This repository has been archived by the owner on Oct 27, 2020. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 52
/
run_soccer_paddpg.py
235 lines (207 loc) · 10.5 KB
/
run_soccer_paddpg.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
import logging
import os
import click
import time
import gym
import gym_soccer
import numpy as np
from gym.wrappers import Monitor
from common import ClickPythonLiteralOption
from common.soccer_domain import SoccerScaledParameterisedActionWrapper, kill_soccer_server
from agents.paddpg import PADDPGAgent
def pad_action(act, act_param):
action = np.zeros((7,))
action[0] = act
if act == 0:
action[[1, 2]] = act_param
elif act == 1:
action[3] = act_param
elif act == 2:
action[[4, 5]] = act_param
elif act == 3:
action[[6]] = act_param
else:
raise ValueError("Unknown action index '{}'".format(act))
return action
def evaluate(env, agent, episodes=10):
returns = []
timesteps = []
goals = []
for _ in range(episodes):
state = env.reset()
terminal = False
t = 0
total_reward = 0.
info = {'status': "NOT_SET"}
while not terminal:
t += 1
state = np.array(state, dtype=np.float32, copy=False)
act, act_param, all_actions, all_action_parameters = agent.act(state)
action = pad_action(act, act_param)
state, reward, terminal, info = env.step(action)
total_reward += reward
# print(info['status'])
goal = info['status'] == 'GOAL'
timesteps.append(t)
returns.append(total_reward)
goals.append(goal)
return np.column_stack((returns, timesteps, goals))
@click.command()
@click.option('--seed', default=0, help='Random seed.', type=int)
@click.option('--episodes', default=20000, help='Number of epsiodes.', type=int)
@click.option('--evaluation-episodes', default=1000, help='Episodes over which to evaluate after training.', type=int)
@click.option('--update-ratio', default=0.1, help='Ratio of updates to samples.', type=float)
@click.option('--batch-size', default=32, help='Minibatch size.', type=int)
@click.option('--gamma', default=0.99, help='Discount factor.', type=float)
@click.option('--beta', default=0.2, help='Averaging factor for on-policy and off-policy targets.', type=float)
@click.option('--inverting-gradients', default=True,
help='Use inverting gradients scheme instead of squashing function.', type=bool)
@click.option('--initial-memory-threshold', default=1000, help='Number of transitions required to start learning.',
type=int)
@click.option('--use-ornstein-noise', default=False,
help='Use Ornstein noise instead of epsilon-greedy with uniform random exploration.', type=bool)
@click.option('--replay-memory-size', default=500000, help='Replay memory size in transitions.', type=int)
@click.option('--epsilon-steps', default=1000, help='Number of episodes over which to linearly anneal epsilon.',
type=int)
@click.option('--epsilon-final', default=0.1, help='Final epsilon value.', type=float)
@click.option('--tau', default=0.001, help='Soft target network update averaging factor.', type=float)
@click.option('--learning-rate-actor', default=0.001, help="Actor network learning rate.", type=float)
@click.option('--learning-rate-critic', default=0.001, help="Critic network learning rate.", type=float)
@click.option('--clip-grad', default=1., help="Gradient clipping.", type=float) # default 10
@click.option('--n-step-returns', default=True, help="Use n-step returns.", type=bool)
@click.option('--scale-actions', default=True, help="Scale actions.", type=bool)
@click.option('--layers', default="[1024,512,256,128]", help='Duplicate action-parameter inputs.',
cls=ClickPythonLiteralOption)
@click.option('--save-dir', default="results/soccer", help='Output directory.', type=str)
@click.option('--title', default="PADDPG", help="Prefix of output files", type=str)
def run(seed, episodes, batch_size, gamma, beta, use_ornstein_noise, inverting_gradients, initial_memory_threshold,
replay_memory_size, tau, learning_rate_actor, learning_rate_critic, epsilon_steps, epsilon_final,
n_step_returns, clip_grad, scale_actions, layers, evaluation_episodes, update_ratio, save_dir, title):
kill_soccer_server()
# env = gym.make('Soccer-v0')
# env = gym.make('SoccerEmptyGoal-v0')
env = gym.make('SoccerScoreGoal-v0')
# env = ScaledStateWrapper(env)
if scale_actions:
env = SoccerScaledParameterisedActionWrapper(env)
dir = os.path.join(save_dir, title)
env = Monitor(env, directory=os.path.join(dir, str(seed)), video_callable=False, write_upon_reset=False, force=True)
# env.seed(seed)
np.random.seed(seed)
agent = PADDPGAgent(env.observation_space, env.action_space,
actor_kwargs={'hidden_layers': layers, 'init_type': "kaiming", 'init_std': 0.01,
'activation': 'leaky_relu'},
critic_kwargs={'hidden_layers': layers, 'init_type': "kaiming", 'init_std': 0.01,
'activation': 'leaky_relu'},
batch_size=batch_size,
learning_rate_actor=learning_rate_actor, # 0.0001
learning_rate_critic=learning_rate_critic, # 0.001
gamma=gamma, # 0.99
tau_actor=tau,
tau_critic=tau,
n_step_returns=n_step_returns,
epsilon_steps=epsilon_steps,
epsilon_final=epsilon_final,
replay_memory_size=replay_memory_size,
inverting_gradients=inverting_gradients,
initial_memory_threshold=initial_memory_threshold,
beta=beta,
clip_grad=clip_grad,
use_ornstein_noise=use_ornstein_noise,
adam_betas=(0.9, 0.999), # default 0.95,0.999
seed=seed)
print(agent)
max_steps = 15000
total_reward = 0.
