sync-o1-baseline-artifacts

.gitattributes

@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/report/O1_baseline_report.pdf filter=lfs diff=lfs merge=lfs -text

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/compiled/manifest.json

@@ -0,0 +1,5 @@
+{
+  "training_runs_csv": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/compiled/training_runs.csv",
+  "training_runs_json": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/compiled/training_runs.json",
+  "table4_metrics_csv": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/compiled/table4_metrics.csv"
+}

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/compiled/table4_metrics.csv

@@ -0,0 +1,7 @@
+model,checkpoint,numel,numel_m,real_params,real_params_m,trainable,mse,rmse,mae,r2,pfa_mse_g2,pfa_rmse_g,pfa_mae_g,pfa_r2,loss,spectral_log_mse,high_freq_log_mse,latency_ms_per_sample
+MLP,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt,9444352,9.444352,9444352,9.444352,9444352,0.7419893067526742,0.8613880117302969,0.5303239485549371,0.04620038990764663,0.06070778257877365,0.24638949364527224,0.21065238288262672,-1.2870233889936116,0.7421306194143092,1.255192926166751,1.0909596579435588,0.07492434233427048
+BiLSTM,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt,8410113,8.410113,8410113,8.410113,8410113,0.009918861484228735,0.09959348113319834,0.04322137715481346,0.9872496731015951,0.0003300793328829806,0.01816808555910558,0.010824900093196154,0.987565034951701,0.009924071305651596,0.18528397405964828,0.22222943357401143,1.0494697839021683
+ResCNN1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt,7355137,7.355137,7355137,7.355137,7355137,0.14809687330360008,0.38483356571848054,0.21798509993385967,0.8096269869021784,0.00298788491202187,0.054661548752499414,0.04117521453913263,0.8874384405566411,0.14820588479615554,0.48400715754387225,0.5338901208788509,0.6065387278795242
+UNet1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D/checkpoints/UNet1D_best.pt,7848321,7.848321,7848321,7.848321,7848321,0.21574773690249133,0.46448653037789084,0.27170887322947895,0.7226643221632085,0.011308721598568394,0.10634247316368184,0.0867897865332614,0.5739704252583557,0.21578753469225645,0.5880364352577403,0.6328961368359572,0.09581409394741058
+Transformer,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer/checkpoints/Transformer_best.pt,7957775,7.957775,7957775,7.957775,7957775,0.021390360317785104,0.14625443691657736,0.06603830133691736,0.9725034887360728,0.0005334253375070953,0.023096002630479052,0.01494890111690535,0.9799044509395886,0.0214709252808199,0.32503730780830337,0.39072677264117744,0.10578092187643051
+FNO-Large,6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt,8505793,8.505793,16927169,16.927169,8505793,0.04488089194907147,0.21185110797225365,0.1050106563313501,0.9423072854931448,0.001087754676069545,0.03298112605824042,0.020487637613904902,0.9590213926454951,0.04496160251073029,0.3416421673399337,0.3977926703209573,0.10789800435304642

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/compiled/training_runs.csv

@@ -0,0 +1,7 @@
+model,run_dir,best_checkpoint,best_val_mse,train_time_s,train_time_h,peak_gpu_gb,params_numel,params_m,batch_size,epochs,split
+BiLSTM,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt,0.010168918287935313,27026.021858215332,7.507228293948703,15.61366891860962,8410113,8.410113,64,50,fixed_474_test_full_57_factors
+FNO-Large-bs64,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt,0.043961879449736535,7784.3661642074585,2.162323934502072,3.2975897789001465,8505793,8.505793,64,50,fixed_474_test_full_57_factors
+MLP,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt,0.7341581120553119,5709.6286454200745,1.5860079570611318,0.24129295349121094,9444352,9.444352,64,50,fixed_474_test_full_57_factors
+ResCNN1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt,0.1459837291314495,22628.45619750023,6.285682277083397,3.494046688079834,7355137,7.355137,64,50,fixed_474_test_full_57_factors
+Transformer,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer/checkpoints/Transformer_best.pt,0.021797189592803894,8430.805659532547,2.341890460981263,2.5364460945129395,7957775,7.957775,64,50,fixed_474_test_full_57_factors
+UNet1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D/checkpoints/UNet1D_best.pt,0.2145316383596796,8155.616280555725,2.265448966821035,3.032552719116211,7848321,7.848321,64,50,fixed_474_test_full_57_factors

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/compiled/training_runs.json

@@ -0,0 +1,86 @@
+[
+  {
+    "model": "BiLSTM",
+    "run_dir": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM",
+    "best_checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt",
+    "best_val_mse": 0.010168918287935313,
+    "train_time_s": 27026.021858215332,
+    "train_time_h": 7.507228293948703,
+    "peak_gpu_gb": 15.61366891860962,
+    "params_numel": 8410113,
+    "params_m": 8.410113,
+    "batch_size": 64,
+    "epochs": 50,
+    "split": "fixed_474_test_full_57_factors"
+  },
+  {
+    "model": "FNO-Large-bs64",
+    "run_dir": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64",
+    "best_checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt",
+    "best_val_mse": 0.043961879449736535,
+    "train_time_s": 7784.3661642074585,
+    "train_time_h": 2.162323934502072,
+    "peak_gpu_gb": 3.2975897789001465,
+    "params_numel": 8505793,
+    "params_m": 8.505793,
+    "batch_size": 64,
+    "epochs": 50,
+    "split": "fixed_474_test_full_57_factors"
+  },
+  {
+    "model": "MLP",
+    "run_dir": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP",
+    "best_checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt",
+    "best_val_mse": 0.7341581120553119,
+    "train_time_s": 5709.6286454200745,
+    "train_time_h": 1.5860079570611318,
+    "peak_gpu_gb": 0.24129295349121094,
+    "params_numel": 9444352,
+    "params_m": 9.444352,
+    "batch_size": 64,
+    "epochs": 50,
+    "split": "fixed_474_test_full_57_factors"
+  },
+  {
+    "model": "ResCNN1D",
+    "run_dir": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D",
+    "best_checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt",
+    "best_val_mse": 0.1459837291314495,
+    "train_time_s": 22628.45619750023,
+    "train_time_h": 6.285682277083397,
+    "peak_gpu_gb": 3.494046688079834,
+    "params_numel": 7355137,
+    "params_m": 7.355137,
+    "batch_size": 64,
+    "epochs": 50,
+    "split": "fixed_474_test_full_57_factors"
+  },
+  {
+    "model": "Transformer",
+    "run_dir": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer",
+    "best_checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer/checkpoints/Transformer_best.pt",
+    "best_val_mse": 0.021797189592803894,
+    "train_time_s": 8430.805659532547,
+    "train_time_h": 2.341890460981263,
+    "peak_gpu_gb": 2.5364460945129395,
+    "params_numel": 7957775,
+    "params_m": 7.957775,
+    "batch_size": 64,
+    "epochs": 50,
+    "split": "fixed_474_test_full_57_factors"
+  },
+  {
+    "model": "UNet1D",
+    "run_dir": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D",
+    "best_checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D/checkpoints/UNet1D_best.pt",
+    "best_val_mse": 0.2145316383596796,
+    "train_time_s": 8155.616280555725,
+    "train_time_h": 2.265448966821035,
+    "peak_gpu_gb": 3.032552719116211,
+    "params_numel": 7848321,
+    "params_m": 7.848321,
+    "batch_size": 64,
+    "epochs": 50,
+    "split": "fixed_474_test_full_57_factors"
+  }
+]

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/evaluation/table4_metrics.csv

@@ -0,0 +1,7 @@
+model,checkpoint,numel,numel_m,real_params,real_params_m,trainable,mse,rmse,mae,r2,pfa_mse_g2,pfa_rmse_g,pfa_mae_g,pfa_r2,loss,spectral_log_mse,high_freq_log_mse,latency_ms_per_sample
+MLP,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt,9444352,9.444352,9444352,9.444352,9444352,0.7419893067526742,0.8613880117302969,0.5303239485549371,0.04620038990764663,0.06070778257877365,0.24638949364527224,0.21065238288262672,-1.2870233889936116,0.7421306194143092,1.255192926166751,1.0909596579435588,0.07492434233427048
+BiLSTM,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt,8410113,8.410113,8410113,8.410113,8410113,0.009918861484228735,0.09959348113319834,0.04322137715481346,0.9872496731015951,0.0003300793328829806,0.01816808555910558,0.010824900093196154,0.987565034951701,0.009924071305651596,0.18528397405964828,0.22222943357401143,1.0494697839021683
+ResCNN1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt,7355137,7.355137,7355137,7.355137,7355137,0.14809687330360008,0.38483356571848054,0.21798509993385967,0.8096269869021784,0.00298788491202187,0.054661548752499414,0.04117521453913263,0.8874384405566411,0.14820588479615554,0.48400715754387225,0.5338901208788509,0.6065387278795242
+UNet1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D/checkpoints/UNet1D_best.pt,7848321,7.848321,7848321,7.848321,7848321,0.21574773690249133,0.46448653037789084,0.27170887322947895,0.7226643221632085,0.011308721598568394,0.10634247316368184,0.0867897865332614,0.5739704252583557,0.21578753469225645,0.5880364352577403,0.6328961368359572,0.09581409394741058
+Transformer,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer/checkpoints/Transformer_best.pt,7957775,7.957775,7957775,7.957775,7957775,0.021390360317785104,0.14625443691657736,0.06603830133691736,0.9725034887360728,0.0005334253375070953,0.023096002630479052,0.01494890111690535,0.9799044509395886,0.0214709252808199,0.32503730780830337,0.39072677264117744,0.10578092187643051
+FNO-Large,6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt,8505793,8.505793,16927169,16.927169,8505793,0.04488089194907147,0.21185110797225365,0.1050106563313501,0.9423072854931448,0.001087754676069545,0.03298112605824042,0.020487637613904902,0.9590213926454951,0.04496160251073029,0.3416421673399337,0.3977926703209573,0.10789800435304642

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/evaluation/table4_metrics.json

@@ -0,0 +1,128 @@
+[
+  {
+    "model": "MLP",
+    "checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt",
+    "numel": 9444352,
+    "numel_m": 9.444352,
+    "real_params": 9444352,
+    "real_params_m": 9.444352,
+    "trainable": 9444352,
+    "mse": 0.7419893067526742,
+    "rmse": 0.8613880117302969,
+    "mae": 0.5303239485549371,
+    "r2": 0.04620038990764663,
+    "pfa_mse_g2": 0.06070778257877365,
+    "pfa_rmse_g": 0.24638949364527224,
+    "pfa_mae_g": 0.21065238288262672,
+    "pfa_r2": -1.2870233889936116,
+    "loss": 0.7421306194143092,
+    "spectral_log_mse": 1.255192926166751,
+    "high_freq_log_mse": 1.0909596579435588,
+    "latency_ms_per_sample": 0.07492434233427048
+  },
+  {
+    "model": "BiLSTM",
+    "checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt",
+    "numel": 8410113,
+    "numel_m": 8.410113,
+    "real_params": 8410113,
+    "real_params_m": 8.410113,
+    "trainable": 8410113,
+    "mse": 0.009918861484228735,
+    "rmse": 0.09959348113319834,
+    "mae": 0.04322137715481346,
+    "r2": 0.9872496731015951,
+    "pfa_mse_g2": 0.0003300793328829806,
+    "pfa_rmse_g": 0.01816808555910558,
+    "pfa_mae_g": 0.010824900093196154,
+    "pfa_r2": 0.987565034951701,
+    "loss": 0.009924071305651596,
+    "spectral_log_mse": 0.18528397405964828,
+    "high_freq_log_mse": 0.22222943357401143,
+    "latency_ms_per_sample": 1.0494697839021683
+  },
+  {
+    "model": "ResCNN1D",
+    "checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt",
+    "numel": 7355137,
+    "numel_m": 7.355137,
+    "real_params": 7355137,
+    "real_params_m": 7.355137,
+    "trainable": 7355137,
+    "mse": 0.14809687330360008,
+    "rmse": 0.38483356571848054,
+    "mae": 0.21798509993385967,
+    "r2": 0.8096269869021784,
+    "pfa_mse_g2": 0.00298788491202187,
+    "pfa_rmse_g": 0.054661548752499414,
+    "pfa_mae_g": 0.04117521453913263,
+    "pfa_r2": 0.8874384405566411,
+    "loss": 0.14820588479615554,
+    "spectral_log_mse": 0.48400715754387225,
+    "high_freq_log_mse": 0.5338901208788509,
+    "latency_ms_per_sample": 0.6065387278795242
+  },
+  {
+    "model": "UNet1D",
+    "checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D/checkpoints/UNet1D_best.pt",
+    "numel": 7848321,
+    "numel_m": 7.848321,
+    "real_params": 7848321,
+    "real_params_m": 7.848321,
+    "trainable": 7848321,
+    "mse": 0.21574773690249133,
+    "rmse": 0.46448653037789084,
+    "mae": 0.27170887322947895,
+    "r2": 0.7226643221632085,
+    "pfa_mse_g2": 0.011308721598568394,
+    "pfa_rmse_g": 0.10634247316368184,
+    "pfa_mae_g": 0.0867897865332614,
+    "pfa_r2": 0.5739704252583557,
+    "loss": 0.21578753469225645,
+    "spectral_log_mse": 0.5880364352577403,
+    "high_freq_log_mse": 0.6328961368359572,
+    "latency_ms_per_sample": 0.09581409394741058
+  },
+  {
+    "model": "Transformer",
+    "checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer/checkpoints/Transformer_best.pt",
+    "numel": 7957775,
+    "numel_m": 7.957775,
+    "real_params": 7957775,
+    "real_params_m": 7.957775,
+    "trainable": 7957775,
+    "mse": 0.021390360317785104,
+    "rmse": 0.14625443691657736,
+    "mae": 0.06603830133691736,
+    "r2": 0.9725034887360728,
+    "pfa_mse_g2": 0.0005334253375070953,
+    "pfa_rmse_g": 0.023096002630479052,
+    "pfa_mae_g": 0.01494890111690535,
+    "pfa_r2": 0.9799044509395886,
+    "loss": 0.0214709252808199,
+    "spectral_log_mse": 0.32503730780830337,
+    "high_freq_log_mse": 0.39072677264117744,
+    "latency_ms_per_sample": 0.10578092187643051
+  },
+  {
+    "model": "FNO-Large",
+    "checkpoint": "6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt",
+    "numel": 8505793,
+    "numel_m": 8.505793,
+    "real_params": 16927169,
+    "real_params_m": 16.927169,
+    "trainable": 8505793,
+    "mse": 0.04488089194907147,
+    "rmse": 0.21185110797225365,
+    "mae": 0.1050106563313501,
+    "r2": 0.9423072854931448,
+    "pfa_mse_g2": 0.001087754676069545,
+    "pfa_rmse_g": 0.03298112605824042,
+    "pfa_mae_g": 0.020487637613904902,
+    "pfa_r2": 0.9590213926454951,
+    "loss": 0.04496160251073029,
+    "spectral_log_mse": 0.3416421673399337,
+    "high_freq_log_mse": 0.3977926703209573,
+    "latency_ms_per_sample": 0.10789800435304642
+  }
+]

