1485 lines
62 KiB
Python
1485 lines
62 KiB
Python
import argparse
|
|
import gc
|
|
import logging
|
|
import math
|
|
import os
|
|
import pickle
|
|
import random
|
|
import shutil
|
|
import sys
|
|
|
|
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
|
|
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.utils.data import RandomSampler
|
|
from torch.utils.tensorboard import SummaryWriter
|
|
from tqdm.auto import tqdm
|
|
from transformers import AutoTokenizer
|
|
from transformers.utils import ContextManagers
|
|
import datasets
|
|
from transformers import Wav2Vec2Model, Wav2Vec2Processor
|
|
import librosa
|
|
from pathlib import Path
|
|
import imageio
|
|
import torchvision
|
|
|
|
from musicfm.model.musicfm_25hz import MusicFM25Hz
|
|
from wan.dataset.talking_video_dataset_fantasy import LargeScaleTalkingFantasyVideos
|
|
from wan.models.wan_fantasy_transformer3d_14B import WanTransformer3DFantasy14BModel
|
|
from wan.models.wan_text_encoder import WanT5EncoderModel
|
|
from wan.models.wan_vae import AutoencoderKLWan
|
|
from wan.models.wan_image_encoder import CLIPModel
|
|
|
|
from wan.utils.discrete_sampler import DiscreteSampling
|
|
from wan.utils.utils import get_image_to_video_latent, save_videos_grid
|
|
|
|
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 save_videos_from_pil(pil_images, path, fps=8):
|
|
save_fmt = Path(path).suffix
|
|
os.makedirs(os.path.dirname(path), exist_ok=True)
|
|
|
|
if save_fmt == ".mp4":
|
|
with imageio.get_writer(path, fps=fps) as writer:
|
|
for img in pil_images:
|
|
img_array = np.array(img) # Convert PIL Image to numpy array
|
|
writer.append_data(img_array)
|
|
|
|
elif save_fmt == ".gif":
|
|
pil_images[0].save(
|
|
fp=path,
|
|
format="GIF",
|
|
append_images=pil_images[1:],
|
|
save_all=True,
|
|
duration=(1 / fps * 1000),
|
|
loop=0,
|
|
optimize=False,
|
|
lossless=True
|
|
)
|
|
else:
|
|
raise ValueError("Unsupported file type. Use .mp4 or .gif.")
|
|
|
|
|
|
def save_videos_grid_png_and_mp4(videos: torch.Tensor, path: str, rescale=False, n_rows=6, fps=8, save_frames_path=None):
|
|
videos = rearrange(videos, "b c t h w -> t b c h w")
|
|
height, width = videos.shape[-2:]
|
|
outputs = []
|
|
|
|
for i, x in enumerate(videos):
|
|
x = torchvision.utils.make_grid(x, nrow=n_rows) # (c h w)
|
|
x = x.transpose(0, 1).transpose(1, 2).squeeze(-1) # (h w c)
|
|
if rescale:
|
|
x = (x + 1.0) / 2.0 # -1,1 -> 0,1
|
|
x = (x * 255).numpy().astype(np.uint8)
|
|
x = Image.fromarray(x)
|
|
outputs.append(x)
|
|
|
|
pil_frames = [Image.fromarray(frame) if isinstance(frame, np.ndarray) else frame for frame in outputs]
|
|
num_frames = len(pil_frames)
|
|
for i in range(num_frames):
|
|
pil_frame = pil_frames[i]
|
|
save_path = os.path.join(save_frames_path, f'frame_{i}.png')
|
|
pil_frame.save(save_path)
|
|
|
|
os.makedirs(os.path.dirname(path), exist_ok=True)
|
|
|
|
save_videos_from_pil(outputs, path, fps)
|
|
|
|
|
|
def log_validation(vae, text_encoder, tokenizer, clip_image_encoder, transformer3d, config, args, accelerator, weight_dtype,
|
|
global_step, height, width, wav2vec_feature_extractor, wav2vec_model, musicfm):
|
|
|
|
logger.info("Running validation... ")
|
|
scheduler = FlowMatchEulerDiscreteScheduler(
|
|
**filter_kwargs(FlowMatchEulerDiscreteScheduler, OmegaConf.to_container(config['scheduler_kwargs']))
|
