class

Arcana::AI::Embed::Provider

Inherits Arcana::AI::Traceable < Reference < Object

Instance methods

batch_embed(request : Request, batch_size : Int32 = 100) : Result

Embed texts in batches, aggregating results.

Source

batch_embed_with_retry(request : Request, batch_size : Int32 = 100, max_retries : Int32 = 3) : Result

Combines batching and retry.

Source

embed(request : Request) : Result

Source

embed_with_retry(request : Request, max_retries : Int32 = 3, base_delay : Float64 = 1.0) : Result

Embed, retrying rate limits, server errors and network errors the way chat calls do (see RetryPolicy), including waiting as long as the provider asks. The result records the retries.

Source

name

Source