dspex

# Generated by SnakeBridge v0.13.0 - DO NOT EDIT MANUALLY
# Regenerate with: mix compile
# Library: dspy 3.1.2
# Python module: dspy.dsp.utils.settings

defmodule Dspy.Dsp.Utils.Settings do
  @moduledoc """
  Submodule bindings for `dspy.dsp.utils.settings`.

  ## Version

  - Requested: 3.1.2
  - Observed at generation: 3.1.2

  ## Runtime Options

  All functions accept a `__runtime__` option for controlling execution behavior:

      Dspy.Dsp.Utils.Settings.some_function(args, __runtime__: [timeout: 120_000])

  ### Supported runtime options

  - `:timeout` - Call timeout in milliseconds (default: 120,000ms / 2 minutes)
  - `:timeout_profile` - Use a named profile (`:default`, `:ml_inference`, `:batch_job`, `:streaming`)
  - `:stream_timeout` - Timeout for streaming operations (default: 1,800,000ms / 30 minutes)
  - `:session_id` - Override the session ID for this call
  - `:pool_name` - Target a specific Snakepit pool (multi-pool setups)
  - `:affinity` - Override session affinity (`:hint`, `:strict_queue`, `:strict_fail_fast`)

  ### Timeout Profiles

  - `:default` - 2 minute timeout for regular calls
  - `:ml_inference` - 10 minute timeout for ML/LLM workloads
  - `:batch_job` - Unlimited timeout for long-running jobs
  - `:streaming` - 2 minute timeout, 30 minute stream_timeout

  ### Example with timeout override

      # For a long-running ML inference call
      Dspy.Dsp.Utils.Settings.predict(data, __runtime__: [timeout_profile: :ml_inference])

      # Or explicit timeout
      Dspy.Dsp.Utils.Settings.predict(data, __runtime__: [timeout: 600_000])

      # Route to a pool and enforce strict affinity
      Dspy.Dsp.Utils.Settings.predict(data, __runtime__: [pool_name: :strict_pool, affinity: :strict_queue])

  See `SnakeBridge.Defaults` for global timeout configuration.

  """

  @doc false
  def __snakebridge_python_name__, do: "dspy.dsp.utils.settings"
  @doc false
  def __snakebridge_library__, do: "dspy"

  @doc """
  A singleton class for DSPy configuration settings.

  Thread-safe global configuration.
  - 'configure' can be called by only one 'owner' thread (the first thread that calls it).
  - Other threads see the configured global values from 'main_thread_config'.
  - 'context' sets thread-local overrides. These overrides propagate to threads spawned
  inside that context block, when (and only when!) using a ParallelExecutor that copies overrides.

  1. Only one unique thread (which can be any thread!) can call dspy.configure.
  2. It affects a global state, visible to all. As a result, user threads work, but they shouldn't be
     mixed with concurrent changes to dspy.configure from the "main" thread.
     (TODO: In the future, add warnings: if there are near-in-time user-thread reads followed by .configure calls.)
  3. Any thread can use dspy.context. It propagates to child threads created with DSPy primitives: Parallel, asyncify, etc.

  ## Returns

  - `Dspy.Dsp.Utils.Settings.Settings.t()`
  """
  @spec settings() :: {:ok, Dspy.Dsp.Utils.Settings.Settings.t()} | {:error, Snakepit.Error.t()}
  def settings() do
    SnakeBridge.Runtime.get_module_attr(__MODULE__, :settings)
  end
end