mirror of
https://github.com/BerriAI/litellm.git
synced 2026-11-07 07:58:09 +00:00
The /chat request errors and deep research notice, and Aria's step limit
and no answer replies, are written into assistant messages that are sent
to the model as history on the next turn. Pin them to English with
i18n.getFixedT("en") so a future UI language never changes the model
payload. English output is unchanged.
|
||
|---|---|---|
| .. | ||
| docker-entrypoint.d | ||
| litellm-dashboard | ||
| Dockerfile | ||
| nginx.conf | ||