Vllm rejecting requests that have empty tools list
Nobody has claimed this yet.
Assessment
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Newbie friendliness
- 76/100
Research direction
Start with lib/Service/OpenAiAPIService.php and inspect the tools handling around the condition reported in the issue. Reproduce by sending a message through Assistant to a vLLM model and check how requests with an empty tools list are formed. Done when the request succeeds without a 400 error and existing tool requests still work.
Written by the indexing model from the issue text.
Description
Which version of integration_openai are you using?
latest
Which version of Nextcloud are you using?
Latest
Which browser are you using? In case you are using the phone App, specify the Android or iOS version and device please.
No response
Describe the Bug
after the update from yesterday, i saw that messages sent through the assistant are all being rejected with 400 error with the error message being that the "tools" parameter being empty. i managed to fix it by changing a line inside the nextcloud-aio-nextcloud contianer in the file /var/www/html/custom_apps/integration_openai/lib/Service/OpenAiAPIService.php.
how it was:
if ($tools !== null) {
after change:
if (!empty($tools)) {
and now it works fine. so i suggest you guys take a look at this (andif i posted in the wrong repo, please tell me so so i can post in the correct one).
Expected Behavior
responds
To Reproduce
send a message through assistant to an Openai supoorted model running on Vllm
- Dominant language
- PHP
- Stars
- 76
- Forks
- 36
- Avg merge
- 2d 16h
- Merged PRs (30d)
- 3
Getting set up
- No Dockerfile or Docker Compose file
- No pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from nextcloud/integration_openai
-
parseStreamChatResponse ignores the [DONE] sentinel and hangs until read timeout on keep-alive serversPossibly taken @mikemikimike claimed this 4 days ago. Openbug
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
nextcloud/integration_openai#446 ·
-
enhancement priority: normal
Difficulty 3/5 1-2 days Newbie friendliness 56/100
nextcloud/integration_openai#443 ·
-
Allow to send extra headers in the requests to the external servicePossibly taken @julien-nc claimed this 39 days ago. Openenhancement
nextcloud/integration_openai#436 · 1 assignee ·
-
bug
Difficulty 3/5 1-2 days Newbie friendliness 68/100
nextcloud/integration_openai#426 · 6 comments ·
-
bug
Difficulty 3/5 1-2 days Newbie friendliness 48/100
nextcloud/integration_openai#407 · 4 comments ·
All issues in nextcloud/integration_openai
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 76/100
Maintainers usually reply within 1 day
-
[2.x] extension-manager: translation keys that changed meaning keep their old namesPossibly taken A pull request linked to this issue is open or already merged. Opentype/bug
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
Maintainers usually reply within 1 day
-
needs-triage Technical TODO
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
Maintainers usually reply within 2 days
-
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
MISP/misp-docker#478 ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
glpi-project/glpi#26045 ·
Maintainers usually reply within 1 day