LiteLLM model list not updating #3613
Comments
|
Roo Code is designed to automatically fetch and list available models from your LiteLLM server, eliminating the need for manual configuration [1][2]. However, if the model list is not updating as expected, it might be due to a caching mechanism or a bug. Unfortunately, the documentation does not specify any caching mechanisms or error handling related to LiteLLM model list updates [1]. To address this issue, you might want to check the Roo Code issue tracker for any reported bugs related to LiteLLM model list updates. If the problem persists, consider submitting a bug report to help the maintainers investigate further. To continue talking to Dosu, mention @dosu. Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other |
|
In my case there is just no model. With OpenAI compatibile provider, litellm works well. |
|
In my case, With OpenAI compatibile provider don't fetch models at all. It used to work. |
|
@slytechnical is working on this |
|
ohhh ok cool thanks @mrubens |
App Version
3.16.6
API Provider
Not Applicable / Other
Model Used
N/A
🔁 Steps to Reproduce
Using LiteLLM as a local LLM proxy, Roo does not query LiteLLM to get updated model lists on VSCode start or when adding a new LiteLLM profile in Roo. Roo seems to query LiteLLM for a model list once and persist that information permanently.
💥 Outcome Summary (Optional)
No response
📄 Relevant Logs or Errors
The text was updated successfully, but these errors were encountered: