News for hackers. Stuff that matters.

A comment on Lucidrains/self-rewarding-lm-PyTorch: Self-Rewarding Language Model, from MetaAI.

Re:Lucidrains/self-rewarding-lm-PyTorch: Self-Rewarding Lang... by frogamel on Wednesday January 24, 2024 @11:14PM (#39126128)
I might be miunderstanding something here, but what complexity here is resolved by making this a framework? Isnt this just:

1. Train model like normal

2. Evaluate model using self

3. Use eval results for DPO finetune

[ Parent | on HN ]
Re:Lucidrains/self-rewarding-lm-PyTorch: Self-Rewarding Lang... by lucidrains on Wednesday January 24, 2024 @11:21PM (#39126162)
No, you aren't wrong. For ML people, it is quite simple and hopefully the final code reflects that

The aim is really to give a good base for follow up research / modifications, which I think there will be many for this paper

[ Parent | on HN ]