mirror of
https://github.com/gryf/coach.git
synced 2025-12-17 11:10:20 +01:00
1b095aeeca
Till now, most of the modules were importing all of the module objects (variables, classes, functions, other imports) into module namespace, which potentially could (and was) cause of unintentional use of class or methods, which was indirect imported. With this patch, all the star imports were substituted with top-level module, which provides desired class or function. Besides, all imports where sorted (where possible) in a way pep8[1] suggests - first are imports from standard library, than goes third party imports (like numpy, tensorflow etc) and finally coach modules. All of those sections are separated by one empty line. [1] https://www.python.org/dev/peps/pep-0008/#imports
298 lines
14 KiB
Python
298 lines
14 KiB
Python
#
|
|
# Copyright (c) 2017 Intel Corporation
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
#
|
|
import collections
|
|
import copy
|
|
|
|
import numpy as np
|
|
|
|
from agents import actor_critic_agent as aca
|
|
from agents import policy_optimization_agent as poa
|
|
from architectures import network_wrapper as nw
|
|
import configurations
|
|
import logger
|
|
import utils
|
|
|
|
|
|
# Proximal Policy Optimization - https://arxiv.org/pdf/1707.06347.pdf
|
|
class PPOAgent(aca.ActorCriticAgent):
|
|
def __init__(self, env, tuning_parameters, replicated_device=None, thread_id=0):
|
|
aca.ActorCriticAgent.__init__(self, env, tuning_parameters, replicated_device, thread_id,
|
|
create_target_network=True)
|
|
self.critic_network = self.main_network
|
|
|
|
# define the policy network
|
|
tuning_parameters.agent.input_types = {'observation': configurations.InputTypes.Observation}
|
|
tuning_parameters.agent.output_types = [configurations.OutputTypes.PPO]
|
|
tuning_parameters.agent.optimizer_type = 'Adam'
|
|
tuning_parameters.agent.l2_regularization = 0
|
|
self.policy_network = nw.NetworkWrapper(tuning_parameters, True, self.has_global, 'policy',
|
|
self.replicated_device, self.worker_device)
|
|
self.networks.append(self.policy_network)
|
|
|
|
# signals definition
|
|
self.value_loss = utils.Signal('Value Loss')
|
|
self.signals.append(self.value_loss)
|
|
self.policy_loss = utils.Signal('Policy Loss')
|
|
self.signals.append(self.policy_loss)
|
|
self.kl_divergence = utils.Signal('KL Divergence')
|
|
self.signals.append(self.kl_divergence)
|
|
self.total_kl_divergence_during_training_process = 0.0
|
|
self.unclipped_grads = utils.Signal('Grads (unclipped)')
|
|
self.signals.append(self.unclipped_grads)
|
|
|
|
self.reset_game(do_not_reset_env=True)
|
|
|
|
def fill_advantages(self, batch):
|
|
current_states, next_states, actions, rewards, game_overs, total_return = self.extract_batch(batch)
|
|
|
|
# * Found not to have any impact *
|
|
# current_states_with_timestep = self.concat_state_and_timestep(batch)
|
|
|
|
current_state_values = self.critic_network.online_network.predict(current_states).squeeze()
|
|
|
|
# calculate advantages
|
|
advantages = []
|
|
if self.policy_gradient_rescaler == poa.PolicyGradientRescaler.A_VALUE:
|
|
advantages = total_return - current_state_values
|
|
elif self.policy_gradient_rescaler == poa.PolicyGradientRescaler.GAE:
|
|
# get bootstraps
|
|
episode_start_idx = 0
|
|
advantages = np.array([])
|
|
# current_state_values[game_overs] = 0
|
|
for idx, game_over in enumerate(game_overs):
|
|
if game_over:
|
|
# get advantages for the rollout
|
|
value_bootstrapping = np.zeros((1,))
|
|
rollout_state_values = np.append(current_state_values[episode_start_idx:idx+1], value_bootstrapping)
|
|
|
|
rollout_advantages, _ = \
|
|
self.get_general_advantage_estimation_values(rewards[episode_start_idx:idx+1],
|
|
rollout_state_values)
|
|
