Training aware quantization or bypass the quanitization while executing 1_quantize_model.sh

While executing " each operation times usage" each operation times usage | Khadas Documentation I want to avaid the quantization step and want to have a training-aware quantization instead. How can i acheive this using different parameters of 0_import_model.sh 1_quantize_model.sh and 2_export_case_code.sh.

Thanks @Frank @numbqq
Regards,

@enggsajjad You can refer to the documentation in the SDK, which describes each parameter of the conversion script in detail

1 Like