Closed rwood-97 closed 5 months ago
This works with running the container!
Annoyingly how it is working is that if ANY request fails then it just skips that part of loading data. So whats currently running has skipped The Turing Way and Hut23 repo (we should see if this is always the two causing issues) but has all the info from RDS courses, wikis, etc etc.
Ideally we should add a boolean flag in llama-index
to be something like fail_on_http_error
and then we can force it to continue even if it fails with one file rather than raise exception. I will make a ticket on llama index about this and mb try address it next friday in OS hack session.
This PR adds try/excepts for the llama index github readers in order to fix the error we are getting loading our data in Azure.
Fixes #157