506 lines
22 KiB
Python
506 lines
22 KiB
Python
import argparse
|
|
import datetime
|
|
import logging
|
|
import inspect
|
|
import math
|
|
import os
|
|
|
|
import folder_paths
|
|
comfy_path = os.path.dirname(folder_paths.__file__)
|
|
|
|
import sys
|
|
SimDA_path=f'{comfy_path}/custom_nodes/ComfyUI-SimDA'
|
|
sys.path.insert(0,SimDA_path)
|
|
|
|
from typing import Dict, Optional, Tuple
|
|
from omegaconf import OmegaConf
|
|
|
|
import torch
|
|
import torch.nn.functional as F
|
|
import torch.utils.checkpoint
|
|
|
|
import diffusers
|
|
import transformers
|
|
from accelerate import Accelerator
|
|
from accelerate.logging import get_logger
|
|
from accelerate.utils import set_seed
|
|
from diffusers import AutoencoderKL, DDPMScheduler, DDIMScheduler
|
|
from diffusers.optimization import get_scheduler
|
|
from diffusers.utils import check_min_version
|
|
from diffusers.utils.import_utils import is_xformers_available
|
|
from tqdm.auto import tqdm
|
|
from transformers import CLIPTextModel, CLIPTokenizer
|
|
|
|
from simda.models.unet import UNet3DConditionModel
|
|
from simda.data.dataset import SimDADataset
|
|
from simda.pipelines.pipeline_simda import SimDAPipeline
|
|
from simda.util import save_videos_grid, ddim_inversion
|
|
from einops import rearrange
|
|
|
|
# Will error if the minimal version of diffusers is not installed. Remove at your own risks.
|
|
check_min_version("0.10.0.dev0")
|
|
|
|
logger = get_logger(__name__, log_level="INFO")
|
|
|
|
import json
|
|
|
|
import comfy.utils
|
|
|
|
def train(
|
|
pretrained_model_path: str,
|
|
output_dir: str,
|
|
train_data: Dict,
|
|
validation_data: Dict,
|
|
validation_steps: int = 100,
|
|
trainable_modules: Tuple[str] = (
|
|
"temp_adapter"
|
|
"adapter_s"
|
|
"adapter_ffn"
|
|
),
|
|
train_batch_size: int = 1,
|
|
max_train_steps: int = 500,
|
|
learning_rate: float = 3e-5,
|
|
scale_lr: bool = False,
|
|
lr_scheduler: str = "constant",
|
|
lr_warmup_steps: int = 0,
|
|
adam_beta1: float = 0.9,
|
|
adam_beta2: float = 0.999,
|
|
adam_weight_decay: float = 1e-2,
|
|
adam_epsilon: float = 1e-08,
|
|
max_grad_norm: float = 1.0,
|
|
gradient_accumulation_steps: int = 1,
|
|
gradient_checkpointing: bool = True,
|
|
checkpointing_steps: int = 500,
|
|
resume_from_checkpoint: Optional[str] = None,
|
|
mixed_precision: Optional[str] = "fp16",
|
|
use_8bit_adam: bool = False,
|
|
enable_xformers_memory_efficient_attention: bool = True,
|
|
seed: Optional[int] = None,
|
|
):
|
|
with torch.inference_mode(False):
|
|
with torch.set_grad_enabled(True):
|
|
*_, config = inspect.getargvalues(inspect.currentframe())
|
|
|
|
accelerator = Accelerator(
|
|
gradient_accumulation_steps=gradient_accumulation_steps,
|
|
mixed_precision=mixed_precision,
|
|
)
|
|
print(f'is_main_process{accelerator.is_main_process}')
|
|
print(f'num_processes{accelerator.num_processes}')
|
|
|
|
# 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:
|
|
transformers.utils.logging.set_verbosity_warning()
|
|
diffusers.utils.logging.set_verbosity_info()
|
|
else:
|
|
transformers.utils.logging.set_verbosity_error()
|
|
diffusers.utils.logging.set_verbosity_error()
|
|
|
|
# If passed along, set the training seed now.
|
|
if seed is not None:
|
|
set_seed(seed)
|
|
|
|
# Handle the output folder creation
|
|
if accelerator.is_main_process:
|
|
# now = datetime.datetime.now().strftime("%Y-%m-%dT%H-%M-%S")
|
|
# output_dir = os.path.join(output_dir, now)
|
|
os.makedirs(output_dir, exist_ok=True)
|
|
os.makedirs(f"{output_dir}/samples", exist_ok=True)
|
|
os.makedirs(f"{output_dir}/inv_latents", exist_ok=True)
|
|
OmegaConf.save(config, os.path.join(output_dir, 'config.yaml'))
|
|
|
|
# Load scheduler, tokenizer and models.
|
|
noise_scheduler = DDPMScheduler.from_pretrained(pretrained_model_path, subfolder="scheduler")
|
|
tokenizer = CLIPTokenizer.from_pretrained(pretrained_model_path, subfolder="tokenizer")
|
|
text_encoder = CLIPTextModel.from_pretrained(pretrained_model_path, subfolder="text_encoder")
|
|
vae = AutoencoderKL.from_pretrained(pretrained_model_path, subfolder="vae")
|
|
unet = UNet3DConditionModel.from_pretrained_2d(pretrained_model_path, subfolder="unet")
|
|
|
|
# Freeze vae and text_encoder
|
|
vae.requires_grad_(False)
|
|
text_encoder.requires_grad_(False)
|
|
|
|
unet.requires_grad_(False)
|
|
for name, module in unet.named_modules():
|
|
if name.endswith(tuple(trainable_modules)):
|
|
for params in module.parameters():
|
|
params.requires_grad = True
|
|
|
|
if enable_xformers_memory_efficient_attention:
|
|
if is_xformers_available():
|
|
unet.enable_xformers_memory_efficient_attention()
|
|
else:
|
|
raise ValueError("xformers is not available. Make sure it is installed correctly")
|
|
|
|
if gradient_checkpointing:
|
|
unet.enable_gradient_checkpointing()
|
|
|
|
if scale_lr:
|
|
learning_rate = (
|
|
learning_rate * gradient_accumulation_steps * train_batch_size * accelerator.num_processes
|
|
)
|
|
|
|
# Initialize the optimizer
|
|
if 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
|
|
else:
|
|
optimizer_cls = torch.optim.AdamW
|
|
|
|
optimizer = optimizer_cls(
|
|
unet.parameters(),
|
|
lr=learning_rate,
|
|
betas=(adam_beta1, adam_beta2),
|
|
weight_decay=adam_weight_decay,
|
|
eps=adam_epsilon,
|
|
)
|
|
|
|
# Get the training dataset
|
|
train_dataset = SimDADataset(**train_data)
|
|
|
|
# Preprocessing the dataset
|
|
train_dataset.prompt_ids = tokenizer(
|
|
train_dataset.prompt, max_length=tokenizer.model_max_length, padding="max_length", truncation=True, return_tensors="pt"
|
|
).input_ids[0]
|
|
|
|
# DataLoaders creation:
|
|
train_dataloader = torch.utils.data.DataLoader(
|
|
train_dataset, batch_size=train_batch_size
|
|
)
|
|
|
|
# Get the validation pipeline
|
|
validation_pipeline = SimDAPipeline(
|
|
vae=vae, text_encoder=text_encoder, tokenizer=tokenizer, unet=unet,
|
|
scheduler=DDIMScheduler.from_pretrained(pretrained_model_path, subfolder="scheduler")
|
|
)
|
|
validation_pipeline.enable_vae_slicing()
|
|
ddim_inv_scheduler = DDIMScheduler.from_pretrained(pretrained_model_path, subfolder='scheduler')
|
|
ddim_inv_scheduler.set_timesteps(validation_data.num_inv_steps)
|
|
|
|
# Scheduler
|
|
lr_scheduler = get_scheduler(
|
|
lr_scheduler,
|
|
optimizer=optimizer,
|
|
num_warmup_steps=lr_warmup_steps * gradient_accumulation_steps,
|
|
num_training_steps=max_train_steps * gradient_accumulation_steps,
|
|
)
|
|
|
|
# Prepare everything with our `accelerator`.
|
|
unet, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
|
|
unet, optimizer, train_dataloader, lr_scheduler
|
|
)
|
|
|
|
# For mixed precision training we cast the text_encoder and vae weights to half-precision
|
|
# as these models are only used for inference, keeping weights in full precision is not required.
|
|
weight_dtype = torch.float32
|
|
if accelerator.mixed_precision == "fp16":
|
|
weight_dtype = torch.float16
|
|
elif accelerator.mixed_precision == "bf16":
|
|
weight_dtype = torch.bfloat16
|
|
|
|
# Move text_encode and vae to gpu and cast to weight_dtype
|
|
text_encoder.to(accelerator.device, dtype=weight_dtype)
|
|
vae.to(accelerator.device, 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) / gradient_accumulation_steps)
|
|
# Afterwards we recalculate our number of training epochs
|
|
num_train_epochs = math.ceil(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("text2video-fine-tune")
