main_class_cond.py

import argparse
import datetime
import json
import numpy as np
import os
import time
from pathlib import Path

import torch
import torch.backends.cudnn as cudnn
from torch.utils.tensorboard import SummaryWriter

import util.lr_decay as lrd
import util.misc as misc
from util.datasets import build_shape_surface_occupancy_dataset
from util.misc import NativeScalerWithGradNormCount as NativeScaler

import models_class_cond, models_ae

from engine_class_cond import train_one_epoch, evaluate

def get_args_parser():
    parser = argparse.ArgumentParser('Latent Diffusion', add_help=False)
    parser.add_argument('--batch_size', default=64, type=int,
                        help='Batch size per GPU (effective batch size is batch_size * accum_iter * # gpus')
    parser.add_argument('--epochs', default=800, type=int)
    parser.add_argument('--accum_iter', default=1, type=int,
                        help='Accumulate gradient iterations (for increasing the effective batch size under memory constraints)')

    # Model parameters
    parser.add_argument('--model', default='kl_d512_m512_l8_edm', type=str, metavar='MODEL',
                        help='Name of model to train')

    parser.add_argument('--ae', default='kl_d512_m512_l8', type=str, metavar='MODEL',
                        help='Name of autoencoder')

    parser.add_argument('--ae-pth', required=True, help='Autoencoder checkpoint')

    parser.add_argument('--point_cloud_size', default=2048, type=int,
                        help='input size')

    # Optimizer parameters
    parser.add_argument('--clip_grad', type=float, default=None, metavar='NORM',
                        help='Clip gradient norm (default: None, no clipping)')
    parser.add_argument('--weight_decay', type=float, default=0.05,
                        help='weight decay (default: 0.05)')

    parser.add_argument('--lr', type=float, default=None, metavar='LR',
                        help='learning rate (absolute lr)')
    parser.add_argument('--blr', type=float, default=1e-4, metavar='LR', # 2e-4
                        help='base learning rate: absolute_lr = base_lr * total_batch_size / 256')
    parser.add_argument('--layer_decay', type=float, default=0.75,
                        help='layer-wise lr decay from ELECTRA/BEiT')

    parser.add_argument('--min_lr', type=float, default=1e-6, metavar='LR',
                        help='lower lr bound for cyclic schedulers that hit 0')

    parser.add_argument('--warmup_epochs', type=int, default=40, metavar='N',
                        help='epochs to warmup LR')


    # Dataset parameters
    parser.add_argument('--data_path', default='/ibex/ai/home/zhanb0b/data', type=str,
                        help='dataset path')

    parser.add_argument('--output_dir', default='./output/',
                        help='path where to save, empty for no saving')
    parser.add_argument('--log_dir', default='./output/',
                        help='path where to tensorboard log')
    parser.add_argument('--device', default='cuda',
                        help='device to use for training / testing')
    parser.add_argument('--seed', default=0, type=int)
    parser.add_argument('--resume', default='',
                        help='resume from checkpoint')

    parser.add_argument('--start_epoch', default=0, type=int, metavar='N',
                        help='start epoch')
    parser.add_argument('--eval', action='store_true',
                        help='Perform evaluation only')
    parser.add_argument('--dist_eval', action='store_true', default=False,
                        help='Enabling distributed evaluation (recommended during training for faster monitor')
    parser.add_argument('--num_workers', default=60, type=int)
    parser.add_argument('--pin_mem', action='store_true',
                        help='Pin CPU memory in DataLoader for more efficient (sometimes) transfer to GPU.')
    parser.add_argument('--no_pin_mem', action='store_false', dest='pin_mem')
    parser.set_defaults(pin_mem=True)

    # distributed training parameters
    parser.add_argument('--world_size', default=1, type=int,
                        help='number of distributed processes')
    parser.add_argument('--local_rank', default=-1, type=int)
    parser.add_argument('--dist_on_itp', action='store_true')
    parser.add_argument('--dist_url', default='env://',
                        help='url used to set up distributed training')

    return parser

def main(args):
    misc.init_distributed_mode(args)

    print('job dir: {}'.format(os.path.dirname(os.path.realpath(__file__))))
    print("{}".format(args).replace(', ', ',\n'))

    device = torch.device(args.device)

    # fix the seed for reproducibility
    seed = args.seed + misc.get_rank()
    torch.manual_seed(seed)
    np.random.seed(seed)

    cudnn.benchmark = True

    dataset_train = build_shape_surface_occupancy_dataset('train', args=args)
    dataset_val = build_shape_surface_occupancy_dataset('val', args=args)

    if True:  # args.distributed:
        num_tasks = misc.get_world_size()
        global_rank = misc.get_rank()
        sampler_train = torch.utils.data.DistributedSampler(
            dataset_train, num_replicas=num_tasks, rank=global_rank, shuffle=True
        )
        print("Sampler_train = %s" % str(sampler_train))
        if args.dist_eval:
            if len(dataset_val) % num_tasks != 0:
                print('Warning: Enabling distributed evaluation with an eval dataset not divisible by process number. '
                      'This will slightly alter validation results as extra duplicate entries are added to achieve '
                      'equal num of samples per-process.')
            sampler_val = torch.utils.data.DistributedSampler(
                dataset_val, num_replicas=num_tasks, rank=global_rank, shuffle=True)  # shuffle=True to reduce monitor bias
        else:
            sampler_val = torch.utils.data.SequentialSampler(dataset_val)
    else:
        sampler_train = torch.utils.data.RandomSampler(dataset_train)
        sampler_val = torch.utils.data.SequentialSampler(dataset_val)

