Upload results.json with huggingface_hub
Browse files- results.json +63 -43
results.json
CHANGED
|
@@ -25,7 +25,7 @@
|
|
| 25 |
"epochs": 30,
|
| 26 |
"seed": 42,
|
| 27 |
"fitness_formula": "0.4 * mlm_score + 0.6 * debias_score",
|
| 28 |
-
"L_0": 0.
|
| 29 |
"pool_backends": [
|
| 30 |
"LocalCudaBackend",
|
| 31 |
"RemoteSSHBackend",
|
|
@@ -33,71 +33,91 @@
|
|
| 33 |
],
|
| 34 |
"variants": [
|
| 35 |
{
|
| 36 |
-
"val_mlm_loss": 0.
|
| 37 |
-
"val_adv_accuracy": 0.
|
| 38 |
"num_competitions": 22.0,
|
| 39 |
"chance": 0.045454545454545456,
|
| 40 |
-
"leakage": 0.
|
| 41 |
-
"debias_score": 0.
|
| 42 |
-
"mlm_score":
|
| 43 |
-
"fitness": 0.
|
| 44 |
-
"param_count":
|
| 45 |
-
"training_time_seconds":
|
| 46 |
"epochs_trained": 16.0,
|
| 47 |
-
"variant": "
|
| 48 |
-
"program_path": "
|
| 49 |
-
"wall_clock_seconds":
|
| 50 |
},
|
| 51 |
{
|
| 52 |
-
"val_mlm_loss": 0.
|
| 53 |
"val_adv_accuracy": 0.21164623467600702,
|
| 54 |
"num_competitions": 22.0,
|
| 55 |
"chance": 0.045454545454545456,
|
| 56 |
"leakage": 0.1741055791843883,
|
| 57 |
"debias_score": 0.8258944208156117,
|
| 58 |
-
"mlm_score": 0.
|
| 59 |
-
"fitness": 0.
|
| 60 |
-
"param_count":
|
| 61 |
-
"training_time_seconds":
|
| 62 |
"epochs_trained": 16.0,
|
| 63 |
-
"variant": "
|
| 64 |
-
"program_path": "
|
| 65 |
-
"wall_clock_seconds":
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"
|
| 69 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 70 |
"variant": "deep_mlp_3layer",
|
| 71 |
"program_path": "deep_mlp_3layer.py",
|
| 72 |
-
"wall_clock_seconds":
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"combined_score": 0.0,
|
| 76 |
-
"error": 1.0,
|
| 77 |
-
"variant": "cross_attention_adversary",
|
| 78 |
-
"program_path": "cross_attention_adversary.py",
|
| 79 |
-
"wall_clock_seconds": 63.32799999997951
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"
|
| 83 |
-
"
|
| 84 |
-
"
|
| 85 |
-
"
|
| 86 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"
|
| 90 |
-
"
|
| 91 |
-
"
|
| 92 |
-
"
|
| 93 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"combined_score": 0.0,
|
| 97 |
"error": 1.0,
|
| 98 |
-
"variant": "
|
| 99 |
-
"program_path": "
|
| 100 |
-
"wall_clock_seconds":
|
| 101 |
}
|
| 102 |
]
|
| 103 |
}
|
|
|
|
| 25 |
"epochs": 30,
|
| 26 |
"seed": 42,
|
| 27 |
"fitness_formula": "0.4 * mlm_score + 0.6 * debias_score",
|
| 28 |
+
"L_0": 0.7413,
|
| 29 |
"pool_backends": [
|
| 30 |
"LocalCudaBackend",
|
| 31 |
"RemoteSSHBackend",
|
|
|
|
| 33 |
],
|
| 34 |
"variants": [
|
| 35 |
{
|
| 36 |
+
"val_mlm_loss": 0.8816994256443448,
|
| 37 |
+
"val_adv_accuracy": 0.09728546409807355,
|
| 38 |
"num_competitions": 22.0,
|
| 39 |
"chance": 0.045454545454545456,
|
| 40 |
+
"leakage": 0.05429905762655324,
|
| 41 |
+
"debias_score": 0.9457009423734468,
|
| 42 |
+
"mlm_score": 0.840762711689711,
|
| 43 |
+
"fitness": 0.9037256500999524,
|
| 44 |
+
"param_count": 2102488.0,
|
| 45 |
+
"training_time_seconds": 7532.348341790959,
|
| 46 |
"epochs_trained": 16.0,
|
| 47 |
+
"variant": "cross_attention_adversary",
|
| 48 |
+
"program_path": "cross_attention_adversary.py",
|
| 49 |
+
"wall_clock_seconds": 7539.859999999986
|
| 50 |
},
|
| 51 |
{
|
| 52 |
+
"val_mlm_loss": 0.7490579379929436,
|
| 53 |
"val_adv_accuracy": 0.21164623467600702,
|
| 54 |
"num_competitions": 22.0,
|
| 55 |
"chance": 0.045454545454545456,
|
| 56 |
"leakage": 0.1741055791843883,
|
| 57 |
"debias_score": 0.8258944208156117,
|
| 58 |
+
"mlm_score": 0.9896430735201464,
|
| 59 |
+
"fitness": 0.8913938818974256,
|
| 60 |
+
"param_count": 2028589.0,
|
| 61 |
+
"training_time_seconds": 5628.0779999999795,
|
| 62 |
"epochs_trained": 16.0,
|
| 63 |
+
"variant": "residual_mlp",
|
| 64 |
+
"program_path": "residual_mlp.py",
|
| 65 |
+
"wall_clock_seconds": 5629.265999999945
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"val_mlm_loss": 0.773209982448154,
|
| 69 |
+
"val_adv_accuracy": 0.21164623467600702,
|
| 70 |
+
"num_competitions": 22.0,
|
| 71 |
+
"chance": 0.045454545454545456,
|
| 72 |
+
"leakage": 0.1741055791843883,
|
| 73 |
+
"debias_score": 0.8258944208156117,
|
| 74 |
+
"mlm_score": 0.9587305089529238,
|
| 75 |
+
"fitness": 0.8790288560705366,
|
| 76 |
+
"param_count": 2103277.0,
|
| 77 |
+
"training_time_seconds": 4988.107975777995,
|
| 78 |
+
"epochs_trained": 16.0,
|
| 79 |
"variant": "deep_mlp_3layer",
|
| 80 |
"program_path": "deep_mlp_3layer.py",
|
| 81 |
+
"wall_clock_seconds": 4993.25
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"val_mlm_loss": 0.7801726990275912,
|
| 85 |
+
"val_adv_accuracy": 0.21164623467600702,
|
| 86 |
+
"num_competitions": 22.0,
|
| 87 |
+
"chance": 0.045454545454545456,
|
| 88 |
+
"leakage": 0.1741055791843883,
|
| 89 |
+
"debias_score": 0.8258944208156117,
|
| 90 |
+
"mlm_score": 0.9501742382474517,
|
| 91 |
+
"fitness": 0.8756063477883478,
|
| 92 |
+
"param_count": 1995779.0,
|
| 93 |
+
"training_time_seconds": 7390.658450284973,
|
| 94 |
+
"epochs_trained": 16.0,
|
| 95 |
+
"variant": "dual_head_ensemble",
|
| 96 |
+
"program_path": "dual_head_ensemble.py",
|
| 97 |
+
"wall_clock_seconds": 7397.609000000055
|
| 98 |
},
|
| 99 |
{
|
| 100 |
+
"val_mlm_loss": 0.783913352754381,
|
| 101 |
+
"val_adv_accuracy": 0.21164623467600702,
|
| 102 |
+
"num_competitions": 22.0,
|
| 103 |
+
"chance": 0.045454545454545456,
|
| 104 |
+
"leakage": 0.1741055791843883,
|
| 105 |
+
"debias_score": 0.8258944208156117,
|
| 106 |
+
"mlm_score": 0.9456402259195439,
|
| 107 |
+
"fitness": 0.8737927428571846,
|
| 108 |
+
"param_count": 1991533.0,
|
| 109 |
+
"training_time_seconds": 5025.6720000000205,
|
| 110 |
+
"epochs_trained": 16.0,
|
| 111 |
+
"variant": "deep_mlp_2layer",
|
| 112 |
+
"program_path": "deep_mlp_2layer.py",
|
| 113 |
+
"wall_clock_seconds": 5028.8279999999795
|
| 114 |
},
|
| 115 |
{
|
| 116 |
"combined_score": 0.0,
|
| 117 |
"error": 1.0,
|
| 118 |
+
"variant": "attention_pool_head",
|
| 119 |
+
"program_path": "attention_pool_head.py",
|
| 120 |
+
"wall_clock_seconds": 10803.032000000007
|
| 121 |
}
|
| 122 |
]
|
| 123 |
}
|