-
Notifications
You must be signed in to change notification settings - Fork 1
/
train.py
593 lines (493 loc) · 24.9 KB
/
train.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
import os
import math
import imageio
import numpy as np
import wandb
import random
import logging
import inspect
import argparse
import datetime
import subprocess
import multiprocessing as mp
from pathlib import Path
from tqdm.auto import tqdm
from einops import rearrange
from omegaconf import OmegaConf
from safetensors import safe_open
from typing import Dict, Optional, Tuple
import torch
import torchvision
import torch.nn.functional as F
import torch.distributed as dist
from torch.optim.swa_utils import AveragedModel
from torch.utils.data.distributed import DistributedSampler
from torch.nn.parallel import DistributedDataParallel as DDP
import diffusers
from diffusers import AutoencoderKL, DDIMScheduler
from diffusers.models import UNet2DConditionModel
from diffusers.pipelines import StableDiffusionPipeline
from diffusers.optimization import get_scheduler
from diffusers.utils import check_min_version
from diffusers.utils.import_utils import is_xformers_available
from diffusers.image_processor import VaeImageProcessor
import transformers
from transformers import CLIPTextModel, CLIPTokenizer
from animatediff.data.dataset import WebVid10M
from animatediff.models.unet import UNet3DConditionModel
from animatediff.models.controlnet import ControlNetModel
from animatediff.pipelines.pipeline_animation import AnimationPipeline
from animatediff.utils.util import save_videos_grid, zero_rank_print
def prepare_image(
image,
width,
height,
batch_size,
num_images_per_prompt,
device,
dtype,
do_classifier_free_guidance=False,
guess_mode=False,
):
control_image_processor = VaeImageProcessor()
image = control_image_processor.preprocess(image, height=height, width=width).to(dtype=torch.float32)
image_batch_size = image.shape[0]
if image_batch_size == 1:
repeat_by = batch_size
else:
# image batch size is the same as prompt batch size
repeat_by = num_images_per_prompt
image = image.repeat_interleave(repeat_by, dim=0)
image = image.to(device=device, dtype=dtype)
if do_classifier_free_guidance and not guess_mode:
image = torch.cat([image] * 2)
return image
def main(
image_finetune: bool,
name: str,
use_wandb: bool,
launcher: str,
output_dir: str,
pretrained_model_path: str,
train_data: Dict,
validation_data: Dict,
cfg_random_null_text: bool = True,
cfg_random_null_text_ratio: float = 0.1,
unet_checkpoint_path: str = "",
unet_additional_kwargs: Dict = {},
ema_decay: float = 0.9999,
noise_scheduler_kwargs = None,
max_train_epoch: int = -1,
max_train_steps: int = 100,
validation_steps: int = 100,
validation_steps_tuple: Tuple = (-1,),
learning_rate: float = 3e-5,
scale_lr: bool = False,
lr_warmup_steps: int = 0,
lr_scheduler: str = "constant",
trainable_modules: Tuple[str] = (None, ),
num_workers: int = 32,
train_batch_size: int = 1,
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 = False,
checkpointing_epochs: int = 5,
checkpointing_steps: int = -1,
mixed_precision_training: bool = True,
enable_xformers_memory_efficient_attention: bool = True,
global_seed: int = 42,
is_debug: bool = False,
):
check_min_version("0.10.0.dev0")
# Initialize distributed training
# local_rank = init_dist(launcher=launcher)
# local_rank = 1
# global_rank = dist.get_rank()
# num_processes = dist.get_world_size()
# is_main_process = global_rank == 0
is_main_process = True
# seed = global_seed + global_rank
seed = 42
torch.manual_seed(seed)
# Logging folder
folder_name = "debug" if is_debug else name + datetime.datetime.now().strftime("-%Y-%m-%dT%H-%M-%S")
output_dir = os.path.join(output_dir, folder_name)
if is_debug and os.path.exists(output_dir):
os.system(f"rm -rf {output_dir}")
*_, config = inspect.getargvalues(inspect.currentframe())
# 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,
)
# 需要设置wandb账号
if is_main_process and (not is_debug) and use_wandb:
run = wandb.init(project="animatediff_pics_controlnetonly", name=folder_name, config=config)
# Handle the output folder creation
if is_main_process:
os.makedirs(output_dir, exist_ok=True)
os.makedirs(f"{output_dir}/samples", exist_ok=True)
os.makedirs(f"{output_dir}/sanity_check", exist_ok=True)
os.makedirs(f"{output_dir}/checkpoints", exist_ok=True)
OmegaConf.save(config, os.path.join(output_dir, 'config.yaml'))
#-----------------------------------------------------------------------------------------------
