logoalt Hacker News

_fizz_buzz_today at 8:08 PM3 repliesview on HN

Can someone explain if i understand this correctly: Are they saying that they started training this new model on August 28th and then started using it on September 1st? Does training a new model only take 3 days?


Replies

tristanjtoday at 8:29 PM

OpenAI finished another pre-train in late August, and they are now building models off that base. He's saying the specific model OpenAI used to solve this problem is currently in post-training, which started on August 28.

gcrtoday at 8:22 PM

it's possible to do a RLHF or RLVR pass pretty quickly. I'm almost certain a full pretraining run isn't possible within that time frame.

lossolotoday at 8:57 PM

Not entirely, it's just a late stage of the overall training process. It's an early checkpoint in post training (you can use the model at different stages of training), so it will probably become even stronger with more post training.