CoolFace
Modelpublic

mirazrafi/NSFW-RP-RolePlay-LoRA-ArliAI-Llama-3.1-8B

sourceHugging Faceapache-2.0updated 7mo agoView on Hugging Face
5likes21downloads
Model Card

ArliAI Llama 3.1 8B RP – NSFW LoRA Adapter

⚠️ Content Warning

This model is intended for research and experimental purposes only. It is fine-tuned for NSFW / adult roleplay content.

Do NOT use this model if you are under the age of 18.

Model Description

This repository contains a LoRA adapter fine-tuned to enhance adult, NSFW, and character-driven roleplay dialogue.

The adapter was trained on top of:

  • —Base model: meta-llama/Llama-3.1-8B
  • —Intermediate model: ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.3

⚠️ This repository does NOT contain any base model weights. Users must obtain the base model separately and comply with all applicable licenses.

Intended Use

This model is intended for:

  • —Research into conversational roleplay behavior
  • —Studying narrative consistency in adult dialogue
  • —Experimentation with NSFW roleplay systems

❌ Not Intended For

  • —Minors or content involving minors
  • —Non-consensual, exploitative, or abusive content
  • —Real-person sexual content
  • —Illegal, harmful, or safety-critical applications
  • —Medical, legal, or financial advice

Ethical & Safety Notice

Users are responsible for ensuring that:

  • —All generated content is consensual
  • —No content involving minors is generated
  • —Usage complies with local laws and platform policies

The author does not endorse or encourage misuse.

How to Use

python
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

base_model = "ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.3"
lora_model = "mirazrafi/NSFW-RP-LoRA-ArliAI-Llama-3.1-8B"

tokenizer = AutoTokenizer.from_pretrained(base_model)
model = AutoModelForCausalLM.from_pretrained(base_model, device_map="auto")
model = PeftModel.from_pretrained(model, lora_model)

Training Details

  • —Fine-tuning method: LoRA (PEFT)
  • —Objective: Improve NSFW roleplay quality and dialogue coherence
  • —Frameworks: Transformers, PEFT, TRL, Unsloth
  • —Precision: FP16

Distribution Notice

  • —This repository distributes LoRA adapter weights only
  • —Merged models or GGUF files are NOT provided
  • —Users may merge and convert models locally for personal use only

License

  • —LoRA adapter: Apache License 2.0
  • —Base model: Llama 3.1 Communuty License Agreement
  • —Intermediate model: ArliAI License

This license applies only to the adapter weights and does not override base model licenses.

Acknowledgements

  • —Meta AI – LLaMA 3.1
  • —ArliAI – RPMax v1.3
  • —Hugging Face, PEFT, and TRL ecosystems