Skip to content

Latest commit

 

History

History
5 lines (4 loc) · 378 Bytes

File metadata and controls

5 lines (4 loc) · 378 Bytes

Fine tuning steps

  • Run finetune_model.py with default configuration to generate QLORA adapters (takes few minutes with very minimal parameters configured by default).
  • Sync the latest adapters from localhost with the huggingface repository (link: https://huggingface.co/tapopadma).
  • Run inference_cli.sh for manual interaction with the latest version of the model.