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/figures/o1_plotting_summary.csv

@@ -0,0 +1,7 @@
+model_plot,model,checkpoint,numel,numel_m,real_params,real_params_m,trainable,mse,rmse,mae,r2,pfa_mse_g2,pfa_rmse_g,pfa_mae_g,pfa_r2,loss,spectral_log_mse,high_freq_log_mse,latency_ms_per_sample,train_time_h,peak_gpu_gb,params_m,best_val_mse
+MLP,MLP,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt,9444352,9.444352,9444352,9.444352,9444352,0.7419893067526742,0.8613880117302969,0.5303239485549371,0.0462003899076466,0.0607077825787736,0.2463894936452722,0.2106523828826267,-1.2870233889936116,0.7421306194143092,1.255192926166751,1.0909596579435588,0.0749243423342704,1.5860079570611318,0.2412929534912109,9.444352,0.7341581120553119
+ResCNN1D,ResCNN1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt,7355137,7.355137,7355137,7.355137,7355137,0.1480968733036,0.3848335657184805,0.2179850999338596,0.8096269869021784,0.0029878849120218,0.0546615487524994,0.0411752145391326,0.8874384405566411,0.1482058847961555,0.4840071575438722,0.5338901208788509,0.6065387278795242,6.285682277083397,3.494046688079834,7.355137,0.1459837291314495
+UNet1D,UNet1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D/checkpoints/UNet1D_best.pt,7848321,7.848321,7848321,7.848321,7848321,0.2157477369024913,0.4644865303778908,0.2717088732294789,0.7226643221632085,0.0113087215985683,0.1063424731636818,0.0867897865332614,0.5739704252583557,0.2157875346922564,0.5880364352577403,0.6328961368359572,0.0958140939474105,2.265448966821035,3.032552719116211,7.848321,0.2145316383596796
+FNO-Large,FNO-Large,6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt,8505793,8.505793,16927169,16.927169,8505793,0.0448808919490714,0.2118511079722536,0.1050106563313501,0.9423072854931448,0.0010877546760695,0.0329811260582404,0.0204876376139049,0.9590213926454952,0.0449616025107302,0.3416421673399337,0.3977926703209573,0.1078980043530464,2.162323934502072,3.2975897789001465,8.505793,0.0439618794497365
+Transformer,Transformer,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer/checkpoints/Transformer_best.pt,7957775,7.957775,7957775,7.957775,7957775,0.0213903603177851,0.1462544369165773,0.0660383013369173,0.9725034887360728,0.000533425337507,0.023096002630479,0.0149489011169053,0.9799044509395886,0.0214709252808199,0.3250373078083033,0.3907267726411774,0.1057809218764305,2.341890460981263,2.5364460945129395,7.957775,0.0217971895928038
+BiLSTM,BiLSTM,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt,8410113,8.410113,8410113,8.410113,8410113,0.0099188614842287,0.0995934811331983,0.0432213771548134,0.9872496731015952,0.0003300793328829,0.0181680855591055,0.0108249000931961,0.987565034951701,0.0099240713056515,0.1852839740596482,0.2222294335740114,1.0494697839021685,7.507228293948703,15.61366891860962,8.410113,0.0101689182879353

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/logs/o1_evaluate_after_training.log

@@ -0,0 +1 @@
+Terminated

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/logs/o1_fno_then_evaluate_after_reruns_20260604_230755.log

@@ -0,0 +1,64 @@
+Train samples=136,800, Val=34,200, Test=27,018
+GM path: /data/home/jason/data/SesimicTransformerData/MDOF/All_GMs/GMs_knet_3474_AF_57.h5
+Building dir: /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno
+FNO-Large-bs64 epoch 001/50 train_mse=0.109347 val_mse=0.075765 val_r2=0.9042 peak=3.30GB
+FNO-Large-bs64 epoch 002/50 train_mse=0.050667 val_mse=0.062758 val_r2=0.9207 peak=3.30GB
+FNO-Large-bs64 epoch 003/50 train_mse=0.039961 val_mse=0.057691 val_r2=0.9271 peak=3.30GB
+FNO-Large-bs64 epoch 004/50 train_mse=0.034654 val_mse=0.053967 val_r2=0.9318 peak=3.30GB
+FNO-Large-bs64 epoch 005/50 train_mse=0.031393 val_mse=0.051442 val_r2=0.9350 peak=3.30GB
+FNO-Large-bs64 epoch 006/50 train_mse=0.028979 val_mse=0.050104 val_r2=0.9367 peak=3.30GB
+FNO-Large-bs64 epoch 007/50 train_mse=0.027356 val_mse=0.049187 val_r2=0.9378 peak=3.30GB
+FNO-Large-bs64 epoch 008/50 train_mse=0.025991 val_mse=0.048714 val_r2=0.9384 peak=3.30GB
+FNO-Large-bs64 epoch 009/50 train_mse=0.025239 val_mse=0.048074 val_r2=0.9392 peak=3.30GB
+FNO-Large-bs64 epoch 010/50 train_mse=0.024089 val_mse=0.047391 val_r2=0.9401 peak=3.30GB
+FNO-Large-bs64 epoch 011/50 train_mse=0.022887 val_mse=0.046664 val_r2=0.9410 peak=3.30GB
+FNO-Large-bs64 epoch 012/50 train_mse=0.022727 val_mse=0.048273 val_r2=0.9390 peak=3.30GB
+FNO-Large-bs64 epoch 013/50 train_mse=0.021990 val_mse=0.048622 val_r2=0.9385 peak=3.30GB
+FNO-Large-bs64 epoch 014/50 train_mse=0.021362 val_mse=0.046148 val_r2=0.9417 peak=3.30GB
+FNO-Large-bs64 epoch 015/50 train_mse=0.020810 val_mse=0.047232 val_r2=0.9403 peak=3.30GB
+FNO-Large-bs64 epoch 016/50 train_mse=0.020371 val_mse=0.046007 val_r2=0.9419 peak=3.30GB
+FNO-Large-bs64 epoch 017/50 train_mse=0.019911 val_mse=0.045929 val_r2=0.9420 peak=3.30GB
+FNO-Large-bs64 epoch 018/50 train_mse=0.019654 val_mse=0.044658 val_r2=0.9436 peak=3.30GB
+FNO-Large-bs64 epoch 019/50 train_mse=0.019239 val_mse=0.045382 val_r2=0.9426 peak=3.30GB
+FNO-Large-bs64 epoch 020/50 train_mse=0.018934 val_mse=0.045127 val_r2=0.9430 peak=3.30GB
+FNO-Large-bs64 epoch 021/50 train_mse=0.015681 val_mse=0.044525 val_r2=0.9437 peak=3.30GB
+FNO-Large-bs64 epoch 022/50 train_mse=0.014857 val_mse=0.043962 val_r2=0.9444 peak=3.30GB
+FNO-Large-bs64 epoch 023/50 train_mse=0.014614 val_mse=0.045200 val_r2=0.9429 peak=3.30GB
+FNO-Large-bs64 epoch 024/50 train_mse=0.014478 val_mse=0.044402 val_r2=0.9439 peak=3.30GB
+FNO-Large-bs64 epoch 025/50 train_mse=0.014195 val_mse=0.044697 val_r2=0.9435 peak=3.30GB
+FNO-Large-bs64 epoch 026/50 train_mse=0.014170 val_mse=0.043984 val_r2=0.9444 peak=3.30GB
+FNO-Large-bs64 epoch 027/50 train_mse=0.014089 val_mse=0.044374 val_r2=0.9439 peak=3.30GB
+FNO-Large-bs64 epoch 028/50 train_mse=0.013878 val_mse=0.044886 val_r2=0.9433 peak=3.30GB
+FNO-Large-bs64 epoch 029/50 train_mse=0.013758 val_mse=0.044486 val_r2=0.9438 peak=3.30GB
+FNO-Large-bs64 epoch 030/50 train_mse=0.013653 val_mse=0.045304 val_r2=0.9427 peak=3.30GB
+FNO-Large-bs64 epoch 031/50 train_mse=0.013572 val_mse=0.045018 val_r2=0.9431 peak=3.30GB
+FNO-Large-bs64 epoch 032/50 train_mse=0.013400 val_mse=0.045657 val_r2=0.9423 peak=3.30GB
+FNO-Large-bs64 epoch 033/50 train_mse=0.013477 val_mse=0.044568 val_r2=0.9437 peak=3.30GB
+FNO-Large-bs64 epoch 034/50 train_mse=0.013120 val_mse=0.045295 val_r2=0.9428 peak=3.30GB
+FNO-Large-bs64 epoch 035/50 train_mse=0.013394 val_mse=0.043994 val_r2=0.9444 peak=3.30GB
+FNO-Large-bs64 epoch 036/50 train_mse=0.013161 val_mse=0.045270 val_r2=0.9428 peak=3.30GB
+FNO-Large-bs64 epoch 037/50 train_mse=0.013047 val_mse=0.044843 val_r2=0.9433 peak=3.30GB
+FNO-Large-bs64 epoch 038/50 train_mse=0.013001 val_mse=0.046831 val_r2=0.9408 peak=3.30GB
+FNO-Large-bs64 epoch 039/50 train_mse=0.012873 val_mse=0.044818 val_r2=0.9434 peak=3.30GB
+FNO-Large-bs64 epoch 040/50 train_mse=0.012790 val_mse=0.044886 val_r2=0.9433 peak=3.30GB
+FNO-Large-bs64 epoch 041/50 train_mse=0.011146 val_mse=0.045627 val_r2=0.9423 peak=3.30GB
+FNO-Large-bs64 epoch 042/50 train_mse=0.010575 val_mse=0.045534 val_r2=0.9425 peak=3.30GB
+FNO-Large-bs64 epoch 043/50 train_mse=0.010580 val_mse=0.045311 val_r2=0.9427 peak=3.30GB
+FNO-Large-bs64 epoch 044/50 train_mse=0.010574 val_mse=0.044932 val_r2=0.9432 peak=3.30GB
+FNO-Large-bs64 epoch 045/50 train_mse=0.010462 val_mse=0.045291 val_r2=0.9428 peak=3.30GB
+FNO-Large-bs64 epoch 046/50 train_mse=0.010377 val_mse=0.044833 val_r2=0.9433 peak=3.30GB
+FNO-Large-bs64 epoch 047/50 train_mse=0.010405 val_mse=0.045327 val_r2=0.9427 peak=3.30GB
+FNO-Large-bs64 epoch 048/50 train_mse=0.010360 val_mse=0.045194 val_r2=0.9429 peak=3.30GB
+FNO-Large-bs64 epoch 049/50 train_mse=0.010323 val_mse=0.045951 val_r2=0.9419 peak=3.30GB
+FNO-Large-bs64 epoch 050/50 train_mse=0.010286 val_mse=0.045115 val_r2=0.9430 peak=3.30GB
+Summary saved to /data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/summary.json
+/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/scripts/baseline_models.py:225: UserWarning: enable_nested_tensor is True, but self.use_nested_tensor is False because encoder_layer.norm_first was True
+  self.encoder = nn.TransformerEncoder(layer, num_layers=n_layers)
+MLP: mse=0.741989, r2=0.0462, pfa_rmse=0.2464g
+BiLSTM: mse=0.009919, r2=0.9872, pfa_rmse=0.0182g
+ResCNN1D: mse=0.148097, r2=0.8096, pfa_rmse=0.0547g
+UNet1D: mse=0.215748, r2=0.7227, pfa_rmse=0.1063g
+Transformer: mse=0.021390, r2=0.9725, pfa_rmse=0.0231g
+FNO-Large: mse=0.044881, r2=0.9423, pfa_rmse=0.0330g
+Evaluation table saved to /data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/evaluation/table4_metrics.csv
+Compiled O1 results written to /data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/compiled

