train.py

# Copyright (c) 2021, NVIDIA CORPORATION & AFFILIATES.  All rights reserved.
#
# NVIDIA CORPORATION and its licensors retain all intellectual property
# and proprietary rights in and to this software, related documentation
# and any modifications thereto.  Any use, reproduction, disclosure or
# distribution of this software and related documentation without an express
# license agreement from NVIDIA CORPORATION is strictly prohibited.

"""Train a GAN using the techniques described in the paper
"Alias-Free Generative Adversarial Networks"."""

import os
import click
import re
import json
import tempfile
import torch

import dnnlib
from training import training_loop
from metrics import metric_main
from torch_utils import training_stats
from torch_utils import custom_ops
from torch_utils import gen_utils


# ----------------------------------------------------------------------------


def subprocess_fn(rank, c, temp_dir):
    dnnlib.util.Logger(file_name=os.path.join(c.run_dir, 'log.txt'), file_mode='a', should_flush=True)

    # Init torch.distributed.
    if c.num_gpus > 1:
        init_file = os.path.abspath(os.path.join(temp_dir, '.torch_distributed_init'))
        if os.name == 'nt':
            init_method = 'file:///' + init_file.replace('\\', '/')
            torch.distributed.init_process_group(backend='gloo', init_method=init_method, rank=rank, world_size=c.num_gpus)
        else:
            init_method = f'file://{init_file}'
            torch.distributed.init_process_group(backend='nccl', init_method=init_method, rank=rank, world_size=c.num_gpus)

    # Init torch_utils.
    sync_device = torch.device('cuda', rank) if c.num_gpus > 1 else None
    training_stats.init_multiprocessing(rank=rank, sync_device=sync_device)
    if rank != 0:
        custom_ops.verbosity = 'none'

    # Execute training loop.
    training_loop.training_loop(rank=rank, **c)


# ----------------------------------------------------------------------------


def launch_training(c, desc, outdir, dry_run):
    dnnlib.util.Logger(should_flush=True)

    # Pick output directory.
    c.run_dir = gen_utils.make_run_dir(outdir=outdir, desc=desc, dry_run=dry_run)

    # Print options.
    print()
    print('Training options:')
    print(json.dumps(c, indent=2))
    print()
    print(f'Output directory:    {c.run_dir}')
    print(f'Number of GPUs:      {c.num_gpus}')
    print(f'Batch size:          {c.batch_size} images')
    print(f'Training duration:   {c.total_kimg} kimg')
    print(f'Dataset path:        {c.training_set_kwargs.path}')
    print(f'Dataset size:        {c.training_set_kwargs.max_size} images')
    print(f'Dataset resolution:  {c.training_set_kwargs.resolution}')
    print(f'Dataset labels:      {c.training_set_kwargs.use_labels}')
    print(f'Dataset x-flips:     {c.training_set_kwargs.xflip}')
    print(f'Dataset y-flips:     {c.training_set_kwargs.yflip}')
    print()

    # Dry run?
    if dry_run:
        print('Dry run; exiting.')
        return

    with open(os.path.join(c.run_dir, 'training_options.json'), 'wt') as f:
        json.dump(c, f, indent=2)

    # Launch processes.
    print('Launching processes...')
    torch.multiprocessing.set_start_method('spawn')
    with tempfile.TemporaryDirectory() as temp_dir:
        if c.num_gpus == 1:
            subprocess_fn(rank=0, c=c, temp_dir=temp_dir)
        else:
            torch.multiprocessing.spawn(fn=subprocess_fn, args=(c, temp_dir), nprocs=c.num_gpus)


# ----------------------------------------------------------------------------


def init_dataset_kwargs(data):
    try:
        dataset_kwargs = dnnlib.EasyDict(class_name='training.dataset.ImageFolderDataset', path=data, use_labels=True, max_size=None, xflip=False, yflip=False)
        dataset_obj = dnnlib.util.construct_class_by_name(**dataset_kwargs) # Subclass of training.dataset.Dataset.
        dataset_kwargs.resolution = dataset_obj.resolution # Be explicit about resolution.
        dataset_kwargs.use_labels = dataset_obj.has_labels # Be explicit about labels.
        dataset_kwargs.max_size = len(dataset_obj) # Be explicit about dataset size.
        return dataset_kwargs, dataset_obj.name
    except IOError as err:
        raise click.ClickException(f'--data: {err}')


