hugpadhugpad

neopolita / Qwen3.6-11B-A3B-Niwaki-4bit-mlx

Text Generation2K460Updated Aug 9, 2026
Hugging Face

Downloads

46

Likes

0

Last updated

Aug 9, 2026

License

apache-2.0

Qwen3.6-35B-A3B pruned to 11B total / 3.05B active parameters, stored at 4-bit — 7.6× less expert memory than the 8-bit reference. Stock mlxlm, no custom code.

Tasks & library

  • text generation
  • mlx

Tags

  • mlx
  • qwen3_5_moe
  • moe
  • pruning
  • quantization
  • mixture-of-experts

Technical specifications

Parameters2K
Librarymlx
LanguagesNot available
Base modelmlx-community/Qwen3.6-35B-A3B-8bit
Licenseapache-2.0