assay-27b / eval-dev-scaled.json
Berk's picture
Upload folder using huggingface_hub
2edbcd9 verified
Raw History Blame Contribute Delete
22.3 kB
Invalid JSON:Unexpected token 'N', ..."fidence": NaN, "... is not valid JSON
{
"overall": {
"count": 5513,
"accuracy": 0.8342100489751496,
"brier": 0.24266185842494856,
"brier_soft": 0.21710945136112134,
"nll": 0.45114049174356935,
"nll_soft": 0.4673940458514404,
"ece": 0.040077075217092,
"mean_confidence": 0.8356972425873123,
"confident_error_rate": 0.020678396517322693,
"score_mae": 0.49627245924055996,
"evidence_auc": 0.9999878244320566
},
"by_type": {
"bool": {
"count": 1620,
"accuracy": 0.8753086419753087,
"brier": 0.1834097224750155,
"brier_soft": 0.17999034550789694,
"nll": 0.287108752618916,
"nll_soft": 0.2929012295966247,
"ece": 0.044949108830161205,
"mean_confidence": 0.9199214072135956,
"confident_error_rate": 0.027777777777777776,
"evidence_auc": 0.9999940298151033
},
"choice": {
"count": 2993,
"accuracy": 0.8529903107250251,
"brier": 0.21721720365580133,
"brier_soft": 0.19167011910193568,
"nll": 0.4385055987416552,
"nll_soft": 0.45237298320260255,
"ece": 0.03450840124038069,
"mean_confidence": 0.871664572050265,
"confident_error_rate": 0.023053792181757436,
"evidence_auc": 0.9999655528763348
},
"score": {
"count": 900,
"accuracy": 0.6977777777777778,
"brier": 0.4339333161615588,
"brier_soft": 0.3685237546210836,
"nll": 0.7884156607843117,
"nll_soft": 0.8314344934522995,
"ece": 0.13329538383179557,
"mean_confidence": 0.5644823939459822,
"confident_error_rate": 0.0,
"score_mae": 0.49627245924055996
}
},
"by_source": {
"ag_news": {
"count": 100,
"accuracy": 0.91,
"brier": 0.14421606169681028,
"brier_soft": 0.14421606169681028,
"nll": 0.27468413989897816,
"nll_soft": 0.27468413989897816,
"ece": 0.0614854983919261,
"mean_confidence": 0.9180683161223662,
"confident_error_rate": 0.03
},
"amazon_polarity": {
"count": 100,
"accuracy": 0.95,
"brier": 0.05888264685815231,
"brier_soft": 0.05888264685815231,
"nll": 0.09794057175036192,
"nll_soft": 0.09794057175036192,
"ece": 0.01887097873912972,
"mean_confidence": 0.9638854049157964,
"confident_error_rate": 0.01
},
"amazon_stars": {
"count": 100,
"accuracy": 0.69,
"brier": 0.4871344044281428,
"brier_soft": 0.4871344044281428,
"nll": 0.8516013229732908,
"nll_soft": 0.8516013229732908,
"ece": 0.17589107107084062,
"mean_confidence": 0.5336082819156109,
"confident_error_rate": 0.0,
"score_mae": 0.4720212013208948
},
"anli": {
"count": 100,
"accuracy": 0.74,
"brier": 0.38152495298281364,
"brier_soft": 0.38152495298281364,
"nll": 0.661801793051201,
"nll_soft": 0.661801793051201,
"ece": 0.10857199956125599,
"mean_confidence": 0.8285262554356989,
"confident_error_rate": 0.03
},
"arc_challenge": {
"count": 100,
"accuracy": 0.98,
"brier": 0.03293778487998213,
"brier_soft": 0.03293778487998213,
"nll": 0.06098433202804499,
"nll_soft": 0.06098433202804499,
"ece": 0.010299708559671938,
"mean_confidence": 0.9801719901814279,
"confident_error_rate": 0.01
},
"arc_easy": {
"count": 100,
"accuracy": 0.98,
"brier": 0.0284256234087672,
"brier_soft": 0.0284256234087672,
"nll": 0.0519709495739374,
"nll_soft": 0.0519709495739374,
"ece": 0.023087990617371268,
"mean_confidence": 0.9835592129939819,
"confident_error_rate": 0.0
},
"banking77": {
"count": 100,
"accuracy": 0.83,
"brier": 0.27660070225206373,
"brier_soft": 0.27660070225206373,
"nll": 0.712525693075277,
"nll_soft": 0.712525693075277,
"ece": 0.10742287422893768,
"mean_confidence": 0.8310200307491299,
"confident_error_rate": 0.03
},
"boolq": {
"count": 125,
"accuracy": 0.72,
"brier": 0.39824933464474527,
"brier_soft": 0.39824933464474527,
"nll": 0.5661039342669112,
"nll_soft": 0.5661039342669112,
"ece": 0.19542354013210722,
"mean_confidence": 0.9012351747266559,
"confident_error_rate": 0.024,
"evidence_auc": 1.0
},
"cb": {
"count": 28,
"accuracy": 0.9642857142857143,
"brier": 0.0779562215366468,
"brier_soft": 0.0779562215366468,
"nll": 0.1411652211306288,
"nll_soft": 0.1411652211306288,
"ece": 0.0595834548195279,
"mean_confidence": 0.9340672209362116,
"confident_error_rate": 0.0
},
"civil_comments_severity": {
"count": 100,
"accuracy": 0.85,
"brier": 0.23714380547223418,
"brier_soft": 0.15459555443470976,
"nll": 0.46339746603398696,
"nll_soft": 0.5929402263240431,
"ece": 0.14721173331831938,
"mean_confidence": 0.7201727500927853,
"confident_error_rate": 0.0,
"score_mae": 0.34746314098192826
},
"civil_comments_toxic": {
"count": 100,
"accuracy": 0.96,
"brier": 0.09337668620304003,
"brier_soft": 0.037982779335719564,
"nll": 0.1861470519288842,
"nll_soft": 0.27998517896776515,
"ece": 0.09812762637818251,
"mean_confidence": 0.8618723736218176,
"confident_error_rate": 0.0
},
"clinc150": {
"count": 100,
"accuracy": 0.88,
"brier": 0.18462976225127634,
"brier_soft": 0.18462976225127634,
"nll": 0.5128245533780458,
"nll_soft": 0.5128245533780458,
"ece": 0.07346104900904875,
"mean_confidence": 0.9269573388899676,
"confident_error_rate": 0.03
},
"cola": {
"count": 100,
"accuracy": 0.82,
"brier": 0.22831615379946643,
"brier_soft": 0.22831615379946643,
"nll": 0.3590869683317728,
"nll_soft": 0.3590869683317728,
"ece": 0.07450945927672789,
"mean_confidence": 0.8555754148249904,
"confident_error_rate": 0.02
},
"commonsense_qa": {
"count": 100,
"accuracy": 0.91,
"brier": 0.15046780462930462,
"brier_soft": 0.15046780462930462,
"nll": 0.3438882238808393,
"nll_soft": 0.3438882238808393,
"ece": 0.03343498566763066,
"mean_confidence": 0.9374398583651028,
"confident_error_rate": 0.04
},
"cosmos_qa": {
"count": 120,
"accuracy": 0.7166666666666667,
"brier": 0.44494135470751833,
"brier_soft": 0.44494135470751833,
"nll": 0.9113581258768009,
"nll_soft": 0.9113581258768009,
"ece": 0.17210482016936163,
"mean_confidence": 0.868085591538144,
"confident_error_rate": 0.09166666666666666,
"evidence_auc": 0.999375
},
"dbpedia": {
"count": 100,
"accuracy": 1.0,
"brier": 0.003741281367652323,
"brier_soft": 0.003741281367652323,
"nll": 0.00958541989430656,
"nll_soft": 0.00958541989430656,
"ece": 0.008060135485155948,
"mean_confidence": 0.9919398645148441,
"confident_error_rate": 0.0
},
"dynasent": {
"count": 100,
