def verify():
logger = LoggerManager().get_logger()
parser = argparse.ArgumentParser('get_network')
parser.add_argument('--config_path', type=str, default='FairNasC')
parser.add_argument('--weights_path', type=str, default='{path_tmp}/s3/')
parser.add_argument('--data_dir',
type=str,
default='{path_data}/ImageNet_ILSVRC2012/')
parser.add_argument('--data_batch_size', type=int, default=128)
parser.add_argument('--data_num_workers', type=int, default=8)
parser.add_argument('--num_batches',
type=int,
default=-1,
help='>0 to stop early')
args, _ = parser.parse_known_args()
# ImageNet with default augmentations / cropping
data_set = get_imagenet(
data_dir=args.data_dir,
num_workers=args.data_num_workers,
batch_size=args.data_batch_size,
aug_dict={
"cls_augmentations": "TimmImagenetAug",
"DartsImagenetAug#0.crop_size": 224,
},
)
# network
network = get_network(args.config_path, data_set.get_data_shape(),
data_set.get_label_shape(), args.weights_path)
network.eval()
network = network.cuda()
# measure the accuracy
top1, top5, num_samples = 0, 0, 0
with torch.no_grad():
for i, (data, targets) in enumerate(data_set.test_loader()):
if i >= args.num_batches > 0:
break
outputs = network(data.cuda())
t1, t5 = accuracy(outputs, targets.cuda(), topk=(1, 5))
n = data.size(0)
top1 += t1 * n
top5 += t5 * n
num_samples += n
logger.info('results:')
logger.info('\ttested images: %d' % num_samples)
logger.info('\ttop1: %.4f (%d/%d)' %
(top1 / num_samples, top1, num_samples))
logger.info('\ttop5: %.4f (%d/%d)' %
(top5 / num_samples, top5, num_samples))
def initialize_weights(self, net: AbstractModule):
logger = LoggerManager().get_logger()
logger.info('Initializing: %s' % self.__class__.__name__)
self._initialize_weights(net, logger)
def __init__(self, data_dir: str, save_dir: Union[str, None],
bs_train: int, bs_test: int,
train_transforms: transforms.Compose, test_transforms: transforms.Compose,
train_batch_aug: Union[BatchAugmentations, None], test_batch_aug: Union[BatchAugmentations, None],
num_workers: int, num_prefetch: int,
valid_split: Union[int, float], valid_shuffle: bool,
fake: bool, download: bool,
**additional_args):
"""
:param data_dir: where to find (or download) the data set
:param save_dir: global save dir, can store and reuse the info which data was used in the random valid split
:param bs_train: batch size for the train loader
:param bs_test: batch size for the test loader, <= 0 to have the same as bs_train
:param train_transforms: train augmentations (on each data point individually)
:param test_transforms: test augmentations (on each data point individually)
:param train_batch_aug: train augmentations (across the entire batch)
:param test_batch_aug: test augmentations (across the entire batch)
:param num_workers: number of workers prefetching data
:param num_prefetch: number of batches prefetched by every worker
:param valid_split: absolute number of data points if int or >1, otherwise a fraction of the training set
:param valid_shuffle: whether to shuffle validation data
:param fake: use fake data instead (no need to provide either real data or enabling downloading)
:param download: whether downloading is allowed
:param additional_args: arguments that are added and used by child classes
"""
super().__init__()
logger = LoggerManager().get_logger()
self.dir = data_dir
self.bs_train = bs_train
self.bs_test = bs_test if bs_test > 0 else self.bs_train
self.num_workers, self.num_prefetch = num_workers, num_prefetch
self.valid_shuffle = valid_shuffle
self.additional_args = additional_args
self.fake = fake
self.download = download and not self.fake
if self.download and (not self.can_download):
LoggerManager().get_logger().warning("The dataset can not be downloaded, but may be asked to.")
