class RubyLLM::EmbeddingRequest

An EmbeddingRequest is an embedding awaiting a provider-side batch. RubyLLM.embed_later stages one; submit an array of them with RubyLLM.batch, and once the batch completes, collecting the results fills in result:

requests = texts.map { |text| RubyLLM.embed_later(text) }
batch = RubyLLM.batch(requests)
# later, once batch.refresh.complete?
batch.results
requests.first.result.vectors

Attributes

dimensions [R]

The requested vector size, or nil for the model’s default.

model [R]

The Model of the embedding model the request targets.

provider [R]

The Provider the request will be submitted through.

result [RW]

The Embedding, once the batch has completed and its results have been collected. nil until then, and for requests that failed. Batch result collection writes it.

text [R]

The text staged for embedding.

Public Instance Methods

render ()

Source

# File lib/ruby_llm/embedding_request.rb, line 45
def render
  @provider.render_embedding(text, model: @model, dimensions: @dimensions)
end

Returns the request payload this embedding request would send to the provider, in the provider’s wire format.