| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 313, | |
| "global_step": 1251, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007993605115907274, | |
| "grad_norm": 0.9691409468650818, | |
| "learning_rate": 2e-05, | |
| "loss": 1.579, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0015987210231814548, | |
| "grad_norm": 1.4015984535217285, | |
| "learning_rate": 4e-05, | |
| "loss": 1.6108, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002398081534772182, | |
| "grad_norm": 1.4863674640655518, | |
| "learning_rate": 6e-05, | |
| "loss": 1.5698, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0031974420463629096, | |
| "grad_norm": 1.6171231269836426, | |
| "learning_rate": 8e-05, | |
| "loss": 1.7706, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.003996802557953637, | |
| "grad_norm": 3.1889543533325195, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7264, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004796163069544364, | |
| "grad_norm": 1.7145410776138306, | |
| "learning_rate": 0.00012, | |
| "loss": 1.7818, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.005595523581135092, | |
| "grad_norm": 0.9461150169372559, | |
| "learning_rate": 0.00014, | |
| "loss": 1.5379, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.006394884092725819, | |
| "grad_norm": 1.3639153242111206, | |
| "learning_rate": 0.00016, | |
| "loss": 1.4683, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.007194244604316547, | |
| "grad_norm": 1.1742007732391357, | |
| "learning_rate": 0.00018, | |
| "loss": 1.5821, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.007993605115907274, | |
| "grad_norm": 0.6428038477897644, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4243, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008792965627498001, | |
| "grad_norm": 0.5557376146316528, | |
| "learning_rate": 0.0001999996795753233, | |
| "loss": 1.3842, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.009592326139088728, | |
| "grad_norm": 0.35388731956481934, | |
| "learning_rate": 0.00019999871830334666, | |
| "loss": 1.4213, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.010391686650679457, | |
| "grad_norm": 0.3517228662967682, | |
| "learning_rate": 0.00019999711619023033, | |
| "loss": 1.3737, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.011191047162270184, | |
| "grad_norm": 0.37514588236808777, | |
| "learning_rate": 0.0001999948732462415, | |
| "loss": 1.6331, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.011990407673860911, | |
| "grad_norm": 0.4272635579109192, | |
| "learning_rate": 0.00019999198948575405, | |
| "loss": 1.4726, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.012789768185451638, | |
| "grad_norm": 0.4880916476249695, | |
| "learning_rate": 0.00019998846492724848, | |
| "loss": 1.4294, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.013589128697042365, | |
| "grad_norm": 0.38794147968292236, | |
| "learning_rate": 0.00019998429959331202, | |
| "loss": 1.4384, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.014388489208633094, | |
| "grad_norm": 0.31120356917381287, | |
| "learning_rate": 0.00019997949351063807, | |
| "loss": 1.3983, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.015187849720223821, | |
| "grad_norm": 0.4935799539089203, | |
| "learning_rate": 0.00019997404671002645, | |
| "loss": 1.263, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.01598721023181455, | |
| "grad_norm": 0.329028844833374, | |
| "learning_rate": 0.00019996795922638293, | |
| "loss": 1.2443, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.016786570743405275, | |
| "grad_norm": 0.2973022758960724, | |
| "learning_rate": 0.00019996123109871906, | |
| "loss": 1.2702, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.017585931254996003, | |
| "grad_norm": 0.32669177651405334, | |
| "learning_rate": 0.00019995386237015207, | |
| "loss": 1.3152, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01838529176658673, | |
| "grad_norm": 0.3220458924770355, | |
| "learning_rate": 0.00019994585308790437, | |
| "loss": 1.4059, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.019184652278177457, | |
| "grad_norm": 0.38787347078323364, | |
| "learning_rate": 0.00019993720330330344, | |
| "loss": 1.1777, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.019984012789768184, | |
| "grad_norm": 0.695192277431488, | |
| "learning_rate": 0.00019992791307178132, | |
| "loss": 1.4574, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.020783373301358914, | |
| "grad_norm": 0.6521245837211609, | |
| "learning_rate": 0.00019991798245287437, | |
| "loss": 1.1948, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.02158273381294964, | |
| "grad_norm": 0.5856667757034302, | |
| "learning_rate": 0.00019990741151022301, | |
| "loss": 1.2414, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02238209432454037, | |
| "grad_norm": 0.5117306709289551, | |
| "learning_rate": 0.00019989620031157096, | |
| "loss": 1.273, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.023181454836131096, | |
| "grad_norm": 0.4806709587574005, | |
| "learning_rate": 0.00019988434892876514, | |
| "loss": 1.2461, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.023980815347721823, | |
| "grad_norm": 0.4751984179019928, | |
| "learning_rate": 0.00019987185743775506, | |
| "loss": 1.3404, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02478017585931255, | |
| "grad_norm": 0.5241938233375549, | |
| "learning_rate": 0.0001998587259185924, | |
| "loss": 1.2368, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.025579536370903277, | |
| "grad_norm": 0.4881473779678345, | |
| "learning_rate": 0.00019984495445543035, | |
| "loss": 1.1433, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.026378896882494004, | |
| "grad_norm": 0.5861169695854187, | |
| "learning_rate": 0.00019983054313652327, | |
| "loss": 1.3639, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.02717825739408473, | |
| "grad_norm": 0.5575733780860901, | |
| "learning_rate": 0.000199815492054226, | |
| "loss": 1.2566, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.027977617905675458, | |
| "grad_norm": 0.5485345721244812, | |
| "learning_rate": 0.00019979980130499332, | |
| "loss": 1.2673, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02877697841726619, | |
| "grad_norm": 0.678505003452301, | |
| "learning_rate": 0.0001997834709893793, | |
| "loss": 1.2663, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.029576338928856916, | |
| "grad_norm": 0.7110869884490967, | |
| "learning_rate": 0.00019976650121203662, | |
| "loss": 1.3822, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.030375699440447643, | |
| "grad_norm": 0.8161877989768982, | |
| "learning_rate": 0.000199748892081716, | |
| "loss": 1.1504, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03117505995203837, | |
| "grad_norm": 0.7314615845680237, | |
| "learning_rate": 0.00019973064371126548, | |
| "loss": 1.3762, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0319744204636291, | |
| "grad_norm": 0.7901433706283569, | |
| "learning_rate": 0.0001997117562176296, | |
| "loss": 1.2416, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.032773780975219824, | |
| "grad_norm": 0.7982814908027649, | |
| "learning_rate": 0.00019969222972184873, | |
| "loss": 1.3356, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03357314148681055, | |
| "grad_norm": 1.3451604843139648, | |
| "learning_rate": 0.0001996720643490583, | |
| "loss": 1.3487, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03437250199840128, | |
| "grad_norm": 0.8946935534477234, | |
| "learning_rate": 0.00019965126022848795, | |
| "loss": 1.3564, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.035171862509992005, | |
| "grad_norm": 1.7600996494293213, | |
| "learning_rate": 0.00019962981749346078, | |
| "loss": 1.2413, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.03597122302158273, | |
| "grad_norm": 1.187146544456482, | |
| "learning_rate": 0.0001996077362813924, | |
| "loss": 1.4331, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03677058353317346, | |
| "grad_norm": 1.0808261632919312, | |
| "learning_rate": 0.00019958501673379017, | |
| "loss": 1.3436, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.037569944044764186, | |
| "grad_norm": 1.1636722087860107, | |
| "learning_rate": 0.0001995616589962521, | |
| "loss": 1.3923, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03836930455635491, | |
| "grad_norm": 1.2179851531982422, | |
| "learning_rate": 0.0001995376632184661, | |
| "loss": 1.4711, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.03916866506794564, | |
| "grad_norm": 1.260860800743103, | |
| "learning_rate": 0.000199513029554209, | |
| "loss": 1.5432, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03996802557953637, | |
| "grad_norm": 1.9464174509048462, | |
| "learning_rate": 0.00019948775816134545, | |
| "loss": 1.511, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0407673860911271, | |
| "grad_norm": 0.4390850067138672, | |
| "learning_rate": 0.00019946184920182703, | |
| "loss": 0.9991, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04156674660271783, | |
| "grad_norm": 0.42749834060668945, | |
| "learning_rate": 0.00019943530284169107, | |
| "loss": 1.1283, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.042366107114308556, | |
| "grad_norm": 0.4460582435131073, | |
| "learning_rate": 0.00019940811925105983, | |
| "loss": 1.2345, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04316546762589928, | |
| "grad_norm": 0.45763692259788513, | |
| "learning_rate": 0.00019938029860413914, | |
| "loss": 1.2941, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04396482813749001, | |
| "grad_norm": 0.46411487460136414, | |
| "learning_rate": 0.00019935184107921746, | |
| "loss": 1.3687, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04476418864908074, | |
| "grad_norm": 0.388842910528183, | |
| "learning_rate": 0.0001993227468586646, | |
| "loss": 1.1788, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.045563549160671464, | |
| "grad_norm": 0.3251805007457733, | |
| "learning_rate": 0.00019929301612893076, | |
| "loss": 1.3171, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.04636290967226219, | |
| "grad_norm": 0.3713827431201935, | |
| "learning_rate": 0.00019926264908054505, | |
| "loss": 1.2828, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.04716227018385292, | |
| "grad_norm": 0.29696419835090637, | |
| "learning_rate": 0.00019923164590811453, | |
| "loss": 1.2708, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.047961630695443645, | |
| "grad_norm": 0.3488282263278961, | |
| "learning_rate": 0.00019920000681032288, | |
| "loss": 1.2673, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04876099120703437, | |
| "grad_norm": 0.3075105547904968, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 1.3265, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0495603517186251, | |
| "grad_norm": 0.3479640781879425, | |
| "learning_rate": 0.0001991348216537659, | |
| "loss": 1.3545, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.050359712230215826, | |
| "grad_norm": 0.3479630947113037, | |
| "learning_rate": 0.00019910127601273922, | |
| "loss": 1.2322, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.051159072741806554, | |
| "grad_norm": 0.32833385467529297, | |
| "learning_rate": 0.000199067095281826, | |
| "loss": 1.3306, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.05195843325339728, | |
| "grad_norm": 0.45465975999832153, | |
| "learning_rate": 0.00019903227968007326, | |
| "loss": 1.1733, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05275779376498801, | |
| "grad_norm": 0.31810107827186584, | |
| "learning_rate": 0.0001989968294305965, | |
| "loss": 1.1675, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.053557154276578735, | |
| "grad_norm": 0.3302905857563019, | |
| "learning_rate": 0.00019896074476057846, | |
| "loss": 1.1695, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.05435651478816946, | |
| "grad_norm": 0.34600090980529785, | |
| "learning_rate": 0.0001989240259012675, | |
| "loss": 1.1706, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.05515587529976019, | |
| "grad_norm": 0.3320174217224121, | |
| "learning_rate": 0.0001988866730879762, | |
| "loss": 1.2779, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.055955235811350916, | |
| "grad_norm": 0.4127947688102722, | |
| "learning_rate": 0.00019884868656007982, | |
| "loss": 1.1394, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05675459632294165, | |
| "grad_norm": 0.4041900038719177, | |
| "learning_rate": 0.00019881006656101479, | |
| "loss": 1.3192, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.05755395683453238, | |
| "grad_norm": 0.390407919883728, | |
| "learning_rate": 0.00019877081333827712, | |
| "loss": 1.261, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.058353317346123104, | |
| "grad_norm": 0.3832196295261383, | |
| "learning_rate": 0.00019873092714342078, | |
| "loss": 1.1741, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.05915267785771383, | |
| "grad_norm": 0.3609188199043274, | |
| "learning_rate": 0.0001986904082320563, | |
| "loss": 1.279, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.05995203836930456, | |
| "grad_norm": 0.3944198191165924, | |
| "learning_rate": 0.00019864925686384873, | |
| "loss": 1.1863, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.060751398880895285, | |
| "grad_norm": 0.4218379855155945, | |
| "learning_rate": 0.00019860747330251646, | |
| "loss": 1.1365, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.06155075939248601, | |
| "grad_norm": 0.4537956416606903, | |
| "learning_rate": 0.0001985650578158291, | |
| "loss": 1.131, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.06235011990407674, | |
| "grad_norm": 0.3956335186958313, | |
| "learning_rate": 0.00019852201067560606, | |
| "loss": 1.2713, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.06314948041566747, | |
| "grad_norm": 0.5430086255073547, | |
| "learning_rate": 0.00019847833215771467, | |
| "loss": 1.2556, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.0639488409272582, | |
| "grad_norm": 0.40932217240333557, | |
| "learning_rate": 0.0001984340225420684, | |
| "loss": 0.9753, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06474820143884892, | |
| "grad_norm": 0.5276725888252258, | |
| "learning_rate": 0.00019838908211262512, | |
| "loss": 1.1394, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.06554756195043965, | |
| "grad_norm": 0.6045296788215637, | |
| "learning_rate": 0.00019834351115738535, | |
| "loss": 1.128, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.06634692246203037, | |
| "grad_norm": 0.6298900842666626, | |
| "learning_rate": 0.0001982973099683902, | |
| "loss": 1.2829, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.0671462829736211, | |
| "grad_norm": 0.6042550206184387, | |
| "learning_rate": 0.0001982504788417197, | |
| "loss": 1.1628, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.06794564348521183, | |
| "grad_norm": 0.6109876036643982, | |
| "learning_rate": 0.00019820301807749085, | |
| "loss": 1.0785, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.06874500399680256, | |
| "grad_norm": 0.6248145699501038, | |
| "learning_rate": 0.0001981549279798556, | |
| "loss": 1.1856, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.06954436450839328, | |
| "grad_norm": 0.74518883228302, | |
| "learning_rate": 0.00019810620885699905, | |
| "loss": 1.1553, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.07034372501998401, | |
| "grad_norm": 0.7344063520431519, | |
| "learning_rate": 0.00019805686102113747, | |
| "loss": 1.2014, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.07114308553157474, | |
| "grad_norm": 0.7360933423042297, | |
| "learning_rate": 0.000198006884788516, | |
| "loss": 1.1937, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.07194244604316546, | |
| "grad_norm": 0.7342174649238586, | |
| "learning_rate": 0.00019795628047940714, | |
| "loss": 1.2054, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07274180655475619, | |
| "grad_norm": 0.7293947339057922, | |
| "learning_rate": 0.00019790504841810818, | |
| "loss": 1.2446, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.07354116706634692, | |
