{ "best_score": 0.6681005201442157, "history": [ { "event": "eval_epoch", "stage": "A", "epoch": 1, "global_step": 1050, "qa_token_acc": 0.0026088659046127876, "qa_em": 0.0, "qa_f1": 0.008460056155250058, "rank_r1": 0.9179775280898876, "rank_r5": 1.0, "rank_mrr": 0.9589887640449438, "dev_score": 0.4837244101000969 }, { "event": "eval_epoch", "stage": "A", "epoch": 2, "global_step": 2100, "qa_token_acc": 0.0002740576395289882, "qa_em": 0.0, "qa_f1": 0.0021882260907039866, "rank_r1": 0.9235955056179775, "rank_r5": 1.0, "rank_mrr": 0.9617977528089887, "dev_score": 0.48199298944984637 }, { "event": "eval_epoch", "stage": "A", "epoch": 3, "global_step": 3150, "qa_token_acc": 0.002442889937898036, "qa_em": 0.0, "qa_f1": 0.010146229290939182, "rank_r1": 0.9258426966292135, "rank_r5": 1.0, "rank_mrr": 0.9629213483146067, "dev_score": 0.486533788802773 }, { "event": "eval_epoch", "stage": "A", "epoch": 4, "global_step": 4200, "qa_token_acc": 0.002491740051239843, "qa_em": 0.0, "qa_f1": 0.00798511925920835, "rank_r1": 0.9089887640449438, "rank_r5": 1.0, "rank_mrr": 0.9544943820224719, "dev_score": 0.4812397506408401 }, { "event": "eval_epoch", "stage": "A", "epoch": 5, "global_step": 5250, "qa_token_acc": 0.0021886772463160474, "qa_em": 0.0, "qa_f1": 0.018776815370558517, "rank_r1": 0.946067415730337, "rank_r5": 1.0, "rank_mrr": 0.9730337078651685, "dev_score": 0.4959052616178635 }, { "event": "eval_epoch", "stage": "A", "epoch": 6, "global_step": 6300, "qa_token_acc": 0.00150491853246987, "qa_em": 0.0, "qa_f1": 0.010527824174669367, "rank_r1": 0.9550561797752809, "rank_r5": 1.0, "rank_mrr": 0.9775280898876404, "dev_score": 0.4940279570311549 }, { "event": "eval_epoch", "stage": "A", "epoch": 7, "global_step": 7350, "qa_token_acc": 0.0007350574409100871, "qa_em": 0.0, "qa_f1": 0.005656881949104641, "rank_r1": 0.9382022471910112, "rank_r5": 1.0, "rank_mrr": 0.9691011235955056, "dev_score": 0.4873790027723051 }, { "event": "eval_epoch", "stage": "A", "epoch": 8, "global_step": 8400, "qa_token_acc": 0.0019801811067466503, "qa_em": 0.0, "qa_f1": 0.016576542164749433, "rank_r1": 0.9426966292134832, "rank_r5": 1.0, "rank_mrr": 0.9713483146067415, "dev_score": 0.4939624283857455 }, { "event": "eval_epoch", "stage": "A", "epoch": 9, "global_step": 9450, "qa_token_acc": 0.003337956813784035, "qa_em": 0.0, "qa_f1": 0.027663325242331938, "rank_r1": 0.950561797752809, "rank_r5": 1.0, "rank_mrr": 0.9752808988764045, "dev_score": 0.5014721120593681 }, { "event": "eval_epoch", "stage": "A", "epoch": 10, "global_step": 10500, "qa_token_acc": 0.002120477749868174, "qa_em": 0.0, "qa_f1": 0.015086279129733322, "rank_r1": 0.9348314606741573, "rank_r5": 1.0, "rank_mrr": 0.9674157303370786, "dev_score": 0.491251004733406 }, { "event": "eval_epoch", "stage": "A", "epoch": 11, "global_step": 11550, "qa_token_acc": 0.004011323542787213, "qa_em": 0.0, "qa_f1": 0.021823077741655643, "rank_r1": 0.9382022471910112, "rank_r5": 1.0, "rank_mrr": 0.9691011235955056, "dev_score": 0.4954621006685806 }, { "event": "eval_epoch", "stage": "A", "epoch": 12, "global_step": 12600, "qa_token_acc": 0.0020072296711053372, "qa_em": 0.0, "qa_f1": 0.019076629476642668, "rank_r1": 0.946067415730337, "rank_r5": 1.0, "rank_mrr": 0.9730337078651685, "dev_score": 0.4960551686709056 }, { "event": "eval_epoch", "stage": "A", "epoch": 13, "global_step": 13650, "qa_token_acc": 