Ming Fang

This PR does not fix the CSRF problem. This issue, https://github.com/oauth2-proxy/oauth2-proxy/issues/1937, is just one of many issues about the CSRF problem. My current workaround is to downgrade to v6.1.1.

> i guess that ideally we can use eg vector search (but maybe also bm25 via pg_search) to be able to form graphs connecting similar entities (instead of exact matching...

First I have to create a function to do the `agtype` conversion ```sql CREATE OR REPLACE FUNCTION public.movies_view_bm25_search(search agtype, limit_rows agtype default '10'::agtype) RETURNS SETOF agtype AS $$ SELECT id::agtype...

+1 I just hit this bug also. Here is a VERY ugly hack that I used to put a line on a 80% mark when displaying disk usage. ``` ```

That's great. How can we join that workspace?

In the openai_api_server, stop works for non-streaming completions, but not for streaming. The problem is the unwanted stop sequence gets streamed out before stopping. https://github.com/lm-sys/FastChat/blob/main/fastchat/serve/openai_api_server.py#L518 As a result, this breaks...

I tested the PR locally and it has the same problem. @merrymercy do you think this problem should be fixed in https://github.com/andy-yang-1/FastChat/blob/langchain-support/fastchat/serve/inference.py#L51 so that it doesn't emit the stop sequence?

The problem happens when the previous generate token is the partial beginning of the stop sequence. It will not match the entire stop sequence until the next few tokens. As...

@merrymercy This is my PR with the stop detection fix https://github.com/lm-sys/FastChat/pull/1392

I have the same setup and am seeing the same error