path: root/logs/ogbg-mollipo_act_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_e100_s0.log

[run] ogbg-mollipo view=gin compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.27165 val_adapt_rmse=1.16086 adapt_steps=5.19 halt=0.18 train_steps=4.00
ep20 val_rmse=1.51709 val_adapt_rmse=1.16539 adapt_steps=5.08 halt=0.19 train_steps=3.67
ep30 val_rmse=3.96000 val_adapt_rmse=2.88970 adapt_steps=4.23 halt=0.21 train_steps=3.41
ep40 val_rmse=1.45958 val_adapt_rmse=1.27486 adapt_steps=6.80 halt=0.17 train_steps=3.91
ep50 val_rmse=1.62306 val_adapt_rmse=1.21914 adapt_steps=3.90 halt=0.21 train_steps=3.43
ep60 val_rmse=0.96105 val_adapt_rmse=0.88005 adapt_steps=3.28 halt=0.29 train_steps=2.74
ep70 val_rmse=0.93501 val_adapt_rmse=0.81324 adapt_steps=2.69 halt=0.29 train_steps=2.70
ep80 val_rmse=0.92098 val_adapt_rmse=0.82825 adapt_steps=2.68 halt=0.33 train_steps=2.45
ep90 val_rmse=0.91291 val_adapt_rmse=0.76874 adapt_steps=2.81 halt=0.36 train_steps=2.21
ep100 val_rmse=0.88727 val_adapt_rmse=0.77048 adapt_steps=2.72 halt=0.36 train_steps=2.25
[ogbg-mollipo_gin_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rmse': np.float32(0.88727427)} test={'rmse': np.float32(1.0246027)} adaptive={'rmse': np.float32(0.86185706)} steps=2.5404761904761903
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_gin_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=gine compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.40935 val_adapt_rmse=1.37490 adapt_steps=6.02 halt=0.16 train_steps=3.96
ep20 val_rmse=1.87982 val_adapt_rmse=1.47829 adapt_steps=5.06 halt=0.19 train_steps=3.73
ep30 val_rmse=1.00892 val_adapt_rmse=0.95657 adapt_steps=4.83 halt=0.24 train_steps=3.32
ep40 val_rmse=1.92570 val_adapt_rmse=1.26685 adapt_steps=4.02 halt=0.23 train_steps=3.29
ep50 val_rmse=0.89318 val_adapt_rmse=0.84550 adapt_steps=4.34 halt=0.21 train_steps=3.32
ep60 val_rmse=0.95712 val_adapt_rmse=0.82385 adapt_steps=3.63 halt=0.29 train_steps=2.75
ep70 val_rmse=0.99810 val_adapt_rmse=0.83945 adapt_steps=3.09 halt=0.30 train_steps=2.58
ep80 val_rmse=0.93375 val_adapt_rmse=0.78058 adapt_steps=2.90 halt=0.34 train_steps=2.41
ep90 val_rmse=0.87733 val_adapt_rmse=0.75005 adapt_steps=2.50 halt=0.36 train_steps=2.25
ep100 val_rmse=0.89479 val_adapt_rmse=0.75851 adapt_steps=2.63 halt=0.37 train_steps=2.22
[ogbg-mollipo_gine_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rmse': np.float32(0.87732804)} test={'rmse': np.float32(0.9709947)} adaptive={'rmse': np.float32(0.82210046)} steps=2.395238095238095
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_gine_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=gcn compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.21491 val_adapt_rmse=1.09189 adapt_steps=6.06 halt=0.20 train_steps=3.72
ep20 val_rmse=0.94709 val_adapt_rmse=0.88665 adapt_steps=4.28 halt=0.20 train_steps=3.65
ep30 val_rmse=1.53588 val_adapt_rmse=1.21575 adapt_steps=5.26 halt=0.23 train_steps=3.20
ep40 val_rmse=1.16658 val_adapt_rmse=0.91900 adapt_steps=2.85 halt=0.29 train_steps=2.66
ep50 val_rmse=1.44945 val_adapt_rmse=0.92876 adapt_steps=3.54 halt=0.28 train_steps=2.66