# Load scheduler, tokenizer and models.
noise_scheduler = DDIMScheduler(**OmegaConf.to_container(noise_scheduler_kwargs))
vae = AutoencoderKL.from_pretrained(pretrained_model_path, subfolder="vae")
tokenizer = CLIPTokenizer.from_pretrained(pretrained_model_path, subfolder="tokenizer")
text_encoder = CLIPTextModel.from_pretrained(pretrained_model_path, subfolder="text_encoder")
unet2d = UNet2DConditionModel.from_pretrained(pretrained_model_path, subfolder="unet")
# controlnet = ControlNetModel.from_unet(unet2d)
controlnet = ControlNetModel()
# unet = UNet3DConditionModel()
if not image_finetune:
unet = UNet3DConditionModel.from_pretrained_2d(
pretrained_model_path, subfolder="unet",
unet_additional_kwargs=OmegaConf.to_container(unet_additional_kwargs)
)
else:
unet = UNet2DConditionModel.from_pretrained(pretrained_model_path, subfolder="unet")
# Load pretrained unet weights
# if unet_checkpoint_path != "":
# zero_rank_print(f"from checkpoint: {unet_checkpoint_path}")
# unet_checkpoint_path = torch.load(unet_checkpoint_path, map_location="cpu")
# if "global_step" in unet_checkpoint_path: zero_rank_print(f"global_step: {unet_checkpoint_path['global_step']}")
# state_dict = unet_checkpoint_path["state_dict"] if "state_dict" in unet_checkpoint_path else unet_checkpoint_path
# m, u = unet.load_state_dict(state_dict, strict=False)
# zero_rank_print(f"missing keys: {len(m)}, unexpected keys: {len(u)}")
# assert len(u) == 0
motion_module_state_dict = torch.load("models/Motion_Module/mm_sd_v15_v2.ckpt", map_location="cpu")
# # print(motion_module_state_dict)
# # if "global_step" in motion_module_state_dict: func_args.update({"global_step": motion_module_state_dict["global_step"]})
missing, unexpected = unet.load_state_dict(motion_module_state_dict, strict=False)
# print(f"### missing keys: {len(missing)}; \n### unexpected keys: {len(unexpected)};")
# print(f"### missing keys:\n{missing}\n### unexpected keys:\n{unexpected}\n")
assert len(unexpected) == 0
controlnet_state_dict = torch.load("/root/lh/AnimateDiffcontrolnet-main/outputs/1/checkpoints/controlnet_checkpoint-epoch-30.ckpt", map_location="cpu")
missing, unexpected = controlnet.load_state_dict(controlnet_state_dict["state_dict"], strict=False)
assert len(unexpected) == 0
#-----------------------------------------------------------------------------------------------
# Freeze vae and text_encoder
vae.requires_grad_(False)
text_encoder.requires_grad_(False)
# controlnet.requires_grad_(False)
# for name, param in controlnet.named_parameters():
# print(name, ": ", param.requires_grad )
# print("---------------------------------------")
# 把这里打上断点 看一下unet的结构
# Set unet trainable parameters
# print(unet)
unet.requires_grad_(False)
# unet.requires_grad_(True)
# for name, param in unet.named_parameters():
# for trainable_module_name in trainable_modules:
# if trainable_module_name in name:
# param.requires_grad = True
# break
# trainable_params = list(filter(lambda p: p.requires_grad, unet.parameters()))
trainable_params = []
for name, param in controlnet.named_parameters():
if (param.requires_grad):
trainable_params.append(param)
# print(name, ": ", param.requires_grad )
# trainable_params.append(list(filter(lambda p: p.requires_grad, controlnet.parameters())))
optimizer = torch.optim.AdamW(
trainable_params,
lr=learning_rate,
betas=(adam_beta1, adam_beta2),
weight_decay=adam_weight_decay,
eps=adam_epsilon,
)
#-----------------------------------------------------------------------------------------------
if is_main_process:
# zero_rank_print(f"trainable params number: {len(trainable_params)}")
# zero_rank_print(f"trainable params scale: {sum(p.numel() for p in trainable_params) / 1e6:.3f} M")
print(f"trainable params number: {len(trainable_params)}")
print(f"trainable params scale: {sum(p.numel() for p in trainable_params) / 1e6:.3f} M")
#-----------------------------------------------------------------------------------------------
# Enable xformers
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")
# Enable gradient checkpointing
if gradient_checkpointing:
unet.enable_gradient_checkpointing()
# Move models to GPU
# vae.to(local_rank)
# text_encoder.to(local_rank)
vae.to("cuda")
controlnet.to("cuda")
text_encoder.to("cuda")
unet.to("cuda")
#-----------------------------------------------------------------------------------------------
# Get the training dataset
train_dataset = WebVid10M(**train_data, is_image=image_finetune)
# distributed_sampler = DistributedSampler(
# train_dataset,
# num_replicas=1,
# rank=0,
# shuffle=True,
# seed=global_seed,
# )
# DataLoaders creation:
train_dataloader = torch.utils.data.DataLoader(
train_dataset,
batch_size=train_batch_size,
shuffle=False,
# sampler=distributed_sampler,
num_workers=0,
pin_memory=True,
drop_last=True,
)
#-----------------------------------------------------------------------------------------------
# Get the training iteration
if max_train_steps == -1:
assert max_train_epoch != -1
max_train_steps = max_train_epoch * len(train_dataloader)
if checkpointing_steps == -1:
assert checkpointing_epochs != -1
checkpointing_steps = checkpointing_epochs * len(train_dataloader)
if scale_lr:
learning_rate = (learning_rate * gradient_accumulation_steps * train_batch_size )
# 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,
)
#-----------------------------------------------------------------------------------------------
# Validation pipeline
if not image_finetune:
validation_pipeline = AnimationPipeline(
unet=unet, vae=vae, tokenizer=tokenizer, text_encoder=text_encoder, scheduler=noise_scheduler, controlnet=controlnet,
).to("cuda")
else:
validation_pipeline = StableDiffusionPipeline.from_pretrained(
pretrained_model_path,
unet=unet, vae=vae, tokenizer=tokenizer, text_encoder=text_encoder, scheduler=noise_scheduler, safety_checker=None,
)
validation_pipeline.enable_vae_slicing()
#-----------------------------------------------------------------------------------------------
# DDP warpper
# unet = DDP(unet, device_ids=["cuda:0"], output_device="cuda:0")
# 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)
# Train!
total_batch_size = train_batch_size * gradient_accumulation_steps
if is_main_process:
logging.info("***** Running training *****")
logging.info(f" Num examples = {len(train_dataset)}")
logging.info(f" Num Epochs = {num_train_epochs}")
logging.info(f" Instantaneous batch size per device = {train_batch_size}")
logging.info(f" Total train batch size (w. parallel, distributed & accumulation) = {total_batch_size}")
logging.info(f" Gradient Accumulation steps = {gradient_accumulation_steps}")
logging.info(f" Total optimization steps = {max_train_steps}")
global_step = 0
first_epoch = 0
# Only show the progress bar once on each machine.
progress_bar = tqdm(range(global_step, max_train_steps), disable=not is_main_process)
progress_bar.set_description("Steps")
# Support mixed-precision training
scaler = torch.cuda.amp.GradScaler() if mixed_precision_training else None
# mp.set_start_method('spawn')
for epoch in range(first_epoch, num_train_epochs):
# train_dataloader.sampler.set_epoch(epoch)
unet.train()
# mp.set_start_method('spawn')
for step, batch in enumerate(train_dataloader):
if cfg_random_null_text:
batch['text'] = [name if random.random() > cfg_random_null_text_ratio else "" for name in batch['text']]
# Data batch sanity check
if epoch == first_epoch and step == 0:
pixel_values, texts = batch['pixel_values'].cpu(), batch['text']
if not image_finetune:
pixel_values = rearrange(pixel_values, "b f c h w -> b c f h w")
for idx, (pixel_value, text) in enumerate(zip(pixel_values, texts)):
pixel_value = pixel_value[None, ...]
