lm-sys / lm-sys/FastChat

"Error! Something went wrong. Please refresh and try again. "

Open
#3,711 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
39.5k
Forks
4.8k
PR merge metrics
No merged PRs in 30d

Description

When I try to talk to any LLM, as soon as I press the Send button, which is the icon that looks like a small airplane, the system will almost always bring up the following error dialog:

Error
Something went wrong. Please refresh and try again.

But when I followed the prompts to refresh the page, it still went wrong. I've tried it again and again and it's still the same.
It is worth mentioning that I also tried to close the page and re-enter it from the main page, and it still looks like this.
Sometimes, however, it works just fine; Sometimes, though, it will break, but I only need to refresh it once, and even worse, I only need to refresh it 2-3 times at most to continue the normal conversation.
I really don't know what to do now.
Maybe it's because the site gets so much traffic?
Perhaps because a LLM is currently unavailable?
When I'm on the Arena(battle) page, can the system at runtime first determine the availability of a LLM and then pre-switch to the available one?
Or, can the site make the service highly available in some other way?

I hope it can get better, and I hope we can all have a better experience.
Thanks!

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the Arena (battle) page and the Send-button flow described in the report, then determine where the intermittent error is surfaced. The issue does not name files or tests; done would require identifying the failure cause and defining reliable behavior when an LLM is unavailable or the request fails.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
ai, api, backend
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.