“They said it could not be done”. We’re releasing Pleias 1.0, the first suite of models trained on open data (either permissibly licensed or uncopyrighted): Pleias-3b, Pleias-1b and Pleias-350m, all based on the two trillion tokens set from Common Corpus.

What are some solid pieces of writing in the vein of situational awareness and machines of loving grace?

It’s time to swap my terminal from simple dark/grayscale to colorful again. The question is, which color scheme?

Good morning everyone what’s the plan for today

Building wheel for flash-attn (setup.py) … /

scraping is hitting a wall

It seems to be real

I’ve been convinced that this is probably the type of method for something like this, sharing for all who haven’t come across it yet arxiv.org/abs/2411.04282

What’s everyone’s favorite papers of the week?

What’s everyone’s favorite sub-10B model at the moment? I’ve still been using deepseek lite on my MacBook mostly

So first version of an ml anon starter pack. go.bsky.app/VgWL5L Kept half-anons (like me and Vic). Not all anime pfp, but generally drawn.

Wow, got a ton of followers out of nowhere, hello

In my opinion you should be able to recreate this graph if you claim to have released a model trained like o1

Anyone know what the best AI research discord communities are?

@bsky.app Please add bookmarks. I desperately need bookmarks.

I’ll post my first blog post as my first post on this app. Hope you enjoy the read: stochasm.blog/posts/scalin...