def main():
parser = argparse.ArgumentParser(
description='''StyleGAN2 generator.
Run 'python %(prog)s <subcommand> --help' for subcommand help.''',
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter
)
subparsers = parser.add_subparsers(help='Sub-commands', dest='command')
parser_generate_images = subparsers.add_parser(
'generate-images', help='Generate images')
parser_generate_images.add_argument(
'--network', help='Network pickle filename', dest='network_pkl', required=True)
parser_generate_images.add_argument(
'--label', type=_parse_label, help='Label vector', default=None, required=False)
parser_generate_images.add_argument(
'--seeds', type=_parse_num_range, help='List of random seeds', required=True)
parser_generate_images.add_argument(
'--truncation-psi', type=float, help='Truncation psi (default: %(default)s)', default=0.5)
parser_generate_images.add_argument(
'--result-dir', help='Root directory for run results (default: %(default)s)', default='results', metavar='DIR')
parser_style_mixing_example = subparsers.add_parser(
'style-mixing-example', help='Generate style mixing video')
parser_style_mixing_example.add_argument(
'--network', help='Network pickle filename', dest='network_pkl', required=True)
parser_style_mixing_example.add_argument(
'--row-seeds', type=_parse_num_range, help='Random seeds to use for image rows', required=True)
parser_style_mixing_example.add_argument(
'--col-seeds', type=_parse_num_range, help='Random seeds to use for image columns', required=True)
parser_style_mixing_example.add_argument(
'--col-styles', type=_parse_num_range, help='Style layer range (default: %(default)s)', default='0-6')
parser_style_mixing_example.add_argument(
'--truncation-psi', type=float, help='Truncation psi (default: %(default)s)', default=0.5)
parser_style_mixing_example.add_argument(
'--result-dir', help='Root directory for run results (default: %(default)s)', default='results', metavar='DIR')
args = parser.parse_args()
kwargs = vars(args)
subcmd = kwargs.pop('command')
if subcmd is None:
print('Error: missing subcommand. Re-run with --help for usage.')
sys.exit(1)
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
sc.run_desc = subcmd
func_name_map = {
'generate-images': 'run_generator.generate_images',
'style-mixing-example': 'run_generator.style_mixing_example'
}
dnnlib.submit_run(sc, func_name_map[subcmd], **kwargs)
def run(data_dir='./../datasets',
result_dir='./../results',
num_gpus=1,
total_kimg=1000,
mirror_augment=True):
train = EasyDict(
run_func_name='classifier_vgg.training_loop-vgg.training_loop')
classifier = EasyDict(
func_name='classifier_vgg.network_classifier-vgg.classifier_vgg')
classifier_opt = EasyDict(beta1=0.0, beta2=0.99, epsilon=1e-8)
classifier_loss = EasyDict(func_name='classifier_vgg.loss.cross_entropy')
sched = EasyDict()
sc = dnnlib.SubmitConfig()
tf_config = {'rnd.np_random_seed': 1000}
# train.resume_pkl = './results/00254-classifier-single_class_model/network-snapshot-001000.pkl'
# train.resume_kimg = 1000.0
sched.minibatch_size_base = 32
sched.minibatch_gpu_base = 8
current_label = 'background'
dataset = 'classifier_oversample_' + current_label
print(dataset)
train.data_dir = data_dir
if not os.path.exists(train.data_dir):
print('Error: dataset root directory does not exist.')
sys.exit(1)
train.total_kimg = total_kimg
train.mirror_augment = mirror_augment
metrics = [
EasyDict(name='acc_' + current_label,
func_name='metrics.accuracy_single_class.ACC',
test_dataset=dataset + '_test',
num_images=5000,
minibatch_per_gpu=8,
test_data_dir=data_dir)
]
desc = 'classifier-single_class_' + current_label
dataset_args = EasyDict(tfrecord_dir=dataset)
assert num_gpus in [1, 2, 4, 8]
sc.num_gpus = num_gpus
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
kwargs = EasyDict(train)
kwargs.update(classifier_args=classifier,
classifier_opt_args=classifier_opt,
classifier_loss_args=classifier_loss)
kwargs.update(dataset_args=dataset_args,
sched_args=sched,
metric_arg_list=metrics,
tf_config=tf_config)
kwargs.submit_config = copy.deepcopy(sc)
kwargs.submit_config.run_dir_root = result_dir
kwargs.submit_config.run_desc = desc
dnnlib.submit_run(**kwargs)
def run(dataset, data_dir, result_dir, config_id, num_gpus, total_kimg, gamma,
mirror_augment, metrics):
train = EasyDict(
run_func_name=
'training.training_loop.rotation.v7_baseline.training_loop')
G = EasyDict(func_name='training.networks.rotation.v7_baseline.G_main')
D = EasyDict(
func_name='training.networks.rotation.v7_baseline.D_stylegan2')
G_opt = EasyDict(beta1=0.0, beta2=0.99, epsilon=1e-8)
D_opt = EasyDict(beta1=0.0, beta2=0.99, epsilon=1e-8)
G_loss = EasyDict(
func_name='training.loss.rotation.v7_baseline.G_logistic_ns_pathreg')
D_loss = EasyDict(
func_name='training.loss.rotation.v7_baseline.D_logistic_r1')
sched = EasyDict()
grid = EasyDict(size='1080p', layout='random')
sc = dnnlib.SubmitConfig()
tf_config = {'rnd.np_random_seed': 1000}
train.data_dir = data_dir
train.total_kimg = total_kimg
train.mirror_augment = mirror_augment
train.image_snapshot_ticks = train.network_snapshot_ticks = 10
sched.G_lrate_base = sched.D_lrate_base = 0.002
sched.minibatch_size_base = 32
sched.minibatch_gpu_base = 4
D_loss.gamma = 10
metrics = [metric_defaults[x] for x in metrics]
desc = 'rotation-v7-baseline_256'
dataset_args = EasyDict(tfrecord_dir=dataset)
assert num_gpus in [1, 2, 4, 8]
sc.num_gpus = num_gpus
assert config_id in _valid_configs
if gamma is not None:
D_loss.gamma = gamma
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
kwargs = EasyDict(train)
kwargs.update(G_args=G,
D_args=D,
G_opt_args=G_opt,
D_opt_args=D_opt,
G_loss_args=G_loss,
D_loss_args=D_loss)
kwargs.update(dataset_args=dataset_args,
sched_args=sched,
grid_args=grid,
metric_arg_list=metrics,
tf_config=tf_config)
kwargs.submit_config = copy.deepcopy(sc)
kwargs.submit_config.run_dir_root = result_dir
kwargs.submit_config.run_desc = desc
dnnlib.submit_run(**kwargs)
def main():
parser = argparse.ArgumentParser(
description='''VC2 projector.
