supernet_train.py

import argparse
import datetime
import os
import sys
import numpy as np
import time
import torch
import torch.backends.cudnn as cudnn
import json
import yaml
from collections import defaultdict
from pathlib import Path
from pprint import pprint
from timm.data import Mixup
from timm.models import create_model
from timm.loss import LabelSmoothingCrossEntropy, SoftTargetCrossEntropy
from timm.scheduler import create_scheduler
from timm.optim import create_optimizer
#from timm.utils import NativeScaler
from lib.cuda import NativeScaler
from lib.datasets import build_dataset
from supernet_engine import train_one_epoch, evaluate
from lib.samplers import RASampler
from lib import utils
from lib.config import cfg, update_config_from_file
from model.supernet_transformer import Vision_TransformerSuper


def get_args_parser():
    parser = argparse.ArgumentParser('AutoFormer training and evaluation script', add_help=False)
    parser.add_argument('--batch-size', default=64, type=int)
    parser.add_argument('--epochs', default=300, type=int)
    # config file
    parser.add_argument('--cfg',help='experiment configure file name',required=True,type=str)

    # custom parameters
    parser.add_argument('--platform', default='pai', type=str, choices=['itp', 'pai', 'aml'],
                        help='Name of model to train')
    parser.add_argument('--teacher_model', default='', type=str,
                        help='Name of teacher model to train')
    parser.add_argument('--relative_position', action='store_true')
    parser.add_argument('--gp', action='store_true')
    parser.add_argument('--change_qkv', action='store_true')
    parser.add_argument('--max_relative_position', type=int, default=14, help='max distance in relative position embedding')

    # Model parameters
    parser.add_argument('--model', default='', type=str, metavar='MODEL',
                        help='Name of model to train')
    # AutoFormer config
    parser.add_argument('--mode', type=str, default='super', choices=['super', 'retrain'], help='mode of AutoFormer')
    parser.add_argument('--input-size', default=224, type=int)
    parser.add_argument('--patch_size', default=16, type=int)

    parser.add_argument('--drop', type=float, default=0.0, metavar='PCT',
                        help='Dropout rate (default: 0.)')
    parser.add_argument('--drop-path', type=float, default=0.1, metavar='PCT',
                        help='Drop path rate (default: 0.1)')
    parser.add_argument('--drop-block', type=float, default=None, metavar='PCT',
                        help='Drop block rate (default: None)')

    parser.add_argument('--model-ema', action='store_true')
    parser.add_argument('--no-model-ema', action='store_false', dest='model_ema')
    # parser.set_defaults(model_ema=True)
    parser.add_argument('--model-ema-decay', type=float, default=0.99996, help='')
    parser.add_argument('--model-ema-force-cpu', action='store_true', default=False, help='')
    parser.add_argument('--rpe_type', type=str, default='bias', choices=['bias', 'direct'])
    parser.add_argument('--post_norm', action='store_true')
    parser.add_argument('--no_abs_pos', action='store_true')

    # Optimizer parameters
    parser.add_argument('--opt', default='adamw', type=str, metavar='OPTIMIZER',
                        help='Optimizer (default: "adamw"')
    parser.add_argument('--opt-eps', default=1e-8, type=float, metavar='EPSILON',
                        help='Optimizer Epsilon (default: 1e-8)')
    parser.add_argument('--opt-betas', default=None, type=float, nargs='+', metavar='BETA',
                        help='Optimizer Betas (default: None, use opt default)')
    parser.add_argument('--clip-grad', type=float, default=None, metavar='NORM',
                        help='Clip gradient norm (default: None, no clipping)')
    parser.add_argument('--momentum', type=float, default=0.9, metavar='M',
                        help='SGD momentum (default: 0.9)')
    parser.add_argument('--weight-decay', type=float, default=0.05,
                        help='weight decay (default: 0.05)')

