Yuge Zhang
Yuge Zhang
Thanks for reaching out! We might already have people working on this. Please hold on for a day or two to see whether they have worked out.
I think it makes sense, though I don't know for sure how `llm_as_judge` works.
I heard that the people from Tencent are implementing ROLL on top of Agent-lightning. I'm sharing the information here and maybe you guys can collaborate.
Normally you would test with "chat completion models", so the "test with text completion models" look very suspicious. Make sure you don't accidentally hit the non-chat mode when testing. I'll...
Also check out this doc: https://microsoft.github.io/poml/stable/python/integration/mcp/ I think you can use MCP in POML in a better way.
I'm sorry to hear that POML 0.0.8 stops to work for your models. Have you checked the chat model vs. text model. Are you sure you are using the chat...
@StefanBelo I'm not familiar with convo-lang, but I know other ts tools that have similar features. Just to clarify, POML only helps you write single prompts. We are not expanding...
I double checked my config. The following works: ``` "poml.languageModel.provider": "openai", "poml.languageModel.model": "gpt-4.1-mini", "poml.languageModel.apiUrl": "https://aihubmix.com/v1", "poml.languageModel.apiKey": "sk-xxxxxxx", ```