380 lines
12 KiB
Python
380 lines
12 KiB
Python
|
|
|
|
import json
|
|
import os
|
|
import time
|
|
import torch
|
|
|
|
|
|
UNET_PATH = "ckpts/t2i/model/pytorch_model_ema.pt"
|
|
|
|
|
|
def set_unet_path(path):
|
|
global UNET_PATH
|
|
UNET_PATH = path
|
|
|
|
|
|
VAE_EMA_PATH = "ckpts/t2i/sdxl-vae-fp16-fix"
|
|
|
|
|
|
def set_vae_ema_path(path):
|
|
global VAE_EMA_PATH
|
|
VAE_EMA_PATH = path
|
|
|
|
|
|
TOKENIZER = "ckpts/t2i/tokenizer"
|
|
|
|
|
|
def set_tokenizer_path(path):
|
|
global TOKENIZER
|
|
TOKENIZER = path
|
|
|
|
|
|
TEXT_ENCODER = 'ckpts/t2i/clip_text_encoder'
|
|
|
|
|
|
def set_text_encoder_path(path):
|
|
global TEXT_ENCODER
|
|
TEXT_ENCODER = path
|
|
|
|
|
|
T5_ENCODER = {
|
|
'MT5': None,
|
|
'attention_mask': True,
|
|
'layer_index': -1,
|
|
'attention_pool': True,
|
|
'torch_dtype': torch.float16,
|
|
'learnable_replace': True
|
|
}
|
|
|
|
|
|
def set_t5_encoder_path(path):
|
|
global T5_ENCODER
|
|
T5_ENCODER['MT5'] = path
|
|
|
|
|
|
global TRAIN_CONFIG
|
|
|
|
|
|
def set_train_config(train_config):
|
|
global TRAIN_CONFIG
|
|
TRAIN_CONFIG = train_config
|
|
|
|
|
|
def easy_sample_images(
|
|
args,
|
|
vae=None,
|
|
text_encoder=None,
|
|
tokenizer=None,
|
|
model=None,
|
|
embedder_t5=None,
|
|
target_height=768,
|
|
target_width=1280,
|
|
prompt="A photo of a girl with a hat on a sunny day",
|
|
negative_prompt="",
|
|
batch_size=1,
|
|
guidance_scale=2.0,
|
|
infer_steps=20,
|
|
sampler='dpmpp_2m_karras',
|
|
train_steps=0,
|
|
seed=0,
|
|
):
|
|
from hydit.diffusion.pipeline import StableDiffusionPipeline
|
|
from diffusers import schedulers
|
|
from hydit.constants import SAMPLER_FACTORY
|
|
from hydit.modules.posemb_layers import get_fill_resize_and_crop, get_2d_rotary_pos_embed
|
|
from hydit.modules.models import HUNYUAN_DIT_CONFIG
|
|
|
|
import traceback
|
|
with torch.cuda.amp.autocast():
|
|
|
|
workspace_dir = TRAIN_CONFIG.get("workspace_dir")
|
|
sample_config_file = TRAIN_CONFIG.get("sample_config_file", None)
|
|
if sample_config_file is None:
|
|
print("sample_config_file is not set.")
|
|
return
|
|
try:
|
|
sample_config = json.load(open(sample_config_file, "r"))
|
|
except Exception as e:
|
|
print(f"Failed to load sample_config_file: {sample_config_file}")
|
|
return
|
|
sample_images_dir = os.path.join(workspace_dir, "sample_images")
|
|
os.makedirs(sample_images_dir, exist_ok=True)
|
|
|
|
sampler_factory = SAMPLER_FACTORY.copy()
|
|
|
|
sampler_factory["uni_pc"] = {
|
|
'scheduler': 'UniPCMultistepScheduler',
|
|
'name': 'UniPCMultistepScheduler',
|
|
'kwargs': {
|
|
'beta_schedule': 'scaled_linear',
|
|
'beta_start': 0.00085,
|
|
'beta_end': 0.03,
|
|
'prediction_type': 'v_prediction',
|
|
'trained_betas': None,
|
|
'solver_order': 2,
|
|
}
|
|
}
|
|
sampler_factory["dpmpp_2m_karras"] = {
|
|
'scheduler': 'DPMSolverMultistepScheduler',
|
|
'name': 'DPMSolverMultistepScheduler',
|
|
'kwargs': {
|
|
'beta_schedule': 'scaled_linear',
|
|
'beta_start': 0.00085,
|
|
'beta_end': 0.03,
|
|
'prediction_type': 'v_prediction',
|
|
'trained_betas': None,
|
|
'solver_order': 2,
|
|
'algorithm_type': 'dpmsolver++',
|
|
"use_karras_sigmas": True,
|
|
}
|
|
}
|
|
|
|
# Load sampler from factory
|
|
kwargs = sampler_factory[sampler]['kwargs']
|
|
scheduler = sampler_factory[sampler]['scheduler']
