def test_fn_multiple_input_sets():
pipe = Pipeline(batch_size = 1, num_threads = 1, device_id = 0)
image1 = np.array([
[1, 2, 3, 4],
[5, 6, 7, 8],
[9, 10, 11, 12]], dtype=np.uint8)[:,:,np.newaxis]
image2 = np.array([
[10, 20],
[30, 40],
[50, 60]], dtype=np.uint8)[:,:,np.newaxis]
batches = [[image1], [image2]]
inputs = fn.external_source(lambda: batches, 2, layout = "HWC")
rotated = fn.rotate(inputs, angle = 90)
pipe.set_outputs(*rotated)
pipe.build()
outs = pipe.run()
arr1 = outs[0].at(0)
arr2 = outs[1].at(0)
ref1 = np.array([
[4, 8, 12],
[3, 7, 11],
[2, 6, 10],
[1, 5, 9]])[:,:,np.newaxis]
ref2 = np.array([
[20, 40, 60],
[10, 30, 50]], dtype=np.uint8)[:,:,np.newaxis]
assert(np.array_equal(arr1, ref1))
assert(np.array_equal(arr2, ref2))
def get_pipeline(folder="train", custom_reader=None):
pipe = Pipeline(batch_size=64, num_threads=1, device_id=1)
if custom_reader:
raw_files, labels = custom_reader
else:
raw_files, labels = fn.file_reader(file_root="%s" % folder,
random_shuffle=True)
decode = fn.image_decoder(raw_files, device="mixed", output_type=types.GRAY)
resize = fn.resize(decode, device="gpu", image_type=types.RGB,
interp_type=types.INTERP_LINEAR, resize_x=WIDTH, resize_y=HEIGHT)
hsv = fn.hsv(resize, hue=fn.uniform(range=(-10, 10)), saturation=fn.uniform(range=(-.5, .5)),
value=fn.uniform(range=(0.9, 1.2)), device="gpu", dtype=types.UINT8)
bc = fn.brightness_contrast(hsv, device="gpu", brightness=fn.uniform(range=(.9, 1.1)))
cmn = fn.crop_mirror_normalize(bc, device="gpu", output_dtype=types.FLOAT,
output_layout=types.NHWC,
image_type=types.GRAY,
mean=[255 // 2],
std=[255 // 2])
rot = fn.rotate(cmn, angle=fn.uniform(range=(-40, 40)), device="gpu", keep_size=True)
tpose = fn.transpose(rot, perm=(2, 0, 1), device="gpu") # Reshaping to a format PyTorch likes
pipe.set_outputs(tpose, labels)
pipe.build()
dali_iter = DALIClassificationIterator([pipe], -1)
return dali_iter
def test_move_to_device_middle():
test_data_shape = [1, 3, 0, 4]
def get_data():
out = [
np.empty(test_data_shape, dtype=np.uint8)
for _ in range(batch_size)
]
return out
pipe = Pipeline(batch_size=batch_size, num_threads=3, device_id=None)
data = fn.external_source(source=get_data)
outs = fn.rotate(data.gpu(), angle=25)
pipe.set_outputs(outs)
assert_raises(RuntimeError, pipe.build)
def _test_fn_rotate(device):
pipe = Pipeline(batch_size=1, num_threads=1, device_id=0)
image = np.array([[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12]],
dtype=np.uint8)[:, :, np.newaxis]
batch = [image]
input = fn.external_source([batch], layout="HWC")
rotated = fn.rotate(input.gpu() if device == "gpu" else input, angle=90)
pipe.set_outputs(rotated)
pipe.build()
outs = pipe.run()
out = outs[0] if device == "cpu" else outs[0].as_cpu()
arr = out.at(0)
ref = np.array([[4, 8, 12], [3, 7, 11], [2, 6, 10], [1, 5, 9]])[:, :,
np.newaxis]
assert (np.array_equal(arr, ref))
def test_move_to_device_middle():
test_data_shape = [1, 3, 0, 4]
def get_data():
out = [
np.empty(test_data_shape, dtype=np.uint8)
for _ in range(batch_size)
]
return out
pipe = Pipeline(batch_size=batch_size, num_threads=3, device_id=None)
data = fn.external_source(source=get_data)
outs = fn.rotate(data.gpu(), angle=25)
pipe.set_outputs(outs)
assert_raises(
RuntimeError,
pipe.build,
glob=
"Cannot add a GPU operator Rotate, device_id should not be equal CPU_ONLY_DEVICE_ID."
)
