CoolFace
Modelpublic

OsaurusAI/Qwen3.5-35B-A3B-JANG_4K

sourceHugging Faceapache-2.0updated 5mo agoView on Hugging Face
1likes16downloads
Model Card

<p align="center"> <a href="https://osaurus.ai"><img src="https://cdn-avatars.huggingface.co/v1/production/uploads/69d00705ce8872981c6c4fce/GWKjOwezSOhW5iuKpDwq_.png" alt="Osaurus AI" width="120"></a> </p>

<h3 align="center">Qwen 3.5 35B-A3B &mdash; JANG_4K (Mixed-Precision, 4-bit)</h3> <p align="center"><b>JANG</b> &mdash; Jang Adaptive N-bit Grading | Mixed-Precision Quantization for Apple Silicon</p>

<p align="center"> <a href="https://osaurus.ai"><img src="https://img.shields.io/badge/Web-osaurus.ai-blue" alt="Website"></a>&nbsp; <a href="https://github.com/jjang-ai/jangq"><img src="https://img.shields.io/badge/GitHub-JANG_Source-blue?logo=github" alt="GitHub"></a>&nbsp; <a href="https://pypi.org/project/jang/"><img src="https://img.shields.io/pypi/v/jang?label=PyPI&color=green" alt="PyPI"></a>&nbsp; <a href="https://huggingface.co/OsaurusAI"><img src="https://img.shields.io/badge/HF-OsaurusAI-yellow?logo=huggingface" alt="OsaurusAI"></a> </p>

Osaurus natively supports JANG models. Download at osaurus.ai.

Model Details

PropertyValue
Base ModelQwen 3.5 VL 35B-A3B
ArchitectureMoE Transformer + Vision
Total Parameters35B (3B active per token)
ProfileJANG_4K
Avg Bits/Weight3.98
Bit Widths Used3, 4, 5, 8
Model Size16.4 GB
VisionYes
FormatJANG v2 (MLX-native safetensors)

Benchmarks

200-question MMLU (20 per subject x 10 subjects). Thinking OFF (enable_thinking=False), greedy decoding (temp=0.0).

ModelMMLUSize
JANG_4K (this)77.5%16.4 GB
MLX 4-bit75.5%18 GB
MLX 2-bit~20%10 GB

JANG_4K beats MLX 4-bit by +2 MMLU while being smaller (16.4 GB vs 18 GB). Budget-neutral bit redistribution boosts attention quality without increasing total size.

JANG_4K Profile

JANG_4K is a balanced 4-bit mixed-precision profile that provides near-original quality. Critical layers (attention, routing, embeddings) are kept at 8-bit, with expert MLP weights at 3-5 bit depending on importance scoring. Best quality-to-size ratio for most use cases.

Usage

bash
# Requires Osaurus (https://osaurus.ai)
osaurus serve OsaurusAI/Qwen3.5-35B-A3B-JANG_4K

Requirements

  • —Apple Silicon Mac with 24+ GB unified memory
  • —MLX framework with Qwen 3.5 MoE support

<p align="center">Quantized by <a href="https://osaurus.ai">Osaurus AI</a> using <a href="https://github.com/jjang-ai/jangq">JANG</a></p>