ray/rllib/agents/ppo/tests/test_appo.py

121 lines
4.2 KiB
Python

import unittest
import ray
import ray.rllib.agents.ppo as ppo
from ray.rllib.policy.sample_batch import DEFAULT_POLICY_ID
from ray.rllib.utils.metrics.learner_info import LEARNER_INFO, \
LEARNER_STATS_KEY
from ray.rllib.utils.test_utils import check_compute_single_action, \
check_train_results, framework_iterator
class TestAPPO(unittest.TestCase):
@classmethod
def setUpClass(cls):
ray.init()
@classmethod
def tearDownClass(cls):
ray.shutdown()
def test_appo_compilation(self):
"""Test whether an APPOTrainer can be built with both frameworks."""
config = ppo.appo.DEFAULT_CONFIG.copy()
config["num_workers"] = 1
num_iterations = 2
for _ in framework_iterator(config, with_eager_tracing=True):
print("w/o v-trace")
_config = config.copy()
_config["vtrace"] = False
trainer = ppo.APPOTrainer(config=_config, env="CartPole-v0")
for i in range(num_iterations):
results = trainer.train()
check_train_results(results)
print(results)
check_compute_single_action(trainer)
trainer.stop()
print("w/ v-trace")
_config = config.copy()
_config["vtrace"] = True
trainer = ppo.APPOTrainer(config=_config, env="CartPole-v0")
for i in range(num_iterations):
results = trainer.train()
check_train_results(results)
print(results)
check_compute_single_action(trainer)
trainer.stop()
def test_appo_two_tf_optimizers(self):
config = ppo.appo.DEFAULT_CONFIG.copy()
config["num_workers"] = 1
# Not explicitly setting this should cause a warning, but not fail.
# config["_tf_policy_handles_more_than_one_loss"] = True
config["_separate_vf_optimizer"] = True
config["_lr_vf"] = 0.0002
# Make sure we have two completely separate models for policy and
# value function.
config["model"]["vf_share_layers"] = False
num_iterations = 2
# Only supported for tf so far.
for _ in framework_iterator(config, frameworks=("tf2", "tf")):
trainer = ppo.APPOTrainer(config=config, env="CartPole-v0")
for i in range(num_iterations):
results = trainer.train()
check_train_results(results)
print(results)
check_compute_single_action(trainer)
trainer.stop()
def test_appo_entropy_coeff_schedule(self):
config = ppo.appo.DEFAULT_CONFIG.copy()
config["num_workers"] = 1
config["num_gpus"] = 0
config["train_batch_size"] = 20
config["batch_mode"] = "truncate_episodes"
config["rollout_fragment_length"] = 10
config["timesteps_per_iteration"] = 20
# 0 metrics reporting delay, this makes sure timestep,
# which entropy coeff depends on, is updated after each worker rollout.
config["min_iter_time_s"] = 0
# Initial lr, doesn't really matter because of the schedule below.
config["entropy_coeff"] = 0.01
schedule = [
[0, 0.01],
[120, 0.0001],
]
config["entropy_coeff_schedule"] = schedule
def _step_n_times(trainer, n: int):
"""Step trainer n times.
Returns:
learning rate at the end of the execution.
"""
for _ in range(n):
results = trainer.train()
return results["info"][LEARNER_INFO][DEFAULT_POLICY_ID][
LEARNER_STATS_KEY]["entropy_coeff"]
for _ in framework_iterator(config):
trainer = ppo.APPOTrainer(config=config, env="CartPole-v0")
coeff = _step_n_times(trainer, 1) # 20 timesteps
# Should be close to the starting coeff of 0.01.
self.assertGreaterEqual(coeff, 0.005)
coeff = _step_n_times(trainer, 10) # 200 timesteps
# Should have annealed to the final coeff of 0.0001.
self.assertLessEqual(coeff, 0.00011)
trainer.stop()
if __name__ == "__main__":
import pytest
import sys
sys.exit(pytest.main(["-v", __file__]))