{ "config": "C4", "description": "RMSNorm instead of LayerNorm", "parameters": 2761728, "windowed": true, "window": { "src_window": 1024, "src_stride": 768, "train_examples": 25654 }, "best_epoch": 95, "best_val_loss": 0.0932135848039557, "test": { "loss": 0.1126672476263856, "ppl": 1.119259434743591, "token_acc": 0.977756783806789, "seq_acc": 0.606318347509113 }, "test_full": { "loss": 7.018010509312258, "ppl": 1116.5630148409782, "token_acc": 0.2683339203945051, "seq_acc": 0.052752293577981654 }, "gen_test_win": { "n": 3292.0, "seq_acc": 0.6995747266099636, "bit_acc": 0.9582018678813191, "lev": 0.9808626974483596, "lev_norm": 0.007909028972323765, "token_acc": 0.8427260560193343, "bleu": 0.9554879432805189, "rouge1": 0.9792703987573325, "rouge2": 0.9625123754993086, "rougeL": 0.9792703987573325 }, "gen_test": { "n": 500.0, "seq_acc": 0.248, "bit_acc": 0.8167653084271503, "lev": 17.93, "lev_norm": 0.026327583729092437, "token_acc": 0.4516043187347932, "bleu": 0.9372483058859963, "rouge1": 0.9687715619097614, "rouge2": 0.9505274209983015, "rougeL": 0.9686895865238564 }, "gen_test_full": { "n": 500.0, "seq_acc": 0.05, "bit_acc": 0.263841918757732, "lev": 491.948, "lev_norm": 0.6245622013006646, "token_acc": 0.15179801598236428, "bleu": 0.11087393700481472, "rouge1": 0.4145830318817317, "rouge2": 0.37154642514857267, "rougeL": 0.4111817651294763 }, "checkpoint": "/home2/winterdew/ANLP_A1/ANLP-M26-Assignment-1/outputs/checkpoints/C4_best.pt", "history": [ { "train/epoch_loss": 6.990067713873048, "epoch": 1, "val/loss": 5.990486202313605, "val/ppl": 399.608853451543, "val/token_acc": 0.11997486810742887, "val/seq_acc": 0.0, "val_full/loss": 6.30906125153289, "val_full/ppl": 549.5288372468726, "val_full/token_acc": 0.06744716319530936, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 5.808596715665518, "epoch": 2, "val/loss": 5.088282229766707, "val/ppl": 162.11115302983052, "val/token_acc": 0.2173808977670513, "val/seq_acc": 0.0, "val_full/loss": 6.501812747298281, "val_full/ppl": 666.3484602485395, "val_full/token_acc": 0.06913334227519209, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 5.090657870371145, "epoch": 3, "val/loss": 4.335293712437731, "val/ppl": 76.34737977198775, "val/token_acc": 0.29339489982148376, "val/seq_acc": 0.0, "val_full/loss": 6.849186146146698, "val_full/ppl": 943.1130380674202, "val_full/token_acc": 0.08084079020483244, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 4.489753499589954, "epoch": 4, "val/loss": 3.621723493590183, "val/ppl": 37.401974367058806, "val/token_acc": 0.365669036909974, "val/seq_acc": 0.0, "val_full/loss": 7.243584840076453, "val_full/ppl": 1399.1005433227, "val_full/token_acc": 0.09680200808599514, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 3.933835908361801, "epoch": 5, "val/loss": 2.952026005618524, "val/ppl": 19.144701736419375, "val/token_acc": 0.4401671470315445, "val/seq_acc": 0.00032123353678124, "val_full/loss": 7.198871080352181, "val_full/ppl": 1337.9195078916825, "val_full/token_acc": 0.11672957903006381, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.0, "gen_val_win/bit_acc": 0.6387846554611425, "gen_val_win/lev": 64.03125, "gen_val_win/lev_norm": 0.4963064714108548, "gen_val_win/token_acc": 0.08706416340004126, "gen_val_win/bleu": 0.09380111127643914, "gen_val_win/rouge1": 0.37930834928566876, "gen_val_win/rouge2": 0.14595221056948834, "gen_val_win/rougeL": 0.3664165562594861, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0, "gen_val/bit_acc": 0.5377282811076338, "gen_val/lev": 441.2265625, "gen_val/lev_norm": 0.5589239589823511, "gen_val/token_acc": 0.017570314036582966, "gen_val/bleu": 0.08385887667364605, "gen_val/rouge1": 0.3727609931593137, "gen_val/rouge2": 0.13819005515740826, "gen_val/rougeL": 0.33527310934752785, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0, "gen_val_full/bit_acc": 0.1435844406213014, "gen_val_full/lev": 540.7578125, "gen_val_full/lev_norm": 0.8030595160097848, "gen_val_full/token_acc": 0.01555299539170507, "gen_val_full/bleu": 0.0041165023106339215, "gen_val_full/rouge1": 0.19130493457033776, "gen_val_full/rouge2": 0.05911226171874653, "gen_val_full/rougeL": 0.16643577359948933 }, { "train/epoch_loss": 3.4454829776078983, "epoch": 6, "val/loss": 2.3897298866945578, "val/ppl": 10.910546461214045, "val/token_acc": 0.5099879327023765, "val/seq_acc": 0.00032123353678124, "val_full/loss": 7.524517558776754, "val_full/ppl": 1852.9190858732256, "val_full/token_acc": 0.13353388645116787, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 2.972313640123591, "epoch": 7, "val/loss": 2.0050376346312433, "val/ppl": 7.426373380322305, "val/token_acc": 0.5756898804240508, "val/seq_acc": 0.0019274012206874397, "val_full/loss": 7.596150005868095, "val_full/ppl": 1990.5176427310705, "val_full/token_acc": 0.1511812834122133, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 2.5695207344920856, "epoch": 8, "val/loss": 1.4335808138655182, "val/ppl": 4.193689158051733, "val/token_acc": 0.6746816128292328, "val/seq_acc": 0.0032123353678123997, "val_full/loss": 7.557392062982621, "val_full/ppl": 1914.8452002764632, "val_full/token_acc": 0.17833259882350688, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 2.2488441728891577, "epoch": 9, "val/loss": 1.1444901273116261, "val/ppl": 3.1408395202175727, "val/token_acc": 0.7309591008367324, "val/seq_acc": 0.007709604882749759, "val_full/loss": 7.449582661695951, "val_full/ppl": 1719.1455303621667, "val_full/token_acc": 0.19458123359328594, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.9930625803155495, "epoch": 10, "val/loss": 0.9739837425829907, "val/ppl": 2.6484743112121603, "val/token_acc": 0.7697639397233498, "val/seq_acc": 0.013491808544812078, "val_full/loss": 7.484455536607331, "val_full/ppl": 1780.15467683183, "val_full/token_acc": 0.20119182203146257, "val_full/seq_acc": 0.0, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.0703125, "gen_val_win/bit_acc": 0.7492851108078248, "gen_val_win/lev": 16.8046875, "gen_val_win/lev_norm": 0.1389714094850395, "gen_val_win/token_acc": 0.305961216183864, "gen_val_win/bleu": 0.5650555368893185, "gen_val_win/rouge1": 0.7565411782605853, "gen_val_win/rouge2": 0.6148089395726135, "gen_val_win/rougeL": 0.7562087314520747, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0, "gen_val/bit_acc": 0.5785370983060754, "gen_val/lev": 206.4921875, "gen_val/lev_norm": 0.26693523873825287, "gen_val/token_acc": 0.05400439384861194, "gen_val/bleu": 0.4526807181423185, "gen_val/rouge1": 0.6902693141789473, "gen_val/rouge2": 0.5442096037348608, "gen_val/rougeL": 0.6835946133167051, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0, "gen_val_full/bit_acc": 0.15023454331379327, "gen_val_full/lev": 511.2578125, "gen_val_full/lev_norm": 0.7361138524453392, "gen_val_full/token_acc": 0.04881352530978779, "gen_val_full/bleu": 0.009854570904565966, "gen_val_full/rouge1": 0.29959395207586836, "gen_val_full/rouge2": 0.21987424189193802, "gen_val_full/rougeL": 0.29394207159326274 }, { "train/epoch_loss": 1.7899117778958822, "epoch": 11, "val/loss": 0.764511505119953, "val/ppl": 2.1479448582685126, "val/token_acc": 0.813684913883376, "val/seq_acc": 0.027947317699967878, "val_full/loss": 7.389033578675583, "val_full/ppl": 1618.1415505418681, "val_full/token_acc": 