def train_eval_online(
# Basic args.
log_dir,
agent_module,
env_name='HalfCheetah-v2',
# Train and eval args.
total_train_steps=int(1e6),
summary_freq=100,
print_freq=1000,
save_freq=int(1e8),
eval_freq=5000,
n_eval_episodes=20,
# For saving a partially trained policy.
eval_target=None, # Target return value to stop training.
eval_target_n=2, # Stop after n consecutive evals above eval_target.
# Agent train args.
initial_explore_steps=10000,
replay_buffer_size=int(1e6),
model_params=(((200, 200),), 2),
optimizers=(('adam', 0.001),),
batch_size=256,
weight_decays=(0.0,),
update_freq=1,
update_rate=0.005,
discount=0.99,
):
"""Training a policy with online interaction."""
# Create tf_env to get specs.
tf_env = train_eval_utils.env_factory(env_name)
tf_env_test = train_eval_utils.env_factory(env_name)
observation_spec = tf_env.observation_spec()
action_spec = tf_env.action_spec()
# Initialize dataset.
with tf.device('/cpu:0'):
train_data = dataset.Dataset(
observation_spec,
action_spec,
replay_buffer_size,
circular=True,
)
data_ckpt = tf.train.Checkpoint(data=train_data)
data_ckpt_name = os.path.join(log_dir, 'replay')
time_st_total = time.time()
time_st = time.time()
timed_at_step = 0
# Collect data from random policy.
explore_policy = policies.ContinuousRandomPolicy(action_spec)
steps_collected = 0
log_freq = 5000
logging.info('Collecting data ...')
collector = train_eval_utils.DataCollector(tf_env, explore_policy, train_data)
while steps_collected < initial_explore_steps:
count = collector.collect_transition()
steps_collected += count
if (steps_collected % log_freq == 0
or steps_collected == initial_explore_steps) and count > 0:
steps_per_sec = ((steps_collected - timed_at_step)
/ (time.time() - time_st))
timed_at_step = steps_collected
time_st = time.time()
logging.info('(%d/%d) steps collected at %.4g steps/s.', steps_collected,
initial_explore_steps, steps_per_sec)
# Construct agent.
agent_flags = utils.Flags(
action_spec=action_spec,
model_params=model_params,
optimizers=optimizers,
batch_size=batch_size,
weight_decays=weight_decays,
update_freq=update_freq,
update_rate=update_rate,
discount=discount,
train_data=train_data)
agent_args = agent_module.Config(agent_flags).agent_args
agent = agent_module.Agent(**vars(agent_args))
# Prepare savers for models and results.
train_summary_dir = os.path.join(log_dir, 'train')
eval_summary_dir = os.path.join(log_dir, 'eval')
train_summary_writer = tf.compat.v2.summary.create_file_writer(
train_summary_dir)
eval_summary_writers = collections.OrderedDict()
for policy_key in agent.test_policies.keys():
eval_summary_writer = tf.compat.v2.summary.create_file_writer(
os.path.join(eval_summary_dir, policy_key))
eval_summary_writers[policy_key] = eval_summary_writer
agent_ckpt_name = os.path.join(log_dir, 'agent')
eval_results = []
# Train agent.
logging.info('Start training ....')
time_st = time.time()
timed_at_step = 0
target_partial_policy_saved = False
collector = train_eval_utils.DataCollector(
tf_env, agent.online_policy, train_data)
for _ in range(total_train_steps):
collector.collect_transition()
agent.train_step()
step = agent.global_step
if step % summary_freq == 0 or step == total_train_steps:
agent.write_train_summary(train_summary_writer)
if step % print_freq == 0 or step == total_train_steps:
agent.print_train_info()
if step % eval_freq == 0 or step == total_train_steps:
time_ed = time.time()
time_cost = time_ed - time_st
logging.info(
'Training at %.4g steps/s.', (step - timed_at_step) / time_cost)
eval_result, eval_infos = train_eval_utils.eval_policies(
tf_env_test, agent.test_policies, n_eval_episodes)
eval_results.append([step] + eval_result)
# Cecide whether to save a partially trained policy based on current model
# performance.
if (eval_target is not None and len(eval_results) >= eval_target_n
and not target_partial_policy_saved):
evals_ = list([eval_results[-(i + 1)][1]
for i in range(eval_target_n)])
evals_ = np.array(evals_)
if np.min(evals_) >= eval_target:
agent.save(agent_ckpt_name + '_partial_target')
dataset.save_copy(train_data, data_ckpt_name + '_partial_target')
logging.info('A partially trained policy was saved at step %d,'
' with episodic return %.4g.', step, evals_[-1])
target_partial_policy_saved = True
logging.info('Testing at step %d:', step)
for policy_key, policy_info in eval_infos.items():
logging.info(utils.get_summary_str(
step=None, info=policy_info, prefix=policy_key + ': '))
utils.write_summary(eval_summary_writers[policy_key], step, policy_info)
time_st = time.time()
timed_at_step = step
if step % save_freq == 0:
agent.save(agent_ckpt_name + '-' + str(step))
# Final save after training.
agent.save(agent_ckpt_name + '_final')
data_ckpt.write(data_ckpt_name + '_final')
time_cost = time.time() - time_st_total
logging.info('Training finished, time cost %.4gs.', time_cost)
return np.array(eval_results)
def print_train_info(self): info = self._train_info step = self._global_step.numpy() summary_str = utils.get_summary_str(step, info) logging.info(summary_str)
def train_eval_offline(
# Basic args.
log_dir,
data_file,
agent_module,
env_name='HalfCheetah-v2',
n_train=int(1e6),
shuffle_steps=0,
seed=0,
use_seed_for_data=False,
# Train and eval args.
total_train_steps=int(1e6),
summary_freq=100,
print_freq=1000,
save_freq=int(2e4),
eval_freq=5000,
n_eval_episodes=20,
# Agent args.
model_params=(((200, 200), ), 2),
behavior_ckpt_file=None,
value_penalty=True,
alpha=1.0,
#model_params=((200, 200),),
optimizers=(('adam', 0.001), ),
batch_size=256,
weight_decays=(0.0, ),
update_freq=1,
update_rate=0.005,
discount=0.99,
):
"""Training a policy with a fixed dataset."""
# Create tf_env to get specs.
print('[train_eval_offline.py] env_name=', env_name)
print('[train_eval_offline.py] data_file=', data_file)
print('[train_eval_offline.py] agent_module=', agent_module)
print('[train_eval_offline.py] model_params=', model_params)
print('[train_eval_offline.py] optimizers=', optimizers)
print('[train_eval_offline.py] bckpt_file=', behavior_ckpt_file)
print('[train_eval_offline.py] value_penalty=', value_penalty)
tf_env = env_factory(env_name)
observation_spec = tf_env.observation_spec()
action_spec = tf_env.action_spec()
# Prepare data.
full_data = get_offline_data(tf_env)
# Split data.
n_train = min(n_train, full_data.size)
logging.info('n_train %s.', n_train)
if use_seed_for_data:
rand = np.random.RandomState(seed)
else:
rand = np.random.RandomState(0)
shuffled_indices = utils.shuffle_indices_with_steps(n=full_data.size,
steps=shuffle_steps,
rand=rand)
train_indices = shuffled_indices[:n_train]
train_data = full_data.create_view(train_indices)
# Create agent.
agent_flags = utils.Flags(observation_spec=observation_spec,
action_spec=action_spec,
model_params=model_params,
optimizers=optimizers,
batch_size=batch_size,
weight_decays=weight_decays,
update_freq=update_freq,
update_rate=update_rate,
discount=discount,
train_data=train_data)
agent_args = agent_module.Config(agent_flags).agent_args
my_agent_arg_dict = {}
for k in vars(agent_args):
my_agent_arg_dict[k] = vars(agent_args)[k]
if 'brac_primal' in agent_module.__name__:
my_agent_arg_dict['behavior_ckpt_file'] = behavior_ckpt_file
my_agent_arg_dict['value_penalty'] = value_penalty
my_agent_arg_dict['alpha'] = alpha
print('agent:', agent_module.__name__)
print('agent_args:', my_agent_arg_dict)
#agent = agent_module.Agent(**vars(agent_args))
agent = agent_module.Agent(**my_agent_arg_dict)
agent_ckpt_name = os.path.join(log_dir, 'agent')
# Restore agent from checkpoint if there exists one.
if tf.io.gfile.exists('{}.index'.format(agent_ckpt_name)):
logging.info('Checkpoint found at %s.', agent_ckpt_name)
agent.restore(agent_ckpt_name)
# Train agent.
train_summary_dir = os.path.join(log_dir, 'train')
eval_summary_dir = os.path.join(log_dir, 'eval')
train_summary_writer = tf0.compat.v2.summary.create_file_writer(
train_summary_dir)
eval_summary_writers = collections.OrderedDict()
for policy_key in agent.test_policies.keys():
eval_summary_writer = tf0.compat.v2.summary.create_file_writer(
os.path.join(eval_summary_dir, policy_key))
eval_summary_writers[policy_key] = eval_summary_writer
eval_results = []
time_st_total = time.time()
time_st = time.time()
step = agent.global_step
timed_at_step = step
while step < total_train_steps:
agent.train_step()
step = agent.global_step
if step % summary_freq == 0 or step == total_train_steps:
agent.write_train_summary(train_summary_writer)
if step % print_freq == 0 or step == total_train_steps:
agent.print_train_info()
if step % eval_freq == 0 or step == total_train_steps:
time_ed = time.time()
time_cost = time_ed - time_st
logging.info('Training at %.4g steps/s.',
(step - timed_at_step) / time_cost)
eval_result, eval_infos = train_eval_utils.eval_policies(
tf_env, agent.test_policies, n_eval_episodes)
eval_results.append([step] + eval_result)
with open(os.path.join(log_dir, 'results.txt'), 'a') as logfile:
logfile.write(str(eval_result) + '\n')
logging.info('Testing at step %d:', step)
for policy_key, policy_info in eval_infos.items():
logging.info(
utils.get_summary_str(step=None,
info=policy_info,
prefix=policy_key + ': '))
utils.write_summary(eval_summary_writers[policy_key], step,
policy_info)
time_st = time.time()
timed_at_step = step
if step % save_freq == 0:
agent.save(agent_ckpt_name)
logging.info('Agent saved at %s.', agent_ckpt_name)
agent.save(agent_ckpt_name)
time_cost = time.time() - time_st_total
logging.info('Training finished, time cost %.4gs.', time_cost)
return np.array(eval_results)
def train_eval_offline(
# Basic args.
log_dir,
data_file,
agent_module,
env_name='HalfCheetah-v2',
n_train=int(1e6),
shuffle_steps=0,
seed=0,
use_seed_for_data=False,
# Train and eval args.
total_train_steps=int(1e6),
summary_freq=100,
print_freq=1000,
save_freq=int(2e4),
eval_freq=5000,
n_eval_episodes=20,
# Agent args.
model_params=(((200, 200),), 2),
optimizers=(('adam', 0.001),),
batch_size=256,
weight_decays=(0.0,),
update_freq=1,
update_rate=0.005,
discount=0.99,
):
"""Training a policy with a fixed dataset."""
# Create tf_env to get specs.
tf_env = train_eval_utils.env_factory(env_name)
observation_spec = tf_env.observation_spec()
action_spec = tf_env.action_spec()
# Prepare data.
logging.info('Loading data from %s ...', data_file)
data_size = utils.load_variable_from_ckpt(data_file, 'data._capacity')
with tf.device('/cpu:0'):
full_data = dataset.Dataset(observation_spec, action_spec, data_size)
data_ckpt = tf.train.Checkpoint(data=full_data)
data_ckpt.restore(data_file)
# Split data.
n_train = min(n_train, full_data.size)
logging.info('n_train %s.', n_train)
if use_seed_for_data:
rand = np.random.RandomState(seed)
else:
rand = np.random.RandomState(0)
shuffled_indices = utils.shuffle_indices_with_steps(
n=full_data.size, steps=shuffle_steps, rand=rand)
train_indices = shuffled_indices[:n_train]
train_data = full_data.create_view(train_indices)
# Create agent.
agent_flags = utils.Flags(
observation_spec=observation_spec,
action_spec=action_spec,
model_params=model_params,
optimizers=optimizers,
batch_size=batch_size,
weight_decays=weight_decays,
update_freq=update_freq,
update_rate=update_rate,
discount=discount,
train_data=train_data)
agent_args = agent_module.Config(agent_flags).agent_args
agent = agent_module.Agent(**vars(agent_args))
agent_ckpt_name = os.path.join(log_dir, 'agent')
# Restore agent from checkpoint if there exists one.
if tf.io.gfile.exists('{}.index'.format(agent_ckpt_name)):
logging.info('Checkpoint found at %s.', agent_ckpt_name)
agent.restore(agent_ckpt_name)
# Train agent.
train_summary_dir = os.path.join(log_dir, 'train')
eval_summary_dir = os.path.join(log_dir, 'eval')
train_summary_writer = tf.compat.v2.summary.create_file_writer(
train_summary_dir)
eval_summary_writers = collections.OrderedDict()
for policy_key in agent.test_policies.keys():
eval_summary_writer = tf.compat.v2.summary.create_file_writer(
os.path.join(eval_summary_dir, policy_key))
eval_summary_writers[policy_key] = eval_summary_writer
eval_results = []
time_st_total = time.time()
time_st = time.time()
step = agent.global_step
timed_at_step = step
while step < total_train_steps:
agent.train_step()
step = agent.global_step
if step % summary_freq == 0 or step == total_train_steps:
agent.write_train_summary(train_summary_writer)
if step % print_freq == 0 or step == total_train_steps:
agent.print_train_info()
if step % eval_freq == 0 or step == total_train_steps:
time_ed = time.time()
time_cost = time_ed - time_st
logging.info(
'Training at %.4g steps/s.', (step - timed_at_step) / time_cost)
eval_result, eval_infos = train_eval_utils.eval_policies(
tf_env, agent.test_policies, n_eval_episodes)
eval_results.append([step] + eval_result)
logging.info('Testing at step %d:', step)
for policy_key, policy_info in eval_infos.items():
logging.info(utils.get_summary_str(
step=None, info=policy_info, prefix=policy_key+': '))
utils.write_summary(eval_summary_writers[policy_key], step, policy_info)
time_st = time.time()
timed_at_step = step
if step % save_freq == 0:
agent.save(agent_ckpt_name)
logging.info('Agent saved at %s.', agent_ckpt_name)
agent.save(agent_ckpt_name)
time_cost = time.time() - time_st_total
logging.info('Training finished, time cost %.4gs.', time_cost)
return np.array(eval_results)