Hacker News Reader

This seems not be a repo ready to open source. You only get weights, very less information about how the weights is trained and finetuned.

But anyway, it always great to see more LLM weigts available.

View on HN

I would argue that there's no bar for open sourcing aside from "do you have the rights to do so." Some source or some public good is certainly better than none, and when the bar is low then you remove barriers to getting started, vs waiting until you have the time someday to "do it right."

Well what constitutes an "open source" model is still controversial and debatable-- lots of people on both sides of that argument.

Open source has had a useful agreed upon meaning for over 25 years. Maybe you're too young to understand why that matters but we're not.

I've been in the open source community for about 25 years so I doubt it.

For what it's worth I would say a model should be fully reproducible to be open source, but that's not a decided consensus -- and AI models are sufficiently different than the source code / binary code distinction as to invoke discussion around defining it.

Reply on news.ycombinator.com