|
)
|
|
pipeline = WanI2VInnerPipeline(
|
|
vae=accelerator.unwrap_model(vae).to(weight_dtype),
|
|
text_encoder=accelerator.unwrap_model(text_encoder),
|
|
tokenizer=tokenizer,
|
|
transformer=accelerator.unwrap_model(transformer3d),
|
|
scheduler=scheduler,
|
|
clip_image_encoder=accelerator.unwrap_model(clip_image_encoder),
|
|
)
|
|
pipeline = pipeline.to(accelerator.device)
|
|
if args.seed is None:
|
|
generator = None
|
|
else:
|
|
generator = torch.Generator(device=accelerator.device).manual_seed(args.seed)
|
|
|
|
with torch.no_grad():
|
|
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
|
|
latent_frames = (video_length - 1) // vae.config.temporal_compression_ratio + 1
|
|
|
|
torch.cuda.empty_cache()
|
|
wav2vec_model.to(accelerator.device)
|
|
musicfm.to(accelerator.device)
|
|
|
|
vocal_sample_rate = 16000
|
|
fps = 30
|
|
vocal_speech_array, vocal_sampling_rate = librosa.load(args.validation_driven_audio_path, sr=vocal_sample_rate)
|
|
vocal_feature = np.squeeze(accelerator.unwrap_model(wav2vec_feature_extractor)(vocal_speech_array, sampling_rate=vocal_sample_rate).input_values)
|
|
vocal_seq_len = int(len(vocal_feature) / vocal_sample_rate * fps)
|
|
clip_length = args.video_sample_n_frames
|
|
vocal_feature = torch.from_numpy(vocal_feature).float()
|
|
if clip_length > 0 and vocal_seq_len % clip_length != 0:
|
|
vocal_feature = torch.nn.functional.pad(vocal_feature, (
|
|
0, int((clip_length - vocal_seq_len % clip_length) * (vocal_sample_rate // fps))), 'constant', 0.0)
|
|
vocal_seq_len += clip_length - vocal_seq_len % clip_length
|
|
vocal_feature = vocal_feature.unsqueeze(0)
|
|
with torch.no_grad():
|
|
vocal_embeddings = accelerator.unwrap_model(wav2vec_model)(vocal_feature.to(accelerator.device), seq_len=vocal_seq_len, output_hidden_states=True)
|
|
assert len(vocal_embeddings) > 0, "Fail to extract audio embedding"
|
|
vocal_embeddings = torch.stack(vocal_embeddings.hidden_states[1:], dim=1).squeeze(0)
|
|
vocal_embeddings = rearrange(vocal_embeddings, "b s d -> s b d")
|
|
concatenated_tensors = []
|
|
for i in range(vocal_embeddings.shape[0]):
|
|
vectors_to_concat = [vocal_embeddings[max(min(i + j, vocal_embeddings.shape[0] - 1), 0)] for j in range(-2, 3)]
|
|
concatenated_tensors.append(torch.stack(vectors_to_concat, dim=0))
|
|
vocal_embeddings = torch.stack(concatenated_tensors, dim=0)
|
|
|
|
audio_sample_rate = 24000
|
|
audio_speech_array, audio_sampling_rate = librosa.load(args.validation_driven_audio_path, sr=audio_sample_rate)
|
|
selected_audio_array = audio_speech_array.reshape(1, -1)
|
|
selected_audio_tensor = torch.from_numpy(selected_audio_array).float().reshape(1, -1)
|
|
audio_embeddings = accelerator.unwrap_model(musicfm).get_latent(selected_audio_tensor.to(accelerator.device), layer_ix=7)
|
|
audio_embeddings = nn.AdaptiveAvgPool1d(video_length)(audio_embeddings.transpose(1, 2))
|
|
audio_embeddings = audio_embeddings.transpose(1, 2)
|
|
audio_embeddings = audio_embeddings.squeeze(0)
|
|
concatenated_tensors = []
|
|
for i in range(audio_embeddings.shape[0]):
|
|
vectors_to_concat = [audio_embeddings[max(min(i + j, audio_embeddings.shape[0] - 1), 0)] for j in range(-2, 3)]
|
|
concatenated_tensors.append(torch.stack(vectors_to_concat, dim=0))
