minhleduc's picture
e2_rl_only_seed42: test_eval_v2.json
e4ed5bd verified
Raw History Blame Contribute Delete
2.94 kB
{
"raw": {
"ece": 0.12679152297228574,
"brier": 0.05535062948582825,
"nll": 0.8679260851321742,
"accuracy": 0.76,
"soft_accuracy": 0.5145708695724607,
"mean_confidence": 0.6345017634481191,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.2358442002814263,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.12754969368378322,
"brier": 0.06870347295043758,
"nll": 0.9975430295554301,
"accuracy": 0.73,
"soft_accuracy": 0.4690246512244145,
"mean_confidence": 0.6079408426582813,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.15719266310334207,
"brier": 0.05798107115449511,
"nll": 1.044197057574056,
"accuracy": 0.71875,
"soft_accuracy": 0.43173864085227254,
"mean_confidence": 0.5640414332598448,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.2358442002814263,
"n": 800
},
"noul": {
"ece": 0.10082647318641345,
"brier": 0.03849053046299644,
"nll": 0.5032811774530759,
"accuracy": 0.845,
"soft_accuracy": 0.6705600595474244,
"mean_confidence": 0.7550097911556561,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.14100187775492665,
"brier": 0.05418899545646501,
"nll": 0.8656314948331565,
"accuracy": 0.76,
"soft_accuracy": 0.5069824938178062,
"mean_confidence": 0.6189981058388948,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.2364325567521155,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.140854502171278,
"brier": 0.06703519375523986,
"nll": 0.9948613907334706,
"accuracy": 0.73,
"soft_accuracy": 0.46041965164244175,
"mean_confidence": 0.5903157998124758,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.1665362735465169,
"brier": 0.05691809966142955,
"nll": 1.0416682769916952,
"accuracy": 0.71875,
"soft_accuracy": 0.4263803570531309,
"mean_confidence": 0.552213729210198,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.2364325567521155,
"n": 800
},
"noul": {
"ece": 0.12232253074645996,
"brier": 0.037703991551070765,
"nll": 0.5016858893881242,
"accuracy": 0.845,
"soft_accuracy": 0.6610148516794045,
"mean_confidence": 0.7367262473702431,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type0_bucket1": 1.0745879411697388,
"type2_bucket0": 1.1491918563842773,
"type1_bucket1": 1.0556284189224243
}
}