{ "config": "C1", "description": "base sinusoidal MHA LayerNorm subword", "parameters": 2763264, "windowed": true, "window": { "src_window": 1024, "src_stride": 768, "train_examples": 25654 }, "best_epoch": 98, "best_val_loss": 0.09419949818377452, "test": { "loss": 0.11081809987924529, "ppl": 1.1171916710708378, "token_acc": 0.9776071238155815, "seq_acc": 0.5972053462940462 }, "test_full": { "loss": 7.01688197593783, "ppl": 1115.3036469670765, "token_acc": 0.27035928143712573, "seq_acc": 0.0481651376146789 }, "gen_test_win": { "n": 3292.0, "seq_acc": 0.7001822600243013, "bit_acc": 0.9585891098472907, "lev": 0.9237545565006076, "lev_norm": 0.007473718377395123, "token_acc": 0.8476717869697116, "bleu": 0.956618354971245, "rouge1": 0.9799931678815064, "rouge2": 0.9636281157124085, "rougeL": 0.9799611924386465 }, "gen_test": { "n": 500.0, "seq_acc": 0.252, "bit_acc": 0.8195124641342711, "lev": 18.962, "lev_norm": 0.02828490693050463, "token_acc": 0.4725456520091935, "bleu": 0.9367783153799967, "rouge1": 0.9691731812226648, "rouge2": 0.9519293494039626, "rougeL": 0.9691285822549072 }, "gen_test_full": { "n": 500.0, "seq_acc": 0.052, "bit_acc": 0.23147938395959428, "lev": 472.024, "lev_norm": 0.6331874815918812, "token_acc": 0.15033397729711984, "bleu": 0.05748892542819064, "rouge1": 0.43040922259512776, "rouge2": 0.3962357629401473, "rougeL": 0.42921030839893004 }, "checkpoint": "/home2/winterdew/ANLP_A1/ANLP-M26-Assignment-1/outputs/checkpoints/C1_best.pt", "history": [ { "train/epoch_loss": 6.984002310736221, "epoch": 1, "val/loss": 5.98103337221754, "val/ppl": 395.8492164553241, "val/token_acc": 0.1203937329836144, "val/seq_acc": 0.0, "val_full/loss": 6.33700835245454, "val_full/ppl": 565.1031904611777, "val_full/token_acc": 0.06376822702101975, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 5.798753053470145, "epoch": 2, "val/loss": 5.077710863520796, "val/ppl": 160.4064430907789, "val/token_acc": 0.21830838427860497, "val/seq_acc": 0.0, "val_full/loss": 6.216300893746767, "val_full/ppl": 500.8471144556947, "val_full/token_acc": 0.08563107168177202, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 5.077572283899397, "epoch": 3, "val/loss": 4.344160065480361, "val/ppl": 77.02731240205254, "val/token_acc": 0.2929461160255707, "val/seq_acc": 0.0, "val_full/loss": 6.818065783139168, "val_full/ppl": 914.2150087654018, "val_full/token_acc": 0.08390657035007378, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 4.477757424190455, "epoch": 4, "val/loss": 3.6354640934785483, "val/ppl": 37.91944698223089, "val/token_acc": 0.3630361719739506, "val/seq_acc": 0.0, "val_full/loss": 7.2917656408678075, "val_full/ppl": 1468.1606545160205, "val_full/token_acc": 0.09666788020464083, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 3.925173870643178, "epoch": 5, "val/loss": 2.956197411070686, "val/ppl": 19.22472884634058, "val/token_acc": 0.4411245524628257, "val/seq_acc": 0.00032123353678124, "val_full/loss": 7.430751241880473, "val_full/ppl": 1687.0744984421979, "val_full/token_acc": 0.11556075034969054, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.0, "gen_val_win/bit_acc": 0.637901376146789, "gen_val_win/lev": 65.015625, "gen_val_win/lev_norm": 0.5051991550055724, "gen_val_win/token_acc": 0.0765774964263835, "gen_val_win/bleu": 0.0900889069734132, "gen_val_win/rouge1": 0.37481460694027796, "gen_val_win/rouge2": 0.13913032898272987, "gen_val_win/rougeL": 0.3592691731648053, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0, "gen_val/bit_acc": 0.5273019193988407, "gen_val/lev": 450.1796875, "gen_val/lev_norm": 0.5554812894715055, "gen_val/token_acc": 0.017688339154676835, "gen_val/bleu": 0.07974401206579905, "gen_val/rouge1": 0.36876009818969385, "gen_val/rouge2": 0.13734676307046037, "gen_val/rougeL": 0.33225961361794976, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0, "gen_val_full/bit_acc": 0.16213275145657377, "gen_val_full/lev": 577.421875, "gen_val_full/lev_norm": 0.7980164241644722, "gen_val_full/token_acc": 0.01256281407035176, "gen_val_full/bleu": 0.011695503776053565, "gen_val_full/rouge1": 0.19282596759503415, "gen_val_full/rouge2": 0.061630050630873574, "gen_val_full/rougeL": 0.16615531265162248 }, { "train/epoch_loss": 3.4305827510624454, "epoch": 6, "val/loss": 2.3556378011487307, "val/ppl": 10.544852236249026, "val/token_acc": 0.5170288518115906, "val/seq_acc": 0.00032123353678124, "val_full/loss": 7.728163876128112, "val_full/ppl": 2271.4277468435957, "val_full/token_acc": 0.1343003314874782, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 2.955749810782454, "epoch": 7, "val/loss": 1.8467047343690786, "val/ppl": 6.338896719939394, "val/token_acc": 0.6020584216772547, "val/seq_acc": 0.0016061676839061998, "val_full/loss": 7.792581859922589, "val_full/ppl": 2422.564236243596, "val_full/token_acc": 0.15353810189886757, "val_full/seq_acc": 0.0023094688221709007, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 2.5541224663988906, "epoch": 8, "val/loss": 1.4247556212507855, "val/ppl": 4.156841874802528, "val/token_acc": 0.676436856119915, "val/seq_acc": 0.00256986829424992, "val_full/loss": 7.737429627771178, "val_full/ppl": 2292.57203982785, "val_full/token_acc": 0.1735806395983828, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 2.23591710504451, "epoch": 9, "val/loss": 1.1423903757438436, "val/ppl": 3.1342514565798587, "val/token_acc": 0.7322356414117741, "val/seq_acc": 0.006103437198843559, "val_full/loss": 7.694882497389297, "val_full/ppl": 2197.0756330371223, "val_full/token_acc": 0.1888520569468662, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.9800643254991184, "epoch": 10, "val/loss": 0.9427200662669011, "val/ppl": 2.5669542163663546, "val/token_acc": 0.7739525884852051, "val/seq_acc": 0.013170575008030838, "val_full/loss": 7.556066801672766, "val_full/ppl": 1912.3092108137005, "val_full/token_acc": 0.19958228745521087, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.0390625, "gen_val_win/bit_acc": 0.7612957673331602, "gen_val_win/lev": 16.546875, "gen_val_win/lev_norm": 0.13655108718827963, "gen_val_win/token_acc": 0.3333333333333333, "gen_val_win/bleu": 0.5727288333029996, "gen_val_win/rouge1": 0.7665771337352694, "gen_val_win/rouge2": 0.6266085287177517, "gen_val_win/rougeL": 0.7654009807556508, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0, "gen_val/bit_acc": 0.5810499292712318, "gen_val/lev": 197.609375, "gen_val/lev_norm": 0.25792806462813633, "gen_val/token_acc": 0.05673045995496944, "gen_val/bleu": 0.45747649207590074, "gen_val/rouge1": 0.6976500757495749, "gen_val/rouge2": 0.5476990016694998, "gen_val/rougeL": 0.6905238439625659, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0, "gen_val_full/bit_acc": 0.14904436261616694, "gen_val_full/lev": 511.3671875, "gen_val_full/lev_norm": 0.7356041195637923, "gen_val_full/token_acc": 0.04243153776707794, "gen_val_full/bleu": 0.00997794972840494, "gen_val_full/rouge1": 0.29445785908991406, "gen_val_full/rouge2": 0.20895489027086148, "gen_val_full/rougeL": 0.2872274582896207 }, { "train/epoch_loss": 1.7839468221117434, "epoch": 11, "val/loss": 0.767583977060837, "val/ppl": 2.1545545073571546, "val/token_acc": 0.8134056706325857, "val/seq_acc": 0.020880179890780597, "val_full/loss": 7.483666337734006, "val_full/ppl": 