def main():
parser = argparse.ArgumentParser('Parse configuration file')
parser.add_argument('--policy', type=str, default='orca')
parser.add_argument('--env_config',
type=str,
default='../configs/env.config')
parser.add_argument('--policy_config',
type=str,
default='../configs/policy.config')
parser.add_argument('--imitate_config',
type=str,
default='../configs/demonstrate.config')
parser.add_argument('--num_frames', type=int, default=1)
parser.add_argument('--output_dir',
type=str,
default='../data/output/temp')
parser.add_argument('--memory_dir',
type=str,
default='../data/memory/temp')
parser.add_argument('--human_type',
type=str,
default='normal',
choices=['aggressive', 'normal', 'cautious'])
parser.add_argument('--expert_file',
type=str,
default='../data/expert/rl_model.pth')
parser.add_argument('--freeze', default=False, action='store_true')
parser.add_argument('--resume', default=False, action='store_true')
parser.add_argument('--gpu', default=False, action='store_true')
parser.add_argument('--debug', default=False, action='store_true')
args = parser.parse_args()
# configure paths
make_new_dir = True
if os.path.exists(args.output_dir):
key = input(
'Output directory already exists! Overwrite the folder? (y/n)')
if key == 'y' and not args.resume:
shutil.rmtree(args.output_dir)
else:
make_new_dir = False
args.env_config = os.path.join(args.output_dir,
os.path.basename(args.env_config))
args.policy_config = os.path.join(
args.output_dir, os.path.basename(args.policy_config))
args.imitate_config = os.path.join(
args.output_dir, os.path.basename(args.imitate_config))
if make_new_dir:
os.makedirs(args.output_dir)
shutil.copy(args.env_config, args.output_dir)
shutil.copy(args.policy_config, args.output_dir)
shutil.copy(args.imitate_config, args.output_dir)
log_file = os.path.join(args.output_dir, 'output.log')
# configure logging
mode = 'a' if args.resume else 'w'
file_handler = logging.FileHandler(log_file, mode=mode)
stdout_handler = logging.StreamHandler(sys.stdout)
level = logging.INFO if not args.debug else logging.DEBUG
logging.basicConfig(level=level,
handlers=[stdout_handler, file_handler],
format='%(asctime)s, %(levelname)s: %(message)s',
datefmt="%Y-%m-%d %H:%M:%S")
repo = git.Repo(search_parent_directories=True)
device = torch.device(
"cuda:0" if torch.cuda.is_available() and args.gpu else "cpu")
logging.info(' =========== Demonstrate %s ============ ', args.policy)
logging.info('Current git head hash code: %s', (repo.head.object.hexsha))
logging.info('Using device: %s', device)
# configure environment
env_config = configparser.RawConfigParser()
env_config.read(args.env_config)
#env = gym.make('CrowdSim-v0')
env = CrowdSim()
env.configure(env_config)
env_mods = None
if args.human_type == 'aggressive':
env_mods = {'safety_space': 0.03, 'time_horizon': 1}
elif args.human_type == 'cautious':
env_mods = {'safety_space': 0.3, 'time_horizon': 8}
modify_env_params(env, mods=env_mods)
robot = Robot(env_config, 'robot')
env.set_robot(robot)
# multi-frame env
if args.num_frames > 1:
logging.info("stack %d frames", args.num_frames)
env = FrameStack(env, args.num_frames)
# configure policy
policy = policy_factory[args.policy]()
if not policy.trainable:
parser.error('Policy has to be trainable')
if args.policy_config is None:
parser.error(
'Policy config has to be specified for a trainable network')
policy_config = configparser.RawConfigParser()
policy_config.read(args.policy_config)
# read training parameters
if args.imitate_config is None:
parser.error(
'Train config has to be specified for a trainable network')
imitate_config = configparser.RawConfigParser()
imitate_config.read(args.imitate_config)
# configure demonstrateion and explorer
memory = ReplayMemory(
10000000) # sufficiently large to store expert demonstration
if not os.path.exists(args.memory_dir):
os.makedirs(args.memory_dir)
if robot.visible:
demonstration_file = os.path.join(args.memory_dir,
'data_imit_visible.pt')
else:
demonstration_file = os.path.join(args.memory_dir,
'data_imit_invisible.pt')
il_episodes = imitate_config.getint('demonstration', 'il_episodes')
expert_policy = imitate_config.get('demonstration', 'il_policy')
expert_policy = policy_factory[expert_policy]()
expert_policy.configure(policy_config)
expert_policy.set_device(device)
expert_policy.set_env(env)
expert_policy.multiagent_training = policy.multiagent_training
counter = imitate_config.getint('demonstration', 'counter_start')
env.set_counter(counter)
if robot.visible:
expert_policy.safety_space = imitate_config.getfloat(
'demonstration', 'safety_space')
else:
expert_policy.safety_space = imitate_config.getfloat(
'demonstration', 'safety_space')
robot.set_policy(expert_policy)
noise_explore = imitate_config.getfloat('demonstration', 'noise_explore')
gamma = policy_config.getfloat('rl', 'gamma')
explorer = Explorer(env,
robot,
device,
args.num_frames,
memory,
gamma=gamma,
target_policy=policy)
if expert_policy.trainable:
if args.expert_file is None:
logging.warning(
'Trainable policy is NOT specified with a model weights directory'
)
else:
try:
expert_policy.get_model().load_state_dict(
torch.load(args.expert_file))
logging.info('Loaded policy from %s', args.expert_file)
except Exception as e:
logging.warning(e)
expert_policy.get_model().load_state_dict(torch.load(
args.expert_file),
strict=False)
logging.info('Loaded policy partially from %s',
args.expert_file)
robot.policy.set_epsilon(-1.0)
# data collection
robot.policy.multiagent_training = True
human_obsv = explorer.run_k_episodes(il_episodes,
'train',
update_memory=True,
imitation_learning=True,
noise_explore=noise_explore)
torch.save(memory.memory, demonstration_file)
logging.info('Save memory to %s', demonstration_file)
obs_file = os.path.join(args.memory_dir, 'data_traj.pt')
torch.save(human_obsv, obs_file)
logging.info('Save #%d episodes of obsv to %s', len(human_obsv), obs_file)
def main():
parser = argparse.ArgumentParser('Parse configuration file')
parser.add_argument('--env_config', type=str, default='configs/env.config')
parser.add_argument('--policy', type=str, default='cadrl')
parser.add_argument('--policy_config', type=str, default='configs/policy.config')
parser.add_argument('--train_config', type=str, default='configs/train.config')
parser.add_argument('--output_dir', type=str, default='data/output')
parser.add_argument('--weights', type=str)
parser.add_argument('--resume', default=True, action='store_true')
parser.add_argument('--gpu', default=False, action='store_true')
parser.add_argument('--debug', default=False, action='store_true')
args = parser.parse_args()
# configure paths
make_new_dir = True
if os.path.exists(args.output_dir):
key = input('Output directory already exists! Overwrite the folder? (y/n)')
if key == 'y' and not args.resume:
shutil.rmtree(args.output_dir)
else:
make_new_dir = False
args.env_config = os.path.join(args.output_dir, os.path.basename(args.env_config))
args.policy_config = os.path.join(args.output_dir, os.path.basename(args.policy_config))
args.train_config = os.path.join(args.output_dir, os.path.basename(args.train_config))
if make_new_dir:
os.makedirs(args.output_dir)
shutil.copy(args.env_config, args.output_dir)
shutil.copy(args.policy_config, args.output_dir)
shutil.copy(args.train_config, args.output_dir)
log_file = os.path.join(args.output_dir, 'output.log')
il_weight_file = os.path.join(args.output_dir, 'il_model.pth')
rl_weight_file = os.path.join(args.output_dir, 'rl_model.pth')
# configure logging
mode = 'a' if args.resume else 'w'
file_handler = logging.FileHandler(log_file, mode=mode)
stdout_handler = logging.StreamHandler(sys.stdout)
level = logging.INFO if not args.debug else logging.DEBUG
logging.basicConfig(level=level, handlers=[stdout_handler, file_handler],
format='%(asctime)s, %(levelname)s: %(message)s', datefmt="%Y-%m-%d %H:%M:%S")
repo = git.Repo(search_parent_directories=True)
logging.info('Current git head hash code: %s'.format(repo.head.object.hexsha))
device = torch.device("cuda:0" if torch.cuda.is_available() and args.gpu else "cpu")
logging.info('Using device: %s', device)
# configure policy
policy = policy_factory[args.policy]()
if not policy.trainable:
parser.error('Policy has to be trainable')
if args.policy_config is None:
parser.error('Policy config has to be specified for a trainable network')
policy_config = configparser.RawConfigParser()
policy_config.read(args.policy_config)
policy.configure(policy_config)
policy.set_device(device)
# configure environment
env_config = configparser.RawConfigParser()
env_config.read(args.env_config)
env = gym.make('CrowdSim-v0')
env.configure(env_config)
robot = Robot(env_config, 'robot')
env.set_robot(robot)
# read training parameters
if args.train_config is None:
parser.error('Train config has to be specified for a trainable network')
train_config = configparser.RawConfigParser()
train_config.read(args.train_config)
rl_learning_rate = train_config.getfloat('train', 'rl_learning_rate')
train_batches = train_config.getint('train', 'train_batches')
train_episodes = train_config.getint('train', 'train_episodes')
sample_episodes = train_config.getint('train', 'sample_episodes')
target_update_interval = train_config.getint('train', 'target_update_interval')
evaluation_interval = train_config.getint('train', 'evaluation_interval')
capacity = train_config.getint('train', 'capacity')
epsilon_start = train_config.getfloat('train', 'epsilon_start')
epsilon_end = train_config.getfloat('train', 'epsilon_end')
epsilon_decay = train_config.getfloat('train', 'epsilon_decay')
checkpoint_interval = train_config.getint('train', 'checkpoint_interval')
# configure trainer and explorer
memory = ReplayMemory(capacity)
model = policy.get_model()
batch_size = train_config.getint('trainer', 'batch_size')
trainer = Trainer(model, memory, device, batch_size)
explorer = Explorer(env, robot, device, memory, policy.gamma, target_policy=policy)
# imitation learning
if args.resume:
if not os.path.exists(rl_weight_file):
logging.error('RL weights does not exist')
model.load_state_dict(torch.load(rl_weight_file))
rl_weight_file = os.path.join(args.output_dir, 'resumed_rl_model.pth')
logging.info('Load reinforcement learning trained weights. Resume training')
elif os.path.exists(il_weight_file):
model.load_state_dict(torch.load(il_weight_file))
logging.info('Load imitation learning trained weights.')
else:
il_episodes = train_config.getint('imitation_learning', 'il_episodes')
il_policy = train_config.get('imitation_learning', 'il_policy')
il_epochs = train_config.getint('imitation_learning', 'il_epochs')
il_learning_rate = train_config.getfloat('imitation_learning', 'il_learning_rate')
trainer.set_learning_rate(il_learning_rate)
if robot.visible:
safety_space = 0
else:
safety_space = train_config.getfloat('imitation_learning', 'safety_space')
il_policy = policy_factory[il_policy]()
il_policy.multiagent_training = policy.multiagent_training
il_policy.safety_space = safety_space
robot.set_policy(il_policy)
explorer.run_k_episodes(il_episodes, 'train', update_memory=True, imitation_learning=True)
trainer.optimize_epoch(il_epochs)
torch.save(model.state_dict(), il_weight_file)
logging.info('Finish imitation learning. Weights saved.')
logging.info('Experience set size: %d/%d', len(memory), memory.capacity)
explorer.update_target_model(model)
# reinforcement learning
policy.set_env(env)
robot.set_policy(policy)
robot.print_info()
trainer.set_learning_rate(rl_learning_rate)
# fill the memory pool with some RL experience
if args.resume:
robot.policy.set_epsilon(epsilon_end)
explorer.run_k_episodes(100, 'train', update_memory=True, episode=0)
logging.info('Experience set size: %d/%d', len(memory), memory.capacity)
episode = 0
while episode < train_episodes:
if args.resume:
epsilon = epsilon_end
else:
if episode < epsilon_decay:
epsilon = epsilon_start + (epsilon_end - epsilon_start) / epsilon_decay * episode
else:
epsilon = epsilon_end
robot.policy.set_epsilon(epsilon)
# evaluate the model
if episode % evaluation_interval == 0:
explorer.run_k_episodes(env.case_size['val'], 'val', episode=episode)
# sample k episodes into memory and optimize over the generated memory
explorer.run_k_episodes(sample_episodes, 'train', update_memory=True, episode=episode)
trainer.optimize_batch(train_batches)
episode += 1
if episode % target_update_interval == 0:
explorer.update_target_model(model)
if episode != 0 and episode % checkpoint_interval == 0:
torch.save(model.state_dict(), rl_weight_file)
# final test
explorer.run_k_episodes(env.case_size['test'], 'test', episode=episode)
def main(args):
set_random_seeds(args.randomseed)
# configure paths
make_new_dir = True
if os.path.exists(args.output_dir):
if args.overwrite:
shutil.rmtree(args.output_dir)
else:
key = input(
'Output directory already exists! Overwrite the folder? (y/n)')
if key == 'y' and not args.resume:
shutil.rmtree(args.output_dir)
else:
make_new_dir = False
if make_new_dir:
os.makedirs(args.output_dir)
shutil.copy(args.config, os.path.join(args.output_dir, 'config.py'))
# # insert the arguments from command line to the config file
# with open(os.path.join(args.output_dir, 'config.py'), 'r') as fo:
# config_text = fo.read()
# search_pairs = {r"gcn.X_dim = \d*": "gcn.X_dim = {}".format(args.X_dim),
# r"gcn.num_layer = \d": "gcn.num_layer = {}".format(args.layers),
# r"gcn.similarity_function = '\w*'": "gcn.similarity_function = '{}'".format(args.sim_func),
# r"gcn.layerwise_graph = \w*": "gcn.layerwise_graph = {}".format(args.layerwise_graph),
# r"gcn.skip_connection = \w*": "gcn.skip_connection = {}".format(args.skip_connection)}
#
# for find, replace in search_pairs.items():
# config_text = re.sub(find, replace, config_text)
#
# with open(os.path.join(args.output_dir, 'config.py'), 'w') as fo:
# fo.write(config_text)
args.config = os.path.join(args.output_dir, 'config.py')
log_file = os.path.join(args.output_dir, 'output.log')
il_weight_file = os.path.join(args.output_dir, 'il_model.pth')
rl_weight_file = os.path.join(args.output_dir, 'rl_model.pth')
spec = importlib.util.spec_from_file_location('config', args.config)
if spec is None:
parser.error('Config file not found.')
config = importlib.util.module_from_spec(spec)
spec.loader.exec_module(config)
# configure logging
mode = 'a' if args.resume else 'w'
file_handler = logging.FileHandler(log_file, mode=mode)
stdout_handler = logging.StreamHandler(sys.stdout)
level = logging.INFO if not args.debug else logging.DEBUG
logging.basicConfig(level=level,
handlers=[stdout_handler, file_handler],
format='%(asctime)s, %(levelname)s: %(message)s',
datefmt="%Y-%m-%d %H:%M:%S")
repo = git.Repo(search_parent_directories=True)
logging.info('Current git head hash code: {}'.format(
repo.head.object.hexsha))
logging.info('Current config content is :{}'.format(config))
device = torch.device(
"cuda:0" if torch.cuda.is_available() and args.gpu else "cpu")
logging.info('Using device: %s', device)
writer = SummaryWriter(log_dir=args.output_dir)
# configure policy
policy_config = config.PolicyConfig()
policy = policy_factory[policy_config.name]()
if not policy.trainable:
parser.error('Policy has to be trainable')
policy.configure(policy_config)
policy.set_device(device)
# configure environment
env_config = config.EnvConfig(args.debug)
env = gym.make('CrowdSim-v0')
env.configure(env_config)
robot = Robot(env_config, 'robot')
robot.time_step = env.time_step
env.set_robot(robot)
# read training parameters
train_config = config.TrainConfig(args.debug)
rl_learning_rate = train_config.train.rl_learning_rate
train_batches = train_config.train.train_batches
train_episodes = train_config.train.train_episodes
sample_episodes = train_config.train.sample_episodes
target_update_interval = train_config.train.target_update_interval
evaluation_interval = train_config.train.evaluation_interval
capacity = train_config.train.capacity
epsilon_start = train_config.train.epsilon_start
epsilon_end = train_config.train.epsilon_end
epsilon_decay = train_config.train.epsilon_decay
checkpoint_interval = train_config.train.checkpoint_interval
# configure trainer and explorer
memory = ReplayMemory(capacity)
model = policy.get_model()
batch_size = train_config.trainer.batch_size
optimizer = train_config.trainer.optimizer
if policy_config.name == 'model_predictive_rl':
trainer = MPRLTrainer(
model,
policy.state_predictor,
memory,
device,
policy,
writer,
batch_size,
optimizer,
env.human_num,
reduce_sp_update_frequency=train_config.train.
reduce_sp_update_frequency,
freeze_state_predictor=train_config.train.freeze_state_predictor,
detach_state_predictor=train_config.train.detach_state_predictor,
share_graph_model=policy_config.model_predictive_rl.
share_graph_model)
else:
trainer = VNRLTrainer(model, memory, device, policy, batch_size,
optimizer, writer)
explorer = Explorer(env,
robot,
device,
writer,
memory,
policy.gamma,
target_policy=policy)
# imitation learning
if args.resume:
if not os.path.exists(rl_weight_file):
logging.error('RL weights does not exist')
model.load_state_dict(torch.load(rl_weight_file))
rl_weight_file = os.path.join(args.output_dir, 'resumed_rl_model.pth')
logging.info(
'Load reinforcement learning trained weights. Resume training')
elif os.path.exists(il_weight_file):
model.load_state_dict(torch.load(il_weight_file))
logging.info('Load imitation learning trained weights.')
else:
il_episodes = train_config.imitation_learning.il_episodes
il_policy = train_config.imitation_learning.il_policy
il_epochs = train_config.imitation_learning.il_epochs
il_learning_rate = train_config.imitation_learning.il_learning_rate
trainer.set_learning_rate(il_learning_rate)
if robot.visible:
safety_space = 0
else:
safety_space = train_config.imitation_learning.safety_space
il_policy = policy_factory[il_policy]()
il_policy.multiagent_training = policy.multiagent_training
il_policy.safety_space = safety_space
robot.set_policy(il_policy)
explorer.run_k_episodes(il_episodes,
'train',
update_memory=True,
imitation_learning=True)
trainer.optimize_epoch(il_epochs)
policy.save_model(il_weight_file)
logging.info('Finish imitation learning. Weights saved.')
