import sys import os from typing import Optional from mikazuki.tasks import tm from mikazuki.log import log def run_train(toml_path: str, trainer_file: str = "./sd-scripts/train_network.py", multi_gpu: bool = False, cpu_threads: Optional[int] = 2): log.info(f"Training started with config file / 训练开始,使用配置文件: {toml_path}") args = [ sys.executable, "-m", "accelerate.commands.launch", # use -m to avoid python script executable error "--num_cpu_threads_per_process", str(cpu_threads), # cpu threads "--quiet", # silence accelerate error message trainer_file, "--config_file", toml_path, ] if multi_gpu: args[3:3] = ["--multi_gpu", "--num_processes", "2"] customize_env = os.environ.copy() customize_env["ACCELERATE_DISABLE_RICH"] = "1" customize_env["PYTHONUNBUFFERED"] = "1" try: task = tm.create_task(args, customize_env) if not task: return result = task.communicate() if result.returncode != 0: log.error(f"Training failed / 训练失败") else: log.info(f"Training finished / 训练完成") except Exception as e: log.error(f"An error occurred when training / 创建训练进程时出现致命错误: {e}")