ray/rllib/examples/two_trainer_workflow.py

226 lines
8 KiB
Python

"""Example of using a custom training workflow.
Here we create a number of CartPole agents, some of which are trained with
DQN, and some of which are trained with PPO. Both are executed concurrently
via a custom training workflow.
"""
import argparse
import os
import ray
from ray import tune
from ray.rllib.agents import with_common_config
from ray.rllib.agents.trainer import Trainer
from ray.rllib.algorithms.dqn.dqn import DEFAULT_CONFIG as DQN_CONFIG
from ray.rllib.algorithms.dqn.dqn_tf_policy import DQNTFPolicy
from ray.rllib.algorithms.dqn.dqn_torch_policy import DQNTorchPolicy
from ray.rllib.algorithms.ppo.ppo import DEFAULT_CONFIG as PPO_CONFIG
from ray.rllib.algorithms.ppo.ppo_tf_policy import PPOTF1Policy
from ray.rllib.algorithms.ppo.ppo_torch_policy import PPOTorchPolicy
from ray.rllib.evaluation.postprocessing import Postprocessing
from ray.rllib.execution.rollout_ops import synchronous_parallel_sample
from ray.rllib.execution.train_ops import train_one_step
from ray.rllib.utils.replay_buffers.multi_agent_replay_buffer import (
MultiAgentReplayBuffer,
)
from ray.rllib.examples.env.multi_agent import MultiAgentCartPole
from ray.rllib.policy.sample_batch import MultiAgentBatch, SampleBatch
from ray.rllib.utils.annotations import override
from ray.rllib.utils.metrics import (
NUM_AGENT_STEPS_SAMPLED,
NUM_ENV_STEPS_SAMPLED,
NUM_TARGET_UPDATES,
LAST_TARGET_UPDATE_TS,
)
from ray.rllib.utils.sgd import standardized
from ray.rllib.utils.test_utils import check_learning_achieved
from ray.rllib.utils.typing import ResultDict, TrainerConfigDict
from ray.tune.registry import register_env
parser = argparse.ArgumentParser()
parser.add_argument("--torch", action="store_true")
parser.add_argument("--mixed-torch-tf", action="store_true")
parser.add_argument(
"--local-mode",
action="store_true",
help="Init Ray in local mode for easier debugging.",
)
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=400, help="Number of iterations to train."
)
parser.add_argument(
"--stop-timesteps", type=int, default=100000, help="Number of timesteps to train."
)
# 600.0 = 4 (num_agents) x 150.0
parser.add_argument(
"--stop-reward", type=float, default=600.0, help="Reward at which we stop training."
)
# Define new Trainer with custom execution_plan/workflow.
class MyTrainer(Trainer):
@classmethod
@override(Trainer)
def get_default_config(cls) -> TrainerConfigDict:
# Run this Trainer with new `training_iteration` API and set some PPO-specific
# parameters.
return with_common_config(
{
"num_sgd_iter": 10,
"sgd_minibatch_size": 128,
}
)
@override(Trainer)
def setup(self, config):
# Call super's `setup` to create rollout workers.
super().setup(config)
# Create local replay buffer.
self.local_replay_buffer = MultiAgentReplayBuffer(
num_shards=1, learning_starts=1000, capacity=50000
)
@override(Trainer)
def training_step(self) -> ResultDict:
# Generate common experiences, collect batch for PPO, store every (DQN) batch
# into replay buffer.
ppo_batches = []
num_env_steps = 0
# PPO batch size fixed at 200.
while num_env_steps < 200:
ma_batches = synchronous_parallel_sample(
worker_set=self.workers, concat=False
)
# Loop through (parallely collected) ma-batches.
for ma_batch in ma_batches:
# Update sampled counters.
self._counters[NUM_ENV_STEPS_SAMPLED] += ma_batch.count
self._counters[NUM_AGENT_STEPS_SAMPLED] += ma_batch.agent_steps()
ppo_batch = ma_batch.policy_batches.pop("ppo_policy")
