ISTA-DASLab / Qwen3.6-35B-A3B-2Bit-GSQ
Vision36B97133Updated Aug 14, 2026
Downloads
971
Likes
33
Last updated
Aug 14, 2026
License
apache-2.0
Low-bit quantization of Qwen/Qwen3.6-35B-A3B (MoE, 35B total / 3B active) produced with GSQ (Gumbel-Softmax Quantization).
Tasks & library
- image text to text
- transformers
Tags
- transformers
- qwen3_5_moe
- image-text-to-text
- gsq
- gumbel-softmax
- quantization
Technical specifications
Parameters36B
Librarytransformers
LanguagesNot available
Base modelQwen/Qwen3.6-35B-A3B
Licenseapache-2.0