millid-medium-v1 / reports /summary.json
byamasupatrick's picture
Evaluated again on baobabtech/millid-dataset
c24c6bc verified
Raw History Blame Contribute Delete
17.7 kB
{
"validation": {
"millid": {
"top1_accuracy": 0.9612852766734016,
"top3_accuracy": 0.990952714592974,
"top5_accuracy": 0.994940689565274,
"passages": 245488,
"languages": 1626,
"macro_precision": 0.95728323988777,
"macro_recall": 0.9621342053274924,
"macro_f1": 0.9583620746332984,
"macro_specificity": 0.9999762003050453,
"weighted_f1": 0.961335355655093,
"median_f1": 0.9850028835063437,
"languages_f1_at_least_0_9": 1462,
"languages_f1_below_0_5": 7,
"mean_fp_per_million": 23.799694954832386,
"micro_f1": 0.9612852766734016,
"top1_by_size": {
"paragraph": 0.9857572558315572,
"sentence": 0.9817676528501283,
"short": 0.9363205702116563
}
}
},
"test": {
"millid": {
"top1_accuracy": 0.9616468462918486,
"top3_accuracy": 0.9910007034602754,
"top5_accuracy": 0.9945180048368097,
"passages": 240241,
"languages": 1621,
"macro_precision": 0.9564228576676835,
"macro_recall": 0.9614016943550959,
"macro_f1": 0.9575764452347921,
"macro_specificity": 0.9999763662862723,
"weighted_f1": 0.9616663642965525,
"median_f1": 0.9841269841269842,
"languages_f1_at_least_0_9": 1451,
"languages_f1_below_0_5": 6,
"mean_fp_per_million": 23.633713727639183,
"micro_f1": 0.9616468462918486,
"top1_by_size": {
"paragraph": 0.9846028678938269,
"sentence": 0.9808588082571843,
"short": 0.9381966542817016
}
},
"glotlid": {
"top1_accuracy": 0.977614145795264,
"top3_accuracy": 0.9977189572137978,
"top5_accuracy": 0.9983766301339072,
"passages": 240241,
"languages": 1621,
"macro_precision": 0.9782517709855841,
"macro_recall": 0.9751782136320077,
"macro_f1": 0.9748179587217549,
"macro_specificity": 0.9999867952157099,
"weighted_f1": 0.9776092036050326,
"median_f1": 1.0,
"languages_f1_at_least_0_9": 1488,
"languages_f1_below_0_5": 6,
"mean_fp_per_million": 13.204784290051938,
"micro_f1": 0.977614145795264,
"top1_by_size": {
"paragraph": 0.9910708837587715,
"sentence": 0.9852877542189529,
"short": 0.9663739311301133
}
},
"glotlid@millid": {
"top1_accuracy": 0.978292631149554,
"top3_accuracy": 0.9977897194900122,
"top5_accuracy": 0.9984557173837938,
"passages": 240241,
"languages": 1621,
"macro_precision": 0.9778492893480586,
"macro_recall": 0.9764998669631992,
"macro_f1": 0.9754780061637239,
"macro_specificity": 0.9999866076235475,
"weighted_f1": 0.9778914257911913,
"median_f1": 1.0,
"languages_f1_at_least_0_9": 1490,
"languages_f1_below_0_5": 5,
"mean_fp_per_million": 13.392376452534537,
"micro_f1": 0.978292631149554,
"top1_by_size": {
"paragraph": 0.9911725821214279,
"sentence": 0.98578410161122,
"short": 0.967431690103287
}
},
"fasttext": {
"top1_accuracy": 0.25110201838986684,
"top3_accuracy": 0.2610045745730329,
"top5_accuracy": 0.262482257399861,
"passages": 240241,
"languages": 1621,
"macro_precision": 0.05202750834127428,
"macro_recall": 0.12048375928156996,
"macro_f1": 0.06567764337804372,
