-
Notifications
You must be signed in to change notification settings - Fork 5
/
main.py
207 lines (176 loc) · 6.06 KB
/
main.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
from __future__ import division
from __future__ import print_function
import argparse
import copy
import os
import random
import numpy as np
import torch
import torch.nn.functional as F
import torch.optim as optim
from utils import load_data
from models import GCN, GAT, MLP, GenGNN
# Training settings
parser = argparse.ArgumentParser()
# General configs.
parser.add_argument("--dataset", default="cora")
parser.add_argument("--model", default="gcn")
parser.add_argument("--num_labels_per_class", type=int, default=20)
parser.add_argument("--result_path", default="results")
parser.add_argument("--seed", type=int, default=0, help="Random seed.")
parser.add_argument(
'--missing_edge',
action='store_true',
default=False,
help='Missing edge in test set.')
parser.add_argument(
"--epochs", type=int, default=2000, help="Number of epochs to train.")
parser.add_argument(
"--patience", type=int, default=200, help="Early stopping patience.")
parser.add_argument("--device", default="cuda")
parser.add_argument("--verbose", type=int, default=1, help="Verbose.")
# Common hyper-parameters.
parser.add_argument(
"--lr", type=float, default=0.01, help="Initial learning rate.")
parser.add_argument(
"--weight_decay",
type=float,
default=5e-4,
help="Weight decay (L2 loss on parameters).")
parser.add_argument(
"--hidden", type=int, default=64, help="Number of hidden units.")
parser.add_argument(
"--dropout",
type=float,
default=0.5,
help="Dropout rate (1 - keep probability).")
parser.add_argument("--activation", default="relu")
# GAT hyper-parameters.
parser.add_argument(
"--num_heads", type=int, default=8, help="Number of heads.")
# Generative model hyper-parameters.
parser.add_argument(
"--lamda",
type=float,
default=1.0,
help="Lambda coefficient for nll_discriminative.")
parser.add_argument(
"--neg_ratio", type=float, default=1.0, help="Negative sample ratio.")
# LSM hyper-parameters.
parser.add_argument(
"--hidden_x",
type=int,
default=2,
help="Number of hidden units for x_enc.")
# SBM hyper-parameters.
parser.add_argument("--p0", type=float, default=0.9, help="p0 in SBM.")
parser.add_argument("--p1", type=float, default=0.1, help="p1 in SBM.")
args = parser.parse_args()
random.seed(args.seed)
np.random.seed(args.seed)
torch.manual_seed(args.seed)
if torch.cuda.is_available():
torch.cuda.manual_seed(args.seed)
data = load_data(
dataset=args.dataset,
num_labels_per_class=args.num_labels_per_class,
missing_edge=args.missing_edge,
verbose=args.verbose).to(args.device)
model_args = {
"num_features": data.num_features,
"num_classes": data.num_classes,
"hidden_size": args.hidden,
"dropout": args.dropout,
"activation": args.activation
}
if args.model == "gcn":
model = GCN(**model_args)
elif args.model == "gat":
model_args["num_heads"] = args.num_heads
model_args["hidden_size"] = int(args.hidden / args.num_heads)
model = GAT(**model_args)
elif args.model == "mlp":
model = MLP(**model_args)
else:
gen_type, post_type = args.model.split("_")
gen_config = copy.deepcopy(model_args)
gen_config["type"] = gen_type
gen_config["neg_ratio"] = args.neg_ratio
if gen_type == "lsm":
gen_config["hidden_x"] = args.hidden_x
if gen_type == "sbm":
gen_config["p0"] = args.p0
gen_config["p1"] = args.p1
post_config = copy.deepcopy(model_args)
post_config["type"] = post_type
if post_type == "gat":
post_config["num_heads"] = args.num_heads
post_config["hidden_size"] = int(args.hidden / args.num_heads)
model = GenGNN(gen_config, post_config)
model = model.to(args.device)
optimizer = optim.Adam(
model.parameters(), lr=args.lr, weight_decay=args.weight_decay)
if hasattr(model, "gen"):
def train_loss_fn(model, data):
post_y_log_prob = model(data)
nll_generative = model.gen.nll_generative(data, post_y_log_prob)
nll_discriminative = F.nll_loss(post_y_log_prob[data.train_mask],
data.y[data.train_mask])
return nll_generative + args.lamda * nll_discriminative
else:
def train_loss_fn(model, data):
return F.nll_loss(
model(data)[data.train_mask], data.y[data.train_mask])
def val_loss_fn(logits, data):
return F.nll_loss(logits[data.val_mask], data.y[data.val_mask]).item()
def train():
model.train()
optimizer.zero_grad()
loss = train_loss_fn(model, data)
loss.backward()
optimizer.step()
def test():
model.eval()
logits = model(data)
val_loss = val_loss_fn(logits, data)
accs = []
for _, mask in data('train_mask', 'val_mask', 'test_mask'):
pred = logits[mask].max(1)[1]
acc = pred.eq(data.y[mask]).sum().item() / mask.sum().item()
accs.append(acc)
return val_loss, accs
# Training.
patience = args.patience
best_val_loss = np.inf
selected_accs = None
for epoch in range(1, args.epochs):
if patience < 0:
break
train()
val_loss, accs = test()
if val_loss < best_val_loss:
best_val_loss = val_loss
selected_accs = accs
patience = args.patience
if args.verbose > 0:
log = 'Epoch: {:03d}, Train: {:.4f}, Val: {:.4f}, Test: {:.4f}'
print(log.format(epoch, *accs))
patience -= 1
# Save results.
if args.verbose < 1:
result_path = os.path.join(
args.result_path,
"%s/nl%d" % (args.dataset, args.num_labels_per_class))
if not os.path.exists(result_path):
os.makedirs(result_path)
results = "vacc_%.4f_tacc_%.4f_seed_%d" % (
selected_accs[1], selected_accs[2], args.seed)
model_settng = "model_%s_lr_%.6f_h_%03d_l2_%.6f" % (
args.model, args.lr, args.hidden, args.weight_decay)
misc_hp = "act_%s_nh_%d_lambda_%.2f_nr_%.2f_hx_%d_p0_%.2f_p1_%.2f" % (
args.activation, args.num_heads, args.lamda, args.neg_ratio,
args.hidden_x, args.p0, args.p1)
fname = os.path.join(result_path,
"_".join([results, model_settng, misc_hp]))
with open(fname, "w") as f:
pass