0.2172680066680718, "val_full/seq_acc": 0.006928406466512702, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.6327564407167887, "epoch": 12, "val/loss": 0.650159046766466, "val/ppl": 1.9158455138177084, "val/token_acc": 0.8405521037987055, "val/seq_acc": 0.037584323803405074, "val_full/loss": 7.266312031151201, "val_full/ppl": 1431.2622570608655, "val_full/token_acc": 0.22638870260016478, "val_full/seq_acc": 0.006928406466512702, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.5073545891745133, "epoch": 13, "val/loss": 0.5588379947130386, "val/ppl": 1.7486393910656248, "val/token_acc": 0.8620737800560481, "val/seq_acc": 0.06264053967234179, "val_full/loss": 7.24454020183851, "val_full/ppl": 1400.4378291772355, "val_full/token_acc": 0.2343405698518845, "val_full/seq_acc": 0.006928406466512702, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.4064734056406187, "epoch": 14, "val/loss": 0.49055496920240677, "val/ppl": 1.6332223566017523, "val/token_acc": 0.8768038615352395, "val/seq_acc": 0.07259877931256023, "val_full/loss": 7.296382124944912, "val_full/ppl": 1474.9540635675344, "val_full/token_acc": 0.23763628350801894, "val_full/seq_acc": 0.011547344110854504, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.3260356609958068, "epoch": 15, "val/loss": 0.44080999293185, "val/ppl": 1.5539654098778293, "val/token_acc": 0.890097834867509, "val/seq_acc": 0.08962415676196595, "val_full/loss": 7.173227955124643, "val_full/ppl": 1304.0472221350244, "val_full/token_acc": 0.24303972101400678, "val_full/seq_acc": 0.011547344110854504, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.171875, "gen_val_win/bit_acc": 0.8253590838509317, "gen_val_win/lev": 6.5859375, "gen_val_win/lev_norm": 0.05457663399349526, "gen_val_win/token_acc": 0.5144337527757217, "gen_val_win/bleu": 0.7674147417541625, "gen_val_win/rouge1": 0.8847424432345669, "gen_val_win/rouge2": 0.7994880673740512, "gen_val_win/rougeL": 0.8847424432345669, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0078125, "gen_val/bit_acc": 0.633121277747699, "gen_val/lev": 114.3515625, "gen_val/lev_norm": 0.15686205728257135, "gen_val/token_acc": 0.11781076066790352, "gen_val/bleu": 0.6704870645429086, "gen_val/rouge1": 0.8318539382184819, "gen_val/rouge2": 0.7420135522432885, "gen_val/rougeL": 0.8301308713615351, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0078125, "gen_val_full/bit_acc": 0.18498270477422887, "gen_val_full/lev": 495.7109375, "gen_val_full/lev_norm": 0.6922923399270288, "gen_val_full/token_acc": 0.08457221620557251, "gen_val_full/bleu": 0.023410420432707323, "gen_val_full/rouge1": 0.3566934145377266, "gen_val_full/rouge2": 0.30034385888740767, "gen_val_full/rougeL": 0.35439762793694857 }, { "train/epoch_loss": 1.263341325476877, "epoch": 16, "val/loss": 0.40030239790889344, "val/ppl": 1.4922758905267268, "val/token_acc": 0.901377267604791, "val/seq_acc": 0.11371667202055895, "val_full/loss": 7.231771182624691, "val_full/ppl": 1382.6692964228841, "val_full/token_acc": 0.24461093333844297, "val_full/seq_acc": 0.009237875288683603, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.2101433101734913, "epoch": 17, "val/loss": 0.36429695081991864, "val/ppl": 1.4395016118807757, "val/token_acc": 0.9093157543058312, "val/seq_acc": 0.13588178605846452, "val_full/loss": 7.036192747154573, "val_full/ppl": 1137.0503172337358, "val_full/token_acc": 0.2512215217766196, "val_full/seq_acc": 0.020785219399538105, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.1677903561223475, "epoch": 18, "val/loss": 0.33787233046198706, "val/ppl": 1.4019615042016376, "val/token_acc": 0.9167256734250182, "val/seq_acc": 0.15772566655958883, "val_full/loss": 7.135746996493514, "val_full/ppl": 1256.0749219630095, "val_full/token_acc": 0.2501293375998774, "val_full/seq_acc": 0.020785219399538105, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.129640483499465, "epoch": 19, "val/loss": 0.3143790873607524, "val/ppl": 1.3694087637168686, "val/token_acc": 0.9222905924943403, "val/seq_acc": 0.16543527144233858, "val_full/loss": 7.132581967703922, "val_full/ppl": 1252.1056933479804, "val_full/token_acc": 0.24913295905267394, "val_full/seq_acc": 0.018475750577367205, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.0997393859592162, "epoch": 20, "val/loss": 0.30170142417466633, "val/ppl": 1.3521574449076452, "val/token_acc": 0.9252625385206091, "val/seq_acc": 0.1824606488917443, "val_full/loss": 7.114956501250264, "val_full/ppl": 1230.230096566517, "val_full/token_acc": 0.2536549847669049, "val_full/seq_acc": 0.020785219399538105, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.2890625, "gen_val_win/bit_acc": 0.8683282655107718, "gen_val_win/lev": 3.7890625, "gen_val_win/lev_norm": 0.030505676094387, "gen_val_win/token_acc": 0.6253083374444992, "gen_val_win/bleu": 0.8433827705015139, "gen_val_win/rouge1": 0.9268961503680291, "gen_val_win/rouge2": 0.8707928876701019, "gen_val_win/rougeL": 0.9268961503680291, "gen_val/n": 128.0, "gen_val/seq_acc": 0.015625, "gen_val/bit_acc": 0.6656033737024222, "gen_val/lev": 78.7421875, "gen_val/lev_norm": 0.10749500842603092, "gen_val/token_acc": 0.1840791504039212, "gen_val/bleu": 0.7588247375591239, "gen_val/rouge1": 0.8818750846760744, "gen_val/rouge2": 0.8151887812217983, "gen_val/rougeL": 0.8813756918160264, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0078125, "gen_val_full/bit_acc": 0.20244844722036334, "gen_val_full/lev": 489.0390625, "gen_val_full/lev_norm": 0.6763109601593527, "gen_val_full/token_acc": 0.10535279805352799, "gen_val_full/bleu": 0.03376119260530743, "gen_val_full/rouge1": 0.37822565989772217, "gen_val_full/rouge2": 0.32997807917962224, "gen_val_full/rougeL": 0.37730803840509763 }, { "train/epoch_loss": 1.0716000804282781, "epoch": 21, "val/loss": 0.2810715791016891, "val/ppl": 1.3245484105275007, "val/token_acc": 0.9319444305930927, "val/seq_acc": 0.21233536781239962, "val_full/loss": 7.099568769879668, "val_full/ppl": 1211.4445504573537, "val_full/token_acc": 0.2570465040525781, "val_full/seq_acc": 0.023094688221709007, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.0505478887486637, "epoch": 22, "val/loss": 0.26504633746258577, "val/ppl": 1.3034913748618013, "val/token_acc": 0.9345772955291161, "val/seq_acc": 0.22165114037905556, "val_full/loss": 7.167601994074422, "val_full/ppl": 1296.7313020928514, "val_full/token_acc": 0.254363946425492, "val_full/seq_acc": 0.03002309468822171, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.0295759485546787, "epoch": 23, "val/loss": 0.2576565782486298, "val/ppl": 1.2938943908377316, "val/token_acc": 0.9369708091073191, "val/seq_acc": 0.2348217153870864, "val_full/loss": 7.04133735976451, "val_full/ppl": 1142.9150736485906, "val_full/token_acc": 0.2592691946578781, "val_full/seq_acc": 0.020785219399538105, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 1.0112593726921557, "epoch": 24, "val/loss": 0.24611412751618894, "val/ppl": 1.2790455395027296, "val/token_acc": 0.9407106740732615, "val/seq_acc": 0.26244779955027303, "val_full/loss": 7.086547138166089, "val_full/ppl": 1195.7718293397481, "val_full/token_acc": 0.2588859721397229, "val_full/seq_acc": 0.020785219399538105, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.9931319905932705, "epoch": 25, "val/loss": 0.22943604752507424, "val/ppl": 1.2578904194453981, "val/token_acc": 0.9448394849956617, "val/seq_acc": 0.2814005782203662, "val_full/loss": 6.964411501944855, "val_full/ppl": 1058.2919313877078, "val_full/token_acc": 0.26122362950046946, "val_full/seq_acc": 0.03002309468822171, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.3203125, "gen_val_win/bit_acc": 0.8913505188067445, "gen_val_win/lev": 2.8203125, "gen_val_win/lev_norm": 0.023024143260475416, "gen_val_win/token_acc": 0.6475267213522247, "gen_val_win/bleu": 0.8634815938944187, "gen_val_win/rouge1": 0.9386105557481881, "gen_val_win/rouge2": 0.8856625970585926, "gen_val_win/rougeL": 0.9386105557481881, "gen_val/n": 128.0, "gen_val/seq_acc": 0.046875, "gen_val/bit_acc": 0.6912871955406081, "gen_val/lev": 58.96875, "gen_val/lev_norm": 0.08226844525398483, "gen_val/token_acc": 0.21943631889306775, "gen_val/bleu": 0.8150753897431956, "gen_val/rouge1": 0.9097833998633333, "gen_val/rouge2": 0.8587412257854484, "gen_val/rougeL": 0.9094444553812215, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0234375, "gen_val_full/bit_acc": 0.20268988905636134, "gen_val_full/lev": 490.4140625, "gen_val_full/lev_norm": 0.6752682424380203, "gen_val_full/token_acc": 0.12101199436701791, "gen_val_full/bleu": 0.03637202072426942, "gen_val_full/rouge1": 0.3856757398366277, "gen_val_full/rouge2": 0.34115292959527105, "gen_val_full/rougeL": 0.3851913497193271 }, { "train/epoch_loss": 0.9760361287064683, "epoch": 26, "val/loss": 0.22184142813794927, "val/ppl": 1.2483734052728483, "val/token_acc": 0.9472230256006223, "val/seq_acc": 0.29489238676517826, "val_full/loss": 7.031865102080898, "val_full/ppl": 1132.1401993169106, "val_full/token_acc": 0.26283316407672114, "val_full/seq_acc": 0.03002309468822171, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.962581092877281, "epoch": 27, "val/loss": 0.20895706102039643, "val/ppl": 1.2323920797357482, "val/token_acc": 0.9510825662454747, "val/seq_acc": 0.31288146482492774, "val_full/loss": 6.912256189642453, "val_full/ppl": 1004.5110549725396, "val_full/token_acc": 0.2634654812316772, "val_full/seq_acc": 0.02771362586605081, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.9470950852605767, "epoch": 28, "val/loss": 0.20178571229438924, "val/ppl": 1.22358578062675, "val/token_acc": 0.9530871338672199, "val/seq_acc": 0.3398650819145519, "val_full/loss": 6.945412197733239, "val_full/ppl": 1038.3749247946582, "val_full/token_acc": 0.2652474659410987, "val_full/seq_acc": 0.03695150115473441, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.934725672080927, "epoch": 29, "val/loss": 0.19083178918017912, "val/ppl": 1.2102558568284472, "val/token_acc": 0.9554507285256953, "val/seq_acc": 0.3472534532605204, "val_full/loss": 6.779224384568587, "val_full/ppl": 879.3863938718173, "val_full/token_acc": 0.26934794688535896, "val_full/seq_acc": 0.03464203233256351, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.921988307389238, "epoch": 30, "val/loss": 0.18518231760142903, "val/ppl": 1.2034378280262246, "val/token_acc": 0.957505161013653, "val/seq_acc": 0.36010279473177, "val_full/loss": 6.807996836019084, "val_full/ppl": 905.0560142822472, "val_full/token_acc": 0.27019103642530035, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.4375, "gen_val_win/bit_acc": 0.8994325551232166, "gen_val_win/lev": 2.171875, "gen_val_win/lev_norm": 0.017877067887702304, "gen_val_win/token_acc": 0.7325697211155379, "gen_val_win/bleu": 0.901279674584352, "gen_val_win/rouge1": 0.9534502977530437, "gen_val_win/rouge2": 0.9157412564720053, "gen_val_win/rougeL": 0.9534502977530437, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0625, "gen_val/bit_acc": 0.7186097393446311, "gen_val/lev": 45.171875, "gen_val/lev_norm": 0.0638238869731173, "gen_val/token_acc": 0.24619108643030044, "gen_val/bleu": 0.857138052597624, "gen_val/rouge1": 0.9330133148475706, "gen_val/rouge2": 0.8940674819910522, "gen_val/rougeL": 0.9326622893626925, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0390625, "gen_val_full/bit_acc": 0.2450701871657754, "gen_val_full/lev": 492.4296875, "gen_val_full/lev_norm": 0.6552991809734925, "gen_val_full/token_acc": 0.13143518518518518, "gen_val_full/bleu": 0.08311012030600878, "gen_val_full/rouge1": 0.39448841705277415, "gen_val_full/rouge2": 0.3507552570648182, "gen_val_full/rougeL": 0.3913003813590169 }, { "train/epoch_loss": 0.9137877008861437, "epoch": 31, "val/loss": 0.1794783621736656, "val/ppl": 1.19659301211557, "val/token_acc": 0.9593002961973053, "val/seq_acc": 0.3893350465788628, "val_full/loss": 6.844233668732874, "val_full/ppl": 938.4538388601528, "val_full/token_acc": 0.26858150184904866, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.9030639915692241, "epoch": 32, "val/loss": 0.17156400974526026, "val/ppl": 1.1871601301178074, "val/token_acc": 0.9616539178825383, "val/seq_acc": 0.4118213941535496, "val_full/loss": 6.791085795138822, "val_full/ppl": 889.8792640374102, "val_full/token_acc": 0.27147483186112015, "val_full/seq_acc": 0.04387990762124711, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8945086822842719, "epoch": 33, "val/loss": 0.17019000776572246, "val/ppl": 1.1855300898447216, "val/token_acc": 0.9612649719260803, "val/seq_acc": 0.3944747831673627, "val_full/loss": 6.652946651832762, "val_full/ppl": 775.0648101313008, "val_full/token_acc": 0.27557531280538045, "val_full/seq_acc": 0.04157043879907621, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8876807639129144, "epoch": 34, "val/loss": 0.1588434852657696, "val/ppl": 1.172154472877116, "val/token_acc": 0.9639277557818312, "val/seq_acc": 0.42274333440411177, "val_full/loss": 6.7804881955488705, "val_full/ppl": 880.4984746342241, "val_full/token_acc": 0.27413822836229856, "val_full/seq_acc": 0.04157043879907621, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.879046155924809, "epoch": 35, "val/loss": 0.15554862238251332, "val/ppl": 1.1682987401544995, "val/token_acc": 0.9655433774471183, "val/seq_acc": 0.4381625441696113, "val_full/loss": 6.855723383399758, "val_full/ppl": 949.2985879286971, "val_full/token_acc": 0.275134606909502, "val_full/seq_acc": 0.050808314087759814, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.5625, "gen_val_win/bit_acc": 0.9325398115047124, "gen_val_win/lev": 1.5859375, "gen_val_win/lev_norm": 0.013151726064149105, "gen_val_win/token_acc": 0.8064355200798204, "gen_val_win/bleu": 0.9270441176249467, "gen_val_win/rouge1": 0.9662520024611324, "gen_val_win/rouge2": 0.9382457410187327, "gen_val_win/rougeL": 0.9662520024611324, "gen_val/n": 128.0, "gen_val/seq_acc": 0.0859375, "gen_val/bit_acc": 0.7606496096197407, "gen_val/lev": 32.453125, "gen_val/lev_norm": 0.05008299398426211, "gen_val/token_acc": 0.3479352813330113, "gen_val/bleu": 0.8938183428026037, "gen_val/rouge1": 0.9466503209763502, "gen_val/rouge2": 0.916952581899266, "gen_val/rougeL": 0.94650813371809, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0234375, "gen_val_full/bit_acc": 0.2777788931464103, "gen_val_full/lev": 500.1640625, "gen_val_full/lev_norm": 0.6418831915247356, "gen_val_full/token_acc": 0.12605360755225894, "gen_val_full/bleu": 0.11029921006728952, "gen_val_full/rouge1": 0.390432164973526, "gen_val_full/rouge2": 0.3431238122984872, "gen_val_full/rougeL": 0.3854217182080359 }, { "train/epoch_loss": 