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-molsider view=gin compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:3
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 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/sider.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, 3.84it/s]
Downloaded 0.00 GB: 50%|█████ | 1/2 [00:00<00:00, 3.84it/s]
Downloaded 0.00 GB: 100%|██████████| 2/2 [00:00<00:00, 7.68it/s]
Processing...
Extracting /home/yurenh2/rrog-gnn-runner/data/ogb/sider.zip
Loading necessary files...
This might take a while.
Processing graphs...
0%| | 0/1427 [00:00<?, ?it/s]
100%|██████████| 1427/1427 [00:00<00:00, 157159.75it/s]
Converting graphs into PyG objects...
0%| | 0/1427 [00:00<?, ?it/s]
100%|██████████| 1427/1427 [00:00<00:00, 79093.89it/s]
Done!
Saving...
ep10 val_rocauc=0.51600 val_adapt_rocauc=0.51600 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.54994 val_adapt_rocauc=0.54994 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.53576 val_adapt_rocauc=0.53552 adapt_steps=7.83 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.55033 val_adapt_rocauc=0.55033 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.55437 val_adapt_rocauc=0.55437 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.60254 val_adapt_rocauc=0.60254 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.58756 val_adapt_rocauc=0.58756 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.60754 val_adapt_rocauc=0.60754 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.59698 val_adapt_rocauc=0.59698 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.59673 val_adapt_rocauc=0.59673 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_gin_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.6075447846476244} test={'rocauc': 0.5810477512085702} adaptive={'rocauc': 0.5810477512085702} steps=8.0
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_gin_rrog-act_T1_ns3_stream_hm8_hmin2_exact_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:3
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 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=0.54584 val_adapt_rocauc=0.54577 adapt_steps=7.98 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.58611 val_adapt_rocauc=0.58611 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.59127 val_adapt_rocauc=0.59127 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.57511 val_adapt_rocauc=0.57438 adapt_steps=7.84 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.58186 val_adapt_rocauc=0.58186 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.56772 val_adapt_rocauc=0.56772 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.59297 val_adapt_rocauc=0.59297 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep80 val_rocauc=0.62339 val_adapt_rocauc=0.62339 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep90 val_rocauc=0.62819 val_adapt_rocauc=0.62813 adapt_steps=7.90 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.63343 val_adapt_rocauc=0.63420 adapt_steps=7.97 halt=0.12 train_steps=4.50
[ogbg-molsider_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.6334288171779305} test={'rocauc': 0.6299766632454395} adaptive={'rocauc': 0.6305843217920319} steps=7.909090909090909
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_exact_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:3
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 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=0.50898 val_adapt_rocauc=0.50898 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.53439 val_adapt_rocauc=0.53439 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.55399 val_adapt_rocauc=0.55399 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.51743 val_adapt_rocauc=0.51743 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.56124 val_adapt_rocauc=0.56124 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.58333 val_adapt_rocauc=0.58333 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.58471 val_adapt_rocauc=0.58471 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep80 val_rocauc=0.59340 val_adapt_rocauc=0.59340 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.59796 val_adapt_rocauc=0.59796 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep100 val_rocauc=0.59492 val_adapt_rocauc=0.59492 adapt_steps=8.00 halt=0.12 train_steps=4.49
[ogbg-molsider_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.5979570667859507} test={'rocauc': 0.6336766486024191} adaptive={'rocauc': 0.6336766486024191} steps=8.0
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_exact_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:3
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 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=0.52583 val_adapt_rocauc=0.52583 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.57520 val_adapt_rocauc=0.57520 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.52761 val_adapt_rocauc=0.52761 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.53501 val_adapt_rocauc=0.53501 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.54257 val_adapt_rocauc=0.54257 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.53427 val_adapt_rocauc=0.53427 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.57027 val_adapt_rocauc=0.57027 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.57984 val_adapt_rocauc=0.57984 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.59046 val_adapt_rocauc=0.59046 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.58588 val_adapt_rocauc=0.58588 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_tag_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rocauc': 0.5904624057788478} test={'rocauc': 0.621622401806624} adaptive={'rocauc': 0.621622401806624} steps=8.0
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_tag_rrog-act_T1_ns3_stream_hm8_hmin2_exact_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:3
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 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=0.52148 val_adapt_rocauc=0.52148 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.53383 val_adapt_rocauc=0.53359 adapt_steps=7.96 halt=0.12 train_steps=4.49
ep30 val_rocauc=0.54880 val_adapt_rocauc=0.54880 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.59286 val_adapt_rocauc=0.59286 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.59415 val_adapt_rocauc=0.59415 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.60132 val_adapt_rocauc=0.60132 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.58871 val_adapt_rocauc=0.58871 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.60004 val_adapt_rocauc=0.60006 adapt_steps=7.99 halt=0.12 train_steps=4.49
ep90 val_rocauc=0.61690 val_adapt_rocauc=0.61690 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.62234 val_adapt_rocauc=0.62234 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.6223441276640819} test={'rocauc': 0.622244619371397} adaptive={'rocauc': 0.622244619371397} steps=8.0
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_exact_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:3
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 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=0.50469 val_adapt_rocauc=0.50469 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.55043 val_adapt_rocauc=0.55043 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep30 val_rocauc=0.56343 val_adapt_rocauc=0.56343 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep40 val_rocauc=0.56853 val_adapt_rocauc=0.56838 adapt_steps=7.95 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.56311 val_adapt_rocauc=0.56311 adapt_steps=8.00 halt=0.12 train_steps=4.48
ep60 val_rocauc=0.57671 val_adapt_rocauc=0.57673 adapt_steps=7.99 halt=0.12 train_steps=4.49
ep70 val_rocauc=0.58004 val_adapt_rocauc=0.57740 adapt_steps=7.78 halt=0.12 train_steps=4.49
ep80 val_rocauc=0.59036 val_adapt_rocauc=0.58962 adapt_steps=7.92 halt=0.12 train_steps=4.48
ep90 val_rocauc=0.59131 val_adapt_rocauc=0.59135 adapt_steps=7.94 halt=0.12 train_steps=4.48
ep100 val_rocauc=0.59335 val_adapt_rocauc=0.59096 adapt_steps=7.85 halt=0.13 train_steps=4.48
[ogbg-molsider_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rocauc': 0.593345692093345} test={'rocauc': 0.5966615384995925} adaptive={'rocauc': 0.5964163493356076} steps=7.888111888111888
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_exact_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:3
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 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=0.51332 val_adapt_rocauc=0.51332 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.49872 val_adapt_rocauc=0.49872 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.52439 val_adapt_rocauc=0.52439 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.51652 val_adapt_rocauc=0.51652 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.54453 val_adapt_rocauc=0.54453 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.56613 val_adapt_rocauc=0.56613 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.58698 val_adapt_rocauc=0.58698 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.56911 val_adapt_rocauc=0.56911 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.57175 val_adapt_rocauc=0.57175 adapt_steps=8.00 halt=0.12 train_steps=4.50
/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=0.58106 val_adapt_rocauc=0.58106 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_pna_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rocauc': 0.5869831648034167} test={'rocauc': 0.5961753244813196} adaptive={'rocauc': 0.5962142801036087} steps=7.958041958041958
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_pna_rrog-act_T1_ns3_stream_hm8_hmin2_exact_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:3
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 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=0.51655 val_adapt_rocauc=0.51655 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep20 val_rocauc=0.53764 val_adapt_rocauc=0.53764 adapt_steps=8.00 halt=0.12 train_steps=4.49
ep30 val_rocauc=0.58175 val_adapt_rocauc=0.58175 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep40 val_rocauc=0.57301 val_adapt_rocauc=0.57301 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep50 val_rocauc=0.56615 val_adapt_rocauc=0.56615 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep60 val_rocauc=0.58257 val_adapt_rocauc=0.58257 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep70 val_rocauc=0.59759 val_adapt_rocauc=0.59759 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep80 val_rocauc=0.62601 val_adapt_rocauc=0.62601 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep90 val_rocauc=0.61042 val_adapt_rocauc=0.61042 adapt_steps=8.00 halt=0.12 train_steps=4.50
ep100 val_rocauc=0.61180 val_adapt_rocauc=0.61180 adapt_steps=8.00 halt=0.12 train_steps=4.50
[ogbg-molsider_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rocauc': 0.6260140078552355} test={'rocauc': 0.6259436225287444} adaptive={'rocauc': 0.6259436225287444} steps=8.0
wrote /home/yurenh2/rrog-gnn-runner/runs/ogbg-molsider_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_exact_lq0.1_hex0.1_qw0_h128_e100_s0.json
|
