{ "config": "C3", "description": "grouped-query attention instead of MHA", "parameters": 2664192, "windowed": true, "window": { "src_window": 1024, "src_stride": 768, "train_examples": 25654 }, "best_epoch": 100, "best_val_loss": 0.09699018683659276, "test": { "loss": 0.1144313706102735, "ppl": 1.121235688702341, "token_acc": 0.9764940276309759, "seq_acc": 0.5841433778857837 }, "test_full": { "loss": 7.072706952822297, "ppl": 1179.336124548187, "token_acc": 0.2687037689327228, "seq_acc": 0.05504587155963303 }, "gen_test_win": { "n": 3292.0, "seq_acc": 0.675273390036452, "bit_acc": 0.9540681959723354, "lev": 1.0814094775212637, "lev_norm": 0.008657267643119223, "token_acc": 0.8318942697353856, "bleu": 0.9509874261074868, "rouge1": 0.9772666938586414, "rouge2": 0.9587682381941536, "rougeL": 0.9772666938586414 }, "gen_test": { "n": 500.0, "seq_acc": 0.234, "bit_acc": 0.8033756827356923, "lev": 20.962, "lev_norm": 0.03107400079250732, "token_acc": 0.4397860179916476, "bleu": 0.929190036701311, "rouge1": 0.9655610236524641, "rouge2": 0.946187698054285, "rougeL": 0.9654813283904791 }, "gen_test_full": { "n": 500.0, "seq_acc": 0.054, "bit_acc": 0.3040029338736997, "lev": 514.444, "lev_norm": 0.6223494821261791, "token_acc": 0.1262062330544299, "bleu": 0.15117062732541714, "rouge1": 0.40921692487062156, "rouge2": 0.36167583779122003, "rougeL": 0.40420492944579967 }, "checkpoint": "/home2/winterdew/ANLP_A1/ANLP-M26-Assignment-1/outputs/checkpoints/C3_best.pt", "history": [ { "train/epoch_loss": 6.989599973483573, "epoch": 1, "val/loss": 5.990274003933715, "val/ppl": 399.5240660964388, "val/token_acc": 0.11756140858274078, "val/seq_acc": 0.0, "val_full/loss": 6.452544253219069, "val_full/ppl": 634.3140972072496, "val_full/token_acc": 0.05217574584682596, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 5.850808373115901, "epoch": 2, "val/loss": 5.175621278250005, "val/ppl": 176.90648829519168, "val/token_acc": 0.2051440595984881, "val/seq_acc": 0.0, "val_full/loss": 6.947227714412999, "val_full/ppl": 1040.261824120783, "val_full/token_acc": 0.05874801203318707, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 5.160893423598901, "epoch": 3, "val/loss": 4.433984833717812, "val/ppl": 84.2665369056405, "val/token_acc": 0.2812378454388607, "val/seq_acc": 0.0, "val_full/loss": 6.874995958200004, "val_full/ppl": 967.7714540381127, "val_full/token_acc": 0.07647205349786354, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 4.584489616669919, "epoch": 4, "val/loss": 3.7977423126686056, "val/ppl": 44.60037703405305, "val/token_acc": 0.34632146881949916, "val/seq_acc": 0.0, "val_full/loss": 7.224059503080151, "val_full/ppl": 1372.0476020835756, "val_full/token_acc": 0.09195424323133228, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 4.070156936930896, "epoch": 5, "val/loss": 3.172344343313994, "val/ppl": 23.863362746675918, "val/token_acc": 0.40933071376569496, "val/seq_acc": 0.0, "val_full/loss": 7.477343016488149, "val_full/ppl": 1767.538211496082, "val_full/token_acc": 0.10760888309797084, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.0, "gen_val_win/bit_acc": 0.6036592754150006, "gen_val_win/lev": 69.84375, "gen_val_win/lev_norm": 0.525023257132679, "gen_val_win/token_acc": 0.07025986525505294, "gen_val_win/bleu": 0.07019977442852068, "gen_val_win/rouge1": 0.35551507592530374, "gen_val_win/rouge2": 0.12485791843316978, "gen_val_win/rougeL": 0.3450646516611862, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0, "gen_val/bit_acc": 0.5221784858948341, "gen_val/lev": 482.515625, "gen_val/lev_norm": 0.5885043305785981, "gen_val/token_acc": 0.016186246162100935, "gen_val/bleu": 0.05964545616706091, "gen_val/rouge1": 0.3455145992929659, "gen_val/rouge2": 0.11337253439032603, "gen_val/rougeL": 0.3069993632651574, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0, "gen_val_full/bit_acc": 0.15350044471811713, "gen_val_full/lev": 576.8671875, "gen_val_full/lev_norm": 0.8074923571419808, "gen_val_full/token_acc": 0.011608838253064043, "gen_val_full/bleu": 0.008959534074905824, "gen_val_full/rouge1": 0.18806005135000617, "gen_val_full/rouge2": 0.050345763578184946, "gen_val_full/rougeL": 0.1554150902560231 }, { "train/epoch_loss": 3.595629543437625, "epoch": 6, "val/loss": 2.6532087690795443, "val/ppl": 14.199528631330784, "val/token_acc": 0.47436447228012085, "val/seq_acc": 0.00032123353678124, "val_full/loss": 7.670631323171549, "val_full/ppl": 2144.4348493954085, "val_full/token_acc": 0.12320603958688613, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 3.170019080811308, "epoch": 7, "val/loss": 2.082639462583742, "val/ppl": 8.025624317664125, "val/token_acc": 0.5620169341085658, "val/seq_acc": 0.0, "val_full/loss": 7.71343772753837, "val_full/ppl": 2238.2234500899194, "val_full/token_acc": 0.14458985609994443, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 2.769275200931806, "epoch": 8, "val/loss": 1.6803616520518458, "val/ppl": 5.3674967863769805, "val/token_acc": 0.6317379900469727, "val/seq_acc": 0.0019274012206874397, "val_full/loss": 8.04885593108701, "val_full/ppl": 3130.2117440021993, "val_full/token_acc": 0.15685297668090978, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 2.417862846369755, "epoch": 9, "val/loss": 1.2903735235422162, "val/ppl": 3.634143740509679, "val/token_acc": 0.7025062081758434, "val/seq_acc": 0.0038548024413748794, "val_full/loss": 7.636971437347909, "val_full/ppl": 2073.4547104056815, "val_full/token_acc": 0.18249056314549042, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 2.1268499670480554, "epoch": 10, "val/loss": 1.0460175324772212, "val/ppl": 2.8462932464827544, "val/token_acc": 0.752440885201105, "val/seq_acc": 0.007067137809187279, "val_full/loss": 7.413242089448926, "val_full/ppl": 1657.7923547272778, "val_full/token_acc": 0.1950602617409799, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.03125, "gen_val_win/bit_acc": 0.7440027916639615, "gen_val_win/lev": 19.96875, "gen_val_win/lev_norm": 0.16611301146738794, "gen_val_win/token_acc": 0.26883022523613465, "gen_val_win/bleu": 0.49429885227764203, "gen_val_win/rouge1": 0.7172496910400237, "gen_val_win/rouge2": 0.5533347647684904, "gen_val_win/rougeL": 0.7159504154335546, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0, "gen_val/bit_acc": 0.5681987141362974, "gen_val/lev": 225.625, "gen_val/lev_norm": 0.2884760047173648, "gen_val/token_acc": 0.054215202369200396, "gen_val/bleu": 0.405683953390014, "gen_val/rouge1": 0.6643176784013426, "gen_val/rouge2": 0.5042827269931021, "gen_val/rougeL": 0.6552923208500258, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0, "gen_val_full/bit_acc": 0.1468951587649527, "gen_val_full/lev": 512.4921875, "gen_val_full/lev_norm": 0.7381816170961426, "gen_val_full/token_acc": 0.04127589146898039, "gen_val_full/bleu": 0.010842703489584136, "gen_val_full/rouge1": 0.2962427389083778, "gen_val_full/rouge2": 0.20960385615883603, "gen_val_full/rougeL": 0.2871395255586265 }, { "train/epoch_loss": 1.9079738789960334, "epoch": 11, "val/loss": 0.8579946087801277, "val/ppl": 2.3584263806610197, "val/token_acc": 0.7918141835625455, "val/seq_acc": 0.01574044330228076, "val_full/loss": 7.7046893262948135, "val_full/ppl": 