huangfu/InteractionImitation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

committing what is hopefully final run of sgail for both experiments, A and B

Browse Source
This commit is contained in:
Arec Jamgochian
2022-02-26 23:48:58 -08:00
parent 3b9051505e
commit 3280041efa
1 changed files with 100 additions and 87 deletions
Download Patch File Download Diff File Expand all files Collapse all files

187
sgail-ppo-options-setobs2.py
View File

@@ -21,11 +21,10 @@ import json
DIR = os.path.dirname(os.path.abspath(__file__)) DIR = os.path.dirname(os.path.abspath(__file__))
option_list = [[(vel, time) for vel in [0, 1, 2, 4, 6, 8, 10] for time in [5]], option_list = [[(vel, time) for vel in [0, 1, 2, 4, 6, 8, 10] for time in [5]],
[(vel, time) for vel in [0, 2.5, 5, 7.5, 10] for time in [5, 10]], [(vel, time) for vel in [0, 1, 2, 5, 7.5, 10] for time in [5, 10]],
[(vel, time) for vel in [0, 2.5, 5, 7.5, 10] for time in [5, 20]],
[(vel, time) for vel in [0, 2.5, 5, 10] for time in [5, 10, 20]],
[(vel, time) for vel in [0, 3, 10] for time in [5, 10, 20]] [(vel, time) for vel in [0, 3, 10] for time in [5, 10, 20]]
] ]
activations = [torch.nn.Tanh, torch.nn.LeakyReLU]
obs_min = np.array([ obs_min = np.array([
[-1000, -1000, 0, -np.pi, -1e-1, 0.], [-1000, -1000, 0, -np.pi, -1e-1, 0.],
@@ -49,27 +48,47 @@ def training_function(config):
np.random.seed(config['seed']) np.random.seed(config['seed'])
torch.manual_seed(config['seed']) torch.manual_seed(config['seed'])
envs = sum([[SafeOptionsEnv(Setobs( if config['experiment'] == 'A':
TransformObservation(CollisionPenaltyWrapper(IntersimpleLidarFlatRandom( envs = [SafeOptionsEnv(Setobs(
n_rays=5, TransformObservation(CollisionPenaltyWrapper(IntersimpleLidarFlatRandom(
reward=functools.partial( n_rays=5,
speed_reward, reward=functools.partial(
collision_penalty=0 speed_reward,
), collision_penalty=0
check_collisions=True, ),
stop_on_collision=config['trainenv']['stop_on_collision'], track=track, check_collisions=True,
), collision_distance=6, collision_penalty=100), lambda obs: (obs - obs_min) / (obs_max - obs_min + 1e-10)) stop_on_collision=config['trainenv']['stop_on_collision'],
), options=option_list[config['policy']['option']], ), collision_distance=6, collision_penalty=100), lambda obs: (obs - obs_min) / (obs_max - obs_min + 1e-10))
safe_actions_collision_method=config['trainenv']['safe_actions_collision_method'], ), options=option_list[config['policy']['option']],
abort_unsafe_collision_method=config['trainenv']['abort_unsafe_collision_method'], safe_actions_collision_method=config['trainenv']['safe_actions_collision_method'],
) for _ in range(20)] for track in range(4)],[]) abort_unsafe_collision_method=config['trainenv']['abort_unsafe_collision_method'],
) for _ in range(60)]
elif config['experiment'] == 'B':
envs = sum([[SafeOptionsEnv(Setobs(
TransformObservation(CollisionPenaltyWrapper(IntersimpleLidarFlatRandom(
n_rays=5,
reward=functools.partial(
speed_reward,
collision_penalty=0
),
check_collisions=True,
stop_on_collision=config['trainenv']['stop_on_collision'], track=track,
), collision_distance=6, collision_penalty=100), lambda obs: (obs - obs_min) / (obs_max - obs_min + 1e-10))
), options=option_list[config['policy']['option']],
safe_actions_collision_method=config['trainenv']['safe_actions_collision_method'],
abort_unsafe_collision_method=config['trainenv']['abort_unsafe_collision_method'],