|
|
|
|
# Train!
|
|
total_batch_size = train_batch_size * accelerator.num_processes * gradient_accumulation_steps
|
|
|
|
logger.info("***** Running training *****")
|
|
logger.info(f" Num examples = {len(train_dataset)}")
|
|
logger.info(f" Num Epochs = {num_train_epochs}")
|
|
logger.info(f" Instantaneous batch size per device = {train_batch_size}")
|
|
logger.info(f" Total train batch size (w. parallel, distributed & accumulation) = {total_batch_size}")
|
|
logger.info(f" Gradient Accumulation steps = {gradient_accumulation_steps}")
|
|
logger.info(f" Total optimization steps = {max_train_steps}")
|
|
global_step = 0
|
|
first_epoch = 0
|
|
|
|
# Potentially load in the weights and states from a previous save
|
|
if resume_from_checkpoint:
|
|
if resume_from_checkpoint != "latest":
|
|
path = os.path.basename(resume_from_checkpoint)
|
|
else:
|
|
# Get the most recent checkpoint
|
|
dirs = os.listdir(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]
|
|
accelerator.print(f"Resuming from checkpoint {path}")
|
|
accelerator.load_state(os.path.join(output_dir, path))
|
|
global_step = int(path.split("-")[1])
|
|
|
|
first_epoch = global_step // num_update_steps_per_epoch
|
|
resume_step = global_step % num_update_steps_per_epoch
|
|
|
|
# Only show the progress bar once on each machine.
|
|
progress_bar = tqdm(range(global_step, max_train_steps), disable=not accelerator.is_local_main_process)
|
|
progress_bar.set_description("Steps")
|
|
pbar = comfy.utils.ProgressBar(max_train_steps)
|
|
|
|
for epoch in range(first_epoch, num_train_epochs):
|
|
unet.train()
|
|
train_loss = 0.0
|
|
for step, batch in enumerate(train_dataloader):
|
|
# Skip steps until we reach the resumed step
|
|
if resume_from_checkpoint and epoch == first_epoch and step < resume_step:
|
|
if step % gradient_accumulation_steps == 0:
|
|
progress_bar.update(1)
|
|
pbar.update(1)
|
|
continue
|
|
|
|
with accelerator.accumulate(unet):
|
|
# Convert videos to latent space
|
|
pixel_values = batch["pixel_values"].to(weight_dtype)
|
|
video_length = pixel_values.shape[1]
|
|
pixel_values = rearrange(pixel_values, "b f c h w -> (b f) c h w")
|
|
latents = vae.encode(pixel_values).latent_dist.sample()
|
|
latents = rearrange(latents, "(b f) c h w -> b c f h w", f=video_length)
|
|
latents = latents * 0.18215
|
|
|
|
# Sample noise that we'll add to the latents
|
|
noise = torch.randn_like(latents)
|
|
bsz = latents.shape[0]
|
|
# Sample a random timestep for each video
|
|
timesteps = torch.randint(0, noise_scheduler.num_train_timesteps, (bsz,), device=latents.device)
|
|
timesteps = timesteps.long()
|
|
|
|
# Add noise to the latents according to the noise magnitude at each timestep
|
|
# (this is the forward diffusion process)
|
|
noisy_latents = noise_scheduler.add_noise(latents, noise, timesteps)
|
|
|
|
# Get the text embedding for conditioning
|
|
encoder_hidden_states = text_encoder(batch["prompt_ids"])[0]
|
|
|
|
# Get the target for loss depending on the prediction type
|
|
if noise_scheduler.prediction_type == "epsilon":
|
|
target = noise
|
|
elif noise_scheduler.prediction_type == "v_prediction":
|
|
target = noise_scheduler.get_velocity(latents, noise, timesteps)
|
|
else:
|
|
raise ValueError(f"Unknown prediction type {noise_scheduler.prediction_type}")
|
|
|
|
# Predict the noise residual and compute loss
|
|
model_pred = unet(noisy_latents, timesteps, encoder_hidden_states).sample
|
|
loss = F.mse_loss(model_pred.float(), target.float(), reduction="mean")