0.004285261014129599, "qa_em": 0.0, "qa_f1": 0.02629726631797427, "rank_r1": 0.9269662921348315, "rank_r5": 1.0, "rank_mrr": 0.9634831460674157, "dev_score": 0.49489020619269497 }, { "event": "eval_epoch", "stage": "A", "epoch": 14, "global_step": 14700, "qa_token_acc": 0.0021624533130300636, "qa_em": 0.0, "qa_f1": 0.013479208688918377, "rank_r1": 0.9325842696629213, "rank_r5": 1.0, "rank_mrr": 0.9662921348314607, "dev_score": 0.48988567176018954 }, { "event": "eval_epoch", "stage": "A", "epoch": 15, "global_step": 15750, "qa_token_acc": 0.0017375697114891482, "qa_em": 0.0, "qa_f1": 0.010337309586139602, "rank_r1": 0.9595505617977528, "rank_r5": 1.0, "rank_mrr": 0.9797752808988764, "dev_score": 0.49505629524250805 }, { "event": "eval_epoch", "stage": "A", "epoch": 16, "global_step": 16800, "qa_token_acc": 0.0031858088179451674, "qa_em": 0.0, "qa_f1": 0.010815200878153929, "rank_r1": 0.9561797752808989, "rank_r5": 1.0, "rank_mrr": 0.9780898876404495, "dev_score": 0.4944525442593017 }, { "event": "eval_epoch", "stage": "A", "epoch": 17, "global_step": 17850, "qa_token_acc": 0.002876572579325557, "qa_em": 0.0, "qa_f1": 0.018080411167543484, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.49752335165118744 }, { "event": "eval_epoch", "stage": "A", "epoch": 18, "global_step": 18900, "qa_token_acc": 0.005014380686699785, "qa_em": 0.0, "qa_f1": 0.01655073734085142, "rank_r1": 0.949438202247191, "rank_r5": 1.0, "rank_mrr": 0.9747191011235955, "dev_score": 0.49563491923222347 }, { "event": "eval_epoch", "stage": "A", "epoch": 19, "global_step": 19950, "qa_token_acc": 0.0024065798941515816, "qa_em": 0.0, "qa_f1": 0.01064849724661803, "rank_r1": 0.9561797752808989, "rank_r5": 1.0, "rank_mrr": 0.9780898876404495, "dev_score": 0.4943691924435337 }, { "event": "eval_epoch", "stage": "A", "epoch": 20, "global_step": 21000, "qa_token_acc": 0.001667128692262136, "qa_em": 0.0, "qa_f1": 0.011459056958696605, "rank_r1": 0.9550561797752809, "rank_r5": 1.0, "rank_mrr": 0.9775280898876404, "dev_score": 0.4944935734231685 }, { "event": "eval_epoch", "stage": "A", "epoch": 21, "global_step": 22050, "qa_token_acc": 0.0010081766771694268, "qa_em": 0.0, "qa_f1": 0.010952496004514445, "rank_r1": 0.9640449438202248, "rank_r5": 1.0, "rank_mrr": 0.9820224719101124, "dev_score": 0.4964874839573134 }, { "event": "eval_epoch", "stage": "A", "epoch": 22, "global_step": 23100, "qa_token_acc": 0.0067095332047939745, "qa_em": 0.0, "qa_f1": 0.02894461489169904, "rank_r1": 0.950561797752809, "rank_r5": 1.0, "rank_mrr": 0.9752808988764045, "dev_score": 0.5021127568840518 }, { "event": "eval_epoch", "stage": "A", "epoch": 23, "global_step": 24150, "qa_token_acc": 0.002842412927427731, "qa_em": 0.0, "qa_f1": 0.015541875253267664, "rank_r1": 0.9651685393258427, "rank_r5": 1.0, "rank_mrr": 0.9825842696629213, "dev_score": 0.4990630724580945 }, { "event": "eval_epoch", "stage": "A", "epoch": 24, "global_step": 25200, "qa_token_acc": 0.0038040943512142413, "qa_em": 0.0, "qa_f1": 0.01523451581945387, "rank_r1": 0.9224719101123595, "rank_r5": 1.0, "rank_mrr": 0.9612359550561798, "dev_score": 0.48823523543781683 }, { "event": "eval_epoch", "stage": "A", "epoch": 25, "global_step": 26250, "qa_token_acc": 0.0020634263952415146, "qa_em": 0.0, "qa_f1": 0.006874889036257684, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.49192059058554455 }, { "event": "eval_epoch", "stage": "B", "epoch": 1, "global_step": 27300, "qa_token_acc": 0.03595962160347243, "qa_em": 0.0, "qa_f1": 0.03595962160347243, "rank_r1": 0.9595505617977528, "rank_r5": 1.0, "rank_mrr": 0.9797752808988764, "dev_score": 0.5078674512511744 }, { "event": "eval_epoch", "stage": "B", "epoch": 2, "global_step": 28350, "qa_token_acc": 0.06349278112664943, "qa_em": 0.0, "qa_f1": 0.06402782542622525, "rank_r1": 0.952808988764045, "rank_r5": 1.0, "rank_mrr": 0.9764044943820225, "dev_score": 0.5202161599041238 }, { "event": "eval_epoch", "stage": "B", "epoch": 3, "global_step": 29400, "qa_token_acc": 0.0644060876516836, "qa_em": 0.0, "qa_f1": 0.06702022663240345, "rank_r1": 0.9584269662921349, "rank_r5": 1.0, "rank_mrr": 0.9792134831460674, "dev_score": 0.5231168548892354 }, { "event": "eval_epoch", "stage": "B", "epoch": 4, "global_step": 30450, "qa_token_acc": 0.06585100411920221, "qa_em": 0.0, "qa_f1": 0.06654645635210447, "rank_r1": 0.9426966292134832, "rank_r5": 1.0, "rank_mrr": 0.9713483146067415, "dev_score": 0.518947385479423 }, { "event": "eval_epoch", "stage": "B", "epoch": 5, "global_step": 31500, "qa_token_acc": 0.06721366773564279, "qa_em": 0.0, "qa_f1": 0.0694767929892264, "rank_r1": 0.949438202247191, "rank_r5": 1.0, "rank_mrr": 0.9747191011235955, "dev_score": 0.522097947056411 }, { "event": "eval_epoch", "stage": "B", "epoch": 6, "global_step": 32550, "qa_token_acc": 0.06689300250838345, "qa_em": 0.0, "qa_f1": 0.06818316411067013, "rank_r1": 0.9471910112359551, "rank_r5": 1.0, "rank_mrr": 0.9735955056179775, "dev_score": 0.5208893348643238 }, { "event": "eval_epoch", "stage": "B", "epoch": 7, "global_step": 33600, "qa_token_acc": 0.06347600901661579, "qa_em": 0.0022471910112359553, "qa_f1": 0.06487733691557757, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.5209218145252045 }, { "event": "eval_epoch", "stage": "B", "epoch": 8, "global_step": 34650, "qa_token_acc": 0.06677638813949753, "qa_em": 0.0, "qa_f1": 0.06790244536850619, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.5224343687516688 }, { "event": "eval_epoch", "stage": "B", "epoch": 9, "global_step": 35700, "qa_token_acc": 0.06686911011489442, "qa_em": 0.0011235955056179776, "qa_f1": 0.0683053846937451, "rank_r1": 0.950561797752809, "rank_r5": 1.0, "rank_mrr": 0.9752808988764045, "dev_score": 0.5217931417850747 }, { "event": "eval_epoch", "stage": "B", "epoch": 10, "global_step": 36750, "qa_token_acc": 0.061308408559934795, "qa_em": 0.0, "qa_f1": 0.06259742774850655, "rank_r1": 0.9550561797752809, "rank_r5": 1.0, "rank_mrr": 0.9775280898876404, "dev_score": 0.5200627588180735 }, { "event": "eval_epoch", "stage": "B", "epoch": 11, "global_step": 37800, "qa_token_acc": 0.06446662016820472, "qa_em": 0.0, "qa_f1": 0.06586377064292695, "rank_r1": 0.949438202247191, "rank_r5": 1.0, "rank_mrr": 0.9747191011235955, "dev_score": 0.5202914358832612 }, { "event": "eval_epoch", "stage": "B", "epoch": 12, "global_step": 38850, "qa_token_acc": 0.0664656804764573, "qa_em": 0.0, "qa_f1": 0.06790622950450839, "rank_r1": 0.9550561797752809, "rank_r5": 1.0, "rank_mrr": 0.9775280898876404, "dev_score": 0.5227171596960744 }, { "event": "eval_epoch", "stage": "B", "epoch": 13, "global_step": 39900, "qa_token_acc": 0.06309952723736646, "qa_em": 0.0022471910112359553, "qa_f1": 0.06405629870588811, "rank_r1": 