"accuracy": 0.73,
"brier": 0.39455497884064294,
"brier_soft": 0.27562515443314395,
"nll": 0.6819105783878199,
"nll_soft": 0.7407221597727708,
"ece": 0.17197017219513994,
"mean_confidence": 0.5922648429725653,
"confident_error_rate": 0.0,
"score_mae": 0.4612412387247546
},
"enron_spam": {
"count": 100,
"accuracy": 0.99,
"brier": 0.020371589208394592,
"brier_soft": 0.020371589208394592,
"nll": 0.04072451183380424,
"nll_soft": 0.04072451183380424,
"ece": 0.02126407007058294,
"mean_confidence": 0.9811057852953994,
"confident_error_rate": 0.0
},
"go_emotions": {
"count": 100,
"accuracy": 0.6,
"brier": 0.6247055007996507,
"brier_soft": 0.2298631426256989,
"nll": 1.5618458274135032,
"nll_soft": 1.9562813612135246,
"ece": 0.2758994442781854,
"mean_confidence": 0.33157592066393576,
"confident_error_rate": 0.0
},
"hellaswag": {
"count": 100,
"accuracy": 0.97,
"brier": 0.04435647337572136,
"brier_soft": 0.04435647337572136,
"nll": 0.08034530834752719,
"nll_soft": 0.08034530834752719,
"ece": 0.03405762729343661,
"mean_confidence": 0.9562575630567488,
"confident_error_rate": 0.0
},
"imdb": {
"count": 100,
"accuracy": 0.98,
"brier": 0.04711217330218374,
"brier_soft": 0.04711217330218374,
"nll": 0.10119131895321937,
"nll_soft": 0.10119131895321937,
"ece": 0.032128265447002514,
"mean_confidence": 0.9619559211052533,
"confident_error_rate": 0.01
},
"ledgar": {
"count": 100,
"accuracy": 0.78,
"brier": 0.29751737785310295,
"brier_soft": 0.29751737785310295,
"nll": 0.7476464446710225,
"nll_soft": 0.7476464446710225,
"ece": 0.061867458329410076,
"mean_confidence": 0.8241415790544235,
"confident_error_rate": 0.03
},
"liar": {
"count": 100,
"accuracy": 0.39,
"brier": 0.7232980473841729,
"brier_soft": 0.7232980473841729,
"nll": 1.4529365835997112,
"nll_soft": 1.4529365835997112,
"ece": 0.09217707472940212,
"mean_confidence": 0.3114025357085896,
"confident_error_rate": 0.0,
"score_mae": 0.9575159525161713
},
"measuring_hate_speech": {
"count": 100,
"accuracy": 0.84,
"brier": 0.2793448439064333,
"brier_soft": 0.1463219404313681,
"nll": 0.5353710733933518,
"nll_soft": 0.6702362126220086,
"ece": 0.16927708181106907,
"mean_confidence": 0.6707229181889309,
"confident_error_rate": 0.0,
"score_mae": 0.5462726519505992
},
"medqa": {
"count": 100,
"accuracy": 0.88,
"brier": 0.1503504657387938,
"brier_soft": 0.1503504657387938,
"nll": 0.2623507353639012,
"nll_soft": 0.2623507353639012,
"ece": 0.08144816899443122,
"mean_confidence": 0.8978548497595163,
"confident_error_rate": 0.0
},
"mnli": {
"count": 100,
"accuracy": 0.97,
"brier": 0.06228021631952076,
"brier_soft": 0.06228021631952076,
"nll": 0.13270113391968807,
"nll_soft": 0.13270113391968807,
"ece": 0.08238346171353027,
"mean_confidence": 0.9110402983650409,
"confident_error_rate": 0.0
},
"mrpc": {
"count": 100,
"accuracy": 0.9,
"brier": 0.16585409895732134,
"brier_soft": 0.16585409895732134,
"nll": 0.2645067375959682,
"nll_soft": 0.2645067375959682,
"ece": 0.061724041454668824,
"mean_confidence": 0.885295208394654,
"confident_error_rate": 0.01
},
"multirc": {
"count": 125,
"accuracy": 0.768,
"brier": 0.39100223311737586,
"brier_soft": 0.39100223311737586,
"nll": 0.61268910167795,
"nll_soft": 0.61268910167795,
"ece": 0.18123682958658455,
"mean_confidence": 0.9092147247418939,
"confident_error_rate": 0.128,
"evidence_auc": 0.9997333333333334
},
"openbookqa": {
"count": 100,
"accuracy": 0.98,
"brier": 0.036709257253909985,
"brier_soft": 0.036709257253909985,
"nll": 0.07752743679878692,
"nll_soft": 0.07752743679878692,
"ece": 0.016796433783453687,
"mean_confidence": 0.9650003552691371,
"confident_error_rate": 0.0
},
"piqa": {
"count": 100,
"accuracy": 0.91,
"brier": 0.12286978168514663,
"brier_soft": 0.12286978168514663,
"nll": 0.2169208854703175,
"nll_soft": 0.2169208854703175,
"ece": 0.07068212570085167,
"mean_confidence": 0.9379244741154052,
"confident_error_rate": 0.03
},
"policy_application": {
"count": 150,
"accuracy": 0.9466666666666667,
"brier": 0.07417797153585332,
"brier_soft": 0.07417797153585332,
"nll": 0.11845208966989801,
"nll_soft": 0.11845208966989801,
"ece": 0.044110458372438946,
"mean_confidence": 0.9800457791058929,
"confident_error_rate": 0.03333333333333333,
"evidence_auc": 1.0
},
"qqp": {
"count": 100,
"accuracy": 0.92,
"brier": 0.13311880331318848,
"brier_soft": 0.13311880331318848,
"nll": 0.2320502253844424,
"nll_soft": 0.2320502253844424,
"ece": 0.07355407226380711,
"mean_confidence": 0.8951777070215553,
"confident_error_rate": 0.02
},
"race": {
"count": 125,
"accuracy": 0.792,
"brier": 0.2674134484204704,
"brier_soft": 0.2674134484204704,
"nll": 0.513583670296167,
"nll_soft": 0.513583670296167,
"ece": 0.07423129886353177,
"mean_confidence": 0.8306386892918173,
"confident_error_rate": 0.016,
"evidence_auc": 1.0
},
"rotten_tomatoes": {
"count": 100,
"accuracy": 0.98,
"brier": 0.03743266007384592,
"brier_soft": 0.03743266007384592,
"nll": 0.08081407315197316,
"nll_soft": 0.08081407315197316,
"ece": 0.0357435013308515,
"mean_confidence": 0.9567344445787098,
"confident_error_rate": 0.01
},
"rte": {
"count": 100,
"accuracy": 0.93,
"brier": 0.11037080326966946,
"brier_soft": 0.11037080326966946,
"nll": 0.18424526282282977,
"nll_soft": 0.18424526282282977,
"ece": 0.03139877402899702,
"mean_confidence": 0.9613987740289969,
"confident_error_rate": 0.02
},
"scotus": {
"count": 100,
"accuracy": 0.61,
"brier": 0.5896053540662425,
"brier_soft": 0.5896053540662425,
"nll": 1.4407866367191966,
"nll_soft": 1.4407866367191966,
"ece": 0.17632924711827827,
"mean_confidence": 0.7863292471182782,
"confident_error_rate": 0.04
},
"shp": {
"count": 100,
"accuracy": 0.65,
"brier": 0.47029834003111476,
"brier_soft": 0.10051645750786714,
"nll": 0.6632380214745927,
"nll_soft": 0.6838533045907272,
"ece": 0.11777214652077381,
"mean_confidence": 0.5322278534792262,
"confident_error_rate": 0.0
},
"sms_spam": {
"count": 100,
"accuracy": 0.99,
"brier": 0.01162780117821516,
"brier_soft": 0.01162780117821516,
"nll": 0.02331012246933196,
"nll_soft": 0.02331012246933196,
"ece": 0.016788442416964124,
"mean_confidence": 0.988267077668862,
"confident_error_rate": 0.0
},
"snli": {
"count": 100,
"accuracy": 0.92,
