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
|
[run] ogbg-molsider view=gin compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view gin --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.52242 val_adapt_rocauc=0.52242 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.57120 val_adapt_rocauc=0.57031 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.52140 val_adapt_rocauc=0.52140 adapt_steps=8.00 halt=0.12 train_steps=4.45
ep40 val_rocauc=0.54349 val_adapt_rocauc=0.54351 adapt_steps=7.93 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.52355 val_adapt_rocauc=0.52355 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.57651 val_adapt_rocauc=0.57763 adapt_steps=7.97 halt=0.12 train_steps=4.47
ep70 val_rocauc=0.59417 val_adapt_rocauc=0.59417 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.57162 val_adapt_rocauc=0.57204 adapt_steps=7.96 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.58977 val_adapt_rocauc=0.58977 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.59364 val_adapt_rocauc=0.59378 adapt_steps=7.99 halt=0.12 train_steps=4.50
[ogbg-molsider_gin_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rocauc': 0.5941723817458207} test={'rocauc': 0.5949106211087469} adaptive={'rocauc': 0.5949106211087469} steps=8.0
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_gin_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=gine compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view gine --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.52714 val_adapt_rocauc=0.52714 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.53410 val_adapt_rocauc=0.53410 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.55999 val_adapt_rocauc=0.55999 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.58039 val_adapt_rocauc=0.58039 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.59855 val_adapt_rocauc=0.59871 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.58511 val_adapt_rocauc=0.58521 adapt_steps=7.98 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.58872 val_adapt_rocauc=0.58872 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.59935 val_adapt_rocauc=0.59935 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.60537 val_adapt_rocauc=0.60495 adapt_steps=7.96 halt=0.12 train_steps=4.47
ep100 val_rocauc=0.61174 val_adapt_rocauc=0.61174 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_gine_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.6117392254931898} test={'rocauc': 0.6143023355631505} adaptive={'rocauc': 0.6147815453760646} steps=7.958041958041958
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_gine_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=gcn compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view gcn --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.53639 val_adapt_rocauc=0.53639 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.54446 val_adapt_rocauc=0.54446 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.56844 val_adapt_rocauc=0.56844 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.59523 val_adapt_rocauc=0.59523 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.58925 val_adapt_rocauc=0.58925 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.61196 val_adapt_rocauc=0.60952 adapt_steps=7.48 halt=0.13 train_steps=4.43
ep70 val_rocauc=0.60679 val_adapt_rocauc=0.60935 adapt_steps=7.81 halt=0.12 train_steps=4.49
ep80 val_rocauc=0.62425 val_adapt_rocauc=0.62328 adapt_steps=7.92 halt=0.12 train_steps=4.49
ep90 val_rocauc=0.63455 val_adapt_rocauc=0.63464 adapt_steps=7.98 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.63261 val_adapt_rocauc=0.63316 adapt_steps=7.90 halt=0.13 train_steps=4.48
[ogbg-molsider_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.634553246023819} test={'rocauc': 0.6445134528927392} adaptive={'rocauc': 0.644509359822436} steps=7.93006993006993
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=graphsage compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view graphsage --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.53993 val_adapt_rocauc=0.53993 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.51721 val_adapt_rocauc=0.51820 adapt_steps=7.81 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.57455 val_adapt_rocauc=0.57455 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.54410 val_adapt_rocauc=0.54400 adapt_steps=7.98 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.57426 val_adapt_rocauc=0.57426 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.57583 val_adapt_rocauc=0.57668 adapt_steps=7.83 halt=0.12 train_steps=4.49
ep70 val_rocauc=0.57077 val_adapt_rocauc=0.57333 adapt_steps=7.81 halt=0.13 train_steps=4.46
ep80 val_rocauc=0.60329 val_adapt_rocauc=0.60324 adapt_steps=7.98 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.60080 val_adapt_rocauc=0.60075 adapt_steps=7.99 halt=0.12 train_steps=4.48
ep100 val_rocauc=0.61496 val_adapt_rocauc=0.61461 adapt_steps=7.97 halt=0.13 train_steps=4.48
[ogbg-molsider_graphsage_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.6149561902831908} test={'rocauc': 0.6082036653138152} adaptive={'rocauc': 0.6072637395395359} steps=7.923076923076923
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_graphsage_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=gatv2 compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view gatv2 --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.51326 val_adapt_rocauc=0.51326 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.53070 val_adapt_rocauc=0.53070 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.52081 val_adapt_rocauc=0.52081 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.53950 val_adapt_rocauc=0.53893 adapt_steps=7.83 halt=0.12 train_steps=4.47
ep50 val_rocauc=0.55260 val_adapt_rocauc=0.55337 adapt_steps=7.90 halt=0.12 train_steps=4.48
ep60 val_rocauc=0.55241 val_adapt_rocauc=0.55241 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.56511 val_adapt_rocauc=0.56511 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.56634 val_adapt_rocauc=0.56626 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.57675 val_adapt_rocauc=0.57675 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep100 val_rocauc=0.57777 val_adapt_rocauc=0.57777 adapt_steps=8.00 halt=0.12 train_steps=4.49
[ogbg-molsider_gatv2_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.5777678714357024} test={'rocauc': 0.6337321824820535} adaptive={'rocauc': 0.6339582691134998} steps=7.958041958041958
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_gatv2_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=graphconv compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view graphconv --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.53249 val_adapt_rocauc=0.52907 adapt_steps=7.41 halt=0.12 train_steps=4.47
ep20 val_rocauc=0.53857 val_adapt_rocauc=0.53838 adapt_steps=7.96 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.55430 val_adapt_rocauc=0.55430 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.55775 val_adapt_rocauc=0.55775 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.53048 val_adapt_rocauc=0.53048 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.58757 val_adapt_rocauc=0.58768 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.58127 val_adapt_rocauc=0.58127 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.58969 val_adapt_rocauc=0.58958 adapt_steps=7.97 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.59525 val_adapt_rocauc=0.59512 adapt_steps=7.94 halt=0.13 train_steps=4.44
ep100 val_rocauc=0.61397 val_adapt_rocauc=0.61302 adapt_steps=7.97 halt=0.12 train_steps=4.45
[ogbg-molsider_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.6139745501285409} test={'rocauc': 0.6068834635828694} adaptive={'rocauc': 0.6064255172013142} steps=7.916083916083916
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=transformer compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view transformer --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.53954 val_adapt_rocauc=0.53954 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.54449 val_adapt_rocauc=0.54449 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.58517 val_adapt_rocauc=0.58517 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.57396 val_adapt_rocauc=0.57671 adapt_steps=7.88 halt=0.12 train_steps=4.45
ep50 val_rocauc=0.58068 val_adapt_rocauc=0.58068 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.60139 val_adapt_rocauc=0.60139 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep70 val_rocauc=0.61355 val_adapt_rocauc=0.61355 adapt_steps=8.00 halt=0.12 train_steps=4.45
ep80 val_rocauc=0.57128 val_adapt_rocauc=0.57180 adapt_steps=7.95 halt=0.12 train_steps=4.48
ep90 val_rocauc=0.62410 val_adapt_rocauc=0.62410 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.62875 val_adapt_rocauc=0.62875 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_transformer_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.6287543036419472} test={'rocauc': 0.621503855814713} adaptive={'rocauc': 0.6212391169096168} steps=7.951048951048951
