Utilities¶

These packages provide experiment setup and deployment helpers.

cli ¶

Utilities for commandline usage of experiments.

This module provides common utilities for commandline usage of experiments. It is intended to be used as a library for writing commandline interfaces for experiments.

Functions:¶

initialize ¶

initialize(**kwargs: Any) -> DictConfig | ListConfig

Initialize the experiment environment.

Loads the configuration and sets up logging according to its settings.

Parameters:

Name	Type	Description	Default
`**kwargs`	`Any`	Additional configuration overrides.	`{}`

Returns:

Type	Description
`DictConfig \| ListConfig`	The initialized configuration object.

utils ¶

Functions:¶

get_seed ¶

get_seed() -> int

Generate a random seed.

Returns:

Type	Description
`int`	A random seed.

initialize_random ¶

initialize_random(seed: int) -> None

Initialize the random number generator with the given seed.

Parameters:

Name	Type	Description	Default
`seed`	`int`	The seed to initialize the random number generator with.	required

build_environments_from_directory ¶

build_environments_from_directory(path: Path | str, builder: EnvironmentBuilder, *, pattern: str = '*', include_files: bool = True, include_folders: bool = True) -> Iterable[OfflineEnvironment]

Build environments from a directory.

This helper function can be used to build environments from multiple files or folders. First all files and folders in the path matching the pattern are selected. These are then passed to the builder function which creates the environment.

Parameters:

Name	Type	Description	Default
`path`	`Path \| str`	The path to the directory from which environments are created.	required
`builder`	`EnvironmentBuilder`	A function building an environment from a path and returning it.	required
`pattern`	`str`	A glob pattern. Matching files and folders will be passed to `builder`.	`'*'`
`include_files`	`bool`	Specify whether to pass files to `builder`.	`True`
`include_folders`	`bool`	Specify whether to pass folders to `builder`.	`True`

tool ¶

Functions:¶

start_prediction_loop ¶

start_prediction_loop(model: Model, adapter: Adapter, *, adapter_to_model_transforms: Transform | None = None, model_to_adapter_transforms: Transform | None = None) -> None

Start a prediction loop with the given model and adapter.