This is a pythia-410m model with LoRA, fine-tuned on CPU using a question-answer dataset for 30 epochs. It was a test to assess fine-tuning with and witout LoRA.
-