"brier": 0.12711939635362396,
"brier_soft": 0.12711939635362396,
"nll": 0.22879034755833566,
"nll_soft": 0.22879034755833566,
"ece": 0.0444196020597182,
"mean_confidence": 0.9308260670907695,
"confident_error_rate": 0.03
},
"social_iqa": {
"count": 100,
"accuracy": 0.84,
"brier": 0.2560615224983722,
"brier_soft": 0.2560615224983722,
"nll": 0.4826577509065636,
"nll_soft": 0.4826577509065636,
"ece": 0.07618387345343328,
"mean_confidence": 0.8898566355070517,
"confident_error_rate": 0.04
},
"sst2": {
"count": 100,
"accuracy": 0.98,
"brier": 0.04105489560088457,
"brier_soft": 0.04105489560088457,
"nll": 0.10883305537700429,
"nll_soft": 0.10883305537700429,
"ece": 0.011059711327280758,
"mean_confidence": 0.9782684225353241,
"confident_error_rate": 0.02
},
"sst5": {
"count": 100,
"accuracy": 0.66,
"brier": 0.497079796594402,
"brier_soft": 0.497079796594402,
"nll": 0.8742453682539675,
"nll_soft": 0.8742453682539675,
"ece": 0.16584622118782563,
"mean_confidence": 0.5079338519681397,
"confident_error_rate": 0.0,
"score_mae": 0.449198760613678
},
"stance_abortion": {
"count": 60,
"accuracy": 0.65,
"brier": 0.5170044623404539,
"brier_soft": 0.5170044623404539,
"nll": 0.8217929893106105,
"nll_soft": 0.8217929893106105,
"ece": 0.19912084090351673,
"mean_confidence": 0.8303765398473246,
"confident_error_rate": 0.06666666666666667
},
"stance_atheism": {
"count": 60,
"accuracy": 0.7833333333333333,
"brier": 0.3001419959341259,
"brier_soft": 0.3001419959341259,
"nll": 0.47098576663159153,
"nll_soft": 0.47098576663159153,
"ece": 0.14294118305587333,
"mean_confidence": 0.8563605992498861,
"confident_error_rate": 0.016666666666666666
},
"stance_feminist": {
"count": 60,
"accuracy": 0.7833333333333333,
"brier": 0.2920330154544224,
"brier_soft": 0.2920330154544224,
"nll": 0.49062532038555723,
"nll_soft": 0.49062532038555723,
"ece": 0.06563462797049141,
"mean_confidence": 0.8395306016627094,
"confident_error_rate": 0.016666666666666666
},
"stance_hillary": {
"count": 60,
"accuracy": 0.7166666666666667,
"brier": 0.36920640328234444,
"brier_soft": 0.36920640328234444,
"nll": 0.5975554352281234,
"nll_soft": 0.5975554352281234,
"ece": 0.08165786929705794,
"mean_confidence": 0.7783727969746086,
"confident_error_rate": 0.03333333333333333
},
"stsb": {
"count": 100,
"accuracy": 0.65,
"brier": 0.4598476608257509,
"brier_soft": 0.20566258588156333,
"nll": 0.8174855825848395,
"nll_soft": 0.8814355956930671,
"ece": 0.12916616422774185,
"mean_confidence": 0.5881235918154974,
"confident_error_rate": 0.0,
"score_mae": 0.4591854752360724
},
"subjectivity": {
"count": 100,
"accuracy": 0.99,
"brier": 0.025245447097925453,
"brier_soft": 0.025245447097925453,
"nll": 0.06443650373540663,
"nll_soft": 0.06443650373540663,
"ece": 0.05522051201824588,
"mean_confidence": 0.9472574338913156,
"confident_error_rate": 0.0
},
"swag": {
"count": 100,
"accuracy": 0.82,
"brier": 0.24280604965542157,
"brier_soft": 0.24280604965542157,
"nll": 0.43879576779754104,
"nll_soft": 0.43879576779754104,
"ece": 0.07809132699403978,
"mean_confidence": 0.8768457190873439,
