Torus.Embeddings.HuggingFace (Torus v0.4.0)

A wrapper around Hugging Face API. It allows you to generate embeddings using a variety of models available on Hugging Face.

To use it:

• Add the following to your config.exs:

  config :torus, embedding_module: Torus.Embeddings.HuggingFace

• Add req to your mix.exs dependencies:

  def deps do
  [
    {:req, "~> 0.5"}
  ]
  end

• Add an API token for hugging face to your runtime.exs. You can get your token here.

  config :torus, Torus.Embeddings.HuggingFace, token: System.get_env("HUGGING_FACE_API_KEY")

By default, it uses sentence-transformers/all-MiniLM-L6-v2 model, but you can specify a different model by explicitly passing model to the config:

config :torus, Torus.Embeddings.HuggingFace, model: "your/model"

See Torus.semantic/5 on how to use this module to introduce semantic search in your application.