Downloads · 30 days
23
1% of all-time downloads
Qwen3.6-35B-A3B-oQ8-fp16-mtp
tfjack/Qwen3.6-35B-A3B-oQ8-fp16-mtp
Qwen3.6-35B-A3B-oQ8-fp16-mtp is a machine learning model from tfjack. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for mlx.
[!WARNING] ⚠️ DEPRECATED — superseded by the official oMLX build. These community oQ quants were converted before the oMLX 0.7.3 oQ improvement, which changed the bit split to lower KL by up to 73% at the same quant l…
All-time downloads
4.2K
Public
Repo size
38.6 GB
Likes
3
Public
Repo makeup
Click a slice to open those files.
.md1.2 KB · 100%
At a glance
- Library
- mlx
- Access
- Public
- Created
- May 13, 2026
- Updated
- Oct 29, 2026
- SHA
- 1e004654
Model card
From the Hugging Face model README
[!WARNING] ⚠️ DEPRECATED — superseded by the official oMLX build.
These community oQ quants were converted before the oMLX 0.7.3 oQ improvement, which changed the bit split to lower KL by up to 73% at the same quant level (oMLX PR #4776). The original oMLX maintainer (
Jundot) now publishes better, officially-maintained quants of this same base model — please use those instead.👉 Official replacement: https://huggingface.co/Jundot/Qwen3.6-35B-A3B-oQ4e-mtp (also oQ2.5e / oQ3e / oQ3.5e / oQ6e / oQ8e)
The weight files on this repository have been removed. This card is kept only as a pointer.
[!IMPORTANT] This quantization was uploaded on 2026-05-20 and replaces a previous version. If you downloaded this model before this date, please re-download for the updated weights.
Qwen3.6-35B-A3B-oQ8-fp16-mtp
This model was quantized using oQ (oMLX v0.3.9rc1) mixed-precision quantization.
Quantization details
- Model type: qwen3_5_moe
- Bits: 8
- Group size: 64
- Format: MLX safetensors
- Library
- mlx
- Created
- May 13, 2026
- Updated
- Oct 29, 2026