Upload benchmarks/business_results.json with huggingface_hub
Browse files- benchmarks/business_results.json +154 -0
benchmarks/business_results.json
ADDED
|
@@ -0,0 +1,154 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model": "/root/anofy/dist/tcpfn/outputs/temporal/final.pt",
|
| 3 |
+
"machine": {
|
| 4 |
+
"timestamp": "2026-04-09T09:49:45.841270+00:00",
|
| 5 |
+
"hostname": "7aa1a49e1d2f",
|
| 6 |
+
"platform": "Linux-6.8.0-106-generic-x86_64-with-glibc2.39",
|
| 7 |
+
"python": "3.12.13",
|
| 8 |
+
"cpu": "x86_64",
|
| 9 |
+
"cpu_count": 384,
|
| 10 |
+
"torch_version": "2.10.0+cu130",
|
| 11 |
+
"gpu": "NVIDIA GeForce RTX 5090",
|
| 12 |
+
"gpu_count": 1,
|
| 13 |
+
"ram_gb": 540.5
|
| 14 |
+
},
|
| 15 |
+
"config": {
|
| 16 |
+
"scenarios": [
|
| 17 |
+
"marketing",
|
| 18 |
+
"pricing",
|
| 19 |
+
"job_training"
|
| 20 |
+
],
|
| 21 |
+
"n_folds": 5,
|
| 22 |
+
"n_tables": 10,
|
| 23 |
+
"device": "cuda:0"
|
| 24 |
+
},
|
| 25 |
+
"total_time_s": 70.37413984700106,
|
| 26 |
+
"results": {
|
| 27 |
+
"marketing": {
|
| 28 |
+
"scenario": "marketing_uplift",
|
| 29 |
+
"description": "E-mail campaign targeting: identify customers who will visit after receiving an email",
|
| 30 |
+
"n_folds": 5,
|
| 31 |
+
"standard_metrics": {
|
| 32 |
+
"ate_error": {
|
| 33 |
+
"mean": NaN,
|
| 34 |
+
"std": NaN
|
| 35 |
+
},
|
| 36 |
+
"pehe": {
|
| 37 |
+
"mean": NaN,
|
| 38 |
+
"std": NaN
|
| 39 |
+
},
|
| 40 |
+
"uplift_auroc": {
|
| 41 |
+
"mean": 0.505058442393027,
|
| 42 |
+
"std": 0.04051704818029946
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"business_metrics": {
|
| 46 |
+
"decision_accuracy": {
|
| 47 |
+
"mean": NaN,
|
| 48 |
+
"std": NaN
|
| 49 |
+
},
|
| 50 |
+
"policy_value": {
|
| 51 |
+
"mean": 0.020291447333237185,
|
| 52 |
+
"std": 0.0001581025200012148
|
| 53 |
+
},
|
| 54 |
+
"policy_value_random": {
|
| 55 |
+
"mean": 0.14670000076293946,
|
| 56 |
+
"std": 0.0004000000655755176
|
| 57 |
+
},
|
| 58 |
+
"policy_value_recommended": {
|
| 59 |
+
"mean": 0.16699144809617666,
|
| 60 |
+
"std": 0.0002913922853402658
|
| 61 |
+
},
|
| 62 |
+
"cost_effectiveness": {
|
| 63 |
+
"mean": 0.12089049530029297,
|
| 64 |
+
"std": 0.01129887460962203
|
| 65 |
+
}
|
| 66 |
+
},
|
| 67 |
+
"time_s": 6.957995977019891
|
| 68 |
+
},
|
| 69 |
+
"pricing": {
|
| 70 |
+
"scenario": "pricing_optimization",
|
| 71 |
+
"description": "Ad targeting: identify users who will visit after ad exposure, optimizing ad spend",
|
| 72 |
+
"n_folds": 5,
|
| 73 |
+
"standard_metrics": {
|
| 74 |
+
"ate_error": {
|
| 75 |
+
"mean": NaN,
|
| 76 |
+
"std": NaN
|
| 77 |
+
},
|
| 78 |
+
"pehe": {
|
| 79 |
+
"mean": NaN,
|
| 80 |
+
"std": NaN
|
| 81 |
+
},
|
| 82 |
+
"uplift_auroc": {
|
| 83 |
+
"mean": 0.5177793054403224,
|
| 84 |
+
"std": 0.16233572232620164
|
| 85 |
+
}
|
| 86 |
+
},
|
| 87 |
+
"business_metrics": {
|
| 88 |
+
"decision_accuracy": {
|
| 89 |
+
"mean": NaN,
|
| 90 |
+
"std": NaN
|
| 91 |
+
},
|
| 92 |
+
"policy_value": {
|
| 93 |
+
"mean": 0.0026842953794115956,
|
| 94 |
+
"std": 0.0023026318725680927
|
| 95 |
+
},
|
| 96 |
+
"policy_value_random": {
|
| 97 |
+
"mean": 0.04699999839067459,
|
| 98 |
+
"std": 0.0
|
| 99 |
+
},
|
| 100 |
+
"policy_value_recommended": {
|
| 101 |
+
"mean": 0.04968429377008619,
|
| 102 |
+
"std": 0.0023026318725680927
|
| 103 |
+
},
|
| 104 |
+
"cost_effectiveness": {
|
| 105 |
+
"mean": 0.00514608790167189,
|
| 106 |
+
"std": 0.0017655088223771072
|
| 107 |
+
}
|
| 108 |
+
},
|
| 109 |
+
"time_s": 35.788739799987525
|
| 110 |
+
},
|
| 111 |
+
"job_training": {
|
| 112 |
+
"scenario": "job_training_roi",
|
| 113 |
+
"description": "Social program ROI: measure whether job training / health interventions deliver value",
|
| 114 |
+
"n_folds": 10,
|
| 115 |
+
"standard_metrics": {
|
| 116 |
+
"ate_error": {
|
| 117 |
+
"mean": 948.0720703125,
|
| 118 |
+
"std": 450.5277766526623
|
| 119 |
+
},
|
| 120 |
+
"pehe": {
|
| 121 |
+
"mean": 18169.43251953125,
|
| 122 |
+
"std": 1294.96700572562
|
| 123 |
+
},
|
| 124 |
+
"uplift_auroc": {
|
| 125 |
+
"mean": NaN,
|
| 126 |
+
"std": NaN
|
| 127 |
+
}
|
| 128 |
+
},
|
| 129 |
+
"business_metrics": {
|
| 130 |
+
"decision_accuracy": {
|
| 131 |
+
"mean": 0.8097014925373134,
|
| 132 |
+
"std": 0.02245016262162883
|
| 133 |
+
},
|
| 134 |
+
"policy_value": {
|
| 135 |
+
"mean": 1026.2975876057826,
|
| 136 |
+
"std": 237.22889389801313
|
| 137 |
+
},
|
| 138 |
+
"policy_value_random": {
|
| 139 |
+
"mean": 19843.8970703125,
|
| 140 |
+
"std": 864.087324165919
|
| 141 |
+
},
|
| 142 |
+
"policy_value_recommended": {
|
| 143 |
+
"mean": 20870.19465791828,
|
| 144 |
+
"std": 892.2700421091743
|
| 145 |
+
},
|
| 146 |
+
"cost_effectiveness": {
|
| 147 |
+
"mean": NaN,
|
| 148 |
+
"std": NaN
|
| 149 |
+
}
|
| 150 |
+
},
|
| 151 |
+
"time_s": 5.124366804026067
|
| 152 |
+
}
|
| 153 |
+
}
|
| 154 |
+
}
|