I mean just use them and compare, the gap is obvious.

otabdeveloper4 · 2026-03-20T05:36:33 1773984993

I did, and I fixed Qwen's issues with trivial sampling and loop detection hacks.

If I can do this, then a company that wants to sell local models seriously could do it too.

ninjagoo · 2026-03-20T13:02:56 1774011776

> I did, and I fixed Qwen's issues with trivial sampling and loop detection hacks.

Wow, that's amazing! Care to share the changes? Would love to try them out.

otabdeveloper4 · 2026-03-20T16:27:33 1774024053

It's not amazing at all.

What's amazing is that LLM technologies are so immature that even basic engineering diligence isn't being done. (Like detecting token loops, for example.)