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/logs/o1_gpu0_bilstm_rerun_20260604_160724.log

@@ -0,0 +1,54 @@
+Train samples=136,800, Val=34,200, Test=27,018
+GM path: /data/home/jason/data/SesimicTransformerData/MDOF/All_GMs/GMs_knet_3474_AF_57.h5
+Building dir: /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno
+BiLSTM epoch 001/50 train_mse=0.172075 val_mse=0.085573 val_r2=0.8918 peak=15.61GB
+BiLSTM epoch 002/50 train_mse=0.078202 val_mse=0.064748 val_r2=0.9182 peak=15.61GB
+BiLSTM epoch 003/50 train_mse=0.062326 val_mse=0.046812 val_r2=0.9408 peak=15.61GB
+BiLSTM epoch 004/50 train_mse=0.040888 val_mse=0.030838 val_r2=0.9610 peak=15.61GB
+BiLSTM epoch 005/50 train_mse=0.030688 val_mse=0.028225 val_r2=0.9643 peak=15.61GB
+BiLSTM epoch 006/50 train_mse=0.026236 val_mse=0.023391 val_r2=0.9704 peak=15.61GB
+BiLSTM epoch 007/50 train_mse=0.023206 val_mse=0.019243 val_r2=0.9757 peak=15.61GB
+BiLSTM epoch 008/50 train_mse=0.021776 val_mse=0.019011 val_r2=0.9760 peak=15.61GB
+BiLSTM epoch 009/50 train_mse=0.020787 val_mse=0.019621 val_r2=0.9752 peak=15.61GB
+BiLSTM epoch 010/50 train_mse=0.019359 val_mse=0.017954 val_r2=0.9773 peak=15.61GB
+BiLSTM epoch 011/50 train_mse=0.018407 val_mse=0.019130 val_r2=0.9758 peak=15.61GB
+BiLSTM epoch 012/50 train_mse=0.017651 val_mse=0.018976 val_r2=0.9760 peak=15.61GB
+BiLSTM epoch 013/50 train_mse=0.017455 val_mse=0.018072 val_r2=0.9772 peak=15.61GB
+BiLSTM epoch 014/50 train_mse=0.016989 val_mse=0.017660 val_r2=0.9777 peak=15.61GB
+BiLSTM epoch 015/50 train_mse=0.016091 val_mse=0.020201 val_r2=0.9745 peak=15.61GB
+BiLSTM epoch 016/50 train_mse=0.015846 val_mse=0.015529 val_r2=0.9804 peak=15.61GB
+BiLSTM epoch 017/50 train_mse=0.015745 val_mse=0.015665 val_r2=0.9802 peak=15.61GB
+BiLSTM epoch 018/50 train_mse=0.015114 val_mse=0.015575 val_r2=0.9803 peak=15.61GB
+BiLSTM epoch 019/50 train_mse=0.014766 val_mse=0.016227 val_r2=0.9795 peak=15.61GB
+BiLSTM epoch 020/50 train_mse=0.014709 val_mse=0.013847 val_r2=0.9825 peak=15.61GB
+BiLSTM epoch 021/50 train_mse=0.011837 val_mse=0.013152 val_r2=0.9834 peak=15.61GB
+BiLSTM epoch 022/50 train_mse=0.011744 val_mse=0.013466 val_r2=0.9830 peak=15.61GB
+BiLSTM epoch 023/50 train_mse=0.011565 val_mse=0.012684 val_r2=0.9840 peak=15.61GB
+BiLSTM epoch 024/50 train_mse=0.011382 val_mse=0.012386 val_r2=0.9843 peak=15.61GB
+BiLSTM epoch 025/50 train_mse=0.011387 val_mse=0.012980 val_r2=0.9836 peak=15.61GB
+BiLSTM epoch 026/50 train_mse=0.011160 val_mse=0.013992 val_r2=0.9823 peak=15.61GB
+BiLSTM epoch 027/50 train_mse=0.010971 val_mse=0.012865 val_r2=0.9837 peak=15.61GB
+BiLSTM epoch 028/50 train_mse=0.011094 val_mse=0.011667 val_r2=0.9853 peak=15.61GB
+BiLSTM epoch 029/50 train_mse=0.010834 val_mse=0.011820 val_r2=0.9851 peak=15.61GB
+BiLSTM epoch 030/50 train_mse=0.010846 val_mse=0.012404 val_r2=0.9843 peak=15.61GB
+BiLSTM epoch 031/50 train_mse=0.010867 val_mse=0.012351 val_r2=0.9844 peak=15.61GB
+BiLSTM epoch 032/50 train_mse=0.010529 val_mse=0.013023 val_r2=0.9835 peak=15.61GB
+BiLSTM epoch 033/50 train_mse=0.010593 val_mse=0.012784 val_r2=0.9838 peak=15.61GB
+BiLSTM epoch 034/50 train_mse=0.010584 val_mse=0.011931 val_r2=0.9849 peak=15.61GB
+BiLSTM epoch 035/50 train_mse=0.010434 val_mse=0.011833 val_r2=0.9850 peak=15.61GB
+BiLSTM epoch 036/50 train_mse=0.010301 val_mse=0.013812 val_r2=0.9825 peak=15.61GB
+BiLSTM epoch 037/50 train_mse=0.010430 val_mse=0.011738 val_r2=0.9852 peak=15.61GB
+BiLSTM epoch 038/50 train_mse=0.010327 val_mse=0.011669 val_r2=0.9853 peak=15.61GB
+BiLSTM epoch 039/50 train_mse=0.010201 val_mse=0.012006 val_r2=0.9848 peak=15.61GB
+BiLSTM epoch 040/50 train_mse=0.010112 val_mse=0.010684 val_r2=0.9865 peak=15.61GB
+BiLSTM epoch 041/50 train_mse=0.008635 val_mse=0.010945 val_r2=0.9862 peak=15.61GB
+BiLSTM epoch 042/50 train_mse=0.008489 val_mse=0.010664 val_r2=0.9865 peak=15.61GB
+BiLSTM epoch 043/50 train_mse=0.008516 val_mse=0.011552 val_r2=0.9854 peak=15.61GB
+BiLSTM epoch 044/50 train_mse=0.008447 val_mse=0.010472 val_r2=0.9868 peak=15.61GB
+BiLSTM epoch 045/50 train_mse=0.008389 val_mse=0.010769 val_r2=0.9864 peak=15.61GB
+BiLSTM epoch 046/50 train_mse=0.008360 val_mse=0.010392 val_r2=0.9869 peak=15.61GB
+BiLSTM epoch 047/50 train_mse=0.008397 val_mse=0.010816 val_r2=0.9863 peak=15.61GB
+BiLSTM epoch 048/50 train_mse=0.008429 val_mse=0.010169 val_r2=0.9871 peak=15.61GB
+BiLSTM epoch 049/50 train_mse=0.008245 val_mse=0.011141 val_r2=0.9859 peak=15.61GB
+BiLSTM epoch 050/50 train_mse=0.008267 val_mse=0.010677 val_r2=0.9865 peak=15.61GB
+Training summary saved to /data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/training_summary.json

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/logs/o1_gpu1_baselines_then_fno.log

@@ -0,0 +1,198 @@
+Train samples=136,800, Val=34,200, Test=27,018
+GM path: /data/home/jason/data/SesimicTransformerData/MDOF/All_GMs/GMs_knet_3474_AF_57.h5
+Building dir: /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno
+MLP epoch 001/50 train_mse=0.573992 val_mse=0.734158 val_r2=0.0721 peak=0.24GB
+MLP epoch 002/50 train_mse=0.481391 val_mse=0.762920 val_r2=0.0358 peak=0.24GB
+MLP epoch 003/50 train_mse=0.441707 val_mse=0.763343 val_r2=0.0352 peak=0.24GB
+MLP epoch 004/50 train_mse=0.429493 val_mse=0.768376 val_r2=0.0289 peak=0.24GB
+MLP epoch 005/50 train_mse=0.420804 val_mse=0.775209 val_r2=0.0202 peak=0.24GB
+MLP epoch 006/50 train_mse=0.406548 val_mse=0.780200 val_r2=0.0139 peak=0.24GB
+MLP epoch 007/50 train_mse=0.391733 val_mse=0.780065 val_r2=0.0141 peak=0.24GB
+MLP epoch 008/50 train_mse=0.379072 val_mse=0.784916 val_r2=0.0080 peak=0.24GB
+MLP epoch 009/50 train_mse=0.368190 val_mse=0.791738 val_r2=-0.0006 peak=0.24GB
+MLP epoch 010/50 train_mse=0.355676 val_mse=0.797634 val_r2=-0.0081 peak=0.24GB
+MLP epoch 011/50 train_mse=0.349777 val_mse=0.809953 val_r2=-0.0237 peak=0.24GB
+MLP epoch 012/50 train_mse=0.332692 val_mse=0.814750 val_r2=-0.0297 peak=0.24GB
+MLP epoch 013/50 train_mse=0.325539 val_mse=0.813150 val_r2=-0.0277 peak=0.24GB
+MLP epoch 014/50 train_mse=0.333468 val_mse=0.808143 val_r2=-0.0214 peak=0.24GB
+MLP epoch 015/50 train_mse=0.330034 val_mse=0.814987 val_r2=-0.0300 peak=0.24GB
+MLP epoch 016/50 train_mse=0.328472 val_mse=0.821593 val_r2=-0.0384 peak=0.24GB
+MLP epoch 017/50 train_mse=0.328651 val_mse=0.847808 val_r2=-0.0715 peak=0.24GB
+MLP epoch 018/50 train_mse=0.327372 val_mse=0.852995 val_r2=-0.0781 peak=0.24GB
+MLP epoch 019/50 train_mse=0.333472 val_mse=0.834577 val_r2=-0.0548 peak=0.24GB
+MLP epoch 020/50 train_mse=0.317803 val_mse=0.829458 val_r2=-0.0483 peak=0.24GB
+MLP epoch 021/50 train_mse=0.248362 val_mse=0.830128 val_r2=-0.0492 peak=0.24GB
+MLP epoch 022/50 train_mse=0.224993 val_mse=0.845820 val_r2=-0.0690 peak=0.24GB
+MLP epoch 023/50 train_mse=0.214885 val_mse=0.867676 val_r2=-0.0966 peak=0.24GB
+MLP epoch 024/50 train_mse=0.206710 val_mse=0.871796 val_r2=-0.1018 peak=0.24GB
+MLP epoch 025/50 train_mse=0.198998 val_mse=0.885137 val_r2=-0.1187 peak=0.24GB
+MLP epoch 026/50 train_mse=0.194856 val_mse=0.894924 val_r2=-0.1311 peak=0.24GB
+MLP epoch 027/50 train_mse=0.191099 val_mse=0.900148 val_r2=-0.1377 peak=0.24GB
+MLP epoch 028/50 train_mse=0.188248 val_mse=0.900271 val_r2=-0.1378 peak=0.24GB
+MLP epoch 029/50 train_mse=0.185885 val_mse=0.895437 val_r2=-0.1317 peak=0.24GB
+MLP epoch 030/50 train_mse=0.185323 val_mse=0.908469 val_r2=-0.1482 peak=0.24GB
+MLP epoch 031/50 train_mse=0.181100 val_mse=0.907792 val_r2=-0.1473 peak=0.24GB
+MLP epoch 032/50 train_mse=0.181872 val_mse=0.889178 val_r2=-0.1238 peak=0.24GB
+MLP epoch 033/50 train_mse=0.179628 val_mse=0.911960 val_r2=-0.1526 peak=0.24GB
+MLP epoch 034/50 train_mse=0.176386 val_mse=0.914456 val_r2=-0.1557 peak=0.24GB
+MLP epoch 035/50 train_mse=0.175493 val_mse=0.916653 val_r2=-0.1585 peak=0.24GB
+MLP epoch 036/50 train_mse=0.176884 val_mse=0.899840 val_r2=-0.1373 peak=0.24GB
+MLP epoch 037/50 train_mse=0.175189 val_mse=0.917892 val_r2=-0.1601 peak=0.24GB
+MLP epoch 038/50 train_mse=0.173540 val_mse=0.910431 val_r2=-0.1507 peak=0.24GB
+MLP epoch 039/50 train_mse=0.174279 val_mse=0.923966 val_r2=-0.1678 peak=0.24GB
+MLP epoch 040/50 train_mse=0.209910 val_mse=0.937192 val_r2=-0.1845 peak=0.24GB
+MLP epoch 041/50 train_mse=0.154239 val_mse=0.953283 val_r2=-0.2048 peak=0.24GB
+MLP epoch 042/50 train_mse=0.147758 val_mse=0.973377 val_r2=-0.2302 peak=0.24GB
+MLP epoch 043/50 train_mse=0.144908 val_mse=0.973160 val_r2=-0.2299 peak=0.24GB
+MLP epoch 044/50 train_mse=0.141619 val_mse=0.969759 val_r2=-0.2256 peak=0.24GB
+MLP epoch 045/50 train_mse=0.139654 val_mse=0.976498 val_r2=-0.2342 peak=0.24GB
+MLP epoch 046/50 train_mse=0.138740 val_mse=0.996403 val_r2=-0.2593 peak=0.24GB
+MLP epoch 047/50 train_mse=0.138481 val_mse=0.976342 val_r2=-0.2340 peak=0.24GB
+MLP epoch 048/50 train_mse=0.136828 val_mse=0.993352 val_r2=-0.2555 peak=0.24GB
+MLP epoch 049/50 train_mse=0.136602 val_mse=0.984006 val_r2=-0.2436 peak=0.24GB
+MLP epoch 050/50 train_mse=0.135365 val_mse=0.973000 val_r2=-0.2297 peak=0.24GB
+ResCNN1D epoch 001/50 train_mse=0.469107 val_mse=0.190547 val_r2=0.7592 peak=3.49GB
+ResCNN1D epoch 002/50 train_mse=0.165536 val_mse=0.182192 val_r2=0.7697 peak=3.49GB
+ResCNN1D epoch 003/50 train_mse=0.154272 val_mse=0.164607 val_r2=0.7920 peak=3.49GB
+ResCNN1D epoch 004/50 train_mse=0.144409 val_mse=0.145984 val_r2=0.8155 peak=3.49GB
+ResCNN1D epoch 005/50 train_mse=0.122611 val_mse=0.157774 val_r2=0.8006 peak=3.49GB
+ResCNN1D epoch 006/50 train_mse=0.099210 val_mse=0.158958 val_r2=0.7991 peak=3.49GB
+ResCNN1D epoch 007/50 train_mse=0.084118 val_mse=0.158893 val_r2=0.7992 peak=3.49GB
+ResCNN1D epoch 008/50 train_mse=0.073337 val_mse=0.167350 val_r2=0.7885 peak=3.49GB
+ResCNN1D epoch 009/50 train_mse=0.068013 val_mse=0.170337 val_r2=0.7847 peak=3.49GB
+ResCNN1D epoch 010/50 train_mse=0.061394 val_mse=0.163127 val_r2=0.7938 peak=3.49GB
+ResCNN1D epoch 011/50 train_mse=0.057482 val_mse=0.170577 val_r2=0.7844 peak=3.49GB
+ResCNN1D epoch 012/50 train_mse=0.054084 val_mse=0.166006 val_r2=0.7902 peak=3.49GB
+ResCNN1D epoch 013/50 train_mse=0.051153 val_mse=0.167997 val_r2=0.7877 peak=3.49GB
+ResCNN1D epoch 014/50 train_mse=0.049454 val_mse=0.164724 val_r2=0.7918 peak=3.49GB
+ResCNN1D epoch 015/50 train_mse=0.047721 val_mse=0.177430 val_r2=0.7758 peak=3.49GB
+ResCNN1D epoch 016/50 train_mse=0.045668 val_mse=0.170451 val_r2=0.7846 peak=3.49GB
+ResCNN1D epoch 017/50 train_mse=0.044752 val_mse=0.170601 val_r2=0.7844 peak=3.49GB
+ResCNN1D epoch 018/50 train_mse=0.042553 val_mse=0.169519 val_r2=0.7858 peak=3.49GB
+ResCNN1D epoch 019/50 train_mse=0.041729 val_mse=0.168241 val_r2=0.7874 peak=3.49GB
+ResCNN1D epoch 020/50 train_mse=0.042131 val_mse=0.171258 val_r2=0.7836 peak=3.49GB
+ResCNN1D epoch 021/50 train_mse=0.035612 val_mse=0.168990 val_r2=0.7864 peak=3.49GB
+ResCNN1D epoch 022/50 train_mse=0.033419 val_mse=0.169909 val_r2=0.7853 peak=3.49GB
+ResCNN1D epoch 023/50 train_mse=0.033344 val_mse=0.169691 val_r2=0.7855 peak=3.49GB
+ResCNN1D epoch 024/50 train_mse=0.032839 val_mse=0.170242 val_r2=0.7848 peak=3.49GB
+ResCNN1D epoch 025/50 train_mse=0.032036 val_mse=0.172842 val_r2=0.7816 peak=3.49GB
+ResCNN1D epoch 026/50 train_mse=0.032480 val_mse=0.170758 val_r2=0.7842 peak=3.49GB
+ResCNN1D epoch 027/50 train_mse=0.031908 val_mse=0.170433 val_r2=0.7846 peak=3.49GB
+ResCNN1D epoch 028/50 train_mse=0.031515 val_mse=0.172284 val_r2=0.7823 peak=3.49GB
+ResCNN1D epoch 029/50 train_mse=0.030883 val_mse=0.169765 val_r2=0.7854 peak=3.49GB
+ResCNN1D epoch 030/50 train_mse=0.030457 val_mse=0.175154 val_r2=0.7786 peak=3.49GB
+ResCNN1D epoch 031/50 train_mse=0.030797 val_mse=0.177349 val_r2=0.7759 peak=3.49GB
+ResCNN1D epoch 032/50 train_mse=0.030221 val_mse=0.175644 val_r2=0.7780 peak=3.49GB
+ResCNN1D epoch 033/50 train_mse=0.029770 val_mse=0.169403 val_r2=0.7859 peak=3.49GB
+ResCNN1D epoch 034/50 train_mse=0.029583 val_mse=0.174598 val_r2=0.7793 peak=3.49GB
+ResCNN1D epoch 035/50 train_mse=0.029964 val_mse=0.172436 val_r2=0.7821 peak=3.49GB
+ResCNN1D epoch 036/50 train_mse=0.029723 val_mse=0.172241 val_r2=0.7823 peak=3.49GB
+ResCNN1D epoch 037/50 train_mse=0.029036 val_mse=0.175143 val_r2=0.7786 peak=3.49GB
+ResCNN1D epoch 038/50 train_mse=0.029193 val_mse=0.176859 val_r2=0.7765 peak=3.49GB
+ResCNN1D epoch 039/50 train_mse=0.028940 val_mse=0.173372 val_r2=0.7809 peak=3.49GB
+ResCNN1D epoch 040/50 train_mse=0.028825 val_mse=0.177999 val_r2=0.7750 peak=3.49GB
+ResCNN1D epoch 041/50 train_mse=0.026301 val_mse=0.176767 val_r2=0.7766 peak=3.49GB
+ResCNN1D epoch 042/50 train_mse=0.025657 val_mse=0.173914 val_r2=0.7802 peak=3.49GB
+ResCNN1D epoch 043/50 train_mse=0.025389 val_mse=0.175338 val_r2=0.7784 peak=3.49GB
+ResCNN1D epoch 044/50 train_mse=0.025592 val_mse=0.175954 val_r2=0.7776 peak=3.49GB
+ResCNN1D epoch 045/50 train_mse=0.025600 val_mse=0.175408 val_r2=0.7783 peak=3.49GB
+ResCNN1D epoch 046/50 train_mse=0.024778 val_mse=0.176913 val_r2=0.7764 peak=3.49GB
+ResCNN1D epoch 047/50 train_mse=0.025347 val_mse=0.180888 val_r2=0.7714 peak=3.49GB
+ResCNN1D epoch 048/50 train_mse=0.024782 val_mse=0.176715 val_r2=0.7767 peak=3.49GB
+ResCNN1D epoch 049/50 train_mse=0.025607 val_mse=0.173001 val_r2=0.7813 peak=3.49GB
+ResCNN1D epoch 050/50 train_mse=0.025550 val_mse=0.173975 val_r2=0.7801 peak=3.49GB
+UNet1D epoch 001/50 train_mse=0.239739 val_mse=0.214532 val_r2=0.7289 peak=3.03GB
+UNet1D epoch 002/50 train_mse=0.211533 val_mse=0.226527 val_r2=0.7137 peak=3.03GB
+UNet1D epoch 003/50 train_mse=0.184933 val_mse=0.235629 val_r2=0.7022 peak=3.03GB
+UNet1D epoch 004/50 train_mse=0.147453 val_mse=0.256907 val_r2=0.6753 peak=3.03GB
+UNet1D epoch 005/50 train_mse=0.121418 val_mse=0.266700 val_r2=0.6629 peak=3.03GB
+UNet1D epoch 006/50 train_mse=0.106983 val_mse=0.279152 val_r2=0.6472 peak=3.03GB
+UNet1D epoch 007/50 train_mse=0.095196 val_mse=0.291865 val_r2=0.6311 peak=3.03GB
+UNet1D epoch 008/50 train_mse=0.088660 val_mse=0.274253 val_r2=0.6534 peak=3.03GB
+UNet1D epoch 009/50 train_mse=0.081545 val_mse=0.284403 val_r2=0.6406 peak=3.03GB
+UNet1D epoch 010/50 train_mse=0.077134 val_mse=0.284246 val_r2=0.6408 peak=3.03GB
+UNet1D epoch 011/50 train_mse=0.072682 val_mse=0.278728 val_r2=0.6477 peak=3.03GB
+UNet1D epoch 012/50 train_mse=0.070811 val_mse=0.276402 val_r2=0.6507 peak=3.03GB
+UNet1D epoch 013/50 train_mse=0.067207 val_mse=0.293130 val_r2=0.6295 peak=3.03GB
+UNet1D epoch 014/50 train_mse=0.065773 val_mse=0.274649 val_r2=0.6529 peak=3.03GB
+UNet1D epoch 015/50 train_mse=0.063175 val_mse=0.282725 val_r2=0.6427 peak=3.03GB
+UNet1D epoch 016/50 train_mse=0.061212 val_mse=0.281930 val_r2=0.6437 peak=3.03GB
+UNet1D epoch 017/50 train_mse=0.059299 val_mse=0.282583 val_r2=0.6429 peak=3.03GB
+UNet1D epoch 018/50 train_mse=0.057764 val_mse=0.292070 val_r2=0.6309 peak=3.03GB
+UNet1D epoch 019/50 train_mse=0.056586 val_mse=0.282744 val_r2=0.6426 peak=3.03GB
+UNet1D epoch 020/50 train_mse=0.055826 val_mse=0.286120 val_r2=0.6384 peak=3.03GB
+UNet1D epoch 021/50 train_mse=0.047203 val_mse=0.285734 val_r2=0.6389 peak=3.03GB
+UNet1D epoch 022/50 train_mse=0.045597 val_mse=0.284260 val_r2=0.6407 peak=3.03GB
+UNet1D epoch 023/50 train_mse=0.044986 val_mse=0.288718 val_r2=0.6351 peak=3.03GB
+UNet1D epoch 024/50 train_mse=0.044418 val_mse=0.286513 val_r2=0.6379 peak=3.03GB
+UNet1D epoch 025/50 train_mse=0.043806 val_mse=0.287607 val_r2=0.6365 peak=3.03GB
+UNet1D epoch 026/50 train_mse=0.042108 val_mse=0.296251 val_r2=0.6256 peak=3.03GB
+UNet1D epoch 027/50 train_mse=0.043165 val_mse=0.294002 val_r2=0.6284 peak=3.03GB
+UNet1D epoch 028/50 train_mse=0.042051 val_mse=0.291547 val_r2=0.6315 peak=3.03GB
+UNet1D epoch 029/50 train_mse=0.041110 val_mse=0.291211 val_r2=0.6319 peak=3.03GB
+UNet1D epoch 030/50 train_mse=0.040932 val_mse=0.287756 val_r2=0.6363 peak=3.03GB
+UNet1D epoch 031/50 train_mse=0.041140 val_mse=0.292280 val_r2=0.6306 peak=3.03GB
+UNet1D epoch 032/50 train_mse=0.040928 val_mse=0.288403 val_r2=0.6355 peak=3.03GB
+UNet1D epoch 033/50 train_mse=0.040636 val_mse=0.286773 val_r2=0.6376 peak=3.03GB
+UNet1D epoch 034/50 train_mse=0.040370 val_mse=0.293447 val_r2=0.6291 peak=3.03GB
+UNet1D epoch 035/50 train_mse=0.039826 val_mse=0.296696 val_r2=0.6250 peak=3.03GB
+UNet1D epoch 036/50 train_mse=0.039400 val_mse=0.288487 val_r2=0.6354 peak=3.03GB
+UNet1D epoch 037/50 train_mse=0.039199 val_mse=0.290246 val_r2=0.6332 peak=3.03GB
+UNet1D epoch 038/50 train_mse=0.038444 val_mse=0.295849 val_r2=0.6261 peak=3.03GB
+UNet1D epoch 039/50 train_mse=0.039511 val_mse=0.292950 val_r2=0.6298 peak=3.03GB
+UNet1D epoch 040/50 train_mse=0.038714 val_mse=0.289714 val_r2=0.6338 peak=3.03GB
+UNet1D epoch 041/50 train_mse=0.033645 val_mse=0.298020 val_r2=0.6233 peak=3.03GB
+UNet1D epoch 042/50 train_mse=0.033762 val_mse=0.300035 val_r2=0.6208 peak=3.03GB
+UNet1D epoch 043/50 train_mse=0.033414 val_mse=0.296293 val_r2=0.6255 peak=3.03GB
+UNet1D epoch 044/50 train_mse=0.033434 val_mse=0.296607 val_r2=0.6251 peak=3.03GB
+UNet1D epoch 045/50 train_mse=0.032771 val_mse=0.301882 val_r2=0.6185 peak=3.03GB
+UNet1D epoch 046/50 train_mse=0.032803 val_mse=0.291851 val_r2=0.6311 peak=3.03GB
+UNet1D epoch 047/50 train_mse=0.032558 val_mse=0.299289 val_r2=0.6217 peak=3.03GB
+UNet1D epoch 048/50 train_mse=0.033069 val_mse=0.295221 val_r2=0.6269 peak=3.03GB
+UNet1D epoch 049/50 train_mse=0.031793 val_mse=0.298443 val_r2=0.6228 peak=3.03GB
+UNet1D epoch 050/50 train_mse=0.032373 val_mse=0.295937 val_r2=0.6260 peak=3.03GB
+Transformer epoch 001/50 train_mse=0.795168 val_mse=0.791456 val_r2=-0.0003 peak=2.51GB
+Transformer epoch 002/50 train_mse=0.791684 val_mse=0.791334 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 003/50 train_mse=0.791530 val_mse=0.791342 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 004/50 train_mse=0.791510 val_mse=0.791307 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 005/50 train_mse=0.791514 val_mse=0.791262 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 006/50 train_mse=0.791497 val_mse=0.791689 val_r2=-0.0006 peak=2.51GB
+Transformer epoch 007/50 train_mse=0.791481 val_mse=0.791814 val_r2=-0.0007 peak=2.51GB
+Transformer epoch 008/50 train_mse=0.791476 val_mse=0.791300 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 009/50 train_mse=0.791474 val_mse=0.791241 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 010/50 train_mse=0.791470 val_mse=0.791250 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 011/50 train_mse=0.791470 val_mse=0.791283 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 012/50 train_mse=0.791475 val_mse=0.791284 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 013/50 train_mse=0.791466 val_mse=0.791243 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 014/50 train_mse=0.791475 val_mse=0.791225 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 015/50 train_mse=0.791479 val_mse=0.791386 val_r2=-0.0002 peak=2.51GB
+Transformer epoch 016/50 train_mse=0.791474 val_mse=0.791352 val_r2=-0.0002 peak=2.51GB
+Transformer epoch 017/50 train_mse=0.791471 val_mse=0.791276 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 018/50 train_mse=0.791466 val_mse=0.791266 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 019/50 train_mse=0.791469 val_mse=0.791297 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 020/50 train_mse=0.791466 val_mse=0.791262 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 021/50 train_mse=0.791459 val_mse=0.791277 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 022/50 train_mse=0.791459 val_mse=0.791232 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 023/50 train_mse=0.791461 val_mse=0.791250 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 024/50 train_mse=0.791460 val_mse=0.791279 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 025/50 train_mse=0.791459 val_mse=0.791242 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 026/50 train_mse=0.791459 val_mse=0.791268 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 027/50 train_mse=0.791459 val_mse=0.791246 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 028/50 train_mse=0.791459 val_mse=0.791233 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 029/50 train_mse=0.791460 val_mse=0.791261 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 030/50 train_mse=0.791460 val_mse=0.791228 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 031/50 train_mse=0.791458 val_mse=0.791255 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 032/50 train_mse=0.791461 val_mse=0.791255 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 033/50 train_mse=0.791460 val_mse=0.791234 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 034/50 train_mse=0.791462 val_mse=0.791265 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 035/50 train_mse=0.791460 val_mse=0.791239 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 036/50 train_mse=0.791460 val_mse=0.791254 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 037/50 train_mse=0.791459 val_mse=0.791247 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 038/50 train_mse=0.791459 val_mse=0.791251 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 039/50 train_mse=0.791461 val_mse=0.791262 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 040/50 train_mse=0.791461 val_mse=0.791262 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 041/50 train_mse=0.791455 val_mse=0.791249 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 042/50 train_mse=0.791455 val_mse=0.791254 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 043/50 train_mse=0.791455 val_mse=0.791297 val_r2=-0.0001 peak=2.51GB
+Transformer epoch 044/50 train_mse=0.791456 val_mse=0.791259 val_r2=-0.0000 peak=2.51GB
+Transformer epoch 045/50 train_mse=0.791456 val_mse=0.791236 val_r2=-0.0000 peak=2.51GB