"confident_error_rate": 0.02
},
"trec": {
"count": 100,
"accuracy": 0.97,
"brier": 0.044109563870200826,
"brier_soft": 0.044109563870200826,
"nll": 0.1119764191390441,
"nll_soft": 0.1119764191390441,
"ece": 0.03157924188503147,
"mean_confidence": 0.9651043539300282,
"confident_error_rate": 0.01
},
"tweet_hate": {
"count": 100,
"accuracy": 0.5,
"brier": 0.6361912887410113,
"brier_soft": 0.6361912887410113,
"nll": 0.87589744678152,
"nll_soft": 0.87589744678152,
"ece": 0.30966203984896934,
"mean_confidence": 0.7996620398489693,
"confident_error_rate": 0.05
},
"tweet_sentiment": {
"count": 100,
"accuracy": 0.74,
"brier": 0.4046324049729258,
"brier_soft": 0.4046324049729258,
"nll": 0.6904489665258657,
"nll_soft": 0.6904489665258657,
"ece": 0.17709361051488434,
"mean_confidence": 0.5807963141231423,
"confident_error_rate": 0.0,
"score_mae": 0.3678134887808973
},
"vitaminc": {
"count": 100,
"accuracy": 0.84,
"brier": 0.22414852755241488,
"brier_soft": 0.22414852755241488,
"nll": 0.4106686914678594,
"nll_soft": 0.4106686914678594,
"ece": 0.05939190838622676,
"mean_confidence": 0.8887477138644124,
"confident_error_rate": 0.03
},
"wic": {
"count": 100,
"accuracy": 0.78,
"brier": 0.32325615773124416,
"brier_soft": 0.32325615773124416,
"nll": 0.49691056539654466,
"nll_soft": 0.49691056539654466,
"ece": 0.11144833190894102,
"mean_confidence": 0.8764042284964352,
"confident_error_rate": 0.06
},
"winogrande": {
"count": 100,
"accuracy": 0.91,
"brier": 0.1699682822388243,
"brier_soft": 0.1699682822388243,
"nll": 0.32014269821420394,
"nll_soft": 0.32014269821420394,
"ece": 0.09579456110528843,
"mean_confidence": 0.9299164455248426,
"confident_error_rate": 0.06
},
"yelp": {
"count": 100,
"accuracy": 0.73,
"brier": 0.4223639030293232,
"brier_soft": 0.4223639030293232,
"nll": 0.7283440053059708,
"nll_soft": 0.7283440053059708,
"ece": 0.1725769529348749,
"mean_confidence": 0.5753164587285784,
"confident_error_rate": 0.0,
"score_mae": 0.40574022304004387
}
},
"reliability": [
{
"bin": "0.0-0.1",
"count": 0,
"confidence": NaN,
"accuracy": NaN
},
{
"bin": "0.1-0.2",
"count": 14,
"confidence": 0.1695548747932898,
"accuracy": 0.14285714285714285
},
{
"bin": "0.2-0.3",
"count": 94,
"confidence": 0.250037071579069,
"accuracy": 0.35106382978723405
},
{
"bin": "0.3-0.4",
"count": 120,
"confidence": 0.3580054189178388,
"accuracy": 0.475
},
{
"bin": "0.4-0.5",
"count": 317,
"confidence": 0.45936274075655936,
"accuracy": 0.5394321766561514
},
{
"bin": "0.5-0.6",
"count": 462,
"confidence": 0.5473715655718976,
"accuracy": 0.6363636363636364
},
{
"bin": "0.6-0.7",
"count": 311,
"confidence": 0.6482043563078829,
"accuracy": 0.7009646302250804
},
{
"bin": "0.7-0.8",
"count": 373,
"confidence": 0.7494447313036959,
"accuracy": 0.7372654155495979
},
{
"bin": "0.8-0.9",
"count": 603,
"confidence": 0.8560699667307652,
"accuracy": 0.736318407960199
},
{
"bin": "0.9-1.0",
"count": 3219,
"confidence": 0.976238805062891,
"accuracy": 0.9645852749301025
}
],
"seconds": 387.77550768852234,
"model": "runs/assay-27b",
"data": "data/v2/dev.jsonl"
}