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

Function check_slice_vs_fused_decoder

dali/test/python/operator_1/test_slice.py:512–536  ·  view source on GitHub ↗
(device, batch_size, axes, axis_names)

Source from the content-addressed store, hash-verified

510
511
512def check_slice_vs_fused_decoder(device, batch_size, axes, axis_names):
513 eii_args = [
514 SliceArgsIterator(batch_size, image_layout="HWC", axes=axes, axis_names=axis_names)
515 for k in range(2)
516 ]
517 compare_pipelines(
518 SlicePipeline(
519 device,
520 batch_size,
521 iter(eii_args[0]),
522 axes=axes,
523 axis_names=axis_names,
524 is_fused_decoder=False,
525 ),
526 SlicePipeline(
527 device,
528 batch_size,
529 iter(eii_args[1]),
530 axes=axes,
531 axis_names=axis_names,
532 is_fused_decoder=True,
533 ),
534 batch_size=batch_size,
535 N_iterations=3,
536 )
537
538
539def test_slice_vs_fused_decoder():

Callers

nothing calls this directly

Calls 3

compare_pipelinesFunction · 0.90
SlicePipelineClass · 0.85
SliceArgsIteratorClass · 0.70

Tested by

no test coverage detected