This is a LORA model, trained on PuffedLIMASub4000 for 4 epochs, BS1.

Base model is LLAMA-2-13b-HF

This model acts as a test for ConvoEvol.