) for _ in range(15)] for track in range(4)],[])
else:
raise NotImplementedError
env_fn = lambda i: envs[i] env_fn = lambda i: envs[i]
policy = SetMaskedDiscretePolicy(env_fn(0).action_space.n, policy = SetMaskedDiscretePolicy(env_fn(0).action_space.n,
n_hidden_layers=config['policy']['n_hidden_layers'], n_hidden_layers=config['policy']['n_hidden_layers'],
hidden_layer_size=config['policy']['hidden_layer_size'], hidden_layer_size=config['policy']['hidden_layer_size'],
activation=config['policy']['activation'] ) # config net architecture activation=activations[config['policy']['activation']] ) # config net architecture
pi_opt = torch.optim.Adam(policy.parameters(), lr=config['policy']['learning_rate']) pi_opt = torch.optim.Adam(policy.parameters(), lr=config['policy']['learning_rate'])
pi_lr_scheduler = torch.optim.lr_scheduler.ExponentialLR(pi_opt, gamma=config['policy']['learning_rate_decay']) pi_lr_scheduler = torch.optim.lr_scheduler.ExponentialLR(pi_opt, gamma=config['policy']['learning_rate_decay'])
@@ -80,21 +99,25 @@ def training_function(config):
n_hidden_layers_element=config['discriminator']['n_hidden_layers_element'], n_hidden_layers_element=config['discriminator']['n_hidden_layers_element'],
n_hidden_layers_global=config['discriminator']['n_hidden_layers_global'], n_hidden_layers_global=config['discriminator']['n_hidden_layers_global'],
hidden_layer_size=config['discriminator']['hidden_layer_size'], hidden_layer_size=config['discriminator']['hidden_layer_size'],
activation=config['discriminator']['activation'], activation=activations[config['discriminator']['activation']],
) )
disc_opt = torch.optim.Adam(discriminator.parameters(), lr=config['discriminator']['learning_rate'], weight_decay=config['discriminator']['weight_decay']) disc_opt = torch.optim.Adam(discriminator.parameters(), lr=config['discriminator']['learning_rate'], weight_decay=config['discriminator']['weight_decay'])
expert_data = [ if config['experiment'] == 'A':
torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track0.pt')), expert_data = torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track0.pt'))
torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track1.pt')), elif config['experiment'] == 'B':
torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track2.pt')), expert_data = [
torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track3.pt')), torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track0.pt')),
] torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track1.pt')),
d0 = [d[0] for d in expert_data] torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track2.pt')),
d1 = [d[1] for d in expert_data] torch.load(os.path.join(DIR, 'intersimple-expert-data-setobs2-loc0-track3.pt')),
d2 = [d[2] for d in expert_data] ]
d3 = [d[3] for d in expert_data] d0 = [d[0] for d in expert_data]
expert_data = (torch.cat(d0), torch.cat(d1), torch.cat(d2), torch.cat(d3)) d1 = [d[1] for d in expert_data]
d2 = [d[2] for d in expert_data]
d3 = [d[3] for d in expert_data]
expert_data = (torch.cat(d0), torch.cat(d1), torch.cat(d2), torch.cat(d3))
expert_data = Buffer(*expert_data) expert_data = Buffer(*expert_data)
def callback(info): def callback(info):
@@ -118,8 +141,8 @@ def training_function(config):
value=value, value=value,
v_opt=v_opt, v_opt=v_opt,
v_iters=config['value']['iterations_per_epoch'], v_iters=config['value']['iterations_per_epoch'],
