forked from pokaxpoka/B_Pref
-
Notifications
You must be signed in to change notification settings - Fork 0
/
utils.py
268 lines (219 loc) · 7.89 KB
/
utils.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
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
import numpy as np
import torch
import torch.nn.functional as F
import gym
import os
import random
import math
import dmc2gym
import metaworld
import metaworld.envs.mujoco.env_dict as _env_dict
from collections import deque
from gym.wrappers.time_limit import TimeLimit
from rlkit.envs.wrappers import NormalizedBoxEnv
from collections import deque
from skimage.util.shape import view_as_windows
from torch import nn
from torch import distributions as pyd
def make_env(cfg):
"""Helper function to create dm_control environment"""
if cfg.env == 'ball_in_cup_catch':
domain_name = 'ball_in_cup'
task_name = 'catch'
else:
domain_name = cfg.env.split('_')[0]
task_name = '_'.join(cfg.env.split('_')[1:])
env = dmc2gym.make(domain_name=domain_name,
task_name=task_name,
seed=cfg.seed,
visualize_reward=False)
env.seed(cfg.seed)
assert env.action_space.low.min() >= -1
assert env.action_space.high.max() <= 1
return env
def ppo_make_env(env_id, seed):
"""Helper function to create dm_control environment"""
if env_id == 'ball_in_cup_catch':
domain_name = 'ball_in_cup'
task_name = 'catch'
else:
domain_name = env_id.split('_')[0]
task_name = '_'.join(env_id.split('_')[1:])
env = dmc2gym.make(domain_name=domain_name,
task_name=task_name,
seed=seed,
visualize_reward=True)
env.seed(seed)
assert env.action_space.low.min() >= -1
assert env.action_space.high.max() <= 1
return env
def tie_weights(src, trg):
assert type(src) == type(trg)
trg.weight = src.weight
trg.bias = src.bias
def make_metaworld_env(cfg):
env_name = cfg.env.replace('metaworld_','')
if env_name in _env_dict.ALL_V2_ENVIRONMENTS:
env_cls = _env_dict.ALL_V2_ENVIRONMENTS[env_name]
else:
env_cls = _env_dict.ALL_V1_ENVIRONMENTS[env_name]
env = env_cls()
env._freeze_rand_vec = False
env._set_task_called = True
env.seed(cfg.seed)
return TimeLimit(NormalizedBoxEnv(env), env.max_path_length)
def ppo_make_metaworld_env(env_id, seed):
env_name = env_id.replace('metaworld_','')
if env_name in _env_dict.ALL_V2_ENVIRONMENTS:
env_cls = _env_dict.ALL_V2_ENVIRONMENTS[env_name]
else:
env_cls = _env_dict.ALL_V1_ENVIRONMENTS[env_name]
env = env_cls()
env._freeze_rand_vec = False
env._set_task_called = True
env.seed(seed)
return TimeLimit(env, env.max_path_length)
class eval_mode(object):
def __init__(self, *models):
self.models = models
def __enter__(self):
self.prev_states = []
for model in self.models:
self.prev_states.append(model.training)
model.train(False)
def __exit__(self, *args):
for model, state in zip(self.models, self.prev_states):
model.train(state)
return False
class train_mode(object):
def __init__(self, *models):
self.models = models
def __enter__(self):
self.prev_states = []
for model in self.models:
self.prev_states.append(model.training)
model.train(True)
def __exit__(self, *args):
for model, state in zip(self.models, self.prev_states):
model.train(state)
return False
def soft_update_params(net, target_net, tau):
for param, target_param in zip(net.parameters(), target_net.parameters()):
target_param.data.copy_(tau * param.data +
(1 - tau) * target_param.data)
def set_seed_everywhere(seed):
torch.manual_seed(seed)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(seed)
np.random.seed(seed)
random.seed(seed)
def make_dir(*path_parts):
dir_path = os.path.join(*path_parts)
try:
os.mkdir(dir_path)
except OSError:
pass
return dir_path
def weight_init(m):
"""Custom weight init for Conv2D and Linear layers."""
