-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathPhase3_lstm-gnn-ppo_adv.py
228 lines (211 loc) · 13.7 KB
/
Phase3_lstm-gnn-ppo_adv.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
import argparse
import random
import torch
from torch.utils.tensorboard import SummaryWriter
from modules.dqn.dqn_utils import seed_everything
from modules.sim.graph_factory import GetWorldSet
from modules.ppo.ppo_custom import *
from modules.ppo.helpfuncs import CreateEnv, evaluate_lstm_ppo
from modules.rl.rl_utils import GetFullCoverageSample
from modules.rl.rl_policy import LSTM_GNN_PPO_Policy, LSTM_GNN_PPO_EMB_Policy
from modules.rl.rl_utils import EvaluatePolicy
from modules.sim.simdata_utils import SimulateAutomaticMode_PPO, SimulateInteractiveMode_PPO
from modules.ppo.ppo_wrappers import PPO_ActWrapper, PPO_ObsFlatWrapper
torch.set_num_threads(1) # Max #threads for torch to avoid inefficient util of cpu cores.
def main(args):
config, hp, tp = GetConfigs(args)
if config['train']:
train_env, _ = make_custom(config, num_envs=hp.parallel_rollouts, asynchronous=tp['asynchronous_environment'])
o=train_env.reset()
hp.max_possible_nodes = int(o[0,-4])
hp.max_possible_edges = int(o[0,-2])
assert int(o[0,-4]) == train_env.envs[0].env.max_possible_num_nodes
assert int(o[0,-2]) == train_env.envs[0].env.max_possible_num_edges
if config['demoruns']:
while True:
a = SimulateInteractiveMode_PPO(train_env.envs[0], filesave_with_time_suffix=False)
if a == 'Q': break
WriteTrainParamsToFile(config,hp,tp)
trainfuncmap={'FE':train_model_FE,'EMB':train_model_EMB,'None':train_model,'Dual':train_model,'DualCC':train_model}
trainfunc = trainfuncmap[config['lstm_type']]
for seed in config['seedrange']:
seed_everything(seed)
train_env.seed(seed)
logdir_=config['logdir']+'/SEED'+str(seed)
tp['writer'] = SummaryWriter(log_dir=f"{logdir_}/logs")
tp["base_checkpoint_path"]=f"{logdir_}/checkpoints/"
tp["seed_path"]=logdir_
#tp["workspace_path"]=logdir_
ppo_model, ppo_optimizer, iteration, stop_conditions = start_or_resume_from_checkpoint(train_env, config, hp, tp)
if seed == config['seed0']: WriteModelParamsToFile(config, ppo_model)
score = trainfunc(train_env, ppo_model, ppo_optimizer, iteration, stop_conditions, hp, tp)
if config['eval']:
train_env, env_all_list = make_custom(config, num_envs=1, asynchronous=tp['asynchronous_environment'])
env_ = train_env.envs[0]
hp.max_possible_nodes = train_env.envs[0].env.max_possible_num_nodes
hp.max_possible_edges = train_env.envs[0].env.max_possible_num_edges
seed = config['seed0']
logdir_=config['logdir']+'/SEED'+str(seed)
tp["base_checkpoint_path"]=f"{logdir_}/checkpoints/"
assert os.path.exists(tp['base_checkpoint_path'])
if config['demoruns']:
ppo_model, ppo_optimizer, max_checkpoint_iteration, stop_conditions = start_or_resume_from_checkpoint(train_env, config, hp, tp)
if config['lstm_type'] == 'EMB':
ppo_policy = LSTM_GNN_PPO_EMB_Policy(None, ppo_model, deterministic=tp['eval_deterministic'])
else:
ppo_policy = LSTM_GNN_PPO_Policy(None, ppo_model, deterministic=tp['eval_deterministic'])
while True:
entries=None#[5012,218,3903]
#demo_env = random.choice(evalenv)
a = SimulateAutomaticMode_PPO(env_, ppo_policy, t_suffix=False, entries=entries)
if a == 'Q': break
evalResults={}
evalName='trainset'+'_evaldet'+str(tp['eval_deterministic'])[0]
evalResults[evalName]={'num_graphs.........':[],'num_graph_instances':[],'avg_return.........':[],'success_rate.......':[],}
for seed in config['seedrange']:
logdir_=config['logdir']+'/SEED'+str(seed)
tp["base_checkpoint_path"]=f"{logdir_}/checkpoints/"
try:
assert os.path.exists(tp['base_checkpoint_path'])
except:
continue
ppo_model, ppo_optimizer, max_checkpoint_iteration, stop_conditions = start_or_resume_from_checkpoint(train_env, config, hp, tp)
if config['lstm_type'] == 'EMB':
ppo_policy = LSTM_GNN_PPO_EMB_Policy(None, ppo_model, deterministic=tp['eval_deterministic'])
else:
ppo_policy = LSTM_GNN_PPO_Policy(None, ppo_model, deterministic=tp['eval_deterministic'])
multiplier=1
if not tp['eval_deterministic']:
k=sum([len(k.world_pool) for k in env_all_list])
multiplier = max(1,20000//k)
result = evaluate_lstm_ppo(logdir=logdir_, config=config, env = env_all_list, ppo_policy=ppo_policy, eval_subdir=evalName, max_num_nodes=hp.max_possible_nodes, multiplier=multiplier)
num_unique_graphs, num_graph_instances, avg_return, success_rate = result
evalResults[evalName]['num_graphs.........'].append(num_unique_graphs)
evalResults[evalName]['num_graph_instances'].append(num_graph_instances)
evalResults[evalName]['avg_return.........'].append(avg_return)
evalResults[evalName]['success_rate.......'].append(success_rate)
for ename, results in evalResults.items():
OF = open(config['logdir']+'/Eval_det'+str(tp['eval_deterministic'])[0]+'_Results_over_seeds_'+ename+'.txt', 'w')
def printing(text):
print(text)
OF.write(text + "\n")
np.set_printoptions(formatter={'float':"{0:0.3f}".format})
printing('Results over seeds for evaluation on '+ename+'\n')
for category,values in results.items():
printing(category)
printing(' avg over seeds: '+str(np.mean(values)))
printing(' std over seeds: '+str(np.std(values)))
printing(' per seed: '+str(np.array(values))+'\n')
if config['test']:
evalResults={}
world_dict={ # [max_nodes,max_edges]
#'Manhattan5x5_DuplicateSetB':[25,300],
#'Manhattan3x3_WalkAround':[9,300],
#'MetroU3_e1t31_FixedEscapeInit':[33, 300],
#'full_solvable_3x3subs':[9,33],
'MemoryTaskU1':[8,16],
#'Manhattan3x3_PredictionExample':[9,9],
#'Manhattan5x5_FixedEscapeInit':[25,105],
#'Manhattan5x5_VariableEscapeInit':[25,105],
# 'MetroU3_e17tborder_FixedEscapeInit':[33,300],
# 'MetroU3_e17tborder_VariableEscapeInit':[33,300],
#'NWB_ROT_FixedEscapeInit':[2602,3640],
#'NWB_ROT_VariableEscapeInit':[2602,3640],
# 'NWB_test_FixedEscapeInit':[975,4000],
# 'NWB_test_VariableEscapeInit':[975,4000],
# 'NWB_UTR_FixedEscapeInit':[1182,4000],
# 'NWB_UTR_VariableEscapeInit':[1182,4000],
# 'SparseManhattan5x5':[25,105],
}
obs_evalmasks = ['None']#'prob_per_u_test','prob_per_u_test','prob_per_u_test','prob_per_u_test','prob_per_u_test'] # ['None']['prob_per_u']
obs_evalrates = [1.0]#0.9,.8,.7,.6,.5] # [1.][0.8]
for obs_mask, obs_rate in zip(obs_evalmasks, obs_evalrates):
for world_name in world_dict.keys():
evalName=world_name+'_obs'+obs_mask+'_evaldet'+str(tp['eval_deterministic'])[0]
if obs_mask != 'None': evalName += str(obs_rate)
if world_name == "full_solvable_3x3subs":
Etest=[0,1,2,3,4,5,6,7,8,9,10]
Utest=[1,2,3]
evalenv, _, _, _ = GetWorldSet('etUte0U0', 'nfm', U=Utest, E=Etest, edge_blocking=config['edge_blocking'], solve_select=config['solve_select'], reject_duplicates=False, nfm_func=modules.gnn.nfm_gen.nfm_funcs[config['nfm_func']], apply_wrappers=False, maxnodes=world_dict[world_name][0], maxedges=world_dict[world_name][1])
for i in range(len(evalenv)):
evalenv[i]=PPO_ObsFlatWrapper(evalenv[i], max_possible_num_nodes=world_dict[world_name][0], max_possible_num_edges=world_dict[world_name][1], obs_mask=obs_mask, obs_rate=obs_rate)
evalenv[i]=PPO_ActWrapper(evalenv[i])
env=evalenv[0]
else:
env = CreateEnv(world_name, max_nodes=world_dict[world_name][0], max_edges = world_dict[world_name][1], nfm_func_name = config['nfm_func'], var_targets=None, remove_world_pool=None, apply_wrappers=True, type_obs_wrap=config['type_obs_wrap'], obs_mask=obs_mask, obs_rate=obs_rate)
evalenv=[env]
hp.max_possible_nodes = world_dict[world_name][0]#env.max_possible_num_nodes
hp.max_possible_edges = world_dict[world_name][1]#env.max_possible_num_edges
def envf():
return env
env_ = SyncVectorEnv([envf])
evalResults[evalName]={'num_graphs.........':[],'num_graph_instances':[],'avg_return.........':[],'success_rate.......':[],}
for seed in config['seedrange']:
logdir_=config['logdir']+'/SEED'+str(seed)
tp["base_checkpoint_path"]=f"{logdir_}/checkpoints/"
try:
assert os.path.exists(tp['base_checkpoint_path'])
except:
continue
ppo_model, ppo_optimizer, max_checkpoint_iteration, stop_conditions = start_or_resume_from_checkpoint(env_, config, hp, tp)
if config['lstm_type'] == 'EMB':
policy = LSTM_GNN_PPO_EMB_Policy(None, ppo_model, deterministic=tp['eval_deterministic'])
else:
ppo_policy = LSTM_GNN_PPO_Policy(env, ppo_model, deterministic=tp['eval_deterministic'])
if config['demoruns']:
while True:
demoenv=random.choice(evalenv)
a = SimulateAutomaticMode_PPO(demoenv, ppo_policy, t_suffix=False, entries=None)
if a == 'Q': break
result = evaluate_lstm_ppo(logdir=logdir_, config=config, env=evalenv, ppo_policy=ppo_policy, eval_subdir=evalName, max_num_nodes=world_dict[world_name][0])
num_unique_graphs, num_graph_instances, avg_return, success_rate = result
evalResults[evalName]['num_graphs.........'].append(num_unique_graphs)
evalResults[evalName]['num_graph_instances'].append(num_graph_instances)
evalResults[evalName]['avg_return.........'].append(avg_return)
evalResults[evalName]['success_rate.......'].append(success_rate)
for ename, results in evalResults.items():
OF = open(config['logdir']+'/Eval_det'+str(tp['eval_deterministic'])[0]+'_Results_over_seeds_'+ename+'.txt', 'w')
def printing(text):
print(text)
OF.write(text + "\n")
np.set_printoptions(formatter={'float':"{0:0.3f}".format})
printing('Results over seeds for evaluation on '+ename+'\n')
for category,values in results.items():
printing(category)
printing(' avg over seeds: '+str(np.mean(values)))
printing(' std over seeds: '+str(np.std(values)))
printing(' per seed: '+str(np.array(values))+'\n')
if __name__ == '__main__':
parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('--train_on', default='None', type=str)
parser.add_argument('--batch_size', default=48, type=int)
parser.add_argument('--lr', default=5e-4, type=float)
parser.add_argument('--recurrent_seq_len', default=2, type=int)
parser.add_argument('--parallel_rollouts', default=1, type=int)
parser.add_argument('--rollout_steps', default=150, type=int)
parser.add_argument('--patience', default=500, type=int)
parser.add_argument('--obs_mask', default='None', type=str, help='U obervation masking type', choices=['None','freq','prob','prob_per_u'])
parser.add_argument('--obs_rate', default=1.0, type=float)
parser.add_argument('--emb_dim', default=64, type=int)
parser.add_argument('--lstm_type', default='None', type=str, choices=['None','EMB','FE','Dual','DualCC'])
parser.add_argument('--lstm_hdim', default=64, type=int)
parser.add_argument('--lstm_layers', default=1, type=int)
#parser.add_argument('--lstm_dropout', default=0.0, type=float)
parser.add_argument('--emb_iterT', default=2, type=int)
parser.add_argument('--nfm_func', default='NFM_ev_ec_t', type=str)
parser.add_argument('--qnet', default='gat2', type=str)
parser.add_argument('--critic', default='q', type=str, choices=['q','v']) # q=v value route, v=single value route
parser.add_argument('--train', type=lambda s: s.lower() in ['true', 't', 'yes', '1'])
parser.add_argument('--eval', type=lambda s: s.lower() in ['true', 't', 'yes', '1'])
parser.add_argument('--test', type=lambda s: s.lower() in ['true', 't', 'yes', '1'])
parser.add_argument('--num_seeds', default=5, type=int)
parser.add_argument('--seed0', default=10, type=int)
parser.add_argument('--demoruns', type=lambda s: s.lower() in ['true', 't', 'yes', '1'])
parser.add_argument('--num_step', default=-1, type=int)
parser.add_argument('--checkpoint_frequency', default=100, type=int)
parser.add_argument('--eval_deter', type=lambda s: s.lower() in ['true', 't', 'yes', '1'],default=True)
parser.add_argument('--type_obs_wrap', default='obs_flat', type=str)
parser.add_argument('--test_heur', type=lambda s: s.lower() in ['true', 't', 'yes', '1'],default=False)
parser.add_argument('--eval_rate', default=-1., type=float)
args=parser.parse_args()
main(args)