train.py

import argparse
import logging
import math
import os
import random

import numpy as np
import torch
import torch.distributed as dist
import torch.multiprocessing as mp
from torchvision.utils import make_grid

import glf.options.options as option
from glf.data import create_dataloader, create_dataset
from glf.data.data_sampler import DistIterSampler
from glf.metrics import InceptionPredictor, frechet_distance, compute_prd_from_embedding, prd_to_max_f_beta_pair
from glf.models import create_model
from glf.utils import util


def init_dist(backend='nccl', **kwargs):
    """initialization for distributed training"""
    if mp.get_start_method(allow_none=True) != 'spawn':
        mp.set_start_method('spawn')
    rank = int(os.environ['RANK'])
    num_gpus = torch.cuda.device_count()
    torch.cuda.set_device(rank % num_gpus)
    dist.init_process_group(backend=backend, **kwargs)


def main():
    #### options
    parser = argparse.ArgumentParser()
    parser.add_argument('-opt', type=str, help='Path to option YAML file.')
    parser.add_argument('--launcher', choices=['none', 'pytorch'], default='none',
                        help='job launcher')
    parser.add_argument('--local_rank', type=int, default=0)
    args = parser.parse_args()
    opt = option.parse(args.opt, is_train=True)

    #### distributed training settings
    if args.launcher == 'none':  # disabled distributed training
        opt['dist'] = False
        rank = -1
        print('Disabled distributed training.')
    else:
        opt['dist'] = True
        init_dist()
        world_size = torch.distributed.get_world_size()
        rank = torch.distributed.get_rank()

    #### loading resume state if exists
    if opt['path'].get('resume_state', None):
        # distributed resuming: all load into default GPU
        device_id = torch.cuda.current_device()
        resume_state = torch.load(opt['path']['resume_state'],
                                  map_location=lambda storage, loc: storage.cuda(device_id))
        option.check_resume(opt, resume_state['iter'])  # check resume options
    else:
        resume_state = None

    #### mkdir and loggers
    if rank <= 0:  # normal training (rank -1) OR distributed training (rank 0)
        if resume_state is None:
            util.mkdir_and_rename(
                opt['path']['experiments_root'])  # rename experiment folder if exists
            util.mkdirs((path for key, path in opt['path'].items() if not key == 'experiments_root'
                         and 'pretrained_' not in key and 'resume' not in key))

        # config loggers. Before it, the log will not work
        util.setup_logger('base', opt['path']['log'], 'train_' + opt['name'], level=logging.INFO,
                          screen=True, tofile=True)
        logger = logging.getLogger('base')
        logger.info(option.dict2str(opt))
        # tensorboard logger
        if opt['use_tb_logger'] and 'debug' not in opt['name']:
            from tensorboardX import SummaryWriter
            tb_logger = SummaryWriter(log_dir='tb_logger/' + opt['name'])
    else:
        util.setup_logger('base', opt['path']['log'], 'train', level=logging.INFO, screen=True)
        logger = logging.getLogger('base')

    # convert to NoneDict, which returns None for missing keys
    opt = option.dict_to_nonedict(opt)

    #### random seed
    seed = opt['train']['manual_seed']
    if seed is None:
        seed = random.randint(1, 10000)
    if rank <= 0:
        logger.info('Random seed: {}'.format(seed))
    util.set_random_seed(seed)

    torch.backends.cudnn.benchmark = True
    # torch.backends.cudnn.deterministic = True

    #### create train and val dataloader
    dataset_ratio = 200  # enlarge the size of each epoch
    for phase, dataset_opt in opt['datasets'].items():
        if phase == 'train':
            train_set = create_dataset(dataset_opt, is_train=True)
            train_size = int(len(train_set) / dataset_opt['batch_size'])
            total_iters = int(opt['train']['niter'])
            total_epochs = int(math.ceil(total_iters / train_size))
            if opt['dist']:
                train_sampler = DistIterSampler(train_set, world_size, rank, dataset_ratio)
                total_epochs = int(math.ceil(total_iters / (train_size * dataset_ratio)))
            else:
                train_sampler = None
            train_loader = create_dataloader(train_set, dataset_opt, opt, train_sampler)
            if rank <= 0:
                logger.info('Number of train images: {:,d}, iters: {:,d}'.format(
                    len(train_set), train_size))
                logger.info('Total epochs needed: {:d} for iters {:,d}'.format(
                    total_epochs, total_iters))
        elif phase == 'val':
            val_set = create_dataset(dataset_opt, is_train=False)
            val_loader = create_dataloader(val_set, dataset_opt, opt)
            if rank <= 0:
                logger.info('Number of val images in [{:s}]: {:d}'.format(
                    dataset_opt['name'], len(val_set)))
        else:
            raise NotImplementedError('Phase [{:s}] is not recognized.'.format(phase))
    assert train_loader is not None