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/logs/o1_gpu1_transformer_rerun_20260604_224843.log

@@ -0,0 +1,56 @@
+/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/scripts/baseline_models.py:225: UserWarning: enable_nested_tensor is True, but self.use_nested_tensor is False because encoder_layer.norm_first was True
+  self.encoder = nn.TransformerEncoder(layer, num_layers=n_layers)
+Train samples=136,800, Val=34,200, Test=27,018
+GM path: /data/home/jason/data/SesimicTransformerData/MDOF/All_GMs/GMs_knet_3474_AF_57.h5
+Building dir: /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno
+Transformer epoch 001/50 train_mse=0.487353 val_mse=0.354366 val_r2=0.5521 peak=2.54GB
+Transformer epoch 002/50 train_mse=0.203208 val_mse=0.077400 val_r2=0.9022 peak=2.54GB
+Transformer epoch 003/50 train_mse=0.088314 val_mse=0.060544 val_r2=0.9235 peak=2.54GB
+Transformer epoch 004/50 train_mse=0.068935 val_mse=0.049628 val_r2=0.9373 peak=2.54GB
+Transformer epoch 005/50 train_mse=0.060136 val_mse=0.050520 val_r2=0.9362 peak=2.54GB
+Transformer epoch 006/50 train_mse=0.053788 val_mse=0.043385 val_r2=0.9452 peak=2.54GB
+Transformer epoch 007/50 train_mse=0.050089 val_mse=0.040144 val_r2=0.9493 peak=2.54GB
+Transformer epoch 008/50 train_mse=0.048317 val_mse=0.037791 val_r2=0.9522 peak=2.54GB
+Transformer epoch 009/50 train_mse=0.045371 val_mse=0.036166 val_r2=0.9543 peak=2.54GB
+Transformer epoch 010/50 train_mse=0.042626 val_mse=0.034969 val_r2=0.9558 peak=2.54GB
+Transformer epoch 011/50 train_mse=0.041674 val_mse=0.033746 val_r2=0.9574 peak=2.54GB
+Transformer epoch 012/50 train_mse=0.039482 val_mse=0.032673 val_r2=0.9587 peak=2.54GB
+Transformer epoch 013/50 train_mse=0.038545 val_mse=0.034781 val_r2=0.9560 peak=2.54GB
+Transformer epoch 014/50 train_mse=0.037823 val_mse=0.032284 val_r2=0.9592 peak=2.54GB
+Transformer epoch 015/50 train_mse=0.036089 val_mse=0.034009 val_r2=0.9570 peak=2.54GB
+Transformer epoch 016/50 train_mse=0.035096 val_mse=0.031875 val_r2=0.9597 peak=2.54GB
+Transformer epoch 017/50 train_mse=0.034933 val_mse=0.030239 val_r2=0.9618 peak=2.54GB
+Transformer epoch 018/50 train_mse=0.033907 val_mse=0.031289 val_r2=0.9605 peak=2.54GB
+Transformer epoch 019/50 train_mse=0.033246 val_mse=0.030087 val_r2=0.9620 peak=2.54GB
+Transformer epoch 020/50 train_mse=0.032652 val_mse=0.027238 val_r2=0.9656 peak=2.54GB
+Transformer epoch 021/50 train_mse=0.027630 val_mse=0.026280 val_r2=0.9668 peak=2.54GB
+Transformer epoch 022/50 train_mse=0.027051 val_mse=0.025751 val_r2=0.9675 peak=2.54GB
+Transformer epoch 023/50 train_mse=0.026689 val_mse=0.025408 val_r2=0.9679 peak=2.54GB
+Transformer epoch 024/50 train_mse=0.026319 val_mse=0.025002 val_r2=0.9684 peak=2.54GB
+Transformer epoch 025/50 train_mse=0.026028 val_mse=0.024969 val_r2=0.9684 peak=2.54GB
+Transformer epoch 026/50 train_mse=0.025801 val_mse=0.025496 val_r2=0.9678 peak=2.54GB
+Transformer epoch 027/50 train_mse=0.025596 val_mse=0.025054 val_r2=0.9683 peak=2.54GB
+Transformer epoch 028/50 train_mse=0.025345 val_mse=0.024592 val_r2=0.9689 peak=2.54GB
+Transformer epoch 029/50 train_mse=0.025148 val_mse=0.025787 val_r2=0.9674 peak=2.54GB
+Transformer epoch 030/50 train_mse=0.025032 val_mse=0.025567 val_r2=0.9677 peak=2.54GB
+Transformer epoch 031/50 train_mse=0.024820 val_mse=0.023949 val_r2=0.9697 peak=2.54GB
+Transformer epoch 032/50 train_mse=0.024560 val_mse=0.024570 val_r2=0.9689 peak=2.54GB
+Transformer epoch 033/50 train_mse=0.024384 val_mse=0.024887 val_r2=0.9685 peak=2.54GB
+Transformer epoch 034/50 train_mse=0.024215 val_mse=0.024489 val_r2=0.9690 peak=2.54GB
+Transformer epoch 035/50 train_mse=0.024170 val_mse=0.023148 val_r2=0.9707 peak=2.54GB
+Transformer epoch 036/50 train_mse=0.023753 val_mse=0.024109 val_r2=0.9695 peak=2.54GB
+Transformer epoch 037/50 train_mse=0.023686 val_mse=0.024587 val_r2=0.9689 peak=2.54GB
+Transformer epoch 038/50 train_mse=0.023533 val_mse=0.023223 val_r2=0.9706 peak=2.54GB
+Transformer epoch 039/50 train_mse=0.023366 val_mse=0.024622 val_r2=0.9689 peak=2.54GB
+Transformer epoch 040/50 train_mse=0.023201 val_mse=0.023320 val_r2=0.9705 peak=2.54GB
+Transformer epoch 041/50 train_mse=0.021118 val_mse=0.021924 val_r2=0.9723 peak=2.54GB
+Transformer epoch 042/50 train_mse=0.020656 val_mse=0.022196 val_r2=0.9719 peak=2.54GB
+Transformer epoch 043/50 train_mse=0.020601 val_mse=0.022267 val_r2=0.9719 peak=2.54GB
+Transformer epoch 044/50 train_mse=0.020526 val_mse=0.022070 val_r2=0.9721 peak=2.54GB
+Transformer epoch 045/50 train_mse=0.020385 val_mse=0.021797 val_r2=0.9725 peak=2.54GB
+Transformer epoch 046/50 train_mse=0.020203 val_mse=0.022567 val_r2=0.9715 peak=2.54GB
+Transformer epoch 047/50 train_mse=0.020253 val_mse=0.022116 val_r2=0.9720 peak=2.54GB
+Transformer epoch 048/50 train_mse=0.020073 val_mse=0.021801 val_r2=0.9724 peak=2.54GB
+Transformer epoch 049/50 train_mse=0.019995 val_mse=0.022298 val_r2=0.9718 peak=2.54GB
+Transformer epoch 050/50 train_mse=0.020025 val_mse=0.022364 val_r2=0.9717 peak=2.54GB
+Training summary saved to /data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/training_summary.json