|
|
|
|
# Build scheduler according to the sampler.
|
|
scheduler_class = getattr(schedulers, scheduler)
|
|
scheduler = scheduler_class(**kwargs)
|
|
|
|
# Set timesteps for inference steps.
|
|
scheduler.set_timesteps(infer_steps, "cuda")
|
|
|
|
def calc_rope(height, width):
|
|
model_config = HUNYUAN_DIT_CONFIG["DiT-g/2"]
|
|
patch_size = model_config['patch_size']
|
|
head_size = model_config['hidden_size'] // model_config['num_heads']
|
|
th = height // 8 // patch_size
|
|
tw = width // 8 // patch_size
|
|
base_size = 512 // 8 // patch_size
|
|
start, stop = get_fill_resize_and_crop((th, tw), base_size)
|
|
sub_args = [start, stop, (th, tw)]
|
|
rope = get_2d_rotary_pos_embed(head_size, *sub_args)
|
|
return rope
|
|
|
|
pipeline = StableDiffusionPipeline(vae=vae,
|
|
text_encoder=text_encoder,
|
|
tokenizer=tokenizer,
|
|
unet=model.module,
|
|
scheduler=scheduler,
|
|
feature_extractor=None,
|
|
safety_checker=None,
|
|
requires_safety_checker=False,
|
|
embedder_t5=embedder_t5,
|
|
)
|
|
pipeline = pipeline.to("cuda")
|
|
# attr _execution_device is not defined
|
|
|
|
style = torch.as_tensor([0, 0] * batch_size, device="cuda")
|
|
|
|
src_size_cond = (target_width, target_height)
|
|
size_cond = list(src_size_cond) + [target_width, target_height, 0, 0]
|
|
image_meta_size = torch.as_tensor(
|
|
[size_cond] * 2 * batch_size, device="cuda",)
|
|
|
|
if type(sample_config) != list:
|
|
sample_config = [sample_config]
|
|
|
|
for i, sample in enumerate(sample_config):
|
|
prompt = sample.get("prompt", "")
|
|
negative_prompt = sample.get("negative_prompt", "")
|
|
guidance_scale = sample.get("cfg", guidance_scale)
|
|
infer_steps = sample.get("steps", infer_steps)
|
|
width = sample.get("width", target_width)
|
|
height = sample.get("height", target_height)
|
|
|
|
freqs_cis_img = calc_rope(height, width)
|
|
|
|
try:
|
|
generator = torch.Generator(device="cuda")
|
|
generator.manual_seed(seed)
|
|
samples = pipeline(
|
|
height=height,
|
|
width=width,
|
|
prompt=prompt,
|
|
negative_prompt=negative_prompt,
|
|
num_images_per_prompt=batch_size,
|
|
guidance_scale=guidance_scale,
|
|
num_inference_steps=infer_steps,
|
|
style=style,
|
|
return_dict=False,
|
|
use_fp16=True,
|
|
learn_sigma=args.learn_sigma,
|
|
freqs_cis_img=freqs_cis_img,
|
|
image_meta_size=image_meta_size,
|
|
generator=generator,
|
|
)[0]
|
|
|
|
pass
|
|
except Exception as e:
|
|
print(f"Failed to sample images: {e} ")
|
|
# 打印堆栈信息
|
|
traceback.print_exc()
|
|
print(f"Failed to sample pipeline: {pipeline} ")
|
|
|
|
# print("samples:",type(samples),)
|
|
# input("Press Enter to continue...")
|
|
# print("samples:",samples,)
|
|
|
|
if type(samples) == list:
|
|
pil_image = samples[0]
|
|
else:
|
|
pil_image = samples
|
|
|
|
sample_filename = f"{args.task_flag}_train_steps_{train_steps:07d}.png"
|
|
sample_filename_path = os.path.join(
|
|
sample_images_dir, sample_filename)
|
|
pil_image.save(sample_filename_path)
|
|
|
|
return None
|
|
|
|
|
|
def model_resume(args, model, ema, logger):
|
|
"""
|
|
Load pretrained weights.
|
|
"""
|
|
start_epoch = 0
|
|
start_epoch_step = 0
|
|
train_steps = 0
|
|
resume_path = UNET_PATH
|
|
|
|
logger.info(f"Resume from checkpoint {resume_path}")
|
|
|
|
if args.resume_split:
|
|
# Resume main model
|
|
|
|
resume_ckpt_module = torch.load(
|
|
resume_path, map_location=lambda storage, loc: storage)
|
|
model.load_state_dict(resume_ckpt_module, strict=False)
|
|
|
|
# Resume ema model
|
|
if args.use_ema:
|
|
if args.module_to_ema:
|
|
if "resume_ckpt_module" in locals():
|
|
logger.info(f" Resume ema model from main states.")