    # Learning rate schedule parameters
    parser.add_argument('--sched', default='cosine', type=str, metavar='SCHEDULER',
                        help='LR scheduler (default: "cosine"')
    parser.add_argument('--lr', type=float, default=5e-4, metavar='LR',
                        help='learning rate (default: 5e-4)')
    parser.add_argument('--lr-noise', type=float, nargs='+', default=None, metavar='pct, pct',
                        help='learning rate noise on/off epoch percentages')
    parser.add_argument('--lr-noise-pct', type=float, default=0.67, metavar='PERCENT',
                        help='learning rate noise limit percent (default: 0.67)')
    parser.add_argument('--lr-noise-std', type=float, default=1.0, metavar='STDDEV',
                        help='learning rate noise std-dev (default: 1.0)')
    parser.add_argument('--warmup-lr', type=float, default=1e-6, metavar='LR',
                        help='warmup learning rate (default: 1e-6)')
    parser.add_argument('--min-lr', type=float, default=1e-5, metavar='LR',
                        help='lower lr bound for cyclic schedulers that hit 0 (1e-5)')
    parser.add_argument('--lr-power', type=float, default=1.0,
                        help='power of the polynomial lr scheduler')

    parser.add_argument('--decay-epochs', type=float, default=30, metavar='N',
                        help='epoch interval to decay LR')
    parser.add_argument('--warmup-epochs', type=int, default=5, metavar='N',
                        help='epochs to warmup LR, if scheduler supports')
    parser.add_argument('--cooldown-epochs', type=int, default=10, metavar='N',
                        help='epochs to cooldown LR at min_lr, after cyclic schedule ends')
    parser.add_argument('--patience-epochs', type=int, default=10, metavar='N',
                        help='patience epochs for Plateau LR scheduler (default: 10')
    parser.add_argument('--decay-rate', '--dr', type=float, default=0.1, metavar='RATE',
                        help='LR decay rate (default: 0.1)')

    # Augmentation parameters
    parser.add_argument('--color-jitter', type=float, default=0.4, metavar='PCT',
                        help='Color jitter factor (default: 0.4)')
    parser.add_argument('--aa', type=str, default='rand-m9-mstd0.5-inc1', metavar='NAME',
                        help='Use AutoAugment policy. "v0" or "original". " + \
                             "(default: rand-m9-mstd0.5-inc1)'),
    parser.add_argument('--smoothing', type=float, default=0.1, help='Label smoothing (default: 0.1)')
    parser.add_argument('--train-interpolation', type=str, default='bicubic',
                        help='Training interpolation (random, bilinear, bicubic default: "bicubic")')

    parser.add_argument('--repeated-aug', action='store_true')
    parser.add_argument('--no-repeated-aug', action='store_false', dest='repeated_aug')


    parser.set_defaults(repeated_aug=True)

    # * Random Erase params
    parser.add_argument('--reprob', type=float, default=0.25, metavar='PCT',
                        help='Random erase prob (default: 0.25)')
    parser.add_argument('--remode', type=str, default='pixel',
                        help='Random erase mode (default: "pixel")')
    parser.add_argument('--recount', type=int, default=1,
                        help='Random erase count (default: 1)')
    parser.add_argument('--resplit', action='store_true', default=False,
                        help='Do not random erase first (clean) augmentation split')

    # * Mixup params
    parser.add_argument('--mixup', type=float, default=0.8,
                        help='mixup alpha, mixup enabled if > 0. (default: 0.8)')
    parser.add_argument('--cutmix', type=float, default=1.0,
                        help='cutmix alpha, cutmix enabled if > 0. (default: 1.0)')
    parser.add_argument('--cutmix-minmax', type=float, nargs='+', default=None,
                        help='cutmix min/max ratio, overrides alpha and enables cutmix if set (default: None)')
    parser.add_argument('--mixup-prob', type=float, default=1.0,
                        help='Probability of performing mixup or cutmix when either/both is enabled')
    parser.add_argument('--mixup-switch-prob', type=float, default=0.5,
                        help='Probability of switching to cutmix when both mixup and cutmix enabled')
    parser.add_argument('--mixup-mode', type=str, default='batch',
                        help='How to apply mixup/cutmix params. Per "batch", "pair", or "elem"')

