train_stylegan2.py

from argparse import ArgumentParser
from pathlib import Path
import shutil

import imageio
def silence_imageio_warning(*args, **kwargs):
    pass
imageio.core.util._precision_warn = silence_imageio_warning

import gin
import numpy as np
import torch
import torch.nn as nn
from torch import autograd
import torch.optim as optim
from torch.utils.data import DataLoader

from evaluate.gan import FIDScore, FixedSampleGeneration, ImageGrid
from datasets import get_dataset
from augment import get_augment
from models.gan import get_architecture
from utils import cycle

from training.gan import setup
from utils import Logger
from utils import count_parameters
from utils import accumulate
from utils import set_grad

# import for gin binding
import penalty


def parse_args():
    """Training script for StyleGAN2."""

    parser = ArgumentParser(description='Training script: StyleGAN2 with DataParallel.')
    parser.add_argument('gin_config', type=str, help='Path to the gin configuration file')
    parser.add_argument('architecture', type=str, help='Architecture')

    parser.add_argument('--mode', default='std', type=str, help='Training mode (default: std)')
    parser.add_argument('--penalty', default='none', type=str, help='Penalty (default: none)')
    parser.add_argument('--aug', default='none', type=str, help='Augmentation (default: hfrt)')
    parser.add_argument('--use_warmup', action='store_true', help='Use warmup strategy on LR')
    parser.add_argument('--workers', default=8, type=int, metavar='N',
                        help='number of data loading workers (default: 0)')

    parser.add_argument('--temp', default=0.1, type=float,
                        help='Temperature hyperparameter for contrastive losses')
    parser.add_argument('--lbd_a', default=1.0, type=float,
                        help='Relative strength of the fake loss of ContraD')

    # Options for StyleGAN2 training
    parser.add_argument('--no_lazy', action='store_true',
                        help='Do not use lazy regularization')
    parser.add_argument("--d_reg_every", type=int, default=16,
                        help='Interval of applying R1 when lazy regularization is used')
    parser.add_argument("--lbd_r1", type=float, default=10, help='R1 regularization')
    parser.add_argument('--style_mix', default=0.9, type=float, help='Style mixing regularization')
    parser.add_argument('--halflife_k', default=20, type=int,
                        help='Half-life of exponential moving average in thousands of images')
    parser.add_argument('--ema_start_k', default=None, type=int,
                        help='When to start the exponential moving average of G (default: halflife_k)')
    parser.add_argument('--halflife_lr', default=0, type=int, help='Apply LR decay when > 0')

    # Options for logging specification
    parser.add_argument('--no_fid', action='store_true',
                        help='Do not track FIDs during training')
    parser.add_argument('--no_gif', action='store_true',
                        help='Do not save GIF of sample generations from a fixed latent periodically during training')
    parser.add_argument('--n_eval_avg', default=3, type=int,
                        help='How many times to average FID and IS')
    parser.add_argument('--print_every', help='', default=50, type=int)
    parser.add_argument('--evaluate_every', help='', default=2000, type=int)
    parser.add_argument('--save_every', help='', default=100000, type=int)
    parser.add_argument('--comment', help='Comment', default='', type=str)

    # Options for resuming / fine-tuning
    parser.add_argument('--resume', default=None, type=str,
                        help='Path to logdir to resume the training')
    parser.add_argument('--finetune', default=None, type=str,
                        help='Path to logdir that contains a pre-trained checkpoint of D')

    return parser.parse_args()


def _update_warmup(optimizer, cur_step, warmup, lr):
    if warmup > 0:
        ratio = min(1., (cur_step + 1) / (warmup + 1e-8))
        lr_w = ratio * lr
        for param_group in optimizer.param_groups:
            param_group['lr'] = lr_w


