python -m experimental_experiment run¶
usage: run [-h] -m MODEL [-v VERBOSE] [-b BATCH] [-r REPEAT] [-w WARMUP]
[-p PROCESSOR] [--validate VALIDATE]
Runs a model with dummy inputs and measures the inference time.
options:
-h, --help show this help message and exit
-m MODEL, --model MODEL
onnx model to optimize
-v VERBOSE, --verbose VERBOSE
verbosity
-b BATCH, --batch BATCH
batch size, if it can be changed
-r REPEAT, --repeat REPEAT
number of time to repeat the measure
-w WARMUP, --warmup WARMUP
number of time to warmup the model
-p PROCESSOR, --processor PROCESSOR
providers to launch, CPU, CUDA, CUDA,CPU.
--validate VALIDATE validate the output with another model
It checks a model runs and the inference time on the same inputs.