Downloads · 30 days
20
6% of all-time downloads
aro-teacher-30b-4bit
ARO-Lang/aro-teacher-30b-4bit
aro-teacher-30b-4bit is a text generation model from ARO-Lang. Use it when you need the model to write or continue text. It is set up for mlx. The card lists the license as mit.
The full 30B MoE teacher model, fine-tuned on ARO training data. This model is used for:
All-time downloads
310
Public
Parameters
30.5B
69.1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors17.2 GB · 100%
Parameter types
How the weights are stored.
U3230.5B · 100%
Try a prompt
Base models
Model card
From the Hugging Face model README
ARO Teacher (30B)
The full 30B MoE teacher model, fine-tuned on ARO training data. This model is used for:
- Distillation — generating high-quality training data for the smaller student model
- Iterative retraining — serving as the starting point for the next training cycle
- High-quality inference — when maximum accuracy is needed (at the cost of speed/memory)
For deployment and everyday use, prefer the distilled 8B student model: ARO-Lang/aro-coder-6bit
| Architecture | Qwen3 30B MoE (3.3B active parameters) |
| Base model | mlx-community/Qwen3-Coder-30B-A3B-Instruct-4bit |
| Quantization | 4-bit (MLX) |
| Size | ~16 GB |
| Training source | dpo |
Usage
from mlx_lm import load, generate
model, tokenizer = load("ARO-Lang/aro-teacher-30b-4bit")
Or as a base for continued fine-tuning:
python -m mlx_lm lora --model ARO-Lang/aro-teacher-30b-4bit --data ./train_data --train
Links
- Distilled student: ARO-Lang/aro-coder-6bit
- Website: arolang.github.io/aro
- GitHub: github.com/arolang/aro
- Language Guide: Wiki
License
MIT License
- Task
- Text Generation
- Library
- mlx
- Type
- qwen3_moe
- License
- mit
- Languages
- en
- Created
- May 5, 2026
- Updated
- Aug 10, 2026