logoalt Hacker News

cronin101 • today at 4:22 PM • 4 replies • view on HN

I'm far behind the curve with local/open weight models, but I thought it was common practice that you should do some last-mile fine-tuning on the particular harness that you're using and would get a reasonable return on investment for doing so (particularly on quantized models)?

Can anyone confirm/deny this is the case?

If so, it seems a shame not to include this practice into the reporting methodology for full transparency.


Replies

bisonbear • today at 5:16 PM

In MiMo's v2.6 public RL training run, they explicitly trained on 23 different harnesses in order to minimize overfitting on any single harness. I imagine that this technique has to be common in RL training runs given that model providers have little control over where their models are used.

https://mimo.xiaomi.com/rl/#overview

hedgehog • today at 5:07 PM

It's typically easier to tune the harness around the models' existing behavior. In my experience the primary failures in models > 5B params on coding tasks are just that they aren't able to translate instructions into a correct description or implementation of the problem (will write wrong code, wrong tests, and wrong documentation). The basic tool calling and syntax of generated code is fine.

ford • today at 4:32 PM

I don't think this is common, though it would be interesting if ThinkingMachines/HuggingFace/OpenRouter/BaseTen offered fine-tuned variants of the most popular open models that worked with specific harnesses

➕ show 1 reply
jrmg • today at 4:26 PM

What does "last-mile fine-tuning on [a] harness" mean?

➕ show 1 reply