Cedar

Cedar provides an easy to setup, in-memory vector database that you can embed in your Rust application.

    // 1. initialize db
    let db = DuckDB::new(Default::default())?;
    db.init()?;

    // 2. initialize embedding function (could be OpenAI, Chrome, etc)
    let embedding_fn = SentenceTransformerEmbeddings::new();
    // Or use OpenAI embeddings:
    let embedding_fn = OpenAIEmbeddingFunction::new(
        "<api_key>".to_string(),
    );

    // 3. initialize client
    let mut client = LocalClient::init(db, embedding_fn)?;

    // 4. create a collection
    let mut collection = client.create_collection("collection1")?;

    // 5. push documents to the store
    let docs = &[
        Document {
            text: "this is about macbooks".to_string(),
            metadata: json!({ "source": "laptops" }),
            id: Uuid::new_v4(),
        },
        Document {
            text: "lychees are better than mangoes".to_string(),
            metadata: json!({ "source": "facts" }),
            id: Uuid::new_v4(),
        },
    ];
    collection.add_documents(docs)?;

    // 6. query the vector store for matching documents
    let k = 1;
    let res = collection.query_documents(&["which one is the better fruit?"], k, json!({ "source": "facts" }))?;

Installation

To use cedar in your project, start with adding it to your Cargo.toml. (Standalone cedar server coming soon!)

[dependencies]
cedar-db = "0.1.0"
uuid = { version = "1.3.2", features = ["v4", "fast-rng"] }
serde_json = "1.0.96"

cedar uses the tch-rs bindings for PyTorch. To set up the bindings, follow these steps:

1. Download libtorch from https://pytorch.org/get-started/locally/. This package requires v2.0.0: if this version is no longer available on the "get started" page, the file should be accessible by modifying the target link, for example https://download.pytorch.org/libtorch/cu118/libtorch-cxx11-abi-shared-with-deps-2.0.0%2Bcu118.zip for a Linux version with CUDA11. NOTE: When using rust-bert as dependency from crates.io, please check the required LIBTORCH on the published package readme as it may differ from the version documented here (applying to the current repository version).

2. Extract the library to a location of your choice

3. Set the following environment variables

Linux:

export LIBTORCH=/path/to/libtorch
export LD_LIBRARY_PATH=${LIBTORCH}/lib:$LD_LIBRARY_PATH

macOS + Homebrew

brew install pytorch jq
export LIBTORCH=$(brew --cellar pytorch)/$(brew info --json pytorch | jq -r '.[0].installed[0].version')
export LD_LIBRARY_PATH=${LIBTORCH}/lib:$LD_LIBRARY_PATH