Run 'python %(prog)s <subcommand> --help' for subcommand help.''',
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter
)
subparsers = parser.add_subparsers(help='Sub-commands', dest='command')
project_generated_images_parser = subparsers.add_parser('project-generated-images', help='Project generated images')
project_generated_images_parser.add_argument('--network', help='Network pickle filename', dest='network_pkl', required=True)
project_generated_images_parser.add_argument('--seeds', type=_parse_num_range, help='List of random seeds', default=range(3))
project_generated_images_parser.add_argument('--num-snapshots', type=int, help='Number of snapshots (default: %(default)s)', default=5)
project_generated_images_parser.add_argument('--result-dir', help='Root directory for run results (default: %(default)s)', default='results', metavar='DIR')
project_generated_images_parser.add_argument('--create_new_G', help='If create a new G for projection.', default=False, type=_str_to_bool)
project_generated_images_parser.add_argument('--new_func_name', help='new G func name if create new G', default='training.vc_networks2.G_main_vc2')
project_real_dataset_images_parser = subparsers.add_parser('project-real-dataset-images', help='Project real images')
project_real_dataset_images_parser.add_argument('--network', help='Network pickle filename', dest='network_pkl', required=True)
project_real_dataset_images_parser.add_argument('--data-dir', help='Dataset root directory', required=True)
project_real_dataset_images_parser.add_argument('--dataset', help='Training dataset', dest='dataset_name', required=True)
project_real_dataset_images_parser.add_argument('--num-snapshots', type=int, help='Number of snapshots (default: %(default)s)', default=5)
project_real_dataset_images_parser.add_argument('--num-images', type=int, help='Number of images to project (default: %(default)s)', default=3)
project_real_dataset_images_parser.add_argument('--result-dir', help='Root directory for run results (default: %(default)s)', default='results', metavar='DIR')
project_real_dataset_images_parser.add_argument('--create_new_G', help='If create a new G for projection.', default=False, type=_str_to_bool)
project_real_dataset_images_parser.add_argument('--new_func_name', help='new G func name if create new G', default='training.vc_networks2.G_main_vc2')
project_real_other_images_parser = subparsers.add_parser('project-real-other-images', help='Project real images')
project_real_other_images_parser.add_argument('--network', help='Network pickle filename', dest='network_pkl', required=True)
project_real_other_images_parser.add_argument('--data-dir', help='Dir of images to project', required=True)
project_real_other_images_parser.add_argument('--num-snapshots', type=int, help='Number of snapshots (default: %(default)s)', default=5)
project_real_other_images_parser.add_argument('--result-dir', help='Root directory for run results (default: %(default)s)', default='results', metavar='DIR')
project_real_other_images_parser.add_argument('--create_new_G', help='If create a new G for projection.', default=False, type=_str_to_bool)
project_real_other_images_parser.add_argument('--new_func_name', help='new G func name if create new G', default='training.vc_networks2.G_main_vc2')
args = parser.parse_args()
subcmd = args.command
if subcmd is None:
print ('Error: missing subcommand. Re-run with --help for usage.')
sys.exit(1)
kwargs = vars(args)
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
sc.run_desc = kwargs.pop('command')
func_name_map = {
'project-generated-images': 'run_projector_vc2.project_generated_images',
'project-real-dataset-images': 'run_projector_vc2.project_real_dataset_images',
'project-real-other-images': 'run_projector_vc2.project_real_other_images'
}
dnnlib.submit_run(sc, func_name_map[subcmd], **kwargs)
def main():
parser = argparse.ArgumentParser(
description='Run StyleGAN2 metrics.',
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument(
'--result-dir',
help='Root directory for run results (default: %(default)s)',
default='results',
metavar='DIR')
parser.add_argument('--network',
help='Network pickle filename',
dest='network_pkl',
required=True)
parser.add_argument('--metrics',
help='Metrics to compute (default: %(default)s)',
default='fid50k',
type=lambda x: x.split(','))
parser.add_argument('--dataset', help='Training dataset')
parser.add_argument('--data-dir', help='Dataset root directory')
parser.add_argument('--mirror-augment',
help='Mirror augment (default: %(default)s)',
default=False,
type=_str_to_bool,
metavar='BOOL')
parser.add_argument('--include_I',
help='If include I for eval',
default=False,
type=_str_to_bool,
metavar='INCLUDE_I')
parser.add_argument('--num-gpus',
help='Number of GPUs to use',
type=int,
default=1,
metavar='N')
parser.add_argument(
'--mapping_nodup',
help='If the mapping layer in G has no duplication operation',
default=False,
type=_str_to_bool,
metavar='MAPPING_NODUP')
parser.add_argument('--avg_mv_for_I',
help='Using moving average for I?',
default=False,
type=_str_to_bool,
metavar='AVG_MV_FOR_I')
args = parser.parse_args()
kwargs = vars(args)
sc = dnnlib.SubmitConfig()
sc.num_gpus = kwargs.pop('num_gpus')
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
sc.run_desc = 'run-metrics'
dnnlib.submit_run(sc, 'run_metrics.run', **kwargs)
def create_training_config(
tfrecord_dir,
checkpoint_path,
run_dir,
G_beta1=0.0,
G_beta2=0.99,
D_beta1=0.0,
D_beta2=0.99,
mirror_augment=False,
style_mixing_probability=0.9,
generator_learning_rate=0.002,
discriminator_learning_rate=0.004,
weight_averaging_half_life=5,
**kwargs
):
train = EasyDict() # Options for training loop.
G = EasyDict(func_name='training.networks_stylegan2.G_main') # Options for generator network.
D = EasyDict(func_name='training.networks_stylegan2.D_stylegan2') # Options for discriminator network.
G_opt = EasyDict(beta1=G_beta1, beta2=G_beta2, epsilon=1e-8) # Options for generator optimizer.
D_opt = EasyDict(beta1=D_beta1, beta2=D_beta2, epsilon=1e-8) # Options for discriminator optimizer.
G_loss = EasyDict(func_name='training.loss.G_logistic_ns_pathreg') # Options for generator loss.
