutoencoderKLWan, CLIPModel, WanT5EncoderModel,
|
|
WanTransformer3DModel)
|
|
from wan.pipeline import WanPipeline, WanI2VPipeline
|
|
from wan.utils.discrete_sampler import DiscreteSampling
|
|
from wan.utils.utils import get_image_to_video_latent, save_videos_grid, split_audio_adapter_sequence, \
|
|
split_tensor_with_padding
|
|
|
|
if is_wandb_available():
|
|
import wandb
|
|
|
|
|
|
def find_replacement(a):
|
|
while a > 0:
|
|
if (a - 1) % 4 == 0:
|
|
return a
|
|
a -= 1
|
|
return 0
|
|
|
|
|
|
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 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)
|
|
|
|
|
|
class CombinedModel(nn.Module):
|
|
def __init__(self, transformer3d, portrait_encoder):
|
|
super().__init__()
|
|
self.transformer3d = transformer3d
|
|
self.portrait_encoder = portrait_encoder
|
|
|
|
|
|
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=42, 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(
|
|
"--selective_ac",
|
|
type=float,
|
|
default=0,
|
|
help="Rate for transformer block apply checkpointing.",
|
|
)
|
|
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=None,
|
|
help=("If you want to load the weight from other transformers, input its path."),
|
|
)
|
|
parser.add_argument(
|
|
"--portrait_encoder_path",
|
|
type=str,
|
|
default=None,
|
|
)
|
|
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(
|
|
'--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(
|
|
'--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(
|
|
"--train_mode",
|
|
type=str,
|
|
default="normal",
|
|
help=(
|
|
'The format of training data. Support `"normal"`'
|
|
' (default), `"i2v"`.'
|
|
),
|
|
)
|
|
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`.",
|
|
)
|
|
|
|
parser.add_argument(
|
|
"--pd_fpg_model_path",
|
|
type=str,
|
|
help="the file path of the pretrained face encoder weights.",
|
|
)
|
|
parser.add_argument(
|
|
"--alignment_model_path",
|
|
type=str,
|
|
help="the file path of the pretrained alignment model weights.",
|
|
)
|
|
parser.add_argument(
|
|
"--det_model_path",
|
|
type=str,
|
|
help="the file path of the pretrained det model weights.",
|
|
)
|
|
parser.add_argument(
|
|
"--train_data_square_dir",
|
|
type=str,
|
|
help="the file path of square video dataset"
|
|
)
|
|
parser.add_argument(
|
|
"--train_data_rec_dir",
|
|
type=str,
|
|
help="the file path of rec video dataset"
|
|
)
|
|
parser.add_argument(
|
|
"--train_data_vec_dir",
|
|
type=str,
|
|
help="the file path of vec video dataset"
|
|
)
|
|
|
|
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
|
|
|
|
|
|
class CombinedModel(nn.Module):
|
|
def __init__(self, transformer3d, portrait_encoder):
|
|
super().__init__()
|
|
self.transformer3d = transformer3d
|
|
self.portrait_encoder = portrait_encoder
|
|
|
|
|
|
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.")
