import torch from diffsynth.utils.data import save_video from diffsynth.pipelines.wan_video import WanVideoPipeline, ModelConfig import pandas as pd import ast import json import os import argparse from util import rgb_to_latent_shot_groups_list, pad_shot_groups_to_4n_plus_1, get_user_wanted_frames, save_video_with_caption import random, sys if __name__ == "__main__": parser = argparse.ArgumentParser() parser.add_argument("--test_csv_path", type=str, default=None) parser.add_argument("--model_path_json", type=str, default=None) parser.add_argument("--output_name", type=str, default="1.3b") parser.add_argument("--target_width", type=int, default=832) parser.add_argument("--target_height", type=int, default=480) parser.add_argument("--use_usp", type=bool, default=False) # [episode run] output named by episode_id, skip finished outputs, claim-based priority-order loop over several GPUs parser.add_argument("--output_folder", type=str, default=None) parser.add_argument("--claim_dir", type=str, default=None) parser.add_argument("--extra_csv_first", type=str, default=None) args = parser.parse_args() 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" # init load_path_json = args.model_path_json with open(load_path_json) as user_file: model_paths = json.load(user_file) pipe = WanVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", use_usp=args.use_usp, # if use usp for inference model_configs=[ ModelConfig(path=model_paths["dit"]), ModelConfig(path=model_paths["t5"]), ModelConfig(path=model_paths["vae"]), ], tokenizer_config=ModelConfig(path=model_paths["tokenizer"]), ) # read test cases df = pd.read_csv(args.test_csv_path) if args.extra_csv_first: # [episode run] process these rows first (smoke episode), then the main priority-ordered CSV df = pd.concat([pd.read_csv(args.extra_csv_first), df], ignore_index=True) output_dir = args.output_folder if args.output_folder else f"output/{args.output_name}" print(f"output_dir: {output_dir}") os.makedirs(output_dir, exist_ok=True) import datetime, subprocess def _ts(): return datetime.datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%S.%fZ") def _nframes(path): r = subprocess.run(["ffprobe", "-v", "error", "-count_frames", "-select_streams", "v:0", "-show_entries", "stream=nb_read_frames", "-of", "csv=p=0", path], capture_output=True, text=True) try: return int(r.stdout.strip()) except ValueError: return -1 os.makedirs(args.claim_dir, exist_ok=True) for index, row in df.iterrows(): episode_id = row["episode_id"] out_path = f"{output_dir}/{episode_id}.mp4" want_frames = ast.literal_eval(row["shot_groups"])[-1][-1] if os.path.exists(out_path) and _nframes(out_path) == want_frames: print(f"{_ts()} [EPISODE] skip {episode_id} (exists, {want_frames} frames)", flush=True) continue try: os.mkdir(os.path.join(args.claim_dir, episode_id)) # atomic claim: another GPU already took it -> next except FileExistsError: continue print(f"{_ts()} [EPISODE] start {episode_id} priority={row['priority']} seed=42", flush=True) torch.cuda.reset_peak_memory_stats() # do not use random seed when use_usp=True seed = 42 # random.randint(0, sys.maxsize) # read shot groups and transform to latent shot groups shot_groups = ast.literal_eval(row["shot_groups"]) padded_shot_groups, save_shot_num_list = pad_shot_groups_to_4n_plus_1(shot_groups) latent_shot_groups = rgb_to_latent_shot_groups_list(padded_shot_groups) # read multi-shot captions with open(row["gemini_caption"]) as user_file: caption_dict = json.load(user_file) num_shots = len(latent_shot_groups) # global caption global_caption = f"Story: {caption_dict['global_caption']} " # per-shot captions now_multishot_video_caption_list = [] for count in range(num_shots): now_shot_caption = global_caption + f"Now: {caption_dict[f'shot{count}']}" now_multishot_video_caption_list.append(now_shot_caption) multishot_negative_prompt = [negative_prompt] * num_shots # Text-to-video video = pipe( width=args.target_width, height=args.target_height, prompt=now_multishot_video_caption_list, negative_prompt=multishot_negative_prompt, seed=seed, tiled=True, num_frames=latent_shot_groups[-1][-1], shot_groups=shot_groups, latent_shot_groups=latent_shot_groups, ) user_wanted_frames = get_user_wanted_frames(video, padded_shot_groups, save_shot_num_list) # save video without caption save_video(user_wanted_frames, out_path, fps=15, quality=5) # save video with caption save_video_with_caption(num_shots, shot_groups, now_multishot_video_caption_list, user_wanted_frames, f"{output_dir}/{episode_id}_with_caption.mp4", args.target_width) print(f"{_ts()} [EPISODE] saved {episode_id} -> {out_path} peak_mem_gb={torch.cuda.max_memory_allocated()/2**30:.1f}", flush=True) print("Enjoy the story") torch.cuda.empty_cache()