The concept of open source doesn't really apply to AI models since their behavior is mostly controlled by the data they were trained on and the complex ways they are trained. Having the source code of the model by itself wouldn't help you.
From a practical POV having all the training data, training infrastructure, and training know-how wouldn't help you either unless you could afford to spend the millions of dollars (hundreds of millions for a SOTA model) in compute to train it each time they released a new training set, in which case you're only talking about the big commercial companies. "open source for the people" just does not apply.
Publishing RL/SFT/self-distillation harnesses would be very impactful even without the data.
Particularly when it comes to tool use w/ self-distillation it can be done without any data... have a tool the model doesn't know? a teacher model RTFMs and the source code, and helps the student learn to get it right.
Thank Apple?
Those are mostly for embedded devices and the current "sponsor" is Apple.
Aren't they only open weights, not true open source?
The concept of open source doesn't really apply to AI models since their behavior is mostly controlled by the data they were trained on and the complex ways they are trained. Having the source code of the model by itself wouldn't help you.
From a practical POV having all the training data, training infrastructure, and training know-how wouldn't help you either unless you could afford to spend the millions of dollars (hundreds of millions for a SOTA model) in compute to train it each time they released a new training set, in which case you're only talking about the big commercial companies. "open source for the people" just does not apply.
If (and that is a big if) the concept of open source doesn't apply, then the term shouldn't be coopted to mean something else though.
But even if I can't build it from source locally, being able to see what went into the model is an important part of what open source is about.
1 reply →
Publishing RL/SFT/self-distillation harnesses would be very impactful even without the data.
Particularly when it comes to tool use w/ self-distillation it can be done without any data... have a tool the model doesn't know? a teacher model RTFMs and the source code, and helps the student learn to get it right.