|
|
ema.load_state_dict(resume_ckpt_module, strict=args.strict)
|
|
else:
|
|
logger.info(f" Resume ema model from module states.")
|
|
resume_ckpt_module = torch.load(
|
|
resume_path, map_location=lambda storage, loc: storage)
|
|
ema.load_state_dict(resume_ckpt_module, strict=args.strict)
|
|
else:
|
|
if "resume_ckpt_ema" in locals():
|
|
logger.info(f" Resume ema model from EMA states.")
|
|
ema.load_state_dict(resume_ckpt_ema, strict=args.strict)
|
|
else:
|
|
logger.info(f" Resume ema model from EMA states.")
|
|
resume_ckpt_ema = torch.load(resume_path,
|
|
map_location=lambda storage, loc: storage)
|
|
ema.load_state_dict(resume_ckpt_ema, strict=args.strict)
|
|
else:
|
|
raise ValueError(
|
|
" “If `resume` is True, then either `resume_split` must be true.”")
|
|
|
|
return model, ema, start_epoch, start_epoch_step, train_steps
|
|
|
|
|
|
import tqdm
|
|
|
|
import requests
|
|
|
|
|
|
def set_master_port(port):
|
|
global master_port
|
|
master_port = port
|
|
|
|
|
|
master_port = 0
|
|
|
|
|
|
def LOG(log):
|
|
if master_port == 0:
|
|
raise Exception("master_port is 0")
|
|
# 发送http
|
|
try:
|
|
resp = requests.request("post", f"http://127.0.0.1:{master_port}/log", data=json.dumps(log), headers={
|
|
"Content-Type": "application/json"})
|
|
except Exception as e:
|
|
return
|
|
if resp.status_code != 200:
|
|
raise Exception(f"LOG failed: {resp.text}")
|
|
|
|
|
|
# with tqdm(total=total_steps, initial=train_steps) as pbar:
|
|
# pbar.update(1)
|
|
# pbar.set_description(
|
|
# f"Epoch {epoch}, step {step}, loss {loss.item():.4f}, mean_loss {mean_loss / step:.4f}")
|
|
class PBar:
|
|
def __init__(self, total):
|
|
self.pbar = tqdm.tqdm(total=total)
|
|
|
|
def step(self, desc, total_steps, train_steps):
|
|
self.pbar.update(1)
|
|
self.pbar.set_description(desc)
|
|
|
|
LOG({
|
|
"type": "sample_images",
|
|
"global_step": train_steps,
|
|
"total_steps": total_steps,
|
|
# "latent": noise_pred_latent_path,
|
|
})
|
|
|
|
|
|
from torch import nn
|
|
|
|
|
|
class CustomizeEmbedsModel(nn.Module):
|
|
dtype = torch.float16
|
|
# x = torch.zeros(1, 1, 256, 2048)
|
|
x = None
|
|
|
|
def __init__(self, *args, **kwargs):
|
|
super().__init__()
|
|
|
|
def to(self, *args, **kwargs):
|
|
self.dtype = torch.float16
|
|
return self
|
|
|
|
def forward(self, *args, **kwargs):
|
|
input_ids = kwargs.get("input_ids", None)
|
|
if self.x is None:
|
|
if input_ids is None:
|
|
batch_size = 1
|
|
else:
|
|
batch_size = input_ids.shape[0]
|
|
self.x = torch.zeros(1, batch_size, 256, 2048, dtype=self.dtype)
|
|
|
|
if kwargs.get("output_hidden_states", False):
|
|
return {
|
|
"hidden_states": self.x.to("cuda"),
|
|
"input_ids": torch.zeros(1, 1),
|
|
}
|
|
return self.x
|
|
|
|
|
|
class CustomizeTokenizer(dict):
|
|
|
|
added_tokens_encoder = []
|
|
input_ids = torch.zeros(1, 256)
|
|
attention_mask = torch.zeros(1, 256)
|
|
|
|
def __init__(self, *args, **kwargs):
|
|
self['added_tokens_encoder'] = self.added_tokens_encoder
|
|
self['input_ids'] = self.input_ids
|
|
self['attention_mask'] = self.attention_mask
|
|
|
|
def tokenize(self, text):
|
|
return text
|
|
|
|
def __call__(self, *args, **kwargs):
|
|
return self
|
|
|
|
|
|
class CustomizeEmbeds():
|
|
def __init__(self):
|
|
super().__init__()
|
|
self.tokenizer = CustomizeTokenizer()
|
|
self.model = CustomizeEmbedsModel().to("cuda")
|
|
self.max_length = 256
|