def main()

in scripts/edm_train.py [0:0]

• 44 lines of code
• 3 McCabe index (conditional complexity)

def main():
    args = create_argparser().parse_args()

    dist_util.setup_dist()
    logger.configure()

    logger.log("creating model and diffusion...")
    model, diffusion = create_model_and_diffusion(
        **args_to_dict(args, model_and_diffusion_defaults().keys())
    )
    model.to(dist_util.dev())
    schedule_sampler = create_named_schedule_sampler(args.schedule_sampler, diffusion)

    logger.log("creating data loader...")
    if args.batch_size == -1:
        batch_size = args.global_batch_size // dist.get_world_size()
        if args.global_batch_size % dist.get_world_size() != 0:
            logger.log(
                f"warning, using smaller global_batch_size of {dist.get_world_size()*batch_size} instead of {args.global_batch_size}"
            )
    else:
        batch_size = args.batch_size

    data = load_data(
        data_dir=args.data_dir,
        batch_size=batch_size,
        image_size=args.image_size,
        class_cond=args.class_cond,
    )

    logger.log("creating data loader...")

    logger.log("training...")
    TrainLoop(
        model=model,
        diffusion=diffusion,
        data=data,
        batch_size=batch_size,
        microbatch=args.microbatch,
        lr=args.lr,
        ema_rate=args.ema_rate,
        log_interval=args.log_interval,
        save_interval=args.save_interval,
        resume_checkpoint=args.resume_checkpoint,
        use_fp16=args.use_fp16,
        fp16_scale_growth=args.fp16_scale_growth,
        schedule_sampler=schedule_sampler,
        weight_decay=args.weight_decay,
        lr_anneal_steps=args.lr_anneal_steps,
    ).run_loop()