Get FrameworkQuantizationCapabilities

model_compression_toolkit.get_target_platform_capabilities(fw_name, target_platform_name, target_platform_version=None)

This is a degenerated function that only returns the MCT default TargetPlatformCapabilities object, to comply with the existing TPC API.

Parameters:

  • fw_name – Framework name of the FrameworkQuantizationCapabilities.

  • target_platform_name – Target platform model name the model will use for inference.

  • target_platform_version – Target platform capabilities version.

Returns:

A default TargetPlatformCapabilities object.

Return type:

TargetPlatformCapabilities

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.