-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathtrain.py
315 lines (232 loc) · 12.2 KB
/
train.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
313
314
315
import torch
print(torch.__version__)
import os
#import torch_geometric
#import torch_scatter
#import torch.nn as nn
#import torch.nn.functional as F
#import torch_geometric.nn as pyg_nn
#import torch_geometric.utils as pyg_utils
#from torch import Tensor
#from typing import Union, Tuple, Optional
#from torch.nn import Parameter, Linear, Sequential, LayerNorm, ReLU
#from torch_sparse import SparseTensor, set_diag
#from torch_geometric.nn.conv import MessagePassing
#from torch_geometric.utils import remove_self_loops, add_self_loops, softmax, degree
import argparse
import random
import pandas as pd
import mesh_model
import stats
import torch.optim as optim
def build_optimizer(args, params):
weight_decay = args.weight_decay
filter_fn = filter(lambda p : p.requires_grad, params)
if args.opt == 'adam':
optimizer = optim.Adam(filter_fn, lr=args.lr, weight_decay=weight_decay)
elif args.opt == 'sgd':
optimizer = optim.SGD(filter_fn, lr=args.lr, momentum=0.95, weight_decay=weight_decay)
elif args.opt == 'rmsprop':
optimizer = optim.RMSprop(filter_fn, lr=args.lr, weight_decay=weight_decay)
elif args.opt == 'adagrad':
optimizer = optim.Adagrad(filter_fn, lr=args.lr, weight_decay=weight_decay)
if args.opt_scheduler == 'none':
return None, optimizer
elif args.opt_scheduler == 'step':
scheduler = optim.lr_scheduler.StepLR(optimizer, step_size=args.opt_decay_step, gamma=args.opt_decay_rate)
elif args.opt_scheduler == 'cos':
scheduler = optim.lr_scheduler.CosineAnnealingLR(optimizer, T_max=args.opt_restart)
return scheduler, optimizer
import time
#import networkx as nx
import numpy as np
#import torch
import torch.optim as optim
from tqdm import trange
import pandas as pd
import copy
#from torch_geometric.datasets import TUDataset
#from torch_geometric.datasets import Planetoid
from torch_geometric.data import DataLoader
#import torch_geometric.nn as pyg_nn
import matplotlib.pyplot as plt
def train(dataset, device, stats_list, args):
df = pd.DataFrame(columns=['epoch','train_loss','test_loss', 'velo_val_loss'])
model_name='model_nl'+str(args.num_layers)+'_bs'+str(args.batch_size) + \
'_hd'+str(args.hidden_dim)+'_ep'+str(args.epochs)+'_wd'+str(args.weight_decay) + \
'_lr'+str(args.lr)+'_shuff_'+str(args.shuffle)+'_tr'+str(args.train_size)+'_te'+str(args.test_size)
loader = DataLoader(dataset[:args.train_size], batch_size=args.batch_size, shuffle=False)
test_loader = DataLoader(dataset[args.train_size:], batch_size=args.batch_size, shuffle=False)
[mean_vec_x,std_vec_x,mean_vec_edge,std_vec_edge,mean_vec_y,std_vec_y] = stats_list
(mean_vec_x,std_vec_x,mean_vec_edge,std_vec_edge,mean_vec_y,std_vec_y)=(mean_vec_x.to(device),
std_vec_x.to(device),mean_vec_edge.to(device),std_vec_edge.to(device),mean_vec_y.to(device),std_vec_y.to(device))
# build model
num_node_features = dataset[0].x.shape[1]
num_edge_features = dataset[0].edge_attr.shape[1]
num_classes = 2 # the dynamic variables have the shape of 2 (velocity)
model = mesh_model.MeshGraphNet(num_node_features, num_edge_features, args.hidden_dim, num_classes,
args).to(device)
scheduler, opt = build_optimizer(args, model.parameters())
# train
losses = []
test_losses = []
velo_val_losses = []
best_test_loss = np.inf
best_model = None
for epoch in trange(args.epochs, desc="Training", unit="Epochs"):
total_loss = 0
model.train()
num_loops=0
for batch in loader:
#Note that normalization must be done before it's called. The unnormalized
#data needs to be preserved in order to correctly calculate the loss
batch=batch.to(device)
opt.zero_grad()
pred = model(batch,mean_vec_x,std_vec_x,mean_vec_edge,std_vec_edge)
#pred = pred[batch.train_mask]
#label = label[batch.train_mask]
loss = model.loss(pred,batch,mean_vec_y,std_vec_y)
loss.backward()
opt.step()
total_loss += loss.item()
num_loops+=1
total_loss /= num_loops
losses.append(total_loss)
if epoch % 10 == 0:
if (args.save_velo_val):
# save velocity evaluation
test_loss, velo_val_rmse = test(test_loader,device,model,mean_vec_x,std_vec_x,mean_vec_edge,
std_vec_edge,mean_vec_y,std_vec_y, args.save_velo_val)
velo_val_losses.append(velo_val_rmse.item())
else:
test_loss, _ = test(test_loader,device,model,mean_vec_x,std_vec_x,mean_vec_edge,
std_vec_edge,mean_vec_y,std_vec_y, args.save_velo_val)
test_losses.append(test_loss.item())
# saving model
if not os.path.isdir( args.checkpoint_dir ):
os.mkdir(args.checkpoint_dir)
PATH = os.path.join(args.checkpoint_dir, model_name+'.csv')
df.to_csv(PATH,index=False)
if test_loss < best_test_loss:
best_test_loss = test_loss
best_model = copy.deepcopy(model)
else:
test_losses.append(test_losses[-1])
velo_val_losses.append(velo_val_losses[-1])
if (args.save_velo_val):
df = df.append({'epoch': epoch,'train_loss': losses[-1],
'test_loss':test_losses[-1],
'velo_val_loss': velo_val_losses[-1]}, ignore_index=True)
else:
df = df.append({'epoch': epoch, 'train_loss': losses[-1], 'test_loss': test_losses[-1]}, ignore_index=True)
if(epoch%100==0):
if (args.save_velo_val):
print("train loss", str(round(total_loss, 2)),
"test loss", str(round(test_loss.item(), 2)),
"velo loss", str(round(velo_val_rmse.item(), 2)))
else:
print("train loss", str(round(total_loss,2)), "test loss", str(round(test_loss.item(),2)))
if(args.save_best_model):
PATH = os.path.join(args.checkpoint_dir, model_name+'.pt')
torch.save(best_model.state_dict(), PATH )
return test_losses, losses, velo_val_losses, best_model, best_test_loss, test_loader
def test(loader,device,test_model,
mean_vec_x,std_vec_x,mean_vec_edge,std_vec_edge,mean_vec_y,std_vec_y, is_validation,
delta_t=0.01, save_model_preds=False, model_type=None):
loss=0
velo_rmse = 0
num_loops=0
for data in loader:
data=data.to(device)
with torch.no_grad():
pred = test_model(data,mean_vec_x,std_vec_x,mean_vec_edge,std_vec_edge)
loss += test_model.loss(pred, data,mean_vec_y,std_vec_y)
if (is_validation):
normal = torch.tensor(0)
outflow = torch.tensor(5)
loss_mask = torch.logical_or((torch.argmax(data.x[:, 2:], dim=1) == torch.tensor(0)),
(torch.argmax(data.x[:, 2:], dim=1) == torch.tensor(5)))
eval_velo = data.x[:, 0:2] + pred[:] * delta_t
gs_velo = data.x[:, 0:2] + data.y[:] * delta_t
error = torch.sum((eval_velo - gs_velo) ** 2, axis=1)
velo_rmse += torch.sqrt(torch.mean(error[loss_mask]))
num_loops+=1
# if velocity is evaluated, return velo_rmse as 0
return loss/num_loops, velo_rmse/num_loops
def save_plots(args, losses, test_losses, velo_val_losses):
model_name='model_nl'+str(args.num_layers)+'_bs'+str(args.batch_size) + \
'_hd'+str(args.hidden_dim)+'_ep'+str(args.epochs)+'_wd'+str(args.weight_decay) + \
'_lr'+str(args.lr)+'_shuff_'+str(args.shuffle)+'_tr'+str(args.train_size)+'_te'+str(args.test_size)
if not os.path.isdir(args.postprocess_dir):
os.mkdir(args.postprocess_dir)
PATH = os.path.join(args.postprocess_dir, model_name + '.pdf')
f = plt.figure()
plt.title(args.dataset)
plt.plot(losses, label="training loss" + " - " + args.model_type)
plt.plot(test_losses, label="test loss" + " - " + args.model_type)
if (args.save_velo_val):
plt.plot(velo_val_losses, label="velocity loss" + " - " + args.model_type)
plt.legend()
plt.show()
f.savefig(PATH, bbox_inches='tight')
class objectview(object):
def __init__(self, d):
self.__dict__ = d
def main(args):
# Get the right path of dataset
DATA_FOLDER_NAME = '01_dataset\\cylinder_flow\\torch_dataset'
root_dir = os.getcwd()
dataset_dir = os.path.join(root_dir, DATA_FOLDER_NAME)
device = 'cuda' if torch.cuda.is_available() else 'cpu'
print('Getting {}...'.format(device))
'''for args in [
{'model_type': 'meshgraphnet', 'dataset': 'mini10', 'num_layers': 10,
'batch_size': 16, 'hidden_dim': 10, 'epochs': 5000,
'opt': 'adam', 'opt_scheduler': 'none', 'opt_restart': 0, 'weight_decay': 5e-4, 'lr': 0.001,
'train_size': 45, 'test_size': 5, 'shuffle': True, 'save_best_model': False, 'checkpoint_dir': './best_models/'},
]:
args = objectview(args)'''
if args.dataset == 'mini10':
file_path = os.path.join(dataset_dir, 'meshgraphnets_miniset5traj_vis.pt')
#stats_path = os.path.join(dataset_dir, 'meshgraphnets_miniset5traj_ms.pt')
dataset = torch.load(file_path)[:(args.train_size+args.test_size)] #, batch_size = args['batch_size'])
if(args.shuffle):
random.shuffle(dataset)
#dataset_stats=torch.load(stats_path)
#import pdb; pdb.set_trace()
else:
raise NotImplementedError("Unknown dataset")
## TODO: CHECK PERFORMANCE OF STAT CHANGES BY ITERATING THROUGH ALL DATASETS AND CHECKING
## THE MEAN AND VAR OF NORMALIZED DATA
stats_list = stats.get_stats(dataset)
test_losses, losses, velo_val_losses, best_model, best_test_loss, test_loader = train(dataset, device, stats_list, args)
print("Min test set loss: {0}".format(min(test_losses)))
print("Minimum loss: {0}".format(min(losses)))
if (args.save_velo_val):
print("Minimum velocity validation loss: {0}".format(min(velo_val_losses)))
# Run test for our best model to save the predictions!
#test(test_loader, best_model, is_validation=False, save_model_preds=True, model_type=model)
#print()
save_plots(args, losses, test_losses, velo_val_losses)
if __name__ == '__main__':
argparser = argparse.ArgumentParser()
argparser.add_argument('--model_type', type=str, help='', default='meshgraphnet')
argparser.add_argument('--dataset', type=str, help='', default='mini10')
argparser.add_argument('--num_layers', type=int, help='', default=10)
argparser.add_argument('--batch_size', type=int, help='', default=16)
argparser.add_argument('--hidden_dim', type=int, help='', default=10)
argparser.add_argument('--epochs', type=int, help='', default=5000)
argparser.add_argument('--opt', type=str, help='', default='adam')
argparser.add_argument('--opt_scheduler', type=str, help='', default='none')
argparser.add_argument('--opt_restart', type=int, help='', default=0)
argparser.add_argument('--weight_decay', type=float, help='', default=5e-4)
argparser.add_argument('--lr', type=float, help='', default=0.001)
argparser.add_argument('--train_size', type=int, help='', default=45)
argparser.add_argument('--test_size', type=int, help='', default=5)
argparser.add_argument('--shuffle', type=bool, help='', default=True)
argparser.add_argument('--save_velo_val', type=bool, help='', default=True)
argparser.add_argument('--save_best_model', type=bool, help='', default=True)
argparser.add_argument('--checkpoint_dir', type=str, help='', default='./best_models/')
argparser.add_argument('--postprocess_dir', type=str, help='', default='./2d_loss_plots/')
args = argparser.parse_args()
main(args)