ray/rllib/examples/rock_paper_scissors_multiagent.py

"""A simple multi-agent env with two agents playing rock paper scissors.

This demonstrates running the following policies in competition:
    (1) heuristic policy of repeating the same move
    (2) heuristic policy of beating the last opponent move
    (3) LSTM/feedforward PG policies
    (4) LSTM policy with custom entropy loss
"""

import argparse
import os
from pettingzoo.classic import rps_v2
import random

from ray import tune
from ray.rllib.agents.pg import PGTrainer, PGTFPolicy, PGTorchPolicy
from ray.rllib.agents.registry import get_trainer_class
from ray.rllib.env import PettingZooEnv
from ray.rllib.examples.policy.rock_paper_scissors_dummies import \
    BeatLastHeuristic, AlwaysSameHeuristic
from ray.rllib.policy.policy import PolicySpec
from ray.rllib.utils.framework import try_import_tf, try_import_torch
from ray.rllib.utils.test_utils import check_learning_achieved
from ray.tune.registry import register_env

tf1, tf, tfv = try_import_tf()
torch, _ = try_import_torch()

parser = argparse.ArgumentParser()
parser.add_argument(
    "--framework",
    choices=["tf", "tf2", "tfe", "torch"],
    default="tf",
    help="The DL framework specifier.")
parser.add_argument(
    "--as-test",
    action="store_true",
    help="Whether this script should be run as a test: --stop-reward must "
    "be achieved within --stop-timesteps AND --stop-iters.")
parser.add_argument(
    "--stop-iters",
    type=int,
    default=150,
    help="Number of iterations to train.")
parser.add_argument(
    "--stop-timesteps",
    type=int,
    default=100000,
    help="Number of timesteps to train.")
parser.add_argument(
    "--stop-reward",
    type=float,
    default=1000.0,
    help="Reward at which we stop training.")


def env_creator(args):
    env = rps_v2.env()
    return env


register_env("RockPaperScissors",
             lambda config: PettingZooEnv(env_creator(config)))


def run_same_policy(args, stop):
    """Use the same policy for both agents (trivial case)."""
    config = {
        "env": "RockPaperScissors",
        "framework": args.framework,
    }

    results = tune.run("PG", config=config, stop=stop, verbose=1)

    if args.as_test:
        # Check vs 0.0 as we are playing a zero-sum game.
        check_learning_achieved(results, 0.0)


def run_heuristic_vs_learned(args, use_lstm=False, trainer="PG"):
    """Run heuristic policies vs a learned agent.

    The learned agent should eventually reach a reward of ~5 with
    use_lstm=False, and ~7 with use_lstm=True. The reason the LSTM policy
    can perform better is since it can distinguish between the always_same vs
    beat_last heuristics.
    """

    def select_policy(agent_id, episode, **kwargs):
        if agent_id == "player_0":
            return "learned"
        else:
            return random.choice(["always_same", "beat_last"])

    config = {
        "env": "RockPaperScissors",
        "gamma": 0.9,
        # Use GPUs iff `RLLIB_NUM_GPUS` env var set to > 0.
        "num_gpus": int(os.environ.get("RLLIB_NUM_GPUS", "0")),
        "num_workers": 0,
        "num_envs_per_worker": 4,
        "rollout_fragment_length": 10,
        "train_batch_size": 200,
        "metrics_num_episodes_for_smoothing": 200,
        "multiagent": {
            "policies_to_train": ["learned"],
            "policies": {
                "always_same": PolicySpec(policy_class=AlwaysSameHeuristic),
                "beat_last": PolicySpec(policy_class=BeatLastHeuristic),
                "learned": PolicySpec(config={
                    "model": {
                        "use_lstm": use_lstm
                    },
                    "framework": args.framework,
                }),
            },
            "policy_mapping_fn": select_policy,
        },
        "framework": args.framework,
    }
    cls = get_trainer_class(trainer) if isinstance(trainer, str) else trainer
    trainer_obj = cls(config=config)
    for _ in range(args.stop_iters):
        results = trainer_obj.train()
        # Timesteps reached.
        if "policy_always_same_reward" not in results["hist_stats"]:
            reward_diff = 0
            continue
        reward_diff = sum(results["hist_stats"]["policy_learned_reward"])
        if results["timesteps_total"] > args.stop_timesteps:
            break
        # Reward (difference) reached -> all good, return.
        elif reward_diff > args.stop_reward:
            return

    # Reward (difference) not reached: Error if `as_test`.
    if args.as_test:
        raise ValueError(
            "Desired reward difference ({}) not reached! Only got to {}.".
            format(args.stop_reward, reward_diff))


def run_with_custom_entropy_loss(args, stop):
    """Example of customizing the loss function of an existing policy.

    This performs about the same as the default loss does."""

    def entropy_policy_gradient_loss(policy, model, dist_class, train_batch):
        logits, _ = model(train_batch)
        action_dist = dist_class(logits, model)
        if args.framework == "torch":
            # Required by PGTorchPolicy's stats fn.
            model.tower_stats["policy_loss"] = torch.tensor([0.0])
            return torch.mean(-0.1 * action_dist.entropy() -
                              (action_dist.logp(train_batch["actions"]) *
                               train_batch["advantages"]))
        else:
            return (-0.1 * action_dist.entropy() - tf.reduce_mean(
                action_dist.logp(train_batch["actions"]) *
                train_batch["advantages"]))

    policy_cls = PGTorchPolicy if args.framework == "torch" \
        else PGTFPolicy
    EntropyPolicy = policy_cls.with_updates(
        loss_fn=entropy_policy_gradient_loss)

    class EntropyLossPG(PGTrainer):
        def get_default_policy_class(self, config):
            return EntropyPolicy

    run_heuristic_vs_learned(args, use_lstm=True, trainer=EntropyLossPG)


if __name__ == "__main__":
    args = parser.parse_args()

    stop = {
        "training_iteration": args.stop_iters,
        "timesteps_total": args.stop_timesteps,
        "episode_reward_mean": args.stop_reward,
    }

    run_same_policy(args, stop=stop)
    print("run_same_policy: ok.")

    run_heuristic_vs_learned(args, use_lstm=False)
    print("run_heuristic_vs_learned(w/o lstm): ok.")

    run_heuristic_vs_learned(args, use_lstm=True)
    print("run_heuristic_vs_learned (w/ lstm): ok.")

    run_with_custom_entropy_loss(args, stop=stop)
    print("run_with_custom_entropy_loss: ok.")