Embeddings

Raw

• Sparse Vectors in Qdrant: Pure Vector-based Hybrid Search
• SPLADE for Sparse Vector Search Explained
• Hybrid Search: SPLADE (Sparse Encoder)

graph LR
    A--"matching<br/>BM25"-->C["1️⃣ sparse = <pre>[{331: 0.5}, {14136: 0.7}]</pre>"]
    A["Vector Search"]--"semantic meaning<br/>cosine similarity"-->B("2️⃣ dense(BERT) = <pre>[0.2, 0.3, 0.5, 0.7, ...]</pre>")

Build

docker build . --platform linux/amd64 -t huggingface/tei

Mac Silicon, from local

# Setup
cargo install --path router -F candle -F metal

# Run
model=intfloat/multilingual-e5-large
revision=main
volume=$PWD/data

text-embeddings-router --model-id $model --revision $revision --port 8081

# Embed
curl 127.0.0.1:8081/embed \
    -X POST \
    -d '{"inputs":"What is Deep Learning?"}' \
    -H 'Content-Type: application/json'

# Classifier model
model=BAAI/bge-large-en-v1.5
revision=refs/pr/5

# Rerank
curl 127.0.0.1:8081/rerank \
    -X POST \
    -d '{"query":"What is Deep Learning?", "texts": ["Deep Learning is not...", "Deep learning is..."]}' \
    -H 'Content-Type: application/json'

# Predict
curl 127.0.0.1:8081/predict \
    -X POST \
    -d '{"inputs":"I like you."}' \
    -H 'Content-Type: application/json'