SmartyPlats-7b An experimental finetune of Mistrel 7b with QLoRA Datasets Trained on alpca style datasets Prompt Template Uses alpaca style prompt template