1778.7503349919184, "val_full/token_acc": 0.21222863055433136, "val_full/seq_acc": 0.0023094688221709007, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.6304147722120594, "epoch": 12, "val/loss": 0.6428174678359926, "val/ppl": 1.9018316875093593, "val/token_acc": 0.8397841848590321, "val/seq_acc": 0.03437198843559268, "val_full/loss": 7.4910218897899945, "val_full/ppl": 1791.882262757104, "val_full/token_acc": 0.22019965893195884, "val_full/seq_acc": 0.004618937644341801, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.5069503243130045, "epoch": 13, "val/loss": 0.566385228477116, "val/ppl": 1.7618867087393306, "val/token_acc": 0.8589821583508691, "val/seq_acc": 0.05493093478959203, "val_full/loss": 7.360876229305984, "val_full/ppl": 1573.2144558052876, "val_full/token_acc": 0.2250091015348062, "val_full/seq_acc": 0.006928406466512702, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.4063936795974312, "epoch": 14, "val/loss": 0.4915677981663832, "val/ppl": 1.634877369490148, "val/token_acc": 0.877571780474913, "val/seq_acc": 0.07516864760681015, "val_full/loss": 7.418828830069555, "val_full/ppl": 1667.079930096576, "val_full/token_acc": 0.23117898407710438, "val_full/seq_acc": 0.013856812933025405, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.330331239319799, "epoch": 15, "val/loss": 0.444097587456937, "val/ppl": 1.5590826251068806, "val/token_acc": 0.8889409699713776, "val/seq_acc": 0.08930292322518471, "val_full/loss": 7.365168171813026, "val_full/ppl": 1579.981112464274, "val_full/token_acc": 0.2383835674184215, "val_full/seq_acc": 0.011547344110854504, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.140625, "gen_val_win/bit_acc": 0.829954954954955, "gen_val_win/lev": 6.75, "gen_val_win/lev_norm": 0.054549695751880595, "gen_val_win/token_acc": 0.46988539380638866, "gen_val_win/bleu": 0.7586045589734727, "gen_val_win/rouge1": 0.879427035672847, "gen_val_win/rouge2": 0.7899234043539008, "gen_val_win/rougeL": 0.879427035672847, "gen_val/n": 128.0, "gen_val/seq_acc": 0.015625, "gen_val/bit_acc": 0.6264025939737963, "gen_val/lev": 114.53125, "gen_val/lev_norm": 0.15199912293698825, "gen_val/token_acc": 0.11446309898565932, "gen_val/bleu": 0.664164425258185, "gen_val/rouge1": 0.8329265285057311, "gen_val/rouge2": 0.7419976960507663, "gen_val/rougeL": 0.8312591387326003, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0078125, "gen_val_full/bit_acc": 0.18785118416837124, "gen_val_full/lev": 499.0234375, "gen_val_full/lev_norm": 0.692340556690393, "gen_val_full/token_acc": 0.07507987220447285, "gen_val_full/bleu": 0.02949611758258529, "gen_val_full/rouge1": 0.34833960494706345, "gen_val_full/rouge2": 0.290113910377186, "gen_val_full/rougeL": 0.34602997620897463 }, { "train/epoch_loss": 1.26865179491162, "epoch": 16, "val/loss": 0.39941281100872883, "val/ppl": 1.4909489717353765, "val/token_acc": 0.8992729702506208, "val/seq_acc": 0.11211050433665275, "val_full/loss": 7.314919316690299, "val_full/ppl": 1502.5505603977872, "val_full/token_acc": 0.2418708923336335, "val_full/seq_acc": 0.011547344110854504, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.216165093412423, "epoch": 17, "val/loss": 0.3674597681348, "val/ppl": 1.4440617000651954, "val/token_acc": 0.9082287002223973, "val/seq_acc": 0.12528107934468358, "val_full/loss": 7.2828523491540365, "val_full/ppl": 1455.1326577400666, "val_full/token_acc": 0.24531989499703002, "val_full/seq_acc": 0.016166281755196306, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.1745105535906746, "epoch": 18, "val/loss": 0.3399719913331746, "val/ppl": 1.4049082404056978, "val/token_acc": 0.9152895652780964, "val/seq_acc": 0.15162222936074526, "val_full/loss": 7.260015131301615, "val_full/ppl": 1422.2780579566363, "val_full/token_acc": 0.24516660598976797, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.1367112198969966, "epoch": 19, "val/loss": 0.3158448645552389, "val/ppl": 1.3714174836609518, "val/token_acc": 0.9208744302939035, "val/seq_acc": 0.17089624156761965, "val_full/loss": 7.345278024942996, "val_full/ppl": 1548.8655286681515, "val_full/token_acc": 0.24828986951273257, "val_full/seq_acc": 0.025404157043879907, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.105731162644384, "epoch": 20, "val/loss": 0.29999073482542, "val/ppl": 1.3498463009564305, "val/token_acc": 0.9256814033967947, "val/seq_acc": 0.18438805011243173, "val_full/loss": 7.237146028138113, "val_full/ppl": 1390.1209380594435, "val_full/token_acc": 0.24986108183716876, "val_full/seq_acc": 0.020785219399538105, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.234375, "gen_val_win/bit_acc": 0.8634353741496599, "gen_val_win/lev": 3.765625, "gen_val_win/lev_norm": 0.03069141229081703, "gen_val_win/token_acc": 0.5879583746283449, "gen_val_win/bleu": 0.8382186552956398, "gen_val_win/rouge1": 0.9247446749599737, "gen_val_win/rouge2": 0.8635844533255597, "gen_val_win/rougeL": 0.9247446749599737, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0234375, "gen_val/bit_acc": 0.660373091016683, "gen_val/lev": 79.9375, "gen_val/lev_norm": 0.10738579394061341, "gen_val/token_acc": 0.15439138534235994, "gen_val/bleu": 0.7563453534813723, "gen_val/rouge1": 0.8810188969481645, "gen_val/rouge2": 0.8140144923263748, "gen_val/rougeL": 0.8799938135497251, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0078125, "gen_val_full/bit_acc": 0.1922525332231115, "gen_val_full/lev": 486.3046875, "gen_val_full/lev_norm": 0.6779508477220664, "gen_val_full/token_acc": 0.10090731364980701, "gen_val_full/bleu": 0.02823774761783071, "gen_val_full/rouge1": 0.37241513065937526, "gen_val_full/rouge2": 0.32244924416498166, "gen_val_full/rougeL": 0.3705662912674661 }, { "train/epoch_loss": 1.0782176945869464, "epoch": 21, "val/loss": 0.2791201578585315, "val/ppl": 1.3219661789518202, "val/token_acc": 0.9309271873223565, "val/seq_acc": 0.21008673305493095, "val_full/loss": 7.399267640810324, "val_full/ppl": 1634.7867403874793, "val_full/token_acc": 0.25198796681292995, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.0565346485361493, "epoch": 22, "val/loss": 0.26585048208320294, "val/ppl": 1.3045399920023264, "val/token_acc": 0.9344875387699335, "val/seq_acc": 0.22165114037905556, "val_full/loss": 7.236046359146563, "val_full/ppl": 1388.5931053784304, "val_full/token_acc": 0.25386575715189025, "val_full/seq_acc": 0.03233256351039261, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.0344027884285945, "epoch": 23, "val/loss": 0.2543554671917229, "val/ppl": 1.2896301440062594, "val/token_acc": 0.9373597550637771, "val/seq_acc": 0.2267908769675554, "val_full/loss": 7.222604529929678, "val_full/ppl": 1370.052761233226, "val_full/token_acc": 0.2567974094157773, "val_full/seq_acc": 0.020785219399538105, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 1.0146796226204184, "epoch": 24, "val/loss": 0.24066011455634917, "val/ppl": 1.2720885974254221, "val/token_acc": 0.9417777822102104, "val/seq_acc": 0.25666559588821075, "val_full/loss": 7.281986805169672, "val_full/ppl": 1453.8737213327208, "val_full/token_acc": 0.25557109735768074, "val_full/seq_acc": 0.03233256351039261, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.9966779969576886, "epoch": 25, "val/loss": 0.2246026765390596, "val/ppl": 1.251825237832509, "val/token_acc": 