simulation/modules/IsaacLab/scripts/reinforcement_learning/skrl/play.py

# Copyright (c) 2022-2026, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
# All rights reserved.
#
# SPDX-License-Identifier: BSD-3-Clause

"""
Script to play a checkpoint of an RL agent from skrl.

Visit the skrl documentation (https://skrl.readthedocs.io) to see the examples structured in
a more user-friendly way.
"""

import warnings

warnings.warn(
    "scripts/reinforcement_learning/skrl/play.py is deprecated. Use "
    "`./isaaclab.sh play --rl_library skrl --task <TASK>` instead. "
    "Example: `./isaaclab.sh play --rl_library skrl --task Isaac-Cartpole-v0`.",
    DeprecationWarning,
    stacklevel=1,
)

import argparse
import contextlib
import os
import random
import sys
import time

import gymnasium as gym
import skrl
import torch
from packaging import version

from isaaclab.envs import DirectMARLEnvCfg
from isaaclab.utils.dict import print_dict
from isaaclab.utils.seed import configure_seed

from isaaclab_rl.utils.pretrained_checkpoint import get_published_pretrained_checkpoint

import isaaclab_tasks  # noqa: F401
from isaaclab_tasks.utils import (
    add_launcher_args,
    get_checkpoint_path,
    launch_simulation,
    resolve_task_config,
    setup_preset_cli,
)

# PLACEHOLDER: Extension template (do not remove this comment)
with contextlib.suppress(ImportError):
    import isaaclab_tasks_experimental  # noqa: F401

SKRL_VERSION = "2.1.0"

# -- argparse ----------------------------------------------------------------
parser = argparse.ArgumentParser(description="Play a checkpoint of an RL agent from skrl.")
parser.add_argument("--video", action="store_true", default=False, help="Record videos during training.")
parser.add_argument("--video_length", type=int, default=200, help="Length of the recorded video (in steps).")
parser.add_argument(
    "--disable_fabric", action="store_true", default=False, help="Disable fabric and use USD I/O operations."
)
parser.add_argument("--num_envs", type=int, default=None, help="Number of environments to simulate.")
parser.add_argument("--task", type=str, default=None, help="Name of the task.")
parser.add_argument(
    "--agent",
    type=str,
    default=None,
    help=(
        "Name of the RL agent configuration entry point. Defaults to None, in which case the argument "
        "--algorithm is used to determine the default agent configuration entry point."
    ),
)
parser.add_argument("--checkpoint", type=str, default=None, help="Path to model checkpoint.")
parser.add_argument("--seed", type=int, default=None, help="Seed used for the environment")
parser.add_argument(
    "--use_pretrained_checkpoint",
    action="store_true",
    help="Use the pre-trained checkpoint from Nucleus.",
)
parser.add_argument(
    "--ml_framework",
    type=str,
    default="torch",
    choices=["torch", "jax"],
    help="The ML framework used for training the skrl agent.",
)
parser.add_argument(
    "--algorithm",
    type=str,
    default="PPO",
    choices=["AMP", "PPO", "IPPO", "MAPPO"],
    help="The RL algorithm used for training the skrl agent.",
)
parser.add_argument("--real-time", action="store_true", default=False, help="Run in real-time, if possible.")
add_launcher_args(parser)
args_cli, hydra_args = setup_preset_cli(parser)
sys.argv = [sys.argv[0]] + hydra_args

if args_cli.video:
    args_cli.enable_cameras = True

# -- check skrl version ------------------------------------------------------
if version.parse(skrl.__version__) < version.parse(SKRL_VERSION):
    skrl.logger.error(
        f"Unsupported skrl version: {skrl.__version__}. "
        f"Install supported version using 'pip install skrl>={SKRL_VERSION}'"
    )
    exit()

# config shortcuts
if args_cli.agent is None:
    algorithm = args_cli.algorithm.lower()
    agent_cfg_entry_point = "skrl_cfg_entry_point" if algorithm in ["ppo"] else f"skrl_{algorithm}_cfg_entry_point"
else:
    agent_cfg_entry_point = args_cli.agent
    algorithm = agent_cfg_entry_point.split("_cfg")[0].split("skrl_")[-1].lower()


def main():
    """Play with skrl agent."""
    env_cfg, experiment_cfg = resolve_task_config(args_cli.task, agent_cfg_entry_point)
    with launch_simulation(env_cfg, args_cli):
        if args_cli.ml_framework.startswith("torch"):
            from skrl.utils.runner.torch import Runner
        elif args_cli.ml_framework.startswith("jax"):
            from skrl.utils.runner.jax import Runner

        from isaaclab_rl.skrl import SkrlVecEnvWrapper

        # grab task name for checkpoint path
        task_name = args_cli.task.split(":")[-1]
        train_task_name = task_name.replace("-Play", "")

