This message was deleted.
# ask-for-helps
Slackbot
01/25/2023, 9:56 AMThis message was deleted.
c
Chaoyu
01/26/2023, 8:58 AMHi Shaked, you should be able to do validation on the input dataframe in the service api definition code, and it’s recommended to validate input before passing it to the runner
Chaoyu
01/26/2023, 9:00 AMI’m not sure I understood your first question. The actual scoring happens inside the runner process. BentoML passes the entire dataframe to the runner. And when adaptive batching is enabled, it may group multiple data frames into one and pass it to the model for scoring
2 Views