|
|
|
|
# Gather the losses across all processes for logging (if we use distributed training).
|
|
avg_loss = accelerator.gather(loss.repeat(train_batch_size)).mean()
|
|
train_loss += avg_loss.item() / gradient_accumulation_steps
|
|
|
|
# Backpropagate
|
|
accelerator.backward(loss)
|
|
if accelerator.sync_gradients:
|
|
accelerator.clip_grad_norm_(unet.parameters(), max_grad_norm)
|
|
optimizer.step()
|
|
lr_scheduler.step()
|
|
optimizer.zero_grad()
|
|
|
|
# Checks if the accelerator has performed an optimization step behind the scenes
|
|
if accelerator.sync_gradients:
|
|
progress_bar.update(1)
|
|
pbar.update(1)
|
|
global_step += 1
|
|
accelerator.log({"train_loss": train_loss}, step=global_step)
|
|
train_loss = 0.0
|
|
|
|
if global_step % checkpointing_steps == 0:
|
|
if accelerator.is_main_process:
|
|
save_path = os.path.join(output_dir, f"checkpoint-{global_step}")
|
|
accelerator.save_state(save_path)
|
|
logger.info(f"Saved state to {save_path}")
|
|
|
|
if global_step % validation_steps == 0:
|
|
if accelerator.is_main_process:
|
|
samples = []
|
|
generator = torch.Generator(device=latents.device)
|
|
generator.manual_seed(seed)
|
|
|
|
ddim_inv_latent = None
|
|
if validation_data.use_inv_latent:
|
|
inv_latents_path = os.path.join(output_dir, f"inv_latents/ddim_latent-{global_step}.pt")
|
|
ddim_inv_latent = ddim_inversion(
|
|
validation_pipeline, ddim_inv_scheduler, video_latent=latents,
|
|
num_inv_steps=validation_data.num_inv_steps, prompt="")[-1].to(weight_dtype)
|
|
torch.save(ddim_inv_latent, inv_latents_path)
|
|
|
|
for idx, prompt in enumerate(validation_data.prompts):
|
|
sample = validation_pipeline(prompt, generator=generator, latents=ddim_inv_latent,
|
|
**validation_data).videos
|
|
save_videos_grid(sample, f"{output_dir}/samples/sample-{global_step}/{prompt}.gif")
|
|
samples.append(sample)
|
|
samples = torch.concat(samples)
|
|
save_path = f"{output_dir}/samples/sample-{global_step}.gif"
|
|
save_videos_grid(samples, save_path)
|
|
logger.info(f"Saved samples to {save_path}")
|
|
|
|
logs = {"step_loss": loss.detach().item(), "lr": lr_scheduler.get_last_lr()[0]}
|
|
progress_bar.set_postfix(**logs)
|
|
|
|
if global_step >= max_train_steps:
|
|
break
|
|
|
|
# Create the pipeline using the trained modules and save it.
|
|
accelerator.wait_for_everyone()
|
|
#if accelerator.is_main_process:
|
|
unet = accelerator.unwrap_model(unet)
|
|
pipeline = SimDAPipeline.from_pretrained(
|
|
pretrained_model_path,
|
|
text_encoder=text_encoder,
|
|
vae=vae,
|
|
unet=unet,
|
|
)
|
|
pipeline.save_pretrained(output_dir)
|
|
|
|
accelerator.end_training()
|
|
|
|
class SimDATrain:
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {
|
|
"pretrained_model_path": (os.listdir(os.path.join(folder_paths.models_dir,"diffusers")), {"default": "stable-diffusion-v1-5"}),
|
|
"output_dir": ("STRING", {"default": "car-turn"}),
|
|
"video_path": ("STRING", {"default": ""}),
|
|
"prompt": ("STRING", {"multiline": True, "default": "a jeep car is moving on the road"}),
|
|
"n_sample_frames":("INT",{"default":24}),
|
|
"width":("INT",{"default":512}),
|
|
"height":("INT",{"default":512}),
|
|
"sample_start_idx":("INT",{"default":16}),
|
|
"sample_frame_rate":("INT",{"default":2}),
|
|
"validation_data_prompts":("STRING",{"multiline": True, "default":'["a jeep car is moving on the beach"]'}),
|
|
"num_inference_steps":("INT",{"default":20}),
|
|
"guidance_scale":("FLOAT",{"default":12.5}),
|
|
"max_train_steps":("INT",{"default":300}),
|
|
},
|
|
"optional": {
|
|
}
|
|
}
|
|
|
|
RETURN_TYPES = ("STRING",)
|
|
FUNCTION = "run"
|
|
CATEGORY = "CameraCtrl"
|
|
|
|
