[ { "model": "TabPFN3", "fit_time_s": 0.6, "predict_time_s": 1.11, "proba_time_s": 0.58, "accuracy": 0.9954, "precision": 0.0, "recall": 0.0, "roc_auc": 0.8708, "n_train": 999, "n_test": 20000, "dataset": "sparknov", "requested_size": 1000 }, { "model": "TabICL", "fit_time_s": 0.21, "predict_time_s": 0.73, "proba_time_s": 0.72, "accuracy": 0.9954, "precision": 0.0, "recall": 0.0, "roc_auc": 0.8571, "n_train": 999, "n_test": 20000, "dataset": "sparknov", "requested_size": 1000 }, { "model": "TabPFN3", "fit_time_s": 0.23, "predict_time_s": 0.9, "proba_time_s": 0.88, "accuracy": 0.9951, "precision": 0.44, "recall": 0.2418, "roc_auc": 0.9098, "n_train": 4999, "n_test": 20000, "dataset": "sparknov", "requested_size": 5000 }, { "model": "TabICL", "fit_time_s": 0.16, "predict_time_s": 0.54, "proba_time_s": 0.53, "accuracy": 0.9957, "precision": 0.6316, "recall": 0.1319, "roc_auc": 0.8712, "n_train": 4999, "n_test": 20000, "dataset": "sparknov", "requested_size": 5000 }, { "model": "TabPFN3", "fit_time_s": 0.2, "predict_time_s": 1.35, "proba_time_s": 1.35, "accuracy": 0.9955, "precision": 0.5082, "recall": 0.3407, "roc_auc": 0.9475, "n_train": 9999, "n_test": 20000, "dataset": "sparknov", "requested_size": 10000 }, { "model": "TabICL", "fit_time_s": 0.16, "predict_time_s": 0.78, "proba_time_s": 0.79, "accuracy": 0.9955, "precision": 0.5106, "recall": 0.2637, "roc_auc": 0.9348, "n_train": 9999, "n_test": 20000, "dataset": "sparknov", "requested_size": 10000 }, { "model": "TabPFN3", "fit_time_s": 0.22, "predict_time_s": 2.64, "proba_time_s": 2.64, "accuracy": 0.9957, "precision": 0.5312, "recall": 0.3736, "roc_auc": 0.9412, "n_train": 19999, "n_test": 20000, "dataset": "sparknov", "requested_size": 20000 }, { "model": "TabICL", "fit_time_s": 0.2, "predict_time_s": 1.42, "proba_time_s": 1.43, "accuracy": 0.9958, "precision": 0.5455, "recall": 0.3956, "roc_auc": 0.9469, "n_train": 19999, "n_test": 20000, "dataset": "sparknov", "requested_size": 20000 }, { "model": "TabPFN3", "fit_time_s": 0.24, "predict_time_s": 8.75, "proba_time_s": 8.76, "accuracy": 0.9959, "precision": 0.5606, "recall": 0.4066, "roc_auc": 0.9577, "n_train": 49999, "n_test": 20000, "dataset": "sparknov", "requested_size": 50000 }, { "model": "TabICL", "fit_time_s": 0.26, "predict_time_s": 4.47, "proba_time_s": 4.48, "accuracy": 0.996, "precision": 0.5909, "recall": 0.4286, "roc_auc": 0.9624, "n_train": 49999, "n_test": 20000, "dataset": "sparknov", "requested_size": 50000 }, { "model": "TabPFN3", "fit_time_s": 0.26, "predict_time_s": 26.33, "proba_time_s": 26.39, "accuracy": 0.996, "precision": 0.6087, "recall": 0.3077, "roc_auc": 0.9692, "n_train": 99999, "n_test": 20000, "dataset": "sparknov", "requested_size": 100000 }, { "model": "TabICL", "fit_time_s": 0.38, "predict_time_s": 13.32, "proba_time_s": 13.35, "accuracy": 0.9964, "precision": 0.6727, "recall": 0.4066, "roc_auc": 0.9673, "n_train": 99999, "n_test": 20000, "dataset": "sparknov", "requested_size": 100000 }, { "model": "TabPFN3", "fit_time_s": 0.23, "predict_time_s": 1.96, "proba_time_s": 1.95, "accuracy": 0.9606, "precision": 0.534, "recall": 0.047, "roc_auc": 0.8333, "n_train": 999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 1000 }, { "model": "TabICL", "fit_time_s": 0.37, "predict_time_s": 1.62, "proba_time_s": 1.6, "accuracy": 0.9614, "precision": 0.5926, "recall": 0.0821, "roc_auc": 0.8247, "n_train": 999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 1000 }, { "model": "TabPFN3", "fit_time_s": 0.21, "predict_time_s": 2.47, "proba_time_s": 2.47, "accuracy": 0.9685, "precision": 0.74, "recall": 0.3165, "roc_auc": 0.8688, "n_train": 4999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 5000 }, { "model": "TabICL", "fit_time_s": 0.43, "predict_time_s": 2.04, "proba_time_s": 2.04, "accuracy": 0.9692, "precision": 0.7594, "recall": 0.3268, "roc_auc": 0.8724, "n_train": 4999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 5000 }, { "model": "TabPFN3", "fit_time_s": 0.22, "predict_time_s": 3.05, "proba_time_s": 3.06, "accuracy": 0.9689, "precision": 0.7466, "recall": 0.3251, "roc_auc": 0.8785, "n_train": 9999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 10000 }, { "model": "TabICL", "fit_time_s": 0.54, "predict_time_s": 2.37, "proba_time_s": 2.37, "accuracy": 0.9705, "precision": 0.8095, "recall": 0.3345, "roc_auc": 0.8836, "n_train": 9999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 10000 }, { "model": "TabPFN3", "fit_time_s": 0.27, "predict_time_s": 4.95, "proba_time_s": 4.93, "accuracy": 0.9704, "precision": 0.7964, "recall": 0.3379, "roc_auc": 0.8884, "n_train": 19999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 20000 }, { "model": "TabICL", "fit_time_s": 0.72, "predict_time_s": 3.69, "proba_time_s": 3.69, "accuracy": 0.9713, "precision": 0.8185, "recall": 0.355, "roc_auc": 0.8879, "n_train": 19999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 20000 }, { "model": "TabPFN3", "fit_time_s": 0.35, "predict_time_s": 12.97, "proba_time_s": 12.99, "accuracy": 0.9735, "precision": 0.8446, "recall": 0.4046, "roc_auc": 0.9092, "n_train": 49999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 50000 }, { "model": "TabICL", "fit_time_s": 1.31, "predict_time_s": 8.9, "proba_time_s": 8.9, "accuracy": 0.9731, "precision": 0.8333, "recall": 0.4021, "roc_auc": 0.9111, "n_train": 49999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 50000 }, { "model": "TabPFN3", "fit_time_s": 0.5, "predict_time_s": 33.76, "proba_time_s": 33.79, "accuracy": 0.9748, "precision": 0.8492, "recall": 0.4431, "roc_auc": 0.923, "n_train": 99999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 100000 }, { "model": "TabICL", "fit_time_s": 2.4, "predict_time_s": 28.8, "proba_time_s": 22.69, "accuracy": 0.9746, "precision": 0.8477, "recall": 0.438, "roc_auc": 0.9223, "n_train": 99999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 100000 }, { "model": "TabPFN3", "fit_time_s": 0.21, "predict_time_s": 1.67, "proba_time_s": 1.67, "accuracy": 0.8633, "precision": 0.8142, "recall": 0.7814, "roc_auc": 0.8965, "n_train": 999, "n_test": 65119, "dataset": "malurl", "requested_size": 1000 }, { "model": "TabICL", "fit_time_s": 0.13, "predict_time_s": 0.9, "proba_time_s": 0.89, "accuracy": 0.8702, "precision": 0.8395, "recall": 0.7702, "roc_auc": 0.9023, "n_train": 999, "n_test": 65119, "dataset": "malurl", "requested_size": 1000 }, { "model": "TabPFN3", "fit_time_s": 0.2, "predict_time_s": 2.31, "proba_time_s": 2.31, "accuracy": 0.8728, "precision": 0.8096, "recall": 0.8244, "roc_auc": 0.908, "n_train": 4999, "n_test": 65119, "dataset": "malurl", "requested_size": 5000 }, { "model": "TabICL", "fit_time_s": 0.13, "predict_time_s": 0.57, "proba_time_s": 0.57, "accuracy": 0.8791, "precision": 0.8219, "recall": 0.8284, "roc_auc": 0.916, "n_train": 4999, "n_test": 65119, "dataset": "malurl", "requested_size": 5000 }, { "model": "TabPFN3", "fit_time_s": 0.24, "predict_time_s": 3.21, "proba_time_s": 3.21, "accuracy": 0.8807, "precision": 0.8322, "recall": 0.8185, "roc_auc": 0.9191, "n_train": 9999, "n_test": 65119, "dataset": "malurl", "requested_size": 10000 }, { "model": "TabICL", "fit_time_s": 0.13, "predict_time_s": 0.78, "proba_time_s": 0.79, "accuracy": 0.8937, "precision": 0.869, "recall": 0.8141, "roc_auc": 0.9303, "n_train": 9999, "n_test": 65119, "dataset": "malurl", "requested_size": 10000 }, { "model": "TabPFN3", "fit_time_s": 0.21, "predict_time_s": 5.37, "proba_time_s": 5.38, "accuracy": 0.8836, "precision": 0.8436, "recall": 0.8127, "roc_auc": 0.9257, "n_train": 19999, "n_test": 65119, "dataset": "malurl", "requested_size": 20000 }, { "model": "TabICL", "fit_time_s": 0.14, "predict_time_s": 1.29, "proba_time_s": 1.29, "accuracy": 0.9, "precision": 0.8784, "recall": 0.8234, "roc_auc": 0.9363, "n_train": 19999, "n_test": 65119, "dataset": "malurl", "requested_size": 20000 }, { "model": "TabPFN3", "fit_time_s": 0.21, "predict_time_s": 14.07, "proba_time_s": 14.08, "accuracy": 0.8866, "precision": 0.8548, "recall": 0.808, "roc_auc": 0.9288, "n_train": 49999, "n_test": 65119, "dataset": "malurl", "requested_size": 50000 }, { "model": "TabICL", "fit_time_s": 0.14, "predict_time_s": 3.37, "proba_time_s": 3.38, "accuracy": 0.9008, "precision": 0.8808, "recall": 0.8235, "roc_auc": 0.9391, "n_train": 49999, "n_test": 65119, "dataset": "malurl", "requested_size": 50000 }, { "model": "TabPFN3", "fit_time_s": 0.21, "predict_time_s": 35.99, "proba_time_s": 36.01, "accuracy": 0.888, "precision": 0.8575, "recall": 0.8094, "roc_auc": 0.9302, "n_train": 99999, "n_test": 65119, "dataset": "malurl", "requested_size": 100000 }, { "model": "TabICL", "fit_time_s": 0.16, "predict_time_s": 8.69, "proba_time_s": 8.7, "accuracy": 0.9038, "precision": 0.892, "recall": 0.8198, "roc_auc": 0.9406, "n_train": 99999, "n_test": 65119, "dataset": "malurl", "requested_size": 100000 }, { "model": "TabPFN3", "fit_time_s": 0.2, "predict_time_s": 0.29, "proba_time_s": 0.28, "accuracy": 0.9581, "precision": 0.9512, "recall": 0.2086, "roc_auc": 0.9024, "n_train": 999, "n_test": 3576, "dataset": "fakejob", "requested_size": 1000 }, { "model": "TabICL", "fit_time_s": 0.16, "predict_time_s": 0.22, "proba_time_s": 0.21, "accuracy": 0.9611, "precision": 0.8333, "recall": 0.3209, "roc_auc": 0.9212, "n_train": 999, "n_test": 3576, "dataset": "fakejob", "requested_size": 1000 }, { "model": "TabPFN3", "fit_time_s": 0.21, "predict_time_s": 0.44, "proba_time_s": 0.44, "accuracy": 0.9768, "precision": 0.9062, "recall": 0.6203, "roc_auc": 0.9806, "n_train": 4999, "n_test": 3576, "dataset": "fakejob", "requested_size": 5000 }, { "model": "TabICL", "fit_time_s": 0.18, "predict_time_s": 0.24, "proba_time_s": 0.24, "accuracy": 0.9785, "precision": 0.9044, "recall": 0.6578, "roc_auc": 0.9813, "n_train": 4999, "n_test": 3576, "dataset": "fakejob", "requested_size": 5000 }, { "model": "TabPFN3", "fit_time_s": 0.22, "predict_time_s": 0.77, "proba_time_s": 0.77, "accuracy": 0.9804, "precision": 0.9398, "recall": 0.6684, "roc_auc": 0.9878, "n_train": 9999, "n_test": 3576, "dataset": "fakejob", "requested_size": 10000 }, { "model": "TabICL", "fit_time_s": 0.2, "predict_time_s": 0.46, "proba_time_s": 0.47, "accuracy": 0.981, "precision": 0.9281, "recall": 0.6898, "roc_auc": 0.9876, "n_train": 9999, "n_test": 3576, "dataset": "fakejob", "requested_size": 10000 }, { "model": "TabPFN3", "fit_time_s": 0.58, "predict_time_s": 92.04, "proba_time_s": 92.16, "accuracy": 0.9958, "precision": 0.5682, "recall": 0.2747, "roc_auc": 0.9722, "n_train": 199999, "n_test": 20000, "dataset": "sparknov", "requested_size": 200000 }, { "model": "TabICL", "fit_time_s": 0.62, "predict_time_s": 44.86, "proba_time_s": 44.92, "accuracy": 0.9962, "precision": 0.6316, "recall": 0.3956, "roc_auc": 0.9686, "n_train": 199999, "n_test": 20000, "dataset": "sparknov", "requested_size": 200000 }, { "model": "TabPFN3", "fit_time_s": 0.76, "predict_time_s": 506.9, "proba_time_s": 507.96, "accuracy": 0.9951, "precision": 0.3871, "recall": 0.1319, "roc_auc": 0.9693, "n_train": 499999, "n_test": 20000, "dataset": "sparknov", "requested_size": 500000 }, { "model": "TabICL", "fit_time_s": 1.29, "predict_time_s": 252.88, "proba_time_s": 253.0, "accuracy": 0.996, "precision": 0.625, "recall": 0.3297, "roc_auc": 0.9701, "n_train": 499999, "n_test": 20000, "dataset": "sparknov", "requested_size": 500000 }, { "model": "TabPFN3", "fit_time_s": 1.13, "predict_time_s": 1956.32, "proba_time_s": 1956.5, "accuracy": 0.9952, "precision": 0.4333, "recall": 0.1429, "roc_auc": 0.9593, "n_train": 999999, "n_test": 20000, "dataset": "sparknov", "requested_size": 1000000 }, { "model": "TabICL", "fit_time_s": 2.62, "predict_time_s": 991.81, "proba_time_s": 986.61, "accuracy": 0.9958, "precision": 0.5745, "recall": 0.2967, "roc_auc": 0.967, "n_train": 999999, "n_test": 20000, "dataset": "sparknov", "requested_size": 1000000 }, { "model": "TabPFN3", "fit_time_s": 0.83, "predict_time_s": 103.23, "proba_time_s": 103.36, "accuracy": 0.9758, "precision": 0.8534, "recall": 0.4679, "roc_auc": 0.9349, "n_train": 199999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 200000 }, { "model": "TabICL", "fit_time_s": 4.35, "predict_time_s": 64.6, "proba_time_s": 62.75, "accuracy": 0.9755, "precision": 0.8474, "recall": 0.4654, "roc_auc": 0.9336, "n_train": 199999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 200000 }, { "model": "TabPFN3", "fit_time_s": 671.92, "predict_time_s": 534.0, "proba_time_s": 535.17, "accuracy": 0.9773, "precision": 0.8611, "recall": 0.509, "roc_auc": 0.952, "n_train": 499999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 500000 }, { "model": "TabICL", "fit_time_s": 10.9, "predict_time_s": 360.89, "proba_time_s": 365.37, "accuracy": 0.9777, "precision": 0.8581, "recall": 0.5227, "roc_auc": 0.9485, "n_train": 499999, "n_test": 29527, "dataset": "ieeecis", "requested_size": 500000 }, { "model": "TabPFN3", "fit_time_s": 1.26, "predict_time_s": 108.35, "proba_time_s": 107.91, "accuracy": 0.8884, "precision": 0.8589, "recall": 0.8088, "roc_auc": 0.9307, "n_train": 199999, "n_test": 65119, "dataset": "malurl", "requested_size": 200000 }, { "model": "TabICL", "fit_time_s": 0.26, "predict_time_s": 26.38, "proba_time_s": 26.35, "accuracy": 0.9052, "precision": 0.8944, "recall": 0.8218, "roc_auc": 0.9406, "n_train": 199999, "n_test": 65119, "dataset": "malurl", "requested_size": 200000 }, { "model": "TabPFN3", "fit_time_s": 0.3, "predict_time_s": 545.89, "proba_time_s": 546.01, "accuracy": 0.8887, "precision": 0.8577, "recall": 0.8113, "roc_auc": 0.9312, "n_train": 499999, "n_test": 65119, "dataset": "malurl", "requested_size": 500000 }, { "model": "TabICL", "fit_time_s": 0.31, "predict_time_s": 135.34, "proba_time_s": 135.31, "accuracy": 0.905, "precision": 0.8947, "recall": 0.8206, "roc_auc": 0.9418, "n_train": 499999, "n_test": 65119, "dataset": "malurl", "requested_size": 500000 }, { "model": "TabFM", "dataset": "fakejob", "requested_size": 500, "n_train": 499, "n_test": 3576, "fit_time_s": 0.0, "proba_time_s": 0.64, "roc_auc": 0.9202, "avg_precision": 0.5488 }, { "model": "TabFM-Ensemble", "dataset": "fakejob", "requested_size": 500, "n_train": 499, "n_test": 3576, "fit_time_s": 10.88, "proba_time_s": 19.5, "roc_auc": 0.9141, "avg_precision": 0.518 }, { "model": "TabFM", "dataset": "fakejob", "requested_size": 1000, "n_train": 999, "n_test": 3576, "fit_time_s": 0.0, "proba_time_s": 0.6, "roc_auc": 0.9519, "avg_precision": 0.6728 }, { "model": "TabFM-Ensemble", "dataset": "fakejob", "requested_size": 1000, "n_train": 999, "n_test": 3576, "fit_time_s": 21.07, "proba_time_s": 22.24, "roc_auc": 0.9507, "avg_precision": 0.6757 }, { "model": "TabFM", "dataset": "fakejob", "requested_size": 2000, "n_train": 1999, "n_test": 3576, "fit_time_s": 0.0, "proba_time_s": 0.81, "roc_auc": 0.9705, "avg_precision": 0.7765 }, { "model": "TabFM-Ensemble", "dataset": "fakejob", "requested_size": 2000, "n_train": 1999, "n_test": 3576, "fit_time_s": 44.74, "proba_time_s": 29.23, "roc_auc": 0.972, "avg_precision": 0.7832 }, { "model": "TabFM", "dataset": "fakejob", "requested_size": 5000, "n_train": 4999, "n_test": 3576, "fit_time_s": 0.0, "proba_time_s": 1.44, "roc_auc": 0.9919, "avg_precision": 0.9243 }, { "model": "TabFM-Ensemble", "dataset": "fakejob", "requested_size": 5000, "n_train": 4999, "n_test": 3576, "fit_time_s": 128.45, "proba_time_s": 50.73, "roc_auc": 0.9919, "avg_precision": 0.9243 }, { "model": "TabFM", "dataset": "fakejob", "requested_size": 10000, "n_train": 9999, "n_test": 3576, "fit_time_s": 0.0, "proba_time_s": 2.73, "roc_auc": 0.9943, "avg_precision": 0.9425 }, { "model": "TabFM-Ensemble", "dataset": "fakejob", "requested_size": 10000, "n_train": 9999, "n_test": 3576, "fit_time_s": 62.61, "proba_time_s": 93.72, "roc_auc": 0.9939, "avg_precision": 0.9422 }, { "model": "TabFM", "dataset": "fakejob", "requested_size": 20000, "n_train": 14304, "n_test": 3576, "fit_time_s": 0.01, "proba_time_s": 4.04, "roc_auc": 0.9953, "avg_precision": 0.9504 }, { "model": "TabFM-Ensemble", "dataset": "fakejob", "requested_size": 20000, "n_train": 14304, "n_test": 3576, "fit_time_s": 102.04, "proba_time_s": 138.43, "roc_auc": 0.9951, "avg_precision": 0.9502 }, { "model": "TabFM", "dataset": "fakejob", "requested_size": 50000, "n_train": 14304, "n_test": 3576, "fit_time_s": 0.01, "proba_time_s": 4.05, "roc_auc": 0.9953, "avg_precision": 0.9504 }, { "model": "TabFM-Ensemble", "dataset": "fakejob", "requested_size": 50000, "n_train": 14304, "n_test": 3576, "fit_time_s": 102.1, "proba_time_s": 138.59, "roc_auc": 0.9951, "avg_precision": 0.9502 }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 500, "error": "CUDA out of memory. Tried to allocate 6.65 GiB. GPU 0 has a total capacity of 31.40 GiB of which 1.32 GiB is free. Including non-PyTorch memory, this process has 30.07 GiB memory in use. Of the allocated memory 26.11 GiB is allocated by PyTorch, and 3.35 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM-Ensemble", "requested_size": 500, "error": "CUDA out of memory. Tried to allocate 8.81 GiB. GPU 0 has a total capacity of 31.40 GiB of which 402.56 MiB is free. Including non-PyTorch memory, this process has 30.99 GiB memory in use. Of the allocated memory 23.78 GiB is allocated by PyTorch, and 6.61 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 1000, "error": "CUDA out of memory. Tried to allocate 6.72 GiB. GPU 0 has a total capacity of 31.40 GiB of which 1.08 GiB is free. Including non-PyTorch memory, this process has 30.30 GiB memory in use. Of the allocated memory 26.33 GiB is allocated by PyTorch, and 3.37 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM-Ensemble", "requested_size": 1000, "error": "CUDA out of memory. Tried to allocate 8.90 GiB. GPU 0 has a total capacity of 31.40 GiB of which 138.56 MiB is free. Including non-PyTorch memory, this process has 31.25 GiB memory in use. Of the allocated memory 23.97 GiB is allocated by PyTorch, and 6.68 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 2000, "error": "CUDA out of memory. Tried to allocate 6.87 GiB. GPU 0 has a total capacity of 31.40 GiB of which 606.56 MiB is free. Including non-PyTorch memory, this process has 30.79 GiB memory in use. Of the allocated memory 26.75 GiB is allocated by PyTorch, and 3.44 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM-Ensemble", "requested_size": 2000, "error": "CUDA out of memory. Tried to allocate 9.09 GiB. GPU 0 has a total capacity of 31.40 GiB of which 8.71 GiB is free. Including non-PyTorch memory, this process has 22.68 GiB memory in use. Of the allocated memory 17.53 GiB is allocated by PyTorch, and 4.54 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 500, "error": "CUDA out of memory. Tried to allocate 6.65 GiB. GPU 0 has a total capacity of 31.40 GiB of which 738.56 MiB is free. Including non-PyTorch memory, this process has 30.67 GiB memory in use. Of the allocated memory 26.11 GiB is allocated by PyTorch, and 3.95 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 1000, "error": "CUDA out of memory. Tried to allocate 6.72 GiB. GPU 0 has a total capacity of 31.40 GiB of which 522.56 MiB is free. Including non-PyTorch memory, this process has 30.88 GiB memory in use. Of the allocated memory 26.33 GiB is allocated by PyTorch, and 3.94 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 2000, "error": "CUDA out of memory. Tried to allocate 6.87 GiB. GPU 0 has a total capacity of 31.40 GiB of which 18.56 MiB is free. Including non-PyTorch memory, this process has 31.37 GiB memory in use. Of the allocated memory 26.75 GiB is allocated by PyTorch, and 4.01 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 5000, "error": "CUDA out of memory. Tried to allocate 7.29 GiB. GPU 0 has a total capacity of 31.40 GiB of which 5.82 GiB is free. Including non-PyTorch memory, this process has 25.56 GiB memory in use. Of the allocated memory 20.73 GiB is allocated by PyTorch, and 4.23 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 10000, "error": "CUDA out of memory. Tried to allocate 7.99 GiB. GPU 0 has a total capacity of 31.40 GiB of which 4.06 GiB is free. Including non-PyTorch memory, this process has 27.33 GiB memory in use. Of the allocated memory 22.14 GiB is allocated by PyTorch, and 4.58 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 20000, "error": "CUDA out of memory. Tried to allocate 9.41 GiB. GPU 0 has a total capacity of 31.40 GiB of which 7.58 GiB is free. Including non-PyTorch memory, this process has 23.80 GiB memory in use. Of the allocated memory 17.92 GiB is allocated by PyTorch, and 5.28 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 50000, "error": "CUDA out of memory. Tried to allocate 10.23 GiB. GPU 0 has a total capacity of 31.40 GiB of which 3.58 GiB is free. Including non-PyTorch memory, this process has 27.80 GiB memory in use. Of the allocated memory 26.65 GiB is allocated by PyTorch, and 557.62 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "model": "TabFM", "dataset": "malurl", "requested_size": 500, "n_train": 499, "n_test": 65119, "fit_time_s": 0.0, "proba_time_s": 32.92, "roc_auc": 0.8929, "avg_precision": 0.8634 }, { "model": "TabFM", "dataset": "malurl", "requested_size": 1000, "n_train": 999, "n_test": 65119, "fit_time_s": 0.0, "proba_time_s": 33.72, "roc_auc": 0.9046, "avg_precision": 0.8795 }, { "model": "TabFM", "dataset": "malurl", "requested_size": 2000, "n_train": 1999, "n_test": 65119, "fit_time_s": 0.0, "proba_time_s": 34.73, "roc_auc": 0.9087, "avg_precision": 0.8861 }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 500, "error": "CUDA out of memory. Tried to allocate 6.65 GiB. GPU 0 has a total capacity of 31.40 GiB of which 4.63 GiB is free. Including non-PyTorch memory, this process has 26.76 GiB memory in use. Of the allocated memory 26.11 GiB is allocated by PyTorch, and 40.80 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 1000, "error": "CUDA out of memory. Tried to allocate 6.72 GiB. GPU 0 has a total capacity of 31.40 GiB of which 4.41 GiB is free. Including non-PyTorch memory, this process has 26.97 GiB memory in use. Of the allocated memory 26.32 GiB is allocated by PyTorch, and 43.91 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 2000, "error": "CUDA out of memory. Tried to allocate 6.87 GiB. GPU 0 has a total capacity of 31.40 GiB of which 3.98 GiB is free. Including non-PyTorch memory, this process has 27.40 GiB memory in use. Of the allocated memory 26.75 GiB is allocated by PyTorch, and 50.13 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 5000, "error": "CUDA out of memory. Tried to allocate 7.29 GiB. GPU 0 has a total capacity of 31.40 GiB of which 2.73 GiB is free. Including non-PyTorch memory, this process has 28.65 GiB memory in use. Of the allocated memory 28.02 GiB is allocated by PyTorch, and 28.80 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 10000, "error": "CUDA out of memory. Tried to allocate 7.99 GiB. GPU 0 has a total capacity of 31.40 GiB of which 6.56 GiB is free. Including non-PyTorch memory, this process has 24.83 GiB memory in use. Of the allocated memory 24.14 GiB is allocated by PyTorch, and 78.51 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 20000, "error": "CUDA out of memory. Tried to allocate 9.41 GiB. GPU 0 has a total capacity of 31.40 GiB of which 5.76 GiB is free. Including non-PyTorch memory, this process has 25.63 GiB memory in use. Of the allocated memory 24.97 GiB is allocated by PyTorch, and 52.25 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 50000, "error": "CUDA out of memory. Tried to allocate 10.23 GiB. GPU 0 has a total capacity of 31.40 GiB of which 4.06 GiB is free. Including non-PyTorch memory, this process has 27.33 GiB memory in use. Of the allocated memory 26.65 GiB is allocated by PyTorch, and 69.62 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "model": "TabFM", "dataset": "vehicleloan", "requested_size": 500, "n_train": 499, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.07, "roc_auc": 0.5786, "avg_precision": 0.2718 }, { "model": "TabFM", "dataset": "vehicleloan", "requested_size": 1000, "n_train": 999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.05, "roc_auc": 0.6057, "avg_precision": 0.2851 }, { "model": "TabFM", "dataset": "vehicleloan", "requested_size": 2000, "n_train": 1999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.29, "roc_auc": 0.6207, "avg_precision": 0.295 }, { "model": "TabFM", "dataset": "vehicleloan", "requested_size": 5000, "n_train": 4999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 2.04, "roc_auc": 0.6475, "avg_precision": 0.3216 }, { "model": "TabFM", "dataset": "vehicleloan", "requested_size": 10000, "n_train": 9999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 3.59, "roc_auc": 0.6634, "avg_precision": 0.3432 }, { "model": "TabFM", "dataset": "vehicleloan", "requested_size": 20000, "n_train": 19999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 7.53, "roc_auc": 0.673, "avg_precision": 0.3548 }, { "dataset": "vehicleloan", "model": "TabFM", "requested_size": 50000, "error": "CUDA out of memory. Tried to allocate 7.76 GiB. GPU 0 has a total capacity of 31.40 GiB of which 7.09 GiB is free. Including non-PyTorch memory, this process has 24.30 GiB memory in use. Of the allocated memory 23.62 GiB is allocated by PyTorch, and 69.21 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "model": "TabFM", "dataset": "malurl", "requested_size": 5000, "n_train": 4999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.56, "roc_auc": 0.9276, "avg_precision": 0.9133 }, { "model": "TabFM", "dataset": "malurl", "requested_size": 10000, "n_train": 9999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 2.71, "roc_auc": 0.9351, "avg_precision": 0.9225 }, { "model": "TabFM", "dataset": "malurl", "requested_size": 20000, "n_train": 19999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 6.06, "roc_auc": 0.9412, "avg_precision": 0.9296 }, { "model": "TabFM", "dataset": "malurl", "requested_size": 50000, "n_train": 49999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 24.3, "roc_auc": 0.9431, "avg_precision": 0.9317 }, { "model": "TabFM", "dataset": "ieeecis", "requested_size": 500, "n_train": 499, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.29, "roc_auc": 0.8056, "avg_precision": 0.3664 }, { "model": "TabFM", "dataset": "ieeecis", "requested_size": 1000, "n_train": 999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.31, "roc_auc": 0.822, "avg_precision": 0.3842 }, { "model": "TabFM", "dataset": "ieeecis", "requested_size": 2000, "n_train": 1999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.6, "roc_auc": 0.844, "avg_precision": 0.425 }, { "model": "TabFM", "dataset": "ieeecis", "requested_size": 5000, "n_train": 4999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 2.47, "roc_auc": 0.8726, "avg_precision": 0.5114 }, { "model": "TabFM", "dataset": "ieeecis", "requested_size": 10000, "n_train": 9999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 4.24, "roc_auc": 0.8811, "avg_precision": 0.5533 }, { "dataset": "ieeecis", "model": "TabFM", "requested_size": 20000, "error": "CUDA out of memory. Tried to allocate 5.82 GiB. GPU 0 has a total capacity of 31.40 GiB of which 5.66 GiB is free. Including non-PyTorch memory, this process has 25.72 GiB memory in use. Of the allocated memory 25.07 GiB is allocated by PyTorch, and 51.71 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "dataset": "ieeecis", "model": "TabFM", "requested_size": 50000, "error": "CUDA out of memory. Tried to allocate 9.76 GiB. GPU 0 has a total capacity of 31.40 GiB of which 5.02 GiB is free. Including non-PyTorch memory, this process has 26.37 GiB memory in use. Of the allocated memory 25.70 GiB is allocated by PyTorch, and 61.44 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "model": "TabFM", "dataset": "ccfraud", "requested_size": 500, "n_train": 500, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.0, "roc_auc": 0.82, "avg_precision": 0.0149 }, { "model": "TabFM", "dataset": "ccfraud", "requested_size": 1000, "n_train": 999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 0.97, "roc_auc": 0.8829, "avg_precision": 0.1378 }, { "model": "TabFM", "dataset": "ccfraud", "requested_size": 2000, "n_train": 1999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.21, "roc_auc": 0.9383, "avg_precision": 0.8339 }, { "model": "TabFM", "dataset": "ccfraud", "requested_size": 5000, "n_train": 4999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.91, "roc_auc": 0.9972, "avg_precision": 0.8446 }, { "model": "TabFM", "dataset": "ccfraud", "requested_size": 10000, "n_train": 9999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 3.4, "roc_auc": 0.9987, "avg_precision": 0.8561 }, { "model": "TabFM", "dataset": "ccfraud", "requested_size": 20000, "n_train": 19999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 7.22, "roc_auc": 0.9989, "avg_precision": 0.8583 }, { "dataset": "ccfraud", "model": "TabFM", "requested_size": 50000, "error": "CUDA out of memory. Tried to allocate 6.09 GiB. GPU 0 has a total capacity of 31.40 GiB of which 4.80 GiB is free. Including non-PyTorch memory, this process has 26.58 GiB memory in use. Of the allocated memory 25.93 GiB is allocated by PyTorch, and 50.07 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf)" }, { "model": "TabFM", "dataset": "fraudecom", "requested_size": 500, "n_train": 499, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 0.78, "roc_auc": 0.4923, "avg_precision": 0.0441 }, { "model": "TabFM", "dataset": "fraudecom", "requested_size": 1000, "n_train": 999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 0.73, "roc_auc": 0.4791, "avg_precision": 0.0441 }, { "model": "TabFM", "dataset": "fraudecom", "requested_size": 2000, "n_train": 1999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 0.93, "roc_auc": 0.5101, "avg_precision": 0.0474 }, { "model": "TabFM", "dataset": "fraudecom", "requested_size": 5000, "n_train": 4999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.52, "roc_auc": 0.5027, "avg_precision": 0.0468 }, { "model": "TabFM", "dataset": "fraudecom", "requested_size": 10000, "n_train": 9999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 2.82, "roc_auc": 0.5394, "avg_precision": 0.05 }, { "model": "TabFM", "dataset": "fraudecom", "requested_size": 20000, "n_train": 19999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 6.27, "roc_auc": 0.5145, "avg_precision": 0.0481 }, { "model": "TabFM", "dataset": "fraudecom", "requested_size": 50000, "n_train": 49999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 24.74, "roc_auc": 0.5233, "avg_precision": 0.0486 }, { "model": "TabFM", "dataset": "sparknov", "requested_size": 500, "n_train": 499, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 0.82, "roc_auc": 0.9289, "avg_precision": 0.3617 }, { "model": "TabFM", "dataset": "sparknov", "requested_size": 1000, "n_train": 999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 0.76, "roc_auc": 0.9446, "avg_precision": 0.4199 }, { "model": "TabFM", "dataset": "sparknov", "requested_size": 2000, "n_train": 1999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 0.96, "roc_auc": 0.9662, "avg_precision": 0.4554 }, { "model": "TabFM", "dataset": "sparknov", "requested_size": 5000, "n_train": 4999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 1.57, "roc_auc": 0.9698, "avg_precision": 0.5518 }, { "model": "TabFM", "dataset": "sparknov", "requested_size": 10000, "n_train": 9999, "n_test": 5000, "fit_time_s": 0.0, "proba_time_s": 2.89, "roc_auc": 0.9816, "avg_precision": 0.6474 }, { "model": "TabFM", "dataset": "sparknov", "requested_size": 20000, "n_train": 19999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 6.36, "roc_auc": 0.9805, "avg_precision": 0.7033 }, { "model": "TabFM", "dataset": "sparknov", "requested_size": 50000, "n_train": 49999, "n_test": 5000, "fit_time_s": 0.01, "proba_time_s": 24.95, "roc_auc": 0.9858, "avg_precision": 0.7072 } ]