I'd take a model whose training data is open source and legitimately obtained. The only ones I know about are Apertus and OLMo, and they aren't really competitive.
- Posts
- 0
- Comments
- 52
- Joined
- 3 yr. ago
- Posts
- 0
- Comments
- 52
- Joined
- 3 yr. ago
Interesting point... In this case, I have a hunch that the requirements in terms of training data and hardware to run the training on would remain prohibitive, but I do wonder how much could be actually achieved using a distributed scheme of sorts.