returns = []
timesteps = []
goals = []
start_time_train = time.time()
from tqdm import tqdm
# for i in tqdm(range(episodes)):
for i in range(episodes):
info = {'status': "NOT_SET"}
state = env.reset()
state = np.array(state, dtype=np.float32, copy=False)
act, act_param, all_actions, all_action_parameters = agent.act(state)
action = pad_action(act, act_param)
episode_reward = 0.
agent.start_episode()
transitions = []
for j in range(max_steps):
ret = env.step(action)
next_state, reward, terminal, info = ret
next_state = np.array(next_state, dtype=np.float32, copy=False)
next_act, next_act_param, next_all_actions, next_all_action_parameters = agent.act(next_state)
next_action = pad_action(next_act, next_act_param)
# don't add individual steps, so we can calculate n-step returns at the end...
if n_step_returns:
transitions.append(
[state, np.concatenate((all_actions.data, all_action_parameters.data)).ravel(), reward,
next_state, np.concatenate((next_all_actions.data,
next_all_action_parameters.data)).ravel(), terminal])
else:
agent.step(state, (act, act_param, all_actions, all_action_parameters), reward, next_state,
(next_act, next_act_param, next_all_actions, next_all_action_parameters), terminal,
optimise=False)
act, act_param, all_actions, all_action_parameters = next_act, next_act_param, next_all_actions, next_all_action_parameters
action = next_action
state = next_state
episode_reward += reward
# env.render()
if terminal:
break
agent.end_episode()
# calculate n-step returns
if n_step_returns:
nsreturns = compute_n_step_returns(transitions, gamma)
for t, nsr in zip(transitions, nsreturns):
t.append(nsr)
agent.replay_memory.append(state=t[0], action=t[1], reward=t[2], next_state=t[3], next_action=t[4],
terminal=t[5], time_steps=None, n_step_return=nsr)
n_updates = int(update_ratio * j)
for _ in range(n_updates):
agent._optimize_td_loss()
returns.append(episode_reward)
timesteps.append(j)
goals.append(info['status'] == 'GOAL')
total_reward += episode_reward
if i % 100 == 0:
print('{0:5s} R:{1:.4f} r:{2:.4f}'.format(str(i + 1), total_reward / (i + 1), episode_reward))
end_time_train = time.time()
returns = env.get_episode_rewards()
np.save(os.path.join(dir, title + "{}".format(str(seed))), np.column_stack((returns, timesteps, goals)))
if evaluation_episodes > 0:
print("Evaluating agent over {} episodes".format(evaluation_episodes))
agent.epsilon_final = 0.
agent.epsilon = 0.
agent.noise = None
agent.actor.eval()
agent.critic.eval()
start_time_eval = time.time()
evaluation_results = evaluate(env, agent, evaluation_episodes) # returns, timesteps, goals
end_time_eval = time.time()
print("Ave. evaluation return =", sum(evaluation_results[:, 0]) / evaluation_results.shape[0])
print("Ave. timesteps =", sum(evaluation_results[:, 1]) / evaluation_results.shape[0])
goal_timesteps = evaluation_results[:, 1][evaluation_results[:, 2] == 1]
if len(goal_timesteps) > 0:
print("Ave. timesteps per goal =", sum(goal_timesteps) / evaluation_results.shape[0])
else:
print("Ave. timesteps per goal =", sum(goal_timesteps) / evaluation_results.shape[0])
print("Ave. goal prob. =", sum(evaluation_results[:, 2]) / evaluation_results.shape[0])
np.save(os.path.join(dir, title + "{}e".format(str(seed))), evaluation_results)
print("Evaluation time: %.2f seconds" % (end_time_eval - start_time_eval))
print("Training time: %.2f seconds" % (end_time_train - start_time_train))
print(agent)
env.close()
def compute_n_step_returns(episode_transitions, gamma):
n = len(episode_transitions)
n_step_returns = np.zeros((n,))
n_step_returns[n - 1] = episode_transitions[n - 1][2] # Q-value is just the final reward
for i in range(n - 2, 0, -1):
reward = episode_transitions[i][2]
target = n_step_returns[i + 1]
n_step_returns[i] = reward + gamma * target
return n_step_returns
if __name__ == '__main__':
run()