epochs=20, #300, #200 FIXME epochs=config['train_epochs'],
rollout_episodes=60, rollout_episodes=60,
rollout_steps=60, rollout_steps=60,
gamma=0.99, gamma=0.99,
gae_lambda=0.9, gae_lambda=0.9,
@@ -134,59 +157,49 @@ def training_function(config):
# save model # save model
torch.save(policy.state_dict(), 'policy_final.pt') torch.save(policy.state_dict(), 'policy_final.pt')
analysis = tune.run( if __name__ == '__main__':
training_function, import argparse
config={ parser = argparse.ArgumentParser()
'trainenv': { parser.add_argument('experiment', choices=['A', 'B'])
'stop_on_collision': False, #tune.grid_search([False, True]), parser.add_argument('--epochs', type=int, default=200)
'safe_actions_collision_method': 'circle', args = parser.parse_args()
'abort_unsafe_collision_method': 'circle',
},
'policy': {
'learning_rate': 3e-4, # tune.grid_search([3e-4]),
'learning_rate_decay': 1.0, #tune.grid_search([1.0]),
'clip_ratio': 0.2, #tune.grid_search([0.2]),
'iterations_per_epoch': 100, #tune.grid_search([100]),
'hidden_layer_size': 10, #tune.grid_search([10, 20, 40]), FIXME
'n_hidden_layers': 2, #tune.grid_search([2, 3, 4]), FIXME
'activation':torch.nn.Tanh,
'option': tune.grid_search(list(range(len(option_list))))
},
'value': {
'learning_rate': 1e-3, # tune.grid_search([1e-3]),
'iterations_per_epoch': 1000, #tune.grid_search([1000]),
},
'discriminator': {
'learning_rate': 1e-3, #tune.grid_search([1e-3]),
'weight_decay': 1e-4, #tune.grid_search([1e-4]),
'iterations_per_epoch': 100, #tune.grid_search([100]),
'n_hidden_layers_element': 3,
'n_hidden_layers_global': 2,
'hidden_layer_size': 10,
'activation': torch.nn.Tanh,
},
'seed': 0,
}
)
print('Best config: ', analysis.get_best_config(metric='gen_collision_rate', mode='min')) print('Running Tuning for Experiment %s'%(args.experiment))
analysis = tune.run(
# %% training_function,
# policy = SetMaskedDiscretePolicy(env_fn(0).action_space.n) config={
# policy(torch.zeros(env_fn(0).observation_space['observation'].shape), torch.zeros(env_fn(0).observation_space['safe_actions'].shape)) 'experiment': args.experiment,
# policy.load_state_dict(torch.load('sgail-ppo-options-setobs2.pt')) 'trainenv': {
'stop_on_collision': False,
# env = env_fn(0) 'safe_actions_collision_method': 'circle',
# obs = env.reset() 'abort_unsafe_collision_method': 'circle',
# env.render(mode='post') },
# for i in range(300): 'policy': {
# action = policy.sample(policy( 'learning_rate': 3e-4,
# torch.tensor(obs['observation'], dtype=torch.float32), 'learning_rate_decay': 1.0,
# torch.tensor(obs['safe_actions'], dtype=torch.float32), 'clip_ratio': 0.2,
# )) 'iterations_per_epoch': 100,
# obs, reward, done, _ = env.step(action, render_mode='post') 'hidden_layer_size': tune.grid_search([20, 40]),
# print('step', i, 'reward', reward, 'safe actions', obs['safe_actions']) 'n_hidden_layers': tune.grid_search([2, 3]),
# if done: 'activation':0,
# break 'option': tune.grid_search(list(range(len(option_list))))
# env.close() },
# %% 'value': {
'learning_rate': 1e-3,
'iterations_per_epoch': 1000,
},
'discriminator': {
'learning_rate': 1e-3,
'weight_decay': 1e-4,
'iterations_per_epoch': 100,
'n_hidden_layers_element': tune.grid_search([3,4]),
'n_hidden_layers_global': tune.grid_search([1,2]),
'hidden_layer_size': 10,
'activation': 0,
},
'train_epochs': args.epochs,
'seed': 0,
}
)
print('Best config: ', analysis.get_best_config(metric='gen_collision_rate', mode='min'))