        # override configurations with non-hydra CLI arguments
        env_cfg.scene.num_envs = args_cli.num_envs if args_cli.num_envs is not None else env_cfg.scene.num_envs
        env_cfg.sim.device = args_cli.device if args_cli.device is not None else env_cfg.sim.device

        # configure the ML framework into the global skrl variable
        if args_cli.ml_framework.startswith("jax"):
            skrl.config.jax.backend = "jax" if args_cli.ml_framework == "jax" else "numpy"

        # randomly sample a seed if seed = -1
        if args_cli.seed == -1:
            args_cli.seed = random.randint(0, 10000)

        # set the agent and environment seed from command line
        experiment_cfg["seed"] = args_cli.seed if args_cli.seed is not None else experiment_cfg["seed"]
        env_cfg.seed = experiment_cfg["seed"]

        # specify directory for logging experiments (load checkpoint)
        log_root_path = os.path.join("logs", "skrl", experiment_cfg["agent"]["experiment"]["directory"])
        log_root_path = os.path.abspath(log_root_path)
        print(f"[INFO] Loading experiment from directory: {log_root_path}")
        # get checkpoint path
        if args_cli.use_pretrained_checkpoint:
            resume_path = get_published_pretrained_checkpoint("skrl", train_task_name)
            if not resume_path:
                print("[INFO] Unfortunately a pre-trained checkpoint is currently unavailable for this task.")
                return
        elif args_cli.checkpoint:
            resume_path = os.path.abspath(args_cli.checkpoint)
        else:
            resume_path = get_checkpoint_path(
                log_root_path, run_dir=f".*_{algorithm}_{args_cli.ml_framework}", other_dirs=["checkpoints"]
            )
        log_dir = os.path.dirname(os.path.dirname(resume_path))

        # set the log directory for the environment
        env_cfg.log_dir = log_dir

        # create isaac environment
        env = gym.make(args_cli.task, cfg=env_cfg, render_mode="rgb_array" if args_cli.video else None)

        # convert to single-agent instance if required by the RL algorithm
        if isinstance(env.unwrapped.cfg, DirectMARLEnvCfg) and algorithm in ["ppo"]:
            from isaaclab.envs import multi_agent_to_single_agent

            env = multi_agent_to_single_agent(env)

        # get environment (step) dt for real-time evaluation
        try:
            dt = env.step_dt
        except AttributeError:
            dt = env.unwrapped.step_dt

        # wrap for video recording
        if args_cli.video:
            video_kwargs = {
                "video_folder": os.path.join(log_dir, "videos", "play"),
                "step_trigger": lambda step: step == 0,
                "video_length": args_cli.video_length,
                "disable_logger": True,
            }
            print("[INFO] Recording videos during training.")
            print_dict(video_kwargs, nesting=4)
            env = gym.wrappers.RecordVideo(env, **video_kwargs)

        # wrap around environment for skrl
        env = SkrlVecEnvWrapper(env, ml_framework=args_cli.ml_framework)

        # configure and instantiate the skrl runner
        experiment_cfg["trainer"]["close_environment_at_exit"] = False
        experiment_cfg["agent"]["experiment"]["write_interval"] = 0
        experiment_cfg["agent"]["experiment"]["checkpoint_interval"] = 0
        runner = Runner(env, experiment_cfg)
        # configure_seed must be called after Runner() so that PyTorch deterministic settings
        # do not interfere with Runner's internal initialization.
        if args_cli.deterministic:
            configure_seed(env_cfg.seed, True)

        print(f"[INFO] Loading model checkpoint from: {resume_path}")
        runner.agent.load(resume_path)
        runner.agent.enable_training_mode(False, apply_to_models=True)

        # reset environment
        obs, _ = env.reset()
        states = env.state()
        timestep = 0
        # simulate environment
        try:
            while True:
                start_time = time.time()

                with torch.inference_mode():
                    outputs = runner.agent.act(obs, states, timestep=0, timesteps=0)
                    if hasattr(env, "possible_agents"):
                        actions = {a: outputs[-1][a].get("mean_actions", outputs[0][a]) for a in env.possible_agents}
                    else:
                        actions = outputs[-1].get("mean_actions", outputs[0])
                    obs, _, _, _, _ = env.step(actions)
                    states = env.state()
                if args_cli.video:
                    timestep += 1
                    if timestep == args_cli.video_length:
                        break

                sleep_time = dt - (time.time() - start_time)
                if args_cli.real_time and sleep_time > 0:
                    time.sleep(sleep_time)

            # close the simulator
            env.close()
        except KeyboardInterrupt:
            pass


if __name__ == "__main__":
    main()