Mengbo-Zhou
Mengbo-Zhou
> for pre-training from scratch you'd need the pretraining data to be specified and available, which it isn't; therefore the code is by definition not complete. Yes, you are right....
Is it supported now?
@snexus We want to use OpenAI's embedding model for testing. How can we modify the source code to achieve this?
@MiaoLu3 Hi, I have the same problem on 8*H100. Have you solved it? Thanks a lot.