maple.cmd.maple_cli.run

maple.cmd.maple_cli.run(policy_id: str = <typer.models.ArgumentInfo object>, env_id: str = <typer.models.ArgumentInfo object>, task: str = <typer.models.OptionInfo object>, instruction: str | None = <typer.models.OptionInfo object>, max_steps: int = <typer.models.OptionInfo object>, seed: int | None = <typer.models.OptionInfo object>, env_kwargs: str = <typer.models.OptionInfo object>, model_kwargs: str = <typer.models.OptionInfo object>, save_video: bool = <typer.models.OptionInfo object>, video_dir: str | None = <typer.models.OptionInfo object>, timeout: int | None = <typer.models.OptionInfo object>, port: int = <typer.models.OptionInfo object>) → None

Run a policy on an environment task.

Executes a single episode of a policy on a specified environment task, communicating with the MAPLE daemon to orchestrate the execution. Displays real-time progress and results including success status, steps taken, rewards, and video paths.

Parameters:

• policy_id – Identifier of the policy container to use.

• env_id – Identifier of the environment container to use.

• task – Task specification string.

• instruction – Optional instruction to override default task instruction.

• max_steps – Maximum number of steps before truncation.

• seed – Random seed for reproducibility.

• env_kwargs – Model-specific parameters.

• model_kwargs – Model-specific parameters.

• save_video – Whether to record and save episode video.

• video_dir – Directory path for saving videos.

• timeout – Timeout multiplier for HTTP request.

• port – Daemon port number.