Llama-3.2-8X4B-MOE-V2-Dark-Champion-Instruct-uncensored-abliterated-21B-Q_6-MLX
mlx-community/Llama-3.2-8X4B-MOE-V2-Dark-Champion-Instruct-uncensored-abliterated-21B-Q_6-MLX
Text generation · mlx
6-bit Open on Hugging Face
Parameters
21B
Weights (est.)
≈ 17.0 GB
Mac memory (est.)
24 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/Llama-3.2-8X4B-MOE-V2-Dark-Champion-Instruct-uncensored-abliterated-21B-Q_6-MLX --prompt "Hello"
Details
- Base model
- DavidAU/L3.2-8X4B-MOE-V2-Dark-Champion-Inst-21B-uncen-ablit
- Vendor
- DavidAU
- Task
- Text generation
- Quantisation
- 6-bit
- Licence
- —
- Downloads
- 2.5K in the last 30 days · 12K all time · 3 likes
- Published
- 2025-06-25 · updated 2025-06-25 (1y ago)
- Tags
- mixtral, Llama 3.2, 8 X 4B, Brainstorm 5x, 128k context, 8 experts, mixture of experts, fine tune
Other MLX builds of DavidAU/L3.2-8X4B-MOE-V2-Dark-Champion-Inst-21B-uncen-ablit
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Llama-3.2-8X4B-MOE-V2-Dark-Champion-Instruct-uncensored-abliterated-21B-Q_6-MLX (this one) | 6-bit | ≈ 17.0 GB | 24 GB+ | 2.5K |
| Llama-3.2-8X4B-MOE-V2-Dark-Champion-Instruct-uncensored-abliterated-21B-MLX | 8-bit | ≈ 22.2 GB | 32 GB+ | 258 |