0.9464451336877063, "val/seq_acc": 0.2846129135881786, "val_full/loss": 7.116387522993351, "val_full/ppl": 1231.9918428316646, "val_full/token_acc": 0.2599589951905574, "val_full/seq_acc": 0.03464203233256351, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.3515625, "gen_val_win/bit_acc": 0.8906118484946585, "gen_val_win/lev": 2.671875, "gen_val_win/lev_norm": 0.021427998740196002, "gen_val_win/token_acc": 0.6612903225806451, "gen_val_win/bleu": 0.8784024424041862, "gen_val_win/rouge1": 0.9449852753643307, "gen_val_win/rouge2": 0.8990182762647996, "gen_val_win/rougeL": 0.9449852753643307, "gen_val/n": 128.0, "gen_val/seq_acc": 0.046875, "gen_val/bit_acc": 0.6874599913872368, "gen_val/lev": 58.109375, "gen_val/lev_norm": 0.07938169352255377, "gen_val/token_acc": 0.18579184260591156, "gen_val/bleu": 0.8196577702686751, "gen_val/rouge1": 0.9160907724183117, "gen_val/rouge2": 0.8688670240363819, "gen_val/rougeL": 0.915463390938692, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.03125, "gen_val_full/bit_acc": 0.23206474057465942, "gen_val_full/lev": 483.1796875, "gen_val_full/lev_norm": 0.6539841533324995, "gen_val_full/token_acc": 0.10666353824495542, "gen_val_full/bleu": 0.05903161788591839, "gen_val_full/rouge1": 0.3759865349791942, "gen_val_full/rouge2": 0.3329023756821975, "gen_val_full/rougeL": 0.37456435402142096 }, { "train/epoch_loss": 0.9793917166621905, "epoch": 26, "val/loss": 0.21895684697643214, "val/ppl": 1.2447775596131456, "val/token_acc": 0.9483798904967538, "val/seq_acc": 0.303886925795053, "val_full/loss": 7.201344437405392, "val_full/ppl": 1341.232756277221, "val_full/token_acc": 0.2597673839314798, "val_full/seq_acc": 0.03926096997690531, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.9648475554815849, "epoch": 27, "val/loss": 0.20409851503353646, "val/ppl": 1.226418968210861, "val/token_acc": 0.9513618094962651, "val/seq_acc": 0.3173787343398651, "val_full/loss": 7.1256100872549775, "val_full/ppl": 1243.406522172011, "val_full/token_acc": 0.2622966525513039, "val_full/seq_acc": 0.03464203233256351, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.9510538398179033, "epoch": 28, "val/loss": 0.19719471821938717, "val/ppl": 1.2179811807441776, "val/token_acc": 0.9541043771379562, "val/seq_acc": 0.3295856087375522, "val_full/loss": 7.073296022030505, "val_full/ppl": 1180.0310398017796, "val_full/token_acc": 0.2622008469217651, "val_full/seq_acc": 0.04157043879907621, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.9381572088398541, "epoch": 29, "val/loss": 0.1882738747841981, "val/ppl": 1.2071640818830376, "val/token_acc": 0.9563981609837341, "val/seq_acc": 0.3504657886283328, "val_full/loss": 6.9559371197714075, "val_full/ppl": 1049.3614546915421, "val_full/token_acc": 0.2663013278660254, "val_full/seq_acc": 0.03926096997690531, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.9259158424903032, "epoch": 30, "val/loss": 0.181659820264768, "val/ppl": 1.199206178844172, "val/token_acc": 0.9579439718363235, "val/seq_acc": 0.3543205910697077, "val_full/loss": 6.999073379926805, "val_full/ppl": 1095.6174667837352, "val_full/token_acc": 0.26467263216386594, "val_full/seq_acc": 0.03926096997690531, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.3984375, "gen_val_win/bit_acc": 0.9014790694660446, "gen_val_win/lev": 2.203125, "gen_val_win/lev_norm": 0.017975127493127647, "gen_val_win/token_acc": 0.6956305858987091, "gen_val_win/bleu": 0.896255462787838, "gen_val_win/rouge1": 0.9513856754915899, "gen_val_win/rouge2": 0.9123063624482264, "gen_val_win/rougeL": 0.9513856754915899, "gen_val/n": 128.0, "gen_val/seq_acc": 0.078125, "gen_val/bit_acc": 0.7226152792213928, "gen_val/lev": 45.4765625, "gen_val/lev_norm": 0.06171135535054479, "gen_val/token_acc": 0.24844632768361583, "gen_val/bleu": 0.8594338472983125, "gen_val/rouge1": 0.9361130067545652, "gen_val/rouge2": 0.9006335239291996, "gen_val/rougeL": 0.9357772854951092, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0390625, "gen_val_full/bit_acc": 0.22359502703770465, "gen_val_full/lev": 477.328125, "gen_val_full/lev_norm": 0.658418478765092, "gen_val_full/token_acc": 0.12446855299809412, "gen_val_full/bleu": 0.05438917441854921, "gen_val_full/rouge1": 0.38872508901583763, "gen_val_full/rouge2": 0.34907696878467487, "gen_val_full/rougeL": 0.38671287626787715 }, { "train/epoch_loss": 0.9156706309377999, "epoch": 31, "val/loss": 0.1739964501194439, "val/ppl": 1.1900513412727471, "val/token_acc": 0.9601679448694039, "val/seq_acc": 0.38644394474783167, "val_full/loss": 6.986951872042001, "val_full/ppl": 1082.4170968437247, "val_full/token_acc": 0.2681407959531702, "val_full/seq_acc": 0.04387990762124711, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.906422534637023, "epoch": 32, "val/loss": 0.17259013352136024, "val/ppl": 1.1883789285654494, "val/token_acc": 0.9610156475950175, "val/seq_acc": 0.3989720526823, "val_full/loss": 6.835739002711299, "val_full/ppl": 930.5157503615784, "val_full/token_acc": 0.2708041924543486, "val_full/seq_acc": 0.04157043879907621, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8981682576443489, "epoch": 33, "val/loss": 0.16830378367860402, "val/ppl": 1.1832960220718471, "val/token_acc": 0.9623520260095142, "val/seq_acc": 0.4115001606167684, "val_full/loss": 7.01746701292418, "val_full/ppl": 1115.956331755318, "val_full/token_acc": 0.2707658702025331, "val_full/seq_acc": 0.04849884526558892, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8888523331901379, "epoch": 34, "val/loss": 0.15626582138186765, "val/ppl": 1.1691369433852152, "val/token_acc": 0.9638878638888612, "val/seq_acc": 0.42081593318342436, "val_full/loss": 7.068620258220123, "val_full/ppl": 1174.5263726613832, "val_full/token_acc": 0.26835156833815554, "val_full/seq_acc": 0.04849884526558892, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8824114710315505, "epoch": 35, "val/loss": 0.15482015402687427, "val/ppl": 1.1674479814053997, "val/token_acc": 0.9650547017582352, "val/seq_acc": 0.43752007709604884, "val_full/loss": 6.989697526657916, "val_full/ppl": 1085.3931240438344, "val_full/token_acc": 0.2699611029144072, "val_full/seq_acc": 0.050808314087759814, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.484375, "gen_val_win/bit_acc": 0.9308343076123669, "gen_val_win/lev": 2.1015625, "gen_val_win/lev_norm": 0.017235736738250036, "gen_val_win/token_acc": 0.7428500373041532, "gen_val_win/bleu": 0.9117037871878669, "gen_val_win/rouge1": 0.9579478151586706, "gen_val_win/rouge2": 0.9244195606437472, "gen_val_win/rougeL": 0.9579478151586706, "gen_val/n": 128.0, "gen_val/seq_acc": 0.125, "gen_val/bit_acc": 0.7590632852041982, "gen_val/lev": 32.59375, "gen_val/lev_norm": 0.044923514189431024, "gen_val/token_acc": 0.3029486747799317, "gen_val/bleu": 0.8894940084077139, "gen_val/rouge1": 0.9515695988537732, "gen_val/rouge2": 0.9226529487191631, "gen_val/rougeL": 0.9515206176625507, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.25446396991204695, "gen_val_full/lev": 493.4375, "gen_val_full/lev_norm": 0.6476487763979369, "gen_val_full/token_acc": 0.11831676665061085, "gen_val_full/bleu": 0.0861712985998642, "gen_val_full/rouge1": 0.38915559189058024, "gen_val_full/rouge2": 0.35058193489371664, "gen_val_full/rougeL": 