self.train_transforms = train_transforms
self.test_transforms = test_transforms
self.train_batch_augmentations = train_batch_aug
self.test_batch_augmentations = test_batch_aug
# load/create meta info dict
if isinstance(save_dir, str) and len(save_dir) > 0:
meta_path = '%s/data.meta.pt' % replace_standard_paths(save_dir)
if os.path.isfile(meta_path):
meta = torch.load(meta_path)
else:
meta = defaultdict(dict)
else:
meta, meta_path = defaultdict(dict), None
# give subclasses a good spot to react to additional arguments
self._before_loading()
# data
if self.fake:
train_data = self._get_fake_train_data(self.train_transforms)
self.test_data = self._get_fake_test_data(self.test_transforms)
else:
train_data = self._get_train_data(self.train_transforms)
self.test_data = self._get_test_data(self.test_transforms)
# split train into train+valid or using stand-alone valid set
if valid_split > 0:
s1 = int(valid_split) if valid_split >= 1 else int(len(train_data)*valid_split)
if s1 >= len(train_data):
logger.warning("Tried to set valid split larger than the training set size, setting to 0.5")
s1 = len(train_data)//2
s0 = len(train_data) - s1
if meta['splits'].get((s0, s1), None) is None:
meta['splits'][(s0, s1)] = torch.randperm(s0+s1).tolist()
indices = meta['splits'][(s0, s1)]
self.valid_data = torch.utils.data.Subset(train_data, np.array(indices[s0:]).astype(np.int))
train_data = torch.utils.data.Subset(train_data, np.array(indices[0:s0]).astype(np.int))
logger.info('Data Set: splitting training set, will use %s data points as validation set' % s1)
if self.length[1] > 0:
logger.info('Data Set: a dedicated validation set exists, but it will be replaced.')
elif self.length[1] > 0:
if self.fake:
self.valid_data = self._get_fake_valid_data(self.test_transforms)
else:
self.valid_data = self._get_valid_data(self.test_transforms)
logger.info('Data Set: using the dedicated validation set with test augmentations')
else:
self.valid_data = None
logger.info('Data Set: not using a validation set at all.')
self.train_data = train_data
# shapes
data, label = self.train_data[0]
self.data_shape = Shape(list(data.shape))
# save meta info dict
if meta_path is not None:
torch.save(meta, meta_path)
def explore(mini: MiniNATSBenchTabularBenchmark):
logger = LoggerManager().get_logger()
# some stats of specific results
logger.info(
mini.get_by_arch_tuple((4, 3, 2, 1, 0, 2)).get_info_str('cifar10'))
logger.info("")
mini.get_by_arch_tuple((1, 2, 1, 2, 3, 4)).print(logger.info)
logger.info("")
mini.get_by_index(1554).print(logger.info)
logger.info("")
mini.get_by_arch_str(
'|avg_pool_3x3~0|+|nor_conv_1x1~0|skip_connect~1|+|nor_conv_1x1~0|skip_connect~1|skip_connect~2|'
).print(logger.info)
logger.info("")
# best results by acc1
rows = [("acc1", "params", "arch tuple", "arch str")]
log_headline(
logger, "highest acc1 topologies (%s, %s, %s)" %
(mini.get_name(), mini.get_default_data_set(),
mini.get_default_result_type()))
for i, r in enumerate(mini.get_all_sorted(['acc1'], [True])):
rows.append(
(r.get_acc1(), r.get_params(), str(r.arch_tuple), r.arch_str))
if i > 8:
break
log_in_columns(logger, rows)
logger.info("")
# best results by acc1
rows = [("acc1", "arch tuple", "arch str")]
c = 0
log_headline(
logger, "highest acc1 topologies without skip (%s, %s, %s)" %
(mini.get_name(), mini.get_default_data_set(),
mini.get_default_result_type()))
for i, r in enumerate(mini.get_all_sorted(['acc1'], [True])):
if 1 not in r.arch_tuple:
rows.append((r.get_acc1(), str(r.arch_tuple), r.arch_str))
c += 1
if c > 9:
break
log_in_columns(logger, rows)
from uninas.utils.paths import replace_standard_paths
from uninas.register import Register
def example_export_network(path: str) -> AbstractUninasNetwork:
""" create a new network and export it, does not require to have onnx installed """
network = get_network("FairNasC",
Shape([3, 224, 224]),
Shape([1000]),
weights_path=None)
network = network.cuda()
network.export_onnx(path, export_params=True)
return network
try:
import onnx
if __name__ == '__main__':
logger = LoggerManager().get_logger()
export_path = replace_standard_paths("{path_tmp}/onnx/FairNasC.onnx")
net1 = example_export_network(export_path)
log_headline(logger, "onnx graph")
net2 = onnx.load(export_path)
onnx.checker.check_model(net2)
logger.info(onnx.helper.printable_graph(net2.graph))
except ImportError as e:
Register.missing_import(e)
