Get TargetPlatformCapabilities¶
- model_compression_toolkit.get_target_platform_capabilities(fw_name, target_platform_name, target_platform_version=None)¶
Get a TargetPlatformCapabilities by the target platform model name and the framework name. For now, it supports frameworks ‘tensorflow’ and ‘pytorch’. For both of them the target platform model can be ‘default’, ‘imx500’, ‘tflite’, or ‘qnnpack’.
- Parameters:
fw_name – Framework name of the TargetPlatformCapabilities.
target_platform_name – Target platform model name the model will use for inference.
target_platform_version – Target platform capabilities version.
- Returns:
A TargetPlatformCapabilities object that models the hardware and attaches a framework information to it.
- Return type:
Note
For now, some fields of OpQuantizationConfig are ignored during
the optimization process such as quantization_preserving, fixed_scale, and fixed_zero_point.
MCT will use more information from
OpQuantizationConfig, in the future.