|
|
audio_embeddings = torch.stack(concatenated_tensors, dim=0)
|
|
audio_embeddings = audio_embeddings.unsqueeze(0)
|
|
|
|
wav2vec_model.to("cpu")
|
|
musicfm.to("cpu")
|
|
torch.cuda.empty_cache()
|
|
|
|
window_size = 0.2
|
|
half_window = window_size / 2
|
|
audio_strength_features = []
|
|
for i in range(video_length):
|
|
center_time = i / fps
|
|
start_sample = int(max((center_time - half_window) * audio_sample_rate, 0))
|
|
end_sample = int(min((center_time + half_window) * audio_sample_rate, len(audio_speech_array)))
|
|
y_slice = audio_speech_array[start_sample:end_sample]
|
|
if len(y_slice) < int(window_size * audio_sample_rate * 0.5):
|
|
# padding if slice too short
|
|
y_slice = np.pad(y_slice, (0, int(window_size * audio_sample_rate) - len(y_slice)), mode='constant')
|
|
rms = librosa.feature.rms(y=y_slice).flatten()
|
|
rms_feat = np.array([np.mean(rms), np.std(rms)])
|
|
mfcc = librosa.feature.mfcc(y=y_slice, sr=audio_sample_rate, n_mfcc=13)
|
|
mfcc_mean = np.mean(mfcc, axis=1)
|
|
mfcc_std = np.std(mfcc, axis=1)
|
|
mfcc_feat = np.concatenate([mfcc_mean, mfcc_std])
|
|
frame_feat = np.concatenate([rms_feat, mfcc_feat])
|
|
audio_strength_features.append(frame_feat)
|
|
audio_strength_features = np.stack(audio_strength_features)
|
|
audio_strength_features = torch.from_numpy(audio_strength_features).float()
|
|
|
|
input_video, input_video_mask, clip_image = get_image_to_video_latent(args.validation_reference_path, None, video_length=video_length, sample_size=[height, width])
|
|
|
|
text_prompt = "The protagonist is singing"
|
|
num_inference_steps = 50
|
|
sample = pipeline(
|
|
text_prompt,
|
|
num_frames=video_length,
|
|
negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
|
|
height=height,
|
|
width=width,
|
|
guidance_scale=6.0,
|
|
generator=generator,
|
|
num_inference_steps=num_inference_steps,
|
|
|
|
video=input_video,
|
|
mask_video=input_video_mask,
|
|
clip_image=clip_image,
|
|
vocal_embeddings=vocal_embeddings,
|
|
audio_embeddings=audio_embeddings,
|
|
audio_strength_features=audio_strength_features,
|
|
multiple_cfg_scale=[6.0, 2.0, 4.5],
|
|
).videos
|
|
os.makedirs(os.path.join(args.output_dir, "sample"), exist_ok=True)
|
|
fps = 16
|
|
video_path = os.path.join(args.output_dir, f"sample/sample-{global_step}.mp4")
|
|
save_videos_grid(sample, video_path, fps=fps)
|
|
|
|
del pipeline
|
|
gc.collect()
|
|
torch.cuda.empty_cache()
|
|
torch.cuda.ipc_collect()
|
|
|
|
|
|
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(
|
|
"--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(
|
|
"--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(
|
|
"--rank",
|
|
type=int,
|
|
default=128,
|
|
help=("The dimension of the LoRA update matrices."),
|
|
)
|
|
parser.add_argument(
|
|
"--network_alpha",
|
|
type=int,
|
|
default=64,
|
|
help=("The dimension of the LoRA update matrices."),
|
|
)
|
|
parser.add_argument(
|
|
"--train_text_encoder",
|
|
action="store_true",
|
|
help="Whether to train the text encoder. If set, the text encoder should be float32 precision.",
|
|
)
|
|
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(
|
|
"--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(
|
|
"--noise_share_in_frames", action="store_true", help="Whether enable noise share in frames."
