minhleduc's picture
e2_rlce_sigma1_fixed_seed44: test_eval_v2.json
b4f00e5 verified
Raw History Blame Contribute Delete
2.94 kB
{
"raw": {
"ece": 0.08006285257637501,
"brier": 0.06528607002967048,
"nll": 0.8893808082595933,
"accuracy": 0.762,
"soft_accuracy": 0.5355462909936904,
"mean_confidence": 0.6835951461046934,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.23656083203619346,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.0807683249314626,
"brier": 0.08273597999323784,
"nll": 1.0243752450713266,
"accuracy": 0.74,
"soft_accuracy": 0.4928574220339457,
"mean_confidence": 0.6668512896696727,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.08953602485358715,
"brier": 0.0701229432496757,
"nll": 1.0710554734291509,
"accuracy": 0.70875,
"soft_accuracy": 0.45566242933273315,
"mean_confidence": 0.6242463922873139,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.23656083203619346,
"n": 800
},
"noul": {
"ece": 0.07552934775749844,
"brier": 0.041386995772762794,
"nll": 0.5121534845551166,
"accuracy": 0.855,
"soft_accuracy": 0.6847469755013784,
"mean_confidence": 0.7794706742962202,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.135570114582777,
"brier": 0.05432780923554507,
"nll": 0.8649242109209299,
"accuracy": 0.762,
"soft_accuracy": 0.5100282768830657,
"mean_confidence": 0.6264995142370462,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.2327637035213411,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.14227418303489683,
"brier": 0.06847539794888387,
"nll": 0.9953496947139502,
"accuracy": 0.74,
"soft_accuracy": 0.4617195088416338,
"mean_confidence": 0.5977258223791917,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.1510212018340826,
"brier": 0.057034744908633,
"nll": 1.0409022262692451,
"accuracy": 0.70875,
"soft_accuracy": 0.42901945907622574,
"mean_confidence": 0.5610361216962337,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.2327637035213411,
"n": 800
},
"noul": {
"ece": 0.11244227200746537,
"brier": 0.03657097295808903,
"nll": 0.499861373330156,
"accuracy": 0.855,
"soft_accuracy": 0.6663488020002842,
"mean_confidence": 0.7425577294826508,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type1_bucket1": 1.3143467903137207,
"type0_bucket1": 1.3149399757385254,
"type2_bucket0": 1.3414469957351685
}
}