D_loss = EasyDict(func_name='training.loss.D_logistic_r1') # Options for discriminator loss.
sched = EasyDict() # Options for TrainingSchedule.
grid = EasyDict(size='8k', layout='random') # Options for setup_snapshot_image_grid().
submit_config = dnnlib.SubmitConfig() # Options for dnnlib.submit_run().
tf_config = {'rnd.np_random_seed': 1000} # Options for tflib.init_tf().
train.mirror_augment = mirror_augment
sched.minibatch_gpu_base = 4
# Uncomment to enable gradient accumulation
sched.minibatch_size_base = 12
train.G_smoothing_kimg = weight_averaging_half_life
D_loss.gamma = 10
metrics = [metric_defaults['fid5k']]
desc = 'stylegan2'
dataset_args = EasyDict(tfrecord_dir=tfrecord_dir)
desc += '-1gpu'
desc += '-config-f'
sched.G_lrate_dict = {128: generator_learning_rate, 256: generator_learning_rate, 512: generator_learning_rate, 1024: generator_learning_rate}
sched.D_lrate_dict = {128: discriminator_learning_rate, 256: discriminator_learning_rate, 512: discriminator_learning_rate, 1024: discriminator_learning_rate}
kwargs = EasyDict(train)
kwargs.update(G_args=G, D_args=D, G_opt_args=G_opt, D_opt_args=D_opt, G_loss_args=G_loss, D_loss_args=D_loss)
kwargs.update(dataset_args=dataset_args, sched_args=sched, grid_args=grid, metric_arg_list=metrics, tf_config=tf_config, resume_pkl=checkpoint_path)
kwargs.submit_config = deepcopy(submit_config)
kwargs.submit_config.run_dir = run_dir
kwargs.submit_config.run_desc = desc
return kwargs
def main():
parser = argparse.ArgumentParser(
description="Run StyleGAN2 metrics.",
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument(
"--result-dir",
help="Root directory for run results (default: %(default)s)",
default="results",
metavar="DIR")
parser.add_argument("--network",
help="Network pickle filename",
default=None,
dest="network_pkl")
parser.add_argument("--metrics",
help="Metrics to compute (default: %(default)s)",
default="fid50k",
type=lambda x: x.split(","))
parser.add_argument("--dataset", help="Training dataset", required=True)
parser.add_argument("--data-dir",
help="Dataset root directory",
required=True)
parser.add_argument("--mirror-augment",
help="Mirror augment (default: %(default)s)",
default=False,
type=_str_to_bool,
metavar="BOOL")
parser.add_argument("--gpus",
help="Number of GPUs to use",
type=str,
default=None)
parser.add_argument("--paths",
help="Image files to run evaluation on",
default=None,
type=str)
args = parser.parse_args()
if not os.path.exists(args.data_dir):
print("Error: dataset root directory does not exist.")
sys.exit(1)
kwargs = vars(args)
sc = dnnlib.SubmitConfig()
# set GPUs
gpus = kwargs.pop("gpus")
sc.num_gpus = len(gpus.split(","))
os.environ["CUDA_VISIBLE_DEVICES"] = gpus
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop("result_dir")
sc.run_desc = "run-metrics"
dnnlib.submit_run(sc, "run_metrics.run", **kwargs)
def run(opt):
submit_config = dnnlib.SubmitConfig()
# Which metrics to evaluate?
metrics = []
if 'FID' in opt.metrics:
metrics.append(metric_base.fid50k)
if 'PPL' in opt.metrics:
metrics.append(metric_base.ppl_zend_v2)
model_pth = find_model(opt.model)
if not model_pth:
# Find the snapshot:
if opt.model.endswith('.pkl'):
model_pth = opt.model
else:
model_pth = misc.find_pkl(
os.path.join(os.getcwd(), config.result_dir), int(opt.model),
opt.snapshot_kimg)
# Define dataset:
dataset = find_dataset(opt.dataset)
if not dataset:
dataset = os.path.join(os.getcwd(), config.data_dir, opt.dataset)
tasks = []
tasks += [
EasyDict(run_func_name='evaluate.run_pickle',
network_pkl=model_pth,
use_RA=True,
dataset_args=EasyDict(tfrecord_dir=dataset,
shuffle_mb=0,
resolution=opt.resolution),
mirror_augment=False)
]
submit_config.num_gpus = opt.num_gpus
# Execute.
submit_config.run_dir_root = dnnlib.submission.submit.get_template_from_path(
config.result_dir)
submit_config.run_dir_ignore += config.run_dir_ignore
for task in tasks:
for metric in metrics:
submit_config.run_desc = '%s-%s' % (task.run_func_name,
metric.name)
if task.run_func_name.endswith('run_snapshot'):
submit_config.run_desc += '-%s-%s' % (task.run_id,
task.snapshot)
if task.run_func_name.endswith('run_all_snapshots'):
submit_config.run_desc += '-%s' % task.run_id
submit_config.run_desc += '-%dgpu' % submit_config.num_gpus
dnnlib.submit_run(submit_config, metric_args=metric, **task)
print('Done with %s.' % model_pth)
def main():
parser = argparse.ArgumentParser(
description='''StyleGAN2 projector.
Run 'python %(prog)s <subcommand> --help' for subcommand help.''',
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter
)
subparsers = parser.add_subparsers(help='Sub-commands', dest='command')
project_generated_images_parser = subparsers.add_parser('project-generated-images', help='Project generated images')
project_generated_images_parser.add_argument('--network', help='Network pickle filename', dest='network_pkl', required=True)
project_generated_images_parser.add_argument('--seeds', type=_parse_num_range, help='List of random seeds', default=range(3))
project_generated_images_parser.add_argument('--num-snapshots', type=int, help='Number of snapshots (default: %(default)s)', default=5)
project_generated_images_parser.add_argument('--num-steps', type=int, help='Number of steps for running projection (default: %(default)s)', default=100)
project_generated_images_parser.add_argument('--truncation-psi', type=float, help='Truncation psi (default: %(default)s)', default=1.0)
project_generated_images_parser.add_argument('--result-dir', help='Root directory for run results (default: %(default)s)', default='results', metavar='DIR')
project_generated_images_parser.add_argument('--save-snapshots', action='store_true', help='Save projection results')
project_generated_images_parser.add_argument('--save-latents', action='store_true', help='If True, save latent vectors')
project_generated_images_parser.add_argument('--save-umap', action='store_true', help='If True, project latents using UMAP embeddings and save.')
