|
|
--- |
|
|
language: |
|
|
- en |
|
|
license: mit |
|
|
pipeline_tag: text-generation |
|
|
tags: |
|
|
- mlx |
|
|
- mixture-of-experts |
|
|
- moe |
|
|
- pruning |
|
|
- reap |
|
|
- minimax |
|
|
- 4bit |
|
|
- quantized |
|
|
- apple-silicon |
|
|
library_name: mlx |
|
|
base_model: Akicou/MiniMax-M2-5-REAP-29 |
|
|
--- |
|
|
|
|
|
<p align="center"> |
|
|
<a href="https://vmlx.net"> |
|
|
<img src="vmlx-logo.png" alt="vMLX" width="120"> |
|
|
</a> |
|
|
</p> |
|
|
|
|
|
# MiniMax-M2.5 REAP-29 — MLX 4-bit |
|
|
|
|
|
MLX 4-bit quantized version of [Akicou/MiniMax-M2-5-REAP-29](https://huggingface.co/Akicou/MiniMax-M2-5-REAP-29) for efficient local inference on Apple Silicon. |
|
|
|
|
|
- **Quantization**: 4-bit (group size 64, affine mode; router gates at 8-bit) |
|
|
- **Architecture**: MiniMax M2.5 MoE — 62 layers, 180 experts (REAP-pruned from 256), 8 active per token |
|
|
- **Context**: 196K tokens |
|
|
- **Size**: ~85 GB |
|
|
- **Pruning**: 29% of experts removed via [REAP](https://github.com/CerebrasResearch/reap) (Router Expert Activation Pruning) |
|
|
|
|
|
## Usage |
|
|
|
|
|
```python |
|
|
from mlx_lm import load, generate |
|
|
|
|
|
model, tokenizer = load("shieldstackllc/MiniMax-M2.5-REAP-29-mlx-4bit") |
|
|
response = generate(model, tokenizer, prompt="Hello!", verbose=True) |
|
|
``` |
|
|
|
|
|
Or with [vMLX](https://vmlx.net) for native macOS inference. |
|
|
|
|
|
## About |
|
|
|
|
|
MiniMax-M2.5 is a large Mixture-of-Experts language model by MiniMax AI. This variant was pruned to 29% fewer experts by [Akicou](https://huggingface.co/Akicou) using REAP (Router Expert Activation Pruning), reducing model size and memory footprint while maintaining strong performance. MLX quantization by [vMLX](https://vmlx.net). |
|
|
|
|
|
## Also Available |
|
|
|
|
|
- [MiniMax-M2.5-REAP-39 MLX 4-bit](https://huggingface.co/shieldstackllc/MiniMax-M2-5-REAP-39-mlx-4bit) (~73 GB) — 39% pruned variant |
|
|
- [MiniMax-M2.5-REAP-39 MLX 8-bit](https://huggingface.co/shieldstackllc/MiniMax-M2-5-REAP-39-mlx-8bit) (~138 GB) — 39% pruned variant |
|
|
|
|
|
## Made for vMLX |
|
|
|
|
|
This model was converted and optimized for [vMLX](https://vmlx.net) — a free, open source macOS native MLX inference engine for Apple Silicon. Download vMLX to run this model locally with zero configuration. |
|
|
|
|
|
## Credits |
|
|
|
|
|
- **Base model**: [MiniMaxAI/MiniMax-M2.5](https://huggingface.co/MiniMaxAI/MiniMax-M2.5) by MiniMax AI |
|
|
- **REAP pruning**: [Akicou/MiniMax-M2-5-REAP-29](https://huggingface.co/Akicou/MiniMax-M2-5-REAP-29) by Akicou |
|
|
- **MLX conversion**: [vMLX](https://vmlx.net) — Run AI locally on Mac. No compromises. |
|
|
|
|
|
## Contact |
|
|
|
|
|
For questions, issues, or collaboration: **admin@vmlx.net** |
|
|
|