| "grad_norm": 0.7864716053009033, | |
| "learning_rate": 0.0001978531889329395, | |
| "loss": 1.182, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.07434052757793765, | |
| "grad_norm": 0.8126608729362488, | |
| "learning_rate": 0.00019780070235624221, | |
| "loss": 1.3634, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07513988808952837, | |
| "grad_norm": 0.797795295715332, | |
| "learning_rate": 0.00019774758902437633, | |
| "loss": 1.0883, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0759392486011191, | |
| "grad_norm": 0.8480005860328674, | |
| "learning_rate": 0.0001976938492777182, | |
| "loss": 1.3654, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07673860911270983, | |
| "grad_norm": 1.0392693281173706, | |
| "learning_rate": 0.00019763948346065863, | |
| "loss": 1.3555, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.07753796962430055, | |
| "grad_norm": 1.147118330001831, | |
| "learning_rate": 0.00019758449192160067, | |
| "loss": 1.3023, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.07833733013589128, | |
| "grad_norm": 1.447702169418335, | |
| "learning_rate": 0.0001975288750129572, | |
| "loss": 1.4948, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.07913669064748201, | |
| "grad_norm": 1.1526646614074707, | |
| "learning_rate": 0.00019747263309114886, | |
| "loss": 1.2744, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.07993605115907274, | |
| "grad_norm": 1.8174890279769897, | |
| "learning_rate": 0.00019741576651660157, | |
| "loss": 1.5239, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08073541167066348, | |
| "grad_norm": 0.33395928144454956, | |
| "learning_rate": 0.00019735827565374448, | |
| "loss": 0.8391, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.0815347721822542, | |
| "grad_norm": 0.48688215017318726, | |
| "learning_rate": 0.00019730016087100738, | |
| "loss": 1.0899, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.08233413269384493, | |
| "grad_norm": 0.3916321098804474, | |
| "learning_rate": 0.0001972414225408185, | |
| "loss": 1.2562, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.08313349320543566, | |
| "grad_norm": 0.38845309615135193, | |
| "learning_rate": 0.00019718206103960199, | |
| "loss": 1.2981, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.08393285371702638, | |
| "grad_norm": 0.35215434432029724, | |
| "learning_rate": 0.0001971220767477757, | |
| "loss": 0.9815, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.08473221422861711, | |
| "grad_norm": 0.32380643486976624, | |
| "learning_rate": 0.00019706147004974858, | |
| "loss": 1.0154, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.08553157474020784, | |
| "grad_norm": 0.35980847477912903, | |
| "learning_rate": 0.00019700024133391825, | |
| "loss": 1.2426, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.08633093525179857, | |
| "grad_norm": 0.3244008421897888, | |
| "learning_rate": 0.00019693839099266854, | |
| "loss": 1.3621, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.08713029576338929, | |
| "grad_norm": 0.4032576382160187, | |
| "learning_rate": 0.00019687591942236694, | |
| "loss": 1.2899, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.08792965627498002, | |
| "grad_norm": 0.36872512102127075, | |
| "learning_rate": 0.00019681282702336213, | |
| "loss": 1.1612, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08872901678657075, | |
| "grad_norm": 0.3233664929866791, | |
| "learning_rate": 0.0001967491141999813, | |
| "loss": 1.2213, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.08952837729816147, | |
| "grad_norm": 0.3443772792816162, | |
| "learning_rate": 0.00019668478136052774, | |
| "loss": 1.1822, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0903277378097522, | |
| "grad_norm": 0.3623986840248108, | |
| "learning_rate": 0.000196619828917278, | |
| "loss": 1.2398, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.09112709832134293, | |
| "grad_norm": 0.2766810357570648, | |
| "learning_rate": 0.0001965542572864794, | |
| "loss": 1.2032, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.09192645883293366, | |
| "grad_norm": 0.3393819332122803, | |
| "learning_rate": 0.00019648806688834727, | |
| "loss": 1.3039, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.09272581934452438, | |
| "grad_norm": 0.33124133944511414, | |
| "learning_rate": 0.0001964212581470624, | |
| "loss": 1.137, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.09352517985611511, | |
| "grad_norm": 0.3681512773036957, | |
| "learning_rate": 0.00019635383149076816, | |
| "loss": 1.1675, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.09432454036770584, | |
| "grad_norm": 0.4348418116569519, | |
| "learning_rate": 0.00019628578735156782, | |
| "loss": 1.1787, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.09512390087929656, | |
| "grad_norm": 0.37171006202697754, | |
| "learning_rate": 0.00019621712616552182, | |
| "loss": 1.137, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.09592326139088729, | |
| "grad_norm": 0.42151620984077454, | |
| "learning_rate": 0.00019614784837264493, | |
| "loss": 1.2804, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09672262190247802, | |
| "grad_norm": 0.47245654463768005, | |
| "learning_rate": 0.00019607795441690343, | |
| "loss": 1.1102, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.09752198241406874, | |
| "grad_norm": 0.42258185148239136, | |
| "learning_rate": 0.00019600744474621232, | |
| "loss": 1.2432, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.09832134292565947, | |
| "grad_norm": 0.3666921555995941, | |
| "learning_rate": 0.0001959363198124323, | |
| "loss": 0.9307, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.0991207034372502, | |
| "grad_norm": 0.4099723696708679, | |
| "learning_rate": 0.00019586458007136713, | |
| "loss": 1.0751, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.09992006394884093, | |
| "grad_norm": 0.4759383201599121, | |
| "learning_rate": 0.00019579222598276042, | |
| "loss": 1.021, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.10071942446043165, | |
| "grad_norm": 0.3923790156841278, | |
| "learning_rate": 0.0001957192580102929, | |
| "loss": 1.0722, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.10151878497202238, | |
| "grad_norm": 0.47697147727012634, | |
| "learning_rate": 0.0001956456766215793, | |
| "loss": 1.1207, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.10231814548361311, | |
| "grad_norm": 0.4863724410533905, | |
| "learning_rate": 0.00019557148228816558, | |
| "loss": 1.0626, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.10311750599520383, | |
| "grad_norm": 0.576336681842804, | |
| "learning_rate": 0.00019549667548552556, | |
| "loss": 1.0729, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.10391686650679456, | |
| "grad_norm": 0.7830796241760254, | |
| "learning_rate": 0.00019542125669305812, | |
| "loss": 1.3441, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10471622701838529, | |
| "grad_norm": 0.5463136434555054, | |
| "learning_rate": 0.0001953452263940842, | |
| "loss": 1.1252, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.10551558752997602, | |
| "grad_norm": 0.7431773543357849, | |
| "learning_rate": 0.00019526858507584347, | |
| "loss": 1.3767, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.10631494804156674, | |
| "grad_norm": 0.6066572666168213, | |
| "learning_rate": 0.00019519133322949122, | |
| "loss": 1.1088, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.10711430855315747, | |
| "grad_norm": 0.5295929312705994, | |
| "learning_rate": 0.00019511347135009555, | |
| "loss": 1.1324, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1079136690647482, | |
| "grad_norm": 0.6992323994636536, | |
| "learning_rate": 0.0001950349999366337, | |
| "loss": 1.2556, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.10871302957633892, | |
| "grad_norm": 0.6171619296073914, | |
| "learning_rate": 0.00019495591949198927, | |
| "loss": 1.1685, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.10951239008792965, | |
| "grad_norm": 0.6525933146476746, | |
| "learning_rate": 0.00019487623052294873, | |
| "loss": 1.3022, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.11031175059952038, | |
| "grad_norm": 0.8878188729286194, | |
| "learning_rate": 0.0001947959335401984, | |
| "loss": 1.2187, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.6956382393836975, | |
| "learning_rate": 0.0001947150290583209, | |
| "loss": 1.3016, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.11191047162270183, | |
| "grad_norm": 0.7935355305671692, | |
| "learning_rate": 0.00019463351759579212, | |
| "loss": 1.2124, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11270983213429256, | |
| "grad_norm": 0.8994975686073303, | |
| "learning_rate": 0.0001945513996749777, | |
| "loss": 1.2633, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.1135091926458833, | |
| "grad_norm": 0.7103203535079956, | |
| "learning_rate": 0.00019446867582212986, | |
| "loss": 1.1257, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.11430855315747403, | |
| "grad_norm": 1.077649712562561, | |
| "learning_rate": 0.00019438534656738386, | |
| "loss": 1.2635, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.11510791366906475, | |
| "grad_norm": 0.9161784648895264, | |
| "learning_rate": 0.0001943014124447547, | |
| "loss": 1.3306, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.11590727418065548, | |
| "grad_norm": 1.036491870880127, | |
| "learning_rate": 0.0001942168739921336, | |
| "loss": 1.5151, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.11670663469224621, | |
| "grad_norm": 0.8974622488021851, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 1.4381, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.11750599520383694, | |
| "grad_norm": 1.1696873903274536, | |
| "learning_rate": 0.0001940459862678416, | |
| "loss": 1.4909, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.11830535571542766, | |
| "grad_norm": 1.1506264209747314, | |
| "learning_rate": 0.00019395963809130354, | |
| "loss": 1.3273, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.11910471622701839, | |
| "grad_norm": 1.7458105087280273, | |
| "learning_rate": 0.00019387268777503236, | |
| "loss": 1.2036, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.11990407673860912, | |
| "grad_norm": 1.8019418716430664, | |
| "learning_rate": 0.00019378513587624854, | |
| "loss": 1.3185, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12070343725019984, | |
| "grad_norm": 0.37937843799591064, | |
| "learning_rate": 0.00019369698295602786, | |
| "loss": 0.9703, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.12150279776179057, | |
| "grad_norm": 0.4317457675933838, | |
| "learning_rate": 0.0001936082295792978, | |
| "loss": 1.1408, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.1223021582733813, | |
| "grad_norm": 0.3610316216945648, | |
| "learning_rate": 0.0001935188763148337, | |
| "loss": 1.0476, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.12310151878497202, | |
| "grad_norm": 0.4185445010662079, | |
| "learning_rate": 0.00019342892373525546, | |
| "loss": 1.1442, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.12390087929656275, | |
| "grad_norm": 0.5041868686676025, | |
| "learning_rate": 0.00019333837241702358, | |
| "loss": 1.2064, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.12470023980815348, | |
| "grad_norm": 0.36838433146476746, | |
| "learning_rate": 0.00019324722294043558, | |
| "loss": 1.1642, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.1254996003197442, | |
| "grad_norm": 0.3606477379798889, | |
| "learning_rate": 0.0001931554758896223, | |
| "loss": 1.1654, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.12629896083133493, | |
| "grad_norm": 0.3449469804763794, | |
| "learning_rate": 0.00019306313185254417, | |
| "loss": 1.1154, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.12709832134292565, | |
| "grad_norm": 0.33281880617141724, | |
| "learning_rate": 0.00019297019142098723, | |
| "loss": 1.3186, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.1278976818545164, | |
| "grad_norm": 0.34775304794311523, | |
| "learning_rate": 0.00019287665519055977, | |
| "loss": 1.1772, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1286970423661071, | |
| "grad_norm": 0.32196998596191406, | |
| "learning_rate": 0.00019278252376068807, | |
| "loss": 1.1682, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.12949640287769784, | |
| "grad_norm": 0.3235302269458771, | |
| "learning_rate": 0.00019268779773461275, | |
| "loss": 1.2322, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.13029576338928858, | |
| "grad_norm": 0.31604263186454773, | |
| "learning_rate": 0.000192592477719385, | |
| "loss": 1.1779, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1310951239008793, | |
| "grad_norm": 0.29469355940818787, | |
| "learning_rate": 0.00019249656432586247, | |
| "loss": 1.1083, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.13189448441247004, | |
| "grad_norm": 0.3672734797000885, | |
| "learning_rate": 0.00019240005816870554, | |
| "loss": 1.1481, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.13269384492406075, | |
| "grad_norm": 0.35141581296920776, | |
| "learning_rate": 0.0001923029598663733, | |
| "loss": 1.0356, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.1334932054356515, | |
| "grad_norm": 0.38158297538757324, | |
| "learning_rate": 0.0001922052700411196, | |
| "loss": 1.1449, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.1342925659472422, | |
| "grad_norm": 0.33295100927352905, | |
| "learning_rate": 0.00019210698931898903, | |
| "loss": 1.1372, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.13509192645883294, | |
| "grad_norm": 0.4519907236099243, | |
| "learning_rate": 0.00019200811832981298, | |
| "loss": 1.1314, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.13589128697042366, | |
| "grad_norm": 0.3855970501899719, | |
| "learning_rate": 0.00019190865770720553, | |
| "loss": 1.0639, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1366906474820144, | |
| "grad_norm": 0.405184268951416, | |
| "learning_rate": 0.00019180860808855945, | |
| "loss": 1.2448, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.1374900079936051, | |
| "grad_norm": 0.41066446900367737, | |
| "learning_rate": 0.00019170797011504206, | |
| "loss": 1.0956, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.13828936850519585, | |
| "grad_norm": 0.3971129059791565, | |
| "learning_rate": 0.00019160674443159122, | |
| "loss": 1.0819, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.13908872901678657, | |
| "grad_norm": 0.43992650508880615, | |
| "learning_rate": 0.00019150493168691094, | |
| "loss": 1.248, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.1398880895283773, | |
| "grad_norm": 0.41885894536972046, | |
| "learning_rate": 0.00019140253253346768, | |
| "loss": 1.1889, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.14068745003996802, | |
| "grad_norm": 0.47921714186668396, | |
| "learning_rate": 0.00019129954762748568, | |
| "loss": 1.093, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.14148681055155876, | |
| "grad_norm": 0.4085664451122284, | |
| "learning_rate": 0.00019119597762894304, | |
| "loss": 1.2167, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.14228617106314947, | |
| "grad_norm": 0.6234189867973328, | |
| "learning_rate": 0.00019109182320156744, | |
| "loss": 1.1884, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.14308553157474022, | |
| "grad_norm": 0.45891621708869934, | |
| "learning_rate": 0.00019098708501283186, | |
| "loss": 1.1723, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.14388489208633093, | |
| "grad_norm": 0.46962589025497437, | |
| "learning_rate": 0.0001908817637339503, | |
| "loss": 1.0018, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14468425259792167, | |
| "grad_norm": 0.5809106230735779, | |
| "learning_rate": 0.0001907758600398735, | |
| "loss": 1.099, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.14548361310951238, | |
| "grad_norm": 0.48514389991760254, | |
| "learning_rate": 0.0001906693746092846, | |
| "loss": 1.0701, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.14628297362110312, | |
| "grad_norm": 0.5460750460624695, | |
| "learning_rate": 0.00019056230812459474, | |
| "loss": 1.2026, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.14708233413269384, | |
| "grad_norm": 0.5718762874603271, | |
| "learning_rate": 0.0001904546612719389, | |
| "loss": 1.1434, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.14788169464428458, | |
| "grad_norm": 0.5084953904151917, | |
| "learning_rate": 0.00019034643474117113, | |
| "loss": 1.0963, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1486810551558753, | |
| "grad_norm": 0.7148297429084778, | |
| "learning_rate": 0.00019023762922586052, | |
| "loss": 1.3716, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.14948041566746603, | |
| "grad_norm": 0.6428340673446655, | |
| "learning_rate": 0.00019012824542328648, | |
| "loss": 1.028, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.15027977617905675, | |
| "grad_norm": 0.7076712846755981, | |
| "learning_rate": 0.00019001828403443445, | |
| "loss": 1.1724, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.1510791366906475, | |
| "grad_norm": 0.7304551601409912, | |
| "learning_rate": 0.00018990774576399123, | |
| "loss": 1.2831, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.1518784972022382, | |
| "grad_norm": 0.673502504825592, | |
| "learning_rate": 0.00018979663132034062, | |
| "loss": 1.2381, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.15267785771382894, | |
| "grad_norm": 0.7513559460639954, | |
| "learning_rate": 0.0001896849414155588, | |
| "loss": 1.1118, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.15347721822541965, | |
| "grad_norm": 0.7825088500976562, | |
| "learning_rate": 0.00018957267676540987, | |
| "loss": 1.2032, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.1542765787370104, | |
| "grad_norm": 0.8123567700386047, | |
| "learning_rate": 0.00018945983808934104, | |
| "loss": 1.2227, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.1550759392486011, | |
| "grad_norm": 0.8477492332458496, | |
| "learning_rate": 0.00018934642611047827, | |
| "loss": 1.3514, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.15587529976019185, | |
| "grad_norm": 0.8480494618415833, | |
| "learning_rate": 0.00018923244155562147, | |
| "loss": 1.1612, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.15667466027178256, | |
| "grad_norm": 1.06630539894104, | |
| "learning_rate": 0.00018911788515523997, | |
| "loss": 1.4464, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.1574740207833733, | |
| "grad_norm": 1.0468109846115112, | |
| "learning_rate": 0.00018900275764346768, | |
| "loss": 1.4918, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.15827338129496402, | |
| "grad_norm": 1.234046459197998, | |
| "learning_rate": 0.0001888870597580985, | |
| "loss": 1.3812, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.15907274180655476, | |
| "grad_norm": 1.3013675212860107, | |
| "learning_rate": 0.00018877079224058165, | |
| "loss": 1.1392, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.15987210231814547, | |
| "grad_norm": 1.3213999271392822, | |
| "learning_rate": 0.00018865395583601667, | |
| "loss": 1.0561, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1606714628297362, | |
| "grad_norm": 0.33066526055336, | |
| "learning_rate": 0.000188536551293149, | |
| "loss": 1.0385, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.16147082334132695, | |
| "grad_norm": 0.41339030861854553, | |
| "learning_rate": 0.0001884185793643648, | |
| "loss": 1.4136, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.16227018385291767, | |
| "grad_norm": 0.37679141759872437, | |
| "learning_rate": 0.0001883000408056865, | |
| "loss": 1.2518, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.1630695443645084, | |
| "grad_norm": 0.4325602352619171, | |
| "learning_rate": 0.0001881809363767676, | |
| "loss": 1.2739, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.16386890487609912, | |
| "grad_norm": 0.3498510718345642, | |
| "learning_rate": 0.00018806126684088815, | |
| "loss": 1.1294, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.16466826538768986, | |
| "grad_norm": 0.4043656885623932, | |
| "learning_rate": 0.00018794103296494955, | |
| "loss": 1.0051, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.16546762589928057, | |
| "grad_norm": 0.3355083465576172, | |
| "learning_rate": 0.0001878202355194698, | |
| "loss": 1.132, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.16626698641087131, | |
| "grad_norm": 0.35696941614151, | |
| "learning_rate": 0.0001876988752785786, | |
| "loss": 1.0992, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.16706634692246203, | |
| "grad_norm": 0.3169572949409485, | |
| "learning_rate": 0.00018757695302001223, | |
| "loss": 1.2305, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.16786570743405277, | |
| "grad_norm": 0.38945549726486206, | |
| "learning_rate": 0.0001874544695251087, | |
| "loss": 1.1978, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.16866506794564348, | |
| "grad_norm": 0.35129424929618835, | |
| "learning_rate": 0.00018733142557880267, | |
| "loss": 1.1808, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.16946442845723422, | |
| "grad_norm": 0.3321034014225006, | |
| "learning_rate": 0.00018720782196962054, | |
| "loss": 1.1648, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.17026378896882494, | |
| "grad_norm": 0.3599945604801178, | |
| "learning_rate": 0.00018708365948967522, | |
| "loss": 1.2779, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.17106314948041568, | |
| "grad_norm": 0.3521081209182739, | |
| "learning_rate": 0.0001869589389346611, | |
| "loss": 1.1971, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.1718625099920064, | |
| "grad_norm": 0.3376477062702179, | |
| "learning_rate": 0.00018683366110384912, | |
| "loss": 1.3075, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.17266187050359713, | |
| "grad_norm": 0.36610740423202515, | |
| "learning_rate": 0.00018670782680008143, | |
| "loss": 1.2168, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.17346123101518784, | |
| "grad_norm": 0.31685808300971985, | |
| "learning_rate": 0.00018658143682976635, | |
| "loss": 1.0374, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.17426059152677859, | |
| "grad_norm": 0.3511636257171631, | |
| "learning_rate": 0.00018645449200287314, | |
| "loss": 0.9931, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.1750599520383693, | |
| "grad_norm": 0.4250391125679016, | |
| "learning_rate": 0.000186326993132927, | |
| "loss": 1.1782, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.17585931254996004, | |
| "grad_norm": 0.3989580571651459, | |
| "learning_rate": 0.00018619894103700352, | |
| "loss": 1.1014, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.17665867306155075, | |
| "grad_norm": 0.409648060798645, | |
| "learning_rate": 0.00018607033653572378, | |
| "loss": 1.0888, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.1774580335731415, | |
| "grad_norm": 0.38411882519721985, | |
| "learning_rate": 0.00018594118045324887, | |
| "loss": 1.0785, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.1782573940847322, | |
| "grad_norm": 0.45235031843185425, | |
| "learning_rate": 0.00018581147361727477, | |
| "loss": 1.2181, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.17905675459632295, | |
| "grad_norm": 0.47184649109840393, | |
| "learning_rate": 0.00018568121685902682, | |
| "loss": 1.0918, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.17985611510791366, | |
| "grad_norm": 0.4382520616054535, | |
| "learning_rate": 0.00018555041101325466, | |
| "loss": 1.1234, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1806554756195044, | |
| "grad_norm": 0.4980821907520294, | |
| "learning_rate": 0.00018541905691822668, | |
| "loss": 1.2183, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.18145483613109512, | |
| "grad_norm": 0.6325019001960754, | |
| "learning_rate": 0.00018528715541572474, | |
| "loss": 0.9963, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.18225419664268586, | |
| "grad_norm": 0.48259237408638, | |
| "learning_rate": 0.0001851547073510388, | |
| "loss": 1.1, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.18305355715427657, | |
| "grad_norm": 0.5209122896194458, | |
| "learning_rate": 0.00018502171357296144, | |
| "loss": 1.0281, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.1838529176658673, | |
| "grad_norm": 0.5251787900924683, | |
| "learning_rate": 0.00018488817493378235, | |
| "loss": 1.1696, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.18465227817745802, | |
| "grad_norm": 0.522381067276001, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 1.0588, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.18545163868904876, | |
| "grad_norm": 0.5586369037628174, | |
| "learning_rate": 0.00018461946649873148, | |
| "loss": 1.0162, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.18625099920063948, | |
| "grad_norm": 0.6529503464698792, | |
| "learning_rate": 0.0001844842984248759, | |
| "loss": 1.1062, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.18705035971223022, | |
| "grad_norm": 0.6378787159919739, | |
| "learning_rate": 0.00018434858893394017, | |
| "loss": 1.1263, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.18784972022382093, | |
| "grad_norm": 0.6238505840301514, | |
| "learning_rate": 0.00018421233889561765, | |
| "loss": 1.1797, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.18864908073541167, | |
| "grad_norm": 0.7397989630699158, | |
| "learning_rate": 0.00018407554918306582, | |
| "loss": 1.1618, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.18944844124700239, | |
| "grad_norm": 0.6958909034729004, | |
| "learning_rate": 0.0001839382206729007, | |
| "loss": 1.1088, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.19024780175859313, | |
| "grad_norm": 0.8362878561019897, | |
| "learning_rate": 0.00018380035424519115, | |
| "loss": 1.1162, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.19104716227018384, | |
| "grad_norm": 0.6862859725952148, | |
| "learning_rate": 0.00018366195078345327, | |
| "loss": 1.1711, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.19184652278177458, | |
| "grad_norm": 0.6498910188674927, | |
| "learning_rate": 0.00018352301117464473, | |
| "loss": 1.1287, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.1926458832933653, | |
| "grad_norm": 0.8801695108413696, | |
| "learning_rate": 0.00018338353630915915, | |
| "loss": 1.3638, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.19344524380495604, | |
| "grad_norm": 0.8688244223594666, | |
| "learning_rate": 0.00018324352708082032, | |
| "loss": 1.1142, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.19424460431654678, | |
| "grad_norm": 1.11106276512146, | |
| "learning_rate": 0.00018310298438687644, | |
| "loss": 1.0854, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.1950439648281375, | |
| "grad_norm": 1.1534092426300049, | |
| "learning_rate": 0.00018296190912799442, | |
| "loss": 1.2603, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.19584332533972823, | |
| "grad_norm": 1.00962233543396, | |
| "learning_rate": 0.00018282030220825423, | |
| "loss": 1.118, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.19664268585131894, | |
| "grad_norm": 1.2045316696166992, | |
| "learning_rate": 0.00018267816453514284, | |
| "loss": 1.3775, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.19744204636290968, | |
| "grad_norm": 0.9561203718185425, | |
| "learning_rate": 0.00018253549701954862, | |
| "loss": 1.114, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.1982414068745004, | |
| "grad_norm": 1.6712232828140259, | |
| "learning_rate": 0.00018239230057575542, | |
| "loss": 1.3807, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.19904076738609114, | |
| "grad_norm": 1.390113115310669, | |
| "learning_rate": 0.00018224857612143675, | |
| "loss": 1.2788, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.19984012789768185, | |
| "grad_norm": 2.1578612327575684, | |
| "learning_rate": 0.00018210432457764982, | |
| "loss": 1.2059, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2006394884092726, | |
| "grad_norm": 0.40902066230773926, | |
| "learning_rate": 0.00018195954686882972, | |
| "loss": 1.147, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.2014388489208633, | |
| "grad_norm": 0.41376692056655884, | |
| "learning_rate": 0.00018181424392278345, | |
| "loss": 1.1842, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.20223820943245405, | |
| "grad_norm": 0.4128184914588928, | |
| "learning_rate": 0.00018166841667068405, | |
| "loss": 1.0763, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.20303756994404476, | |
| "grad_norm": 0.4061759114265442, | |
| "learning_rate": 0.00018152206604706445, | |
| "loss": 1.2217, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.2038369304556355, | |
| "grad_norm": 0.42886969447135925, | |
| "learning_rate": 0.00018137519298981172, | |
| "loss": 1.2359, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.20463629096722621, | |
| "grad_norm": 0.32949739694595337, | |
| "learning_rate": 0.00018122779844016086, | |
| "loss": 0.9807, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.20543565147881696, | |
| "grad_norm": 0.39663881063461304, | |
| "learning_rate": 0.00018107988334268896, | |
| "loss": 1.144, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.20623501199040767, | |
| "grad_norm": 0.41834643483161926, | |
| "learning_rate": 0.0001809314486453089, | |
| "loss": 1.0347, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2070343725019984, | |
| "grad_norm": 0.34301498532295227, | |
| "learning_rate": 0.0001807824952992635, | |
| "loss": 1.1805, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.20783373301358912, | |
| "grad_norm": 0.3798671364784241, | |
| "learning_rate": 0.00018063302425911935, | |
| "loss": 1.2937, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.20863309352517986, | |
| "grad_norm": 0.3300327658653259, | |
| "learning_rate": 0.00018048303648276058, | |
| "loss": 1.2345, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.20943245403677058, | |
| "grad_norm": 0.30229341983795166, | |
| "learning_rate": 0.0001803325329313829, | |
| "loss": 1.1839, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.21023181454836132, | |
| "grad_norm": 0.33945149183273315, | |
| "learning_rate": 0.00018018151456948738, | |
| "loss": 1.0572, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.21103117505995203, | |
| "grad_norm": 0.3481135666370392, | |
| "learning_rate": 0.00018002998236487421, | |
| "loss": 1.1325, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.21183053557154277, | |
| "grad_norm": 0.33425942063331604, | |
| "learning_rate": 0.00017987793728863651, | |
| "loss": 1.1208, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.21262989608313348, | |
| "grad_norm": 0.36472010612487793, | |
| "learning_rate": 0.00017972538031515417, | |
| "loss": 1.1532, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.21342925659472423, | |
| "grad_norm": 0.5549772381782532, | |
| "learning_rate": 0.0001795723124220876, | |
| "loss": 1.2021, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.21422861710631494, | |
| "grad_norm": 0.37048134207725525, | |
| "learning_rate": 0.00017941873459037142, | |
| "loss": 1.1668, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.21502797761790568, | |
| "grad_norm": 0.41433608531951904, | |
| "learning_rate": 0.0001792646478042081, | |
| "loss": 1.1087, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.2158273381294964, | |
| "grad_norm": 0.4452265202999115, | |
| "learning_rate": 0.00017911005305106187, | |
| "loss": 1.1752, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.21662669864108713, | |
| "grad_norm": 0.3953174948692322, | |
| "learning_rate": 0.0001789549513216522, | |
| "loss": 1.0982, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.21742605915267785, | |
| "grad_norm": 0.39090120792388916, | |
| "learning_rate": 0.0001787993436099475, | |
| "loss": 1.0464, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.2182254196642686, | |
| "grad_norm": 0.4702832102775574, | |
| "learning_rate": 0.0001786432309131588, | |
| "loss": 1.1884, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.2190247801758593, | |
| "grad_norm": 0.377530574798584, | |
| "learning_rate": 0.00017848661423173326, | |
| "loss": 1.2785, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.21982414068745004, | |
| "grad_norm": 0.39601218700408936, | |
| "learning_rate": 0.00017832949456934798, | |
| "loss": 1.145, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.22062350119904076, | |
| "grad_norm": 0.4364507794380188, | |
| "learning_rate": 0.00017817187293290318, | |
| "loss": 1.2276, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.2214228617106315, | |
| "grad_norm": 0.6152124404907227, | |
| "learning_rate": 0.00017801375033251618, | |
| "loss": 1.2185, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.42372921109199524, | |
| "learning_rate": 0.00017785512778151457, | |
| "loss": 0.8677, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.22302158273381295, | |
| "grad_norm": 0.5250888466835022, | |
| "learning_rate": 0.00017769600629643004, | |
| "loss": 1.0941, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.22382094324540366, | |
| "grad_norm": 0.48516589403152466, | |
| "learning_rate": 0.0001775363868969915, | |
| "loss": 1.0313, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2246203037569944, | |
| "grad_norm": 0.4571071267127991, | |
| "learning_rate": 0.0001773762706061189, | |
| "loss": 1.111, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.22541966426858512, | |
| "grad_norm": 0.595092236995697, | |
| "learning_rate": 0.00017721565844991643, | |
| "loss": 1.1941, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.22621902478017586, | |
| "grad_norm": 0.5247442126274109, | |
| "learning_rate": 0.00017705455145766607, | |
| "loss": 1.2034, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.2270183852917666, | |
| "grad_norm": 0.5719393491744995, | |
| "learning_rate": 0.0001768929506618209, | |
| "loss": 0.9624, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.2278177458033573, | |
| "grad_norm": 0.6198736429214478, | |
| "learning_rate": 0.00017673085709799866, | |
| "loss": 1.1507, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.22861710631494805, | |
| "grad_norm": 0.5668182969093323, | |
| "learning_rate": 0.00017656827180497482, | |
| "loss": 1.0515, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.22941646682653877, | |
| "grad_norm": 0.5910491347312927, | |
| "learning_rate": 0.0001764051958246762, | |
| "loss": 1.0355, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.2302158273381295, | |
| "grad_norm": 0.6988612413406372, | |
| "learning_rate": 0.0001762416302021742, | |
| "loss": 1.1333, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.23101518784972022, | |
| "grad_norm": 0.6751168370246887, | |
| "learning_rate": 0.000176077575985678, | |
| "loss": 1.2745, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.23181454836131096, | |
| "grad_norm": 0.6833542585372925, | |
| "learning_rate": 0.00017591303422652803, | |
| "loss": 1.3063, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.23261390887290168, | |
| "grad_norm": 0.7865142226219177, | |
| "learning_rate": 0.00017574800597918908, | |
| "loss": 1.2927, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.23341326938449242, | |
| "grad_norm": 0.8319119215011597, | |
| "learning_rate": 0.0001755824923012436, | |
| "loss": 1.0652, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.23421262989608313, | |
| "grad_norm": 0.801260769367218, | |
| "learning_rate": 0.0001754164942533849, | |
| "loss": 1.3349, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.23501199040767387, | |
| "grad_norm": 0.9116886854171753, | |
| "learning_rate": 0.00017525001289941044, | |
| "loss": 1.373, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.23581135091926458, | |
| "grad_norm": 0.9009039998054504, | |
| "learning_rate": 0.00017508304930621488, | |
| "loss": 1.1461, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.23661071143085532, | |
| "grad_norm": 0.9449632167816162, | |
| "learning_rate": 0.00017491560454378332, | |
| "loss": 1.0613, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.23741007194244604, | |
| "grad_norm": 0.9248090982437134, | |
| "learning_rate": 0.00017474767968518446, | |
| "loss": 1.3297, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.23820943245403678, | |
| "grad_norm": 1.0803725719451904, | |
| "learning_rate": 0.00017457927580656366, | |
| "loss": 1.278, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.2390087929656275, | |
| "grad_norm": 1.1624330282211304, | |
| "learning_rate": 0.00017441039398713608, | |
| "loss": 1.3367, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.23980815347721823, | |
| "grad_norm": 1.5282617807388306, | |
| "learning_rate": 0.00017424103530917976, | |
| "loss": 1.433, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24060751398880895, | |
| "grad_norm": 0.3481718897819519, | |
| "learning_rate": 0.00017407120085802873, | |
| "loss": 1.0721, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.2414068745003997, | |
| "grad_norm": 0.3966723084449768, | |
| "learning_rate": 0.00017390089172206592, | |
| "loss": 1.0614, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.2422062350119904, | |
| "grad_norm": 0.3270397484302521, | |
| "learning_rate": 0.00017373010899271636, | |
| "loss": 1.1234, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.24300559552358114, | |
| "grad_norm": 0.3478359580039978, | |
| "learning_rate": 0.00017355885376444007, | |
| "loss": 1.0732, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.24380495603517185, | |
| "grad_norm": 0.38680601119995117, | |
| "learning_rate": 0.00017338712713472508, | |
| "loss": 1.0783, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.2446043165467626, | |
| "grad_norm": 0.35394465923309326, | |
| "learning_rate": 0.00017321493020408033, | |
| "loss": 1.2562, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.2454036770583533, | |
| "grad_norm": 0.36669933795928955, | |
| "learning_rate": 0.0001730422640760288, | |
| "loss": 1.2158, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.24620303756994405, | |
| "grad_norm": 0.37237656116485596, | |
| "learning_rate": 0.00017286912985710023, | |
| "loss": 1.1921, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.24700239808153476, | |
| "grad_norm": 0.35484257340431213, | |
| "learning_rate": 0.00017269552865682413, | |
| "loss": 1.1135, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.2478017585931255, | |
| "grad_norm": 0.3214118480682373, | |
| "learning_rate": 0.00017252146158772269, | |
| "loss": 1.1479, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.24860111910471622, | |
| "grad_norm": 0.3264860212802887, | |
| "learning_rate": 0.00017234692976530357, | |
| "loss": 1.2161, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.24940047961630696, | |
| "grad_norm": 0.33951637148857117, | |
| "learning_rate": 0.00017217193430805287, | |
| "loss": 1.1978, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.2501998401278977, | |
| "grad_norm": 0.30299174785614014, | |
| "learning_rate": 0.00017199647633742783, | |
| "loss": 1.1297, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2501998401278977, | |
| "eval_loss": 1.1418739557266235, | |
| "eval_runtime": 39.9979, | |
| "eval_samples_per_second": 13.176, | |
| "eval_steps_per_second": 6.6, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2509992006394884, | |
| "grad_norm": 0.3521738052368164, | |
| "learning_rate": 0.0001718205569778497, | |
| "loss": 1.189, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.2517985611510791, | |
| "grad_norm": 0.40434741973876953, | |
| "learning_rate": 0.0001716441773566966, | |
| "loss": 1.1654, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.25259792166266987, | |
| "grad_norm": 0.3887425661087036, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 1.1563, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2533972821742606, | |
| "grad_norm": 0.36217692494392395, | |
| "learning_rate": 0.0001712900418539183, | |
| "loss": 1.1309, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2541966426858513, | |
| "grad_norm": 0.3586144745349884, | |
| "learning_rate": 0.00017111228824176825, | |
| "loss": 1.1123, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.25499600319744203, | |
| "grad_norm": 0.3718055784702301, | |
| "learning_rate": 0.0001709340789069788, | |
| "loss": 1.1042, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.2557953637090328, | |
| "grad_norm": 0.4410824179649353, | |
| "learning_rate": 0.00017075541499160334, | |
| "loss": 1.1729, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2565947242206235, | |
| "grad_norm": 0.42812585830688477, | |
| "learning_rate": 0.0001705762976406084, | |
| "loss": 1.1608, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.2573940847322142, | |
| "grad_norm": 0.46935486793518066, | |
| "learning_rate": 0.00017039672800186638, | |
| "loss": 1.0939, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.25819344524380494, | |
| "grad_norm": 0.4472469091415405, | |
| "learning_rate": 0.00017021670722614817, | |
| "loss": 1.0144, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.2589928057553957, | |
| "grad_norm": 0.4422782361507416, | |
| "learning_rate": 0.0001700362364671157, | |
| "loss": 1.2359, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.2597921662669864, | |
| "grad_norm": 0.4598246216773987, | |
| "learning_rate": 0.0001698553168813147, | |
| "loss": 1.3228, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.26059152677857716, | |
| "grad_norm": 0.5052394866943359, | |
| "learning_rate": 0.00016967394962816713, | |
| "loss": 1.2048, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.26139088729016785, | |
| "grad_norm": 0.43572553992271423, | |
| "learning_rate": 0.0001694921358699639, | |
| "loss": 0.8828, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.2621902478017586, | |
| "grad_norm": 0.558404803276062, | |
| "learning_rate": 0.00016930987677185727, | |
| "loss": 1.2108, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.26298960831334933, | |
| "grad_norm": 0.4867321848869324, | |
| "learning_rate": 0.00016912717350185355, | |
| "loss": 1.1716, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.2637889688249401, | |
| "grad_norm": 0.5654073357582092, | |
| "learning_rate": 0.00016894402723080537, | |
| "loss": 1.2348, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.26458832933653076, | |
| "grad_norm": 0.524683952331543, | |
| "learning_rate": 0.00016876043913240452, | |
| "loss": 1.1331, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.2653876898481215, | |
| "grad_norm": 0.5388863682746887, | |
| "learning_rate": 0.00016857641038317404, | |
| "loss": 1.1401, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.26618705035971224, | |
| "grad_norm": 0.52965247631073, | |
| "learning_rate": 0.00016839194216246108, | |
| "loss": 1.0995, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.266986410871303, | |
| "grad_norm": 0.5100024342536926, | |
| "learning_rate": 0.00016820703565242898, | |
| "loss": 1.0813, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.26778577138289367, | |
| "grad_norm": 0.6176280379295349, | |
| "learning_rate": 0.0001680216920380499, | |
| "loss": 0.8405, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.2685851318944844, | |
| "grad_norm": 0.661065399646759, | |
| "learning_rate": 0.00016783591250709728, | |
| "loss": 1.0367, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.26938449240607515, | |
| "grad_norm": 0.6393107175827026, | |
| "learning_rate": 0.00016764969825013793, | |
| "loss": 0.8476, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.2701838529176659, | |
| "grad_norm": 0.8756284713745117, | |
| "learning_rate": 0.0001674630504605248, | |
| "loss": 1.1601, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.2709832134292566, | |
| "grad_norm": 0.6742208003997803, | |
| "learning_rate": 0.00016727597033438902, | |
| "loss": 1.0125, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.2717825739408473, | |
| "grad_norm": 0.8255277276039124, | |
| "learning_rate": 0.00016708845907063236, | |
| "loss": 1.1828, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.27258193445243806, | |
| "grad_norm": 0.8534565567970276, | |
| "learning_rate": 0.0001669005178709195, | |
| "loss": 1.1659, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.2733812949640288, | |
| "grad_norm": 0.9478426575660706, | |
| "learning_rate": 0.00016671214793967048, | |
| "loss": 1.133, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.2741806554756195, | |
| "grad_norm": 1.1164189577102661, | |
| "learning_rate": 0.00016652335048405274, | |
| "loss": 1.125, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.2749800159872102, | |
| "grad_norm": 0.866569459438324, | |
| "learning_rate": 0.00016633412671397357, | |
| "loss": 0.9521, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.27577937649880097, | |
| "grad_norm": 0.9649004340171814, | |
| "learning_rate": 0.00016614447784207227, | |
| "loss": 1.264, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.2765787370103917, | |
| "grad_norm": 1.1662349700927734, | |
| "learning_rate": 0.00016595440508371237, | |
| "loss": 1.4375, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.2773780975219824, | |
| "grad_norm": 1.5530859231948853, | |
| "learning_rate": 0.00016576390965697405, | |
| "loss": 0.9643, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.27817745803357313, | |
| "grad_norm": 1.300599455833435, | |
| "learning_rate": 0.00016557299278264584, | |
| "loss": 1.6229, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.2789768185451639, | |
| "grad_norm": 1.559091567993164, | |
| "learning_rate": 0.0001653816556842174, | |
| "loss": 1.3086, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.2797761790567546, | |
| "grad_norm": 2.289126396179199, | |
| "learning_rate": 0.00016518989958787126, | |
| "loss": 0.8696, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2805755395683453, | |
| "grad_norm": 0.3264933228492737, | |
| "learning_rate": 0.00016499772572247515, | |
| "loss": 0.8783, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.28137490007993604, | |
| "grad_norm": 0.37487271428108215, | |
| "learning_rate": 0.000164805135319574, | |
| "loss": 1.0175, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.2821742605915268, | |
| "grad_norm": 0.43567776679992676, | |
| "learning_rate": 0.0001646121296133822, | |
| "loss": 1.1585, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.2829736211031175, | |
| "grad_norm": 0.3546614646911621, | |
| "learning_rate": 0.00016441870984077554, | |
| "loss": 1.095, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.2837729816147082, | |
| "grad_norm": 0.39957696199417114, | |
| "learning_rate": 0.00016422487724128342, | |
| "loss": 1.0681, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.28457234212629895, | |
| "grad_norm": 0.6007651686668396, | |
| "learning_rate": 0.00016403063305708078, | |
| "loss": 1.023, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.2853717026378897, | |
| "grad_norm": 0.4657421112060547, | |
| "learning_rate": 0.0001638359785329802, | |
| "loss": 1.1031, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.28617106314948043, | |
| "grad_norm": 0.3539028465747833, | |
| "learning_rate": 0.00016364091491642398, | |
| "loss": 1.2028, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.2869704236610711, | |
| "grad_norm": 0.3230942189693451, | |
| "learning_rate": 0.000163445443457476, | |
| "loss": 1.07, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.28776978417266186, | |
| "grad_norm": 0.3208749294281006, | |
| "learning_rate": 0.00016324956540881384, | |
| "loss": 1.0636, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2885691446842526, | |
| "grad_norm": 0.2991190254688263, | |
| "learning_rate": 0.00016305328202572076, | |
| "loss": 1.0855, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.28936850519584334, | |
| "grad_norm": 0.3368874490261078, | |
| "learning_rate": 0.0001628565945660775, | |
| "loss": 1.0894, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.290167865707434, | |
| "grad_norm": 0.34719499945640564, | |
| "learning_rate": 0.00016265950429035443, | |
| "loss": 1.0933, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.29096722621902477, | |
| "grad_norm": 0.34301942586898804, | |
| "learning_rate": 0.00016246201246160327, | |
| "loss": 1.1828, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.2917665867306155, | |
| "grad_norm": 0.46968960762023926, | |
