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-molhiv view=gin compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.50465 val_adapt_rocauc=0.51605 adapt_steps=2.17 halt=0.44 train_steps=1.86
ep20 val_rocauc=0.68965 val_adapt_rocauc=0.70338 adapt_steps=2.05 halt=0.43 train_steps=1.93
ep30 val_rocauc=0.74705 val_adapt_rocauc=0.71674 adapt_steps=2.16 halt=0.44 train_steps=1.88
ep40 val_rocauc=0.66882 val_adapt_rocauc=0.69730 adapt_steps=2.23 halt=0.44 train_steps=1.90
ep50 val_rocauc=0.69645 val_adapt_rocauc=0.75311 adapt_steps=2.17 halt=0.44 train_steps=1.88
ep60 val_rocauc=0.72560 val_adapt_rocauc=0.75246 adapt_steps=2.12 halt=0.44 train_steps=1.87
ep70 val_rocauc=0.73678 val_adapt_rocauc=0.76610 adapt_steps=2.13 halt=0.44 train_steps=1.86
ep80 val_rocauc=0.73638 val_adapt_rocauc=0.77730 adapt_steps=2.07 halt=0.44 train_steps=1.85
ep90 val_rocauc=0.76228 val_adapt_rocauc=0.78163 adapt_steps=2.11 halt=0.44 train_steps=1.85
ep100 val_rocauc=0.75586 val_adapt_rocauc=0.78034 adapt_steps=2.10 halt=0.44 train_steps=1.85
[ogbg-molhiv_gin_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.7622844405251813} test={'rocauc': 0.728463276617934} adaptive={'rocauc': 0.7414685490256668} steps=2.1835643082907854
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_gin_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=gine compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.69659 val_adapt_rocauc=0.67019 adapt_steps=2.27 halt=0.43 train_steps=1.91
ep20 val_rocauc=0.71128 val_adapt_rocauc=0.74182 adapt_steps=2.25 halt=0.44 train_steps=1.90
ep30 val_rocauc=0.71784 val_adapt_rocauc=0.71432 adapt_steps=2.09 halt=0.43 train_steps=1.92
ep40 val_rocauc=0.57809 val_adapt_rocauc=0.61863 adapt_steps=2.07 halt=0.43 train_steps=1.95
ep50 val_rocauc=0.71898 val_adapt_rocauc=0.72970 adapt_steps=2.15 halt=0.44 train_steps=1.90
ep60 val_rocauc=0.74272 val_adapt_rocauc=0.75436 adapt_steps=2.17 halt=0.44 train_steps=1.91
ep70 val_rocauc=0.64453 val_adapt_rocauc=0.67305 adapt_steps=2.09 halt=0.43 train_steps=1.92
ep80 val_rocauc=0.66148 val_adapt_rocauc=0.71639 adapt_steps=2.11 halt=0.44 train_steps=1.89
ep90 val_rocauc=0.68795 val_adapt_rocauc=0.73859 adapt_steps=2.10 halt=0.44 train_steps=1.89
ep100 val_rocauc=0.69479 val_adapt_rocauc=0.73030 adapt_steps=2.10 halt=0.44 train_steps=1.88
[ogbg-molhiv_gine_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=60 val={'rocauc': 0.7427218058005095} test={'rocauc': 0.7216728789663763} adaptive={'rocauc': 0.7219567778442999} steps=2.1913445173839046
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_gine_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=gcn compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.65285 val_adapt_rocauc=0.67983 adapt_steps=2.17 halt=0.43 train_steps=1.95
ep20 val_rocauc=0.63964 val_adapt_rocauc=0.67770 adapt_steps=2.12 halt=0.43 train_steps=1.91
ep30 val_rocauc=0.70439 val_adapt_rocauc=0.72607 adapt_steps=2.28 halt=0.43 train_steps=1.93
ep40 val_rocauc=0.67857 val_adapt_rocauc=0.67646 adapt_steps=2.09 halt=0.44 train_steps=1.90
ep50 val_rocauc=0.66299 val_adapt_rocauc=0.70493 adapt_steps=2.10 halt=0.44 train_steps=1.89
ep60 val_rocauc=0.72309 val_adapt_rocauc=0.75250 adapt_steps=2.10 halt=0.43 train_steps=1.91
ep70 val_rocauc=0.69147 val_adapt_rocauc=0.71552 adapt_steps=2.06 halt=0.44 train_steps=1.88
ep80 val_rocauc=0.75410 val_adapt_rocauc=0.76395 adapt_steps=2.12 halt=0.44 train_steps=1.88
ep90 val_rocauc=0.76238 val_adapt_rocauc=0.75313 adapt_steps=2.09 halt=0.44 train_steps=1.85
ep100 val_rocauc=0.76521 val_adapt_rocauc=0.75985 adapt_steps=2.09 halt=0.44 train_steps=1.87
[ogbg-molhiv_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.76521164021164} test={'rocauc': 0.7279630738330211} adaptive={'rocauc': 0.7195040460418316} steps=2.1412594213469487
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=graphsage compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.60161 val_adapt_rocauc=0.64669 adapt_steps=2.08 halt=0.43 train_steps=1.92
ep20 val_rocauc=0.71865 val_adapt_rocauc=0.72325 adapt_steps=2.51 halt=0.43 train_steps=1.95
ep30 val_rocauc=0.72758 val_adapt_rocauc=0.72496 adapt_steps=2.13 halt=0.43 train_steps=1.95
ep40 val_rocauc=0.75529 val_adapt_rocauc=0.73154 adapt_steps=2.08 halt=0.43 train_steps=1.93
ep50 val_rocauc=0.74199 val_adapt_rocauc=0.74141 adapt_steps=2.12 halt=0.43 train_steps=1.96
ep60 val_rocauc=0.70099 val_adapt_rocauc=0.73010 adapt_steps=2.10 halt=0.43 train_steps=1.95
ep70 val_rocauc=0.76185 val_adapt_rocauc=0.77202 adapt_steps=2.10 halt=0.44 train_steps=1.91
ep80 val_rocauc=0.77242 val_adapt_rocauc=0.76566 adapt_steps=2.14 halt=0.44 train_steps=1.91
ep90 val_rocauc=0.75081 val_adapt_rocauc=0.75828 adapt_steps=2.14 halt=0.44 train_steps=1.90
ep100 val_rocauc=0.74508 val_adapt_rocauc=0.74839 adapt_steps=2.13 halt=0.44 train_steps=1.90
[ogbg-molhiv_graphsage_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.7724224720752498} test={'rocauc': 0.7540721141775624} adaptive={'rocauc': 0.7493028061569362} steps=2.225382932166302
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_graphsage_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=gatv2 compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.63791 val_adapt_rocauc=0.65941 adapt_steps=2.16 halt=0.44 train_steps=1.86
ep20 val_rocauc=0.65958 val_adapt_rocauc=0.68114 adapt_steps=2.14 halt=0.43 train_steps=1.98
ep30 val_rocauc=0.70813 val_adapt_rocauc=0.71648 adapt_steps=2.10 halt=0.44 train_steps=1.91
ep40 val_rocauc=0.71160 val_adapt_rocauc=0.71489 adapt_steps=2.12 halt=0.43 train_steps=1.92
ep50 val_rocauc=0.74629 val_adapt_rocauc=0.72922 adapt_steps=2.12 halt=0.43 train_steps=1.93
ep60 val_rocauc=0.77221 val_adapt_rocauc=0.77362 adapt_steps=2.10 halt=0.44 train_steps=1.91
ep70 val_rocauc=0.73550 val_adapt_rocauc=0.75380 adapt_steps=2.22 halt=0.43 train_steps=1.96
ep80 val_rocauc=0.75176 val_adapt_rocauc=0.76963 adapt_steps=2.17 halt=0.43 train_steps=1.95
ep90 val_rocauc=0.77064 val_adapt_rocauc=0.77482 adapt_steps=2.17 halt=0.44 train_steps=1.89
ep100 val_rocauc=0.77092 val_adapt_rocauc=0.77462 adapt_steps=2.15 halt=0.44 train_steps=1.89
[ogbg-molhiv_gatv2_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=60 val={'rocauc': 0.7722111992945326} test={'rocauc': 0.7005330346279379} adaptive={'rocauc': 0.7209988605419185} steps=2.161196207148067
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_gatv2_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=graphconv compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.62286 val_adapt_rocauc=0.63631 adapt_steps=2.12 halt=0.43 train_steps=1.92
ep20 val_rocauc=0.50681 val_adapt_rocauc=0.62452 adapt_steps=2.03 halt=0.44 train_steps=1.88
ep30 val_rocauc=0.71981 val_adapt_rocauc=0.71587 adapt_steps=2.12 halt=0.44 train_steps=1.92
ep40 val_rocauc=0.63676 val_adapt_rocauc=0.69827 adapt_steps=2.10 halt=0.44 train_steps=1.89
ep50 val_rocauc=0.71987 val_adapt_rocauc=0.71579 adapt_steps=2.10 halt=0.43 train_steps=1.95
ep60 val_rocauc=0.77652 val_adapt_rocauc=0.77421 adapt_steps=2.13 halt=0.43 train_steps=1.94
ep70 val_rocauc=0.73578 val_adapt_rocauc=0.73905 adapt_steps=2.14 halt=0.44 train_steps=1.90
ep80 val_rocauc=0.77909 val_adapt_rocauc=0.77380 adapt_steps=2.21 halt=0.43 train_steps=1.93
ep90 val_rocauc=0.77284 val_adapt_rocauc=0.76992 adapt_steps=2.20 halt=0.43 train_steps=1.95
ep100 val_rocauc=0.75856 val_adapt_rocauc=0.75794 adapt_steps=2.19 halt=0.43 train_steps=1.95
[ogbg-molhiv_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.7790944052518126} test={'rocauc': 0.7510651036134339} adaptive={'rocauc': 0.7493443287819387} steps=2.2287867736445417
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=transformer compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.69482 val_adapt_rocauc=0.69337 adapt_steps=2.09 halt=0.43 train_steps=1.95
ep20 val_rocauc=0.69048 val_adapt_rocauc=0.68828 adapt_steps=2.11 halt=0.43 train_steps=1.92
ep30 val_rocauc=0.70734 val_adapt_rocauc=0.68263 adapt_steps=2.08 halt=0.43 train_steps=1.96
ep40 val_rocauc=0.75896 val_adapt_rocauc=0.75541 adapt_steps=2.17 halt=0.43 train_steps=1.96
ep50 val_rocauc=0.63094 val_adapt_rocauc=0.71781 adapt_steps=2.15 halt=0.43 train_steps=2.00
ep60 val_rocauc=0.78134 val_adapt_rocauc=0.78253 adapt_steps=2.15 halt=0.43 train_steps=1.95
ep70 val_rocauc=0.74767 val_adapt_rocauc=0.77658 adapt_steps=2.14 halt=0.43 train_steps=1.93
ep80 val_rocauc=0.75221 val_adapt_rocauc=0.76723 adapt_steps=2.19 halt=0.43 train_steps=1.98
ep90 val_rocauc=0.73516 val_adapt_rocauc=0.74942 adapt_steps=2.14 halt=0.43 train_steps=1.92
ep100 val_rocauc=0.74673 val_adapt_rocauc=0.75054 adapt_steps=2.13 halt=0.44 train_steps=1.91
[ogbg-molhiv_transformer_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=60 val={'rocauc': 0.7813418577307466} test={'rocauc': 0.7294279534174086} adaptive={'rocauc': 0.7396338283860252} steps=2.215171407731583
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_transformer_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=pna compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.65392 val_adapt_rocauc=0.65972 adapt_steps=2.26 halt=0.45 train_steps=1.76
/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=0.70167 val_adapt_rocauc=0.73359 adapt_steps=2.14 halt=0.45 train_steps=1.83
/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=0.57289 val_adapt_rocauc=0.60013 adapt_steps=2.02 halt=0.45 train_steps=1.83
/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=0.74876 val_adapt_rocauc=0.72038 adapt_steps=2.03 halt=0.44 train_steps=1.89
/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=0.69108 val_adapt_rocauc=0.72229 adapt_steps=2.11 halt=0.43 train_steps=1.92
/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=0.69969 val_adapt_rocauc=0.70883 adapt_steps=2.12 halt=0.44 train_steps=1.89
/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=0.72681 val_adapt_rocauc=0.77015 adapt_steps=2.16 halt=0.43 train_steps=1.93
/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=0.74562 val_adapt_rocauc=0.77410 adapt_steps=2.13 halt=0.43 train_steps=1.92
/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=0.71370 val_adapt_rocauc=0.78801 adapt_steps=2.16 halt=0.44 train_steps=1.90
/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=0.71775 val_adapt_rocauc=0.78617 adapt_steps=2.14 halt=0.44 train_steps=1.89
[ogbg-molhiv_pna_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=40 val={'rocauc': 0.7487599206349206} test={'rocauc': 0.6730373317368046} adaptive={'rocauc': 0.6877382722725429} steps=2.0588378312667155
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_pna_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=gen compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.58244 val_adapt_rocauc=0.62459 adapt_steps=2.17 halt=0.43 train_steps=1.95
ep20 val_rocauc=0.65886 val_adapt_rocauc=0.66235 adapt_steps=2.06 halt=0.44 train_steps=1.89
ep30 val_rocauc=0.69707 val_adapt_rocauc=0.69637 adapt_steps=2.06 halt=0.44 train_steps=1.92
ep40 val_rocauc=0.73515 val_adapt_rocauc=0.77200 adapt_steps=2.17 halt=0.44 train_steps=1.91
ep50 val_rocauc=0.72530 val_adapt_rocauc=0.75137 adapt_steps=2.12 halt=0.43 train_steps=1.97
ep60 val_rocauc=0.69888 val_adapt_rocauc=0.73379 adapt_steps=2.09 halt=0.43 train_steps=1.98
ep70 val_rocauc=0.71051 val_adapt_rocauc=0.74897 adapt_steps=2.12 halt=0.43 train_steps=1.94
ep80 val_rocauc=0.76901 val_adapt_rocauc=0.77657 adapt_steps=2.10 halt=0.43 train_steps=1.93
ep90 val_rocauc=0.73056 val_adapt_rocauc=0.76530 adapt_steps=2.11 halt=0.44 train_steps=1.92
ep100 val_rocauc=0.73376 val_adapt_rocauc=0.75580 adapt_steps=2.13 halt=0.43 train_steps=1.93
[ogbg-molhiv_gen_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.769011488340192} test={'rocauc': 0.7434075590490352} adaptive={'rocauc': 0.7426736707931786} steps=2.178215414539266
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_gen_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=film compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.60803 val_adapt_rocauc=0.69045 adapt_steps=2.06 halt=0.47 train_steps=1.67
ep20 val_rocauc=0.64308 val_adapt_rocauc=0.78621 adapt_steps=2.01 halt=0.45 train_steps=1.79
ep30 val_rocauc=0.74069 val_adapt_rocauc=0.75513 adapt_steps=2.21 halt=0.45 train_steps=1.83
ep40 val_rocauc=0.75046 val_adapt_rocauc=0.78035 adapt_steps=2.23 halt=0.43 train_steps=1.93
ep50 val_rocauc=0.71486 val_adapt_rocauc=0.75589 adapt_steps=2.34 halt=0.43 train_steps=2.01
ep60 val_rocauc=0.77074 val_adapt_rocauc=0.80191 adapt_steps=2.10 halt=0.43 train_steps=1.99
ep70 val_rocauc=0.78228 val_adapt_rocauc=0.80466 adapt_steps=2.21 halt=0.43 train_steps=1.97
ep80 val_rocauc=0.81468 val_adapt_rocauc=0.81506 adapt_steps=2.13 halt=0.43 train_steps=1.96
ep90 val_rocauc=0.81502 val_adapt_rocauc=0.81015 adapt_steps=2.14 halt=0.43 train_steps=1.95
ep100 val_rocauc=0.81203 val_adapt_rocauc=0.80753 adapt_steps=2.11 halt=0.43 train_steps=1.95
[ogbg-molhiv_film_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.8150199637468156} test={'rocauc': 0.7415303501419496} adaptive={'rocauc': 0.7291546766063463} steps=2.2144420131291027
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_film_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=resgated compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.57563 val_adapt_rocauc=0.57248 adapt_steps=2.08 halt=0.43 train_steps=1.96
ep20 val_rocauc=0.68901 val_adapt_rocauc=0.69659 adapt_steps=2.07 halt=0.43 train_steps=1.97
ep30 val_rocauc=0.72785 val_adapt_rocauc=0.73945 adapt_steps=2.17 halt=0.43 train_steps=1.95
ep40 val_rocauc=0.80307 val_adapt_rocauc=0.79528 adapt_steps=2.10 halt=0.43 train_steps=1.93
ep50 val_rocauc=0.74046 val_adapt_rocauc=0.74882 adapt_steps=2.15 halt=0.43 train_steps=1.96
ep60 val_rocauc=0.76188 val_adapt_rocauc=0.76877 adapt_steps=2.13 halt=0.43 train_steps=1.95
ep70 val_rocauc=0.74584 val_adapt_rocauc=0.74880 adapt_steps=2.14 halt=0.43 train_steps=1.99
ep80 val_rocauc=0.72576 val_adapt_rocauc=0.72937 adapt_steps=2.09 halt=0.43 train_steps=1.95
ep90 val_rocauc=0.72962 val_adapt_rocauc=0.72854 adapt_steps=2.17 halt=0.43 train_steps=1.98
ep100 val_rocauc=0.73262 val_adapt_rocauc=0.73243 adapt_steps=2.13 halt=0.43 train_steps=1.97
[ogbg-molhiv_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=40 val={'rocauc': 0.8030723349010386} test={'rocauc': 0.7502983835145524} adaptive={'rocauc': 0.7566658297765504} steps=2.2205203014831025
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=tag compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.55912 val_adapt_rocauc=0.62342 adapt_steps=2.04 halt=0.43 train_steps=1.90
ep20 val_rocauc=0.68402 val_adapt_rocauc=0.67609 adapt_steps=2.15 halt=0.44 train_steps=1.84
ep30 val_rocauc=0.70234 val_adapt_rocauc=0.71080 adapt_steps=2.18 halt=0.44 train_steps=1.90
ep40 val_rocauc=0.71746 val_adapt_rocauc=0.70275 adapt_steps=2.42 halt=0.44 train_steps=1.86
ep50 val_rocauc=0.74377 val_adapt_rocauc=0.74631 adapt_steps=2.14 halt=0.44 train_steps=1.90
ep60 val_rocauc=0.68302 val_adapt_rocauc=0.74882 adapt_steps=2.18 halt=0.44 train_steps=1.91
ep70 val_rocauc=0.70337 val_adapt_rocauc=0.73207 adapt_steps=2.09 halt=0.44 train_steps=1.90
ep80 val_rocauc=0.75679 val_adapt_rocauc=0.76447 adapt_steps=2.14 halt=0.44 train_steps=1.90
ep90 val_rocauc=0.74719 val_adapt_rocauc=0.76515 adapt_steps=2.12 halt=0.44 train_steps=1.88
ep100 val_rocauc=0.74130 val_adapt_rocauc=0.75597 adapt_steps=2.12 halt=0.44 train_steps=1.89
[ogbg-molhiv_tag_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.7567852243778168} test={'rocauc': 0.7480619556190734} adaptive={'rocauc': 0.7477104617702157} steps=2.225869195234622
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_tag_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=sgc compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.62000 val_adapt_rocauc=0.63880 adapt_steps=2.22 halt=0.44 train_steps=1.89
ep20 val_rocauc=0.59820 val_adapt_rocauc=0.61098 adapt_steps=2.16 halt=0.44 train_steps=1.89
ep30 val_rocauc=0.67764 val_adapt_rocauc=0.68890 adapt_steps=2.08 halt=0.44 train_steps=1.86
ep40 val_rocauc=0.66809 val_adapt_rocauc=0.66128 adapt_steps=2.09 halt=0.44 train_steps=1.89
ep50 val_rocauc=0.73528 val_adapt_rocauc=0.71478 adapt_steps=2.09 halt=0.44 train_steps=1.91
ep60 val_rocauc=0.69053 val_adapt_rocauc=0.68013 adapt_steps=2.16 halt=0.44 train_steps=1.89
ep70 val_rocauc=0.68528 val_adapt_rocauc=0.72841 adapt_steps=2.10 halt=0.44 train_steps=1.89
ep80 val_rocauc=0.68138 val_adapt_rocauc=0.69448 adapt_steps=2.11 halt=0.44 train_steps=1.88
ep90 val_rocauc=0.68754 val_adapt_rocauc=0.69915 adapt_steps=2.12 halt=0.43 train_steps=1.93
ep100 val_rocauc=0.69254 val_adapt_rocauc=0.70377 adapt_steps=2.12 halt=0.44 train_steps=1.88
[ogbg-molhiv_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=50 val={'rocauc': 0.7352782676856752} test={'rocauc': 0.6961702620753587} adaptive={'rocauc': 0.7016995306977732} steps=2.0795040116703136
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=cheb compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.63285 val_adapt_rocauc=0.63797 adapt_steps=2.09 halt=0.43 train_steps=1.94
ep20 val_rocauc=0.59546 val_adapt_rocauc=0.62621 adapt_steps=2.09 halt=0.43 train_steps=1.94
ep30 val_rocauc=0.64157 val_adapt_rocauc=0.64537 adapt_steps=2.07 halt=0.43 train_steps=1.95
ep40 val_rocauc=0.78253 val_adapt_rocauc=0.79697 adapt_steps=2.21 halt=0.43 train_steps=1.94
ep50 val_rocauc=0.78099 val_adapt_rocauc=0.78545 adapt_steps=2.23 halt=0.42 train_steps=2.02
ep60 val_rocauc=0.75496 val_adapt_rocauc=0.75580 adapt_steps=2.16 halt=0.43 train_steps=1.92
ep70 val_rocauc=0.76923 val_adapt_rocauc=0.76459 adapt_steps=2.17 halt=0.43 train_steps=1.94
ep80 val_rocauc=0.75888 val_adapt_rocauc=0.75674 adapt_steps=2.10 halt=0.45 train_steps=1.83
ep90 val_rocauc=0.73937 val_adapt_rocauc=0.73410 adapt_steps=2.10 halt=0.45 train_steps=1.77
ep100 val_rocauc=0.71941 val_adapt_rocauc=0.71715 adapt_steps=2.07 halt=0.46 train_steps=1.74
[ogbg-molhiv_cheb_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=40 val={'rocauc': 0.7825329463060945} test={'rocauc': 0.7205932907163135} adaptive={'rocauc': 0.7218775951640628} steps=2.2939460247994163
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_cheb_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=arma compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.58405 val_adapt_rocauc=0.62652 adapt_steps=2.13 halt=0.44 train_steps=1.87
ep20 val_rocauc=0.63653 val_adapt_rocauc=0.65469 adapt_steps=2.04 halt=0.44 train_steps=1.83
ep30 val_rocauc=0.68206 val_adapt_rocauc=0.66472 adapt_steps=2.09 halt=0.44 train_steps=1.85
ep40 val_rocauc=0.73261 val_adapt_rocauc=0.71919 adapt_steps=2.08 halt=0.44 train_steps=1.82
ep50 val_rocauc=0.76797 val_adapt_rocauc=0.79653 adapt_steps=2.13 halt=0.44 train_steps=1.86
ep60 val_rocauc=0.74831 val_adapt_rocauc=0.72643 adapt_steps=2.04 halt=0.45 train_steps=1.77
ep70 val_rocauc=0.73358 val_adapt_rocauc=0.77502 adapt_steps=2.03 halt=0.46 train_steps=1.73
ep80 val_rocauc=0.77783 val_adapt_rocauc=0.77223 adapt_steps=2.06 halt=0.45 train_steps=1.75
