FSC_pretrain.py

import argparse
import datetime
import json

import PIL.Image
import numpy as np
import os
import time
import random
from pathlib import Path
import math
import sys
from PIL import Image

import torch
import torch.backends.cudnn as cudnn
from torch.utils.tensorboard import SummaryWriter
import torch.nn.functional as F
from torch.utils.data import Dataset
import wandb
import timm

assert "0.4.5" <= timm.__version__ <= "0.4.9"  # version check
import timm.optim.optim_factory as optim_factory

import util.misc as misc
from util.misc import NativeScalerWithGradNormCount as NativeScaler
import util.lr_sched as lr_sched
from util.FSC147 import transform_pre_train
import models_mae_noct


def get_args_parser():
    parser = argparse.ArgumentParser('MAE pre-training', add_help=False)
    parser.add_argument('--batch_size', default=8, type=int,
                        help='Batch size per GPU (effective batch size is batch_size * accum_iter * # gpus')
    parser.add_argument('--epochs', default=200, 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='mae_vit_base_patch16', type=str, metavar='MODEL',
                        help='Name of model to train')

    parser.add_argument('--mask_ratio', default=0.5, type=float,
                        help='Masking ratio (percentage of removed patches).')

    parser.add_argument('--norm_pix_loss', action='store_true',
                        help='Use (per-patch) normalized pixels as targets for computing loss')
    parser.set_defaults(norm_pix_loss=False)

    # Optimizer parameters
    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-3, metavar='LR',
                        help='base learning rate: absolute_lr = base_lr * total_batch_size / 256')
    parser.add_argument('--min_lr', type=float, default=0., metavar='LR',
                        help='lower lr bound for cyclic schedulers that hit 0')
    parser.add_argument('--warmup_epochs', type=int, default=10, metavar='N',
                        help='epochs to warmup LR')

    # Dataset parameters
    parser.add_argument('--data_path', default='./data/FSC147/', type=str,
                        help='dataset path')
    parser.add_argument('--anno_file', default='annotation_FSC147_384.json', type=str,
                        help='annotation json file')
    parser.add_argument('--data_split_file', default='Train_Test_Val_FSC_147.json', type=str,
                        help='data split json file')
    parser.add_argument('--im_dir', default='images_384_VarV2', type=str,
                        help='images directory')
    parser.add_argument('--gt_dir', default='gt_density_map_adaptive_384_VarV2', type=str,
                        help='ground truth directory')
    parser.add_argument('--output_dir', default='./data/out/pre_4_dir',
                        help='path where to save, empty for no saving')
    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='./weights/mae_pretrain_vit_base_full.pth',  # mae_visualize_vit_base
                        help='resume from checkpoint')

    # Training parameters
    parser.add_argument('--start_epoch', default=0, type=int, metavar='N',
                        help='start epoch')
    parser.add_argument('--num_workers', default=10, 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')

    # Logging parameters
    parser.add_argument('--log_dir', default='./logs/pre_4_dir',
                        help='path where to tensorboard log')
    parser.add_argument("--title", default="CounTR_pretraining", type=str)
    parser.add_argument("--wandb", default="counting", type=str)
    parser.add_argument("--team", default="wsense", type=str)
    parser.add_argument("--wandb_id", default=None, type=str)

    return parser


os.environ["CUDA_LAUNCH_BLOCKING"] = '1'


class TrainData(Dataset):
    def __init__(self):
        self.img = data_split['train']
        random.shuffle(self.img)
        self.img_dir = im_dir
        self.TransformPreTrain = transform_pre_train(data_path)

    def __len__(self):
        return len(self.img)

    def __getitem__(self, idx):
        im_id = self.img[idx]
        anno = annotations[im_id]
        bboxes = anno['box_examples_coordinates']

        rects = list()
        for bbox in bboxes:
            x1 = bbox[0][0]
            y1 = bbox[0][1]
            x2 = bbox[2][0]
            y2 = bbox[2][1]
            rects.append([y1, x1, y2, x2])

        image = Image.open('{}/{}'.format(im_dir, im_id))
        image.load()
        density_path = gt_dir / (im_id.split(".jpg")[0] + ".npy")
        density = np.load(density_path).astype('float32')
        sample = {'image': image, 'lines_boxes': rects, 'gt_density': density}
        sample = self.TransformPreTrain(sample)
        return sample['image']


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 = TrainData()
    print(dataset_train)

    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))
    else:
        sampler_train = torch.utils.data.RandomSampler(dataset_train)

    if global_rank == 0:
        if args.log_dir is not None:
            os.makedirs(args.log_dir, exist_ok=True)
            log_writer = SummaryWriter(log_dir=args.log_dir)
        else:
            log_writer = None
        if args.wandb is not None:
            wandb_run = wandb.init(
                config=args,
                resume="allow",
                project=args.wandb,
                name=args.title,
                entity=args.team,
                tags=["CounTR", "pretraining"],
                id=args.wandb_id,
            )
        else:
            wandb_run = 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=False,
    )

    # define the model
    model = models_mae_noct.__dict__[args.model](norm_pix_loss=args.norm_pix_loss)

    model.to(device)

    model_without_ddp = model

    print("Model = %s" % str(model_without_ddp))