logging.info('Experience set size: %d/%d', len(memory),
memory.capacity)
trainer.update_target_model(model)
# reinforcement learning
policy.set_env(env)
robot.set_policy(policy)
robot.print_info()
trainer.set_learning_rate(rl_learning_rate)
# fill the memory pool with some RL experience
if args.resume:
robot.policy.set_epsilon(epsilon_end)
explorer.run_k_episodes(100, 'train', update_memory=True, episode=0)
logging.info('Experience set size: %d/%d', len(memory),
memory.capacity)
episode = 0
best_val_reward = -1
best_val_model = None
# evaluate the model after imitation learning
if episode % evaluation_interval == 0:
logging.info(
'Evaluate the model instantly after imitation learning on the validation cases'
)
explorer.run_k_episodes(env.case_size['val'], 'val', episode=episode)
explorer.log('val', episode // evaluation_interval)
if args.test_after_every_eval:
explorer.run_k_episodes(env.case_size['test'],
'test',
episode=episode,
print_failure=True)
explorer.log('test', episode // evaluation_interval)
episode = 0
while episode < train_episodes:
if args.resume:
epsilon = epsilon_end
else:
if episode < epsilon_decay:
epsilon = epsilon_start + (
epsilon_end - epsilon_start) / epsilon_decay * episode
else:
epsilon = epsilon_end
robot.policy.set_epsilon(epsilon)
# sample k episodes into memory and optimize over the generated memory
explorer.run_k_episodes(sample_episodes,
'train',
update_memory=True,
episode=episode)
explorer.log('train', episode)
trainer.optimize_batch(train_batches, episode)
episode += 1
if episode % target_update_interval == 0:
trainer.update_target_model(model)
# evaluate the model
if episode % evaluation_interval == 0:
_, _, _, reward, _ = explorer.run_k_episodes(env.case_size['val'],
'val',
episode=episode)
explorer.log('val', episode // evaluation_interval)
if episode % checkpoint_interval == 0 and reward > best_val_reward:
best_val_reward = reward
best_val_model = copy.deepcopy(policy.get_state_dict())
# test after every evaluation to check how the generalization performance evolves
if args.test_after_every_eval:
explorer.run_k_episodes(env.case_size['test'],
'test',
episode=episode,
print_failure=True)
explorer.log('test', episode // evaluation_interval)
if episode != 0 and episode % checkpoint_interval == 0:
current_checkpoint = episode // checkpoint_interval - 1
save_every_checkpoint_rl_weight_file = rl_weight_file.split(
'.')[0] + '_' + str(current_checkpoint) + '.pth'
policy.save_model(save_every_checkpoint_rl_weight_file)
# # test with the best val model
if best_val_model is not None:
policy.load_state_dict(best_val_model)
torch.save(best_val_model, os.path.join(args.output_dir,
'best_val.pth'))
logging.info('Save the best val model with the reward: {}'.format(
best_val_reward))
explorer.run_k_episodes(env.case_size['test'],
'test',
episode=episode,
print_failure=True)
def main():
parser = argparse.ArgumentParser('Parse configuration file')
parser.add_argument('--env_config', type=str, default='configs/env.config')
parser.add_argument('--policy_config',
type=str,
default='configs/policy.config')
parser.add_argument('--train_config',
type=str,
default='configs/train.config')
parser.add_argument('--output_dir', type=str, default='data')
parser.add_argument('--resume', default=False, action='store_true')
parser.add_argument('--gpu', default=False, action='store_true')
parser.add_argument('--debug', default=False, action='store_true')
parser.add_argument('--gpu_num', default="0", type=str)
args = parser.parse_args()
os.environ["CUDA_VISIBLE_DEVICES"] = args.gpu_num
########## CONFIGURE PATHS ##########
make_new_dir = True
outputs = os.listdir(args.output_dir)
# Find the 'output' files in output directory
output_nums = [
name[name.find("output") + 6:] for name in outputs if 'output' in name
and os.path.isdir(os.path.join(args.output_dir, name))
]
num = 0
if output_nums != []:
num = max([int(num) if num != '' else 0 for num in output_nums])
if num == 0:
num = ''
key = input('Continue from last output ? (y/n)')
if key == 'y' and not args.resume:
args.output_dir = os.path.join(args.output_dir,
"output" + str(num))
args.env_config = os.path.join(args.output_dir,
os.path.basename(args.env_config))
args.policy_config = os.path.join(
args.output_dir, os.path.basename(args.policy_config))
args.train_config = os.path.join(
args.output_dir, os.path.basename(args.train_config))
else:
key = input(
'Overwrite last output directory instead of new ? (y/n)')
if key == 'y' and not args.resume:
args.output_dir = os.path.join(args.output_dir,
'output' + str(num))
shutil.rmtree(args.output_dir)
os.makedirs(args.output_dir)
else:
num = num + 1 if num else 1
args.output_dir = os.path.join(args.output_dir,
'output' + str(num))
os.makedirs(args.output_dir)
shutil.copy(args.env_config, args.output_dir)
shutil.copy(args.policy_config, args.output_dir)
shutil.copy(args.train_config, args.output_dir)
else:
args.output_dir = os.path.join(args.output_dir, 'output')
os.makedirs(args.output_dir)
shutil.copy(args.env_config, args.output_dir)
shutil.copy(args.policy_config, args.output_dir)
shutil.copy(args.train_config, args.output_dir)
log_file = os.path.join(args.output_dir, 'output.log')
il_weight_file = os.path.join(args.output_dir, 'il_model.pth')
rl_weight_file = os.path.join(args.output_dir, 'rl_model.pth')
########## CONFIGURE LOGGING ##########
mode = 'a' if args.resume else 'w'
logger = logging.getLogger('train_sgan')
logger.propogate = False
level = logging.INFO if not args.debug else logging.DEBUG
logger.setLevel(level)
file_handler = logging.FileHandler(log_file, mode=mode)
file_handler.setLevel(level)
stdout_handler = logging.StreamHandler(sys.stdout)
stdout_handler.setLevel(level)
formatter = logging.Formatter('%(asctime)s, %(levelname)s: %(message)s')
stdout_handler.setFormatter(formatter)
file_handler.setFormatter(formatter)
logger.addHandler(file_handler)
logger.addHandler(stdout_handler)
device = torch.device(
"cuda:0" if torch.cuda.is_available() and args.gpu else "cpu")
########## CONFIGURE POLICY ##########
policy = TLSGAN(device, logger)
if args.policy_config is None:
parser.error(
'Policy config has to be specified for a trainable network')
policy_config = configparser.RawConfigParser()
policy_config.read(args.policy_config)
policy.configure(policy_config)
policy.set_device(device)
########## CONFIGURE ENVIRONMENT ##########
env_config = configparser.RawConfigParser()
env_config.read(args.env_config)
env = gym.make('CrowdSim-v0')
env.configure(env_config)
robot = Robot(env_config, 'robot')
env.set_robot(robot)
########## READ RL PARAMETERS ##########
if args.train_config is None:
parser.error(
'Train config has to be specified for a trainable network')
train_config = configparser.RawConfigParser()
train_config.read(args.train_config)
rl_learning_rate = train_config.getfloat('train', 'rl_learning_rate')
train_batches = train_config.getint('train', 'train_batches')
train_episodes = train_config.getint('train', 'train_episodes')
sample_episodes = train_config.getint('train', 'sample_episodes')
target_update_interval = train_config.getint('train',
'target_update_interval')
evaluation_interval = train_config.getint('train', 'evaluation_interval')
capacity = train_config.getint('train', 'capacity')
epsilon_start = train_config.getfloat('train', 'epsilon_start')
epsilon_end = train_config.getfloat('train', 'epsilon_end')
epsilon_decay = train_config.getfloat('train', 'epsilon_decay')
checkpoint_interval = train_config.getint('train', 'checkpoint_interval')
########## CONFIGURE MEMORY, TRAINER, EXPLORER ##########
train_memory = ReplayMemory(capacity)
val_memory = ReplayMemory(capacity)
model = policy.get_model()
batch_size = train_config.getint('trainer', 'batch_size')
trainer = Trainer(model, train_memory, val_memory, device, batch_size,
args.output_dir, logger)
explorer = Explorer(env, robot, policy.policy_learning, device,
train_memory, val_memory, policy.gamma, logger, policy,
policy.obs_len)
logger.info('Using device: %s', device)
########## IMITATION LEARNING ##########
if args.resume:
if not os.path.exists(rl_weight_file):
logger.error('RL weights does not exist')
model.load_state_dict(torch.load(rl_weight_file, map_location=device))
rl_weight_file = os.path.join(args.output_dir, 'resumed_rl_model.pth')
logger.info(
'Load reinforcement learning trained weights. Resume training')
elif os.path.exists(il_weight_file):
model.load_state_dict(torch.load(il_weight_file, map_location=device),
strict=False)
logger.info('Load imitation learning trained weights.')
if robot.visible:
safety_space = 0
else:
safety_space = train_config.getfloat('imitation_learning',
'safety_space')
il_policy = train_config.get('imitation_learning', 'il_policy')
il_policy = policy_factory[il_policy]()
il_policy.multiagent_training = policy.multiagent_training
il_policy.safety_space = safety_space
robot.set_test_policy(il_policy)
robot.set_policy(policy)
policy.set_env(env)
robot.policy.set_epsilon(epsilon_end)
explorer.run_k_episodes(100,
0,
'val',
update_memory=True,
imitation_learning=True,
with_render=False)
explorer.run_k_episodes(100,
0,
'val',
update_memory=True,
imitation_learning=True,
with_render=False)
explorer.run_k_episodes(100,
0,
'val',
update_memory=True,
imitation_learning=True,
with_render=False)
explorer.run_k_episodes(100,
0,
'val',
update_memory=True,
imitation_learning=True,
with_render=False)
explorer.run_k_episodes(100,
0,
'val',
update_memory=True,
imitation_learning=True,
with_render=False)
logger.info('Experience training set size: %d/%d', len(train_memory),
train_memory.capacity)
else:
train_episodes = train_config.getint('imitation_learning',
'train_episodes')
validation_episodes = train_config.getint('imitation_learning',
'validation_episodes')
logger.info(
'Starting imitation learning on %d training episodes and %d validation episodes',
train_episodes, validation_episodes)
il_policy = train_config.get('imitation_learning', 'il_policy')
il_epochs = train_config.getint('imitation_learning', 'il_epochs')
il_learning_rate = train_config.getfloat('imitation_learning',
'il_learning_rate')
trainer.set_learning_rate(il_learning_rate)
if robot.visible:
safety_space = 0
else:
safety_space = train_config.getfloat('imitation_learning',
'safety_space')
il_policy = policy_factory[il_policy]()
il_policy.multiagent_training = policy.multiagent_training
il_policy.safety_space = safety_space
robot.set_test_policy(il_policy)
robot.set_policy(il_policy)
explorer.run_k_episodes(train_episodes,
validation_episodes,
'train',
update_memory=True,
imitation_learning=True,
with_render=False)
trainer.optimize_epoch_il(il_epochs, robot, with_validation=True)
torch.save(model.state_dict(), il_weight_file)
logger.info('Finish imitation learning. Weights saved.')
logger.info('Experience train_set size: %d/%d', len(train_memory),
train_memory.capacity)
logger.info('Experience validation_set size: %d/%d', len(val_memory),
val_memory.capacity)
########## REINFORCEMENT LEARNING ##########
raise NotImplementedError
policy.set_env(env)
robot.set_policy(policy)
robot.print_info()
trainer.set_learning_rate(rl_learning_rate)
explorer.update_target_model(model)
robot.policy.set_epsilon(epsilon_end)
rl_memory = ReplayMemory(capacity)
explorer.train_memory = rl_memory
trainer.train_memory = rl_memory
episode = 0
while episode < train_episodes:
if args.resume:
epsilon = epsilon_end
else:
if episode < epsilon_decay:
epsilon = epsilon_start + (
epsilon_end - epsilon_start) / epsilon_decay * episode
else:
epsilon = epsilon_end
robot.policy.set_epsilon(epsilon)
# evaluate the model
if episode % evaluation_interval == 0:
explorer.run_k_episodes(env.case_size['val'],
0,
'val',
episode=episode,
with_render=False)
# sample k episodes into memory and optimize over the generated memory
explorer.run_k_episodes(sample_episodes,
0,
'train',
update_memory=True,
episode=episode)
trainer.optimize_batch_rl(train_batches)
rl_memory.clear()
episode += 1
if episode % target_update_interval == 0:
explorer.update_target_model(model)
if episode != 0 and episode % checkpoint_interval == 0:
torch.save(model.state_dict(), rl_weight_file)
# final test
explorer.run_k_episodes(env.case_size['test'], 0, 'test', episode=episode)
policy.model.load_state_dict(torch.load(weights_path))
# domain/env config (initially same as used in training)
env_config = configparser.RawConfigParser()
env_config.read('configs/env.config')
if args.policy == 'orca':
env = CrowdSim()
else:
env = CrowdSim_SF()
env.configure(env_config)
robot = Robot(env_config, 'robot')
robot.set_policy(policy)
env.set_robot(robot)
# explorer, memory is not used as we will only use the explorer in test mode
memory = ReplayMemory(100000)
explorer = Explorer(env,
robot,
device,
memory,
policy.gamma,
target_policy=policy)
policy.set_env(env)
robot.set_policy(policy)
mods = {
'neighbor_dist': args.neighd,
'max_neighbors': args.maxn,
'time_horizon': args.timeh,
'max_speed': 1
def main(args):
set_random_seeds(args.randomseed)
# configure paths
make_new_dir = True
if os.path.exists(args.output_dir):
if args.overwrite:
shutil.rmtree(args.output_dir)
else:
shutil.rmtree(args.output_dir)
if make_new_dir:
os.makedirs(args.output_dir)
shutil.copy(args.config, os.path.join(args.output_dir, 'config.py'))
# # insert the arguments from command line to the config file
# with open(os.path.join(args.output_dir, 'config.py'), 'r') as fo:
# config_text = fo.read()
# search_pairs = {r"gcn.X_dim = \d*": "gcn.X_dim = {}".format(args.X_dim),
# r"gcn.num_layer = \d": "gcn.num_layer = {}".format(args.layers),
# r"gcn.similarity_function = '\w*'": "gcn.similarity_function = '{}'".format(args.sim_func),
# r"gcn.layerwise_graph = \w*": "gcn.layerwise_graph = {}".format(args.layerwise_graph),
# r"gcn.skip_connection = \w*": "gcn.skip_connection = {}".format(args.skip_connection)}
#
# for find, replace in search_pairs.items():
# config_text = re.sub(find, replace, config_text)
#
# with open(os.path.join(args.output_dir, 'config.py'), 'w') as fo:
# fo.write(config_text)
args.config = os.path.join(args.output_dir, 'config.py')
log_file = os.path.join(args.output_dir, 'output.log')
in_weight_file = os.path.join(args.output_dir, 'in_model.pth')
il_weight_file = os.path.join(args.output_dir, 'il_model.pth')
rl_weight_file = os.path.join(args.output_dir, 'rl_model.pth')
spec = importlib.util.spec_from_file_location('config', args.config)
if spec is None:
parser.error('Config file not found.')
