- open a jupyter server and run
pip install -r requirements.txt - run
pip install intel_extension_for_pytorch==2.0.110+xpu -f https://developer.intel.com/ipex-whl-stable-xpu(if using xpu enabled device) - run notebooks using shift+enter to see results, instruction are there on notebook
- preferred order train and quantize => submission generation