0.8721080425374228, "epoch": 36, "val/loss": 0.1536751798195737, "val/ppl": 1.1661120485286156, "val/token_acc": 0.9661716747613966, "val/seq_acc": 0.4416961130742049, "val_full/loss": 6.884481763398417, "val_full/ppl": 976.9952238061296, "val_full/token_acc": 0.2718388932533676, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8669680997617821, "epoch": 37, "val/loss": 0.15108451801793896, "val/ppl": 1.1630949564081623, "val/token_acc": 0.9667501072094623, "val/seq_acc": 0.45872149052361066, "val_full/loss": 6.763667047294449, "val_full/ppl": 865.8113526921411, "val_full/token_acc": 0.2743681618731917, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8606726491540447, "epoch": 38, "val/loss": 0.14761672792391797, "val/ppl": 1.1590685726018253, "val/token_acc": 0.9673285396575281, "val/seq_acc": 0.4596851911339544, "val_full/loss": 6.796287816278334, "val_full/ppl": 894.5204962081091, "val_full/token_acc": 0.277357297514802, "val_full/seq_acc": 0.04849884526558892, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8550009199805985, "epoch": 39, "val/loss": 0.14455289163802618, "val/ppl": 1.155522810841721, "val/token_acc": 0.9679967288647765, "val/seq_acc": 0.4616125923546418, "val_full/loss": 6.9220916105405355, "val_full/ppl": 1014.4395896168472, "val_full/token_acc": 0.2733717833259882, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8486695164754207, "epoch": 40, "val/loss": 0.14110447896284078, "val/ppl": 1.1515449539223652, "val/token_acc": 0.9693131613327881, "val/seq_acc": 0.4783167362672663, "val_full/loss": 6.805751241880473, "val_full/ppl": 903.0259060541298, "val_full/token_acc": 0.27726149188526317, "val_full/seq_acc": 0.04849884526558892, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.5703125, "gen_val_win/bit_acc": 0.9290323104927838, "gen_val_win/lev": 1.421875, "gen_val_win/lev_norm": 0.012020312107521698, "gen_val_win/token_acc": 0.780603340812765, "gen_val_win/bleu": 0.9302935749766778, "gen_val_win/rouge1": 0.9670626681356478, "gen_val_win/rouge2": 0.9402155495834914, "gen_val_win/rougeL": 0.9670626681356478, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1171875, "gen_val/bit_acc": 0.7647158097763848, "gen_val/lev": 31.0078125, "gen_val/lev_norm": 0.04805006222021078, "gen_val/token_acc": 0.3425086772078673, "gen_val/bleu": 0.9023945030093932, "gen_val/rouge1": 0.9518489310293933, "gen_val/rouge2": 0.9251854675072462, "gen_val/rougeL": 0.9517225669277901, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.03125, "gen_val_full/bit_acc": 0.25490635928584576, "gen_val_full/lev": 506.390625, "gen_val_full/lev_norm": 0.6556178206024416, "gen_val_full/token_acc": 0.1345693519606563, "gen_val_full/bleu": 0.10329667206422248, "gen_val_full/rouge1": 0.3968283980348358, "gen_val_full/rouge2": 0.35719301093260636, "gen_val_full/rougeL": 0.39434770467619223 }, { "train/epoch_loss": 0.8443385738684352, "epoch": 41, "val/loss": 0.13863290924709273, "val/ppl": 1.1487023445885434, "val/token_acc": 0.9704600532556771, "val/seq_acc": 0.49405717956954703, "val_full/loss": 6.887876351458162, "val_full/ppl": 980.3173555715614, "val_full/token_acc": 0.27488551227270114, "val_full/seq_acc": 0.050808314087759814, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8392858692535439, "epoch": 42, "val/loss": 0.14005327605645732, "val/ppl": 1.1503350825415386, "val/token_acc": 0.9703004856837969, "val/seq_acc": 0.49341471249598456, "val_full/loss": 6.787500718542222, "val_full/ppl": 886.6946905948521, "val_full/token_acc": 0.2773381363888942, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8346499131802014, "epoch": 43, "val/loss": 0.1349753907024269, "val/ppl": 1.144508618451578, "val/token_acc": 0.9707592424529525, "val/seq_acc": 0.49405717956954703, "val_full/loss": 6.784884401526184, "val_full/ppl": 884.3778483063571, "val_full/token_acc": 0.2769357527448313, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8297731333837247, "epoch": 44, "val/loss": 0.12998312917330948, "val/ppl": 1.138809170510405, "val/token_acc": 0.9721255397871768, "val/seq_acc": 0.5056215868936716, "val_full/loss": 6.794875955661155, "val_full/ppl": 893.258449075217, "val_full/token_acc": 0.278066259173389, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8255203184938787, "epoch": 45, "val/loss": 0.12887180933761302, "val/ppl": 1.1375442922624674, "val/token_acc": 0.9722551884393295, "val/seq_acc": 0.5075489881143591, "val_full/loss": 6.758241754129223, "val_full/ppl": 861.1267913277483, "val_full/token_acc": 0.27856444844699074, "val_full/seq_acc": 0.046189376443418015, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6171875, "gen_val_win/bit_acc": 0.9323052159209158, "gen_val_win/lev": 1.359375, "gen_val_win/lev_norm": 0.01138481483285739, "gen_val_win/token_acc": 0.7803295057413879, "gen_val_win/bleu": 0.9315021809021891, "gen_val_win/rouge1": 0.9675047946447325, "gen_val_win/rouge2": 0.9421216436300084, "gen_val_win/rougeL": 0.9675047946447325, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1328125, "gen_val/bit_acc": 0.7850964918660751, "gen_val/lev": 26.2109375, "gen_val/lev_norm": 0.0381783158658728, "gen_val/token_acc": 0.4119619084637062, "gen_val/bleu": 0.9146404948905696, "gen_val/rouge1": 0.9578586821061348, "gen_val/rouge2": 0.9334976589909693, "gen_val/rougeL": 0.9576914054599553, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0390625, "gen_val_full/bit_acc": 0.25087987919333876, "gen_val_full/lev": 481.1953125, "gen_val_full/lev_norm": 0.6476865247857184, "gen_val_full/token_acc": 0.14947564670240038, "gen_val_full/bleu": 0.08066496293439519, "gen_val_full/rouge1": 0.4046582131040958, "gen_val_full/rouge2": 0.3634622979946576, "gen_val_full/rougeL": 0.4008218178497883 }, { "train/epoch_loss": 0.8222747914808944, "epoch": 46, "val/loss": 0.12569449061110796, "val/ppl": 1.1339356873602442, "val/token_acc": 0.9732325398170957, "val/seq_acc": 0.5184709283649213, "val_full/loss": 6.7449742462492095, "val_full/ppl": 849.7772414442645, "val_full/token_acc": 0.28044223878595104, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8187131611189046, "epoch": 47, "val/loss": 0.1255246817805875, "val/ppl": 1.1337431514149026, "val/token_acc": 0.9733621884692484, "val/seq_acc": 0.5197558625120463, "val_full/loss": 6.71542764040315, "val_full/ppl": 825.0365104143483, "val_full/token_acc": 0.2793692157351166, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8147504094830178, "epoch": 48, "val/loss": 0.12274895509852869, "val/ppl": 1.1306005538126465, "val/token_acc": 0.9736713506397663, "val/seq_acc": 0.5245743655637649, "val_full/loss": 6.86805708578436, "val_full/ppl": 961.0794557028462, "val_full/token_acc": 0.2769740749966468, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8106390549952253, "epoch": 49, "val/loss": 0.12299048530508688, "val/ppl": 1.130873660978431, "val/token_acc": 0.973900729024344, "val/seq_acc": 0.5226469643430774, "val_full/loss": 6.860290916788021, "val_full/ppl": 953.6444583231315, "val_full/token_acc": 0.277357297514802, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.8075192716353552, "epoch": 50, "val/loss": 0.12012071325404443, "val/ppl": 1.1276329636083524, "val/token_acc": 0.9741500533554068, "val/seq_acc": 0.5248955991005461, "val_full/loss": 6.774583649691506, "val_full/ppl": 875.3148495494207, "val_full/token_acc": 0.2793692157351166, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.640625, "gen_val_win/bit_acc": 0.9531483770246536, "gen_val_win/lev": 1.1171875, "gen_val_win/lev_norm": 0.009315030828151301, "gen_val_win/token_acc": 0.8449825262106839, "gen_val_win/bleu": 0.9430299212089701, "gen_val_win/rouge1": 0.9730548479415811, "gen_val_win/rouge2": 0.9521518225861325, "gen_val_win/rougeL": 0.9730548479415811, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1796875, "gen_val/bit_acc": 0.7906203279405866, "gen_val/lev": 24.4140625, "gen_val/lev_norm": 0.03772097419657422, "gen_val/token_acc": 0.42660103323910714, "gen_val/bleu": 0.922705189396041, "gen_val/rouge1": 0.9620870458131665, "gen_val/rouge2": 0.9415936317770713, "gen_val/rougeL": 0.9619467157692413, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.25293307040539276, "gen_val_full/lev": 499.6640625, "gen_val_full/lev_norm": 0.6499442482745769, "gen_val_full/token_acc": 0.14533563448087927, "gen_val_full/bleu": 0.08698750913622745, "gen_val_full/rouge1": 0.40105203647039417, "gen_val_full/rouge2": 0.36045596768241667, "gen_val_full/rougeL": 0.3970530593932996 }, { "train/epoch_loss": 0.8036111270697635, "epoch": 51, "val/loss": 0.11705588645785546, "val/ppl": 1.1241822544991804, "val/token_acc": 0.9757557020474514, "val/seq_acc": 0.5460970125281079, "val_full/loss": 6.720507958453889, "val_full/ppl": 829.2386232877545, "val_full/token_acc": 0.28197512885857173, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.802009754347385, "epoch": 52, "val/loss": 0.11703468110697354, "val/ppl": 1.1241584160727705, "val/token_acc": 0.9754265939304485, "val/seq_acc": 0.5435271442338581, "val_full/loss": 6.857643612638679, "val_full/ppl": 951.1232101194863, "val_full/token_acc": 0.28007817739370366, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7984260820093892, "epoch": 53, "val/loss": 0.11540293127402085, "val/ppl": 1.1223255665477645, "val/token_acc": 0.9759451885390592, "val/seq_acc": 0.5512367491166078, "val_full/loss": 6.755989797898025, "val_full/ppl": 859.1897533651303, "val_full/token_acc": 0.282032612236295, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7952976674213077, "epoch": 54, "val/loss": 0.1126369097363333, "val/ppl": 1.1192254792889906, "val/token_acc": 0.9768028642379153, "val/seq_acc": 0.5656922582717636, "val_full/loss": 6.887612287191746, "val_full/ppl": 980.058522963933, "val_full/token_acc": 0.27942669911283985, "val_full/seq_acc": 0.05542725173210162, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7932472062527093, "epoch": 55, "val/loss": 0.11222307121674871, "val/ppl": 1.118762396500917, "val/token_acc": 0.9766931615322476, "val/seq_acc": 0.5489881143591391, "val_full/loss": 6.857043255642951, "val_full/ppl": 950.5523680181624, "val_full/token_acc": 0.28026978865278124, "val_full/seq_acc": 0.04849884526558892, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6015625, "gen_val_win/bit_acc": 0.9434130957799597, "gen_val_win/lev": 1.015625, "gen_val_win/lev_norm": 0.008404055441860925, "gen_val_win/token_acc": 0.827508736894658, "gen_val_win/bleu": 0.9447877671870976, "gen_val_win/rouge1": 0.9736760000347455, "gen_val_win/rouge2": 0.9526720945687155, "gen_val_win/rougeL": 0.9736760000347455, "gen_val/n": 128.0, "gen_val/seq_acc": 0.1796875, "gen_val/bit_acc": 0.8026254845079349, "gen_val/lev": 21.2109375, "gen_val/lev_norm": 0.033298694550533796, "gen_val/token_acc": 0.44966344746853965, "gen_val/bleu": 0.9283722944468638, "gen_val/rouge1": 0.9649679552018245, "gen_val/rouge2": 0.9460301560892727, "gen_val/rougeL": 0.9649679552018245, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.2768017234821729, "gen_val_full/lev": 526.359375, "gen_val_full/lev_norm": 0.6382252887415918, "gen_val_full/token_acc": 0.1395535893826664, "gen_val_full/bleu": 0.13738766267312283, "gen_val_full/rouge1": 0.396691173858192, "gen_val_full/rouge2": 0.3501046502117075, "gen_val_full/rougeL": 0.3911794123701447 }, { "train/epoch_loss": 0.7909607488912834, "epoch": 56, "val/loss": 0.11199673482031663, "val/ppl": 1.1185092085055395, "val/token_acc": 0.9770023237027655, "val/seq_acc": 0.5615162222936074, "val_full/loss": 6.779683652805189, "val_full/ppl": 879.7903608677269, "val_full/token_acc": 0.2812086838222614, "val_full/seq_acc": 0.050808314087759814, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7882721620307599, "epoch": 57, "val/loss": 0.11073786029849343, "val/ppl": 1.1171020316758935, "val/token_acc": 0.9772516480338284, "val/seq_acc": 0.5631223899775136, "val_full/loss": 6.796833534125965, "val_full/ppl": 895.0087852298605, "val_full/token_acc": 0.28026978865278124, "val_full/seq_acc": 0.050808314087759814, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7850316878863404, "epoch": 58, "val/loss": 0.11057474171143662, "val/ppl": 1.1169198264318232, "val/token_acc": 0.9772416750605858, "val/seq_acc": 0.5605525216832637, "val_full/loss": 6.9179297541148514, "val_full/ppl": 1010.2264110957698, "val_full/token_acc": 0.277357297514802, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7827186526503052, "epoch": 59, "val/loss": 0.11041448752372282, "val/ppl": 1.116740849693566, "val/token_acc": 0.977830080481894, "val/seq_acc": 0.576935432059107, "val_full/loss": 6.806891927657169, "val_full/ppl": 904.0565625770178, "val_full/token_acc": 0.27967579374964074, "val_full/seq_acc": 0.053117782909930716, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7811763912365026, "epoch": 60, "val/loss": 0.10726854023204538, "val/ppl": 1.1132331622348575, "val/token_acc": 0.9784284588764448, "val/seq_acc": 0.5833601027947317, "val_full/loss": 6.743684462961544, "val_full/ppl": 848.6819194758385, "val_full/token_acc": 0.2826074460135278, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.65625, "gen_val_win/bit_acc": 0.95302734375, "gen_val_win/lev": 1.0546875, "gen_val_win/lev_norm": 0.00873492671117449, "gen_val_win/token_acc": 0.8637044433349975, "gen_val_win/bleu": 0.9486185951902287, "gen_val_win/rouge1": 0.9755020740121405, "gen_val_win/rouge2": 0.9564016840093721, "gen_val_win/rougeL": 0.9755020740121405, "gen_val/n": 128.0, "gen_val/seq_acc": 0.234375, "gen_val/bit_acc": 0.8127208048604574, "gen_val/lev": 19.203125, "gen_val/lev_norm": 0.029506916085575666, "gen_val/token_acc": 0.4906334067009049, "gen_val/bleu": 0.9358831378116098, "gen_val/rouge1": 0.9693193715448124, "gen_val/rouge2": 0.9520082403892176, "gen_val/rougeL": 0.9691837745935589, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.2777634326203327, "gen_val_full/lev": 513.265625, "gen_val_full/lev_norm": 0.6369153323199476, "gen_val_full/token_acc": 0.14151257704481093, "gen_val_full/bleu": 0.13503748617078795, "gen_val_full/rouge1": 0.39628284323102375, "gen_val_full/rouge2": 0.3437121409297294, "gen_val_full/rougeL": 0.39008121267803997 }, { "train/epoch_loss": 0.7783261029203039, "epoch": 61, "val/loss": 0.10788866267873048, "val/ppl": 1.113923717199341, "val/token_acc": 0.9782589183313222, "val/seq_acc": 0.5788628332797944, "val_full/loss": 6.769276466903945, "val_full/ppl": 870.6816990094395, "val_full/token_acc": 0.28220506236946485, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7761228485892241, "epoch": 62, "val/loss": 0.1064259923752383, "val/ppl": 