0.949438202247191, "rank_r5": 1.0, "rank_mrr": 0.9747191011235955, "dev_score": 0.5193876999147418 }, { "event": "eval_epoch", "stage": "B", "epoch": 14, "global_step": 40950, "qa_token_acc": 0.06719656732612586, "qa_em": 0.0, "qa_f1": 0.06777480643511719, "rank_r1": 0.950561797752809, "rank_r5": 1.0, "rank_mrr": 0.9752808988764045, "dev_score": 0.5215278526557608 }, { "event": "eval_epoch", "stage": "B", "epoch": 15, "global_step": 42000, "qa_token_acc": 0.0670964259986907, "qa_em": 0.0, "qa_f1": 0.06829975286082439, "rank_r1": 0.9573033707865168, "rank_r5": 1.0, "rank_mrr": 0.9786516853932584, "dev_score": 0.5234757191270414 }, { "event": "eval_epoch", "stage": "B", "epoch": 16, "global_step": 43050, "qa_token_acc": 0.06484622750046914, "qa_em": 0.007865168539325843, "qa_f1": 0.06521137736749497, "rank_r1": 0.9606741573033708, "rank_r5": 1.0, "rank_mrr": 0.9803370786516854, "dev_score": 0.5227742280095902 }, { "event": "eval_epoch", "stage": "B", "epoch": 17, "global_step": 44100, "qa_token_acc": 0.07489982579562145, "qa_em": 0.0044943820224719105, "qa_f1": 0.07553913227954111, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.5262527122071863 }, { "event": "eval_epoch", "stage": "B", "epoch": 18, "global_step": 45150, "qa_token_acc": 0.0797927738995402, "qa_em": 0.0022471910112359553, "qa_f1": 0.08073487259905027, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.5288505823669408 }, { "event": "eval_epoch", "stage": "B", "epoch": 19, "global_step": 46200, "qa_token_acc": 0.08249868470286494, "qa_em": 0.0022471910112359553, "qa_f1": 0.08344526659295631, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.5302057793638939 }, { "event": "eval_epoch", "stage": "B", "epoch": 20, "global_step": 47250, "qa_token_acc": 0.08896931529260833, "qa_em": 0.0056179775280898875, "qa_f1": 0.08979104532890575, "rank_r1": 0.9606741573033708, "rank_r5": 1.0, "rank_mrr": 0.9803370786516854, "dev_score": 0.5350640619902955 }, { "event": "eval_epoch", "stage": "B", "epoch": 21, "global_step": 48300, "qa_token_acc": 0.11453380242677842, "qa_em": 0.010112359550561797, "qa_f1": 0.11563916112067416, "rank_r1": 0.949438202247191, "rank_r5": 1.0, "rank_mrr": 0.9747191011235955, "dev_score": 0.5451791311221348 }, { "event": "eval_epoch", "stage": "B", "epoch": 22, "global_step": 49350, "qa_token_acc": 0.13940242582300655, "qa_em": 0.0044943820224719105, "qa_f1": 0.14130278684123287, "rank_r1": 0.9516853932584269, "rank_r5": 1.0, "rank_mrr": 0.9758426966292135, "dev_score": 0.5585727417352232 }, { "event": "eval_epoch", "stage": "B", "epoch": 23, "global_step": 50400, "qa_token_acc": 0.21105027623330114, "qa_em": 0.012359550561797753, "qa_f1": 0.2131529528282703, "rank_r1": 0.949438202247191, "rank_r5": 1.0, "rank_mrr": 0.9747191011235955, "dev_score": 0.5939360269759328 }, { "event": "eval_epoch", "stage": "B", "epoch": 24, "global_step": 51450, "qa_token_acc": 0.2707055699335783, "qa_em": 0.017977528089887642, "qa_f1": 0.27222918334792845, "rank_r1": 0.952808988764045, "rank_r5": 1.0, "rank_mrr": 0.9764044943820225, "dev_score": 0.6243168388649755 }, { "event": "eval_epoch", "stage": "B", "epoch": 25, "global_step": 52500, "qa_token_acc": 0.35802289772288703, "qa_em": 0.019101123595505618, "qa_f1": 0.35923474815360007, "rank_r1": 0.9539325842696629, "rank_r5": 1.0, "rank_mrr": 0.9769662921348314, "dev_score": 0.6681005201442157 } ] }