[Feature] Requesting support for Multimodal

[wolfgangchen]

MaxKB Version

v1.10.2-lts

Please describe your needs or suggestions for improvements

Currently MaxKB only support stream model, which limits the business competition of MaxKB. Most high value scenarios needs handle text/pic mixed corpus.
Typically, qwen-omni-turbo from "百炼" is a multimodal, MaxKB cannot leverage its supper ability.

Please describe the solution you suggest

No response

Additional Information

No response

[baixin513]

感谢反馈，应该是该模型不支持非流式输出，添加模型时我们会进行一次对话来校验，该校验使用了非流式对话，所以校验失败了，后续版本考虑一下，添加模型时支持不验证。

[shaohuzhang1]

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

---

Thanks for the feedback, it should be that this model does not support non-streaming output. When adding the model, we will conduct a conversation to verify. This verification uses non-streaming conversation, so the verification failed. Consider it later versions. It supports no verification when adding the model.

[baixin513]

该问题在 V1.10.3-lts版本中解决。

[shaohuzhang1]

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

---

This issue is resolved in V1.10.3-lts version.