I tried to follow this tutorial but looks like the autogen cannot access to the model.
https://microsoft.github.io/autogen/docs/topics/non-openai-models/local-litellm-ollama/
I am struggling this problem. Could you help me to deal with this?
#{294} WARNING - Model ollama_chat/llama2 is not found.
jagged pelican
runic kite
Hi @jagged pelican, the warning you're getting there is just saying that the pricing can't be determined because AutoGen doesn't know how much it costs per token. You can ignore the costing notice.
jagged pelican
Hi @runic kite Oh, my bad. I missed one step, an attached image is correct one. I am entirely not sure but it said '{294} WARNING - Model ollama_chat/llama3 is not found.' I thought this means that the code cannot get the replay from ollama, right? Also I just followed and this doesn't have such warnings though... https://microsoft.github.io/autogen/docs/topics/non-openai-models/local-litellm-ollama/
runic kite
Ah, it does look like the model is working correctly... I think there's a way to pass in the costs, which will be zero.... Let me have a look at that.
Okay, found it. In section 3 of this notebook you can add the "price" key to your config, so you could add:
"price": [0]
and that should set the cost to zero and hopefully the warning won't show.
https://github.com/microsoft/autogen/blob/main/notebook/agentchat_cost_token_tracking.ipynb
static jackal
https://github.com/microsoft/autogen/pull/2902
in the next release everyone will be enabled to pass in the pricing param in the config_list. this should fix the pricing issue.