2218.7279747357975, "val_full/token_acc": 0.20188162256414188, "val_full/seq_acc": 0.0023094688221709007, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.7427265025731036, "epoch": 12, "val/loss": 0.7416713511559455, "val/ppl": 2.099441487779417, "val/token_acc": 0.8173350220901358, "val/seq_acc": 0.022165114037905557, "val_full/loss": 7.504850459387994, "val_full/ppl": 1816.8335540562318, "val_full/token_acc": 0.21044664584490985, "val_full/seq_acc": 0.004618937644341801, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.6153624997174652, "epoch": 13, "val/loss": 0.6337494883494635, "val/ppl": 1.8846638731143608, "val/token_acc": 0.8424968335809955, "val/seq_acc": 0.04079665917121748, "val_full/loss": 7.501646734105846, "val_full/ppl": 1811.0222323752907, "val_full/token_acc": 0.21941405276974074, "val_full/seq_acc": 0.004618937644341801, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.5089074813219674, "epoch": 14, "val/loss": 0.5701952463462688, "val/ppl": 1.768612332821123, "val/token_acc": 0.8580746177858005, "val/seq_acc": 0.05942820430452939, "val_full/loss": 7.528978583250302, "val_full/ppl": 1861.2034679396102, "val_full/token_acc": 0.22184751576002606, "val_full/seq_acc": 0.006928406466512702, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.4236534477172051, "epoch": 15, "val/loss": 0.49503823520096474, "val/ppl": 1.6405609650360593, "val/token_acc": 0.8752680236558925, "val/seq_acc": 0.07516864760681015, "val_full/loss": 7.415242331358141, "val_full/ppl": 1661.1116590711063, "val_full/token_acc": 0.23340167468240433, "val_full/seq_acc": 0.004618937644341801, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.140625, "gen_val_win/bit_acc": 0.8135277616560264, "gen_val_win/lev": 7.8203125, "gen_val_win/lev_norm": 0.06329580938465548, "gen_val_win/token_acc": 0.49032574087680625, "gen_val_win/bleu": 0.7353962257751552, "gen_val_win/rouge1": 0.8714817151223528, "gen_val_win/rouge2": 0.7773797943198335, "gen_val_win/rougeL": 0.8714817151223528, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0, "gen_val/bit_acc": 0.6095861418172014, "gen_val/lev": 135.1484375, "gen_val/lev_norm": 0.1781086903715838, "gen_val/token_acc": 0.10005983928876731, "gen_val/bleu": 0.6256925958806794, "gen_val/rouge1": 0.8080047761340744, "gen_val/rouge2": 0.7107960052352724, "gen_val/rougeL": 0.8060544540305521, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0, "gen_val_full/bit_acc": 0.1701948798719968, "gen_val_full/lev": 499.03125, "gen_val_full/lev_norm": 0.7031187460979907, "gen_val_full/token_acc": 0.06618264434237171, "gen_val_full/bleu": 0.022303699612794235, "gen_val_full/rouge1": 0.3507267530203208, "gen_val_full/rouge2": 0.29420275504308835, "gen_val_full/rougeL": 0.3477867655418387 }, { "train/epoch_loss": 1.3528264306132634, "epoch": 16, "val/loss": 0.4480674072482323, "val/ppl": 1.5652842035983476, "val/token_acc": 0.8863081050353542, "val/seq_acc": 0.09476389335046578, "val_full/loss": 7.397080802467953, "val_full/ppl": 1631.2156322033163, "val_full/token_acc": 0.234934564755025, "val_full/seq_acc": 0.004618937644341801, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.2913542623234509, "epoch": 17, "val/loss": 0.40650290897702385, "val/ppl": 1.50155750935578, "val/token_acc": 0.8975775647993937, "val/seq_acc": 0.11146803726309026, "val_full/loss": 7.4587507365057775, "val_full/ppl": 1734.9792566805615, "val_full/token_acc": 0.23809615052980512, "val_full/seq_acc": 0.009237875288683603, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.2382299448189296, "epoch": 18, "val/loss": 0.374204581122268, "val/ppl": 1.4538345471381873, "val/token_acc": 0.9063837001725324, "val/seq_acc": 0.13395438483777705, "val_full/loss": 7.399187702988177, "val_full/ppl": 1634.6560643188286, "val_full/token_acc": 0.23978232960968787, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.1952774215517497, "epoch": 19, "val/loss": 0.34625202322798965, "val/ppl": 1.413758870997769, "val/token_acc": 0.9128162679139532, "val/seq_acc": 0.1551557982653389, "val_full/loss": 7.362065491530783, "val_full/ppl": 1575.0865333039621, "val_full/token_acc": 0.24482170572342832, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.1590502074829063, "epoch": 20, "val/loss": 0.3230982060466721, "val/ppl": 1.3814010061761226, "val/token_acc": 0.9186704032073082, "val/seq_acc": 0.1680051397365885, "val_full/loss": 7.344470563121539, "val_full/ppl": 1547.6153836772812, "val_full/token_acc": 0.24809825825365497, "val_full/seq_acc": 0.016166281755196306, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.2109375, "gen_val_win/bit_acc": 0.8629504431649091, "gen_val_win/lev": 4.5859375, "gen_val_win/lev_norm": 0.03685677940057642, "gen_val_win/token_acc": 0.603708281829419, "gen_val_win/bleu": 0.8208011392014744, "gen_val_win/rouge1": 0.9160634376987054, "gen_val_win/rouge2": 0.8502941556641527, "gen_val_win/rougeL": 0.9160634376987054, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0078125, "gen_val/bit_acc": 0.6568317828498149, "gen_val/lev": 89.3828125, "gen_val/lev_norm": 0.12194520731317908, "gen_val/token_acc": 0.16936566162655453, "gen_val/bleu": 0.7386865045711205, "gen_val/rouge1": 0.870665861224225, "gen_val/rouge2": 0.8007925323721086, "gen_val/rougeL": 0.8697057923330963, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0078125, "gen_val_full/bit_acc": 0.18173107033256092, "gen_val_full/lev": 491.140625, "gen_val_full/lev_norm": 0.6852363477125506, "gen_val_full/token_acc": 0.07901580316063213, "gen_val_full/bleu": 0.025232565097805543, "gen_val_full/rouge1": 0.3728066963008474, "gen_val_full/rouge2": 0.32242031350294886, "gen_val_full/rougeL": 0.37052920225074204 }, { "train/epoch_loss": 1.127733638756293, "epoch": 21, "val/loss": 0.30722854832934, "val/ppl": 1.35965167868408, "val/token_acc": 0.9227992141297084, "val/seq_acc": 0.17475104400899455, "val_full/loss": 7.514302957160512, "val_full/ppl": 1834.088592306763, "val_full/token_acc": 0.24543486175247659, "val_full/seq_acc": 0.025404157043879907, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.0995229063485923, "epoch": 22, "val/loss": 0.28644887309005873, "val/ppl": 1.331690080954848, "val/token_acc": 0.9280051061623001, "val/seq_acc": 0.19948602634115, "val_full/loss": 7.390057875582019, "val_full/ppl": 1619.7998570804027, "val_full/token_acc": 0.24993772634079978, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.0749349121440972, "epoch": 23, "val/loss": 0.2756799358771908, "val/ppl": 1.3174261357003612, "val/token_acc": 0.9319444305930927, "val/seq_acc": 0.2184388050112432, "val_full/loss": 7.453724009729061, "val_full/ppl": 1726.2798730243228, "val_full/token_acc": 0.25271608959742475, "val_full/seq_acc": 0.02771362586605081, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.0557607749751083, "epoch": 24, "val/loss": 0.2697615612203623, "val/ppl": 1.3096521416430247, "val/token_acc": 0.9346770252615412, "val/seq_acc": 0.22968197879858657, "val_full/loss": 7.464223109156144, "val_full/ppl": 1744.4997357594107, "val_full/token_acc": 0.2529077008565023, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.0338933228554572, "epoch": 25, "val/loss": 0.2501835199947985, "val/ppl": 1.2842610826495797, "val/token_acc": 