mirror of
https://github.com/vale981/ray
synced 2025-03-06 18:41:40 -05:00
026bf01071
* Fix QMix, SAC, and MADDPA too. * Unpin gym and deprecate pendulum v0 Many tests in rllib depended on pendulum v0, however in gym 0.21, pendulum v0 was deprecated in favor of pendulum v1. This may change reward thresholds, so will have to potentially rerun all of the pendulum v1 benchmarks, or use another environment in favor. The same applies to frozen lake v0 and frozen lake v1 Lastly, all of the RLlib tests and have been moved to python 3.7 * Add gym installation based on python version. Pin python<= 3.6 to gym 0.19 due to install issues with atari roms in gym 0.20 * Reformatting * Fixing tests * Move atari-py install conditional to req.txt * migrate to new ale install method * Fix QMix, SAC, and MADDPA too. * Unpin gym and deprecate pendulum v0 Many tests in rllib depended on pendulum v0, however in gym 0.21, pendulum v0 was deprecated in favor of pendulum v1. This may change reward thresholds, so will have to potentially rerun all of the pendulum v1 benchmarks, or use another environment in favor. The same applies to frozen lake v0 and frozen lake v1 Lastly, all of the RLlib tests and have been moved to python 3.7 * Add gym installation based on python version. Pin python<= 3.6 to gym 0.19 due to install issues with atari roms in gym 0.20 Move atari-py install conditional to req.txt migrate to new ale install method Make parametric_actions_cartpole return float32 actions/obs Adding type conversions if obs/actions don't match space Add utils to make elements match gym space dtypes Co-authored-by: Jun Gong <jungong@anyscale.com> Co-authored-by: sven1977 <svenmika1977@gmail.com>
75 lines
2.7 KiB
Python
75 lines
2.7 KiB
Python
import unittest
|
|
import pytest
|
|
|
|
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, check_compute_single_action, \
|
|
check_train_results, framework_iterator
|
|
|
|
|
|
class TestDDPPO(unittest.TestCase):
|
|
@classmethod
|
|
def setUpClass(cls):
|
|
ray.init()
|
|
|
|
@classmethod
|
|
def tearDownClass(cls):
|
|
ray.shutdown()
|
|
|
|
def test_ddppo_compilation(self):
|
|
"""Test whether a DDPPOTrainer can be built with both frameworks."""
|
|
config = ppo.ddppo.DEFAULT_CONFIG.copy()
|
|
config["num_gpus_per_worker"] = 0
|
|
num_iterations = 2
|
|
|
|
for _ in framework_iterator(config, frameworks="torch"):
|
|
trainer = ppo.ddppo.DDPPOTrainer(config=config, env="CartPole-v0")
|
|
for i in range(num_iterations):
|
|
results = trainer.train()
|
|
check_train_results(results)
|
|
print(results)
|
|
# Make sure, weights on all workers are the same (including
|
|
# local one).
|
|
weights = trainer.workers.foreach_worker(
|
|
lambda w: w.get_weights())
|
|
for w in weights[1:]:
|
|
check(w, weights[0])
|
|
|
|
check_compute_single_action(trainer)
|
|
trainer.stop()
|
|
|
|
def test_ddppo_schedule(self):
|
|
"""Test whether lr_schedule will anneal lr to 0"""
|
|
config = ppo.ddppo.DEFAULT_CONFIG.copy()
|
|
config["num_gpus_per_worker"] = 0
|
|
config["lr_schedule"] = [[0, config["lr"]], [1000, 0.0]]
|
|
num_iterations = 3
|
|
|
|
for _ in framework_iterator(config, "torch"):
|
|
trainer = ppo.ddppo.DDPPOTrainer(config=config, env="CartPole-v0")
|
|
for _ in range(num_iterations):
|
|
result = trainer.train()
|
|
lr = result["info"][LEARNER_INFO][DEFAULT_POLICY_ID][
|
|
LEARNER_STATS_KEY]["cur_lr"]
|
|
trainer.stop()
|
|
assert lr == 0.0, "lr should anneal to 0.0"
|
|
|
|
def test_validate_config(self):
|
|
"""Test if DDPPO will raise errors after invalid configs are passed."""
|
|
config = ppo.ddppo.DEFAULT_CONFIG.copy()
|
|
config["kl_coeff"] = 1.
|
|
msg = "DDPPO doesn't support KL penalties like PPO-1"
|
|
with pytest.raises(ValueError, match=msg):
|
|
ppo.ddppo.DDPPOTrainer(config=config, env="CartPole-v0")
|
|
config["kl_coeff"] = 0.
|
|
config["kl_target"] = 1.
|
|
with pytest.raises(ValueError, match=msg):
|
|
ppo.ddppo.DDPPOTrainer(config=config, env="CartPole-v0")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
import sys
|
|
sys.exit(pytest.main(["-v", __file__]))
|