[Bug] dataprep/retriever is not working as expected in ChatQnA with Qdrant #1473
Description
Priority
P2 - High
OS type
Ubuntu
Hardware type
Xeon-GNR
Installation method
- Pull docker images from hub.docker.com
- Build docker images from source
- Other
Deploy method
- Docker
- Docker Compose
- Kubernetes Helm Charts
- Kubernetes GMC
- Other
Running nodes
Single Node
What's the version?
git commit ed16308
Description
When launch ChatQnA with Qdran vector DB, after successfully uploading the document, we can not retrieve the previously uploaded document. The return of the retriever is always containing 0 document, not as what we've expected.
The retrieval result is:
{"id":"6d08404f2c77900d78baeb58bbb97d6c","retrieved_docs":[],"initial_query":"What is the revenue of Nike in 2023?","top_n":1,"metadata":[]}
Expected result should be something valid docs in the field retrieved_docs.
Checking the qdrant-vector-db log, finding that dataprep and retriever are not using the same collection in Qdrant.
Reproduce steps
cd GenAIExamples/ChatQnA/docker_compose/intel/cpu/xeon
source ./set_env.sh
docker compose -f compose_qdrant.yaml up
upload the document
curl http://localhost:6043/v1/dataprep/ingest -X POST -H "Content-Type: multipart/form-data" -F "files=@./nike.txt"
retrieve the document
export your_embedding=$(python3 -c "import random; embedding = [random.uniform(-1, 1) for _ in range(768)]; print(embedding)")
curl http://localhost:6045/v1/retrieval -X POST -H 'Content-Type: application/json' -d "{"text":"What is the revenue of Nike in 2023?","embedding":${your_embedding}}"
Raw log
docker compose -f compose_qdrant.yaml logs qdrant-vector-db
qdrant-vector-db | 2025-01-26T08:27:08.808448Z INFO actix_server::builder: Starting 151 workers
qdrant-vector-db | 2025-01-26T08:27:08.808458Z INFO actix_server::server: Actix runtime found; starting in Actix runtime
qdrant-vector-db | 2025-01-26T08:27:35.611864Z INFO actix_web::middleware::logger: 172.20.0.5 "GET /collections/rag-qdrant HTTP/1.1" 404 92 "-
" "qdrant-client/1.13.2 python/3.11.11" 0.000409
qdrant-vector-db | 2025-01-26T08:27:35.612829Z INFO storage::content_manager::toc::collection_meta_ops: Deleting collection rag-qdrant
qdrant-vector-db | 2025-01-26T08:27:35.612918Z INFO actix_web::middleware::logger: 172.20.0.5 "DELETE /collections/rag-qdrant HTTP/1.1" 200 58
"-" "qdrant-client/1.13.2 python/3.11.11" 0.000167
qdrant-vector-db | 2025-01-26T08:27:35.614417Z INFO storage::content_manager::toc::collection_meta_ops: Creating collection rag-qdrant
qdrant-vector-db | 2025-01-26T08:27:35.637944Z INFO actix_web::middleware::logger: 172.20.0.5 "PUT /collections/rag-qdrant HTTP/1.1" 200 57 "-
" "qdrant-client/1.13.2 python/3.11.11" 0.023616
qdrant-vector-db | 2025-01-26T08:27:35.652448Z INFO actix_web::middleware::logger: 172.20.0.5 "PUT /collections/rag-qdrant/points?wait=true HT
TP/1.1" 200 79 "-" "qdrant-client/1.13.2 python/3.11.11" 0.000784
qdrant-vector-db | 2025-01-26T08:28:03.911038Z INFO actix_web::middleware::logger: 172.20.0.4 "GET /collections/rag_qdrant/exists HTTP/1.1" 200 66 "-" "qdrant-client/1.13.2 python/3.11.11" 0.000387
qdrant-vector-db | 2025-01-26T08:28:03.913743Z INFO actix_web::middleware::logger: 172.20.0.4 "GET /collections/rag_qdrant/exists HTTP/1.1" 200 66 "-" "qdrant-client/1.13.2 python/3.11.11" 0.000072
qdrant-vector-db | 2025-01-26T08:28:03.914621Z INFO storage::content_manager::toc::collection_meta_ops: Creating collection rag_qdrant
qdrant-vector-db | 2025-01-26T08:28:03.938203Z INFO actix_web::middleware::logger: 172.20.0.4 "PUT /collections/rag_qdrant HTTP/1.1" 200 57 "-" "qdrant-client/1.13.2 python/3.11.11" 0.023654
qdrant-vector-db | 2025-01-26T08:28:03.945781Z INFO actix_web::middleware::logger: 172.20.0.4 "POST /collections/rag_qdrant/points/query HTTP/1.1" 200 66 "-" "qdrant-client/1.13.2 python/3.11.11" 0.00130Attachments
No response