Trained on Pygamlion-Vicuna-7b on 2epochs of CheeseFire dataset, lora settings are in the file provided
-