embeddingRoles

Map of role to provider to embedding model name, for deployments whose provider is not fixed

• a bring-your-own-key application, or one configured for failover across providers.

embabel:
  models:
    embedding-roles:
      documents:
        openai:  text-embedding-3-small
        mistral: mistral-embed

Takes precedence over embeddingServices for the active provider.

Flatter than roles, which maps to a whole LlmOptions. An embedding model takes text and returns a vector: there is no temperature, no token budget, and the one parameter that does vary - the width - is the model's to state and the index's to follow, not the caller's to ask for. A name is the whole of what a deployment has to say here.

Declared last, despite belonging with embeddingServices, so that adding it does not renumber the existing parameters for anyone constructing this positionally.