# ----------------------------------------------------------------------------


def parse_comma_separated_list(s):
    if isinstance(s, list):
        return s
    if s is None or s.lower() == 'none' or s == '':
        return []
    return s.split(',')


# ----------------------------------------------------------------------------


@click.command()
# Required.
@click.option('--cfg',          help='Base configuration',                                      type=click.Choice(['stylegan3-t', 'stylegan3-r', 'stylegan2']), required=True)
@click.option('--data',         help='Training data', metavar='[ZIP|DIR]',                      type=click.Path(exists=True, dir_okay=False), required=True)
@click.option('--gpus',         help='Number of GPUs to use', metavar='INT',                    type=click.IntRange(min=1), required=True)
@click.option('--batch',        help='Total batch size', metavar='INT',                         type=click.IntRange(min=1), required=True)
# Optional features.
@click.option('--cond',         help='Train conditional model', metavar='BOOL',                 type=bool, default=False, show_default=True)
@click.option('--cond-D-nofix', help='For old cond models trained w/o fix in D conditioning',   type=bool, default=False, show_default=True)
@click.option('--mirror',       help='Enable dataset x-flips', metavar='BOOL',                  type=bool, default=False, show_default=True)
@click.option('--mirror-y',     help='Enable dataset y-flips', metavar='BOOL',                  type=bool, default=False, show_default=True)
@click.option('--aug',          help='Augmentation mode',                                       type=click.Choice(['noaug', 'ada', 'fixed']), default='ada', show_default=True)
@click.option('--augpipe',      help='Augmentation pipeline',                                   type=click.Choice(['blit', 'geom', 'color', 'filter', 'noise', 'cutout', 'bg', 'bgc', 'bgcf', 'bgcfn', 'bgcfnc']), default='bgc', show_default=True)
@click.option('--resume',       help='Resume from given network pickle', metavar='[PATH|URL]',  type=str)
@click.option('--initstrength', help='Override ADA augment strength at the beginning',          type=click.FloatRange(min=0.0), )
@click.option('--freezeD',      help='Freeze first layers of D', metavar='INT',                 type=click.IntRange(min=0), default=0, show_default=True)
@click.option('--freezeM',      help='Freeze first layers of the Mapping Network Gm',           type=click.IntRange(min=0), default=0, show_default=True)
@click.option('--freezeE',      help='Freeze the embedding layer for conditional models',       type=bool, default=False, show_default=True)
# Misc hyperparameters.
@click.option('--gamma',        help='R1 regularization weight', metavar='FLOAT',               type=click.FloatRange(min=0), default=None)
@click.option('--p',            help='Probability for --aug=fixed', metavar='FLOAT',            type=click.FloatRange(min=0, max=1), default=0.2, show_default=True)
@click.option('--target',       help='Target value for --aug=ada', metavar='FLOAT',             type=click.FloatRange(min=0, max=1), default=0.6, show_default=True)
@click.option('--batch-gpu',    help='Limit batch size per GPU', metavar='INT',                 type=click.IntRange(min=1))
@click.option('--cbase',        help='Capacity multiplier', metavar='INT',                      type=click.IntRange(min=1), default=32768, show_default=True)
@click.option('--cmax',         help='Max. feature maps', metavar='INT',                        type=click.IntRange(min=1), default=512, show_default=True)
@click.option('--glr',          help='G learning rate  [default: varies]', metavar='FLOAT',     type=click.FloatRange(min=0))
@click.option('--dlr',          help='D learning rate', metavar='FLOAT',                        type=click.FloatRange(min=0), default=0.002, show_default=True)
@click.option('--map-depth',    help='Mapping network depth  [default: varies]', metavar='INT', type=click.IntRange(min=1))
@click.option('--mbstd-group',  help='Minibatch std group size', metavar='INT',                 type=click.IntRange(min=1), default=4, show_default=True)
# Misc settings.
@click.option('--outdir',       help='Where to save the results', metavar='DIR',                type=click.Path(file_okay=False), default=os.path.join(os.getcwd(), 'training-runs'))
@click.option('--desc',         help='String to include in result dir name', metavar='STR',     type=str)
@click.option('--metrics',      help='Quality metrics', metavar='[NAME|A,B,C|none]',            type=parse_comma_separated_list, default='none', show_default=True)
@click.option('--kimg',         help='Total training duration', metavar='KIMG',                 type=click.IntRange(min=1), default=25000, show_default=True)
@click.option('--resume-kimg',  help='Number of kimg images to resume from', metavar='RKIMG',   type=click.IntRange(min=0), default=0, show_default=True)
@click.option('--tick',         help='How often to print progress', metavar='KIMG',             type=click.IntRange(min=1), default=4, show_default=True)
@click.option('--snap',         help='How often to save model snapshots', metavar='TICKS',      type=click.IntRange(min=1), default=50, show_default=True)
@click.option('--img-snap',     help='How often to save image snapshots', metavar='INT',        type=click.IntRange(min=1), default=50, show_default=True)
@click.option('--snap-res',     help='Screen resolution to save image snapshot',                type=click.Choice(['1080p', '4k', '8k']), default='4k', show_default=True)
@click.option('--seed',         help='Random seed', metavar='INT',                              type=click.IntRange(min=0), default=0, show_default=True)
@click.option('--fp32',         help='Disable mixed-precision', metavar='BOOL',                 type=bool, default=False, show_default=True)
@click.option('--nobench',      help='Disable cuDNN benchmarking', metavar='BOOL',              type=bool, default=False, show_default=True)
@click.option('--workers',      help='DataLoader worker processes', metavar='INT',              type=click.IntRange(min=1), default=3, show_default=True)
@click.option('-n','--dry-run', help='Print training options and exit',                         is_flag=True)
def main(**kwargs):
    """Train a GAN using the techniques described in the paper
    "Alias-Free Generative Adversarial Networks".

