Hy3-REAP-200B-21B
A 33%-expert-pruned version of Tencent Hunyuan Hy3 (295B / A21B), compressed with REAP (Router-weighted Expert Activation Pruning) —...
A 33%-expert-pruned version of Tencent Hunyuan Hy3 (295B / A21B), compressed with REAP (Router-weighted Expert Activation Pruning) —...
This repository contains a REAM-compressed version of Tencent Hy3, produced with Akicou/ream — a REAM/REAP-style Mixture-of-Experts compression framework....
Pre-converted weights for colibrì — a pure-C engine that runs huge MoE models on consumer hardware by keeping...
The high-memory Apple Silicon edition of SuperHY3, with the same fused OBLITERATUS and SuperTune behavior as the NVFP4...
Quantized tencent/Hy3 for Apple Silicon MLX / JANG runtimes — a 295B-total / 21B-active text MoE, packed to...
Version 26.05.01 Calibration STEM and Agentic Languages EN ZH HI AR RU JA KO NL FR ES Model...
INVALID LANGUAGE PAIR SPECIFIED. EXAMPLE: LANGPAIR=EN|IT USING 2 LETTER ISO OR RFC3066 LIKE ZH-CN. ALMOST ALL LANGUAGES SUPPORTED...
This model mlx-community/Hy3-preview-4bit was converted to MLX format from tencent/Hy3-preview using mlx-lm version 0.31.3. Модальности:Генерация текста Области применения:Диалог...
> Выпущено 16 июля 2026 г. Все этапы пройдены: проверка артефакта (—check-mtp), > качество по сравнению с FP8...
Экспертно-обрезанная версия tencent/Hy3, созданная с помощью REAP (экспертная обрезка активации с учетом маршрутизатора). 192 маршрутизируемых эксперта каждого уровня...