1.1122956050437542, "val/token_acc": 0.9784683507694149, "val/seq_acc": 0.5846450369418568, "val_full/loss": 6.780887660115158, "val_full/ppl": 880.8502728362906, "val_full/token_acc": 0.2829140240280519, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7745619712030501, "epoch": 63, "val/loss": 0.10581883733509923, "val/ppl": 1.1116204741363338, "val/token_acc": 0.9787475940202053, "val/seq_acc": 0.5843238034050755, "val_full/loss": 6.857020052717048, "val_full/ppl": 950.5303126778753, "val_full/token_acc": 0.27956082699419416, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7732487248066358, "epoch": 64, "val/loss": 0.1057837293070788, "val/ppl": 1.1115814480186488, "val/token_acc": 0.9788273778061454, "val/seq_acc": 0.5910697076774816, "val_full/loss": 6.8926346727758725, "val_full/ppl": 984.9931361536419, "val_full/token_acc": 0.2792734101055778, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.770642676555605, "epoch": 65, "val/loss": 0.10219489288946389, "val/ppl": 1.1075993139248546, "val/token_acc": 0.9797448913444565, "val/seq_acc": 0.5991005460970126, "val_full/loss": 6.853444182442661, "val_full/ppl": 947.1374094940631, "val_full/token_acc": 0.28178351759949416, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6640625, "gen_val_win/bit_acc": 0.94716796875, "gen_val_win/lev": 0.9609375, "gen_val_win/lev_norm": 0.008012357196860212, "gen_val_win/token_acc": 0.8406991260923845, "gen_val_win/bleu": 0.9488457621544722, "gen_val_win/rouge1": 0.9759582423491148, "gen_val_win/rouge2": 0.9571459817078646, "gen_val_win/rougeL": 0.9759582423491148, "gen_val/n": 128.0, "gen_val/seq_acc": 0.203125, "gen_val/bit_acc": 0.8022057747646686, "gen_val/lev": 20.6875, "gen_val/lev_norm": 0.032646855331402165, "gen_val/token_acc": 0.46047691032330423, "gen_val/bleu": 0.9349888475799477, "gen_val/rouge1": 0.9677987347197661, "gen_val/rouge2": 0.9510535290071539, "gen_val/rougeL": 0.9677987347197661, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0390625, "gen_val_full/bit_acc": 0.2726492281871784, "gen_val_full/lev": 521.8828125, "gen_val_full/lev_norm": 0.6413107590960745, "gen_val_full/token_acc": 0.14461985874532612, "gen_val_full/bleu": 0.1328266569388491, "gen_val_full/rouge1": 0.3926870150503518, "gen_val_full/rouge2": 0.3462960574696285, "gen_val_full/rougeL": 0.3887805928512462 }, { "train/epoch_loss": 0.7692587909258511, "epoch": 66, "val/loss": 0.10406785868657754, "val/ppl": 1.1096757534998816, "val/token_acc": 0.9792761616020584, "val/seq_acc": 0.5952457436556376, "val_full/loss": 6.883722878025063, "val_full/ppl": 976.254077678993, "val_full/token_acc": 0.2807488168004752, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7676588149736647, "epoch": 67, "val/loss": 0.10332856144662418, "val/ppl": 1.1088556764557753, "val/token_acc": 0.9794556751204236, "val/seq_acc": 0.6019916479280437, "val_full/loss": 6.912414643171933, "val_full/ppl": 1004.6702359056585, "val_full/token_acc": 0.2805572055413976, "val_full/seq_acc": 0.057736720554272515, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7655511186307208, "epoch": 68, "val/loss": 0.10274870638594312, "val/ppl": 1.108212887260512, "val/token_acc": 0.9797648372909415, "val/seq_acc": 0.6016704143912625, "val_full/loss": 6.902308047433367, "val_full/ppl": 994.5675777149717, "val_full/token_acc": 0.280844622430014, "val_full/seq_acc": 0.06004618937644342, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7639158102045035, "epoch": 69, "val/loss": 0.10019125087516736, "val/ppl": 1.1053823031940893, "val/token_acc": 0.9800939454079445, "val/seq_acc": 0.6129135881786059, "val_full/loss": 6.834015848646266, "val_full/ppl": 928.9137090421781, "val_full/token_acc": 0.2828948629021441, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7631545408705523, "epoch": 70, "val/loss": 0.10234444427026099, "val/ppl": 1.1077649693183096, "val/token_acc": 0.9800041886487618, "val/seq_acc": 0.6080950851268873, "val_full/loss": 6.938692256150722, "val_full/ppl": 1031.4204987979276, "val_full/token_acc": 0.2798865661346261, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.640625, "gen_val_win/bit_acc": 0.9406687276277254, "gen_val_win/lev": 1.140625, "gen_val_win/lev_norm": 0.00953148366853247, "gen_val_win/token_acc": 0.8224275724275725, "gen_val_win/bleu": 0.9432549842361063, "gen_val_win/rouge1": 0.9735918431566635, "gen_val_win/rouge2": 0.9523991281585007, "gen_val_win/rougeL": 0.9735918431566635, "gen_val/n": 128.0, "gen_val/seq_acc": 0.203125, "gen_val/bit_acc": 0.8020629584352078, "gen_val/lev": 19.6640625, "gen_val/lev_norm": 0.030503058844672484, "gen_val/token_acc": 0.44701320375006137, "gen_val/bleu": 0.9370248231135917, "gen_val/rouge1": 0.969306447534143, "gen_val/rouge2": 0.9519395212698342, "gen_val/rougeL": 0.9692721068748023, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.0546875, "gen_val_full/bit_acc": 0.28398101096549877, "gen_val_full/lev": 540.5703125, "gen_val_full/lev_norm": 0.6432798659738262, "gen_val_full/token_acc": 0.1397216058099657, "gen_val_full/bleu": 0.1492476040614511, "gen_val_full/rouge1": 0.3895888579477387, "gen_val_full/rouge2": 0.34444585875889433, "gen_val_full/rougeL": 0.386186285588132 }, { "train/epoch_loss": 0.7617306089460701, "epoch": 71, "val/loss": 0.09945536487719842, "val/ppl": 1.1045691670592088, "val/token_acc": 0.9803632156854923, "val/seq_acc": 0.6103437198843559, "val_full/loss": 6.923220170917243, "val_full/ppl": 1015.5850922050876, "val_full/token_acc": 0.2790626377205925, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.760312094652742, "epoch": 72, "val/loss": 0.09993753483540561, "val/ppl": 1.1051018855484316, "val/token_acc": 0.9802535129798247, "val/seq_acc": 0.6093800192740122, "val_full/loss": 6.87795799881201, "val_full/ppl": 970.6422820437074, "val_full/token_acc": 0.2807488168004752, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7584700630490024, "epoch": 73, "val/loss": 0.09888564029139722, "val/ppl": 1.1039400460778794, "val/token_acc": 0.9805427292038575, "val/seq_acc": 0.6138772887889495, "val_full/loss": 6.902672183673762, "val_full/ppl": 994.9298017589881, "val_full/token_acc": 0.28069133342275193, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.757090363716544, "epoch": 74, "val/loss": 0.09934481980407801, "val/ppl": 1.1044470691286563, "val/token_acc": 0.98051281028413, "val/seq_acc": 0.6151622229360745, "val_full/loss": 6.869902541723351, "val_full/ppl": 962.854723077238, "val_full/token_acc": 0.2803272720305045, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7558433368913551, "epoch": 75, "val/loss": 0.09762332196203367, "val/ppl": 1.1025474014885182, "val/token_acc": 0.9810613238124682, "val/seq_acc": 0.6254416961130742, "val_full/loss": 6.849906035634904, "val_full/ppl": 943.7922196682308, "val_full/token_acc": 0.2812086838222614, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.671875, "gen_val_win/bit_acc": 0.9422832877782117, "gen_val_win/lev": 0.9921875, "gen_val_win/lev_norm": 0.00819606567965318, "gen_val_win/token_acc": 0.8276292780414689, "gen_val_win/bleu": 0.947414316592444, "gen_val_win/rouge1": 0.9764729090833626, "gen_val_win/rouge2": 0.9580003342554922, "gen_val_win/rougeL": 0.9764729090833626, "gen_val/n": 128.0, "gen_val/seq_acc": 0.25, "gen_val/bit_acc": 0.8232695207620729, "gen_val/lev": 