episode_start_idx = idx + 1
|
|
advantages = np.append(advantages, rollout_advantages)
|
|
else:
|
|
logger.screen.warning("WARNING: The requested policy gradient rescaler is not available")
|
|
|
|
# standardize
|
|
advantages = (advantages - np.mean(advantages)) / np.std(advantages)
|
|
|
|
for transition, advantage in zip(self.memory.transitions, advantages):
|
|
transition.info['advantage'] = advantage
|
|
|
|
self.action_advantages.add_sample(advantages)
|
|
|
|
def train_value_network(self, dataset, epochs):
|
|
loss = []
|
|
current_states, _, _, _, _, total_return = self.extract_batch(dataset)
|
|
|
|
# * Found not to have any impact *
|
|
# add a timestep to the observation
|
|
# current_states_with_timestep = self.concat_state_and_timestep(dataset)
|
|
|
|
total_return = np.expand_dims(total_return, -1)
|
|
mix_fraction = self.tp.agent.value_targets_mix_fraction
|
|
for j in range(epochs):
|
|
batch_size = len(dataset)
|
|
if self.critic_network.online_network.optimizer_type != 'LBFGS':
|
|
batch_size = self.tp.batch_size
|
|
for i in range(len(dataset) // batch_size):
|
|
# split to batches for first order optimization techniques
|
|
current_states_batch = {
|
|
k: v[i * batch_size:(i + 1) * batch_size]
|
|
for k, v in current_states.items()
|
|
}
|
|
total_return_batch = total_return[i * batch_size:(i + 1) * batch_size]
|
|
old_policy_values = utils.force_list(self.critic_network.target_network.predict(
|
|
current_states_batch).squeeze())
|
|
if self.critic_network.online_network.optimizer_type != 'LBFGS':
|
|
targets = total_return_batch
|
|
else:
|
|
current_values = self.critic_network.online_network.predict(current_states_batch)
|
|
targets = current_values * (1 - mix_fraction) + total_return_batch * mix_fraction
|
|
|
|
inputs = copy.copy(current_states_batch)
|
|
for input_index, input in enumerate(old_policy_values):
|
|
name = 'output_0_{}'.format(input_index)
|
|
if name in self.critic_network.online_network.inputs:
|
|
inputs[name] = input
|
|
|
|
value_loss = self.critic_network.online_network.accumulate_gradients(inputs, targets)
|
|
self.critic_network.apply_gradients_to_online_network()
|
|
if self.tp.distributed:
|
|
self.critic_network.apply_gradients_to_global_network()
|
|
self.critic_network.online_network.reset_accumulated_gradients()
|
|
|
|
loss.append([value_loss[0]])
|
|
loss = np.mean(loss, 0)
|
|
return loss
|
|
|
|
def concat_state_and_timestep(self, dataset):
|
|
current_states_with_timestep = [np.append(transition.state['observation'], transition.info['timestep'])
|
|
for transition in dataset]
|
|
current_states_with_timestep = np.expand_dims(current_states_with_timestep, -1)
|
|
return current_states_with_timestep
|
|
|
|
def train_policy_network(self, dataset, epochs):
|
|
loss = []
|
|
for j in range(epochs):
|
|
loss = {
|
|
'total_loss': [],
|
|
'policy_losses': [],
|
|
'unclipped_grads': [],
|
|
'fetch_result': []
|
|
}
|
|
#shuffle(dataset)
|
|
for i in range(len(dataset) // self.tp.batch_size):
|
|
batch = dataset[i * self.tp.batch_size:(i + 1) * self.tp.batch_size]
|
|
current_states, _, actions, _, _, total_return = self.extract_batch(batch)
|
|
advantages = np.array([t.info['advantage'] for t in batch])
|
|
if not self.tp.env_instance.discrete_controls and len(actions.shape) == 1:
|
|
actions = np.expand_dims(actions, -1)
|
|
|
|
# get old policy probabilities and distribution
|
|
old_policy = utils.force_list(self.policy_network.target_network.predict(current_states))
|
|
|
|
# calculate gradients and apply on both the local policy network and on the global policy network
|
|
fetches = [self.policy_network.online_network.output_heads[0].kl_divergence,
|
|
self.policy_network.online_network.output_heads[0].entropy]
|
|
|
|
inputs = copy.copy(current_states)