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/monitors/o1_status_monitor.log

@@ -0,0 +1,24 @@
+===== 2026-06-04 16:47:41 HKT =====
+[BiLSTM training_log tail]
+2,0.07820247465434828,0.2796470537201282,0.1557729295875594,0.9011912377009019,0.06474816788879752,0.254456612979104,0.1407365133292215,0.9181671082534892,0.05326596666151942,0.0003,393.2450256347656,552.4883005619049,15.61366891860962
+3,0.062326362649720544,0.24965248376437305,0.13602147245827242,0.921250692247271,0.04681216217933343,0.21636118454873884,0.10932230883741471,0.9408357838538273,0.037439060737486575,0.0003,130.40016174316406,551.8074929714203,15.61366891860962
+4,0.04088781967906692,0.20220736801379646,0.10259190409699379,0.9483382735082833,0.030837602496230812,0.17560638512375001,0.08693451443618269,0.9610254580310279,0.034270109928641344,0.0003,44.51044464111328,552.1496667861938,15.61366891860962
+BiLSTM nonfinite_report: none
+[O1 baseline processes]
+2670839 2670837    06:16:03  264  0.5 /data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/train_baselines.py --device cuda --batch_size 64 --epochs 50 --models MLP ResCNN1D UNet1D Transformer --base_data_dir /data/home/jason/data/SesimicTransformerData --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno --no_progress --log_interval 100
+2975011 2975009       40:17  171  1.2 /data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/train_baselines.py --device cuda --batch_size 64 --epochs 50 --models BiLSTM --base_data_dir /data/home/jason/data/SesimicTransformerData --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno --no_progress --log_interval 100
+[GPU summary]
+0, NVIDIA RTX PRO 6000 Blackwell Workstation Edition, 98 %, 20461 MiB, 97887 MiB, 80, 402.08 W
+1, NVIDIA RTX PRO 6000 Blackwell Workstation Edition, 56 %, 4859 MiB, 97887 MiB, 78, 465.30 W
+[Recent O1 run files]
+2026-06-04 10:34 113345561 6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt
+2026-06-04 12:06 11293 6-code/O1-baseline-comparison/results/runs/MLP/training_log.csv
+2026-06-04 12:06 408 6-code/O1-baseline-comparison/results/runs/MLP/summary.json
+2026-06-04 12:36 88398415 6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt
+2026-06-04 13:11 100940852 6-code/O1-baseline-comparison/results/runs/BiLSTM_nan_20260604_160510/checkpoints/BiLSTM_best.pt
+2026-06-04 15:53 5697 6-code/O1-baseline-comparison/results/runs/BiLSTM_nan_20260604_160510/training_log.csv
+2026-06-04 16:05 0 6-code/O1-baseline-comparison/results/runs/BiLSTM_starttest_20260604_160724/training_log.csv
+2026-06-04 16:44 100940980 6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt
+2026-06-04 16:44 1104 6-code/O1-baseline-comparison/results/runs/BiLSTM/training_log.csv
+2026-06-04 16:45 8374 6-code/O1-baseline-comparison/results/runs/ResCNN1D/training_log.csv
+

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/monitors/o1_status_monitor.sh

@@ -0,0 +1,29 @@
+#!/usr/bin/env bash
+set -uo pipefail
+cd /data/home/jason/Seismic-FNO
+OUT="6-code/O1-baseline-comparison/results/monitors/o1_status_monitor.log"
+while true; do
+  {
+    echo "===== $(date '+%Y-%m-%d %H:%M:%S %Z') ====="
+    if [[ -f 6-code/O1-baseline-comparison/results/runs/BiLSTM/training_log.csv ]]; then
+      echo "[BiLSTM training_log tail]"
+      tail -n 3 6-code/O1-baseline-comparison/results/runs/BiLSTM/training_log.csv
+    else
+      echo "BiLSTM training_log.csv not found"
+    fi
+    if [[ -f 6-code/O1-baseline-comparison/results/runs/BiLSTM/nonfinite_report.json ]]; then
+      echo "[BiLSTM nonfinite_report]"
+      cat 6-code/O1-baseline-comparison/results/runs/BiLSTM/nonfinite_report.json
+    else
+      echo "BiLSTM nonfinite_report: none"
+    fi
+    echo "[O1 baseline processes]"
+    ps -eo pid,ppid,etime,pcpu,pmem,args --sort=-pcpu | grep -E 'O1-baseline-comparison/scripts/(train_baselines|train_fno_large_bs64|evaluate_baselines)\.py' | grep -v grep || true
+    echo "[GPU summary]"
+    nvidia-smi --query-gpu=index,name,utilization.gpu,memory.used,memory.total,temperature.gpu,power.draw --format=csv,noheader || true
+    echo "[Recent O1 run files]"
+    find 6-code/O1-baseline-comparison/results/runs -maxdepth 3 -type f \( -name 'training_log.csv' -o -name 'summary.json' -o -name '*_best.pt' -o -name 'nonfinite_report.json' \) -printf '%TY-%Tm-%Td %TH:%TM %s %p\n' | sort || true
+    echo
+  } >> "$OUT" 2>&1
+  sleep 7200
+done

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/report/O1_baseline_report.md

@@ -0,0 +1,43 @@
+# O1 Baseline Comparison Report
+
+**Protocol.** Full-data O1 comparison on the canonical fixed split: 2,400 train GMs, 600 validation GMs, and 474 held-out test GMs. All models use batch size 64 and 50 epochs. Metrics below are from the fixed 474-GM test set.
+
+## Executive Summary
+
+- BiLSTM achieved the lowest test MSE (0.0099) and highest R2 (0.9872), but required the longest training time (7.51 h) and largest memory footprint (15.61 GB).
+- The stabilized pre-norm Transformer ranked second in MSE (0.0214) with substantially lower memory than BiLSTM (2.54 GB).
+- FNO-Large was resource efficient but underperformed BiLSTM/Transformer on this O1 fixed 474-GM test split (MSE 0.0449).
+- The original Transformer run collapsed to a mean-prediction regime and was excluded after replacing it with a standard pre-norm Transformer and a lower learning rate.
+
+## Main Figures
+
+![Fixed 474-GM Test Accuracy](figures/fig1_test_accuracy.png)
+
+![Accuracy/resource tradeoffs](figures/fig2_tradeoff_bubbles.png)
+
+![Validation MSE curves](figures/fig3_training_curves.png)
+
+![Normalized accuracy/resource profile](figures/fig4_normalized_profile.png)
+
+![Result table](figures/fig5_results_table.png)
+
+## Test Results Table
+
+|Model|MSE ↓|R2 ↑|PFA RMSE ↓|Spec ↓|High-f ↓|Train h ↓|GPU GB ↓|Latency ms ↓|
+|---|---|---|---|---|---|---|---|---|
+|BiLSTM|0.0099|0.9872|0.0182|0.185|0.222|7.51|15.61|1.049|
+|Transformer|0.0214|0.9725|0.0231|0.325|0.391|2.34|2.54|0.106|
+|FNO-Large|0.0449|0.9423|0.0330|0.342|0.398|2.16|3.30|0.108|
+|ResCNN1D|0.1481|0.8096|0.0547|0.484|0.534|6.29|3.49|0.607|
+|UNet1D|0.2157|0.7227|0.1063|0.588|0.633|2.27|3.03|0.096|
+|MLP|0.7420|0.0462|0.2464|1.255|1.091|1.59|0.24|0.075|
+
+## Interpretation
+
+BiLSTM is the strongest model in raw test accuracy, but this result comes with a clear compute cost: it used roughly 15.6 GB peak memory and 7.5 hours of training. Transformer is a strong second after pre-norm stabilization, with much better memory efficiency. FNO-Large remains competitive and efficient, but it does not dominate this particular full-data fixed-test protocol. This should be reported transparently, with additional OOD/classical-GM analysis used to assess whether FNO has advantages outside the i.i.d. split.
+
+## Artifacts
+
+- Metrics CSV: `../compiled/table4_metrics.csv`
+- Training summary: `../compiled/training_runs.csv`
+- Plotting summary: `o1_report_summary.csv`

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/report/O1_baseline_report.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fa741988ea251482c251f731f5f05946585f935ac9f6cdcbff66d0f74b7f5b2
+size 712304

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/report/O1_baseline_report.tex

@@ -0,0 +1,65 @@
+
+\documentclass[10pt]{article}
+\usepackage[margin=0.75in]{geometry}
+\usepackage{graphicx}
+\usepackage{booktabs}
+\usepackage{caption}
+\usepackage{xcolor}
+\usepackage{hyperref}
+\usepackage{amsmath}
+\title{O1 Baseline Comparison on the Fixed 474-GM Test Set}
+\author{Seismic-FNO Internal Report}
+\date{June 2026}
+\begin{document}
+\maketitle
+\section*{Protocol}
+Full-data O1 comparison on the canonical fixed split: 2,400 train ground motions, 600 validation ground motions, and 474 held-out test ground motions. All models use batch size 64 and 50 epochs. Reported metrics are measured on the fixed 474-GM test set.
+\section*{Executive Summary}
+\begin{itemize}
+  \item BiLSTM achieved the lowest test MSE (0.0099) and highest R2 (0.9872), but required the longest training time (7.51 h) and largest memory footprint (15.61 GB).
+  \item The stabilized pre-norm Transformer ranked second in MSE (0.0214) with substantially lower memory than BiLSTM (2.54 GB).
+  \item FNO-Large was resource efficient but underperformed BiLSTM/Transformer on this O1 fixed 474-GM test split (MSE 0.0449).
+  \item The original Transformer run collapsed to a mean-prediction regime and was excluded after replacing it with a standard pre-norm Transformer and a lower learning rate.
+
+\end{itemize}
+\section*{Quantitative Results}
+\begin{table}[h]
+\centering
+\small
+\caption{Fixed 474-GM test accuracy and resource costs.}
+\begin{tabular}{lrrrrrrrr}
+\toprule
+Model & MSE $\downarrow$ & $R^2$ $\uparrow$ & PFA RMSE $\downarrow$ & Spec $\downarrow$ & High-f $\downarrow$ & Train h $\downarrow$ & GPU GB $\downarrow$ & Lat. ms $\downarrow$ \\
+\midrule
+BiLSTM & 0.0099 & 0.9872 & 0.0182 & 0.185 & 0.222 & 7.51 & 15.61 & 1.049 \\
+Transformer & 0.0214 & 0.9725 & 0.0231 & 0.325 & 0.391 & 2.34 & 2.54 & 0.106 \\
+FNO-Large & 0.0449 & 0.9423 & 0.0330 & 0.342 & 0.398 & 2.16 & 3.30 & 0.108 \\
+ResCNN1D & 0.1481 & 0.8096 & 0.0547 & 0.484 & 0.534 & 6.29 & 3.49 & 0.607 \\
+UNet1D & 0.2157 & 0.7227 & 0.1063 & 0.588 & 0.633 & 2.27 & 3.03 & 0.096 \\
+MLP & 0.7420 & 0.0462 & 0.2464 & 1.255 & 1.091 & 1.59 & 0.24 & 0.075 \\
+\bottomrule
+\end{tabular}
+\end{table}
+\begin{figure}[h]
+\centering
+\includegraphics[width=0.98\linewidth]{figures/fig1_test_accuracy.pdf}
+\caption{Main test-set accuracy metrics.}
+\end{figure}
+\begin{figure}[h]
+\centering
+\includegraphics[width=0.98\linewidth]{figures/fig2_tradeoff_bubbles.pdf}
+\caption{Accuracy-resource tradeoffs. Bubble area scales with parameter count.}
+\end{figure}
+\begin{figure}[h]
+\centering
+\includegraphics[width=0.98\linewidth]{figures/fig3_training_curves.pdf}
+\caption{Validation MSE curves with final test MSE markers.}
+\end{figure}
+\begin{figure}[h]
+\centering
+\includegraphics[width=0.98\linewidth]{figures/fig4_normalized_profile.pdf}
+\caption{Normalized error and cost profile. Lower is better for all columns.}
+\end{figure}
+\section*{Interpretation}
+BiLSTM is the strongest model in raw test accuracy, but it is also the most expensive model to train and deploy in memory. The stabilized pre-norm Transformer is a strong second and provides a better memory/latency tradeoff. FNO-Large remains competitive and resource efficient, but it does not dominate this particular in-distribution fixed-test protocol. OOD and classical-GM experiments should be used to assess whether FNO has advantages outside the i.i.d. split.
+\end{document}

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/report/o1_report_summary.csv

@@ -0,0 +1,7 @@
+model,checkpoint,numel,numel_m,real_params,real_params_m,trainable,mse,rmse,mae,r2,pfa_mse_g2,pfa_rmse_g,pfa_mae_g,pfa_r2,loss,spectral_log_mse,high_freq_log_mse,latency_ms_per_sample,best_val_mse,train_time_h,peak_gpu_gb,params_m,batch_size,epochs,rank_mse,rank_r2,rank_pfa,mse_per_hour,r2_per_gb,generalization_gap
+BiLSTM,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt,8410113,8.410113,8410113,8.410113,8410113,0.0099188614842287,0.0995934811331983,0.0432213771548134,0.9872496731015952,0.0003300793328829,0.0181680855591055,0.0108249000931961,0.987565034951701,0.0099240713056515,0.1852839740596482,0.2222294335740114,1.0494697839021685,0.0101689182879353,7.507228293948703,15.61366891860962,8.410113,64,50,1,1,1,0.0013212414883165234,0.06322983267084087,-0.00025005680370659977
+Transformer,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/Transformer/checkpoints/Transformer_best.pt,7957775,7.957775,7957775,7.957775,7957775,0.0213903603177851,0.1462544369165773,0.0660383013369173,0.9725034887360728,0.000533425337507,0.023096002630479,0.0149489011169053,0.9799044509395886,0.0214709252808199,0.3250373078083033,0.3907267726411774,0.1057809218764305,0.0217971895928038,2.341890460981263,2.5364460945129395,7.957775,64,50,2,2,2,0.009133800523198869,0.3834118496899567,-0.0004068292750186997
+FNO-Large,6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt,8505793,8.505793,16927169,16.927169,8505793,0.0448808919490714,0.2118511079722536,0.1050106563313501,0.9423072854931448,0.0010877546760695,0.0329811260582404,0.0204876376139049,0.9590213926454952,0.0449616025107302,0.3416421673399337,0.3977926703209573,0.1078980043530464,0.0439618794497365,2.162323934502072,3.2975897789001465,8.505793,64,50,3,3,3,0.020755859579108957,0.28575637015936983,0.0009190124993348966
+ResCNN1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/ResCNN1D/checkpoints/ResCNN1D_best.pt,7355137,7.355137,7355137,7.355137,7355137,0.1480968733036,0.3848335657184805,0.2179850999338596,0.8096269869021784,0.0029878849120218,0.0546615487524994,0.0411752145391326,0.8874384405566411,0.1482058847961555,0.4840071575438722,0.5338901208788509,0.6065387278795242,0.1459837291314495,6.285682277083397,3.494046688079834,7.355137,64,50,4,4,4,0.023560986186581173,0.23171613294815813,0.0021131441721504995
+UNet1D,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/UNet1D/checkpoints/UNet1D_best.pt,7848321,7.848321,7848321,7.848321,7848321,0.2157477369024913,0.4644865303778908,0.2717088732294789,0.7226643221632085,0.0113087215985683,0.1063424731636818,0.0867897865332614,0.5739704252583557,0.2157875346922564,0.5880364352577403,0.6328961368359572,0.0958140939474105,0.2145316383596796,2.265448966821035,3.032552719116211,7.848321,64,50,5,5,5,0.09523398675593951,0.23830231131936183,0.001216098542811711
+MLP,/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/MLP/checkpoints/MLP_best.pt,9444352,9.444352,9444352,9.444352,9444352,0.7419893067526742,0.8613880117302969,0.5303239485549371,0.0462003899076466,0.0607077825787736,0.2463894936452722,0.2106523828826267,-1.2870233889936116,0.7421306194143092,1.255192926166751,1.0909596579435588,0.0749243423342704,0.7341581120553119,1.5860079570611318,0.2412929534912109,9.444352,64,50,6,6,6,0.467834542348437,0.19147011646707474,0.007831194697362265

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/run_scripts/run_evaluate_after_training.sh

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+set -euo pipefail
+cd /data/home/jason/Seismic-FNO
+export PYTHONDONTWRITEBYTECODE=1
+GPU0_PID=$(cat 6-code/O1-baseline-comparison/results/pids/o1_gpu0_bilstm.pid)
+GPU1_PID=$(cat 6-code/O1-baseline-comparison/results/pids/o1_gpu1_baselines_then_fno.pid)
+while kill -0 "$GPU0_PID" 2>/dev/null || kill -0 "$GPU1_PID" 2>/dev/null; do
+  sleep 300
+done
+export CUDA_VISIBLE_DEVICES=1
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/evaluate_baselines.py \
+  --device cuda \
+  --batch_size 64 \
+  --base_data_dir /data/home/jason/data/SesimicTransformerData \
+  --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno \
+  --fno_checkpoint 6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/collect_o1_results.py

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/run_scripts/run_fno_then_evaluate_after_reruns.sh

@@ -0,0 +1,32 @@
+#!/usr/bin/env bash
+set -euo pipefail
+cd /data/home/jason/Seismic-FNO
+export PYTHONDONTWRITEBYTECODE=1
+
+BILSTM_PID=$(cat 6-code/O1-baseline-comparison/results/pids/o1_gpu0_bilstm_rerun.pid)
+TRANSFORMER_PID=$(cat 6-code/O1-baseline-comparison/results/pids/o1_gpu1_transformer_rerun.pid)
+
+while kill -0 "$BILSTM_PID" 2>/dev/null || kill -0 "$TRANSFORMER_PID" 2>/dev/null; do
+  sleep 300
+done
+
+export CUDA_VISIBLE_DEVICES=1
+FNO_CKPT=6-code/O1-baseline-comparison/results/runs/FNO-Large-bs64/checkpoints/FNO-Large-bs64_best.pt
+if [[ ! -f "$FNO_CKPT" ]]; then
+  /data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/train_fno_large_bs64.py \
+    --device cuda \
+    --batch_size 64 \
+    --epochs 50 \
+    --base_data_dir /data/home/jason/data/SesimicTransformerData \
+    --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno \
+    --no_progress \
+    --log_interval 100
+fi
+
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/evaluate_baselines.py \
+  --device cuda \
+  --batch_size 64 \
+  --base_data_dir /data/home/jason/data/SesimicTransformerData \
+  --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno \
+  --fno_checkpoint "$FNO_CKPT"
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/collect_o1_results.py

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/run_scripts/run_gpu0_bilstm.sh

@@ -0,0 +1,14 @@
+#!/usr/bin/env bash
+set -euo pipefail
+cd /data/home/jason/Seismic-FNO
+export PYTHONDONTWRITEBYTECODE=1
+export CUDA_VISIBLE_DEVICES=0
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/train_baselines.py \
+  --device cuda \
+  --batch_size 64 \
+  --epochs 50 \
+  --models BiLSTM \
+  --base_data_dir /data/home/jason/data/SesimicTransformerData \
+  --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno \
+  --no_progress \
+  --log_interval 100

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/run_scripts/run_gpu1_baselines_then_fno.sh

@@ -0,0 +1,22 @@
+#!/usr/bin/env bash
+set -euo pipefail
+cd /data/home/jason/Seismic-FNO
+export PYTHONDONTWRITEBYTECODE=1
+export CUDA_VISIBLE_DEVICES=1
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/train_baselines.py \
+  --device cuda \
+  --batch_size 64 \
+  --epochs 50 \
+  --models MLP ResCNN1D UNet1D Transformer \
+  --base_data_dir /data/home/jason/data/SesimicTransformerData \
+  --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno \
+  --no_progress \
+  --log_interval 100
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/train_fno_large_bs64.py \
+  --device cuda \
+  --batch_size 64 \
+  --epochs 50 \
+  --base_data_dir /data/home/jason/data/SesimicTransformerData \
+  --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno \
+  --no_progress \
+  --log_interval 100

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/run_scripts/run_gpu1_transformer_rerun.sh

@@ -0,0 +1,16 @@
+#!/usr/bin/env bash
+set -euo pipefail
+cd /data/home/jason/Seismic-FNO
+export PYTHONDONTWRITEBYTECODE=1
+export CUDA_VISIBLE_DEVICES=1
+/data/home/jason/miniconda3/envs/sefno/bin/python 6-code/O1-baseline-comparison/scripts/train_baselines.py \
+  --device cuda \
+  --batch_size 64 \
+  --epochs 50 \
+  --models Transformer \
+  --transformer_lr 0.0003 \
+  --grad_clip 1.0 \
+  --base_data_dir /data/home/jason/data/SesimicTransformerData \
+  --building_dir /data/home/jason/data/SesimicTransformerData/MDOF/knet-250/Data/fno \
+  --no_progress \
+  --log_interval 100

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/runs/BiLSTM/checkpoints/BiLSTM_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4de6f5eb8d10c37f688d085fd4201a9b53c8abd5ce5417d52bb07c8f0cedcfbe
+size 100940980

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/runs/BiLSTM/config.json

@@ -0,0 +1,21 @@
+{
+  "model": "BiLSTM",
+  "seed": 42,
+  "epochs": 50,
+  "batch_size": 64,
+  "lr": 0.0003,
+  "base_lr": 0.001,
+  "bilstm_lr": 0.0003,
+  "weight_decay": 0.0001,
+  "scheduler_step": 20,
+  "scheduler_gamma": 0.5,
+  "grad_clip": 1.0,
+  "split": "fixed_474_test_full_57_factors",
+  "params": {
+    "numel": 8410113,
+    "numel_m": 8.410113,
+    "real_params": 8410113,
+    "real_params_m": 8.410113,
+    "trainable": 8410113
+  }
+}

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/runs/BiLSTM/summary.json

@@ -0,0 +1,14 @@
+{
+  "model": "BiLSTM",
+  "best_val_mse": 0.010168918287935313,
+  "train_time_s": 27026.021858215332,
+  "peak_gpu_gb": 15.61366891860962,
+  "best_checkpoint": "/data/home/jason/Seismic-FNO/6-code/O1-baseline-comparison/results/runs/BiLSTM/checkpoints/BiLSTM_best.pt",
+  "params": {
+    "numel": 8410113,
+    "numel_m": 8.410113,
+    "real_params": 8410113,
+    "real_params_m": 8.410113,
+    "trainable": 8410113
+  }
+}

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/runs/BiLSTM/training_log.csv

@@ -0,0 +1,51 @@
+epoch,train_mse,train_rmse,train_mae,train_r2,val_mse,val_rmse,val_mae,val_r2,val_pfa_rmse_g,lr,train_max_grad_norm,epoch_s,gpu_peak_gb
+1,0.17207485024138036,0.41481905723023427,0.23663428507249026,0.7825835685561484,0.08557287870827242,0.29252842376130295,0.16889224710267423,0.891847501665249,0.062053152707117434,0.0003,87.55697631835938,553.1554102897644,15.61366891860962
+2,0.07820247465434828,0.2796470537201282,0.1557729295875594,0.9011912377009019,0.06474816788879752,0.254456612979104,0.1407365133292215,0.9181671082534892,0.05326596666151942,0.0003,393.2450256347656,552.4883005619049,15.61366891860962
+3,0.062326362649720544,0.24965248376437305,0.13602147245827242,0.921250692247271,0.04681216217933343,0.21636118454873884,0.10932230883741471,0.9408357838538273,0.037439060737486575,0.0003,130.40016174316406,551.8074929714203,15.61366891860962
+4,0.04088781967906692,0.20220736801379646,0.10259190409699379,0.9483382735082833,0.030837602496230812,0.17560638512375001,0.08693451443618269,0.9610254580310279,0.034270109928641344,0.0003,44.51044464111328,552.1496667861938,15.61366891860962
+5,0.030687664422524137,0.1751789497129268,0.08851397506862588,0.9612261612690818,0.028224547466553212,0.1680016293568405,0.08779470005444384,0.9643280047492383,0.03241088135468313,0.0003,332.8004150390625,553.1752779483795,15.61366891860962
+6,0.026235554861027827,0.16197393265901716,0.08241845321090133,0.9668513980136207,0.02339132823207922,0.15294223822109843,0.07865044828597101,0.9704365375355427,0.030126346796048965,0.0003,2321.9580078125,552.7198297977448,15.61366891860962
+7,0.023206317354912413,0.15233619843921672,0.07698157368858888,0.9706788371700855,0.01924310572574013,0.13871952179033825,0.06669962409482365,0.9756793274764828,0.02667533069867842,0.0003,63.98646926879883,551.8243319988251,15.61366891860962
+8,0.02177611140825595,0.14756731144889762,0.07445870659603246,0.9724859012174911,0.019011426404689023,0.1378819292173163,0.06981341599849232,0.9759721386774428,0.025830060918657586,0.0003,6.219137668609619,549.75452876091,15.61366891860962
+9,0.02078723863122059,0.14417780214450693,0.07285866195827433,0.9737353411269044,0.019620552691186382,0.1400733832360252,0.06399445783648575,0.9752022857674965,0.026191281994320078,0.0003,8.779004096984863,550.7502152919769,15.61366891860962
+10,0.019359440921240614,0.1391382079848688,0.06973348341200086,0.9755393623492989,0.017953624786317464,0.1339911369692692,0.06577417593258864,0.9773090562790985,0.02461549578301737,0.0003,4.036068916320801,561.89670753479,15.61366891860962
+11,0.018406539198074192,0.13567070132520945,0.06813129394412273,0.9767433529209213,0.019130299081783778,0.1383123244030834,0.06539384769432262,0.9758218997558873,0.026322183977645235,0.0003,5.770036697387695,562.0843281745911,15.61366891860962
+12,0.017651037056906183,0.13285720551368746,0.06627685446101556,0.9776979292079717,0.018976129141467356,0.13775387160246116,0.06470196146766112,0.97601674962497,0.022980993509382295,0.0003,62.99094772338867,557.7918729782104,15.61366891860962
+13,0.017454917250664843,0.132117058893486,0.0661125298132348,0.977945726445593,0.01807208177949485,0.13443244318056133,0.07075910585518701,0.9771593427255667,0.023060836130341458,0.0003,5.197931289672852,561.0134081840515,15.61366891860962
+14,0.016989237708673607,0.1303427700667498,0.06475850709435535,0.9785341121887563,0.017660157841316207,0.13289152659713188,0.06419469443397448,0.977679958646293,0.02260687581659268,0.0003,4.526665210723877,558.1419236660004,15.61366891860962
+15,0.016091491024322214,0.12685224091170882,0.06330188571268122,0.9796684143608573,0.02020123865971556,0.1421310615583925,0.07137851512529697,0.9744683775574142,0.023599449053765827,0.0003,8.005902290344238,562.7496387958527,15.61366891860962
+16,0.015846276291518184,0.1258819935158249,0.06255892001324927,0.9799782429773907,0.01552947549384937,0.12461731618779699,0.05934699321070145,0.980372851797894,0.02200200209071282,0.0003,2.8729448318481445,560.3456118106842,15.61366891860962
+17,0.01574501089174148,0.12547912532266664,0.062303752108270896,0.9801061917097231,0.015665315686201723,0.1251611588560993,0.058260369870667336,0.9802011682411533,0.020337146054607583,0.0003,6.478118419647217,556.2974791526794,15.61366891860962
+18,0.015113573759499116,0.12293727571204396,0.06081780514008817,0.9809040120483631,0.015575286795037996,0.12480098875825463,0.06626637572052418,0.9803149525341284,0.02199726000572434,0.0003,10.412088394165039,550.1747784614563,15.61366891860962
+19,0.014766259466844229,0.12151649874335678,0.06022447817915829,0.9813428433318856,0.016227261269255222,0.12738626797757763,0.06571704073959857,0.9794909453334654,0.02028726520318195,0.0003,49.98328399658203,538.2235124111176,15.61366891860962
+20,0.014708558822542604,0.1212788473829736,0.059804201098585225,0.9814157480417048,0.013846861443175675,0.11767268775368257,0.054642928245499815,0.9824994474676961,0.01947560310501034,0.00015,9.356063842773438,531.6026237010956,15.61366891860962
+21,0.01183709588539763,0.1087984185794887,0.052219862388923156,0.9850438391201525,0.013151589070816486,0.11468037788050965,0.0515361669921875,0.983378177331981,0.019003010265720446,0.00015,4.093421936035156,531.3247518539429,15.61366891860962
+22,0.01174447589271947,0.10837193314100967,0.052257262409054046,0.9851608643755331,0.013466143303586725,0.11604371289986685,0.05013281923171373,0.9829806234965905,0.019728167983838498,0.00015,7.250269889831543,531.4999833106995,15.61366891860962
+23,0.011564689539561726,0.10753924650824799,0.051574795813012074,0.9853880242566636,0.012683614089103238,0.11262155250707227,0.04997526297704983,0.9839696341602944,0.019117205453633004,0.00015,9.744721412658691,532.6182827949524,15.61366891860962
+24,0.01138211274204663,0.10668698487653792,0.05112212401391935,0.985618709926478,0.012385876994588454,0.11129185502357508,0.05018708650049875,0.9843459334166099,0.018736014857563234,0.00015,11.520026206970215,532.3354380130768,15.61366891860962
+25,0.011387145550971376,0.10671056906872615,0.05136132493641874,0.9856123509782768,0.012980026881048786,0.11392992092092746,0.05498469361065424,0.983595008642593,0.018844369660464225,0.00015,17.79848861694336,533.6690957546234,15.61366891860962
+26,0.011159899359511586,0.10564042483591017,0.05075171758045695,0.9858994763417467,0.013992438299762808,0.11828963733042218,0.0628879584180449,0.9823154580895511,0.018878037808015127,0.00015,11.512173652648926,532.4485032558441,15.61366891860962
+27,0.010970771435297023,0.10474145041623695,0.0503415414696409,0.9861384393377133,0.01286471925523546,0.11342274575778644,0.051048521669027166,0.9837407418234443,0.018830923998297222,0.00015,6.900726318359375,531.5452935695648,15.61366891860962
+28,0.011093538632755391,0.10532586877284893,0.0506937338369864,0.9859833230836083,0.011667141421134012,0.10801454263724868,0.051272597304078335,0.9852543175808982,0.018798423631315456,0.00015,5.414413928985596,533.6138067245483,15.61366891860962
+29,0.010833991134375857,0.10408645989933492,0.04967032655236317,0.9863112611367879,0.011820104898077237,0.10872030582222089,0.04863193864313018,0.9850609925176878,0.018463439228445185,0.00015,5.334768295288086,535.6726810932159,15.61366891860962
+30,0.01084551792747096,0.10414181642102734,0.04974949384960747,0.9862966970458547,0.012403674238476372,0.11137178385244789,0.05028121586076465,0.9843234401171089,0.019225600319013467,0.00015,10.574443817138672,534.8414218425751,15.61366891860962
+31,0.010867470088906688,0.10424715866107186,0.04975006031415616,0.9862689604822306,0.012351168545133654,0.11113581126321818,0.04957446837308114,0.9843898001834936,0.01886151943132265,0.00015,9.412710189819336,538.1961479187012,15.61366891860962
+32,0.010529413123177274,0.10261292863561235,0.0490348576702988,0.9866960951877036,0.013022996462102521,0.11411834410865994,0.053481162337810674,0.9835407009271869,0.018997931170279803,0.00015,20.489402770996094,531.7264304161072,15.61366891860962
+33,0.010592538603844001,0.10292005928799304,0.04912573545995047,0.9866163361887371,0.012783800535443698,0.11306547012878732,0.04963184263352065,0.9838430120969192,0.01913747273349189,0.00015,12.580561637878418,531.5507717132568,15.61366891860962
+34,0.01058365534553751,0.10287689412855303,0.04907868236400463,0.9866275602117497,0.011931246101126568,0.1092302435277271,0.050259027944345454,0.9849205251294315,0.018190955764006346,0.00015,7.177252769470215,532.3788435459137,15.61366891860962
+35,0.010434384283768503,0.10214883398144349,0.04851404713532149,0.9868161640639155,0.01183321956433748,0.10878060288644056,0.04891940743072688,0.9850444173604384,0.018616918577541615,0.00015,9.626569747924805,531.3226070404053,15.61366891860962
+36,0.01030102634199414,0.10149397194904798,0.048454770092583074,0.9869846617255614,0.013811774556715818,0.11752350640070189,0.05368427133303637,0.9825437925275644,0.02058679591057512,0.00015,6.123040199279785,531.1708281040192,15.61366891860962
+37,0.010430023481580945,0.10212748641566062,0.04878799512004295,0.9868216739490977,0.011738169856973095,0.10834283482064282,0.04873321113452577,0.9851645472832903,0.018255200576627825,0.00015,10.532815933227539,532.7471764087677,15.61366891860962
+38,0.01032712747116535,0.10162247522652335,0.04825099668387549,0.9869516829741778,0.011669270233094808,0.10802439647179153,0.05015753272103055,0.9852516270516614,0.018489158196144415,0.00015,7.455319881439209,531.7821385860443,15.61366891860962
+39,0.010200631628798695,0.10099817636372795,0.04832264464975101,0.9871115103360979,0.012005633285254762,0.10957022079586572,0.04885756744325277,0.9848265098300868,0.017998728809005902,0.00015,19.690031051635742,531.2974264621735,15.61366891860962
+40,0.010111673074019582,0.10055681515451641,0.0477584238343118,0.9872239093988103,0.010683551145259864,0.10336126520732931,0.047339039865832114,0.9864974421231519,0.017894600845949595,7.5e-05,9.210453987121582,530.4533870220184,15.61366891860962
+41,0.00863529350444355,0.09292627994514549,0.04290549063342357,0.9890893137878084,0.010945453581335949,0.1046205217982397,0.04418458711679916,0.9861664330089425,0.01778884588330159,7.5e-05,15.215879440307617,531.2262489795685,15.61366891860962
+42,0.00848942618342171,0.09213808215619484,0.04275957156770643,0.9892736170130209,0.010663886136952896,0.10326609383990902,0.04317725809376142,0.9865222960232459,0.017997760401286533,7.5e-05,15.141629219055176,530.781108379364,15.61366891860962
+43,0.008516182029382062,0.09228316222032089,0.042765503281859164,0.9892398110348263,0.01155202952275267,0.10748036808065307,0.048631642407274155,0.9853998033888542,0.01802388738832502,7.5e-05,16.626554489135742,531.5453591346741,15.61366891860962
+44,0.008446617250349554,0.09190547997997482,0.04287548606589989,0.9893277060767426,0.010471863752814994,0.10233212473517295,0.04609285024423581,0.9867649862411538,0.01773814277203301,7.5e-05,6.203566074371338,533.7978627681732,15.61366891860962
+45,0.00838924279538279,0.09159280973626036,0.04235905202229818,0.9894001986415055,0.01076850360952158,0.10377140073026662,0.04447352124071028,0.9863900737444286,0.018010859694853887,7.5e-05,5.629856109619141,531.0890798568726,15.61366891860962
+46,0.008360426455938328,0.09143536764260494,0.04248183742233187,0.9894366080608675,0.010392227583619355,0.10194227574279159,0.042537693618268764,0.9868656355448389,0.017609251985833248,7.5e-05,8.28220272064209,529.6962485313416,15.61366891860962
+47,0.008397068752117789,0.0916355212355874,0.04255814020405968,0.9893903105518753,0.010816038316574246,0.10400018421413611,0.044581657557794224,0.9863299963296803,0.01769391946185467,7.5e-05,6.721210479736328,531.6631908416748,15.61366891860962
+48,0.008428662986234847,0.09180775014253888,0.04266411737215217,0.9893503912523457,0.010168918287935313,0.10084105457568021,0.043735230441623266,0.987147868170341,0.01781722473341725,7.5e-05,9.564264297485352,528.5599226951599,15.61366891860962
+49,0.008245496613203201,0.0908047169105394,0.04208973167136863,0.9895818218226001,0.011141420545670954,0.10555292769824508,0.04655334806505467,0.9859187573773189,0.018009201017080232,7.5e-05,10.830031394958496,526.5827569961548,15.61366891860962
+50,0.008266888935319629,0.09092243361964983,0.04223130043609798,0.9895547926320609,0.010677166836387232,0.10333037712302821,0.043909925634670444,0.986505511022611,0.017487494745196026,7.5e-05,7.556849002838135,528.1909773349762,15.61366891860962

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/runs/BiLSTM_nan_20260604_160510/checkpoints/BiLSTM_best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bec1c7d7924d69f85129d3fe6fd91294d65707a2d9891fefcd3c3a608322121
+size 100940852

experiments/O1_baseline/2026-06-09_xgb2_full_bs64/artifact/runs/BiLSTM_nan_20260604_160510/config.json

@@ -0,0 +1,18 @@
+{
+  "model": "BiLSTM",
+  "seed": 42,
+  "epochs": 50,
+  "batch_size": 64,
+  "lr": 0.001,
+  "weight_decay": 0.0001,
+  "scheduler_step": 20,
+  "scheduler_gamma": 0.5,
+  "split": "fixed_474_test_full_57_factors",
+  "params": {
+    "numel": 8410113,
+    "numel_m": 8.410113,
+    "real_params": 8410113,
+    "real_params_m": 8.410113,
+    "trainable": 8410113
+  }
+}