Embeddings

The Alith SDK supports embeddings, which are numerical representations of text that capture semantic meaning. Embeddings are useful for tasks like semantic search, clustering, and similarity comparisons. Below, you’ll find examples of how to generate and use embeddings in Rust, Python, and Node.js.

Rust

Large Language Embeddings Model

Here we take the OpenAI embeddings model as the example.


use alith::{Agent, EmbeddingsBuilder, LLM};
 
#[tokio::main]
async fn main() -> Result<(), anyhow::Error> {
    let model = LLM::from_model_name("gpt-4")?;
    let embeddings_model = model.embeddings_model("text-embedding-ada-002");
    let data = EmbeddingsBuilder::new(embeddings_model.clone())
        .documents(vec!["doc0", "doc1", "doc2"])
        .unwrap()
        .build()
        .await?;
}

Local Fast Embedding Model


use alith::{EmbeddingsBuilder, FastEmbeddingsModel};
 
#[tokio::main]
async fn main() -> Result<(), anyhow::Error> {
    let embeddings_model = FastEmbeddingsModel::try_default().unwrap();
    let data = EmbeddingsBuilder::new(embeddings_model.clone())
        .documents(vec!["doc0", "doc1", "doc2"])
        .unwrap()
        .build()
        .await?;
    println!("{:?}", data);
    Ok(())
}

Note that running this program will pull the embeddings model from Hugging Face and start the inference engine locally for inference, so we need to turn on the inference feature.