| "learning_rate": 0.00016226412034544912, | |
| "loss": 1.1122, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.29256594724220625, | |
| "grad_norm": 0.3344099521636963, | |
| "learning_rate": 0.00016206582921008233, | |
| "loss": 1.0661, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.293365307753797, | |
| "grad_norm": 0.3538711965084076, | |
| "learning_rate": 0.00016186714032625035, | |
| "loss": 1.0411, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.2941646682653877, | |
| "grad_norm": 0.4503370523452759, | |
| "learning_rate": 0.00016166805496724965, | |
| "loss": 1.1247, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.2949640287769784, | |
| "grad_norm": 0.43027833104133606, | |
| "learning_rate": 0.00016146857440891744, | |
| "loss": 1.1507, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.29576338928856916, | |
| "grad_norm": 0.5363488793373108, | |
| "learning_rate": 0.00016126869992962357, | |
| "loss": 1.0835, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2965627498001599, | |
| "grad_norm": 0.3750304877758026, | |
| "learning_rate": 0.00016106843281026237, | |
| "loss": 1.2127, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.2973621103117506, | |
| "grad_norm": 0.4115138053894043, | |
| "learning_rate": 0.00016086777433424435, | |
| "loss": 1.081, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.2981614708233413, | |
| "grad_norm": 0.4057779610157013, | |
| "learning_rate": 0.00016066672578748814, | |
| "loss": 1.0342, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.29896083133493206, | |
| "grad_norm": 0.3976861238479614, | |
| "learning_rate": 0.00016046528845841196, | |
| "loss": 0.9994, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.2997601918465228, | |
| "grad_norm": 0.4205765426158905, | |
| "learning_rate": 0.00016026346363792567, | |
| "loss": 0.885, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3005595523581135, | |
| "grad_norm": 0.44076278805732727, | |
| "learning_rate": 0.00016006125261942229, | |
| "loss": 1.0035, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.30135891286970423, | |
| "grad_norm": 0.44243985414505005, | |
| "learning_rate": 0.00015985865669876988, | |
| "loss": 0.9194, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.302158273381295, | |
| "grad_norm": 0.5455163717269897, | |
| "learning_rate": 0.00015965567717430304, | |
| "loss": 1.0047, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.3029576338928857, | |
| "grad_norm": 0.47469913959503174, | |
| "learning_rate": 0.00015945231534681483, | |
| "loss": 1.0338, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.3037569944044764, | |
| "grad_norm": 0.5423705577850342, | |
| "learning_rate": 0.00015924857251954806, | |
| "loss": 1.059, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.30455635491606714, | |
| "grad_norm": 0.5439475178718567, | |
| "learning_rate": 0.00015904444999818745, | |
| "loss": 1.0005, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3053557154276579, | |
| "grad_norm": 0.6124017834663391, | |
| "learning_rate": 0.0001588399490908508, | |
| "loss": 0.8819, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.3061550759392486, | |
| "grad_norm": 0.7043859958648682, | |
| "learning_rate": 0.00015863507110808086, | |
| "loss": 1.4554, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3069544364508393, | |
| "grad_norm": 0.5058495998382568, | |
| "learning_rate": 0.00015842981736283686, | |
| "loss": 0.8867, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.30775379696243005, | |
| "grad_norm": 0.6541773080825806, | |
| "learning_rate": 0.0001582241891704861, | |
| "loss": 1.2015, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3085531574740208, | |
| "grad_norm": 0.6508180499076843, | |
| "learning_rate": 0.0001580181878487955, | |
| "loss": 1.0498, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.30935251798561153, | |
| "grad_norm": 0.6798310875892639, | |
| "learning_rate": 0.00015781181471792322, | |
| "loss": 1.1951, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.3101518784972022, | |
| "grad_norm": 0.6557720303535461, | |
| "learning_rate": 0.00015760507110041015, | |
| "loss": 1.0498, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.31095123900879296, | |
| "grad_norm": 0.8685877323150635, | |
| "learning_rate": 0.00015739795832117143, | |
| "loss": 1.1951, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3117505995203837, | |
| "grad_norm": 0.8445819020271301, | |
| "learning_rate": 0.00015719047770748788, | |
| "loss": 0.8783, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.31254996003197444, | |
| "grad_norm": 1.0820878744125366, | |
| "learning_rate": 0.00015698263058899778, | |
| "loss": 1.1618, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.3133493205435651, | |
| "grad_norm": 0.8614665269851685, | |
| "learning_rate": 0.000156774418297688, | |
| "loss": 1.2297, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.31414868105515587, | |
| "grad_norm": 0.8265265822410583, | |
| "learning_rate": 0.00015656584216788567, | |
| "loss": 1.2369, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.3149480415667466, | |
| "grad_norm": 1.281562328338623, | |
| "learning_rate": 0.00015635690353624955, | |
| "loss": 1.1882, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.31574740207833735, | |
| "grad_norm": 1.058127999305725, | |
| "learning_rate": 0.0001561476037417615, | |
| "loss": 1.3153, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.31654676258992803, | |
| "grad_norm": 1.0368887186050415, | |
| "learning_rate": 0.00015593794412571796, | |
| "loss": 1.3065, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.3173461231015188, | |
| "grad_norm": 1.2579238414764404, | |
| "learning_rate": 0.00015572792603172115, | |
| "loss": 1.1121, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.3181454836131095, | |
| "grad_norm": 1.089697003364563, | |
| "learning_rate": 0.00015551755080567075, | |
| "loss": 1.2048, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.31894484412470026, | |
| "grad_norm": 1.2540833950042725, | |
| "learning_rate": 0.00015530681979575496, | |
| "loss": 1.2861, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.31974420463629094, | |
| "grad_norm": 1.39395010471344, | |
| "learning_rate": 0.00015509573435244214, | |
| "loss": 1.1636, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3205435651478817, | |
| "grad_norm": 0.3763245940208435, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.8533, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.3213429256594724, | |
| "grad_norm": 0.4181549847126007, | |
| "learning_rate": 0.00015467250557884684, | |
| "loss": 1.0928, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.32214228617106316, | |
| "grad_norm": 0.38936108350753784, | |
| "learning_rate": 0.0001544603649608232, | |
| "loss": 1.1645, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.3229416466826539, | |
| "grad_norm": 0.4016236364841461, | |
| "learning_rate": 0.0001542478753339029, | |
| "loss": 1.0839, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.3237410071942446, | |
| "grad_norm": 0.43020734190940857, | |
| "learning_rate": 0.0001540350380598243, | |
| "loss": 0.968, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.32454036770583533, | |
| "grad_norm": 0.38603150844573975, | |
| "learning_rate": 0.00015382185450255365, | |
| "loss": 1.1274, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.32533972821742607, | |
| "grad_norm": 0.44878101348876953, | |
| "learning_rate": 0.00015360832602827644, | |
| "loss": 1.1192, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.3261390887290168, | |
| "grad_norm": 0.35593459010124207, | |
| "learning_rate": 0.00015339445400538852, | |
| "loss": 1.0117, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.3269384492406075, | |
| "grad_norm": 0.3497203588485718, | |
| "learning_rate": 0.00015318023980448735, | |
| "loss": 1.2231, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.32773780975219824, | |
| "grad_norm": 0.40464532375335693, | |
| "learning_rate": 0.00015296568479836325, | |
| "loss": 1.2276, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.328537170263789, | |
| "grad_norm": 0.3322058320045471, | |
| "learning_rate": 0.00015275079036199062, | |
| "loss": 1.1565, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.3293365307753797, | |
| "grad_norm": 0.30930295586586, | |
| "learning_rate": 0.000152535557872519, | |
| "loss": 1.0212, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.3301358912869704, | |
| "grad_norm": 0.3241097927093506, | |
| "learning_rate": 0.00015231998870926446, | |
| "loss": 1.1983, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.33093525179856115, | |
| "grad_norm": 0.31953859329223633, | |
| "learning_rate": 0.0001521040842537006, | |
| "loss": 1.1207, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.3317346123101519, | |
| "grad_norm": 0.35638073086738586, | |
| "learning_rate": 0.0001518878458894497, | |
| "loss": 1.1693, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.33253397282174263, | |
| "grad_norm": 0.48129066824913025, | |
| "learning_rate": 0.00015167127500227397, | |
| "loss": 1.0441, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.3575907051563263, | |
| "learning_rate": 0.0001514543729800664, | |
| "loss": 0.885, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.33413269384492406, | |
| "grad_norm": 0.35546669363975525, | |
| "learning_rate": 0.0001512371412128424, | |
| "loss": 0.9673, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.3349320543565148, | |
| "grad_norm": 0.36895594000816345, | |
| "learning_rate": 0.00015101958109273015, | |
| "loss": 1.0111, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.33573141486810554, | |
| "grad_norm": 0.39300018548965454, | |
| "learning_rate": 0.0001508016940139624, | |
| "loss": 1.1185, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3365307753796962, | |
| "grad_norm": 0.45900556445121765, | |
| "learning_rate": 0.00015058348137286703, | |
| "loss": 1.3187, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.33733013589128696, | |
| "grad_norm": 0.3960709273815155, | |
| "learning_rate": 0.00015036494456785836, | |
| "loss": 1.1019, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.3381294964028777, | |
| "grad_norm": 0.43770337104797363, | |
| "learning_rate": 0.00015014608499942808, | |
| "loss": 1.0425, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.33892885691446845, | |
| "grad_norm": 0.39608171582221985, | |
| "learning_rate": 0.00014992690407013634, | |
| "loss": 0.9861, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.33972821742605913, | |
| "grad_norm": 0.37925106287002563, | |
| "learning_rate": 0.0001497074031846027, | |
| "loss": 1.0193, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.3405275779376499, | |
| "grad_norm": 0.40427684783935547, | |
| "learning_rate": 0.00014948758374949713, | |
| "loss": 1.1671, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.3413269384492406, | |
| "grad_norm": 0.5301655530929565, | |
| "learning_rate": 0.0001492674471735311, | |
| "loss": 0.9758, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.34212629896083135, | |
| "grad_norm": 0.5658469796180725, | |
| "learning_rate": 0.00014904699486744847, | |
| "loss": 1.2819, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.34292565947242204, | |
| "grad_norm": 0.5009921789169312, | |
| "learning_rate": 0.0001488262282440163, | |
| "loss": 1.0126, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.3437250199840128, | |
| "grad_norm": 0.49652495980262756, | |
| "learning_rate": 0.00014860514871801618, | |
| "loss": 1.1613, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3445243804956035, | |
| "grad_norm": 0.6930481195449829, | |
| "learning_rate": 0.0001483837577062348, | |
| "loss": 0.9825, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.34532374100719426, | |
| "grad_norm": 0.4906662106513977, | |
| "learning_rate": 0.00014816205662745497, | |
| "loss": 1.1218, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.34612310151878495, | |
| "grad_norm": 0.5959265828132629, | |
| "learning_rate": 0.00014794004690244669, | |
| "loss": 1.1655, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.3469224620303757, | |
| "grad_norm": 0.6609960794448853, | |
| "learning_rate": 0.00014771772995395777, | |
| "loss": 1.1025, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.34772182254196643, | |
| "grad_norm": 0.6050605177879333, | |
| "learning_rate": 0.00014749510720670506, | |
| "loss": 0.9879, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.34852118305355717, | |
| "grad_norm": 0.7246385812759399, | |
| "learning_rate": 0.00014727218008736486, | |
| "loss": 1.1155, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.34932054356514786, | |
| "grad_norm": 0.6366634368896484, | |
| "learning_rate": 0.00014704895002456424, | |
| "loss": 1.0055, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.3501199040767386, | |
| "grad_norm": 0.7289115786552429, | |
| "learning_rate": 0.00014682541844887164, | |
| "loss": 1.2236, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.35091926458832934, | |
| "grad_norm": 0.7496143579483032, | |
| "learning_rate": 0.00014660158679278765, | |
| "loss": 1.1094, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.3517186250999201, | |
| "grad_norm": 0.7350897789001465, | |
| "learning_rate": 0.00014637745649073602, | |
| "loss": 1.2462, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.35251798561151076, | |
| "grad_norm": 0.7020392417907715, | |
| "learning_rate": 0.00014615302897905434, | |
| "loss": 1.1669, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.3533173461231015, | |
| "grad_norm": 0.7738919854164124, | |
| "learning_rate": 0.0001459283056959849, | |
| "loss": 0.989, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.35411670663469225, | |
| "grad_norm": 0.8123486638069153, | |
| "learning_rate": 0.00014570328808166539, | |
| "loss": 1.0462, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.354916067146283, | |
| "grad_norm": 0.8612303137779236, | |
| "learning_rate": 0.0001454779775781197, | |
| "loss": 1.3888, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.35571542765787373, | |
| "grad_norm": 0.8494888544082642, | |
| "learning_rate": 0.00014525237562924877, | |
| "loss": 1.2268, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.3565147881694644, | |
| "grad_norm": 1.2113531827926636, | |
| "learning_rate": 0.00014502648368082123, | |
| "loss": 1.3974, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.35731414868105515, | |
| "grad_norm": 1.109397053718567, | |
| "learning_rate": 0.00014480030318046416, | |
| "loss": 1.3267, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.3581135091926459, | |
| "grad_norm": 1.2456046342849731, | |
| "learning_rate": 0.00014457383557765386, | |
| "loss": 1.4305, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.35891286970423664, | |
| "grad_norm": 1.5046449899673462, | |
| "learning_rate": 0.0001443470823237064, | |
| "loss": 0.9759, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.3597122302158273, | |
| "grad_norm": 1.4025421142578125, | |
| "learning_rate": 0.0001441200448717687, | |
| "loss": 0.8929, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.36051159072741806, | |
| "grad_norm": 0.30919286608695984, | |
| "learning_rate": 0.0001438927246768087, | |
| "loss": 0.8245, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.3613109512390088, | |
| "grad_norm": 0.3811976909637451, | |
| "learning_rate": 0.0001436651231956064, | |
| "loss": 1.1732, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.36211031175059955, | |
| "grad_norm": 0.3917795717716217, | |
| "learning_rate": 0.00014343724188674447, | |
| "loss": 1.1951, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.36290967226219023, | |
| "grad_norm": 0.4045757055282593, | |
| "learning_rate": 0.00014320908221059875, | |
| "loss": 1.0743, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.36370903277378097, | |
| "grad_norm": 0.3281409442424774, | |
| "learning_rate": 0.00014298064562932904, | |
| "loss": 1.0291, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.3645083932853717, | |
| "grad_norm": 0.46397122740745544, | |
| "learning_rate": 0.0001427519336068698, | |
| "loss": 1.1945, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.36530775379696245, | |
