log = test_metrics.result()
print(log)
# summary(model, (1,496, 384))
time_results = compute_precise_time(model, [496, 384], 96, loss_fn, device)
print(time_results)
reset_bn_stats(model)
return
if __name__ == '__main__':
args = argparse.ArgumentParser(description='PyTorch Template')
args.add_argument('-c',
'--config',
default=None,
type=str,
help='config file path (default: None)')
args.add_argument('-r',
'--resume',
default=None,
type=str,
help='path to latest checkpoint (default: None)')
args.add_argument('-d',
'--device',
default=None,
type=str,
help='indices of GPUs to enable (default: all)')
config = ConfigParser.from_args(args, mode='test')
main(config)
CustomArgs(['--gamma'], type=float, target='lr_scheduler;args;gamma'),
CustomArgs(['--save_period'], type=int, target='trainer;save_period'),
CustomArgs(['--reduce_dimension'],
type=int,
target='arch;args;reduce_dimension'),
CustomArgs(['--layer2_dimension'],
type=int,
target='arch;args;layer2_output_dim'),
CustomArgs(['--layer3_dimension'],
type=int,
target='arch;args;layer3_output_dim'),
CustomArgs(['--layer4_dimension'],
type=int,
target='arch;args;layer4_output_dim'),
CustomArgs(['--num_experts'], type=int,
target='arch;args;num_experts'),
CustomArgs(['--distribution_aware_diversity_factor'],
type=float,
target='loss;args;additional_diversity_factor'),
CustomArgs(['--pos_weight'], type=float,
target='arch;args;pos_weight'),
CustomArgs(['--collaborative_loss'],
type=int,
target='loss;args;collaborative_loss'),
CustomArgs(['--distill_checkpoint'],
type=str,
target='distill_checkpoint')
]
config = ConfigParser.from_args(args, options)
main(config)
metavar='N', help='mini-batch size (default: 256)')
parser.add_argument('--lr', '--learning-rate', default=0.001, type=float,
metavar='LR', help='initial learning rate')
parser.add_argument('--momentum', default=0.9, type=float, metavar='M',
help='momentum')
parser.add_argument('--weight-decay', '--wd', default=5e-3, type=float,
metavar='W', help='weight decay (default: 5e-3)')
parser.add_argument('--categorical', default=True, action="store_true")
parser.add_argument('--continuous', default=False, action="store_true")
# ========================= Monitor Configs ==========================
parser.add_argument('--print-freq', '-p', default=20, type=int,
metavar='N', help='print frequency (default: 10)')
parser.add_argument('--eval-freq', '-ef', default=5, type=int,
metavar='N', help='evaluation frequency (default: 5)')
# ========================= Runtime Configs ==========================
parser.add_argument('-j', '--workers', default=4, type=int, metavar='N',
help='number of data loading workers (default: 4)')
CustomArgs = collections.namedtuple('CustomArgs', 'flags type target')
options = [
CustomArgs(['--exp_name'], type=str, target='name'),
]
config = ConfigParser.from_args(parser, options)
print(config)
args = parser.parse_args()
main(args, config)
optimizer,
config=config,
data_loader=data_loader,
valid_data_loader=valid_data_loader,
lr_scheduler=lr_scheduler)
trainer.train()
if __name__ == '__main__':
args = argparse.ArgumentParser()
args.add_argument('-c',
'--config',
default=None,
type=str,
required=True,
help='path to config file (default: None)')
args.add_argument('-r',
'--resume',
default=None,
type=str,
help='path to latest checkpoint (default: None)')
args.add_argument('-d',
'--device',
default=None,
type=str,
help='indices of GPUs to enable (default: all)')
args.add_argument('-s', '--seed', default=1234, type=str)
config = ConfigParser.from_args(args)
main(config)
from data_loader.data_loaders import *
from train_test import *
from parse_config import ConfigParser
def main(config):
data = load_data(config)
train_test(data, config)
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='AnchorKG')
parser.add_argument('-c',
'--config',
default="./config.json",
type=str,
help='config file path (default: None)')
parser.add_argument('-r',
'--resume',
default=None,
type=str,
help='path to latest checkpoint (default: None)')
parser.add_argument('-d',
'--device',
default=None,
type=str,
help='indices of GPUs to enable (default: all)')
config = ConfigParser.from_args(parser)
main(config)
yticklabels=tgt_seq)
fig.xaxis.set_label_position('top')
fig.figure.show()
if __name__ == '__main__':
args = argparse.ArgumentParser()
args.add_argument('-c',
'--config',
default=None,
type=str,
help='config file path (default: None)')
args.add_argument('--model-path',
type=str,
required=True,
help='path to model.pth to test (default: None')
args.add_argument('-r',
'--resume',
default=None,
type=str,
help='path to latest checkpoint (default: None)')
args.add_argument('-d',
'--device',
default=None,
type=str,
help='indices of GPUs to enable (default: all)')
