Maksim Proshin comments

Results 47 comments of


                                            Maksim Proshin

The question about function create_compressed_model()：RuntimeError: CUDA error: device-side assert triggered

Closed issue as no answer for a long time period. Feel free to re-open it if needed.

untimeError: CUDA is available for PyTorch, but NNcF could not compile GPu quan tization extensions.

@huangqiu15444 , do you have any update here? Did you check what @vshampor recommended above about the compatibility?

untimeError: CUDA is available for PyTorch, but NNcF could not compile GPu quan tization extensions.

@huangqiu15444 , please give us an update if any.

untimeError: CUDA is available for PyTorch, but NNcF could not compile GPu quan tization extensions.

@huangqiu15444 , do you mean the issue is resolved now? Can we close it?

When will there be a c++ interface?

Hi @qq978358810 , we don't have plans for C++ interface. How that will help you compared to Python? Model compression is an offline step, so usually you don't integrate it...

When will there be a c++ interface?

@qq978358810 , I close this issue. If you want us to consider it in future development, please provide more details on how you plan to use it and why the...

Tiny Llama ONNX llm_compression example missing dependancies

@andrey-churkin , please take a look.

ov.compile_model(quantized_model, device_name="GPU") is not working

@paguilomanas , please note that the quantization actually failed, but I think it's not due to NNCF but some issue with your code or settings. Can you please try to...

ov.compile_model(quantized_model, device_name="GPU") is not working

@paguilomanas , thanks for your feedback! It means that there is some issue in GPU Plugin of OpenVINO and if you want to follow it up, feel free to open...

Algorithm.py refactoring

@alexsu52 ,can this PR be closed?