| "grad_norm": 0.35933932662010193, | |
| "learning_rate": 0.0001425229476089204, | |
| "loss": 1.1349, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.36610711430855314, | |
| "grad_norm": 0.42518380284309387, | |
| "learning_rate": 0.00014229368910293627, | |
| "loss": 1.0901, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.3669064748201439, | |
| "grad_norm": 0.3412696123123169, | |
| "learning_rate": 0.00014206415955811898, | |
| "loss": 1.0984, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.3677058353317346, | |
| "grad_norm": 0.32191070914268494, | |
| "learning_rate": 0.00014183436044540713, | |
| "loss": 1.0522, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.36850519584332536, | |
| "grad_norm": 0.34181419014930725, | |
| "learning_rate": 0.00014160429323746687, | |
| "loss": 1.0657, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.36930455635491605, | |
| "grad_norm": 0.3284541368484497, | |
| "learning_rate": 0.00014137395940868244, | |
| "loss": 1.0296, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.3701039168665068, | |
| "grad_norm": 0.31918108463287354, | |
| "learning_rate": 0.00014114336043514661, | |
| "loss": 1.1504, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.37090327737809753, | |
| "grad_norm": 0.2991093397140503, | |
| "learning_rate": 0.0001409124977946515, | |
| "loss": 1.2299, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.37170263788968827, | |
| "grad_norm": 0.3887624442577362, | |
| "learning_rate": 0.00014068137296667884, | |
| "loss": 1.3608, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.37250199840127896, | |
| "grad_norm": 0.42528679966926575, | |
| "learning_rate": 0.0001404499874323905, | |
| "loss": 1.0458, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.3733013589128697, | |
| "grad_norm": 0.32530438899993896, | |
| "learning_rate": 0.00014021834267461932, | |
| "loss": 1.2109, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.37410071942446044, | |
| "grad_norm": 0.3990713953971863, | |
| "learning_rate": 0.00013998644017785913, | |
| "loss": 1.0526, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.3749000799360512, | |
| "grad_norm": 0.3738067150115967, | |
| "learning_rate": 0.0001397542814282556, | |
| "loss": 1.2217, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.37569944044764186, | |
| "grad_norm": 0.37012550234794617, | |
| "learning_rate": 0.0001395218679135966, | |
| "loss": 0.9932, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.3764988009592326, | |
| "grad_norm": 0.3903065323829651, | |
| "learning_rate": 0.00013928920112330262, | |
| "loss": 1.122, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.37729816147082335, | |
| "grad_norm": 0.4195614457130432, | |
| "learning_rate": 0.00013905628254841729, | |
| "loss": 1.193, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.3780975219824141, | |
| "grad_norm": 0.35402435064315796, | |
| "learning_rate": 0.00013882311368159782, | |
| "loss": 0.9951, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.37889688249400477, | |
| "grad_norm": 0.36403220891952515, | |
| "learning_rate": 0.00013858969601710533, | |
| "loss": 0.9178, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.3796962430055955, | |
| "grad_norm": 0.40461957454681396, | |
| "learning_rate": 0.00013835603105079539, | |
| "loss": 0.9439, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.38049560351718625, | |
| "grad_norm": 0.42805224657058716, | |
| "learning_rate": 0.00013812212028010854, | |
| "loss": 1.0269, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.381294964028777, | |
| "grad_norm": 0.4904803931713104, | |
| "learning_rate": 0.00013788796520406034, | |
| "loss": 1.0917, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.3820943245403677, | |
| "grad_norm": 0.4846837818622589, | |
| "learning_rate": 0.0001376535673232321, | |
| "loss": 1.0697, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.3828936850519584, | |
| "grad_norm": 0.46769338846206665, | |
| "learning_rate": 0.00013741892813976116, | |
| "loss": 1.142, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.38369304556354916, | |
| "grad_norm": 0.5554665327072144, | |
| "learning_rate": 0.0001371840491573312, | |
| "loss": 1.184, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.3844924060751399, | |
| "grad_norm": 0.46426481008529663, | |
| "learning_rate": 0.0001369489318811626, | |
| "loss": 1.1334, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.3852917665867306, | |
| "grad_norm": 0.5384045243263245, | |
| "learning_rate": 0.000136713577818003, | |
| "loss": 1.0772, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.38609112709832133, | |
| "grad_norm": 0.5401681065559387, | |
| "learning_rate": 0.0001364779884761173, | |
| "loss": 1.1188, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.38689048760991207, | |
| "grad_norm": 0.5268040299415588, | |
| "learning_rate": 0.0001362421653652784, | |
| "loss": 1.0124, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.3876898481215028, | |
| "grad_norm": 0.6549224853515625, | |
| "learning_rate": 0.00013600610999675702, | |
| "loss": 1.1054, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.38848920863309355, | |
| "grad_norm": 0.5715364217758179, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.9463, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.38928856914468424, | |
| "grad_norm": 0.5503459572792053, | |
| "learning_rate": 0.00013553330853918304, | |
| "loss": 1.1581, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.390087929656275, | |
| "grad_norm": 0.6885532736778259, | |
| "learning_rate": 0.0001352965654800755, | |
| "loss": 1.0947, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.3908872901678657, | |
| "grad_norm": 0.606448233127594, | |
| "learning_rate": 0.0001350595962231563, | |
| "loss": 1.1791, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.39168665067945646, | |
| "grad_norm": 0.748115062713623, | |
| "learning_rate": 0.00013482240228704142, | |
| "loss": 1.191, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.39248601119104715, | |
| "grad_norm": 0.7352818250656128, | |
| "learning_rate": 0.00013458498519178664, | |
| "loss": 1.1051, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.3932853717026379, | |
| "grad_norm": 0.8258904814720154, | |
| "learning_rate": 0.00013434734645887788, | |
| "loss": 0.8982, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.39408473221422863, | |
| "grad_norm": 0.8865075707435608, | |
| "learning_rate": 0.00013410948761122143, | |
| "loss": 1.0176, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.39488409272581937, | |
| "grad_norm": 0.8664267063140869, | |
| "learning_rate": 0.00013387141017313418, | |
| "loss": 1.1245, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.39568345323741005, | |
| "grad_norm": 0.8627355098724365, | |
| "learning_rate": 0.00013363311567033382, | |
| "loss": 1.1, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.3964828137490008, | |
| "grad_norm": 0.766534149646759, | |
| "learning_rate": 0.0001333946056299292, | |
| "loss": 1.0663, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.39728217426059154, | |
| "grad_norm": 0.9121149182319641, | |
| "learning_rate": 0.00013315588158041032, | |
| "loss": 1.1432, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.3980815347721823, | |
| "grad_norm": 1.1074888706207275, | |
| "learning_rate": 0.0001329169450516387, | |
| "loss": 0.9393, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.39888089528377296, | |
| "grad_norm": 1.1374878883361816, | |
| "learning_rate": 0.0001326777975748376, | |
| "loss": 1.1407, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.3996802557953637, | |
| "grad_norm": 1.6384693384170532, | |
| "learning_rate": 0.000132438440682582, | |
| "loss": 1.2904, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.40047961630695444, | |
| "grad_norm": 0.3537842929363251, | |
| "learning_rate": 0.00013219887590878903, | |
| "loss": 1.0584, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.4012789768185452, | |
| "grad_norm": 0.3787943422794342, | |
| "learning_rate": 0.00013195910478870803, | |
| "loss": 1.0717, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.40207833733013587, | |
| "grad_norm": 0.40689560770988464, | |
| "learning_rate": 0.00013171912885891063, | |
| "loss": 1.1355, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.4028776978417266, | |
| "grad_norm": 0.40294623374938965, | |
| "learning_rate": 0.000131478949657281, | |
| "loss": 1.2585, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.40367705835331735, | |
| "grad_norm": 0.33191537857055664, | |
| "learning_rate": 0.00013123856872300607, | |
| "loss": 0.8516, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.4044764188649081, | |
| "grad_norm": 0.3889703154563904, | |
| "learning_rate": 0.00013099798759656545, | |
| "loss": 1.1512, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.4052757793764988, | |
| "grad_norm": 0.36418843269348145, | |
| "learning_rate": 0.00013075720781972175, | |
| "loss": 1.2602, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.4060751398880895, | |
| "grad_norm": 0.3365548551082611, | |
| "learning_rate": 0.00013051623093551063, | |
| "loss": 1.1464, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.40687450039968026, | |
| "grad_norm": 0.38964754343032837, | |
| "learning_rate": 0.00013027505848823084, | |
| "loss": 1.09, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.407673860911271, | |
| "grad_norm": 0.43493184447288513, | |
| "learning_rate": 0.00013003369202343453, | |
| "loss": 1.1176, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4084732214228617, | |
| "grad_norm": 0.3899182677268982, | |
| "learning_rate": 0.0001297921330879171, | |
| "loss": 1.2212, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.40927258193445243, | |
| "grad_norm": 0.35002589225769043, | |
| "learning_rate": 0.00012955038322970736, | |
| "loss": 1.1259, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.41007194244604317, | |
| "grad_norm": 0.3097720146179199, | |
| "learning_rate": 0.0001293084439980578, | |
| "loss": 1.1399, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.4108713029576339, | |
| "grad_norm": 0.34566211700439453, | |
| "learning_rate": 0.00012906631694343436, | |
| "loss": 1.0455, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.4116706634692246, | |
| "grad_norm": 0.3389396369457245, | |
| "learning_rate": 0.00012882400361750675, | |
| "loss": 1.2381, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.41247002398081534, | |
| "grad_norm": 0.3703395128250122, | |
| "learning_rate": 0.0001285815055731383, | |
| "loss": 1.1114, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.4132693844924061, | |
| "grad_norm": 0.36339160799980164, | |
| "learning_rate": 0.00012833882436437624, | |
| "loss": 1.0136, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.4140687450039968, | |
| "grad_norm": 0.35047227144241333, | |
| "learning_rate": 0.00012809596154644148, | |
| "loss": 1.0354, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.4148681055155875, | |
| "grad_norm": 0.3567598760128021, | |
| "learning_rate": 0.00012785291867571883, | |
| "loss": 1.1561, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.41566746602717825, | |
| "grad_norm": 0.3623413145542145, | |
| "learning_rate": 0.00012760969730974694, | |
| "loss": 1.0828, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.416466826538769, | |
| "grad_norm": 0.41852647066116333, | |
| "learning_rate": 0.0001273662990072083, | |
| "loss": 1.1553, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.4172661870503597, | |
| "grad_norm": 0.41852867603302, | |
| "learning_rate": 0.00012712272532791946, | |
| "loss": 1.145, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.4180655475619504, | |
| "grad_norm": 0.4308965802192688, | |
| "learning_rate": 0.0001268789778328207, | |
| "loss": 1.0496, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.41886490807354115, | |
| "grad_norm": 0.4174908399581909, | |
| "learning_rate": 0.0001266350580839664, | |
| "loss": 0.9394, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.4196642685851319, | |
| "grad_norm": 0.4295555353164673, | |
| "learning_rate": 0.0001263909676445145, | |
| "loss": 1.1062, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.42046362909672264, | |
| "grad_norm": 0.43274104595184326, | |
| "learning_rate": 0.0001261467080787171, | |
| "loss": 1.0089, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.4212629896083134, | |
| "grad_norm": 0.4415391683578491, | |
| "learning_rate": 0.0001259022809519101, | |
| "loss": 0.9486, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.42206235011990406, | |
| "grad_norm": 0.4199742376804352, | |
| "learning_rate": 0.00012565768783050305, | |
| "loss": 1.1883, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.4228617106314948, | |
| "grad_norm": 0.454895943403244, | |
| "learning_rate": 0.00012541293028196944, | |
| "loss": 1.1424, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.42366107114308554, | |
| "grad_norm": 0.4642922878265381, | |
| "learning_rate": 0.0001251680098748365, | |
| "loss": 1.1524, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4244604316546763, | |
| "grad_norm": 0.499031126499176, | |
| "learning_rate": 0.00012492292817867497, | |
| "loss": 0.888, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.42525979216626697, | |
| "grad_norm": 0.5840815305709839, | |
| "learning_rate": 0.00012467768676408935, | |
| "loss": 1.0772, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.4260591526778577, | |
| "grad_norm": 0.5054577589035034, | |
| "learning_rate": 0.00012443228720270767, | |
| "loss": 0.9563, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.42685851318944845, | |
| "grad_norm": 0.5701196193695068, | |
| "learning_rate": 0.00012418673106717143, | |
| "loss": 1.2029, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.4276578737010392, | |
| "grad_norm": 0.5654375553131104, | |
| "learning_rate": 0.00012394101993112557, | |
| "loss": 1.2564, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4284572342126299, | |
| "grad_norm": 0.7642629742622375, | |
| "learning_rate": 0.00012369515536920825, | |
| "loss": 1.0069, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.4292565947242206, | |
| "grad_norm": 0.583177387714386, | |
| "learning_rate": 0.00012344913895704097, | |
| "loss": 0.893, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.43005595523581136, | |
| "grad_norm": 0.5912151336669922, | |
| "learning_rate": 0.00012320297227121833, | |
| "loss": 1.1477, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.4308553157474021, | |
| "grad_norm": 0.6719471216201782, | |
| "learning_rate": 0.0001229566568892979, | |
| "loss": 1.1085, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.4316546762589928, | |
| "grad_norm": 0.8098368644714355, | |
| "learning_rate": 0.00012271019438979023, | |
| "loss": 1.181, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4324540367705835, | |
| "grad_norm": 0.7867881655693054, | |
| "learning_rate": 0.00012246358635214868, | |
| "loss": 1.0798, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.43325339728217427, | |
| "grad_norm": 0.6670185327529907, | |
| "learning_rate": 0.0001222168343567592, | |
| "loss": 1.0568, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.434052757793765, | |
| "grad_norm": 0.7349308729171753, | |
| "learning_rate": 0.00012196993998493043, | |
| "loss": 1.1429, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.4348521183053557, | |
| "grad_norm": 0.7176403999328613, | |
| "learning_rate": 0.00012172290481888331, | |
| "loss": 1.0956, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.43565147881694644, | |
| "grad_norm": 0.8897795677185059, | |
| "learning_rate": 0.00012147573044174112, | |
| "loss": 1.2973, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.4364508393285372, | |
| "grad_norm": 0.7887820601463318, | |
| "learning_rate": 0.00012122841843751924, | |
| "loss": 1.1597, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.4372501998401279, | |
| "grad_norm": 0.8651790618896484, | |
| "learning_rate": 0.00012098097039111512, | |
| "loss": 1.0137, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.4380495603517186, | |
| "grad_norm": 1.0698950290679932, | |
| "learning_rate": 0.00012073338788829787, | |