if isinstance(m, nn.Linear):
nn.init.orthogonal_(m.weight.data)
if hasattr(m.bias, 'data'):
m.bias.data.fill_(0.0)
class MLP(nn.Module):
def __init__(self,
input_dim,
hidden_dim,
output_dim,
hidden_depth,
output_mod=None):
super().__init__()
self.trunk = mlp(input_dim, hidden_dim, output_dim, hidden_depth,
output_mod)
self.apply(weight_init)
def forward(self, x):
return self.trunk(x)
class TanhTransform(pyd.transforms.Transform):
domain = pyd.constraints.real
codomain = pyd.constraints.interval(-1.0, 1.0)
bijective = True
sign = +1
def __init__(self, cache_size=1):
super().__init__(cache_size=cache_size)
@staticmethod
def atanh(x):
return 0.5 * (x.log1p() - (-x).log1p())
def __eq__(self, other):
return isinstance(other, TanhTransform)
def _call(self, x):
return x.tanh()
def _inverse(self, y):
# We do not clamp to the boundary here as it may degrade the performance of certain algorithms.
# one should use `cache_size=1` instead
return self.atanh(y)
def log_abs_det_jacobian(self, x, y):
# We use a formula that is more numerically stable, see details in the following link
# https://github.com/tensorflow/probability/commit/ef6bb176e0ebd1cf6e25c6b5cecdd2428c22963f#diff-e120f70e92e6741bca649f04fcd907b7
return 2.0 * (math.log(2.0) - x - F.softplus(-2.0 * x))
class SquashedNormal(pyd.transformed_distribution.TransformedDistribution):
def __init__(self, loc, scale):
self.loc = loc
self.scale = scale
self.base_dist = pyd.Normal(loc, scale)
transforms = [TanhTransform()]
super().__init__(self.base_dist, transforms)
@property
def mean(self):
mu = self.loc
for tr in self.transforms:
mu = tr(mu)
return mu
class TorchRunningMeanStd:
def __init__(self, epsilon=1e-4, shape=(), device=None):
self.mean = torch.zeros(shape, device=device)
self.var = torch.ones(shape, device=device)
self.count = epsilon
def update(self, x):
with torch.no_grad():
batch_mean = torch.mean(x, axis=0)
batch_var = torch.var(x, axis=0)
batch_count = x.shape[0]
self.update_from_moments(batch_mean, batch_var, batch_count)
def update_from_moments(self, batch_mean, batch_var, batch_count):
self.mean, self.var, self.count = update_mean_var_count_from_moments(
self.mean, self.var, self.count, batch_mean, batch_var, batch_count
)
@property
def std(self):
return torch.sqrt(self.var)
def update_mean_var_count_from_moments(
mean, var, count, batch_mean, batch_var, batch_count
):
delta = batch_mean - mean
tot_count = count + batch_count
new_mean = mean + delta + batch_count / tot_count
m_a = var * count
m_b = batch_var * batch_count
M2 = m_a + m_b + torch.pow(delta, 2) * count * batch_count / tot_count
new_var = M2 / tot_count
new_count = tot_count
return new_mean, new_var, new_count
def mlp(input_dim, hidden_dim, output_dim, hidden_depth, output_mod=None):
if hidden_depth == 0:
mods = [nn.Linear(input_dim, output_dim)]
else:
mods = [nn.Linear(input_dim, hidden_dim), nn.ReLU(inplace=True)]
for i in range(hidden_depth - 1):
mods += [nn.Linear(hidden_dim, hidden_dim), nn.ReLU(inplace=True)]
mods.append(nn.Linear(hidden_dim, output_dim))
if output_mod is not None:
mods.append(output_mod)
trunk = nn.Sequential(*mods)
return trunk
def to_np(t):
if t is None:
return None
elif t.nelement() == 0:
return np.array([])
else:
return t.cpu().detach().numpy()