0.38681440719709487 }, { "train/epoch_loss": 0.8742316589688423, "epoch": 36, "val/loss": 0.15262643972140189, "val/ppl": 1.1648897411177896, "val/token_acc": 0.9656131882598159, "val/seq_acc": 0.43944747831673625, "val_full/loss": 6.9694766258215335, "val_full/ppl": 1063.6659095611283, "val_full/token_acc": 0.26948207476671326, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8685504611293574, "epoch": 37, "val/loss": 0.14920651916850258, "val/ppl": 1.1609127151828729, "val/token_acc": 0.966839863968645, "val/seq_acc": 0.448442017346611, "val_full/loss": 6.827416038341413, "val_full/ppl": 922.8032409178775, "val_full/token_acc": 0.2746364176359003, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8622932013430797, "epoch": 38, "val/loss": 0.1467437934359994, "val/ppl": 1.158057223156066, "val/token_acc": 0.9674482153364382, "val/seq_acc": 0.455187921619017, "val_full/loss": 6.851437503592711, "val_full/ppl": 945.2387145590523, "val_full/token_acc": 0.2743873229990994, "val_full/seq_acc": 0.04387990762124711, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8569676853772114, "epoch": 39, "val/loss": 0.14298660328994753, "val/ppl": 1.1537143455859091, "val/token_acc": 0.9685651883395997, "val/seq_acc": 0.4664310954063604, "val_full/loss": 6.90554231075993, "val_full/ppl": 997.7894785867333, "val_full/token_acc": 0.27356339458506584, "val_full/seq_acc": 0.050808314087759814, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8512335027245215, "epoch": 40, "val/loss": 0.14007736406774166, "val/ppl": 1.150362792159721, "val/token_acc": 0.9691236748411804, "val/seq_acc": 0.4702858978477353, "val_full/loss": 6.785554891236659, "val_full/ppl": 884.9710133871387, "val_full/token_acc": 0.2772423307593554, "val_full/seq_acc": 0.04849884526558892, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.5078125, "gen_val_win/bit_acc": 0.9173999870113002, "gen_val_win/lev": 1.7265625, "gen_val_win/lev_norm": 0.014513703180466476, "gen_val_win/token_acc": 0.7325523429710867, "gen_val_win/bleu": 0.917084303052808, "gen_val_win/rouge1": 0.9594692429684545, "gen_val_win/rouge2": 0.9269216146462497, "gen_val_win/rougeL": 0.9594692429684545, "gen_val/n": 128.0, "gen_val/seq_acc": 0.125, "gen_val/bit_acc": 0.7652119190142965, "gen_val/lev": 30.328125, "gen_val/lev_norm": 0.043247885824583616, "gen_val/token_acc": 0.361323032351521, "gen_val/bleu": 0.9052091652905336, "gen_val/rouge1": 0.9566517099456627, "gen_val/rouge2": 0.9320639203128686, "gen_val/rougeL": 0.9564223790591462, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.21251617392153513, "gen_val_full/lev": 488.0078125, "gen_val_full/lev_norm": 0.6578779553011542, "gen_val_full/token_acc": 0.12489704956155225, "gen_val_full/bleu": 0.049929408199586, "gen_val_full/rouge1": 0.4107873808386542, "gen_val_full/rouge2": 0.378455877108134, "gen_val_full/rougeL": 0.4092350802313837 }, { "train/epoch_loss": 0.8460126861075213, "epoch": 41, "val/loss": 0.1405003839436436, "val/ppl": 1.1508495214261214, "val/token_acc": 0.9693829721454857, "val/seq_acc": 0.4744619338258914, "val_full/loss": 6.7984301199726, "val_full/ppl": 896.4388849236499, "val_full/token_acc": 0.2767441414857537, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8413778880885117, "epoch": 42, "val/loss": 0.1342328940487782, "val/ppl": 1.1436591400387714, "val/token_acc": 0.9709387559713177, "val/seq_acc": 0.49662704786379697, "val_full/loss": 6.737154036889958, "val_full/ppl": 843.1577222554625, "val_full/token_acc": 0.2786794152024373, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8364077290691937, "epoch": 43, "val/loss": 0.1335013461176462, "val/ppl": 1.1428228045080175, "val/token_acc": 0.9709188100248327, "val/seq_acc": 0.4892386765178285, "val_full/loss": 6.816333497600069, "val_full/ppl": 912.6326982286059, "val_full/token_acc": 0.2762267910862442, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8323588939081701, "epoch": 44, "val/loss": 0.12981260511796236, "val/ppl": 1.1386149927088467, "val/token_acc": 0.9721255397871768, "val/seq_acc": 0.5053003533568905, "val_full/loss": 6.766124835932859, "val_full/ppl": 867.9419512056744, "val_full/token_acc": 0.2794458602387476, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8287213367416971, "epoch": 45, "val/loss": 0.12679556340549533, "val/ppl": 1.1351849207184035, "val/token_acc": 0.9725444046633623, "val/seq_acc": 0.5130099582396402, "val_full/loss": 6.791365876908928, "val_full/ppl": 890.1285379037652, "val_full/token_acc": 0.27745310314434074, "val_full/seq_acc": 0.04849884526558892, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.578125, "gen_val_win/bit_acc": 0.9343016451004617, "gen_val_win/lev": 1.4140625, "gen_val_win/lev_norm": 0.011782779034388364, "gen_val_win/token_acc": 0.7893686049413526, "gen_val_win/bleu": 0.933227582164221, "gen_val_win/rouge1": 0.9690531563329465, "gen_val_win/rouge2": 0.9430003341910561, "gen_val_win/rougeL": 0.9690531563329465, "gen_val/n": 128.0, "gen_val/seq_acc": 0.140625, "gen_val/bit_acc": 0.7848099353357104, "gen_val/lev": 27.6171875, "gen_val/lev_norm": 0.03986961318439014, "gen_val/token_acc": 0.3932807280824902, "gen_val/bleu": 0.9159549322699244, "gen_val/rouge1": 0.9605696778513796, "gen_val/rouge2": 0.9378652308904575, "gen_val/rougeL": 0.9605696778513796, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.2185478684028494, "gen_val_full/lev": 490.34375, "gen_val_full/lev_norm": 0.6588991424259258, "gen_val_full/token_acc": 0.13908332534279413, "gen_val_full/bleu": 0.05773467665606936, "gen_val_full/rouge1": 0.40528572243611893, "gen_val_full/rouge2": 0.3716209301269883, "gen_val_full/rougeL": 0.40357714803306044 }, { "train/epoch_loss": 0.8250167219716117, "epoch": 46, "val/loss": 0.12611736242025323, "val/ppl": 1.1344152981956164, "val/token_acc": 0.9731727019776406, "val/seq_acc": 0.5216832637327337, "val_full/loss": 6.802635088931575, "val_full/ppl": 900.2163190389992, "val_full/token_acc": 0.27998237176416485, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8211745716389872, "epoch": 47, "val/loss": 0.12300918295598615, "val/ppl": 1.1308948058570343, "val/token_acc": 0.9733422425227634, "val/seq_acc": 0.5191133954384838, "val_full/loss": 6.8057073808657, "val_full/ppl": 902.9862992901267, "val_full/token_acc": 0.28021230527505797, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8174262425845995, "epoch": 48, "val/loss": 0.12292003672488962, "val/ppl": 1.130793995340831, "val/token_acc": 0.9731926479241256, "val/seq_acc": 0.5152585929971089, "val_full/loss": 6.9009790437161085, "val_full/ppl": 993.2466716460823, "val_full/token_acc": 0.2754603460499339, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.812278275477916, "epoch": 49, "val/loss": 0.12178783749822161, "val/ppl": 1.129514435748662, "val/token_acc": 0.9741699993018919, "val/seq_acc": 0.5252168326373273, "val_full/loss": 6.919303217632068, "val_full/ppl": 1011.6148734984997, "val_full/token_acc": 0.27712736400390886, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8098846071973406, "epoch": 50, "val/loss": 0.12083248856658027, "val/ppl": 1.1284358706242488, "val/token_acc": 0.973551674960856, "val/seq_acc": 0.5281079344683585, "val_full/loss": 6.881966117141543, "val_full/ppl": 974.5405382831361, "val_full/token_acc": 0.27866025407652956, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.578125, "gen_val_win/bit_acc": 0.9368864183083024, "gen_val_win/lev": 1.3984375, "gen_val_win/lev_norm": 0.011525111958033919, "gen_val_win/token_acc": 0.7784983786480419, "gen_val_win/bleu": 0.9324601863584853, "gen_val_win/rouge1": 0.96818064759939, "gen_val_win/rouge2": 0.9428303629822684, "gen_val_win/rougeL": 0.96818064759939, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1640625, "gen_val/bit_acc": 0.7724822099527612, "gen_val/lev": 27.7890625, "gen_val/lev_norm": 0.03770991940904738, "gen_val/token_acc": 0.3615764500823085, "gen_val/bleu": 0.9140015755681794, "gen_val/rouge1": 0.9621254574622405, "gen_val/rouge2": 0.9411123589598672, "gen_val/rougeL": 0.9621254574622405, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.23854308312229444, "gen_val_full/lev": 469.140625, "gen_val_full/lev_norm": 0.6497838885509514, "gen_val_full/token_acc": 0.13850007367025194, "gen_val_full/bleu": 0.04929951284392544, "gen_val_full/rouge1": 0.4112228741559003, "gen_val_full/rouge2": 0.3750524297507468, "gen_val_full/rougeL": 0.4090682272171236 }, { "train/epoch_loss": 0.8071269010962393, "epoch": 51, "val/loss": 0.11746961495455713, "val/ppl": 1.1246474569604747, "val/token_acc": 0.9746885939105026, "val/seq_acc": 0.5335689045936396, "val_full/loss": 7.029665689249651, "val_full/ppl": 1129.6528919463135, "val_full/token_acc": 0.27655253022667614, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8043096029847637, "epoch": 52, "val/loss": 0.1202792218366547, "val/ppl": 1.1278117172777395, "val/token_acc": 0.9746686479640175, "val/seq_acc": 0.5284291680051397, "val_full/loss": 6.875676327866025, "val_full/ppl": 968.4301204219872, "val_full/token_acc": 0.2813428117036157, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.8011098281106449, "epoch": 53, "val/loss": 0.1161432288395908, "val/ppl": 1.1231567290482853, "val/token_acc": 0.9756061074488137, "val/seq_acc": 0.5451333119177643, "val_full/loss": 6.900267686916783, "val_full/ppl": 992.5403701187809, "val_full/token_acc": 0.2817068730958631, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7980428501554855, "epoch": 54, "val/loss": 0.11542519394024225, "val/ppl": 1.1223505527853734, "val/token_acc": 0.9757856209671789, "val/seq_acc": 0.5486668808223578, "val_full/loss": 6.884415073698481, "val_full/ppl": 976.9300704603672, "val_full/token_acc": 0.27860277069880623, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7956641035484256, "epoch": 55, "val/loss": 0.11362590436557554, "val/ppl": 1.1203329348204372, "val/token_acc": 0.9756360263685413, "val/seq_acc": 0.5464182460648892, "val_full/loss": 6.70676928148173, "val_full/ppl": 817.9238844525337, "val_full/token_acc": 0.28323976316848376, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.5625, "gen_val_win/bit_acc": 0.9245338184653369, "gen_val_win/lev": 1.3671875, "gen_val_win/lev_norm": 0.011252138313495592, "gen_val_win/token_acc": 0.8012980529206191, "gen_val_win/bleu": 0.9334476386155498, "gen_val_win/rouge1": 0.9686966012150466, "gen_val_win/rouge2": 0.9433306062794882, "gen_val_win/rougeL": 0.9686966012150466, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1796875, "gen_val/bit_acc": 0.8005998639984457, "gen_val/lev": 23.9140625, "gen_val/lev_norm": 0.03381555990721037, "gen_val/token_acc": 0.42408300453017683, "gen_val/bleu": 0.9243783846420387, "gen_val/rouge1": 0.9660114466639844, "gen_val/rouge2": 0.9467287778457162, "gen_val/rougeL": 0.9658799603488537, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.22798498607277132, "gen_val_full/lev": 471.296875, "gen_val_full/lev_norm": 0.6503911209146307, "gen_val_full/token_acc": 0.13822584630590562, "gen_val_full/bleu": 0.04608787360585937, "gen_val_full/rouge1": 0.4157333718696041, "gen_val_full/rouge2": 0.3803645305417662, "gen_val_full/rougeL": 0.4135861160547048 }, { "train/epoch_loss": 0.7926069560490939, "epoch": 56, "val/loss": 0.11021173498116003, "val/ppl": 1.1165144505991784, "val/token_acc": 0.9772815669535558, "val/seq_acc": 0.567619659492451, "val_full/loss": 6.834271529920098, "val_full/ppl": 929.1512452480719, "val_full/token_acc": 0.2817451953476786, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7896959133576277, "epoch": 57, "val/loss": 0.11451808098831463, "val/ppl": 1.1213329156880147, "val/token_acc": 0.9759252425925741, "val/seq_acc": 0.5522004497269515, "val_full/loss": 6.838782852349154, "val_full/ppl": 933.3524153921015, "val_full/token_acc": 0.2815727452145088, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.787601800184892, "epoch": 58, "val/loss": 0.11016986206759564, "val/ppl": 1.1164676998648966, "val/token_acc": 0.9775508372311037, "val/seq_acc": 0.5766141985223258, "val_full/loss": 6.835710859807622, "val_full/ppl": 930.4895633149376, "val_full/token_acc": 0.282032612236295, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7850704155955232, "epoch": 59, "val/loss": 0.11107015603559081, "val/ppl": 1.1174733016012126, "val/token_acc": 0.9769125669435829, "val/seq_acc": 0.5647285576614198, "val_full/loss": 6.780155645226963, "val_full/ppl": 880.2057132646873, "val_full/token_acc": 0.28233919025081916, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.783727664751305, "epoch": 60, "val/loss": 0.10900171244545447, "val/ppl": 1.1151642599977807, "val/token_acc": 0.9774909993916486, "val/seq_acc": 0.5717956954706072, "val_full/loss": 6.902688276025599, "val_full/ppl": 994.945812648237, "val_full/token_acc": 0.277912970166127, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.578125, "gen_val_win/bit_acc": 0.938020020800832, "gen_val_win/lev": 1.1640625, "gen_val_win/lev_norm": 0.009588767776364098, "gen_val_win/token_acc": 0.7939635819406335, "gen_val_win/bleu": 0.940691768154969, "gen_val_win/rouge1": 0.9716373786563488, "gen_val_win/rouge2": 0.9492735742171591, "gen_val_win/rougeL": 0.9716373786563488, "gen_val/n": 128.0, "gen_val/seq_acc": 0.203125, "gen_val/bit_acc": 0.8121411192214112, "gen_val/lev": 22.375, "gen_val/lev_norm": 0.03272813063869122, "gen_val/token_acc": 0.4550282595985188, "gen_val/bleu": 0.9291785902814881, "gen_val/rouge1": 0.9673117615588889, "gen_val/rouge2": 0.9491464941255068, "gen_val/rougeL": 0.9671864068265679, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.21297833030625665, "gen_val_full/lev": 477.8671875, "gen_val_full/lev_norm": 0.6598604662721664, "gen_val_full/token_acc": 0.14494934296318587, "gen_val_full/bleu": 0.035802724055477714, "gen_val_full/rouge1": 0.41812380775289, "gen_val_full/rouge2": 0.3875853743023587, "gen_val_full/rougeL": 0.41658939958675123 }, { "train/epoch_loss": 0.7810055152733725, "epoch": 61, "val/loss": 0.10764384350319813, "val/ppl": 1.113651040692871, "val/token_acc": 0.9775109453381337, "val/seq_acc": 0.5692258271763572, "val_full/loss": 6.863788944988407, "val_full/ppl": 956.986174833291, "val_full/token_acc": 0.2814961007108778, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7789576417787414, "epoch": 62, "val/loss": 0.10761799824538489, "val/ppl": 1.113622258466554, "val/token_acc": 0.9779597291340467, "val/seq_acc": 0.5833601027947317, "val_full/loss": 6.912707224582767, "val_full/ppl": 1004.9642267467332, "val_full/token_acc": 0.28040391653413554, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7767515249383122, "epoch": 