    #### create model
    model = create_model(opt)
    if opt['datasets'].get('val', None) and opt['train']['val_calculate_fid_prd'] and \
            opt['datasets']['val']['name'] in ['CIFAR-10', 'CelebA', 'dots']:
        predictor_device = torch.device('cuda' if opt['gpu_ids'] is not None else 'cpu')
        predictor_dim = 2048
        predictor = InceptionPredictor(output_dim=predictor_dim).to(predictor_device)
    else:
        predictor = None

    #### resume training
    if resume_state:
        logger.info('Resuming training from epoch: {}, iter: {}.'.format(
            resume_state['epoch'], resume_state['iter']))

        start_epoch = resume_state['epoch']
        current_step = resume_state['iter']
        model.resume_training(resume_state)  # handle optimizers and schedulers
    else:
        current_step = 0
        start_epoch = 0

    #### training
    logger.info('Start training from epoch: {:d}, iter: {:d}'.format(start_epoch, current_step))
    for epoch in range(start_epoch, total_epochs + 1):

        if opt['dist']:
            train_sampler.set_epoch(epoch)
        for _, train_data in enumerate(train_loader):
            current_step += 1
            if current_step > total_iters:
                break
            #### update learning rate
            model.update_learning_rate(current_step, warmup_iter=opt['train']['warmup_iter'])

            #### training
            model.feed_data(train_data)
            model.optimize_parameters(current_step)

            #### log
            if current_step % opt['logger']['print_freq'] == 0:
                logs = model.get_current_log()
                message = '[epoch:{:3d}, iter:{:8,d}, lr:('.format(epoch, current_step)
                for v in model.get_current_learning_rate():
                    message += '{:.3e},'.format(v)
                message += ')] '
                for k, v in logs.items():
                    message += '{:s}: {:.4e} '.format(k, v)
                    # tensorboard logger
                    if opt['use_tb_logger'] and 'debug' not in opt['name']:
                        if rank <= 0:
                            tb_logger.add_scalar(k, v, current_step)
                if rank <= 0:
                    logger.info(message)

            #### validation
            if opt['datasets'].get('val', None) and (
                    current_step % opt['train']['val_freq'] == 0 or current_step == total_iters):
                if rank <= 0:
                    # save 60 samples generated from noise
                    samples = model.sample_images(60)
                    grid = make_grid(samples, nrow=6)
                    grid = util.tensor2img(grid)
                    sample_name = 'latest.png' if current_step == total_iters else '{:d}.png'.format(current_step)
                    util.save_img(grid, os.path.join(opt['path']['samples'], sample_name))
                    del samples, grid

                    # calculate FID
                    if predictor is not None:
                        art_samples = []
                        true_samples = []
                        logger.info('Calculating FID and PRD:')

                        if opt['train']['val_num_batches'] is None or current_step == total_iters:
                            num_val_batches = len(val_loader)
                        else:
                            num_val_batches = int(opt['train']['val_num_batches'])
                        pbar = util.ProgressBar(num_val_batches)
                        for k, (val_data, _) in enumerate(val_loader):
                            if k >= num_val_batches:
                                break

                            samples = model.sample_images(val_data.size(0)).to(predictor_device)
                            val_data = val_data.to(predictor_device)

                            art_samples.append(predictor(samples).detach().cpu().numpy())
                            true_samples.append(predictor(val_data).detach().cpu().numpy())

                            pbar.update('batch #{}'.format(k))

                            del samples, val_data

                        art_samples = np.concatenate(art_samples, axis=0)
                        true_samples = np.concatenate(true_samples, axis=0)

                        art_samples = art_samples.reshape(-1, predictor_dim)
                        true_samples = true_samples.reshape(-1, predictor_dim)
                        FID = frechet_distance(true_samples, art_samples)
                        precision, recall = compute_prd_from_embedding(true_samples, art_samples)
                        f_8, f_1_8 = prd_to_max_f_beta_pair(precision, recall, beta=8)

                        # log
                        if current_step == total_iters:
                            logger.info('# Final Validation # FID: {:.4e}'.format(FID))
                            logger.info('# Final Validation # F8: {:.4e} F1/8: {:.4e}'.format(f_8, f_1_8))
                        else:
                            logger.info('# Validation # FID: {:.4e}'.format(FID))
                            logger.info('# Validation # F8: {:.4e} F1/8: {:.4e}'.format(f_8, f_1_8))

                        # tensorboard logger
                        if opt['use_tb_logger'] and 'debug' not in opt['name']:
                            tb_logger.add_scalar('metrics/fid', FID, current_step)
                            tb_logger.add_scalar('metrics/F8', f_8, current_step)
                            tb_logger.add_scalar('metrics/F1_8', f_8, f_1_8)

                        del art_samples, true_samples, FID

            #### save models and training states
            if opt['logger']['save_checkpoint_freq'] and current_step % opt['logger']['save_checkpoint_freq'] == 0:
                if rank <= 0:
                    logger.info('Saving models and training states.')
                    model.save(current_step)
                    model.save_training_state(epoch, current_step)

    if rank <= 0:
        logger.info('Saving the final model.')
        model.save('latest')
        logger.info('End of training.')
        tb_logger.close()


if __name__ == '__main__':
    main()