"macro_specificity": 0.9995388522643763,
"weighted_f1": 0.14972995387859817,
"median_f1": 0.0,
"languages_f1_at_least_0_9": 25,
"languages_f1_below_0_5": 1518,
"mean_fp_per_million": 461.14773562356646,
"micro_f1": 0.25110201838986684,
"top1_by_size": {
"paragraph": 0.19578968778602665,
"sentence": 0.2500699977091659,
"short": 0.2759950934205614
}
},
"fasttext@millid": {
"top1_accuracy": 0.2514599922577745,
"top3_accuracy": 0.2610753368492472,
"top5_accuracy": 0.2625197197813862,
"passages": 240241,
"languages": 1621,
"macro_precision": 0.05205991842533006,
"macro_recall": 0.12089826722608317,
"macro_f1": 0.06593662879256669,
"macro_specificity": 0.9995376539627219,
"weighted_f1": 0.1498625447047829,
"median_f1": 0.0,
"languages_f1_at_least_0_9": 26,
"languages_f1_below_0_5": 1517,
"mean_fp_per_million": 462.34603727812436,
"micro_f1": 0.2514599922577745,
"top1_by_size": {
"paragraph": 0.19587104647615172,
"sentence": 0.25056634510143305,
"short": 0.27637730884784273
}
}
},
"test-common": {
"millid": {
"top1_accuracy": 0.9424741087047386,
"top3_accuracy": 0.988172825363407,
"top5_accuracy": 0.9923525860334071,
"passages": 64597,
"languages": 213,
"macro_precision": 0.9732280825173869,
"macro_recall": 0.9360723187927141,
"macro_f1": 0.9527046866271709,
"macro_specificity": 0.9999026288484889,
"weighted_f1": 0.9592398159015416,
"median_f1": 0.9738219895287958,
"languages_f1_at_least_0_9": 191,
"languages_f1_below_0_5": 0,
"mean_fp_per_million": 97.37115151108515,
"micro_f1": 0.9424741087047386,
"top1_by_size": {
"paragraph": 0.9737317024263084,
"sentence": 0.9696320809811174,
"short": 0.9169479090242113
}
},
"glotlid": {
"top1_accuracy": 0.9738223137297398,
"top3_accuracy": 0.9924609501989257,
"top5_accuracy": 0.9946437141043701,
"passages": 64597,
"languages": 213,
"macro_precision": 0.9924377815827973,
"macro_recall": 0.9770740040297234,
"macro_f1": 0.9835053354686013,
"macro_specificity": 0.9999660547959016,
"weighted_f1": 0.9823407826125089,
"median_f1": 0.9982788296041308,
"languages_f1_at_least_0_9": 201,
"languages_f1_below_0_5": 0,
"mean_fp_per_million": 33.94520409829502,
"micro_f1": 0.9738223137297398,
"top1_by_size": {
"paragraph": 0.9946861840786043,
"sentence": 0.9876386996301343,
"short": 0.9593837123991196
}
},
"glotlid@millid": {
"top1_accuracy": 0.9749369165750732,
"top3_accuracy": 0.9926931591250368,
"top5_accuracy": 0.994829481245259,
"passages": 64597,
"languages": 213,
"macro_precision": 0.9922871273519716,
"macro_recall": 0.9790381188248578,
"macro_f1": 0.9845885184031669,
"macro_specificity": 0.9999656167342897,
"weighted_f1": 0.9829375630604269,
"median_f1": 0.9983471074380166,
"languages_f1_at_least_0_9": 203,
"languages_f1_below_0_5": 0,
"mean_fp_per_million": 34.38326571019212,
"micro_f1": 0.9749369165750732,
"top1_by_size": {
"paragraph": 0.9948867054341287,
"sentence": 0.9885146972941405,
"short": 0.9609097578870139
}
},
"fasttext": {
"top1_accuracy": 0.9338668978435531,
"top3_accuracy": 0.9706952335247767,
