Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

OpenSource strategy #50

Open
sebastianlutter opened this issue Mar 13, 2024 · 3 comments
Open

OpenSource strategy #50

sebastianlutter opened this issue Mar 13, 2024 · 3 comments

Comments

@sebastianlutter
Copy link

sebastianlutter commented Mar 13, 2024

AlephAlpha has repeatedly emphasised in various publications and statements that it is committed to OpenSource. Looking at the activity in this repository and the available trained weights, the reality is different.

Apart from positive marketing, what role does OpenSource play at AlephAlpha? The French company Mistral has shown how important the OpenSource idea is, and has set international standards with its released models.

All I see are academic advances that are used as arguments for further funding rounds. The model code in this repository is two years out of date, as are the available checkpoints. There are no publications on Huggingface nor any other engagement in the community of LLM's tools. Once again it seems to be a special German way of compartmentalisation, combined with the belief that enough money will take care of it.

So what is AlpehAlpha's actual strategy regarding OpenSource?

@bm777
Copy link

bm777 commented Mar 27, 2024

They will not

@stefan-it
Copy link

stefan-it commented Aug 26, 2024

After two years, signs of life w.r.t. open weights:

  • Aleph-Alpha/Pharia-1-LLM-7B-control see here
  • Aleph-Alpha/Pharia-1-LLM-7B-control-aligned see here

@bm777
Copy link

bm777 commented Aug 26, 2024

@stefan-it do you work for Aleph-Alpha? 😆

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants