/
4frame_get_avg_score.py
200 lines (170 loc) · 8.1 KB
/
4frame_get_avg_score.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
# Baby Advantage Actor-Critic | Sam Greydanus | October 2017 | MIT License
from __future__ import print_function
import torch, os, gym, time, glob, argparse, sys
import numpy as np
from scipy.signal import lfilter
from scipy.misc import imsave
from scipy.misc import imresize
import torch.nn as nn
from torch.autograd import Variable
import torch.nn.functional as F
import torch.multiprocessing as mp
os.environ['OMP_NUM_THREADS'] = '1'
from collections import deque
def immsave(file, pixels, size=400):
imsave(file, imresize(pixels,size))
def get_args():
parser = argparse.ArgumentParser(description=None)
parser.add_argument('--env', default='SpaceInvaders-v0', type=str, help='gym environment')
parser.add_argument('--missing', default='none', type=str, help='gym environment')
parser.add_argument('--processes', default=20, type=int, help='number of processes to train with')
parser.add_argument('--render', default=0, type=int, help='renders the atari environment')
parser.add_argument('--test', default=False, type=bool, help='test mode sets lr=0, chooses most likely actions')
parser.add_argument('--lstm_steps', default=20, type=int, help='steps to train LSTM over')
parser.add_argument('--gpu', default=7, type=int, help='')
parser.add_argument('--lr', default=1e-4, type=float, help='learning rate')
parser.add_argument('--seed', default=1, type=int, help='seed random # generators (for reproducibility)')
parser.add_argument('--game_runs', default=30, type=int, help='# of games to play')
parser.add_argument('--fskip', default=8, type=int)
parser.add_argument('--latent', default=256, type=int)
parser.add_argument('--gamma', default=0.99, type=float, help='discount for gamma-discounted rewards')
parser.add_argument('--tau', default=1.0, type=float, help='discount for generalized advantage estimation')
parser.add_argument('--horizon', default=0.99, type=float, help='horizon for running averages')
parser.add_argument('--agent_file', default="", type=str, help='file to test')
parser.add_argument('--img_dir', type=str, default='4frame_vid')
return parser.parse_args()
discount = lambda x, gamma: lfilter([1],[1,-gamma],x[::-1])[::-1] # discounted rewards one liner
prepro = lambda img: imresize(img[35:195], (80,80)).astype(np.float32).mean(2).reshape(1,80,80)/255.
def printlog(args, s, end='\n', mode='a'):
print(s, end=end) ; f=open(args.save_dir+'log.txt',mode) ; f.write(s+'\n') ; f.close()
class NNPolicy(torch.nn.Module): # an actor-critic neural network
def __init__(self, num_actions, latent_size):
super(NNPolicy, self).__init__()
channels = 4
self.conv1 = nn.Conv2d(channels, 32, 3, stride=2, padding=1)
self.conv2 = nn.Conv2d(32, 32, 3, stride=2, padding=1)
self.conv3 = nn.Conv2d(32, 32, 3, stride=2, padding=1)
self.conv4 = nn.Conv2d(32, 32, 3, stride=2, padding=1)
self.linear = nn.Linear(32 * 5 * 5, latent_size)
self.critic_linear, self.actor_linear = nn.Linear(latent_size, 1), nn.Linear(latent_size, num_actions)
def forward(self, inputs):
x = F.elu(self.conv1(inputs)) ; x = F.elu(self.conv2(x))
x = F.elu(self.conv3(x)) ; x = F.elu(self.conv4(x))
x = self.linear(x.view(-1, 32 * 5 * 5))
return self.critic_linear(x), self.actor_linear(x)
def try_load(self, save_dir):
paths = glob.glob(save_dir + '*.tar')
step = 0
if len(paths) > 0:
ckpts = [int(s.split('.')[-2]) for s in paths]
ix = np.argmax(ckpts)
step = ckpts[ix]
self.load_state_dict(torch.load(paths[ix]))
print("\tno saved models") if step is 0 else print("\tloaded model: {}".format(paths[ix]))
return step
def ablate_screen(orig_img, section):
img = orig_img
#height first, then width
if section == "none":
return orig_img
if section == "bottom":
img[:,40:] = 0
elif section == "top":
img[:,:40] = 0
elif section == "barrier":
img[:,60:70] = 0
elif section == "agent":
img[:,70:] = 0
elif section == "left":
img[:,:,:40] = 0
elif section == "right":
img[:,:,40:] = 0
elif section == "center_column":
img[:,:,20:60] = 0
elif section == "center_row":
img[:,20:60,:] = 0
elif section == "stripe_column":
img[:,:,5:10 ] = 0
img[:,:,15:20] = 0
img[:,:,25:30] = 0
img[:,:,35:40] = 0
img[:,:,45:50] = 0
img[:,:,55:60] = 0
img[:,:,65:70] = 0
img[:,:,75:80] = 0
elif section == "stripe_row":
img[:,5:10 ,:] = 0
img[:,15:20,:] = 0
img[:,25:30,:] = 0
img[:,35:40,:] = 0
img[:,45:50,:] = 0
img[:,55:60,:] = 0
img[:,65:70,:] = 0
img[:,75:80,:] = 0
else:
raise "hey, you tried to ablate a screen that didnt exist"
return img
def train(model, args):
env = gym.make(args.env) # make a local (unshared) environment
env.unwrapped.frameskip = args.fskip
env.seed(args.seed )
torch.manual_seed(args.seed)
if args.render == 1: os.makedirs(args.img_dir, exist_ok=True)
bs = 1
img = prepro(env.reset())
img = ablate_screen(img, args.missing)
state = Variable(torch.Tensor(img).view(1,1,80,80)).cuda()
state_history = deque([state, state.clone(), state.clone(),state.clone()], maxlen=4)
start_time = last_disp_time = time.time()
episode_length, epr, eploss, done = 0, 0, 0, False # bookkeeping
np.set_printoptions(precision=4)
epr_list = []
clipped_reward = []
cur_clipped_reward = 0
while len(epr_list) < args.game_runs: # openai baselines uses 40M frames...we'll use 80M
state = torch.cat(list(state_history), dim=1)
episode_length += 1
value, logit = model(state)
logp = F.log_softmax(logit, dim=1)
action = logp.max(1)[1].data
#if episode_length == 432 or episode_length == 433:
# action = 5
new_frame, reward, done, _ = env.step(action)
#if args.render == 1: immsave(os.path.join(args.img_dir, "{:05d}.png".format(episode_length)),new_frame)
img = ablate_screen(prepro(new_frame), args.missing)
state_history.append(Variable(torch.Tensor(img).view(1,1,80,80)).cuda())
epr += reward
reward = np.clip(reward, -1, 1) # reward
if reward < 0: print("????")
cur_clipped_reward += reward
done = done or episode_length >= 1e4 # keep agent from playing one episode too long
if env.unwrapped.ale.lives() < 3: done = True
if done: # update shared data. maybe print info.
if args.render == 1:
print("{}, {}, {}, {}".format(args.seed, epr, cur_clipped_reward, episode_length))
exit()
epr_list.append(epr)
clipped_reward.append(cur_clipped_reward)
episode_length, epr, eploss = 0, 0, 0
cur_clipped_reward = 0
img = ablate_screen(prepro(env.reset()), args.missing)
state = Variable(torch.Tensor(img).view(bs,1,80,80).cuda())
state_history = deque([state, state.clone(), state.clone(),state.clone()], maxlen=4)
#print("latent size: {}".format(args.latent))
output = np.array(epr_list)
output2 = np.array(clipped_reward)
print("{},{},{},{},{}".format(args.latent,np.mean(output),np.std(output),np.mean(output2),np.std(output2)))
#print("scores: mean: {}, std: {}".format(np.mean(output),np.std(output)))
#print("rewards: mean: {}, std: {}".format(np.mean(output2),np.std(output2)))
if __name__ == "__main__":
args = get_args()
args.num_actions = gym.make(args.env).action_space.n # get the action space of this game
args.img_dir += os.path.basename(args.agent_file)
if args.fskip % 2 ==0 and args.env == 'SpaceInvaders-v0':
print("SpaceInvaders needs odd frameskip due to bullet alternations")
args.fskip = args.fskip - 1
torch.cuda.set_device(args.gpu)
torch.manual_seed(args.seed)
model = NNPolicy(args.num_actions, args.latent).cuda()
model.load_state_dict(torch.load(args.agent_file))
train(model, args)