    Examples:

    \b
    # Train StyleGAN3-T for AFHQv2 using 8 GPUs.
    python train.py --outdir=~/training-runs --cfg=stylegan3-t --data=~/datasets/afhqv2-512x512.zip \\
        --gpus=8 --batch=32 --gamma=8.2 --mirror=1

    \b
    # Fine-tune StyleGAN3-R for MetFaces-U using 1 GPU, starting from the pre-trained FFHQ-U pickle.
    python train.py --outdir=~/training-runs --cfg=stylegan3-r --data=~/datasets/metfacesu-1024x1024.zip \\
        --gpus=8 --batch=32 --gamma=6.6 --mirror=1 --kimg=5000 --snap=5 \\
        --resume=https://api.ngc.nvidia.com/v2/models/nvidia/research/stylegan3/versions/1/files/stylegan3-r-ffhqu-1024x1024.pkl

    \b
    # Train StyleGAN2 for FFHQ at 1024x1024 resolution using 8 GPUs.
    python train.py --outdir=~/training-runs --cfg=stylegan2 --data=~/datasets/ffhq-1024x1024.zip \\
        --gpus=8 --batch=32 --gamma=10 --mirror=1 --aug=noaug
    """

    # Initialize config.
    opts = dnnlib.EasyDict(kwargs) # Command line arguments.
    c = dnnlib.EasyDict() # Main config dict.
    c.G_kwargs = dnnlib.EasyDict(class_name=None, z_dim=512, w_dim=512, mapping_kwargs=dnnlib.EasyDict())
    c.D_kwargs = dnnlib.EasyDict(class_name='training.networks_stylegan2.Discriminator', block_kwargs=dnnlib.EasyDict(), mapping_kwargs=dnnlib.EasyDict(), epilogue_kwargs=dnnlib.EasyDict())
    c.G_opt_kwargs = dnnlib.EasyDict(class_name='torch.optim.Adam', betas=[0,0.99], eps=1e-8)
    c.D_opt_kwargs = dnnlib.EasyDict(class_name='torch.optim.Adam', betas=[0,0.99], eps=1e-8)  # TODO: Use ComplexSGD: https://arxiv.org/abs/2102.08431
    c.loss_kwargs = dnnlib.EasyDict(class_name='training.loss.StyleGAN2Loss')
    c.data_loader_kwargs = dnnlib.EasyDict(pin_memory=True, prefetch_factor=2)

