def create_learner(self, env, parameters):
if (self.trpo() or self.ppo()) and not issubclass(type(env), VecEnv):
env = DummyVecEnv([lambda: env])
if self.trpo():
model = TRPO(MlpPolicy, env, **parameters["common"],
**parameters[str(self)])
interface = TRPOInterface(model, env.observation_space.shape[0])
elif self.ppo():
model = PPO2(MlpPolicy, env, **parameters["common"],
**parameters[str(self)])
interface = PPOInterface(model, env.observation_space.shape[0])
else:
model = SAC(SACMlpPolicy, env, **parameters["common"],
**parameters[str(self)])
interface = SACInterface(model, env.observation_space.shape[0])
if "pretrain_data_path" in parameters:
data_path = parameters["pretrain_data_path"]
model.pretrain(ExpertDataset(expert_path=data_path, verbose=0),
n_epochs=25)
return model, interface
class GAIL(ActorCriticRLModel):
"""
Generative Adversarial Imitation Learning (GAIL)
.. warning::
Images are not yet handled properly by the current implementation
:param policy: (ActorCriticPolicy or str) The policy model to use (MlpPolicy, CnnPolicy, CnnLstmPolicy, ...)
:param env: (Gym environment or str) The environment to learn from (if registered in Gym, can be str)
:param expert_dataset: (ExpertDataset) the dataset manager
:param gamma: (float) the discount value
:param timesteps_per_batch: (int) the number of timesteps to run per batch (horizon)
:param max_kl: (float) the kullback leiber loss threashold
:param cg_iters: (int) the number of iterations for the conjugate gradient calculation
:param lam: (float) GAE factor
:param entcoeff: (float) the weight for the entropy loss
:param cg_damping: (float) the compute gradient dampening factor
:param vf_stepsize: (float) the value function stepsize
:param vf_iters: (int) the value function's number iterations for learning
:param hidden_size: ([int]) the hidden dimension for the MLP
:param g_step: (int) number of steps to train policy in each epoch
:param d_step: (int) number of steps to train discriminator in each epoch
:param d_stepsize: (float) the reward giver stepsize
:param verbose: (int) the verbosity level: 0 none, 1 training information, 2 tensorflow debug
:param _init_setup_model: (bool) Whether or not to build the network at the creation of the instance
:param full_tensorboard_log: (bool) enable additional logging when using tensorboard
WARNING: this logging can take a lot of space quickly
"""
def __init__(self, policy, env, expert_dataset=None,
hidden_size_adversary=100, adversary_entcoeff=1e-3,
g_step=3, d_step=1, d_stepsize=3e-4, verbose=0,
_init_setup_model=True, **kwargs):
super().__init__(policy=policy, env=env, verbose=verbose, requires_vec_env=False,
_init_setup_model=_init_setup_model)
self.trpo = TRPO(policy, env, verbose=verbose, _init_setup_model=False, **kwargs)
self.trpo.using_gail = True
self.trpo.expert_dataset = expert_dataset
self.trpo.g_step = g_step
self.trpo.d_step = d_step
self.trpo.d_stepsize = d_stepsize
self.trpo.hidden_size_adversary = hidden_size_adversary
self.trpo.adversary_entcoeff = adversary_entcoeff
self.env = self.trpo.env
if _init_setup_model:
self.setup_model()
def _get_pretrain_placeholders(self):
pass
def pretrain(self, dataset, n_epochs=10, learning_rate=1e-4,
adam_epsilon=1e-8, val_interval=None):
self.trpo.pretrain(dataset, n_epochs=n_epochs, learning_rate=learning_rate,
adam_epsilon=adam_epsilon, val_interval=val_interval)
return self
def set_env(self, env):
self.trpo.set_env(env)
self.env = self.trpo.env
def setup_model(self):
assert issubclass(self.policy, ActorCriticPolicy), "Error: the input policy for the GAIL model must be an " \
"instance of common.policies.ActorCriticPolicy."
self.trpo.setup_model()
def learn(self, total_timesteps, callback=None, seed=None, log_interval=100, tb_log_name="GAIL",
reset_num_timesteps=True):
assert self.trpo.expert_dataset is not None, "You must pass an expert dataset to GAIL for training"
self.trpo.learn(total_timesteps, callback, seed, log_interval, tb_log_name, reset_num_timesteps)
return self
def predict(self, observation, state=None, mask=None, deterministic=False):
return self.trpo.predict(observation, state=state, mask=mask, deterministic=deterministic)
def action_probability(self, observation, state=None, mask=None, actions=None):
return self.trpo.action_probability(observation, state=state, mask=mask, actions=actions)
def save(self, save_path):
self.trpo.save(save_path)
@classmethod
def load(cls, load_path, env=None, **kwargs):
data, params = cls._load_from_file(load_path)
model = cls(policy=data["policy"], env=None, _init_setup_model=False)
model.trpo.__dict__.update(data)
model.trpo.__dict__.update(kwargs)
model.set_env(env)
model.setup_model()
restores = []
for param, loaded_p in zip(model.trpo.params, params):
restores.append(param.assign(loaded_p))
model.trpo.sess.run(restores)
return model