63, "val/loss": 0.10578877643563482, "val/ppl": 1.1115870583272753, "val/token_acc": 0.9784484048229298, "val/seq_acc": 0.5817539351108255, "val_full/loss": 6.866050631478855, "val_full/ppl": 959.1530269829126, "val_full/token_acc": 0.2813428117036157, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7750067960591684, "epoch": 64, "val/loss": 0.1043469808817616, "val/ppl": 1.1099855318631382, "val/token_acc": 0.9789669994315405, "val/seq_acc": 0.5933183424349502, "val_full/loss": 6.958347379835789, "val_full/ppl": 1051.893739205093, "val_full/token_acc": 0.2801739830232424, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.773212856931282, "epoch": 65, "val/loss": 0.10502199797520886, "val/ppl": 1.1107350440089205, "val/token_acc": 0.9786678102342652, "val/seq_acc": 0.5872149052361066, "val_full/loss": 6.872812113783556, "val_full/ppl": 965.6602978092244, "val_full/token_acc": 0.2822242234953726, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6171875, "gen_val_win/bit_acc": 0.9453617629851134, "gen_val_win/lev": 1.2421875, "gen_val_win/lev_norm": 0.010281928570813866, "gen_val_win/token_acc": 0.8282147315855181, "gen_val_win/bleu": 0.9418137357482569, "gen_val_win/rouge1": 0.9725544904451727, "gen_val_win/rouge2": 0.950202918292495, "gen_val_win/rougeL": 0.9725544904451727, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2109375, "gen_val/bit_acc": 0.8043572766185125, "gen_val/lev": 22.640625, "gen_val/lev_norm": 0.033196262197214076, "gen_val/token_acc": 0.4307707298060996, "gen_val/bleu": 0.9292221174997839, "gen_val/rouge1": 0.9670763794403113, "gen_val/rouge2": 0.9492955565709471, "gen_val/rougeL": 0.9670023273076098, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.22169309187456165, "gen_val_full/lev": 475.546875, "gen_val_full/lev_norm": 0.6555373834376012, "gen_val_full/token_acc": 0.15454454253611558, "gen_val_full/bleu": 0.03498185149945715, "gen_val_full/rouge1": 0.4163295378717208, "gen_val_full/rouge2": 0.3870074054727635, "gen_val_full/rougeL": 0.4146622911427836 }, { "train/epoch_loss": 0.770219087600708, "epoch": 66, "val/loss": 0.10270934588980173, "val/ppl": 1.1081692683098765, "val/token_acc": 0.9789071615920855, "val/seq_acc": 0.5958882107292002, "val_full/loss": 6.7476459508948246, "val_full/ppl": 852.0506308081428, "val_full/token_acc": 0.2828757017762364, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7695136395773091, "epoch": 67, "val/loss": 0.10249686111562957, "val/ppl": 1.1079338242281824, "val/token_acc": 0.9789470534850555, "val/seq_acc": 0.5897847735303565, "val_full/loss": 6.823881184852172, "val_full/ppl": 919.5470251752109, "val_full/token_acc": 0.2799057272605338, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7675651704284021, "epoch": 68, "val/loss": 0.10315197228474701, "val/ppl": 1.108659881849339, "val/token_acc": 0.9794457021471812, "val/seq_acc": 0.5978156119498875, "val_full/loss": 6.876186043754431, "val_full/ppl": 968.923870466608, "val_full/token_acc": 0.27892850983923817, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7661644361560184, "epoch": 69, "val/loss": 0.10053432976784282, "val/ppl": 1.1057616015915688, "val/token_acc": 0.9796252156655464, "val/seq_acc": 0.6007067137809188, "val_full/loss": 6.889591272226906, "val_full/ppl": 981.999964522737, "val_full/token_acc": 0.2792734101055778, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7648059015262156, "epoch": 70, "val/loss": 0.10301445250728923, "val/ppl": 1.1085074296719455, "val/token_acc": 0.9794457021471812, "val/seq_acc": 0.6016704143912625, "val_full/loss": 6.940487938071241, "val_full/ppl": 1033.274265829926, "val_full/token_acc": 0.27785548678840366, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.59375, "gen_val_win/bit_acc": 0.9391946810585864, "gen_val_win/lev": 1.265625, "gen_val_win/lev_norm": 0.010328533726839211, "gen_val_win/token_acc": 0.8033483258370815, "gen_val_win/bleu": 0.938008330578863, "gen_val_win/rouge1": 0.9714033530879299, "gen_val_win/rouge2": 0.9482913011947609, "gen_val_win/rougeL": 0.9714033530879299, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2109375, "gen_val/bit_acc": 0.8127124995430117, "gen_val/lev": 21.078125, "gen_val/lev_norm": 0.030384673034467572, "gen_val/token_acc": 0.43574846206425155, "gen_val/bleu": 0.9337218708154749, "gen_val/rouge1": 0.9698726033156129, "gen_val/rouge2": 0.953315243246635, "gen_val/rougeL": 0.9698234680954871, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.22683614581507724, "gen_val_full/lev": 473.53125, "gen_val_full/lev_norm": 0.6547270387112042, "gen_val_full/token_acc": 0.15837149114075255, "gen_val_full/bleu": 0.039209644314140445, "gen_val_full/rouge1": 0.41601138208210403, "gen_val_full/rouge2": 0.38557293246915453, "gen_val_full/rougeL": 0.41427622562892696 }, { "train/epoch_loss": 0.7633525406630557, "epoch": 71, "val/loss": 0.10103169528619434, "val/ppl": 1.106311706073878, "val/token_acc": 0.979734918371214, "val/seq_acc": 0.607773851590106, "val_full/loss": 6.825737718317078, "val_full/ppl": 921.2557806901024, "val_full/token_acc": 0.2811895226963536, "val_full/seq_acc": 0.07621247113163972, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.761559683783096, "epoch": 72, "val/loss": 0.09984984324986611, "val/ppl": 1.105004981660795, "val/token_acc": 0.9800540535149744, "val/seq_acc": 0.6068101509797623, "val_full/loss": 6.900935931182817, "val_full/ppl": 993.2038511889399, "val_full/token_acc": 0.2799248883864416, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7608953455737106, "epoch": 73, "val/loss": 0.09975712660121734, "val/ppl": 1.1049025340515282, "val/token_acc": 0.9801138913544295, "val/seq_acc": 0.607773851590106, "val_full/loss": 6.916914064745445, "val_full/ppl": 1009.2008557803075, "val_full/token_acc": 0.2796183103719175, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7595704202342807, "epoch": 74, "val/loss": 0.09861468147455649, "val/ppl": 1.103640964310382, "val/token_acc": 0.9802036481136122, "val/seq_acc": 0.6161259235464183, "val_full/loss": 6.89143582998812, "val_full/ppl": 983.8129917815077, "val_full/token_acc": 0.28040391653413554, "val_full/seq_acc": 0.07621247113163972, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7577390655912366, "epoch": 75, "val/loss": 0.09781357404855692, "val/ppl": 1.1027571833872314, "val/token_acc": 0.9803033778460373, "val/seq_acc": 0.6129135881786059, "val_full/loss": 6.873157762531377, "val_full/ppl": 965.9941347738234, "val_full/token_acc": 0.2799248883864416, "val_full/seq_acc": 0.07621247113163972, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6484375, "gen_val_win/bit_acc": 0.9477277160132631, "gen_val_win/lev": 1.2265625, "gen_val_win/lev_norm": 0.009933992764308992, "gen_val_win/token_acc": 0.8263368315842079, "gen_val_win/bleu": 0.948940788847061, "gen_val_win/rouge1": 0.975223076168123, "gen_val_win/rouge2": 0.9558755429763787, "gen_val_win/rougeL": 0.975223076168123, "gen_val/n": 128.0, "gen_val/seq_acc": 0.265625, "gen_val/bit_acc": 0.827022685349198, "gen_val/lev": 19.4921875, "gen_val/lev_norm": 0.028093445624839997, "gen_val/token_acc": 0.4537181996086106, "gen_val/bleu": 0.9395818107737776, "gen_val/rouge1": 0.9727136218976249, "gen_val/rouge2": 0.9578099264852933, "gen_val/rougeL": 0.9726374023854297, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0625, "gen_val_full/bit_acc": 0.21014270463943457, "gen_val_full/lev": 489.9609375, "gen_val_full/lev_norm": 0.66450092980431, "gen_val_full/token_acc": 0.15215470697948366, "gen_val_full/bleu": 0.04617370962311767, "gen_val_full/rouge1": 0.4170446543161026, "gen_val_full/rouge2": 0.38553693213167584, "gen_val_full/rougeL": 0.41477079029860786 }, { "train/epoch_loss": 0.7566669477786209, "epoch": 76, "val/loss": 0.09789315239757873, "val/ppl": 1.1028449424750726, "val/token_acc": 0.9803332967657648, "val/seq_acc": 0.6103437198843559, "val_full/loss": 6.994019109031597, "val_full/ppl": 1090.093889894653, "val_full/token_acc": 0.27912012109831574, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7555111049119374, "epoch": 77, "val/loss": 0.09780210671195319, "val/ppl": 1.1027445377719232, "val/token_acc": 0.9801837021671271, "val/seq_acc": 0.6080950851268873, "val_full/loss": 6.951348628901684, "val_full/ppl": 1044.5574991194273, "val_full/token_acc": 0.27925424897967005, "val_full/seq_acc": 0.07621247113163972, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7553404783667471, "epoch": 78, "val/loss": 0.09664009059922707, "val/ppl": 1.1014638750702632, "val/token_acc": 0.9807621346151928, "val/seq_acc": 0.6161259235464183, "val_full/loss": 6.919909712175937, "val_full/ppl": 1012.2285984913908, "val_full/token_acc": 0.2790817988465002, "val_full/seq_acc": 0.07621247113163972, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7534938090459962, "epoch": 79, "val/loss": 0.09773809523128357, "val/ppl": 1.102673951720443, "val/token_acc": 0.98051281028413, "val/seq_acc": 0.6135560552521683, "val_full/loss": 6.890389452878001, "val_full/ppl": 982.7840907893716, "val_full/token_acc": 0.2804805610377666, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7526507014942883, "epoch": 80, "val/loss": 0.09675926189245894, "val/ppl": 1.1015951457663975, "val/token_acc": 0.9809615940800431, "val/seq_acc": 0.6231930613556055, "val_full/loss": 6.919802679324187, "val_full/ppl": 1012.1202625757253, "val_full/token_acc": 0.27892850983923817, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.609375, "gen_val_win/bit_acc": 0.9441672625625691, "gen_val_win/lev": 1.1015625, "gen_val_win/lev_norm": 0.008989873971881496, "gen_val_win/token_acc": 0.8328753434923807, "gen_val_win/bleu": 0.946943297086583, "gen_val_win/rouge1": 0.9749287667537709, "gen_val_win/rouge2": 0.955063203209249, "gen_val_win/rougeL": 0.9749287667537709, "gen_val/n": 128.0, "gen_val/seq_acc": 0.265625, "gen_val/bit_acc": 0.8182594325790529, "gen_val/lev": 19.109375, "gen_val/lev_norm": 0.02715051442269032, "gen_val/token_acc": 0.4533705389925099, "gen_val/bleu": 0.9404578053678531, "gen_val/rouge1": 0.9727231661591267, "gen_val/rouge2": 0.9571244233976623, "gen_val/rougeL": 0.9727231661591267, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.22054357961377077, "gen_val_full/lev": 486.0859375, "gen_val_full/lev_norm": 0.6598224035758996, "gen_val_full/token_acc": 0.14803816570927855, "gen_val_full/bleu": 0.05276136759418174, "gen_val_full/rouge1": 0.4132071764317172, "gen_val_full/rouge2": 0.38286578743053684, "gen_val_full/rougeL": 0.41167112979211257 }, { "train/epoch_loss": 0.7518132294799919, "epoch": 81, "val/loss": 0.09610618934236072, "val/ppl": 1.1008759590814785, "val/token_acc": 0.9809217021870731, "val/seq_acc": 0.6225505942820431, "val_full/loss": 6.88060500356876, "val_full/ppl": 973.214980251413, "val_full/token_acc": 0.28153442296269326, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7509538390095395, "epoch": 82, "val/loss": 0.09580683929558709, "val/ppl": 1.1005464611317093, "val/token_acc": 0.9810712967857107, "val/seq_acc": 0.6228718278188243, "val_full/loss": 6.879593206303053, "val_full/ppl": 972.2307819839707, "val_full/token_acc": 0.2807104945486597, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7502511255164396, "epoch": 83, "val/loss": 0.09586762260139281, "val/ppl": 1.100613358016896, "val/token_acc": 0.9811710265181358, "val/seq_acc": 0.6228718278188243, "val_full/loss": 6.861478906594302, "val_full/ppl": 954.778051433644, "val_full/token_acc": 0.2818601621031252, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7495843078430157, "epoch": 84, "val/loss": 0.09611286429218711, "val/ppl": 1.1008833073977953, "val/token_acc": 0.9812508103040759, "val/seq_acc": 0.623835528429168, "val_full/loss": 6.905877705311464, "val_full/ppl": 998.1241878681257, "val_full/token_acc": 0.28038475540822777, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7487430059701724, "epoch": 85, "val/loss": 0.09677635241795515, "val/ppl": 1.1016139727672039, "val/token_acc": 0.9810413778659832, "val/seq_acc": 0.6206231930613556, "val_full/loss": 6.920329460590354, "val_full/ppl": 1012.653569024749, "val_full/token_acc": 0.2796374714978252, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6171875, "gen_val_win/bit_acc": 0.9393328564926198, "gen_val_win/lev": 1.1796875, "gen_val_win/lev_norm": 0.009645241015390197, "gen_val_win/token_acc": 0.8133899575318511, "gen_val_win/bleu": 0.9448595993938573, "gen_val_win/rouge1": 0.9734506608756318, "gen_val_win/rouge2": 0.9526852076518891, "gen_val_win/rougeL": 0.9734506608756318, "gen_val/n": 128.0, "gen_val/seq_acc": 0.234375, "gen_val/bit_acc": 0.8268218177815286, "gen_val/lev": 18.5390625, "gen_val/lev_norm": 0.026290384138187156, "gen_val/token_acc": 0.46954774854230974, "gen_val/bleu": 0.9403833721542829, "gen_val/rouge1": 0.973054856354748, "gen_val/rouge2": 0.9578737121374634, "gen_val/rougeL": 0.9729291279973673, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0625, "gen_val_full/bit_acc": 0.22329975119804882, "gen_val_full/lev": 484.0390625, "gen_val_full/lev_norm": 0.6541975507644586, "gen_val_full/token_acc": 0.15453790137334442, "gen_val_full/bleu": 0.05572146709220061, "gen_val_full/rouge1": 0.41206496777430407, "gen_val_full/rouge2": 0.3784958360749287, "gen_val_full/rougeL": 0.4091047743639005 }, { "train/epoch_loss": 0.7480881860725898, "epoch": 86, "val/loss": 0.09574928028417014, "val/ppl": 1.1004831165884303, "val/token_acc": 0.9812807292238035, "val/seq_acc": 0.6280115644073241, "val_full/loss": 6.897095996642013, "val_full/ppl": 989.3973264948235, "val_full/token_acc": 0.28044223878595104, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7478880351618341, "epoch": 87, "val/loss": 0.09570968604646837, "val/ppl": 1.1004395446609296, "val/token_acc": 0.9814004049027136, "val/seq_acc": 0.6280115644073241, "val_full/loss": 6.902811401229187, "val_full/ppl": 995.0683230958849, "val_full/token_acc": 0.27996321063825713, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7470159013372407, "epoch": 88, "val/loss": 0.09502952671478068, "val/ppl": 1.0996913249188351, "val/token_acc": 0.9813006751702885, "val/seq_acc": 0.6254416961130742, "val_full/loss": 6.884519711409492, "val_full/ppl": 977.0322995351734, "val_full/token_acc": 0.2811703615704459, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7463050220078067, "epoch": 89, "val/loss": 0.09567432609191591, "val/ppl": 1.1004006338565888, "val/token_acc": 0.9812707562505609, "val/seq_acc": 0.6244779955027305, "val_full/loss": 6.896156203294756, "val_full/ppl": 988.4679342563153, "val_full/token_acc": 