|
|
)
|
|
parser.add_argument(
|
|
"--noise_share_in_frames_ratio", type=float, default=0.5, help="Noise share ratio.",
|
|
)
|
|
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(
|
|
"--keep_all_node_same_token_length",
|
|
action="store_true",
|
|
help="Reference of the length token.",
|
|
)
|
|
parser.add_argument(
|
|
"--train_sampling_steps",
|
|
type=int,
|
|
default=1000,
|
|
help="Run train_sampling_steps.",
|
|
)
|
|
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(
|
|
"--image_repeat_in_forward",
|
|
type=int,
|
|
default=0,
|
|
help="Num of repeat image in forward.",
|
|
)
|
|
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(
|
|
"--vae_path",
|
|
type=str,
|
|
default=None,
|
|
help=("If you want to load the weight from other vaes, input its path."),
|
|
)
|
|
parser.add_argument("--save_state", action="store_true", help="Whether or not to save state.")
|
|
|
|
parser.add_argument(
|
|
'--tokenizer_max_length',
|
|
type=int,
|
|
default=226,
|
|
help='Max length of tokenizer'
|
|
)
|
|
parser.add_argument(
|
|
"--use_deepspeed", action="store_true", help="Whether or not to use deepspeed."
|
|
)
|
|
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), `"inpaint"`.'
|
|
),
|
|
)
|
|
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(
|
|
"--vocal_window_size",
|
|
type=int,
|
|
default=5,
|
|
)
|
|
parser.add_argument(
|
|
"--pretrained_wav2vec_path",
|
|
type=str,
|
|
)
|
|
parser.add_argument(
|
|
"--pretrained_musicfm_stat_path",
|
|
type=str,
|
|
)
|
|
parser.add_argument(
|
|
"--pretrained_musicfm_model_path",
|
|
type=str,
|
|
)
|
|
parser.add_argument(
|
|
"--validation_reference_path",
|
|
type=str,
|
|
)
|
|
parser.add_argument(
|
|
"--validation_driven_audio_path",
|
|
type=str,
|
|
)
|
|
parser.add_argument(
|
|
"--train_data_rec_dir",
|
|
type=str,
|
|
default=None,
|
|
help=(
|
|
"A folder containing the training data. "
|
|
),
|
|
)
|
|
parser.add_argument(
|
|
"--train_data_square_dir",
|
|
type=str,
|
|
default=None,
|
|
help=(
|
|
"A folder containing the training data. "
|
|
),
|
|
)
|
|
parser.add_argument(
|
|
"--train_data_vec_dir",
|
|
type=str,
|
|
default=None,
|
|
help=(
|
|
"A folder containing the training data. "
|
|
),
|
|
)
|
|
|
|
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
|
|
|
|
|
|
# This is for training using deepspeed.
|
|
# Since now the DeepSpeed only supports trainging with only one model
|
|
# So we create a virtual wrapper to contail all the models
|
|
class DeepSpeedWrapperModel(nn.Module):
|
|
def __init__(self, **kwargs):
|
|
super().__init__()
|
|
for name, value in kwargs.items():
|
|
assert isinstance(value, nn.Module)
|
|
self.register_module(name, value)
|
|
|
|
|
|
class CombinedModel(nn.Module):
|
|
def __init__(self, transformer3d, lora_network):
|
|
super().__init__()
|
|
self.transformer3d = transformer3d
|
|
self.lora_network = lora_network
|
|
|
|
# Function for unwrapping if model was compiled with `torch.compile`.
|
|
def unwrap_model(accelerator, model):
|
|
model = accelerator.unwrap_model(model)
|
|
model = model._orig_mod if is_compiled_module(model) else model
|
|
return model
|
|
|
|
def main():
|
|
args = parse_args()
|
|
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 deepspeed_plugin is not None:
|
|
zero_stage = int(deepspeed_plugin.zero_stage)
|
|
print(f"Using DeepSpeed Zero stage: {zero_stage}")
|
|
else:
|
|
zero_stage = 0
|
|
print("DeepSpeed is not enabled.")
