有人能告诉一下安装的本地的huggingface rerank大模型，怎么在模型供应商里添加不？

[RexLyu]

如题。使用docker的TEI，部署了BAAI/bge-reranker-v2-m3，怎么都配置不上去。谁能给个配置说明不？

[dosubot[bot]]

Hello @RexLyu! I'm Dosu, and I'm here to help you with any bugs, questions, or guidance you might need while waiting for a human maintainer. Let's tackle this together!

要在RAGFlow中添加本地的Hugging Face模型，您需要将RAGFlow连接到本地部署的Hugging Face模型，该模型通过FastAPI运行，并使用Docker主机IP地址'172.17.0.1'而不是localhost。完整的端点URL应为'http://172.17.0.1:8970'。此外，您可以使用`rag/llm/embedding_model.py`文件中的`HuggingFaceEmbed`类来处理嵌入过程。此类需要模型名称和基础URL，如果未提供，默认值为'http://127.0.0.1:8080'。您可以修改基础URL以匹配您的本地设置^[1][2]。

对于BAAI/bge-reranker-v2-m3的配置，您需要在customer_service.json文件中进行如下配置：

{
  "id": "retrieval:0",
  "type": "ragNode",
  "position": {
    "x": 0,
    "y": 0
  },
  "data": {
    "label": "Retrieval",
    "name": "ShinyPathsDraw",
    "form": {
      "similarity_threshold": 0.2,
      "keywords_similarity_weight": 0.3,
      "top_n": 6,
      "top_k": 1024,
      "rerank_id": "BAAI/bge-reranker-v2-m3",
      "kb_ids": ["869a236818b811ef91dffa163e197198"]
    }
  },
  "sourcePosition": "left",
  "targetPosition": "right"
}

此配置包括similarity_threshold、keywords_similarity_weight、top_n、top_k和rerank_id设置为"BAAI/bge-reranker-v2-m3"等设置^[3]。

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}