"top5_accuracy": 0.9761908447760732,
"passages": 64597,
"languages": 213,
"macro_precision": 0.9356570616737796,
"macro_recall": 0.9169210037343893,
"macro_f1": 0.9148824189777787,
"macro_specificity": 0.9996959024393602,
"weighted_f1": 0.9352000319739436,
"median_f1": 0.9680365296803652,
"languages_f1_at_least_0_9": 165,
"languages_f1_below_0_5": 8,
"mean_fp_per_million": 304.0975606397608,
"micro_f1": 0.9338668978435531,
"top1_by_size": {
"paragraph": 0.9651092841387607,
"sentence": 0.9562487833365778,
"short": 0.9112252384446075
}
},
"fasttext@millid": {
"top1_accuracy": 0.9351982290199236,
"top3_accuracy": 0.9709584036410359,
"top5_accuracy": 0.9763301701317398,
"passages": 64597,
"languages": 213,
"macro_precision": 0.9353710554982398,
"macro_recall": 0.9200755454154029,
"macro_f1": 0.9173068426975766,
"macro_specificity": 0.9996942142238651,
"weighted_f1": 0.9361359929041825,
"median_f1": 0.9680365296803652,
"languages_f1_at_least_0_9": 166,
"languages_f1_below_0_5": 7,
"mean_fp_per_million": 305.78577613496856,
"micro_f1": 0.9351982290199236,
"top1_by_size": {
"paragraph": 0.9655103268498095,
"sentence": 0.958146778275258,
"short": 0.9124871606749817
}
}
},
"udhr": {
"millid": {
"top1_accuracy": 0.8524028910773913,
"top3_accuracy": 0.9194386914070756,
"top5_accuracy": 0.9340208800033813,
"passages": 23659,
"languages": 358,
"macro_precision": 0.9186545076419138,
"macro_recall": 0.8769388469921001,
"macro_f1": 0.8779986710808113,
"macro_specificity": 0.9998608120725921,
"weighted_f1": 0.8598114286766851,
"median_f1": 0.9909909909909909,
"languages_f1_at_least_0_9": 282,
"languages_f1_below_0_5": 35,
"mean_fp_per_million": 139.18792740793418,
"micro_f1": 0.8524028910773913
},
"glotlid": {
"top1_accuracy": 0.8598419206221734,
"top3_accuracy": 0.9279766684982459,
"top5_accuracy": 0.9420516505346802,
"passages": 23659,
"languages": 358,
"macro_precision": 0.9261030959501081,
"macro_recall": 0.8825915640670888,
"macro_f1": 0.8864649884094992,
"macro_specificity": 0.999887104475242,
"weighted_f1": 0.8701800337416647,
"median_f1": 0.9915248150542268,
"languages_f1_at_least_0_9": 287,
"languages_f1_below_0_5": 32,
"mean_fp_per_million": 112.89552475798357,
"micro_f1": 0.8598419206221734
},
"glotlid@millid": {
"top1_accuracy": 0.8614480747284331,
"top3_accuracy": 0.9294137537512152,
"top5_accuracy": 0.943319666934359,
"passages": 23659,
"languages": 358,
"macro_precision": 0.9257695969546421,
"macro_recall": 0.8844312457708232,
"macro_f1": 0.8873151713293488,
"macro_specificity": 0.9998860393705818,
"weighted_f1": 0.8709300475765809,
"median_f1": 0.9915248150542268,
"languages_f1_at_least_0_9": 291,
"languages_f1_below_0_5": 32,
"mean_fp_per_million": 113.96062941822754,
"micro_f1": 0.8614480747284331
},
"fasttext": {
"top1_accuracy": 0.4624878481761697,
"top3_accuracy": 0.4856502810769686,
"top5_accuracy": 0.49034194175578005,
"passages": 23659,
"languages": 358,
"macro_precision": 0.31521227915727285,