    # Dataset parameters
    parser.add_argument('--data-path', default='./data/imagenet/', type=str,
                        help='dataset path')
    parser.add_argument('--data-set', default='IMNET', choices=['CIFAR', 'IMNET', 'INAT', 'INAT19', 'EVO_IMNET'],
                        type=str, help='Image Net dataset path')
    parser.add_argument('--inat-category', default='name',
                        choices=['kingdom', 'phylum', 'class', 'order', 'supercategory', 'family', 'genus', 'name'],
                        type=str, help='semantic granularity')

    parser.add_argument('--output_dir', default='./logs/',
                        help='path where to save, empty for no saving')
    parser.add_argument('--task', default='', help='task prefix')
    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('--num_workers', default=10, type=int)
    parser.add_argument('--dist-eval', action='store_true', default=False, help='Enabling distributed evaluation')
    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',
                        help='')
    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('--dist_url', default='env://', help='url used to set up distributed training')

    parser.add_argument('--amp', action='store_true')
    parser.add_argument('--no-amp', action='store_false', dest='amp')
    parser.set_defaults(amp=True)

    parser.add_argument('--print2file', action='store_true', default=False, help='save stdout to file')
    parser.add_argument('--candfile', default='', type=str, help='candidates json file')
    parser.add_argument('--group-by-dim', action='store_true', default=False, help='group candidates by embed_dim')
    parser.add_argument('--group-by-depth', action='store_true', default=False, help='group candidates by depth')
    parser.add_argument('--sandwich', default=0, type=int, help='number of interlayers in sandwich, default 0 to turn off')
    parser.add_argument('--no-sandwich-base', action='store_true', default=False, help='remove the base layer of sandwich')
    parser.add_argument('--no-sandwich-top', action='store_true', default=False, help='remove the top layer of sandwich')
    parser.add_argument('--switch-ln', action='store_true', default=False, help='Enabling switchable layernorm')
    parser.add_argument('--scale-attn', action='store_true', default=False, help='scale attention')
    parser.add_argument('--scale-mlp', action='store_true', default=False, help='scale mlp')
    parser.add_argument('--scale-embed', action='store_true', default=False, help='scale embed dim')
    parser.add_argument('--shuffle', action='store_true', default=False, help='shuffle chosen candidate')
    parser.add_argument('--eval-crops', default=1, type=int, choices=[1, 5, 10], help='number of crops for evaluation')

    return parser

def main(args):

    utils.init_distributed_mode(args)
    update_config_from_file(args.cfg)

    print(args)
    args_text = yaml.safe_dump(args.__dict__, default_flow_style=False)

    if args.print2file:
        sys._stdout = sys.stdout
        sys.stdout = open(os.path.join(args.output_dir, f'out{utils.get_rank()}'), 'w', buffering=1)

    device = torch.device(args.device)

    # fix the seed for reproducibility
    seed = args.seed + utils.get_rank()
    torch.manual_seed(seed)
    np.random.seed(seed)
    # random.seed(seed)
    cudnn.benchmark = True

    dataset_train, args.nb_classes = build_dataset(is_train=True, args=args, folder_name="subImageNet" if args.data_set == "EVO_IMNET" else "train")
    dataset_val, _ = build_dataset(is_train=False, args=args, folder_name="val")

    if args.distributed:
        num_tasks = utils.get_world_size()
        global_rank = utils.get_rank()
        if args.repeated_aug:
            sampler_train = RASampler(
                dataset_train, num_replicas=num_tasks, rank=global_rank, shuffle=True
            )
        else:
            sampler_train = torch.utils.data.DistributedSampler(
                dataset_train, num_replicas=num_tasks, rank=global_rank, shuffle=True
            )
        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=False)
        else:
            sampler_val = torch.utils.data.SequentialSampler(dataset_val)
    else:
        sampler_val = torch.utils.data.SequentialSampler(dataset_val)
        sampler_train = torch.utils.data.RandomSampler(dataset_train)

    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,
    )

    data_loader_val = torch.utils.data.DataLoader(
        dataset_val, batch_size=int(2 * args.batch_size),
        sampler=sampler_val, num_workers=args.num_workers,
        pin_memory=args.pin_mem, drop_last=False
    )

    mixup_fn = None
    mixup_active = args.mixup > 0 or args.cutmix > 0. or args.cutmix_minmax is not None
    if mixup_active:
        mixup_fn = Mixup(
            mixup_alpha=args.mixup, cutmix_alpha=args.cutmix, cutmix_minmax=args.cutmix_minmax,
            prob=args.mixup_prob, switch_prob=args.mixup_switch_prob, mode=args.mixup_mode,
            label_smoothing=args.smoothing, num_classes=args.nb_classes)

