自定义模型无效

[jli113]

ERROR 01-09 08:22:48 [serving_chat.py:179] Error with model error=ErrorInfo(message='The model gpt-4o does not exist.', type='NotFoundError', param=None, code=404)

[pzp5700]

We recommend that you confirm whether the model name in the model lists provided by your LLM server service provider is gpt-4o.

[jli113]

I only deployed one model, and it is not gpt-4o.

[pzp5700]

In the execution scripts under the script directory, you can configure your API URL, API key, and model name. Please ensure that these configurations are consistent with those of your LLM server.

BTW, gpt-4o is the default model.

[jli113]

So when deployed through Docker Compose, parameters from web-ui by ./gradio_app/app.py are not used?