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_transformer_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=pna compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view pna --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep10 val_rocauc=0.53705 val_adapt_rocauc=0.53705 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep20 val_rocauc=0.49938 val_adapt_rocauc=0.49938 adapt_steps=8.00 halt=0.12 train_steps=4.43
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep30 val_rocauc=0.50914 val_adapt_rocauc=0.50914 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep40 val_rocauc=0.52533 val_adapt_rocauc=0.52533 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep50 val_rocauc=0.56368 val_adapt_rocauc=0.56368 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep60 val_rocauc=0.58102 val_adapt_rocauc=0.58102 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep70 val_rocauc=0.55427 val_adapt_rocauc=0.55427 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep80 val_rocauc=0.56861 val_adapt_rocauc=0.56861 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep90 val_rocauc=0.54785 val_adapt_rocauc=0.54785 adapt_steps=8.00 halt=0.12 train_steps=4.50
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='min')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
/orion/u/oscarwan/rrog-gnn-runner/.venv/lib/python3.13/site-packages/torch_geometric/utils/_scatter.py:91: UserWarning: The usage of `scatter(reduce='max')` can be accelerated via the 'torch-scatter' package, but it was not found
warnings.warn(
ep100 val_rocauc=0.56012 val_adapt_rocauc=0.56012 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_pna_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=60 val={'rocauc': 0.5810212759423172} test={'rocauc': 0.5397701280540309} adaptive={'rocauc': 0.5397701280540309} steps=8.0
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_pna_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=gen compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view gen --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.52015 val_adapt_rocauc=0.51409 adapt_steps=7.45 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.56979 val_adapt_rocauc=0.56338 adapt_steps=7.21 halt=0.13 train_steps=4.43
ep30 val_rocauc=0.57163 val_adapt_rocauc=0.57163 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.50300 val_adapt_rocauc=0.50300 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.55048 val_adapt_rocauc=0.55135 adapt_steps=7.97 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.55929 val_adapt_rocauc=0.55929 adapt_steps=8.00 halt=0.13 train_steps=4.50
ep70 val_rocauc=0.61128 val_adapt_rocauc=0.61133 adapt_steps=7.96 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.61143 val_adapt_rocauc=0.61143 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.59636 val_adapt_rocauc=0.59635 adapt_steps=7.96 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.59800 val_adapt_rocauc=0.59800 adapt_steps=8.00 halt=0.12 train_steps=4.49
[ogbg-molsider_gen_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.6114344272964439} test={'rocauc': 0.6071293608886619} adaptive={'rocauc': 0.6071293608886619} steps=8.0
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_gen_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=film compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view film --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.50663 val_adapt_rocauc=0.50663 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.53836 val_adapt_rocauc=0.53836 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.54072 val_adapt_rocauc=0.54072 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep40 val_rocauc=0.52242 val_adapt_rocauc=0.52242 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.52826 val_adapt_rocauc=0.52826 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.52041 val_adapt_rocauc=0.52041 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.53993 val_adapt_rocauc=0.53993 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.52758 val_adapt_rocauc=0.52758 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.53864 val_adapt_rocauc=0.53864 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.54079 val_adapt_rocauc=0.54079 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_film_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.5407876481859505} test={'rocauc': 0.5825278400429263} adaptive={'rocauc': 0.5825278400429263} steps=8.0