project_real_images_parser = subparsers.add_parser('project-real-images', help='Project real images')
project_real_images_parser.add_argument('--network', help='Network pickle filename', dest='network_pkl', required=True)
project_real_images_parser.add_argument('--data-dir', help='Dataset root directory', required=True)
project_real_images_parser.add_argument('--dataset', help='Training dataset', dest='dataset_name', required=True)
project_real_images_parser.add_argument('--num-snapshots', type=int, help='Number of snapshots (default: %(default)s)', default=5)
project_real_images_parser.add_argument('--num-steps', type=int, help='Number of steps for running projection (default: %(default)s)', default=100)
project_real_images_parser.add_argument('--num-images', type=int, help='Number of images to project (default: %(default)s)', default=3)
project_real_images_parser.add_argument('--result-dir', help='Root directory for run results (default: %(default)s)', default='results', metavar='DIR')
project_real_images_parser.add_argument('--save-snapshots', action='store_true', help='If True, save projection results')
project_real_images_parser.add_argument('--save-latents', action='store_true', help='If True, save latent vectors')
project_real_images_parser.add_argument('--save-umap', action='store_true', help='If True, project latents using UMAP embeddings and save.')
project_real_images_parser.add_argument('--save-tiles', action='store_true', help='If True, stores images as 2048-by-2048 texture map.')
args = parser.parse_args()
subcmd = args.command
if subcmd is None:
print ('Error: missing subcommand. Re-run with --help for usage.')
sys.exit(1)
kwargs = vars(args)
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
sc.run_desc = kwargs.pop('command')
func_name_map = {
'project-generated-images': 'run_projector.project_generated_images',
'project-real-images': 'run_projector.project_real_images'
}
dnnlib.submit_run(sc, func_name_map[subcmd], **kwargs)
def main():
print()
print()
print()
print('GENERATOR STARTED')
parser = argparse.ArgumentParser(
description='''StyleGAN2 generator.
Run 'python %(prog)s <subcommand> --help' for subcommand help.''',
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter)
subparsers = parser.add_subparsers(help='Sub-commands', dest='command')
parser_generate_images = subparsers.add_parser('generate-images',
help='Generate images')
parser_generate_images.add_argument('--network',
help='Network pickle filename',
dest='network_pkl',
default='results/002332.pkl')
parser_generate_images.add_argument(
'--truncation-psi',
type=float,
help='Truncation psi (default: %(default)s)',
default=0.5)
parser_generate_images.add_argument(
'--result-dir',
help='Root directory for run results (default: %(default)s)',
default='results',
metavar='DIR')
args = parser.parse_args()
kwargs = vars(args)
subcmd = kwargs.pop('command')
if subcmd is None:
print('Error: missing subcommand. Re-run with --help for usage.')
sys.exit(1)
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
sc.run_desc = subcmd
func_name_map = {
'generate-images': 'TD_listen.generate_images',
'style-mixing-example': 'TD_listen.style_mixing_example'
}
dnnlib.submit_run(sc, func_name_map[subcmd], **kwargs)
def main():
submit_config = dnnlib.SubmitConfig()
# Which metrics to evaluate?
metrics = []
metrics += [metric_base.fid50k]
#metrics += [metric_base.ppl_zfull]
#metrics += [metric_base.ppl_wfull]
#metrics += [metric_base.ppl_zend]
#metrics += [metric_base.ppl_wend]
#metrics += [metric_base.ls]
metrics += [metric_base.lm_hd]
metrics += [metric_base.csim]
#metrics += [metric_base.dummy]
# Which networks to evaluate them on?
tasks = []
tasks += [
EasyDict(
run_func_name='run_metrics.run_pickle',
network_pkl=
'/content/gdrive/My Drive/Public/tensorboards_shared/run/t/58_rignet_fixed_pose_only_larger_adv_scaling/snapshots/network-snapshot-01560192.pkl',
dataset_args=EasyDict(
tfrecord_dir=
'/content/gdrive/My Drive/Public/tensorboards_shared/split_dataset/test',
shuffle_mb=0),
mirror_augment=True)
]
#tasks += [EasyDict(run_func_name='run_metrics.run_snapshot', run_id=100, snapshot=25000)]
#tasks += [EasyDict(run_func_name='run_metrics.run_all_snapshots', run_id=100)]
# How many GPUs to use?
submit_config.num_gpus = 1
#submit_config.num_gpus = 2
#submit_config.num_gpus = 4
#submit_config.num_gpus = 8
# Execute.
submit_config.run_dir_root = dnnlib.submission.submit.get_template_from_path(
config.result_dir)
submit_config.run_dir_ignore += config.run_dir_ignore
for task in tasks:
for metric in metrics:
submit_config.run_desc = '%s-%s' % (task.run_func_name,
metric.name)
if task.run_func_name.endswith('run_snapshot'):
submit_config.run_desc += '-%s-%s' % (task.run_id,
task.snapshot)
if task.run_func_name.endswith('run_all_snapshots'):
submit_config.run_desc += '-%s' % task.run_id
submit_config.run_desc += '-%dgpu' % submit_config.num_gpus
dnnlib.submit_run(submit_config, metric_args=metric, **task)
def run(data_dir='./../datasets',
result_dir='./../results',
num_gpus=2,
total_kimg=1000):
train = EasyDict(
run_func_name='classifier.training_loop-single_class.training_loop')
classifier = EasyDict(
func_name='classifier.network_classifier-new_label.classifier')
classifier_opt = EasyDict(beta1=0.0, beta2=0.99, epsilon=1e-8)
classifier_loss = EasyDict(func_name='classifier.loss.euclidean')
sched = EasyDict()
sc = dnnlib.SubmitConfig()
tf_config = {'rnd.np_random_seed': 1000}
dataset = 'classifier_oversample_rotation_v7'
sched.minibatch_size_base = 8
sched.minibatch_gpu_base = 4
train.data_dir = data_dir
train.total_kimg = total_kimg
desc = 'classifier_rotation_v7'
metrics = [
EasyDict(
name='ACCv7',
func_name='metrics.accuracy_v7.ACCv7',
num_images=5000,
minibatch_per_gpu=8,
test_dataset='./../datasets/classifier_oversample_rotation_v7_test'
)
]
dataset_args = EasyDict(tfrecord_dir=dataset)
assert num_gpus in [1, 2, 4, 8]
sc.num_gpus = num_gpus
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
kwargs = EasyDict(train)
kwargs.update(classifier_args=classifier,
classifier_opt_args=classifier_opt,
classifier_loss_args=classifier_loss)
kwargs.update(dataset_args=dataset_args,
sched_args=sched,
metric_arg_list=metrics,
tf_config=tf_config)
kwargs.submit_config = copy.deepcopy(sc)
kwargs.submit_config.run_dir_root = result_dir
kwargs.submit_config.run_desc = desc
dnnlib.submit_run(**kwargs)
def run_generate_pic(seeds, nums, psi=0.5, result_dir='results'):
dict = {}
dict['network_pkl'] = 'gdrive:networks/stylegan2-ffhq-config-f.pkl'
