Questions about future plans about the reward model #2017

Dear team,

I've read the guides/developers and reward/instructor , I have some questions regarding the training process of the reward model, and I hope you can help me with it.

1. reward/instructor#dataset mentioned that Once open-asisstant dataset are available it will be added here.. Has the task of supporting OPEN-ASSISTANT's own dataset been scheduled? How long will it take to complete it?
2. Is the process of training the reward model currently done manually? Will it be integrated into the data loop for automatic execution in the future?( The idea here is that the reward model will automatically perform an initial ranking of the data, and then user feedback will be used as further training data for it.)

Thank you for your time and assistance.