|
|
if accelerator.is_main_process:
|
|
writer = SummaryWriter(log_dir=logging_dir)
|
|
|
|
# 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 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')), )
|
|
|
|
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()
|
|
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']),
|
|
)
|
|
wav2vec_processor = Wav2Vec2Processor.from_pretrained(args.pretrained_wav2vec_path)
|
|
wav2vec = Wav2Vec2Model.from_pretrained(args.pretrained_wav2vec_path).to("cpu")
|
|
|
|
transformer3d = WanTransformer3DFantasy14BModel.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)
|
|
|
|
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()
|
|
|
|
vae.requires_grad_(False)
|
|
text_encoder.requires_grad_(False)
|
|
transformer3d.requires_grad_(False)
|
|
clip_image_encoder.requires_grad_(False)
|
|
wav2vec.requires_grad_(False)
|
|
|
|
transformer3d.train()
|
|
if args.gradient_checkpointing:
|
|
transformer3d.enable_gradient_checkpointing()
|
|
|
|
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_optim = []
|
|
for name, para in transformer3d.named_parameters():
|
|
if "vocal_projector" in name or "audio_projector" in name or "vocal" in name or "audio" in name or "attn" in name or "blocks" in name:
|
|
para.requires_grad = True
|
|
trainable_params_optim.append({"params": para, "lr": args.learning_rate})
|
|
trainable_params = list(filter(lambda p: p.requires_grad, transformer3d.parameters()))
|
|
|
|
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:
|
|
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,
|
|
)
|
|
print(1 / 0)
|
|
|
|
sample_n_frames_bucket_interval = vae.config.temporal_compression_ratio
|
|
|
|
train_rec_dataset = LargeScaleTalkingFantasyVideos(
|
|
txt_path=args.train_data_rec_dir,
|
|
# width=1280,
|
|
# height=720,
|
|
width=832,
|
|
height=480,
|
|
n_sample_frames=args.video_sample_n_frames,
|
|
wav2vec_processor=wav2vec_processor,
|
|
sample_frame_rate=1,
|
|
only_last_features=False,
|
|
vocal_sample_rate=16000,
|
|
audio_sample_rate=24000,
|
|
enable_inpaint=True,
|
|
vae_stride=(4, 8, 8),
|
|
patch_size=(1, 2, 2)
|
|
)
|
|
train_rec_dataloader = torch.utils.data.DataLoader(
|
|
train_rec_dataset,
|
|
batch_size=args.train_batch_size,
|
|
num_workers=args.dataloader_num_workers,
|
|
shuffle=True,
|
|
)
|
|
train_square_dataset = LargeScaleTalkingFantasyVideos(
|
|
txt_path=args.train_data_square_dir,
|
|
width=512,
|
|
height=512,
|
|
n_sample_frames=args.video_sample_n_frames,
|
|
wav2vec_processor=wav2vec_processor,
|
|
sample_frame_rate=1,
|
|
only_last_features=False,
|
|
vocal_sample_rate=16000,
|
|
audio_sample_rate=24000,
|
|
enable_inpaint=True,
|
|
vae_stride=(4, 8, 8),
|
|
patch_size=(1, 2, 2)
|
|
)
|
|
train_square_dataloader = torch.utils.data.DataLoader(
|
|
train_square_dataset,
|
|
batch_size=args.train_batch_size,
|
|
num_workers=args.dataloader_num_workers,
|
|
shuffle=True,
|
|
)
|
|
train_vec_dataset = LargeScaleTalkingFantasyVideos(
|
|
txt_path=args.train_data_vec_dir,
|
|
# width=720,
|
|
# height=1280,
|
|
width=480,
|
|
height=832,
|
|
n_sample_frames=args.video_sample_n_frames,
|
|
wav2vec_processor=wav2vec_processor,
|
|
sample_frame_rate=1,
|
|
only_last_features=False,
|
|
vocal_sample_rate=16000,
|
|
audio_sample_rate=24000,
|
|
enable_inpaint=True,
|
|
vae_stride=(4, 8, 8),
|
|