| "loss": 1.0543, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.43884892086330934, | |
| "grad_norm": 1.1988959312438965, | |
| "learning_rate": 0.0001204856725156984, | |
| "loss": 1.1408, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.4396482813749001, | |
| "grad_norm": 1.5389310121536255, | |
| "learning_rate": 0.00012023782586079908, | |
| "loss": 1.1225, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4404476418864908, | |
| "grad_norm": 0.41004931926727295, | |
| "learning_rate": 0.00011998984951192361, | |
| "loss": 1.0031, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.4412470023980815, | |
| "grad_norm": 0.4041696786880493, | |
| "learning_rate": 0.00011974174505822677, | |
| "loss": 0.9878, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.44204636290967225, | |
| "grad_norm": 0.40297624468803406, | |
| "learning_rate": 0.00011949351408968439, | |
| "loss": 1.0132, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.442845723421263, | |
| "grad_norm": 0.3916489779949188, | |
| "learning_rate": 0.000119245158197083, | |
| "loss": 1.0056, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.44364508393285373, | |
| "grad_norm": 0.3816269636154175, | |
| "learning_rate": 0.00011899667897200977, | |
| "loss": 1.0741, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.36485084891319275, | |
| "learning_rate": 0.00011874807800684215, | |
| "loss": 1.0796, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.44524380495603516, | |
| "grad_norm": 0.36104485392570496, | |
| "learning_rate": 0.00011849935689473786, | |
| "loss": 1.2108, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.4460431654676259, | |
| "grad_norm": 0.6658501029014587, | |
| "learning_rate": 0.00011825051722962453, | |
| "loss": 1.2172, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.44684252597921664, | |
| "grad_norm": 0.39362889528274536, | |
| "learning_rate": 0.00011800156060618955, | |
| "loss": 1.1477, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.44764188649080733, | |
| "grad_norm": 0.361454039812088, | |
| "learning_rate": 0.00011775248861986978, | |
| "loss": 1.0736, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.44844124700239807, | |
| "grad_norm": 0.3387874960899353, | |
| "learning_rate": 0.00011750330286684153, | |
| "loss": 1.1703, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.4492406075139888, | |
| "grad_norm": 0.3527488112449646, | |
| "learning_rate": 0.00011725400494401, | |
| "loss": 1.1032, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.45003996802557955, | |
| "grad_norm": 0.35559266805648804, | |
| "learning_rate": 0.00011700459644899937, | |
| "loss": 1.2311, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.45083932853717024, | |
| "grad_norm": 0.36314061284065247, | |
| "learning_rate": 0.00011675507898014235, | |
| "loss": 1.1108, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.451638689048761, | |
| "grad_norm": 0.3469599187374115, | |
| "learning_rate": 0.00011650545413647, | |
| "loss": 1.1626, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.4524380495603517, | |
| "grad_norm": 0.3834267258644104, | |
| "learning_rate": 0.00011625572351770155, | |
| "loss": 1.1673, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.45323741007194246, | |
| "grad_norm": 0.3510596752166748, | |
| "learning_rate": 0.00011600588872423405, | |
| "loss": 1.1076, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.4540367705835332, | |
| "grad_norm": 0.3876965641975403, | |
| "learning_rate": 0.0001157559513571322, | |
| "loss": 1.1704, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.4548361310951239, | |
| "grad_norm": 0.3952408730983734, | |
| "learning_rate": 0.00011550591301811788, | |
| "loss": 1.0154, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.4556354916067146, | |
| "grad_norm": 0.38112473487854004, | |
| "learning_rate": 0.00011525577530956029, | |
| "loss": 1.0749, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.45643485211830537, | |
| "grad_norm": 0.3879120945930481, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 1.1225, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.4572342126298961, | |
| "grad_norm": 0.4533351957798004, | |
| "learning_rate": 0.00011475520819646526, | |
| "loss": 1.046, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.4580335731414868, | |
| "grad_norm": 0.5451374650001526, | |
| "learning_rate": 0.0001145047819998089, | |
| "loss": 0.9717, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.45883293365307753, | |
| "grad_norm": 0.5647904276847839, | |
| "learning_rate": 0.00011425426284935091, | |
| "loss": 1.075, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.4596322941646683, | |
| "grad_norm": 0.4953969717025757, | |
| "learning_rate": 0.0001140036523505416, | |
| "loss": 1.186, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.460431654676259, | |
| "grad_norm": 0.3667427897453308, | |
| "learning_rate": 0.00011375295210941674, | |
| "loss": 1.0549, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.4612310151878497, | |
| "grad_norm": 0.4673584997653961, | |
| "learning_rate": 0.00011350216373258722, | |
| "loss": 1.004, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.46203037569944044, | |
| "grad_norm": 0.6425716876983643, | |
| "learning_rate": 0.00011325128882722875, | |
| "loss": 1.2121, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.4628297362110312, | |
| "grad_norm": 0.4969245493412018, | |
| "learning_rate": 0.00011300032900107147, | |
| "loss": 0.9811, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.4636290967226219, | |
| "grad_norm": 0.4420992136001587, | |
| "learning_rate": 0.00011274928586238986, | |
| "loss": 1.0611, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4644284572342126, | |
| "grad_norm": 0.5042086243629456, | |
| "learning_rate": 0.00011249816101999227, | |
| "loss": 0.9749, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.46522781774580335, | |
| "grad_norm": 0.5414336919784546, | |
| "learning_rate": 0.00011224695608321056, | |
| "loss": 1.072, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.4660271782573941, | |
| "grad_norm": 0.6206878423690796, | |
| "learning_rate": 0.00011199567266188999, | |
| "loss": 1.3321, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.46682653876898483, | |
| "grad_norm": 0.6050058603286743, | |
| "learning_rate": 0.00011174431236637874, | |
| "loss": 1.0738, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.4676258992805755, | |
| "grad_norm": 0.6047278642654419, | |
| "learning_rate": 0.00011149287680751763, | |
| "loss": 1.1115, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.46842525979216626, | |
| "grad_norm": 0.648656964302063, | |
| "learning_rate": 0.0001112413675966298, | |
| "loss": 1.046, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.469224620303757, | |
| "grad_norm": 0.6471766829490662, | |
| "learning_rate": 0.0001109897863455104, | |
| "loss": 1.1142, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.47002398081534774, | |
| "grad_norm": 0.7656954526901245, | |
| "learning_rate": 0.00011073813466641632, | |
| "loss": 0.9949, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.4708233413269384, | |
| "grad_norm": 0.7737781405448914, | |
| "learning_rate": 0.00011048641417205563, | |
| "loss": 1.1068, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.47162270183852917, | |
| "grad_norm": 0.6723043322563171, | |
| "learning_rate": 0.00011023462647557751, | |
| "loss": 1.0541, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.4724220623501199, | |
| "grad_norm": 0.8895704746246338, | |
| "learning_rate": 0.0001099827731905618, | |
| "loss": 1.0259, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.47322142286171065, | |
| "grad_norm": 0.8586205840110779, | |
| "learning_rate": 0.0001097308559310087, | |
| "loss": 1.0622, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.47402078337330134, | |
| "grad_norm": 0.9140827059745789, | |
| "learning_rate": 0.00010947887631132823, | |
| "loss": 1.0914, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.4748201438848921, | |
| "grad_norm": 0.9299436211585999, | |
| "learning_rate": 0.00010922683594633021, | |
| "loss": 1.1697, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.4756195043964828, | |
| "grad_norm": 1.0227932929992676, | |
| "learning_rate": 0.00010897473645121372, | |
| "loss": 1.1914, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.47641886490807356, | |
| "grad_norm": 0.9583956003189087, | |
| "learning_rate": 0.00010872257944155669, | |
| "loss": 1.1075, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.47721822541966424, | |
| "grad_norm": 0.8131706714630127, | |
| "learning_rate": 0.00010847036653330568, | |
| "loss": 0.9999, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.478017585931255, | |
| "grad_norm": 1.167009711265564, | |
| "learning_rate": 0.00010821809934276554, | |
| "loss": 1.147, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.4788169464428457, | |
| "grad_norm": 1.1924556493759155, | |
| "learning_rate": 0.00010796577948658892, | |
| "loss": 1.0572, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.47961630695443647, | |
| "grad_norm": 1.5610358715057373, | |
| "learning_rate": 0.00010771340858176594, | |
| "loss": 1.3864, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.48041566746602715, | |
| "grad_norm": 0.38183894753456116, | |
| "learning_rate": 0.00010746098824561397, | |
| "loss": 0.9494, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.4812150279776179, | |
| "grad_norm": 0.404624879360199, | |
| "learning_rate": 0.00010720852009576705, | |
| "loss": 0.8817, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.48201438848920863, | |
| "grad_norm": 0.4218512773513794, | |
| "learning_rate": 0.00010695600575016571, | |
| "loss": 0.9529, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.4828137490007994, | |
| "grad_norm": 0.37252408266067505, | |
| "learning_rate": 0.0001067034468270465, | |
| "loss": 0.9314, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.48361310951239006, | |
| "grad_norm": 0.36367425322532654, | |
| "learning_rate": 0.00010645084494493165, | |
| "loss": 1.1066, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.4844124700239808, | |
| "grad_norm": 0.4049261510372162, | |
| "learning_rate": 0.00010619820172261867, | |
| "loss": 1.3076, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.48521183053557154, | |
| "grad_norm": 0.4004517197608948, | |
| "learning_rate": 0.00010594551877917003, | |
| "loss": 0.9514, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.4860111910471623, | |
| "grad_norm": 0.4588954746723175, | |
| "learning_rate": 0.00010569279773390273, | |
| "loss": 1.0399, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.486810551558753, | |
| "grad_norm": 0.3579251766204834, | |
| "learning_rate": 0.00010544004020637799, | |
| "loss": 1.0246, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.4876099120703437, | |
| "grad_norm": 0.3608669936656952, | |
| "learning_rate": 0.00010518724781639073, | |
| "loss": 1.2545, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.48840927258193445, | |
| "grad_norm": 0.3573903441429138, | |
| "learning_rate": 0.00010493442218395936, | |
| "loss": 1.1229, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.4892086330935252, | |
| "grad_norm": 0.35174012184143066, | |
| "learning_rate": 0.0001046815649293154, | |
| "loss": 1.0862, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.49000799360511593, | |
| "grad_norm": 0.33587905764579773, | |
| "learning_rate": 0.00010442867767289281, | |
| "loss": 1.0393, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.4908073541167066, | |
| "grad_norm": 0.363019198179245, | |
| "learning_rate": 0.000104175762035318, | |
| "loss": 1.1546, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.49160671462829736, | |
| "grad_norm": 0.3578447997570038, | |
| "learning_rate": 0.00010392281963739918, | |
| "loss": 1.1706, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.4924060751398881, | |
| "grad_norm": 0.41954872012138367, | |
| "learning_rate": 0.0001036698521001161, | |
| "loss": 1.0753, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.49320543565147884, | |
| "grad_norm": 0.36402976512908936, | |
| "learning_rate": 0.00010341686104460954, | |
| "loss": 1.0745, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.4940047961630695, | |
| "grad_norm": 0.38607266545295715, | |
| "learning_rate": 0.00010316384809217108, | |
| "loss": 1.0168, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.49480415667466027, | |
| "grad_norm": 0.38932761549949646, | |
| "learning_rate": 0.0001029108148642326, | |
| "loss": 1.0627, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.495603517186251, | |
| "grad_norm": 0.3459804058074951, | |
| "learning_rate": 0.00010265776298235586, | |
| "loss": 1.1029, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.49640287769784175, | |
| "grad_norm": 0.3945534825325012, | |
| "learning_rate": 0.00010240469406822221, | |
| "loss": 0.9676, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.49720223820943243, | |
| "grad_norm": 0.4144132733345032, | |
| "learning_rate": 0.00010215160974362223, | |
| "loss": 1.1176, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.4980015987210232, | |
| "grad_norm": 0.4400719404220581, | |
| "learning_rate": 0.0001018985116304451, | |
| "loss": 0.968, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.4988009592326139, | |
| "grad_norm": 0.4012643098831177, | |
| "learning_rate": 0.00010164540135066845, | |
| "loss": 1.0228, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.49960031974420466, | |
| "grad_norm": 0.40959033370018005, | |
| "learning_rate": 0.00010139228052634789, | |
| "loss": 1.0099, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.5003996802557954, | |
| "grad_norm": 0.3941745162010193, | |
| "learning_rate": 0.00010113915077960658, | |
| "loss": 1.1203, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.5003996802557954, | |
| "eval_loss": 1.0932515859603882, | |
| "eval_runtime": 40.3343, | |
| "eval_samples_per_second": 13.066, | |
| "eval_steps_per_second": 6.545, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.5011990407673861, | |
| "grad_norm": 0.4395071268081665, | |
| "learning_rate": 0.00010088601373262486, | |
| "loss": 0.9102, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.5019984012789768, | |
| "grad_norm": 0.4352056384086609, | |
| "learning_rate": 0.00010063287100762985, | |
| "loss": 1.109, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.5027977617905676, | |
| "grad_norm": 0.46828141808509827, | |
| "learning_rate": 0.00010037972422688511, | |
| "loss": 1.0963, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.5035971223021583, | |
| "grad_norm": 0.45365041494369507, | |
| "learning_rate": 0.0001001265750126801, | |
| "loss": 0.9817, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.504396482813749, | |
| "grad_norm": 0.5557889342308044, | |
| "learning_rate": 9.987342498731992e-05, | |
| "loss": 1.0607, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.5051958433253397, | |
| "grad_norm": 0.4783964157104492, | |
| "learning_rate": 9.962027577311491e-05, | |
| "loss": 1.0659, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.5059952038369304, | |
| "grad_norm": 0.5517037510871887, | |
| "learning_rate": 9.936712899237017e-05, | |
| "loss": 1.0411, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.5067945643485212, | |
| "grad_norm": 0.5102303624153137, | |
| "learning_rate": 9.911398626737517e-05, | |
| "loss": 0.7951, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.5075939248601119, | |
| "grad_norm": 0.547484278678894, | |
| "learning_rate": 9.886084922039344e-05, | |
| "loss": 0.9594, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.5083932853717026, | |
| "grad_norm": 0.719420850276947, | |
| "learning_rate": 9.860771947365214e-05, | |
| "loss": 1.0052, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.5091926458832934, | |
| "grad_norm": 0.6604675650596619, | |
| "learning_rate": 9.835459864933156e-05, | |
| "loss": 1.0055, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.5099920063948841, | |
| "grad_norm": 0.6773166656494141, | |
| "learning_rate": 9.810148836955492e-05, | |
| "loss": 1.1806, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.5107913669064749, | |
