inference-optimization/Qwen3-30B-A3B-Instruct-2507-7-bits-mode-noise-per-tensor 26B • Updated 4 days ago • 17