ep60 val_rmse=1.13499 val_adapt_rmse=0.82488 adapt_steps=2.63 halt=0.38 train_steps=2.07
ep70 val_rmse=0.91036 val_adapt_rmse=0.73458 adapt_steps=2.57 halt=0.41 train_steps=1.95
ep80 val_rmse=1.03989 val_adapt_rmse=0.75262 adapt_steps=2.41 halt=0.42 train_steps=1.81
ep90 val_rmse=0.94022 val_adapt_rmse=0.73763 adapt_steps=2.20 halt=0.44 train_steps=1.74
ep100 val_rmse=0.93598 val_adapt_rmse=0.74143 adapt_steps=2.24 halt=0.44 train_steps=1.75
[ogbg-mollipo_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rmse': np.float32(0.91035604)} test={'rmse': np.float32(0.9299253)} adaptive={'rmse': np.float32(0.80096245)} steps=2.4785714285714286
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_gcn_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=graphsage compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.07418 val_adapt_rmse=0.96580 adapt_steps=5.63 halt=0.17 train_steps=4.01
ep20 val_rmse=1.90831 val_adapt_rmse=1.30970 adapt_steps=4.28 halt=0.20 train_steps=3.73
ep30 val_rmse=1.69953 val_adapt_rmse=1.09305 adapt_steps=3.80 halt=0.28 train_steps=2.82
ep40 val_rmse=1.27229 val_adapt_rmse=0.92711 adapt_steps=3.40 halt=0.32 train_steps=2.51
ep50 val_rmse=1.35372 val_adapt_rmse=0.91724 adapt_steps=2.40 halt=0.37 train_steps=2.18
ep60 val_rmse=1.15361 val_adapt_rmse=0.77828 adapt_steps=2.54 halt=0.37 train_steps=2.12
ep70 val_rmse=1.14055 val_adapt_rmse=0.76690 adapt_steps=2.25 halt=0.40 train_steps=1.94
ep80 val_rmse=1.12685 val_adapt_rmse=0.76632 adapt_steps=2.16 halt=0.42 train_steps=1.82
ep90 val_rmse=1.16042 val_adapt_rmse=0.74041 adapt_steps=2.18 halt=0.43 train_steps=1.75
ep100 val_rmse=1.14354 val_adapt_rmse=0.74383 adapt_steps=2.17 halt=0.44 train_steps=1.72
[ogbg-mollipo_graphsage_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=10 val={'rmse': np.float32(1.0741799)} test={'rmse': np.float32(1.1165451)} adaptive={'rmse': np.float32(0.9747765)} steps=5.461904761904762
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_graphsage_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=gatv2 compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.08473 val_adapt_rmse=1.00704 adapt_steps=4.76 halt=0.23 train_steps=3.42
ep20 val_rmse=1.22357 val_adapt_rmse=1.10102 adapt_steps=5.70 halt=0.18 train_steps=3.84
ep30 val_rmse=1.41681 val_adapt_rmse=1.07035 adapt_steps=3.46 halt=0.24 train_steps=3.17
ep40 val_rmse=1.58552 val_adapt_rmse=1.03190 adapt_steps=3.37 halt=0.32 train_steps=2.55
ep50 val_rmse=0.95944 val_adapt_rmse=0.79974 adapt_steps=4.35 halt=0.32 train_steps=2.42
ep60 val_rmse=0.88782 val_adapt_rmse=0.74478 adapt_steps=2.66 halt=0.36 train_steps=2.22
ep70 val_rmse=0.88373 val_adapt_rmse=0.74463 adapt_steps=2.40 halt=0.41 train_steps=1.95
ep80 val_rmse=0.89006 val_adapt_rmse=0.73390 adapt_steps=2.38 halt=0.41 train_steps=1.89
ep90 val_rmse=0.93277 val_adapt_rmse=0.72185 adapt_steps=2.30 halt=0.44 train_steps=1.78
ep100 val_rmse=0.89800 val_adapt_rmse=0.71467 adapt_steps=2.27 halt=0.44 train_steps=1.75
[ogbg-mollipo_gatv2_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rmse': np.float32(0.88372606)} test={'rmse': np.float32(0.9250613)} adaptive={'rmse': np.float32(0.79357445)} steps=2.2928571428571427
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_gatv2_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=graphconv compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.53265 val_adapt_rmse=1.42937 adapt_steps=6.60 halt=0.17 train_steps=3.92
ep20 val_rmse=1.29141 val_adapt_rmse=0.99349 adapt_steps=4.15 halt=0.19 train_steps=3.62
ep30 val_rmse=1.13429 val_adapt_rmse=0.86465 adapt_steps=3.24 halt=0.24 train_steps=3.34
ep40 val_rmse=1.92000 val_adapt_rmse=1.25179 adapt_steps=3.38 halt=0.31 train_steps=2.61
ep50 val_rmse=1.34785 val_adapt_rmse=0.94338 adapt_steps=3.02 halt=0.37 train_steps=2.19
ep60 val_rmse=1.07604 val_adapt_rmse=0.77996 adapt_steps=2.36 halt=0.39 train_steps=1.97
ep70 val_rmse=1.03247 val_adapt_rmse=0.74302 adapt_steps=2.47 halt=0.39 train_steps=1.97
ep80 val_rmse=1.07814 val_adapt_rmse=0.77095 adapt_steps=2.34 halt=0.43 train_steps=1.78
ep90 val_rmse=0.98119 val_adapt_rmse=0.74076 adapt_steps=2.20 halt=0.44 train_steps=1.73
ep100 val_rmse=0.96236 val_adapt_rmse=0.73650 adapt_steps=2.20 halt=0.44 train_steps=1.71
[ogbg-mollipo_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rmse': np.float32(0.96235555)} test={'rmse': np.float32(0.9985843)} adaptive={'rmse': np.float32(0.76624817)} steps=2.123809523809524
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_graphconv_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=transformer compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=2.42198 val_adapt_rmse=1.36438 adapt_steps=3.38 halt=0.21 train_steps=3.54
ep20 val_rmse=1.25519 val_adapt_rmse=0.94352 adapt_steps=3.34 halt=0.23 train_steps=3.29
ep30 val_rmse=0.88739 val_adapt_rmse=0.77527 adapt_steps=3.58 halt=0.29 train_steps=2.79
ep40 val_rmse=0.94143 val_adapt_rmse=0.72441 adapt_steps=2.57 halt=0.36 train_steps=2.34
ep50 val_rmse=1.17491 val_adapt_rmse=0.78578 adapt_steps=2.60 halt=0.39 train_steps=1.98
ep60 val_rmse=1.07281 val_adapt_rmse=0.74421 adapt_steps=2.29 halt=0.42 train_steps=1.87
ep70 val_rmse=0.86657 val_adapt_rmse=0.73868 adapt_steps=2.39 halt=0.43 train_steps=1.77
ep80 val_rmse=1.02592 val_adapt_rmse=0.70370 adapt_steps=2.29 halt=0.45 train_steps=1.68
ep90 val_rmse=0.92289 val_adapt_rmse=0.68310 adapt_steps=2.12 halt=0.45 train_steps=1.67
ep100 val_rmse=0.96157 val_adapt_rmse=0.68717 adapt_steps=2.11 halt=0.45 train_steps=1.66
[ogbg-mollipo_transformer_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rmse': np.float32(0.8665701)} test={'rmse': np.float32(0.8929515)} adaptive={'rmse': np.float32(0.7886162)} steps=2.316666666666667
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_transformer_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=pna compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --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_rmse=1.28346 val_adapt_rmse=1.19172 adapt_steps=7.30 halt=0.17 train_steps=4.01
ep20 val_rmse=1.19132 val_adapt_rmse=1.15026 adapt_steps=5.02 halt=0.16 train_steps=4.12
ep30 val_rmse=1.24669 val_adapt_rmse=1.23446 adapt_steps=7.86 halt=0.17 train_steps=3.95
ep40 val_rmse=1.09641 val_adapt_rmse=1.01818 adapt_steps=6.98 halt=0.17 train_steps=3.91