17.078125, "gen_val/lev_norm": 0.027155623468979973, "gen_val/token_acc": 0.5156119388307027, "gen_val/bleu": 0.9442810293234564, "gen_val/rouge1": 0.9720860858655934, "gen_val/rouge2": 0.956508183308305, "gen_val/rougeL": 0.9720517452062527, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.25740549577550126, "gen_val_full/lev": 519.9609375, "gen_val_full/lev_norm": 0.648417922157906, "gen_val_full/token_acc": 0.1487769414984605, "gen_val_full/bleu": 0.12166800634455158, "gen_val_full/rouge1": 0.39725764411295683, "gen_val_full/rouge2": 0.3554125757910429, "gen_val_full/rougeL": 0.3937009832241912 }, { "train/epoch_loss": 0.7550903430900668, "epoch": 76, "val/loss": 0.0970157071997241, "val/ppl": 1.1018776808978912, "val/token_acc": 0.9810014859730132, "val/seq_acc": 0.6257629296498555, "val_full/loss": 6.878550347055893, "val_full/ppl": 971.2174106162977, "val_full/token_acc": 0.28025062752687346, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7538298772754812, "epoch": 77, "val/loss": 0.0969697945593808, "val/ppl": 1.1018270919455708, "val/token_acc": 0.9809715670532856, "val/seq_acc": 0.6199807259877931, "val_full/loss": 6.892423675846443, "val_full/ppl": 984.7853275506636, "val_full/token_acc": 0.2793692157351166, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7530528736530694, "epoch": 78, "val/loss": 0.0965790942477283, "val/ppl": 1.101396691841563, "val/token_acc": 0.9811311346251658, "val/seq_acc": 0.6228718278188243, "val_full/loss": 6.895507868636111, "val_full/ppl": 987.8272839359221, "val_full/token_acc": 0.28007817739370366, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.751869860432689, "epoch": 79, "val/loss": 0.09688268519728528, "val/ppl": 1.1017311166706851, "val/token_acc": 0.9814103778759562, "val/seq_acc": 0.6283327979441053, "val_full/loss": 6.917226780308111, "val_full/ppl": 1009.5164979443022, "val_full/token_acc": 0.2785836095728985, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7509182062826847, "epoch": 80, "val/loss": 0.09529541679414509, "val/ppl": 1.0999837608086722, "val/token_acc": 0.9814901616618963, "val/seq_acc": 0.6305814327015741, "val_full/loss": 6.916158472690605, "val_full/ppl": 1008.4385996455477, "val_full/token_acc": 0.2793308934833011, "val_full/seq_acc": 0.06928406466512702, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6796875, "gen_val_win/bit_acc": 0.9463134765625, "gen_val_win/lev": 1.03125, "gen_val_win/lev_norm": 0.00840752607775145, "gen_val_win/token_acc": 0.8440779610194903, "gen_val_win/bleu": 0.9484006916644606, "gen_val_win/rouge1": 0.9752649798115214, "gen_val_win/rouge2": 0.9575720654460987, "gen_val_win/rougeL": 0.9752649798115214, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2578125, "gen_val/bit_acc": 0.8277827796560341, "gen_val/lev": 17.578125, "gen_val/lev_norm": 0.02862162289325215, "gen_val/token_acc": 0.5, "gen_val/bleu": 0.9436785839609088, "gen_val/rouge1": 0.9712670856476661, "gen_val/rouge2": 0.9558578356795889, "gen_val/rougeL": 0.9712327449883253, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.26057411602926095, "gen_val_full/lev": 514.6640625, "gen_val_full/lev_norm": 0.6444729621561217, "gen_val_full/token_acc": 0.15183111832827656, "gen_val_full/bleu": 0.11805424657010553, "gen_val_full/rouge1": 0.4007690189704065, "gen_val_full/rouge2": 0.3598147292839778, "gen_val_full/rougeL": 0.39727718902750403 }, { "train/epoch_loss": 0.7507990524358583, "epoch": 81, "val/loss": 0.09466045776930845, "val/ppl": 1.0992855378875044, "val/token_acc": 0.9816696751802615, "val/seq_acc": 0.633151300995824, "val_full/loss": 6.895651726776715, "val_full/ppl": 987.9694011543417, "val_full/token_acc": 0.2792350878537623, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7488932466863694, "epoch": 82, "val/loss": 0.09387006661775689, "val/ppl": 1.0984170156067075, "val/token_acc": 0.9819588914042944, "val/seq_acc": 0.6379698040475426, "val_full/loss": 6.8829558342035675, "val_full/ppl": 975.5055351396653, "val_full/token_acc": 0.27948418249056317, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7485113114192896, "epoch": 83, "val/loss": 0.09439190545574618, "val/ppl": 1.098990361849917, "val/token_acc": 0.9815599724745938, "val/seq_acc": 0.6305814327015741, "val_full/loss": 6.91077868719941, "val_full/ppl": 1003.0279833255418, "val_full/token_acc": 0.28067217229684416, "val_full/seq_acc": 0.06235565819861432, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.747908340576581, "epoch": 84, "val/loss": 0.09449007295453929, "val/ppl": 1.0990982522805177, "val/token_acc": 0.9817993238324142, "val/seq_acc": 0.6341150016061677, "val_full/loss": 6.901587783704421, "val_full/ppl": 993.8514846816345, "val_full/token_acc": 0.28053804441548985, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7474257183491143, "epoch": 85, "val/loss": 0.09467065984275874, "val/ppl": 1.099296752936513, "val/token_acc": 0.9817494589662016, "val/seq_acc": 0.6337937680693865, "val_full/loss": 6.911398953802525, "val_full/ppl": 1003.6503210732963, "val_full/token_acc": 0.2797715993791795, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.671875, "gen_val_win/bit_acc": 0.9461205493718675, "gen_val_win/lev": 0.984375, "gen_val_win/lev_norm": 0.008073753998034325, "gen_val_win/token_acc": 0.8438280859570215, "gen_val_win/bleu": 0.9504047711393092, "gen_val_win/rouge1": 0.9756339928696618, "gen_val_win/rouge2": 0.9577301872106105, "gen_val_win/rougeL": 0.9756339928696618, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2578125, "gen_val/bit_acc": 0.8263365802193489, "gen_val/lev": 16.3671875, "gen_val/lev_norm": 0.026195040386058396, "gen_val/token_acc": 0.4932828108852911, "gen_val/bleu": 0.9447168631706051, "gen_val/rouge1": 0.9723404306268314, "gen_val/rouge2": 0.9572176248150943, "gen_val/rougeL": 0.9723404306268314, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.27094894400871206, "gen_val_full/lev": 495.28125, "gen_val_full/lev_norm": 0.6361664286096078, "gen_val_full/token_acc": 0.1528590135312091, "gen_val_full/bleu": 0.11087106212334848, "gen_val_full/rouge1": 0.4078374448731216, "gen_val_full/rouge2": 0.36316155601903416, "gen_val_full/rougeL": 0.4038701031651404 }, { "train/epoch_loss": 0.7465092236561668, "epoch": 86, "val/loss": 0.09443948373297037, "val/ppl": 1.0990426511619276, "val/token_acc": 0.9816896211267465, "val/seq_acc": 0.6309026662383553, "val_full/loss": 6.893623566508268, "val_full/ppl": 985.9676714689517, "val_full/token_acc": 0.28069133342275193, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7463586770388254, "epoch": 87, "val/loss": 0.09405464651487827, "val/ppl": 1.0986197800189859, "val/token_acc": 0.9819090265380818, "val/seq_acc": 0.6353999357532927, "val_full/loss": 6.9136479909559485, "val_full/ppl": 1005.9101081544558, "val_full/token_acc": 0.27967579374964074, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7457975926541925, "epoch": 88, "val/loss": 0.09399636139808677, "val/ppl": 1.0985557487028526, "val/token_acc": 0.9820685941099621, "val/seq_acc": 0.641182139415355, "val_full/loss": 6.927047980058058, "val_full/ppl": 1019.4800078454203, "val_full/token_acc": 0.2784878039433597, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7448093894116599, "epoch": 89, "val/loss": 0.09390369131513948, "val/ppl": 1.0984539501674102, "val/token_acc": 