0.28141945620724673, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7465946442468505, "epoch": 90, "val/loss": 0.09506868299140774, "val/ppl": 1.0997343855796, "val/token_acc": 0.9813006751702885, "val/seq_acc": 0.6247992290395117, "val_full/loss": 6.906040050944643, "val_full/ppl": 998.2862421254414, "val_full/token_acc": 0.28080630017819846, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6328125, "gen_val_win/bit_acc": 0.948234376016128, "gen_val_win/lev": 1.1484375, "gen_val_win/lev_norm": 0.009260526504873589, "gen_val_win/token_acc": 0.8229270729270729, "gen_val_win/bleu": 0.9468972326978264, "gen_val_win/rouge1": 0.9742900944698534, "gen_val_win/rouge2": 0.9546527250468961, "gen_val_win/rougeL": 0.9742900944698534, "gen_val/n": 128.0, "gen_val/seq_acc": 0.265625, "gen_val/bit_acc": 0.8273257521073688, "gen_val/lev": 18.1015625, "gen_val/lev_norm": 0.026206693651223804, "gen_val/token_acc": 0.4726399843237153, "gen_val/bleu": 0.9417559536654114, "gen_val/rouge1": 0.9730018711800703, "gen_val/rouge2": 0.957789926367469, "gen_val/rougeL": 0.9729252780428154, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.22266309172484944, "gen_val_full/lev": 494.6328125, "gen_val_full/lev_norm": 0.6550999153123273, "gen_val_full/token_acc": 0.148574430488562, "gen_val_full/bleu": 0.06439043101259838, "gen_val_full/rouge1": 0.4115580029978681, "gen_val_full/rouge2": 0.37889859722086316, "gen_val_full/rougeL": 0.40949815517329763 }, { "train/epoch_loss": 0.7456189798298025, "epoch": 91, "val/loss": 0.09507677279289921, "val/ppl": 1.0997432822484587, "val/token_acc": 0.9816995940999891, "val/seq_acc": 0.633151300995824, "val_full/loss": 6.899730501758033, "val_full/ppl": 992.0073353444957, "val_full/token_acc": 0.28044223878595104, "val_full/seq_acc": 0.07852193995381063, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7457025095708947, "epoch": 92, "val/loss": 0.09457407618651793, "val/ppl": 1.099190583963997, "val/token_acc": 0.9816297832872914, "val/seq_acc": 0.6328300674590427, "val_full/loss": 6.905121589319588, "val_full/ppl": 997.3697754553878, "val_full/token_acc": 0.2804805610377666, "val_full/seq_acc": 0.07621247113163972, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7453115549765323, "epoch": 93, "val/loss": 0.09485687859387389, "val/ppl": 1.0995014816665223, "val/token_acc": 0.9814203508491987, "val/seq_acc": 0.6270478637969804, "val_full/loss": 6.906126949644561, "val_full/ppl": 998.3729956713582, "val_full/token_acc": 0.28057636666730534, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7451746613902047, "epoch": 94, "val/loss": 0.09463875001102968, "val/ppl": 1.099261675121773, "val/token_acc": 0.9814303238224412, "val/seq_acc": 0.6273690973337617, "val_full/loss": 6.889909825945122, "val_full/ppl": 982.3128340929605, "val_full/token_acc": 0.28140029508133896, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7441930962619638, "epoch": 95, "val/loss": 0.0942737746624616, "val/ppl": 1.0988605449144992, "val/token_acc": 0.9816597022070189, "val/seq_acc": 0.6341150016061677, "val_full/loss": 6.899455210269405, "val_full/ppl": 991.7342817548076, "val_full/token_acc": 0.28082546130410624, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.640625, "gen_val_win/bit_acc": 0.9460986795030247, "gen_val_win/lev": 1.046875, "gen_val_win/lev_norm": 0.00861263128527809, "gen_val_win/token_acc": 0.8261304021983512, "gen_val_win/bleu": 0.9483143445977356, "gen_val_win/rouge1": 0.9750161739508517, "gen_val_win/rouge2": 0.9557847030923851, "gen_val_win/rougeL": 0.9750161739508517, "gen_val/n": 128.0, "gen_val/seq_acc": 0.265625, "gen_val/bit_acc": 0.82669561013817, "gen_val/lev": 17.625, "gen_val/lev_norm": 0.025413935762036446, "gen_val/token_acc": 0.4777794110441472, "gen_val/bleu": 0.9439005564007774, "gen_val/rouge1": 0.9738912170828397, "gen_val/rouge2": 0.9588290590817293, "gen_val/rougeL": 0.9738149975706445, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.22933777150644621, "gen_val_full/lev": 489.8984375, "gen_val_full/lev_norm": 0.6522661916190245, "gen_val_full/token_acc": 0.14838494202967697, "gen_val_full/bleu": 0.06345583442366877, "gen_val_full/rouge1": 0.4121691395479764, "gen_val_full/rouge2": 0.37892516646498936, "gen_val_full/rougeL": 0.4102100214860513 }, { "train/epoch_loss": 0.7443976519708324, "epoch": 96, "val/loss": 0.09471141776647103, "val/ppl": 1.0993415589027993, "val/token_acc": 0.9816198103140489, "val/seq_acc": 0.6302601991647928, "val_full/loss": 6.905406311674874, "val_full/ppl": 997.6537893575809, "val_full/token_acc": 0.28049972216367436, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7445039637665499, "epoch": 97, "val/loss": 0.09430421752241855, "val/ppl": 1.0988939978813796, "val/token_acc": 0.9816597022070189, "val/seq_acc": 0.6299389656280115, "val_full/loss": 6.9064326294813085, "val_full/ppl": 998.6782248145091, "val_full/token_acc": 0.28023146640096575, "val_full/seq_acc": 0.07390300230946882, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7445861585716952, "epoch": 98, "val/loss": 0.09419949818377452, "val/ppl": 1.098778928453784, "val/token_acc": 0.9818192697788992, "val/seq_acc": 0.6309026662383553, "val_full/loss": 6.926152871055203, "val_full/ppl": 1018.5678704042873, "val_full/token_acc": 0.2800398551418881, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7440448505920068, "epoch": 99, "val/loss": 0.09448873190798932, "val/ppl": 1.0990967783395866, "val/token_acc": 0.9817893508591716, "val/seq_acc": 0.6325088339222615, "val_full/loss": 6.908056983990879, "val_full/ppl": 1000.3017505262758, "val_full/token_acc": 0.28021230527505797, "val_full/seq_acc": 0.07159353348729793, "system/peak_gpu_mem_gb": 1.0640783309936523 }, { "train/epoch_loss": 0.7439111903718583, "epoch": 100, "val/loss": 0.09459350283412615, "val/ppl": 1.0992119377595415, "val/token_acc": 0.9816995940999891, "val/seq_acc": 0.6309026662383553, "val_full/loss": 6.9084148329868365, "val_full/ppl": 1000.6597715582687, "val_full/token_acc": 0.28069133342275193, "val_full/seq_acc": 0.07621247113163972, "system/peak_gpu_mem_gb": 1.0640783309936523, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.640625, "gen_val_win/bit_acc": 0.9482520325203252, "gen_val_win/lev": 1.1015625, "gen_val_win/lev_norm": 0.009058394145180868, "gen_val_win/token_acc": 0.8271296527604297, "gen_val_win/bleu": 0.9478921762989313, "gen_val_win/rouge1": 0.9744954786963448, "gen_val_win/rouge2": 0.9552503553129715, "gen_val_win/rougeL": 0.9744954786963448, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2734375, "gen_val/bit_acc": 0.8298359229487305, "gen_val/lev": 16.8984375, "gen_val/lev_norm": 0.02450121651292563, "gen_val/token_acc": 0.47211184694628405, "gen_val/bleu": 0.944144459080651, "gen_val/rouge1": 0.9744453304041033, "gen_val/rouge2": 0.9598925526179642, "gen_val/rougeL": 0.9743196020467227, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0625, "gen_val_full/bit_acc": 0.22729673904011205, "gen_val_full/lev": 499.2578125, "gen_val_full/lev_norm": 0.6536408560029752, "gen_val_full/token_acc": 0.14381632078129403, "gen_val_full/bleu": 0.07404126646290975, "gen_val_full/rouge1": 0.41020808499516986, "gen_val_full/rouge2": 0.37856127839435855, "gen_val_full/rougeL": 0.40831992325988636 } ] }