ThakiCloud/Muse-Glimmer-30B-NVFP4-GPTQ-mm
Image-Text-to-Text • 19B • Updated • 60 • 1
Quantization of multimodal (image + text) checkpoints, calibrated to keep the vision tower intact.
Note Muse-Glimmer 30B NVFP4
Note Qwen3.8-27B NVFP4 — multimodal calibration