0.938017971297783, "val/seq_acc": 0.24156761965949244, "val_full/loss": 7.373464415394048, "val_full/ppl": 1593.1435445029879, "val_full/token_acc": 0.25288853973059455, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.3046875, "gen_val_win/bit_acc": 0.8987543734611896, "gen_val_win/lev": 3.0234375, "gen_val_win/lev_norm": 0.02485262001207127, "gen_val_win/token_acc": 0.6712057439960386, "gen_val_win/bleu": 0.858645459061528, "gen_val_win/rouge1": 0.9340992734099315, "gen_val_win/rouge2": 0.8803333475786087, "gen_val_win/rougeL": 0.9340992734099315, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0390625, "gen_val/bit_acc": 0.6774444137550062, "gen_val/lev": 68.0, "gen_val/lev_norm": 0.08974779532846075, "gen_val/token_acc": 0.18498512245365073, "gen_val/bleu": 0.7958589390813121, "gen_val/rouge1": 0.9055358504009412, "gen_val/rouge2": 0.8516807112664563, "gen_val/rougeL": 0.9046970030456197, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0234375, "gen_val_full/bit_acc": 0.229586730237298, "gen_val_full/lev": 484.625, "gen_val_full/lev_norm": 0.6573923606536735, "gen_val_full/token_acc": 0.08771044543327676, "gen_val_full/bleu": 0.07428542996131729, "gen_val_full/rouge1": 0.3755869223139731, "gen_val_full/rouge2": 0.3288902465083256, "gen_val_full/rougeL": 0.3732770282789671 }, { "train/epoch_loss": 1.0178972708614094, "epoch": 26, "val/loss": 0.2462288763612719, "val/ppl": 1.2791923169223045, "val/token_acc": 0.9397931605349503, "val/seq_acc": 0.26341150016061676, "val_full/loss": 7.3820554109822, "val_full/ppl": 1606.8891935285726, "val_full/token_acc": 0.25583935312038936, "val_full/seq_acc": 0.03464203233256351, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 1.000409460127205, "epoch": 27, "val/loss": 0.23729652084988354, "val/ppl": 1.2678169962178472, "val/token_acc": 0.9424459714174587, "val/seq_acc": 0.26823000321233537, "val_full/loss": 7.3220146965835715, "val_full/ppl": 1513.2496395617113, "val_full/token_acc": 0.25685489279350054, "val_full/seq_acc": 0.03002309468822171, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.9873875840346414, "epoch": 28, "val/loss": 0.22455890630142314, "val/ppl": 1.2517704463434969, "val/token_acc": 0.9448694039153893, "val/seq_acc": 0.2839704465146161, "val_full/loss": 7.272783476882101, "val_full/ppl": 1440.5546281918132, "val_full/token_acc": 0.2579470769702428, "val_full/seq_acc": 0.03233256351039261, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.9726274816175351, "epoch": 29, "val/loss": 0.21345565413481418, "val/ppl": 1.2379485991491357, "val/token_acc": 0.9484098094164813, "val/seq_acc": 0.30742049469964666, "val_full/loss": 7.19820433304911, "val_full/ppl": 1337.0277509893597, "val_full/token_acc": 0.25959493379831, "val_full/seq_acc": 0.03926096997690531, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.9600888559051285, "epoch": 30, "val/loss": 0.20765087757197526, "val/ppl": 1.2307834004430809, "val/token_acc": 0.9509828365130496, "val/seq_acc": 0.31962736909733375, "val_full/loss": 7.3021666892209085, "val_full/ppl": 1483.5107546574454, "val_full/token_acc": 0.2605529900936979, "val_full/seq_acc": 0.02771362586605081, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.3828125, "gen_val_win/bit_acc": 0.8974579572755016, "gen_val_win/lev": 2.390625, "gen_val_win/lev_norm": 0.019784462880069027, "gen_val_win/token_acc": 0.6616035856573705, "gen_val_win/bleu": 0.8878941878218652, "gen_val_win/rouge1": 0.9467021643176814, "gen_val_win/rouge2": 0.9065743961645639, "gen_val_win/rougeL": 0.9467021643176814, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0703125, "gen_val/bit_acc": 0.6997807325698494, "gen_val/lev": 54.375, "gen_val/lev_norm": 0.07490780581180576, "gen_val/token_acc": 0.21767341984875363, "gen_val/bleu": 0.8355757707495678, "gen_val/rouge1": 0.9232232207954487, "gen_val/rouge2": 0.8791442454850718, "gen_val/rougeL": 0.9226268194751988, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.03125, "gen_val_full/bit_acc": 0.2536090553408127, "gen_val_full/lev": 488.1484375, "gen_val_full/lev_norm": 0.6420661298538113, "gen_val_full/token_acc": 0.09642747015549975, "gen_val_full/bleu": 0.10218164183324352, "gen_val_full/rouge1": 0.38488713260792456, "gen_val_full/rouge2": 0.33679224344990644, "gen_val_full/rougeL": 0.38187573967898775 }, { "train/epoch_loss": 0.9481533981320864, "epoch": 31, "val/loss": 0.19604998262326284, "val/ppl": 1.2165877120599418, "val/token_acc": 0.9536755392885281, "val/seq_acc": 0.3366527465467395, "val_full/loss": 7.445736739303301, "val_full/ppl": 1712.5465278304107, "val_full/token_acc": 0.2591733890283393, "val_full/seq_acc": 0.04157043879907621, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.9368665063767659, "epoch": 32, "val/loss": 0.1882823868726883, "val/ppl": 1.2071743574142577, "val/token_acc": 0.955540485284878, "val/seq_acc": 0.3536781239961452, "val_full/loss": 7.332897392769549, "val_full/ppl": 1529.8078109978937, "val_full/token_acc": 0.2597482228055721, "val_full/seq_acc": 0.04387990762124711, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.9245470563075192, "epoch": 33, "val/loss": 0.188316835448201, "val/ppl": 1.2072159435675538, "val/token_acc": 0.9562884582780664, "val/seq_acc": 0.36395759717314485, "val_full/loss": 7.277306475622257, "val_full/ppl": 1447.0850122868178, "val_full/token_acc": 0.2576021767039031, "val_full/seq_acc": 0.03464203233256351, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.9167766255928097, "epoch": 34, "val/loss": 0.17551936084266326, "val/ppl": 1.1918650639398616, "val/token_acc": 0.9590709178127275, "val/seq_acc": 0.37744940571795693, "val_full/loss": 7.092753247212056, "val_full/ppl": 1203.2159954669517, "val_full/token_acc": 0.26356128686121594, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.9071165656211073, "epoch": 35, "val/loss": 0.17466107933316144, "val/ppl": 1.1908425470600958, "val/token_acc": 0.9594000259297304, "val/seq_acc": 0.37905557340186313, "val_full/loss": 7.123949431393589, "val_full/ppl": 1241.343365413703, "val_full/token_acc": 0.2632930310985073, "val_full/seq_acc": 0.04387990762124711, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.4296875, "gen_val_win/bit_acc": 0.8995026568170036, "gen_val_win/lev": 2.4609375, "gen_val_win/lev_norm": 0.02030042998078787, "gen_val_win/token_acc": 0.6849962677282906, "gen_val_win/bleu": 0.8935636755079926, "gen_val_win/rouge1": 0.9496295003446326, "gen_val_win/rouge2": 0.9087427336687283, "gen_val_win/rougeL": 0.9496295003446326, "gen_val/n": 128.0, "gen_val/seq_acc": 0.09375, "gen_val/bit_acc": 0.7275839065464895, "gen_val/lev": 44.671875, "gen_val/lev_norm": 0.0607550933357241, "gen_val/token_acc": 0.2460539413186709, "gen_val/bleu": 0.860614537150288, "gen_val/rouge1": 0.9372103644788088, "gen_val/rouge2": 0.9016292745925857, "gen_val/rougeL": 0.9365655646486407, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.21380614775659493, "gen_val_full/lev": 476.015625, "gen_val_full/lev_norm": 0.659540394018064, "gen_val_full/token_acc": 0.13644803532895067, "gen_val_full/bleu": 0.03953044845339991, "gen_val_full/rouge1": 0.4031077808673979, "gen_val_full/rouge2": 0.36516244955261856, "gen_val_full/rougeL": 0.40142488755217415 }, { "train/epoch_loss": 0.8991394277819968, "epoch": 36, "val/loss": 0.16474125962148436, "val/ppl": 1.1790880015642018, "val/token_acc": 0.9621425935714214, "val/seq_acc": 0.41278509476389336, "val_full/loss": 7.189634744031309, "val_full/ppl": 1325.6189269444867, "val_full/token_acc": 0.2660330721033168, "val_full/seq_acc": 0.04387990762124711, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8905899891829551, "epoch": 37, "val/loss": 0.16329940015390323, "val/ppl": 1.1773891474148255, "val/token_acc": 0.9624118638489693, "val/seq_acc": 0.4044330228075811, "val_full/loss": 7.107422885449999, "val_full/ppl": 1220.9968392493697, "val_full/token_acc": 0.26833240721224777, "val_full/seq_acc": 0.04387990762124711, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8835558582125161, "epoch": 38, "val/loss": 0.158727331384341, "val/ppl": 1.1720183304923437, "val/token_acc": 0.9633293773872804, "val/seq_acc": 0.4098939929328622, "val_full/loss": 7.1334000579624055, "val_full/ppl": 1253.1304479320177, "val_full/token_acc": 0.2667611948878116, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8764812345813932, "epoch": 39, "val/loss": 0.15226682620031357, "val/ppl": 1.1644709063301466, "val/token_acc": 0.9653838098752381, "val/seq_acc": 0.4349502088017989, "val_full/loss": 7.217386341590182, "val_full/ppl": 1362.9221884062017, "val_full/token_acc": 0.2680066680718159, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8709905346730106, "epoch": 40, "val/loss": 0.14978427443561543, "val/ppl": 1.161583632413047, "val/token_acc": 0.9663811071994893, "val/seq_acc": 0.43912624477995504, "val_full/loss": 7.151483445386001, "val_full/ppl": 1275.9974243726945, "val_full/token_acc": 0.2660330721033168, "val_full/seq_acc": 0.050808314087759814, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.515625, "gen_val_win/bit_acc": 0.9180745987393593, "gen_val_win/lev": 1.8125, "gen_val_win/lev_norm": 0.014676212894125386, "gen_val_win/token_acc": 0.7444472173696032, "gen_val_win/bleu": 0.9170337296382548, "gen_val_win/rouge1": 0.9630370497463334, "gen_val_win/rouge2": 0.9316800265092104, "gen_val_win/rougeL": 0.9630370497463334, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1171875, "gen_val/bit_acc": 0.7612049753041802, "gen_val/lev": 31.6796875, "gen_val/lev_norm": 0.04610325164722047, "gen_val/token_acc": 0.33399295332786333, "gen_val/bleu": 0.8953249344427726, "gen_val/rouge1": 0.9509976892667833, "gen_val/rouge2": 0.9232984114211248, "gen_val/rougeL": 0.9508806769609053, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.23482250969926113, "gen_val_full/lev": 477.0703125, "gen_val_full/lev_norm": 0.652337028903153, "gen_val_full/token_acc": 0.13237907680484173, "gen_val_full/bleu": 0.05830825512469474, "gen_val_full/rouge1": 0.3996018975104587, "gen_val_full/rouge2": 0.35933917788508163, "gen_val_full/rougeL": 0.3965440679391057 }, { "train/epoch_loss": 0.8649544650479742, "epoch": 41, "val/loss": 0.14653525715939686, "val/ppl": 1.1578157513933425, "val/token_acc": 0.9670094045137677, "val/seq_acc": 0.4542242210086733, "val_full/loss": 7.25342085975972, "val_full/ppl": 1412.9300258399235, "val_full/token_acc": 0.26733602866504436, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8585872424213666, "epoch": 42, "val/loss": 0.14374816692442194, "val/ppl": 1.1545933071261698, "val/token_acc": 0.9675279991223783, "val/seq_acc": 0.45390298747189206, "val_full/loss": 7.027556468436836, "val_full/ppl": 1127.2727155965244, "val_full/token_acc": 0.27053593669164, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8537254562401712, "epoch": 43, "val/loss": 0.14356002992510414, "val/ppl": 1.1543761058383746, "val/token_acc": 0.968455485633932, "val/seq_acc": 0.4709283649212978, "val_full/loss": 7.015169025081914, "val_full/ppl": 1113.394821958853, "val_full/token_acc": 0.2733717833259882, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8484716263792461, "epoch": 44, "val/loss": 0.13710990600117237, "val/ppl": 1.146954198742945, "val/token_acc": 0.9693729991722432, "val/seq_acc": 0.4763893350465789, "val_full/loss": 7.04258612626703, "val_full/ppl": 1144.3431992198882, "val_full/token_acc": 0.2740807449845753, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8441733285375961, "epoch": 45, "val/loss": 0.1360174527311747, "val/ppl": 1.1457018890474495, "val/token_acc": 0.9699115397273389, "val/seq_acc": 0.4760681015097976, "val_full/loss": 7.074859674691985, "val_full/ppl": 1181.8776418236316, "val_full/token_acc": 0.27249037153423133, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.5078125, "gen_val_win/bit_acc": 0.9189896036387264, "gen_val_win/lev": 1.59375, "gen_val_win/lev_norm": 0.013447091084903304, "gen_val_win/token_acc": 0.7301745635910224, "gen_val_win/bleu": 0.9215820149903344, "gen_val_win/rouge1": 0.9630205767515715, "gen_val_win/rouge2": 0.9340606154481862, "gen_val_win/rougeL": 0.9630205767515715, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1015625, "gen_val/bit_acc": 0.7606132500992314, "gen_val/lev": 30.9609375, "gen_val/lev_norm": 0.04498149540982823, "gen_val/token_acc": 0.3138903569884207, "gen_val/bleu": 0.9025650486403642, "gen_val/rouge1": 0.9523255437381359, "gen_val/rouge2": 0.926374370699363, "gen_val/rougeL": 0.9522049728607912, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0390625, "gen_val_full/bit_acc": 0.3093394927120395, "gen_val_full/lev": 486.3671875, "gen_val_full/lev_norm": 0.6231400267833306, "gen_val_full/token_acc": 0.1185331522410174, "gen_val_full/bleu": 0.13290797493409356, "gen_val_full/rouge1": 0.38819537038851504, "gen_val_full/rouge2": 0.342392423858554, "gen_val_full/rougeL": 0.38469243954602167 }, { "train/epoch_loss": 0.8397917191583915, "epoch": 46, "val/loss": 0.13419244351956588, "val/ppl": 1.1436128793569593, "val/token_acc": 0.9709487289445602, "val/seq_acc": 0.48217153870864116, "val_full/loss": 7.048060594665542, "val_full/ppl": 1150.6250491066048, "val_full/token_acc": 0.27298856080783307, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8343660966713827, "epoch": 47, "val/loss": 0.12970648584085176, "val/ppl": 1.1384941701198295, "val/token_acc": 0.9716368640982936, "val/seq_acc": 0.49502088017989077, "val_full/loss": 7.031459499966468, "val_full/ppl": 1131.68109397158, "val_full/token_acc": 0.273390944451896, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8307965989422025, "epoch": 48, "val/loss": 0.12999792976195643, "val/ppl": 1.138826025681218, "val/token_acc": 0.9708789181318627, "val/seq_acc": 0.487311275297141, "val_full/loss": 7.029193771676024, "val_full/ppl": 1129.1199146650697, "val_full/token_acc": 0.27189637663109084, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8285939009706873, "epoch": 49, "val/loss": 0.12634814644987863, "val/ppl": 1.1346771333419254, "val/token_acc": 0.9722352424928444, "val/seq_acc": 0.5030517185994218, "val_full/loss": 6.995170872334208, "val_full/ppl": 1091.3501433478734, "val_full/token_acc": 0.27574776293855024, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8234937889617577, "epoch": 50, "val/loss": 0.12595068642664903, "val/ppl": 1.1342262341552858, "val/token_acc": 0.9722352424928444, "val/seq_acc": 