|
|
|
|
|
|
# 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))
|
|
print(f"Init rng with seed {args.seed + accelerator.process_index}. Process_index is {accelerator.process_index}")
|
|
|
|
# Handle the repository creation
|
|
if accelerator.is_main_process:
|
|
if args.output_dir is not None:
|
|
os.makedirs(args.output_dir, 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
|
|
vae = AutoencoderKLWan.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()
|
|
|
|
pd_fpg_motion = FanEncoder()
|
|
pd_fpg_checkpoint = torch.load(args.pd_fpg_model_path, map_location="cpu")
|
|
m, u = pd_fpg_motion.load_state_dict(pd_fpg_checkpoint, strict=False)
|
|
print(f"### pd_fpg_motion missing keys: {len(m)}; \n### unexpected keys: {len(u)};")
|
|
pd_fpg_motion = pd_fpg_motion.eval()
|
|
|
|
face_aligner = FaceModel(
|
|
face_alignment_module=FaceAlignment(
|
|
gpu_id=None,
|
|
alignment_model_path=args.alignment_model_path,
|
|
det_model_path=args.det_model_path,
|
|
),
|
|
reset=False,
|
|
)
|
|
|
|
# Get Clip Image Encoder
|
|
if args.train_mode != "normal":
|
|
clip_image_encoder = CLIPModel.from_pretrained(os.path.join(args.pretrained_model_name_or_path, config['image_encoder_kwargs'].get('image_encoder_subpath', 'image_encoder')),)
|
|
clip_image_encoder = clip_image_encoder.eval()
|
|
|
|
portrait_encoder = PortraitEncoder(adapter_in_dim=768, adapter_proj_dim=2048)
|
|
if args.portrait_encoder_path is not None:
|
|
portrait_encoder_state_dict = torch.load(args.portrait_encoder_path, map_location="cpu", weights_only=True)
|
|
proj_prefix = "proj_model."
|
|
mouth_prefix = "mouth_proj_model."
|
|
emo_prefix = "emo_proj_model."
|
|
portrait_encoder_state_dict_sub_proj_state = {}
|
|
portrait_encoder_state_dict_sub_mouth_state = {}
|
|
portrait_encoder_state_dict_sub_emo_state = {}
|
|
for k, v in portrait_encoder_state_dict.items():
|
|
if k.startswith(proj_prefix):
|
|
new_key = k[len(proj_prefix):] # remove prefix + dot
|
|
portrait_encoder_state_dict_sub_proj_state[new_key] = v
|
|
elif k.startswith(mouth_prefix):
|
|
new_key = k[len(mouth_prefix):] # remove prefix + dot
|
|
portrait_encoder_state_dict_sub_mouth_state[new_key] = v
|
|
elif k.startswith(emo_prefix):
|
|
new_key = k[len(emo_prefix):] # remove prefix + dot
|
|
portrait_encoder_state_dict_sub_emo_state[new_key] = v
|
|
portrait_encoder.proj_model.load_state_dict(portrait_encoder_state_dict_sub_proj_state)
|
|
portrait_encoder.mouth_proj_model.load_state_dict(portrait_encoder_state_dict_sub_mouth_state)
|
|
portrait_encoder.emo_proj_model.load_state_dict(portrait_encoder_state_dict_sub_emo_state)
|
|
|
|
|
|
# Get Transformer
|
|
transformer3d = WanTransformer3DModel.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)
|
|
|
|
# Freeze vae and text_encoder and set transformer3d to trainable
|
|
pd_fpg_motion.requires_grad_(False)
|
|
vae.requires_grad_(False)
|
|
text_encoder.requires_grad_(False)
|
|
transformer3d.requires_grad_(False)
|
|
portrait_encoder.requires_grad_(False)
|
|
if args.train_mode != "normal":
|
|
clip_image_encoder.requires_grad_(False)
|
|
|
|
if args.transformer_path is not None:
|
|
transformer_state_dict = torch.load(args.transformer_path, map_location="cpu", weights_only=True)
|
|
transformer_state_dict = transformer_state_dict["state_dict"] if "state_dict" in transformer_state_dict else transformer_state_dict
|
|
m, u = transformer3d.load_state_dict(transformer_state_dict, strict=False)
|
|
print(f"portrait transformer missing keys: {len(m)}, unexpected keys: {len(u)}")
|
|
print(f"portrait transformer missing keys: {m}")
|
|
print(f"portrait transformer unexpected keys: {u}")
|
|
|
|
# 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']
|
|
|
|
combined_model = CombinedModel(transformer3d=transformer3d, portrait_encoder=portrait_encoder)
|
|
combined_model.transformer3d.train()
|
|
# trainable_modules = ["emo_k_proj", "emo_v_proj"]
|
|
trainable_modules = ["attn", "portrait_encoder"]
|
|
for name, param in combined_model.transformer3d.named_parameters():
|
|
if "attn" in name and "blocks" in name:
|
|
param.requires_grad = True
|
|
for name, param in combined_model.portrait_encoder.named_parameters():