ep50 val_rmse=1.74926 val_adapt_rmse=1.34685 adapt_steps=6.52 halt=0.19 train_steps=3.68
ep60 val_rmse=1.01835 val_adapt_rmse=1.01362 adapt_steps=7.48 halt=0.19 train_steps=3.73
ep70 val_rmse=0.91474 val_adapt_rmse=0.86541 adapt_steps=4.82 halt=0.22 train_steps=3.43
ep80 val_rmse=1.25632 val_adapt_rmse=0.98700 adapt_steps=3.31 halt=0.29 train_steps=2.85
ep90 val_rmse=1.07636 val_adapt_rmse=0.88479 adapt_steps=3.19 halt=0.30 train_steps=2.75
ep100 val_rmse=0.96630 val_adapt_rmse=0.82440 adapt_steps=3.11 halt=0.32 train_steps=2.60
[ogbg-mollipo_pna_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rmse': np.float32(0.91474026)} test={'rmse': np.float32(0.92669034)} adaptive={'rmse': np.float32(0.87501276)} steps=4.614285714285714
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_pna_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=gen compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.86424 val_adapt_rmse=1.27606 adapt_steps=3.81 halt=0.18 train_steps=4.03
ep20 val_rmse=1.04319 val_adapt_rmse=1.04292 adapt_steps=7.96 halt=0.22 train_steps=3.37
ep30 val_rmse=0.92998 val_adapt_rmse=0.93388 adapt_steps=4.37 halt=0.24 train_steps=3.26
ep40 val_rmse=1.51121 val_adapt_rmse=1.15834 adapt_steps=4.63 halt=0.23 train_steps=3.36
ep50 val_rmse=0.96101 val_adapt_rmse=0.84809 adapt_steps=3.66 halt=0.26 train_steps=3.06
ep60 val_rmse=1.14213 val_adapt_rmse=1.05520 adapt_steps=3.24 halt=0.28 train_steps=2.91
ep70 val_rmse=0.93218 val_adapt_rmse=0.81774 adapt_steps=3.45 halt=0.33 train_steps=2.58
ep80 val_rmse=0.95239 val_adapt_rmse=0.77677 adapt_steps=2.99 halt=0.35 train_steps=2.34
ep90 val_rmse=0.93136 val_adapt_rmse=0.76093 adapt_steps=2.85 halt=0.36 train_steps=2.30
ep100 val_rmse=0.93118 val_adapt_rmse=0.76794 adapt_steps=2.76 halt=0.36 train_steps=2.30
[ogbg-mollipo_gen_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=30 val={'rmse': np.float32(0.9299843)} test={'rmse': np.float32(0.9385538)} adaptive={'rmse': np.float32(0.9375044)} steps=4.252380952380952
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_gen_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=film compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.88452 val_adapt_rmse=1.87707 adapt_steps=7.73 halt=0.15 train_steps=4.10
ep20 val_rmse=1.09441 val_adapt_rmse=1.07942 adapt_steps=4.85 halt=0.16 train_steps=4.14
ep30 val_rmse=1.46053 val_adapt_rmse=1.45161 adapt_steps=6.21 halt=0.18 train_steps=3.76
ep40 val_rmse=1.06170 val_adapt_rmse=1.01696 adapt_steps=6.72 halt=0.19 train_steps=3.72
ep50 val_rmse=1.10540 val_adapt_rmse=0.99523 adapt_steps=4.06 halt=0.22 train_steps=3.40
ep60 val_rmse=1.28399 val_adapt_rmse=1.00615 adapt_steps=3.50 halt=0.25 train_steps=2.98
ep70 val_rmse=0.98637 val_adapt_rmse=0.86637 adapt_steps=3.27 halt=0.28 train_steps=2.68
ep80 val_rmse=1.05416 val_adapt_rmse=0.83985 adapt_steps=3.08 halt=0.33 train_steps=2.34
ep90 val_rmse=1.01988 val_adapt_rmse=0.82922 adapt_steps=2.91 halt=0.34 train_steps=2.30
ep100 val_rmse=1.01144 val_adapt_rmse=0.82713 adapt_steps=2.86 halt=0.35 train_steps=2.21
[ogbg-mollipo_film_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rmse': np.float32(0.9863713)} test={'rmse': np.float32(0.97120994)} adaptive={'rmse': np.float32(0.8344992)} steps=3.211904761904762
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_film_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=resgated compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.34779 val_adapt_rmse=0.93483 adapt_steps=4.54 halt=0.20 train_steps=3.73
ep20 val_rmse=1.15086 val_adapt_rmse=0.84117 adapt_steps=3.96 halt=0.23 train_steps=3.41
ep30 val_rmse=1.08380 val_adapt_rmse=0.86139 adapt_steps=3.40 halt=0.31 train_steps=2.63
ep40 val_rmse=1.24278 val_adapt_rmse=1.00814 adapt_steps=2.49 halt=0.35 train_steps=2.27
ep50 val_rmse=0.85133 val_adapt_rmse=0.72631 adapt_steps=2.55 halt=0.35 train_steps=2.27
ep60 val_rmse=0.84346 val_adapt_rmse=0.71899 adapt_steps=2.54 halt=0.39 train_steps=2.02
ep70 val_rmse=0.83666 val_adapt_rmse=0.72927 adapt_steps=2.23 halt=0.43 train_steps=1.81
ep80 val_rmse=0.83940 val_adapt_rmse=0.68725 adapt_steps=2.22 halt=0.45 train_steps=1.70
ep90 val_rmse=0.86161 val_adapt_rmse=0.68760 adapt_steps=2.14 halt=0.45 train_steps=1.67
ep100 val_rmse=0.83073 val_adapt_rmse=0.67974 adapt_steps=2.15 halt=0.46 train_steps=1.66
[ogbg-mollipo_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=100 val={'rmse': np.float32(0.8307315)} test={'rmse': np.float32(0.88247204)} adaptive={'rmse': np.float32(0.7250288)} steps=2.1166666666666667
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_resgated_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=tag compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.43440 val_adapt_rmse=1.35579 adapt_steps=6.75 halt=0.16 train_steps=4.16
ep20 val_rmse=1.18041 val_adapt_rmse=1.01839 adapt_steps=4.81 halt=0.23 train_steps=3.30
ep30 val_rmse=1.33342 val_adapt_rmse=1.19676 adapt_steps=4.62 halt=0.24 train_steps=3.29
ep40 val_rmse=1.14385 val_adapt_rmse=0.94329 adapt_steps=3.83 halt=0.30 train_steps=2.73
ep50 val_rmse=0.98573 val_adapt_rmse=0.81495 adapt_steps=3.43 halt=0.29 train_steps=2.84
ep60 val_rmse=1.09272 val_adapt_rmse=0.87901 adapt_steps=3.29 halt=0.35 train_steps=2.31
ep70 val_rmse=0.98565 val_adapt_rmse=0.80267 adapt_steps=2.66 halt=0.37 train_steps=2.19
ep80 val_rmse=0.90149 val_adapt_rmse=0.75615 adapt_steps=2.45 halt=0.38 train_steps=2.15
ep90 val_rmse=0.86151 val_adapt_rmse=0.74536 adapt_steps=2.43 halt=0.39 train_steps=2.05
ep100 val_rmse=0.86321 val_adapt_rmse=0.74945 adapt_steps=2.36 halt=0.40 train_steps=2.01
[ogbg-mollipo_tag_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rmse': np.float32(0.86150765)} test={'rmse': np.float32(0.9196055)} adaptive={'rmse': np.float32(0.80384284)} steps=2.35
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_tag_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=sgc compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.56494 val_adapt_rmse=1.41555 adapt_steps=7.49 halt=0.17 train_steps=3.92
ep20 val_rmse=2.31785 val_adapt_rmse=1.61681 adapt_steps=4.02 halt=0.21 train_steps=3.49
ep30 val_rmse=1.08211 val_adapt_rmse=1.00973 adapt_steps=3.67 halt=0.27 train_steps=3.19
ep40 val_rmse=1.31061 val_adapt_rmse=0.94954 adapt_steps=3.96 halt=0.26 train_steps=2.95
ep50 val_rmse=1.32023 val_adapt_rmse=0.95793 adapt_steps=3.31 halt=0.26 train_steps=2.99
ep60 val_rmse=1.03568 val_adapt_rmse=0.85814 adapt_steps=3.84 halt=0.33 train_steps=2.48
ep70 val_rmse=1.04043 val_adapt_rmse=0.80950 adapt_steps=2.95 halt=0.33 train_steps=2.40
ep80 val_rmse=0.91964 val_adapt_rmse=0.75463 adapt_steps=2.46 halt=0.39 train_steps=2.06
ep90 val_rmse=1.05229 val_adapt_rmse=0.75106 adapt_steps=2.40 halt=0.40 train_steps=1.98
ep100 val_rmse=0.99474 val_adapt_rmse=0.76454 adapt_steps=2.36 halt=0.40 train_steps=1.96
[ogbg-mollipo_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=80 val={'rmse': np.float32(0.9196411)} test={'rmse': np.float32(1.0144817)} adaptive={'rmse': np.float32(0.84780806)} steps=2.3452380952380953
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_sgc_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=cheb compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.42499 val_adapt_rmse=1.05416 adapt_steps=3.42 halt=0.19 train_steps=3.92
ep20 val_rmse=1.07841 val_adapt_rmse=0.88516 adapt_steps=3.39 halt=0.32 train_steps=2.56
ep30 val_rmse=1.03923 val_adapt_rmse=0.87261 adapt_steps=4.42 halt=0.31 train_steps=2.51
ep40 val_rmse=1.21034 val_adapt_rmse=0.82150 adapt_steps=2.78 halt=0.33 train_steps=2.51
ep50 val_rmse=1.11343 val_adapt_rmse=0.92757 adapt_steps=2.13 halt=0.42 train_steps=1.90
ep60 val_rmse=0.96797 val_adapt_rmse=0.88460 adapt_steps=2.27 halt=0.43 train_steps=1.82
ep70 val_rmse=0.81329 val_adapt_rmse=0.73885 adapt_steps=2.09 halt=0.44 train_steps=1.74
ep80 val_rmse=0.79543 val_adapt_rmse=0.73836 adapt_steps=2.12 halt=0.45 train_steps=1.70
ep90 val_rmse=0.74588 val_adapt_rmse=0.70559 adapt_steps=2.05 halt=0.46 train_steps=1.64
ep100 val_rmse=0.75996 val_adapt_rmse=0.71414 adapt_steps=2.04 halt=0.47 train_steps=1.61
[ogbg-mollipo_cheb_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=90 val={'rmse': np.float32(0.74587685)} test={'rmse': np.float32(0.80546176)} adaptive={'rmse': np.float32(0.74045366)} steps=2.033333333333333
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_cheb_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=arma compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=108.34754 val_adapt_rmse=108.19418 adapt_steps=6.18 halt=0.21 train_steps=3.49
ep20 val_rmse=2266.93286 val_adapt_rmse=2266.91235 adapt_steps=4.57 halt=0.23 train_steps=3.37
ep30 val_rmse=2111.57764 val_adapt_rmse=2111.54224 adapt_steps=3.60 halt=0.26 train_steps=3.07
ep40 val_rmse=108348.03906 val_adapt_rmse=108348.01562 adapt_steps=3.31 halt=0.31 train_steps=2.56
ep50 val_rmse=182760.04688 val_adapt_rmse=182759.93750 adapt_steps=3.01 halt=0.35 train_steps=2.25
ep60 val_rmse=164947.39062 val_adapt_rmse=164947.28125 adapt_steps=3.77 halt=0.34 train_steps=2.36
ep70 val_rmse=15163.92676 val_adapt_rmse=15163.80664 adapt_steps=2.50 halt=0.41 train_steps=1.95
ep80 val_rmse=25785.35156 val_adapt_rmse=25785.21094 adapt_steps=2.29 halt=0.40 train_steps=1.97
ep90 val_rmse=17197.17773 val_adapt_rmse=17197.09961 adapt_steps=2.17 halt=0.44 train_steps=1.78
