Trained on 5 epochs of the full GGB dataset, on Pygamlion-vicuna-7b. Training settings are in the files
-