Closed raspawar closed 2 months ago
New implementation:
This can be simplified by using the first available model if none is provided, e.g.
llm = ChatNVIDIA(base_url="http://localhost:1234/v1") llm.chat(...)
llm._client.model => default model => first available model in local NIM
Fixes:
New implementation:
This can be simplified by using the first available model if none is provided, e.g.
llm = ChatNVIDIA(base_url="http://localhost:1234/v1") llm.chat(...)
llm._client.model => default model => first available model in local NIM
Fixes: