hugpadhugpad

neopolita / Qwen3.8-Flash-Next-119B-A5B-Niwaki-v2.4-3bit-mlx

Text Generation118B7092Updated Sep 20, 2026
Hugging Face

Downloads

709

Likes

2

Last updated

Sep 20, 2026

License

other

Qwen3.8-Flash-Next pruned to 119B total / ~5.6B active parameters, stored at 3-bit — 44.8 GB on disk, 10 GB below the unpruned model's best 2-bit cast at the same perplexity and task average — for 64 GB Macs. The program-state behaviours we test (variable tracking, two-step ar…

Tasks & library

  • text generation
  • mlx

Tags

  • mlx
  • qwen4_exp
  • moe
  • pruning
  • quantization
  • mixture-of-experts

Technical specifications

Parameters118B
Librarymlx
LanguagesNot available
Base modelmlx-community/Qwen3.8-Flash-Next-4bit
Licenseother