patch_size=(1, 2, 2)
|
|
)
|
|
train_vec_dataloader = torch.utils.data.DataLoader(
|
|
train_vec_dataset,
|
|
batch_size=args.train_batch_size,
|
|
num_workers=args.dataloader_num_workers,
|
|
shuffle=True,
|
|
)
|
|
|
|
overrode_max_train_steps = False
|
|
num_update_steps_per_epoch = math.ceil((len(train_rec_dataloader)+len(train_square_dataloader)+len(train_vec_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:
|
|
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,
|
|
)
|
|
print(1 / 0)
|
|
|
|
transformer3d, optimizer, train_rec_dataloader, train_square_dataloader, train_vec_dataloader, lr_scheduler = accelerator.prepare(transformer3d, optimizer, train_rec_dataloader, train_square_dataloader, train_vec_dataloader, lr_scheduler)
|
|
|
|
vae.to(accelerator.device, dtype=weight_dtype)
|
|
clip_image_encoder.to(accelerator.device, dtype=weight_dtype)
|
|
text_encoder.to(accelerator.device)
|
|
|
|
num_update_steps_per_epoch = math.ceil((len(train_rec_dataloader)+len(train_square_dataloader)+len(train_vec_dataloader)) / args.gradient_accumulation_steps)
|
|
if overrode_max_train_steps:
|
|
args.max_train_steps = args.num_train_epochs * num_update_steps_per_epoch
|
|
# Afterwards we recalculate our number of training epochs
|
|
args.num_train_epochs = math.ceil(args.max_train_steps / num_update_steps_per_epoch)
|
|
|
|
# We need to initialize the trackers we use, and also store our configuration.
|
|
# The trackers initializes automatically on the main process.
|
|
if accelerator.is_main_process:
|
|
accelerator.init_trackers("Talking_Face", config=vars(args))
|
|
|
|
|
|
total_batch_size = args.train_batch_size * accelerator.num_processes * args.gradient_accumulation_steps
|
|
|
|
len_zero = len(train_rec_dataloader)
|
|
len_one = len(train_square_dataloader)
|
|
len_two = len(train_vec_dataloader)
|
|
|
|
logger.info("***** Running training *****")
|
|
logger.info(f" Num examples = {len(train_rec_dataset)+len(train_square_dataset)+len(train_vec_dataset)}")
|
|
logger.info(f" Num Epochs = {args.num_train_epochs}")
|
|
logger.info(f" Instantaneous batch size per device = {args.train_batch_size}")
|
|
logger.info(f" Total train batch size (w. parallel, distributed & accumulation) = {total_batch_size}")
|
|
logger.info(f" Gradient Accumulation steps = {args.gradient_accumulation_steps}")
|
|
logger.info(f" Total optimization steps = {args.max_train_steps}")
|
|
global_step = 0
|
|
first_epoch = 0
|
|
|
|
# Potentially load in the weights and states from a previous save
|
|
if args.resume_from_checkpoint:
|
|
if args.resume_from_checkpoint != "latest":
|
|
path = os.path.basename(args.resume_from_checkpoint)
|
|
else:
|
|
# Get the most recent checkpoint
|
|
dirs = os.listdir(args.output_dir)
|
|
dirs = [d for d in dirs if d.startswith("checkpoint")]
|
|
dirs = sorted(dirs, key=lambda x: int(x.split("-")[1]))
|
|
path = dirs[-1] if len(dirs) > 0 else None
|
|
|
|
if path is None:
|
|
accelerator.print(
|
|
f"Checkpoint '{args.resume_from_checkpoint}' does not exist. Starting a new training run."
|
|
)
|
|
args.resume_from_checkpoint = None
|
|
initial_global_step = 0
|
|
else:
|
|
global_step = int(path.split("-")[1])
|
|
initial_global_step = global_step
|
|
first_epoch = global_step // num_update_steps_per_epoch
|
|
print(f"Resuming from checkpoint {path}. Get first_epoch = {first_epoch}.")