|
|
param.requires_grad = True
|
|
|
|
# Create EMA for the transformer3d.
|
|
if args.use_ema:
|
|
if zero_stage == 3:
|
|
raise NotImplementedError("FSDP does not support EMA.")
|
|
|
|
ema_transformer3d = WanTransformer3DModel.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=WanTransformer3DModel, model_config=ema_transformer3d.config)
|
|
|
|
if args.gradient_checkpointing:
|
|
combined_model.transformer3d.enable_gradient_checkpointing()
|
|
elif args.selective_ac > 0:
|
|
from wan.utils.ac_handle import apply_checkpointing, partial
|
|
from wan.models.wan_transformer3d import WanAttentionBlock
|
|
apply_selective_ac = partial(apply_checkpointing, block=WanAttentionBlock)
|
|
apply_selective_ac(combined_model.transformer3d, p=args.selective_ac)
|
|
|
|
# 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
|
|
|
|
trainable_params = list(filter(lambda p: p.requires_grad, combined_model.parameters()))
|
|
trainable_params_optim = [
|
|
{'params': [], 'lr': args.learning_rate},
|
|
{'params': [], 'lr': args.learning_rate / 2},
|
|
]
|
|
in_already = []
|
|
for name, param in combined_model.named_parameters():
|
|
high_lr_flag = False
|
|
if name in in_already:
|
|
continue
|
|
for trainable_module_name in trainable_modules:
|
|
if trainable_module_name in name and ("blocks" in name or "portrait_encoder" 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 and "blocks" in name:
|
|
in_already.append(name)
|
|
trainable_params_optim[1]['params'].append(param)
|
|
if accelerator.is_main_process:
|
|
print(f"Set {name} to lr : {args.learning_rate / 2}")
|
|
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)
|
|
)
|
|
elif accelerator.state.deepspeed_plugin is not None and "optimizer" in accelerator.state.deepspeed_plugin.deepspeed_config:
|
|
from accelerate.utils import DummyOptim
|
|
optimizer = DummyOptim(
|
|
trainable_params_optim,
|
|
lr=args.learning_rate,
|
|
betas=(args.adam_beta1, args.adam_beta2),
|
|
eps=args.adam_epsilon,
|
|
weight_decay=args.adam_weight_decay,
|
|
)
|
|
else:
|
|
print(1/0)
|
|
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
|
|
|
|
# Get the dataset
|
|
train_square_dataset = LargeScalePortraitVideos(
|
|
txt_path=args.train_data_square_dir,
|
|
width=720,
|
|
height=720,
|
|
# width=512,
|
|
# height=512,
|
|
n_sample_frames=args.video_sample_n_frames,
|
|
sample_frame_rate=1,
|
|
enable_inpaint=True,
|
|
face_aligner=face_aligner,
|
|
)
|
|
train_rec_dataset = LargeScalePortraitVideos(
|
|
txt_path=args.train_data_rec_dir,
|
|
# width=1280,
|
|
# height=720,
|
|
width=832,
|
|
height=480,
|
|
n_sample_frames=args.video_sample_n_frames,
|
|
sample_frame_rate=1,
|
|
enable_inpaint=True,
|
|
face_aligner=face_aligner,
|
|
)
|
|
train_vec_dataset = LargeScalePortraitVideos(
|
|
txt_path=args.train_data_vec_dir,
|
|
# width=720,
|
|
# height=1280,
|
|
width=480,
|
|
height=832,
|
|
n_sample_frames=args.video_sample_n_frames,
|
|
sample_frame_rate=1,
|
|
enable_inpaint=True,
|
|
face_aligner=face_aligner,
|
|
)
|
|
train_dataset_list = []
|
|
train_dataset_list.append(train_square_dataset)
|
|
train_dataset_list.append(train_rec_dataset)
|
|
train_dataset_list.append(train_vec_dataset)
|
|
train_dataset_all = ConcatDataset(train_dataset_list)
|
|
train_dataloader = torch.utils.data.DataLoader(
|
|
train_dataset_all,
|
|
batch_size=args.train_batch_size,
|
|
persistent_workers=True if args.dataloader_num_workers != 0 else False,
|
|
num_workers=args.dataloader_num_workers,
|
|
shuffle=True,
|
|
)
