---
license: apache-2.0
---

Train 4Chan Dataset, fp16 weights and more info on [branch Llama-3-8B-4Chan_SD_QLoRa](https://huggingface.co/Nekochu/Luminia-8B-v4-Chan/tree/Llama-3-8B-4Chan_SD_QLoRa) (other branch are failed attempt)

- System Prompts: `write a response like a 4chan user`
- Note: Fine-Tuned of Llama-3-8B, attempt to replicate [ykilcher/gpt-4chan](https://huggingface.co/ykilcher/gpt-4chan).