diff --git "a/checkpoint-44000/trainer_state.json" "b/checkpoint-44000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-44000/trainer_state.json" @@ -0,0 +1,15631 @@ +{ + "best_metric": 0.22153388779557867, + "best_model_checkpoint": "./checkpoints/w2v-multilingual-v1.3/checkpoint-30000", + "epoch": 3.6249450790861157, + "eval_steps": 2000, + "global_step": 44000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001647627416520211, + "grad_norm": 0.2767059803009033, + "learning_rate": 4.9431537320810675e-08, + "loss": 0.1923, + "step": 20 + }, + { + "epoch": 0.003295254833040422, + "grad_norm": 2.006911039352417, + "learning_rate": 1.977261492832427e-07, + "loss": 0.2947, + "step": 40 + }, + { + "epoch": 0.004942882249560633, + "grad_norm": 0.3714193105697632, + "learning_rate": 3.624979403526116e-07, + "loss": 0.2955, + "step": 60 + }, + { + "epoch": 0.006590509666080844, + "grad_norm": 0.3106062412261963, + "learning_rate": 5.272697314219806e-07, + "loss": 0.2754, + "step": 80 + }, + { + "epoch": 0.008238137082601054, + "grad_norm": 3.2078254222869873, + "learning_rate": 6.838029329378811e-07, + "loss": 0.3177, + "step": 100 + }, + { + "epoch": 0.009885764499121265, + "grad_norm": 0.38503074645996094, + "learning_rate": 8.485747240072501e-07, + "loss": 0.1905, + "step": 120 + }, + { + "epoch": 0.011533391915641476, + "grad_norm": 0.15724419057369232, + "learning_rate": 1.013346515076619e-06, + "loss": 0.2942, + "step": 140 + }, + { + "epoch": 0.013181019332161687, + "grad_norm": 0.5840986967086792, + "learning_rate": 1.1781183061459877e-06, + "loss": 0.2907, + "step": 160 + }, + { + "epoch": 0.014828646748681899, + "grad_norm": 0.23122897744178772, + "learning_rate": 1.3346515076618883e-06, + "loss": 0.2811, + "step": 180 + }, + { + "epoch": 0.016476274165202108, + "grad_norm": 2.7248637676239014, + "learning_rate": 1.4994232987312573e-06, + "loss": 0.3168, + "step": 200 + }, + { + "epoch": 0.01812390158172232, + "grad_norm": 1.0846350193023682, + "learning_rate": 1.6641950898006263e-06, + "loss": 0.1979, + "step": 220 + }, + { + "epoch": 0.01977152899824253, + "grad_norm": 1.7905633449554443, + "learning_rate": 1.828966880869995e-06, + "loss": 0.2656, + "step": 240 + }, + { + "epoch": 0.021419156414762743, + "grad_norm": 0.31305554509162903, + "learning_rate": 1.993738671939364e-06, + "loss": 0.272, + "step": 260 + }, + { + "epoch": 0.023066783831282953, + "grad_norm": 0.31258106231689453, + "learning_rate": 2.158510463008733e-06, + "loss": 0.2693, + "step": 280 + }, + { + "epoch": 0.024714411247803162, + "grad_norm": 1.783349871635437, + "learning_rate": 2.323282254078102e-06, + "loss": 0.3116, + "step": 300 + }, + { + "epoch": 0.026362038664323375, + "grad_norm": 0.6936488747596741, + "learning_rate": 2.488054045147471e-06, + "loss": 0.1931, + "step": 320 + }, + { + "epoch": 0.028009666080843584, + "grad_norm": 0.6185577511787415, + "learning_rate": 2.65282583621684e-06, + "loss": 0.258, + "step": 340 + }, + { + "epoch": 0.029657293497363797, + "grad_norm": 0.3592207729816437, + "learning_rate": 2.8175976272862085e-06, + "loss": 0.2401, + "step": 360 + }, + { + "epoch": 0.03130492091388401, + "grad_norm": 1.2324920892715454, + "learning_rate": 2.982369418355578e-06, + "loss": 0.2371, + "step": 380 + }, + { + "epoch": 0.032952548330404216, + "grad_norm": 1.5560214519500732, + "learning_rate": 3.147141209424947e-06, + "loss": 0.2414, + "step": 400 + }, + { + "epoch": 0.03460017574692443, + "grad_norm": 0.8133947253227234, + "learning_rate": 3.3119130004943154e-06, + "loss": 0.1744, + "step": 420 + }, + { + "epoch": 0.03624780316344464, + "grad_norm": 0.8599107265472412, + "learning_rate": 3.4766847915636844e-06, + "loss": 0.2536, + "step": 440 + }, + { + "epoch": 0.03789543057996485, + "grad_norm": 0.3699595034122467, + "learning_rate": 3.641456582633054e-06, + "loss": 0.2552, + "step": 460 + }, + { + "epoch": 0.03954305799648506, + "grad_norm": 1.2955116033554077, + "learning_rate": 3.797989784148954e-06, + "loss": 0.2532, + "step": 480 + }, + { + "epoch": 0.041190685413005274, + "grad_norm": 11.518170356750488, + "learning_rate": 3.962761575218322e-06, + "loss": 0.2507, + "step": 500 + }, + { + "epoch": 0.042838312829525486, + "grad_norm": 0.9779248833656311, + "learning_rate": 4.127533366287692e-06, + "loss": 0.1733, + "step": 520 + }, + { + "epoch": 0.04448594024604569, + "grad_norm": 0.5386682152748108, + "learning_rate": 4.29230515735706e-06, + "loss": 0.2415, + "step": 540 + }, + { + "epoch": 0.046133567662565905, + "grad_norm": 2.7366793155670166, + "learning_rate": 4.457076948426429e-06, + "loss": 0.2409, + "step": 560 + }, + { + "epoch": 0.04778119507908612, + "grad_norm": 1.9912066459655762, + "learning_rate": 4.621848739495799e-06, + "loss": 0.2134, + "step": 580 + }, + { + "epoch": 0.049428822495606324, + "grad_norm": 4.058573246002197, + "learning_rate": 4.786620530565167e-06, + "loss": 0.2456, + "step": 600 + }, + { + "epoch": 0.05107644991212654, + "grad_norm": 0.37456750869750977, + "learning_rate": 4.951392321634536e-06, + "loss": 0.1589, + "step": 620 + }, + { + "epoch": 0.05272407732864675, + "grad_norm": 0.7950440645217896, + "learning_rate": 5.116164112703905e-06, + "loss": 0.2485, + "step": 640 + }, + { + "epoch": 0.05437170474516696, + "grad_norm": 1.884665846824646, + "learning_rate": 5.280935903773274e-06, + "loss": 0.2197, + "step": 660 + }, + { + "epoch": 0.05601933216168717, + "grad_norm": 0.32170844078063965, + "learning_rate": 5.445707694842643e-06, + "loss": 0.2111, + "step": 680 + }, + { + "epoch": 0.05766695957820738, + "grad_norm": 1.8230172395706177, + "learning_rate": 5.610479485912012e-06, + "loss": 0.2379, + "step": 700 + }, + { + "epoch": 0.059314586994727594, + "grad_norm": 1.2472524642944336, + "learning_rate": 5.775251276981381e-06, + "loss": 0.1684, + "step": 720 + }, + { + "epoch": 0.0609622144112478, + "grad_norm": 0.29022061824798584, + "learning_rate": 5.94002306805075e-06, + "loss": 0.2201, + "step": 740 + }, + { + "epoch": 0.06260984182776802, + "grad_norm": 0.49721184372901917, + "learning_rate": 6.104794859120118e-06, + "loss": 0.2297, + "step": 760 + }, + { + "epoch": 0.06425746924428823, + "grad_norm": 0.6133716106414795, + "learning_rate": 6.269566650189487e-06, + "loss": 0.3173, + "step": 780 + }, + { + "epoch": 0.06590509666080843, + "grad_norm": 0.9667792320251465, + "learning_rate": 6.434338441258857e-06, + "loss": 0.2349, + "step": 800 + }, + { + "epoch": 0.06755272407732865, + "grad_norm": 0.3177216053009033, + "learning_rate": 6.599110232328226e-06, + "loss": 0.1635, + "step": 820 + }, + { + "epoch": 0.06920035149384886, + "grad_norm": 0.8457621335983276, + "learning_rate": 6.763882023397594e-06, + "loss": 0.2777, + "step": 840 + }, + { + "epoch": 0.07084797891036906, + "grad_norm": 0.3330087661743164, + "learning_rate": 6.928653814466963e-06, + "loss": 0.2154, + "step": 860 + }, + { + "epoch": 0.07249560632688928, + "grad_norm": 0.5845814943313599, + "learning_rate": 7.093425605536333e-06, + "loss": 0.2164, + "step": 880 + }, + { + "epoch": 0.07414323374340949, + "grad_norm": 2.325303554534912, + "learning_rate": 7.258197396605701e-06, + "loss": 0.2068, + "step": 900 + }, + { + "epoch": 0.0757908611599297, + "grad_norm": 0.21893823146820068, + "learning_rate": 7.414730598121602e-06, + "loss": 0.1621, + "step": 920 + }, + { + "epoch": 0.07743848857644992, + "grad_norm": 0.5854327082633972, + "learning_rate": 7.5795023891909705e-06, + "loss": 0.2281, + "step": 940 + }, + { + "epoch": 0.07908611599297012, + "grad_norm": 0.2406030148267746, + "learning_rate": 7.74427418026034e-06, + "loss": 0.2342, + "step": 960 + }, + { + "epoch": 0.08073374340949033, + "grad_norm": 1.3764126300811768, + "learning_rate": 7.90904597132971e-06, + "loss": 0.2018, + "step": 980 + }, + { + "epoch": 0.08238137082601055, + "grad_norm": 0.9587862491607666, + "learning_rate": 8.073817762399077e-06, + "loss": 0.2489, + "step": 1000 + }, + { + "epoch": 0.08402899824253075, + "grad_norm": 0.9726558923721313, + "learning_rate": 8.238589553468447e-06, + "loss": 0.154, + "step": 1020 + }, + { + "epoch": 0.08567662565905097, + "grad_norm": 1.9828400611877441, + "learning_rate": 8.403361344537817e-06, + "loss": 0.2336, + "step": 1040 + }, + { + "epoch": 0.08732425307557118, + "grad_norm": 0.14761961996555328, + "learning_rate": 8.568133135607183e-06, + "loss": 0.1949, + "step": 1060 + }, + { + "epoch": 0.08897188049209138, + "grad_norm": 0.39077144861221313, + "learning_rate": 8.732904926676553e-06, + "loss": 0.2067, + "step": 1080 + }, + { + "epoch": 0.0906195079086116, + "grad_norm": 2.3257837295532227, + "learning_rate": 8.897676717745921e-06, + "loss": 0.2184, + "step": 1100 + }, + { + "epoch": 0.09226713532513181, + "grad_norm": 1.0966060161590576, + "learning_rate": 9.062448508815291e-06, + "loss": 0.1655, + "step": 1120 + }, + { + "epoch": 0.09391476274165202, + "grad_norm": 0.5648412704467773, + "learning_rate": 9.227220299884661e-06, + "loss": 0.2097, + "step": 1140 + }, + { + "epoch": 0.09556239015817224, + "grad_norm": 0.49816444516181946, + "learning_rate": 9.39199209095403e-06, + "loss": 0.2052, + "step": 1160 + }, + { + "epoch": 0.09721001757469244, + "grad_norm": 1.1164054870605469, + "learning_rate": 9.556763882023397e-06, + "loss": 0.2298, + "step": 1180 + }, + { + "epoch": 0.09885764499121265, + "grad_norm": 0.9453270435333252, + "learning_rate": 9.721535673092767e-06, + "loss": 0.2157, + "step": 1200 + }, + { + "epoch": 0.10050527240773287, + "grad_norm": 0.4588276743888855, + "learning_rate": 9.886307464162135e-06, + "loss": 0.1433, + "step": 1220 + }, + { + "epoch": 0.10215289982425307, + "grad_norm": 0.47535696625709534, + "learning_rate": 1.0051079255231505e-05, + "loss": 0.2115, + "step": 1240 + }, + { + "epoch": 0.10380052724077328, + "grad_norm": 0.585959792137146, + "learning_rate": 1.0215851046300875e-05, + "loss": 0.2245, + "step": 1260 + }, + { + "epoch": 0.1054481546572935, + "grad_norm": 0.24861204624176025, + "learning_rate": 1.0380622837370241e-05, + "loss": 0.1981, + "step": 1280 + }, + { + "epoch": 0.1070957820738137, + "grad_norm": 31.602128982543945, + "learning_rate": 1.0545394628439611e-05, + "loss": 0.2265, + "step": 1300 + }, + { + "epoch": 0.10874340949033393, + "grad_norm": 0.6234269142150879, + "learning_rate": 1.0710166419508981e-05, + "loss": 0.154, + "step": 1320 + }, + { + "epoch": 0.11039103690685413, + "grad_norm": 1.2423540353775024, + "learning_rate": 1.087493821057835e-05, + "loss": 0.2062, + "step": 1340 + }, + { + "epoch": 0.11203866432337434, + "grad_norm": 0.2090279757976532, + "learning_rate": 1.1039710001647719e-05, + "loss": 0.2214, + "step": 1360 + }, + { + "epoch": 0.11368629173989456, + "grad_norm": 0.6145613193511963, + "learning_rate": 1.1204481792717087e-05, + "loss": 0.1942, + "step": 1380 + }, + { + "epoch": 0.11533391915641476, + "grad_norm": 0.9004138708114624, + "learning_rate": 1.1369253583786455e-05, + "loss": 0.2271, + "step": 1400 + }, + { + "epoch": 0.11698154657293497, + "grad_norm": 1.609165906906128, + "learning_rate": 1.1534025374855825e-05, + "loss": 0.1606, + "step": 1420 + }, + { + "epoch": 0.11862917398945519, + "grad_norm": 0.8725568652153015, + "learning_rate": 1.1698797165925195e-05, + "loss": 0.204, + "step": 1440 + }, + { + "epoch": 0.1202768014059754, + "grad_norm": 1.8169455528259277, + "learning_rate": 1.1863568956994563e-05, + "loss": 0.207, + "step": 1460 + }, + { + "epoch": 0.1219244288224956, + "grad_norm": 0.37334388494491577, + "learning_rate": 1.2028340748063933e-05, + "loss": 0.2033, + "step": 1480 + }, + { + "epoch": 0.12357205623901582, + "grad_norm": 15.727474212646484, + "learning_rate": 1.2193112539133301e-05, + "loss": 0.3298, + "step": 1500 + }, + { + "epoch": 0.12521968365553604, + "grad_norm": 4.70903205871582, + "learning_rate": 1.2357884330202669e-05, + "loss": 0.159, + "step": 1520 + }, + { + "epoch": 0.12686731107205623, + "grad_norm": 0.4843326807022095, + "learning_rate": 1.2522656121272039e-05, + "loss": 0.2141, + "step": 1540 + }, + { + "epoch": 0.12851493848857645, + "grad_norm": 0.572084367275238, + "learning_rate": 1.2687427912341407e-05, + "loss": 0.2182, + "step": 1560 + }, + { + "epoch": 0.13016256590509667, + "grad_norm": 0.31078797578811646, + "learning_rate": 1.2852199703410777e-05, + "loss": 0.1994, + "step": 1580 + }, + { + "epoch": 0.13181019332161686, + "grad_norm": 9.014205932617188, + "learning_rate": 1.3016971494480145e-05, + "loss": 0.2089, + "step": 1600 + }, + { + "epoch": 0.13345782073813708, + "grad_norm": 1.0150245428085327, + "learning_rate": 1.3181743285549513e-05, + "loss": 0.1521, + "step": 1620 + }, + { + "epoch": 0.1351054481546573, + "grad_norm": 0.25271451473236084, + "learning_rate": 1.3346515076618885e-05, + "loss": 0.1996, + "step": 1640 + }, + { + "epoch": 0.1367530755711775, + "grad_norm": 0.47118502855300903, + "learning_rate": 1.3511286867688253e-05, + "loss": 0.2059, + "step": 1660 + }, + { + "epoch": 0.13840070298769772, + "grad_norm": 0.5134350657463074, + "learning_rate": 1.3676058658757621e-05, + "loss": 0.1935, + "step": 1680 + }, + { + "epoch": 0.14004833040421794, + "grad_norm": 1.0354816913604736, + "learning_rate": 1.384083044982699e-05, + "loss": 0.2103, + "step": 1700 + }, + { + "epoch": 0.14169595782073813, + "grad_norm": 0.5588876605033875, + "learning_rate": 1.4005602240896359e-05, + "loss": 0.1598, + "step": 1720 + }, + { + "epoch": 0.14334358523725835, + "grad_norm": 0.7309175133705139, + "learning_rate": 1.4170374031965727e-05, + "loss": 0.2204, + "step": 1740 + }, + { + "epoch": 0.14499121265377857, + "grad_norm": 0.6155902743339539, + "learning_rate": 1.4335145823035099e-05, + "loss": 0.2133, + "step": 1760 + }, + { + "epoch": 0.14663884007029876, + "grad_norm": 0.7660940885543823, + "learning_rate": 1.4499917614104467e-05, + "loss": 0.2065, + "step": 1780 + }, + { + "epoch": 0.14828646748681898, + "grad_norm": 1.1954026222229004, + "learning_rate": 1.4664689405173835e-05, + "loss": 0.2147, + "step": 1800 + }, + { + "epoch": 0.1499340949033392, + "grad_norm": 0.4249323606491089, + "learning_rate": 1.4829461196243205e-05, + "loss": 0.1553, + "step": 1820 + }, + { + "epoch": 0.1515817223198594, + "grad_norm": 2.9014129638671875, + "learning_rate": 1.4994232987312573e-05, + "loss": 0.2208, + "step": 1840 + }, + { + "epoch": 0.1532293497363796, + "grad_norm": 1.6474498510360718, + "learning_rate": 1.5159004778381941e-05, + "loss": 0.209, + "step": 1860 + }, + { + "epoch": 0.15487697715289983, + "grad_norm": 0.1585623174905777, + "learning_rate": 1.532377656945131e-05, + "loss": 0.1873, + "step": 1880 + }, + { + "epoch": 0.15652460456942002, + "grad_norm": 1.171941876411438, + "learning_rate": 1.548854836052068e-05, + "loss": 0.2389, + "step": 1900 + }, + { + "epoch": 0.15817223198594024, + "grad_norm": 0.48890382051467896, + "learning_rate": 1.5653320151590047e-05, + "loss": 0.1679, + "step": 1920 + }, + { + "epoch": 0.15981985940246046, + "grad_norm": 0.5568016767501831, + "learning_rate": 1.581809194265942e-05, + "loss": 0.1968, + "step": 1940 + }, + { + "epoch": 0.16146748681898065, + "grad_norm": 0.9775394797325134, + "learning_rate": 1.5982863733728787e-05, + "loss": 0.2208, + "step": 1960 + }, + { + "epoch": 0.16311511423550087, + "grad_norm": 0.60302734375, + "learning_rate": 1.6147635524798155e-05, + "loss": 0.1929, + "step": 1980 + }, + { + "epoch": 0.1647627416520211, + "grad_norm": 1.7513552904129028, + "learning_rate": 1.6312407315867526e-05, + "loss": 0.2055, + "step": 2000 + }, + { + "epoch": 0.1647627416520211, + "eval_loss": 0.5699400305747986, + "eval_runtime": 686.8117, + "eval_samples_per_second": 31.725, + "eval_steps_per_second": 7.932, + "eval_wer": 0.23469093535410654, + "step": 2000 + }, + { + "epoch": 0.16641036906854131, + "grad_norm": 1.049734354019165, + "learning_rate": 1.6477179106936894e-05, + "loss": 0.1608, + "step": 2020 + }, + { + "epoch": 0.1680579964850615, + "grad_norm": 1.7113618850708008, + "learning_rate": 1.6641950898006263e-05, + "loss": 0.2008, + "step": 2040 + }, + { + "epoch": 0.16970562390158173, + "grad_norm": 0.3202134370803833, + "learning_rate": 1.6806722689075634e-05, + "loss": 0.1973, + "step": 2060 + }, + { + "epoch": 0.17135325131810195, + "grad_norm": 0.3612610697746277, + "learning_rate": 1.6971494480145e-05, + "loss": 0.1732, + "step": 2080 + }, + { + "epoch": 0.17300087873462214, + "grad_norm": 1.8115849494934082, + "learning_rate": 1.7136266271214367e-05, + "loss": 0.2138, + "step": 2100 + }, + { + "epoch": 0.17464850615114236, + "grad_norm": 0.7046949863433838, + "learning_rate": 1.7301038062283735e-05, + "loss": 0.1571, + "step": 2120 + }, + { + "epoch": 0.17629613356766258, + "grad_norm": 0.5983096957206726, + "learning_rate": 1.7465809853353107e-05, + "loss": 0.2092, + "step": 2140 + }, + { + "epoch": 0.17794376098418277, + "grad_norm": 0.17064958810806274, + "learning_rate": 1.7630581644422475e-05, + "loss": 0.2083, + "step": 2160 + }, + { + "epoch": 0.179591388400703, + "grad_norm": 1.422013759613037, + "learning_rate": 1.7795353435491843e-05, + "loss": 0.2087, + "step": 2180 + }, + { + "epoch": 0.1812390158172232, + "grad_norm": 0.982097864151001, + "learning_rate": 1.7960125226561214e-05, + "loss": 0.2161, + "step": 2200 + }, + { + "epoch": 0.1828866432337434, + "grad_norm": 0.2690947949886322, + "learning_rate": 1.8124897017630583e-05, + "loss": 0.1693, + "step": 2220 + }, + { + "epoch": 0.18453427065026362, + "grad_norm": 0.24867244064807892, + "learning_rate": 1.828966880869995e-05, + "loss": 0.2058, + "step": 2240 + }, + { + "epoch": 0.18618189806678384, + "grad_norm": 0.9435555934906006, + "learning_rate": 1.8454440599769322e-05, + "loss": 0.1991, + "step": 2260 + }, + { + "epoch": 0.18782952548330403, + "grad_norm": 0.23964335024356842, + "learning_rate": 1.861921239083869e-05, + "loss": 0.1932, + "step": 2280 + }, + { + "epoch": 0.18947715289982425, + "grad_norm": 2.1007418632507324, + "learning_rate": 1.878398418190806e-05, + "loss": 0.2075, + "step": 2300 + }, + { + "epoch": 0.19112478031634447, + "grad_norm": 0.31368857622146606, + "learning_rate": 1.8948755972977427e-05, + "loss": 0.1557, + "step": 2320 + }, + { + "epoch": 0.19277240773286466, + "grad_norm": 0.4029647409915924, + "learning_rate": 1.9113527764046795e-05, + "loss": 0.1952, + "step": 2340 + }, + { + "epoch": 0.19442003514938488, + "grad_norm": 0.38545289635658264, + "learning_rate": 1.9278299555116163e-05, + "loss": 0.1998, + "step": 2360 + }, + { + "epoch": 0.1960676625659051, + "grad_norm": 0.44573166966438293, + "learning_rate": 1.9443071346185534e-05, + "loss": 0.2078, + "step": 2380 + }, + { + "epoch": 0.1977152899824253, + "grad_norm": 9.132265090942383, + "learning_rate": 1.9607843137254903e-05, + "loss": 0.2175, + "step": 2400 + }, + { + "epoch": 0.19936291739894552, + "grad_norm": 0.42929643392562866, + "learning_rate": 1.977261492832427e-05, + "loss": 0.1762, + "step": 2420 + }, + { + "epoch": 0.20101054481546574, + "grad_norm": 0.6267173886299133, + "learning_rate": 1.9937386719393642e-05, + "loss": 0.2101, + "step": 2440 + }, + { + "epoch": 0.20265817223198593, + "grad_norm": 9.70997142791748, + "learning_rate": 2.010215851046301e-05, + "loss": 0.2101, + "step": 2460 + }, + { + "epoch": 0.20430579964850615, + "grad_norm": 0.47748956084251404, + "learning_rate": 2.026693030153238e-05, + "loss": 0.2039, + "step": 2480 + }, + { + "epoch": 0.20595342706502637, + "grad_norm": 1.3222582340240479, + "learning_rate": 2.043170209260175e-05, + "loss": 0.213, + "step": 2500 + }, + { + "epoch": 0.20760105448154656, + "grad_norm": 0.4152863919734955, + "learning_rate": 2.0596473883671115e-05, + "loss": 0.148, + "step": 2520 + }, + { + "epoch": 0.20924868189806678, + "grad_norm": 0.7384160757064819, + "learning_rate": 2.0761245674740483e-05, + "loss": 0.2138, + "step": 2540 + }, + { + "epoch": 0.210896309314587, + "grad_norm": 0.27651092410087585, + "learning_rate": 2.0926017465809854e-05, + "loss": 0.2046, + "step": 2560 + }, + { + "epoch": 0.21254393673110722, + "grad_norm": 0.226897194981575, + "learning_rate": 2.1090789256879222e-05, + "loss": 0.1904, + "step": 2580 + }, + { + "epoch": 0.2141915641476274, + "grad_norm": 1.2391464710235596, + "learning_rate": 2.125556104794859e-05, + "loss": 0.204, + "step": 2600 + }, + { + "epoch": 0.21583919156414763, + "grad_norm": 1.6048617362976074, + "learning_rate": 2.1420332839017962e-05, + "loss": 0.1548, + "step": 2620 + }, + { + "epoch": 0.21748681898066785, + "grad_norm": 0.28409889340400696, + "learning_rate": 2.158510463008733e-05, + "loss": 0.201, + "step": 2640 + }, + { + "epoch": 0.21913444639718804, + "grad_norm": 0.354885995388031, + "learning_rate": 2.17498764211567e-05, + "loss": 0.2083, + "step": 2660 + }, + { + "epoch": 0.22078207381370826, + "grad_norm": 0.2778099775314331, + "learning_rate": 2.191464821222607e-05, + "loss": 0.1891, + "step": 2680 + }, + { + "epoch": 0.22242970123022848, + "grad_norm": 1.007686734199524, + "learning_rate": 2.2079420003295438e-05, + "loss": 0.2152, + "step": 2700 + }, + { + "epoch": 0.22407732864674867, + "grad_norm": 0.9725649952888489, + "learning_rate": 2.2244191794364806e-05, + "loss": 0.1581, + "step": 2720 + }, + { + "epoch": 0.2257249560632689, + "grad_norm": 0.2451123297214508, + "learning_rate": 2.2408963585434174e-05, + "loss": 0.2019, + "step": 2740 + }, + { + "epoch": 0.22737258347978911, + "grad_norm": 0.3667006194591522, + "learning_rate": 2.2573735376503542e-05, + "loss": 0.2083, + "step": 2760 + }, + { + "epoch": 0.2290202108963093, + "grad_norm": 3.1283884048461914, + "learning_rate": 2.273850716757291e-05, + "loss": 0.194, + "step": 2780 + }, + { + "epoch": 0.23066783831282953, + "grad_norm": 0.7148507237434387, + "learning_rate": 2.2903278958642282e-05, + "loss": 0.1931, + "step": 2800 + }, + { + "epoch": 0.23231546572934975, + "grad_norm": 0.5805519223213196, + "learning_rate": 2.306805074971165e-05, + "loss": 0.1552, + "step": 2820 + }, + { + "epoch": 0.23396309314586994, + "grad_norm": 0.8168196082115173, + "learning_rate": 2.323282254078102e-05, + "loss": 0.2107, + "step": 2840 + }, + { + "epoch": 0.23561072056239016, + "grad_norm": 0.17171867191791534, + "learning_rate": 2.339759433185039e-05, + "loss": 0.2195, + "step": 2860 + }, + { + "epoch": 0.23725834797891038, + "grad_norm": 0.6692082285881042, + "learning_rate": 2.3562366122919758e-05, + "loss": 0.1975, + "step": 2880 + }, + { + "epoch": 0.23890597539543057, + "grad_norm": 1.5185160636901855, + "learning_rate": 2.3727137913989126e-05, + "loss": 0.2086, + "step": 2900 + }, + { + "epoch": 0.2405536028119508, + "grad_norm": 0.8978987336158752, + "learning_rate": 2.3891909705058498e-05, + "loss": 0.1537, + "step": 2920 + }, + { + "epoch": 0.242201230228471, + "grad_norm": 1.1462221145629883, + "learning_rate": 2.4056681496127866e-05, + "loss": 0.215, + "step": 2940 + }, + { + "epoch": 0.2438488576449912, + "grad_norm": 0.2455727905035019, + "learning_rate": 2.422145328719723e-05, + "loss": 0.2137, + "step": 2960 + }, + { + "epoch": 0.24549648506151142, + "grad_norm": 0.21464231610298157, + "learning_rate": 2.4386225078266602e-05, + "loss": 0.203, + "step": 2980 + }, + { + "epoch": 0.24714411247803164, + "grad_norm": 0.8478316068649292, + "learning_rate": 2.455099686933597e-05, + "loss": 0.2085, + "step": 3000 + }, + { + "epoch": 0.24879173989455183, + "grad_norm": 0.37225690484046936, + "learning_rate": 2.4715768660405338e-05, + "loss": 0.1645, + "step": 3020 + }, + { + "epoch": 0.2504393673110721, + "grad_norm": 1.3999593257904053, + "learning_rate": 2.488054045147471e-05, + "loss": 0.2129, + "step": 3040 + }, + { + "epoch": 0.2520869947275923, + "grad_norm": 2.0909199714660645, + "learning_rate": 2.5045312242544078e-05, + "loss": 0.1926, + "step": 3060 + }, + { + "epoch": 0.25373462214411246, + "grad_norm": 0.19655053317546844, + "learning_rate": 2.5210084033613446e-05, + "loss": 0.2007, + "step": 3080 + }, + { + "epoch": 0.2553822495606327, + "grad_norm": 1.2680870294570923, + "learning_rate": 2.5374855824682814e-05, + "loss": 0.2107, + "step": 3100 + }, + { + "epoch": 0.2570298769771529, + "grad_norm": 0.19821316003799438, + "learning_rate": 2.5539627615752182e-05, + "loss": 0.1511, + "step": 3120 + }, + { + "epoch": 0.2586775043936731, + "grad_norm": 0.22427937388420105, + "learning_rate": 2.5704399406821554e-05, + "loss": 0.2139, + "step": 3140 + }, + { + "epoch": 0.26032513181019334, + "grad_norm": 0.4531656503677368, + "learning_rate": 2.5869171197890922e-05, + "loss": 0.1995, + "step": 3160 + }, + { + "epoch": 0.26197275922671354, + "grad_norm": 0.3967747986316681, + "learning_rate": 2.603394298896029e-05, + "loss": 0.1981, + "step": 3180 + }, + { + "epoch": 0.26362038664323373, + "grad_norm": 1.0957462787628174, + "learning_rate": 2.6198714780029658e-05, + "loss": 0.2092, + "step": 3200 + }, + { + "epoch": 0.265268014059754, + "grad_norm": 0.5567193627357483, + "learning_rate": 2.6363486571099026e-05, + "loss": 0.1642, + "step": 3220 + }, + { + "epoch": 0.26691564147627417, + "grad_norm": 0.3523741066455841, + "learning_rate": 2.6528258362168395e-05, + "loss": 0.2059, + "step": 3240 + }, + { + "epoch": 0.26856326889279436, + "grad_norm": 0.40257710218429565, + "learning_rate": 2.669303015323777e-05, + "loss": 0.195, + "step": 3260 + }, + { + "epoch": 0.2702108963093146, + "grad_norm": 0.3187640309333801, + "learning_rate": 2.6849563354753665e-05, + "loss": 0.1827, + "step": 3280 + }, + { + "epoch": 0.2718585237258348, + "grad_norm": 0.7375414967536926, + "learning_rate": 2.701433514582304e-05, + "loss": 0.2247, + "step": 3300 + }, + { + "epoch": 0.273506151142355, + "grad_norm": 0.45597076416015625, + "learning_rate": 2.7179106936892408e-05, + "loss": 0.1654, + "step": 3320 + }, + { + "epoch": 0.27515377855887524, + "grad_norm": 0.21507132053375244, + "learning_rate": 2.7343878727961776e-05, + "loss": 0.1858, + "step": 3340 + }, + { + "epoch": 0.27680140597539543, + "grad_norm": 0.7203060388565063, + "learning_rate": 2.7508650519031144e-05, + "loss": 0.1908, + "step": 3360 + }, + { + "epoch": 0.2784490333919156, + "grad_norm": 0.8007901906967163, + "learning_rate": 2.7673422310100512e-05, + "loss": 0.1793, + "step": 3380 + }, + { + "epoch": 0.28009666080843587, + "grad_norm": 3.210064649581909, + "learning_rate": 2.783819410116988e-05, + "loss": 0.206, + "step": 3400 + }, + { + "epoch": 0.28174428822495606, + "grad_norm": 0.8950255513191223, + "learning_rate": 2.8002965892239252e-05, + "loss": 0.1531, + "step": 3420 + }, + { + "epoch": 0.28339191564147626, + "grad_norm": 0.4942973256111145, + "learning_rate": 2.816773768330862e-05, + "loss": 0.1916, + "step": 3440 + }, + { + "epoch": 0.2850395430579965, + "grad_norm": 0.31426137685775757, + "learning_rate": 2.8332509474377988e-05, + "loss": 0.2013, + "step": 3460 + }, + { + "epoch": 0.2866871704745167, + "grad_norm": 0.47154414653778076, + "learning_rate": 2.8497281265447356e-05, + "loss": 0.1831, + "step": 3480 + }, + { + "epoch": 0.2883347978910369, + "grad_norm": 0.8456603288650513, + "learning_rate": 2.8662053056516724e-05, + "loss": 0.2043, + "step": 3500 + }, + { + "epoch": 0.28998242530755713, + "grad_norm": 0.29031482338905334, + "learning_rate": 2.8826824847586092e-05, + "loss": 0.1607, + "step": 3520 + }, + { + "epoch": 0.2916300527240773, + "grad_norm": 0.3170378804206848, + "learning_rate": 2.8991596638655467e-05, + "loss": 0.1916, + "step": 3540 + }, + { + "epoch": 0.2932776801405975, + "grad_norm": 0.3800877332687378, + "learning_rate": 2.9156368429724836e-05, + "loss": 0.2051, + "step": 3560 + }, + { + "epoch": 0.29492530755711777, + "grad_norm": 0.5847609639167786, + "learning_rate": 2.9321140220794204e-05, + "loss": 0.196, + "step": 3580 + }, + { + "epoch": 0.29657293497363796, + "grad_norm": 1.0933667421340942, + "learning_rate": 2.9485912011863572e-05, + "loss": 0.2154, + "step": 3600 + }, + { + "epoch": 0.29822056239015815, + "grad_norm": 4.349573135375977, + "learning_rate": 2.9650683802932937e-05, + "loss": 0.1606, + "step": 3620 + }, + { + "epoch": 0.2998681898066784, + "grad_norm": 0.4264489710330963, + "learning_rate": 2.9815455594002305e-05, + "loss": 0.2117, + "step": 3640 + }, + { + "epoch": 0.3015158172231986, + "grad_norm": 0.47935691475868225, + "learning_rate": 2.998022738507168e-05, + "loss": 0.1901, + "step": 3660 + }, + { + "epoch": 0.3031634446397188, + "grad_norm": 0.7258153557777405, + "learning_rate": 3.0144999176141048e-05, + "loss": 0.189, + "step": 3680 + }, + { + "epoch": 0.30481107205623903, + "grad_norm": 2.0093941688537598, + "learning_rate": 3.0309770967210416e-05, + "loss": 0.2104, + "step": 3700 + }, + { + "epoch": 0.3064586994727592, + "grad_norm": 1.1718699932098389, + "learning_rate": 3.0474542758279784e-05, + "loss": 0.1577, + "step": 3720 + }, + { + "epoch": 0.3081063268892794, + "grad_norm": 0.19388867914676666, + "learning_rate": 3.063931454934915e-05, + "loss": 0.2011, + "step": 3740 + }, + { + "epoch": 0.30975395430579966, + "grad_norm": 0.2112320065498352, + "learning_rate": 3.080408634041852e-05, + "loss": 0.2077, + "step": 3760 + }, + { + "epoch": 0.31140158172231985, + "grad_norm": 1.9554697275161743, + "learning_rate": 3.096885813148789e-05, + "loss": 0.1861, + "step": 3780 + }, + { + "epoch": 0.31304920913884005, + "grad_norm": 0.7065563201904297, + "learning_rate": 3.113362992255726e-05, + "loss": 0.2071, + "step": 3800 + }, + { + "epoch": 0.3146968365553603, + "grad_norm": 0.4599238634109497, + "learning_rate": 3.129840171362663e-05, + "loss": 0.1655, + "step": 3820 + }, + { + "epoch": 0.3163444639718805, + "grad_norm": 0.5441445708274841, + "learning_rate": 3.1463173504695996e-05, + "loss": 0.1971, + "step": 3840 + }, + { + "epoch": 0.3179920913884007, + "grad_norm": 1.1993862390518188, + "learning_rate": 3.1627945295765364e-05, + "loss": 0.2048, + "step": 3860 + }, + { + "epoch": 0.3196397188049209, + "grad_norm": 0.3095191717147827, + "learning_rate": 3.179271708683473e-05, + "loss": 0.2009, + "step": 3880 + }, + { + "epoch": 0.3212873462214411, + "grad_norm": 1.0743999481201172, + "learning_rate": 3.195748887790411e-05, + "loss": 0.2371, + "step": 3900 + }, + { + "epoch": 0.3229349736379613, + "grad_norm": 0.5000220537185669, + "learning_rate": 3.2122260668973475e-05, + "loss": 0.1722, + "step": 3920 + }, + { + "epoch": 0.32458260105448156, + "grad_norm": 1.1417018175125122, + "learning_rate": 3.2287032460042844e-05, + "loss": 0.2079, + "step": 3940 + }, + { + "epoch": 0.32623022847100175, + "grad_norm": 1.099433422088623, + "learning_rate": 3.245180425111221e-05, + "loss": 0.2009, + "step": 3960 + }, + { + "epoch": 0.32787785588752194, + "grad_norm": 0.3827146589756012, + "learning_rate": 3.261657604218158e-05, + "loss": 0.2171, + "step": 3980 + }, + { + "epoch": 0.3295254833040422, + "grad_norm": 1.1845418214797974, + "learning_rate": 3.278134783325095e-05, + "loss": 0.2195, + "step": 4000 + }, + { + "epoch": 0.3295254833040422, + "eval_loss": 0.5868579149246216, + "eval_runtime": 260.92, + "eval_samples_per_second": 83.508, + "eval_steps_per_second": 20.88, + "eval_wer": 0.2350240606008012, + "step": 4000 + }, + { + "epoch": 0.3311731107205624, + "grad_norm": 1.6073412895202637, + "learning_rate": 3.2946119624320316e-05, + "loss": 0.1695, + "step": 4020 + }, + { + "epoch": 0.33282073813708263, + "grad_norm": 2.9597036838531494, + "learning_rate": 3.311089141538969e-05, + "loss": 0.2484, + "step": 4040 + }, + { + "epoch": 0.3344683655536028, + "grad_norm": 1.4150543212890625, + "learning_rate": 3.327566320645906e-05, + "loss": 0.2087, + "step": 4060 + }, + { + "epoch": 0.336115992970123, + "grad_norm": 0.255397230386734, + "learning_rate": 3.344043499752843e-05, + "loss": 0.2048, + "step": 4080 + }, + { + "epoch": 0.33776362038664326, + "grad_norm": 1.9232691526412964, + "learning_rate": 3.3605206788597795e-05, + "loss": 0.2274, + "step": 4100 + }, + { + "epoch": 0.33941124780316345, + "grad_norm": 1.0172194242477417, + "learning_rate": 3.3769978579667164e-05, + "loss": 0.1779, + "step": 4120 + }, + { + "epoch": 0.34105887521968364, + "grad_norm": 1.3862395286560059, + "learning_rate": 3.393475037073653e-05, + "loss": 0.2095, + "step": 4140 + }, + { + "epoch": 0.3427065026362039, + "grad_norm": 0.3353387117385864, + "learning_rate": 3.40995221618059e-05, + "loss": 0.2021, + "step": 4160 + }, + { + "epoch": 0.3443541300527241, + "grad_norm": 0.9549083709716797, + "learning_rate": 3.426429395287527e-05, + "loss": 0.1997, + "step": 4180 + }, + { + "epoch": 0.3460017574692443, + "grad_norm": 1.6077580451965332, + "learning_rate": 3.4429065743944636e-05, + "loss": 0.2562, + "step": 4200 + }, + { + "epoch": 0.3476493848857645, + "grad_norm": 5.387716770172119, + "learning_rate": 3.4593837535014004e-05, + "loss": 0.1726, + "step": 4220 + }, + { + "epoch": 0.3492970123022847, + "grad_norm": 0.5455642342567444, + "learning_rate": 3.475860932608337e-05, + "loss": 0.2342, + "step": 4240 + }, + { + "epoch": 0.3509446397188049, + "grad_norm": 0.18990729749202728, + "learning_rate": 3.492338111715274e-05, + "loss": 0.2334, + "step": 4260 + }, + { + "epoch": 0.35259226713532515, + "grad_norm": 0.4878564476966858, + "learning_rate": 3.5088152908222115e-05, + "loss": 0.2065, + "step": 4280 + }, + { + "epoch": 0.35423989455184535, + "grad_norm": 1.3400063514709473, + "learning_rate": 3.5252924699291483e-05, + "loss": 0.237, + "step": 4300 + }, + { + "epoch": 0.35588752196836554, + "grad_norm": 0.5822551250457764, + "learning_rate": 3.541769649036085e-05, + "loss": 0.1787, + "step": 4320 + }, + { + "epoch": 0.3575351493848858, + "grad_norm": 0.4629223644733429, + "learning_rate": 3.558246828143022e-05, + "loss": 0.2129, + "step": 4340 + }, + { + "epoch": 0.359182776801406, + "grad_norm": 1.4195072650909424, + "learning_rate": 3.574724007249959e-05, + "loss": 0.2215, + "step": 4360 + }, + { + "epoch": 0.36083040421792617, + "grad_norm": 0.3443647623062134, + "learning_rate": 3.5912011863568956e-05, + "loss": 0.2002, + "step": 4380 + }, + { + "epoch": 0.3624780316344464, + "grad_norm": 3.2932443618774414, + "learning_rate": 3.607678365463833e-05, + "loss": 0.2172, + "step": 4400 + }, + { + "epoch": 0.3641256590509666, + "grad_norm": 0.4463866055011749, + "learning_rate": 3.62415554457077e-05, + "loss": 0.1889, + "step": 4420 + }, + { + "epoch": 0.3657732864674868, + "grad_norm": 0.4873151183128357, + "learning_rate": 3.640632723677707e-05, + "loss": 0.2139, + "step": 4440 + }, + { + "epoch": 0.36742091388400705, + "grad_norm": 1.6354761123657227, + "learning_rate": 3.6571099027846435e-05, + "loss": 0.1986, + "step": 4460 + }, + { + "epoch": 0.36906854130052724, + "grad_norm": 0.5571808815002441, + "learning_rate": 3.6735870818915803e-05, + "loss": 0.2047, + "step": 4480 + }, + { + "epoch": 0.37071616871704743, + "grad_norm": 0.7461993098258972, + "learning_rate": 3.690064260998517e-05, + "loss": 0.2196, + "step": 4500 + }, + { + "epoch": 0.3723637961335677, + "grad_norm": 0.2061534970998764, + "learning_rate": 3.7065414401054546e-05, + "loss": 0.1657, + "step": 4520 + }, + { + "epoch": 0.3740114235500879, + "grad_norm": 1.0977954864501953, + "learning_rate": 3.7230186192123915e-05, + "loss": 0.2062, + "step": 4540 + }, + { + "epoch": 0.37565905096660807, + "grad_norm": 2.38232684135437, + "learning_rate": 3.739495798319328e-05, + "loss": 0.2319, + "step": 4560 + }, + { + "epoch": 0.3773066783831283, + "grad_norm": 0.6182531118392944, + "learning_rate": 3.755972977426265e-05, + "loss": 0.202, + "step": 4580 + }, + { + "epoch": 0.3789543057996485, + "grad_norm": 28.744009017944336, + "learning_rate": 3.772450156533201e-05, + "loss": 0.2235, + "step": 4600 + }, + { + "epoch": 0.3806019332161687, + "grad_norm": 0.5899057984352112, + "learning_rate": 3.788927335640138e-05, + "loss": 0.1577, + "step": 4620 + }, + { + "epoch": 0.38224956063268895, + "grad_norm": 0.5423290133476257, + "learning_rate": 3.8054045147470755e-05, + "loss": 0.2098, + "step": 4640 + }, + { + "epoch": 0.38389718804920914, + "grad_norm": 0.23139849305152893, + "learning_rate": 3.8218816938540123e-05, + "loss": 0.1989, + "step": 4660 + }, + { + "epoch": 0.38554481546572933, + "grad_norm": 0.3539600670337677, + "learning_rate": 3.838358872960949e-05, + "loss": 0.1968, + "step": 4680 + }, + { + "epoch": 0.3871924428822496, + "grad_norm": 0.7127693295478821, + "learning_rate": 3.854836052067886e-05, + "loss": 0.2225, + "step": 4700 + }, + { + "epoch": 0.38884007029876977, + "grad_norm": 0.3489459753036499, + "learning_rate": 3.871313231174823e-05, + "loss": 0.1606, + "step": 4720 + }, + { + "epoch": 0.39048769771528996, + "grad_norm": 0.27798184752464294, + "learning_rate": 3.886966551326413e-05, + "loss": 0.2096, + "step": 4740 + }, + { + "epoch": 0.3921353251318102, + "grad_norm": 0.4229481816291809, + "learning_rate": 3.90344373043335e-05, + "loss": 0.2195, + "step": 4760 + }, + { + "epoch": 0.3937829525483304, + "grad_norm": 0.25523656606674194, + "learning_rate": 3.9199209095402866e-05, + "loss": 0.2106, + "step": 4780 + }, + { + "epoch": 0.3954305799648506, + "grad_norm": 0.9676795601844788, + "learning_rate": 3.936398088647224e-05, + "loss": 0.2267, + "step": 4800 + }, + { + "epoch": 0.39707820738137084, + "grad_norm": 0.40875479578971863, + "learning_rate": 3.952875267754161e-05, + "loss": 0.155, + "step": 4820 + }, + { + "epoch": 0.39872583479789103, + "grad_norm": 0.335033655166626, + "learning_rate": 3.969352446861098e-05, + "loss": 0.2081, + "step": 4840 + }, + { + "epoch": 0.4003734622144112, + "grad_norm": 0.3303823173046112, + "learning_rate": 3.9858296259680345e-05, + "loss": 0.2282, + "step": 4860 + }, + { + "epoch": 0.40202108963093147, + "grad_norm": 0.20934663712978363, + "learning_rate": 4.0023068050749714e-05, + "loss": 0.1965, + "step": 4880 + }, + { + "epoch": 0.40366871704745166, + "grad_norm": 1.4093564748764038, + "learning_rate": 4.018783984181908e-05, + "loss": 0.22, + "step": 4900 + }, + { + "epoch": 0.40531634446397186, + "grad_norm": 0.4276560842990875, + "learning_rate": 4.035261163288846e-05, + "loss": 0.1705, + "step": 4920 + }, + { + "epoch": 0.4069639718804921, + "grad_norm": 0.4243983030319214, + "learning_rate": 4.0517383423957825e-05, + "loss": 0.2285, + "step": 4940 + }, + { + "epoch": 0.4086115992970123, + "grad_norm": 0.16951896250247955, + "learning_rate": 4.068215521502719e-05, + "loss": 0.2234, + "step": 4960 + }, + { + "epoch": 0.4102592267135325, + "grad_norm": 0.30336254835128784, + "learning_rate": 4.084692700609656e-05, + "loss": 0.2152, + "step": 4980 + }, + { + "epoch": 0.41190685413005274, + "grad_norm": 1.0761586427688599, + "learning_rate": 4.101169879716593e-05, + "loss": 0.2093, + "step": 5000 + }, + { + "epoch": 0.4135544815465729, + "grad_norm": 0.7221740484237671, + "learning_rate": 4.11764705882353e-05, + "loss": 0.1538, + "step": 5020 + }, + { + "epoch": 0.4152021089630931, + "grad_norm": 0.4801746904850006, + "learning_rate": 4.1341242379304665e-05, + "loss": 0.1911, + "step": 5040 + }, + { + "epoch": 0.41684973637961337, + "grad_norm": 0.311234712600708, + "learning_rate": 4.1506014170374034e-05, + "loss": 0.2015, + "step": 5060 + }, + { + "epoch": 0.41849736379613356, + "grad_norm": 0.6403760313987732, + "learning_rate": 4.16707859614434e-05, + "loss": 0.2014, + "step": 5080 + }, + { + "epoch": 0.42014499121265375, + "grad_norm": 1.2653217315673828, + "learning_rate": 4.183555775251277e-05, + "loss": 0.2388, + "step": 5100 + }, + { + "epoch": 0.421792618629174, + "grad_norm": 0.5536401867866516, + "learning_rate": 4.200032954358214e-05, + "loss": 0.1553, + "step": 5120 + }, + { + "epoch": 0.4234402460456942, + "grad_norm": 0.4605076014995575, + "learning_rate": 4.2165101334651506e-05, + "loss": 0.2105, + "step": 5140 + }, + { + "epoch": 0.42508787346221444, + "grad_norm": 0.28758004307746887, + "learning_rate": 4.232987312572088e-05, + "loss": 0.2067, + "step": 5160 + }, + { + "epoch": 0.42673550087873463, + "grad_norm": 0.36655622720718384, + "learning_rate": 4.249464491679025e-05, + "loss": 0.1976, + "step": 5180 + }, + { + "epoch": 0.4283831282952548, + "grad_norm": 0.9053062796592712, + "learning_rate": 4.265941670785962e-05, + "loss": 0.2102, + "step": 5200 + }, + { + "epoch": 0.43003075571177507, + "grad_norm": 0.5088081359863281, + "learning_rate": 4.2824188498928985e-05, + "loss": 0.1639, + "step": 5220 + }, + { + "epoch": 0.43167838312829526, + "grad_norm": 0.7981218695640564, + "learning_rate": 4.2988960289998354e-05, + "loss": 0.2024, + "step": 5240 + }, + { + "epoch": 0.43332601054481545, + "grad_norm": 0.7993011474609375, + "learning_rate": 4.315373208106772e-05, + "loss": 0.204, + "step": 5260 + }, + { + "epoch": 0.4349736379613357, + "grad_norm": 0.5183406472206116, + "learning_rate": 4.331850387213709e-05, + "loss": 0.2041, + "step": 5280 + }, + { + "epoch": 0.4366212653778559, + "grad_norm": 1.2263312339782715, + "learning_rate": 4.3483275663206465e-05, + "loss": 0.2182, + "step": 5300 + }, + { + "epoch": 0.4382688927943761, + "grad_norm": 0.7018775343894958, + "learning_rate": 4.364804745427583e-05, + "loss": 0.1685, + "step": 5320 + }, + { + "epoch": 0.43991652021089633, + "grad_norm": 0.7381752133369446, + "learning_rate": 4.38128192453452e-05, + "loss": 0.2073, + "step": 5340 + }, + { + "epoch": 0.4415641476274165, + "grad_norm": 0.4658122956752777, + "learning_rate": 4.397759103641457e-05, + "loss": 0.2031, + "step": 5360 + }, + { + "epoch": 0.4432117750439367, + "grad_norm": 0.34789761900901794, + "learning_rate": 4.414236282748394e-05, + "loss": 0.2023, + "step": 5380 + }, + { + "epoch": 0.44485940246045697, + "grad_norm": 0.9787063598632812, + "learning_rate": 4.4307134618553305e-05, + "loss": 0.2264, + "step": 5400 + }, + { + "epoch": 0.44650702987697716, + "grad_norm": 0.3786025047302246, + "learning_rate": 4.4471906409622673e-05, + "loss": 0.1656, + "step": 5420 + }, + { + "epoch": 0.44815465729349735, + "grad_norm": 0.4397692084312439, + "learning_rate": 4.463667820069204e-05, + "loss": 0.2023, + "step": 5440 + }, + { + "epoch": 0.4498022847100176, + "grad_norm": 0.20323756337165833, + "learning_rate": 4.480144999176141e-05, + "loss": 0.2146, + "step": 5460 + }, + { + "epoch": 0.4514499121265378, + "grad_norm": 0.2108180820941925, + "learning_rate": 4.496622178283078e-05, + "loss": 0.2012, + "step": 5480 + }, + { + "epoch": 0.453097539543058, + "grad_norm": 1.6603738069534302, + "learning_rate": 4.5130993573900146e-05, + "loss": 0.2257, + "step": 5500 + }, + { + "epoch": 0.45474516695957823, + "grad_norm": 1.6646907329559326, + "learning_rate": 4.5295765364969514e-05, + "loss": 0.1636, + "step": 5520 + }, + { + "epoch": 0.4563927943760984, + "grad_norm": 0.98222416639328, + "learning_rate": 4.546053715603889e-05, + "loss": 0.2018, + "step": 5540 + }, + { + "epoch": 0.4580404217926186, + "grad_norm": 0.6065666079521179, + "learning_rate": 4.562530894710826e-05, + "loss": 0.2351, + "step": 5560 + }, + { + "epoch": 0.45968804920913886, + "grad_norm": 0.4321737587451935, + "learning_rate": 4.5790080738177625e-05, + "loss": 0.2139, + "step": 5580 + }, + { + "epoch": 0.46133567662565905, + "grad_norm": 2.530203342437744, + "learning_rate": 4.5954852529246993e-05, + "loss": 0.232, + "step": 5600 + }, + { + "epoch": 0.46298330404217924, + "grad_norm": 0.8252795934677124, + "learning_rate": 4.611962432031636e-05, + "loss": 0.1792, + "step": 5620 + }, + { + "epoch": 0.4646309314586995, + "grad_norm": 4.3282880783081055, + "learning_rate": 4.628439611138573e-05, + "loss": 0.2079, + "step": 5640 + }, + { + "epoch": 0.4662785588752197, + "grad_norm": 0.2798108756542206, + "learning_rate": 4.6449167902455105e-05, + "loss": 0.2114, + "step": 5660 + }, + { + "epoch": 0.4679261862917399, + "grad_norm": 0.1499057412147522, + "learning_rate": 4.661393969352447e-05, + "loss": 0.2026, + "step": 5680 + }, + { + "epoch": 0.4695738137082601, + "grad_norm": 0.8664823770523071, + "learning_rate": 4.677871148459384e-05, + "loss": 0.2208, + "step": 5700 + }, + { + "epoch": 0.4712214411247803, + "grad_norm": 0.32247450947761536, + "learning_rate": 4.694348327566321e-05, + "loss": 0.175, + "step": 5720 + }, + { + "epoch": 0.4728690685413005, + "grad_norm": 0.4217327833175659, + "learning_rate": 4.710825506673258e-05, + "loss": 0.2207, + "step": 5740 + }, + { + "epoch": 0.47451669595782076, + "grad_norm": 0.9544996023178101, + "learning_rate": 4.7273026857801945e-05, + "loss": 0.2269, + "step": 5760 + }, + { + "epoch": 0.47616432337434095, + "grad_norm": 0.3182899057865143, + "learning_rate": 4.743779864887132e-05, + "loss": 0.224, + "step": 5780 + }, + { + "epoch": 0.47781195079086114, + "grad_norm": 0.669552743434906, + "learning_rate": 4.760257043994069e-05, + "loss": 0.2332, + "step": 5800 + }, + { + "epoch": 0.4794595782073814, + "grad_norm": 0.5297145247459412, + "learning_rate": 4.7767342231010056e-05, + "loss": 0.2233, + "step": 5820 + }, + { + "epoch": 0.4811072056239016, + "grad_norm": 0.5998116731643677, + "learning_rate": 4.7932114022079425e-05, + "loss": 0.2282, + "step": 5840 + }, + { + "epoch": 0.48275483304042177, + "grad_norm": 0.4391906261444092, + "learning_rate": 4.809688581314879e-05, + "loss": 0.2094, + "step": 5860 + }, + { + "epoch": 0.484402460456942, + "grad_norm": 0.559304416179657, + "learning_rate": 4.826165760421816e-05, + "loss": 0.2106, + "step": 5880 + }, + { + "epoch": 0.4860500878734622, + "grad_norm": 2.914066791534424, + "learning_rate": 4.842642939528753e-05, + "loss": 0.2484, + "step": 5900 + }, + { + "epoch": 0.4876977152899824, + "grad_norm": 0.29989850521087646, + "learning_rate": 4.85912011863569e-05, + "loss": 0.1631, + "step": 5920 + }, + { + "epoch": 0.48934534270650265, + "grad_norm": 2.1414361000061035, + "learning_rate": 4.8755972977426265e-05, + "loss": 0.2315, + "step": 5940 + }, + { + "epoch": 0.49099297012302284, + "grad_norm": 0.5668668746948242, + "learning_rate": 4.892074476849563e-05, + "loss": 0.218, + "step": 5960 + }, + { + "epoch": 0.49264059753954303, + "grad_norm": 0.34356266260147095, + "learning_rate": 4.9077277970011535e-05, + "loss": 0.2121, + "step": 5980 + }, + { + "epoch": 0.4942882249560633, + "grad_norm": 0.7654836773872375, + "learning_rate": 4.9242049761080904e-05, + "loss": 0.212, + "step": 6000 + }, + { + "epoch": 0.4942882249560633, + "eval_loss": 0.5638302564620972, + "eval_runtime": 264.9088, + "eval_samples_per_second": 82.251, + "eval_steps_per_second": 20.566, + "eval_wer": 0.24026168527782366, + "step": 6000 + }, + { + "epoch": 0.4959358523725835, + "grad_norm": 1.2249395847320557, + "learning_rate": 4.940682155215027e-05, + "loss": 0.1786, + "step": 6020 + }, + { + "epoch": 0.49758347978910367, + "grad_norm": 0.834241509437561, + "learning_rate": 4.957159334321964e-05, + "loss": 0.2193, + "step": 6040 + }, + { + "epoch": 0.4992311072056239, + "grad_norm": 1.9017058610916138, + "learning_rate": 4.9736365134289015e-05, + "loss": 0.2338, + "step": 6060 + }, + { + "epoch": 0.5008787346221442, + "grad_norm": 3.1669952869415283, + "learning_rate": 4.990113692535838e-05, + "loss": 0.2245, + "step": 6080 + }, + { + "epoch": 0.5025263620386643, + "grad_norm": 2.9794270992279053, + "learning_rate": 4.9999997353508944e-05, + "loss": 0.239, + "step": 6100 + }, + { + "epoch": 0.5041739894551845, + "grad_norm": 1.2019259929656982, + "learning_rate": 4.999996758049098e-05, + "loss": 0.1871, + "step": 6120 + }, + { + "epoch": 0.5058216168717048, + "grad_norm": 2.175334930419922, + "learning_rate": 4.999990472638076e-05, + "loss": 0.2265, + "step": 6140 + }, + { + "epoch": 0.5074692442882249, + "grad_norm": 1.4895635843276978, + "learning_rate": 4.999980879126146e-05, + "loss": 0.2259, + "step": 6160 + }, + { + "epoch": 0.5091168717047452, + "grad_norm": 3.355076789855957, + "learning_rate": 4.9999679775260015e-05, + "loss": 0.2316, + "step": 6180 + }, + { + "epoch": 0.5107644991212654, + "grad_norm": 3.545771598815918, + "learning_rate": 4.999951767854715e-05, + "loss": 0.2426, + "step": 6200 + }, + { + "epoch": 0.5124121265377856, + "grad_norm": 2.085190773010254, + "learning_rate": 4.999932250133736e-05, + "loss": 0.1884, + "step": 6220 + }, + { + "epoch": 0.5140597539543058, + "grad_norm": 1.6277663707733154, + "learning_rate": 4.999909424388892e-05, + "loss": 0.2332, + "step": 6240 + }, + { + "epoch": 0.5157073813708261, + "grad_norm": 3.9234516620635986, + "learning_rate": 4.9998832906503856e-05, + "loss": 0.2346, + "step": 6260 + }, + { + "epoch": 0.5173550087873462, + "grad_norm": 1.7027534246444702, + "learning_rate": 4.9998538489527984e-05, + "loss": 0.2339, + "step": 6280 + }, + { + "epoch": 0.5190026362038664, + "grad_norm": 2.7195184230804443, + "learning_rate": 4.9998210993350895e-05, + "loss": 0.2511, + "step": 6300 + }, + { + "epoch": 0.5206502636203867, + "grad_norm": 0.7267903089523315, + "learning_rate": 4.9997850418405945e-05, + "loss": 0.1765, + "step": 6320 + }, + { + "epoch": 0.5222978910369068, + "grad_norm": 0.751308798789978, + "learning_rate": 4.999745676517027e-05, + "loss": 0.2393, + "step": 6340 + }, + { + "epoch": 0.5239455184534271, + "grad_norm": 1.0236356258392334, + "learning_rate": 4.999703003416476e-05, + "loss": 0.2292, + "step": 6360 + }, + { + "epoch": 0.5255931458699473, + "grad_norm": 0.7693842649459839, + "learning_rate": 4.999657022595409e-05, + "loss": 0.2285, + "step": 6380 + }, + { + "epoch": 0.5272407732864675, + "grad_norm": 3.4103684425354004, + "learning_rate": 4.999607734114669e-05, + "loss": 0.2635, + "step": 6400 + }, + { + "epoch": 0.5288884007029877, + "grad_norm": 0.5887142419815063, + "learning_rate": 4.999555138039478e-05, + "loss": 0.1938, + "step": 6420 + }, + { + "epoch": 0.530536028119508, + "grad_norm": 1.6326848268508911, + "learning_rate": 4.999499234439433e-05, + "loss": 0.2112, + "step": 6440 + }, + { + "epoch": 0.5321836555360281, + "grad_norm": 0.4003719985485077, + "learning_rate": 4.9994400233885086e-05, + "loss": 0.2011, + "step": 6460 + }, + { + "epoch": 0.5338312829525483, + "grad_norm": 1.0633249282836914, + "learning_rate": 4.999377504965055e-05, + "loss": 0.2388, + "step": 6480 + }, + { + "epoch": 0.5354789103690686, + "grad_norm": 3.87857985496521, + "learning_rate": 4.999311679251799e-05, + "loss": 0.2222, + "step": 6500 + }, + { + "epoch": 0.5371265377855887, + "grad_norm": 1.1982243061065674, + "learning_rate": 4.999242546335845e-05, + "loss": 0.1736, + "step": 6520 + }, + { + "epoch": 0.538774165202109, + "grad_norm": 1.4149129390716553, + "learning_rate": 4.999170106308673e-05, + "loss": 0.2044, + "step": 6540 + }, + { + "epoch": 0.5404217926186292, + "grad_norm": 0.9540938138961792, + "learning_rate": 4.999094359266138e-05, + "loss": 0.2343, + "step": 6560 + }, + { + "epoch": 0.5420694200351494, + "grad_norm": 0.8150781989097595, + "learning_rate": 4.999015305308472e-05, + "loss": 0.2027, + "step": 6580 + }, + { + "epoch": 0.5437170474516696, + "grad_norm": 2.2410573959350586, + "learning_rate": 4.998932944540284e-05, + "loss": 0.2191, + "step": 6600 + }, + { + "epoch": 0.5453646748681898, + "grad_norm": 2.924471616744995, + "learning_rate": 4.998847277070556e-05, + "loss": 0.1776, + "step": 6620 + }, + { + "epoch": 0.54701230228471, + "grad_norm": 2.5102221965789795, + "learning_rate": 4.9987583030126484e-05, + "loss": 0.2172, + "step": 6640 + }, + { + "epoch": 0.5486599297012302, + "grad_norm": 1.2676066160202026, + "learning_rate": 4.998666022484295e-05, + "loss": 0.2235, + "step": 6660 + }, + { + "epoch": 0.5503075571177505, + "grad_norm": 1.7459567785263062, + "learning_rate": 4.998570435607605e-05, + "loss": 0.2114, + "step": 6680 + }, + { + "epoch": 0.5519551845342706, + "grad_norm": 2.4994170665740967, + "learning_rate": 4.998476565684759e-05, + "loss": 0.2321, + "step": 6700 + }, + { + "epoch": 0.5536028119507909, + "grad_norm": 1.436635136604309, + "learning_rate": 4.998374531796601e-05, + "loss": 0.1715, + "step": 6720 + }, + { + "epoch": 0.5552504393673111, + "grad_norm": 5.170281410217285, + "learning_rate": 4.9982691919458215e-05, + "loss": 0.2227, + "step": 6740 + }, + { + "epoch": 0.5568980667838312, + "grad_norm": 2.329806327819824, + "learning_rate": 4.9981605462718097e-05, + "loss": 0.2103, + "step": 6760 + }, + { + "epoch": 0.5585456942003515, + "grad_norm": 0.5105818510055542, + "learning_rate": 4.998048594918331e-05, + "loss": 0.2297, + "step": 6780 + }, + { + "epoch": 0.5601933216168717, + "grad_norm": 2.74815034866333, + "learning_rate": 4.997933338033525e-05, + "loss": 0.2454, + "step": 6800 + }, + { + "epoch": 0.5618409490333919, + "grad_norm": 1.0433367490768433, + "learning_rate": 4.997814775769904e-05, + "loss": 0.175, + "step": 6820 + }, + { + "epoch": 0.5634885764499121, + "grad_norm": 0.5902445316314697, + "learning_rate": 4.997692908284356e-05, + "loss": 0.2153, + "step": 6840 + }, + { + "epoch": 0.5651362038664324, + "grad_norm": 0.9332594275474548, + "learning_rate": 4.997567735738141e-05, + "loss": 0.2246, + "step": 6860 + }, + { + "epoch": 0.5667838312829525, + "grad_norm": 1.2199161052703857, + "learning_rate": 4.9974392582968934e-05, + "loss": 0.2281, + "step": 6880 + }, + { + "epoch": 0.5684314586994728, + "grad_norm": 2.2492709159851074, + "learning_rate": 4.9973074761306186e-05, + "loss": 0.238, + "step": 6900 + }, + { + "epoch": 0.570079086115993, + "grad_norm": 9.928802490234375, + "learning_rate": 4.997172389413699e-05, + "loss": 0.1731, + "step": 6920 + }, + { + "epoch": 0.5717267135325131, + "grad_norm": 1.4187493324279785, + "learning_rate": 4.997033998324886e-05, + "loss": 0.2267, + "step": 6940 + }, + { + "epoch": 0.5733743409490334, + "grad_norm": 1.0871655941009521, + "learning_rate": 4.996892303047306e-05, + "loss": 0.2478, + "step": 6960 + }, + { + "epoch": 0.5750219683655536, + "grad_norm": 0.9939442873001099, + "learning_rate": 4.996747303768456e-05, + "loss": 0.1992, + "step": 6980 + }, + { + "epoch": 0.5766695957820738, + "grad_norm": 48.66152572631836, + "learning_rate": 4.996599000680206e-05, + "loss": 0.2341, + "step": 7000 + }, + { + "epoch": 0.578317223198594, + "grad_norm": 0.8890938758850098, + "learning_rate": 4.996447393978797e-05, + "loss": 0.1687, + "step": 7020 + }, + { + "epoch": 0.5799648506151143, + "grad_norm": 0.8341367840766907, + "learning_rate": 4.996292483864843e-05, + "loss": 0.2175, + "step": 7040 + }, + { + "epoch": 0.5816124780316344, + "grad_norm": 1.6562838554382324, + "learning_rate": 4.996134270543326e-05, + "loss": 0.2136, + "step": 7060 + }, + { + "epoch": 0.5832601054481547, + "grad_norm": 0.6537133455276489, + "learning_rate": 4.9959727542236025e-05, + "loss": 0.2139, + "step": 7080 + }, + { + "epoch": 0.5849077328646749, + "grad_norm": 2.101271390914917, + "learning_rate": 4.9958079351193976e-05, + "loss": 0.2359, + "step": 7100 + }, + { + "epoch": 0.586555360281195, + "grad_norm": 0.9616307616233826, + "learning_rate": 4.995639813448808e-05, + "loss": 0.1774, + "step": 7120 + }, + { + "epoch": 0.5882029876977153, + "grad_norm": 1.0992920398712158, + "learning_rate": 4.9954683894343e-05, + "loss": 0.1962, + "step": 7140 + }, + { + "epoch": 0.5898506151142355, + "grad_norm": 0.6329948902130127, + "learning_rate": 4.995293663302709e-05, + "loss": 0.2145, + "step": 7160 + }, + { + "epoch": 0.5914982425307557, + "grad_norm": 1.0156137943267822, + "learning_rate": 4.9951156352852415e-05, + "loss": 0.2161, + "step": 7180 + }, + { + "epoch": 0.5931458699472759, + "grad_norm": 2.6131973266601562, + "learning_rate": 4.994943450511368e-05, + "loss": 0.2234, + "step": 7200 + }, + { + "epoch": 0.5947934973637962, + "grad_norm": 3.4186134338378906, + "learning_rate": 4.99475898449799e-05, + "loss": 0.171, + "step": 7220 + }, + { + "epoch": 0.5964411247803163, + "grad_norm": 4.896924018859863, + "learning_rate": 4.9945712173062477e-05, + "loss": 0.2073, + "step": 7240 + }, + { + "epoch": 0.5980887521968365, + "grad_norm": 1.2636560201644897, + "learning_rate": 4.994380149184601e-05, + "loss": 0.2144, + "step": 7260 + }, + { + "epoch": 0.5997363796133568, + "grad_norm": 1.4861242771148682, + "learning_rate": 4.99418578038588e-05, + "loss": 0.2024, + "step": 7280 + }, + { + "epoch": 0.6013840070298769, + "grad_norm": 6.527903079986572, + "learning_rate": 4.993988111167284e-05, + "loss": 0.2172, + "step": 7300 + }, + { + "epoch": 0.6030316344463972, + "grad_norm": 1.9219621419906616, + "learning_rate": 4.993787141790375e-05, + "loss": 0.1865, + "step": 7320 + }, + { + "epoch": 0.6046792618629174, + "grad_norm": 0.7625167369842529, + "learning_rate": 4.9935828725210874e-05, + "loss": 0.2238, + "step": 7340 + }, + { + "epoch": 0.6063268892794376, + "grad_norm": 1.3375577926635742, + "learning_rate": 4.9933753036297196e-05, + "loss": 0.1878, + "step": 7360 + }, + { + "epoch": 0.6079745166959578, + "grad_norm": 1.1067700386047363, + "learning_rate": 4.993164435390935e-05, + "loss": 0.2111, + "step": 7380 + }, + { + "epoch": 0.6096221441124781, + "grad_norm": 4.094808578491211, + "learning_rate": 4.992950268083764e-05, + "loss": 0.2073, + "step": 7400 + }, + { + "epoch": 0.6112697715289982, + "grad_norm": 2.683678150177002, + "learning_rate": 4.992732801991602e-05, + "loss": 0.172, + "step": 7420 + }, + { + "epoch": 0.6129173989455184, + "grad_norm": 0.9752517342567444, + "learning_rate": 4.992512037402212e-05, + "loss": 0.1924, + "step": 7440 + }, + { + "epoch": 0.6145650263620387, + "grad_norm": 0.8463476896286011, + "learning_rate": 4.9922879746077176e-05, + "loss": 0.2021, + "step": 7460 + }, + { + "epoch": 0.6162126537785588, + "grad_norm": 3.053812026977539, + "learning_rate": 4.992060613904611e-05, + "loss": 0.2086, + "step": 7480 + }, + { + "epoch": 0.6178602811950791, + "grad_norm": 2.9483275413513184, + "learning_rate": 4.991829955593744e-05, + "loss": 0.2154, + "step": 7500 + }, + { + "epoch": 0.6195079086115993, + "grad_norm": 1.3794467449188232, + "learning_rate": 4.9915959999803365e-05, + "loss": 0.1599, + "step": 7520 + }, + { + "epoch": 0.6211555360281195, + "grad_norm": 1.0529364347457886, + "learning_rate": 4.9913587473739666e-05, + "loss": 0.1891, + "step": 7540 + }, + { + "epoch": 0.6228031634446397, + "grad_norm": 0.9026776552200317, + "learning_rate": 4.991118198088579e-05, + "loss": 0.2074, + "step": 7560 + }, + { + "epoch": 0.62445079086116, + "grad_norm": 0.4504956603050232, + "learning_rate": 4.9908743524424806e-05, + "loss": 0.1931, + "step": 7580 + }, + { + "epoch": 0.6260984182776801, + "grad_norm": 4.0763726234436035, + "learning_rate": 4.9906272107583366e-05, + "loss": 0.2192, + "step": 7600 + }, + { + "epoch": 0.6277460456942003, + "grad_norm": 1.012601375579834, + "learning_rate": 4.990376773363178e-05, + "loss": 0.1604, + "step": 7620 + }, + { + "epoch": 0.6293936731107206, + "grad_norm": 0.6183291673660278, + "learning_rate": 4.990123040588395e-05, + "loss": 0.2015, + "step": 7640 + }, + { + "epoch": 0.6310413005272407, + "grad_norm": 0.718015730381012, + "learning_rate": 4.989866012769736e-05, + "loss": 0.2155, + "step": 7660 + }, + { + "epoch": 0.632688927943761, + "grad_norm": 1.5611259937286377, + "learning_rate": 4.989605690247315e-05, + "loss": 0.1975, + "step": 7680 + }, + { + "epoch": 0.6343365553602812, + "grad_norm": 3.2936527729034424, + "learning_rate": 4.9893420733656e-05, + "loss": 0.2305, + "step": 7700 + }, + { + "epoch": 0.6359841827768014, + "grad_norm": 1.3771491050720215, + "learning_rate": 4.9890751624734225e-05, + "loss": 0.1575, + "step": 7720 + }, + { + "epoch": 0.6376318101933216, + "grad_norm": 1.0577577352523804, + "learning_rate": 4.98880495792397e-05, + "loss": 0.2045, + "step": 7740 + }, + { + "epoch": 0.6392794376098418, + "grad_norm": 3.6020660400390625, + "learning_rate": 4.988531460074791e-05, + "loss": 0.2034, + "step": 7760 + }, + { + "epoch": 0.640927065026362, + "grad_norm": 0.7804440855979919, + "learning_rate": 4.9882546692877885e-05, + "loss": 0.1953, + "step": 7780 + }, + { + "epoch": 0.6425746924428822, + "grad_norm": 57.011390686035156, + "learning_rate": 4.987974585929226e-05, + "loss": 0.2207, + "step": 7800 + }, + { + "epoch": 0.6442223198594025, + "grad_norm": 1.5725603103637695, + "learning_rate": 4.987691210369721e-05, + "loss": 0.1664, + "step": 7820 + }, + { + "epoch": 0.6458699472759226, + "grad_norm": 0.6347635984420776, + "learning_rate": 4.98740454298425e-05, + "loss": 0.2134, + "step": 7840 + }, + { + "epoch": 0.6475175746924429, + "grad_norm": 1.3581891059875488, + "learning_rate": 4.987114584152145e-05, + "loss": 0.217, + "step": 7860 + }, + { + "epoch": 0.6491652021089631, + "grad_norm": 5.659873008728027, + "learning_rate": 4.986821334257091e-05, + "loss": 0.1977, + "step": 7880 + }, + { + "epoch": 0.6508128295254832, + "grad_norm": 11.7850980758667, + "learning_rate": 4.986524793687131e-05, + "loss": 0.218, + "step": 7900 + }, + { + "epoch": 0.6524604569420035, + "grad_norm": 1.2856826782226562, + "learning_rate": 4.986224962834659e-05, + "loss": 0.1675, + "step": 7920 + }, + { + "epoch": 0.6541080843585237, + "grad_norm": 1.055503249168396, + "learning_rate": 4.985921842096427e-05, + "loss": 0.2113, + "step": 7940 + }, + { + "epoch": 0.6557557117750439, + "grad_norm": 1.5587449073791504, + "learning_rate": 4.9856154318735374e-05, + "loss": 0.1939, + "step": 7960 + }, + { + "epoch": 0.6574033391915641, + "grad_norm": 2.40641450881958, + "learning_rate": 4.985305732571446e-05, + "loss": 0.206, + "step": 7980 + }, + { + "epoch": 0.6590509666080844, + "grad_norm": 4.06686544418335, + "learning_rate": 4.98499274459996e-05, + "loss": 0.2107, + "step": 8000 + }, + { + "epoch": 0.6590509666080844, + "eval_loss": 0.5655311346054077, + "eval_runtime": 835.4807, + "eval_samples_per_second": 26.08, + "eval_steps_per_second": 6.521, + "eval_wer": 0.2457764477651215, + "step": 8000 + }, + { + "epoch": 0.6606985940246046, + "grad_norm": 0.8585990071296692, + "learning_rate": 4.984676468373241e-05, + "loss": 0.1667, + "step": 8020 + }, + { + "epoch": 0.6623462214411248, + "grad_norm": 1.2851459980010986, + "learning_rate": 4.984356904309799e-05, + "loss": 0.2184, + "step": 8040 + }, + { + "epoch": 0.663993848857645, + "grad_norm": 0.8445897698402405, + "learning_rate": 4.984034052832496e-05, + "loss": 0.2169, + "step": 8060 + }, + { + "epoch": 0.6656414762741653, + "grad_norm": 1.4265161752700806, + "learning_rate": 4.983707914368544e-05, + "loss": 0.2004, + "step": 8080 + }, + { + "epoch": 0.6672891036906854, + "grad_norm": 3.0620169639587402, + "learning_rate": 4.983378489349504e-05, + "loss": 0.2467, + "step": 8100 + }, + { + "epoch": 0.6689367311072056, + "grad_norm": 1.1079747676849365, + "learning_rate": 4.983045778211286e-05, + "loss": 0.1587, + "step": 8120 + }, + { + "epoch": 0.6705843585237259, + "grad_norm": 0.8565286993980408, + "learning_rate": 4.982709781394148e-05, + "loss": 0.2101, + "step": 8140 + }, + { + "epoch": 0.672231985940246, + "grad_norm": 2.957345962524414, + "learning_rate": 4.982370499342698e-05, + "loss": 0.212, + "step": 8160 + }, + { + "epoch": 0.6738796133567663, + "grad_norm": 1.744805932044983, + "learning_rate": 4.982027932505887e-05, + "loss": 0.209, + "step": 8180 + }, + { + "epoch": 0.6755272407732865, + "grad_norm": 3.963977336883545, + "learning_rate": 4.9816820813370166e-05, + "loss": 0.214, + "step": 8200 + }, + { + "epoch": 0.6771748681898067, + "grad_norm": 1.3982582092285156, + "learning_rate": 4.981332946293733e-05, + "loss": 0.1671, + "step": 8220 + }, + { + "epoch": 0.6788224956063269, + "grad_norm": 0.5959907174110413, + "learning_rate": 4.9809805278380264e-05, + "loss": 0.2061, + "step": 8240 + }, + { + "epoch": 0.6804701230228472, + "grad_norm": 1.2237777709960938, + "learning_rate": 4.980624826436233e-05, + "loss": 0.2188, + "step": 8260 + }, + { + "epoch": 0.6821177504393673, + "grad_norm": 2.6652753353118896, + "learning_rate": 4.9802658425590344e-05, + "loss": 0.1964, + "step": 8280 + }, + { + "epoch": 0.6837653778558875, + "grad_norm": 31.890893936157227, + "learning_rate": 4.979903576681453e-05, + "loss": 0.2133, + "step": 8300 + }, + { + "epoch": 0.6854130052724078, + "grad_norm": 1.1191130876541138, + "learning_rate": 4.979538029282855e-05, + "loss": 0.1623, + "step": 8320 + }, + { + "epoch": 0.6870606326889279, + "grad_norm": 1.4404329061508179, + "learning_rate": 4.9791692008469514e-05, + "loss": 0.2087, + "step": 8340 + }, + { + "epoch": 0.6887082601054482, + "grad_norm": 1.4656809568405151, + "learning_rate": 4.9787970918617914e-05, + "loss": 0.2134, + "step": 8360 + }, + { + "epoch": 0.6903558875219684, + "grad_norm": 2.310316801071167, + "learning_rate": 4.978421702819767e-05, + "loss": 0.215, + "step": 8380 + }, + { + "epoch": 0.6920035149384886, + "grad_norm": 21.141889572143555, + "learning_rate": 4.978043034217609e-05, + "loss": 0.2206, + "step": 8400 + }, + { + "epoch": 0.6936511423550088, + "grad_norm": 1.2799972295761108, + "learning_rate": 4.977680261809319e-05, + "loss": 0.1553, + "step": 8420 + }, + { + "epoch": 0.695298769771529, + "grad_norm": 0.927029550075531, + "learning_rate": 4.97729519951006e-05, + "loss": 0.2205, + "step": 8440 + }, + { + "epoch": 0.6969463971880492, + "grad_norm": 1.8685028553009033, + "learning_rate": 4.976906859141309e-05, + "loss": 0.1938, + "step": 8460 + }, + { + "epoch": 0.6985940246045694, + "grad_norm": 0.38425010442733765, + "learning_rate": 4.976515241216936e-05, + "loss": 0.1964, + "step": 8480 + }, + { + "epoch": 0.7002416520210897, + "grad_norm": 2.6303980350494385, + "learning_rate": 4.976120346255146e-05, + "loss": 0.1984, + "step": 8500 + }, + { + "epoch": 0.7018892794376098, + "grad_norm": 7.791947841644287, + "learning_rate": 4.975722174778482e-05, + "loss": 0.1678, + "step": 8520 + }, + { + "epoch": 0.7035369068541301, + "grad_norm": 1.193328857421875, + "learning_rate": 4.9753207273138245e-05, + "loss": 0.2182, + "step": 8540 + }, + { + "epoch": 0.7051845342706503, + "grad_norm": 1.1064016819000244, + "learning_rate": 4.974916004392385e-05, + "loss": 0.2065, + "step": 8560 + }, + { + "epoch": 0.7068321616871704, + "grad_norm": 3.304832696914673, + "learning_rate": 4.974508006549711e-05, + "loss": 0.1872, + "step": 8580 + }, + { + "epoch": 0.7084797891036907, + "grad_norm": 2.5478882789611816, + "learning_rate": 4.974096734325686e-05, + "loss": 0.2295, + "step": 8600 + }, + { + "epoch": 0.7101274165202109, + "grad_norm": 14.596063613891602, + "learning_rate": 4.9736821882645226e-05, + "loss": 0.1628, + "step": 8620 + }, + { + "epoch": 0.7117750439367311, + "grad_norm": 0.9999619126319885, + "learning_rate": 4.973264368914766e-05, + "loss": 0.2107, + "step": 8640 + }, + { + "epoch": 0.7134226713532513, + "grad_norm": 1.053412675857544, + "learning_rate": 4.972843276829296e-05, + "loss": 0.2085, + "step": 8660 + }, + { + "epoch": 0.7150702987697716, + "grad_norm": 2.1676464080810547, + "learning_rate": 4.9724189125653195e-05, + "loss": 0.2048, + "step": 8680 + }, + { + "epoch": 0.7167179261862917, + "grad_norm": 10.32970142364502, + "learning_rate": 4.9719912766843746e-05, + "loss": 0.2224, + "step": 8700 + }, + { + "epoch": 0.718365553602812, + "grad_norm": 1.9829623699188232, + "learning_rate": 4.971560369752328e-05, + "loss": 0.1686, + "step": 8720 + }, + { + "epoch": 0.7200131810193322, + "grad_norm": 2.7383005619049072, + "learning_rate": 4.971126192339377e-05, + "loss": 0.2088, + "step": 8740 + }, + { + "epoch": 0.7216608084358523, + "grad_norm": 0.6413734555244446, + "learning_rate": 4.970688745020043e-05, + "loss": 0.2111, + "step": 8760 + }, + { + "epoch": 0.7233084358523726, + "grad_norm": 0.8469798564910889, + "learning_rate": 4.970248028373178e-05, + "loss": 0.2057, + "step": 8780 + }, + { + "epoch": 0.7249560632688928, + "grad_norm": 8.44117259979248, + "learning_rate": 4.969804042981956e-05, + "loss": 0.2179, + "step": 8800 + }, + { + "epoch": 0.726603690685413, + "grad_norm": 1.40958571434021, + "learning_rate": 4.969356789433881e-05, + "loss": 0.1604, + "step": 8820 + }, + { + "epoch": 0.7282513181019332, + "grad_norm": 5.328885078430176, + "learning_rate": 4.968906268320777e-05, + "loss": 0.2075, + "step": 8840 + }, + { + "epoch": 0.7298989455184535, + "grad_norm": 0.6396345496177673, + "learning_rate": 4.9684524802387956e-05, + "loss": 0.2057, + "step": 8860 + }, + { + "epoch": 0.7315465729349736, + "grad_norm": 0.7661327123641968, + "learning_rate": 4.967995425788409e-05, + "loss": 0.2064, + "step": 8880 + }, + { + "epoch": 0.7331942003514939, + "grad_norm": 2.8300130367279053, + "learning_rate": 4.9675351055744134e-05, + "loss": 0.2109, + "step": 8900 + }, + { + "epoch": 0.7348418277680141, + "grad_norm": 1.3380045890808105, + "learning_rate": 4.9670715202059235e-05, + "loss": 0.1492, + "step": 8920 + }, + { + "epoch": 0.7364894551845342, + "grad_norm": 1.478825569152832, + "learning_rate": 4.9666046702963784e-05, + "loss": 0.2133, + "step": 8940 + }, + { + "epoch": 0.7381370826010545, + "grad_norm": 1.3596333265304565, + "learning_rate": 4.9661345564635356e-05, + "loss": 0.1999, + "step": 8960 + }, + { + "epoch": 0.7397847100175747, + "grad_norm": 0.3950871229171753, + "learning_rate": 4.965661179329468e-05, + "loss": 0.1925, + "step": 8980 + }, + { + "epoch": 0.7414323374340949, + "grad_norm": 3.47658634185791, + "learning_rate": 4.965184539520572e-05, + "loss": 0.2016, + "step": 9000 + }, + { + "epoch": 0.7430799648506151, + "grad_norm": 1.6241281032562256, + "learning_rate": 4.9647046376675586e-05, + "loss": 0.1522, + "step": 9020 + }, + { + "epoch": 0.7447275922671354, + "grad_norm": 0.582032322883606, + "learning_rate": 4.964221474405456e-05, + "loss": 0.2161, + "step": 9040 + }, + { + "epoch": 0.7463752196836555, + "grad_norm": 1.1760663986206055, + "learning_rate": 4.963735050373608e-05, + "loss": 0.1986, + "step": 9060 + }, + { + "epoch": 0.7480228471001757, + "grad_norm": 1.9206633567810059, + "learning_rate": 4.963245366215672e-05, + "loss": 0.195, + "step": 9080 + }, + { + "epoch": 0.749670474516696, + "grad_norm": 2.0300755500793457, + "learning_rate": 4.9627524225796206e-05, + "loss": 0.2044, + "step": 9100 + }, + { + "epoch": 0.7513181019332161, + "grad_norm": 0.91159987449646, + "learning_rate": 4.962256220117739e-05, + "loss": 0.156, + "step": 9120 + }, + { + "epoch": 0.7529657293497364, + "grad_norm": 0.6553214192390442, + "learning_rate": 4.961756759486625e-05, + "loss": 0.1918, + "step": 9140 + }, + { + "epoch": 0.7546133567662566, + "grad_norm": 2.1197195053100586, + "learning_rate": 4.961254041347189e-05, + "loss": 0.1942, + "step": 9160 + }, + { + "epoch": 0.7562609841827768, + "grad_norm": 0.5275347232818604, + "learning_rate": 4.9607480663646487e-05, + "loss": 0.1975, + "step": 9180 + }, + { + "epoch": 0.757908611599297, + "grad_norm": 2.276308536529541, + "learning_rate": 4.9602388352085337e-05, + "loss": 0.2035, + "step": 9200 + }, + { + "epoch": 0.7595562390158173, + "grad_norm": 1.4071913957595825, + "learning_rate": 4.9597263485526826e-05, + "loss": 0.1636, + "step": 9220 + }, + { + "epoch": 0.7612038664323374, + "grad_norm": 5.9329023361206055, + "learning_rate": 4.959210607075239e-05, + "loss": 0.1954, + "step": 9240 + }, + { + "epoch": 0.7628514938488576, + "grad_norm": 1.0558881759643555, + "learning_rate": 4.958691611458657e-05, + "loss": 0.2089, + "step": 9260 + }, + { + "epoch": 0.7644991212653779, + "grad_norm": 1.1962648630142212, + "learning_rate": 4.958169362389695e-05, + "loss": 0.1966, + "step": 9280 + }, + { + "epoch": 0.766146748681898, + "grad_norm": 27.050457000732422, + "learning_rate": 4.957643860559417e-05, + "loss": 0.2254, + "step": 9300 + }, + { + "epoch": 0.7677943760984183, + "grad_norm": 1.5262819528579712, + "learning_rate": 4.95711510666319e-05, + "loss": 0.1672, + "step": 9320 + }, + { + "epoch": 0.7694420035149385, + "grad_norm": 1.5432629585266113, + "learning_rate": 4.956583101400685e-05, + "loss": 0.2201, + "step": 9340 + }, + { + "epoch": 0.7710896309314587, + "grad_norm": 1.4080497026443481, + "learning_rate": 4.956047845475877e-05, + "loss": 0.2034, + "step": 9360 + }, + { + "epoch": 0.7727372583479789, + "grad_norm": 1.6339974403381348, + "learning_rate": 4.9555093395970396e-05, + "loss": 0.1918, + "step": 9380 + }, + { + "epoch": 0.7743848857644992, + "grad_norm": 3.812486410140991, + "learning_rate": 4.954967584476748e-05, + "loss": 0.2221, + "step": 9400 + }, + { + "epoch": 0.7760325131810193, + "grad_norm": 2.0563764572143555, + "learning_rate": 4.954422580831879e-05, + "loss": 0.1611, + "step": 9420 + }, + { + "epoch": 0.7776801405975395, + "grad_norm": 0.7051644325256348, + "learning_rate": 4.9538743293836046e-05, + "loss": 0.2053, + "step": 9440 + }, + { + "epoch": 0.7793277680140598, + "grad_norm": 0.837563157081604, + "learning_rate": 4.9533228308573966e-05, + "loss": 0.2078, + "step": 9460 + }, + { + "epoch": 0.7809753954305799, + "grad_norm": 1.1410398483276367, + "learning_rate": 4.952768085983023e-05, + "loss": 0.2071, + "step": 9480 + }, + { + "epoch": 0.7826230228471002, + "grad_norm": 2.372843027114868, + "learning_rate": 4.952210095494546e-05, + "loss": 0.2154, + "step": 9500 + }, + { + "epoch": 0.7842706502636204, + "grad_norm": 25.939308166503906, + "learning_rate": 4.9516488601303255e-05, + "loss": 0.1618, + "step": 9520 + }, + { + "epoch": 0.7859182776801406, + "grad_norm": 1.780004858970642, + "learning_rate": 4.951084380633013e-05, + "loss": 0.1993, + "step": 9540 + }, + { + "epoch": 0.7875659050966608, + "grad_norm": 1.1589267253875732, + "learning_rate": 4.9505166577495546e-05, + "loss": 0.1939, + "step": 9560 + }, + { + "epoch": 0.789213532513181, + "grad_norm": 1.0380936861038208, + "learning_rate": 4.949945692231185e-05, + "loss": 0.1863, + "step": 9580 + }, + { + "epoch": 0.7908611599297012, + "grad_norm": 3.860309600830078, + "learning_rate": 4.949371484833433e-05, + "loss": 0.2199, + "step": 9600 + }, + { + "epoch": 0.7925087873462214, + "grad_norm": 2.6430704593658447, + "learning_rate": 4.9487940363161155e-05, + "loss": 0.161, + "step": 9620 + }, + { + "epoch": 0.7941564147627417, + "grad_norm": 2.1502444744110107, + "learning_rate": 4.948213347443339e-05, + "loss": 0.2101, + "step": 9640 + }, + { + "epoch": 0.7958040421792618, + "grad_norm": 0.9621193408966064, + "learning_rate": 4.9476294189834974e-05, + "loss": 0.2105, + "step": 9660 + }, + { + "epoch": 0.7974516695957821, + "grad_norm": 0.8426349759101868, + "learning_rate": 4.9470422517092696e-05, + "loss": 0.1821, + "step": 9680 + }, + { + "epoch": 0.7990992970123023, + "grad_norm": 2.1861345767974854, + "learning_rate": 4.9464518463976246e-05, + "loss": 0.2157, + "step": 9700 + }, + { + "epoch": 0.8007469244288224, + "grad_norm": 1.074504017829895, + "learning_rate": 4.945858203829812e-05, + "loss": 0.1593, + "step": 9720 + }, + { + "epoch": 0.8023945518453427, + "grad_norm": 1.4774161577224731, + "learning_rate": 4.945261324791367e-05, + "loss": 0.1981, + "step": 9740 + }, + { + "epoch": 0.8040421792618629, + "grad_norm": 0.9464648962020874, + "learning_rate": 4.944661210072107e-05, + "loss": 0.2107, + "step": 9760 + }, + { + "epoch": 0.8056898066783831, + "grad_norm": 0.34157031774520874, + "learning_rate": 4.94405786046613e-05, + "loss": 0.209, + "step": 9780 + }, + { + "epoch": 0.8073374340949033, + "grad_norm": 3.6989963054656982, + "learning_rate": 4.943451276771818e-05, + "loss": 0.221, + "step": 9800 + }, + { + "epoch": 0.8089850615114236, + "grad_norm": 0.6388671398162842, + "learning_rate": 4.942841459791828e-05, + "loss": 0.1661, + "step": 9820 + }, + { + "epoch": 0.8106326889279437, + "grad_norm": 0.6647291779518127, + "learning_rate": 4.9422284103330985e-05, + "loss": 0.192, + "step": 9840 + }, + { + "epoch": 0.812280316344464, + "grad_norm": 0.540134847164154, + "learning_rate": 4.941612129206844e-05, + "loss": 0.2126, + "step": 9860 + }, + { + "epoch": 0.8139279437609842, + "grad_norm": 0.5511148571968079, + "learning_rate": 4.940992617228556e-05, + "loss": 0.2018, + "step": 9880 + }, + { + "epoch": 0.8155755711775043, + "grad_norm": 2.1634738445281982, + "learning_rate": 4.9403698752180006e-05, + "loss": 0.2087, + "step": 9900 + }, + { + "epoch": 0.8172231985940246, + "grad_norm": 1.3764668703079224, + "learning_rate": 4.939743903999218e-05, + "loss": 0.1561, + "step": 9920 + }, + { + "epoch": 0.8188708260105448, + "grad_norm": 0.8314900398254395, + "learning_rate": 4.939114704400523e-05, + "loss": 0.2226, + "step": 9940 + }, + { + "epoch": 0.820518453427065, + "grad_norm": 1.0714060068130493, + "learning_rate": 4.9384822772544994e-05, + "loss": 0.191, + "step": 9960 + }, + { + "epoch": 0.8221660808435852, + "grad_norm": 0.5301602482795715, + "learning_rate": 4.937846623398003e-05, + "loss": 0.2069, + "step": 9980 + }, + { + "epoch": 0.8238137082601055, + "grad_norm": 2.4229841232299805, + "learning_rate": 4.9372077436721634e-05, + "loss": 0.2132, + "step": 10000 + }, + { + "epoch": 0.8238137082601055, + "eval_loss": 0.5383469462394714, + "eval_runtime": 248.9799, + "eval_samples_per_second": 87.513, + "eval_steps_per_second": 21.881, + "eval_wer": 0.23184117484036582, + "step": 10000 + }, + { + "epoch": 0.8254613356766256, + "grad_norm": 1.1694583892822266, + "learning_rate": 4.936565638922372e-05, + "loss": 0.1624, + "step": 10020 + }, + { + "epoch": 0.8271089630931459, + "grad_norm": 2.124119281768799, + "learning_rate": 4.9359203099982924e-05, + "loss": 0.2062, + "step": 10040 + }, + { + "epoch": 0.8287565905096661, + "grad_norm": 0.7251669764518738, + "learning_rate": 4.935271757753852e-05, + "loss": 0.1981, + "step": 10060 + }, + { + "epoch": 0.8304042179261862, + "grad_norm": 1.2785292863845825, + "learning_rate": 4.934619983047246e-05, + "loss": 0.2094, + "step": 10080 + }, + { + "epoch": 0.8320518453427065, + "grad_norm": 3.0615155696868896, + "learning_rate": 4.933964986740931e-05, + "loss": 0.2159, + "step": 10100 + }, + { + "epoch": 0.8336994727592267, + "grad_norm": 2.058302402496338, + "learning_rate": 4.933306769701629e-05, + "loss": 0.1585, + "step": 10120 + }, + { + "epoch": 0.8353471001757469, + "grad_norm": 1.0357364416122437, + "learning_rate": 4.9326453328003217e-05, + "loss": 0.187, + "step": 10140 + }, + { + "epoch": 0.8369947275922671, + "grad_norm": 0.6068626642227173, + "learning_rate": 4.931980676912252e-05, + "loss": 0.1971, + "step": 10160 + }, + { + "epoch": 0.8386423550087874, + "grad_norm": 3.377023220062256, + "learning_rate": 4.931312802916925e-05, + "loss": 0.1999, + "step": 10180 + }, + { + "epoch": 0.8402899824253075, + "grad_norm": 3.0354530811309814, + "learning_rate": 4.9306417116980996e-05, + "loss": 0.2008, + "step": 10200 + }, + { + "epoch": 0.8419376098418277, + "grad_norm": 2.531010150909424, + "learning_rate": 4.929967404143796e-05, + "loss": 0.1688, + "step": 10220 + }, + { + "epoch": 0.843585237258348, + "grad_norm": 0.526393711566925, + "learning_rate": 4.929289881146286e-05, + "loss": 0.1844, + "step": 10240 + }, + { + "epoch": 0.8452328646748682, + "grad_norm": 0.9323037266731262, + "learning_rate": 4.9286091436021015e-05, + "loss": 0.1968, + "step": 10260 + }, + { + "epoch": 0.8468804920913884, + "grad_norm": 1.0275589227676392, + "learning_rate": 4.927925192412024e-05, + "loss": 0.182, + "step": 10280 + }, + { + "epoch": 0.8485281195079086, + "grad_norm": 4.738681793212891, + "learning_rate": 4.927238028481089e-05, + "loss": 0.214, + "step": 10300 + }, + { + "epoch": 0.8501757469244289, + "grad_norm": 1.584230899810791, + "learning_rate": 4.926547652718583e-05, + "loss": 0.1582, + "step": 10320 + }, + { + "epoch": 0.851823374340949, + "grad_norm": 0.9227916598320007, + "learning_rate": 4.9258540660380434e-05, + "loss": 0.1921, + "step": 10340 + }, + { + "epoch": 0.8534710017574693, + "grad_norm": 7.2498250007629395, + "learning_rate": 4.925157269357254e-05, + "loss": 0.2197, + "step": 10360 + }, + { + "epoch": 0.8551186291739895, + "grad_norm": 0.8496150970458984, + "learning_rate": 4.924457263598248e-05, + "loss": 0.1798, + "step": 10380 + }, + { + "epoch": 0.8567662565905096, + "grad_norm": 9.715910911560059, + "learning_rate": 4.9237540496873064e-05, + "loss": 0.2241, + "step": 10400 + }, + { + "epoch": 0.8584138840070299, + "grad_norm": 2.707458257675171, + "learning_rate": 4.923047628554952e-05, + "loss": 0.1581, + "step": 10420 + }, + { + "epoch": 0.8600615114235501, + "grad_norm": 0.9239126443862915, + "learning_rate": 4.9223380011359544e-05, + "loss": 0.1964, + "step": 10440 + }, + { + "epoch": 0.8617091388400703, + "grad_norm": 1.0885734558105469, + "learning_rate": 4.9216251683693246e-05, + "loss": 0.2139, + "step": 10460 + }, + { + "epoch": 0.8633567662565905, + "grad_norm": 0.8583139777183533, + "learning_rate": 4.920909131198315e-05, + "loss": 0.2007, + "step": 10480 + }, + { + "epoch": 0.8650043936731108, + "grad_norm": 2.6485893726348877, + "learning_rate": 4.920189890570419e-05, + "loss": 0.2065, + "step": 10500 + }, + { + "epoch": 0.8666520210896309, + "grad_norm": 1.3636205196380615, + "learning_rate": 4.919467447437368e-05, + "loss": 0.1583, + "step": 10520 + }, + { + "epoch": 0.8682996485061512, + "grad_norm": 1.2751201391220093, + "learning_rate": 4.918741802755132e-05, + "loss": 0.2044, + "step": 10540 + }, + { + "epoch": 0.8699472759226714, + "grad_norm": 0.8484062552452087, + "learning_rate": 4.918012957483916e-05, + "loss": 0.192, + "step": 10560 + }, + { + "epoch": 0.8715949033391915, + "grad_norm": 0.5631062984466553, + "learning_rate": 4.917280912588163e-05, + "loss": 0.1997, + "step": 10580 + }, + { + "epoch": 0.8732425307557118, + "grad_norm": 2.789642572402954, + "learning_rate": 4.916545669036545e-05, + "loss": 0.2309, + "step": 10600 + }, + { + "epoch": 0.874890158172232, + "grad_norm": 7.149659156799316, + "learning_rate": 4.915807227801973e-05, + "loss": 0.1559, + "step": 10620 + }, + { + "epoch": 0.8765377855887522, + "grad_norm": 1.080868124961853, + "learning_rate": 4.915065589861584e-05, + "loss": 0.2, + "step": 10640 + }, + { + "epoch": 0.8781854130052724, + "grad_norm": 0.49652764201164246, + "learning_rate": 4.914320756196748e-05, + "loss": 0.2034, + "step": 10660 + }, + { + "epoch": 0.8798330404217927, + "grad_norm": 0.9888033270835876, + "learning_rate": 4.913572727793062e-05, + "loss": 0.1952, + "step": 10680 + }, + { + "epoch": 0.8814806678383128, + "grad_norm": 13.734339714050293, + "learning_rate": 4.9128215056403507e-05, + "loss": 0.2042, + "step": 10700 + }, + { + "epoch": 0.883128295254833, + "grad_norm": 0.7586619257926941, + "learning_rate": 4.912067090732667e-05, + "loss": 0.1449, + "step": 10720 + }, + { + "epoch": 0.8847759226713533, + "grad_norm": 0.673252284526825, + "learning_rate": 4.911309484068285e-05, + "loss": 0.2138, + "step": 10740 + }, + { + "epoch": 0.8864235500878734, + "grad_norm": 1.2692677974700928, + "learning_rate": 4.910548686649706e-05, + "loss": 0.1994, + "step": 10760 + }, + { + "epoch": 0.8880711775043937, + "grad_norm": 0.7501096725463867, + "learning_rate": 4.9097846994836505e-05, + "loss": 0.1949, + "step": 10780 + }, + { + "epoch": 0.8897188049209139, + "grad_norm": 3.0218920707702637, + "learning_rate": 4.909017523581062e-05, + "loss": 0.2137, + "step": 10800 + }, + { + "epoch": 0.8913664323374341, + "grad_norm": 1.1619638204574585, + "learning_rate": 4.9082471599571015e-05, + "loss": 0.1509, + "step": 10820 + }, + { + "epoch": 0.8930140597539543, + "grad_norm": 1.3160550594329834, + "learning_rate": 4.907512362815835e-05, + "loss": 0.1879, + "step": 10840 + }, + { + "epoch": 0.8946616871704746, + "grad_norm": 2.0613820552825928, + "learning_rate": 4.9067357860710327e-05, + "loss": 0.2094, + "step": 10860 + }, + { + "epoch": 0.8963093145869947, + "grad_norm": 0.6547063589096069, + "learning_rate": 4.905956024624158e-05, + "loss": 0.1879, + "step": 10880 + }, + { + "epoch": 0.897956942003515, + "grad_norm": 4.540472507476807, + "learning_rate": 4.905173079507026e-05, + "loss": 0.2022, + "step": 10900 + }, + { + "epoch": 0.8996045694200352, + "grad_norm": 2.7642438411712646, + "learning_rate": 4.904386951755665e-05, + "loss": 0.1566, + "step": 10920 + }, + { + "epoch": 0.9012521968365553, + "grad_norm": 1.2250425815582275, + "learning_rate": 4.903597642410316e-05, + "loss": 0.2049, + "step": 10940 + }, + { + "epoch": 0.9028998242530756, + "grad_norm": 0.6964439749717712, + "learning_rate": 4.902805152515427e-05, + "loss": 0.2003, + "step": 10960 + }, + { + "epoch": 0.9045474516695958, + "grad_norm": 0.8881003260612488, + "learning_rate": 4.902049342086357e-05, + "loss": 0.1969, + "step": 10980 + }, + { + "epoch": 0.906195079086116, + "grad_norm": 3.828152656555176, + "learning_rate": 4.901250653139905e-05, + "loss": 0.2199, + "step": 11000 + }, + { + "epoch": 0.9078427065026362, + "grad_norm": 1.2089974880218506, + "learning_rate": 4.900448786749557e-05, + "loss": 0.157, + "step": 11020 + }, + { + "epoch": 0.9094903339191565, + "grad_norm": 1.148004412651062, + "learning_rate": 4.8996437439763784e-05, + "loss": 0.2013, + "step": 11040 + }, + { + "epoch": 0.9111379613356766, + "grad_norm": 1.6459652185440063, + "learning_rate": 4.89883552588564e-05, + "loss": 0.1928, + "step": 11060 + }, + { + "epoch": 0.9127855887521968, + "grad_norm": 0.7444145083427429, + "learning_rate": 4.898024133546811e-05, + "loss": 0.1965, + "step": 11080 + }, + { + "epoch": 0.9144332161687171, + "grad_norm": 7.0522260665893555, + "learning_rate": 4.897209568033564e-05, + "loss": 0.2181, + "step": 11100 + }, + { + "epoch": 0.9160808435852372, + "grad_norm": 1.5349088907241821, + "learning_rate": 4.896391830423768e-05, + "loss": 0.1575, + "step": 11120 + }, + { + "epoch": 0.9177284710017575, + "grad_norm": 0.5103209018707275, + "learning_rate": 4.895570921799491e-05, + "loss": 0.2057, + "step": 11140 + }, + { + "epoch": 0.9193760984182777, + "grad_norm": 0.4984442889690399, + "learning_rate": 4.8947468432469955e-05, + "loss": 0.1897, + "step": 11160 + }, + { + "epoch": 0.9210237258347979, + "grad_norm": 0.5621454119682312, + "learning_rate": 4.893919595856742e-05, + "loss": 0.1912, + "step": 11180 + }, + { + "epoch": 0.9226713532513181, + "grad_norm": 5.476040363311768, + "learning_rate": 4.8930891807233794e-05, + "loss": 0.2166, + "step": 11200 + }, + { + "epoch": 0.9243189806678384, + "grad_norm": 1.3259927034378052, + "learning_rate": 4.892255598945753e-05, + "loss": 0.1561, + "step": 11220 + }, + { + "epoch": 0.9259666080843585, + "grad_norm": 0.8870605230331421, + "learning_rate": 4.891418851626893e-05, + "loss": 0.2003, + "step": 11240 + }, + { + "epoch": 0.9276142355008787, + "grad_norm": 16.381240844726562, + "learning_rate": 4.890578939874025e-05, + "loss": 0.1916, + "step": 11260 + }, + { + "epoch": 0.929261862917399, + "grad_norm": 1.6785776615142822, + "learning_rate": 4.889735864798556e-05, + "loss": 0.1825, + "step": 11280 + }, + { + "epoch": 0.9309094903339191, + "grad_norm": 5.951013565063477, + "learning_rate": 4.8888896275160816e-05, + "loss": 0.2045, + "step": 11300 + }, + { + "epoch": 0.9325571177504394, + "grad_norm": 3.8519906997680664, + "learning_rate": 4.888040229146382e-05, + "loss": 0.1573, + "step": 11320 + }, + { + "epoch": 0.9342047451669596, + "grad_norm": 1.0654627084732056, + "learning_rate": 4.887187670813419e-05, + "loss": 0.223, + "step": 11340 + }, + { + "epoch": 0.9358523725834798, + "grad_norm": 1.6851961612701416, + "learning_rate": 4.8863319536453364e-05, + "loss": 0.2058, + "step": 11360 + }, + { + "epoch": 0.9375, + "grad_norm": 0.6880635023117065, + "learning_rate": 4.88547307877446e-05, + "loss": 0.1974, + "step": 11380 + }, + { + "epoch": 0.9391476274165202, + "grad_norm": 5.7095866203308105, + "learning_rate": 4.884611047337289e-05, + "loss": 0.2169, + "step": 11400 + }, + { + "epoch": 0.9407952548330404, + "grad_norm": 8.853272438049316, + "learning_rate": 4.883745860474505e-05, + "loss": 0.1584, + "step": 11420 + }, + { + "epoch": 0.9424428822495606, + "grad_norm": 1.7952899932861328, + "learning_rate": 4.882877519330961e-05, + "loss": 0.1937, + "step": 11440 + }, + { + "epoch": 0.9440905096660809, + "grad_norm": 5.514512538909912, + "learning_rate": 4.882006025055685e-05, + "loss": 0.1989, + "step": 11460 + }, + { + "epoch": 0.945738137082601, + "grad_norm": 0.776798665523529, + "learning_rate": 4.88113137880188e-05, + "loss": 0.1862, + "step": 11480 + }, + { + "epoch": 0.9473857644991213, + "grad_norm": 2.824843645095825, + "learning_rate": 4.880253581726916e-05, + "loss": 0.2168, + "step": 11500 + }, + { + "epoch": 0.9490333919156415, + "grad_norm": 0.9011817574501038, + "learning_rate": 4.879372634992335e-05, + "loss": 0.1601, + "step": 11520 + }, + { + "epoch": 0.9506810193321616, + "grad_norm": 5.290555477142334, + "learning_rate": 4.8784885397638445e-05, + "loss": 0.1892, + "step": 11540 + }, + { + "epoch": 0.9523286467486819, + "grad_norm": 1.164330005645752, + "learning_rate": 4.87760129721132e-05, + "loss": 0.2112, + "step": 11560 + }, + { + "epoch": 0.9539762741652021, + "grad_norm": 0.8354160189628601, + "learning_rate": 4.876710908508801e-05, + "loss": 0.2176, + "step": 11580 + }, + { + "epoch": 0.9556239015817223, + "grad_norm": 3.980334758758545, + "learning_rate": 4.8758173748344904e-05, + "loss": 0.2484, + "step": 11600 + }, + { + "epoch": 0.9572715289982425, + "grad_norm": 1.7610282897949219, + "learning_rate": 4.874920697370753e-05, + "loss": 0.1618, + "step": 11620 + }, + { + "epoch": 0.9589191564147628, + "grad_norm": 1.3336617946624756, + "learning_rate": 4.874020877304113e-05, + "loss": 0.1895, + "step": 11640 + }, + { + "epoch": 0.9605667838312829, + "grad_norm": 0.9175180792808533, + "learning_rate": 4.873117915825252e-05, + "loss": 0.1945, + "step": 11660 + }, + { + "epoch": 0.9622144112478032, + "grad_norm": 0.49205484986305237, + "learning_rate": 4.8722118141290105e-05, + "loss": 0.1935, + "step": 11680 + }, + { + "epoch": 0.9638620386643234, + "grad_norm": 3.554715394973755, + "learning_rate": 4.871302573414384e-05, + "loss": 0.2135, + "step": 11700 + }, + { + "epoch": 0.9655096660808435, + "grad_norm": 1.0651150941848755, + "learning_rate": 4.8703901948845205e-05, + "loss": 0.1564, + "step": 11720 + }, + { + "epoch": 0.9671572934973638, + "grad_norm": 1.88383150100708, + "learning_rate": 4.869474679746721e-05, + "loss": 0.1994, + "step": 11740 + }, + { + "epoch": 0.968804920913884, + "grad_norm": 0.6271897554397583, + "learning_rate": 4.868556029212435e-05, + "loss": 0.2043, + "step": 11760 + }, + { + "epoch": 0.9704525483304042, + "grad_norm": 0.616521954536438, + "learning_rate": 4.867634244497265e-05, + "loss": 0.2059, + "step": 11780 + }, + { + "epoch": 0.9721001757469244, + "grad_norm": 3.2947463989257812, + "learning_rate": 4.8667093268209575e-05, + "loss": 0.205, + "step": 11800 + }, + { + "epoch": 0.9737478031634447, + "grad_norm": 0.7675888538360596, + "learning_rate": 4.865781277407405e-05, + "loss": 0.1436, + "step": 11820 + }, + { + "epoch": 0.9753954305799648, + "grad_norm": 1.3581780195236206, + "learning_rate": 4.8648500974846445e-05, + "loss": 0.1964, + "step": 11840 + }, + { + "epoch": 0.977043057996485, + "grad_norm": 0.7175058722496033, + "learning_rate": 4.8639157882848566e-05, + "loss": 0.197, + "step": 11860 + }, + { + "epoch": 0.9786906854130053, + "grad_norm": 0.7383044362068176, + "learning_rate": 4.8629783510443616e-05, + "loss": 0.1904, + "step": 11880 + }, + { + "epoch": 0.9803383128295254, + "grad_norm": 5.0527777671813965, + "learning_rate": 4.8620377870036184e-05, + "loss": 0.2117, + "step": 11900 + }, + { + "epoch": 0.9819859402460457, + "grad_norm": 1.5865782499313354, + "learning_rate": 4.861094097407224e-05, + "loss": 0.1515, + "step": 11920 + }, + { + "epoch": 0.9836335676625659, + "grad_norm": 0.5650818943977356, + "learning_rate": 4.860147283503912e-05, + "loss": 0.1926, + "step": 11940 + }, + { + "epoch": 0.9852811950790861, + "grad_norm": 1.3663684129714966, + "learning_rate": 4.8591973465465493e-05, + "loss": 0.2066, + "step": 11960 + }, + { + "epoch": 0.9869288224956063, + "grad_norm": 1.2778149843215942, + "learning_rate": 4.8582442877921344e-05, + "loss": 0.2022, + "step": 11980 + }, + { + "epoch": 0.9885764499121266, + "grad_norm": 4.921967029571533, + "learning_rate": 4.8572881085017996e-05, + "loss": 0.2158, + "step": 12000 + }, + { + "epoch": 0.9885764499121266, + "eval_loss": 0.5375325679779053, + "eval_runtime": 252.7679, + "eval_samples_per_second": 86.202, + "eval_steps_per_second": 21.553, + "eval_wer": 0.2286638878235724, + "step": 12000 + }, + { + "epoch": 0.9902240773286467, + "grad_norm": 1.044043779373169, + "learning_rate": 4.856328809940804e-05, + "loss": 0.1584, + "step": 12020 + }, + { + "epoch": 0.991871704745167, + "grad_norm": 2.4731016159057617, + "learning_rate": 4.855366393378535e-05, + "loss": 0.2071, + "step": 12040 + }, + { + "epoch": 0.9935193321616872, + "grad_norm": 0.6640995144844055, + "learning_rate": 4.854400860088506e-05, + "loss": 0.2119, + "step": 12060 + }, + { + "epoch": 0.9951669595782073, + "grad_norm": 0.43121984601020813, + "learning_rate": 4.853432211348356e-05, + "loss": 0.1918, + "step": 12080 + }, + { + "epoch": 0.9968145869947276, + "grad_norm": 5.402212619781494, + "learning_rate": 4.8524604484398414e-05, + "loss": 0.2181, + "step": 12100 + }, + { + "epoch": 0.9984622144112478, + "grad_norm": 1.417004108428955, + "learning_rate": 4.851485572648846e-05, + "loss": 0.1493, + "step": 12120 + }, + { + "epoch": 1.000109841827768, + "grad_norm": 0.9582122564315796, + "learning_rate": 4.8505075852653684e-05, + "loss": 0.207, + "step": 12140 + }, + { + "epoch": 1.0017574692442883, + "grad_norm": 1.74813973903656, + "learning_rate": 4.8495264875835264e-05, + "loss": 0.1468, + "step": 12160 + }, + { + "epoch": 1.0034050966608083, + "grad_norm": 3.2926461696624756, + "learning_rate": 4.8485422809015535e-05, + "loss": 0.2009, + "step": 12180 + }, + { + "epoch": 1.0050527240773286, + "grad_norm": 2.141998291015625, + "learning_rate": 4.8475549665217956e-05, + "loss": 0.1878, + "step": 12200 + }, + { + "epoch": 1.0067003514938488, + "grad_norm": 1.7473055124282837, + "learning_rate": 4.846564545750712e-05, + "loss": 0.1916, + "step": 12220 + }, + { + "epoch": 1.008347978910369, + "grad_norm": 0.20854823291301727, + "learning_rate": 4.845571019898874e-05, + "loss": 0.19, + "step": 12240 + }, + { + "epoch": 1.0099956063268893, + "grad_norm": 1.6081265211105347, + "learning_rate": 4.844574390280958e-05, + "loss": 0.1623, + "step": 12260 + }, + { + "epoch": 1.0116432337434096, + "grad_norm": 2.013803482055664, + "learning_rate": 4.843574658215751e-05, + "loss": 0.1916, + "step": 12280 + }, + { + "epoch": 1.0132908611599296, + "grad_norm": 1.2559956312179565, + "learning_rate": 4.842571825026143e-05, + "loss": 0.1759, + "step": 12300 + }, + { + "epoch": 1.0149384885764499, + "grad_norm": 1.103947401046753, + "learning_rate": 4.8415658920391296e-05, + "loss": 0.202, + "step": 12320 + }, + { + "epoch": 1.01658611599297, + "grad_norm": 0.5012603998184204, + "learning_rate": 4.840556860585805e-05, + "loss": 0.198, + "step": 12340 + }, + { + "epoch": 1.0182337434094904, + "grad_norm": 1.4309896230697632, + "learning_rate": 4.839544732001368e-05, + "loss": 0.163, + "step": 12360 + }, + { + "epoch": 1.0198813708260106, + "grad_norm": 0.6756483912467957, + "learning_rate": 4.838529507625111e-05, + "loss": 0.1831, + "step": 12380 + }, + { + "epoch": 1.0215289982425309, + "grad_norm": 1.141714096069336, + "learning_rate": 4.8375111888004263e-05, + "loss": 0.1888, + "step": 12400 + }, + { + "epoch": 1.0231766256590509, + "grad_norm": 0.471711665391922, + "learning_rate": 4.836489776874799e-05, + "loss": 0.1892, + "step": 12420 + }, + { + "epoch": 1.0248242530755711, + "grad_norm": 0.28789934515953064, + "learning_rate": 4.835465273199807e-05, + "loss": 0.1795, + "step": 12440 + }, + { + "epoch": 1.0264718804920914, + "grad_norm": 1.1327400207519531, + "learning_rate": 4.834437679131121e-05, + "loss": 0.169, + "step": 12460 + }, + { + "epoch": 1.0281195079086116, + "grad_norm": 1.0975223779678345, + "learning_rate": 4.8334069960285e-05, + "loss": 0.1875, + "step": 12480 + }, + { + "epoch": 1.0297671353251319, + "grad_norm": 0.5748992562294006, + "learning_rate": 4.832373225255791e-05, + "loss": 0.1888, + "step": 12500 + }, + { + "epoch": 1.0314147627416521, + "grad_norm": 0.5393682718276978, + "learning_rate": 4.831336368180927e-05, + "loss": 0.1808, + "step": 12520 + }, + { + "epoch": 1.0330623901581721, + "grad_norm": 0.22509616613388062, + "learning_rate": 4.830296426175922e-05, + "loss": 0.1803, + "step": 12540 + }, + { + "epoch": 1.0347100175746924, + "grad_norm": 1.0403659343719482, + "learning_rate": 4.829253400616876e-05, + "loss": 0.1594, + "step": 12560 + }, + { + "epoch": 1.0363576449912126, + "grad_norm": 1.5973280668258667, + "learning_rate": 4.828207292883968e-05, + "loss": 0.1907, + "step": 12580 + }, + { + "epoch": 1.0380052724077329, + "grad_norm": 0.9520076513290405, + "learning_rate": 4.8271581043614555e-05, + "loss": 0.1852, + "step": 12600 + }, + { + "epoch": 1.0396528998242531, + "grad_norm": 3.2169032096862793, + "learning_rate": 4.826105836437672e-05, + "loss": 0.1796, + "step": 12620 + }, + { + "epoch": 1.0413005272407734, + "grad_norm": 0.3701666295528412, + "learning_rate": 4.825050490505025e-05, + "loss": 0.1894, + "step": 12640 + }, + { + "epoch": 1.0429481546572934, + "grad_norm": 1.573367953300476, + "learning_rate": 4.823992067959998e-05, + "loss": 0.1565, + "step": 12660 + }, + { + "epoch": 1.0445957820738137, + "grad_norm": 0.6586406230926514, + "learning_rate": 4.8229305702031426e-05, + "loss": 0.183, + "step": 12680 + }, + { + "epoch": 1.046243409490334, + "grad_norm": 1.1278094053268433, + "learning_rate": 4.821865998639081e-05, + "loss": 0.175, + "step": 12700 + }, + { + "epoch": 1.0478910369068541, + "grad_norm": 0.9823224544525146, + "learning_rate": 4.820798354676504e-05, + "loss": 0.1939, + "step": 12720 + }, + { + "epoch": 1.0495386643233744, + "grad_norm": 0.6462484002113342, + "learning_rate": 4.819727639728164e-05, + "loss": 0.1652, + "step": 12740 + }, + { + "epoch": 1.0511862917398946, + "grad_norm": 1.3370243310928345, + "learning_rate": 4.818653855210882e-05, + "loss": 0.1605, + "step": 12760 + }, + { + "epoch": 1.0528339191564147, + "grad_norm": 0.7861006259918213, + "learning_rate": 4.817577002545538e-05, + "loss": 0.1858, + "step": 12780 + }, + { + "epoch": 1.054481546572935, + "grad_norm": 2.107189655303955, + "learning_rate": 4.816497083157071e-05, + "loss": 0.1933, + "step": 12800 + }, + { + "epoch": 1.0561291739894552, + "grad_norm": 0.9943327307701111, + "learning_rate": 4.81541409847448e-05, + "loss": 0.1949, + "step": 12820 + }, + { + "epoch": 1.0577768014059754, + "grad_norm": 0.5097872018814087, + "learning_rate": 4.8143280499308205e-05, + "loss": 0.1879, + "step": 12840 + }, + { + "epoch": 1.0594244288224957, + "grad_norm": 2.121750593185425, + "learning_rate": 4.8132389389631995e-05, + "loss": 0.1616, + "step": 12860 + }, + { + "epoch": 1.061072056239016, + "grad_norm": 0.86652010679245, + "learning_rate": 4.81214676701278e-05, + "loss": 0.2113, + "step": 12880 + }, + { + "epoch": 1.062719683655536, + "grad_norm": 2.3862969875335693, + "learning_rate": 4.811051535524772e-05, + "loss": 0.1904, + "step": 12900 + }, + { + "epoch": 1.0643673110720562, + "grad_norm": 0.5471189618110657, + "learning_rate": 4.8099532459484356e-05, + "loss": 0.182, + "step": 12920 + }, + { + "epoch": 1.0660149384885764, + "grad_norm": 1.5992333889007568, + "learning_rate": 4.8088518997370794e-05, + "loss": 0.1821, + "step": 12940 + }, + { + "epoch": 1.0676625659050967, + "grad_norm": 0.6893438696861267, + "learning_rate": 4.807747498348052e-05, + "loss": 0.1648, + "step": 12960 + }, + { + "epoch": 1.069310193321617, + "grad_norm": 4.395170211791992, + "learning_rate": 4.806640043242748e-05, + "loss": 0.1759, + "step": 12980 + }, + { + "epoch": 1.0709578207381372, + "grad_norm": 0.5166418552398682, + "learning_rate": 4.805529535886605e-05, + "loss": 0.1956, + "step": 13000 + }, + { + "epoch": 1.0726054481546572, + "grad_norm": 0.8624104261398315, + "learning_rate": 4.804415977749094e-05, + "loss": 0.1839, + "step": 13020 + }, + { + "epoch": 1.0742530755711774, + "grad_norm": 0.6681638360023499, + "learning_rate": 4.803299370303725e-05, + "loss": 0.2012, + "step": 13040 + }, + { + "epoch": 1.0759007029876977, + "grad_norm": 1.1625574827194214, + "learning_rate": 4.802179715028047e-05, + "loss": 0.165, + "step": 13060 + }, + { + "epoch": 1.077548330404218, + "grad_norm": 3.3364834785461426, + "learning_rate": 4.801057013403636e-05, + "loss": 0.1905, + "step": 13080 + }, + { + "epoch": 1.0791959578207382, + "grad_norm": 0.9237430691719055, + "learning_rate": 4.799931266916103e-05, + "loss": 0.1923, + "step": 13100 + }, + { + "epoch": 1.0808435852372584, + "grad_norm": 2.341661214828491, + "learning_rate": 4.798858988805215e-05, + "loss": 0.1918, + "step": 13120 + }, + { + "epoch": 1.0824912126537785, + "grad_norm": 0.2139461189508438, + "learning_rate": 4.797727309122835e-05, + "loss": 0.2039, + "step": 13140 + }, + { + "epoch": 1.0841388400702987, + "grad_norm": 1.3084605932235718, + "learning_rate": 4.796592588983348e-05, + "loss": 0.1808, + "step": 13160 + }, + { + "epoch": 1.085786467486819, + "grad_norm": 0.5879707336425781, + "learning_rate": 4.7954548298882685e-05, + "loss": 0.1879, + "step": 13180 + }, + { + "epoch": 1.0874340949033392, + "grad_norm": 9.557912826538086, + "learning_rate": 4.794314033343129e-05, + "loss": 0.1667, + "step": 13200 + }, + { + "epoch": 1.0890817223198594, + "grad_norm": 2.7290220260620117, + "learning_rate": 4.793170200857485e-05, + "loss": 0.1848, + "step": 13220 + }, + { + "epoch": 1.0907293497363797, + "grad_norm": 0.4870263338088989, + "learning_rate": 4.792023333944907e-05, + "loss": 0.183, + "step": 13240 + }, + { + "epoch": 1.0923769771528997, + "grad_norm": 1.1112992763519287, + "learning_rate": 4.790873434122982e-05, + "loss": 0.1628, + "step": 13260 + }, + { + "epoch": 1.09402460456942, + "grad_norm": 1.2795594930648804, + "learning_rate": 4.789720502913309e-05, + "loss": 0.1687, + "step": 13280 + }, + { + "epoch": 1.0956722319859402, + "grad_norm": 1.0361709594726562, + "learning_rate": 4.788564541841498e-05, + "loss": 0.196, + "step": 13300 + }, + { + "epoch": 1.0973198594024605, + "grad_norm": 1.4173661470413208, + "learning_rate": 4.787405552437171e-05, + "loss": 0.191, + "step": 13320 + }, + { + "epoch": 1.0989674868189807, + "grad_norm": 3.3120641708374023, + "learning_rate": 4.786243536233954e-05, + "loss": 0.1881, + "step": 13340 + }, + { + "epoch": 1.100615114235501, + "grad_norm": 1.3123334646224976, + "learning_rate": 4.785078494769481e-05, + "loss": 0.1701, + "step": 13360 + }, + { + "epoch": 1.1022627416520212, + "grad_norm": 1.0791329145431519, + "learning_rate": 4.783910429585386e-05, + "loss": 0.1883, + "step": 13380 + }, + { + "epoch": 1.1039103690685412, + "grad_norm": 4.972422122955322, + "learning_rate": 4.782739342227308e-05, + "loss": 0.1977, + "step": 13400 + }, + { + "epoch": 1.1055579964850615, + "grad_norm": 1.7628355026245117, + "learning_rate": 4.781565234244881e-05, + "loss": 0.1828, + "step": 13420 + }, + { + "epoch": 1.1072056239015817, + "grad_norm": 0.7487472295761108, + "learning_rate": 4.7803881071917386e-05, + "loss": 0.1903, + "step": 13440 + }, + { + "epoch": 1.108853251318102, + "grad_norm": 1.0923112630844116, + "learning_rate": 4.77920796262551e-05, + "loss": 0.1597, + "step": 13460 + }, + { + "epoch": 1.1105008787346222, + "grad_norm": 2.634671211242676, + "learning_rate": 4.778024802107814e-05, + "loss": 0.201, + "step": 13480 + }, + { + "epoch": 1.1121485061511422, + "grad_norm": 0.6952950358390808, + "learning_rate": 4.7768386272042646e-05, + "loss": 0.1887, + "step": 13500 + }, + { + "epoch": 1.1137961335676625, + "grad_norm": 1.2554984092712402, + "learning_rate": 4.77564943948446e-05, + "loss": 0.1862, + "step": 13520 + }, + { + "epoch": 1.1154437609841827, + "grad_norm": 0.42830732464790344, + "learning_rate": 4.774457240521989e-05, + "loss": 0.1892, + "step": 13540 + }, + { + "epoch": 1.117091388400703, + "grad_norm": 1.1879973411560059, + "learning_rate": 4.7732620318944224e-05, + "loss": 0.1576, + "step": 13560 + }, + { + "epoch": 1.1187390158172232, + "grad_norm": 0.9063071012496948, + "learning_rate": 4.772063815183315e-05, + "loss": 0.1941, + "step": 13580 + }, + { + "epoch": 1.1203866432337435, + "grad_norm": 1.071677803993225, + "learning_rate": 4.770862591974202e-05, + "loss": 0.1779, + "step": 13600 + }, + { + "epoch": 1.1220342706502637, + "grad_norm": 1.2179434299468994, + "learning_rate": 4.769658363856595e-05, + "loss": 0.1838, + "step": 13620 + }, + { + "epoch": 1.1236818980667838, + "grad_norm": 0.8977575898170471, + "learning_rate": 4.768451132423985e-05, + "loss": 0.2017, + "step": 13640 + }, + { + "epoch": 1.125329525483304, + "grad_norm": 1.6982157230377197, + "learning_rate": 4.767240899273835e-05, + "loss": 0.1608, + "step": 13660 + }, + { + "epoch": 1.1269771528998243, + "grad_norm": 0.7634503245353699, + "learning_rate": 4.7660276660075804e-05, + "loss": 0.1718, + "step": 13680 + }, + { + "epoch": 1.1286247803163445, + "grad_norm": 0.8203310966491699, + "learning_rate": 4.764811434230626e-05, + "loss": 0.1831, + "step": 13700 + }, + { + "epoch": 1.1302724077328647, + "grad_norm": 0.9817870259284973, + "learning_rate": 4.763592205552345e-05, + "loss": 0.1858, + "step": 13720 + }, + { + "epoch": 1.1319200351493848, + "grad_norm": 0.333933562040329, + "learning_rate": 4.762369981586077e-05, + "loss": 0.1785, + "step": 13740 + }, + { + "epoch": 1.133567662565905, + "grad_norm": 1.1885998249053955, + "learning_rate": 4.761144763949124e-05, + "loss": 0.1508, + "step": 13760 + }, + { + "epoch": 1.1352152899824253, + "grad_norm": 4.4533820152282715, + "learning_rate": 4.759916554262749e-05, + "loss": 0.1855, + "step": 13780 + }, + { + "epoch": 1.1368629173989455, + "grad_norm": 0.6248905062675476, + "learning_rate": 4.7586853541521744e-05, + "loss": 0.1938, + "step": 13800 + }, + { + "epoch": 1.1385105448154658, + "grad_norm": 0.7727264761924744, + "learning_rate": 4.757451165246583e-05, + "loss": 0.1809, + "step": 13820 + }, + { + "epoch": 1.140158172231986, + "grad_norm": 0.3115682005882263, + "learning_rate": 4.7562139891791046e-05, + "loss": 0.188, + "step": 13840 + }, + { + "epoch": 1.1418057996485063, + "grad_norm": 3.9221365451812744, + "learning_rate": 4.754973827586832e-05, + "loss": 0.1556, + "step": 13860 + }, + { + "epoch": 1.1434534270650263, + "grad_norm": 0.8228521347045898, + "learning_rate": 4.7537306821108e-05, + "loss": 0.185, + "step": 13880 + }, + { + "epoch": 1.1451010544815465, + "grad_norm": 3.115772247314453, + "learning_rate": 4.752484554395995e-05, + "loss": 0.1786, + "step": 13900 + }, + { + "epoch": 1.1467486818980668, + "grad_norm": 0.7074133157730103, + "learning_rate": 4.7512354460913524e-05, + "loss": 0.175, + "step": 13920 + }, + { + "epoch": 1.148396309314587, + "grad_norm": 0.4195045828819275, + "learning_rate": 4.7499833588497464e-05, + "loss": 0.1868, + "step": 13940 + }, + { + "epoch": 1.1500439367311073, + "grad_norm": 1.2703948020935059, + "learning_rate": 4.7487282943279965e-05, + "loss": 0.1517, + "step": 13960 + }, + { + "epoch": 1.1516915641476273, + "grad_norm": 2.4614219665527344, + "learning_rate": 4.747470254186862e-05, + "loss": 0.1926, + "step": 13980 + }, + { + "epoch": 1.1533391915641475, + "grad_norm": 2.0017571449279785, + "learning_rate": 4.746209240091038e-05, + "loss": 0.1852, + "step": 14000 + }, + { + "epoch": 1.1533391915641475, + "eval_loss": 0.5445581078529358, + "eval_runtime": 260.7987, + "eval_samples_per_second": 83.547, + "eval_steps_per_second": 20.89, + "eval_wer": 0.22884024824829308, + "step": 14000 + }, + { + "epoch": 1.1550417398945518, + "grad_norm": 0.7891082763671875, + "learning_rate": 4.744945253709156e-05, + "loss": 0.1683, + "step": 14020 + }, + { + "epoch": 1.156689367311072, + "grad_norm": 1.2621268033981323, + "learning_rate": 4.743678296713779e-05, + "loss": 0.2037, + "step": 14040 + }, + { + "epoch": 1.1583369947275923, + "grad_norm": 1.7061976194381714, + "learning_rate": 4.7424083707814035e-05, + "loss": 0.18, + "step": 14060 + }, + { + "epoch": 1.1599846221441124, + "grad_norm": 4.809337615966797, + "learning_rate": 4.7411354775924523e-05, + "loss": 0.1772, + "step": 14080 + }, + { + "epoch": 1.1616322495606326, + "grad_norm": 4.00537109375, + "learning_rate": 4.739859618831276e-05, + "loss": 0.2029, + "step": 14100 + }, + { + "epoch": 1.1632798769771528, + "grad_norm": 3.184988498687744, + "learning_rate": 4.7385807961861486e-05, + "loss": 0.1857, + "step": 14120 + }, + { + "epoch": 1.164927504393673, + "grad_norm": 0.24155808985233307, + "learning_rate": 4.737299011349265e-05, + "loss": 0.1808, + "step": 14140 + }, + { + "epoch": 1.1665751318101933, + "grad_norm": 1.2904281616210938, + "learning_rate": 4.736014266016743e-05, + "loss": 0.1728, + "step": 14160 + }, + { + "epoch": 1.1682227592267136, + "grad_norm": 0.8972460031509399, + "learning_rate": 4.734726561888616e-05, + "loss": 0.1712, + "step": 14180 + }, + { + "epoch": 1.1698703866432338, + "grad_norm": 0.48299163579940796, + "learning_rate": 4.7334359006688314e-05, + "loss": 0.1872, + "step": 14200 + }, + { + "epoch": 1.1715180140597539, + "grad_norm": 1.942762017250061, + "learning_rate": 4.732142284065251e-05, + "loss": 0.1742, + "step": 14220 + }, + { + "epoch": 1.1731656414762741, + "grad_norm": 0.8175634741783142, + "learning_rate": 4.730845713789647e-05, + "loss": 0.175, + "step": 14240 + }, + { + "epoch": 1.1748132688927944, + "grad_norm": 3.179103136062622, + "learning_rate": 4.7295461915577e-05, + "loss": 0.1544, + "step": 14260 + }, + { + "epoch": 1.1764608963093146, + "grad_norm": 0.7264633178710938, + "learning_rate": 4.7282437190889975e-05, + "loss": 0.1854, + "step": 14280 + }, + { + "epoch": 1.1781085237258349, + "grad_norm": 1.372662901878357, + "learning_rate": 4.7269382981070296e-05, + "loss": 0.193, + "step": 14300 + }, + { + "epoch": 1.179756151142355, + "grad_norm": 1.8890334367752075, + "learning_rate": 4.7256299303391895e-05, + "loss": 0.1908, + "step": 14320 + }, + { + "epoch": 1.1814037785588751, + "grad_norm": 0.4149373769760132, + "learning_rate": 4.724318617516768e-05, + "loss": 0.1838, + "step": 14340 + }, + { + "epoch": 1.1830514059753954, + "grad_norm": 2.029543876647949, + "learning_rate": 4.723004361374953e-05, + "loss": 0.1656, + "step": 14360 + }, + { + "epoch": 1.1846990333919156, + "grad_norm": 10.552154541015625, + "learning_rate": 4.721687163652829e-05, + "loss": 0.1828, + "step": 14380 + }, + { + "epoch": 1.1863466608084359, + "grad_norm": 0.5331339240074158, + "learning_rate": 4.7203670260933725e-05, + "loss": 0.1807, + "step": 14400 + }, + { + "epoch": 1.1879942882249561, + "grad_norm": 2.227116584777832, + "learning_rate": 4.719043950443448e-05, + "loss": 0.1903, + "step": 14420 + }, + { + "epoch": 1.1896419156414764, + "grad_norm": 0.36228641867637634, + "learning_rate": 4.717717938453811e-05, + "loss": 0.1775, + "step": 14440 + }, + { + "epoch": 1.1912895430579964, + "grad_norm": 2.8174803256988525, + "learning_rate": 4.7163889918790984e-05, + "loss": 0.1672, + "step": 14460 + }, + { + "epoch": 1.1929371704745166, + "grad_norm": 1.1632130146026611, + "learning_rate": 4.7150571124778344e-05, + "loss": 0.1798, + "step": 14480 + }, + { + "epoch": 1.1945847978910369, + "grad_norm": 2.7487854957580566, + "learning_rate": 4.713722302012421e-05, + "loss": 0.1833, + "step": 14500 + }, + { + "epoch": 1.1962324253075571, + "grad_norm": 0.7984181642532349, + "learning_rate": 4.712384562249141e-05, + "loss": 0.1934, + "step": 14520 + }, + { + "epoch": 1.1978800527240774, + "grad_norm": 0.22223146259784698, + "learning_rate": 4.711043894958153e-05, + "loss": 0.1883, + "step": 14540 + }, + { + "epoch": 1.1995276801405976, + "grad_norm": 5.944990158081055, + "learning_rate": 4.7097003019134876e-05, + "loss": 0.1584, + "step": 14560 + }, + { + "epoch": 1.2011753075571177, + "grad_norm": 0.5789378881454468, + "learning_rate": 4.708353784893049e-05, + "loss": 0.1837, + "step": 14580 + }, + { + "epoch": 1.202822934973638, + "grad_norm": 0.4476211369037628, + "learning_rate": 4.707004345678609e-05, + "loss": 0.1878, + "step": 14600 + }, + { + "epoch": 1.2044705623901582, + "grad_norm": 1.0182552337646484, + "learning_rate": 4.705651986055807e-05, + "loss": 0.1968, + "step": 14620 + }, + { + "epoch": 1.2061181898066784, + "grad_norm": 0.5336174368858337, + "learning_rate": 4.7042967078141466e-05, + "loss": 0.1849, + "step": 14640 + }, + { + "epoch": 1.2077658172231986, + "grad_norm": 0.6163532733917236, + "learning_rate": 4.702938512746994e-05, + "loss": 0.1571, + "step": 14660 + }, + { + "epoch": 1.209413444639719, + "grad_norm": 1.7392462491989136, + "learning_rate": 4.701577402651574e-05, + "loss": 0.205, + "step": 14680 + }, + { + "epoch": 1.2110610720562391, + "grad_norm": 0.9206298589706421, + "learning_rate": 4.70021337932897e-05, + "loss": 0.1872, + "step": 14700 + }, + { + "epoch": 1.2127086994727592, + "grad_norm": 1.0826961994171143, + "learning_rate": 4.6988464445841186e-05, + "loss": 0.185, + "step": 14720 + }, + { + "epoch": 1.2143563268892794, + "grad_norm": 0.3322293758392334, + "learning_rate": 4.6974766002258105e-05, + "loss": 0.1746, + "step": 14740 + }, + { + "epoch": 1.2160039543057997, + "grad_norm": 1.4564270973205566, + "learning_rate": 4.696103848066686e-05, + "loss": 0.1581, + "step": 14760 + }, + { + "epoch": 1.21765158172232, + "grad_norm": 0.7961132526397705, + "learning_rate": 4.6947281899232333e-05, + "loss": 0.1774, + "step": 14780 + }, + { + "epoch": 1.2192992091388402, + "grad_norm": 2.5025506019592285, + "learning_rate": 4.693349627615784e-05, + "loss": 0.2024, + "step": 14800 + }, + { + "epoch": 1.2209468365553602, + "grad_norm": 0.9771924614906311, + "learning_rate": 4.692037305103247e-05, + "loss": 0.1857, + "step": 14820 + }, + { + "epoch": 1.2225944639718804, + "grad_norm": 0.3456748127937317, + "learning_rate": 4.690653084926293e-05, + "loss": 0.1828, + "step": 14840 + }, + { + "epoch": 1.2242420913884007, + "grad_norm": 0.9823417067527771, + "learning_rate": 4.689265965977708e-05, + "loss": 0.1599, + "step": 14860 + }, + { + "epoch": 1.225889718804921, + "grad_norm": 1.0269907712936401, + "learning_rate": 4.68787595009299e-05, + "loss": 0.1797, + "step": 14880 + }, + { + "epoch": 1.2275373462214412, + "grad_norm": 0.7121568918228149, + "learning_rate": 4.686483039111472e-05, + "loss": 0.1851, + "step": 14900 + }, + { + "epoch": 1.2291849736379614, + "grad_norm": 0.5829301476478577, + "learning_rate": 4.6850872348763166e-05, + "loss": 0.1816, + "step": 14920 + }, + { + "epoch": 1.2308326010544817, + "grad_norm": 0.23825562000274658, + "learning_rate": 4.683688539234515e-05, + "loss": 0.1687, + "step": 14940 + }, + { + "epoch": 1.2324802284710017, + "grad_norm": 1.7142807245254517, + "learning_rate": 4.682286954036886e-05, + "loss": 0.1618, + "step": 14960 + }, + { + "epoch": 1.234127855887522, + "grad_norm": 2.196317195892334, + "learning_rate": 4.6808824811380704e-05, + "loss": 0.1868, + "step": 14980 + }, + { + "epoch": 1.2357754833040422, + "grad_norm": 0.6108691096305847, + "learning_rate": 4.67947512239653e-05, + "loss": 0.195, + "step": 15000 + }, + { + "epoch": 1.2374231107205624, + "grad_norm": 0.5913376808166504, + "learning_rate": 4.678135460276387e-05, + "loss": 0.1855, + "step": 15020 + }, + { + "epoch": 1.2390707381370827, + "grad_norm": 0.6738501191139221, + "learning_rate": 4.676722479501395e-05, + "loss": 0.1843, + "step": 15040 + }, + { + "epoch": 1.2407183655536027, + "grad_norm": 1.7320432662963867, + "learning_rate": 4.675306618388381e-05, + "loss": 0.1567, + "step": 15060 + }, + { + "epoch": 1.242365992970123, + "grad_norm": 1.1772425174713135, + "learning_rate": 4.6738878788108776e-05, + "loss": 0.1927, + "step": 15080 + }, + { + "epoch": 1.2440136203866432, + "grad_norm": 0.8685306906700134, + "learning_rate": 4.672466262646225e-05, + "loss": 0.2057, + "step": 15100 + }, + { + "epoch": 1.2456612478031635, + "grad_norm": 0.5060691833496094, + "learning_rate": 4.6710417717755695e-05, + "loss": 0.1846, + "step": 15120 + }, + { + "epoch": 1.2473088752196837, + "grad_norm": 0.2376640886068344, + "learning_rate": 4.6696144080838624e-05, + "loss": 0.1835, + "step": 15140 + }, + { + "epoch": 1.248956502636204, + "grad_norm": 1.7704603672027588, + "learning_rate": 4.6681841734598576e-05, + "loss": 0.1661, + "step": 15160 + }, + { + "epoch": 1.2506041300527242, + "grad_norm": 0.9057344198226929, + "learning_rate": 4.6667510697961045e-05, + "loss": 0.1942, + "step": 15180 + }, + { + "epoch": 1.2522517574692442, + "grad_norm": 0.7479068636894226, + "learning_rate": 4.6653150989889525e-05, + "loss": 0.1807, + "step": 15200 + }, + { + "epoch": 1.2538993848857645, + "grad_norm": 0.8971445560455322, + "learning_rate": 4.663876262938543e-05, + "loss": 0.192, + "step": 15220 + }, + { + "epoch": 1.2555470123022847, + "grad_norm": 0.6187984347343445, + "learning_rate": 4.6624345635488085e-05, + "loss": 0.1781, + "step": 15240 + }, + { + "epoch": 1.257194639718805, + "grad_norm": 1.5628702640533447, + "learning_rate": 4.6609900027274714e-05, + "loss": 0.1511, + "step": 15260 + }, + { + "epoch": 1.2588422671353252, + "grad_norm": 1.0656757354736328, + "learning_rate": 4.659542582386041e-05, + "loss": 0.1888, + "step": 15280 + }, + { + "epoch": 1.2604898945518452, + "grad_norm": 0.9140082597732544, + "learning_rate": 4.658092304439808e-05, + "loss": 0.1791, + "step": 15300 + }, + { + "epoch": 1.2621375219683655, + "grad_norm": 0.7536066770553589, + "learning_rate": 4.6566391708078484e-05, + "loss": 0.1945, + "step": 15320 + }, + { + "epoch": 1.2637851493848857, + "grad_norm": 0.6287294030189514, + "learning_rate": 4.655183183413012e-05, + "loss": 0.1844, + "step": 15340 + }, + { + "epoch": 1.265432776801406, + "grad_norm": 1.3471739292144775, + "learning_rate": 4.653724344181929e-05, + "loss": 0.1718, + "step": 15360 + }, + { + "epoch": 1.2670804042179262, + "grad_norm": 0.8668192625045776, + "learning_rate": 4.652262655045001e-05, + "loss": 0.1816, + "step": 15380 + }, + { + "epoch": 1.2687280316344465, + "grad_norm": 0.7043120861053467, + "learning_rate": 4.6507981179364015e-05, + "loss": 0.1835, + "step": 15400 + }, + { + "epoch": 1.2703756590509667, + "grad_norm": 2.341987133026123, + "learning_rate": 4.6493307347940735e-05, + "loss": 0.1779, + "step": 15420 + }, + { + "epoch": 1.2720232864674867, + "grad_norm": 0.7353787422180176, + "learning_rate": 4.6478605075597236e-05, + "loss": 0.1995, + "step": 15440 + }, + { + "epoch": 1.273670913884007, + "grad_norm": 1.131020188331604, + "learning_rate": 4.646387438178824e-05, + "loss": 0.1626, + "step": 15460 + }, + { + "epoch": 1.2753185413005272, + "grad_norm": 5.1669206619262695, + "learning_rate": 4.644911528600607e-05, + "loss": 0.1739, + "step": 15480 + }, + { + "epoch": 1.2769661687170475, + "grad_norm": 1.9862631559371948, + "learning_rate": 4.643432780778064e-05, + "loss": 0.1811, + "step": 15500 + }, + { + "epoch": 1.2786137961335677, + "grad_norm": 4.627335548400879, + "learning_rate": 4.641951196667939e-05, + "loss": 0.1755, + "step": 15520 + }, + { + "epoch": 1.2802614235500878, + "grad_norm": 1.6128586530685425, + "learning_rate": 4.6404667782307334e-05, + "loss": 0.1758, + "step": 15540 + }, + { + "epoch": 1.281909050966608, + "grad_norm": 0.8337876200675964, + "learning_rate": 4.638979527430696e-05, + "loss": 0.1543, + "step": 15560 + }, + { + "epoch": 1.2835566783831283, + "grad_norm": 1.7509500980377197, + "learning_rate": 4.6374894462358246e-05, + "loss": 0.1866, + "step": 15580 + }, + { + "epoch": 1.2852043057996485, + "grad_norm": 1.9350312948226929, + "learning_rate": 4.635996536617863e-05, + "loss": 0.1898, + "step": 15600 + }, + { + "epoch": 1.2868519332161688, + "grad_norm": 0.9639732241630554, + "learning_rate": 4.6345008005522966e-05, + "loss": 0.1757, + "step": 15620 + }, + { + "epoch": 1.288499560632689, + "grad_norm": 0.22245891392230988, + "learning_rate": 4.633002240018351e-05, + "loss": 0.1725, + "step": 15640 + }, + { + "epoch": 1.2901471880492092, + "grad_norm": 1.5688083171844482, + "learning_rate": 4.6315008569989905e-05, + "loss": 0.1682, + "step": 15660 + }, + { + "epoch": 1.2917948154657293, + "grad_norm": 0.9785603880882263, + "learning_rate": 4.629996653480912e-05, + "loss": 0.184, + "step": 15680 + }, + { + "epoch": 1.2934424428822495, + "grad_norm": 1.3254144191741943, + "learning_rate": 4.628489631454548e-05, + "loss": 0.1862, + "step": 15700 + }, + { + "epoch": 1.2950900702987698, + "grad_norm": 2.9795165061950684, + "learning_rate": 4.6269797929140554e-05, + "loss": 0.1836, + "step": 15720 + }, + { + "epoch": 1.29673769771529, + "grad_norm": 0.510050356388092, + "learning_rate": 4.625467139857323e-05, + "loss": 0.194, + "step": 15740 + }, + { + "epoch": 1.2983853251318103, + "grad_norm": 1.456040859222412, + "learning_rate": 4.6239516742859625e-05, + "loss": 0.1679, + "step": 15760 + }, + { + "epoch": 1.3000329525483303, + "grad_norm": 0.7667500376701355, + "learning_rate": 4.622433398205305e-05, + "loss": 0.1955, + "step": 15780 + }, + { + "epoch": 1.3016805799648505, + "grad_norm": 5.522034168243408, + "learning_rate": 4.620912313624404e-05, + "loss": 0.1766, + "step": 15800 + }, + { + "epoch": 1.3033282073813708, + "grad_norm": 1.5902382135391235, + "learning_rate": 4.619388422556026e-05, + "loss": 0.1843, + "step": 15820 + }, + { + "epoch": 1.304975834797891, + "grad_norm": 0.3163457214832306, + "learning_rate": 4.617861727016655e-05, + "loss": 0.1876, + "step": 15840 + }, + { + "epoch": 1.3066234622144113, + "grad_norm": 1.0746263265609741, + "learning_rate": 4.616332229026482e-05, + "loss": 0.1727, + "step": 15860 + }, + { + "epoch": 1.3082710896309315, + "grad_norm": 7.499575614929199, + "learning_rate": 4.6147999306094095e-05, + "loss": 0.191, + "step": 15880 + }, + { + "epoch": 1.3099187170474518, + "grad_norm": 1.3614699840545654, + "learning_rate": 4.613264833793044e-05, + "loss": 0.1979, + "step": 15900 + }, + { + "epoch": 1.3115663444639718, + "grad_norm": 2.371039390563965, + "learning_rate": 4.611726940608694e-05, + "loss": 0.1868, + "step": 15920 + }, + { + "epoch": 1.313213971880492, + "grad_norm": 0.4890645444393158, + "learning_rate": 4.6101862530913715e-05, + "loss": 0.181, + "step": 15940 + }, + { + "epoch": 1.3148615992970123, + "grad_norm": 1.2924984693527222, + "learning_rate": 4.6086427732797836e-05, + "loss": 0.1826, + "step": 15960 + }, + { + "epoch": 1.3165092267135325, + "grad_norm": 0.8311454653739929, + "learning_rate": 4.6070965032163316e-05, + "loss": 0.1757, + "step": 15980 + }, + { + "epoch": 1.3181568541300528, + "grad_norm": 4.735937595367432, + "learning_rate": 4.605547444947112e-05, + "loss": 0.1934, + "step": 16000 + }, + { + "epoch": 1.3181568541300528, + "eval_loss": 0.5345502495765686, + "eval_runtime": 711.0231, + "eval_samples_per_second": 30.645, + "eval_steps_per_second": 7.662, + "eval_wer": 0.23376714265318863, + "step": 16000 + }, + { + "epoch": 1.3198044815465728, + "grad_norm": 0.9712895750999451, + "learning_rate": 4.603995600521908e-05, + "loss": 0.1792, + "step": 16020 + }, + { + "epoch": 1.321452108963093, + "grad_norm": 0.3473972976207733, + "learning_rate": 4.6024409719941905e-05, + "loss": 0.1926, + "step": 16040 + }, + { + "epoch": 1.3230997363796133, + "grad_norm": 1.2832224369049072, + "learning_rate": 4.600883561421116e-05, + "loss": 0.1696, + "step": 16060 + }, + { + "epoch": 1.3247473637961336, + "grad_norm": 1.0073579549789429, + "learning_rate": 4.599401446384175e-05, + "loss": 0.1816, + "step": 16080 + }, + { + "epoch": 1.3263949912126538, + "grad_norm": 1.0329769849777222, + "learning_rate": 4.597838616753484e-05, + "loss": 0.1834, + "step": 16100 + }, + { + "epoch": 1.328042618629174, + "grad_norm": 2.175204038619995, + "learning_rate": 4.596273011167479e-05, + "loss": 0.1936, + "step": 16120 + }, + { + "epoch": 1.3296902460456943, + "grad_norm": 0.29955798387527466, + "learning_rate": 4.594704631697842e-05, + "loss": 0.1735, + "step": 16140 + }, + { + "epoch": 1.3313378734622145, + "grad_norm": 5.876577377319336, + "learning_rate": 4.593133480419924e-05, + "loss": 0.1642, + "step": 16160 + }, + { + "epoch": 1.3329855008787346, + "grad_norm": 1.0501422882080078, + "learning_rate": 4.591559559412743e-05, + "loss": 0.2176, + "step": 16180 + }, + { + "epoch": 1.3346331282952548, + "grad_norm": 2.9819207191467285, + "learning_rate": 4.5899828707589844e-05, + "loss": 0.1841, + "step": 16200 + }, + { + "epoch": 1.336280755711775, + "grad_norm": 0.8568416237831116, + "learning_rate": 4.588403416544992e-05, + "loss": 0.1916, + "step": 16220 + }, + { + "epoch": 1.3379283831282953, + "grad_norm": 1.058608889579773, + "learning_rate": 4.586821198860772e-05, + "loss": 0.1881, + "step": 16240 + }, + { + "epoch": 1.3395760105448153, + "grad_norm": 1.0041903257369995, + "learning_rate": 4.585236219799988e-05, + "loss": 0.1696, + "step": 16260 + }, + { + "epoch": 1.3412236379613356, + "grad_norm": 1.351030945777893, + "learning_rate": 4.5836484814599544e-05, + "loss": 0.183, + "step": 16280 + }, + { + "epoch": 1.3428712653778558, + "grad_norm": 0.7650487422943115, + "learning_rate": 4.58205798594164e-05, + "loss": 0.1775, + "step": 16300 + }, + { + "epoch": 1.344518892794376, + "grad_norm": 1.0618997812271118, + "learning_rate": 4.5804647353496604e-05, + "loss": 0.1846, + "step": 16320 + }, + { + "epoch": 1.3461665202108963, + "grad_norm": 1.2706764936447144, + "learning_rate": 4.578868731792277e-05, + "loss": 0.1994, + "step": 16340 + }, + { + "epoch": 1.3478141476274166, + "grad_norm": 0.9554028511047363, + "learning_rate": 4.5772699773813936e-05, + "loss": 0.1564, + "step": 16360 + }, + { + "epoch": 1.3494617750439368, + "grad_norm": 0.8473567366600037, + "learning_rate": 4.575668474232556e-05, + "loss": 0.1924, + "step": 16380 + }, + { + "epoch": 1.351109402460457, + "grad_norm": 1.2312932014465332, + "learning_rate": 4.574064224464944e-05, + "loss": 0.194, + "step": 16400 + }, + { + "epoch": 1.352757029876977, + "grad_norm": 0.5687373280525208, + "learning_rate": 4.5724572302013754e-05, + "loss": 0.1828, + "step": 16420 + }, + { + "epoch": 1.3544046572934973, + "grad_norm": 0.2973005473613739, + "learning_rate": 4.570847493568297e-05, + "loss": 0.1774, + "step": 16440 + }, + { + "epoch": 1.3560522847100176, + "grad_norm": 22.61744499206543, + "learning_rate": 4.5692350166957864e-05, + "loss": 0.16, + "step": 16460 + }, + { + "epoch": 1.3576999121265378, + "grad_norm": 0.5010253190994263, + "learning_rate": 4.567619801717545e-05, + "loss": 0.1842, + "step": 16480 + }, + { + "epoch": 1.3593475395430579, + "grad_norm": 1.405232548713684, + "learning_rate": 4.566001850770899e-05, + "loss": 0.194, + "step": 16500 + }, + { + "epoch": 1.3609951669595781, + "grad_norm": 0.8372935652732849, + "learning_rate": 4.564381165996796e-05, + "loss": 0.1866, + "step": 16520 + }, + { + "epoch": 1.3626427943760984, + "grad_norm": 0.2158607393503189, + "learning_rate": 4.5627577495397974e-05, + "loss": 0.1802, + "step": 16540 + }, + { + "epoch": 1.3642904217926186, + "grad_norm": 7.98114013671875, + "learning_rate": 4.561131603548082e-05, + "loss": 0.1832, + "step": 16560 + }, + { + "epoch": 1.3659380492091389, + "grad_norm": 0.6528833508491516, + "learning_rate": 4.5595027301734413e-05, + "loss": 0.1919, + "step": 16580 + }, + { + "epoch": 1.367585676625659, + "grad_norm": 0.5290044546127319, + "learning_rate": 4.557871131571274e-05, + "loss": 0.1693, + "step": 16600 + }, + { + "epoch": 1.3692333040421794, + "grad_norm": 1.5808736085891724, + "learning_rate": 4.5562368099005854e-05, + "loss": 0.1813, + "step": 16620 + }, + { + "epoch": 1.3708809314586996, + "grad_norm": 0.17612561583518982, + "learning_rate": 4.554599767323984e-05, + "loss": 0.1879, + "step": 16640 + }, + { + "epoch": 1.3725285588752196, + "grad_norm": 1.369644284248352, + "learning_rate": 4.552960006007678e-05, + "loss": 0.1602, + "step": 16660 + }, + { + "epoch": 1.3741761862917399, + "grad_norm": 1.1678745746612549, + "learning_rate": 4.551317528121476e-05, + "loss": 0.1897, + "step": 16680 + }, + { + "epoch": 1.3758238137082601, + "grad_norm": 0.5510971546173096, + "learning_rate": 4.549672335838777e-05, + "loss": 0.196, + "step": 16700 + }, + { + "epoch": 1.3774714411247804, + "grad_norm": 0.9784049391746521, + "learning_rate": 4.548024431336576e-05, + "loss": 0.1865, + "step": 16720 + }, + { + "epoch": 1.3791190685413004, + "grad_norm": 0.2733114957809448, + "learning_rate": 4.546373816795455e-05, + "loss": 0.1811, + "step": 16740 + }, + { + "epoch": 1.3807666959578206, + "grad_norm": 1.3034899234771729, + "learning_rate": 4.5447204943995816e-05, + "loss": 0.1528, + "step": 16760 + }, + { + "epoch": 1.382414323374341, + "grad_norm": 1.3286523818969727, + "learning_rate": 4.543064466336706e-05, + "loss": 0.1832, + "step": 16780 + }, + { + "epoch": 1.3840619507908611, + "grad_norm": 0.747806191444397, + "learning_rate": 4.541405734798161e-05, + "loss": 0.1672, + "step": 16800 + }, + { + "epoch": 1.3857095782073814, + "grad_norm": 0.5867149233818054, + "learning_rate": 4.539744301978856e-05, + "loss": 0.1814, + "step": 16820 + }, + { + "epoch": 1.3873572056239016, + "grad_norm": 0.43317413330078125, + "learning_rate": 4.5380801700772726e-05, + "loss": 0.1825, + "step": 16840 + }, + { + "epoch": 1.3890048330404219, + "grad_norm": 1.011514663696289, + "learning_rate": 4.536413341295467e-05, + "loss": 0.1553, + "step": 16860 + }, + { + "epoch": 1.3906524604569421, + "grad_norm": 0.7447202205657959, + "learning_rate": 4.5347438178390624e-05, + "loss": 0.1832, + "step": 16880 + }, + { + "epoch": 1.3923000878734622, + "grad_norm": 4.756796360015869, + "learning_rate": 4.533071601917248e-05, + "loss": 0.1923, + "step": 16900 + }, + { + "epoch": 1.3939477152899824, + "grad_norm": 1.5507986545562744, + "learning_rate": 4.5313966957427755e-05, + "loss": 0.1754, + "step": 16920 + }, + { + "epoch": 1.3955953427065027, + "grad_norm": 0.557498037815094, + "learning_rate": 4.529719101531957e-05, + "loss": 0.178, + "step": 16940 + }, + { + "epoch": 1.397242970123023, + "grad_norm": 1.611181378364563, + "learning_rate": 4.5280388215046624e-05, + "loss": 0.1557, + "step": 16960 + }, + { + "epoch": 1.398890597539543, + "grad_norm": 0.6798775792121887, + "learning_rate": 4.5263558578843136e-05, + "loss": 0.1741, + "step": 16980 + }, + { + "epoch": 1.4005382249560632, + "grad_norm": 10.896245956420898, + "learning_rate": 4.524670212897884e-05, + "loss": 0.2016, + "step": 17000 + }, + { + "epoch": 1.4021858523725834, + "grad_norm": 1.8802106380462646, + "learning_rate": 4.5229818887758957e-05, + "loss": 0.177, + "step": 17020 + }, + { + "epoch": 1.4038334797891037, + "grad_norm": 0.2604859471321106, + "learning_rate": 4.5212908877524165e-05, + "loss": 0.1791, + "step": 17040 + }, + { + "epoch": 1.405481107205624, + "grad_norm": 1.834112524986267, + "learning_rate": 4.519597212065056e-05, + "loss": 0.1664, + "step": 17060 + }, + { + "epoch": 1.4071287346221442, + "grad_norm": 0.9913097620010376, + "learning_rate": 4.517900863954962e-05, + "loss": 0.1955, + "step": 17080 + }, + { + "epoch": 1.4087763620386644, + "grad_norm": 2.1117942333221436, + "learning_rate": 4.51620184566682e-05, + "loss": 0.1798, + "step": 17100 + }, + { + "epoch": 1.4104239894551847, + "grad_norm": 5.098875999450684, + "learning_rate": 4.5145001594488466e-05, + "loss": 0.1908, + "step": 17120 + }, + { + "epoch": 1.4120716168717047, + "grad_norm": 0.7090994715690613, + "learning_rate": 4.512795807552792e-05, + "loss": 0.1727, + "step": 17140 + }, + { + "epoch": 1.413719244288225, + "grad_norm": 3.752267599105835, + "learning_rate": 4.511088792233932e-05, + "loss": 0.1542, + "step": 17160 + }, + { + "epoch": 1.4153668717047452, + "grad_norm": 1.5449800491333008, + "learning_rate": 4.5093791157510656e-05, + "loss": 0.1651, + "step": 17180 + }, + { + "epoch": 1.4170144991212654, + "grad_norm": 0.9400858283042908, + "learning_rate": 4.507666780366514e-05, + "loss": 0.1844, + "step": 17200 + }, + { + "epoch": 1.4186621265377855, + "grad_norm": 0.37838345766067505, + "learning_rate": 4.5059517883461196e-05, + "loss": 0.1733, + "step": 17220 + }, + { + "epoch": 1.4203097539543057, + "grad_norm": 0.239340141415596, + "learning_rate": 4.50432008728471e-05, + "loss": 0.1855, + "step": 17240 + }, + { + "epoch": 1.421957381370826, + "grad_norm": 1.4485529661178589, + "learning_rate": 4.502599921354852e-05, + "loss": 0.1515, + "step": 17260 + }, + { + "epoch": 1.4236050087873462, + "grad_norm": 0.6260993480682373, + "learning_rate": 4.500877105493846e-05, + "loss": 0.1781, + "step": 17280 + }, + { + "epoch": 1.4252526362038664, + "grad_norm": 2.114168882369995, + "learning_rate": 4.499151641981402e-05, + "loss": 0.1732, + "step": 17300 + }, + { + "epoch": 1.4269002636203867, + "grad_norm": 0.7402538061141968, + "learning_rate": 4.497423533100732e-05, + "loss": 0.1769, + "step": 17320 + }, + { + "epoch": 1.428547891036907, + "grad_norm": 0.7155308723449707, + "learning_rate": 4.4956927811385475e-05, + "loss": 0.1649, + "step": 17340 + }, + { + "epoch": 1.4301955184534272, + "grad_norm": 0.8413863182067871, + "learning_rate": 4.4939593883850576e-05, + "loss": 0.1563, + "step": 17360 + }, + { + "epoch": 1.4318431458699472, + "grad_norm": 2.0796799659729004, + "learning_rate": 4.492223357133967e-05, + "loss": 0.1797, + "step": 17380 + }, + { + "epoch": 1.4334907732864675, + "grad_norm": 1.1681219339370728, + "learning_rate": 4.490484689682471e-05, + "loss": 0.1819, + "step": 17400 + }, + { + "epoch": 1.4351384007029877, + "grad_norm": 1.0417784452438354, + "learning_rate": 4.488743388331254e-05, + "loss": 0.1861, + "step": 17420 + }, + { + "epoch": 1.436786028119508, + "grad_norm": 0.17414021492004395, + "learning_rate": 4.486999455384483e-05, + "loss": 0.1734, + "step": 17440 + }, + { + "epoch": 1.4384336555360282, + "grad_norm": 1.723633885383606, + "learning_rate": 4.485252893149812e-05, + "loss": 0.1629, + "step": 17460 + }, + { + "epoch": 1.4400812829525482, + "grad_norm": 0.5598846673965454, + "learning_rate": 4.4835037039383686e-05, + "loss": 0.1749, + "step": 17480 + }, + { + "epoch": 1.4417289103690685, + "grad_norm": 0.617661714553833, + "learning_rate": 4.481751890064761e-05, + "loss": 0.1734, + "step": 17500 + }, + { + "epoch": 1.4433765377855887, + "grad_norm": 0.4917828440666199, + "learning_rate": 4.479997453847069e-05, + "loss": 0.1724, + "step": 17520 + }, + { + "epoch": 1.445024165202109, + "grad_norm": 0.27072271704673767, + "learning_rate": 4.4782403976068417e-05, + "loss": 0.1664, + "step": 17540 + }, + { + "epoch": 1.4466717926186292, + "grad_norm": 1.5086781978607178, + "learning_rate": 4.476480723669095e-05, + "loss": 0.1614, + "step": 17560 + }, + { + "epoch": 1.4483194200351495, + "grad_norm": 1.092875599861145, + "learning_rate": 4.4747184343623113e-05, + "loss": 0.1827, + "step": 17580 + }, + { + "epoch": 1.4499670474516697, + "grad_norm": 0.9140943288803101, + "learning_rate": 4.4729535320184304e-05, + "loss": 0.1865, + "step": 17600 + }, + { + "epoch": 1.4516146748681897, + "grad_norm": 0.9031366109848022, + "learning_rate": 4.471186018972852e-05, + "loss": 0.1767, + "step": 17620 + }, + { + "epoch": 1.45326230228471, + "grad_norm": 0.4150620102882385, + "learning_rate": 4.46941589756443e-05, + "loss": 0.1814, + "step": 17640 + }, + { + "epoch": 1.4549099297012302, + "grad_norm": 2.178504228591919, + "learning_rate": 4.467643170135467e-05, + "loss": 0.1608, + "step": 17660 + }, + { + "epoch": 1.4565575571177505, + "grad_norm": 1.05979323387146, + "learning_rate": 4.4658678390317186e-05, + "loss": 0.175, + "step": 17680 + }, + { + "epoch": 1.4582051845342707, + "grad_norm": 1.3193062543869019, + "learning_rate": 4.464089906602383e-05, + "loss": 0.2092, + "step": 17700 + }, + { + "epoch": 1.4598528119507908, + "grad_norm": 0.6934455037117004, + "learning_rate": 4.462309375200101e-05, + "loss": 0.1695, + "step": 17720 + }, + { + "epoch": 1.461500439367311, + "grad_norm": 0.34769323468208313, + "learning_rate": 4.460526247180953e-05, + "loss": 0.1844, + "step": 17740 + }, + { + "epoch": 1.4631480667838312, + "grad_norm": 2.567417621612549, + "learning_rate": 4.4587405249044546e-05, + "loss": 0.1657, + "step": 17760 + }, + { + "epoch": 1.4647956942003515, + "grad_norm": 0.7568791508674622, + "learning_rate": 4.4569522107335535e-05, + "loss": 0.1794, + "step": 17780 + }, + { + "epoch": 1.4664433216168717, + "grad_norm": 0.6661089062690735, + "learning_rate": 4.455161307034631e-05, + "loss": 0.1812, + "step": 17800 + }, + { + "epoch": 1.468090949033392, + "grad_norm": 0.6958649754524231, + "learning_rate": 4.4533678161774894e-05, + "loss": 0.1714, + "step": 17820 + }, + { + "epoch": 1.4697385764499122, + "grad_norm": 0.3765020966529846, + "learning_rate": 4.45157174053536e-05, + "loss": 0.176, + "step": 17840 + }, + { + "epoch": 1.4713862038664323, + "grad_norm": 1.378657579421997, + "learning_rate": 4.4497730824848885e-05, + "loss": 0.1645, + "step": 17860 + }, + { + "epoch": 1.4730338312829525, + "grad_norm": 0.8017684817314148, + "learning_rate": 4.447971844406143e-05, + "loss": 0.1795, + "step": 17880 + }, + { + "epoch": 1.4746814586994728, + "grad_norm": 1.0388306379318237, + "learning_rate": 4.4461680286826024e-05, + "loss": 0.1941, + "step": 17900 + }, + { + "epoch": 1.476329086115993, + "grad_norm": 0.5598440766334534, + "learning_rate": 4.4443616377011586e-05, + "loss": 0.1879, + "step": 17920 + }, + { + "epoch": 1.4779767135325133, + "grad_norm": 0.18948721885681152, + "learning_rate": 4.4425526738521096e-05, + "loss": 0.1837, + "step": 17940 + }, + { + "epoch": 1.4796243409490333, + "grad_norm": 1.4350675344467163, + "learning_rate": 4.440741139529159e-05, + "loss": 0.1578, + "step": 17960 + }, + { + "epoch": 1.4812719683655535, + "grad_norm": 0.7614567279815674, + "learning_rate": 4.4389270371294104e-05, + "loss": 0.1878, + "step": 17980 + }, + { + "epoch": 1.4829195957820738, + "grad_norm": 0.5294657945632935, + "learning_rate": 4.437110369053368e-05, + "loss": 0.1792, + "step": 18000 + }, + { + "epoch": 1.4829195957820738, + "eval_loss": 0.539612352848053, + "eval_runtime": 824.8651, + "eval_samples_per_second": 26.415, + "eval_steps_per_second": 6.605, + "eval_wer": 0.2317739899166627, + "step": 18000 + }, + { + "epoch": 1.484567223198594, + "grad_norm": 0.9461752772331238, + "learning_rate": 4.4352911377049266e-05, + "loss": 0.1759, + "step": 18020 + }, + { + "epoch": 1.4862148506151143, + "grad_norm": 0.35723596811294556, + "learning_rate": 4.433469345491379e-05, + "loss": 0.1883, + "step": 18040 + }, + { + "epoch": 1.4878624780316345, + "grad_norm": 1.2138187885284424, + "learning_rate": 4.4316449948234007e-05, + "loss": 0.1486, + "step": 18060 + }, + { + "epoch": 1.4895101054481548, + "grad_norm": 0.9856829643249512, + "learning_rate": 4.4298180881150575e-05, + "loss": 0.1985, + "step": 18080 + }, + { + "epoch": 1.4911577328646748, + "grad_norm": 0.697496771812439, + "learning_rate": 4.4279886277837945e-05, + "loss": 0.1799, + "step": 18100 + }, + { + "epoch": 1.492805360281195, + "grad_norm": 0.8957895040512085, + "learning_rate": 4.426156616250436e-05, + "loss": 0.1755, + "step": 18120 + }, + { + "epoch": 1.4944529876977153, + "grad_norm": 0.3005205988883972, + "learning_rate": 4.424322055939183e-05, + "loss": 0.1754, + "step": 18140 + }, + { + "epoch": 1.4961006151142355, + "grad_norm": 0.5195316672325134, + "learning_rate": 4.42248494927761e-05, + "loss": 0.1566, + "step": 18160 + }, + { + "epoch": 1.4977482425307558, + "grad_norm": 1.898149013519287, + "learning_rate": 4.42064529869666e-05, + "loss": 0.177, + "step": 18180 + }, + { + "epoch": 1.4993958699472758, + "grad_norm": 2.6191961765289307, + "learning_rate": 4.4188031066306406e-05, + "loss": 0.1794, + "step": 18200 + }, + { + "epoch": 1.501043497363796, + "grad_norm": 0.5942957997322083, + "learning_rate": 4.416958375517225e-05, + "loss": 0.1805, + "step": 18220 + }, + { + "epoch": 1.5026911247803163, + "grad_norm": 0.2722390294075012, + "learning_rate": 4.415111107797445e-05, + "loss": 0.1704, + "step": 18240 + }, + { + "epoch": 1.5043387521968365, + "grad_norm": 1.5523074865341187, + "learning_rate": 4.41326130591569e-05, + "loss": 0.1683, + "step": 18260 + }, + { + "epoch": 1.5059863796133568, + "grad_norm": 0.6534819006919861, + "learning_rate": 4.4114089723197e-05, + "loss": 0.1734, + "step": 18280 + }, + { + "epoch": 1.507634007029877, + "grad_norm": 0.3791787922382355, + "learning_rate": 4.409554109460568e-05, + "loss": 0.1713, + "step": 18300 + }, + { + "epoch": 1.5092816344463973, + "grad_norm": 0.8552284836769104, + "learning_rate": 4.407696719792734e-05, + "loss": 0.1816, + "step": 18320 + }, + { + "epoch": 1.5109292618629175, + "grad_norm": 1.9957807064056396, + "learning_rate": 4.405836805773979e-05, + "loss": 0.1676, + "step": 18340 + }, + { + "epoch": 1.5125768892794376, + "grad_norm": 3.502870559692383, + "learning_rate": 4.4039743698654244e-05, + "loss": 0.1557, + "step": 18360 + }, + { + "epoch": 1.5142245166959578, + "grad_norm": 2.0844368934631348, + "learning_rate": 4.402109414531532e-05, + "loss": 0.1744, + "step": 18380 + }, + { + "epoch": 1.515872144112478, + "grad_norm": 2.32145357131958, + "learning_rate": 4.4002419422400954e-05, + "loss": 0.1853, + "step": 18400 + }, + { + "epoch": 1.517519771528998, + "grad_norm": 4.515986919403076, + "learning_rate": 4.398371955462237e-05, + "loss": 0.1737, + "step": 18420 + }, + { + "epoch": 1.5191673989455183, + "grad_norm": 0.3069842457771301, + "learning_rate": 4.396499456672409e-05, + "loss": 0.1863, + "step": 18440 + }, + { + "epoch": 1.5208150263620386, + "grad_norm": 2.3054933547973633, + "learning_rate": 4.394624448348388e-05, + "loss": 0.1503, + "step": 18460 + }, + { + "epoch": 1.5224626537785588, + "grad_norm": 0.7396153211593628, + "learning_rate": 4.392746932971269e-05, + "loss": 0.1882, + "step": 18480 + }, + { + "epoch": 1.524110281195079, + "grad_norm": 0.6996922492980957, + "learning_rate": 4.390866913025466e-05, + "loss": 0.1812, + "step": 18500 + }, + { + "epoch": 1.5257579086115993, + "grad_norm": 0.8116775751113892, + "learning_rate": 4.3889843909987064e-05, + "loss": 0.182, + "step": 18520 + }, + { + "epoch": 1.5274055360281196, + "grad_norm": 0.23936544358730316, + "learning_rate": 4.387099369382029e-05, + "loss": 0.1837, + "step": 18540 + }, + { + "epoch": 1.5290531634446398, + "grad_norm": 0.7272351980209351, + "learning_rate": 4.3852118506697795e-05, + "loss": 0.1601, + "step": 18560 + }, + { + "epoch": 1.53070079086116, + "grad_norm": 0.8234082460403442, + "learning_rate": 4.3833218373596094e-05, + "loss": 0.1628, + "step": 18580 + }, + { + "epoch": 1.53234841827768, + "grad_norm": 0.8772994875907898, + "learning_rate": 4.38142933195247e-05, + "loss": 0.1683, + "step": 18600 + }, + { + "epoch": 1.5339960456942003, + "grad_norm": 0.4307045340538025, + "learning_rate": 4.379534336952609e-05, + "loss": 0.1983, + "step": 18620 + }, + { + "epoch": 1.5356436731107206, + "grad_norm": 0.24042223393917084, + "learning_rate": 4.377636854867572e-05, + "loss": 0.163, + "step": 18640 + }, + { + "epoch": 1.5372913005272406, + "grad_norm": 1.3200021982192993, + "learning_rate": 4.3757368882081905e-05, + "loss": 0.1574, + "step": 18660 + }, + { + "epoch": 1.5389389279437609, + "grad_norm": 0.6988914608955383, + "learning_rate": 4.37383443948859e-05, + "loss": 0.1656, + "step": 18680 + }, + { + "epoch": 1.540586555360281, + "grad_norm": 0.46211034059524536, + "learning_rate": 4.371929511226175e-05, + "loss": 0.1843, + "step": 18700 + }, + { + "epoch": 1.5422341827768014, + "grad_norm": 0.7007827758789062, + "learning_rate": 4.370022105941633e-05, + "loss": 0.1642, + "step": 18720 + }, + { + "epoch": 1.5438818101933216, + "grad_norm": 0.3869803249835968, + "learning_rate": 4.3681122261589305e-05, + "loss": 0.1561, + "step": 18740 + }, + { + "epoch": 1.5455294376098418, + "grad_norm": 0.945253312587738, + "learning_rate": 4.366199874405306e-05, + "loss": 0.161, + "step": 18760 + }, + { + "epoch": 1.547177065026362, + "grad_norm": 1.2405415773391724, + "learning_rate": 4.36428505321127e-05, + "loss": 0.1671, + "step": 18780 + }, + { + "epoch": 1.5488246924428823, + "grad_norm": 0.9262678027153015, + "learning_rate": 4.3623677651106025e-05, + "loss": 0.1819, + "step": 18800 + }, + { + "epoch": 1.5504723198594026, + "grad_norm": 0.5617028474807739, + "learning_rate": 4.360448012640345e-05, + "loss": 0.1823, + "step": 18820 + }, + { + "epoch": 1.5521199472759226, + "grad_norm": 0.24927659332752228, + "learning_rate": 4.358525798340801e-05, + "loss": 0.1802, + "step": 18840 + }, + { + "epoch": 1.5537675746924429, + "grad_norm": 1.1216681003570557, + "learning_rate": 4.356601124755532e-05, + "loss": 0.142, + "step": 18860 + }, + { + "epoch": 1.5554152021089631, + "grad_norm": 1.321250319480896, + "learning_rate": 4.3546739944313544e-05, + "loss": 0.1849, + "step": 18880 + }, + { + "epoch": 1.5570628295254831, + "grad_norm": 1.2743161916732788, + "learning_rate": 4.352744409918332e-05, + "loss": 0.1802, + "step": 18900 + }, + { + "epoch": 1.5587104569420034, + "grad_norm": 0.6473543047904968, + "learning_rate": 4.350812373769781e-05, + "loss": 0.1807, + "step": 18920 + }, + { + "epoch": 1.5603580843585236, + "grad_norm": 0.6407447457313538, + "learning_rate": 4.348877888542259e-05, + "loss": 0.1825, + "step": 18940 + }, + { + "epoch": 1.5620057117750439, + "grad_norm": 1.5329558849334717, + "learning_rate": 4.346940956795564e-05, + "loss": 0.1555, + "step": 18960 + }, + { + "epoch": 1.5636533391915641, + "grad_norm": 0.9263864755630493, + "learning_rate": 4.345001581092733e-05, + "loss": 0.1847, + "step": 18980 + }, + { + "epoch": 1.5653009666080844, + "grad_norm": 0.6662928462028503, + "learning_rate": 4.343059764000035e-05, + "loss": 0.1947, + "step": 19000 + }, + { + "epoch": 1.5669485940246046, + "grad_norm": 3.533243417739868, + "learning_rate": 4.3411155080869704e-05, + "loss": 0.189, + "step": 19020 + }, + { + "epoch": 1.5685962214411249, + "grad_norm": 0.34724509716033936, + "learning_rate": 4.339168815926268e-05, + "loss": 0.1751, + "step": 19040 + }, + { + "epoch": 1.5702438488576451, + "grad_norm": 1.1154448986053467, + "learning_rate": 4.337219690093879e-05, + "loss": 0.1615, + "step": 19060 + }, + { + "epoch": 1.5718914762741654, + "grad_norm": 2.445345163345337, + "learning_rate": 4.335268133168974e-05, + "loss": 0.1927, + "step": 19080 + }, + { + "epoch": 1.5735391036906854, + "grad_norm": 1.5590624809265137, + "learning_rate": 4.3333141477339426e-05, + "loss": 0.1985, + "step": 19100 + }, + { + "epoch": 1.5751867311072056, + "grad_norm": 1.7810287475585938, + "learning_rate": 4.3313577363743874e-05, + "loss": 0.1825, + "step": 19120 + }, + { + "epoch": 1.5768343585237259, + "grad_norm": 0.18706871569156647, + "learning_rate": 4.32939890167912e-05, + "loss": 0.1884, + "step": 19140 + }, + { + "epoch": 1.578481985940246, + "grad_norm": 5.887953281402588, + "learning_rate": 4.327437646240161e-05, + "loss": 0.1584, + "step": 19160 + }, + { + "epoch": 1.5801296133567662, + "grad_norm": 0.7244459390640259, + "learning_rate": 4.32547397265273e-05, + "loss": 0.1821, + "step": 19180 + }, + { + "epoch": 1.5817772407732864, + "grad_norm": 0.7429137229919434, + "learning_rate": 4.323507883515252e-05, + "loss": 0.18, + "step": 19200 + }, + { + "epoch": 1.5834248681898067, + "grad_norm": 0.573156476020813, + "learning_rate": 4.321539381429343e-05, + "loss": 0.1777, + "step": 19220 + }, + { + "epoch": 1.585072495606327, + "grad_norm": 0.19367003440856934, + "learning_rate": 4.3195684689998155e-05, + "loss": 0.1765, + "step": 19240 + }, + { + "epoch": 1.5867201230228472, + "grad_norm": 1.1158256530761719, + "learning_rate": 4.31759514883467e-05, + "loss": 0.1667, + "step": 19260 + }, + { + "epoch": 1.5883677504393674, + "grad_norm": 0.7506932616233826, + "learning_rate": 4.3156194235450936e-05, + "loss": 0.1657, + "step": 19280 + }, + { + "epoch": 1.5900153778558876, + "grad_norm": 4.906134605407715, + "learning_rate": 4.313641295745455e-05, + "loss": 0.1837, + "step": 19300 + }, + { + "epoch": 1.591663005272408, + "grad_norm": 2.754065990447998, + "learning_rate": 4.311660768053304e-05, + "loss": 0.1786, + "step": 19320 + }, + { + "epoch": 1.593310632688928, + "grad_norm": 0.19174990057945251, + "learning_rate": 4.309677843089364e-05, + "loss": 0.1813, + "step": 19340 + }, + { + "epoch": 1.5949582601054482, + "grad_norm": 4.260095119476318, + "learning_rate": 4.307692523477531e-05, + "loss": 0.1542, + "step": 19360 + }, + { + "epoch": 1.5966058875219684, + "grad_norm": 1.4479669332504272, + "learning_rate": 4.305704811844871e-05, + "loss": 0.1803, + "step": 19380 + }, + { + "epoch": 1.5982535149384884, + "grad_norm": 1.0454448461532593, + "learning_rate": 4.303714710821613e-05, + "loss": 0.1877, + "step": 19400 + }, + { + "epoch": 1.5999011423550087, + "grad_norm": 1.397001028060913, + "learning_rate": 4.301722223041151e-05, + "loss": 0.182, + "step": 19420 + }, + { + "epoch": 1.601548769771529, + "grad_norm": 0.5425103902816772, + "learning_rate": 4.299827151317217e-05, + "loss": 0.1793, + "step": 19440 + }, + { + "epoch": 1.6031963971880492, + "grad_norm": 1.6720526218414307, + "learning_rate": 4.297830016946455e-05, + "loss": 0.1804, + "step": 19460 + }, + { + "epoch": 1.6048440246045694, + "grad_norm": 0.6204736232757568, + "learning_rate": 4.295830503605383e-05, + "loss": 0.1906, + "step": 19480 + }, + { + "epoch": 1.6064916520210897, + "grad_norm": 2.814265727996826, + "learning_rate": 4.293828613939845e-05, + "loss": 0.1687, + "step": 19500 + }, + { + "epoch": 1.60813927943761, + "grad_norm": 2.273043394088745, + "learning_rate": 4.2918243505988354e-05, + "loss": 0.1837, + "step": 19520 + }, + { + "epoch": 1.6097869068541302, + "grad_norm": 1.504699945449829, + "learning_rate": 4.289817716234485e-05, + "loss": 0.1715, + "step": 19540 + }, + { + "epoch": 1.6114345342706504, + "grad_norm": 2.6365151405334473, + "learning_rate": 4.287808713502063e-05, + "loss": 0.1655, + "step": 19560 + }, + { + "epoch": 1.6130821616871704, + "grad_norm": 0.9557230472564697, + "learning_rate": 4.2857973450599744e-05, + "loss": 0.1628, + "step": 19580 + }, + { + "epoch": 1.6147297891036907, + "grad_norm": 0.9700766801834106, + "learning_rate": 4.283783613569752e-05, + "loss": 0.1912, + "step": 19600 + }, + { + "epoch": 1.616377416520211, + "grad_norm": 1.2748225927352905, + "learning_rate": 4.281767521696057e-05, + "loss": 0.1788, + "step": 19620 + }, + { + "epoch": 1.618025043936731, + "grad_norm": 0.2875102460384369, + "learning_rate": 4.279749072106675e-05, + "loss": 0.1726, + "step": 19640 + }, + { + "epoch": 1.6196726713532512, + "grad_norm": 2.1846017837524414, + "learning_rate": 4.2777282674725086e-05, + "loss": 0.1544, + "step": 19660 + }, + { + "epoch": 1.6213202987697715, + "grad_norm": 1.5533403158187866, + "learning_rate": 4.2757051104675786e-05, + "loss": 0.1621, + "step": 19680 + }, + { + "epoch": 1.6229679261862917, + "grad_norm": 6.069032669067383, + "learning_rate": 4.273780934867807e-05, + "loss": 0.1855, + "step": 19700 + }, + { + "epoch": 1.624615553602812, + "grad_norm": 1.1619160175323486, + "learning_rate": 4.271753198442825e-05, + "loss": 0.1821, + "step": 19720 + }, + { + "epoch": 1.6262631810193322, + "grad_norm": 0.8923582434654236, + "learning_rate": 4.269723117553562e-05, + "loss": 0.1861, + "step": 19740 + }, + { + "epoch": 1.6279108084358525, + "grad_norm": 4.761159420013428, + "learning_rate": 4.2676906948863134e-05, + "loss": 0.147, + "step": 19760 + }, + { + "epoch": 1.6295584358523727, + "grad_norm": 3.785452127456665, + "learning_rate": 4.265655933130473e-05, + "loss": 0.1789, + "step": 19780 + }, + { + "epoch": 1.631206063268893, + "grad_norm": 7.1180806159973145, + "learning_rate": 4.26361883497853e-05, + "loss": 0.1867, + "step": 19800 + }, + { + "epoch": 1.632853690685413, + "grad_norm": 0.7360396981239319, + "learning_rate": 4.2615794031260655e-05, + "loss": 0.1789, + "step": 19820 + }, + { + "epoch": 1.6345013181019332, + "grad_norm": 0.38821306824684143, + "learning_rate": 4.259537640271749e-05, + "loss": 0.1824, + "step": 19840 + }, + { + "epoch": 1.6361489455184535, + "grad_norm": 1.545905590057373, + "learning_rate": 4.2574935491173325e-05, + "loss": 0.1436, + "step": 19860 + }, + { + "epoch": 1.6377965729349735, + "grad_norm": 2.462200880050659, + "learning_rate": 4.2554471323676505e-05, + "loss": 0.1876, + "step": 19880 + }, + { + "epoch": 1.6394442003514937, + "grad_norm": 1.7179991006851196, + "learning_rate": 4.253398392730614e-05, + "loss": 0.1744, + "step": 19900 + }, + { + "epoch": 1.641091827768014, + "grad_norm": 1.1609716415405273, + "learning_rate": 4.2513473329172106e-05, + "loss": 0.1865, + "step": 19920 + }, + { + "epoch": 1.6427394551845342, + "grad_norm": 0.445671945810318, + "learning_rate": 4.249293955641493e-05, + "loss": 0.1842, + "step": 19940 + }, + { + "epoch": 1.6443870826010545, + "grad_norm": 1.0819984674453735, + "learning_rate": 4.2472382636205845e-05, + "loss": 0.1583, + "step": 19960 + }, + { + "epoch": 1.6460347100175747, + "grad_norm": 2.5188121795654297, + "learning_rate": 4.24518025957467e-05, + "loss": 0.1861, + "step": 19980 + }, + { + "epoch": 1.647682337434095, + "grad_norm": 0.8718958497047424, + "learning_rate": 4.243119946226996e-05, + "loss": 0.1842, + "step": 20000 + }, + { + "epoch": 1.647682337434095, + "eval_loss": 0.5287063121795654, + "eval_runtime": 251.8074, + "eval_samples_per_second": 86.53, + "eval_steps_per_second": 21.636, + "eval_wer": 0.23113573314148306, + "step": 20000 + }, + { + "epoch": 1.6493299648506152, + "grad_norm": 1.2543200254440308, + "learning_rate": 4.2410573263038583e-05, + "loss": 0.1729, + "step": 20020 + }, + { + "epoch": 1.6509775922671355, + "grad_norm": 0.2878034710884094, + "learning_rate": 4.238992402534613e-05, + "loss": 0.1789, + "step": 20040 + }, + { + "epoch": 1.6526252196836555, + "grad_norm": 2.4613733291625977, + "learning_rate": 4.236925177651658e-05, + "loss": 0.1631, + "step": 20060 + }, + { + "epoch": 1.6542728471001757, + "grad_norm": 0.9280514121055603, + "learning_rate": 4.234855654390442e-05, + "loss": 0.1741, + "step": 20080 + }, + { + "epoch": 1.655920474516696, + "grad_norm": 1.553101897239685, + "learning_rate": 4.232783835489451e-05, + "loss": 0.1665, + "step": 20100 + }, + { + "epoch": 1.657568101933216, + "grad_norm": 0.5047245621681213, + "learning_rate": 4.23070972369021e-05, + "loss": 0.1882, + "step": 20120 + }, + { + "epoch": 1.6592157293497363, + "grad_norm": 0.32549336552619934, + "learning_rate": 4.228633321737276e-05, + "loss": 0.1716, + "step": 20140 + }, + { + "epoch": 1.6608633567662565, + "grad_norm": 2.7314305305480957, + "learning_rate": 4.226554632378241e-05, + "loss": 0.1551, + "step": 20160 + }, + { + "epoch": 1.6625109841827768, + "grad_norm": 1.0868523120880127, + "learning_rate": 4.2244736583637214e-05, + "loss": 0.1674, + "step": 20180 + }, + { + "epoch": 1.664158611599297, + "grad_norm": 0.6131607294082642, + "learning_rate": 4.222390402447353e-05, + "loss": 0.1832, + "step": 20200 + }, + { + "epoch": 1.6658062390158173, + "grad_norm": 0.8340827226638794, + "learning_rate": 4.220304867385798e-05, + "loss": 0.1695, + "step": 20220 + }, + { + "epoch": 1.6674538664323375, + "grad_norm": 0.22361576557159424, + "learning_rate": 4.21821705593873e-05, + "loss": 0.1935, + "step": 20240 + }, + { + "epoch": 1.6691014938488578, + "grad_norm": 1.1089187860488892, + "learning_rate": 4.216126970868836e-05, + "loss": 0.1536, + "step": 20260 + }, + { + "epoch": 1.670749121265378, + "grad_norm": 0.504047691822052, + "learning_rate": 4.214034614941811e-05, + "loss": 0.1758, + "step": 20280 + }, + { + "epoch": 1.672396748681898, + "grad_norm": 0.8742786049842834, + "learning_rate": 4.211939990926355e-05, + "loss": 0.1843, + "step": 20300 + }, + { + "epoch": 1.6740443760984183, + "grad_norm": 0.856685221195221, + "learning_rate": 4.209843101594171e-05, + "loss": 0.1708, + "step": 20320 + }, + { + "epoch": 1.6756920035149385, + "grad_norm": 0.20526465773582458, + "learning_rate": 4.2077439497199576e-05, + "loss": 0.1699, + "step": 20340 + }, + { + "epoch": 1.6773396309314585, + "grad_norm": 0.9350545406341553, + "learning_rate": 4.2056425380814074e-05, + "loss": 0.1503, + "step": 20360 + }, + { + "epoch": 1.6789872583479788, + "grad_norm": 2.845334529876709, + "learning_rate": 4.203538869459204e-05, + "loss": 0.1708, + "step": 20380 + }, + { + "epoch": 1.680634885764499, + "grad_norm": 1.2838850021362305, + "learning_rate": 4.2014329466370176e-05, + "loss": 0.18, + "step": 20400 + }, + { + "epoch": 1.6822825131810193, + "grad_norm": 1.8057399988174438, + "learning_rate": 4.1993247724015e-05, + "loss": 0.173, + "step": 20420 + }, + { + "epoch": 1.6839301405975395, + "grad_norm": 0.3127855658531189, + "learning_rate": 4.1972143495422844e-05, + "loss": 0.1728, + "step": 20440 + }, + { + "epoch": 1.6855777680140598, + "grad_norm": 2.700399398803711, + "learning_rate": 4.195101680851977e-05, + "loss": 0.1578, + "step": 20460 + }, + { + "epoch": 1.68722539543058, + "grad_norm": 0.7250802516937256, + "learning_rate": 4.192986769126158e-05, + "loss": 0.1823, + "step": 20480 + }, + { + "epoch": 1.6888730228471003, + "grad_norm": 1.1070244312286377, + "learning_rate": 4.1908696171633734e-05, + "loss": 0.189, + "step": 20500 + }, + { + "epoch": 1.6905206502636205, + "grad_norm": 1.0246353149414062, + "learning_rate": 4.188750227765135e-05, + "loss": 0.1902, + "step": 20520 + }, + { + "epoch": 1.6921682776801406, + "grad_norm": 0.6496437191963196, + "learning_rate": 4.186628603735916e-05, + "loss": 0.1747, + "step": 20540 + }, + { + "epoch": 1.6938159050966608, + "grad_norm": 0.8663806319236755, + "learning_rate": 4.184504747883145e-05, + "loss": 0.1519, + "step": 20560 + }, + { + "epoch": 1.695463532513181, + "grad_norm": 0.7036964893341064, + "learning_rate": 4.182378663017205e-05, + "loss": 0.1829, + "step": 20580 + }, + { + "epoch": 1.697111159929701, + "grad_norm": 0.7158936858177185, + "learning_rate": 4.180250351951427e-05, + "loss": 0.1672, + "step": 20600 + }, + { + "epoch": 1.6987587873462213, + "grad_norm": 0.5382769703865051, + "learning_rate": 4.17811981750209e-05, + "loss": 0.1804, + "step": 20620 + }, + { + "epoch": 1.7004064147627416, + "grad_norm": 0.5654181838035583, + "learning_rate": 4.175987062488413e-05, + "loss": 0.1585, + "step": 20640 + }, + { + "epoch": 1.7020540421792618, + "grad_norm": 1.4323805570602417, + "learning_rate": 4.1738520897325544e-05, + "loss": 0.1532, + "step": 20660 + }, + { + "epoch": 1.703701669595782, + "grad_norm": 0.7951692938804626, + "learning_rate": 4.171714902059608e-05, + "loss": 0.1755, + "step": 20680 + }, + { + "epoch": 1.7053492970123023, + "grad_norm": 0.7378861308097839, + "learning_rate": 4.169575502297597e-05, + "loss": 0.1814, + "step": 20700 + }, + { + "epoch": 1.7069969244288226, + "grad_norm": 0.6902081966400146, + "learning_rate": 4.1674338932774724e-05, + "loss": 0.1639, + "step": 20720 + }, + { + "epoch": 1.7086445518453428, + "grad_norm": 0.16693931818008423, + "learning_rate": 4.1652900778331076e-05, + "loss": 0.1788, + "step": 20740 + }, + { + "epoch": 1.710292179261863, + "grad_norm": 2.9214699268341064, + "learning_rate": 4.1631440588012974e-05, + "loss": 0.1562, + "step": 20760 + }, + { + "epoch": 1.711939806678383, + "grad_norm": 0.8093401193618774, + "learning_rate": 4.1609958390217516e-05, + "loss": 0.1714, + "step": 20780 + }, + { + "epoch": 1.7135874340949033, + "grad_norm": 1.0199005603790283, + "learning_rate": 4.158845421337093e-05, + "loss": 0.1869, + "step": 20800 + }, + { + "epoch": 1.7152350615114236, + "grad_norm": 0.5902179479598999, + "learning_rate": 4.1566928085928504e-05, + "loss": 0.1868, + "step": 20820 + }, + { + "epoch": 1.7168826889279436, + "grad_norm": 0.30310845375061035, + "learning_rate": 4.154538003637459e-05, + "loss": 0.1851, + "step": 20840 + }, + { + "epoch": 1.7185303163444638, + "grad_norm": 4.026705741882324, + "learning_rate": 4.1523810093222547e-05, + "loss": 0.1667, + "step": 20860 + }, + { + "epoch": 1.720177943760984, + "grad_norm": 0.82432621717453, + "learning_rate": 4.1502218285014714e-05, + "loss": 0.1785, + "step": 20880 + }, + { + "epoch": 1.7218255711775043, + "grad_norm": 1.0857402086257935, + "learning_rate": 4.148060464032234e-05, + "loss": 0.1867, + "step": 20900 + }, + { + "epoch": 1.7234731985940246, + "grad_norm": 1.142042636871338, + "learning_rate": 4.145896918774558e-05, + "loss": 0.1744, + "step": 20920 + }, + { + "epoch": 1.7251208260105448, + "grad_norm": 0.4160640835762024, + "learning_rate": 4.143731195591345e-05, + "loss": 0.1801, + "step": 20940 + }, + { + "epoch": 1.726768453427065, + "grad_norm": 1.8969520330429077, + "learning_rate": 4.141563297348378e-05, + "loss": 0.1536, + "step": 20960 + }, + { + "epoch": 1.7284160808435853, + "grad_norm": 0.7605903744697571, + "learning_rate": 4.1393932269143206e-05, + "loss": 0.1777, + "step": 20980 + }, + { + "epoch": 1.7300637082601056, + "grad_norm": 1.4995478391647339, + "learning_rate": 4.137220987160705e-05, + "loss": 0.184, + "step": 21000 + }, + { + "epoch": 1.7317113356766256, + "grad_norm": 1.7079527378082275, + "learning_rate": 4.13504658096194e-05, + "loss": 0.1799, + "step": 21020 + }, + { + "epoch": 1.7333589630931459, + "grad_norm": 0.3677631616592407, + "learning_rate": 4.132870011195298e-05, + "loss": 0.1718, + "step": 21040 + }, + { + "epoch": 1.735006590509666, + "grad_norm": 5.215697288513184, + "learning_rate": 4.1306912807409134e-05, + "loss": 0.1515, + "step": 21060 + }, + { + "epoch": 1.7366542179261861, + "grad_norm": 2.6421728134155273, + "learning_rate": 4.1285103924817836e-05, + "loss": 0.1877, + "step": 21080 + }, + { + "epoch": 1.7383018453427064, + "grad_norm": 2.3778457641601562, + "learning_rate": 4.126327349303758e-05, + "loss": 0.1818, + "step": 21100 + }, + { + "epoch": 1.7399494727592266, + "grad_norm": 2.7478296756744385, + "learning_rate": 4.124142154095539e-05, + "loss": 0.1791, + "step": 21120 + }, + { + "epoch": 1.7415971001757469, + "grad_norm": 0.5207391381263733, + "learning_rate": 4.121954809748675e-05, + "loss": 0.1644, + "step": 21140 + }, + { + "epoch": 1.7432447275922671, + "grad_norm": 3.1791322231292725, + "learning_rate": 4.119765319157559e-05, + "loss": 0.1652, + "step": 21160 + }, + { + "epoch": 1.7448923550087874, + "grad_norm": 4.9425764083862305, + "learning_rate": 4.117573685219427e-05, + "loss": 0.1841, + "step": 21180 + }, + { + "epoch": 1.7465399824253076, + "grad_norm": 0.698239266872406, + "learning_rate": 4.115379910834345e-05, + "loss": 0.1775, + "step": 21200 + }, + { + "epoch": 1.7481876098418279, + "grad_norm": 3.20509934425354, + "learning_rate": 4.113183998905218e-05, + "loss": 0.1703, + "step": 21220 + }, + { + "epoch": 1.749835237258348, + "grad_norm": 0.2055424153804779, + "learning_rate": 4.110985952337775e-05, + "loss": 0.1679, + "step": 21240 + }, + { + "epoch": 1.7514828646748684, + "grad_norm": 2.4991447925567627, + "learning_rate": 4.108785774040571e-05, + "loss": 0.1528, + "step": 21260 + }, + { + "epoch": 1.7531304920913884, + "grad_norm": 1.0346381664276123, + "learning_rate": 4.106583466924981e-05, + "loss": 0.1735, + "step": 21280 + }, + { + "epoch": 1.7547781195079086, + "grad_norm": 1.605350136756897, + "learning_rate": 4.104379033905199e-05, + "loss": 0.1765, + "step": 21300 + }, + { + "epoch": 1.7564257469244289, + "grad_norm": 0.4309673309326172, + "learning_rate": 4.102172477898232e-05, + "loss": 0.1841, + "step": 21320 + }, + { + "epoch": 1.758073374340949, + "grad_norm": 0.5647534132003784, + "learning_rate": 4.099963801823892e-05, + "loss": 0.1606, + "step": 21340 + }, + { + "epoch": 1.7597210017574691, + "grad_norm": 2.7488176822662354, + "learning_rate": 4.0977530086048025e-05, + "loss": 0.1562, + "step": 21360 + }, + { + "epoch": 1.7613686291739894, + "grad_norm": 0.6944543719291687, + "learning_rate": 4.095540101166383e-05, + "loss": 0.1757, + "step": 21380 + }, + { + "epoch": 1.7630162565905096, + "grad_norm": 2.5942211151123047, + "learning_rate": 4.093325082436854e-05, + "loss": 0.1852, + "step": 21400 + }, + { + "epoch": 1.76466388400703, + "grad_norm": 0.6985669732093811, + "learning_rate": 4.09110795534723e-05, + "loss": 0.1801, + "step": 21420 + }, + { + "epoch": 1.7663115114235501, + "grad_norm": 0.5788651704788208, + "learning_rate": 4.088888722831312e-05, + "loss": 0.1841, + "step": 21440 + }, + { + "epoch": 1.7679591388400704, + "grad_norm": 1.5852128267288208, + "learning_rate": 4.0866673878256925e-05, + "loss": 0.1632, + "step": 21460 + }, + { + "epoch": 1.7696067662565906, + "grad_norm": 1.0001317262649536, + "learning_rate": 4.0844439532697386e-05, + "loss": 0.1828, + "step": 21480 + }, + { + "epoch": 1.7712543936731109, + "grad_norm": 3.1319148540496826, + "learning_rate": 4.0822184221056036e-05, + "loss": 0.1797, + "step": 21500 + }, + { + "epoch": 1.772902021089631, + "grad_norm": 1.2414968013763428, + "learning_rate": 4.0799907972782105e-05, + "loss": 0.172, + "step": 21520 + }, + { + "epoch": 1.7745496485061512, + "grad_norm": 0.37049078941345215, + "learning_rate": 4.077761081735252e-05, + "loss": 0.1836, + "step": 21540 + }, + { + "epoch": 1.7761972759226714, + "grad_norm": 2.273507595062256, + "learning_rate": 4.0755292784271896e-05, + "loss": 0.1663, + "step": 21560 + }, + { + "epoch": 1.7778449033391914, + "grad_norm": 1.669340968132019, + "learning_rate": 4.073295390307248e-05, + "loss": 0.1789, + "step": 21580 + }, + { + "epoch": 1.7794925307557117, + "grad_norm": 0.8448110222816467, + "learning_rate": 4.071059420331408e-05, + "loss": 0.1927, + "step": 21600 + }, + { + "epoch": 1.781140158172232, + "grad_norm": 2.4922406673431396, + "learning_rate": 4.0688213714584076e-05, + "loss": 0.1852, + "step": 21620 + }, + { + "epoch": 1.7827877855887522, + "grad_norm": 0.36666399240493774, + "learning_rate": 4.066581246649734e-05, + "loss": 0.1869, + "step": 21640 + }, + { + "epoch": 1.7844354130052724, + "grad_norm": 1.3697789907455444, + "learning_rate": 4.0643390488696225e-05, + "loss": 0.159, + "step": 21660 + }, + { + "epoch": 1.7860830404217927, + "grad_norm": 0.6621468663215637, + "learning_rate": 4.062094781085051e-05, + "loss": 0.1827, + "step": 21680 + }, + { + "epoch": 1.787730667838313, + "grad_norm": 3.3689136505126953, + "learning_rate": 4.059848446265736e-05, + "loss": 0.1829, + "step": 21700 + }, + { + "epoch": 1.7893782952548332, + "grad_norm": 1.2748357057571411, + "learning_rate": 4.0576000473841295e-05, + "loss": 0.1814, + "step": 21720 + }, + { + "epoch": 1.7910259226713534, + "grad_norm": 0.7753732800483704, + "learning_rate": 4.0553495874154167e-05, + "loss": 0.1837, + "step": 21740 + }, + { + "epoch": 1.7926735500878734, + "grad_norm": 4.699821472167969, + "learning_rate": 4.053097069337507e-05, + "loss": 0.1608, + "step": 21760 + }, + { + "epoch": 1.7943211775043937, + "grad_norm": 0.6081763505935669, + "learning_rate": 4.050842496131034e-05, + "loss": 0.1802, + "step": 21780 + }, + { + "epoch": 1.795968804920914, + "grad_norm": 1.5527398586273193, + "learning_rate": 4.0485858707793534e-05, + "loss": 0.186, + "step": 21800 + }, + { + "epoch": 1.797616432337434, + "grad_norm": 0.7068532109260559, + "learning_rate": 4.046327196268533e-05, + "loss": 0.1653, + "step": 21820 + }, + { + "epoch": 1.7992640597539542, + "grad_norm": 1.8269671201705933, + "learning_rate": 4.044066475587353e-05, + "loss": 0.1799, + "step": 21840 + }, + { + "epoch": 1.8009116871704745, + "grad_norm": 3.3052139282226562, + "learning_rate": 4.041803711727303e-05, + "loss": 0.1502, + "step": 21860 + }, + { + "epoch": 1.8025593145869947, + "grad_norm": 0.9614394307136536, + "learning_rate": 4.039538907682574e-05, + "loss": 0.1739, + "step": 21880 + }, + { + "epoch": 1.804206942003515, + "grad_norm": 1.014877438545227, + "learning_rate": 4.0372720664500575e-05, + "loss": 0.1829, + "step": 21900 + }, + { + "epoch": 1.8058545694200352, + "grad_norm": 1.322003960609436, + "learning_rate": 4.035003191029341e-05, + "loss": 0.1935, + "step": 21920 + }, + { + "epoch": 1.8075021968365554, + "grad_norm": 0.6618562340736389, + "learning_rate": 4.0328458779473224e-05, + "loss": 0.1741, + "step": 21940 + }, + { + "epoch": 1.8091498242530757, + "grad_norm": 1.402084469795227, + "learning_rate": 4.0305730444973686e-05, + "loss": 0.1621, + "step": 21960 + }, + { + "epoch": 1.810797451669596, + "grad_norm": 1.3006558418273926, + "learning_rate": 4.0282981857236654e-05, + "loss": 0.1674, + "step": 21980 + }, + { + "epoch": 1.812445079086116, + "grad_norm": 1.6449670791625977, + "learning_rate": 4.026021304636408e-05, + "loss": 0.1845, + "step": 22000 + }, + { + "epoch": 1.812445079086116, + "eval_loss": 0.5395076274871826, + "eval_runtime": 246.9697, + "eval_samples_per_second": 88.225, + "eval_steps_per_second": 22.059, + "eval_wer": 0.2258113279380107, + "step": 22000 + }, + { + "epoch": 1.8140927065026362, + "grad_norm": 0.5984182953834534, + "learning_rate": 4.0237424042484694e-05, + "loss": 0.18, + "step": 22020 + }, + { + "epoch": 1.8157403339191565, + "grad_norm": 1.0855306386947632, + "learning_rate": 4.0214614875753955e-05, + "loss": 0.1651, + "step": 22040 + }, + { + "epoch": 1.8173879613356765, + "grad_norm": 1.4949283599853516, + "learning_rate": 4.0191785576353956e-05, + "loss": 0.1563, + "step": 22060 + }, + { + "epoch": 1.8190355887521967, + "grad_norm": 0.9411268830299377, + "learning_rate": 4.016893617449349e-05, + "loss": 0.1807, + "step": 22080 + }, + { + "epoch": 1.820683216168717, + "grad_norm": 3.1195433139801025, + "learning_rate": 4.014606670040791e-05, + "loss": 0.1611, + "step": 22100 + }, + { + "epoch": 1.8223308435852372, + "grad_norm": 1.749937653541565, + "learning_rate": 4.012317718435915e-05, + "loss": 0.1809, + "step": 22120 + }, + { + "epoch": 1.8239784710017575, + "grad_norm": 0.41568905115127563, + "learning_rate": 4.010026765663566e-05, + "loss": 0.1754, + "step": 22140 + }, + { + "epoch": 1.8256260984182777, + "grad_norm": 1.6782418489456177, + "learning_rate": 4.007733814755235e-05, + "loss": 0.1577, + "step": 22160 + }, + { + "epoch": 1.827273725834798, + "grad_norm": 1.470269799232483, + "learning_rate": 4.005438868745059e-05, + "loss": 0.1804, + "step": 22180 + }, + { + "epoch": 1.8289213532513182, + "grad_norm": 0.8983897566795349, + "learning_rate": 4.003141930669818e-05, + "loss": 0.1771, + "step": 22200 + }, + { + "epoch": 1.8305689806678385, + "grad_norm": 0.6398609280586243, + "learning_rate": 4.000843003568921e-05, + "loss": 0.1937, + "step": 22220 + }, + { + "epoch": 1.8322166080843585, + "grad_norm": 0.20628367364406586, + "learning_rate": 3.998542090484414e-05, + "loss": 0.1804, + "step": 22240 + }, + { + "epoch": 1.8338642355008787, + "grad_norm": 1.640540361404419, + "learning_rate": 3.9962391944609704e-05, + "loss": 0.1576, + "step": 22260 + }, + { + "epoch": 1.835511862917399, + "grad_norm": 0.5554460883140564, + "learning_rate": 3.9939343185458856e-05, + "loss": 0.1629, + "step": 22280 + }, + { + "epoch": 1.837159490333919, + "grad_norm": 0.5767085552215576, + "learning_rate": 3.991627465789076e-05, + "loss": 0.1779, + "step": 22300 + }, + { + "epoch": 1.8388071177504393, + "grad_norm": 0.3550276458263397, + "learning_rate": 3.989318639243075e-05, + "loss": 0.1812, + "step": 22320 + }, + { + "epoch": 1.8404547451669595, + "grad_norm": 0.23671790957450867, + "learning_rate": 3.9870078419630276e-05, + "loss": 0.1717, + "step": 22340 + }, + { + "epoch": 1.8421023725834798, + "grad_norm": 1.2688475847244263, + "learning_rate": 3.984695077006683e-05, + "loss": 0.1606, + "step": 22360 + }, + { + "epoch": 1.84375, + "grad_norm": 2.13102388381958, + "learning_rate": 3.982380347434398e-05, + "loss": 0.1652, + "step": 22380 + }, + { + "epoch": 1.8453976274165202, + "grad_norm": 0.6820785403251648, + "learning_rate": 3.980063656309128e-05, + "loss": 0.1832, + "step": 22400 + }, + { + "epoch": 1.8470452548330405, + "grad_norm": 1.0797230005264282, + "learning_rate": 3.9777450066964236e-05, + "loss": 0.1673, + "step": 22420 + }, + { + "epoch": 1.8486928822495607, + "grad_norm": 0.22749900817871094, + "learning_rate": 3.9754244016644286e-05, + "loss": 0.1935, + "step": 22440 + }, + { + "epoch": 1.850340509666081, + "grad_norm": 21.42913246154785, + "learning_rate": 3.973101844283871e-05, + "loss": 0.1616, + "step": 22460 + }, + { + "epoch": 1.851988137082601, + "grad_norm": 1.0109410285949707, + "learning_rate": 3.970777337628066e-05, + "loss": 0.1733, + "step": 22480 + }, + { + "epoch": 1.8536357644991213, + "grad_norm": 1.8864178657531738, + "learning_rate": 3.968450884772905e-05, + "loss": 0.191, + "step": 22500 + }, + { + "epoch": 1.8552833919156415, + "grad_norm": 1.0088257789611816, + "learning_rate": 3.9661224887968564e-05, + "loss": 0.179, + "step": 22520 + }, + { + "epoch": 1.8569310193321615, + "grad_norm": 0.27392879128456116, + "learning_rate": 3.9637921527809594e-05, + "loss": 0.1883, + "step": 22540 + }, + { + "epoch": 1.8585786467486818, + "grad_norm": 1.0473995208740234, + "learning_rate": 3.9614598798088205e-05, + "loss": 0.1568, + "step": 22560 + }, + { + "epoch": 1.860226274165202, + "grad_norm": 2.516446352005005, + "learning_rate": 3.959125672966609e-05, + "loss": 0.179, + "step": 22580 + }, + { + "epoch": 1.8618739015817223, + "grad_norm": 0.5339066982269287, + "learning_rate": 3.956789535343054e-05, + "loss": 0.1886, + "step": 22600 + }, + { + "epoch": 1.8635215289982425, + "grad_norm": 3.4075698852539062, + "learning_rate": 3.954451470029439e-05, + "loss": 0.1723, + "step": 22620 + }, + { + "epoch": 1.8651691564147628, + "grad_norm": 0.4262460172176361, + "learning_rate": 3.952111480119597e-05, + "loss": 0.1774, + "step": 22640 + }, + { + "epoch": 1.866816783831283, + "grad_norm": 1.0418535470962524, + "learning_rate": 3.94976956870991e-05, + "loss": 0.1613, + "step": 22660 + }, + { + "epoch": 1.8684644112478033, + "grad_norm": 1.1704654693603516, + "learning_rate": 3.947425738899302e-05, + "loss": 0.1879, + "step": 22680 + }, + { + "epoch": 1.8701120386643235, + "grad_norm": 1.1728986501693726, + "learning_rate": 3.9450799937892324e-05, + "loss": 0.1655, + "step": 22700 + }, + { + "epoch": 1.8717596660808435, + "grad_norm": 1.5375456809997559, + "learning_rate": 3.9427323364837e-05, + "loss": 0.1865, + "step": 22720 + }, + { + "epoch": 1.8734072934973638, + "grad_norm": 0.16443948447704315, + "learning_rate": 3.940382770089232e-05, + "loss": 0.179, + "step": 22740 + }, + { + "epoch": 1.875054920913884, + "grad_norm": 13.896778106689453, + "learning_rate": 3.9380312977148795e-05, + "loss": 0.1495, + "step": 22760 + }, + { + "epoch": 1.876702548330404, + "grad_norm": 2.690234422683716, + "learning_rate": 3.935677922472218e-05, + "loss": 0.1715, + "step": 22780 + }, + { + "epoch": 1.8783501757469243, + "grad_norm": 2.224005937576294, + "learning_rate": 3.933322647475342e-05, + "loss": 0.1753, + "step": 22800 + }, + { + "epoch": 1.8799978031634446, + "grad_norm": 1.617680311203003, + "learning_rate": 3.9309654758408556e-05, + "loss": 0.1773, + "step": 22820 + }, + { + "epoch": 1.8816454305799648, + "grad_norm": 0.17472465336322784, + "learning_rate": 3.928606410687877e-05, + "loss": 0.174, + "step": 22840 + }, + { + "epoch": 1.883293057996485, + "grad_norm": 2.1024489402770996, + "learning_rate": 3.926245455138029e-05, + "loss": 0.1486, + "step": 22860 + }, + { + "epoch": 1.8849406854130053, + "grad_norm": 3.535928726196289, + "learning_rate": 3.923882612315434e-05, + "loss": 0.1917, + "step": 22880 + }, + { + "epoch": 1.8865883128295255, + "grad_norm": 1.4998210668563843, + "learning_rate": 3.9215178853467135e-05, + "loss": 0.1786, + "step": 22900 + }, + { + "epoch": 1.8882359402460458, + "grad_norm": 1.4746694564819336, + "learning_rate": 3.9191512773609806e-05, + "loss": 0.1731, + "step": 22920 + }, + { + "epoch": 1.889883567662566, + "grad_norm": 0.23350854218006134, + "learning_rate": 3.91678279148984e-05, + "loss": 0.1719, + "step": 22940 + }, + { + "epoch": 1.891531195079086, + "grad_norm": 1.7336878776550293, + "learning_rate": 3.9144124308673784e-05, + "loss": 0.1495, + "step": 22960 + }, + { + "epoch": 1.8931788224956063, + "grad_norm": 1.0310170650482178, + "learning_rate": 3.912040198630166e-05, + "loss": 0.1682, + "step": 22980 + }, + { + "epoch": 1.8948264499121266, + "grad_norm": 0.8213064074516296, + "learning_rate": 3.9096660979172464e-05, + "loss": 0.1812, + "step": 23000 + }, + { + "epoch": 1.8964740773286466, + "grad_norm": 0.4071774184703827, + "learning_rate": 3.907290131870139e-05, + "loss": 0.1697, + "step": 23020 + }, + { + "epoch": 1.8981217047451668, + "grad_norm": 0.33335015177726746, + "learning_rate": 3.9049123036328296e-05, + "loss": 0.1645, + "step": 23040 + }, + { + "epoch": 1.899769332161687, + "grad_norm": 1.0531188249588013, + "learning_rate": 3.9025326163517676e-05, + "loss": 0.1484, + "step": 23060 + }, + { + "epoch": 1.9014169595782073, + "grad_norm": 1.8591293096542358, + "learning_rate": 3.900151073175864e-05, + "loss": 0.1911, + "step": 23080 + }, + { + "epoch": 1.9030645869947276, + "grad_norm": 1.4219475984573364, + "learning_rate": 3.897767677256485e-05, + "loss": 0.1809, + "step": 23100 + }, + { + "epoch": 1.9047122144112478, + "grad_norm": 1.1440541744232178, + "learning_rate": 3.8955017379019396e-05, + "loss": 0.174, + "step": 23120 + }, + { + "epoch": 1.906359841827768, + "grad_norm": 0.2897416353225708, + "learning_rate": 3.8931147382061806e-05, + "loss": 0.176, + "step": 23140 + }, + { + "epoch": 1.9080074692442883, + "grad_norm": 1.9240251779556274, + "learning_rate": 3.890725895077744e-05, + "loss": 0.1549, + "step": 23160 + }, + { + "epoch": 1.9096550966608086, + "grad_norm": 0.8471350073814392, + "learning_rate": 3.888335211677656e-05, + "loss": 0.1813, + "step": 23180 + }, + { + "epoch": 1.9113027240773286, + "grad_norm": 1.2786263227462769, + "learning_rate": 3.8859426911693756e-05, + "loss": 0.1754, + "step": 23200 + }, + { + "epoch": 1.9129503514938488, + "grad_norm": 2.0185673236846924, + "learning_rate": 3.8835483367187974e-05, + "loss": 0.1832, + "step": 23220 + }, + { + "epoch": 1.914597978910369, + "grad_norm": 0.2411476969718933, + "learning_rate": 3.8811521514942374e-05, + "loss": 0.1714, + "step": 23240 + }, + { + "epoch": 1.9162456063268891, + "grad_norm": 1.0712230205535889, + "learning_rate": 3.8787541386664374e-05, + "loss": 0.1563, + "step": 23260 + }, + { + "epoch": 1.9178932337434094, + "grad_norm": 0.9587976932525635, + "learning_rate": 3.876354301408558e-05, + "loss": 0.178, + "step": 23280 + }, + { + "epoch": 1.9195408611599296, + "grad_norm": 1.298622965812683, + "learning_rate": 3.873952642896171e-05, + "loss": 0.1695, + "step": 23300 + }, + { + "epoch": 1.9211884885764499, + "grad_norm": 4.192885398864746, + "learning_rate": 3.871549166307261e-05, + "loss": 0.1721, + "step": 23320 + }, + { + "epoch": 1.92283611599297, + "grad_norm": 0.1895458996295929, + "learning_rate": 3.869143874822218e-05, + "loss": 0.1775, + "step": 23340 + }, + { + "epoch": 1.9244837434094904, + "grad_norm": 1.721372127532959, + "learning_rate": 3.8667367716238326e-05, + "loss": 0.1597, + "step": 23360 + }, + { + "epoch": 1.9261313708260106, + "grad_norm": 4.6658525466918945, + "learning_rate": 3.864327859897291e-05, + "loss": 0.1728, + "step": 23380 + }, + { + "epoch": 1.9277789982425309, + "grad_norm": 0.3837490379810333, + "learning_rate": 3.861917142830177e-05, + "loss": 0.1788, + "step": 23400 + }, + { + "epoch": 1.929426625659051, + "grad_norm": 0.4308772087097168, + "learning_rate": 3.859504623612461e-05, + "loss": 0.1747, + "step": 23420 + }, + { + "epoch": 1.9310742530755711, + "grad_norm": 0.3830674886703491, + "learning_rate": 3.857090305436496e-05, + "loss": 0.1703, + "step": 23440 + }, + { + "epoch": 1.9327218804920914, + "grad_norm": 0.9429712891578674, + "learning_rate": 3.854674191497019e-05, + "loss": 0.1552, + "step": 23460 + }, + { + "epoch": 1.9343695079086116, + "grad_norm": 0.5584468245506287, + "learning_rate": 3.85225628499114e-05, + "loss": 0.1817, + "step": 23480 + }, + { + "epoch": 1.9360171353251316, + "grad_norm": 1.4192808866500854, + "learning_rate": 3.849836589118345e-05, + "loss": 0.1776, + "step": 23500 + }, + { + "epoch": 1.937664762741652, + "grad_norm": 0.453254371881485, + "learning_rate": 3.847415107080485e-05, + "loss": 0.1721, + "step": 23520 + }, + { + "epoch": 1.9393123901581721, + "grad_norm": 0.37701448798179626, + "learning_rate": 3.8449918420817734e-05, + "loss": 0.1724, + "step": 23540 + }, + { + "epoch": 1.9409600175746924, + "grad_norm": 0.9709746241569519, + "learning_rate": 3.8425667973287847e-05, + "loss": 0.1512, + "step": 23560 + }, + { + "epoch": 1.9426076449912126, + "grad_norm": 3.227597951889038, + "learning_rate": 3.8401399760304504e-05, + "loss": 0.1702, + "step": 23580 + }, + { + "epoch": 1.9442552724077329, + "grad_norm": 1.5262757539749146, + "learning_rate": 3.837711381398048e-05, + "loss": 0.1773, + "step": 23600 + }, + { + "epoch": 1.9459028998242531, + "grad_norm": 0.9164407849311829, + "learning_rate": 3.835281016645206e-05, + "loss": 0.168, + "step": 23620 + }, + { + "epoch": 1.9475505272407734, + "grad_norm": 0.3824036121368408, + "learning_rate": 3.832848884987892e-05, + "loss": 0.178, + "step": 23640 + }, + { + "epoch": 1.9491981546572936, + "grad_norm": 0.8727893233299255, + "learning_rate": 3.8304149896444145e-05, + "loss": 0.1525, + "step": 23660 + }, + { + "epoch": 1.9508457820738139, + "grad_norm": 1.189348578453064, + "learning_rate": 3.827979333835413e-05, + "loss": 0.1655, + "step": 23680 + }, + { + "epoch": 1.952493409490334, + "grad_norm": 0.6089633107185364, + "learning_rate": 3.825541920783858e-05, + "loss": 0.179, + "step": 23700 + }, + { + "epoch": 1.9541410369068541, + "grad_norm": 5.746613025665283, + "learning_rate": 3.8231027537150464e-05, + "loss": 0.1729, + "step": 23720 + }, + { + "epoch": 1.9557886643233744, + "grad_norm": 0.3469756841659546, + "learning_rate": 3.820661835856594e-05, + "loss": 0.2064, + "step": 23740 + }, + { + "epoch": 1.9574362917398944, + "grad_norm": 1.2260953187942505, + "learning_rate": 3.818219170438434e-05, + "loss": 0.1463, + "step": 23760 + }, + { + "epoch": 1.9590839191564147, + "grad_norm": 0.5917554497718811, + "learning_rate": 3.815774760692813e-05, + "loss": 0.1633, + "step": 23780 + }, + { + "epoch": 1.960731546572935, + "grad_norm": 0.7345995903015137, + "learning_rate": 3.8133286098542835e-05, + "loss": 0.1662, + "step": 23800 + }, + { + "epoch": 1.9623791739894552, + "grad_norm": 0.6015161275863647, + "learning_rate": 3.810880721159704e-05, + "loss": 0.1782, + "step": 23820 + }, + { + "epoch": 1.9640268014059754, + "grad_norm": 0.1269834190607071, + "learning_rate": 3.808431097848233e-05, + "loss": 0.177, + "step": 23840 + }, + { + "epoch": 1.9656744288224957, + "grad_norm": 3.330901861190796, + "learning_rate": 3.805979743161322e-05, + "loss": 0.1551, + "step": 23860 + }, + { + "epoch": 1.967322056239016, + "grad_norm": 1.1339266300201416, + "learning_rate": 3.8035266603427144e-05, + "loss": 0.1705, + "step": 23880 + }, + { + "epoch": 1.9689696836555362, + "grad_norm": 0.560189962387085, + "learning_rate": 3.801071852638443e-05, + "loss": 0.1862, + "step": 23900 + }, + { + "epoch": 1.9706173110720564, + "grad_norm": 0.4056016206741333, + "learning_rate": 3.798615323296819e-05, + "loss": 0.1811, + "step": 23920 + }, + { + "epoch": 1.9722649384885764, + "grad_norm": 0.24118812382221222, + "learning_rate": 3.796157075568433e-05, + "loss": 0.1726, + "step": 23940 + }, + { + "epoch": 1.9739125659050967, + "grad_norm": 1.474798560142517, + "learning_rate": 3.793697112706153e-05, + "loss": 0.1446, + "step": 23960 + }, + { + "epoch": 1.975560193321617, + "grad_norm": 1.19919753074646, + "learning_rate": 3.791235437965109e-05, + "loss": 0.1664, + "step": 23980 + }, + { + "epoch": 1.977207820738137, + "grad_norm": 0.38849976658821106, + "learning_rate": 3.788772054602705e-05, + "loss": 0.1769, + "step": 24000 + }, + { + "epoch": 1.977207820738137, + "eval_loss": 0.5249012112617493, + "eval_runtime": 254.2293, + "eval_samples_per_second": 85.706, + "eval_steps_per_second": 21.429, + "eval_wer": 0.22245208175285466, + "step": 24000 + }, + { + "epoch": 1.9788554481546572, + "grad_norm": 0.323760062456131, + "learning_rate": 3.7863069658786e-05, + "loss": 0.1633, + "step": 24020 + }, + { + "epoch": 1.9805030755711774, + "grad_norm": 0.3223556876182556, + "learning_rate": 3.7838401750547115e-05, + "loss": 0.1713, + "step": 24040 + }, + { + "epoch": 1.9821507029876977, + "grad_norm": 1.7246328592300415, + "learning_rate": 3.7813716853952086e-05, + "loss": 0.1459, + "step": 24060 + }, + { + "epoch": 1.983798330404218, + "grad_norm": 0.7763593196868896, + "learning_rate": 3.77890150016651e-05, + "loss": 0.1636, + "step": 24080 + }, + { + "epoch": 1.9854459578207382, + "grad_norm": 0.703742265701294, + "learning_rate": 3.7764296226372756e-05, + "loss": 0.1739, + "step": 24100 + }, + { + "epoch": 1.9870935852372584, + "grad_norm": 0.5700841546058655, + "learning_rate": 3.773956056078407e-05, + "loss": 0.1784, + "step": 24120 + }, + { + "epoch": 1.9887412126537787, + "grad_norm": 0.1853460967540741, + "learning_rate": 3.7714808037630396e-05, + "loss": 0.173, + "step": 24140 + }, + { + "epoch": 1.990388840070299, + "grad_norm": 0.5716795921325684, + "learning_rate": 3.76900386896654e-05, + "loss": 0.1502, + "step": 24160 + }, + { + "epoch": 1.992036467486819, + "grad_norm": 0.789965033531189, + "learning_rate": 3.766525254966499e-05, + "loss": 0.1886, + "step": 24180 + }, + { + "epoch": 1.9936840949033392, + "grad_norm": 2.330145835876465, + "learning_rate": 3.764044965042734e-05, + "loss": 0.1851, + "step": 24200 + }, + { + "epoch": 1.9953317223198594, + "grad_norm": 4.897230625152588, + "learning_rate": 3.761563002477274e-05, + "loss": 0.1855, + "step": 24220 + }, + { + "epoch": 1.9969793497363795, + "grad_norm": 0.2082502692937851, + "learning_rate": 3.759203591747027e-05, + "loss": 0.1807, + "step": 24240 + }, + { + "epoch": 1.9986269771528997, + "grad_norm": 1.0545985698699951, + "learning_rate": 3.7567183769785896e-05, + "loss": 0.1403, + "step": 24260 + }, + { + "epoch": 2.00027460456942, + "grad_norm": 0.38820046186447144, + "learning_rate": 3.754231499263332e-05, + "loss": 0.1762, + "step": 24280 + }, + { + "epoch": 2.00192223198594, + "grad_norm": 1.5918121337890625, + "learning_rate": 3.751742961892006e-05, + "loss": 0.154, + "step": 24300 + }, + { + "epoch": 2.0035698594024605, + "grad_norm": 1.0402336120605469, + "learning_rate": 3.749252768157555e-05, + "loss": 0.1637, + "step": 24320 + }, + { + "epoch": 2.0052174868189807, + "grad_norm": 0.6094750761985779, + "learning_rate": 3.746760921355119e-05, + "loss": 0.1579, + "step": 24340 + }, + { + "epoch": 2.006865114235501, + "grad_norm": 0.7969854474067688, + "learning_rate": 3.744267424782019e-05, + "loss": 0.1586, + "step": 24360 + }, + { + "epoch": 2.008512741652021, + "grad_norm": 0.7175785899162292, + "learning_rate": 3.741772281737766e-05, + "loss": 0.1691, + "step": 24380 + }, + { + "epoch": 2.0101603690685415, + "grad_norm": 2.261885643005371, + "learning_rate": 3.739275495524046e-05, + "loss": 0.1655, + "step": 24400 + }, + { + "epoch": 2.0118079964850617, + "grad_norm": 0.9677129983901978, + "learning_rate": 3.736777069444719e-05, + "loss": 0.1628, + "step": 24420 + }, + { + "epoch": 2.0134556239015815, + "grad_norm": 0.5170992016792297, + "learning_rate": 3.7342770068058164e-05, + "loss": 0.1736, + "step": 24440 + }, + { + "epoch": 2.0151032513181018, + "grad_norm": 1.3504503965377808, + "learning_rate": 3.731775310915534e-05, + "loss": 0.1666, + "step": 24460 + }, + { + "epoch": 2.016750878734622, + "grad_norm": 0.5912557244300842, + "learning_rate": 3.729271985084231e-05, + "loss": 0.1591, + "step": 24480 + }, + { + "epoch": 2.0183985061511422, + "grad_norm": 2.212615728378296, + "learning_rate": 3.72676703262442e-05, + "loss": 0.1522, + "step": 24500 + }, + { + "epoch": 2.0200461335676625, + "grad_norm": 0.7549368143081665, + "learning_rate": 3.724260456850769e-05, + "loss": 0.162, + "step": 24520 + }, + { + "epoch": 2.0216937609841827, + "grad_norm": 0.7109091281890869, + "learning_rate": 3.721752261080093e-05, + "loss": 0.1605, + "step": 24540 + }, + { + "epoch": 2.023341388400703, + "grad_norm": 0.46456268429756165, + "learning_rate": 3.719242448631349e-05, + "loss": 0.1597, + "step": 24560 + }, + { + "epoch": 2.0249890158172232, + "grad_norm": 0.6666226387023926, + "learning_rate": 3.716731022825636e-05, + "loss": 0.1527, + "step": 24580 + }, + { + "epoch": 2.0266366432337435, + "grad_norm": 1.9217432737350464, + "learning_rate": 3.714217986986187e-05, + "loss": 0.1519, + "step": 24600 + }, + { + "epoch": 2.0282842706502637, + "grad_norm": 0.7165806889533997, + "learning_rate": 3.711703344438364e-05, + "loss": 0.1475, + "step": 24620 + }, + { + "epoch": 2.029931898066784, + "grad_norm": 3.279149055480957, + "learning_rate": 3.7091870985096574e-05, + "loss": 0.1805, + "step": 24640 + }, + { + "epoch": 2.0315795254833042, + "grad_norm": 0.8710042238235474, + "learning_rate": 3.7066692525296766e-05, + "loss": 0.1779, + "step": 24660 + }, + { + "epoch": 2.0332271528998245, + "grad_norm": 0.5144171714782715, + "learning_rate": 3.704149809830151e-05, + "loss": 0.1779, + "step": 24680 + }, + { + "epoch": 2.0348747803163443, + "grad_norm": 3.780592441558838, + "learning_rate": 3.701628773744921e-05, + "loss": 0.1646, + "step": 24700 + }, + { + "epoch": 2.0365224077328645, + "grad_norm": 2.9000730514526367, + "learning_rate": 3.699106147609936e-05, + "loss": 0.162, + "step": 24720 + }, + { + "epoch": 2.0381700351493848, + "grad_norm": 0.6916472315788269, + "learning_rate": 3.69658193476325e-05, + "loss": 0.1755, + "step": 24740 + }, + { + "epoch": 2.039817662565905, + "grad_norm": 1.2293349504470825, + "learning_rate": 3.694056138545015e-05, + "loss": 0.1679, + "step": 24760 + }, + { + "epoch": 2.0414652899824253, + "grad_norm": 0.5820478200912476, + "learning_rate": 3.691528762297481e-05, + "loss": 0.174, + "step": 24780 + }, + { + "epoch": 2.0431129173989455, + "grad_norm": 1.158288598060608, + "learning_rate": 3.6889998093649855e-05, + "loss": 0.159, + "step": 24800 + }, + { + "epoch": 2.0447605448154658, + "grad_norm": 1.0274357795715332, + "learning_rate": 3.686469283093954e-05, + "loss": 0.1652, + "step": 24820 + }, + { + "epoch": 2.046408172231986, + "grad_norm": 0.7709755897521973, + "learning_rate": 3.683937186832894e-05, + "loss": 0.1699, + "step": 24840 + }, + { + "epoch": 2.0480557996485063, + "grad_norm": 1.5248311758041382, + "learning_rate": 3.681403523932391e-05, + "loss": 0.161, + "step": 24860 + }, + { + "epoch": 2.0497034270650265, + "grad_norm": 0.45869287848472595, + "learning_rate": 3.6788682977451016e-05, + "loss": 0.1563, + "step": 24880 + }, + { + "epoch": 2.0513510544815468, + "grad_norm": 1.448834776878357, + "learning_rate": 3.6763315116257524e-05, + "loss": 0.1497, + "step": 24900 + }, + { + "epoch": 2.052998681898067, + "grad_norm": 0.9293612837791443, + "learning_rate": 3.6737931689311344e-05, + "loss": 0.1546, + "step": 24920 + }, + { + "epoch": 2.054646309314587, + "grad_norm": 1.0912532806396484, + "learning_rate": 3.671253273020098e-05, + "loss": 0.1598, + "step": 24940 + }, + { + "epoch": 2.056293936731107, + "grad_norm": 0.8708963990211487, + "learning_rate": 3.668711827253548e-05, + "loss": 0.1656, + "step": 24960 + }, + { + "epoch": 2.0579415641476273, + "grad_norm": 0.5793485045433044, + "learning_rate": 3.666168834994442e-05, + "loss": 0.1746, + "step": 24980 + }, + { + "epoch": 2.0595891915641475, + "grad_norm": 1.7373342514038086, + "learning_rate": 3.663624299607782e-05, + "loss": 0.1607, + "step": 25000 + }, + { + "epoch": 2.061236818980668, + "grad_norm": 0.9185656309127808, + "learning_rate": 3.661078224460613e-05, + "loss": 0.1708, + "step": 25020 + }, + { + "epoch": 2.062884446397188, + "grad_norm": 1.290109634399414, + "learning_rate": 3.6585306129220175e-05, + "loss": 0.1598, + "step": 25040 + }, + { + "epoch": 2.0645320738137083, + "grad_norm": 1.8787568807601929, + "learning_rate": 3.6559814683631096e-05, + "loss": 0.1665, + "step": 25060 + }, + { + "epoch": 2.0661797012302285, + "grad_norm": 0.3839080333709717, + "learning_rate": 3.653430794157035e-05, + "loss": 0.1739, + "step": 25080 + }, + { + "epoch": 2.067827328646749, + "grad_norm": 1.4831798076629639, + "learning_rate": 3.6508785936789604e-05, + "loss": 0.157, + "step": 25100 + }, + { + "epoch": 2.069474956063269, + "grad_norm": 2.0582001209259033, + "learning_rate": 3.648324870306073e-05, + "loss": 0.158, + "step": 25120 + }, + { + "epoch": 2.0711225834797893, + "grad_norm": 9.493792533874512, + "learning_rate": 3.645769627417577e-05, + "loss": 0.1641, + "step": 25140 + }, + { + "epoch": 2.0727702108963095, + "grad_norm": 0.920673668384552, + "learning_rate": 3.643212868394684e-05, + "loss": 0.161, + "step": 25160 + }, + { + "epoch": 2.0744178383128293, + "grad_norm": 1.4923367500305176, + "learning_rate": 3.6406545966206154e-05, + "loss": 0.1646, + "step": 25180 + }, + { + "epoch": 2.0760654657293496, + "grad_norm": 2.3396904468536377, + "learning_rate": 3.638094815480592e-05, + "loss": 0.1602, + "step": 25200 + }, + { + "epoch": 2.07771309314587, + "grad_norm": 1.5938606262207031, + "learning_rate": 3.635533528361832e-05, + "loss": 0.1858, + "step": 25220 + }, + { + "epoch": 2.07936072056239, + "grad_norm": 0.813642680644989, + "learning_rate": 3.632970738653546e-05, + "loss": 0.1671, + "step": 25240 + }, + { + "epoch": 2.0810083479789103, + "grad_norm": 0.6709133982658386, + "learning_rate": 3.630406449746935e-05, + "loss": 0.1515, + "step": 25260 + }, + { + "epoch": 2.0826559753954306, + "grad_norm": 0.36236572265625, + "learning_rate": 3.6278406650351834e-05, + "loss": 0.1755, + "step": 25280 + }, + { + "epoch": 2.084303602811951, + "grad_norm": 1.213829517364502, + "learning_rate": 3.6252733879134516e-05, + "loss": 0.158, + "step": 25300 + }, + { + "epoch": 2.085951230228471, + "grad_norm": 3.2344813346862793, + "learning_rate": 3.622704621778879e-05, + "loss": 0.1657, + "step": 25320 + }, + { + "epoch": 2.0875988576449913, + "grad_norm": 2.512044668197632, + "learning_rate": 3.6201343700305725e-05, + "loss": 0.1677, + "step": 25340 + }, + { + "epoch": 2.0892464850615116, + "grad_norm": 1.1710649728775024, + "learning_rate": 3.6175626360696076e-05, + "loss": 0.1755, + "step": 25360 + }, + { + "epoch": 2.090894112478032, + "grad_norm": 0.6621778607368469, + "learning_rate": 3.6149894232990176e-05, + "loss": 0.1591, + "step": 25380 + }, + { + "epoch": 2.092541739894552, + "grad_norm": 9.004714012145996, + "learning_rate": 3.612414735123795e-05, + "loss": 0.1559, + "step": 25400 + }, + { + "epoch": 2.094189367311072, + "grad_norm": 6.198091983795166, + "learning_rate": 3.609838574950885e-05, + "loss": 0.1685, + "step": 25420 + }, + { + "epoch": 2.095836994727592, + "grad_norm": 0.4765397012233734, + "learning_rate": 3.6072609461891785e-05, + "loss": 0.1748, + "step": 25440 + }, + { + "epoch": 2.0974846221441124, + "grad_norm": 3.423109531402588, + "learning_rate": 3.604681852249512e-05, + "loss": 0.1725, + "step": 25460 + }, + { + "epoch": 2.0991322495606326, + "grad_norm": 0.9131501317024231, + "learning_rate": 3.6021012965446594e-05, + "loss": 0.1718, + "step": 25480 + }, + { + "epoch": 2.100779876977153, + "grad_norm": 0.9718762636184692, + "learning_rate": 3.599519282489329e-05, + "loss": 0.1683, + "step": 25500 + }, + { + "epoch": 2.102427504393673, + "grad_norm": 0.8705006837844849, + "learning_rate": 3.596935813500159e-05, + "loss": 0.1727, + "step": 25520 + }, + { + "epoch": 2.1040751318101933, + "grad_norm": 0.5502598881721497, + "learning_rate": 3.5943508929957115e-05, + "loss": 0.1855, + "step": 25540 + }, + { + "epoch": 2.1057227592267136, + "grad_norm": 1.4949660301208496, + "learning_rate": 3.5917645243964726e-05, + "loss": 0.1705, + "step": 25560 + }, + { + "epoch": 2.107370386643234, + "grad_norm": 0.46673768758773804, + "learning_rate": 3.589176711124843e-05, + "loss": 0.1585, + "step": 25580 + }, + { + "epoch": 2.109018014059754, + "grad_norm": 1.2299778461456299, + "learning_rate": 3.586587456605133e-05, + "loss": 0.1541, + "step": 25600 + }, + { + "epoch": 2.1106656414762743, + "grad_norm": 0.7118710875511169, + "learning_rate": 3.5839967642635636e-05, + "loss": 0.1755, + "step": 25620 + }, + { + "epoch": 2.1123132688927946, + "grad_norm": 0.45549798011779785, + "learning_rate": 3.5814046375282554e-05, + "loss": 0.1689, + "step": 25640 + }, + { + "epoch": 2.1139608963093144, + "grad_norm": 1.2367688417434692, + "learning_rate": 3.578811079829227e-05, + "loss": 0.1671, + "step": 25660 + }, + { + "epoch": 2.1156085237258346, + "grad_norm": 0.7065935730934143, + "learning_rate": 3.5762160945983944e-05, + "loss": 0.1671, + "step": 25680 + }, + { + "epoch": 2.117256151142355, + "grad_norm": 1.5369142293930054, + "learning_rate": 3.573619685269558e-05, + "loss": 0.1499, + "step": 25700 + }, + { + "epoch": 2.118903778558875, + "grad_norm": 0.8731626272201538, + "learning_rate": 3.571021855278405e-05, + "loss": 0.1631, + "step": 25720 + }, + { + "epoch": 2.1205514059753954, + "grad_norm": 0.6267048120498657, + "learning_rate": 3.568422608062502e-05, + "loss": 0.1768, + "step": 25740 + }, + { + "epoch": 2.1221990333919156, + "grad_norm": 0.6229351758956909, + "learning_rate": 3.565821947061291e-05, + "loss": 0.1702, + "step": 25760 + }, + { + "epoch": 2.123846660808436, + "grad_norm": 0.5864184498786926, + "learning_rate": 3.563219875716084e-05, + "loss": 0.1669, + "step": 25780 + }, + { + "epoch": 2.125494288224956, + "grad_norm": 1.3878580331802368, + "learning_rate": 3.5606163974700616e-05, + "loss": 0.1563, + "step": 25800 + }, + { + "epoch": 2.1271419156414764, + "grad_norm": 1.8174159526824951, + "learning_rate": 3.558011515768264e-05, + "loss": 0.1583, + "step": 25820 + }, + { + "epoch": 2.1287895430579966, + "grad_norm": 0.8748142123222351, + "learning_rate": 3.5554052340575876e-05, + "loss": 0.1836, + "step": 25840 + }, + { + "epoch": 2.130437170474517, + "grad_norm": 0.6941066384315491, + "learning_rate": 3.552797555786784e-05, + "loss": 0.1671, + "step": 25860 + }, + { + "epoch": 2.132084797891037, + "grad_norm": 0.7565058469772339, + "learning_rate": 3.5501884844064506e-05, + "loss": 0.1663, + "step": 25880 + }, + { + "epoch": 2.133732425307557, + "grad_norm": 2.014808177947998, + "learning_rate": 3.54757802336903e-05, + "loss": 0.1577, + "step": 25900 + }, + { + "epoch": 2.135380052724077, + "grad_norm": 1.0949345827102661, + "learning_rate": 3.544966176128802e-05, + "loss": 0.1474, + "step": 25920 + }, + { + "epoch": 2.1370276801405974, + "grad_norm": 1.0987663269042969, + "learning_rate": 3.5423529461418816e-05, + "loss": 0.1705, + "step": 25940 + }, + { + "epoch": 2.1386753075571177, + "grad_norm": 0.8172488212585449, + "learning_rate": 3.539738336866214e-05, + "loss": 0.163, + "step": 25960 + }, + { + "epoch": 2.140322934973638, + "grad_norm": 0.37782424688339233, + "learning_rate": 3.537122351761568e-05, + "loss": 0.178, + "step": 25980 + }, + { + "epoch": 2.141970562390158, + "grad_norm": 1.336592435836792, + "learning_rate": 3.534504994289535e-05, + "loss": 0.1706, + "step": 26000 + }, + { + "epoch": 2.141970562390158, + "eval_loss": 0.5332348942756653, + "eval_runtime": 245.7071, + "eval_samples_per_second": 88.679, + "eval_steps_per_second": 22.173, + "eval_wer": 0.2268666911145139, + "step": 26000 + }, + { + "epoch": 2.1436181898066784, + "grad_norm": 0.7879346609115601, + "learning_rate": 3.53188626791352e-05, + "loss": 0.1631, + "step": 26020 + }, + { + "epoch": 2.1452658172231986, + "grad_norm": 0.4960474967956543, + "learning_rate": 3.529266176098741e-05, + "loss": 0.1843, + "step": 26040 + }, + { + "epoch": 2.146913444639719, + "grad_norm": 0.8829359412193298, + "learning_rate": 3.5266447223122234e-05, + "loss": 0.166, + "step": 26060 + }, + { + "epoch": 2.148561072056239, + "grad_norm": 2.7551321983337402, + "learning_rate": 3.524021910022793e-05, + "loss": 0.1662, + "step": 26080 + }, + { + "epoch": 2.1502086994727594, + "grad_norm": 2.773979663848877, + "learning_rate": 3.5213977427010744e-05, + "loss": 0.1536, + "step": 26100 + }, + { + "epoch": 2.1518563268892796, + "grad_norm": 1.1753557920455933, + "learning_rate": 3.518772223819485e-05, + "loss": 0.1671, + "step": 26120 + }, + { + "epoch": 2.1535039543057994, + "grad_norm": 0.704517662525177, + "learning_rate": 3.51614535685223e-05, + "loss": 0.1834, + "step": 26140 + }, + { + "epoch": 2.1551515817223197, + "grad_norm": 1.0736467838287354, + "learning_rate": 3.5135171452753e-05, + "loss": 0.1673, + "step": 26160 + }, + { + "epoch": 2.15679920913884, + "grad_norm": 0.44670116901397705, + "learning_rate": 3.510887592566463e-05, + "loss": 0.1579, + "step": 26180 + }, + { + "epoch": 2.15844683655536, + "grad_norm": 1.965410828590393, + "learning_rate": 3.508256702205263e-05, + "loss": 0.1622, + "step": 26200 + }, + { + "epoch": 2.1600944639718804, + "grad_norm": 0.6488901972770691, + "learning_rate": 3.505624477673013e-05, + "loss": 0.1706, + "step": 26220 + }, + { + "epoch": 2.1617420913884007, + "grad_norm": 1.937798023223877, + "learning_rate": 3.5029909224527936e-05, + "loss": 0.1686, + "step": 26240 + }, + { + "epoch": 2.163389718804921, + "grad_norm": 0.9314224720001221, + "learning_rate": 3.500356040029443e-05, + "loss": 0.1586, + "step": 26260 + }, + { + "epoch": 2.165037346221441, + "grad_norm": 0.41909438371658325, + "learning_rate": 3.497719833889558e-05, + "loss": 0.158, + "step": 26280 + }, + { + "epoch": 2.1666849736379614, + "grad_norm": 1.581127643585205, + "learning_rate": 3.495082307521486e-05, + "loss": 0.1477, + "step": 26300 + }, + { + "epoch": 2.1683326010544817, + "grad_norm": 3.524155855178833, + "learning_rate": 3.492443464415322e-05, + "loss": 0.1651, + "step": 26320 + }, + { + "epoch": 2.169980228471002, + "grad_norm": 0.7497128248214722, + "learning_rate": 3.489803308062902e-05, + "loss": 0.1726, + "step": 26340 + }, + { + "epoch": 2.171627855887522, + "grad_norm": 0.5789768695831299, + "learning_rate": 3.4871618419578024e-05, + "loss": 0.1636, + "step": 26360 + }, + { + "epoch": 2.1732754833040424, + "grad_norm": 0.4211706221103668, + "learning_rate": 3.4845190695953294e-05, + "loss": 0.1604, + "step": 26380 + }, + { + "epoch": 2.174923110720562, + "grad_norm": 5.170593738555908, + "learning_rate": 3.4818749944725214e-05, + "loss": 0.1465, + "step": 26400 + }, + { + "epoch": 2.1765707381370825, + "grad_norm": 0.47005754709243774, + "learning_rate": 3.479229620088137e-05, + "loss": 0.1585, + "step": 26420 + }, + { + "epoch": 2.1782183655536027, + "grad_norm": 0.8993769884109497, + "learning_rate": 3.4765829499426554e-05, + "loss": 0.1969, + "step": 26440 + }, + { + "epoch": 2.179865992970123, + "grad_norm": 0.8027138710021973, + "learning_rate": 3.473934987538273e-05, + "loss": 0.1534, + "step": 26460 + }, + { + "epoch": 2.181513620386643, + "grad_norm": 0.3248322308063507, + "learning_rate": 3.471285736378891e-05, + "loss": 0.1519, + "step": 26480 + }, + { + "epoch": 2.1831612478031635, + "grad_norm": 0.6821154356002808, + "learning_rate": 3.468635199970121e-05, + "loss": 0.145, + "step": 26500 + }, + { + "epoch": 2.1848088752196837, + "grad_norm": 1.0828319787979126, + "learning_rate": 3.465983381819272e-05, + "loss": 0.1639, + "step": 26520 + }, + { + "epoch": 2.186456502636204, + "grad_norm": 0.5787882804870605, + "learning_rate": 3.4633302854353505e-05, + "loss": 0.175, + "step": 26540 + }, + { + "epoch": 2.188104130052724, + "grad_norm": 1.9811298847198486, + "learning_rate": 3.460675914329055e-05, + "loss": 0.1739, + "step": 26560 + }, + { + "epoch": 2.1897517574692444, + "grad_norm": 0.5545487403869629, + "learning_rate": 3.458020272012768e-05, + "loss": 0.1542, + "step": 26580 + }, + { + "epoch": 2.1913993848857647, + "grad_norm": 2.1951963901519775, + "learning_rate": 3.455363362000558e-05, + "loss": 0.1532, + "step": 26600 + }, + { + "epoch": 2.1930470123022845, + "grad_norm": 0.9359253644943237, + "learning_rate": 3.452705187808167e-05, + "loss": 0.1622, + "step": 26620 + }, + { + "epoch": 2.1946946397188047, + "grad_norm": 0.5310496687889099, + "learning_rate": 3.450045752953012e-05, + "loss": 0.175, + "step": 26640 + }, + { + "epoch": 2.196342267135325, + "grad_norm": 0.44132381677627563, + "learning_rate": 3.44738506095418e-05, + "loss": 0.1718, + "step": 26660 + }, + { + "epoch": 2.1979898945518452, + "grad_norm": 0.6888401508331299, + "learning_rate": 3.444723115332418e-05, + "loss": 0.1837, + "step": 26680 + }, + { + "epoch": 2.1996375219683655, + "grad_norm": 1.194044589996338, + "learning_rate": 3.442059919610133e-05, + "loss": 0.1585, + "step": 26700 + }, + { + "epoch": 2.2012851493848857, + "grad_norm": 0.7281453013420105, + "learning_rate": 3.439395477311388e-05, + "loss": 0.1678, + "step": 26720 + }, + { + "epoch": 2.202932776801406, + "grad_norm": 0.5641632676124573, + "learning_rate": 3.4367297919618925e-05, + "loss": 0.1839, + "step": 26740 + }, + { + "epoch": 2.2045804042179262, + "grad_norm": 10.77109146118164, + "learning_rate": 3.434062867089003e-05, + "loss": 0.1674, + "step": 26760 + }, + { + "epoch": 2.2062280316344465, + "grad_norm": 0.7375435829162598, + "learning_rate": 3.431394706221715e-05, + "loss": 0.1657, + "step": 26780 + }, + { + "epoch": 2.2078756590509667, + "grad_norm": 2.0869717597961426, + "learning_rate": 3.4288588117737054e-05, + "loss": 0.158, + "step": 26800 + }, + { + "epoch": 2.209523286467487, + "grad_norm": 1.758624792098999, + "learning_rate": 3.4261882508738067e-05, + "loss": 0.1579, + "step": 26820 + }, + { + "epoch": 2.211170913884007, + "grad_norm": 0.6693238019943237, + "learning_rate": 3.4235164643995594e-05, + "loss": 0.1886, + "step": 26840 + }, + { + "epoch": 2.2128185413005275, + "grad_norm": 1.1407721042633057, + "learning_rate": 3.4208434558863925e-05, + "loss": 0.1755, + "step": 26860 + }, + { + "epoch": 2.2144661687170473, + "grad_norm": 1.6515735387802124, + "learning_rate": 3.418169228871353e-05, + "loss": 0.1667, + "step": 26880 + }, + { + "epoch": 2.2161137961335675, + "grad_norm": 3.605949878692627, + "learning_rate": 3.4154937868931e-05, + "loss": 0.1542, + "step": 26900 + }, + { + "epoch": 2.2177614235500878, + "grad_norm": 0.5841062068939209, + "learning_rate": 3.4128171334918986e-05, + "loss": 0.154, + "step": 26920 + }, + { + "epoch": 2.219409050966608, + "grad_norm": 0.9051983952522278, + "learning_rate": 3.410139272209619e-05, + "loss": 0.157, + "step": 26940 + }, + { + "epoch": 2.2210566783831283, + "grad_norm": 1.0805680751800537, + "learning_rate": 3.407460206589728e-05, + "loss": 0.1773, + "step": 26960 + }, + { + "epoch": 2.2227043057996485, + "grad_norm": 2.126185655593872, + "learning_rate": 3.4047799401772874e-05, + "loss": 0.1693, + "step": 26980 + }, + { + "epoch": 2.2243519332161688, + "grad_norm": 1.8613393306732178, + "learning_rate": 3.402098476518947e-05, + "loss": 0.1583, + "step": 27000 + }, + { + "epoch": 2.225999560632689, + "grad_norm": 1.6910747289657593, + "learning_rate": 3.399415819162941e-05, + "loss": 0.1845, + "step": 27020 + }, + { + "epoch": 2.2276471880492092, + "grad_norm": 1.2789907455444336, + "learning_rate": 3.396866192245471e-05, + "loss": 0.1829, + "step": 27040 + }, + { + "epoch": 2.2292948154657295, + "grad_norm": 1.3591282367706299, + "learning_rate": 3.394181217390606e-05, + "loss": 0.1657, + "step": 27060 + }, + { + "epoch": 2.2309424428822497, + "grad_norm": 0.36001384258270264, + "learning_rate": 3.3914950593145534e-05, + "loss": 0.1608, + "step": 27080 + }, + { + "epoch": 2.2325900702987695, + "grad_norm": 3.2652955055236816, + "learning_rate": 3.388807721571758e-05, + "loss": 0.1536, + "step": 27100 + }, + { + "epoch": 2.23423769771529, + "grad_norm": 1.4743667840957642, + "learning_rate": 3.3861192077182284e-05, + "loss": 0.1593, + "step": 27120 + }, + { + "epoch": 2.23588532513181, + "grad_norm": 4.98727560043335, + "learning_rate": 3.3834295213115274e-05, + "loss": 0.1608, + "step": 27140 + }, + { + "epoch": 2.2375329525483303, + "grad_norm": 2.447120189666748, + "learning_rate": 3.380738665910771e-05, + "loss": 0.1712, + "step": 27160 + }, + { + "epoch": 2.2391805799648505, + "grad_norm": 1.3015066385269165, + "learning_rate": 3.3780466450766215e-05, + "loss": 0.1785, + "step": 27180 + }, + { + "epoch": 2.240828207381371, + "grad_norm": 2.257363796234131, + "learning_rate": 3.375353462371281e-05, + "loss": 0.1678, + "step": 27200 + }, + { + "epoch": 2.242475834797891, + "grad_norm": 1.4966301918029785, + "learning_rate": 3.372659121358493e-05, + "loss": 0.1719, + "step": 27220 + }, + { + "epoch": 2.2441234622144113, + "grad_norm": 0.6006611585617065, + "learning_rate": 3.369963625603533e-05, + "loss": 0.1759, + "step": 27240 + }, + { + "epoch": 2.2457710896309315, + "grad_norm": 1.252107858657837, + "learning_rate": 3.367266978673201e-05, + "loss": 0.1698, + "step": 27260 + }, + { + "epoch": 2.2474187170474518, + "grad_norm": 1.2595499753952026, + "learning_rate": 3.364569184135824e-05, + "loss": 0.1687, + "step": 27280 + }, + { + "epoch": 2.249066344463972, + "grad_norm": 1.3134331703186035, + "learning_rate": 3.361870245561244e-05, + "loss": 0.1586, + "step": 27300 + }, + { + "epoch": 2.2507139718804923, + "grad_norm": 1.1949400901794434, + "learning_rate": 3.359170166520822e-05, + "loss": 0.1564, + "step": 27320 + }, + { + "epoch": 2.2523615992970125, + "grad_norm": 1.6761950254440308, + "learning_rate": 3.3564689505874236e-05, + "loss": 0.1755, + "step": 27340 + }, + { + "epoch": 2.2540092267135323, + "grad_norm": 0.8743230104446411, + "learning_rate": 3.3537666013354216e-05, + "loss": 0.1771, + "step": 27360 + }, + { + "epoch": 2.2556568541300526, + "grad_norm": 2.021210193634033, + "learning_rate": 3.351063122340686e-05, + "loss": 0.1639, + "step": 27380 + }, + { + "epoch": 2.257304481546573, + "grad_norm": 1.8056607246398926, + "learning_rate": 3.3483585171805825e-05, + "loss": 0.1647, + "step": 27400 + }, + { + "epoch": 2.258952108963093, + "grad_norm": 4.348305702209473, + "learning_rate": 3.345652789433969e-05, + "loss": 0.1604, + "step": 27420 + }, + { + "epoch": 2.2605997363796133, + "grad_norm": 1.9006876945495605, + "learning_rate": 3.342945942681186e-05, + "loss": 0.1757, + "step": 27440 + }, + { + "epoch": 2.2622473637961336, + "grad_norm": 3.284151315689087, + "learning_rate": 3.340237980504058e-05, + "loss": 0.1914, + "step": 27460 + }, + { + "epoch": 2.263894991212654, + "grad_norm": 1.1403635740280151, + "learning_rate": 3.337528906485881e-05, + "loss": 0.1663, + "step": 27480 + }, + { + "epoch": 2.265542618629174, + "grad_norm": 1.159006953239441, + "learning_rate": 3.334818724211426e-05, + "loss": 0.1658, + "step": 27500 + }, + { + "epoch": 2.2671902460456943, + "grad_norm": 0.8393998742103577, + "learning_rate": 3.33210743726693e-05, + "loss": 0.1664, + "step": 27520 + }, + { + "epoch": 2.2688378734622145, + "grad_norm": 0.6777483224868774, + "learning_rate": 3.329395049240089e-05, + "loss": 0.1835, + "step": 27540 + }, + { + "epoch": 2.270485500878735, + "grad_norm": 1.037467360496521, + "learning_rate": 3.3266815637200605e-05, + "loss": 0.1755, + "step": 27560 + }, + { + "epoch": 2.2721331282952546, + "grad_norm": 0.3177587389945984, + "learning_rate": 3.3239669842974495e-05, + "loss": 0.1611, + "step": 27580 + }, + { + "epoch": 2.2737807557117753, + "grad_norm": 3.4033076763153076, + "learning_rate": 3.321251314564312e-05, + "loss": 0.1668, + "step": 27600 + }, + { + "epoch": 2.275428383128295, + "grad_norm": 0.5792399048805237, + "learning_rate": 3.318534558114146e-05, + "loss": 0.1765, + "step": 27620 + }, + { + "epoch": 2.2770760105448153, + "grad_norm": 0.6022660732269287, + "learning_rate": 3.3158167185418846e-05, + "loss": 0.1721, + "step": 27640 + }, + { + "epoch": 2.2787236379613356, + "grad_norm": 2.6053919792175293, + "learning_rate": 3.313097799443898e-05, + "loss": 0.1774, + "step": 27660 + }, + { + "epoch": 2.280371265377856, + "grad_norm": 1.1545339822769165, + "learning_rate": 3.310377804417985e-05, + "loss": 0.1595, + "step": 27680 + }, + { + "epoch": 2.282018892794376, + "grad_norm": 1.8960176706314087, + "learning_rate": 3.3076567370633636e-05, + "loss": 0.1611, + "step": 27700 + }, + { + "epoch": 2.2836665202108963, + "grad_norm": 2.5447661876678467, + "learning_rate": 3.3049346009806766e-05, + "loss": 0.1634, + "step": 27720 + }, + { + "epoch": 2.2853141476274166, + "grad_norm": 3.0229389667510986, + "learning_rate": 3.302211399771975e-05, + "loss": 0.1873, + "step": 27740 + }, + { + "epoch": 2.286961775043937, + "grad_norm": 1.5011436939239502, + "learning_rate": 3.299487137040724e-05, + "loss": 0.1653, + "step": 27760 + }, + { + "epoch": 2.288609402460457, + "grad_norm": 3.6717607975006104, + "learning_rate": 3.2967618163917926e-05, + "loss": 0.1905, + "step": 27780 + }, + { + "epoch": 2.2902570298769773, + "grad_norm": 1.347084403038025, + "learning_rate": 3.2940354414314465e-05, + "loss": 0.157, + "step": 27800 + }, + { + "epoch": 2.2919046572934976, + "grad_norm": 1.2442917823791504, + "learning_rate": 3.2913080157673516e-05, + "loss": 0.156, + "step": 27820 + }, + { + "epoch": 2.2935522847100174, + "grad_norm": 1.3263580799102783, + "learning_rate": 3.288579543008559e-05, + "loss": 0.1801, + "step": 27840 + }, + { + "epoch": 2.2951999121265376, + "grad_norm": 0.6720674633979797, + "learning_rate": 3.285850026765509e-05, + "loss": 0.176, + "step": 27860 + }, + { + "epoch": 2.296847539543058, + "grad_norm": 1.0861464738845825, + "learning_rate": 3.2831194706500215e-05, + "loss": 0.1623, + "step": 27880 + }, + { + "epoch": 2.298495166959578, + "grad_norm": 0.9440199136734009, + "learning_rate": 3.2803878782752904e-05, + "loss": 0.1497, + "step": 27900 + }, + { + "epoch": 2.3001427943760984, + "grad_norm": 2.2923052310943604, + "learning_rate": 3.277655253255886e-05, + "loss": 0.17, + "step": 27920 + }, + { + "epoch": 2.3017904217926186, + "grad_norm": 0.5938337445259094, + "learning_rate": 3.2749215992077394e-05, + "loss": 0.184, + "step": 27940 + }, + { + "epoch": 2.303438049209139, + "grad_norm": 0.8251938223838806, + "learning_rate": 3.272186919748147e-05, + "loss": 0.1765, + "step": 27960 + }, + { + "epoch": 2.305085676625659, + "grad_norm": 2.4454212188720703, + "learning_rate": 3.2694512184957596e-05, + "loss": 0.1649, + "step": 27980 + }, + { + "epoch": 2.3067333040421794, + "grad_norm": 11.248618125915527, + "learning_rate": 3.266714499070582e-05, + "loss": 0.1558, + "step": 28000 + }, + { + "epoch": 2.3067333040421794, + "eval_loss": 0.5309145450592041, + "eval_runtime": 249.2844, + "eval_samples_per_second": 87.406, + "eval_steps_per_second": 21.855, + "eval_wer": 0.22905859925032823, + "step": 28000 + }, + { + "epoch": 2.3083809314586996, + "grad_norm": 2.0679521560668945, + "learning_rate": 3.2639767650939665e-05, + "loss": 0.1621, + "step": 28020 + }, + { + "epoch": 2.31002855887522, + "grad_norm": 1.7116316556930542, + "learning_rate": 3.261238020188605e-05, + "loss": 0.1709, + "step": 28040 + }, + { + "epoch": 2.3116761862917397, + "grad_norm": 1.4451134204864502, + "learning_rate": 3.258498267978531e-05, + "loss": 0.1652, + "step": 28060 + }, + { + "epoch": 2.3133238137082603, + "grad_norm": 0.7450625896453857, + "learning_rate": 3.255757512089108e-05, + "loss": 0.1599, + "step": 28080 + }, + { + "epoch": 2.31497144112478, + "grad_norm": 2.857830762863159, + "learning_rate": 3.2530157561470295e-05, + "loss": 0.1548, + "step": 28100 + }, + { + "epoch": 2.3166190685413004, + "grad_norm": 1.0902551412582397, + "learning_rate": 3.25027300378031e-05, + "loss": 0.1593, + "step": 28120 + }, + { + "epoch": 2.3182666959578206, + "grad_norm": 1.3042412996292114, + "learning_rate": 3.247529258618285e-05, + "loss": 0.1732, + "step": 28140 + }, + { + "epoch": 2.319914323374341, + "grad_norm": 0.9760801196098328, + "learning_rate": 3.244784524291603e-05, + "loss": 0.1776, + "step": 28160 + }, + { + "epoch": 2.321561950790861, + "grad_norm": 0.6647989749908447, + "learning_rate": 3.2420388044322204e-05, + "loss": 0.1681, + "step": 28180 + }, + { + "epoch": 2.3232095782073814, + "grad_norm": 2.4924325942993164, + "learning_rate": 3.239292102673398e-05, + "loss": 0.1659, + "step": 28200 + }, + { + "epoch": 2.3248572056239016, + "grad_norm": 1.96873939037323, + "learning_rate": 3.2365444226496977e-05, + "loss": 0.174, + "step": 28220 + }, + { + "epoch": 2.326504833040422, + "grad_norm": 1.3874828815460205, + "learning_rate": 3.2337957679969736e-05, + "loss": 0.1817, + "step": 28240 + }, + { + "epoch": 2.328152460456942, + "grad_norm": 1.0978938341140747, + "learning_rate": 3.2310461423523706e-05, + "loss": 0.1703, + "step": 28260 + }, + { + "epoch": 2.3298000878734624, + "grad_norm": 0.40774619579315186, + "learning_rate": 3.228295549354317e-05, + "loss": 0.1709, + "step": 28280 + }, + { + "epoch": 2.3314477152899826, + "grad_norm": 2.165424346923828, + "learning_rate": 3.225543992642524e-05, + "loss": 0.1606, + "step": 28300 + }, + { + "epoch": 2.3330953427065024, + "grad_norm": 1.4774760007858276, + "learning_rate": 3.222791475857976e-05, + "loss": 0.1602, + "step": 28320 + }, + { + "epoch": 2.3347429701230227, + "grad_norm": 0.60544353723526, + "learning_rate": 3.2200380026429274e-05, + "loss": 0.1806, + "step": 28340 + }, + { + "epoch": 2.336390597539543, + "grad_norm": 0.8085589408874512, + "learning_rate": 3.2172835766409e-05, + "loss": 0.1744, + "step": 28360 + }, + { + "epoch": 2.338038224956063, + "grad_norm": 0.43836772441864014, + "learning_rate": 3.2145282014966736e-05, + "loss": 0.1671, + "step": 28380 + }, + { + "epoch": 2.3396858523725834, + "grad_norm": 2.028322219848633, + "learning_rate": 3.211771880856288e-05, + "loss": 0.1567, + "step": 28400 + }, + { + "epoch": 2.3413334797891037, + "grad_norm": 0.6231321692466736, + "learning_rate": 3.2090146183670305e-05, + "loss": 0.1718, + "step": 28420 + }, + { + "epoch": 2.342981107205624, + "grad_norm": 1.3398654460906982, + "learning_rate": 3.206256417677437e-05, + "loss": 0.1705, + "step": 28440 + }, + { + "epoch": 2.344628734622144, + "grad_norm": 1.2689118385314941, + "learning_rate": 3.203497282437282e-05, + "loss": 0.1622, + "step": 28460 + }, + { + "epoch": 2.3462763620386644, + "grad_norm": 0.6729577779769897, + "learning_rate": 3.2007372162975805e-05, + "loss": 0.1616, + "step": 28480 + }, + { + "epoch": 2.3479239894551847, + "grad_norm": 2.0656583309173584, + "learning_rate": 3.197976222910577e-05, + "loss": 0.1654, + "step": 28500 + }, + { + "epoch": 2.349571616871705, + "grad_norm": 1.439696192741394, + "learning_rate": 3.195214305929742e-05, + "loss": 0.166, + "step": 28520 + }, + { + "epoch": 2.3512192442882247, + "grad_norm": 0.6270071268081665, + "learning_rate": 3.1924514690097706e-05, + "loss": 0.183, + "step": 28540 + }, + { + "epoch": 2.3528668717047454, + "grad_norm": 0.5838283896446228, + "learning_rate": 3.1896877158065747e-05, + "loss": 0.1716, + "step": 28560 + }, + { + "epoch": 2.354514499121265, + "grad_norm": 0.424365371465683, + "learning_rate": 3.186923049977276e-05, + "loss": 0.1776, + "step": 28580 + }, + { + "epoch": 2.3561621265377855, + "grad_norm": 1.257899284362793, + "learning_rate": 3.1841574751802076e-05, + "loss": 0.1535, + "step": 28600 + }, + { + "epoch": 2.3578097539543057, + "grad_norm": 1.2399417161941528, + "learning_rate": 3.181390995074903e-05, + "loss": 0.1676, + "step": 28620 + }, + { + "epoch": 2.359457381370826, + "grad_norm": 1.031891942024231, + "learning_rate": 3.178623613322094e-05, + "loss": 0.1682, + "step": 28640 + }, + { + "epoch": 2.361105008787346, + "grad_norm": 2.5640039443969727, + "learning_rate": 3.175855333583706e-05, + "loss": 0.1656, + "step": 28660 + }, + { + "epoch": 2.3627526362038664, + "grad_norm": 0.32436710596084595, + "learning_rate": 3.173086159522853e-05, + "loss": 0.1617, + "step": 28680 + }, + { + "epoch": 2.3644002636203867, + "grad_norm": 1.4761956930160522, + "learning_rate": 3.1703160948038315e-05, + "loss": 0.1558, + "step": 28700 + }, + { + "epoch": 2.366047891036907, + "grad_norm": 1.316041350364685, + "learning_rate": 3.1675451430921166e-05, + "loss": 0.1596, + "step": 28720 + }, + { + "epoch": 2.367695518453427, + "grad_norm": 0.48719462752342224, + "learning_rate": 3.164773308054357e-05, + "loss": 0.176, + "step": 28740 + }, + { + "epoch": 2.3693431458699474, + "grad_norm": 0.4793241322040558, + "learning_rate": 3.162000593358372e-05, + "loss": 0.162, + "step": 28760 + }, + { + "epoch": 2.3709907732864677, + "grad_norm": 0.25977346301078796, + "learning_rate": 3.159227002673143e-05, + "loss": 0.1477, + "step": 28780 + }, + { + "epoch": 2.3726384007029875, + "grad_norm": 3.16365385055542, + "learning_rate": 3.156452539668811e-05, + "loss": 0.1599, + "step": 28800 + }, + { + "epoch": 2.3742860281195077, + "grad_norm": 0.5996500849723816, + "learning_rate": 3.153677208016672e-05, + "loss": 0.1545, + "step": 28820 + }, + { + "epoch": 2.375933655536028, + "grad_norm": 1.808449387550354, + "learning_rate": 3.15090101138917e-05, + "loss": 0.1565, + "step": 28840 + }, + { + "epoch": 2.3775812829525482, + "grad_norm": 0.822944700717926, + "learning_rate": 3.148123953459894e-05, + "loss": 0.1655, + "step": 28860 + }, + { + "epoch": 2.3792289103690685, + "grad_norm": 0.3074828088283539, + "learning_rate": 3.145346037903574e-05, + "loss": 0.1617, + "step": 28880 + }, + { + "epoch": 2.3808765377855887, + "grad_norm": 3.4006128311157227, + "learning_rate": 3.142567268396075e-05, + "loss": 0.1755, + "step": 28900 + }, + { + "epoch": 2.382524165202109, + "grad_norm": 0.9167118668556213, + "learning_rate": 3.1397876486143895e-05, + "loss": 0.1648, + "step": 28920 + }, + { + "epoch": 2.384171792618629, + "grad_norm": 0.6109206676483154, + "learning_rate": 3.137007182236637e-05, + "loss": 0.1637, + "step": 28940 + }, + { + "epoch": 2.3858194200351495, + "grad_norm": 0.45387306809425354, + "learning_rate": 3.1342258729420564e-05, + "loss": 0.1736, + "step": 28960 + }, + { + "epoch": 2.3874670474516697, + "grad_norm": 0.3555924892425537, + "learning_rate": 3.131443724411003e-05, + "loss": 0.1664, + "step": 28980 + }, + { + "epoch": 2.38911467486819, + "grad_norm": 1.275738000869751, + "learning_rate": 3.128660740324943e-05, + "loss": 0.1553, + "step": 29000 + }, + { + "epoch": 2.39076230228471, + "grad_norm": 0.5454599261283875, + "learning_rate": 3.125876924366445e-05, + "loss": 0.163, + "step": 29020 + }, + { + "epoch": 2.3924099297012305, + "grad_norm": 0.4194986820220947, + "learning_rate": 3.12323153203915e-05, + "loss": 0.1833, + "step": 29040 + }, + { + "epoch": 2.3940575571177503, + "grad_norm": 0.4429135024547577, + "learning_rate": 3.1204461045255604e-05, + "loss": 0.1668, + "step": 29060 + }, + { + "epoch": 2.3957051845342705, + "grad_norm": 0.3895147442817688, + "learning_rate": 3.117659856009512e-05, + "loss": 0.1609, + "step": 29080 + }, + { + "epoch": 2.3973528119507908, + "grad_norm": 3.838486909866333, + "learning_rate": 3.114872790177897e-05, + "loss": 0.157, + "step": 29100 + }, + { + "epoch": 2.399000439367311, + "grad_norm": 0.43187758326530457, + "learning_rate": 3.112084910718688e-05, + "loss": 0.1643, + "step": 29120 + }, + { + "epoch": 2.4006480667838312, + "grad_norm": 0.3776147663593292, + "learning_rate": 3.109296221320932e-05, + "loss": 0.1791, + "step": 29140 + }, + { + "epoch": 2.4022956942003515, + "grad_norm": 0.5783634185791016, + "learning_rate": 3.1065067256747495e-05, + "loss": 0.1644, + "step": 29160 + }, + { + "epoch": 2.4039433216168717, + "grad_norm": 0.8182345032691956, + "learning_rate": 3.1037164274713286e-05, + "loss": 0.1492, + "step": 29180 + }, + { + "epoch": 2.405590949033392, + "grad_norm": 2.9094724655151367, + "learning_rate": 3.100925330402919e-05, + "loss": 0.1555, + "step": 29200 + }, + { + "epoch": 2.4072385764499122, + "grad_norm": 0.932719349861145, + "learning_rate": 3.0981334381628256e-05, + "loss": 0.1575, + "step": 29220 + }, + { + "epoch": 2.4088862038664325, + "grad_norm": 0.6541189551353455, + "learning_rate": 3.095340754445409e-05, + "loss": 0.1681, + "step": 29240 + }, + { + "epoch": 2.4105338312829527, + "grad_norm": 1.609031319618225, + "learning_rate": 3.092547282946073e-05, + "loss": 0.1605, + "step": 29260 + }, + { + "epoch": 2.4121814586994725, + "grad_norm": 0.24770651757717133, + "learning_rate": 3.089753027361267e-05, + "loss": 0.1618, + "step": 29280 + }, + { + "epoch": 2.413829086115993, + "grad_norm": 5.404471397399902, + "learning_rate": 3.0869579913884775e-05, + "loss": 0.1633, + "step": 29300 + }, + { + "epoch": 2.415476713532513, + "grad_norm": 2.0868561267852783, + "learning_rate": 3.0841621787262213e-05, + "loss": 0.1533, + "step": 29320 + }, + { + "epoch": 2.4171243409490333, + "grad_norm": 1.2306647300720215, + "learning_rate": 3.0813655930740466e-05, + "loss": 0.1672, + "step": 29340 + }, + { + "epoch": 2.4187719683655535, + "grad_norm": 1.3059163093566895, + "learning_rate": 3.078568238132521e-05, + "loss": 0.163, + "step": 29360 + }, + { + "epoch": 2.4204195957820738, + "grad_norm": 0.6795551180839539, + "learning_rate": 3.075770117603232e-05, + "loss": 0.1696, + "step": 29380 + }, + { + "epoch": 2.422067223198594, + "grad_norm": 3.5448899269104004, + "learning_rate": 3.07297123518878e-05, + "loss": 0.1523, + "step": 29400 + }, + { + "epoch": 2.4237148506151143, + "grad_norm": 0.5173410177230835, + "learning_rate": 3.070171594592773e-05, + "loss": 0.1685, + "step": 29420 + }, + { + "epoch": 2.4253624780316345, + "grad_norm": 0.46082064509391785, + "learning_rate": 3.067371199519822e-05, + "loss": 0.1729, + "step": 29440 + }, + { + "epoch": 2.4270101054481548, + "grad_norm": 0.4469384551048279, + "learning_rate": 3.0645700536755384e-05, + "loss": 0.1704, + "step": 29460 + }, + { + "epoch": 2.428657732864675, + "grad_norm": 0.2371663600206375, + "learning_rate": 3.0617681607665256e-05, + "loss": 0.1697, + "step": 29480 + }, + { + "epoch": 2.4303053602811953, + "grad_norm": 2.3724477291107178, + "learning_rate": 3.058965524500374e-05, + "loss": 0.171, + "step": 29500 + }, + { + "epoch": 2.4319529876977155, + "grad_norm": 0.954786479473114, + "learning_rate": 3.056162148585659e-05, + "loss": 0.1641, + "step": 29520 + }, + { + "epoch": 2.4336006151142353, + "grad_norm": 0.5289298892021179, + "learning_rate": 3.053358036731937e-05, + "loss": 0.1664, + "step": 29540 + }, + { + "epoch": 2.4352482425307556, + "grad_norm": 0.7976502180099487, + "learning_rate": 3.0505531926497343e-05, + "loss": 0.1745, + "step": 29560 + }, + { + "epoch": 2.436895869947276, + "grad_norm": 0.3595167100429535, + "learning_rate": 3.047747620050549e-05, + "loss": 0.151, + "step": 29580 + }, + { + "epoch": 2.438543497363796, + "grad_norm": 19.019508361816406, + "learning_rate": 3.0449413226468417e-05, + "loss": 0.1803, + "step": 29600 + }, + { + "epoch": 2.4401911247803163, + "grad_norm": 7.179177761077881, + "learning_rate": 3.0421343041520322e-05, + "loss": 0.1552, + "step": 29620 + }, + { + "epoch": 2.4418387521968365, + "grad_norm": 0.5879939794540405, + "learning_rate": 3.039326568280495e-05, + "loss": 0.1721, + "step": 29640 + }, + { + "epoch": 2.443486379613357, + "grad_norm": 0.8371209502220154, + "learning_rate": 3.036518118747554e-05, + "loss": 0.1691, + "step": 29660 + }, + { + "epoch": 2.445134007029877, + "grad_norm": 0.41010937094688416, + "learning_rate": 3.0337089592694772e-05, + "loss": 0.1709, + "step": 29680 + }, + { + "epoch": 2.4467816344463973, + "grad_norm": 1.2957394123077393, + "learning_rate": 3.0308990935634718e-05, + "loss": 0.1466, + "step": 29700 + }, + { + "epoch": 2.4484292618629175, + "grad_norm": 0.7702599167823792, + "learning_rate": 3.0280885253476797e-05, + "loss": 0.1652, + "step": 29720 + }, + { + "epoch": 2.450076889279438, + "grad_norm": 0.7201776504516602, + "learning_rate": 3.025277258341172e-05, + "loss": 0.1675, + "step": 29740 + }, + { + "epoch": 2.4517245166959576, + "grad_norm": 0.37506094574928284, + "learning_rate": 3.0224652962639454e-05, + "loss": 0.163, + "step": 29760 + }, + { + "epoch": 2.4533721441124783, + "grad_norm": 0.3698022961616516, + "learning_rate": 3.0196526428369155e-05, + "loss": 0.1634, + "step": 29780 + }, + { + "epoch": 2.455019771528998, + "grad_norm": 2.1484596729278564, + "learning_rate": 3.016839301781913e-05, + "loss": 0.1422, + "step": 29800 + }, + { + "epoch": 2.4566673989455183, + "grad_norm": 1.094212293624878, + "learning_rate": 3.014025276821679e-05, + "loss": 0.1539, + "step": 29820 + }, + { + "epoch": 2.4583150263620386, + "grad_norm": 1.0083136558532715, + "learning_rate": 3.011210571679859e-05, + "loss": 0.1652, + "step": 29840 + }, + { + "epoch": 2.459962653778559, + "grad_norm": 0.9524624943733215, + "learning_rate": 3.0083951900809983e-05, + "loss": 0.1549, + "step": 29860 + }, + { + "epoch": 2.461610281195079, + "grad_norm": 0.9156121611595154, + "learning_rate": 3.005579135750538e-05, + "loss": 0.1493, + "step": 29880 + }, + { + "epoch": 2.4632579086115993, + "grad_norm": 1.3590614795684814, + "learning_rate": 3.002762412414808e-05, + "loss": 0.1754, + "step": 29900 + }, + { + "epoch": 2.4649055360281196, + "grad_norm": 0.9748837947845459, + "learning_rate": 2.9999450238010264e-05, + "loss": 0.1662, + "step": 29920 + }, + { + "epoch": 2.46655316344464, + "grad_norm": 0.43516382575035095, + "learning_rate": 2.9971269736372874e-05, + "loss": 0.1816, + "step": 29940 + }, + { + "epoch": 2.46820079086116, + "grad_norm": 0.7476974129676819, + "learning_rate": 2.994308265652565e-05, + "loss": 0.1624, + "step": 29960 + }, + { + "epoch": 2.4698484182776803, + "grad_norm": 0.2631094455718994, + "learning_rate": 2.9914889035767014e-05, + "loss": 0.1598, + "step": 29980 + }, + { + "epoch": 2.4714960456942006, + "grad_norm": 7.102409839630127, + "learning_rate": 2.9886688911404033e-05, + "loss": 0.1574, + "step": 30000 + }, + { + "epoch": 2.4714960456942006, + "eval_loss": 0.5125442147254944, + "eval_runtime": 246.1596, + "eval_samples_per_second": 88.516, + "eval_steps_per_second": 22.132, + "eval_wer": 0.22153388779557867, + "step": 30000 + }, + { + "epoch": 2.4731436731107204, + "grad_norm": 0.5292249917984009, + "learning_rate": 2.98584823207524e-05, + "loss": 0.1598, + "step": 30020 + }, + { + "epoch": 2.4747913005272406, + "grad_norm": 0.8891049027442932, + "learning_rate": 2.983026930113635e-05, + "loss": 0.1769, + "step": 30040 + }, + { + "epoch": 2.476438927943761, + "grad_norm": 0.5276570916175842, + "learning_rate": 2.9802049889888638e-05, + "loss": 0.156, + "step": 30060 + }, + { + "epoch": 2.478086555360281, + "grad_norm": 0.30637362599372864, + "learning_rate": 2.977382412435047e-05, + "loss": 0.1549, + "step": 30080 + }, + { + "epoch": 2.4797341827768014, + "grad_norm": 7.410438537597656, + "learning_rate": 2.974559204187146e-05, + "loss": 0.1503, + "step": 30100 + }, + { + "epoch": 2.4813818101933216, + "grad_norm": 3.507681369781494, + "learning_rate": 2.9717353679809584e-05, + "loss": 0.1577, + "step": 30120 + }, + { + "epoch": 2.483029437609842, + "grad_norm": 0.5745053291320801, + "learning_rate": 2.968910907553112e-05, + "loss": 0.164, + "step": 30140 + }, + { + "epoch": 2.484677065026362, + "grad_norm": 0.6689228415489197, + "learning_rate": 2.9660858266410613e-05, + "loss": 0.1603, + "step": 30160 + }, + { + "epoch": 2.4863246924428823, + "grad_norm": 0.7255517244338989, + "learning_rate": 2.963260128983081e-05, + "loss": 0.1567, + "step": 30180 + }, + { + "epoch": 2.4879723198594026, + "grad_norm": 5.331938743591309, + "learning_rate": 2.9604338183182633e-05, + "loss": 0.1504, + "step": 30200 + }, + { + "epoch": 2.489619947275923, + "grad_norm": 3.504755973815918, + "learning_rate": 2.957606898386511e-05, + "loss": 0.1538, + "step": 30220 + }, + { + "epoch": 2.4912675746924426, + "grad_norm": 0.5787059664726257, + "learning_rate": 2.9547793729285328e-05, + "loss": 0.1868, + "step": 30240 + }, + { + "epoch": 2.4929152021089633, + "grad_norm": 0.596563458442688, + "learning_rate": 2.9519512456858384e-05, + "loss": 0.1732, + "step": 30260 + }, + { + "epoch": 2.494562829525483, + "grad_norm": 0.5609906315803528, + "learning_rate": 2.9491225204007338e-05, + "loss": 0.1564, + "step": 30280 + }, + { + "epoch": 2.4962104569420034, + "grad_norm": 17.780019760131836, + "learning_rate": 2.9462932008163175e-05, + "loss": 0.1568, + "step": 30300 + }, + { + "epoch": 2.4978580843585236, + "grad_norm": 0.8004500865936279, + "learning_rate": 2.9434632906764737e-05, + "loss": 0.1605, + "step": 30320 + }, + { + "epoch": 2.499505711775044, + "grad_norm": 1.081133246421814, + "learning_rate": 2.9406327937258677e-05, + "loss": 0.1782, + "step": 30340 + }, + { + "epoch": 2.501153339191564, + "grad_norm": 3.350252389907837, + "learning_rate": 2.9378017137099417e-05, + "loss": 0.1673, + "step": 30360 + }, + { + "epoch": 2.5028009666080844, + "grad_norm": 0.6096969842910767, + "learning_rate": 2.935111651042648e-05, + "loss": 0.1636, + "step": 30380 + }, + { + "epoch": 2.5044485940246046, + "grad_norm": 4.936652660369873, + "learning_rate": 2.932279444825093e-05, + "loss": 0.1585, + "step": 30400 + }, + { + "epoch": 2.506096221441125, + "grad_norm": 1.6837114095687866, + "learning_rate": 2.9294466665957475e-05, + "loss": 0.1644, + "step": 30420 + }, + { + "epoch": 2.507743848857645, + "grad_norm": 0.9502339959144592, + "learning_rate": 2.926613320103074e-05, + "loss": 0.1791, + "step": 30440 + }, + { + "epoch": 2.5093914762741654, + "grad_norm": 1.6495269536972046, + "learning_rate": 2.9237794090962834e-05, + "loss": 0.1724, + "step": 30460 + }, + { + "epoch": 2.5110391036906856, + "grad_norm": 1.6112371683120728, + "learning_rate": 2.9209449373253372e-05, + "loss": 0.1719, + "step": 30480 + }, + { + "epoch": 2.5126867311072054, + "grad_norm": 1.7970635890960693, + "learning_rate": 2.918109908540937e-05, + "loss": 0.1553, + "step": 30500 + }, + { + "epoch": 2.514334358523726, + "grad_norm": 0.9640928506851196, + "learning_rate": 2.9152743264945205e-05, + "loss": 0.1651, + "step": 30520 + }, + { + "epoch": 2.515981985940246, + "grad_norm": 1.5328330993652344, + "learning_rate": 2.9124381949382584e-05, + "loss": 0.1851, + "step": 30540 + }, + { + "epoch": 2.517629613356766, + "grad_norm": 1.3880538940429688, + "learning_rate": 2.909601517625051e-05, + "loss": 0.1693, + "step": 30560 + }, + { + "epoch": 2.5192772407732864, + "grad_norm": 0.763766348361969, + "learning_rate": 2.906764298308516e-05, + "loss": 0.1702, + "step": 30580 + }, + { + "epoch": 2.5209248681898067, + "grad_norm": 1.9157522916793823, + "learning_rate": 2.9039265407429928e-05, + "loss": 0.1686, + "step": 30600 + }, + { + "epoch": 2.522572495606327, + "grad_norm": 1.0073531866073608, + "learning_rate": 2.9010882486835288e-05, + "loss": 0.1687, + "step": 30620 + }, + { + "epoch": 2.524220123022847, + "grad_norm": 0.8558169007301331, + "learning_rate": 2.8982494258858828e-05, + "loss": 0.1721, + "step": 30640 + }, + { + "epoch": 2.5258677504393674, + "grad_norm": 2.5032193660736084, + "learning_rate": 2.895410076106514e-05, + "loss": 0.1641, + "step": 30660 + }, + { + "epoch": 2.5275153778558876, + "grad_norm": 0.6763248443603516, + "learning_rate": 2.8925702031025775e-05, + "loss": 0.168, + "step": 30680 + }, + { + "epoch": 2.529163005272408, + "grad_norm": 5.686913013458252, + "learning_rate": 2.8897298106319243e-05, + "loss": 0.1576, + "step": 30700 + }, + { + "epoch": 2.5308106326889277, + "grad_norm": 0.7178692817687988, + "learning_rate": 2.8868889024530888e-05, + "loss": 0.1597, + "step": 30720 + }, + { + "epoch": 2.5324582601054484, + "grad_norm": 1.8655472993850708, + "learning_rate": 2.88404748232529e-05, + "loss": 0.1634, + "step": 30740 + }, + { + "epoch": 2.534105887521968, + "grad_norm": 0.9312283396720886, + "learning_rate": 2.881205554008425e-05, + "loss": 0.1678, + "step": 30760 + }, + { + "epoch": 2.5357535149384884, + "grad_norm": 1.1616804599761963, + "learning_rate": 2.8783631212630618e-05, + "loss": 0.1518, + "step": 30780 + }, + { + "epoch": 2.5374011423550087, + "grad_norm": 2.426833391189575, + "learning_rate": 2.8755201878504363e-05, + "loss": 0.1612, + "step": 30800 + }, + { + "epoch": 2.539048769771529, + "grad_norm": 0.6256110072135925, + "learning_rate": 2.872676757532447e-05, + "loss": 0.15, + "step": 30820 + }, + { + "epoch": 2.540696397188049, + "grad_norm": 0.48603343963623047, + "learning_rate": 2.869832834071651e-05, + "loss": 0.1692, + "step": 30840 + }, + { + "epoch": 2.5423440246045694, + "grad_norm": 1.9028749465942383, + "learning_rate": 2.866988421231256e-05, + "loss": 0.1625, + "step": 30860 + }, + { + "epoch": 2.5439916520210897, + "grad_norm": 0.6048976182937622, + "learning_rate": 2.864143522775118e-05, + "loss": 0.1539, + "step": 30880 + }, + { + "epoch": 2.54563927943761, + "grad_norm": 0.9251933693885803, + "learning_rate": 2.861298142467737e-05, + "loss": 0.1484, + "step": 30900 + }, + { + "epoch": 2.54728690685413, + "grad_norm": 2.2003366947174072, + "learning_rate": 2.8584522840742485e-05, + "loss": 0.1696, + "step": 30920 + }, + { + "epoch": 2.5489345342706504, + "grad_norm": 0.9549096822738647, + "learning_rate": 2.8556059513604215e-05, + "loss": 0.1821, + "step": 30940 + }, + { + "epoch": 2.5505821616871707, + "grad_norm": 1.7979339361190796, + "learning_rate": 2.852759148092653e-05, + "loss": 0.1712, + "step": 30960 + }, + { + "epoch": 2.5522297891036905, + "grad_norm": 0.9172550439834595, + "learning_rate": 2.8499118780379625e-05, + "loss": 0.1699, + "step": 30980 + }, + { + "epoch": 2.553877416520211, + "grad_norm": 1.2865842580795288, + "learning_rate": 2.8470641449639868e-05, + "loss": 0.1538, + "step": 31000 + }, + { + "epoch": 2.555525043936731, + "grad_norm": 1.5538517236709595, + "learning_rate": 2.8442159526389755e-05, + "loss": 0.1582, + "step": 31020 + }, + { + "epoch": 2.557172671353251, + "grad_norm": 0.45951759815216064, + "learning_rate": 2.841367304831786e-05, + "loss": 0.1701, + "step": 31040 + }, + { + "epoch": 2.5588202987697715, + "grad_norm": 1.1140258312225342, + "learning_rate": 2.8385182053118793e-05, + "loss": 0.1665, + "step": 31060 + }, + { + "epoch": 2.5604679261862917, + "grad_norm": 0.5959802269935608, + "learning_rate": 2.8356686578493112e-05, + "loss": 0.1731, + "step": 31080 + }, + { + "epoch": 2.562115553602812, + "grad_norm": 0.9773458242416382, + "learning_rate": 2.8328186662147345e-05, + "loss": 0.1665, + "step": 31100 + }, + { + "epoch": 2.563763181019332, + "grad_norm": 1.550663948059082, + "learning_rate": 2.829968234179386e-05, + "loss": 0.1703, + "step": 31120 + }, + { + "epoch": 2.5654108084358525, + "grad_norm": 1.0775203704833984, + "learning_rate": 2.8271173655150878e-05, + "loss": 0.1783, + "step": 31140 + }, + { + "epoch": 2.5670584358523727, + "grad_norm": 0.7373536825180054, + "learning_rate": 2.8242660639942388e-05, + "loss": 0.1727, + "step": 31160 + }, + { + "epoch": 2.568706063268893, + "grad_norm": 1.5068947076797485, + "learning_rate": 2.8215569300525145e-05, + "loss": 0.1738, + "step": 31180 + }, + { + "epoch": 2.5703536906854128, + "grad_norm": 1.7037030458450317, + "learning_rate": 2.8187047953139167e-05, + "loss": 0.1617, + "step": 31200 + }, + { + "epoch": 2.5720013181019334, + "grad_norm": 3.2985782623291016, + "learning_rate": 2.8158522388506632e-05, + "loss": 0.173, + "step": 31220 + }, + { + "epoch": 2.5736489455184532, + "grad_norm": 0.8943614363670349, + "learning_rate": 2.812999264437386e-05, + "loss": 0.1845, + "step": 31240 + }, + { + "epoch": 2.5752965729349735, + "grad_norm": 2.3433315753936768, + "learning_rate": 2.810145875849269e-05, + "loss": 0.1716, + "step": 31260 + }, + { + "epoch": 2.5769442003514937, + "grad_norm": 0.7237234711647034, + "learning_rate": 2.807292076862047e-05, + "loss": 0.1827, + "step": 31280 + }, + { + "epoch": 2.578591827768014, + "grad_norm": 1.9707437753677368, + "learning_rate": 2.804437871251996e-05, + "loss": 0.1644, + "step": 31300 + }, + { + "epoch": 2.5802394551845342, + "grad_norm": 5.560729503631592, + "learning_rate": 2.8015832627959298e-05, + "loss": 0.1614, + "step": 31320 + }, + { + "epoch": 2.5818870826010545, + "grad_norm": 1.6304755210876465, + "learning_rate": 2.7987282552711973e-05, + "loss": 0.1732, + "step": 31340 + }, + { + "epoch": 2.5835347100175747, + "grad_norm": 0.6963209509849548, + "learning_rate": 2.7958728524556727e-05, + "loss": 0.1793, + "step": 31360 + }, + { + "epoch": 2.585182337434095, + "grad_norm": 0.8520829677581787, + "learning_rate": 2.793017058127756e-05, + "loss": 0.1813, + "step": 31380 + }, + { + "epoch": 2.5868299648506152, + "grad_norm": 1.0092051029205322, + "learning_rate": 2.7901608760663623e-05, + "loss": 0.1655, + "step": 31400 + }, + { + "epoch": 2.5884775922671355, + "grad_norm": 1.0574760437011719, + "learning_rate": 2.7873043100509228e-05, + "loss": 0.1779, + "step": 31420 + }, + { + "epoch": 2.5901252196836557, + "grad_norm": 1.1205741167068481, + "learning_rate": 2.784447363861375e-05, + "loss": 0.1747, + "step": 31440 + }, + { + "epoch": 2.5917728471001755, + "grad_norm": 0.9736700654029846, + "learning_rate": 2.7815900412781594e-05, + "loss": 0.1703, + "step": 31460 + }, + { + "epoch": 2.593420474516696, + "grad_norm": 0.5332132577896118, + "learning_rate": 2.7787323460822146e-05, + "loss": 0.1648, + "step": 31480 + }, + { + "epoch": 2.595068101933216, + "grad_norm": 2.0113890171051025, + "learning_rate": 2.7758742820549738e-05, + "loss": 0.1647, + "step": 31500 + }, + { + "epoch": 2.5967157293497363, + "grad_norm": 7.613828659057617, + "learning_rate": 2.7730158529783566e-05, + "loss": 0.1549, + "step": 31520 + }, + { + "epoch": 2.5983633567662565, + "grad_norm": 1.237125277519226, + "learning_rate": 2.770157062634766e-05, + "loss": 0.1739, + "step": 31540 + }, + { + "epoch": 2.6000109841827768, + "grad_norm": 1.3644907474517822, + "learning_rate": 2.767297914807084e-05, + "loss": 0.1746, + "step": 31560 + }, + { + "epoch": 2.601658611599297, + "grad_norm": 1.010188341140747, + "learning_rate": 2.7644384132786637e-05, + "loss": 0.1577, + "step": 31580 + }, + { + "epoch": 2.6033062390158173, + "grad_norm": 1.2578411102294922, + "learning_rate": 2.761578561833328e-05, + "loss": 0.1569, + "step": 31600 + }, + { + "epoch": 2.6049538664323375, + "grad_norm": 0.8691573143005371, + "learning_rate": 2.758718364255361e-05, + "loss": 0.17, + "step": 31620 + }, + { + "epoch": 2.6066014938488578, + "grad_norm": 4.4133405685424805, + "learning_rate": 2.7558578243295087e-05, + "loss": 0.178, + "step": 31640 + }, + { + "epoch": 2.608249121265378, + "grad_norm": 0.5621764659881592, + "learning_rate": 2.7529969458409644e-05, + "loss": 0.1619, + "step": 31660 + }, + { + "epoch": 2.609896748681898, + "grad_norm": 1.044129729270935, + "learning_rate": 2.750135732575374e-05, + "loss": 0.1662, + "step": 31680 + }, + { + "epoch": 2.6115443760984185, + "grad_norm": 4.191971302032471, + "learning_rate": 2.7472741883188248e-05, + "loss": 0.1698, + "step": 31700 + }, + { + "epoch": 2.6131920035149383, + "grad_norm": 1.4464364051818848, + "learning_rate": 2.7444123168578418e-05, + "loss": 0.1691, + "step": 31720 + }, + { + "epoch": 2.6148396309314585, + "grad_norm": 1.0693591833114624, + "learning_rate": 2.7415501219793833e-05, + "loss": 0.1636, + "step": 31740 + }, + { + "epoch": 2.616487258347979, + "grad_norm": 0.7950401306152344, + "learning_rate": 2.738687607470835e-05, + "loss": 0.1763, + "step": 31760 + }, + { + "epoch": 2.618134885764499, + "grad_norm": 1.1098366975784302, + "learning_rate": 2.7358247771200073e-05, + "loss": 0.1771, + "step": 31780 + }, + { + "epoch": 2.6197825131810193, + "grad_norm": 0.9066492319107056, + "learning_rate": 2.732961634715126e-05, + "loss": 0.1582, + "step": 31800 + }, + { + "epoch": 2.6214301405975395, + "grad_norm": 1.0293641090393066, + "learning_rate": 2.730098184044832e-05, + "loss": 0.1752, + "step": 31820 + }, + { + "epoch": 2.62307776801406, + "grad_norm": 0.5921834111213684, + "learning_rate": 2.7272344288981726e-05, + "loss": 0.1781, + "step": 31840 + }, + { + "epoch": 2.62472539543058, + "grad_norm": 3.703381061553955, + "learning_rate": 2.7243703730645992e-05, + "loss": 0.1809, + "step": 31860 + }, + { + "epoch": 2.6263730228471003, + "grad_norm": 1.139620065689087, + "learning_rate": 2.721506020333961e-05, + "loss": 0.166, + "step": 31880 + }, + { + "epoch": 2.6280206502636205, + "grad_norm": 0.6249837279319763, + "learning_rate": 2.7186413744964982e-05, + "loss": 0.1528, + "step": 31900 + }, + { + "epoch": 2.6296682776801408, + "grad_norm": 0.8693757057189941, + "learning_rate": 2.715776439342842e-05, + "loss": 0.1691, + "step": 31920 + }, + { + "epoch": 2.6313159050966606, + "grad_norm": 7.358799934387207, + "learning_rate": 2.7129112186640027e-05, + "loss": 0.1675, + "step": 31940 + }, + { + "epoch": 2.6329635325131813, + "grad_norm": 1.3295739889144897, + "learning_rate": 2.7100457162513715e-05, + "loss": 0.1754, + "step": 31960 + }, + { + "epoch": 2.634611159929701, + "grad_norm": 0.8648635149002075, + "learning_rate": 2.7071799358967116e-05, + "loss": 0.1611, + "step": 31980 + }, + { + "epoch": 2.6362587873462213, + "grad_norm": 11.785006523132324, + "learning_rate": 2.7043138813921527e-05, + "loss": 0.1681, + "step": 32000 + }, + { + "epoch": 2.6362587873462213, + "eval_loss": 0.5284088253974915, + "eval_runtime": 251.5884, + "eval_samples_per_second": 86.606, + "eval_steps_per_second": 21.654, + "eval_wer": 0.22814880340851512, + "step": 32000 + }, + { + "epoch": 2.6379064147627416, + "grad_norm": 0.9631918668746948, + "learning_rate": 2.7014475565301893e-05, + "loss": 0.1791, + "step": 32020 + }, + { + "epoch": 2.639554042179262, + "grad_norm": 1.5670260190963745, + "learning_rate": 2.698580965103672e-05, + "loss": 0.1808, + "step": 32040 + }, + { + "epoch": 2.641201669595782, + "grad_norm": 30.275543212890625, + "learning_rate": 2.695714110905805e-05, + "loss": 0.1803, + "step": 32060 + }, + { + "epoch": 2.6428492970123023, + "grad_norm": 0.5388301014900208, + "learning_rate": 2.6928469977301397e-05, + "loss": 0.1691, + "step": 32080 + }, + { + "epoch": 2.6444969244288226, + "grad_norm": 1.3333371877670288, + "learning_rate": 2.6899796293705705e-05, + "loss": 0.1516, + "step": 32100 + }, + { + "epoch": 2.646144551845343, + "grad_norm": 1.3209874629974365, + "learning_rate": 2.6871120096213286e-05, + "loss": 0.1551, + "step": 32120 + }, + { + "epoch": 2.647792179261863, + "grad_norm": 1.3346821069717407, + "learning_rate": 2.684244142276981e-05, + "loss": 0.1624, + "step": 32140 + }, + { + "epoch": 2.649439806678383, + "grad_norm": 1.642838478088379, + "learning_rate": 2.6813760311324176e-05, + "loss": 0.1808, + "step": 32160 + }, + { + "epoch": 2.6510874340949035, + "grad_norm": 0.6234583258628845, + "learning_rate": 2.6785076799828546e-05, + "loss": 0.1544, + "step": 32180 + }, + { + "epoch": 2.6527350615114234, + "grad_norm": 1.2087680101394653, + "learning_rate": 2.6756390926238235e-05, + "loss": 0.1571, + "step": 32200 + }, + { + "epoch": 2.6543826889279436, + "grad_norm": 0.8298318386077881, + "learning_rate": 2.672770272851171e-05, + "loss": 0.1613, + "step": 32220 + }, + { + "epoch": 2.656030316344464, + "grad_norm": 0.6398922204971313, + "learning_rate": 2.6699012244610477e-05, + "loss": 0.1685, + "step": 32240 + }, + { + "epoch": 2.657677943760984, + "grad_norm": 0.7772256731987, + "learning_rate": 2.6670319512499104e-05, + "loss": 0.1572, + "step": 32260 + }, + { + "epoch": 2.6593255711775043, + "grad_norm": 0.9730490446090698, + "learning_rate": 2.6641624570145102e-05, + "loss": 0.1633, + "step": 32280 + }, + { + "epoch": 2.6609731985940246, + "grad_norm": 2.9239614009857178, + "learning_rate": 2.6612927455518928e-05, + "loss": 0.1694, + "step": 32300 + }, + { + "epoch": 2.662620826010545, + "grad_norm": 0.7175331711769104, + "learning_rate": 2.6584228206593913e-05, + "loss": 0.1618, + "step": 32320 + }, + { + "epoch": 2.664268453427065, + "grad_norm": 0.7444982528686523, + "learning_rate": 2.6555526861346203e-05, + "loss": 0.181, + "step": 32340 + }, + { + "epoch": 2.6659160808435853, + "grad_norm": 3.278308153152466, + "learning_rate": 2.652682345775472e-05, + "loss": 0.1682, + "step": 32360 + }, + { + "epoch": 2.6675637082601056, + "grad_norm": 1.4923603534698486, + "learning_rate": 2.64981180338011e-05, + "loss": 0.1698, + "step": 32380 + }, + { + "epoch": 2.669211335676626, + "grad_norm": 6.6563897132873535, + "learning_rate": 2.6469410627469683e-05, + "loss": 0.1513, + "step": 32400 + }, + { + "epoch": 2.6708589630931456, + "grad_norm": 3.57399845123291, + "learning_rate": 2.6440701276747404e-05, + "loss": 0.1505, + "step": 32420 + }, + { + "epoch": 2.6725065905096663, + "grad_norm": 0.5343198776245117, + "learning_rate": 2.641199001962377e-05, + "loss": 0.181, + "step": 32440 + }, + { + "epoch": 2.674154217926186, + "grad_norm": 0.8804404735565186, + "learning_rate": 2.638327689409083e-05, + "loss": 0.1773, + "step": 32460 + }, + { + "epoch": 2.6758018453427064, + "grad_norm": 1.2259153127670288, + "learning_rate": 2.63545619381431e-05, + "loss": 0.1703, + "step": 32480 + }, + { + "epoch": 2.6774494727592266, + "grad_norm": 1.8356521129608154, + "learning_rate": 2.6325845189777502e-05, + "loss": 0.1566, + "step": 32500 + }, + { + "epoch": 2.679097100175747, + "grad_norm": 1.4201345443725586, + "learning_rate": 2.6297126686993357e-05, + "loss": 0.1648, + "step": 32520 + }, + { + "epoch": 2.680744727592267, + "grad_norm": 0.9832906126976013, + "learning_rate": 2.626840646779228e-05, + "loss": 0.1673, + "step": 32540 + }, + { + "epoch": 2.6823923550087874, + "grad_norm": 1.2985649108886719, + "learning_rate": 2.623968457017817e-05, + "loss": 0.1626, + "step": 32560 + }, + { + "epoch": 2.6840399824253076, + "grad_norm": 1.3062632083892822, + "learning_rate": 2.6210961032157145e-05, + "loss": 0.1699, + "step": 32580 + }, + { + "epoch": 2.685687609841828, + "grad_norm": 2.640608787536621, + "learning_rate": 2.6182235891737493e-05, + "loss": 0.1629, + "step": 32600 + }, + { + "epoch": 2.687335237258348, + "grad_norm": 1.0804749727249146, + "learning_rate": 2.6153509186929635e-05, + "loss": 0.1662, + "step": 32620 + }, + { + "epoch": 2.6889828646748684, + "grad_norm": 1.186482548713684, + "learning_rate": 2.612478095574604e-05, + "loss": 0.182, + "step": 32640 + }, + { + "epoch": 2.6906304920913886, + "grad_norm": 1.7019797563552856, + "learning_rate": 2.6096051236201217e-05, + "loss": 0.1627, + "step": 32660 + }, + { + "epoch": 2.6922781195079084, + "grad_norm": 1.0984563827514648, + "learning_rate": 2.606732006631163e-05, + "loss": 0.1804, + "step": 32680 + }, + { + "epoch": 2.693925746924429, + "grad_norm": 12.436518669128418, + "learning_rate": 2.6038587484095673e-05, + "loss": 0.174, + "step": 32700 + }, + { + "epoch": 2.695573374340949, + "grad_norm": 0.690076470375061, + "learning_rate": 2.6009853527573585e-05, + "loss": 0.1741, + "step": 32720 + }, + { + "epoch": 2.697221001757469, + "grad_norm": 7.712978363037109, + "learning_rate": 2.5981118234767467e-05, + "loss": 0.186, + "step": 32740 + }, + { + "epoch": 2.6988686291739894, + "grad_norm": 5.45370626449585, + "learning_rate": 2.5952381643701147e-05, + "loss": 0.1873, + "step": 32760 + }, + { + "epoch": 2.7005162565905096, + "grad_norm": 0.6914022564888, + "learning_rate": 2.5923643792400194e-05, + "loss": 0.1692, + "step": 32780 + }, + { + "epoch": 2.70216388400703, + "grad_norm": 2.832401752471924, + "learning_rate": 2.5894904718891834e-05, + "loss": 0.1604, + "step": 32800 + }, + { + "epoch": 2.70381151142355, + "grad_norm": 1.0212492942810059, + "learning_rate": 2.5867601501626415e-05, + "loss": 0.169, + "step": 32820 + }, + { + "epoch": 2.7054591388400704, + "grad_norm": 2.72938871383667, + "learning_rate": 2.583886015419551e-05, + "loss": 0.1496, + "step": 32840 + }, + { + "epoch": 2.7071067662565906, + "grad_norm": 1.3014295101165771, + "learning_rate": 2.581011769674676e-05, + "loss": 0.1735, + "step": 32860 + }, + { + "epoch": 2.708754393673111, + "grad_norm": 0.7176096439361572, + "learning_rate": 2.5781374167313492e-05, + "loss": 0.1582, + "step": 32880 + }, + { + "epoch": 2.7104020210896307, + "grad_norm": 2.523735523223877, + "learning_rate": 2.5752629603930457e-05, + "loss": 0.161, + "step": 32900 + }, + { + "epoch": 2.7120496485061514, + "grad_norm": 4.890264987945557, + "learning_rate": 2.572388404463374e-05, + "loss": 0.1665, + "step": 32920 + }, + { + "epoch": 2.713697275922671, + "grad_norm": 1.3927950859069824, + "learning_rate": 2.5695137527460795e-05, + "loss": 0.178, + "step": 32940 + }, + { + "epoch": 2.7153449033391914, + "grad_norm": 1.3219293355941772, + "learning_rate": 2.566639009045031e-05, + "loss": 0.1692, + "step": 32960 + }, + { + "epoch": 2.7169925307557117, + "grad_norm": 4.5115790367126465, + "learning_rate": 2.5637641771642196e-05, + "loss": 0.1666, + "step": 32980 + }, + { + "epoch": 2.718640158172232, + "grad_norm": 2.417607307434082, + "learning_rate": 2.560889260907754e-05, + "loss": 0.1567, + "step": 33000 + }, + { + "epoch": 2.720287785588752, + "grad_norm": 6.914716720581055, + "learning_rate": 2.558014264079853e-05, + "loss": 0.1693, + "step": 33020 + }, + { + "epoch": 2.7219354130052724, + "grad_norm": 5.992996692657471, + "learning_rate": 2.5551391904848448e-05, + "loss": 0.1729, + "step": 33040 + }, + { + "epoch": 2.7235830404217927, + "grad_norm": 