0.5081914551879216, "val_full/loss": 7.024475494117534, "val_full/ppl": 1123.8049620817935, "val_full/token_acc": 0.27342926670371154, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.53125, "gen_val_win/bit_acc": 0.9263139089196388, "gen_val_win/lev": 1.7265625, "gen_val_win/lev_norm": 0.014082697474906695, "gen_val_win/token_acc": 0.7457036114570361, "gen_val_win/bleu": 0.9286855647369958, "gen_val_win/rouge1": 0.9669730997324508, "gen_val_win/rouge2": 0.9399787584580754, "gen_val_win/rougeL": 0.9669730997324508, "gen_val/n": 128.0, "gen_val/seq_acc": 0.125, "gen_val/bit_acc": 0.7676625028636194, "gen_val/lev": 29.5390625, "gen_val/lev_norm": 0.04243871533849176, "gen_val/token_acc": 0.32651978980861013, "gen_val/bleu": 0.9093670119304996, "gen_val/rouge1": 0.9580463065841937, "gen_val/rouge2": 0.9349438930719784, "gen_val/rougeL": 0.9580130619033426, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.2653244358642375, "gen_val_full/lev": 497.78125, "gen_val_full/lev_norm": 0.6385245514725162, "gen_val_full/token_acc": 0.13323566165381134, "gen_val_full/bleu": 0.11306855868577965, "gen_val_full/rouge1": 0.39961402059075135, "gen_val_full/rouge2": 0.35710937556903904, "gen_val_full/rougeL": 0.3967295042250338 }, { "train/epoch_loss": 0.8208263587773292, "epoch": 51, "val/loss": 0.12440784771365303, "val/ppl": 1.1324776552464046, "val/token_acc": 0.9731727019776406, "val/seq_acc": 0.5184709283649213, "val_full/loss": 6.980418302348196, "val_full/ppl": 1075.3681019288363, "val_full/token_acc": 0.2729310774301098, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8166668803020011, "epoch": 52, "val/loss": 0.12330046541889059, "val/ppl": 1.1312242636616812, "val/token_acc": 0.9731926479241256, "val/seq_acc": 0.52071956312239, "val_full/loss": 6.954803843961371, "val_full/ppl": 1048.1729123393093, "val_full/token_acc": 0.27570944068673475, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8141154854969491, "epoch": 53, "val/loss": 0.12225551159921066, "val/ppl": 1.130042803939384, "val/token_acc": 0.9741500533554068, "val/seq_acc": 0.5371024734982333, "val_full/loss": 6.948484714211807, "val_full/ppl": 1041.5702552009307, "val_full/token_acc": 0.27639924121941406, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8096887772517312, "epoch": 54, "val/loss": 0.11932988825015653, "val/ppl": 1.1267415557857776, "val/token_acc": 0.9745290263386224, "val/seq_acc": 0.5387086411821395, "val_full/loss": 6.870936419192741, "val_full/ppl": 963.8507116580822, "val_full/token_acc": 0.2786794152024373, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8073471624060462, "epoch": 55, "val/loss": 0.12065567768500683, "val/ppl": 1.1282363685207557, "val/token_acc": 0.9744691884991673, "val/seq_acc": 0.5309990362993896, "val_full/loss": 6.957529439273602, "val_full/ppl": 1051.0337044253001, "val_full/token_acc": 0.27843032056563644, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.546875, "gen_val_win/bit_acc": 0.933475732570983, "gen_val_win/lev": 1.6640625, "gen_val_win/lev_norm": 0.013839482376344307, "gen_val_win/token_acc": 0.7663178873941205, "gen_val_win/bleu": 0.9322082762348067, "gen_val_win/rouge1": 0.9668210706687936, "gen_val_win/rouge2": 0.9413098245496214, "gen_val_win/rougeL": 0.9668210706687936, "gen_val/n": 128.0, "gen_val/seq_acc": 0.140625, "gen_val/bit_acc": 0.7805413557515263, "gen_val/lev": 26.7265625, "gen_val/lev_norm": 0.03801630189622856, "gen_val/token_acc": 0.3863250351822196, "gen_val/bleu": 0.9156504639929312, "gen_val/rouge1": 0.9586333868092295, "gen_val/rouge2": 0.9366220379922797, "gen_val/rougeL": 0.9585567936719745, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.29388764938478873, "gen_val_full/lev": 513.7109375, "gen_val_full/lev_norm": 0.632981294675191, "gen_val_full/token_acc": 0.12693907701877843, "gen_val_full/bleu": 0.144878931685497, "gen_val_full/rouge1": 0.40207429271220163, "gen_val_full/rouge2": 0.35482217511074904, "gen_val_full/rougeL": 0.3979313163717163 }, { "train/epoch_loss": 0.8044558020303969, "epoch": 56, "val/loss": 0.11727545439261944, "val/ppl": 1.1244291159755395, "val/token_acc": 0.97471851283023, "val/seq_acc": 0.5428846771602955, "val_full/loss": 6.872435028813543, "val_full/ppl": 965.2962304712572, "val_full/token_acc": 0.27860277069880623, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.8018194896324614, "epoch": 57, "val/loss": 0.11493732554236029, "val/ppl": 1.1218031269660607, "val/token_acc": 0.9755462696093586, "val/seq_acc": 0.5470607131384516, "val_full/loss": 6.876011946962003, "val_full/ppl": 968.7551986116936, "val_full/token_acc": 0.2772040085075399, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7985977537911432, "epoch": 58, "val/loss": 0.11409033180237806, "val/ppl": 1.1208533690163265, "val/token_acc": 0.975825512860149, "val/seq_acc": 0.5489881143591391, "val_full/loss": 7.032326990002683, "val_full/ppl": 1132.6632419849886, "val_full/token_acc": 0.27360171683688134, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7964224613218236, "epoch": 59, "val/loss": 0.11439675070524698, "val/ppl": 1.1211968723011996, "val/token_acc": 0.9758354858333915, "val/seq_acc": 0.5531641503372953, "val_full/loss": 7.12281428437985, "val_full/ppl": 1239.935057668819, "val_full/token_acc": 0.27243288815650807, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7937823178762212, "epoch": 60, "val/loss": 0.11233342386395563, "val/ppl": 1.1188858617051973, "val/token_acc": 0.976274296656062, "val/seq_acc": 0.5589463539993575, "val_full/loss": 7.0887693796825, "val_full/ppl": 1198.4320779056939, "val_full/token_acc": 0.2771848473816321, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.59375, "gen_val_win/bit_acc": 0.9359351629056383, "gen_val_win/lev": 1.203125, "gen_val_win/lev_norm": 0.009813192725213648, "gen_val_win/token_acc": 0.7863738457699027, "gen_val_win/bleu": 0.937542852744623, "gen_val_win/rouge1": 0.9723619405030127, "gen_val_win/rouge2": 0.9483160381848725, "gen_val_win/rougeL": 0.9723619405030127, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1328125, "gen_val/bit_acc": 0.7929439746300211, "gen_val/lev": 23.46875, "gen_val/lev_norm": 0.03506297233047032, "gen_val/token_acc": 0.4057604359248808, "gen_val/bleu": 0.9220486900184519, "gen_val/rouge1": 0.9625118970209101, "gen_val/rouge2": 0.9422479262296738, "gen_val/rougeL": 0.962374847468115, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.29463140024755, "gen_val_full/lev": 497.6875, "gen_val_full/lev_norm": 0.6315682614090774, "gen_val_full/token_acc": 0.13808296616510102, "gen_val_full/bleu": 0.1400583795604914, "gen_val_full/rouge1": 0.4055928447192218, "gen_val_full/rouge2": 0.3578277179736621, "gen_val_full/rougeL": 0.400730196566443 }, { "train/epoch_loss": 0.7907380693571229, "epoch": 61, "val/loss": 0.11094982448029707, "val/ppl": 1.117338842390837, "val/token_acc": 0.9765335939603674, "val/seq_acc": 0.5589463539993575, "val_full/loss": 7.0543479888003215, "val_full/ppl": 1157.8822728975465, "val_full/token_acc": 0.2771848473816321, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7890451688718915, "epoch": 62, "val/loss": 0.11016168126515814, "val/ppl": 1.1164585663005762, "val/token_acc": 0.9768327831576428, "val/seq_acc": 0.5682621265660135, "val_full/loss": 7.0252810098871405, "val_full/ppl": 1124.7105693922297, "val_full/token_acc": 0.2749238345245167, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7872676192376382, "epoch": 63, "val/loss": 0.10965380452883818, "val/ppl": 1.1158916869322937, "val/token_acc": 0.9768128372111577, "val/seq_acc": 0.5637648570510762, "val_full/loss": 6.9669353066498685, "val_full/ppl": 1060.9662268236104, "val_full/token_acc": 0.27768303665523386, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7847744113489279, "epoch": 64, "val/loss": 0.10876622832348944, "val/ppl": 1.114901687438216, "val/token_acc": 0.9772516480338284, "val/seq_acc": 0.5692258271763572, "val_full/loss": 6.980946505729177, "val_full/ppl": 1075.936265035702, "val_full/token_acc": 0.27798961466975797, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7830557331182713, "epoch": 65, "val/loss": 0.10645102258110015, "val/ppl": 1.1123234463801634, "val/token_acc": 0.977810134535409, "val/seq_acc": 0.5734018631545134, "val_full/loss": 7.054159221770871, "val_full/ppl": 1157.6637235285452, "val_full/token_acc": 0.2764567245971373, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6015625, "gen_val_win/bit_acc": 0.9402124406272366, "gen_val_win/lev": 1.2578125, "gen_val_win/lev_norm": 0.010319764987177463, "gen_val_win/token_acc": 0.802247191011236, "gen_val_win/bleu": 0.9408222745459279, "gen_val_win/rouge1": 0.9723646467290687, "gen_val_win/rouge2": 0.9498936886696873, "gen_val_win/rougeL": 0.9723646467290687, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1640625, "gen_val/bit_acc": 0.8042992682096665, "gen_val/lev": 22.1953125, "gen_val/lev_norm": 0.03289171167731446, "gen_val/token_acc": 0.4387715370388397, "gen_val/bleu": 0.9307432526155675, "gen_val/rouge1": 0.9663413005339677, "gen_val/rouge2": 0.9492914775610083, "gen_val/rougeL": 0.9663413005339677, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.2765949084048587, "gen_val_full/lev": 493.265625, "gen_val_full/lev_norm": 0.6362559467386152, "gen_val_full/token_acc": 0.14097345883151022, "gen_val_full/bleu": 0.09673504561378343, "gen_val_full/rouge1": 0.4032993868775779, "gen_val_full/rouge2": 0.36106457149352444, "gen_val_full/rougeL": 0.4005574710969387 }, { "train/epoch_loss": 0.7807564665551792, "epoch": 66, "val/loss": 0.11015447742939875, "val/ppl": 1.1164505235454019, "val/token_acc": 0.9773613507394959, "val/seq_acc": 0.5727593960809508, "val_full/loss": 7.119707038360574, "val_full/ppl": 1236.0882539745162, "val_full/token_acc": 0.27557531280538045, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7793064249721251, "epoch": 67, "val/loss": 0.10633744472645147, "val/ppl": 1.1121971182436232, "val/token_acc": 0.977770242642439, "val/seq_acc": 0.5820751686476068, "val_full/loss": 7.019108881900401, "val_full/ppl": 1117.790090819359, "val_full/token_acc": 0.27643756347122955, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7768991897231028, "epoch": 68, "val/loss": 0.10601249100046296, "val/ppl": 1.1118357643608414, "val/token_acc": 0.9781891075186245, "val/seq_acc": 0.5801477674269194, "val_full/loss": 7.048800393761137, "val_full/ppl": 1151.4765954250634, "val_full/token_acc": 0.27475138439134683, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7758400914377702, "epoch": 69, "val/loss": 0.10457366941228918, "val/ppl": 1.1102371813742145, "val/token_acc": 0.9782489453580796, "val/seq_acc": 0.5833601027947317, "val_full/loss": 6.991210357426852, "val_full/ppl": 1087.0363828311026, "val_full/token_acc": 0.2753645404203951, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7739744387065384, "epoch": 70, "val/loss": 0.10573894611325962, "val/ppl": 1.1115316689658572, "val/token_acc": 0.9783586480637473, "val/seq_acc": 0.5846450369418568, "val_full/loss": 6.96796993260074, "val_full/ppl": 1062.0644980668744, "val_full/token_acc": 0.2790626377205925, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.65625, "gen_val_win/bit_acc": 0.9419245270753429, "gen_val_win/lev": 1.2265625, "gen_val_win/lev_norm": 0.009946800146149582, "gen_val_win/token_acc": 0.827508736894658, "gen_val_win/bleu": 0.9451463515498472, "gen_val_win/rouge1": 0.9743838578409949, "gen_val_win/rouge2": 0.9537120775392683, "gen_val_win/rougeL": 0.9743838578409949, "gen_val/n": 128.0, "gen_val/seq_acc": 0.203125, "gen_val/bit_acc": 0.7995197386795546, "gen_val/lev": 23.0234375, "gen_val/lev_norm": 0.03421892503360193, "gen_val/token_acc": 0.42862003596947457, "gen_val/bleu": 0.9313711910781673, "gen_val/rouge1": 0.967647715508924, "gen_val/rouge2": 0.9522383546620025, "gen_val/rougeL": 0.9676096984772938, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0625, "gen_val_full/bit_acc": 0.3143617203466268, "gen_val_full/lev": 494.84375, "gen_val_full/lev_norm": 0.6316105063997238, "gen_val_full/token_acc": 0.1377857284834029, "gen_val_full/bleu": 0.13270216591039943, "gen_val_full/rouge1": 0.40943278048118475, "gen_val_full/rouge2": 0.3658221510259879, "gen_val_full/rougeL": 0.40381763336122867 }, { "train/epoch_loss": 0.772312005559108, "epoch": 71, "val/loss": 0.1051495013333203, "val/ppl": 1.1108766754860564, "val/token_acc": 0.9786179453680526, "val/seq_acc": 0.5881786058464503, "val_full/loss": 7.0460855017340815, "val_full/ppl": 1148.3547005181283, "val_full/token_acc": 0.2768974304930158, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7710337933162204, "epoch": 72, "val/loss": 0.10477765427201075, "val/ppl": 1.1104636760498716, "val/token_acc": 0.9785182156356275, "val/seq_acc": 0.5939608095085127, "val_full/loss": 7.020232951388223, "val_full/ppl": 1119.0472710008069, "val_full/token_acc": 0.27555615167947267, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7694247332297062, "epoch": 73, "val/loss": 0.10346171766784962, "val/ppl": 1.1090033373182997, "val/token_acc": 0.9788174048329028, "val/seq_acc": 0.5926758753613878, "val_full/loss": 7.010280467986549, "val_full/ppl": 1107.9652100988787, "val_full/token_acc": 0.27810458142520456, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7678132699314792, "epoch": 74, "val/loss": 0.10348112049999288, "val/ppl": 1.1090248553326545, "val/token_acc": 0.9791764318696333, "val/seq_acc": 0.6039190491487312, "val_full/loss": 7.011301247030025, "val_full/ppl": 1109.0967752069664, "val_full/token_acc": 0.27942669911283985, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7668327173389996, "epoch": 75, "val/loss": 0.1011526638481103, "val/ppl": 1.106445543104866, "val/token_acc": 0.9796052697190614, "val/seq_acc": 0.6019916479280437, "val_full/loss": 6.98935135397306, "val_full/ppl": 1085.017455618785, "val_full/token_acc": 0.27764471440341837, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6015625, "gen_val_win/bit_acc": 0.9409238776837996, "gen_val_win/lev": 1.203125, "gen_val_win/lev_norm": 0.009910526271676068, "gen_val_win/token_acc": 0.800749063670412, "gen_val_win/bleu": 0.9417021666457165, "gen_val_win/rouge1": 0.9725673683788281, "gen_val_win/rouge2": 0.9505006370978257, "gen_val_win/rougeL": 0.9725673683788281, "gen_val/n": 128.0, "gen_val/seq_acc": 0.171875, "gen_val/bit_acc": 0.8131958317092185, "gen_val/lev": 20.984375, "gen_val/lev_norm": 