|
|
|
|
# Scheduler and math around the number of training steps.
|
|
overrode_max_train_steps = False
|
|
num_update_steps_per_epoch = math.ceil(len(train_dataloader) / args.gradient_accumulation_steps)
|
|
if args.max_train_steps is None:
|
|
args.max_train_steps = args.num_train_epochs * num_update_steps_per_epoch
|
|
overrode_max_train_steps = True
|
|
|
|
if accelerator.state.deepspeed_plugin is not None and "scheduler" in accelerator.state.deepspeed_plugin.deepspeed_config:
|
|
from accelerate.utils import DummyScheduler
|
|
lr_scheduler = DummyScheduler(
|
|
optimizer=optimizer,
|
|
warmup_num_steps=args.lr_warmup_steps * accelerator.num_processes,
|
|
total_num_steps=args.max_train_steps * accelerator.num_processes,
|
|
)
|
|
else:
|
|
print(1/0)
|
|
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`.
|
|
combined_model, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(combined_model, optimizer, train_dataloader, lr_scheduler)
|
|
|
|
if fsdp_stage != 0:
|
|
from functools import partial
|
|
from wan.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)
|
|
pd_fpg_motion.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")
|
|
if args.train_mode != "normal":
|
|
clip_image_encoder.to(accelerator.device if not args.low_vram else "cpu", dtype=weight_dtype)
|
|
|
|
# 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("trainable_modules")
|
|
tracker_config.pop("trainable_modules_low_learning_rate")
|
|
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_all)}")
|
|
logger.info(f" Num Epochs = {args.num_train_epochs}")
|
|
logger.info(f" Instantaneous batch size per device = {args.train_batch_size}")
|
|
logger.info(f" Total train batch size (w. parallel, distributed & accumulation) = {total_batch_size}")
|
|
logger.info(f" Gradient Accumulation steps = {args.gradient_accumulation_steps}")
|
|
logger.info(f" Total optimization steps = {args.max_train_steps}")
|
|
global_step = 0
|
|
first_epoch = 0
|
|
|
|
# Potentially load in the weights and states from a previous save
|
|
if args.resume_from_checkpoint:
|
|
if args.resume_from_checkpoint != "latest":
|
|
path = os.path.basename(args.resume_from_checkpoint)
|
|
else:
|
|
# Get the most recent checkpoint
|
|
dirs = os.listdir(args.output_dir)
|
|
dirs = [d for d in dirs if d.startswith("checkpoint")]
|
|
dirs = sorted(dirs, key=lambda x: int(x.split("-")[1]))
|
|
path = dirs[-1] if len(dirs) > 0 else None
|
|
|
|
if path is None:
|
|
accelerator.print(f"Checkpoint '{args.resume_from_checkpoint}' does not exist. Starting a new training run.")
