[Deepspeed-chat] Can you release model weights for different stages to HF-hub?

[kouroshHakha]

Hello, I have successfully ran through the three stages. But I had to make some cuts on the batch size / lora training. I don't have a good baseline on what the performance of each step should be before moving on to the next stage. For example I would like to know the accuracy of the reward model that you put on your readmes? Mine gets to like ~%69, but it is not consistent. is that good enough for the next stage? I would like to run the evaluation scripts against the pre-verified models if possible.

[yaozhewei]

Hi, we plan to upload training log into our repo. Will keep you posted

[yaozhewei]

@kouroshHakha please check the repo for the training logs and let us know if it resolves your concern

[yaozhewei]

Closed as no followup