    if global_rank == 0 and args.log_dir is not None and not args.eval:
        os.makedirs(args.log_dir, exist_ok=True)
        log_writer = SummaryWriter(log_dir=args.log_dir)
    else:
        log_writer = None

    data_loader_train = torch.utils.data.DataLoader(
        dataset_train, sampler=sampler_train,
        batch_size=args.batch_size,
        num_workers=args.num_workers,
        pin_memory=args.pin_mem,
        drop_last=True,
        # prefetch_factor=2,
    )

    data_loader_val = torch.utils.data.DataLoader(
        dataset_val, sampler=sampler_val,
        # batch_size=args.batch_size,
        batch_size=1,
        num_workers=args.num_workers,
        pin_memory=args.pin_mem,
        drop_last=False
    )

    ae = models_ae.__dict__[args.ae]()
    ae.eval()
    print("Loading autoencoder %s" % args.ae_pth)
    ae.load_state_dict(torch.load(args.ae_pth, map_location='cpu')['model'])
    
    ae.to(device)

    model = models_class_cond.__dict__[args.model]()
    model.to(device)

    model_without_ddp = model
    n_parameters = sum(p.numel() for p in model.parameters() if p.requires_grad)

    print("Model = %s" % str(model_without_ddp))
    print('number of params (M): %.2f' % (n_parameters / 1.e6))

    eff_batch_size = args.batch_size * args.accum_iter * misc.get_world_size()
    
    if args.lr is None:  # only base_lr is specified
        args.lr = args.blr * eff_batch_size / 256

    print("base lr: %.2e" % (args.lr * 256 / eff_batch_size))
    print("actual lr: %.2e" % args.lr)

    print("accumulate grad iterations: %d" % args.accum_iter)
    print("effective batch size: %d" % eff_batch_size)

    if args.distributed:
        model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.gpu], find_unused_parameters=False)
        model_without_ddp = model.module

    # # build optimizer with layer-wise lr decay (lrd)
    # param_groups = lrd.param_groups_lrd(model_without_ddp, args.weight_decay,
    #     no_weight_decay_list=model_without_ddp.no_weight_decay(),
    #     layer_decay=args.layer_decay
    # )
    optimizer = torch.optim.AdamW(model_without_ddp.parameters(), lr=args.lr)
    loss_scaler = NativeScaler()

    criterion = models_class_cond.__dict__['EDMLoss']()

    print("criterion = %s" % str(criterion))

    misc.load_model(args=args, model_without_ddp=model_without_ddp, optimizer=optimizer, loss_scaler=loss_scaler)

    if args.eval:
        test_stats = evaluate(data_loader_val, model, device)
        print(f"loss of the network on the {len(dataset_val)} test images: {test_stats['loss']:.3f}")
        exit(0)

    print(f"Start training for {args.epochs} epochs")
    start_time = time.time()
    for epoch in range(args.start_epoch, args.epochs):
        if args.distributed:
            data_loader_train.sampler.set_epoch(epoch)
        train_stats = train_one_epoch(
            model, ae, criterion, data_loader_train,
            optimizer, device, epoch, loss_scaler,
            args.clip_grad,
            log_writer=log_writer,
            args=args
        )
        if args.output_dir and (epoch % 10 == 0 or epoch + 1 == args.epochs):
            misc.save_model(
                args=args, model=model, model_without_ddp=model_without_ddp, optimizer=optimizer,
                loss_scaler=loss_scaler, epoch=epoch)

        if epoch % 5 == 0 or epoch + 1 == args.epochs:
            test_stats = evaluate(data_loader_val, model, ae, criterion, device)
            print(f"loss of the network on the {len(dataset_val)} test images: {test_stats['loss']:.3f}")

            if log_writer is not None:
                log_writer.add_scalar('perf/test_loss', test_stats['loss'], epoch)

            log_stats = {**{f'train_{k}': v for k, v in train_stats.items()},
                            **{f'test_{k}': v for k, v in test_stats.items()},
                            'epoch': epoch,
                            'n_parameters': n_parameters}

            # if args.output_dir and misc.is_main_process():
            #     if log_writer is not None:
            #         log_writer.flush()
            #     with open(os.path.join(args.output_dir, "log.txt"), mode="a", encoding="utf-8") as f:
            #         f.write(json.dumps(log_stats) + "\n")
        
        else:
            log_stats = {**{f'train_{k}': v for k, v in train_stats.items()},
                            'epoch': epoch,
                            'n_parameters': n_parameters}

        if args.output_dir and misc.is_main_process():
            if log_writer is not None:
                log_writer.flush()
            with open(os.path.join(args.output_dir, "log.txt"), mode="a", encoding="utf-8") as f:
                f.write(json.dumps(log_stats) + "\n")

    total_time = time.time() - start_time
    total_time_str = str(datetime.timedelta(seconds=int(total_time)))
    print('Training time {}'.format(total_time_str))

if __name__ == '__main__':
    args = get_args_parser()
    args = args.parse_args()
    if args.output_dir:
        Path(args.output_dir).mkdir(parents=True, exist_ok=True)
    main(args)