Unity 机器学习代理工具包 (ML-Agents) 是一个开源项目,它使游戏和模拟能够作为训练智能代理的环境。
您最多选择25个主题 主题必须以中文或者字母或数字开头,可以包含连字符 (-),并且长度不得超过35个字符
 
 
 
 
 

405 行
15 KiB

from unittest import mock
import pytest
import numpy as np
from mlagents.tf_utils import tf
import yaml
from mlagents.trainers.ppo.trainer import PPOTrainer, discount_rewards
from mlagents.trainers.ppo.optimizer import PPOOptimizer
from mlagents.trainers.policy.nn_policy import NNPolicy
from mlagents.trainers.brain import BrainParameters
from mlagents.trainers.agent_processor import AgentManagerQueue
from mlagents.trainers.tests import mock_brain as mb
from mlagents.trainers.tests.mock_brain import make_brain_parameters
from mlagents.trainers.tests.test_trajectory import make_fake_trajectory
from mlagents.trainers.exception import UnityTrainerException
from mlagents.trainers.tests.test_reward_signals import ( # noqa: F401; pylint: disable=unused-variable
curiosity_dummy_config,
gail_dummy_config,
)
@pytest.fixture
def dummy_config():
return yaml.safe_load(
"""
trainer: ppo
batch_size: 32
beta: 5.0e-3
buffer_size: 512
epsilon: 0.2
hidden_units: 128
lambd: 0.95
learning_rate: 3.0e-4
max_steps: 5.0e4
normalize: true
num_epoch: 5
num_layers: 2
time_horizon: 64
sequence_length: 16
summary_freq: 1000
use_recurrent: false
normalize: true
memory_size: 10
curiosity_strength: 0.0
curiosity_enc_size: 1
summary_path: test
model_path: test
reward_signals:
extrinsic:
strength: 1.0
gamma: 0.99
"""
)
VECTOR_ACTION_SPACE = [2]
VECTOR_OBS_SPACE = 8
DISCRETE_ACTION_SPACE = [3, 3, 3, 2]
BUFFER_INIT_SAMPLES = 64
NUM_AGENTS = 12
def _create_ppo_optimizer_ops_mock(dummy_config, use_rnn, use_discrete, use_visual):
mock_brain = mb.setup_mock_brain(
use_discrete,
use_visual,
vector_action_space=VECTOR_ACTION_SPACE,
vector_obs_space=VECTOR_OBS_SPACE,
discrete_action_space=DISCRETE_ACTION_SPACE,
)
trainer_parameters = dummy_config
model_path = "testmodel"
trainer_parameters["model_path"] = model_path
trainer_parameters["keep_checkpoints"] = 3
trainer_parameters["use_recurrent"] = use_rnn
policy = NNPolicy(
0, mock_brain, trainer_parameters, False, False, create_tf_graph=False
)
optimizer = PPOOptimizer(policy, trainer_parameters)
return optimizer
def _create_fake_trajectory(use_discrete, use_visual, time_horizon):
if use_discrete:
act_space = DISCRETE_ACTION_SPACE
else:
act_space = VECTOR_ACTION_SPACE
if use_visual:
num_vis_obs = 1
vec_obs_size = 0
else:
num_vis_obs = 0
vec_obs_size = VECTOR_OBS_SPACE
trajectory = make_fake_trajectory(
length=time_horizon,
max_step_complete=True,
vec_obs_size=vec_obs_size,
num_vis_obs=num_vis_obs,
action_space=act_space,
)
return trajectory
@pytest.mark.parametrize("discrete", [True, False], ids=["discrete", "continuous"])
@pytest.mark.parametrize("visual", [True, False], ids=["visual", "vector"])
@pytest.mark.parametrize("rnn", [True, False], ids=["rnn", "no_rnn"])
def test_ppo_optimizer_update(dummy_config, rnn, visual, discrete):
# Test evaluate
tf.reset_default_graph()
optimizer = _create_ppo_optimizer_ops_mock(
dummy_config, use_rnn=rnn, use_discrete=discrete, use_visual=visual
)
# Test update
update_buffer = mb.simulate_rollout(BUFFER_INIT_SAMPLES, optimizer.policy.brain)
# Mock out reward signal eval
update_buffer["advantages"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_returns"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_value_estimates"] = update_buffer["environment_rewards"]
optimizer.update(
update_buffer,
num_sequences=update_buffer.num_experiences // optimizer.policy.sequence_length,
)
@pytest.mark.parametrize("discrete", [True, False], ids=["discrete", "continuous"])
@pytest.mark.parametrize("visual", [True, False], ids=["visual", "vector"])
@pytest.mark.parametrize("rnn", [True, False], ids=["rnn", "no_rnn"])
# We need to test this separately from test_reward_signals.py to ensure no interactions
def test_ppo_optimizer_update_curiosity(
curiosity_dummy_config, dummy_config, rnn, visual, discrete # noqa: F811
):
# Test evaluate
tf.reset_default_graph()
dummy_config["reward_signals"].update(curiosity_dummy_config)
optimizer = _create_ppo_optimizer_ops_mock(
dummy_config, use_rnn=rnn, use_discrete=discrete, use_visual=visual
)
# Test update
update_buffer = mb.simulate_rollout(BUFFER_INIT_SAMPLES, optimizer.policy.brain)
# Mock out reward signal eval
update_buffer["advantages"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_returns"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_value_estimates"] = update_buffer["environment_rewards"]
update_buffer["curiosity_returns"] = update_buffer["environment_rewards"]
update_buffer["curiosity_value_estimates"] = update_buffer["environment_rewards"]
optimizer.update(
update_buffer,
num_sequences=update_buffer.num_experiences // optimizer.policy.sequence_length,
)
# We need to test this separately from test_reward_signals.py to ensure no interactions
def test_ppo_optimizer_update_gail(gail_dummy_config, dummy_config): # noqa: F811
# Test evaluate
tf.reset_default_graph()
dummy_config["reward_signals"].update(gail_dummy_config)
optimizer = _create_ppo_optimizer_ops_mock(
dummy_config, use_rnn=False, use_discrete=False, use_visual=False
)
# Test update
update_buffer = mb.simulate_rollout(BUFFER_INIT_SAMPLES, optimizer.policy.brain)
# Mock out reward signal eval
update_buffer["advantages"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_returns"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_value_estimates"] = update_buffer["environment_rewards"]
update_buffer["gail_returns"] = update_buffer["environment_rewards"]
update_buffer["gail_value_estimates"] = update_buffer["environment_rewards"]
optimizer.update(
update_buffer,
num_sequences=update_buffer.num_experiences // optimizer.policy.sequence_length,
)
# Check if buffer size is too big
update_buffer = mb.simulate_rollout(3000, optimizer.policy.brain)
# Mock out reward signal eval
update_buffer["advantages"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_returns"] = update_buffer["environment_rewards"]
update_buffer["extrinsic_value_estimates"] = update_buffer["environment_rewards"]
update_buffer["gail_returns"] = update_buffer["environment_rewards"]
update_buffer["gail_value_estimates"] = update_buffer["environment_rewards"]
optimizer.update(
update_buffer,
num_sequences=update_buffer.num_experiences // optimizer.policy.sequence_length,
)
@pytest.mark.parametrize("discrete", [True, False], ids=["discrete", "continuous"])
@pytest.mark.parametrize("visual", [True, False], ids=["visual", "vector"])
@pytest.mark.parametrize("rnn", [True, False], ids=["rnn", "no_rnn"])
def test_ppo_get_value_estimates(dummy_config, rnn, visual, discrete):
tf.reset_default_graph()
optimizer = _create_ppo_optimizer_ops_mock(
dummy_config, use_rnn=rnn, use_discrete=discrete, use_visual=visual
)
time_horizon = 15
trajectory = _create_fake_trajectory(discrete, visual, time_horizon)
run_out, final_value_out = optimizer.get_trajectory_value_estimates(
trajectory.to_agentbuffer(), trajectory.next_obs, done=False
)
for key, val in run_out.items():
assert type(key) is str
assert len(val) == 15
run_out, final_value_out = optimizer.get_trajectory_value_estimates(
trajectory.to_agentbuffer(), trajectory.next_obs, done=True
)
for key, val in final_value_out.items():
assert type(key) is str
assert val == 0.0
# Check if we ignore terminal states properly
optimizer.reward_signals["extrinsic"].use_terminal_states = False
run_out, final_value_out = optimizer.get_trajectory_value_estimates(
trajectory.to_agentbuffer(), trajectory.next_obs, done=False
)
for key, val in final_value_out.items():
assert type(key) is str
assert val != 0.0
def test_rl_functions():
rewards = np.array([0.0, 0.0, 0.0, 1.0], dtype=np.float32)
gamma = 0.9
returns = discount_rewards(rewards, gamma, 0.0)
np.testing.assert_array_almost_equal(
returns, np.array([0.729, 0.81, 0.9, 1.0], dtype=np.float32)
)
@mock.patch("mlagents.trainers.ppo.trainer.PPOOptimizer")
def test_trainer_increment_step(ppo_optimizer, dummy_config):
trainer_params = dummy_config
mock_optimizer = mock.Mock()
mock_optimizer.reward_signals = {}
ppo_optimizer.return_value = mock_optimizer
brain_params = BrainParameters(
brain_name="test_brain",
vector_observation_space_size=1,
camera_resolutions=[],
vector_action_space_size=[2],
vector_action_descriptions=[],
vector_action_space_type=0,
)
trainer = PPOTrainer(
brain_params.brain_name, 0, trainer_params, True, False, 0, "0"
)
policy_mock = mock.Mock(spec=NNPolicy)
policy_mock.get_current_step.return_value = 0
step_count = (
5
) # 10 hacked because this function is no longer called through trainer
policy_mock.increment_step = mock.Mock(return_value=step_count)
trainer.add_policy("testbehavior", policy_mock)
trainer._increment_step(5, "testbehavior")
policy_mock.increment_step.assert_called_with(5)
assert trainer.step == step_count
@pytest.mark.parametrize("use_discrete", [True, False])
def test_trainer_update_policy(dummy_config, use_discrete):
mock_brain = mb.setup_mock_brain(
use_discrete,
False,
vector_action_space=VECTOR_ACTION_SPACE,
vector_obs_space=VECTOR_OBS_SPACE,
discrete_action_space=DISCRETE_ACTION_SPACE,
)
trainer_params = dummy_config
trainer_params["use_recurrent"] = True
# Test curiosity reward signal
trainer_params["reward_signals"]["curiosity"] = {}
trainer_params["reward_signals"]["curiosity"]["strength"] = 1.0
trainer_params["reward_signals"]["curiosity"]["gamma"] = 0.99
trainer_params["reward_signals"]["curiosity"]["encoding_size"] = 128
trainer = PPOTrainer(mock_brain.brain_name, 0, trainer_params, True, False, 0, "0")
policy = trainer.create_policy(mock_brain.brain_name, mock_brain)
trainer.add_policy(mock_brain.brain_name, policy)
# Test update with sequence length smaller than batch size
buffer = mb.simulate_rollout(BUFFER_INIT_SAMPLES, mock_brain)
# Mock out reward signal eval
buffer["extrinsic_rewards"] = buffer["environment_rewards"]
buffer["extrinsic_returns"] = buffer["environment_rewards"]
buffer["extrinsic_value_estimates"] = buffer["environment_rewards"]
buffer["curiosity_rewards"] = buffer["environment_rewards"]
buffer["curiosity_returns"] = buffer["environment_rewards"]
buffer["curiosity_value_estimates"] = buffer["environment_rewards"]
buffer["advantages"] = buffer["environment_rewards"]
trainer.update_buffer = buffer
trainer._update_policy()
def test_process_trajectory(dummy_config):
brain_params = BrainParameters(
brain_name="test_brain",
vector_observation_space_size=1,
camera_resolutions=[],
vector_action_space_size=[2],
vector_action_descriptions=[],
vector_action_space_type=0,
)
dummy_config["summary_path"] = "./summaries/test_trainer_summary"
dummy_config["model_path"] = "./models/test_trainer_models/TestModel"
trainer = PPOTrainer(brain_params, 0, dummy_config, True, False, 0, "0")
policy = trainer.create_policy(brain_params.brain_name, brain_params)
trainer.add_policy(brain_params.brain_name, policy)
trajectory_queue = AgentManagerQueue("testbrain")
trainer.subscribe_trajectory_queue(trajectory_queue)
time_horizon = 15
trajectory = make_fake_trajectory(
length=time_horizon,
max_step_complete=True,
vec_obs_size=1,
num_vis_obs=0,
action_space=[2],
)
trajectory_queue.put(trajectory)
trainer.advance()
# Check that trainer put trajectory in update buffer
assert trainer.update_buffer.num_experiences == 15
# Check that GAE worked
assert (
"advantages" in trainer.update_buffer
and "discounted_returns" in trainer.update_buffer
)
# Check that the stats are being collected as episode isn't complete
for reward in trainer.collected_rewards.values():
for agent in reward.values():
assert agent > 0
# Add a terminal trajectory
trajectory = make_fake_trajectory(
length=time_horizon + 1,
max_step_complete=False,
vec_obs_size=1,
num_vis_obs=0,
action_space=[2],
)
trajectory_queue.put(trajectory)
trainer.advance()
# Check that the stats are reset as episode is finished
for reward in trainer.collected_rewards.values():
for agent in reward.values():
assert agent == 0
assert trainer.stats_reporter.get_stats_summaries("Policy/Extrinsic Reward").num > 0
@mock.patch("mlagents.trainers.ppo.trainer.PPOOptimizer")
def test_add_get_policy(ppo_optimizer, dummy_config):
brain_params = make_brain_parameters(
discrete_action=False, visual_inputs=0, vec_obs_size=6
)
mock_optimizer = mock.Mock()
mock_optimizer.reward_signals = {}
ppo_optimizer.return_value = mock_optimizer
dummy_config["summary_path"] = "./summaries/test_trainer_summary"
dummy_config["model_path"] = "./models/test_trainer_models/TestModel"
trainer = PPOTrainer(brain_params, 0, dummy_config, True, False, 0, "0")
policy = mock.Mock(spec=NNPolicy)
policy.get_current_step.return_value = 2000
trainer.add_policy(brain_params.brain_name, policy)
assert trainer.get_policy(brain_params.brain_name) == policy
# Make sure the summary steps were loaded properly
assert trainer.get_step == 2000
assert trainer.next_summary_step > 2000
# Test incorrect class of policy
policy = mock.Mock()
with pytest.raises(RuntimeError):
trainer.add_policy(brain_params, policy)
def test_bad_config(dummy_config):
brain_params = make_brain_parameters(
discrete_action=False, visual_inputs=0, vec_obs_size=6
)
# Test that we throw an error if we have sequence length greater than batch size
dummy_config["sequence_length"] = 64
dummy_config["batch_size"] = 32
dummy_config["use_recurrent"] = True
with pytest.raises(UnityTrainerException):
_ = PPOTrainer(brain_params, 0, dummy_config, True, False, 0, "0")
if __name__ == "__main__":
pytest.main()