MCPcopy Create free account
hub / github.com/NVIDIA/DALI / check_slice_vs_numpy

Function check_slice_vs_numpy

dali/test/python/operator_1/test_slice.py:546–556  ·  view source on GitHub ↗
(device, batch_size, axes, axis_names)

Source from the content-addressed store, hash-verified

544
545
546def check_slice_vs_numpy(device, batch_size, axes, axis_names):
547 eii_args = [
548 SliceArgsIterator(batch_size, image_layout="HWC", axes=axes, axis_names=axis_names)
549 for k in range(2)
550 ]
551 compare_pipelines(
552 SlicePipeline(device, batch_size, iter(eii_args[0]), axes=axes, axis_names=axis_names),
553 SlicePythonOp(batch_size, iter(eii_args[1]), axes=axes, axis_names=axis_names),
554 batch_size=batch_size,
555 N_iterations=3,
556 )
557
558
559def test_slice_vs_numpy():

Callers

nothing calls this directly

Calls 4

compare_pipelinesFunction · 0.90
SlicePipelineClass · 0.85
SlicePythonOpClass · 0.85
SliceArgsIteratorClass · 0.70

Tested by

no test coverage detected