    print(f"Creating SuperVisionTransformer")
    print(cfg)

    choices = {'num_heads': cfg.SEARCH_SPACE.NUM_HEADS, 'mlp_ratio': cfg.SEARCH_SPACE.MLP_RATIO,
               'embed_dim': cfg.SEARCH_SPACE.EMBED_DIM , 'depth': cfg.SEARCH_SPACE.DEPTH}

    model = Vision_TransformerSuper(img_size=args.input_size,
                                    patch_size=args.patch_size,
                                    embed_dim=cfg.SUPERNET.EMBED_DIM, depth=cfg.SUPERNET.DEPTH,
                                    num_heads=cfg.SUPERNET.NUM_HEADS,mlp_ratio=cfg.SUPERNET.MLP_RATIO,
                                    qkv_bias=True, drop_rate=args.drop,
                                    drop_path_rate=args.drop_path,
                                    gp=args.gp,
                                    num_classes=args.nb_classes,
                                    max_relative_position=args.max_relative_position,
                                    relative_position=args.relative_position,
                                    change_qkv=args.change_qkv, abs_pos=not args.no_abs_pos,
                                    choices=choices, switch_ln=args.switch_ln,
                                    scale_attn=args.scale_attn, scale_mlp=args.scale_mlp,
                                    scale_embed=args.scale_embed)

    model.to(device)
    if args.teacher_model:
        teacher_model = create_model(
            args.teacher_model,
            pretrained=True,
            num_classes=args.nb_classes,
        )
        teacher_model.to(device)
        teacher_loss = LabelSmoothingCrossEntropy(smoothing=args.smoothing)
    else:
        teacher_model = None
        teacher_loss = None

    model_ema = None

    print(model)

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

    n_parameters = sum(p.numel() for p in model.parameters() if p.requires_grad)
    print('number of params:', n_parameters)

    linear_scaled_lr = args.lr * args.batch_size * utils.get_world_size() / 512.0
    args.lr = linear_scaled_lr
    optimizer = create_optimizer(args, model_without_ddp)
    loss_scaler = NativeScaler()
    lr_scheduler, _ = create_scheduler(args, optimizer)

    # criterion = LabelSmoothingCrossEntropy()

    if args.mixup > 0.:
        # smoothing is handled with mixup label transform
        criterion = SoftTargetCrossEntropy()
    elif args.smoothing:
        criterion = LabelSmoothingCrossEntropy(smoothing=args.smoothing)
    else:
        criterion = torch.nn.CrossEntropyLoss()

    output_dir = Path(args.output_dir)

    if not output_dir.exists():
        output_dir.mkdir(parents=True)
    # save config for later experiments
    if args.output_dir and utils.is_main_process():
        with open(output_dir / "config.yaml", 'w') as f:
            f.write(args_text)
    if args.resume:
        if args.resume.startswith('https'):
            checkpoint = torch.hub.load_state_dict_from_url(
                args.resume, map_location='cpu', check_hash=True)
        else:
            checkpoint = torch.load(args.resume, map_location='cpu')
        model_without_ddp.load_state_dict(checkpoint['model'])
        if not args.eval and 'optimizer' in checkpoint and 'lr_scheduler' in checkpoint and 'epoch' in checkpoint:
            optimizer.load_state_dict(checkpoint['optimizer'])
            lr_scheduler.load_state_dict(checkpoint['lr_scheduler'])
            args.start_epoch = checkpoint['epoch'] + 1
            if 'scaler' in checkpoint:
                loss_scaler.load_state_dict(checkpoint['scaler'])
            if args.model_ema:
                utils._load_checkpoint_for_ema(model_ema, checkpoint['model_ema'])

    retrain_config = None
    if args.mode == 'retrain' and "RETRAIN" in cfg:
        retrain_config = {'layer_num': cfg.RETRAIN.DEPTH, 'embed_dim': [cfg.RETRAIN.EMBED_DIM]*cfg.RETRAIN.DEPTH,
                          'num_heads': cfg.RETRAIN.NUM_HEADS,'mlp_ratio': cfg.RETRAIN.MLP_RATIO}
    candidates = None
    cand_index_group = None
    if args.candfile:
        cand_dict = json.load(open(args.candfile))
        candidates = [cand for cand_list in cand_dict.values() for cand in cand_list]
        if args.group_by_dim or args.group_by_depth:
            cand_index_group = defaultdict(list)
            for idx, cand in enumerate(candidates):
                k = ()
                k = k + (cand['embed_dim'][0],) if args.group_by_dim else k
                k = k + (cand['layer_num'],) if args.group_by_depth else k
                cand_index_group[k].append(idx)
            cand_index_group = list(cand_index_group.values())