# custom cli options to modify configuration from default values given in json file.
CustomArgs = collections.namedtuple('CustomArgs', 'flags type target help')
config, args = ConfigParser.from_args(args)
main(config, args.model_path)
config.log_dir))
logger.info(log)
if __name__ == '__main__':
args = argparse.ArgumentParser(description='PyTorch Template')
args.add_argument('-c',
'--config',
default=None,
type=str,
help='config file path (default: None)')
args.add_argument('-r',
'--resume',
default=None,
type=str,
help='path to latest checkpoint (default: None)')
args.add_argument('-d',
'--device',
default=None,
type=str,
help='indices of GPUs to enable (default: all)')
CustomArgs = collections.namedtuple('CustomArgs', 'flags type target help')
options = [
CustomArgs(['-x', '--extract'],
type=str,
target=('extract'),
help='extract parameters of the model (default: False)')
]
config = ConfigParser.from_args(args, options=options, test=True)
main(config)
batch_size = data.shape[0]
total_loss += (loss_s.item() - loss_b.item()) * batch_size
n_samples = len(data_loader.sampler)
log = {'loss': total_loss / n_samples}
logger.info(log)
if __name__ == '__main__':
args = argparse.ArgumentParser(description='Test code for evaluation.')
# args.add_argument('-c', '--config', default=None, type=str,
# help='config file path (default: None)')
args.add_argument('-c',
'--config',
default='./configs/fmnist_glow_config_test.json',
type=str,
help='config file path (default: None)')
args.add_argument('-r',
'--resume',
default=None,
type=str,
help='path to latest checkpoint (default: None)')
args.add_argument('-d',
'--device',
default=None,
type=str,
help='indices of GPUs to enable (default: all)')
config_test = ConfigParser.from_args(args)
main(config_test)
def parse_args():
global config
args = argparse.ArgumentParser(
description='MASTER PyTorch Distributed Training')
args.add_argument('-c',
'--config',
default=None,
type=str,
help='config file path (default: None)')
args.add_argument('-r',
'--resume',
default=None,
type=str,
help='path to latest checkpoint (default: None)')
args.add_argument('-d',
'--device',
default=None,
type=str,
help='indices of GPUs to be available (default: all)')
# custom cli options to modify configuration from default values given in json file.
CustomArgs = collections.namedtuple('CustomArgs',
'flags default type target help')
options = [
# CustomArgs(['--lr', '--learning_rate'], default=0.0001, type=float, target='optimizer;args;lr',
# help='learning rate (default: 0.0001)'),
CustomArgs(
['-dist', '--distributed'],
default='true',
type=str,
target='distributed',
help='run distributed training, true or false, (default: true).'
' turn off distributed mode can debug code on one gpu/cpu'),
CustomArgs(
['--local_world_size'],
default=1,
type=int,
target='local_world_size',
help=
'the number of processes running on each node, this is passed in explicitly '
'and is typically either $1$ or the number of GPUs per node. (default: 1)'
),
CustomArgs(
['--local_rank'],
default=0,
type=int,
target='local_rank',
help=
'this is automatically passed in via torch.distributed.launch.py, '
'process will be assigned a local rank ID in [0,local_world_size-1]. (default: 0)'
),
CustomArgs(
['--finetune'],
default='false',
type=str,
target='finetune',
help=
'finetune mode will load resume checkpoint, but do not use previous config and optimizer '
'(default: false), so there has three running mode: normal, resume, finetune'
)
]
config = ConfigParser.from_args(args, options)
return config
