mlabonne
Llama 3.1 Abliterated 8B
A reduced-refusal Llama model adapted by mlabonne. Experimental; answer quality can differ from the original.
About Llama 3.1 Abliterated 8B
A community adaptation of Llama 3.1 8B by mlabonne with refusal behaviour reduced. Experimental, and it needs the same flagship memory as the base 8B model.
Every model in this catalog runs locally in the Offline AI Chat app. After the download finishes, Llama 3.1 Abliterated answers on your phone with no connection, and your prompts and replies are never uploaded.
- Download size
- 4.9 GB (Q4_K_M GGUF)
- Memory needed
- 10 GB minimum · 12 GB recommended
- Context in app
- 2,048 tokens
- Runs on
- Android 8+, 64-bit · iOS in private testing
- Access
- Offline Plus subscription
- License
- Llama 3.1 Community License
- Model source
- mlabonne on Hugging Face