AnyangAngus

### Describe your problem I can see that deepdoc use many model from https://huggingface.co/InfiniFlow/deepdoc/tree/main so which base model does deepdoc select? det.onnx, rec.onnx from paddle? layout.onnx, tsr.onnx base on which...

[Feature] Minicpm-2.6 support

2

### Motivation Is there any plan to support Minicpmv 2.6 update? ### Related resources https://github.com/OpenBMB/MiniCPM-V?tab=readme-ov-file#minicpm-v-26 https://github.com/vllm-project/vllm/blob/9118217f58c39040aa935b7c85250c7364ffa72d/vllm/model_executor/models/minicpmv.py#L913C15-L913C26 ### Additional context Appreciate

Support bm25 SPARSE_INVERTED_INDEX and hybird search benchmark?

4

We use milvus dense and sparse vector and hybird search: https://milvus.io/docs/zh/multi-vector-search.md Dose this benchmark repo support hybird search? How can i config the test?

[Question] Different performance of vLLM server and Triton Inference Server

1

I deployed the Qwen-7B-VL model using both vLLM serve and Triton Inference Server VLLM backend. I used the same model, GPU resources, dataset, and stress test parameters, but got different...