    # Training set.
    c.training_set_kwargs, dataset_name = init_dataset_kwargs(data=opts.data)
    if opts.cond and not c.training_set_kwargs.use_labels:
        raise click.ClickException('--cond=True requires labels specified in dataset.json')
    c.training_set_kwargs.use_labels = opts.cond
    c.training_set_kwargs.xflip = opts.mirror
    c.training_set_kwargs.yflip = opts.mirror_y

    # Hyperparameters & settings.
    c.num_gpus = opts.gpus
    c.batch_size = opts.batch
    c.batch_gpu = opts.batch_gpu or opts.batch // opts.gpus
    c.G_kwargs.channel_base = c.D_kwargs.channel_base = opts.cbase
    c.G_kwargs.channel_max = c.D_kwargs.channel_max = opts.cmax
    c.G_kwargs.mapping_kwargs.num_layers = (8 if opts.cfg == 'stylegan2' else 2) if opts.map_depth is None else opts.map_depth
    if not opts.cond_d_nofix:
        c.D_kwargs.mapping_kwargs.num_layers = c.G_kwargs.mapping_kwargs.num_layers
    c.G_kwargs.mapping_kwargs.freeze_layers = opts.freezem
    c.G_kwargs.mapping_kwargs.freeze_embed = opts.freezee
    c.D_kwargs.block_kwargs.freeze_layers = opts.freezed
    c.D_kwargs.epilogue_kwargs.mbstd_group_size = opts.mbstd_group
    if opts.gamma is not None:
        c.loss_kwargs.r1_gamma = float(opts.gamma)
    else:
        # Use heuristic from StyleGAN2-ADA
        c.loss_kwargs.r1_gamma = 0.0002 * c.training_set_kwargs.resolution ** 2 / c.batch_size
        print(f'Using heuristic, R1 gamma set at: {c.loss_kwargs.r1_gamma:.4f}')
    c.G_opt_kwargs.lr = (0.002 if opts.cfg == 'stylegan2' else 0.0025) if opts.glr is None else opts.glr
    c.D_opt_kwargs.lr = opts.dlr
    c.metrics = opts.metrics
    c.total_kimg = opts.kimg
    c.resume_kimg = opts.resume_kimg
    c.kimg_per_tick = opts.tick
    c.network_snapshot_ticks = opts.snap
    c.image_snapshot_ticks = opts.img_snap
    c.snap_res = opts.snap_res
    c.random_seed = c.training_set_kwargs.random_seed = opts.seed
    c.data_loader_kwargs.num_workers = opts.workers

    # Sanity checks.
    if c.batch_size % c.num_gpus != 0:
        raise click.ClickException('--batch must be a multiple of --gpus')
    if c.batch_size % (c.num_gpus * c.batch_gpu) != 0:
        raise click.ClickException('--batch must be a multiple of --gpus times --batch-gpu')
    if c.batch_gpu < c.D_kwargs.epilogue_kwargs.mbstd_group_size:
        raise click.ClickException('--batch-gpu cannot be smaller than --mbstd')
    if any(not metric_main.is_valid_metric(metric) for metric in c.metrics):
        raise click.ClickException('\n'.join(['--metrics can only contain the following values:'] + metric_main.list_valid_metrics()))

    # Base configuration.
    c.ema_kimg = c.batch_size * 10 / 32
    if opts.cfg == 'stylegan2':
        c.G_kwargs.class_name = 'training.networks_stylegan2.Generator'
        c.loss_kwargs.style_mixing_prob = 0.9 # Enable style mixing regularization.
        c.loss_kwargs.pl_weight = 2 # Enable path length regularization.
        c.G_reg_interval = 4 # Enable lazy regularization for G.
        c.G_kwargs.fused_modconv_default = 'inference_only' # Speed up training by using regular convolutions instead of grouped convolutions.
        c.loss_kwargs.pl_no_weight_grad = True # Speed up path length regularization by skipping gradient computation wrt. conv2d weights.
    else:
        c.G_kwargs.class_name = 'training.networks_stylegan3.Generator'
        c.G_kwargs.magnitude_ema_beta = 0.5 ** (c.batch_size / (20 * 1e3))
        if opts.cfg == 'stylegan3-r':
            c.G_kwargs.conv_kernel = 1 # Use 1x1 convolutions.
            c.G_kwargs.channel_base *= 2 # Double the number of feature maps.
            c.G_kwargs.channel_max *= 2
            c.G_kwargs.use_radial_filters = True # Use radially symmetric downsampling filters.
            c.loss_kwargs.blur_init_sigma = 10 # Blur the images seen by the discriminator.
            c.loss_kwargs.blur_fade_kimg = c.batch_size * 200 / 32 # Fade out the blur during the first N kimg.