0.9340835213661194, + "learning_rate": 2.552264043927155e-05, + "loss": 0.1823, + "step": 33060 + }, + { + "epoch": 2.725230667838313, + "grad_norm": 0.47994711995124817, + "learning_rate": 2.5493888282113092e-05, + "loss": 0.1607, + "step": 33080 + }, + { + "epoch": 2.726878295254833, + "grad_norm": 2.1940431594848633, + "learning_rate": 2.5465135471419236e-05, + "loss": 0.1561, + "step": 33100 + }, + { + "epoch": 2.7285259226713534, + "grad_norm": 0.7191601991653442, + "learning_rate": 2.5436382045237008e-05, + "loss": 0.1865, + "step": 33120 + }, + { + "epoch": 2.7301735500878737, + "grad_norm": 1.24649977684021, + "learning_rate": 2.5407628041614234e-05, + "loss": 0.1733, + "step": 33140 + }, + { + "epoch": 2.7318211775043935, + "grad_norm": 1.3951886892318726, + "learning_rate": 2.5378873498599535e-05, + "loss": 0.1781, + "step": 33160 + }, + { + "epoch": 2.733468804920914, + "grad_norm": 0.28106406331062317, + "learning_rate": 2.5350118454242217e-05, + "loss": 0.1859, + "step": 33180 + }, + { + "epoch": 2.735116432337434, + "grad_norm": 1.036487102508545, + "learning_rate": 2.532136294659226e-05, + "loss": 0.1637, + "step": 33200 + }, + { + "epoch": 2.736764059753954, + "grad_norm": 0.9746518135070801, + "learning_rate": 2.5292607013700258e-05, + "loss": 0.1625, + "step": 33220 + }, + { + "epoch": 2.7384116871704745, + "grad_norm": 31.79758644104004, + "learning_rate": 2.526385069361738e-05, + "loss": 0.1615, + "step": 33240 + }, + { + "epoch": 2.7400593145869947, + "grad_norm": 0.9092070460319519, + "learning_rate": 2.5235094024395273e-05, + "loss": 0.1646, + "step": 33260 + }, + { + "epoch": 2.741706942003515, + "grad_norm": 0.34284865856170654, + "learning_rate": 2.520633704408608e-05, + "loss": 0.1563, + "step": 33280 + }, + { + "epoch": 2.743354569420035, + "grad_norm": 0.9926866888999939, + "learning_rate": 2.517757979074235e-05, + "loss": 0.1501, + "step": 33300 + }, + { + "epoch": 2.7450021968365554, + "grad_norm": 1.1938321590423584, + "learning_rate": 2.514882230241697e-05, + "loss": 0.1544, + "step": 33320 + }, + { + "epoch": 2.7466498242530757, + "grad_norm": 0.7486788630485535, + "learning_rate": 2.5120064617163162e-05, + "loss": 0.1733, + "step": 33340 + }, + { + "epoch": 2.748297451669596, + "grad_norm": 0.8478560447692871, + "learning_rate": 2.5091306773034405e-05, + "loss": 0.1635, + "step": 33360 + }, + { + "epoch": 2.7499450790861157, + "grad_norm": 0.4918525516986847, + "learning_rate": 2.5062548808084374e-05, + "loss": 0.1574, + "step": 33380 + }, + { + "epoch": 2.7515927065026364, + "grad_norm": 2.3312413692474365, + "learning_rate": 2.5033790760366925e-05, + "loss": 0.1493, + "step": 33400 + }, + { + "epoch": 2.7532403339191562, + "grad_norm": 1.4235730171203613, + "learning_rate": 2.5005032667936006e-05, + "loss": 0.1566, + "step": 33420 + }, + { + "epoch": 2.7548879613356765, + "grad_norm": 1.1899724006652832, + "learning_rate": 2.4976274568845632e-05, + "loss": 0.1996, + "step": 33440 + }, + { + "epoch": 2.7565355887521967, + "grad_norm": 1.3901513814926147, + "learning_rate": 2.494751650114983e-05, + "loss": 0.1829, + "step": 33460 + }, + { + "epoch": 2.758183216168717, + "grad_norm": 3.0261080265045166, + "learning_rate": 2.491875850290257e-05, + "loss": 0.1772, + "step": 33480 + }, + { + "epoch": 2.7598308435852372, + "grad_norm": 1.2606502771377563, + "learning_rate": 2.489000061215775e-05, + "loss": 0.161, + "step": 33500 + }, + { + "epoch": 2.7614784710017575, + "grad_norm": 0.659880518913269, + "learning_rate": 2.486124286696911e-05, + "loss": 0.164, + "step": 33520 + }, + { + "epoch": 2.7631260984182777, + "grad_norm": 0.6361564993858337, + "learning_rate": 2.4832485305390213e-05, + "loss": 0.1718, + "step": 33540 + }, + { + "epoch": 2.764773725834798, + "grad_norm": 0.8560569286346436, + "learning_rate": 2.4803727965474376e-05, + "loss": 0.1758, + "step": 33560 + }, + { + "epoch": 2.766421353251318, + "grad_norm": 0.9023476839065552, + "learning_rate": 2.4774970885274598e-05, + "loss": 0.1525, + "step": 33580 + }, + { + "epoch": 2.7680689806678385, + "grad_norm": 1.1586657762527466, + "learning_rate": 2.474621410284356e-05, + "loss": 0.1567, + "step": 33600 + }, + { + "epoch": 2.7697166080843587, + "grad_norm": 2.3108408451080322, + "learning_rate": 2.4717457656233545e-05, + "loss": 0.1645, + "step": 33620 + }, + { + "epoch": 2.7713642355008785, + "grad_norm": 0.7011807560920715, + "learning_rate": 2.4688701583496392e-05, + "loss": 0.1519, + "step": 33640 + }, + { + "epoch": 2.773011862917399, + "grad_norm": 2.491318702697754, + "learning_rate": 2.4659945922683446e-05, + "loss": 0.1806, + "step": 33660 + }, + { + "epoch": 2.774659490333919, + "grad_norm": 1.146855354309082, + "learning_rate": 2.4631190711845484e-05, + "loss": 0.1637, + "step": 33680 + }, + { + "epoch": 2.7763071177504393, + "grad_norm": 1.3814749717712402, + "learning_rate": 2.4602435989032718e-05, + "loss": 0.159, + "step": 33700 + }, + { + "epoch": 2.7779547451669595, + "grad_norm": 2.1869618892669678, + "learning_rate": 2.4573681792294706e-05, + "loss": 0.1643, + "step": 33720 + }, + { + "epoch": 2.7796023725834798, + "grad_norm": 0.6119910478591919, + "learning_rate": 2.4544928159680303e-05, + "loss": 0.1843, + "step": 33740 + }, + { + "epoch": 2.78125, + "grad_norm": 2.1698923110961914, + "learning_rate": 2.4516175129237627e-05, + "loss": 0.1701, + "step": 33760 + }, + { + "epoch": 2.7828976274165202, + "grad_norm": 1.134415626525879, + "learning_rate": 2.448742273901399e-05, + "loss": 0.1674, + "step": 33780 + }, + { + "epoch": 2.7845452548330405, + "grad_norm": 2.005388021469116, + "learning_rate": 2.445867102705585e-05, + "loss": 0.1645, + "step": 33800 + }, + { + "epoch": 2.7861928822495607, + "grad_norm": 1.6660879850387573, + "learning_rate": 2.442992003140879e-05, + "loss": 0.1693, + "step": 33820 + }, + { + "epoch": 2.787840509666081, + "grad_norm": 1.1064176559448242, + "learning_rate": 2.4401169790117427e-05, + "loss": 0.175, + "step": 33840 + }, + { + "epoch": 2.789488137082601, + "grad_norm": 3.4418516159057617, + "learning_rate": 2.4372420341225395e-05, + "loss": 0.1818, + "step": 33860 + }, + { + "epoch": 2.7911357644991215, + "grad_norm": 1.3131994009017944, + "learning_rate": 2.434367172277526e-05, + "loss": 0.1754, + "step": 33880 + }, + { + "epoch": 2.7927833919156413, + "grad_norm": 2.1762378215789795, + "learning_rate": 2.43149239728085e-05, + "loss": 0.1644, + "step": 33900 + }, + { + "epoch": 2.7944310193321615, + "grad_norm": 1.5846805572509766, + "learning_rate": 2.4286177129365446e-05, + "loss": 0.1762, + "step": 33920 + }, + { + "epoch": 2.796078646748682, + "grad_norm": 0.6951166987419128, + "learning_rate": 2.4257431230485228e-05, + "loss": 0.1704, + "step": 33940 + }, + { + "epoch": 2.797726274165202, + "grad_norm": 0.9165875911712646, + "learning_rate": 2.4228686314205736e-05, + "loss": 0.1813, + "step": 33960 + }, + { + "epoch": 2.7993739015817223, + "grad_norm": 1.009450912475586, + "learning_rate": 2.419994241856352e-05, + "loss": 0.1599, + "step": 33980 + }, + { + "epoch": 2.8010215289982425, + "grad_norm": 0.898774266242981, + "learning_rate": 2.4171199581593825e-05, + "loss": 0.1596, + "step": 34000 + }, + { + "epoch": 2.8010215289982425, + "eval_loss": 0.5283127427101135, + "eval_runtime": 250.9148, + "eval_samples_per_second": 86.838, + "eval_steps_per_second": 21.713, + "eval_wer": 0.22828317325592137, + "step": 34000 + }, + { + "epoch": 2.8026691564147628, + "grad_norm": 23.361583709716797, + "learning_rate": 2.4142457841330478e-05, + "loss": 0.1676, + "step": 34020 + }, + { + "epoch": 2.804316783831283, + "grad_norm": 6.748179912567139, + "learning_rate": 2.4113717235805854e-05, + "loss": 0.1755, + "step": 34040 + }, + { + "epoch": 2.8059644112478033, + "grad_norm": 1.5261645317077637, + "learning_rate": 2.408497780305083e-05, + "loss": 0.1895, + "step": 34060 + }, + { + "epoch": 2.8076120386643235, + "grad_norm": 0.5314382314682007, + "learning_rate": 2.405623958109472e-05, + "loss": 0.1584, + "step": 34080 + }, + { + "epoch": 2.8092596660808438, + "grad_norm": 1.132767915725708, + "learning_rate": 2.402750260796525e-05, + "loss": 0.1689, + "step": 34100 + }, + { + "epoch": 2.8109072934973636, + "grad_norm": 1.2487684488296509, + "learning_rate": 2.3998766921688485e-05, + "loss": 0.1749, + "step": 34120 + }, + { + "epoch": 2.8125549209138843, + "grad_norm": 0.5913691520690918, + "learning_rate": 2.3970032560288788e-05, + "loss": 0.179, + "step": 34140 + }, + { + "epoch": 2.814202548330404, + "grad_norm": 1.5123908519744873, + "learning_rate": 2.3941299561788783e-05, + "loss": 0.1617, + "step": 34160 + }, + { + "epoch": 2.8158501757469243, + "grad_norm": 0.2831098735332489, + "learning_rate": 2.3912567964209264e-05, + "loss": 0.1548, + "step": 34180 + }, + { + "epoch": 2.8174978031634446, + "grad_norm": 4.424227714538574, + "learning_rate": 2.3883837805569188e-05, + "loss": 0.1512, + "step": 34200 + }, + { + "epoch": 2.819145430579965, + "grad_norm": 1.4950264692306519, + "learning_rate": 2.3855109123885612e-05, + "loss": 0.1693, + "step": 34220 + }, + { + "epoch": 2.820793057996485, + "grad_norm": 1.211940884590149, + "learning_rate": 2.3826381957173638e-05, + "loss": 0.1671, + "step": 34240 + }, + { + "epoch": 2.8224406854130053, + "grad_norm": 1.2161052227020264, + "learning_rate": 2.3797656343446338e-05, + "loss": 0.168, + "step": 34260 + }, + { + "epoch": 2.8240883128295255, + "grad_norm": 0.32065290212631226, + "learning_rate": 2.3768932320714758e-05, + "loss": 0.1751, + "step": 34280 + }, + { + "epoch": 2.825735940246046, + "grad_norm": 2.1875534057617188, + "learning_rate": 2.3740209926987828e-05, + "loss": 0.1666, + "step": 34300 + }, + { + "epoch": 2.827383567662566, + "grad_norm": 1.0044718980789185, + "learning_rate": 2.3711489200272326e-05, + "loss": 0.1624, + "step": 34320 + }, + { + "epoch": 2.829031195079086, + "grad_norm": 2.3182806968688965, + "learning_rate": 2.368277017857283e-05, + "loss": 0.1698, + "step": 34340 + }, + { + "epoch": 2.8306788224956065, + "grad_norm": 1.2233198881149292, + "learning_rate": 2.365405289989163e-05, + "loss": 0.1624, + "step": 34360 + }, + { + "epoch": 2.8323264499121263, + "grad_norm": 0.30370575189590454, + "learning_rate": 2.3625337402228754e-05, + "loss": 0.1532, + "step": 34380 + }, + { + "epoch": 2.8339740773286466, + "grad_norm": 2.319612979888916, + "learning_rate": 2.3596623723581842e-05, + "loss": 0.155, + "step": 34400 + }, + { + "epoch": 2.835621704745167, + "grad_norm": 3.8146913051605225, + "learning_rate": 2.3567911901946143e-05, + "loss": 0.173, + "step": 34420 + }, + { + "epoch": 2.837269332161687, + "grad_norm": 3.995103120803833, + "learning_rate": 2.3539201975314447e-05, + "loss": 0.1643, + "step": 34440 + }, + { + "epoch": 2.8389169595782073, + "grad_norm": 1.1264597177505493, + "learning_rate": 2.351049398167702e-05, + "loss": 0.1811, + "step": 34460 + }, + { + "epoch": 2.8405645869947276, + "grad_norm": 1.4618791341781616, + "learning_rate": 2.3481787959021586e-05, + "loss": 0.1668, + "step": 34480 + }, + { + "epoch": 2.842212214411248, + "grad_norm": 1.8342463970184326, + "learning_rate": 2.3453083945333266e-05, + "loss": 0.1635, + "step": 34500 + }, + { + "epoch": 2.843859841827768, + "grad_norm": 1.134645700454712, + "learning_rate": 2.3424381978594505e-05, + "loss": 0.1748, + "step": 34520 + }, + { + "epoch": 2.8455074692442883, + "grad_norm": 0.6016003489494324, + "learning_rate": 2.3395682096785065e-05, + "loss": 0.1651, + "step": 34540 + }, + { + "epoch": 2.8471550966608086, + "grad_norm": 0.5989654064178467, + "learning_rate": 2.3366984337881908e-05, + "loss": 0.1748, + "step": 34560 + }, + { + "epoch": 2.848802724077329, + "grad_norm": 0.23500819504261017, + "learning_rate": 2.333828873985923e-05, + "loss": 0.1656, + "step": 34580 + }, + { + "epoch": 2.8504503514938486, + "grad_norm": 3.5726726055145264, + "learning_rate": 2.3309595340688338e-05, + "loss": 0.1519, + "step": 34600 + }, + { + "epoch": 2.8520979789103693, + "grad_norm": 0.5327156782150269, + "learning_rate": 2.3280904178337648e-05, + "loss": 0.155, + "step": 34620 + }, + { + "epoch": 2.853745606326889, + "grad_norm": 0.4773898422718048, + "learning_rate": 2.3252215290772616e-05, + "loss": 0.182, + "step": 34640 + }, + { + "epoch": 2.8553932337434094, + "grad_norm": 0.9713151454925537, + "learning_rate": 2.3223528715955665e-05, + "loss": 0.1806, + "step": 34660 + }, + { + "epoch": 2.8570408611599296, + "grad_norm": 0.49397534132003784, + "learning_rate": 2.3194844491846178e-05, + "loss": 0.1587, + "step": 34680 + }, + { + "epoch": 2.85868848857645, + "grad_norm": 5.4063029289245605, + "learning_rate": 2.3166162656400418e-05, + "loss": 0.1492, + "step": 34700 + }, + { + "epoch": 2.86033611599297, + "grad_norm": 1.0393853187561035, + "learning_rate": 2.3137483247571505e-05, + "loss": 0.1624, + "step": 34720 + }, + { + "epoch": 2.8619837434094904, + "grad_norm": 1.0829802751541138, + "learning_rate": 2.3108806303309337e-05, + "loss": 0.173, + "step": 34740 + }, + { + "epoch": 2.8636313708260106, + "grad_norm": 1.1949620246887207, + "learning_rate": 2.3081565523627518e-05, + "loss": 0.1661, + "step": 34760 + }, + { + "epoch": 2.865278998242531, + "grad_norm": 0.9706825613975525, + "learning_rate": 2.305289349441147e-05, + "loss": 0.153, + "step": 34780 + }, + { + "epoch": 2.866926625659051, + "grad_norm": 1.9239734411239624, + "learning_rate": 2.302422404169516e-05, + "loss": 0.1571, + "step": 34800 + }, + { + "epoch": 2.868574253075571, + "grad_norm": 6.778972148895264, + "learning_rate": 2.2995557203415314e-05, + "loss": 0.1624, + "step": 34820 + }, + { + "epoch": 2.8702218804920916, + "grad_norm": 0.6842126250267029, + "learning_rate": 2.2966893017505192e-05, + "loss": 0.1684, + "step": 34840 + }, + { + "epoch": 2.8718695079086114, + "grad_norm": 1.1653012037277222, + "learning_rate": 2.2938231521894544e-05, + "loss": 0.1662, + "step": 34860 + }, + { + "epoch": 2.8735171353251316, + "grad_norm": 1.9263895750045776, + "learning_rate": 2.2909572754509558e-05, + "loss": 0.1609, + "step": 34880 + }, + { + "epoch": 2.875164762741652, + "grad_norm": 3.092329978942871, + "learning_rate": 2.2880916753272824e-05, + "loss": 0.1626, + "step": 34900 + }, + { + "epoch": 2.876812390158172, + "grad_norm": 0.6692825555801392, + "learning_rate": 2.2852263556103263e-05, + "loss": 0.1648, + "step": 34920 + }, + { + "epoch": 2.8784600175746924, + "grad_norm": 1.1220498085021973, + "learning_rate": 2.2823613200916095e-05, + "loss": 0.1957, + "step": 34940 + }, + { + "epoch": 2.8801076449912126, + "grad_norm": 1.7044694423675537, + "learning_rate": 2.279496572562275e-05, + "loss": 0.1757, + "step": 34960 + }, + { + "epoch": 2.881755272407733, + "grad_norm": 0.3736783266067505, + "learning_rate": 2.2766321168130876e-05, + "loss": 0.1637, + "step": 34980 + }, + { + "epoch": 2.883402899824253, + "grad_norm": 3.305779457092285, + "learning_rate": 2.273767956634426e-05, + "loss": 0.1617, + "step": 35000 + }, + { + "epoch": 2.8850505272407734, + "grad_norm": 0.9354811310768127, + "learning_rate": 2.2709040958162764e-05, + "loss": 0.1711, + "step": 35020 + }, + { + "epoch": 2.8866981546572936, + "grad_norm": 0.9158414006233215, + "learning_rate": 2.2680405381482305e-05, + "loss": 0.1756, + "step": 35040 + }, + { + "epoch": 2.888345782073814, + "grad_norm": 1.0217347145080566, + "learning_rate": 2.265177287419476e-05, + "loss": 0.1711, + "step": 35060 + }, + { + "epoch": 2.8899934094903337, + "grad_norm": 0.9840256571769714, + "learning_rate": 2.262314347418797e-05, + "loss": 0.1628, + "step": 35080 + }, + { + "epoch": 2.8916410369068544, + "grad_norm": 4.251138210296631, + "learning_rate": 2.259451721934566e-05, + "loss": 0.1616, + "step": 35100 + }, + { + "epoch": 2.893288664323374, + "grad_norm": 6.069693565368652, + "learning_rate": 2.2565894147547397e-05, + "loss": 0.1628, + "step": 35120 + }, + { + "epoch": 2.8949362917398944, + "grad_norm": 0.44294020533561707, + "learning_rate": 2.2537274296668513e-05, + "loss": 0.1791, + "step": 35140 + }, + { + "epoch": 2.8965839191564147, + "grad_norm": 1.3097333908081055, + "learning_rate": 2.2508657704580104e-05, + "loss": 0.1795, + "step": 35160 + }, + { + "epoch": 2.898231546572935, + "grad_norm": 0.8062418103218079, + "learning_rate": 2.2480044409148947e-05, + "loss": 0.1886, + "step": 35180 + }, + { + "epoch": 2.899879173989455, + "grad_norm": 1.6019326448440552, + "learning_rate": 2.2451434448237448e-05, + "loss": 0.1498, + "step": 35200 + }, + { + "epoch": 2.9015268014059754, + "grad_norm": 1.6042932271957397, + "learning_rate": 2.242282785970361e-05, + "loss": 0.1691, + "step": 35220 + }, + { + "epoch": 2.9031744288224957, + "grad_norm": 0.9575341939926147, + "learning_rate": 2.2394224681400962e-05, + "loss": 0.1647, + "step": 35240 + }, + { + "epoch": 2.904822056239016, + "grad_norm": 0.8834397792816162, + "learning_rate": 2.2365624951178535e-05, + "loss": 0.1652, + "step": 35260 + }, + { + "epoch": 2.906469683655536, + "grad_norm": 0.6135185360908508, + "learning_rate": 2.2337028706880792e-05, + "loss": 0.1763, + "step": 35280 + }, + { + "epoch": 2.9081173110720564, + "grad_norm": 1.6442698240280151, + "learning_rate": 2.2308435986347577e-05, + "loss": 0.1683, + "step": 35300 + }, + { + "epoch": 2.9097649384885766, + "grad_norm": 1.1094133853912354, + "learning_rate": 2.2279846827414088e-05, + "loss": 0.1597, + "step": 35320 + }, + { + "epoch": 2.9114125659050965, + "grad_norm": 0.7289668917655945, + "learning_rate": 2.2251261267910777e-05, + "loss": 0.1767, + "step": 35340 + }, + { + "epoch": 2.913060193321617, + "grad_norm": 0.8773306012153625, + "learning_rate": 2.222267934566337e-05, + "loss": 0.1597, + "step": 35360 + }, + { + "epoch": 2.914707820738137, + "grad_norm": 0.398133248090744, + "learning_rate": 2.2194101098492758e-05, + "loss": 0.1705, + "step": 35380 + }, + { + "epoch": 2.916355448154657, + "grad_norm": 1.7705035209655762, + "learning_rate": 2.216552656421498e-05, + "loss": 0.1636, + "step": 35400 + }, + { + "epoch": 2.9180030755711774, + "grad_norm": 0.6762025952339172, + "learning_rate": 2.2136955780641165e-05, + "loss": 0.1661, + "step": 35420 + }, + { + "epoch": 2.9196507029876977, + "grad_norm": 0.7774632573127747, + "learning_rate": 2.2108388785577448e-05, + "loss": 0.1737, + "step": 35440 + }, + { + "epoch": 2.921298330404218, + "grad_norm": 0.9466442465782166, + "learning_rate": 2.2079825616824988e-05, + "loss": 0.1779, + "step": 35460 + }, + { + "epoch": 2.922945957820738, + "grad_norm": 0.8889173269271851, + "learning_rate": 2.2051266312179865e-05, + "loss": 0.1699, + "step": 35480 + }, + { + "epoch": 2.9245935852372584, + "grad_norm": 1.440914273262024, + "learning_rate": 2.202271090943305e-05, + "loss": 0.16, + "step": 35500 + }, + { + "epoch": 2.9262412126537787, + "grad_norm": 0.5880303978919983, + "learning_rate": 2.1994159446370356e-05, + "loss": 0.1718, + "step": 35520 + }, + { + "epoch": 2.927888840070299, + "grad_norm": 1.3823106288909912, + "learning_rate": 2.1965611960772354e-05, + "loss": 0.1811, + "step": 35540 + }, + { + "epoch": 2.9295364674868187, + "grad_norm": 1.034883975982666, + "learning_rate": 2.193706849041439e-05, + "loss": 0.1904, + "step": 35560 + }, + { + "epoch": 2.9311840949033394, + "grad_norm": 1.2182835340499878, + "learning_rate": 2.1908529073066476e-05, + "loss": 0.1596, + "step": 35580 + }, + { + "epoch": 2.9328317223198592, + "grad_norm": 1.7504252195358276, + "learning_rate": 2.1879993746493262e-05, + "loss": 0.1579, + "step": 35600 + }, + { + "epoch": 2.9344793497363795, + "grad_norm": 0.9394094347953796, + "learning_rate": 2.1851462548454e-05, + "loss": 0.1782, + "step": 35620 + }, + { + "epoch": 2.9361269771528997, + "grad_norm": 0.9647118449211121, + "learning_rate": 2.182293551670245e-05, + "loss": 0.1822, + "step": 35640 + }, + { + "epoch": 2.93777460456942, + "grad_norm": 1.628814935684204, + "learning_rate": 2.179441268898688e-05, + "loss": 0.174, + "step": 35660 + }, + { + "epoch": 2.93942223198594, + "grad_norm": 0.8225361704826355, + "learning_rate": 2.1765894103049992e-05, + "loss": 0.162, + "step": 35680 + }, + { + "epoch": 2.9410698594024605, + "grad_norm": 1.2492133378982544, + "learning_rate": 2.1737379796628882e-05, + "loss": 0.1508, + "step": 35700 + }, + { + "epoch": 2.9427174868189807, + "grad_norm": 0.734495222568512, + "learning_rate": 2.1708869807454974e-05, + "loss": 0.1672, + "step": 35720 + }, + { + "epoch": 2.944365114235501, + "grad_norm": 0.8144159317016602, + "learning_rate": 2.1680364173253964e-05, + "loss": 0.1717, + "step": 35740 + }, + { + "epoch": 2.946012741652021, + "grad_norm": 0.5943552255630493, + "learning_rate": 2.1651862931745815e-05, + "loss": 0.1584, + "step": 35760 + }, + { + "epoch": 2.9476603690685415, + "grad_norm": 0.5486462116241455, + "learning_rate": 2.162336612064466e-05, + "loss": 0.1537, + "step": 35780 + }, + { + "epoch": 2.9493079964850617, + "grad_norm": 1.7498178482055664, + "learning_rate": 2.1594873777658772e-05, + "loss": 0.1583, + "step": 35800 + }, + { + "epoch": 2.9509556239015815, + "grad_norm": 0.6559889316558838, + "learning_rate": 2.156638594049052e-05, + "loss": 0.1789, + "step": 35820 + }, + { + "epoch": 2.952603251318102, + "grad_norm": 1.0158253908157349, + "learning_rate": 2.1537902646836298e-05, + "loss": 0.1746, + "step": 35840 + }, + { + "epoch": 2.954250878734622, + "grad_norm": 2.991706132888794, + "learning_rate": 2.150942393438649e-05, + "loss": 0.1604, + "step": 35860 + }, + { + "epoch": 2.9558985061511422, + "grad_norm": 1.1813044548034668, + "learning_rate": 2.148094984082543e-05, + "loss": 0.1537, + "step": 35880 + }, + { + "epoch": 2.9575461335676625, + "grad_norm": 1.1172895431518555, + "learning_rate": 2.1452480403831323e-05, + "loss": 0.1494, + "step": 35900 + }, + { + "epoch": 2.9591937609841827, + "grad_norm": 5.81532096862793, + "learning_rate": 2.142401566107624e-05, + "loss": 0.1671, + "step": 35920 + }, + { + "epoch": 2.960841388400703, + "grad_norm": 9.261301040649414, + "learning_rate": 2.1395555650225993e-05, + "loss": 0.1748, + "step": 35940 + }, + { + "epoch": 2.9624890158172232, + "grad_norm": 0.48542654514312744, + "learning_rate": 2.1367100408940176e-05, + "loss": 0.1812, + "step": 35960 + }, + { + "epoch": 2.9641366432337435, + "grad_norm": 0.28661516308784485, + "learning_rate": 2.133864997487205e-05, + "loss": 0.157, + "step": 35980 + }, + { + "epoch": 2.9657842706502637, + "grad_norm": 0.7356573343276978, + "learning_rate": 2.1310204385668527e-05, + "loss": 0.148, + "step": 36000 + }, + { + "epoch": 2.9657842706502637, + "eval_loss": 0.5161551833152771, + "eval_runtime": 250.2579, + "eval_samples_per_second": 87.066, + "eval_steps_per_second": 21.77, + "eval_wer": 0.22239329494461443, + "step": 36000 + }, + { + "epoch": 2.967486818980668, + "grad_norm": 1.018925428390503, + "learning_rate": 2.1281763678970113e-05, + "loss": 0.1475, + "step": 36020 + }, + { + "epoch": 2.969134446397188, + "grad_norm": 0.5352170467376709, + "learning_rate": 2.1254749564304577e-05, + "loss": 0.1641, + "step": 36040 + }, + { + "epoch": 2.9707820738137083, + "grad_norm": 1.6044977903366089, + "learning_rate": 2.122631848673005e-05, + "loss": 0.1644, + "step": 36060 + }, + { + "epoch": 2.9724297012302285, + "grad_norm": 0.3798460066318512, + "learning_rate": 2.1197892402662252e-05, + "loss": 0.1449, + "step": 36080 + }, + { + "epoch": 2.974077328646749, + "grad_norm": 1.5500409603118896, + "learning_rate": 2.1169471349715874e-05, + "loss": 0.14, + "step": 36100 + }, + { + "epoch": 2.975724956063269, + "grad_norm": 1.157693862915039, + "learning_rate": 2.114105536549893e-05, + "loss": 0.1439, + "step": 36120 + }, + { + "epoch": 2.977372583479789, + "grad_norm": 0.46837174892425537, + "learning_rate": 2.1112644487612743e-05, + "loss": 0.1555, + "step": 36140 + }, + { + "epoch": 2.9790202108963095, + "grad_norm": 4.228370189666748, + "learning_rate": 2.108423875365188e-05, + "loss": 0.1484, + "step": 36160 + }, + { + "epoch": 2.9806678383128293, + "grad_norm": 0.4278930425643921, + "learning_rate": 2.10558382012041e-05, + "loss": 0.1428, + "step": 36180 + }, + { + "epoch": 2.9823154657293496, + "grad_norm": 1.5068235397338867, + "learning_rate": 2.102744286785031e-05, + "loss": 0.1393, + "step": 36200 + }, + { + "epoch": 2.98396309314587, + "grad_norm": 0.6257343888282776, + "learning_rate": 2.0999052791164483e-05, + "loss": 0.1385, + "step": 36220 + }, + { + "epoch": 2.98561072056239, + "grad_norm": 0.6275610327720642, + "learning_rate": 2.0970668008713677e-05, + "loss": 0.162, + "step": 36240 + }, + { + "epoch": 2.9872583479789103, + "grad_norm": 2.056185722351074, + "learning_rate": 2.0942288558057915e-05, + "loss": 0.1513, + "step": 36260 + }, + { + "epoch": 2.9889059753954306, + "grad_norm": 0.30075156688690186, + "learning_rate": 2.091391447675018e-05, + "loss": 0.1463, + "step": 36280 + }, + { + "epoch": 2.990553602811951, + "grad_norm": 3.691248655319214, + "learning_rate": 2.0885545802336355e-05, + "loss": 0.154, + "step": 36300 + }, + { + "epoch": 2.992201230228471, + "grad_norm": 1.2786400318145752, + "learning_rate": 2.0857182572355133e-05, + "loss": 0.1622, + "step": 36320 + }, + { + "epoch": 2.9938488576449913, + "grad_norm": 1.1107524633407593, + "learning_rate": 2.082882482433804e-05, + "loss": 0.1603, + "step": 36340 + }, + { + "epoch": 2.9954964850615116, + "grad_norm": 0.4886188507080078, + "learning_rate": 2.080047259580934e-05, + "loss": 0.1627, + "step": 36360 + }, + { + "epoch": 2.997144112478032, + "grad_norm": 0.2950489819049835, + "learning_rate": 2.0772125924285983e-05, + "loss": 0.1525, + "step": 36380 + }, + { + "epoch": 2.9987917398945516, + "grad_norm": 5.7747111320495605, + "learning_rate": 2.0743784847277586e-05, + "loss": 0.1378, + "step": 36400 + }, + { + "epoch": 3.000439367311072, + "grad_norm": 3.0797595977783203, + "learning_rate": 2.0715449402286327e-05, + "loss": 0.153, + "step": 36420 + }, + { + "epoch": 3.002086994727592, + "grad_norm": 3.7219903469085693, + "learning_rate": 2.0687119626806965e-05, + "loss": 0.1586, + "step": 36440 + }, + { + "epoch": 3.0037346221441124, + "grad_norm": 0.7922471165657043, + "learning_rate": 2.0658795558326743e-05, + "loss": 0.1308, + "step": 36460 + }, + { + "epoch": 3.0053822495606326, + "grad_norm": 0.8190672397613525, + "learning_rate": 2.063047723432536e-05, + "loss": 0.1516, + "step": 36480 + }, + { + "epoch": 3.007029876977153, + "grad_norm": 0.46464812755584717, + "learning_rate": 2.0602164692274904e-05, + "loss": 0.1593, + "step": 36500 + }, + { + "epoch": 3.008677504393673, + "grad_norm": 1.922914981842041, + "learning_rate": 2.0573857969639813e-05, + "loss": 0.1499, + "step": 36520 + }, + { + "epoch": 3.0103251318101933, + "grad_norm": 1.359255313873291, + "learning_rate": 2.0545557103876832e-05, + "loss": 0.1452, + "step": 36540 + }, + { + "epoch": 3.0119727592267136, + "grad_norm": 1.6646554470062256, + "learning_rate": 2.0517262132434947e-05, + "loss": 0.134, + "step": 36560 + }, + { + "epoch": 3.013620386643234, + "grad_norm": 0.5390244722366333, + "learning_rate": 2.0488973092755355e-05, + "loss": 0.1459, + "step": 36580 + }, + { + "epoch": 3.015268014059754, + "grad_norm": 0.6744359731674194, + "learning_rate": 2.0460690022271405e-05, + "loss": 0.1543, + "step": 36600 + }, + { + "epoch": 3.0169156414762743, + "grad_norm": 0.5361171960830688, + "learning_rate": 2.0432412958408524e-05, + "loss": 0.1612, + "step": 36620 + }, + { + "epoch": 3.0185632688927946, + "grad_norm": 1.7773494720458984, + "learning_rate": 2.040414193858421e-05, + "loss": 0.1637, + "step": 36640 + }, + { + "epoch": 3.0202108963093144, + "grad_norm": 0.7252753376960754, + "learning_rate": 2.0375877000207963e-05, + "loss": 0.1275, + "step": 36660 + }, + { + "epoch": 3.0218585237258346, + "grad_norm": 0.7316433787345886, + "learning_rate": 2.0347618180681232e-05, + "loss": 0.1707, + "step": 36680 + }, + { + "epoch": 3.023506151142355, + "grad_norm": 0.9963041543960571, + "learning_rate": 2.0319365517397385e-05, + "loss": 0.1606, + "step": 36700 + }, + { + "epoch": 3.025153778558875, + "grad_norm": 1.5264005661010742, + "learning_rate": 2.02911190477416e-05, + "loss": 0.1484, + "step": 36720 + }, + { + "epoch": 3.0268014059753954, + "grad_norm": 4.306487560272217, + "learning_rate": 2.026287880909091e-05, + "loss": 0.1769, + "step": 36740 + }, + { + "epoch": 3.0284490333919156, + "grad_norm": 0.7492676377296448, + "learning_rate": 2.023464483881407e-05, + "loss": 0.1336, + "step": 36760 + }, + { + "epoch": 3.030096660808436, + "grad_norm": 3.7590513229370117, + "learning_rate": 2.0206417174271557e-05, + "loss": 0.1541, + "step": 36780 + }, + { + "epoch": 3.031744288224956, + "grad_norm": 0.990483283996582, + "learning_rate": 2.0178195852815508e-05, + "loss": 0.1557, + "step": 36800 + }, + { + "epoch": 3.0333919156414764, + "grad_norm": 2.186178207397461, + "learning_rate": 2.014998091178964e-05, + "loss": 0.1451, + "step": 36820 + }, + { + "epoch": 3.0350395430579966, + "grad_norm": 3.0323667526245117, + "learning_rate": 2.0121772388529247e-05, + "loss": 0.1503, + "step": 36840 + }, + { + "epoch": 3.036687170474517, + "grad_norm": 0.8073984980583191, + "learning_rate": 2.009357032036113e-05, + "loss": 0.1241, + "step": 36860 + }, + { + "epoch": 3.038334797891037, + "grad_norm": 0.6493812799453735, + "learning_rate": 2.0065374744603555e-05, + "loss": 0.1553, + "step": 36880 + }, + { + "epoch": 3.039982425307557, + "grad_norm": 0.4456873834133148, + "learning_rate": 2.0037185698566175e-05, + "loss": 0.1546, + "step": 36900 + }, + { + "epoch": 3.041630052724077, + "grad_norm": 0.3840373456478119, + "learning_rate": 2.0009003219550018e-05, + "loss": 0.1624, + "step": 36920 + }, + { + "epoch": 3.0432776801405974, + "grad_norm": 1.3345998525619507, + "learning_rate": 1.9980827344847425e-05, + "loss": 0.1505, + "step": 36940 + }, + { + "epoch": 3.0449253075571177, + "grad_norm": 1.5938456058502197, + "learning_rate": 1.9952658111741994e-05, + "loss": 0.1283, + "step": 36960 + }, + { + "epoch": 3.046572934973638, + "grad_norm": 0.8933133482933044, + "learning_rate": 1.992449555750854e-05, + "loss": 0.1597, + "step": 36980 + }, + { + "epoch": 3.048220562390158, + "grad_norm": 0.4933435022830963, + "learning_rate": 1.989633971941301e-05, + "loss": 0.1621, + "step": 37000 + }, + { + "epoch": 3.0498681898066784, + "grad_norm": 0.9858216643333435, + "learning_rate": 1.9868190634712504e-05, + "loss": 0.1589, + "step": 37020 + }, + { + "epoch": 3.0515158172231986, + "grad_norm": 3.218104124069214, + "learning_rate": 1.984004834065516e-05, + "loss": 0.1659, + "step": 37040 + }, + { + "epoch": 3.053163444639719, + "grad_norm": 0.7501934766769409, + "learning_rate": 1.9811912874480154e-05, + "loss": 0.1294, + "step": 37060 + }, + { + "epoch": 3.054811072056239, + "grad_norm": 0.9431729912757874, + "learning_rate": 1.9783784273417612e-05, + "loss": 0.1668, + "step": 37080 + }, + { + "epoch": 3.0564586994727594, + "grad_norm": 1.0041712522506714, + "learning_rate": 1.9755662574688555e-05, + "loss": 0.1619, + "step": 37100 + }, + { + "epoch": 3.0581063268892796, + "grad_norm": 1.4895607233047485, + "learning_rate": 1.9727547815504903e-05, + "loss": 0.1646, + "step": 37120 + }, + { + "epoch": 3.0597539543057994, + "grad_norm": 2.0483157634735107, + "learning_rate": 1.969944003306939e-05, + "loss": 0.1601, + "step": 37140 + }, + { + "epoch": 3.0614015817223197, + "grad_norm": 1.0434845685958862, + "learning_rate": 1.96713392645755e-05, + "loss": 0.1326, + "step": 37160 + }, + { + "epoch": 3.06304920913884, + "grad_norm": 0.5273759961128235, + "learning_rate": 1.9643245547207453e-05, + "loss": 0.1556, + "step": 37180 + }, + { + "epoch": 3.06469683655536, + "grad_norm": 0.9159584045410156, + "learning_rate": 1.9615158918140126e-05, + "loss": 0.1678, + "step": 37200 + }, + { + "epoch": 3.0663444639718804, + "grad_norm": 0.8162606358528137, + "learning_rate": 1.9587079414539022e-05, + "loss": 0.1401, + "step": 37220 + }, + { + "epoch": 3.0679920913884007, + "grad_norm": 16.133909225463867, + "learning_rate": 1.9559007073560217e-05, + "loss": 0.1578, + "step": 37240 + }, + { + "epoch": 3.069639718804921, + "grad_norm": 0.6620104312896729, + "learning_rate": 1.953094193235031e-05, + "loss": 0.1345, + "step": 37260 + }, + { + "epoch": 3.071287346221441, + "grad_norm": 0.8160563707351685, + "learning_rate": 1.950288402804637e-05, + "loss": 0.1483, + "step": 37280 + }, + { + "epoch": 3.0729349736379614, + "grad_norm": 0.6311087012290955, + "learning_rate": 1.9474833397775887e-05, + "loss": 0.1596, + "step": 37300 + }, + { + "epoch": 3.0745826010544817, + "grad_norm": 1.5575555562973022, + "learning_rate": 1.944679007865672e-05, + "loss": 0.1478, + "step": 37320 + }, + { + "epoch": 3.076230228471002, + "grad_norm": 1.3235657215118408, + "learning_rate": 1.9420155731246147e-05, + "loss": 0.1656, + "step": 37340 + }, + { + "epoch": 3.077877855887522, + "grad_norm": 2.973172903060913, + "learning_rate": 1.9392126775595654e-05, + "loss": 0.1311, + "step": 37360 + }, + { + "epoch": 3.079525483304042, + "grad_norm": 0.46286314725875854, + "learning_rate": 1.936410524053765e-05, + "loss": 0.1577, + "step": 37380 + }, + { + "epoch": 3.081173110720562, + "grad_norm": 1.0705723762512207, + "learning_rate": 1.933609116315149e-05, + "loss": 0.1565, + "step": 37400 + }, + { + "epoch": 3.0828207381370825, + "grad_norm": 0.6402031183242798, + "learning_rate": 1.9308084580506677e-05, + "loss": 0.1449, + "step": 37420 + }, + { + "epoch": 3.0844683655536027, + "grad_norm": 5.519851207733154, + "learning_rate": 1.9280085529662798e-05, + "loss": 0.1702, + "step": 37440 + }, + { + "epoch": 3.086115992970123, + "grad_norm": 1.0702592134475708, + "learning_rate": 1.9252094047669464e-05, + "loss": 0.1253, + "step": 37460 + }, + { + "epoch": 3.087763620386643, + "grad_norm": 0.9042578935623169, + "learning_rate": 1.9224110171566288e-05, + "loss": 0.1471, + "step": 37480 + }, + { + "epoch": 3.0894112478031635, + "grad_norm": 0.34733909368515015, + "learning_rate": 1.9196133938382785e-05, + "loss": 0.1534, + "step": 37500 + }, + { + "epoch": 3.0910588752196837, + "grad_norm": 1.277514100074768, + "learning_rate": 1.9168165385138385e-05, + "loss": 0.149, + "step": 37520 + }, + { + "epoch": 3.092706502636204, + "grad_norm": 1.3620398044586182, + "learning_rate": 1.9140204548842348e-05, + "loss": 0.1538, + "step": 37540 + }, + { + "epoch": 3.094354130052724, + "grad_norm": 0.6622684001922607, + "learning_rate": 1.9112251466493725e-05, + "loss": 0.1328, + "step": 37560 + }, + { + "epoch": 3.0960017574692444, + "grad_norm": 7.100107669830322, + "learning_rate": 1.9084306175081303e-05, + "loss": 0.1455, + "step": 37580 + }, + { + "epoch": 3.0976493848857647, + "grad_norm": 0.45634210109710693, + "learning_rate": 1.9056368711583558e-05, + "loss": 0.1499, + "step": 37600 + }, + { + "epoch": 3.0992970123022845, + "grad_norm": 1.493963360786438, + "learning_rate": 1.902843911296861e-05, + "loss": 0.1531, + "step": 37620 + }, + { + "epoch": 3.1009446397188047, + "grad_norm": 1.5713478326797485, + "learning_rate": 1.9000517416194174e-05, + "loss": 0.1624, + "step": 37640 + }, + { + "epoch": 3.102592267135325, + "grad_norm": 0.6847428679466248, + "learning_rate": 1.8972603658207518e-05, + "loss": 0.1328, + "step": 37660 + }, + { + "epoch": 3.1042398945518452, + "grad_norm": 0.9931829571723938, + "learning_rate": 1.8944697875945374e-05, + "loss": 0.1476, + "step": 37680 + }, + { + "epoch": 3.1058875219683655, + "grad_norm": 0.6386281847953796, + "learning_rate": 1.8916800106333946e-05, + "loss": 0.1618, + "step": 37700 + }, + { + "epoch": 3.1075351493848857, + "grad_norm": 0.6640763282775879, + "learning_rate": 1.8888910386288832e-05, + "loss": 0.1491, + "step": 37720 + }, + { + "epoch": 3.109182776801406, + "grad_norm": 1.5718262195587158, + "learning_rate": 1.886102875271498e-05, + "loss": 0.157, + "step": 37740 + }, + { + "epoch": 3.1108304042179262, + "grad_norm": 1.4136067628860474, + "learning_rate": 1.8833155242506635e-05, + "loss": 0.1376, + "step": 37760 + }, + { + "epoch": 3.1124780316344465, + "grad_norm": 1.0528970956802368, + "learning_rate": 1.880528989254728e-05, + "loss": 0.1644, + "step": 37780 + }, + { + "epoch": 3.1141256590509667, + "grad_norm": 0.6771901845932007, + "learning_rate": 1.8777432739709616e-05, + "loss": 0.1622, + "step": 37800 + }, + { + "epoch": 3.115773286467487, + "grad_norm": 0.9342694282531738, + "learning_rate": 1.8749583820855493e-05, + "loss": 0.1425, + "step": 37820 + }, + { + "epoch": 3.117420913884007, + "grad_norm": 11.285189628601074, + "learning_rate": 1.8721743172835872e-05, + "loss": 0.161, + "step": 37840 + }, + { + "epoch": 3.1190685413005275, + "grad_norm": 1.2589502334594727, + "learning_rate": 1.869391083249075e-05, + "loss": 0.1379, + "step": 37860 + }, + { + "epoch": 3.1207161687170473, + "grad_norm": 0.5863010287284851, + "learning_rate": 1.8666086836649156e-05, + "loss": 0.1573, + "step": 37880 + }, + { + "epoch": 3.1223637961335675, + "grad_norm": 0.6433321833610535, + "learning_rate": 1.8638271222129056e-05, + "loss": 0.1447, + "step": 37900 + }, + { + "epoch": 3.1240114235500878, + "grad_norm": 0.7037969827651978, + "learning_rate": 1.8610464025737336e-05, + "loss": 0.1445, + "step": 37920 + }, + { + "epoch": 3.125659050966608, + "grad_norm": 3.2603673934936523, + "learning_rate": 1.858266528426974e-05, + "loss": 0.1718, + "step": 37940 + }, + { + "epoch": 3.1273066783831283, + "grad_norm": 0.9228240847587585, + "learning_rate": 1.855487503451084e-05, + "loss": 0.1432, + "step": 37960 + }, + { + "epoch": 3.1289543057996485, + "grad_norm": 0.524641752243042, + "learning_rate": 1.8527093313233927e-05, + "loss": 0.1568, + "step": 37980 + }, + { + "epoch": 3.1306019332161688, + "grad_norm": 0.6379425525665283, + "learning_rate": 1.8499320157201057e-05, + "loss": 0.1576, + "step": 38000 + }, + { + "epoch": 3.1306019332161688, + "eval_loss": 0.5186827778816223, + "eval_runtime": 810.9319, + "eval_samples_per_second": 26.869, + "eval_steps_per_second": 6.718, + "eval_wer": 0.2234430593774757, + "step": 38000 + }, + { + "epoch": 3.132304481546573, + "grad_norm": 0.6488630771636963, + "learning_rate": 1.8471555603162927e-05, + "loss": 0.1522, + "step": 38020 + }, + { + "epoch": 3.133952108963093, + "grad_norm": 0.13922233879566193, + "learning_rate": 1.844379968785885e-05, + "loss": 0.159, + "step": 38040 + }, + { + "epoch": 3.1355997363796133, + "grad_norm": 1.2710275650024414, + "learning_rate": 1.8416052448016736e-05, + "loss": 0.1262, + "step": 38060 + }, + { + "epoch": 3.1372473637961336, + "grad_norm": 1.2879269123077393, + "learning_rate": 1.8388313920352956e-05, + "loss": 0.1641, + "step": 38080 + }, + { + "epoch": 3.138894991212654, + "grad_norm": 0.41094842553138733, + "learning_rate": 1.8360584141572414e-05, + "loss": 0.1609, + "step": 38100 + }, + { + "epoch": 3.140542618629174, + "grad_norm": 0.3641243875026703, + "learning_rate": 1.8332863148368406e-05, + "loss": 0.1506, + "step": 38120 + }, + { + "epoch": 3.1421902460456943, + "grad_norm": 0.9135380983352661, + "learning_rate": 1.830515097742261e-05, + "loss": 0.1666, + "step": 38140 + }, + { + "epoch": 3.1438378734622145, + "grad_norm": 2.1218717098236084, + "learning_rate": 1.827744766540503e-05, + "loss": 0.1353, + "step": 38160 + }, + { + "epoch": 3.145485500878735, + "grad_norm": 1.940061092376709, + "learning_rate": 1.824975324897394e-05, + "loss": 0.1658, + "step": 38180 + }, + { + "epoch": 3.1471331282952546, + "grad_norm": 0.5710387229919434, + "learning_rate": 1.8222067764775856e-05, + "loss": 0.1621, + "step": 38200 + }, + { + "epoch": 3.148780755711775, + "grad_norm": 1.434187412261963, + "learning_rate": 1.819439124944547e-05, + "loss": 0.1555, + "step": 38220 + }, + { + "epoch": 3.150428383128295, + "grad_norm": 0.399093896150589, + "learning_rate": 1.8166723739605598e-05, + "loss": 0.1665, + "step": 38240 + }, + { + "epoch": 3.1520760105448153, + "grad_norm": 1.5391314029693604, + "learning_rate": 1.8139065271867152e-05, + "loss": 0.136, + "step": 38260 + }, + { + "epoch": 3.1537236379613356, + "grad_norm": 0.7850829362869263, + "learning_rate": 1.811141588282906e-05, + "loss": 0.1549, + "step": 38280 + }, + { + "epoch": 3.155371265377856, + "grad_norm": 0.7228546738624573, + "learning_rate": 1.8083775609078254e-05, + "loss": 0.141, + "step": 38300 + }, + { + "epoch": 3.157018892794376, + "grad_norm": 0.5662539601325989, + "learning_rate": 1.8056144487189602e-05, + "loss": 0.1729, + "step": 38320 + }, + { + "epoch": 3.1586665202108963, + "grad_norm": 0.20174540579319, + "learning_rate": 1.8028522553725863e-05, + "loss": 0.1549, + "step": 38340 + }, + { + "epoch": 3.1603141476274166, + "grad_norm": 0.5369256734848022, + "learning_rate": 1.8000909845237638e-05, + "loss": 0.1321, + "step": 38360 + }, + { + "epoch": 3.161961775043937, + "grad_norm": 2.0894880294799805, + "learning_rate": 1.7973306398263297e-05, + "loss": 0.1592, + "step": 38380 + }, + { + "epoch": 3.163609402460457, + "grad_norm": 0.35338452458381653, + "learning_rate": 1.7945712249328982e-05, + "loss": 0.1482, + "step": 38400 + }, + { + "epoch": 3.1652570298769773, + "grad_norm": 0.660490095615387, + "learning_rate": 1.7918127434948533e-05, + "loss": 0.147, + "step": 38420 + }, + { + "epoch": 3.1669046572934976, + "grad_norm": 0.15226125717163086, + "learning_rate": 1.789055199162342e-05, + "loss": 0.1602, + "step": 38440 + }, + { + "epoch": 3.1685522847100174, + "grad_norm": 1.184643268585205, + "learning_rate": 1.7862985955842736e-05, + "loss": 0.1352, + "step": 38460 + }, + { + "epoch": 3.1701999121265376, + "grad_norm": 6.8397536277771, + "learning_rate": 1.783542936408309e-05, + "loss": 0.149, + "step": 38480 + }, + { + "epoch": 3.171847539543058, + "grad_norm": 0.5208004713058472, + "learning_rate": 1.780788225280862e-05, + "loss": 0.138, + "step": 38500 + }, + { + "epoch": 3.173495166959578, + "grad_norm": 5.069784164428711, + "learning_rate": 1.7780344658470928e-05, + "loss": 0.144, + "step": 38520 + }, + { + "epoch": 3.1751427943760984, + "grad_norm": 0.22637605667114258, + "learning_rate": 1.7752816617509e-05, + "loss": 0.1561, + "step": 38540 + }, + { + "epoch": 3.1767904217926186, + "grad_norm": 1.149791955947876, + "learning_rate": 1.7725298166349193e-05, + "loss": 0.1294, + "step": 38560 + }, + { + "epoch": 3.178438049209139, + "grad_norm": 0.4961896240711212, + "learning_rate": 1.7697789341405163e-05, + "loss": 0.1557, + "step": 38580 + }, + { + "epoch": 3.180085676625659, + "grad_norm": 1.0814521312713623, + "learning_rate": 1.7670290179077844e-05, + "loss": 0.1595, + "step": 38600 + }, + { + "epoch": 3.1817333040421794, + "grad_norm": 4.159520626068115, + "learning_rate": 1.7642800715755376e-05, + "loss": 0.1617, + "step": 38620 + }, + { + "epoch": 3.1833809314586996, + "grad_norm": 0.23039385676383972, + "learning_rate": 1.761532098781306e-05, + "loss": 0.1539, + "step": 38640 + }, + { + "epoch": 3.18502855887522, + "grad_norm": 0.8867893218994141, + "learning_rate": 1.7587851031613337e-05, + "loss": 0.1391, + "step": 38660 + }, + { + "epoch": 3.18667618629174, + "grad_norm": 0.7063521146774292, + "learning_rate": 1.7560390883505672e-05, + "loss": 0.1482, + "step": 38680 + }, + { + "epoch": 3.18832381370826, + "grad_norm": 0.4100753366947174, + "learning_rate": 1.7532940579826598e-05, + "loss": 0.1629, + "step": 38700 + }, + { + "epoch": 3.18997144112478, + "grad_norm": 1.997273325920105, + "learning_rate": 1.75055001568996e-05, + "loss": 0.1573, + "step": 38720 + }, + { + "epoch": 3.1916190685413004, + "grad_norm": 0.39883023500442505, + "learning_rate": 1.74780696510351e-05, + "loss": 0.153, + "step": 38740 + }, + { + "epoch": 3.1932666959578206, + "grad_norm": 1.3483376502990723, + "learning_rate": 1.7450649098530368e-05, + "loss": 0.139, + "step": 38760 + }, + { + "epoch": 3.194914323374341, + "grad_norm": 3.0959737300872803, + "learning_rate": 1.7423238535669538e-05, + "loss": 0.1582, + "step": 38780 + }, + { + "epoch": 3.196561950790861, + "grad_norm": 0.8072210550308228, + "learning_rate": 1.7395837998723512e-05, + "loss": 0.1654, + "step": 38800 + }, + { + "epoch": 3.1982095782073814, + "grad_norm": 1.0747969150543213, + "learning_rate": 1.7368447523949922e-05, + "loss": 0.1615, + "step": 38820 + }, + { + "epoch": 3.1998572056239016, + "grad_norm": 1.8250178098678589, + "learning_rate": 1.734106714759309e-05, + "loss": 0.1572, + "step": 38840 + }, + { + "epoch": 3.201504833040422, + "grad_norm": 0.7575979232788086, + "learning_rate": 1.731369690588397e-05, + "loss": 0.1329, + "step": 38860 + }, + { + "epoch": 3.203152460456942, + "grad_norm": 1.538978934288025, + "learning_rate": 1.72877045964653e-05, + "loss": 0.1588, + "step": 38880 + }, + { + "epoch": 3.2048000878734624, + "grad_norm": 0.6645594239234924, + "learning_rate": 1.7260354221477675e-05, + "loss": 0.1628, + "step": 38900 + }, + { + "epoch": 3.2064477152899826, + "grad_norm": 1.7454744577407837, + "learning_rate": 1.723301408794078e-05, + "loss": 0.1528, + "step": 38920 + }, + { + "epoch": 3.2080953427065024, + "grad_norm": 0.19629539549350739, + "learning_rate": 1.7205684232032316e-05, + "loss": 0.1772, + "step": 38940 + }, + { + "epoch": 3.2097429701230227, + "grad_norm": 13.83620548248291, + "learning_rate": 1.717836468991638e-05, + "loss": 0.146, + "step": 38960 + }, + { + "epoch": 3.211390597539543, + "grad_norm": 1.3969836235046387, + "learning_rate": 1.7151055497743437e-05, + "loss": 0.1644, + "step": 38980 + }, + { + "epoch": 3.213038224956063, + "grad_norm": 0.7652454972267151, + "learning_rate": 1.712375669165025e-05, + "loss": 0.1585, + "step": 39000 + }, + { + "epoch": 3.2146858523725834, + "grad_norm": 1.5630168914794922, + "learning_rate": 1.7096468307759844e-05, + "loss": 0.1556, + "step": 39020 + }, + { + "epoch": 3.2163334797891037, + "grad_norm": 0.46949490904808044, + "learning_rate": 1.706919038218145e-05, + "loss": 0.1495, + "step": 39040 + }, + { + "epoch": 3.217981107205624, + "grad_norm": 4.1307692527771, + "learning_rate": 1.7041922951010436e-05, + "loss": 0.1408, + "step": 39060 + }, + { + "epoch": 3.219628734622144, + "grad_norm": 0.4786194860935211, + "learning_rate": 1.7014666050328325e-05, + "loss": 0.1751, + "step": 39080 + }, + { + "epoch": 3.2212763620386644, + "grad_norm": 0.9991576075553894, + "learning_rate": 1.6987419716202672e-05, + "loss": 0.1508, + "step": 39100 + }, + { + "epoch": 3.2229239894551847, + "grad_norm": 0.6878542304039001, + "learning_rate": 1.696018398468708e-05, + "loss": 0.1464, + "step": 39120 + }, + { + "epoch": 3.224571616871705, + "grad_norm": 0.37046971917152405, + "learning_rate": 1.6932958891821106e-05, + "loss": 0.1502, + "step": 39140 + }, + { + "epoch": 3.226219244288225, + "grad_norm": 1.1257407665252686, + "learning_rate": 1.690574447363021e-05, + "loss": 0.1369, + "step": 39160 + }, + { + "epoch": 3.2278668717047454, + "grad_norm": 11.745176315307617, + "learning_rate": 1.687854076612575e-05, + "loss": 0.172, + "step": 39180 + }, + { + "epoch": 3.229514499121265, + "grad_norm": 0.8210144639015198, + "learning_rate": 1.6851347805304917e-05, + "loss": 0.1706, + "step": 39200 + }, + { + "epoch": 3.2311621265377855, + "grad_norm": 1.5889972448349, + "learning_rate": 1.682416562715067e-05, + "loss": 0.1449, + "step": 39220 + }, + { + "epoch": 3.2328097539543057, + "grad_norm": 0.25009989738464355, + "learning_rate": 1.6796994267631704e-05, + "loss": 0.167, + "step": 39240 + }, + { + "epoch": 3.234457381370826, + "grad_norm": 0.7307600975036621, + "learning_rate": 1.676983376270238e-05, + "loss": 0.1413, + "step": 39260 + }, + { + "epoch": 3.236105008787346, + "grad_norm": 1.3502663373947144, + "learning_rate": 1.6742684148302727e-05, + "loss": 0.1693, + "step": 39280 + }, + { + "epoch": 3.2377526362038664, + "grad_norm": 0.6214455366134644, + "learning_rate": 1.671554546035834e-05, + "loss": 0.1497, + "step": 39300 + }, + { + "epoch": 3.2394002636203867, + "grad_norm": 0.5916234850883484, + "learning_rate": 1.6688417734780364e-05, + "loss": 0.1701, + "step": 39320 + }, + { + "epoch": 3.241047891036907, + "grad_norm": 0.2480258345603943, + "learning_rate": 1.6661301007465452e-05, + "loss": 0.1605, + "step": 39340 + }, + { + "epoch": 3.242695518453427, + "grad_norm": 1.1683536767959595, + "learning_rate": 1.6634195314295665e-05, + "loss": 0.1475, + "step": 39360 + }, + { + "epoch": 3.2443431458699474, + "grad_norm": 6.31130313873291, + "learning_rate": 1.6607100691138494e-05, + "loss": 0.1799, + "step": 39380 + }, + { + "epoch": 3.2459907732864677, + "grad_norm": 0.7771806120872498, + "learning_rate": 1.6580017173846784e-05, + "loss": 0.1494, + "step": 39400 + }, + { + "epoch": 3.2476384007029875, + "grad_norm": 1.8961833715438843, + "learning_rate": 1.6552944798258665e-05, + "loss": 0.153, + "step": 39420 + }, + { + "epoch": 3.2492860281195077, + "grad_norm": 2.086655616760254, + "learning_rate": 1.652588360019755e-05, + "loss": 0.1708, + "step": 39440 + }, + { + "epoch": 3.250933655536028, + "grad_norm": 1.0704363584518433, + "learning_rate": 1.649883361547202e-05, + "loss": 0.1441, + "step": 39460 + }, + { + "epoch": 3.2525812829525482, + "grad_norm": 0.9307258129119873, + "learning_rate": 1.647179487987586e-05, + "loss": 0.1579, + "step": 39480 + }, + { + "epoch": 3.2542289103690685, + "grad_norm": 1.0694314241409302, + "learning_rate": 1.6444767429187947e-05, + "loss": 0.1651, + "step": 39500 + }, + { + "epoch": 3.2558765377855887, + "grad_norm": 1.8976612091064453, + "learning_rate": 1.6417751299172233e-05, + "loss": 0.1552, + "step": 39520 + }, + { + "epoch": 3.257524165202109, + "grad_norm": 0.33035072684288025, + "learning_rate": 1.6390746525577687e-05, + "loss": 0.1633, + "step": 39540 + }, + { + "epoch": 3.259171792618629, + "grad_norm": 0.8933309316635132, + "learning_rate": 1.6363753144138254e-05, + "loss": 0.14, + "step": 39560 + }, + { + "epoch": 3.2608194200351495, + "grad_norm": 0.6716617345809937, + "learning_rate": 1.6336771190572797e-05, + "loss": 0.1644, + "step": 39580 + }, + { + "epoch": 3.2624670474516697, + "grad_norm": 5.440345764160156, + "learning_rate": 1.6309800700585065e-05, + "loss": 0.1657, + "step": 39600 + }, + { + "epoch": 3.26411467486819, + "grad_norm": 3.83003830909729, + "learning_rate": 1.6282841709863644e-05, + "loss": 0.1784, + "step": 39620 + }, + { + "epoch": 3.26576230228471, + "grad_norm": 0.3602890968322754, + "learning_rate": 1.6255894254081876e-05, + "loss": 0.1852, + "step": 39640 + }, + { + "epoch": 3.2674099297012305, + "grad_norm": 1.2716591358184814, + "learning_rate": 1.6228958368897866e-05, + "loss": 0.1426, + "step": 39660 + }, + { + "epoch": 3.2690575571177503, + "grad_norm": 0.9488288760185242, + "learning_rate": 1.6202034089954406e-05, + "loss": 0.1551, + "step": 39680 + }, + { + "epoch": 3.2707051845342705, + "grad_norm": 0.6714811325073242, + "learning_rate": 1.617512145287892e-05, + "loss": 0.1541, + "step": 39700 + }, + { + "epoch": 3.2723528119507908, + "grad_norm": 0.9304949641227722, + "learning_rate": 1.614822049328344e-05, + "loss": 0.1701, + "step": 39720 + }, + { + "epoch": 3.274000439367311, + "grad_norm": 0.3308255076408386, + "learning_rate": 1.6121331246764522e-05, + "loss": 0.1621, + "step": 39740 + }, + { + "epoch": 3.2756480667838312, + "grad_norm": 1.1160569190979004, + "learning_rate": 1.6094453748903243e-05, + "loss": 0.1395, + "step": 39760 + }, + { + "epoch": 3.2772956942003515, + "grad_norm": 0.5838717222213745, + "learning_rate": 1.6067588035265128e-05, + "loss": 0.1556, + "step": 39780 + }, + { + "epoch": 3.2789433216168717, + "grad_norm": 1.3350988626480103, + "learning_rate": 1.6040734141400116e-05, + "loss": 0.1496, + "step": 39800 + }, + { + "epoch": 3.280590949033392, + "grad_norm": 0.5808613300323486, + "learning_rate": 1.60138921028425e-05, + "loss": 0.1516, + "step": 39820 + }, + { + "epoch": 3.2822385764499122, + "grad_norm": 0.6595000624656677, + "learning_rate": 1.5987061955110866e-05, + "loss": 0.1521, + "step": 39840 + }, + { + "epoch": 3.2838862038664325, + "grad_norm": 0.8309194445610046, + "learning_rate": 1.5960243733708102e-05, + "loss": 0.1376, + "step": 39860 + }, + { + "epoch": 3.2855338312829527, + "grad_norm": 1.0434929132461548, + "learning_rate": 1.59334374741213e-05, + "loss": 0.1601, + "step": 39880 + }, + { + "epoch": 3.2871814586994725, + "grad_norm": 2.53883957862854, + "learning_rate": 1.5907982639453698e-05, + "loss": 0.1461, + "step": 39900 + }, + { + "epoch": 3.288829086115993, + "grad_norm": 6.17413330078125, + "learning_rate": 1.5881199807417752e-05, + "loss": 0.147, + "step": 39920 + }, + { + "epoch": 3.290476713532513, + "grad_norm": 0.4980841279029846, + "learning_rate": 1.5854429041792265e-05, + "loss": 0.1493, + "step": 39940 + }, + { + "epoch": 3.2921243409490333, + "grad_norm": 0.8692514896392822, + "learning_rate": 1.5827670378001537e-05, + "loss": 0.1343, + "step": 39960 + }, + { + "epoch": 3.2937719683655535, + "grad_norm": 1.5790104866027832, + "learning_rate": 1.5800923851453842e-05, + "loss": 0.1609, + "step": 39980 + }, + { + "epoch": 3.2954195957820738, + "grad_norm": 2.524847984313965, + "learning_rate": 1.5774189497541406e-05, + "loss": 0.1503, + "step": 40000 + }, + { + "epoch": 3.2954195957820738, + "eval_loss": 0.5555781722068787, + "eval_runtime": 807.4041, + "eval_samples_per_second": 26.986, + "eval_steps_per_second": 6.748, + "eval_wer": 0.22916497537952482, + "step": 40000 + }, + { + "epoch": 3.297067223198594, + "grad_norm": 1.8827567100524902, + "learning_rate": 1.5747467351640345e-05, + "loss": 0.1538, + "step": 40020 + }, + { + "epoch": 3.2987148506151143, + "grad_norm": 0.4699152410030365, + "learning_rate": 1.5720757449110597e-05, + "loss": 0.1671, + "step": 40040 + }, + { + "epoch": 3.3003624780316345, + "grad_norm": 0.644223690032959, + "learning_rate": 1.5694059825295928e-05, + "loss": 0.135, + "step": 40060 + }, + { + "epoch": 3.3020101054481548, + "grad_norm": 0.40487343072891235, + "learning_rate": 1.5667374515523843e-05, + "loss": 0.1394, + "step": 40080 + }, + { + "epoch": 3.303657732864675, + "grad_norm": 0.7618485689163208, + "learning_rate": 1.564070155510556e-05, + "loss": 0.1663, + "step": 40100 + }, + { + "epoch": 3.3053053602811953, + "grad_norm": 0.6829048991203308, + "learning_rate": 1.561404097933596e-05, + "loss": 0.146, + "step": 40120 + }, + { + "epoch": 3.3069529876977155, + "grad_norm": 1.3032153844833374, + "learning_rate": 1.5587392823493506e-05, + "loss": 0.1531, + "step": 40140 + }, + { + "epoch": 3.3086006151142353, + "grad_norm": 1.1882566213607788, + "learning_rate": 1.556075712284026e-05, + "loss": 0.143, + "step": 40160 + }, + { + "epoch": 3.3102482425307556, + "grad_norm": 9.589151382446289, + "learning_rate": 1.5534133912621786e-05, + "loss": 0.1644, + "step": 40180 + }, + { + "epoch": 3.311895869947276, + "grad_norm": 1.2138134241104126, + "learning_rate": 1.550752322806713e-05, + "loss": 0.1495, + "step": 40200 + }, + { + "epoch": 3.313543497363796, + "grad_norm": 0.6604083776473999, + "learning_rate": 1.5480925104388762e-05, + "loss": 0.1519, + "step": 40220 + }, + { + "epoch": 3.3151911247803163, + "grad_norm": 0.5243780612945557, + "learning_rate": 1.5454339576782517e-05, + "loss": 0.1624, + "step": 40240 + }, + { + "epoch": 3.3168387521968365, + "grad_norm": 0.8697063326835632, + "learning_rate": 1.542776668042758e-05, + "loss": 0.1324, + "step": 40260 + }, + { + "epoch": 3.318486379613357, + "grad_norm": 2.33467960357666, + "learning_rate": 1.5401206450486406e-05, + "loss": 0.1718, + "step": 40280 + }, + { + "epoch": 3.320134007029877, + "grad_norm": 2.60256028175354, + "learning_rate": 1.5374658922104704e-05, + "loss": 0.1613, + "step": 40300 + }, + { + "epoch": 3.3217816344463973, + "grad_norm": 1.0596078634262085, + "learning_rate": 1.5348124130411363e-05, + "loss": 0.1687, + "step": 40320 + }, + { + "epoch": 3.3234292618629175, + "grad_norm": 0.6154633164405823, + "learning_rate": 1.532160211051843e-05, + "loss": 0.1694, + "step": 40340 + }, + { + "epoch": 3.325076889279438, + "grad_norm": 0.6055323481559753, + "learning_rate": 1.529509289752104e-05, + "loss": 0.1398, + "step": 40360 + }, + { + "epoch": 3.3267245166959576, + "grad_norm": 0.6304402947425842, + "learning_rate": 1.526859652649739e-05, + "loss": 0.1613, + "step": 40380 + }, + { + "epoch": 3.3283721441124783, + "grad_norm": 0.7361418604850769, + "learning_rate": 1.5242113032508697e-05, + "loss": 0.1659, + "step": 40400 + }, + { + "epoch": 3.330019771528998, + "grad_norm": 0.3940108120441437, + "learning_rate": 1.5215642450599099e-05, + "loss": 0.1483, + "step": 40420 + }, + { + "epoch": 3.3316673989455183, + "grad_norm": 0.16881057620048523, + "learning_rate": 1.5189184815795676e-05, + "loss": 0.1516, + "step": 40440 + }, + { + "epoch": 3.3333150263620386, + "grad_norm": 0.9273294806480408, + "learning_rate": 1.516274016310838e-05, + "loss": 0.1542, + "step": 40460 + }, + { + "epoch": 3.334962653778559, + "grad_norm": 0.9351533651351929, + "learning_rate": 1.5136308527529974e-05, + "loss": 0.1544, + "step": 40480 + }, + { + "epoch": 3.336610281195079, + "grad_norm": 0.6827693581581116, + "learning_rate": 1.5109889944036005e-05, + "loss": 0.1545, + "step": 40500 + }, + { + "epoch": 3.3382579086115993, + "grad_norm": 0.5103474259376526, + "learning_rate": 1.5083484447584722e-05, + "loss": 0.1535, + "step": 40520 + }, + { + "epoch": 3.3399055360281196, + "grad_norm": 0.4932575821876526, + "learning_rate": 1.5057092073117097e-05, + "loss": 0.163, + "step": 40540 + }, + { + "epoch": 3.34155316344464, + "grad_norm": 0.8227238655090332, + "learning_rate": 1.5030712855556705e-05, + "loss": 0.1447, + "step": 40560 + }, + { + "epoch": 3.34320079086116, + "grad_norm": 1.6172962188720703, + "learning_rate": 1.5004346829809729e-05, + "loss": 0.1559, + "step": 40580 + }, + { + "epoch": 3.3448484182776803, + "grad_norm": 2.6368677616119385, + "learning_rate": 1.49779940307649e-05, + "loss": 0.1573, + "step": 40600 + }, + { + "epoch": 3.3464960456942006, + "grad_norm": 0.5566956996917725, + "learning_rate": 1.4951654493293427e-05, + "loss": 0.1645, + "step": 40620 + }, + { + "epoch": 3.3481436731107204, + "grad_norm": 0.9300214052200317, + "learning_rate": 1.4925328252248982e-05, + "loss": 0.1735, + "step": 40640 + }, + { + "epoch": 3.3497913005272406, + "grad_norm": 2.0181519985198975, + "learning_rate": 1.4899015342467649e-05, + "loss": 0.1469, + "step": 40660 + }, + { + "epoch": 3.351438927943761, + "grad_norm": 1.1045427322387695, + "learning_rate": 1.4872715798767861e-05, + "loss": 0.1666, + "step": 40680 + }, + { + "epoch": 3.353086555360281, + "grad_norm": 1.1991610527038574, + "learning_rate": 1.4846429655950383e-05, + "loss": 0.1648, + "step": 40700 + }, + { + "epoch": 3.3547341827768014, + "grad_norm": 0.6012402772903442, + "learning_rate": 1.4820156948798205e-05, + "loss": 0.1479, + "step": 40720 + }, + { + "epoch": 3.3563818101933216, + "grad_norm": 0.15366359055042267, + "learning_rate": 1.4793897712076581e-05, + "loss": 0.1608, + "step": 40740 + }, + { + "epoch": 3.358029437609842, + "grad_norm": 1.1802315711975098, + "learning_rate": 1.4767651980532926e-05, + "loss": 0.1384, + "step": 40760 + }, + { + "epoch": 3.359677065026362, + "grad_norm": 0.8219059109687805, + "learning_rate": 1.4741419788896784e-05, + "loss": 0.1594, + "step": 40780 + }, + { + "epoch": 3.3613246924428823, + "grad_norm": 1.291498064994812, + "learning_rate": 1.4715201171879789e-05, + "loss": 0.1646, + "step": 40800 + }, + { + "epoch": 3.3629723198594026, + "grad_norm": 1.1582722663879395, + "learning_rate": 1.4688996164175589e-05, + "loss": 0.1539, + "step": 40820 + }, + { + "epoch": 3.364619947275923, + "grad_norm": 0.22327923774719238, + "learning_rate": 1.4662804800459846e-05, + "loss": 0.173, + "step": 40840 + }, + { + "epoch": 3.3662675746924426, + "grad_norm": 0.767255425453186, + "learning_rate": 1.4636627115390167e-05, + "loss": 0.1339, + "step": 40860 + }, + { + "epoch": 3.3679152021089633, + "grad_norm": 0.8810155987739563, + "learning_rate": 1.4610463143606046e-05, + "loss": 0.1579, + "step": 40880 + }, + { + "epoch": 3.369562829525483, + "grad_norm": 0.5456433296203613, + "learning_rate": 1.4584312919728853e-05, + "loss": 0.15, + "step": 40900 + }, + { + "epoch": 3.3712104569420034, + "grad_norm": 1.5852445363998413, + "learning_rate": 1.4558176478361735e-05, + "loss": 0.1619, + "step": 40920 + }, + { + "epoch": 3.3728580843585236, + "grad_norm": 0.3770715296268463, + "learning_rate": 1.4532053854089625e-05, + "loss": 0.1533, + "step": 40940 + }, + { + "epoch": 3.374505711775044, + "grad_norm": 0.9395153522491455, + "learning_rate": 1.4505945081479161e-05, + "loss": 0.1415, + "step": 40960 + }, + { + "epoch": 3.376153339191564, + "grad_norm": 0.5108890533447266, + "learning_rate": 1.4479850195078664e-05, + "loss": 0.1675, + "step": 40980 + }, + { + "epoch": 3.3778009666080844, + "grad_norm": 0.7402895092964172, + "learning_rate": 1.445376922941808e-05, + "loss": 0.1633, + "step": 41000 + }, + { + "epoch": 3.3794485940246046, + "grad_norm": 2.410801887512207, + "learning_rate": 1.4427702219008901e-05, + "loss": 0.1661, + "step": 41020 + }, + { + "epoch": 3.381096221441125, + "grad_norm": 0.3239726722240448, + "learning_rate": 1.4401649198344196e-05, + "loss": 0.1576, + "step": 41040 + }, + { + "epoch": 3.382743848857645, + "grad_norm": 0.6572871208190918, + "learning_rate": 1.4375610201898505e-05, + "loss": 0.1386, + "step": 41060 + }, + { + "epoch": 3.3843914762741654, + "grad_norm": 0.9796928763389587, + "learning_rate": 1.4350886176591104e-05, + "loss": 0.1531, + "step": 41080 + }, + { + "epoch": 3.3860391036906856, + "grad_norm": 1.2608447074890137, + "learning_rate": 1.4324874626459562e-05, + "loss": 0.1589, + "step": 41100 + }, + { + "epoch": 3.3876867311072054, + "grad_norm": 1.1734158992767334, + "learning_rate": 1.4298877202138631e-05, + "loss": 0.155, + "step": 41120 + }, + { + "epoch": 3.3893343585237257, + "grad_norm": 0.28572332859039307, + "learning_rate": 1.4272893938029281e-05, + "loss": 0.1594, + "step": 41140 + }, + { + "epoch": 3.390981985940246, + "grad_norm": 22.722564697265625, + "learning_rate": 1.4246924868513762e-05, + "loss": 0.1429, + "step": 41160 + }, + { + "epoch": 3.392629613356766, + "grad_norm": 0.7266652584075928, + "learning_rate": 1.4220970027955499e-05, + "loss": 0.1656, + "step": 41180 + }, + { + "epoch": 3.3942772407732864, + "grad_norm": 1.7261922359466553, + "learning_rate": 1.4195029450699135e-05, + "loss": 0.1553, + "step": 41200 + }, + { + "epoch": 3.3959248681898067, + "grad_norm": 1.3773916959762573, + "learning_rate": 1.4169103171070414e-05, + "loss": 0.1568, + "step": 41220 + }, + { + "epoch": 3.397572495606327, + "grad_norm": 0.42017877101898193, + "learning_rate": 1.4143191223376162e-05, + "loss": 0.1491, + "step": 41240 + }, + { + "epoch": 3.399220123022847, + "grad_norm": 1.750963807106018, + "learning_rate": 1.4117293641904256e-05, + "loss": 0.1386, + "step": 41260 + }, + { + "epoch": 3.4008677504393674, + "grad_norm": 1.1979619264602661, + "learning_rate": 1.4091410460923554e-05, + "loss": 0.1728, + "step": 41280 + }, + { + "epoch": 3.4025153778558876, + "grad_norm": 0.625075101852417, + "learning_rate": 1.4065541714683836e-05, + "loss": 0.1458, + "step": 41300 + }, + { + "epoch": 3.404163005272408, + "grad_norm": 1.3867195844650269, + "learning_rate": 1.4039687437415813e-05, + "loss": 0.1607, + "step": 41320 + }, + { + "epoch": 3.4058106326889277, + "grad_norm": 0.3580436110496521, + "learning_rate": 1.4013847663331042e-05, + "loss": 0.176, + "step": 41340 + }, + { + "epoch": 3.4074582601054484, + "grad_norm": 1.7142635583877563, + "learning_rate": 1.398802242662188e-05, + "loss": 0.1395, + "step": 41360 + }, + { + "epoch": 3.409105887521968, + "grad_norm": 0.6943619847297668, + "learning_rate": 1.3962211761461466e-05, + "loss": 0.168, + "step": 41380 + }, + { + "epoch": 3.4107535149384884, + "grad_norm": 1.8670823574066162, + "learning_rate": 1.3936415702003624e-05, + "loss": 0.1596, + "step": 41400 + }, + { + "epoch": 3.4124011423550087, + "grad_norm": 2.691347360610962, + "learning_rate": 1.3910634282382878e-05, + "loss": 0.1528, + "step": 41420 + }, + { + "epoch": 3.414048769771529, + "grad_norm": 0.25475576519966125, + "learning_rate": 1.388486753671438e-05, + "loss": 0.1516, + "step": 41440 + }, + { + "epoch": 3.415696397188049, + "grad_norm": 1.1902556419372559, + "learning_rate": 1.3859115499093849e-05, + "loss": 0.1311, + "step": 41460 + }, + { + "epoch": 3.4173440246045694, + "grad_norm": 0.6261383891105652, + "learning_rate": 1.383337820359757e-05, + "loss": 0.1641, + "step": 41480 + }, + { + "epoch": 3.4189916520210897, + "grad_norm": 0.8379786610603333, + "learning_rate": 1.3807655684282272e-05, + "loss": 0.1674, + "step": 41500 + }, + { + "epoch": 3.42063927943761, + "grad_norm": 1.3956279754638672, + "learning_rate": 1.3781947975185178e-05, + "loss": 0.1628, + "step": 41520 + }, + { + "epoch": 3.42228690685413, + "grad_norm": 0.22908486425876617, + "learning_rate": 1.3756255110323891e-05, + "loss": 0.159, + "step": 41540 + }, + { + "epoch": 3.4239345342706504, + "grad_norm": 1.428760290145874, + "learning_rate": 1.3730577123696376e-05, + "loss": 0.1397, + "step": 41560 + }, + { + "epoch": 3.4255821616871707, + "grad_norm": 3.128469467163086, + "learning_rate": 1.3704914049280926e-05, + "loss": 0.1554, + "step": 41580 + }, + { + "epoch": 3.4272297891036905, + "grad_norm": 0.8418864011764526, + "learning_rate": 1.3679265921036062e-05, + "loss": 0.1482, + "step": 41600 + }, + { + "epoch": 3.4288774165202107, + "grad_norm": 1.0914111137390137, + "learning_rate": 1.365363277290056e-05, + "loss": 0.148, + "step": 41620 + }, + { + "epoch": 3.430525043936731, + "grad_norm": 0.5711450576782227, + "learning_rate": 1.3628014638793369e-05, + "loss": 0.158, + "step": 41640 + }, + { + "epoch": 3.432172671353251, + "grad_norm": 1.1503074169158936, + "learning_rate": 1.360241155261357e-05, + "loss": 0.1343, + "step": 41660 + }, + { + "epoch": 3.4338202987697715, + "grad_norm": 1.1104477643966675, + "learning_rate": 1.3576823548240333e-05, + "loss": 0.1539, + "step": 41680 + }, + { + "epoch": 3.4354679261862917, + "grad_norm": 3.4182581901550293, + "learning_rate": 1.3551250659532852e-05, + "loss": 0.1602, + "step": 41700 + }, + { + "epoch": 3.437115553602812, + "grad_norm": 0.9579629302024841, + "learning_rate": 1.352569292033034e-05, + "loss": 0.1449, + "step": 41720 + }, + { + "epoch": 3.438763181019332, + "grad_norm": 0.18893243372440338, + "learning_rate": 1.3500150364451963e-05, + "loss": 0.1503, + "step": 41740 + }, + { + "epoch": 3.4404108084358525, + "grad_norm": 1.7320404052734375, + "learning_rate": 1.3474623025696792e-05, + "loss": 0.1315, + "step": 41760 + }, + { + "epoch": 3.4420584358523727, + "grad_norm": 0.9133217930793762, + "learning_rate": 1.3449110937843751e-05, + "loss": 0.1573, + "step": 41780 + }, + { + "epoch": 3.443706063268893, + "grad_norm": 0.9956775903701782, + "learning_rate": 1.3423614134651602e-05, + "loss": 0.1461, + "step": 41800 + }, + { + "epoch": 3.445353690685413, + "grad_norm": 3.1847634315490723, + "learning_rate": 1.3398132649858868e-05, + "loss": 0.1528, + "step": 41820 + }, + { + "epoch": 3.4470013181019334, + "grad_norm": 0.3741768002510071, + "learning_rate": 1.3372666517183813e-05, + "loss": 0.1561, + "step": 41840 + }, + { + "epoch": 3.4486489455184532, + "grad_norm": 0.5312712788581848, + "learning_rate": 1.3347215770324372e-05, + "loss": 0.137, + "step": 41860 + }, + { + "epoch": 3.4502965729349735, + "grad_norm": 0.594749927520752, + "learning_rate": 1.3321780442958143e-05, + "loss": 0.1611, + "step": 41880 + }, + { + "epoch": 3.4519442003514937, + "grad_norm": 0.9457196593284607, + "learning_rate": 1.3296360568742282e-05, + "loss": 0.1535, + "step": 41900 + }, + { + "epoch": 3.453591827768014, + "grad_norm": 1.3242007493972778, + "learning_rate": 1.327095618131354e-05, + "loss": 0.1513, + "step": 41920 + }, + { + "epoch": 3.4552394551845342, + "grad_norm": 0.14979644119739532, + "learning_rate": 1.3245567314288148e-05, + "loss": 0.1538, + "step": 41940 + }, + { + "epoch": 3.4568870826010545, + "grad_norm": 3.511759042739868, + "learning_rate": 1.3220194001261813e-05, + "loss": 0.1442, + "step": 41960 + }, + { + "epoch": 3.4585347100175747, + "grad_norm": 2.007502794265747, + "learning_rate": 1.3194836275809658e-05, + "loss": 0.1631, + "step": 41980 + }, + { + "epoch": 3.460182337434095, + "grad_norm": 3.720379590988159, + "learning_rate": 1.3169494171486164e-05, + "loss": 0.1442, + "step": 42000 + }, + { + "epoch": 3.460182337434095, + "eval_loss": 0.5300702452659607, + "eval_runtime": 254.8955, + "eval_samples_per_second": 85.482, + "eval_steps_per_second": 21.373, + "eval_wer": 0.22274321642223485, + "step": 42000 + }, + { + "epoch": 3.4618299648506152, + "grad_norm": 1.0895417928695679, + "learning_rate": 1.3144167721825157e-05, + "loss": 0.1459, + "step": 42020 + }, + { + "epoch": 3.4634775922671355, + "grad_norm": 0.23793169856071472, + "learning_rate": 1.311885696033975e-05, + "loss": 0.1583, + "step": 42040 + }, + { + "epoch": 3.4651252196836557, + "grad_norm": 0.7614142298698425, + "learning_rate": 1.3093561920522293e-05, + "loss": 0.1396, + "step": 42060 + }, + { + "epoch": 3.4667728471001755, + "grad_norm": 0.9006688594818115, + "learning_rate": 1.3068282635844337e-05, + "loss": 0.1558, + "step": 42080 + }, + { + "epoch": 3.4684204745166958, + "grad_norm": 0.690870463848114, + "learning_rate": 1.3043019139756563e-05, + "loss": 0.1487, + "step": 42100 + }, + { + "epoch": 3.470068101933216, + "grad_norm": 0.4665576219558716, + "learning_rate": 1.3017771465688782e-05, + "loss": 0.1546, + "step": 42120 + }, + { + "epoch": 3.4717157293497363, + "grad_norm": 0.12971307337284088, + "learning_rate": 1.2992539647049867e-05, + "loss": 0.1659, + "step": 42140 + }, + { + "epoch": 3.4733633567662565, + "grad_norm": 1.8812086582183838, + "learning_rate": 1.2967323717227715e-05, + "loss": 0.141, + "step": 42160 + }, + { + "epoch": 3.4750109841827768, + "grad_norm": 0.7102630734443665, + "learning_rate": 1.2942123709589165e-05, + "loss": 0.1554, + "step": 42180 + }, + { + "epoch": 3.476658611599297, + "grad_norm": 0.522306501865387, + "learning_rate": 1.2916939657480028e-05, + "loss": 0.158, + "step": 42200 + }, + { + "epoch": 3.4783062390158173, + "grad_norm": 1.0656942129135132, + "learning_rate": 1.2891771594224983e-05, + "loss": 0.1546, + "step": 42220 + }, + { + "epoch": 3.4799538664323375, + "grad_norm": 0.21948251128196716, + "learning_rate": 1.2866619553127551e-05, + "loss": 0.1635, + "step": 42240 + }, + { + "epoch": 3.4816014938488578, + "grad_norm": 0.8842756748199463, + "learning_rate": 1.284148356747007e-05, + "loss": 0.1368, + "step": 42260 + }, + { + "epoch": 3.483249121265378, + "grad_norm": 0.8186292052268982, + "learning_rate": 1.2816363670513592e-05, + "loss": 0.1429, + "step": 42280 + }, + { + "epoch": 3.4848967486818982, + "grad_norm": 1.9236111640930176, + "learning_rate": 1.279125989549792e-05, + "loss": 0.1467, + "step": 42300 + }, + { + "epoch": 3.4865443760984185, + "grad_norm": 0.48028749227523804, + "learning_rate": 1.2766172275641507e-05, + "loss": 0.1589, + "step": 42320 + }, + { + "epoch": 3.4881920035149383, + "grad_norm": 0.4948166608810425, + "learning_rate": 1.274110084414143e-05, + "loss": 0.1437, + "step": 42340 + }, + { + "epoch": 3.4898396309314585, + "grad_norm": 1.7658034563064575, + "learning_rate": 1.2716045634173345e-05, + "loss": 0.1518, + "step": 42360 + }, + { + "epoch": 3.491487258347979, + "grad_norm": 0.6642500162124634, + "learning_rate": 1.2691006678891446e-05, + "loss": 0.1595, + "step": 42380 + }, + { + "epoch": 3.493134885764499, + "grad_norm": 0.4344884157180786, + "learning_rate": 1.2665984011428414e-05, + "loss": 0.1488, + "step": 42400 + }, + { + "epoch": 3.4947825131810193, + "grad_norm": 0.3623495399951935, + "learning_rate": 1.2640977664895387e-05, + "loss": 0.149, + "step": 42420 + }, + { + "epoch": 3.4964301405975395, + "grad_norm": 0.13554301857948303, + "learning_rate": 1.2615987672381884e-05, + "loss": 0.1578, + "step": 42440 + }, + { + "epoch": 3.49807776801406, + "grad_norm": 3.0575642585754395, + "learning_rate": 1.2591014066955803e-05, + "loss": 0.1246, + "step": 42460 + }, + { + "epoch": 3.49972539543058, + "grad_norm": 0.6213861107826233, + "learning_rate": 1.2566056881663352e-05, + "loss": 0.1578, + "step": 42480 + }, + { + "epoch": 3.5013730228471003, + "grad_norm": 0.3637482523918152, + "learning_rate": 1.2541116149529015e-05, + "loss": 0.1567, + "step": 42500 + }, + { + "epoch": 3.5030206502636205, + "grad_norm": 0.5865427851676941, + "learning_rate": 1.2516191903555496e-05, + "loss": 0.1512, + "step": 42520 + }, + { + "epoch": 3.5046682776801408, + "grad_norm": 0.20063956081867218, + "learning_rate": 1.2491284176723697e-05, + "loss": 0.1599, + "step": 42540 + }, + { + "epoch": 3.5063159050966606, + "grad_norm": 0.615571141242981, + "learning_rate": 1.2466393001992658e-05, + "loss": 0.1288, + "step": 42560 + }, + { + "epoch": 3.5079635325131813, + "grad_norm": 0.4955737590789795, + "learning_rate": 1.2441518412299491e-05, + "loss": 0.1531, + "step": 42580 + }, + { + "epoch": 3.509611159929701, + "grad_norm": 0.3354365825653076, + "learning_rate": 1.2416660440559396e-05, + "loss": 0.1545, + "step": 42600 + }, + { + "epoch": 3.5112587873462213, + "grad_norm": 0.47431784868240356, + "learning_rate": 1.2391819119665568e-05, + "loss": 0.1463, + "step": 42620 + }, + { + "epoch": 3.5129064147627416, + "grad_norm": 0.18025411665439606, + "learning_rate": 1.236699448248917e-05, + "loss": 0.1519, + "step": 42640 + }, + { + "epoch": 3.514554042179262, + "grad_norm": 0.9464624524116516, + "learning_rate": 1.2342186561879296e-05, + "loss": 0.1344, + "step": 42660 + }, + { + "epoch": 3.516201669595782, + "grad_norm": 0.6174651384353638, + "learning_rate": 1.23173953906629e-05, + "loss": 0.155, + "step": 42680 + }, + { + "epoch": 3.5178492970123023, + "grad_norm": 0.641559362411499, + "learning_rate": 1.2292621001644795e-05, + "loss": 0.148, + "step": 42700 + }, + { + "epoch": 3.5194969244288226, + "grad_norm": 1.0432337522506714, + "learning_rate": 1.226786342760757e-05, + "loss": 0.1529, + "step": 42720 + }, + { + "epoch": 3.521144551845343, + "grad_norm": 0.2608916461467743, + "learning_rate": 1.2243122701311577e-05, + "loss": 0.1533, + "step": 42740 + }, + { + "epoch": 3.522792179261863, + "grad_norm": 12.690485000610352, + "learning_rate": 1.2218398855494879e-05, + "loss": 0.1404, + "step": 42760 + }, + { + "epoch": 3.524439806678383, + "grad_norm": 0.7469035983085632, + "learning_rate": 1.2193691922873166e-05, + "loss": 0.156, + "step": 42780 + }, + { + "epoch": 3.5260874340949035, + "grad_norm": 0.37507572770118713, + "learning_rate": 1.2169001936139788e-05, + "loss": 0.1555, + "step": 42800 + }, + { + "epoch": 3.5277350615114234, + "grad_norm": 1.2525111436843872, + "learning_rate": 1.2144328927965656e-05, + "loss": 0.1596, + "step": 42820 + }, + { + "epoch": 3.5293826889279436, + "grad_norm": 0.3763274550437927, + "learning_rate": 1.2119672930999212e-05, + "loss": 0.1434, + "step": 42840 + }, + { + "epoch": 3.531030316344464, + "grad_norm": 0.7608032822608948, + "learning_rate": 1.2095033977866404e-05, + "loss": 0.1346, + "step": 42860 + }, + { + "epoch": 3.532677943760984, + "grad_norm": 5.237201690673828, + "learning_rate": 1.2070412101170594e-05, + "loss": 0.1401, + "step": 42880 + }, + { + "epoch": 3.5343255711775043, + "grad_norm": 0.3807220160961151, + "learning_rate": 1.2045807333492574e-05, + "loss": 0.1649, + "step": 42900 + }, + { + "epoch": 3.5359731985940246, + "grad_norm": 0.5126063227653503, + "learning_rate": 1.202121970739049e-05, + "loss": 0.1431, + "step": 42920 + }, + { + "epoch": 3.537620826010545, + "grad_norm": 0.4814915359020233, + "learning_rate": 1.1996649255399809e-05, + "loss": 0.1523, + "step": 42940 + }, + { + "epoch": 3.539268453427065, + "grad_norm": 0.8156435489654541, + "learning_rate": 1.197209601003327e-05, + "loss": 0.1371, + "step": 42960 + }, + { + "epoch": 3.5409160808435853, + "grad_norm": 0.6996150612831116, + "learning_rate": 1.1947560003780836e-05, + "loss": 0.1628, + "step": 42980 + }, + { + "epoch": 3.5425637082601056, + "grad_norm": 0.7120909690856934, + "learning_rate": 1.1923041269109663e-05, + "loss": 0.1583, + "step": 43000 + }, + { + "epoch": 3.544211335676626, + "grad_norm": 0.6141802668571472, + "learning_rate": 1.1898539838464057e-05, + "loss": 0.1408, + "step": 43020 + }, + { + "epoch": 3.5458589630931456, + "grad_norm": 0.7216193675994873, + "learning_rate": 1.1874055744265427e-05, + "loss": 0.1559, + "step": 43040 + }, + { + "epoch": 3.5475065905096663, + "grad_norm": 0.7931749820709229, + "learning_rate": 1.1849589018912236e-05, + "loss": 0.1387, + "step": 43060 + }, + { + "epoch": 3.549154217926186, + "grad_norm": 3.7226061820983887, + "learning_rate": 1.1825139694779964e-05, + "loss": 0.1621, + "step": 43080 + }, + { + "epoch": 3.5508018453427064, + "grad_norm": 1.089087963104248, + "learning_rate": 1.1800707804221067e-05, + "loss": 0.1552, + "step": 43100 + }, + { + "epoch": 3.5524494727592266, + "grad_norm": 0.803062379360199, + "learning_rate": 1.1776293379564934e-05, + "loss": 0.1549, + "step": 43120 + }, + { + "epoch": 3.554097100175747, + "grad_norm": 0.412349671125412, + "learning_rate": 1.175189645311785e-05, + "loss": 0.1546, + "step": 43140 + }, + { + "epoch": 3.555744727592267, + "grad_norm": 0.5987644195556641, + "learning_rate": 1.1727517057162919e-05, + "loss": 0.127, + "step": 43160 + }, + { + "epoch": 3.5573923550087874, + "grad_norm": 0.8935190439224243, + "learning_rate": 1.1703155223960067e-05, + "loss": 0.1631, + "step": 43180 + }, + { + "epoch": 3.5590399824253076, + "grad_norm": 0.720613420009613, + "learning_rate": 1.1678810985745983e-05, + "loss": 0.169, + "step": 43200 + }, + { + "epoch": 3.560687609841828, + "grad_norm": 1.8782857656478882, + "learning_rate": 1.1654484374734067e-05, + "loss": 0.1601, + "step": 43220 + }, + { + "epoch": 3.562335237258348, + "grad_norm": 0.4977351725101471, + "learning_rate": 1.1630175423114407e-05, + "loss": 0.1579, + "step": 43240 + }, + { + "epoch": 3.5639828646748684, + "grad_norm": 1.4265191555023193, + "learning_rate": 1.1605884163053688e-05, + "loss": 0.131, + "step": 43260 + }, + { + "epoch": 3.5656304920913886, + "grad_norm": 4.71371603012085, + "learning_rate": 1.1581610626695222e-05, + "loss": 0.1656, + "step": 43280 + }, + { + "epoch": 3.5672781195079084, + "grad_norm": 0.9297062158584595, + "learning_rate": 1.1557354846158855e-05, + "loss": 0.1639, + "step": 43300 + }, + { + "epoch": 3.568925746924429, + "grad_norm": 0.9973880648612976, + "learning_rate": 1.1533116853540932e-05, + "loss": 0.1631, + "step": 43320 + }, + { + "epoch": 3.570573374340949, + "grad_norm": 0.2250540405511856, + "learning_rate": 1.150889668091428e-05, + "loss": 0.1651, + "step": 43340 + }, + { + "epoch": 3.572221001757469, + "grad_norm": 3.2225499153137207, + "learning_rate": 1.1484694360328108e-05, + "loss": 0.1396, + "step": 43360 + }, + { + "epoch": 3.5738686291739894, + "grad_norm": 0.9798160195350647, + "learning_rate": 1.1460509923808038e-05, + "loss": 0.1687, + "step": 43380 + }, + { + "epoch": 3.5755162565905096, + "grad_norm": 0.8608407974243164, + "learning_rate": 1.1436343403356017e-05, + "loss": 0.1574, + "step": 43400 + }, + { + "epoch": 3.57716388400703, + "grad_norm": 0.802926242351532, + "learning_rate": 1.1412194830950276e-05, + "loss": 0.1612, + "step": 43420 + }, + { + "epoch": 3.57881151142355, + "grad_norm": 0.17060858011245728, + "learning_rate": 1.1389270340647518e-05, + "loss": 0.1472, + "step": 43440 + }, + { + "epoch": 3.5804591388400704, + "grad_norm": 1.1347203254699707, + "learning_rate": 1.1365156858819454e-05, + "loss": 0.133, + "step": 43460 + }, + { + "epoch": 3.5821067662565906, + "grad_norm": 1.255826711654663, + "learning_rate": 1.1341061419234941e-05, + "loss": 0.1598, + "step": 43480 + }, + { + "epoch": 3.583754393673111, + "grad_norm": 1.792175054550171, + "learning_rate": 1.1316984053778159e-05, + "loss": 0.162, + "step": 43500 + }, + { + "epoch": 3.5854020210896307, + "grad_norm": 2.900879383087158, + "learning_rate": 1.1292924794309374e-05, + "loss": 0.158, + "step": 43520 + }, + { + "epoch": 3.5870496485061514, + "grad_norm": 0.2784268856048584, + "learning_rate": 1.1268883672664893e-05, + "loss": 0.1599, + "step": 43540 + }, + { + "epoch": 3.588697275922671, + "grad_norm": 0.8298107981681824, + "learning_rate": 1.1244860720657006e-05, + "loss": 0.1342, + "step": 43560 + }, + { + "epoch": 3.5903449033391914, + "grad_norm": 0.5499345064163208, + "learning_rate": 1.1220855970073981e-05, + "loss": 0.16, + "step": 43580 + }, + { + "epoch": 3.5919925307557117, + "grad_norm": 1.1263962984085083, + "learning_rate": 1.1196869452679997e-05, + "loss": 0.1605, + "step": 43600 + }, + { + "epoch": 3.593640158172232, + "grad_norm": 1.825993299484253, + "learning_rate": 1.1172901200215103e-05, + "loss": 0.1553, + "step": 43620 + }, + { + "epoch": 3.595287785588752, + "grad_norm": 0.2598525285720825, + "learning_rate": 1.115014830715152e-05, + "loss": 0.1541, + "step": 43640 + }, + { + "epoch": 3.5969354130052724, + "grad_norm": 1.0398223400115967, + "learning_rate": 1.1126215762499096e-05, + "loss": 0.1403, + "step": 43660 + }, + { + "epoch": 3.5985830404217927, + "grad_norm": 1.0068581104278564, + "learning_rate": 1.1102301576267929e-05, + "loss": 0.1651, + "step": 43680 + }, + { + "epoch": 3.600230667838313, + "grad_norm": 1.069786548614502, + "learning_rate": 1.1078405780102358e-05, + "loss": 0.1624, + "step": 43700 + }, + { + "epoch": 3.601878295254833, + "grad_norm": 1.3623285293579102, + "learning_rate": 1.1054528405622395e-05, + "loss": 0.1509, + "step": 43720 + }, + { + "epoch": 3.6035259226713534, + "grad_norm": 0.17843444645404816, + "learning_rate": 1.1030669484423648e-05, + "loss": 0.1789, + "step": 43740 + }, + { + "epoch": 3.6051735500878737, + "grad_norm": 2.411144733428955, + "learning_rate": 1.1006829048077333e-05, + "loss": 0.1272, + "step": 43760 + }, + { + "epoch": 3.6068211775043935, + "grad_norm": 1.3706731796264648, + "learning_rate": 1.0983007128130198e-05, + "loss": 0.1472, + "step": 43780 + }, + { + "epoch": 3.608468804920914, + "grad_norm": 1.3497474193572998, + "learning_rate": 1.095920375610449e-05, + "loss": 0.1573, + "step": 43800 + }, + { + "epoch": 3.610116432337434, + "grad_norm": 0.8513794541358948, + "learning_rate": 1.0935418963497915e-05, + "loss": 0.1448, + "step": 43820 + }, + { + "epoch": 3.611764059753954, + "grad_norm": 0.19558091461658478, + "learning_rate": 1.0911652781783581e-05, + "loss": 0.1562, + "step": 43840 + }, + { + "epoch": 3.6134116871704745, + "grad_norm": 1.4188400506973267, + "learning_rate": 1.088790524240999e-05, + "loss": 0.1317, + "step": 43860 + }, + { + "epoch": 3.6150593145869947, + "grad_norm": 0.6879696249961853, + "learning_rate": 1.0864176376800956e-05, + "loss": 0.1655, + "step": 43880 + }, + { + "epoch": 3.616706942003515, + "grad_norm": 1.0637201070785522, + "learning_rate": 1.0840466216355594e-05, + "loss": 0.1627, + "step": 43900 + }, + { + "epoch": 3.618354569420035, + "grad_norm": 1.584519386291504, + "learning_rate": 1.0816774792448276e-05, + "loss": 0.1543, + "step": 43920 + }, + { + "epoch": 3.6200021968365554, + "grad_norm": 0.3126054108142853, + "learning_rate": 1.0793102136428548e-05, + "loss": 0.145, + "step": 43940 + }, + { + "epoch": 3.6216498242530757, + "grad_norm": 1.0722142457962036, + "learning_rate": 1.0769448279621158e-05, + "loss": 0.1298, + "step": 43960 + }, + { + "epoch": 3.623297451669596, + "grad_norm": 2.167898416519165, + "learning_rate": 1.0745813253325957e-05, + "loss": 0.1636, + "step": 43980 + }, + { + "epoch": 3.6249450790861157, + "grad_norm": 0.8137891292572021, + "learning_rate": 1.072219708881789e-05, + "loss": 0.1522, + "step": 44000 + }, + { + "epoch": 3.6249450790861157, + "eval_loss": 0.5213801860809326, + "eval_runtime": 254.0846, + "eval_samples_per_second": 85.755, + "eval_steps_per_second": 21.442, + "eval_wer": 0.2228915831287459, + "step": 44000 + } + ], + "logging_steps": 20, + "max_steps": 60690, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 2000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.3715074404589576e+21, + "train_batch_size": 24, + "trial_name": null, + "trial_params": null +}