dict['seeds'] = range(seeds, seeds + nums)
dict['truncation_psi'] = psi
#dict['run_dir']=result_dir
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = result_dir
sc.run_desc = 'generate-images'
dnnlib.submit_run(sc, 'run_generator.generate_images', **dict)
def main():
parser = argparse.ArgumentParser(
description='Run CoModGAN metrics.',
formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument(
'--result-dir',
help='Root directory for run results (default: %(default)s)',
default='results',
metavar='DIR')
parser.add_argument('--network',
help='Network pickle filename',
dest='network_pkls',
required=True)
parser.add_argument('--metrics',
help='Metrics to compute (default: %(default)s)',
default='ids10k',
type=lambda x: x.split(','))
parser.add_argument('--dataset', help='Training dataset', required=True)
parser.add_argument('--data-dir',
help='Dataset root directory',
required=True)
parser.add_argument('--mirror-augment',
help='Mirror augment (default: %(default)s)',
default=False,
type=_str_to_bool,
metavar='BOOL')
parser.add_argument('--num-gpus',
help='Number of GPUs to use',
type=int,
default=1,
metavar='N')
parser.add_argument('--num-repeats', type=int, default=1)
parser.add_argument('--truncation', type=str, default=None)
parser.add_argument('--resume-with-new-nets',
default=False,
action='store_true')
args = parser.parse_args()
if not os.path.exists(args.data_dir):
print('Error: dataset root directory does not exist.')
sys.exit(1)
kwargs = vars(args)
sc = dnnlib.SubmitConfig()
sc.num_gpus = kwargs.pop('num_gpus')
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
sc.run_desc = 'run-metrics'
dnnlib.submit_run(sc, 'run_metrics.run', **kwargs)
def main():
parser = argparse.ArgumentParser(
description='VC-GAN and INFO-GAN image-pair generator.',
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument('--network_pkl',
help='Network pickle filename',
required=True)
parser.add_argument('--n_imgs',
type=int,
help='Number of image pairs to generate',
required=True)
parser.add_argument('--n_discrete',
type=int,
help='Number of discrete latents',
default=0)
parser.add_argument('--n_continuous',
type=int,
help='Number of continuous latents',
default=14)
parser.add_argument('--batch_size',
type=int,
help='Batch size for generation',
default=10)
parser.add_argument('--latent_type',
type=str,
help='What type of latent difference to use',
default='onedim',
choices=['onedim', 'fulldim'])
parser.add_argument('--model_type',
type=str,
help='Which model is this pkl',
default='vc_gan_with_vc_head',
choices=['info_gan', 'vc_gan', 'vc_gan_with_vc_head'])
parser.add_argument('--result-dir',
help='Root directory to store this dataset',
required=True,
metavar='DIR')
args = parser.parse_args()
kwargs = vars(args)
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs['result_dir']
dnnlib.submit_run(sc, 'run_pair_generator_vc.generate_image_pairs',
**kwargs)
def run_auto(dataset, data_dir, result_dir, config_id, num_gpus, resolution,
total_kimg, gamma, mirror_augment, metrics, train_auto):
train = EasyDict(run_func_name='training.training_loop.training_auto_loop'
) # Options for training loop.
Enc = EasyDict(func_name='training.networks_stylegan2.Encoder'
) # Options for encoder network.
Dec = EasyDict(func_name='training.networks_stylegan2.Decoder'
) # Options for decoder network.
opt = EasyDict(beta1=0.0, beta2=0.99,
epsilon=1e-8) # Options for autoencoder optimizer.
loss = EasyDict(
func_name='training.loss.auto_l1') # Options for autoencoder loss.
sched = EasyDict() # Options for TrainingSchedule.
grid = EasyDict(
size='1080p',
layout='random') # Options for setup_snapshot_image_grid().
sc = dnnlib.SubmitConfig() # Options for dnnlib.submit_run().
tf_config = {'rnd.np_random_seed': 1000} # Options for tflib.init_tf().
train.data_dir = data_dir
train.total_kimg = total_kimg
train.image_snapshot_ticks = 10
train.network_snapshot_ticks = 125
sched.lrate = 0.003
sched.minibatch_size = 64
sched.minibatch_gpu = 64
desc = 'stylegan2-hrae'
desc += '-' + dataset
dataset_args = EasyDict(tfrecord_dir=dataset)
dataset_args.resolution = resolution
dataset_args.num_threads = 4
assert num_gpus in [1, 2, 4, 8]
sc.num_gpus = num_gpus
desc += '-%dgpu' % num_gpus
desc += '-auto'
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
kwargs = EasyDict(train)
kwargs.update(Enc_args=Enc, Dec_args=Dec, opt_args=opt, loss_args=loss)
kwargs.update(dataset_args=dataset_args,
sched_args=sched,
grid_args=grid,
tf_config=tf_config)
kwargs.submit_config = copy.deepcopy(sc)
kwargs.submit_config.run_dir_root = result_dir
kwargs.submit_config.run_desc = desc
dnnlib.submit_run(**kwargs)
def generate_images(model_path, dest_path, seed_range):
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = './stylegan2/results'
sc.run_desc = 'generate_images'
dnnlib.submit_run(sc,
'run_generator.generate_images',
network_pkl=model_path,
seeds=range(seed_range),
truncation_psi=1.0,
dest=dest_path)
def main():
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = "D:\PythonProjectsDDrive\stylegan2-master"
sc.run_desc = 'generate-images'
network_pkl = 'D:\PythonProjectsDDrive\stylegan2-master\TrainedGANs\\network-snapshot-018708.pkl'
print('Loading networks from "%s"...' % network_pkl)
_G, _D, Gs = pretrained_networks.load_networks(network_pkl)
vector_size = Gs.input_shape[1:][0]
vec1 = np.load(
"D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\ToEditImages\image39_19039.npy",
mmap_mode=None,
allow_pickle=True,
fix_imports=True,
encoding='ASCII')
vec2 = np.load(
"D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\ToEditImages\image40_19040.npy",
mmap_mode=None,
allow_pickle=True,
fix_imports=True,
encoding='ASCII')
vec3 = np.load(
"D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\ToEditImages\image41_19041.npy",
mmap_mode=None,
allow_pickle=True,
fix_imports=True,
encoding='ASCII')
vec4 = np.load(
"D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\ToEditImages\image42_19042.npy",
mmap_mode=None,
allow_pickle=True,
fix_imports=True,
encoding='ASCII')
displayList = [vec1, vec2, vec3, vec4]
#LoadImageFromVector(displayList, Gs)
avrgVecBack = LoadVectorAverage(
"D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\BackImages"
)
avrgVecFront = LoadVectorAverage(
"D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\FrontImages"
)
PerformVectorOperationAddSub(
"D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\ToEditImages",
avrgVecFront, avrgVecBack, Gs)
def main():
print("run_metrics.py main()")
submit_config = dnnlib.SubmitConfig()
