forked from tensorpack/tensorpack
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsimulator.py
312 lines (246 loc) · 9.45 KB
/
simulator.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
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
#!/usr/bin/env python
# -*- coding: utf-8 -*-
# File: simulator.py
# Author: Yuxin Wu <[email protected]>
import tensorflow as tf
import multiprocessing as mp
import time
import os
import threading
from abc import abstractmethod, ABCMeta
from collections import defaultdict
import six
from six.moves import queue
import zmq
from tensorpack.models.common import disable_layer_logging
from tensorpack.callbacks import Callback
from tensorpack.tfutils.varmanip import SessionUpdate
from tensorpack.predict import OfflinePredictor
from tensorpack.utils import logger
from tensorpack.utils.serialize import loads, dumps
from tensorpack.utils.concurrency import LoopThread, ensure_proc_terminate
__all__ = ['SimulatorProcess', 'SimulatorMaster',
'SimulatorProcessStateExchange',
'TransitionExperience']
class TransitionExperience(object):
""" A transition of state, or experience"""
def __init__(self, state, action, reward, **kwargs):
""" kwargs: whatever other attribute you want to save"""
self.state = state
self.action = action
self.reward = reward
for k, v in six.iteritems(kwargs):
setattr(self, k, v)
@six.add_metaclass(ABCMeta)
class SimulatorProcessBase(mp.Process):
def __init__(self, idx):
super(SimulatorProcessBase, self).__init__()
self.idx = int(idx)
self.name = u'simulator-{}'.format(self.idx)
self.identity = self.name.encode('utf-8')
@abstractmethod
def _build_player(self):
pass
class SimulatorProcessStateExchange(SimulatorProcessBase):
"""
A process that simulates a player and communicates to master to
send states and receive the next action
"""
def __init__(self, idx, pipe_c2s, pipe_s2c):
"""
:param idx: idx of this process
"""
super(SimulatorProcessStateExchange, self).__init__(idx)
self.c2s = pipe_c2s
self.s2c = pipe_s2c
def run(self):
player = self._build_player()
context = zmq.Context()
c2s_socket = context.socket(zmq.PUSH)
c2s_socket.setsockopt(zmq.IDENTITY, self.identity)
c2s_socket.set_hwm(2)
c2s_socket.connect(self.c2s)
s2c_socket = context.socket(zmq.DEALER)
s2c_socket.setsockopt(zmq.IDENTITY, self.identity)
# s2c_socket.set_hwm(5)
s2c_socket.connect(self.s2c)
state = player.current_state()
reward, isOver = 0, False
while True:
c2s_socket.send(dumps(
(self.identity, state, reward, isOver)),
copy=False)
action = loads(s2c_socket.recv(copy=False).bytes)
reward, isOver = player.action(action)
state = player.current_state()
# compatibility
SimulatorProcess = SimulatorProcessStateExchange
class SimulatorMaster(threading.Thread):
""" A base thread to communicate with all StateExchangeSimulatorProcess.
It should produce action for each simulator, as well as
defining callbacks when a transition or an episode is finished.
"""
class ClientState(object):
def __init__(self):
self.memory = [] # list of Experience
def __init__(self, pipe_c2s, pipe_s2c):
super(SimulatorMaster, self).__init__()
assert os.name != 'nt', "Doesn't support windows!"
self.daemon = True
self.name = 'SimulatorMaster'
self.context = zmq.Context()
self.c2s_socket = self.context.socket(zmq.PULL)
self.c2s_socket.bind(pipe_c2s)
self.c2s_socket.set_hwm(10)
self.s2c_socket = self.context.socket(zmq.ROUTER)
self.s2c_socket.bind(pipe_s2c)
self.s2c_socket.set_hwm(10)
# queueing messages to client
self.send_queue = queue.Queue(maxsize=100)
def f():
msg = self.send_queue.get()
self.s2c_socket.send_multipart(msg, copy=False)
self.send_thread = LoopThread(f)
self.send_thread.daemon = True
self.send_thread.start()
# make sure socket get closed at the end
def clean_context(soks, context):
for s in soks:
s.close()
context.term()
import atexit
atexit.register(clean_context, [self.c2s_socket, self.s2c_socket], self.context)
def run(self):
self.clients = defaultdict(self.ClientState)
try:
while True:
msg = loads(self.c2s_socket.recv(copy=False).bytes)
ident, state, reward, isOver = msg
# TODO check history and warn about dead client
client = self.clients[ident]
# check if reward&isOver is valid
# in the first message, only state is valid
if len(client.memory) > 0:
client.memory[-1].reward = reward
if isOver:
self._on_episode_over(ident)
else:
self._on_datapoint(ident)
# feed state and return action
self._on_state(state, ident)
except zmq.ContextTerminated:
logger.info("[Simulator] Context was terminated.")
@abstractmethod
def _on_state(self, state, ident):
"""response to state sent by ident. Preferrably an async call"""
@abstractmethod
def _on_episode_over(self, client):
""" callback when the client just finished an episode.
You may want to clear the client's memory in this callback.
"""
def _on_datapoint(self, client):
""" callback when the client just finished a transition
"""
def __del__(self):
self.context.destroy(linger=0)
# ------------------- the following code are not used at all. Just experimental
class SimulatorProcessDF(SimulatorProcessBase):
""" A simulator which contains a forward model itself, allowing
it to produce data points directly """
def __init__(self, idx, pipe_c2s):
super(SimulatorProcessDF, self).__init__(idx)
self.pipe_c2s = pipe_c2s
def run(self):
self.player = self._build_player()
self.ctx = zmq.Context()
self.c2s_socket = self.ctx.socket(zmq.PUSH)
self.c2s_socket.setsockopt(zmq.IDENTITY, self.identity)
self.c2s_socket.set_hwm(5)
self.c2s_socket.connect(self.pipe_c2s)
self._prepare()
for dp in self.get_data():
self.c2s_socket.send(dumps(dp), copy=False)
@abstractmethod
def _prepare(self):
pass
@abstractmethod
def get_data(self):
pass
class SimulatorProcessSharedWeight(SimulatorProcessDF):
""" A simulator process with an extra thread waiting for event,
and take shared weight from shm.
Start me under some CUDA_VISIBLE_DEVICES set!
"""
def __init__(self, idx, pipe_c2s, condvar, shared_dic, pred_config):
super(SimulatorProcessSharedWeight, self).__init__(idx, pipe_c2s)
self.condvar = condvar
self.shared_dic = shared_dic
self.pred_config = pred_config
def _prepare(self):
disable_layer_logging()
self.predictor = OfflinePredictor(self.pred_config)
with self.predictor.graph.as_default():
vars_to_update = self._params_to_update()
self.sess_updater = SessionUpdate(
self.predictor.session, vars_to_update)
# TODO setup callback for explore?
self.predictor.graph.finalize()
self.weight_lock = threading.Lock()
# start a thread to wait for notification
def func():
self.condvar.acquire()
while True:
self.condvar.wait()
self._trigger_evt()
self.evt_th = threading.Thread(target=func)
self.evt_th.daemon = True
self.evt_th.start()
def _trigger_evt(self):
with self.weight_lock:
self.sess_updater.update(self.shared_dic['params'])
logger.info("Updated.")
def _params_to_update(self):
# can be overwritten to update more params
return tf.trainable_variables()
class WeightSync(Callback):
""" Sync weight from main process to shared_dic and notify"""
def __init__(self, condvar, shared_dic):
self.condvar = condvar
self.shared_dic = shared_dic
def _setup_graph(self):
self.vars = self._params_to_update()
def _params_to_update(self):
# can be overwritten to update more params
return tf.trainable_variables()
def _before_train(self):
self._sync()
def _trigger_epoch(self):
self._sync()
def _sync(self):
logger.info("Updating weights ...")
dic = {v.name: v.eval() for v in self.vars}
self.shared_dic['params'] = dic
self.condvar.acquire()
self.condvar.notify_all()
self.condvar.release()
if __name__ == '__main__':
import random
from tensorpack.RL import NaiveRLEnvironment
class NaiveSimulator(SimulatorProcess):
def _build_player(self):
return NaiveRLEnvironment()
class NaiveActioner(SimulatorMaster):
def _get_action(self, state):
time.sleep(1)
return random.randint(1, 12)
def _on_episode_over(self, client):
# print("Over: ", client.memory)
client.memory = []
client.state = 0
name = 'ipc://whatever'
procs = [NaiveSimulator(k, name) for k in range(10)]
[k.start() for k in procs]
th = NaiveActioner(name)
ensure_proc_terminate(procs)
th.start()
time.sleep(100)