MCPcopy
hub / github.com/OpenPPL/ppq / enable_cuda_kernel.py

File enable_cuda_kernel.py

ppq/samples/enable_cuda_kernel.py:None–None  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

1# Since ppq 0.6.4, PPQ_CONFIG.USING_CUDA_KERNEL = False is the defualt execution option in ppq.
2# However you should notice that if you are able to compile ppq kernel functions, the execution speed wiil boost at least 3x.
3# This example will show you how to enable kernel function within ppq.
4# if you want to use kernel function everywhere, just rewrite ppq.core.config.PPQ_CONFIG.USING_CUDA_KERNEL = True

Callers

nothing calls this directly

Calls 6

ENABLE_CUDA_KERNELClass · 0.90
quantize_torch_modelFunction · 0.90
export_ppq_graphFunction · 0.90
toMethod · 0.80
pplcuda_settingMethod · 0.80
load_calibration_datasetFunction · 0.70

Tested by

no test coverage detected