|
|
accelerator.load_state(os.path.join(args.output_dir, path))
|
|
else:
|
|
initial_global_step = 0
|
|
|
|
# function for saving/removing
|
|
def save_model(ckpt_file, unwrapped_nw):
|
|
os.makedirs(args.output_dir, exist_ok=True)
|
|
accelerator.print(f"\nsaving checkpoint: {ckpt_file}")
|
|
unwrapped_nw.save_weights(ckpt_file, weight_dtype, None)
|
|
|
|
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
|
|
|
|
iter1 = iter(train_rec_dataloader)
|
|
iter2 = iter(train_square_dataloader)
|
|
iter3 = iter(train_vec_dataloader)
|
|
iter_list = [0] * len_zero + [1] * len_one + [2] * len_two
|
|
random.shuffle(iter_list)
|
|
|
|
for step in range(0, len(iter_list)):
|
|
|
|
current_idx = iter_list[step]
|
|
if current_idx == 0:
|
|
try:
|
|
batch = next(iter1)
|
|
except StopIteration:
|
|
iter1 = iter(train_rec_dataloader)
|
|
batch = next(iter1)
|
|
elif current_idx == 1:
|
|
try:
|
|
batch = next(iter2)
|
|
except StopIteration:
|
|
iter2 = iter(train_square_dataloader)
|
|
batch = next(iter2)
|
|
elif current_idx == 2:
|
|
try:
|
|
batch = next(iter3)
|
|
except StopIteration:
|
|
iter3 = iter(train_vec_dataloader)
|
|
batch = next(iter3)
|
|
else:
|
|
batch = None
|
|
print(1 / 0)
|
|
|
|
with accelerator.accumulate(transformer3d):
|
|
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)
|
|
clip_pixel_values = batch["clip_pixel_values"].to(weight_dtype)
|
|
vocal_input_values = batch["vocal_input_values"].squeeze(0)
|
|
tgt_face_masks = batch["tgt_face_masks"].to(weight_dtype)
|
|
tgt_lip_masks = batch["tgt_lip_masks"].to(weight_dtype)
|
|
|
|
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)
|
|
clip_image_encoder.to(accelerator.device)
|
|
text_encoder.to("cpu")
|
|
wav2vec.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)
|
|
else:
|
|
latents = _batch_encode_vae(pixel_values)
|
|
|
|
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)
|
|
|
|
masked_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, masked_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, :, :]])
|
|
if rng is None:
|
|
zero_init_clip_in = np.random.choice([True, False], p=[0.1, 0.9])
|
|
else:
|
|
zero_init_clip_in = rng.choice([True, False], p=[0.1, 0.9])
|
|
clip_context.append(_clip_context if not zero_init_clip_in else torch.zeros_like(_clip_context))
|
|
clip_context = torch.cat(clip_context)
|
|
|
|
if vae_stream_1 is not None:
|
|
torch.cuda.current_stream().wait_stream(vae_stream_1)
|
|
|
|
if args.low_vram:
|
|
vae.to('cpu')
|
|
clip_image_encoder.to('cpu')
|
|
torch.cuda.empty_cache()
|
|
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]
|
|
# print(f"The size of prompt_embeds: {prompt_embeds.size()}") # [1, 226, 4096]
|
|
# print(f"seq_lens: {seq_lens}") # [1]
|
|
prompt_embeds = [u[:v] for u, v in zip(prompt_embeds, seq_lens)]
|
|
|
|
if args.low_vram:
|
|
text_encoder.to('cpu')
|
|
torch.cuda.empty_cache()
|
|
wav2vec.to(accelerator.device)
|
|
|
|
with torch.no_grad():
|
|
try:
|
|
vocal_embeddings = wav2vec(vocal_input_values).last_hidden_state
|
|
except Exception as e:
|
|
audio_path = batch["audio_path"]
|
|
print(f"{audio_path} cannot be loaded !!!")