    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=True)
        model_without_ddp = model.module

    # following timm: set wd as 0 for bias and norm layers
    param_groups = optim_factory.add_weight_decay(model_without_ddp, args.weight_decay)
    optimizer = torch.optim.AdamW(param_groups, lr=args.lr, betas=(0.9, 0.95))
    print(optimizer)
    loss_scaler = NativeScaler()

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

    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 one epoch
        model.train(True)
        metric_logger = misc.MetricLogger(delimiter="  ")
        metric_logger.add_meter('lr', misc.SmoothedValue(window_size=1, fmt='{value:.6f}'))
        header = 'Epoch: [{}]'.format(epoch)
        print_freq = 20
        accum_iter = args.accum_iter

        optimizer.zero_grad()

        if log_writer is not None:
            print('log_dir: {}'.format(log_writer.log_dir))

        model_ = getattr(models_mae_noct, args.model)()

        for data_iter_step, samples in enumerate(metric_logger.log_every(data_loader_train, print_freq, header)):
            epoch_1000x = int((data_iter_step / len(data_loader_train) + epoch) * 1000)

            if data_iter_step % accum_iter == 0:
                lr_sched.adjust_learning_rate(optimizer, data_iter_step / len(data_loader_train) + epoch, args)

            samples = samples.to(device, non_blocking=True)

            with torch.cuda.amp.autocast():
                loss, pred, mask = model(samples, mask_ratio=args.mask_ratio)

            loss_value = loss.item()

            if data_iter_step % 2000 == 0:
                preds = model_.unpatchify(pred)
                preds = preds.float()
                preds = torch.einsum('nchw->nhwc', preds)
                preds = torch.clip(preds, 0, 1)

                if log_writer is not None:
                    log_writer.add_images('reconstruction', preds, int(epoch), dataformats='NHWC')

                if wandb_run is not None:
                    wandb_images = []
                    w_samples = torch.einsum('nchw->nhwc', samples.float()).clip(0, 1)
                    masks = F.interpolate(
                        mask.reshape(shape=(mask.shape[0], 1, int(mask.shape[1] ** .5), int(mask.shape[1] ** .5))),
                        size=(preds.shape[1], preds.shape[2]))
                    masks = torch.einsum('nchw->nhwc', masks.float())
                    combos = (w_samples + masks.repeat(1, 1, 1, 3)).clip(0, 1)
                    w_images = (torch.cat([w_samples, combos, preds], dim=2) * 255).detach().cpu()
                    print("w_images:", w_samples.shape, combos.shape, preds.shape, "-->", w_images.shape)

                    for i in range(w_images.shape[0]):
                        wi = w_images[i, :, :, :]
                        wandb_images += [wandb.Image(wi.numpy().astype(np.uint8),
                                                     caption=f"Prediction {i} at epoch {epoch}")]
                    wandb.log({f"reconstruction": wandb_images}, step=epoch_1000x, commit=False)

            if not math.isfinite(loss_value):
                print("Loss is {}, stopping training".format(loss_value))
                sys.exit(1)

            loss /= accum_iter
            loss_scaler(loss, optimizer, parameters=model.parameters(),
                        update_grad=(data_iter_step + 1) % accum_iter == 0)
            if (data_iter_step + 1) % accum_iter == 0:
                optimizer.zero_grad()

            torch.cuda.synchronize()

            metric_logger.update(loss=loss_value)

            lr = optimizer.param_groups[0]["lr"]
            metric_logger.update(lr=lr)

            loss_value_reduce = misc.all_reduce_mean(loss_value)
            if (data_iter_step + 1) % accum_iter == 0:
                if log_writer is not None:
                    """ We use epoch_1000x as the x-axis in tensorboard.
                    This calibrates different curves when batch size changes.
                    """
                    log_writer.add_scalar('train_loss', loss_value_reduce, epoch_1000x)
                    log_writer.add_scalar('lr', lr, epoch_1000x)
                if wandb_run is not None:
                    log = {"train/loss": loss_value_reduce, "train/lr": lr}
                    wandb.log(log, step=epoch_1000x, commit=True if data_iter_step == 0 else False)

        metric_logger.synchronize_between_processes()
        print("Averaged stats:", metric_logger)
        train_stats = {k: meter.global_avg for k, meter in metric_logger.meters.items()}

        # save train status and model
        if args.output_dir and (epoch % 100 == 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, suffix=f"pretraining_{epoch}")

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

        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))
    wandb.run.finish()


if __name__ == '__main__':
    args = get_args_parser()
    args = args.parse_args()

    # load data
    data_path = Path(args.data_path)
    anno_file = data_path / args.anno_file
    data_split_file = data_path / args.data_split_file
    im_dir = data_path / args.im_dir
    gt_dir = data_path / args.gt_dir
    with open(anno_file) as f:
        annotations = json.load(f)
    with open(data_split_file) as f:
        data_split = json.load(f)

    if args.output_dir:
        Path(args.output_dir).mkdir(parents=True, exist_ok=True)
    main(args)