# Which metrics to evaluate?
metrics = []
metrics += [metric_base.fid50k]
#metrics += [metric_base.ppl_zfull]
#metrics += [metric_base.ppl_wfull]
#metrics += [metric_base.ppl_zend]
#metrics += [metric_base.ppl_wend]
#metrics += [metric_base.ls]
#metrics += [metric_base.dummy]
# Which networks to evaluate them on?
tasks = []
tasks += [
EasyDict(
run_func_name='run_metrics.run_pickle',
network_pkl=
'https://drive.google.com/uc?id=1MEGjdvVpUsu1jB4zrXZN7Y4kBBOzizDQ',
dataset_args=EasyDict(tfrecord_dir='ffhq', shuffle_mb=0),
mirror_augment=True)
] # karras2019stylegan-ffhq-1024x1024.pkl
#tasks += [EasyDict(run_func_name='run_metrics.run_snapshot', run_id=100, snapshot=25000)]
#tasks += [EasyDict(run_func_name='run_metrics.run_all_snapshots', run_id=100)]
# How many GPUs to use?
submit_config.num_gpus = 1
#submit_config.num_gpus = 2
#submit_config.num_gpus = 4
#submit_config.num_gpus = 8
# Execute.
submit_config.run_dir_root = dnnlib.submission.submit.get_template_from_path(
config.result_dir)
submit_config.run_dir_ignore += config.run_dir_ignore
for task in tasks:
for metric in metrics:
submit_config.run_desc = '%s-%s' % (task.run_func_name,
metric.name)
if task.run_func_name.endswith('run_snapshot'):
submit_config.run_desc += '-%s-%s' % (task.run_id,
task.snapshot)
if task.run_func_name.endswith('run_all_snapshots'):
submit_config.run_desc += '-%s' % task.run_id
submit_config.run_desc += '-%dgpu' % submit_config.num_gpus
dnnlib.submit_run(submit_config, metric_args=metric, **task)
def main():
parser = argparse.ArgumentParser(
description='''VC-GAN and INFO-GAN generator.
Run 'python %(prog)s --help' for subcommand help.''',
epilog=_examples,
formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument('--network_pkl',
help='Network pickle filename',
required=True)
parser.add_argument('--n_imgs',
type=int,
help='Number of images to generate',
required=True)
parser.add_argument('--n_discrete',
type=int,
help='Number of discrete latents',
default=0)
parser.add_argument('--n_continuous',
type=int,
help='Number of continuous latents',
default=14)
parser.add_argument('--n_samples_per',
type=int,
help='Number of samples per row',
default=10)
parser.add_argument('--model_type',
type=str,
help='Which model is this pkl',
default='vc_gan_with_vc_head',
choices=['info_gan', 'vc_gan', 'vc_gan_with_vc_head'])
parser.add_argument(
'--result-dir',
help='Root directory for run results (default: %(default)s)',
default='results',
metavar='DIR')
args = parser.parse_args()
kwargs = vars(args)
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
dnnlib.submit_run(sc, 'run_generator_vc.generate_images', **kwargs)
def main():
submit_config = dnnlib.SubmitConfig()
# Which metrics to evaluate?
metrics = []
metric_base.fid50k.update({
"inception_net_path":
os.path.join(config.result_dir, "inception_network",
"inception_v3_features.pkl")
})
metrics += [metric_base.fid50k]
# Which networks to evaluate them on?
tasks = []
tasks += [
EasyDict(
run_func_name="run_metrics.run_pickle",
network_pkl=
"/home/karnewar/msg-stylegan/00002-msg-stylegan-indian_celebs-4gpu/network-snapshot.pkl",
dataset_args=EasyDict(tfrecord_dir="indian_celebs/tfrecords",
shuffle_mb=0),
mirror_augment=True,
)
]
# tasks += [EasyDict(run_func_name='run_metrics.run_snapshot', run_id=100, snapshot=25000)]
# tasks += [EasyDict(run_func_name='run_metrics.run_all_snapshots', run_id=100)]
# How many GPUs to use?
submit_config.num_gpus = 1
# submit_config.num_gpus = 2
# submit_config.num_gpus = 4
# submit_config.num_gpus = 8
# Execute.
submit_config.run_dir_root = dnnlib.submission.submit.get_template_from_path(
config.result_dir)
submit_config.run_dir_ignore += config.run_dir_ignore
for task in tasks:
for metric in metrics:
submit_config.run_desc = "%s-%s" % (task.run_func_name,
metric.name)
if task.run_func_name.endswith("run_snapshot"):
submit_config.run_desc += "-%s-%s" % (task.run_id,
task.snapshot)
if task.run_func_name.endswith("run_all_snapshots"):
submit_config.run_desc += "-%s" % task.run_id
submit_config.run_desc += "-%dgpu" % submit_config.num_gpus
dnnlib.submit_run(submit_config, metric_args=metric, **task)
def main():
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = "D:\PythonProjectsDDrive\stylegan2-master"
sc.run_desc = 'generate-images'
network_pkl = 'D:\PythonProjectsDDrive\stylegan2-master\TrainedGANs\\network-snapshot-018708.pkl'
print('Loading networks from "%s"...' % network_pkl)
_G, _D, Gs = pretrained_networks.load_networks(network_pkl)
vector_size = Gs.input_shape[1:][0]
#
seedData = range(18000, 19000)
seeds = expand_seed(seedData, vector_size)
generate_images(Gs, seeds, seedData, "ArithmaticBaseImages", truncation_psi=0.5)
def loadingNetworks():
#loading network
latentCloatingPicePath,latentGenratedImagePath,clothingPicesPath,startingClothinPicePath,personsFolderPath,startingPersonPath,networkPath= loadPath("1024")
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = "/content/drive/My Drive/projects/stylegan2"
sc.run_desc = 'generate-images'
network_pkl = networkPath
print('Loading networks from "%s"...' % network_pkl)
_G, _D, Gs = pretrained_networks.load_networks(network_pkl)
return Gs,_D
def run(dataset, data_dir, result_dir, config_id, num_gpus, total_kimg, gamma, mirror_augment, metrics):
train = EasyDict(run_func_name='training.training_loop.conditional.v5_baseline.training_loop') # Options for training loop.