# Add collected batches (only for DQN policy) to replay buffer.
self.local_replay_buffer.add(ma_batch)
ppo_batches.append(ppo_batch)
num_env_steps += ppo_batch.count
# DQN sub-flow.
dqn_train_results = {}
dqn_train_batch = self.local_replay_buffer.sample(num_items=64)
if dqn_train_batch is not None:
dqn_train_results = train_one_step(self, dqn_train_batch, ["dqn_policy"])
self._counters["agent_steps_trained_DQN"] += dqn_train_batch.agent_steps()
print(
"DQN policy learning on samples from",
"agent steps trained",
dqn_train_batch.agent_steps(),
)
# Update DQN's target net every 500 train steps.
if (
self._counters["agent_steps_trained_DQN"]
- self._counters[LAST_TARGET_UPDATE_TS]
>= 500
):
self.workers.local_worker().get_policy("dqn_policy").update_target()
self._counters[NUM_TARGET_UPDATES] += 1
self._counters[LAST_TARGET_UPDATE_TS] = self._counters[
"agent_steps_trained_DQN"
]
# PPO sub-flow.
ppo_train_batch = SampleBatch.concat_samples(ppo_batches)
self._counters["agent_steps_trained_PPO"] += ppo_train_batch.agent_steps()
# Standardize advantages.
ppo_train_batch[Postprocessing.ADVANTAGES] = standardized(
ppo_train_batch[Postprocessing.ADVANTAGES]
)
print(
"PPO policy learning on samples from",
"agent steps trained",
ppo_train_batch.agent_steps(),
)
ppo_train_batch = MultiAgentBatch(
{"ppo_policy": ppo_train_batch}, ppo_train_batch.count
)
ppo_train_results = train_one_step(self, ppo_train_batch, ["ppo_policy"])
# Combine results for PPO and DQN into one results dict.
results = dict(ppo_train_results, **dqn_train_results)
return results
if __name__ == "__main__":
args = parser.parse_args()
assert not (
args.torch and args.mixed_torch_tf
), "Use either --torch or --mixed-torch-tf, not both!"
ray.init(local_mode=args.local_mode)
# Simple environment with 4 independent cartpole entities
register_env(
"multi_agent_cartpole", lambda _: MultiAgentCartPole({"num_agents": 4})
)
# framework can be changed, so removed the hardcoded framework key
# from policy configs.
ppo_config = PPO_CONFIG
del ppo_config["framework"]
dqn_config = DQN_CONFIG
del dqn_config["framework"]
# Note that since the trainer below does not include a default policy or
# policy configs, we have to explicitly set it in the multiagent config:
policies = {
"ppo_policy": (
PPOTorchPolicy if args.torch or args.mixed_torch_tf else PPOTF1Policy,
None,
None,
ppo_config,
),
"dqn_policy": (
DQNTorchPolicy if args.torch else DQNTFPolicy,
None,
None,
dqn_config,
),
}
def policy_mapping_fn(agent_id, episode, worker, **kwargs):
if agent_id % 2 == 0:
return "ppo_policy"
else:
return "dqn_policy"
config = {
"rollout_fragment_length": 50,
"num_workers": 0,
"env": "multi_agent_cartpole",
"multiagent": {
"policies": policies,
"policy_mapping_fn": policy_mapping_fn,
"policies_to_train": ["dqn_policy", "ppo_policy"],
},
# Use GPUs iff `RLLIB_NUM_GPUS` env var set to > 0.
"num_gpus": int(os.environ.get("RLLIB_NUM_GPUS", "0")),
"framework": "torch" if args.torch else "tf",
}
stop = {
"training_iteration": args.stop_iters,
"timesteps_total": args.stop_timesteps,
"episode_reward_mean": args.stop_reward,
}
results = tune.run(MyTrainer, config=config, stop=stop)
if args.as_test:
check_learning_achieved(results, args.stop_reward)
ray.shutdown()