[Feature] Hope to support the embedding and reranker interfaces of the text embedding inference framework

[linfengjjj]

MaxKB Version

v1.9.1

Please describe your needs or suggestions for improvements

希望模型管理可以支持对接text-embeddings-inference推理框架的embeding和reranker接口
1、框架地址：https://github.com/huggingface/text-embeddings-inference
2、embeding和reranker接口示例

Please describe the solution you suggest

No response

Additional Information

No response

[liuruibin]

You can add it to the system after deploying vllm or xinference

[linfengjjj]

Thanks for your reply.
Compared to the vllm and xinference, it focuses more on text embedding task optimization, deeply optimized the inference process of the embedded model, supporting high concurrency and low latency request processing, and more mature hardware acceleration for embedded tasks.
This is the reason why the potential customer are so demanding, hope you can consider this requirement, sincerely.