def test_pytorch_iterator_not_fill_last_batch_pad_last_batch():
from nvidia.dali.plugin.pytorch import DALIGenericIterator as PyTorchIterator
num_gpus = 1
batch_size = 100
pipes, data_size = create_pipeline(
lambda gpu: COCOReaderPipeline(batch_size=batch_size,
num_threads=4,
shard_id=gpu,
num_gpus=num_gpus,
data_paths=data_sets[0],
random_shuffle=False,
stick_to_shard=False,
shuffle_after_epoch=False,
pad_last_batch=True), batch_size,
num_gpus)
dali_train_iter = PyTorchIterator(pipes,
output_map=["data"],
size=pipes[0].epoch_size("Reader"),
fill_last_batch=False,
last_batch_padded=True)
img_ids_list, img_ids_list_set, mirrored_data, _, _ = \
gather_ids(dali_train_iter, lambda x: x["data"].squeeze().numpy(), lambda x: 0, data_size)
assert len(img_ids_list) == data_size
assert len(img_ids_list_set) == data_size
assert len(set(mirrored_data)) != 1
dali_train_iter.reset()
next_img_ids_list, next_img_ids_list_set, next_mirrored_data, _, _ = \
gather_ids(dali_train_iter, lambda x: x["data"].squeeze().numpy(), lambda x: 0, data_size)
# there is no mirroring as data in the output is just cut off,
# in the mirrored_data there is real data
assert len(next_img_ids_list) == data_size
assert len(next_img_ids_list_set) == data_size
assert len(set(next_mirrored_data)) != 1
def test_pytorch_iterator_last_batch_pad_last_batch():
num_gpus = 1
batch_size = 100
iters = 0
pipes, data_size = create_pipeline(
lambda gpu: COCOReaderPipeline(batch_size=batch_size,
num_threads=4,
shard_id=gpu,
num_gpus=num_gpus,
data_paths=data_sets[0],
random_shuffle=True,
stick_to_shard=False,
shuffle_after_epoch=False,
pad_last_batch=True), batch_size,
num_gpus)
dali_train_iter = PyTorchIterator(pipes,
output_map=["data"],
size=pipes[0].epoch_size("Reader"),
fill_last_batch=True)
img_ids_list, img_ids_list_set, mirrored_data, _, _ = \
gather_ids(dali_train_iter, lambda x: x["data"].squeeze().numpy(), lambda x: 0, data_size)
assert len(img_ids_list) > data_size
assert len(img_ids_list_set) == data_size
assert len(set(mirrored_data)) == 1
dali_train_iter.reset()
next_img_ids_list, next_img_ids_list_set, next_mirrored_data, _, _ = \
gather_ids(dali_train_iter, lambda x: x["data"].squeeze().numpy(), lambda x: 0, data_size)
assert len(next_img_ids_list) > data_size
assert len(next_img_ids_list_set) == data_size
assert len(set(next_mirrored_data)) == 1
eii = CXRImageIterator(batch_size,
0,
1,
train_ids,
train_labels,
args.image_dir,
image_format=args.image_format,
encoding=args.label_encoding)
pipe = CXRImagePipeline(batch_size=batch_size,
num_threads=2,
device_id=0,
num_gpus=4,
external_data=eii)
pii = PyTorchIterator(pipe,
size=eii.size,
last_batch_padded=True,
fill_last_batch=False)
times = []
for e in range(1):
for i, data in enumerate(pii):
if i == 100: break
if i != 0:
stop = time()
times.append(stop - start)
print("epoch: {}, iter {}, real batch size: {}, data shape: {}".
format(e, i, len(data[0]["data"]), data[0]['data'].shape))
img = data[0]['data'].squeeze().cpu().numpy()
print(np.max(img))
print(np.min(img))
print(img.shape)
def test_stop_iteration_pytorch():
from nvidia.dali.plugin.pytorch import DALIGenericIterator as PyTorchIterator
test_stop_iteration(lambda pipe, size, auto_reset : PyTorchIterator(pipe, output_map=["data"], size=size, auto_reset=auto_reset), "PyTorchIterator")
def test_stop_iteration_pytorch_fail_single():
from nvidia.dali.plugin.pytorch import DALIGenericIterator as PyTorchIterator
fw_iter = lambda pipe, size, auto_reset: PyTorchIterator(
pipe, output_map=["data"], size=size, auto_reset=auto_reset)
check_stop_iter_fail_single(fw_iter)
def test_stop_iteration_pytorch():
from nvidia.dali.plugin.pytorch import DALIGenericIterator as PyTorchIterator
fw_iter = lambda pipe, size, auto_reset : PyTorchIterator(pipe, output_map=["data"], size=size, auto_reset=auto_reset)
iter_name = "PyTorchIterator"
for batch_size, epochs, iter_num, auto_reset, infinite in stop_teration_case_generator():
yield check_stop_iter, fw_iter, iter_name, batch_size, epochs, iter_num, auto_reset, infinite
