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.
Organizations
models 54
bearzi/GLM-5.2-JANGTQ
Text Generation • 64B • Updated • 196 • 3
bearzi/GLM-5.2-JANGTQ4
Text Generation • 109B • Updated • 122 • 1
bearzi/GLM-5.2-JANGTQ_K
Text Generation • 79B • Updated • 194 • 3
bearzi/Kimi-K2.7-Code-JANGTQ_K
Text Generation • 97B • Updated • 213 • 1
bearzi/Kimi-K2.7-Code-JANGTQ1
Image-Text-to-Text • 68B • Updated • 17
bearzi/Kimi-K2.7-Code-JANGTQ2
Image-Text-to-Text • 76B • Updated • 13
bearzi/MiMo-V2.5-MLX
Updated • 104
bearzi/MiMo-V2-Flash-JANG_2S
Text Generation • 474B • Updated • 40
bearzi/MiMo-V2-Flash-JANG_2M
Text Generation • 480B • Updated • 36
bearzi/MiMo-V2-Flash-JANG_2L
Text Generation • 481B • Updated • 37
datasets 0
None public yet