config = importlib.util.module_from_spec(spec)
spec.loader.exec_module(config)
# configure logging
mode = 'a' if args.resume else 'w'
file_handler = logging.FileHandler(log_file, mode=mode)
stdout_handler = logging.StreamHandler(sys.stdout)
level = logging.INFO if not args.debug else logging.DEBUG
logging.basicConfig(level=level,
handlers=[stdout_handler, file_handler],
format='%(asctime)s, %(levelname)s: %(message)s',
datefmt="%Y-%m-%d %H:%M:%S")
repo = git.Repo(search_parent_directories=True)
logging.info('Current git head hash code: {}'.format(
repo.head.object.hexsha))
logging.info('Current random seed: {}'.format(sys_args.randomseed))
logging.info('Current safe_weight: {}'.format(sys_args.safe_weight))
logging.info('Current goal_weight: {}'.format(sys_args.goal_weight))
logging.info('Current re_collision: {}'.format(sys_args.re_collision))
logging.info('Current re_arrival: {}'.format(sys_args.re_arrival))
logging.info('Current config content is :{}'.format(config))
device = torch.device(
"cuda:0" if torch.cuda.is_available() and args.gpu else "cpu")
logging.info('Using device: %s', device)
writer = SummaryWriter(log_dir=args.output_dir)
# configure policy
policy_config = config.PolicyConfig()
policy = policy_factory[policy_config.name]()
if not policy.trainable:
parser.error('Policy has to be trainable')
policy.configure(policy_config, device)
policy.set_device(device)
# configure environment
env_config = config.EnvConfig(args.debug)
env_config.reward.collision_penalty = args.re_collision
env_config.reward.success_reward = args.re_arrival
env_config.reward.goal_factor = args.goal_weight
env_config.reward.discomfort_penalty_factor = args.safe_weight
env_config.sim.human_num = args.human_num
env = gym.make('CrowdSim-v0')
env.configure(env_config)
robot = Robot(env_config, 'robot')
robot.time_step = env.time_step
env.set_robot(robot)
reward_estimator = Reward_Estimator()
reward_estimator.configure(env_config)
policy.reward_estimator = reward_estimator
# for continous action
action_dim = env.action_space.shape[0]
max_action = env.action_space.high
min_action = env.action_space.low
# read training parameters
train_config = config.TrainConfig(args.debug)
rl_learning_rate = train_config.train.rl_learning_rate
train_batches = train_config.train.train_batches
train_episodes = train_config.train.train_episodes
sample_episodes = train_config.train.sample_episodes
target_update_interval = train_config.train.target_update_interval
evaluation_interval = train_config.train.evaluation_interval
capacity = train_config.train.capacity
epsilon_start = train_config.train.epsilon_start
epsilon_end = train_config.train.epsilon_end
epsilon_decay = train_config.train.epsilon_decay
checkpoint_interval = train_config.train.checkpoint_interval
# configure trainer and explorer
memory = ReplayMemory(capacity)
model = policy.get_model()
batch_size = train_config.trainer.batch_size
optimizer = train_config.trainer.optimizer
print(policy_config.name)
if policy_config.name == 'model_predictive_rl':
trainer = MPRLTrainer(
model,
policy.state_predictor,
memory,
device,
policy,
writer,
batch_size,
optimizer,
env.human_num,
reduce_sp_update_frequency=train_config.train.
reduce_sp_update_frequency,
freeze_state_predictor=train_config.train.freeze_state_predictor,
detach_state_predictor=train_config.train.detach_state_predictor,
share_graph_model=policy_config.model_predictive_rl.
share_graph_model)
elif policy_config.name == 'tree_search_rl':
trainer = TSRLTrainer(
model,
policy.state_predictor,
memory,
device,
policy,
writer,
batch_size,
optimizer,
env.human_num,
reduce_sp_update_frequency=train_config.train.
reduce_sp_update_frequency,
freeze_state_predictor=train_config.train.freeze_state_predictor,
detach_state_predictor=train_config.train.detach_state_predictor,
share_graph_model=policy_config.model_predictive_rl.
share_graph_model)
elif policy_config.name == 'gat_predictive_rl':
trainer = MPRLTrainer(
model,
policy.state_predictor,
memory,
device,
policy,
writer,
batch_size,
optimizer,
env.human_num,
reduce_sp_update_frequency=train_config.train.
reduce_sp_update_frequency,
freeze_state_predictor=train_config.train.freeze_state_predictor,
detach_state_predictor=train_config.train.detach_state_predictor,
share_graph_model=policy_config.model_predictive_rl.
share_graph_model)
elif policy_config.name == 'td3_rl':
policy.set_action(action_dim, max_action, min_action)
trainer = TD3RLTrainer(
policy.actor,
policy.critic,
policy.state_predictor,
memory,
device,
policy,
writer,
batch_size,
optimizer,
env.human_num,
reduce_sp_update_frequency=train_config.train.
reduce_sp_update_frequency,
freeze_state_predictor=train_config.train.freeze_state_predictor,
detach_state_predictor=train_config.train.detach_state_predictor,
share_graph_model=policy_config.model_predictive_rl.
share_graph_model)
else:
trainer = VNRLTrainer(model, memory, device, policy, batch_size,
optimizer, writer)
explorer = Explorer(env,
robot,
device,
writer,
memory,
policy.gamma,
target_policy=policy)
policy.save_model(in_weight_file)