    if args.eval:
        if args.candfile:
            batch_stats = []
            for retrain_config in candidates:
                test_stats = evaluate(data_loader_val, model, device,  mode = args.mode, retrain_config=retrain_config, print2file=args.print2file, eval_crops=args.eval_crops)
                print(f"Accuracy of the network on the {len(dataset_val)} test images: {test_stats['acc1']:.1f}%")
                test_stats['config'] = retrain_config
                batch_stats.append(test_stats)
            if args.output_dir and utils.is_main_process():
                json.dump(batch_stats, open(output_dir / 'results.json', 'w'), indent=2)
        else:
            test_stats = evaluate(data_loader_val, model, device,  mode = args.mode, retrain_config=retrain_config, print2file=args.print2file, eval_crops=args.eval_crops)
            print(f"Accuracy of the network on the {len(dataset_val)} test images: {test_stats['acc1']:.1f}%")

        return

    print("Start training")
    start_time = time.time()
    max_accuracy = 0.0

    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, criterion, data_loader_train,
            optimizer, device, epoch, loss_scaler,
            args.clip_grad, model_ema, mixup_fn,
            amp=args.amp, teacher_model=teacher_model,
            teach_loss=teacher_loss,
            choices=choices, mode = args.mode, retrain_config=retrain_config,
            print2file=args.print2file, candidates=candidates, sandwich=args.sandwich,
            sandwich_base=not args.no_sandwich_base, sandwich_top=not args.no_sandwich_top,
            shuffle=args.shuffle, grouping=cand_index_group,
        )

        lr_scheduler.step(epoch)
        if args.output_dir and (epoch + 1) % 10 == 0:
            checkpoint_paths = [output_dir / 'checkpoint-{}.pth'.format(epoch+1)]
            for checkpoint_path in checkpoint_paths:
                utils.save_on_master({
                    'model': model_without_ddp.state_dict(),
                    'optimizer': optimizer.state_dict(),
                    'lr_scheduler': lr_scheduler.state_dict(),
                    'epoch': epoch,
                    # 'model_ema': get_state_dict(model_ema),
                    'scaler': loss_scaler.state_dict(),
                    'args': args,
                }, checkpoint_path)

        test_stats = evaluate(data_loader_val, model, device, amp=args.amp, choices=choices, mode = args.mode, retrain_config=retrain_config, print2file=args.print2file, candidates=candidates, eval_crops=args.eval_crops)
        print(f"Accuracy of the network on the {len(dataset_val)} test images: {test_stats['acc1']:.1f}%")
        max_accuracy = max(max_accuracy, test_stats["acc1"])
        print(f'Max accuracy: {max_accuracy:.2f}%')

        log_stats = {'datetime': datetime.datetime.now().strftime("%m/%d %H:%M"),
                     **{f'train_{k}': v for k, v in train_stats.items()},
                     **{f'test_{k}': v for k, v in test_stats.items()},
                     'epoch': epoch}

        if args.output_dir and utils.is_main_process():
            with (output_dir / "log.txt").open("a") 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__':
    now = datetime.datetime.now().strftime("%Y_%m_%d-%H_%M_%S") 
    parser = argparse.ArgumentParser('AutoFormer training and evaluation script', parents=[get_args_parser()])
    args = parser.parse_args()
    if args.task:
        suffix = args.task
    elif args.resume:
        resume_folder = os.path.basename(os.path.dirname(os.path.normpath(args.resume)))
        suffix = resume_folder.partition('@')[-1]
    else:
        suffix = ''
    sep = '@' if suffix else ''
    args.output_dir = os.path.join(args.output_dir, 'test' if args.eval else 'train', now+sep+suffix)
    Path(args.output_dir).mkdir(parents=True, exist_ok=True)
    main(args)