Benchmarker
Benchmark Model
benchmark_model(self, input_model_path: str, target_device_name: DeviceName, target_software_version: str | SoftwareVersion | None = None, target_hardware_type: str | HardwareType | None = None, wait_until_done: bool = True, sleep_interval: int = 30) → BenchmarkerMetadata
Benchmark the specified model on the specified device.
- Parameters:
- input_model_path (str) – The file path where the model is located.
- target_device_name (DeviceName) – Target device name.
- target_software_version (Union *[*str , SoftwareVersion ] , optional) – Target software version. Required if target_device_name is one of the Jetson devices.
- target_hardware_type (Union *[*str , HardwareType ] , optional) – Hardware type. Acceleration options for processing the model inference.
- wait_until_done (bool) – If True, wait for the benchmark result before returning the function.
If False, request the benchmark and return the function immediately.
- Raises:
e – If an error occurs during the benchmarking of the model. - Returns:
Benchmark metadata. - Return type:
BenchmarkerMetadata
Example
from netspresso import NetsPresso
from netspresso.enums import DeviceName, SoftwareVersion
netspresso = NetsPresso(email="YOUR_EMAIL", password="YOUR_PASSWORD")
benchmarker = netspresso.benchmarker_v2()
benchmark_task = benchmarker.benchmark_model(
input_model_path="./outputs/converted/TENSORRT_JETSON_AGX_ORIN_JETPACK_5_0_1/TENSORRT_JETSON_AGX_ORIN_JETPACK_5_0_1.trt",
target_device_name=DeviceName.JETSON_AGX_ORIN,
target_software_version=SoftwareVersion.JETPACK_5_0_1,
)
Updated about 1 month ago