|
|
# TODO: why is this output 0 and not output 1?
|
|
inputs['output_0_0'] = actions
|
|
# TODO: does old_policy_distribution really need to be represented as a list?
|
|
# A: yes it does, in the event of discrete controls, it has just a mean
|
|
# otherwise, it has both a mean and standard deviation
|
|
for input_index, input in enumerate(old_policy):
|
|
inputs['output_0_{}'.format(input_index + 1)] = input
|
|
total_loss, policy_losses, unclipped_grads, fetch_result =\
|
|
self.policy_network.online_network.accumulate_gradients(
|
|
inputs, [advantages], additional_fetches=fetches)
|
|
|
|
self.policy_network.apply_gradients_to_online_network()
|
|
if self.tp.distributed:
|
|
self.policy_network.apply_gradients_to_global_network()
|
|
|
|
self.policy_network.online_network.reset_accumulated_gradients()
|
|
|
|
loss['total_loss'].append(total_loss)
|
|
loss['policy_losses'].append(policy_losses)
|
|
loss['unclipped_grads'].append(unclipped_grads)
|
|
loss['fetch_result'].append(fetch_result)
|
|
|
|
self.unclipped_grads.add_sample(unclipped_grads)
|
|
|
|
for key in loss.keys():
|
|
loss[key] = np.mean(loss[key], 0)
|
|
|
|
if self.tp.learning_rate_decay_rate != 0:
|
|
curr_learning_rate = self.main_network.online_network.get_variable_value(self.tp.learning_rate)
|
|
self.curr_learning_rate.add_sample(curr_learning_rate)
|
|
else:
|
|
curr_learning_rate = self.tp.learning_rate
|
|
|
|
# log training parameters
|
|
logger.screen.log_dict(
|
|
collections.OrderedDict([
|
|
("Surrogate loss", loss['policy_losses'][0]),
|
|
("KL divergence", loss['fetch_result'][0]),
|
|
("Entropy", loss['fetch_result'][1]),
|
|
("training epoch", j),
|
|
("learning_rate", curr_learning_rate)
|
|
]),
|
|
prefix="Policy training"
|
|
)
|
|
|
|
self.total_kl_divergence_during_training_process = loss['fetch_result'][0]
|
|
self.entropy.add_sample(loss['fetch_result'][1])
|
|
self.kl_divergence.add_sample(loss['fetch_result'][0])
|
|
return loss['total_loss']
|
|
|
|
def update_kl_coefficient(self):
|
|
# John Schulman takes the mean kl divergence only over the last epoch which is strange but we will follow
|
|
# his implementation for now because we know it works well
|
|
logger.screen.log_title("KL = {}".format(self.total_kl_divergence_during_training_process))
|
|
|
|
# update kl coefficient
|
|
kl_target = self.tp.agent.target_kl_divergence
|
|
kl_coefficient = self.policy_network.online_network.get_variable_value(
|
|
self.policy_network.online_network.output_heads[0].kl_coefficient)
|
|
new_kl_coefficient = kl_coefficient
|
|
if self.total_kl_divergence_during_training_process > 1.3 * kl_target:
|
|
# kl too high => increase regularization
|
|
new_kl_coefficient *= 1.5
|
|
elif self.total_kl_divergence_during_training_process < 0.7 * kl_target:
|
|
# kl too low => decrease regularization
|
|
new_kl_coefficient /= 1.5
|
|
|
|
# update the kl coefficient variable
|
|
if kl_coefficient != new_kl_coefficient:
|
|
self.policy_network.online_network.set_variable_value(
|
|
self.policy_network.online_network.output_heads[0].assign_kl_coefficient,
|
|
new_kl_coefficient,
|
|
self.policy_network.online_network.output_heads[0].kl_coefficient_ph)
|
|
|
|
logger.screen.log_title("KL penalty coefficient change = {} -> {}".format(kl_coefficient, new_kl_coefficient))
|
|
|
|
def post_training_commands(self):
|
|
if self.tp.agent.use_kl_regularization:
|
|
self.update_kl_coefficient()
|
|
|
|
# clean memory
|
|
self.memory.clean()
|
|
|
|
def train(self):
|
|
self.policy_network.sync()
|
|
self.critic_network.sync()
|
|
|
|
dataset = self.memory.transitions
|
|
|
|
self.fill_advantages(dataset)
|
|
|
|
# take only the requested number of steps
|
|
dataset = dataset[:self.tp.agent.num_consecutive_playing_steps]
|
|
|
|
value_loss = self.train_value_network(dataset, 1)
|
|
policy_loss = self.train_policy_network(dataset, 10)
|
|
|
|
self.value_loss.add_sample(value_loss)
|
|
self.policy_loss.add_sample(policy_loss)
|
|
self.update_log() # should be done in order to update the data that has been accumulated * while not playing *
|
|
return np.append(value_loss, policy_loss)
|
|
|
|
def choose_action(self, curr_state, phase=utils.RunPhase.TRAIN):
|
|
if self.env.discrete_controls:
|
|
# DISCRETE
|
|
action_values = self.policy_network.online_network.predict(self.tf_input_state(curr_state)).squeeze()
|
|
|
|
if phase == utils.RunPhase.TRAIN:
|
|
action = self.exploration_policy.get_action(action_values)
|
|
else:
|
|
action = np.argmax(action_values)
|
|
action_info = {"action_probability": action_values[action]}
|
|
# self.entropy.add_sample(-np.sum(action_values * np.log(action_values)))
|
|
else:
|
|
# CONTINUOUS
|
|
action_values_mean, action_values_std = self.policy_network.online_network.predict(self.tf_input_state(curr_state))
|
|
action_values_mean = action_values_mean.squeeze()
|
|
action_values_std = action_values_std.squeeze()
|
|
if phase == utils.RunPhase.TRAIN:
|
|
action = np.squeeze(np.random.randn(1, self.action_space_size) * action_values_std + action_values_mean)
|
|
else:
|
|
action = action_values_mean
|
|
action_info = {"action_probability": action_values_mean}
|
|
|
|
return action, action_info
|