# imitation learning
# if args.resume:
# if not os.path.exists(rl_weight_file):
# logging.error('RL weights does not exist')
# policy.load_state_dict(torch.load(rl_weight_file))
# model = policy.get_model()
# rl_weight_file = os.path.join(args.output_dir, 'resumed_rl_model.pth')
# logging.info('Load reinforcement learning trained weights. Resume training')
# elif os.path.exists(il_weight_file):
# policy.load_state_dict(torch.load(rl_weight_file))
# model = policy.get_model()
# logging.info('Load imitation learning trained weights.')
# else:
# il_episodes = train_config.imitation_learning.il_episodes
# il_policy = train_config.imitation_learning.il_policy
# il_epochs = train_config.imitation_learning.il_epochs
# il_learning_rate = train_config.imitation_learning.il_learning_rate
# trainer.set_learning_rate(il_learning_rate)
# if robot.visible:
# safety_space = 0
# else:
# safety_space = train_config.imitation_learning.safety_space
# il_policy = policy_factory[il_policy]()
# il_policy.set_common_parameters(policy_config)
# il_policy.multiagent_training = policy.multiagent_training
# il_policy.safety_space = safety_space
# robot.set_policy(il_policy)
# explorer.run_k_episodes(il_episodes, 'train', update_memory=True, imitation_learning=True)
# trainer.optimize_epoch(il_epochs)
# policy.save_model(il_weight_file)
# logging.info('Finish imitation learning. Weights saved.')
# logging.info('Experience set size: %d/%d', len(memory), memory.capacity)
trainer.update_target_model(model)
# reinforcement learning
policy.set_env(env)
robot.set_policy(policy)
robot.print_info()
trainer.set_rl_learning_rate(rl_learning_rate)
# fill the memory pool with some RL experience
if args.resume:
robot.policy.set_epsilon(epsilon_end)
explorer.run_k_episodes(100, 'train', update_memory=True, episode=0)
logging.info('Experience set size: %d/%d', len(memory),
memory.capacity)
episode = 0
best_val_reward = -1
best_val_return = -1
best_val_model = None
# evaluate the model after imitation learning
if episode % evaluation_interval == 0:
logging.info(
'Evaluate the model instantly after imitation learning on the validation cases'
)
explorer.run_k_episodes(env.case_size['val'], 'val', episode=episode)
explorer.log('val', episode // evaluation_interval)
if args.test_after_every_eval:
explorer.run_k_episodes(env.case_size['test'],
'test',
episode=episode,
print_failure=True)
explorer.log('test', episode // evaluation_interval)
episode = 0
reward_rec = []
return_rec = []
discom_tim_rec = []
nav_time_rec = []
total_time_rec = []
reward_in_last_interval = 0
return_in_last_interval = 0
nav_time__in_last_interval = 0
discom_time_in_last_interval = 0
total_time_in_last_interval = 0
eps_count = 0
fw = open(sys_args.output_dir + '/data.txt', 'w')
print("%f %f %f %f %f" % (0, 0, 0, 0, 0), file=fw)
while episode < train_episodes:
if args.resume:
epsilon = epsilon_end
else:
if episode < epsilon_decay:
epsilon = epsilon_start + (
epsilon_end - epsilon_start) / epsilon_decay * episode
else:
epsilon = epsilon_end
robot.policy.set_epsilon(epsilon)
# sample k episodes into memory and optimize over the generated memory
_, _, nav_time, sum_reward, ave_return, discom_time, total_time = \
explorer.run_k_episodes(sample_episodes, 'train', update_memory=True, episode=episode)
eps_count = eps_count + 1
reward_in_last_interval = reward_in_last_interval + sum_reward
return_in_last_interval = return_in_last_interval + ave_return
nav_time__in_last_interval = nav_time__in_last_interval + nav_time
discom_time_in_last_interval = discom_time_in_last_interval + discom_time
total_time_in_last_interval = total_time_in_last_interval + total_time
interval = 100
if eps_count % interval == 0:
reward_rec.append(reward_in_last_interval / 100.0)
return_rec.append(return_in_last_interval / 100.0)
discom_tim_rec.append(discom_time_in_last_interval / 100.0)
nav_time_rec.append(nav_time__in_last_interval / 100.0)
total_time_rec.append(total_time_in_last_interval / 100.0)
logging.info(
'Train in episode %d reward in last 100 episodes %f %f %f %f %f',
eps_count, reward_rec[-1], return_rec[-1], discom_tim_rec[-1],
nav_time_rec[-1], total_time_rec[-1])
print("%f %f %f %f %f" %
(reward_rec[-1], return_rec[-1], discom_tim_rec[-1],
nav_time_rec[-1], total_time_rec[-1]),
file=fw)
reward_in_last_interval = 0
return_in_last_interval = 0
nav_time__in_last_interval = 0
discom_time_in_last_interval = 0
total_time_in_last_interval = 0
min_reward = (np.min(reward_rec) // 5.0) * 5.0
max_reward = (np.max(reward_rec) // 5.0 + 1) * 5.0
pos = np.array(range(1, len(reward_rec) + 1)) * interval
plt.plot(pos,
reward_rec,
color='r',
marker='.',
linestyle='dashed')
plt.axis([0, eps_count, min_reward, max_reward])
savefig(args.output_dir + "/reward_record.jpg")
explorer.log('train', episode)
trainer.optimize_batch(train_batches, episode)
episode += 1
if episode % target_update_interval == 0:
trainer.update_target_model(model)
# evaluate the model
if episode % evaluation_interval == 0:
_, _, _, reward, average_return, _, _ = explorer.run_k_episodes(
env.case_size['val'], 'val', episode=episode)
explorer.log('val', episode // evaluation_interval)
if episode % checkpoint_interval == 0 and average_return > best_val_return:
best_val_return = average_return
best_val_model = copy.deepcopy(policy.get_state_dict())
# test after every evaluation to check how the generalization performance evolves
if args.test_after_every_eval:
explorer.run_k_episodes(env.case_size['test'],
'test',
episode=episode,
print_failure=True)
explorer.log('test', episode // evaluation_interval)
if episode != 0 and episode % checkpoint_interval == 0:
current_checkpoint = episode // checkpoint_interval - 1
save_every_checkpoint_rl_weight_file = rl_weight_file.split(
'.')[0] + '_' + str(current_checkpoint) + '.pth'
policy.save_model(save_every_checkpoint_rl_weight_file)
# # test with the best val model
fw.close()
if best_val_model is not None:
policy.load_state_dict(best_val_model)
torch.save(best_val_model, os.path.join(args.output_dir,
'best_val.pth'))
logging.info('Save the best val model with the return: {}'.format(
best_val_return))
explorer.run_k_episodes(env.case_size['test'],
'test',
episode=episode,
print_failure=True)
def main():
parser = argparse.ArgumentParser("Parse configuration file")
parser.add_argument("--env_config", type=str, default="configs/env.config")
parser.add_argument("--policy", type=str, default="cadrl")
parser.add_argument("--policy_config",
type=str,
default="configs/policy.config")
# parser.add_argument('--train_config', type=str, default='configs/train-test.config')
parser.add_argument("--train_config",
type=str,
default="configs/train.config")
parser.add_argument("--output_dir",
type=str,
default="data2/actenvcarl_alltf")
parser.add_argument("--weights", type=str)
parser.add_argument("--resume", default=False, action="store_true")
parser.add_argument("--gpu", default=True, action="store_true")
parser.add_argument("--debug", default=True, action="store_true")
parser.add_argument("--phase", type=str, default="train")
parser.add_argument("--test_policy_flag", type=str, default="1")
parser.add_argument("--optimizer", type=str, default="SGD")
parser.add_argument("--multi_process", type=str, default="average")
# reward parameters
parser.add_argument("--agent_timestep", type=float, default=0.4)
parser.add_argument("--human_timestep", type=float, default=0.5)
parser.add_argument("--reward_increment", type=float, default=4.0)
parser.add_argument("--position_variance", type=float, default=4.0)
parser.add_argument("--direction_variance", type=float, default=4.0)
# visable or not
parser.add_argument("--visible", default=False, action="store_true")