|
|
args.resume_from_checkpoint = None
|
|
initial_global_step = 0
|
|
else:
|
|
global_step = int(path.split("-")[1])
|
|
initial_global_step = global_step
|
|
first_epoch = global_step // num_update_steps_per_epoch
|
|
accelerator.print(f"Resuming from checkpoint {path}")
|
|
accelerator.load_state(os.path.join(args.output_dir, 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
|
|
|
|
idx_sampling = DiscreteSampling(args.train_sampling_steps, uniform_sampling=args.uniform_sampling)
|
|
|
|
for epoch in range(first_epoch, args.num_train_epochs):
|
|
train_loss = 0.0
|
|
for step, batch in enumerate(train_dataloader):
|
|
with accelerator.accumulate(combined_model):
|
|
|
|
pixel_values = batch["pixel_values"].to(weight_dtype)
|
|
masked_pixel_values = batch["masked_pixel_values"].to(weight_dtype)
|
|
pixel_value_masks = batch["pixel_value_masks"].to(weight_dtype)
|
|
tgt_face_masks = batch["tgt_face_masks"].to(weight_dtype)
|
|
tgt_lip_masks = batch["tgt_lip_masks"].to(weight_dtype)
|
|
clip_pixel_values = batch["clip_pixel_values"].to(weight_dtype)
|
|
emo_list = batch["emo_list"].to(weight_dtype) # [1, 81, 3, 224, 224]
|
|
|
|
# Make the inpaint latents to be zeros.
|
|
if args.train_mode != "normal":
|
|
t2v_flag = [(_mask == 1).all() for _mask in pixel_value_masks]
|
|
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)
|
|
pd_fpg_motion.to(accelerator.device)
|
|
if args.train_mode != "normal":
|
|
clip_image_encoder.to(accelerator.device)
|
|
torch.cuda.empty_cache()
|
|
if not args.enable_text_encoder_in_dataloader:
|
|
text_encoder.to("cpu")
|
|
|
|
with torch.no_grad():
|
|
emo_len = emo_list.size()[1]
|
|
emo_list_pd = []
|
|
for i in range(emo_len):
|
|
input_emo_tensor = emo_list[0, i].unsqueeze(0)
|
|
headpose_emb, eye_embed, emo_embed, mouth_feat = pd_fpg_motion(input_emo_tensor)
|
|
emotion = {
|
|
"headpose_emb": headpose_emb,
|
|
"eye_embed": eye_embed,
|
|
"emo_embed": emo_embed,
|
|
"mouth_feat": mouth_feat,
|
|
}
|
|
emo_list_pd.append(emotion)
|
|
head_emo_feat_list = []
|
|
for emo in emo_list_pd:
|
|
headpose_emb = emo["headpose_emb"]
|
|
eye_embed = emo["eye_embed"]
|
|
emo_embed = emo["emo_embed"]
|
|
mouth_feat = emo["mouth_feat"]
|
|
emo_feat = torch.cat([eye_embed, emo_embed, mouth_feat], dim=1)
|
|
head_emo_feat = torch.cat([headpose_emb, emo_feat], dim=1)
|
|
head_emo_feat_list.append(head_emo_feat)
|
|
head_emo_feat_all = torch.cat(head_emo_feat_list, dim=0).unsqueeze(0)
|
|
num_emo_frames = pixel_values.size()[1]
|
|
torch.cuda.empty_cache()
|
|
|
|
# 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)
|
|
del pixel_values_bs
|
|
if args.low_vram:
|
|
torch.cuda.empty_cache()
|
|
result = torch.cat(new_pixel_values, dim=0)
|
|
del new_pixel_values
|
|
return result
|
|
|
|
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)
|
|
else:
|
|
latents = _batch_encode_vae(pixel_values)
|
|
|
|
if args.train_mode != "normal":
|
|
pixel_value_masks = rearrange(pixel_value_masks, "b f c h w -> b c f h w")
|
|
pixel_value_masks = torch.concat(
|
|
[
|
|
torch.repeat_interleave(pixel_value_masks[:, :, 0:1], repeats=4, dim=2),
|
|
pixel_value_masks[:, :, 1:]
|
|
], dim=2
|
|
)
|
|
pixel_value_masks = pixel_value_masks.view(pixel_value_masks.shape[0], pixel_value_masks.shape[2] // 4, 4, pixel_value_masks.shape[3], pixel_value_masks.shape[4])
|
|
pixel_value_masks = pixel_value_masks.transpose(1, 2)
|
|
pixel_value_masks = resize_mask(1 - pixel_value_masks, latents)