0.9820386751902345, "val/seq_acc": 0.6366848699004176, "val_full/loss": 6.915798004009465, "val_full/ppl": 1008.0751546227325, "val_full/token_acc": 0.2790051543428692, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.744969092996935, "epoch": 90, "val/loss": 0.09396284368007718, "val/ppl": 1.0985189282381222, "val/token_acc": 0.9821184589761746, "val/seq_acc": 0.6370061034371989, "val_full/loss": 6.916054658309222, "val_full/ppl": 1008.3339146501609, "val_full/token_acc": 0.27843032056563644, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.671875, "gen_val_win/bit_acc": 0.9469842458173297, "gen_val_win/lev": 0.921875, "gen_val_win/lev_norm": 0.0076333968113760035, "gen_val_win/token_acc": 0.8429070929070929, "gen_val_win/bleu": 0.9518338842915179, "gen_val_win/rouge1": 0.9765398144347615, "gen_val_win/rouge2": 0.9595123282565482, "gen_val_win/rougeL": 0.9765398144347615, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2734375, "gen_val/bit_acc": 0.8333814035690615, "gen_val/lev": 15.8203125, "gen_val/lev_norm": 0.02494144594680893, "gen_val/token_acc": 0.5343902318940476, "gen_val/bleu": 0.9471141224581772, "gen_val/rouge1": 0.9736141858183894, "gen_val/rouge2": 0.9589001042069789, "gen_val/rougeL": 0.9736141858183894, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.27698332520495134, "gen_val_full/lev": 492.984375, "gen_val_full/lev_norm": 0.6425150535082252, "gen_val_full/token_acc": 0.1512491853139257, "gen_val_full/bleu": 0.1106422770318846, "gen_val_full/rouge1": 0.40028403190140727, "gen_val_full/rouge2": 0.35550888541493625, "gen_val_full/rougeL": 0.396394024880765 }, { "train/epoch_loss": 0.7436553470214406, "epoch": 91, "val/loss": 0.0937822974227149, "val/ppl": 1.0983206126600935, "val/token_acc": 0.9821184589761746, "val/seq_acc": 0.6395759717314488, "val_full/loss": 6.916037892324053, "val_full/ppl": 1008.3170090804219, "val_full/token_acc": 0.2788710264615149, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7441859234895493, "epoch": 92, "val/loss": 0.09341365357948807, "val/ppl": 1.0979157981491203, "val/token_acc": 0.9820885400564471, "val/seq_acc": 0.6360424028268551, "val_full/loss": 6.888767568357316, "val_full/ppl": 981.1914203982003, "val_full/token_acc": 0.2792350878537623, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.743897993962961, "epoch": 93, "val/loss": 0.09332327845518137, "val/ppl": 1.0978165783559524, "val/token_acc": 0.982437594119935, "val/seq_acc": 0.6437520077096048, "val_full/loss": 6.903468118880415, "val_full/ppl": 995.7220166503945, "val_full/token_acc": 0.27841115943972866, "val_full/seq_acc": 0.06697459584295612, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7440287855795197, "epoch": 94, "val/loss": 0.09330615087495212, "val/ppl": 1.097797775575453, "val/token_acc": 0.9823578103339948, "val/seq_acc": 0.6415033729521362, "val_full/loss": 6.894526833959263, "val_full/ppl": 986.8586663170666, "val_full/token_acc": 0.2790434765946847, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7428029515499486, "epoch": 95, "val/loss": 0.0932135848039557, "val/ppl": 1.0976961614517022, "val/token_acc": 0.9822979724945398, "val/seq_acc": 0.6447157083199486, "val_full/loss": 6.889717466204564, "val_full/ppl": 982.1238948237842, "val_full/token_acc": 0.2791392822242235, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.671875, "gen_val_win/bit_acc": 0.9478576544898092, "gen_val_win/lev": 0.90625, "gen_val_win/lev_norm": 0.007477396981291025, "gen_val_win/token_acc": 0.8376217836622533, "gen_val_win/bleu": 0.9518338842915179, "gen_val_win/rouge1": 0.9769652824175764, "gen_val_win/rouge2": 0.9599654522901929, "gen_val_win/rougeL": 0.9769652824175764, "gen_val/n": 128.0, "gen_val/seq_acc": 0.265625, "gen_val/bit_acc": 0.8294901943929365, "gen_val/lev": 15.5625, "gen_val/lev_norm": 0.024555963825244963, "gen_val/token_acc": 0.513325779595054, "gen_val/bleu": 0.9482070055086043, "gen_val/rouge1": 0.9740928407914825, "gen_val/rouge2": 0.9593441806355478, "gen_val/rougeL": 0.9740928407914825, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.270390880608576, "gen_val_full/lev": 496.0078125, "gen_val_full/lev_norm": 0.6427870772959738, "gen_val_full/token_acc": 0.15639998238738936, "gen_val_full/bleu": 0.1128420212627896, "gen_val_full/rouge1": 0.4008848461905546, "gen_val_full/rouge2": 0.3553601486269709, "gen_val_full/rougeL": 0.39781992688738604 }, { "train/epoch_loss": 0.7428625163592008, "epoch": 96, "val/loss": 0.09367556906318544, "val/ppl": 1.098203396958098, "val/token_acc": 0.9821683238423872, "val/seq_acc": 0.6386122711211051, "val_full/loss": 6.919083164076721, "val_full/ppl": 1011.392288540148, "val_full/token_acc": 0.27869857632834505, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7432246505470942, "epoch": 97, "val/loss": 0.09347515743626277, "val/ppl": 1.0979833262817194, "val/token_acc": 0.9823278914142674, "val/seq_acc": 0.6415033729521362, "val_full/loss": 6.918384157221828, "val_full/ppl": 1010.6855654283987, "val_full/token_acc": 0.2786410929506218, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7426908750486493, "epoch": 98, "val/loss": 0.09355550883679935, "val/ppl": 1.0980715543243278, "val/token_acc": 0.9821084860029321, "val/seq_acc": 0.6376485705107613, "val_full/loss": 6.906848261247581, "val_full/ppl": 999.0933934816932, "val_full/token_acc": 0.27935005460920886, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7424187783588495, "epoch": 99, "val/loss": 0.09336124761077168, "val/ppl": 1.097858262315773, "val/token_acc": 0.9822780265480547, "val/seq_acc": 0.6431095406360424, "val_full/loss": 6.9131777948897275, "val_full/ppl": 1005.4372443566903, "val_full/token_acc": 0.2787560597060683, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648 }, { "train/epoch_loss": 0.7429585312667333, "epoch": 100, "val/loss": 0.09325608924582449, "val/ppl": 1.0977428194059644, "val/token_acc": 0.9822879995212973, "val/seq_acc": 0.6437520077096048, "val_full/loss": 6.905993645092836, "val_full/ppl": 998.2399168769175, "val_full/token_acc": 0.2794650213646554, "val_full/seq_acc": 0.06466512702078522, "system/peak_gpu_mem_gb": 1.0646429061889648, "gen_val_win/n": 128.0, "gen_val_win/seq_acc": 0.6796875, "gen_val_win/bit_acc": 0.9479946611107494, "gen_val_win/lev": 0.921875, "gen_val_win/lev_norm": 0.0076366944103690385, "gen_val_win/token_acc": 0.843906093906094, "gen_val_win/bleu": 0.9518607673132358, "gen_val_win/rouge1": 0.9767753379144052, "gen_val_win/rouge2": 0.9597567560068014, "gen_val_win/rougeL": 0.9767753379144052, "gen_val/n": 128.0, "gen_val/seq_acc": 0.2578125, "gen_val/bit_acc": 0.8311946006334864, "gen_val/lev": 15.7890625, "gen_val/lev_norm": 0.024648775069649208, "gen_val/token_acc": 0.5136173356316178, "gen_val/bleu": 0.9468208028852879, "gen_val/rouge1": 0.9731890311710304, "gen_val/rouge2": 0.9581393845511863, "gen_val/rougeL": 0.9731890311710304, "gen_val_full/n": 128.0, "gen_val_full/seq_acc": 0.046875, "gen_val_full/bit_acc": 0.2707485691618972, "gen_val_full/lev": 505.265625, "gen_val_full/lev_norm": 0.6417719213032872, "gen_val_full/token_acc": 0.15068609903690447, "gen_val_full/bleu": 0.11966627133448503, "gen_val_full/rouge1": 0.4006572313888738, "gen_val_full/rouge2": 0.35643417678722816, "gen_val_full/rougeL": 0.39757239123784605 } ] }