LlmCore.Pipelines.InferencePipeline (llm_core v0.3.0)

ALF pipeline that normalizes a request, resolves routing, and dispatches it to the selected provider using either blocking or streaming mode.

Summary

alf_components()

components()

Executes the inference pipeline for the given mode (:send or :stream).

start()

start(opts)

started?()

stop()

@spec call(any(), Keyword.t()) :: any() | [any()] | nil

@spec call(any(), Keyword.t()) :: reference()

@spec components() :: [map()]

@spec execute(
  :send | :stream,
  String.t() | [map()] | map(),
  String.t() | atom() | nil,
  keyword()
) :: {:ok, term()} | {:error, term()}

Executes the inference pipeline for the given mode (:send or :stream).

Normalizes the request, resolves routing, dispatches to the provider, and optionally applies structured output extraction.

@spec flow(map(), list(), Keyword.t()) :: Enumerable.t()

@spec start() :: :ok

@spec start(list()) :: :ok

@spec started?() :: true | false

@spec stop() :: :ok | {:exit, {atom(), any()}}

@spec stream(Enumerable.t(), Keyword.t()) :: Enumerable.t()