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
|
[run] ogbg-molclintox view=gin compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
Downloading http://snap.stanford.edu/ogb/data/graphproppred/csv_mol_download/clintox.zip
0%| | 0/2 [00:00<?, ?it/s]
Downloaded 0.00 GB: 0%| | 0/2 [00:00<?, ?it/s]
Downloaded 0.00 GB: 50%|█████ | 1/2 [00:00<00:00, 4.81it/s]
Downloaded 0.00 GB: 50%|█████ | 1/2 [00:00<00:00, 4.81it/s]
Downloaded 0.00 GB: 100%|██████████| 2/2 [00:00<00:00, 9.61it/s]
Processing...
Extracting /home/yurenh2/rrog-gnn-runner/data/ogb/clintox.zip
Loading necessary files...
This might take a while.
Processing graphs...
0%| | 0/1477 [00:00<?, ?it/s]
100%|██████████| 1477/1477 [00:00<00:00, 169141.78it/s]
Converting graphs into PyG objects...
0%| | 0/1477 [00:00<?, ?it/s]
100%|██████████| 1477/1477 [00:00<00:00, 83199.98it/s]
Done!
Saving...
ep10 val_rocauc=0.60624 val_adapt_rocauc=0.74118 adapt_steps=2.01 halt=0.40 train_steps=2.12
ep20 val_rocauc=0.78931 val_adapt_rocauc=0.78193 adapt_steps=2.51 halt=0.46 train_steps=1.67
ep30 val_rocauc=0.87347 val_adapt_rocauc=0.88131 adapt_steps=2.23 halt=0.43 train_steps=1.76
ep40 val_rocauc=0.90655 val_adapt_rocauc=0.89215 adapt_steps=2.11 halt=0.43 train_steps=1.90
ep50 val_rocauc=0.90059 val_adapt_rocauc=0.89949 adapt_steps=2.04 halt=0.45 train_steps=1.68
ep60 val_rocauc=0.92569 val_adapt_rocauc=0.91470 adapt_steps=2.04 halt=0.46 train_steps=1.73
ep70 val_rocauc=0.85343 val_adapt_rocauc=0.84051 adapt_steps=2.00 halt=0.47 train_steps=1.58
ep80 val_rocauc=0.90430 val_adapt_rocauc=0.88959 adapt_steps=2.00 halt=0.47 train_steps=1.61
ep90 val_rocauc=0.91698 val_adapt_rocauc=0.90556 adapt_steps=2.01 halt=0.47 train_steps=1.59
ep100 val_rocauc=0.89472 val_adapt_rocauc=0.88990 adapt_steps=2.01 halt=0.46 train_steps=1.67
[ogbg-molclintox_gin_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=60 val={'rocauc': 0.925691092944614} test={'rocauc': 0.880287943558197} adaptive={'rocauc': 0.8739417161922636} steps=2.0675675675675675
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_gin_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molclintox view=gcn compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
ep10 val_rocauc=0.82183 val_adapt_rocauc=0.86297 adapt_steps=2.22 halt=0.43 train_steps=1.73
ep20 val_rocauc=0.76252 val_adapt_rocauc=0.86505 adapt_steps=2.26 halt=0.47 train_steps=1.63
ep30 val_rocauc=0.83688 val_adapt_rocauc=0.89876 adapt_steps=2.09 halt=0.45 train_steps=1.62
ep40 val_rocauc=0.81513 val_adapt_rocauc=0.86423 adapt_steps=2.01 halt=0.46 train_steps=1.65
ep50 val_rocauc=0.93016 val_adapt_rocauc=0.95113 adapt_steps=2.00 halt=0.47 train_steps=1.62
ep60 val_rocauc=0.96179 val_adapt_rocauc=0.96556 adapt_steps=2.00 halt=0.47 train_steps=1.64
ep70 val_rocauc=0.94241 val_adapt_rocauc=0.94934 adapt_steps=2.00 halt=0.47 train_steps=1.64
ep80 val_rocauc=0.97166 val_adapt_rocauc=0.96956 adapt_steps=2.00 halt=0.46 train_steps=1.63
ep90 val_rocauc=0.97588 val_adapt_rocauc=0.97344 adapt_steps=2.00 halt=0.46 train_steps=1.62
ep100 val_rocauc=0.97353 val_adapt_rocauc=0.96875 adapt_steps=2.00 halt=0.45 train_steps=1.63
[ogbg-molclintox_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.9758757674250633} test={'rocauc': 0.8739825530879644} adaptive={'rocauc': 0.8682749452611824} steps=2.054054054054054
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molclintox view=sgc compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
ep10 val_rocauc=0.68124 val_adapt_rocauc=0.73447 adapt_steps=2.00 halt=0.43 train_steps=1.98
ep20 val_rocauc=0.74983 val_adapt_rocauc=0.82573 adapt_steps=2.00 halt=0.45 train_steps=1.75
ep30 val_rocauc=0.83839 val_adapt_rocauc=0.86462 adapt_steps=2.07 halt=0.45 train_steps=1.69
ep40 val_rocauc=0.85645 val_adapt_rocauc=0.90685 adapt_steps=2.11 halt=0.47 train_steps=1.61
ep50 val_rocauc=0.82385 val_adapt_rocauc=0.82321 adapt_steps=2.01 halt=0.47 train_steps=1.56
ep60 val_rocauc=0.87257 val_adapt_rocauc=0.89450 adapt_steps=2.02 halt=0.47 train_steps=1.61
ep70 val_rocauc=0.93091 val_adapt_rocauc=0.93933 adapt_steps=2.01 halt=0.47 train_steps=1.61
ep80 val_rocauc=0.91830 val_adapt_rocauc=0.92507 adapt_steps=2.03 halt=0.47 train_steps=1.67
ep90 val_rocauc=0.91036 val_adapt_rocauc=0.91478 adapt_steps=2.01 halt=0.47 train_steps=1.67
ep100 val_rocauc=0.91573 val_adapt_rocauc=0.91911 adapt_steps=2.01 halt=0.47 train_steps=1.63
[ogbg-molclintox_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rocauc': 0.9309095833743721} test={'rocauc': 0.8809821707851111} adaptive={'rocauc': 0.9033008375907969} steps=2.0878378378378377
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molclintox view=tag compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
ep10 val_rocauc=0.64072 val_adapt_rocauc=0.74493 adapt_steps=2.03 halt=0.41 train_steps=2.16
ep20 val_rocauc=0.69510 val_adapt_rocauc=0.88005 adapt_steps=2.13 halt=0.47 train_steps=1.63
ep30 val_rocauc=0.86787 val_adapt_rocauc=0.86841 adapt_steps=2.00 halt=0.46 train_steps=1.64
ep40 val_rocauc=0.79619 val_adapt_rocauc=0.83481 adapt_steps=2.00 halt=0.46 train_steps=1.68
ep50 val_rocauc=0.81643 val_adapt_rocauc=0.85514 adapt_steps=2.04 halt=0.44 train_steps=1.79
ep60 val_rocauc=0.85361 val_adapt_rocauc=0.91327 adapt_steps=2.00 halt=0.48 train_steps=1.59
ep70 val_rocauc=0.85772 val_adapt_rocauc=0.91744 adapt_steps=2.00 halt=0.47 train_steps=1.59
ep80 val_rocauc=0.84607 val_adapt_rocauc=0.91703 adapt_steps=2.00 halt=0.46 train_steps=1.63
ep90 val_rocauc=0.86604 val_adapt_rocauc=0.92838 adapt_steps=2.00 halt=0.47 train_steps=1.61
ep100 val_rocauc=0.85561 val_adapt_rocauc=0.91784 adapt_steps=2.00 halt=0.46 train_steps=1.62
[ogbg-molclintox_tag_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=30 val={'rocauc': 0.8678740273810697} test={'rocauc': 0.8432636499496056} adaptive={'rocauc': 0.8344932749452612} steps=2.0
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_tag_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molclintox view=graphconv compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
ep10 val_rocauc=0.71091 val_adapt_rocauc=0.73332 adapt_steps=2.07 halt=0.42 train_steps=1.94
ep20 val_rocauc=0.77725 val_adapt_rocauc=0.79820 adapt_steps=2.14 halt=0.46 train_steps=1.73
ep30 val_rocauc=0.85712 val_adapt_rocauc=0.83995 adapt_steps=2.12 halt=0.44 train_steps=1.69
ep40 val_rocauc=0.86597 val_adapt_rocauc=0.88025 adapt_steps=2.14 halt=0.45 train_steps=1.71
ep50 val_rocauc=0.83623 val_adapt_rocauc=0.85307 adapt_steps=2.01 halt=0.46 train_steps=1.61
ep60 val_rocauc=0.84827 val_adapt_rocauc=0.85414 adapt_steps=2.03 halt=0.47 train_steps=1.66
ep70 val_rocauc=0.86190 val_adapt_rocauc=0.88018 adapt_steps=2.00 halt=0.47 train_steps=1.62
ep80 val_rocauc=0.81287 val_adapt_rocauc=0.85623 adapt_steps=2.04 halt=0.45 train_steps=1.78
ep90 val_rocauc=0.81184 val_adapt_rocauc=0.85519 adapt_steps=2.03 halt=0.46 train_steps=1.73
ep100 val_rocauc=0.80130 val_adapt_rocauc=0.84348 adapt_steps=2.05 halt=0.45 train_steps=1.75
[ogbg-molclintox_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=40 val={'rocauc': 0.8659657244164286} test={'rocauc': 0.805447815660515} adaptive={'rocauc': 0.8240981128140966} steps=2.2094594594594597
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molclintox view=appnp compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
ep10 val_rocauc=0.76585 val_adapt_rocauc=0.86739 adapt_steps=2.02 halt=0.45 train_steps=1.70
ep20 val_rocauc=0.86039 val_adapt_rocauc=0.86324 adapt_steps=2.00 halt=0.47 train_steps=1.62
ep30 val_rocauc=0.90459 val_adapt_rocauc=0.89663 adapt_steps=2.02 halt=0.47 train_steps=1.59
ep40 val_rocauc=0.90971 val_adapt_rocauc=0.88686 adapt_steps=2.00 halt=0.45 train_steps=1.69
ep50 val_rocauc=0.93832 val_adapt_rocauc=0.92655 adapt_steps=2.00 halt=0.46 train_steps=1.61
ep60 val_rocauc=0.89696 val_adapt_rocauc=0.91244 adapt_steps=2.03 halt=0.47 train_steps=1.65
