Yohann Bearzi
bearzi
AI & ML interests
Uploading mixed-precision MLX quantizations of popular open-weight models, produced with oMLX’s oQ and JANG on apple silicon. Benchmarks and comparisons are welcome.
Recent Activity
updated a model about 8 hours ago
bearzi/Kimi-K2.7-Code-JANGTQ_K updated a collection about 8 hours ago
Kimi-K2.7-Code-JANGTQ updated a collection about 8 hours ago
Kimi-K2.7-Code-JANGTQOrganizations
MiniMax-M2.7-JANG
gemma-4-26B-A4B-it-oQ
Qwen3.6-27B-JANG
Qwen3.6-35B-A3B-JANG
Gemma-E2B-it-oQ
Qwen3-Coder-Next-oQ