When I return to JupyterLab, nothing has changed. Unfortunately, I’ve had this occur on both my machines. I’ve used JupyterLab under Anaconda Navigator for over a year previously with no such difficulties. The problem seems to be with the the most recent version of Anaconda.
GenAI indicates that this is a known problem, but I haven’t found any reference to this here or elsewhere (though I might have missed it). I’m hoping someone has a suggestion, other than using Anaconda desktop or using a GenAI assistant separately from within JupyterLab.
You may think that you have a daily limit of 30 questions for Anaconda Assistant but if you are using desktop Navigator then it is a lifetime limit and your problem is almost certainly that you are now locked out. The daily limit only applies to use of the web version which comes with privacy implications. They are basically looking at everything you do when you use the free web version.
It took me months to find the answer in a forum thread that was 10months old with hundreds of complaints before one of the users found the answer. I am afraid I seem to have lost the link to the thread but do have the text of the answer posted below.
"I was also getting the same error. The issue is indeed with a local installation of Anaconda Assistant which only comes with 30 free questions to the assistant. Once you use these it’s over for you unless you switch over to using the cloud version or upgrading to the starter paid version (which uses the cloud). I went ahead and purchased the starter version for $15/month, and once I paid I was able to use the assistant again.
Only downside is that the interface is slightly different than the local version I was using, and I no longer see an option to fire up Jupyterlab (sad) but the interface is not that much different.
Either way, I don’t think Anaconda v4.0.15 has any meaningful way to tell you your local usage of the assistant is used up at 30 uses, and the connection error is in fact b/c you need to switch over to the cloud version. It’s not really a connection error, so don’t wait for a fix; switch over to the cloud version.
If you read the documentation that has been linked by the staff, this will become clear
cheers,
Todd"
Until I found that I wasted a lot of time on responses telling me to make sure everything was up to date. Gemini helped me work through a lot of suggested solutions but did not actually find the answer. At the end it suggested I could use it alongside JupyterLab if all else failed and I found that useful.