G = EasyDict(func_name='training.networks.conditional.baseline.G_main') # Options for generator network.
D = EasyDict(func_name='training.networks.conditional.baseline.D_stylegan2') # Options for discriminator network.
G_opt = EasyDict(beta1=0.0, beta2=0.99, epsilon=1e-8) # Options for generator optimizer.
D_opt = EasyDict(beta1=0.0, beta2=0.99, epsilon=1e-8) # Options for discriminator optimizer.
G_loss = EasyDict(func_name='training.loss.conditional.label_dropout.G_logistic_ns_pathreg') # Options for generator loss.
D_loss = EasyDict(func_name='training.loss.conditional.label_dropout.D_logistic_r1') # Options for discriminator loss.
sched = EasyDict() # Options for TrainingSchedule.
grid = EasyDict(size='1080p', layout='random') # Options for setup_snapshot_image_grid().
sc = dnnlib.SubmitConfig() # Options for dnnlib.submit_run().
tf_config = {'rnd.np_random_seed': 1000} # Options for tflib.init_tf().
# train.resume_pkl = './../results/00326-conditional_label_dropout_25/network-snapshot-000887.pkl'
# train.resume_kimg = 887.0
train.data_dir = data_dir
train.total_kimg = total_kimg
train.mirror_augment = mirror_augment
train.image_snapshot_ticks = train.network_snapshot_ticks = 10
sched.G_lrate_base = sched.D_lrate_base = 0.002
sched.minibatch_size_base = 32
sched.minibatch_gpu_base = 4
D_loss.gamma = 10
metrics = [metric_defaults[x] for x in metrics]
desc = 'conditional_label_dropout_25'
G_loss.label_dropout_prob = 0.5
D_loss.label_dropout_prob = 0.5
dataset_args = EasyDict(tfrecord_dir=dataset)
assert num_gpus in [1, 2, 4, 8]
sc.num_gpus = num_gpus
if gamma is not None:
D_loss.gamma = gamma
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
kwargs = EasyDict(train)
kwargs.update(G_args=G, D_args=D, G_opt_args=G_opt, D_opt_args=D_opt, G_loss_args=G_loss, D_loss_args=D_loss)
kwargs.update(dataset_args=dataset_args, sched_args=sched, grid_args=grid, metric_arg_list=metrics, tf_config=tf_config)
kwargs.submit_config = copy.deepcopy(sc)
kwargs.submit_config.run_dir_root = result_dir
kwargs.submit_config.run_desc = desc
dnnlib.submit_run(**kwargs)
def main():
submit_config = dnnlib.SubmitConfig()
# Which metrics to evaluate?
metrics = []
metrics += [metric_base.fid50k]
#metrics += [metric_base.ppl_zfull]
#metrics += [metric_base.ppl_wfull]
#metrics += [metric_base.ppl_zend]
#metrics += [metric_base.ppl_wend]
#metrics += [metric_base.ls]
#metrics += [metric_base.dummy]
# Which networks to evaluate them on?
tasks = []
tasks += [
EasyDict(
run_func_name='run_metrics.run_pickle',
network_pkl=
'./results/00010-sgan-logos-1gpu-cond/network-snapshot-020400.pkl',
dataset_args=EasyDict(tfrecord_dir='resnet_conditions',
shuffle_mb=0),
mirror_augment=True)
]
# How many GPUs to use?
submit_config.num_gpus = 1
#submit_config.num_gpus = 2
#submit_config.num_gpus = 4
#submit_config.num_gpus = 8
# Execute.
submit_config.run_dir_root = dnnlib.submission.submit.get_template_from_path(
config.result_dir)
submit_config.run_dir_ignore += config.run_dir_ignore
for task in tasks:
for metric in metrics:
submit_config.run_desc = '%s-%s' % (task.run_func_name,
metric.name)
if task.run_func_name.endswith('run_snapshot'):
submit_config.run_desc += '-%s-%s' % (task.run_id,
task.snapshot)
if task.run_func_name.endswith('run_all_snapshots'):
submit_config.run_desc += '-%s' % task.run_id
submit_config.run_desc += '-%dgpu' % submit_config.num_gpus
dnnlib.submit_run(submit_config, metric_args=metric, **task)
def main():
submit_config = dnnlib.SubmitConfig()
# Which metrics to evaluate?
metrics = []
metrics += [metric_base.fid50k]
#metrics += [metric_base.ppl_zfull]
#metrics += [metric_base.ppl_wfull]
#metrics += [metric_base.ppl_zend]
#metrics += [metric_base.ppl_wend]
#metrics += [metric_base.ls]
#metrics += [metric_base.dummy]
# Which networks to evaluate them on?
tasks = []
# tasks += [EasyDict(run_func_name='run_metrics.run_pickle', network_pkl='../model_results/stylegan/network-snapshot-008908.pkl', dataset_args=EasyDict(tfrecord_dir='ffhq_128x128_tf_stylegan', shuffle_mb=0), mirror_augment=True)] # karras2019stylegan-ffhq-1024x1024.pkl
tasks += [EasyDict(run_func_name='run_metrics.run_pickle', network_pkl='/home/hxz/GraduationProject/newBeholder-GAN/results/beholder-6001/network-final.pkl', dataset_args=EasyDict(tfrecord_dir='tf_files', shuffle_mb=0), mirror_augment=True)] # karras2019stylegan-ffhq-1024x1024.pkl
# tasks += [EasyDict(run_func_name='run_metrics.run_pickle', network_pkl='../model_results/stylegan/karras2019stylegan-ffhq-1024x1024.pkl', dataset_args=EasyDict(tfrecord_dir='ffhq', shuffle_mb=0), mirror_augment=True)] # karras2019stylegan-ffhq-1024x1024.pkl
#tasks += [EasyDict(run_func_name='run_metrics.run_pickle', network_pkl='../model_results/beholdergan-original/network-final.pkl', dataset_args=EasyDict(tfrecord_dir='ffhq_128x128_tf', shuffle_mb=0), mirror_augment=True)] # beholdergan
# tasks += [EasyDict(run_func_name='run_metrics.run_pickle', network_pkl='../model_results/beholdergan-id/first/network-final.pkl', dataset_args=EasyDict(tfrecord_dir='ffhq_128x128_id_tf', shuffle_mb=0), mirror_augment=True)] # beholdergan-id
# tasks += [EasyDict(run_func_name='run_metrics.run_pickle', network_pkl='../model_results/beholdergan-id/second/network-final.pkl', dataset_args=EasyDict(tfrecord_dir='ffhq_128x128_id_tf', shuffle_mb=0), mirror_augment=True)] # beholdergan-id
#tasks += [EasyDict(run_func_name='run_metrics.run_snapshot', run_id=100, snapshot=25000)]
#tasks += [EasyDict(run_func_name='run_metrics.run_all_snapshots', run_id=100)]
# How many GPUs to use?
submit_config.num_gpus = 1
#submit_config.num_gpus = 2
#submit_config.num_gpus = 4
#submit_config.num_gpus = 8
# Execute.
submit_config.run_dir_root = dnnlib.submission.submit.get_template_from_path(config.result_dir)
submit_config.run_dir_ignore += config.run_dir_ignore
for task in tasks:
for metric in metrics:
submit_config.run_desc = '%s-%s' % (task.run_func_name, metric.name)
if task.run_func_name.endswith('run_snapshot'):
submit_config.run_desc += '-%s-%s' % (task.run_id, task.snapshot)
if task.run_func_name.endswith('run_all_snapshots'):
submit_config.run_desc += '-%s' % task.run_id
submit_config.run_desc += '-%dgpu' % submit_config.num_gpus
dnnlib.submit_run(submit_config, metric_args=metric, **task)
def __init__(self, num_gpus, results_dir_root, network_pkl):
self.results_dir_root = results_dir_root
sc = dnnlib.SubmitConfig()
sc.num_gpus = num_gpus
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = results_dir_root
sc.run_desc = 'generate-images'
self.sc = sc
self.network_pkl = network_pkl
print('Loading networks from "%s"...' % self.network_pkl)
# _G = Instantaneous snapshot of the generator. Mainly useful for resuming a previous training run.
# _D = Instantaneous snapshot of the discriminator. Mainly useful for resuming a previous training run.
# Gs = Long-term average of the generator. Yields higher-quality results than the instantaneous snapshot.
dnnlib.tflib.init_tf()
self._G, self._D, self.Gs = pretrained_networks.load_networks(
self.network_pkl)
def morph():
image_1 = request.json['image1']
image_2 = request.json['image2']
fc = request.json['frame_count']
no_cache = 'no_cache' in request.json and request.json['no_cache'] is True
seeded = 'seeded' in request.json and request.json['seeded'] is True
if fc is None:
fc = 9
# floor the resolution
frames = int(sqrt(fc)) ** 2
# ensure that our morphs are always alphabetically sorted
image_1, image_2 = sorted([image_1, image_2])
output_path = "spritesheets/{}_{}_{}.jpg".format(image_1, image_2, frames)
# check if sheet exists to return early
bucket_blob = bucket.blob(output_path)
if not no_cache and bucket_blob.exists():
return bucket_blob.public_url
if seeded:
kwargs = kwards_from_seeds(image_1, image_2, frames)
else:
try:
kwargs = kwargs_from_projections(image_1, image_2, frames)
except IndexError:
return "projection for one or both images not found", 400
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = kwargs.pop('result_dir')
sc.run_desc = "serve-latent-walk"
result_path, _ = dnnlib.submit_run(
sc, 'run_generator.generate_latent_walk', **kwargs)
morph_pattern = "{}/*.png".format(result_path)
sheet = make_spritesheet(morph_pattern, output_path)
# delete result_path since we have our sheet
rmtree(result_path)
# upload to bucket instead of storing locally
bucket_blob.upload_from_filename(sheet)
remove(sheet)
# return the file
return bucket_blob.public_url
def main():
args = parse_args()
kwargs = vars(args)
subcmd = kwargs.pop('command')
if subcmd is None:
print('Error: missing subcommand. Re-run with --help for usage.')
sys.exit(1)
widget = get_widget(args.psi_steps)
canvas_elem = widget.FindElement('-CANVAS-')
canvas = canvas_elem.TKCanvas
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1])
ax.axis("off")
fig.tight_layout()
fig_agg = draw_figure(canvas, fig)
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_desc = subcmd
func_name_map = {
'latent-walk': 'latent_explorer.generate_images',
}
result = dnnlib.submit_run(sc, func_name_map[subcmd], **kwargs)
images = result.return_value
while True:
event, values = widget.Read(timeout=10)
if event in ('Exit', None):
exit(0)
ax.cla()
psi_idx = int(values['-PSI-'])
ax.imshow(images[psi_idx], aspect='auto')
fig_agg.draw()
def main():
sc = dnnlib.SubmitConfig()
sc.num_gpus = 1
sc.submit_target = dnnlib.SubmitTarget.LOCAL
sc.local.do_not_copy_source_files = True
sc.run_dir_root = "D:\PythonProjectsDDrive\stylegan2-master"
sc.run_desc = 'generate-images'
#network_pkl = 'D:\PythonProjectsDDrive\stylegan2-master\TrainedGANs\\network-snapshot-018708.pkl'
network_pkl = 'D:\PythonProjectsDDrive\stylegan2-master\TrainedGANs\\256-model-network-snapshot-018708.pkl'
example_img_pkl = "D:\PythonProjectsDDrive\MasterThesisClothesFittingwGANs\LatentSpaceImages\ToEditImagesDL\seed8007.npy"
print('Loading networks from "%s"...' % network_pkl)
_G, _D, Gs = pretrained_networks.load_networks(network_pkl)
vector_size = Gs.input_shape[1:][0]
#
#seedData = range(18000,19000)
#startSeed = 190000
#seeds = expand_seed(seedData, vector_size)
#generate_images(Gs, seeds, seedData, "ArithmaticBaseImages", truncation_psi=0.5)
generateDisplayVectors(example_img_pkl, 10, 1, Gs)