ep100 val_rmse=18469.90625 val_adapt_rmse=18469.83008 adapt_steps=2.14 halt=0.44 train_steps=1.76
[ogbg-mollipo_arma_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=10 val={'rmse': np.float32(108.34754)} test={'rmse': np.float32(1.5814604)} adaptive={'rmse': np.float32(1.2714993)} steps=6.019047619047619
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_arma_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=mf compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.22657 val_adapt_rmse=1.19078 adapt_steps=6.04 halt=0.16 train_steps=4.01
ep20 val_rmse=1.00560 val_adapt_rmse=0.97855 adapt_steps=4.33 halt=0.23 train_steps=3.33
ep30 val_rmse=1.42232 val_adapt_rmse=1.00428 adapt_steps=3.31 halt=0.28 train_steps=2.89
ep40 val_rmse=1.33162 val_adapt_rmse=0.99679 adapt_steps=3.43 halt=0.32 train_steps=2.59
ep50 val_rmse=1.29004 val_adapt_rmse=0.84554 adapt_steps=3.06 halt=0.35 train_steps=2.23
ep60 val_rmse=1.26334 val_adapt_rmse=0.87993 adapt_steps=2.58 halt=0.40 train_steps=2.00
ep70 val_rmse=1.29285 val_adapt_rmse=0.77047 adapt_steps=2.37 halt=0.41 train_steps=1.92
ep80 val_rmse=1.30170 val_adapt_rmse=0.73633 adapt_steps=2.18 halt=0.42 train_steps=1.83
ep90 val_rmse=1.13537 val_adapt_rmse=0.75800 adapt_steps=2.15 halt=0.44 train_steps=1.74
ep100 val_rmse=1.07797 val_adapt_rmse=0.75675 adapt_steps=2.15 halt=0.44 train_steps=1.75
[ogbg-mollipo_mf_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=20 val={'rmse': np.float32(1.005599)} test={'rmse': np.float32(0.98203164)} adaptive={'rmse': np.float32(0.9414124)} steps=4.321428571428571
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_mf_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json
[run] ogbg-mollipo view=appnp compute=rrog-act mode=stream T=1 ns=3 seed=0 device=cuda:1
python3 rrog/train_ogb_graphprop.py --dataset ogbg-mollipo --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:1 --num_workers 0
ep10 val_rmse=1.16326 val_adapt_rmse=1.09274 adapt_steps=5.73 halt=0.21 train_steps=3.61
ep20 val_rmse=2.63258 val_adapt_rmse=1.56333 adapt_steps=4.45 halt=0.24 train_steps=3.16
ep30 val_rmse=1.78657 val_adapt_rmse=1.24891 adapt_steps=4.25 halt=0.27 train_steps=3.00
ep40 val_rmse=1.38779 val_adapt_rmse=1.02154 adapt_steps=2.87 halt=0.32 train_steps=2.53
ep50 val_rmse=1.18581 val_adapt_rmse=0.83394 adapt_steps=2.96 halt=0.34 train_steps=2.37
ep60 val_rmse=1.45301 val_adapt_rmse=0.94618 adapt_steps=2.93 halt=0.35 train_steps=2.25
ep70 val_rmse=1.13892 val_adapt_rmse=0.85011 adapt_steps=2.33 halt=0.43 train_steps=1.82
ep80 val_rmse=1.14148 val_adapt_rmse=0.83007 adapt_steps=2.34 halt=0.43 train_steps=1.77
ep90 val_rmse=1.20479 val_adapt_rmse=0.82034 adapt_steps=2.19 halt=0.46 train_steps=1.65
ep100 val_rmse=1.18848 val_adapt_rmse=0.81683 adapt_steps=2.25 halt=0.45 train_steps=1.66
[ogbg-mollipo_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0] best_ep=70 val={'rmse': np.float32(1.1389227)} test={'rmse': np.float32(1.0732719)} adaptive={'rmse': np.float32(0.83319956)} steps=2.2404761904761905
  wrote /orion/u/oscarwan/rrog-gnn-runner/runs/ogbg-mollipo_appnp_rrog-act_T1_ns3_stream_hm8_hmin2_loss0.2_lq0.1_hex0.1_qw0_h128_e100_s0.json