---
language: en
tags:
- obliteratus
- abliteration
- uncensored
- obliterate
- mlx
- mlx-my-repo
base_model: usermma/GRM-Kerlin-3b-Abliterated
---

# usermma/GRM-Kerlin-3b-Abliterated-mlx-8Bit

The Model [usermma/GRM-Kerlin-3b-Abliterated-mlx-8Bit](https://huggingface.co/usermma/GRM-Kerlin-3b-Abliterated-mlx-8Bit) was converted to MLX format from [usermma/GRM-Kerlin-3b-Abliterated](https://huggingface.co/usermma/GRM-Kerlin-3b-Abliterated) using mlx-lm version **0.31.2**.

## Use with mlx

```bash
pip install mlx-lm
```

```python
from mlx_lm import load, generate

model, tokenizer = load("usermma/GRM-Kerlin-3b-Abliterated-mlx-8Bit")

prompt="hello"

if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)
```