0.03203312336490747, "gen_val/token_acc": 0.4456686003997465, "gen_val/bleu": 0.9351877411006548, "gen_val/rouge1": 0.9676545189628474, "gen_val/rouge2": 0.9514111632302684, "gen_val/rougeL": 0.9676545189628474, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.26838069788648666, "gen_val_full/lev": 501.8984375, "gen_val_full/lev_norm": 0.6427860547151769, "gen_val_full/token_acc": 0.1451008463514253, "gen_val_full/bleu": 0.10803582749515844, "gen_val_full/rouge1": 0.40431753566921874, "gen_val_full/rouge2": 0.36399098353234705, "gen_val_full/rougeL": 0.40028478769857007 }, { "train/epoch_loss": 0.7651353652340516, "epoch": 76, "val/loss": 0.10142115133491168, "val/ppl": 1.1067426497709432, "val/token_acc": 0.9794656480936662, "val/seq_acc": 0.6068101509797623, "val_full/loss": 7.03116631977045, "val_full/ppl": 1131.3493561184798, "val_full/token_acc": 0.27785548678840366, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7640917625213205, "epoch": 77, "val/loss": 0.10122045795505995, "val/ppl": 1.106520556135041, "val/token_acc": 0.9797847832374266, "val/seq_acc": 0.6042402826855123, "val_full/loss": 7.068036517513269, "val_full/ppl": 1173.8409538792728, "val_full/token_acc": 0.2760926632048899, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7630953344323689, "epoch": 78, "val/loss": 0.1007125227833499, "val/ppl": 1.1059586581423175, "val/token_acc": 0.9798346481036392, "val/seq_acc": 0.6106649534211371, "val_full/loss": 6.95840418958018, "val_full/ppl": 1051.9534987169882, "val_full/token_acc": 0.2786410929506218, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7618435771388009, "epoch": 79, "val/loss": 0.10049521107856303, "val/ppl": 1.1057183464931055, "val/token_acc": 0.9799742697290343, "val/seq_acc": 0.6103437198843559, "val_full/loss": 7.0375905363199145, "val_full/ppl": 1138.6407851580732, "val_full/token_acc": 0.27860277069880623, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7610475468516648, "epoch": 80, "val/loss": 0.10011893768862944, "val/ppl": 1.1053023723674638, "val/token_acc": 0.9798346481036392, "val/seq_acc": 0.6071313845165436, "val_full/loss": 7.028768559346797, "val_full/ppl": 1128.6399010172713, "val_full/token_acc": 0.2776063921516028, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6328125, "gen_val_win/bit_acc": 0.9416607015032212, "gen_val_win/lev": 1.203125, "gen_val_win/lev_norm": 0.010130014586885836, "gen_val_win/token_acc": 0.8159340659340659, "gen_val_win/bleu": 0.9433120489715164, "gen_val_win/rouge1": 0.9729463479767222, "gen_val_win/rouge2": 0.9521642468627991, "gen_val_win/rougeL": 0.9729463479767222, "gen_val/n": 128.0, "gen_val/seq_acc": 0.203125, "gen_val/bit_acc": 0.8114976033954946, "gen_val/lev": 20.296875, "gen_val/lev_norm": 0.03033220533638265, "gen_val/token_acc": 0.44190140845070425, "gen_val/bleu": 0.9355313252714937, "gen_val/rouge1": 0.9681861965727557, "gen_val/rouge2": 0.9517188443046432, "gen_val/rougeL": 0.9681861965727557, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.31610552790370816, "gen_val_full/lev": 531.640625, "gen_val_full/lev_norm": 0.6330868539493608, "gen_val_full/token_acc": 0.1200580983420717, "gen_val_full/bleu": 0.1563287324590722, "gen_val_full/rouge1": 0.38919547233238444, "gen_val_full/rouge2": 0.3424790163074015, "gen_val_full/rougeL": 0.38470223548468035 }, { "train/epoch_loss": 0.7604821041932427, "epoch": 81, "val/loss": 0.09921897805019686, "val/ppl": 1.1043080923171344, "val/token_acc": 0.9801537832473995, "val/seq_acc": 0.6148409893992933, "val_full/loss": 7.001320920117266, "val_full/ppl": 1098.0826803692694, "val_full/token_acc": 0.2782195481806511, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7593906780430801, "epoch": 82, "val/loss": 0.09945412890782776, "val/ppl": 1.1045678018463942, "val/token_acc": 0.9800540535149744, "val/seq_acc": 0.6100224863475747, "val_full/loss": 6.987200966678802, "val_full/ppl": 1082.6867547211023, "val_full/token_acc": 0.27837283718791317, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7583864650821448, "epoch": 83, "val/loss": 0.09982666859590296, "val/ppl": 1.1049793738494447, "val/token_acc": 0.9797748102641841, "val/seq_acc": 0.6032765820751687, "val_full/loss": 7.013561361709364, "val_full/ppl": 1111.6062959435021, "val_full/token_acc": 0.276801624863477, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7572603337188016, "epoch": 84, "val/loss": 0.099075026423266, "val/ppl": 1.1041491368118388, "val/token_acc": 0.9800939454079445, "val/seq_acc": 0.6103437198843559, "val_full/loss": 7.002755160330721, "val_full/ppl": 1099.65872465003, "val_full/token_acc": 0.27737645864070976, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7569487954018419, "epoch": 85, "val/loss": 0.09825877160066629, "val/ppl": 1.103248237485777, "val/token_acc": 0.9802634859530672, "val/seq_acc": 0.6109861869579184, "val_full/loss": 7.0303007009379375, "val_full/ppl": 1130.3704625451523, "val_full/token_acc": 0.2771082028780011, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.640625, "gen_val_win/bit_acc": 0.9446866662328366, "gen_val_win/lev": 1.1484375, "gen_val_win/lev_norm": 0.009763818906267789, "gen_val_win/token_acc": 0.8214285714285714, "gen_val_win/bleu": 0.9453573986319382, "gen_val_win/rouge1": 0.9734190274754968, "gen_val_win/rouge2": 0.9535350289776457, "gen_val_win/rougeL": 0.9734190274754968, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2578125, "gen_val/bit_acc": 0.8249741137775612, "gen_val/lev": 20.734375, "gen_val/lev_norm": 0.029876197612941666, "gen_val/token_acc": 0.4683174108013258, "gen_val/bleu": 0.9375790423570625, "gen_val/rouge1": 0.9703548314468113, "gen_val/rouge2": 0.9562056387128605, "gen_val/rougeL": 0.9702975970145769, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.29718050557604253, "gen_val_full/lev": 513.109375, "gen_val_full/lev_norm": 0.6357929992260594, "gen_val_full/token_acc": 0.12981775882124855, "gen_val_full/bleu": 0.1367566133666529, "gen_val_full/rouge1": 0.3971780451367779, "gen_val_full/rouge2": 0.353656808113383, "gen_val_full/rougeL": 0.3928044742530124 }, { "train/epoch_loss": 0.7564138046226596, "epoch": 86, "val/loss": 0.09867334451918991, "val/ppl": 1.1037057091485765, "val/token_acc": 0.9803133508192797, "val/seq_acc": 0.6151622229360745, "val_full/loss": 7.034119303756539, "val_full/ppl": 1134.6951502567163, "val_full/token_acc": 0.2774722642702485, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7555149182714429, "epoch": 87, "val/loss": 0.09888196388361271, "val/ppl": 1.1039359875515609, "val/token_acc": 0.9801837021671271, "val/seq_acc": 0.6129135881786059, "val_full/loss": 7.0174621477945545, "val_full/ppl": 1115.9509024963147, "val_full/token_acc": 0.2773381363888942, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7550272677010135, "epoch": 88, "val/loss": 0.09768425592157948, "val/ppl": 1.1026145861141696, "val/token_acc": 0.9803432697390073, "val/seq_acc": 0.6148409893992933, "val_full/loss": 7.026221327171434, "val_full/ppl": 1125.768651568974, "val_full/token_acc": 0.2771082028780011, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7552429828560561, "epoch": 89, "val/loss": 0.09785207361183516, "val/ppl": 1.1027996398744666, "val/token_acc": 