"macro_recall": 0.4478627321056064,
"macro_f1": 0.3496561644788506,
"macro_specificity": 0.9986965572826259,
"weighted_f1": 0.3709499799103122,
"median_f1": 0.0,
"languages_f1_at_least_0_9": 60,
"languages_f1_below_0_5": 220,
"mean_fp_per_million": 1303.4427173741606,
"micro_f1": 0.4624878481761697
},
"fasttext@millid": {
"top1_accuracy": 0.4628682530960734,
"top3_accuracy": 0.4856502810769686,
"top5_accuracy": 0.4904264761824253,
"passages": 23659,
"languages": 358,
"macro_precision": 0.3151627394343268,
"macro_recall": 0.4483116546914005,
"macro_f1": 0.34986047374808604,
"macro_specificity": 0.9986947805299987,
"weighted_f1": 0.37111316345366563,
"median_f1": 0.0,
"languages_f1_at_least_0_9": 61,
"languages_f1_below_0_5": 220,
"mean_fp_per_million": 1305.2194700013772,
"micro_f1": 0.4628682530960734
}
},
"udhr-common": {
"millid": {
"top1_accuracy": 0.9120906801007557,
"top3_accuracy": 0.9633081444164567,
"top5_accuracy": 0.9688497061293031,
"passages": 11910,
"languages": 170,
"macro_precision": 0.9770201360820081,
"macro_recall": 0.9382111759192128,
"macro_f1": 0.947989665742032,
"macro_specificity": 0.9998980853793291,
"weighted_f1": 0.9268326524022035,
"median_f1": 0.9955099164401491,
"languages_f1_at_least_0_9": 153,
"languages_f1_below_0_5": 5,
"mean_fp_per_million": 101.91462067096401,
"micro_f1": 0.9120906801007557
},
"glotlid": {
"top1_accuracy": 0.9335852225020991,
"top3_accuracy": 0.9677581863979848,
"top5_accuracy": 0.973215785054576,
"passages": 11910,
"languages": 170,
"macro_precision": 0.9887806114106193,
"macro_recall": 0.9575995164374027,
"macro_f1": 0.9657221446088623,
"macro_specificity": 0.9999338492838657,
"weighted_f1": 0.9467136106728533,
"median_f1": 1.0,
"languages_f1_at_least_0_9": 155,
"languages_f1_below_0_5": 3,
"mean_fp_per_million": 66.15071613431165,
"micro_f1": 0.9335852225020991
},
"glotlid@millid": {
"top1_accuracy": 0.9338371116708648,
"top3_accuracy": 0.9682619647355164,
"top5_accuracy": 0.9735516372795969,
"passages": 11910,
"languages": 170,
"macro_precision": 0.9887806114106193,
"macro_recall": 0.9578604445634422,
"macro_f1": 0.9658801542070193,
"macro_specificity": 0.9999338492838657,
"weighted_f1": 0.9468741038422634,
"median_f1": 1.0,
"languages_f1_at_least_0_9": 156,
"languages_f1_below_0_5": 3,
"mean_fp_per_million": 66.15071613431165,
"micro_f1": 0.9338371116708648
},
"fasttext": {
"top1_accuracy": 0.9187237615449202,
"top3_accuracy": 0.964735516372796,
"top5_accuracy": 0.9740554156171285,
"passages": 11910,
"languages": 170,
"macro_precision": 0.9473808585265043,
"macro_recall": 0.9431462240812183,
"macro_f1": 0.9323102708833079,
"macro_specificity": 0.9996006150778249,
"weighted_f1": 0.917761071560676,
"median_f1": 1.0,
"languages_f1_at_least_0_9": 141,
"languages_f1_below_0_5": 6,
"mean_fp_per_million": 399.3849221751375,
"micro_f1": 0.9187237615449202
},
"fasttext@millid": {
"top1_accuracy": 0.9194794290512175,
"top3_accuracy": 0.964735516372796,
"top5_accuracy": 0.974223341729639,