# save_videos_grid(pixel_value, f"{output_dir}/sanity_check/{'-'.join(text.replace('/', '').split()[:10]) if not text == '' else f'{global_seed}-{idx}'}.gif", rescale=True)
else:
for idx, (pixel_value, text) in enumerate(zip(pixel_values, texts)):
pixel_value = pixel_value / 2. + 0.5
torchvision.utils.save_image(pixel_value, f"{output_dir}/sanity_check/{'-'.join(text.replace('/', '').split()[:10]) if not text == '' else f'{global_seed}-{idx}'}.png")
### >>>> Training >>>> ###
# Convert videos to latent space
orig_img = batch['image'].squeeze(1)
orig_img = orig_img / 2. + 0.5
pixel_values = batch["pixel_values"].to("cuda")
video_length = pixel_values.shape[1]
with torch.no_grad():
if not image_finetune:
pixel_values = rearrange(pixel_values, "b f c h w -> (b f) c h w")
latents = vae.encode(pixel_values).latent_dist
latents = latents.sample()
latents = rearrange(latents, "(b f) c h w -> b c f h w", f=video_length)
else:
latents = vae.encode(pixel_values).latent_dist
latents = latents.sample()
latents = latents * 0.18215
#--------------------------------------
image = prepare_image(
image=orig_img,
width=orig_img.shape[-1],
height=orig_img.shape[-2],
batch_size=orig_img.shape[0],
num_images_per_prompt=1,
device="cuda",
dtype=controlnet.dtype,
)
#--------------------------------------
# Sample noise that we'll add to the latents 如果要加原图信号的话 就是在这里加
noise = torch.randn_like(latents) # latents shape为 [4, 4, 16, 32, 32]
bsz = latents.shape[0]
# Sample a random timestep for each video
timesteps = torch.randint(0, noise_scheduler.config.num_train_timesteps, (bsz,), device=latents.device) # shape [4]
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) # shape[4, 4, 16, 32, 32]
# Get the text embedding for conditioning
with torch.no_grad():
prompt_ids = tokenizer(
batch['text'], max_length=tokenizer.model_max_length, padding="max_length", truncation=True, return_tensors="pt"
).input_ids.to(latents.device) # shape [4, 77]
encoder_hidden_states = text_encoder(prompt_ids)[0] # shape [4, 77, 768]
# Get the target for loss depending on the prediction type
if noise_scheduler.config.prediction_type == "epsilon":
target = noise
elif noise_scheduler.config.prediction_type == "v_prediction":
raise NotImplementedError
else:
raise ValueError(f"Unknown prediction type {noise_scheduler.config.prediction_type}")
#------------------------------------------------
down_block_res_samples, mid_block_res_sample = controlnet(
sample=noisy_latents[:,:,0,:,:],
timestep=timesteps,
encoder_hidden_states=encoder_hidden_states, # [4,77,768]
controlnet_cond=image,
return_dict=False,
)
# down_block_additional_residuals
# mid_block_additional_residual
#------------------------------------------------
# Predict the noise residual and compute loss
# Mixed-precision training
with torch.cuda.amp.autocast(enabled=mixed_precision_training):
# noisy_latents shape [4, 4, 16, 32, 32]
# encoder_hidden_states [4, 77, 768]
model_pred = unet(sample=noisy_latents,
timestep=timesteps,
encoder_hidden_states=encoder_hidden_states,
down_block_additional_residuals=down_block_res_samples,
mid_block_additional_residual=mid_block_res_sample,
).sample
loss = F.mse_loss(model_pred.float(), target.float(), reduction="mean")
optimizer.zero_grad()
# Backpropagate
if mixed_precision_training:
scaler.scale(loss).backward()
""" >>> gradient clipping >>> """
scaler.unscale_(optimizer)
torch.nn.utils.clip_grad_norm_(unet.parameters(), max_grad_norm)
""" <<< gradient clipping <<< """
scaler.step(optimizer)
scaler.update()
else:
loss.backward()
""" >>> gradient clipping >>> """
torch.nn.utils.clip_grad_norm_(unet.parameters(), max_grad_norm)
""" <<< gradient clipping <<< """
optimizer.step()
lr_scheduler.step()
progress_bar.update(1)
global_step += 1
### <<<< Training <<<< ###
# Wandb logging
if is_main_process and (not is_debug) and use_wandb:
wandb.log({"train_loss": loss.item()}, step=global_step)
# Save checkpoint
if is_main_process and (global_step % checkpointing_steps == 0 or step == len(train_dataloader) - 1):
save_path = os.path.join(output_dir, f"checkpoints")
controlnet_state_dict = {
"epoch": epoch,
"global_step": global_step,
"state_dict": controlnet.state_dict(),
}
if step == len(train_dataloader) - 1:
torch.save(controlnet_state_dict, os.path.join(save_path, f"controlnet_checkpoint-epoch-{epoch+1}.ckpt"))
else:
torch.save(controlnet_state_dict, os.path.join(save_path, f"controlnet_checkpoint.ckpt"))
logging.info(f"Saved state to {save_path} (global_step: {global_step})")
# Periodically validation
if is_main_process and (global_step % validation_steps == 0 or global_step in validation_steps_tuple):
samples = []
generator = torch.Generator(device=latents.device)
generator.manual_seed(global_seed)
height = train_data.sample_size[0] if not isinstance(train_data.sample_size, int) else train_data.sample_size
width = train_data.sample_size[1] if not isinstance(train_data.sample_size, int) else train_data.sample_size
# prompts = validation_data.prompts[:2] if global_step < 1000 and (not image_finetune) else validation_data.prompts
prompts = batch['text']
init_images = batch['image'].squeeze(1) #[b, 1, c, h, w]
controlnet_images = init_images / 2. + 0.5
init_images = init_images.permute(0,2,3,1) # [b, 1, h, w, c]
init_images = np.array(init_images.cpu())
for idx, prompt in enumerate(prompts):
if not image_finetune:
controlnet_image = controlnet_images[idx, :, :, :]
sample = validation_pipeline(
prompt,
generator = generator,
video_length = train_data.sample_n_frames,
height = height,
width = width,
controlnet_image=controlnet_image,
**validation_data,
).videos
init_image = init_images[idx, :, :, :]
save_videos_grid(sample, f"{output_dir}/samples/sample-{global_step}/{idx}.gif")
imageio.imsave(f"{output_dir}/samples/sample-{global_step}/{idx}.jpg", init_image)
samples.append(sample)
else:
sample = validation_pipeline(
prompt,
generator = generator,
height = height,
width = width,
num_inference_steps = validation_data.get("num_inference_steps", 25),
guidance_scale = validation_data.get("guidance_scale", 8.),
).images[0]
sample = torchvision.transforms.functional.to_tensor(sample)
samples.append(sample)
if not image_finetune:
samples = torch.concat(samples)
save_path = f"{output_dir}/samples/sample-{global_step}.gif"
save_videos_grid(samples, save_path)
else:
samples = torch.stack(samples)
save_path = f"{output_dir}/samples/sample-{global_step}.png"
torchvision.utils.save_image(samples, save_path, nrow=4)
logging.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
# dist.destroy_process_group()
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--config", type=str, default="./configs/training/training.yaml")
parser.add_argument("--launcher", type=str, choices=["pytorch", "slurm"], default="pytorch")
# parser.add_argument("--wandb", action="store_true")
parser.add_argument("--wandb", type=bool, default=True)
args = parser.parse_args()
name = Path(args.config).stem
config = OmegaConf.load(args.config)
main(name=name, launcher=args.launcher, use_wandb=args.wandb, **config)