ERR: Unable to process query

AI & Data Science NVIDIA AI Workbench
1

When i Submit a promt to NVIDIA endpoint it returns:
*** ERR: Unable to process query. ***
Check Output > Chat on the AI Workbench application for full logs.

Submission Type
Bug or Error

Workbench Version
Desktop App v0.44.8

Host Machine operating system and location
Local Windows 11

Logs:

{“level”:“info”,“time”:“2024/05/21 - 09:25:57”,“status”:200,“latency”:“13.692µs”,“client-ip”:“127.0.0.1”,“method”:“OPTIONS”,“path”:“/v1/query”,“time”:“2024-05-21T09:25:57+10:00”,“message”:“GIN-Request”}
{“level”:“info”,“time”:“2024-05-21T09:25:57+10:00”,“message”:“Processing git status output”}
{“level”:“info”,“time”:“2024-05-21T09:25:57+10:00”,“message”:“Processing git diff output”}
{“level”:“info”,“time”:“2024-05-21T09:25:57+10:00”,“message”:“Processing git diff output”}
{“level”:“info”,“time”:“2024-05-21T09:25:57+10:00”,“message”:“Processing git status output”}
{“level”:“info”,“time”:“2024-05-21T09:25:57+10:00”,“message”:“Processing git diff output”}
{“level”:“info”,“time”:“2024-05-21T09:25:57+10:00”,“message”:“Processing git diff output”}
{“level”:“info”,“time”:“2024/05/21 - 09:25:58”,“status”:200,“latency”:“1.174839448s”,“client-ip”:“127.0.0.1”,“method”:“POST”,“path”:“/v1/query”,“time”:“2024-05-21T09:25:58+10:00”,“message”:“GIN-Request”}

2

hi alex

i’m assuming you are using the hybrid-rag project from github.

i’m also assuming that you selected the cloud option for inference.

you can see the logs for the chat app in the workbench UI using the output widget at the bottom right.

you will need to select chat to see those logs.

can you report back what they say?

3

Please see my response here for instructions on how to access the logs for the Chat application. You will need to select the right logs from the dropdown.

The most common issue we are seeing on the cloud endpoint inference mode is an improperly configured API run key. If you are using 0.44.8, there is a known issue with configuring the key with the auto-prompt when you click Open Chat where the key you paste in that field can get truncated erroneously.

I recommend either upgrading to 0.50.16, and/or going into Environment > Secrets and reconfiguring your Run key. Your full key should get pasted properly in this method. Click the reveal characters button to ensure your full key gets pasted properly before confirming.