    # Augmentation.
    augpipe_specs = {
        'blit': dict(xflip=1, rotate90=1, xint=1),
        'geom': dict(scale=1, rotate=1, aniso=1, xfrac=1),
        'color': dict(brightness=1, contrast=1, lumaflip=1, hue=1, saturation=1),
        'filter': dict(imgfilter=1),
        'noise': dict(noise=1),
        'cutout': dict(cutout=1),
        'bg': dict(xflip=1, rotate90=1, xint=1, scale=1, rotate=1, aniso=1, xfrac=1),
        'bgc': dict(xflip=1, rotate90=1, xint=1, scale=1, rotate=1, aniso=1, xfrac=1, brightness=1, contrast=1, lumaflip=1, hue=1, saturation=1),
        'bgcf': dict(xflip=1, rotate90=1, xint=1, scale=1, rotate=1, aniso=1, xfrac=1, brightness=1, contrast=1, lumaflip=1, hue=1, saturation=1, imgfilter=1),
        'bgcfn': dict(xflip=1, rotate90=1, xint=1, scale=1, rotate=1, aniso=1, xfrac=1, brightness=1, contrast=1, lumaflip=1, hue=1, saturation=1, imgfilter=1, noise=1),
        'bgcfnc': dict(xflip=1, rotate90=1, xint=1, scale=1, rotate=1, aniso=1, xfrac=1, brightness=1, contrast=1, lumaflip=1, hue=1, saturation=1, imgfilter=1, noise=1, cutout=1)
    }

    if opts.aug != 'noaug':
        c.augment_kwargs = dnnlib.EasyDict(class_name='training.augment.AugmentPipe', **augpipe_specs[opts.augpipe])
        if opts.aug == 'ada':
            c.ada_target = opts.target
        if opts.aug == 'fixed':
            c.augment_p = opts.p

    # Initial augmentation strength for opts.aug == 'ada'; thanks to @dvschultz
    if opts.initstrength is not None:
         c.augment_p = opts.initstrength

    # Resume.

    if opts.resume is None:
        resume_desc = 'no_resume'
    else:
        if opts.resume in gen_utils.resume_specs[opts.cfg]:
            c.resume_pkl = gen_utils.resume_specs[opts.cfg][opts.resume]
            resume_desc = f'resume_{opts.resume}'
        else:  # A local file
            c.resume_pkl = opts.resume
            resume_desc = f'resume_custom'
        c.ada_kimg = 100 # Make ADA react faster at the beginning.
        c.ema_rampup = None # Disable EMA rampup.
        c.loss_kwargs.blur_init_sigma = 0 # Disable blur rampup.

    # Performance-related toggles.
    if opts.fp32:
        c.G_kwargs.num_fp16_res = c.D_kwargs.num_fp16_res = 0
        c.G_kwargs.conv_clamp = c.D_kwargs.conv_clamp = None
    if opts.nobench:
        c.cudnn_benchmark = False

    # Description string.
    desc = f'{opts.cfg:s}-{dataset_name:s}-gpus{c.num_gpus:d}-batch{c.batch_size:d}-gamma{c.loss_kwargs.r1_gamma:g}-{resume_desc}'
    if opts.desc is not None:
        desc += f'-{opts.desc}'

    # Launch.
    launch_training(c=c, desc=desc, outdir=opts.outdir, dry_run=opts.dry_run)


# ----------------------------------------------------------------------------


if __name__ == "__main__":
    main()  # pylint: disable=no-value-for-parameter


# ----------------------------------------------------------------------------