def run(self,pretrained_model_path,output_dir,video_path,prompt,n_sample_frames,width,height,sample_start_idx,sample_frame_rate,validation_data_prompts,num_inference_steps,guidance_scale,max_train_steps):
|
|
pretrained_model_path=os.path.join(os.path.join(folder_paths.models_dir,"diffusers"),pretrained_model_path)
|
|
output_dir=os.path.join(folder_paths.output_directory,output_dir)
|
|
|
|
config = OmegaConf.load(os.path.join(SimDA_path,"configs/car-turn.yaml"))
|
|
OmegaConf.update(config, "pretrained_model_path", pretrained_model_path)
|
|
OmegaConf.update(config, "output_dir", output_dir)
|
|
|
|
OmegaConf.update(config, "train_data.video_path", video_path)
|
|
OmegaConf.update(config, "train_data.prompt", prompt)
|
|
OmegaConf.update(config, "train_data.n_sample_frames", n_sample_frames)
|
|
OmegaConf.update(config, "train_data.width", width)
|
|
OmegaConf.update(config, "train_data.height", height)
|
|
OmegaConf.update(config, "train_data.sample_start_idx", sample_start_idx)
|
|
OmegaConf.update(config, "train_data.sample_frame_rate", sample_frame_rate)
|
|
|
|
OmegaConf.update(config, "validation_data.prompts", json.loads(validation_data_prompts))
|
|
OmegaConf.update(config, "validation_data.num_inference_steps", num_inference_steps)
|
|
OmegaConf.update(config, "validation_data.guidance_scale", guidance_scale)
|
|
|
|
OmegaConf.update(config, "max_train_steps", max_train_steps)
|
|
|
|
train(**config)
|
|
|
|
return (output_dir,)
|
|
|
|
class SimDALoader:
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {
|
|
"pretrained_model_path": (os.listdir(os.path.join(folder_paths.models_dir,"diffusers")), {"default": "stable-diffusion-v1-5"}),
|
|
"my_model_path": ("STRING", {"default": ""}),
|
|
},
|
|
"optional": {
|
|
}
|
|
}
|
|
|
|
RETURN_TYPES = ("SimDAPipe",)
|
|
FUNCTION = "run"
|
|
CATEGORY = "CameraCtrl"
|
|
|
|
def run(self,pretrained_model_path,my_model_path):
|
|
pretrained_model_path=os.path.join(os.path.join(folder_paths.models_dir,"diffusers"),pretrained_model_path)
|
|
#my_model_path = "./outputs/car-turn"
|
|
unet = UNet3DConditionModel.from_pretrained(my_model_path, subfolder='unet', torch_dtype=torch.float16).to('cuda')
|
|
pipe = SimDAPipeline.from_pretrained(pretrained_model_path, unet=unet, torch_dtype=torch.float16).to("cuda")
|
|
pipe.enable_xformers_memory_efficient_attention()
|
|
pipe.enable_vae_slicing()
|
|
|
|
return (pipe,)
|
|
|
|
|
|
class SimDARun:
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {
|
|
"pipe": ("SimDAPipe",),
|
|
"prompt": ("STRING", {"default": "a jeep car is moving on the road"}),
|
|
"video_length":("INT",{"default":24}),
|
|
"width":("INT",{"default":512}),
|
|
"height":("INT",{"default":512}),
|
|
"num_inference_steps":("INT",{"default":50}),
|
|
"guidance_scale":("FLOAT",{"default":12.5}),
|
|
},
|
|
"optional": {
|
|
}
|
|
}
|
|
|
|
RETURN_TYPES = ("IMAGE",)
|
|
FUNCTION = "run"
|
|
CATEGORY = "CameraCtrl"
|
|
|
|
def run(self,pipe,prompt,video_length,width,height,num_inference_steps,guidance_scale):
|
|
video = pipe(prompt, latents=None, video_length=video_length, height=height, width=width, num_inference_steps=num_inference_steps, guidance_scale=guidance_scale).videos
|
|
|
|
video = rearrange(video, "b c t h w -> b t h w c")
|
|
return video
|
|
|
|
class VHS_FILENAMES_STRING_SimDA:
|
|
@classmethod
|
|
def INPUT_TYPES(s):
|
|
return {
|
|
"required": {
|
|
"filenames": ("VHS_FILENAMES",),
|
|
}
|
|
}
|
|
|
|
RETURN_TYPES = ("STRING",)
|
|
CATEGORY = "StreamingT2V"
|
|
FUNCTION = "run"
|
|
|
|
def run(self, filenames):
|
|
return (filenames[1][-1],)
|
|
|
|
NODE_CLASS_MAPPINGS = {
|
|
"SimDATrain":SimDATrain,
|
|
"SimDALoader":SimDALoader,
|
|
"SimDARun":SimDARun,
|
|
"VHS_FILENAMES_STRING_SimDA":VHS_FILENAMES_STRING_SimDA,
|
|
} |