Your browser does not seem to support JavaScript. As a result, your viewing experience will be diminished, and you have been placed in read-only mode.
Please download a browser that supports JavaScript, or enable it if it's disabled (i.e. NoScript).
https://www.reddit.com/r/LocalLLaMA/comments/1sw5fb7/qwen36_35b_a3b_heretic_kld_00015_incredible_model/
這個技術可以把整個qwen 3.6 35bA3B Q8量化+256k上下文塞進去24g vram～有點想跑看看