|
|
|
|
# Encode inpaint latents.
|
|
mask_latents = _batch_encode_vae(masked_pixel_values)
|
|
if vae_stream_2 is not None:
|
|
torch.cuda.current_stream().wait_stream(vae_stream_2)
|
|
|
|
inpaint_latents = torch.concat([pixel_value_masks, mask_latents], dim=1)
|
|
inpaint_latents = t2v_flag[:, None, None, None, None] * inpaint_latents
|
|
|
|
clip_context = []
|
|
for clip_pixel_value in clip_pixel_values:
|
|
clip_image = Image.fromarray(np.uint8(clip_pixel_value.float().cpu().numpy()))
|
|
clip_image = TF.to_tensor(clip_image).sub_(0.5).div_(0.5).to(clip_image_encoder.device, weight_dtype)
|
|
_clip_context = clip_image_encoder([clip_image[:, None, :, :]])
|
|
clip_context.append(_clip_context)
|
|
clip_context = torch.cat(clip_context)
|
|
|
|
# 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')
|
|
pd_fpg_motion.to('cpu')
|
|
if args.train_mode != "normal":
|
|
clip_image_encoder.to('cpu')
|
|
torch.cuda.empty_cache()
|
|
if not args.enable_text_encoder_in_dataloader:
|
|
text_encoder.to(accelerator.device)
|
|
|
|
with torch.no_grad():
|
|
prompt_ids = tokenizer(
|
|
batch['text_prompt'],
|
|
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 = torch.randn(latents.size(), device=latents.device, generator=torch_rng, dtype=weight_dtype)
|
|
|
|
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
|
|
|
|
# Add noise
|
|
target = noise - latents
|
|
|
|
target_shape = (vae.latent_channels, num_frames, width, height)
|
|
seq_len = math.ceil(
|
|
(target_shape[2] * target_shape[3]) /
|
|
(accelerator.unwrap_model(combined_model).transformer3d.config.patch_size[1] *
|
|
accelerator.unwrap_model(combined_model).transformer3d.config.patch_size[2]) *
|
|
target_shape[1]
|
|
)
|
|
|
|
# Predict the noise residual
|
|
with torch.cuda.amp.autocast(dtype=weight_dtype), torch.cuda.device(device=accelerator.device):
|
|
|
|
emo_proj = combined_model.portrait_encoder.get_adapter_proj(head_emo_feat_all)
|
|
pos_idx_range = split_audio_adapter_sequence(emo_proj.size(1), num_frames=num_emo_frames)
|
|
emo_proj_split, emo_context_lens = split_tensor_with_padding(emo_proj, pos_idx_range, expand_length=0)
|
|
if random.random() < 0.1:
|
|
emo_proj_split = torch.zeros_like(emo_proj_split)
|
|
|
|
latents_num_frames = latents.size()[2]
|
|
noise_pred = combined_model.transformer3d(
|
|
x=noisy_latents,
|
|
context=prompt_embeds,
|
|
t=timesteps,
|
|
seq_len=seq_len,
|
|
y=inpaint_latents if args.train_mode != "normal" else None,
|
|
clip_fea=clip_context if args.train_mode != "normal" else None,
|
|
emo_proj=emo_proj_split,
|
|
emo_context_lens=emo_context_lens,
|
|
latents_num_frames=latents_num_frames,
|
|
ip_scale=1.0,
|
|
)
|
|
|
|
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_loss_flag = torch.rand(1).item()
|
|
if mask_loss_flag >= 0.5 and mask_loss_flag < 0.7:
|
|
mse_loss = mse_loss * tgt_face_masks
|
|
elif mask_loss_flag >= 0.7:
|
|
mse_loss = mse_loss * tgt_lip_masks
|
|
else:
|
|
mse_loss = mse_loss * (1 + tgt_face_masks + tgt_lip_masks)
|
|
|
|
if weighting is not None:
|
|
mse_loss = mse_loss * weighting
|
|
final_loss = mse_loss.mean()
|
|
return final_loss
|
|
|
|
tgt_face_masks = F.interpolate(tgt_face_masks, size=(target.size()[-3], target.size()[-2], target.size()[-1]), mode='trilinear', align_corners=False)
|
|
tgt_lip_masks = F.interpolate(tgt_lip_masks, size=(target.size()[-3], target.size()[-2], target.size()[-1]), mode='trilinear', align_corners=False)
|
|
|
|
weighting = compute_loss_weighting_for_sd3(weighting_scheme=args.weighting_scheme, sigmas=sigmas)
|
|
loss = custom_mse_loss(noise_pred.float(), target.float(), weighting.float())
|
|
loss = loss.mean()
|
|
|
|
if args.motion_sub_loss and noise_pred.size()[1] > 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()
|
|
train_loss += avg_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
|
|
norm_sum = accelerator.clip_grad_norm_(trainable_params, actual_max_grad_norm)
|
|
|
|
optimizer.step()
|
|
lr_scheduler.step()
|
|
optimizer.zero_grad()
|
|
|
|
if step % 2 == 0:
|
|
gc.collect()
|
|
torch.cuda.empty_cache()
|
|
torch.cuda.ipc_collect()
|
|
|
|
# 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
|
|
accelerator.log({"train_loss": train_loss}, step=global_step)
|
|
train_loss = 0.0
|
|
|
|
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:
|
|
checkpoints = os.listdir(args.output_dir)
|
|
checkpoints = [d for d in checkpoints if d.startswith("checkpoint")]
|
|
checkpoints = sorted(checkpoints, key=lambda x: int(x.split("-")[1]))
|
|
|
|
# before we save the new checkpoint, we need to have at _most_ `checkpoints_total_limit - 1` checkpoints
|
|
if len(checkpoints) >= args.checkpoints_total_limit:
|
|
num_to_remove = len(checkpoints) - args.checkpoints_total_limit + 1
|
|
removing_checkpoints = checkpoints[0:num_to_remove]
|
|
|
|
logger.info(
|
|
f"{len(checkpoints)} checkpoints already exist, removing {len(removing_checkpoints)} checkpoints"
|
|
)
|
|
logger.info(f"removing checkpoints: {', '.join(removing_checkpoints)}")
|
|
|
|
for removing_checkpoint in removing_checkpoints:
|
|
removing_checkpoint = os.path.join(args.output_dir, removing_checkpoint)
|
|
shutil.rmtree(removing_checkpoint)
|
|
|
|
save_path = os.path.join(args.output_dir, f"checkpoint-{global_step}")
|
|
accelerator.save_state(save_path)
|
|
logger.info(f"Saved state to {save_path}")
|
|
torch.cuda.empty_cache()
|
|
|
|
logs = {"step_loss": loss.detach().item(), "lr": lr_scheduler.get_last_lr()[0]}
|
|
del noise_pred, target, loss
|
|
progress_bar.set_postfix(**logs)
|
|
|
|
if global_step >= args.max_train_steps:
|
|
break
|
|
|
|
# Create the pipeline using the trained modules and save it.
|
|
accelerator.wait_for_everyone()
|
|
if accelerator.is_main_process:
|
|
combined_model = unwrap_model(combined_model)
|
|
if args.use_ema:
|
|
ema_transformer3d.copy_to(transformer3d.parameters())
|
|
|
|
if args.use_deepspeed or args.use_fsdp or accelerator.is_main_process:
|
|
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()
|