def _update_lr(optimizer, cur_step, batch_size, halflife_lr, lr, mult=1.0):
    if halflife_lr > 0 and (cur_step > 0) and (cur_step % 1000 == 0):
        ratio = (cur_step * batch_size) / halflife_lr
        lr_mul = 0.5 ** ratio
        lr_w = lr_mul * lr * mult
        for param_group in optimizer.param_groups:
            param_group['lr'] = lr_w
        return lr_w
    return None


def r1_loss(D, images, augment_fn):
    images_aug = augment_fn(images).detach()
    images_aug.requires_grad = True
    d_real = D(images_aug)
    grad_real, = autograd.grad(outputs=d_real.sum(), inputs=images_aug,
                               create_graph=True, retain_graph=True)
    grad_penalty = grad_real.pow(2).reshape(grad_real.shape[0], -1).sum(1).mean()
    return grad_penalty


def _sample_generator(G, num_samples, style_mix=0.9, enable_grad=True):
    latent_samples = G.sample_latent(num_samples)
    if enable_grad:
        generated_data = G(latent_samples, style_mix=style_mix)
    else:
        with torch.no_grad():
            generated_data = G(latent_samples, style_mix=style_mix)
    return generated_data


@gin.configurable("options")
def get_options_dict(dataset=gin.REQUIRED,
                     loss=gin.REQUIRED,
                     batch_size=32, fid_size=10000,
                     max_steps=800000, warmup=0, n_critic=1,
                     lr=0.002, lr_d=None, beta=(.0, .99),
                     lbd=10., lbd2=10.):
    if lr_d is None:
        lr_d = lr
    return {
        "dataset": dataset,
        "batch_size": batch_size,
        "fid_size": fid_size,
        "loss": loss,
        "max_steps": max_steps, "warmup": warmup,
        "n_critic": n_critic,
        "lr": lr, "lr_d": lr_d, "beta": beta,
        "lbd": lbd, "lbd2": lbd2
    }


def train(P, opt, train_fn, models, optimizers, train_loader, logger):
    generator, discriminator, g_ema = models
    opt_G, opt_D = optimizers

    losses = {'G_loss': [], 'D_loss': [], 'D_penalty': [],
              'D_real': [], 'D_gen': [], 'D_r1': []}
    metrics = {}

    metrics['image_grid'] = ImageGrid(volatile=P.no_gif)
    metrics['fixed_gen'] = FixedSampleGeneration(g_ema, volatile=P.no_gif)
    if not P.no_fid:
        metrics['fid_score'] = FIDScore(opt['dataset'], opt['fid_size'], P.n_eval_avg)

    logger.log_dirname("Steps {}".format(P.starting_step))

    for step in range(P.starting_step, opt['max_steps'] + 1):
        d_regularize = (step % P.d_reg_every == 0) and (P.lbd_r1 > 0)

        if P.use_warmup:
            _update_warmup(opt_G, step, opt["warmup"], opt["lr"])
            _update_warmup(opt_D, step, opt["warmup"], opt["lr_d"])
        if (not P.use_warmup) or step > opt["warmup"]:
            cur_lr_g = _update_lr(opt_G, step, opt["batch_size"], P.halflife_lr, opt["lr"])
            cur_lr_d = _update_lr(opt_D, step, opt["batch_size"], P.halflife_lr, opt["lr_d"])
            if cur_lr_d and cur_lr_g:
                logger.log('LR Updated: [G %.5f] [D %.5f]' % (cur_lr_g, cur_lr_d))

        do_ema = (step * opt['batch_size']) > (P.ema_start_k * 1000)
        accum = P.accum if do_ema else 0
        accumulate(g_ema, generator, accum)

        generator.train()
        discriminator.train()

        images, labels = next(train_loader)
        images = images.cuda()

        set_grad(generator, True)
        set_grad(discriminator, False)

        gen_images = _sample_generator(generator, images.size(0),
                                       style_mix=P.style_mix, enable_grad=True)
        g_loss = train_fn["G"](P, discriminator, opt, images, gen_images)

        opt_G.zero_grad()
        g_loss.backward()
        opt_G.step()
        losses['G_loss'].append(g_loss.item())

        set_grad(generator, False)
        set_grad(discriminator, True)

        if d_regularize:
            images.requires_grad = True
        d_loss, aux = train_fn["D"](P, discriminator, opt, images, gen_images)
        loss = d_loss + aux['penalty']

        if d_regularize:
            r1 = r1_loss(discriminator, images, P.augment_fn)
            lazy_r1 = (0.5 * P.lbd_r1) * r1 * P.d_reg_every
            loss = loss + lazy_r1
            losses['D_r1'].append(r1.item())

        opt_D.zero_grad()
        loss.backward()
        opt_D.step()
        losses['D_loss'].append(d_loss.item())
        losses['D_real'].append(aux['d_real'].item())
        losses['D_gen'].append(aux['d_gen'].item())
        losses['D_penalty'].append(aux['penalty'].item())

        for i in range(opt['n_critic'] - 1):
            images, labels = next(train_loader)
            images = images.cuda()
            gen_images = _sample_generator(generator, images.size(0),
                                           style_mix=P.style_mix, enable_grad=False)

            d_loss, aux = train_fn["D"](P, discriminator, opt, images, gen_images)
            loss = d_loss + aux['penalty']

            opt_D.zero_grad()
            loss.backward()
            opt_D.step()

        generator.eval()
        discriminator.eval()

        if step % P.print_every == 0:
            logger.log('[Steps %7d] [G %.3f] [D %.3f]' %
                       (step, losses['G_loss'][-1], losses['D_loss'][-1]))
            for name in losses:
                values = losses[name]
                if len(values) > 0:
                    logger.scalar_summary('gan/train/' + name, values[-1], step)

        if step % P.evaluate_every == 0:
            logger.log_dirname("Steps {}".format(step + 1))
            fid_score = metrics.get('fid_score')
            fixed_gen = metrics.get('fixed_gen')
            image_grid = metrics.get('image_grid')

            if fid_score:
                fid_avg = fid_score.update(step, g_ema)
                fid_score.save(logger.logdir + f'/results_fid_{P.eval_seed}.csv')
                logger.scalar_summary('gan/test/fid', fid_avg, step)
                logger.scalar_summary('gan/test/fid/best', fid_score.best, step)

            if not P.no_gif:
                _ = fixed_gen.update(step)
                imageio.mimsave(logger.logdir + f'/training_progress_{P.eval_seed}.gif',
                                fixed_gen.summary())
            aug_grid = image_grid.update(step, P.augment_fn(images))
            imageio.imsave(logger.logdir + f'/real_augment_{P.eval_seed}.jpg', aug_grid)

            G_state_dict = generator.module.state_dict()
            D_state_dict = discriminator.module.state_dict()
            Ge_state_dict = g_ema.state_dict()
            torch.save(G_state_dict, logger.logdir + '/gen.pt')
            torch.save(D_state_dict, logger.logdir + '/dis.pt')
            torch.save(Ge_state_dict, logger.logdir + '/gen_ema.pt')
            if fid_score and fid_score.is_best:
                torch.save(G_state_dict, logger.logdir + '/gen_best.pt')
                torch.save(D_state_dict, logger.logdir + '/dis_best.pt')
                torch.save(Ge_state_dict, logger.logdir + '/gen_ema_best.pt')
            if step % P.save_every == 0:
                torch.save(G_state_dict, logger.logdir + f'/gen_{step}.pt')
                torch.save(D_state_dict, logger.logdir + f'/dis_{step}.pt')
                torch.save(Ge_state_dict, logger.logdir + f'/gen_ema_{step}.pt')
            torch.save({
                'epoch': step,
                'optim_G': opt_G.state_dict(),
                'optim_D': opt_D.state_dict(),
            }, logger.logdir + '/optim.pt')


def worker(P):
    gin.parse_config_files_and_bindings(['configs/defaults/gan.gin',
                                         'configs/defaults/augment.gin',
                                         P.gin_config], [])
    options = get_options_dict()

    train_set, _, image_size = get_dataset(dataset=options['dataset'])
    train_loader = DataLoader(train_set, shuffle=True, pin_memory=True, num_workers=P.workers,
                              batch_size=options['batch_size'], drop_last=True)
    train_loader = cycle(train_loader)

    if P.no_lazy:
        P.d_reg_every = 1
    if P.ema_start_k is None:
        P.ema_start_k = P.halflife_k

    P.accum = 0.5 ** (options['batch_size'] / (P.halflife_k * 1000))

    generator, discriminator = get_architecture(P.architecture, image_size, P=P)
    g_ema, _ = get_architecture(P.architecture, image_size, P=P)
    if P.resume:
        print(f"=> Loading checkpoint from '{P.resume}'")
        state_G = torch.load(f"{P.resume}/gen.pt")
        state_D = torch.load(f"{P.resume}/dis.pt")
        state_Ge = torch.load(f"{P.resume}/gen_ema.pt")
        generator.load_state_dict(state_G)
        discriminator.load_state_dict(state_D)
        g_ema.load_state_dict(state_Ge)
    if P.finetune:
        print(f"=> Loading checkpoint for fine-tuning: '{P.finetune}'")
        state_D = torch.load(f"{P.finetune}/dis.pt")
        discriminator.load_state_dict(state_D, strict=False)
        discriminator.reset_parameters(discriminator.linear)
        P.comment += 'ft'

    generator = generator.cuda()
    discriminator = discriminator.cuda()
    g_ema = g_ema.cuda()
    g_ema.eval()

    G_optimizer = optim.Adam(generator.parameters(),
                             lr=options["lr"], betas=options["beta"])
    D_optimizer = optim.Adam(discriminator.parameters(),
                             lr=options["lr_d"], betas=options["beta"])

    if P.resume:
        logger = Logger(None, resume=P.resume)
    else:
        _desc = f"R{P.lbd_r1}_mix{P.style_mix}_H{P.halflife_k}"
        if P.halflife_lr > 0:
            _desc += f"_lr{P.halflife_lr / 1000000:.1f}M"
        _desc += f"_NoLazy" if P.no_lazy else "_Lazy"

        logger = Logger(f'{P.filename}_{_desc}{P.comment}', subdir=f'gan_dp/st_{P.gin_stem}/{P.architecture}')
        shutil.copy2(P.gin_config, f"{logger.logdir}/config.gin")
    P.logdir = logger.logdir
    P.eval_seed = np.random.randint(10000)

    if P.resume:
        opt = torch.load(f"{P.resume}/optim.pt")
        G_optimizer.load_state_dict(opt['optim_G'])
        D_optimizer.load_state_dict(opt['optim_D'])
        logger.log(f"Checkpoint loaded from '{P.resume}'")
        P.starting_step = opt['epoch'] + 1
    else:
        logger.log(generator)
        logger.log(discriminator)
        logger.log(f"# Params - G: {count_parameters(generator)}, D: {count_parameters(discriminator)}")
        logger.log(options)
        P.starting_step = 1
    logger.log(f"Use G moving average: {P.accum}")

    if P.finetune:
        logger.log(f"Checkpoint loaded from '{P.finetune}'")

    P.augment_fn = get_augment(mode=P.aug).cuda()
    generator = nn.DataParallel(generator)
    generator.sample_latent = generator.module.sample_latent
    discriminator = nn.DataParallel(discriminator)

    train(P, options, P.train_fn,
          models=(generator, discriminator, g_ema),
          optimizers=(G_optimizer, D_optimizer),
          train_loader=train_loader, logger=logger)


if __name__ == '__main__':
    P = parse_args()
    if P.comment:
        P.comment = '_' + P.comment
    P.gin_stem = Path(P.gin_config).stem
    P = setup(P)
    P.distributed = False
    worker(P)