File size: 981 Bytes
2f674a7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 |
---
license: unlicense
language:
- en
base_model:
- ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.2
tags:
- roleplay
- rp
- sillytavern
- gguf
- llama
inference: false
---
My GGUF-IQ-ARM-Imatrix quants for [**ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.2**](https://huggingface.co/ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.2).
"RPMax is a series of models that are trained on a diverse set of curated creative writing and RP datasets with a focus on variety and deduplication. This model is designed to be highly creative and non-repetitive by making sure no two entries in the dataset have repeated characters or situations, which makes sure the model does not latch on to a certain personality and be capable of understanding and acting appropriately to any characters or situations. Early tests by users mentioned that these models does not feel like any other RP models..."
> [!NOTE]
> Use the **Llama 3 Instruct** prompt format.
<img src="https://iili.io/2f5bgWb.png" alt="anime-girl-model-card"> |