{ "validation": { "millid": { "top1_accuracy": 0.9577820504464577, "top3_accuracy": 0.9888874405266245, "top5_accuracy": 0.9932746203480415, "passages": 245488, "languages": 1626, "macro_precision": 0.9538820330004751, "macro_recall": 0.9590954075456577, "macro_f1": 0.9550536411106632, "macro_specificity": 0.9999740290832557, "weighted_f1": 0.9579611590588513, "median_f1": 0.9822485207100591, "languages_f1_at_least_0_9": 1435, "languages_f1_below_0_5": 8, "mean_fp_per_million": 25.97091674438312, "micro_f1": 0.9577820504464577, "top1_by_size": { "paragraph": 0.984960458955001, "sentence": 0.9788390119887341, "short": 0.9312356034595158 } } }, "test": { "millid": { "top1_accuracy": 0.9587247805328816, "top3_accuracy": 0.9890484971341278, "top5_accuracy": 0.9932401213781161, "passages": 240241, "languages": 1621, "macro_precision": 0.9549770065582561, "macro_recall": 0.9588167712047285, "macro_f1": 0.9553762683202616, "macro_specificity": 0.9999745397211366, "weighted_f1": 0.9589133936122667, "median_f1": 0.9824561403508771, "languages_f1_at_least_0_9": 1446, "languages_f1_below_0_5": 8, "mean_fp_per_million": 25.46027886332467, "micro_f1": 0.9587247805328816, "top1_by_size": { "paragraph": 0.9832807891792942, "sentence": 0.9797006643418943, "short": 0.933343407228316 } }, "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": 0.9395947180209607, "top3_accuracy": 0.9856649689614069, "top5_accuracy": 0.9908664489062959, "passages": 64597, "languages": 213, "macro_precision": 0.970922775379856, "macro_recall": 0.9353270477650256, "macro_f1": 0.951259698841108, "macro_specificity": 0.9998952436560236, "weighted_f1": 0.9570662852696624, "median_f1": 0.9726495726495726, "languages_f1_at_least_0_9": 188, "languages_f1_below_0_5": 0, "mean_fp_per_million": 104.75634397647721, "micro_f1": 0.9395947180209607, "top1_by_size": { "paragraph": 0.9725285742931622, "sentence": 0.9684640840957758, "short": 0.9125458547322084 } }, "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": 0.8492328500781944, "top3_accuracy": 0.911703791369035, "top5_accuracy": 0.9256519717655015, "passages": 23659, "languages": 358, "macro_precision": 0.914803800630921, "macro_recall": 0.8736275490421017, "macro_f1": 0.8764285216392155, "macro_specificity": 0.9998609432354121, "weighted_f1": 0.8584797362391883, "median_f1": 0.9906542056074767, "languages_f1_at_least_0_9": 274, "languages_f1_below_0_5": 32, "mean_fp_per_million": 139.05676458798044, "micro_f1": 0.8492328500781944 }, "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": 0.9123425692695214, "top3_accuracy": 0.9606213266162889, "top5_accuracy": 0.9654911838790932, "passages": 11910, "languages": 170, "macro_precision": 0.9794271783100589, "macro_recall": 0.9393117513722192, "macro_f1": 0.9512636942896991, "macro_specificity": 0.9999174355030307, "weighted_f1": 0.92868171659548, "median_f1": 1.0, "languages_f1_at_least_0_9": 149, "languages_f1_below_0_5": 4, "mean_fp_per_million": 82.56449696922371, "micro_f1": 0.9123425692695214 }, "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": 0.7273648770951743, "top3_accuracy": 0.7808710132860297, "top5_accuracy": 0.7928175795532708, "passages": 27623, "languages": 429, "macro_precision": 0.7340309394117082, "macro_recall": 0.7290411714617072, "macro_f1": 0.7148579318843858, "macro_specificity": 0.9997903094250377, "weighted_f1": 0.7196404590366388, "median_f1": 0.9649122807017544, "languages_f1_at_least_0_9": 254, "languages_f1_below_0_5": 105, "mean_fp_per_million": 209.69057496245543, "micro_f1": 0.7273648770951743 }, "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 } } }