|
|
vocal_embeddings = torch.zeros([1, 168, 768])
|
|
if random.random() < 0.1:
|
|
vocal_embeddings = torch.zeros_like(vocal_embeddings)
|
|
if random.random() < 0.3:
|
|
is_clip_level_modeling = True
|
|
else:
|
|
is_clip_level_modeling = False
|
|
|
|
if args.low_vram:
|
|
wav2vec.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(transformer3d).config.patch_size[1] *
|
|
accelerator.unwrap_model(transformer3d).config.patch_size[2]) *
|
|
target_shape[1]
|
|
)
|
|
|
|
# Predict the noise residual
|
|
with torch.cuda.amp.autocast(dtype=weight_dtype):
|
|
|
|
noise_pred = transformer3d(
|
|
x=noisy_latents,
|
|
context=prompt_embeds,
|
|
t=timesteps,
|
|
seq_len=seq_len,
|
|
y=inpaint_latents,
|
|
clip_fea=clip_context,
|
|
vocal_embeddings=vocal_embeddings,
|
|
is_clip_level_modeling=is_clip_level_modeling,
|
|
)
|
|
|
|
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)
|
|
|
|
def custom_mse_loss(noise_pred, target, weighting=None, threshold=50, tgt_face_masks=None, tgt_lip_masks=None):
|
|
noise_pred = noise_pred.float()
|
|
target = target.float()
|
|
diff = noise_pred - target
|
|
mse_loss = F.mse_loss(noise_pred, target, reduction='none')
|
|
|
|
# training setting 1.0
|
|
mask_loss_flag = torch.rand(1).item()
|
|
if mask_loss_flag >= 0.4 and mask_loss_flag < 0.5:
|
|
mse_loss = mse_loss * tgt_face_masks
|
|
elif mask_loss_flag >= 0.5:
|
|
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
|
|
|
|
weighting = compute_loss_weighting_for_sd3(weighting_scheme=args.weighting_scheme, sigmas=sigmas)
|
|
loss = custom_mse_loss(noise_pred.float(), target.float(), weighting.float(), tgt_face_masks=tgt_face_masks, tgt_lip_masks=tgt_lip_masks)
|
|
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:
|
|
accelerator.clip_grad_norm_(trainable_params, args.max_grad_norm)
|
|
optimizer.step()
|
|
lr_scheduler.step()
|
|
optimizer.zero_grad()
|
|
|
|
with torch.cuda.device(accelerator.device):
|
|
torch.cuda.empty_cache()
|
|
|
|
# Checks if the accelerator has performed an optimization step behind the scenes
|
|
if accelerator.sync_gradients:
|
|
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 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)
|
|
|
|
accelerator_save_path = os.path.join(args.output_dir, f"checkpoint-{global_step}")
|
|
os.makedirs(accelerator_save_path, exist_ok=True)
|
|
accelerator.save_state(accelerator_save_path)
|
|
logger.info(f"Saved state to {accelerator_save_path}")
|
|
|
|
transformer3d_saved_path = os.path.join(args.output_dir, f"checkpoint-{global_step}", f"transformer3d-checkpoint-{global_step}.pt")
|
|
unwrap_transformer3d = accelerator.unwrap_model(transformer3d)
|
|
unwrap_transformer3d_state_dict = unwrap_transformer3d.state_dict()
|
|
torch.save(unwrap_transformer3d_state_dict, transformer3d_saved_path)
|
|
print(f"Saved transformer to {transformer3d_saved_path}")
|
|
|
|
|
|
torch.cuda.empty_cache()
|
|
logs = {"step_loss": loss.detach().item(), "lr": lr_scheduler.get_last_lr()[0]}
|
|
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:
|
|
accelerator_save_path = os.path.join(args.output_dir, f"checkpoint-{global_step}")
|
|
os.makedirs(accelerator_save_path, exist_ok=True)
|
|
accelerator.save_state(accelerator_save_path)
|
|
logger.info(f"Saved state to {accelerator_save_path}")
|
|
accelerator.end_training()
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|
|
|
|
|