ep70 val_rocauc=0.92770 val_adapt_rocauc=0.92621 adapt_steps=2.01 halt=0.47 train_steps=1.62
ep80 val_rocauc=0.93654 val_adapt_rocauc=0.94478 adapt_steps=2.01 halt=0.46 train_steps=1.69
ep90 val_rocauc=0.93512 val_adapt_rocauc=0.94803 adapt_steps=2.01 halt=0.46 train_steps=1.63
ep100 val_rocauc=0.93911 val_adapt_rocauc=0.94431 adapt_steps=2.00 halt=0.46 train_steps=1.64
[ogbg-molclintox_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.9391148757345941} test={'rocauc': 0.8412244117749279} adaptive={'rocauc': 0.9026361519480067} steps=2.0405405405405403
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molclintox view=pna compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.73449 val_adapt_rocauc=0.82290 adapt_steps=2.02 halt=0.44 train_steps=1.88
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.72647 val_adapt_rocauc=0.81333 adapt_steps=2.02 halt=0.41 train_steps=1.96
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.69121 val_adapt_rocauc=0.81663 adapt_steps=2.03 halt=0.45 train_steps=1.67
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.68050 val_adapt_rocauc=0.79323 adapt_steps=2.05 halt=0.39 train_steps=2.17
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.78235 val_adapt_rocauc=0.82074 adapt_steps=2.27 halt=0.42 train_steps=1.74
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.77786 val_adapt_rocauc=0.80819 adapt_steps=2.03 halt=0.44 train_steps=1.77
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.80327 val_adapt_rocauc=0.80844 adapt_steps=2.07 halt=0.45 train_steps=1.65
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.82184 val_adapt_rocauc=0.81128 adapt_steps=2.07 halt=0.46 train_steps=1.66
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.81683 val_adapt_rocauc=0.80189 adapt_steps=2.05 halt=0.47 train_steps=1.62
/home/yurenh2/miniconda3/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(
/home/yurenh2/miniconda3/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=0.82947 val_adapt_rocauc=0.80329 adapt_steps=2.05 halt=0.46 train_steps=1.66
[ogbg-molclintox_pna_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.8294650185495256} test={'rocauc': 0.9034242171480207} adaptive={'rocauc': 0.9194748548986897} steps=2.22972972972973
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_pna_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-molclintox view=resgated compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
python3 rrog/train_ogb_graphprop.py --dataset ogbg-molclintox --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 exact --halt_loss_threshold 0.2 --halt_exploration_prob 0.1 --act_train_mode stream --q_warmup_epochs 0 --device cuda:3 --num_workers 0
ep10 val_rocauc=0.86017 val_adapt_rocauc=0.85805 adapt_steps=2.00 halt=0.44 train_steps=1.96
ep20 val_rocauc=0.81251 val_adapt_rocauc=0.81417 adapt_steps=2.05 halt=0.46 train_steps=1.74
ep30 val_rocauc=0.83266 val_adapt_rocauc=0.88528 adapt_steps=2.09 halt=0.44 train_steps=1.83
ep40 val_rocauc=0.89601 val_adapt_rocauc=0.90433 adapt_steps=2.14 halt=0.47 train_steps=1.62
ep50 val_rocauc=0.84727 val_adapt_rocauc=0.87116 adapt_steps=2.09 halt=0.46 train_steps=1.60
ep60 val_rocauc=0.85582 val_adapt_rocauc=0.90812 adapt_steps=2.07 halt=0.47 train_steps=1.67
ep70 val_rocauc=0.91150 val_adapt_rocauc=0.90620 adapt_steps=2.02 halt=0.46 train_steps=1.65
ep80 val_rocauc=0.91162 val_adapt_rocauc=0.89879 adapt_steps=2.03 halt=0.47 train_steps=1.61
ep90 val_rocauc=0.89291 val_adapt_rocauc=0.89228 adapt_steps=2.01 halt=0.47 train_steps=1.63
ep100 val_rocauc=0.89000 val_adapt_rocauc=0.88702 adapt_steps=2.02 halt=0.45 train_steps=1.68
[ogbg-molclintox_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.9116172559834532} test={'rocauc': 0.8402903763945366} adaptive={'rocauc': 0.8415754353039308} steps=2.0337837837837838
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molclintox_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
|