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_film_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=resgated compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view resgated --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.51625 val_adapt_rocauc=0.51625 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.55876 val_adapt_rocauc=0.55876 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.58353 val_adapt_rocauc=0.58374 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.59270 val_adapt_rocauc=0.59269 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.59117 val_adapt_rocauc=0.59117 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.59394 val_adapt_rocauc=0.59559 adapt_steps=7.96 halt=0.12 train_steps=4.48
ep70 val_rocauc=0.57628 val_adapt_rocauc=0.57628 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.59810 val_adapt_rocauc=0.59810 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.59935 val_adapt_rocauc=0.59919 adapt_steps=7.96 halt=0.12 train_steps=4.48
ep100 val_rocauc=0.60973 val_adapt_rocauc=0.60972 adapt_steps=7.97 halt=0.12 train_steps=4.48
[ogbg-molsider_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.609727680398977} test={'rocauc': 0.6245839801623748} adaptive={'rocauc': 0.624652550767547} steps=7.951048951048951
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=tag compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view tag --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.53875 val_adapt_rocauc=0.53875 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.55745 val_adapt_rocauc=0.55745 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep30 val_rocauc=0.52523 val_adapt_rocauc=0.52508 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.57020 val_adapt_rocauc=0.57005 adapt_steps=7.99 halt=0.12 train_steps=4.43
ep50 val_rocauc=0.57818 val_adapt_rocauc=0.57818 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.58430 val_adapt_rocauc=0.58478 adapt_steps=7.88 halt=0.13 train_steps=4.46
ep70 val_rocauc=0.62524 val_adapt_rocauc=0.62524 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep80 val_rocauc=0.58782 val_adapt_rocauc=0.58782 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.63291 val_adapt_rocauc=0.63291 adapt_steps=8.00 halt=0.13 train_steps=4.50
ep100 val_rocauc=0.63016 val_adapt_rocauc=0.63032 adapt_steps=7.97 halt=0.13 train_steps=4.47
[ogbg-molsider_tag_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.6329113613630306} test={'rocauc': 0.6060675375955825} adaptive={'rocauc': 0.6066839663477379} steps=7.909090909090909
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_tag_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=sgc compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view sgc --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.51769 val_adapt_rocauc=0.51769 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.54554 val_adapt_rocauc=0.54554 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.52433 val_adapt_rocauc=0.52433 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.54839 val_adapt_rocauc=0.54757 adapt_steps=7.82 halt=0.12 train_steps=4.40
ep50 val_rocauc=0.56667 val_adapt_rocauc=0.56667 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.56001 val_adapt_rocauc=0.56001 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.56325 val_adapt_rocauc=0.56263 adapt_steps=7.93 halt=0.12 train_steps=4.38
ep80 val_rocauc=0.57410 val_adapt_rocauc=0.57405 adapt_steps=7.99 halt=0.12 train_steps=4.49
ep90 val_rocauc=0.58333 val_adapt_rocauc=0.58333 adapt_steps=7.92 halt=0.13 train_steps=4.47
ep100 val_rocauc=0.58659 val_adapt_rocauc=0.58673 adapt_steps=7.91 halt=0.13 train_steps=4.47
[ogbg-molsider_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.5865936638036943} test={'rocauc': 0.624583829207758} adaptive={'rocauc': 0.6241791163198843} steps=7.881118881118881
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=cheb compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view cheb --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.52959 val_adapt_rocauc=0.53014 adapt_steps=7.83 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.59000 val_adapt_rocauc=0.59000 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.57720 val_adapt_rocauc=0.57720 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep40 val_rocauc=0.58010 val_adapt_rocauc=0.58016 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.56105 val_adapt_rocauc=0.56396 adapt_steps=7.42 halt=0.12 train_steps=4.47
ep60 val_rocauc=0.60272 val_adapt_rocauc=0.60144 adapt_steps=7.73 halt=0.13 train_steps=4.46
ep70 val_rocauc=0.63704 val_adapt_rocauc=0.63704 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep80 val_rocauc=0.63367 val_adapt_rocauc=0.63378 adapt_steps=7.95 halt=0.12 train_steps=4.46
ep90 val_rocauc=0.64091 val_adapt_rocauc=0.64157 adapt_steps=7.88 halt=0.13 train_steps=4.47
ep100 val_rocauc=0.63703 val_adapt_rocauc=0.63925 adapt_steps=7.80 halt=0.13 train_steps=4.46
