This LORA was trained 17k lines of discord texts from GiftedGummyBee, trained on 100 epochs periodically. The base model is StableLM 3B finetuned