🌟 Reranking Microservice with TEI

July 16, 2025 · View on GitHub

Text Embeddings Inference (TEI) is a comprehensive toolkit designed for efficient deployment and serving of open source text embeddings models. It enables us to host our own reranker endpoint seamlessly.

This README provides set-up instructions and comprehensive details regarding the reranking microservice via TEI.

📦 1. Start Microservice with Docker

🔹 1.1 Start Reranking Service with TEI

Start the TEI service:

For Gaudi HPU:

  export HF_TOKEN=${your_hf_api_token}
  export RERANK_MODEL_ID="BAAI/bge-reranker-base"
  export volume=$PWD/data

  docker run -p 12005:80 -v $volume:/data --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none -e MAX_WARMUP_SEQUENCE_LENGTH=512 --cap-add=sys_nice --ipc=host ghcr.io/huggingface/tei-gaudi:latest --model-id $RERANK_MODEL_ID --hf-api-token $HF_TOKEN --auto-truncate

For Xeon CPU:

  export HF_TOKEN=${your_hf_api_token}
  export RERANK_MODEL_ID="BAAI/bge-reranker-base"
  export volume=$PWD/data

  docker run -d -p 12005:80 -v $volume:/data -e http_proxy=$http_proxy -e https_proxy=$https_proxy --pull always ghcr.io/huggingface/text-embeddings-inference:cpu-1.6 --model-id $RERANK_MODEL_ID --hf-api-token $HF_TOKEN --auto-truncate

Verify the TEI Service:

Run the following command to check if the service is up and running.

  export host_ip=$(hostname -I | awk '{print \$1}')
  curl $host_ip:12005/rerank \
      -X POST \
      -d '{"query":"What is Deep Learning?", "texts": ["Deep Learning is not...", "Deep learning is..."]}' \
      -H 'Content-Type: application/json'

🔹 1.2 Build Docker Image and Run Docker with CLI

Build the Docker image for the reranking microservice:

 docker build --no-cache \
   -t opea/reranking:comps \
   --build-arg https_proxy=$https_proxy \
   --build-arg http_proxy=$http_proxy \
   --build-arg SERVICE=tei \
   -f comps/rerankings/src/Dockerfile .

Run the reranking microservice and connect it to the TEI service:

 export TEI_RERANKING_PORT=12005
 export HF_TOKEN=${your_hf_api_token}
 export host_ip=$(hostname -I | awk '{print \$1}')
 export TEI_RERANKING_ENDPOINT="http://${host_ip}:${TEI_RERANKING_PORT}"

docker run -d --name="reranking-tei-server" -e LOGFLAG=True  -p 8000:8000 --ipc=host -e http_proxy=$http_proxy -e https_proxy=$https_proxy -e TEI_RERANKING_ENDPOINT=$TEI_RERANKING_ENDPOINT -e HF_TOKEN=$HF_TOKEN  -e RERANK_COMPONENT_NAME="OPEA_TEI_RERANKING"  opea/reranking:comps

📦 2. Start Microservice with docker compose

Deploy both the TEI Reranking Service and the Reranking Microservice using Docker Compose.

🔹 Steps:

Set environment variables:

 export RERANK_MODEL_ID="BAAI/bge-reranker-base"
 export TEI_RERANKING_PORT=12005
 export RERANK_PORT=8000
 export host_ip=$(hostname -I | awk '{print \$1}')
 export TEI_RERANKING_ENDPOINT="http://${host_ip}:${TEI_RERANKING_PORT}"
 export TAG=comps

Navigate to the Docker Compose directory:

cd comps/rerankings/deployment/docker_compose/

Start the services:

For Gaudi HPU:

 docker compose up reranking-tei-gaudi -d

For Xeon CPU:
```
 docker compose up reranking-tei -d
```

📦 3. Consume Reranking Service

🔹 3.1 Check Service Status

Verify the reranking service is running:

  curl http://localhost:8000/v1/health_check \
  -X GET \
  -H 'Content-Type: application/json'

🔹 3.2 Use the Reranking Service API

Execute reranking process by providing query and documents

curl http://localhost:8000/v1/reranking \
  -X POST \
  -d '{"initial_query":"What is Deep Learning?", "retrieved_docs": [{"text":"Deep Learning is not..."}, {"text":"Deep learning is..."}]}' \
  -H 'Content-Type: application/json'

You can add the parameter top_n to specify the return number of the reranker model, default value is 1.

curl http://localhost:8000/v1/reranking \
  -X POST \
  -d '{"initial_query":"What is Deep Learning?", "retrieved_docs": [{"text":"Deep Learning is not..."}, {"text":"Deep learning is..."}], "top_n":2}' \
  -H 'Content-Type: application/json'

✨ Tips for Better Understanding:

Port Mapping: Ensure the ports are correctly mapped to avoid conflicts with other services.
Model Selection: Choose a model appropriate for your use case, like "BAAI/bge-reranker-base".
Environment Variables: Use http_proxy and https_proxy for proxy setup if necessary.
Data Volume: The -v ./data:/data flag ensures the data directory is correctly mounted.