0.9803233237925223, "val/seq_acc": 0.6154834564728557, "val_full/loss": 7.041651572290138, "val_full/ppl": 1143.2742483060929, "val_full/token_acc": 0.2776255532775106, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7543259109939423, "epoch": 90, "val/loss": 0.09752594258012934, "val/ppl": 1.1024400413314586, "val/token_acc": 0.9803033778460373, "val/seq_acc": 0.6138772887889495, "val_full/loss": 7.01955108475924, "val_full/ppl": 1118.2844900974326, "val_full/token_acc": 0.2769740749966468, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.625, "gen_val_win/bit_acc": 0.9387200676730869, "gen_val_win/lev": 1.265625, "gen_val_win/lev_norm": 0.010707005743345951, "gen_val_win/token_acc": 0.7999500499500499, "gen_val_win/bleu": 0.9436598616706124, "gen_val_win/rouge1": 0.9731206235255984, "gen_val_win/rouge2": 0.9524116594743423, "gen_val_win/rougeL": 0.9731206235255984, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2421875, "gen_val/bit_acc": 0.8212076803080898, "gen_val/lev": 20.515625, "gen_val/lev_norm": 0.030115650628926727, "gen_val/token_acc": 0.4650595354284599, "gen_val/bleu": 0.9373279416546282, "gen_val/rouge1": 0.9698777642129366, "gen_val/rouge2": 0.9553416417989058, "gen_val/rougeL": 0.9698777642129366, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.30722775239338557, "gen_val_full/lev": 518.40625, "gen_val_full/lev_norm": 0.629464694120204, "gen_val_full/token_acc": 0.12698472420799156, "gen_val_full/bleu": 0.14496082643669014, "gen_val_full/rouge1": 0.396843977411071, "gen_val_full/rouge2": 0.3505116328255736, "gen_val_full/rougeL": 0.39195707285895226 }, { "train/epoch_loss": 0.7537691541146162, "epoch": 91, "val/loss": 0.09784929930038522, "val/ppl": 1.1027965803690427, "val/token_acc": 0.9804828913644025, "val/seq_acc": 0.6151622229360745, "val_full/loss": 7.0174765934871335, "val_full/ppl": 1115.9670232964231, "val_full/token_acc": 0.2780854202992968, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7539387119381208, "epoch": 92, "val/loss": 0.09733558960125287, "val/ppl": 1.102230208557339, "val/token_acc": 0.9807820805616778, "val/seq_acc": 0.6206231930613556, "val_full/loss": 7.000471169092146, "val_full/ppl": 1097.1499798237048, "val_full/token_acc": 0.27839199831382094, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7530239126331492, "epoch": 93, "val/loss": 0.09749438305638984, "val/ppl": 1.1024052493978143, "val/token_acc": 0.9804928643376449, "val/seq_acc": 0.615804690009637, "val_full/loss": 7.016848168435877, "val_full/ppl": 1115.2659419742133, "val_full/token_acc": 0.27745310314434074, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.752830046965297, "epoch": 94, "val/loss": 0.09725568810756736, "val/ppl": 1.102142142235652, "val/token_acc": 0.9806424589362827, "val/seq_acc": 0.6190170253774494, "val_full/loss": 7.035196967392554, "val_full/ppl": 1135.918629089026, "val_full/token_acc": 0.27745310314434074, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7526666074619626, "epoch": 95, "val/loss": 0.0974356839461641, "val/ppl": 1.1023405410897449, "val/token_acc": 0.9805726481235851, "val/seq_acc": 0.6186957918406681, "val_full/loss": 7.034542495185767, "val_full/ppl": 1135.1754451402633, "val_full/token_acc": 0.27754890877387955, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6484375, "gen_val_win/bit_acc": 0.9433038328886575, "gen_val_win/lev": 1.1484375, "gen_val_win/lev_norm": 0.009791012537938229, "gen_val_win/token_acc": 0.8164335664335665, "gen_val_win/bleu": 0.9468014920515613, "gen_val_win/rouge1": 0.9746591377799843, "gen_val_win/rouge2": 0.9552154590548152, "gen_val_win/rougeL": 0.9746591377799843, "gen_val/n": 128.0, "gen_val/seq_acc": 0.234375, "gen_val/bit_acc": 0.8205073126142596, "gen_val/lev": 20.3359375, "gen_val/lev_norm": 0.029706804663936343, "gen_val/token_acc": 0.46780571149621675, "gen_val/bleu": 0.9379902756641028, "gen_val/rouge1": 0.9701759319437169, "gen_val/rouge2": 0.9557873857637655, "gen_val/rougeL": 0.9701759319437169, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.32589548918182326, "gen_val_full/lev": 503.1640625, "gen_val_full/lev_norm": 0.6235041491932111, "gen_val_full/token_acc": 0.1317401727615206, "gen_val_full/bleu": 0.1440345293709632, "gen_val_full/rouge1": 0.4011776847726795, "gen_val_full/rouge2": 0.35511867984083206, "gen_val_full/rougeL": 0.39681370631635776 }, { "train/epoch_loss": 0.752360462399195, "epoch": 96, "val/loss": 0.0970352539137979, "val/ppl": 1.1018992191963648, "val/token_acc": 0.9806225129897976, "val/seq_acc": 0.6209444265981369, "val_full/loss": 7.015264606167008, "val_full/ppl": 1113.5012465300817, "val_full/token_acc": 0.2776638755293261, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7521154284477234, "epoch": 97, "val/loss": 0.09730644616784802, "val/ppl": 1.1021980862527383, "val/token_acc": 0.9806225129897976, "val/seq_acc": 0.6190170253774494, "val_full/loss": 7.026946680574451, "val_full/ppl": 1126.5855279176023, "val_full/token_acc": 0.27800877579566574, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7520241318200889, "epoch": 98, "val/loss": 0.09717217031541787, "val/ppl": 1.1020500976010335, "val/token_acc": 0.9806723778560102, "val/seq_acc": 0.6206231930613556, "val_full/loss": 7.01809409070877, "val_full/ppl": 1116.6563426372322, "val_full/token_acc": 0.2782195481806511, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.7518042974043962, "epoch": 99, "val/loss": 0.09711618107031994, "val/ppl": 1.1019883963753274, "val/token_acc": 0.9806923238024953, "val/seq_acc": 0.6206231930613556, "val_full/loss": 7.012527409391826, "val_full/ppl": 1110.4575420183533, "val_full/token_acc": 0.2787560597060683, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602 }, { "train/epoch_loss": 0.75164632577254, "epoch": 100, "val/loss": 0.09699018683659276, "val/ppl": 1.1018495609381658, "val/token_acc": 0.9805925940700702, "val/seq_acc": 0.6180533247671057, "val_full/loss": 7.003005677082335, "val_full/ppl": 1099.93424209104, "val_full/token_acc": 0.2787177374542528, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0454950332641602, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.625, "gen_val_win/bit_acc": 0.9405052378163836, "gen_val_win/lev": 1.2265625, "gen_val_win/lev_norm": 0.01040267898356598, "gen_val_win/token_acc": 0.8099400599400599, "gen_val_win/bleu": 0.9416873344713449, "gen_val_win/rouge1": 0.9726281858244996, "gen_val_win/rouge2": 0.9509208029259197, "gen_val_win/rougeL": 0.9726281858244996, "gen_val/n": 128.0, "gen_val/seq_acc": 0.234375, "gen_val/bit_acc": 0.8242630212776328, "gen_val/lev": 20.546875, "gen_val/lev_norm": 0.029680341459339253, "gen_val/token_acc": 0.4637716516223469, "gen_val/bleu": 0.9376694342844308, "gen_val/rouge1": 0.9701342845623324, "gen_val/rouge2": 0.9557410524227979, "gen_val/rougeL": 0.9701342845623324, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.30792562027758424, "gen_val_full/lev": 529.7578125, "gen_val_full/lev_norm": 0.6370845639085457, "gen_val_full/token_acc": 0.12720306513409962, "gen_val_full/bleu": 0.15298623544400525, "gen_val_full/rouge1": 0.3947661668018885, "gen_val_full/rouge2": 0.346543388928826, "gen_val_full/rougeL": 0.38871372941533305 } ] }