ep90 val_rocauc=0.78442 val_adapt_rocauc=0.76721 adapt_steps=2.05 halt=0.46 train_steps=1.71
ep100 val_rocauc=0.78612 val_adapt_rocauc=0.76507 adapt_steps=2.06 halt=0.46 train_steps=1.73
[ogbg-molhiv_arma_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.7861153978052127} test={'rocauc': 0.7306939106587612} adaptive={'rocauc': 0.7378782904266208} steps=2.085825431558473
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_arma_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=mf compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.65109 val_adapt_rocauc=0.65071 adapt_steps=2.09 halt=0.43 train_steps=1.91
ep20 val_rocauc=0.66377 val_adapt_rocauc=0.70133 adapt_steps=2.15 halt=0.43 train_steps=1.92
ep30 val_rocauc=0.69912 val_adapt_rocauc=0.72873 adapt_steps=2.15 halt=0.43 train_steps=1.91
ep40 val_rocauc=0.75265 val_adapt_rocauc=0.77675 adapt_steps=2.15 halt=0.43 train_steps=1.95
ep50 val_rocauc=0.71979 val_adapt_rocauc=0.76668 adapt_steps=2.10 halt=0.45 train_steps=1.82
ep60 val_rocauc=0.78688 val_adapt_rocauc=0.78537 adapt_steps=2.27 halt=0.44 train_steps=1.89
ep70 val_rocauc=0.80465 val_adapt_rocauc=0.78273 adapt_steps=2.15 halt=0.43 train_steps=1.93
ep80 val_rocauc=0.77804 val_adapt_rocauc=0.78100 adapt_steps=2.15 halt=0.43 train_steps=1.94
ep90 val_rocauc=0.76305 val_adapt_rocauc=0.77873 adapt_steps=2.18 halt=0.44 train_steps=1.91
ep100 val_rocauc=0.76926 val_adapt_rocauc=0.78354 adapt_steps=2.14 halt=0.44 train_steps=1.90
[ogbg-molhiv_mf_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rocauc': 0.8046522878698804} test={'rocauc': 0.7416423646652118} adaptive={'rocauc': 0.7333127329612392} steps=2.1964502796012644
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_mf_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molhiv view=appnp compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:0
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molhiv --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=0.62676 val_adapt_rocauc=0.67960 adapt_steps=2.09 halt=0.44 train_steps=1.87
ep20 val_rocauc=0.65528 val_adapt_rocauc=0.72533 adapt_steps=2.03 halt=0.44 train_steps=1.88
ep30 val_rocauc=0.72398 val_adapt_rocauc=0.73609 adapt_steps=2.22 halt=0.43 train_steps=1.92
ep40 val_rocauc=0.67653 val_adapt_rocauc=0.71372 adapt_steps=2.12 halt=0.44 train_steps=1.88
ep50 val_rocauc=0.72748 val_adapt_rocauc=0.72608 adapt_steps=2.14 halt=0.44 train_steps=1.90
ep60 val_rocauc=0.69598 val_adapt_rocauc=0.70466 adapt_steps=2.12 halt=0.44 train_steps=1.87
ep70 val_rocauc=0.73788 val_adapt_rocauc=0.71230 adapt_steps=2.10 halt=0.44 train_steps=1.88
ep80 val_rocauc=0.72564 val_adapt_rocauc=0.74313 adapt_steps=2.06 halt=0.45 train_steps=1.82
ep90 val_rocauc=0.73463 val_adapt_rocauc=0.75132 adapt_steps=2.07 halt=0.45 train_steps=1.78
ep100 val_rocauc=0.72183 val_adapt_rocauc=0.73975 adapt_steps=2.05 halt=0.45 train_steps=1.76
[ogbg-molhiv_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rocauc': 0.7378778414658045} test={'rocauc': 0.6883099325981575} adaptive={'rocauc': 0.7290619749319223} steps=2.102358375881352
wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-molhiv_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
|