# act step cnt
parser.add_argument("--act_steps", type=int, default=1)
parser.add_argument("--act_fixed", default=False, action="store_true")
args = parser.parse_args()
agent_timestep = args.agent_timestep
human_timestep = args.human_timestep
reward_increment = args.reward_increment
position_variance = args.position_variance
direction_variance = args.direction_variance
# agent_visable = args.visable
optimizer_type = args.optimizer
# configure paths
make_new_dir = True
if os.path.exists(args.output_dir):
key = input(
"Output directory already exists! Overwrite the folder? (y/n)")
if key == "y" and not args.resume:
shutil.rmtree(args.output_dir)
else:
make_new_dir = False
args.env_config = os.path.join(args.output_dir,
os.path.basename(args.env_config))
args.policy_config = os.path.join(
args.output_dir, os.path.basename(args.policy_config))
args.train_config = os.path.join(
args.output_dir, os.path.basename(args.train_config))
shutil.copy("utils/trainer.py", args.output_dir)
shutil.copy("policy/" + args.policy + ".py", args.output_dir)
if make_new_dir:
os.makedirs(args.output_dir)
shutil.copy(args.env_config, args.output_dir)
shutil.copy(args.policy_config, args.output_dir)
shutil.copy(args.train_config, args.output_dir)
shutil.copy("utils/trainer.py", args.output_dir)
shutil.copy("policy/" + args.policy + ".py", args.output_dir)
log_file = os.path.join(args.output_dir, "output.log")
il_weight_file = os.path.join(args.output_dir, "il_model.pth")
rl_weight_file = os.path.join(args.output_dir, "rl_model.pth")
# configure logging
mode = "a" if args.resume else "w"
file_handler = logging.FileHandler(log_file, mode=mode)
stdout_handler = logging.StreamHandler(sys.stdout)
level = logging.INFO if not args.debug else logging.DEBUG
logging.basicConfig(
level=level,
handlers=[stdout_handler, file_handler],
format="%(asctime)s, %(levelname)s: %(message)s",
datefmt="%Y-%m-%d %H:%M:%S",
)
# repo = git.Repo(search_parent_directories=True)
# logging.info('Current git head hash code: %s'.format(repo.head.object.hexsha))
device = torch.device(
"cuda:0" if torch.cuda.is_available() and args.gpu else "cpu")
logging.info("Using device: %s", device)
# configure policy
policy = policy_factory[args.policy]()
# policy.set_phase(args.phase)
print("policy type: ", type(policy))
if not policy.trainable:
parser.error("Policy has to be trainable")
if args.policy_config is None:
parser.error(
"Policy config has to be specified for a trainable network")
policy_config = configparser.RawConfigParser()
policy_config.read(args.policy_config)
policy_config.set("actenvcarl", "test_policy_flag", args.test_policy_flag)
policy_config.set("actenvcarl", "multi_process", args.multi_process)
policy_config.set("actenvcarl", "act_steps", args.act_steps)
policy_config.set("actenvcarl", "act_fixed", args.act_fixed)
policy.configure(policy_config)
policy.set_device(device)
# configure environment
env_config = configparser.RawConfigParser()
env_config.read(args.env_config)
env_config.set("reward", "agent_timestep", agent_timestep)
env_config.set("reward", "human_timestep", human_timestep)
env_config.set("reward", "reward_increment", reward_increment)
env_config.set("reward", "position_variance", position_variance)
env_config.set("reward", "direction_variance", direction_variance)
env = gym.make("CrowdSim-v0")
env.configure(env_config)
# 只是为了获取部分的环境配置信息,如半径速度啥的
robot = Robot(env_config, "robot")
env.set_robot(robot)
# read training parameters
if args.train_config is None:
parser.error(
"Train config has to be specified for a trainable network")
train_config = configparser.RawConfigParser()
train_config.read(args.train_config)
rl_learning_rate = train_config.getfloat("train", "rl_learning_rate")
train_batches = train_config.getint("train", "train_batches")
train_episodes = train_config.getint("train", "train_episodes")
sample_episodes = train_config.getint("train", "sample_episodes")
target_update_interval = train_config.getint("train",
"target_update_interval")
evaluation_interval = train_config.getint("train", "evaluation_interval")
capacity = train_config.getint("train", "capacity")
epsilon_start = train_config.getfloat("train", "epsilon_start")
epsilon_end = train_config.getfloat("train", "epsilon_end")
epsilon_decay = train_config.getfloat("train", "epsilon_decay")
checkpoint_interval = train_config.getint("train", "checkpoint_interval")
# configure trainer and explorer
memory = ReplayMemory(capacity)
model = policy.get_model()
batch_size = train_config.getint("trainer", "batch_size")
trainer = Trainer(model, memory, device, batch_size)
explorer = Explorer(env,
robot,
device,
memory,
policy.gamma,
target_policy=policy)
# imitation learning
if args.resume:
if not os.path.exists(rl_weight_file):
logging.error("RL weights does not exist")
model.load_state_dict(torch.load(rl_weight_file))
rl_weight_file = os.path.join(args.output_dir, "resumed_rl_model.pth")
logging.info(
"Load reinforcement learning trained weights. Resume training")
elif os.path.exists(il_weight_file):
model.load_state_dict(torch.load(il_weight_file))
logging.info("Load imitation learning trained weights.")
else:
il_episodes = train_config.getint("imitation_learning", "il_episodes")
il_policy = train_config.get("imitation_learning", "il_policy")
il_epochs = train_config.getint("imitation_learning", "il_epochs")
il_learning_rate = train_config.getfloat("imitation_learning",
"il_learning_rate")
trainer.set_learning_rate(il_learning_rate, optimizer_type)
if robot.visible:
safety_space = 0
else:
safety_space = train_config.getfloat("imitation_learning",
"safety_space")
il_policy = policy_factory[il_policy]()
il_policy.multiagent_training = policy.multiagent_training
il_policy.safety_space = safety_space
print("il_policy: ", type(il_policy))
robot.set_policy(il_policy)
explorer.run_k_episodes(200,
"train",
update_memory=True,
imitation_learning=True)
trainer.optimize_epoch(il_epochs)
torch.save(model.state_dict(), il_weight_file)
logging.info("Finish imitation learning. Weights saved.")
logging.info("Experience set size: %d/%d", len(memory),
memory.capacity)
explorer.update_target_model(model)
# reinforcement learning
policy.set_env(env)
robot.set_policy(policy)
robot.print_info()
trainer.set_learning_rate(rl_learning_rate, optimizer_type)
# fill the memory pool with some RL experienceo
if args.resume:
robot.policy.set_epsilon(epsilon_end)
explorer.run_k_episodes(100, "train", update_memory=True, episode=0)
logging.info("Experience set size: %d/%d", len(memory),
memory.capacity)
episode = 0
while episode < train_episodes:
if args.resume:
epsilon = epsilon_end
else:
if episode < epsilon_decay:
epsilon = epsilon_start + (
epsilon_end - epsilon_start) / epsilon_decay * episode
else:
epsilon = epsilon_end
robot.policy.set_epsilon(epsilon)
# evaluate the model
if episode % evaluation_interval == 0:
explorer.run_k_episodes(env.case_size["val"],
"val",
episode=episode)
# sample k episodes into memory and optimize over the generated memory
explorer.run_k_episodes(sample_episodes,
"train",
update_memory=True,
episode=episode)
trainer.optimize_batch(train_batches)
episode += 1
if episode % target_update_interval == 0:
explorer.update_target_model(model)
if episode != 0 and episode % checkpoint_interval == 0:
# rl_weight_file_name = 'rl_model_' + str(episode) + '.pth'
rl_weight_file_name = "rl_model_{:d}.pth".format(episode)
rl_weight_file = os.path.join(args.output_dir, rl_weight_file_name)
torch.save(model.state_dict(),
rl_weight_file,
_use_new_zipfile_serialization=False)
# final test
explorer.run_k_episodes(env.case_size["test"], "test", episode=episode)