[ogbg-molsider_cheb_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.6409076943311921} test={'rocauc': 0.6149744329414113} adaptive={'rocauc': 0.6144652607315165} steps=7.8671328671328675
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_cheb_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=arma compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view arma --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.51335 val_adapt_rocauc=0.51335 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.51105 val_adapt_rocauc=0.50798 adapt_steps=7.59 halt=0.13 train_steps=4.47
ep30 val_rocauc=0.54401 val_adapt_rocauc=0.54401 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.57421 val_adapt_rocauc=0.57421 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.58087 val_adapt_rocauc=0.58094 adapt_steps=7.98 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.58740 val_adapt_rocauc=0.58740 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.62738 val_adapt_rocauc=0.62725 adapt_steps=7.99 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.66049 val_adapt_rocauc=0.66028 adapt_steps=7.93 halt=0.13 train_steps=4.46
ep90 val_rocauc=0.63467 val_adapt_rocauc=0.63512 adapt_steps=7.97 halt=0.13 train_steps=4.45
ep100 val_rocauc=0.64614 val_adapt_rocauc=0.64687 adapt_steps=7.94 halt=0.12 train_steps=4.50
[ogbg-molsider_arma_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.6604919913889579} test={'rocauc': 0.5589855993934405} adaptive={'rocauc': 0.5580832949005652} steps=7.895104895104895
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_arma_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=mf compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view mf --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.52188 val_adapt_rocauc=0.52188 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep20 val_rocauc=0.55853 val_adapt_rocauc=0.55853 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.55126 val_adapt_rocauc=0.55126 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.55645 val_adapt_rocauc=0.55527 adapt_steps=7.83 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.55799 val_adapt_rocauc=0.55799 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.59604 val_adapt_rocauc=0.59534 adapt_steps=7.92 halt=0.12 train_steps=4.49
ep70 val_rocauc=0.57920 val_adapt_rocauc=0.57921 adapt_steps=7.96 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.63001 val_adapt_rocauc=0.62920 adapt_steps=7.96 halt=0.12 train_steps=4.49
ep90 val_rocauc=0.61962 val_adapt_rocauc=0.61885 adapt_steps=7.95 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.61522 val_adapt_rocauc=0.61430 adapt_steps=7.93 halt=0.12 train_steps=4.47
[ogbg-molsider_mf_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.630007512812611} test={'rocauc': 0.5843445908586726} adaptive={'rocauc': 0.5856297203566897} steps=7.8671328671328675
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_mf_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molsider view=appnp compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molsider --view appnp --compute rrog-act --T 1 --n_sup 3 --hidden 128 --bs 128 --epochs 100 --eval_every 10 --agg_layers 5 --compute_layers 2 --seed 0 --lam_q 0.1 --halt_max_steps 8 --halt_min_steps 2 --halt_target loss --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:0 --num_workers 0
ep10 val_rocauc=0.52120 val_adapt_rocauc=0.52120 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.54017 val_adapt_rocauc=0.53843 adapt_steps=7.87 halt=0.12 train_steps=4.49
ep30 val_rocauc=0.53233 val_adapt_rocauc=0.53233 adapt_steps=8.00 halt=0.12 train_steps=4.48
ep40 val_rocauc=0.53931 val_adapt_rocauc=0.53735 adapt_steps=7.72 halt=0.13 train_steps=4.46
ep50 val_rocauc=0.56152 val_adapt_rocauc=0.56188 adapt_steps=7.75 halt=0.13 train_steps=4.47
ep60 val_rocauc=0.57531 val_adapt_rocauc=0.57388 adapt_steps=7.59 halt=0.13 train_steps=4.41
ep70 val_rocauc=0.55453 val_adapt_rocauc=0.56066 adapt_steps=7.09 halt=0.13 train_steps=4.44
ep80 val_rocauc=0.58010 val_adapt_rocauc=0.57429 adapt_steps=7.18 halt=0.13 train_steps=4.45
ep90 val_rocauc=0.56976 val_adapt_rocauc=0.56750 adapt_steps=7.34 halt=0.13 train_steps=4.44
ep100 val_rocauc=0.57081 val_adapt_rocauc=0.56938 adapt_steps=7.11 halt=0.13 train_steps=4.39
[ogbg-molsider_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.58009596629467} test={'rocauc': 0.6009364115000575} adaptive={'rocauc': 0.5959003553252203} steps=7.27972027972028
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molsider_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
|