"passages": 11910,
"languages": 170,
"macro_precision": 0.9473808585265043,
"macro_recall": 0.9440916022324787,
"macro_f1": 0.9328542536140149,
"macro_specificity": 0.9996006150778249,
"weighted_f1": 0.9181958923496206,
"median_f1": 1.0,
"languages_f1_at_least_0_9": 142,
"languages_f1_below_0_5": 6,
"mean_fp_per_million": 399.3849221751375,
"micro_f1": 0.9194794290512175
}
},
"udhr-all": {
"millid": {
"top1_accuracy": 0.7300800057922745,
"top3_accuracy": 0.7874959273069544,
"top5_accuracy": 0.7999855193136155,
"passages": 27623,
"languages": 429,
"macro_precision": 0.7415286668963516,
"macro_recall": 0.731804445741659,
"macro_f1": 0.718364827833673,
"macro_specificity": 0.9998029838447381,
"weighted_f1": 0.722839631537149,
"median_f1": 0.9672131147540983,
"languages_f1_at_least_0_9": 266,
"languages_f1_below_0_5": 108,
"mean_fp_per_million": 197.01615526182962,
"micro_f1": 0.7300800057922745
},
"glotlid": {
"top1_accuracy": 0.750244361582739,
"top3_accuracy": 0.8144662056981501,
"top5_accuracy": 0.8281866560474966,
"passages": 27623,
"languages": 429,
"macro_precision": 0.7694575973118034,
"macro_recall": 0.7529055089081532,
"macro_f1": 0.7454327849166208,
"macro_specificity": 0.9998410395298513,
"weighted_f1": 0.7485822543592835,
"median_f1": 0.9795918367346939,
"languages_f1_at_least_0_9": 277,
"languages_f1_below_0_5": 97,
"mean_fp_per_million": 158.96047014874904,
"micro_f1": 0.750244361582739
},
"glotlid@millid": {
"top1_accuracy": 0.7378271730080006,
"top3_accuracy": 0.7960395322738297,
"top5_accuracy": 0.8079498968251095,
"passages": 27623,
"languages": 429,
"macro_precision": 0.7425329775774382,
"macro_recall": 0.7380568437901042,
"macro_f1": 0.724506747472521,
"macro_specificity": 0.999809579502659,
"weighted_f1": 0.7305913692421664,
"median_f1": 0.9747899159663865,
"languages_f1_at_least_0_9": 272,
"languages_f1_below_0_5": 105,
"mean_fp_per_million": 190.42049734090864,
"micro_f1": 0.7378271730080006
},
"fasttext": {
"top1_accuracy": 0.3961191760489447,
"top3_accuracy": 0.4159577163957572,
"top5_accuracy": 0.4199761068674655,
"passages": 27623,
"languages": 429,
"macro_precision": 0.24791077144530446,
"macro_recall": 0.3737409279575923,
"macro_f1": 0.279501714390051,
"macro_specificity": 0.9988210740543896,
"weighted_f1": 0.30525958931365826,
"median_f1": 0.0,
"languages_f1_at_least_0_9": 54,
"languages_f1_below_0_5": 299,
"mean_fp_per_million": 1178.9259456102943,
"micro_f1": 0.3961191760489447
},
"fasttext@millid": {
"top1_accuracy": 0.39644499149259677,
"top3_accuracy": 0.4159577163957572,
"top5_accuracy": 0.42004851029938817,
"passages": 27623,
"languages": 429,
"macro_precision": 0.2478791005299098,
"macro_recall": 0.3741155533322177,
"macro_f1": 0.27967804305179117,
"macro_specificity": 0.998819804723614,
"weighted_f1": 0.3054055953686982,
"median_f1": 0.0,
"languages_f1_at_least_0_9": 55,
"languages_f1_below_0_5": 299,
"mean_fp_per_million": 1180.1952763859204,
"micro_f1": 0.39644499149259677
}
}
}