请问有Yi-VL可以实现few-shot（in-context）数据的推理或微调吗？

[waltonfuture]

Reminder

• [X] I have searched the Github Discussion and issues and have not found anything similar to this.

Motivation

希望用Yi-VL实现多模态few-shot的训练或推理

Solution

No response

Alternatives

No response

Anything Else?

No response

Are you willing to submit a PR?

• [ ] I'm willing to submit a PR!

[Yimi81]

目前不支持多图输入，多模态few-shot的推理应该是不可以，已记录反馈此需求