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
When generating a testset, it takes around 45-60 minutes to go thru a doc and produce the topics.
Then it starts generating the testset but... it's failing around 80% because of JSONDecode error and I have to restart again, including all the topic generation part.We should implement a solution to save the progress, since it is very likely to have error like this when generating testset from very long docs.
Standalone code OR list down the steps to reproduce the issue
TODO
Relevant log output
No response
The text was updated successfully, but these errors were encountered:
Thanks for registering this issue. I've seen this behavior happening in a lot of projects when creating large datasets, assuming that the main process won't fail.
Besides catching the error and ignoring those to have the process finish, I think it could be very useful to append the created results to a local file, so that if the process fails somehow, you have some output as a result plus it can be easier to pinpoint the problem in the input data. Otherwise, it just stays in memory and when it fails... it's gone.
Issue Type
Bug
Source
source
Giskard Library Version
2.15.0
OS Platform and Distribution
Macos
Python version
3.12.4
Installed python packages
No response
Current Behaviour?
Standalone code OR list down the steps to reproduce the issue
Relevant log output
No response
The text was updated successfully, but these errors were encountered: