def train_and_assert(self, agent_type, is_v1: bool, num_iterations=100):
logger = logging.warning
v2_backends = [b for b in get_backends(agent_type, skip_v1=True)]
v1_backends = [
b for b in get_backends(agent_type) if (not b in v2_backends)
]
backends = v1_backends if is_v1 else v2_backends
for backend in backends:
logger(
f'backend={backend} agent={agent_type}, num_iterations={num_iterations}'
)
cem_agent: CemAgent = agent_type('CartPole-v0',
fc_layers=(100, ),
backend=backend)
tc: core.TrainContext = cem_agent.train(
[log.Duration(),
log.Iteration(eval_only=True),
log.Agent()],
num_iterations=num_iterations,
num_iterations_between_eval=10,
max_steps_per_episode=200,
default_plots=False)
(min_steps, avg_steps,
max_steps) = tc.eval_steps[tc.episodes_done_in_training]
assert max_steps >= 100
assert avg_steps >= 50
def train_and_assert(self, agent_type, is_v1: bool, num_iterations=10000):
logger = logging.warning
v2_backends = [b for b in get_backends(agent_type, skip_v1=True) if b != 'default']
v1_backends = [b for b in get_backends(agent_type) if (not b in v2_backends) and b != 'default']
backends = v1_backends if is_v1 else v2_backends
for backend in backends:
current_num_iterations = num_iterations
if backend == 'tensorforce':
current_num_iterations = num_iterations * 3
logger(f'backend={backend} agent={agent_type}, num_iterations={current_num_iterations}')
max_avg_steps = self.train_and_eval(agent_type=agent_type, backend=backend,
num_iterations=current_num_iterations)
assert max_avg_steps >= 80, f'agent_type={agent_type} backend={backend} num_iterations={num_iterations}'
def test_train(self):
for backend in get_backends(SacAgent):
sac_agent: SacAgent = SacAgent('CartPole-v0', backend=backend)
sac_agent.train([log.Duration(), log.Iteration(), log.Agent()],
num_iterations=10,
max_steps_per_episode=200,
default_plots=False)
def test_train_single_episode(self):
for backend in get_backends(PpoAgent):
ppo = agents.PpoAgent(gym_env_name=_env_name, backend=backend)
count = log._CallbackCounts()
ppo.train([log.Agent(), count, duration._SingleEpisode()])
assert count.gym_init_begin_count == count.gym_init_end_count == 1
assert count.gym_step_begin_count == count.gym_step_end_count
assert count.gym_step_begin_count < 10 + count.gym_reset_begin_count
def test_train(self):
for backend in get_backends(RandomAgent):
reinforce_agent: ReinforceAgent = ReinforceAgent('CartPole-v0', backend=backend)
tc: core.TrainContext = reinforce_agent.train([log.Duration(), log.Iteration()],
num_iterations=10,
max_steps_per_episode=200,
default_plots=False)
(min_steps, avg_steps, max_steps) = tc.eval_steps[tc.episodes_done_in_training]
assert avg_steps >= 10
def test_play_single_episode(self):
for backend in get_backends(PpoAgent):
ppo = agents.PpoAgent(gym_env_name=_env_name, backend=backend)
count = log._CallbackCounts()
cb = [log.Agent(), count, duration._SingleEpisode()]
ppo.train(duration._SingleEpisode())
ppo.play(cb)
assert count.gym_init_begin_count == count.gym_init_end_count == 1
assert count.gym_step_begin_count == count.gym_step_end_count <= 10
def test_train_cartpole(self):
for backend in get_backends(PpoAgent):
ppo = PpoAgent(gym_env_name="CartPole-v0", backend=backend)
tc = core.PpoTrainContext()
tc.num_iterations = 3
tc.num_episodes_per_iteration = 10
tc.max_steps_per_episode = 500
tc.num_epochs_per_iteration = 5
tc.num_iterations_between_eval = 2
tc.num_episodes_per_eval = 5
ppo.train([log.Iteration()], train_context=tc)
def test_getbackends_randomagent(self):
assert agents._backends is not None
backends = agents.get_backends(agents.RandomAgent)
assert 'default' in backends
assert 'tfagents' in backends
assert 'tensorforce' in backends
def test_getbackends(self):
assert agents._backends is not None
assert agents.get_backends() is not None
def test_callback_single(self):
for backend in get_backends(PpoAgent):
env._StepCountEnv.clear()
agent = PpoAgent(_env_name, backend=backend)
agent.train(duration._SingleEpisode())
assert env._StepCountEnv.reset_count <= 2
def test_getbackends_ppoagent(self):
assert agents._backends is not None
backends = agents.get_backends(agents.PpoAgent)
assert 'default' in backends
assert 'tfagents' in backends
def get_backends(agent: Optional[Type[EasyAgent]] = None):
result = [b for b in agents.get_backends(agent) if b != 'default']
assert result, f'no backend found for agent {agent}.'
return result
