You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There is some limit of how big a request you can send to Airtable / how much data you can put in a field. If an application is very long, and then the AI evaluator copies it 5 times to evaluate it for 5 different criteria, and then tries to put it in the one notes field (with the (optional) Logs field thing), the API rejects it.
It would be nice to either:
store the logs without copying the application so much
if it will be too long, truncating it before putting it in the field / not setting it at all? (this seems better than just silently failing - especially because the errors are not obvious, see Better surface errors to users #4. and maybe better than noisily failing, because then you need to re-run the evaluator)
The workaround for now is to run it on everyone with the logs field set, and this will work for everyone apart from the long applications. Then for the long applications we re-run evaluations with the logs field unset. But this adds extra effort and means we have no logs for these people.
The text was updated successfully, but these errors were encountered:
There is some limit of how big a request you can send to Airtable / how much data you can put in a field. If an application is very long, and then the AI evaluator copies it 5 times to evaluate it for 5 different criteria, and then tries to put it in the one notes field (with the
(optional) Logs field
thing), the API rejects it.It would be nice to either:
The workaround for now is to run it on everyone with the logs field set, and this will work for everyone apart from the long applications. Then for the long applications we re-run evaluations with the logs field unset. But this adds extra effort and means we have no logs for these people.
The text was updated successfully, but these errors were encountered: