minhleduc's picture
E2 run e2_rlce_sigma1_fixed_seed44 (epoch_4.pt)
c4c806b verified
Raw History Blame Contribute Delete
2.62 kB
{
"raw": {
"ece": 0.07792394097149372,
"brier": 0.0651938930230446,
"nll": 0.889134444310097,
"accuracy": 0.761,
"soft_accuracy": 0.5354805619120598,
"mean_confidence": 0.6834244523942471,
"score_mae": 0.23642302882857621,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.07670323486129442,
"brier": 0.0824239765906547,
"nll": 1.0238185043431198,
"accuracy": 0.7383333333333333,
"soft_accuracy": 0.4928270164380471,
"mean_confidence": 0.6666522144277891,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.08631576582789423,
"brier": 0.07005521307971321,
"nll": 1.0707776059722527,
"accuracy": 0.71,
"soft_accuracy": 0.4555477532558143,
"mean_confidence": 0.6238887127488851,
"score_mae": 0.23642302882857621,
"n": 800
},
"noul": {
"ece": 0.07208900928497315,
"brier": 0.04148204937987631,
"nll": 0.5122595020608666,
"accuracy": 0.8516666666666667,
"soft_accuracy": 0.6847111855943998,
"mean_confidence": 0.7795776765545209,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.1343130425512791,
"brier": 0.05431590324572952,
"nll": 0.8648814527792856,
"accuracy": 0.761,
"soft_accuracy": 0.5101195223405958,
"mean_confidence": 0.6266869630515576,
"score_mae": 0.23274893001886085,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.1416614914437135,
"brier": 0.06833023387277838,
"nll": 0.9951477930446466,
"accuracy": 0.7383333333333333,
"soft_accuracy": 0.4618977982302507,
"mean_confidence": 0.5980274300773939,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.15051539737731218,
"brier": 0.05705952123160387,
"nll": 1.0408778808638453,
"accuracy": 0.71,
"soft_accuracy": 0.42907225193455817,
"mean_confidence": 0.5610981351137161,
"score_mae": 0.23274893001886085,
"n": 800
},
"noul": {
"ece": 0.10886838714281719,
"brier": 0.03664341530418151,
"nll": 0.4999532084011783,
"accuracy": 0.8516666666666667,
"soft_accuracy": 0.6664042736589909,
"mean_confidence": 0.7427982666095098,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type1_bucket1": 1.3121464252471924,
"type0_bucket1": 1.3120508193969727,
"type2_bucket0": 1.3396928310394287
}
}