Trying out the new assistants features in the playground, here’s some issues ive been facing
Few bugs:
1.) Cannot delete files in the Playground
2.) Cannot upload .json, .db, .csv files
Trying out the new assistants features in the playground, here’s some issues ive been facing
Few bugs:
1.) Cannot delete files in the Playground
2.) Cannot upload .json, .db, .csv files
I just get this message whenever I try to do anything with the assistant. Is anyone else getting this? " Run failed - We’re currently processing too many requests - please try again later."
I haven’t seen anyone else mention getting this error, and I don’t know how people are testing stuff seeing as this is all I get when I try it in the playground.
Hit and miss for me
Also can’t delete files - it says file not found
Getting mixed errors, sometimes can’t access the file, sometimes works after a few tries, sometimes says run failed
Created two assistants. One is a 750 page labor agreement. The other was 12,000 pages of Biblical scripture. As you can imagine, the assistant works well with the labor agreement. After 5 minutes, it still can’t answer a question on the scriptures. Haven’t tested much else, but so far, so good.
I am facing the same situation. It says “There was an error in deleting the file: ‘No such File object: file-xxx’”
Same with me. I keep getting this error: Run failed - We’re currently processing too many requests - please try again later."
“We’re currently processing too many requests - please try again later”. I am in Australia, so I am concerned the model doesn’t work upside down, or in metric?
When it does work - wow!
It looks like files are not indexed and then not accessed. They’re probably experiencing a huge amount of requests.
same here. any clues? i just uploaded a half page pdf and said “hi”
but keep getting “Run failed - We’re currently processing too many requests - please try again later.” Am i missing something?
Guys, just imagine the amount of requests they’re getting.
Give them some time, they literally launched less than 24 hours ago (and the functionality is, clearly stated, in BETA)
+1 The same.
Tier-1, current usage is 82 requests a day and 164k tokens, around 1 message per 5 minutes. Prepaid credits.
I can’t even add a function since it keeps telling me the JSON is incorrect - even though it’s perfectly fine and I verified with the JSON parser
Yes, me too. Been getting this message last few hours.
Same here, I have removed the files from the assistant, but it keeps showing in the file section. When tried to delete there it says: There was an error deleting the file: No such File object
I haven’t been able to get a large file (within the file token limit) to open in the assistant. The file is there, but if I get this message when the thread tries to access it:
There was an issue fetching items: An error occurred. If this issue persists please contact us through our help center at help.openai.com.
Me too, anytime my message is even slightly complex I get:
" Run failed We’re currently processing too many requests - please try again later."
As a former programmer who has been to many dev conferences, I see the “dog and pony” show is still thriving … as in, you get excited, open wallet, then get home to find that nothing works like it did on “stage” and I suppose that AI and Houdini have a lot in common. Of course, I really do want AI to succeed because it has so much promise, but failures like this are not a good sign going forward … hoping for better.
You can also not upload .zip
files, I made a separate thread for asking about that but ig you could add that to the list as well. (edit I managed to get it working by disabling Retrieval
which was somehow still enabled after submitting with it off anyways)
when i upload pdfs for retrieval it gives me an error when i try to save the assistant.
Be mindful of the restrictions on content length as they seem to apply to uploads.
It gives an error if your files are above the 2,000,000 token limit, but it seems to stall and fail even on this 3mb file, when it tries to access it from the thread.