| "grad_norm": 0.877698540687561, | |
| "learning_rate": 9.784839025637778e-05, | |
| "loss": 1.1224, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.5115907274180655, | |
| "grad_norm": 0.7317598462104797, | |
| "learning_rate": 9.759530593177777e-05, | |
| "loss": 1.0127, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5123900879296562, | |
| "grad_norm": 0.8157115578651428, | |
| "learning_rate": 9.734223701764415e-05, | |
| "loss": 1.2444, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.513189448441247, | |
| "grad_norm": 0.9267081022262573, | |
| "learning_rate": 9.708918513576743e-05, | |
| "loss": 1.1454, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.5139888089528377, | |
| "grad_norm": 0.762115478515625, | |
| "learning_rate": 9.683615190782892e-05, | |
| "loss": 1.0811, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.5147881694644284, | |
| "grad_norm": 0.9379759430885315, | |
| "learning_rate": 9.658313895539045e-05, | |
| "loss": 1.0864, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.5155875299760192, | |
| "grad_norm": 1.0492953062057495, | |
| "learning_rate": 9.63301478998839e-05, | |
| "loss": 1.1515, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.5163868904876099, | |
| "grad_norm": 1.03159499168396, | |
| "learning_rate": 9.607718036260083e-05, | |
| "loss": 1.2317, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.5171862509992007, | |
| "grad_norm": 1.086949348449707, | |
| "learning_rate": 9.582423796468201e-05, | |
| "loss": 1.1982, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.5179856115107914, | |
| "grad_norm": 1.140920877456665, | |
| "learning_rate": 9.557132232710724e-05, | |
| "loss": 1.0614, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.518784972022382, | |
| "grad_norm": 1.3680213689804077, | |
| "learning_rate": 9.531843507068465e-05, | |
| "loss": 1.3409, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.5195843325339728, | |
| "grad_norm": 1.9634144306182861, | |
| "learning_rate": 9.506557781604066e-05, | |
| "loss": 1.0384, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5203836930455635, | |
| "grad_norm": 0.35742971301078796, | |
| "learning_rate": 9.481275218360932e-05, | |
| "loss": 0.9856, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.5211830535571543, | |
| "grad_norm": 0.39657095074653625, | |
| "learning_rate": 9.455995979362206e-05, | |
| "loss": 0.9763, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.521982414068745, | |
| "grad_norm": 0.3931129276752472, | |
| "learning_rate": 9.43072022660973e-05, | |
| "loss": 0.9867, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.5227817745803357, | |
| "grad_norm": 0.35188305377960205, | |
| "learning_rate": 9.405448122083002e-05, | |
| "loss": 1.0704, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.5235811350919265, | |
| "grad_norm": 0.35903307795524597, | |
| "learning_rate": 9.380179827738136e-05, | |
| "loss": 1.144, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.5243804956035172, | |
| "grad_norm": 0.36907997727394104, | |
| "learning_rate": 9.354915505506839e-05, | |
| "loss": 1.0573, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.5251798561151079, | |
| "grad_norm": 0.3568347096443176, | |
| "learning_rate": 9.329655317295354e-05, | |
| "loss": 0.8698, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.5259792166266987, | |
| "grad_norm": 0.368610143661499, | |
| "learning_rate": 9.304399424983432e-05, | |
| "loss": 1.0712, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.5267785771382894, | |
| "grad_norm": 0.4084762930870056, | |
| "learning_rate": 9.279147990423296e-05, | |
| "loss": 1.1245, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.5275779376498801, | |
| "grad_norm": 0.35369983315467834, | |
| "learning_rate": 9.253901175438607e-05, | |
| "loss": 1.0057, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5283772981614708, | |
| "grad_norm": 0.38250893354415894, | |
| "learning_rate": 9.228659141823406e-05, | |
| "loss": 1.1446, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.5291766586730615, | |
| "grad_norm": 0.3296395540237427, | |
| "learning_rate": 9.203422051341109e-05, | |
| "loss": 1.1662, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.5299760191846523, | |
| "grad_norm": 0.31364706158638, | |
| "learning_rate": 9.178190065723447e-05, | |
| "loss": 1.0356, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.530775379696243, | |
| "grad_norm": 0.3382437527179718, | |
| "learning_rate": 9.152963346669433e-05, | |
| "loss": 1.1452, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.5315747402078337, | |
| "grad_norm": 0.3270893692970276, | |
| "learning_rate": 9.127742055844334e-05, | |
| "loss": 0.9644, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.5323741007194245, | |
| "grad_norm": 0.38368576765060425, | |
| "learning_rate": 9.102526354878631e-05, | |
| "loss": 1.0437, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.5331734612310152, | |
| "grad_norm": 0.4607539772987366, | |
| "learning_rate": 9.077316405366981e-05, | |
| "loss": 1.1424, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.533972821742606, | |
| "grad_norm": 0.3516378700733185, | |
| "learning_rate": 9.052112368867181e-05, | |
| "loss": 0.9903, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.5347721822541966, | |
| "grad_norm": 0.4142468571662903, | |
| "learning_rate": 9.026914406899134e-05, | |
| "loss": 1.0158, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.5355715427657873, | |
| "grad_norm": 0.5051367282867432, | |
| "learning_rate": 9.00172268094382e-05, | |
| "loss": 1.0143, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5363709032773781, | |
| "grad_norm": 0.39056387543678284, | |
| "learning_rate": 8.97653735244225e-05, | |
| "loss": 1.1368, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.5371702637889688, | |
| "grad_norm": 0.4168921411037445, | |
| "learning_rate": 8.951358582794438e-05, | |
| "loss": 1.1265, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.5379696243005595, | |
| "grad_norm": 0.43402183055877686, | |
| "learning_rate": 8.92618653335837e-05, | |
| "loss": 1.1295, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.5387689848121503, | |
| "grad_norm": 0.43355968594551086, | |
| "learning_rate": 8.901021365448958e-05, | |
| "loss": 1.0647, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.539568345323741, | |
| "grad_norm": 0.4178990125656128, | |
| "learning_rate": 8.87586324033702e-05, | |
| "loss": 1.0942, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.5403677058353318, | |
| "grad_norm": 0.4249877333641052, | |
| "learning_rate": 8.850712319248236e-05, | |
| "loss": 0.9605, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.5411670663469225, | |
| "grad_norm": 0.46403768658638, | |
| "learning_rate": 8.825568763362127e-05, | |
| "loss": 1.1758, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.5419664268585132, | |
| "grad_norm": 0.5160415768623352, | |
| "learning_rate": 8.800432733811001e-05, | |
| "loss": 1.2294, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.542765787370104, | |
| "grad_norm": 0.3881043791770935, | |
| "learning_rate": 8.775304391678942e-05, | |
| "loss": 1.078, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.5435651478816946, | |
| "grad_norm": 0.5268367528915405, | |
| "learning_rate": 8.750183898000774e-05, | |
| "loss": 0.92, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5443645083932853, | |
| "grad_norm": 0.45545732975006104, | |
| "learning_rate": 8.725071413761012e-05, | |
| "loss": 0.9236, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.5451638689048761, | |
| "grad_norm": 0.5602855682373047, | |
| "learning_rate": 8.699967099892852e-05, | |
| "loss": 1.2108, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.5459632294164668, | |
| "grad_norm": 0.5163658261299133, | |
| "learning_rate": 8.67487111727713e-05, | |
| "loss": 0.9604, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.5467625899280576, | |
| "grad_norm": 0.5549715757369995, | |
| "learning_rate": 8.64978362674128e-05, | |
| "loss": 0.9453, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.5475619504396483, | |
| "grad_norm": 0.5914304256439209, | |
| "learning_rate": 8.624704789058331e-05, | |
| "loss": 0.9931, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.548361310951239, | |
| "grad_norm": 0.6247863173484802, | |
| "learning_rate": 8.599634764945845e-05, | |
| "loss": 1.0431, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.5491606714628298, | |
| "grad_norm": 0.7612833380699158, | |
| "learning_rate": 8.574573715064913e-05, | |
| "loss": 0.9153, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.5499600319744204, | |
| "grad_norm": 0.6234817504882812, | |
| "learning_rate": 8.549521800019113e-05, | |
| "loss": 1.0812, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.5507593924860112, | |
| "grad_norm": 0.6706773638725281, | |
| "learning_rate": 8.524479180353478e-05, | |
| "loss": 0.8517, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.5515587529976019, | |
| "grad_norm": 0.761870801448822, | |
| "learning_rate": 8.499446016553474e-05, | |
| "loss": 1.1793, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5523581135091926, | |
| "grad_norm": 0.7513264417648315, | |
| "learning_rate": 8.474422469043973e-05, | |
| "loss": 1.3344, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.5531574740207834, | |
| "grad_norm": 0.7695756554603577, | |
| "learning_rate": 8.449408698188213e-05, | |
| "loss": 0.9859, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.5539568345323741, | |
| "grad_norm": 0.8410113453865051, | |
| "learning_rate": 8.424404864286784e-05, | |
| "loss": 0.9965, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.5547561950439648, | |
| "grad_norm": 0.9044365286827087, | |
| "learning_rate": 8.399411127576597e-05, | |
| "loss": 1.0025, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.9618779420852661, | |
| "learning_rate": 8.374427648229846e-05, | |
| "loss": 1.1298, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.5563549160671463, | |
| "grad_norm": 1.178658127784729, | |
| "learning_rate": 8.349454586353e-05, | |
| "loss": 1.0035, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.5571542765787371, | |
| "grad_norm": 1.1606285572052002, | |
| "learning_rate": 8.324492101985768e-05, | |
| "loss": 1.1512, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.5579536370903277, | |
| "grad_norm": 1.2002030611038208, | |
| "learning_rate": 8.299540355100065e-05, | |
| "loss": 1.2104, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.5587529976019184, | |
| "grad_norm": 1.7446719408035278, | |
| "learning_rate": 8.274599505599e-05, | |
| "loss": 1.4774, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.5595523581135092, | |
| "grad_norm": 1.540687084197998, | |
| "learning_rate": 8.249669713315848e-05, | |
| "loss": 1.1088, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5603517186250999, | |
| "grad_norm": 0.32139134407043457, | |
| "learning_rate": 8.224751138013023e-05, | |
| "loss": 0.8488, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.5611510791366906, | |
| "grad_norm": 0.3686629831790924, | |
| "learning_rate": 8.199843939381049e-05, | |
| "loss": 1.0761, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.5619504396482814, | |
| "grad_norm": 0.4355558454990387, | |
| "learning_rate": 8.174948277037548e-05, | |
| "loss": 1.1466, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.5627498001598721, | |
| "grad_norm": 0.3635938763618469, | |
| "learning_rate": 8.150064310526217e-05, | |
| "loss": 1.0464, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.5635491606714629, | |
| "grad_norm": 0.34856486320495605, | |
| "learning_rate": 8.125192199315787e-05, | |
| "loss": 0.9698, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.5643485211830536, | |
| "grad_norm": 0.39139944314956665, | |
| "learning_rate": 8.100332102799025e-05, | |
| "loss": 1.081, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.5651478816946442, | |
| "grad_norm": 0.4321231544017792, | |
| "learning_rate": 8.075484180291701e-05, | |
| "loss": 1.0198, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.565947242206235, | |
| "grad_norm": 0.3629261255264282, | |
| "learning_rate": 8.050648591031562e-05, | |
| "loss": 1.0162, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.5667466027178257, | |
| "grad_norm": 0.39623329043388367, | |
| "learning_rate": 8.025825494177323e-05, | |
| "loss": 1.0521, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.5675459632294164, | |
| "grad_norm": 0.39008572697639465, | |
| "learning_rate": 8.00101504880764e-05, | |
| "loss": 1.2911, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.5683453237410072, | |
| "grad_norm": 0.3346228003501892, | |
| "learning_rate": 7.976217413920093e-05, | |
| "loss": 1.0933, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.5691446842525979, | |
| "grad_norm": 0.34785088896751404, | |
| "learning_rate": 7.95143274843016e-05, | |
| "loss": 1.1413, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.5699440447641887, | |
| "grad_norm": 0.313931405544281, | |
| "learning_rate": 7.926661211170213e-05, | |
| "loss": 1.1282, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.5707434052757794, | |
| "grad_norm": 0.39661192893981934, | |
| "learning_rate": 7.901902960888491e-05, | |
| "loss": 1.241, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.5715427657873701, | |
| "grad_norm": 0.38687777519226074, | |
| "learning_rate": 7.877158156248074e-05, | |
| "loss": 1.2047, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.5723421262989609, | |
| "grad_norm": 0.33725664019584656, | |
| "learning_rate": 7.852426955825887e-05, | |
| "loss": 1.0205, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.5731414868105515, | |
| "grad_norm": 0.3965463936328888, | |
| "learning_rate": 7.827709518111673e-05, | |
| "loss": 1.017, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.5739408473221422, | |
| "grad_norm": 0.5649027228355408, | |
| "learning_rate": 7.803006001506961e-05, | |
| "loss": 1.2147, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.574740207833733, | |
| "grad_norm": 0.35998645424842834, | |
| "learning_rate": 7.778316564324085e-05, | |
| "loss": 1.0301, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.5755395683453237, | |
| "grad_norm": 0.4070900082588196, | |
| "learning_rate": 7.753641364785138e-05, | |
| "loss": 1.1825, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5763389288569145, | |
| "grad_norm": 0.4043642580509186, | |
| "learning_rate": 7.728980561020979e-05, | |
| "loss": 1.2504, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.5771382893685052, | |
| "grad_norm": 0.4256274402141571, | |
| "learning_rate": 7.704334311070215e-05, | |
| "loss": 1.0733, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.5779376498800959, | |
| "grad_norm": 0.38983389735221863, | |
| "learning_rate": 7.679702772878172e-05, | |
| "loss": 0.9375, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.5787370103916867, | |
| "grad_norm": 0.449326753616333, | |
| "learning_rate": 7.655086104295904e-05, | |
| "loss": 1.1876, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.5795363709032774, | |
| "grad_norm": 0.42246901988983154, | |
| "learning_rate": 7.630484463079177e-05, | |
| "loss": 1.0497, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.580335731414868, | |
| "grad_norm": 0.3945838510990143, | |
| "learning_rate": 7.605898006887447e-05, | |
| "loss": 1.0072, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.5811350919264588, | |
| "grad_norm": 0.4597523510456085, | |
| "learning_rate": 7.581326893282858e-05, | |
| "loss": 1.034, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.5819344524380495, | |
| "grad_norm": 0.4484693706035614, | |
| "learning_rate": 7.556771279729236e-05, | |
| "loss": 1.025, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.5827338129496403, | |
| "grad_norm": 0.47144320607185364, | |
| "learning_rate": 7.532231323591067e-05, | |
| "loss": 1.0669, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.583533173461231, | |
| "grad_norm": 0.44177764654159546, | |
| "learning_rate": 7.507707182132506e-05, | |
| "loss": 1.0858, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.5843325339728217, | |
| "grad_norm": 0.5104259848594666, | |
| "learning_rate": 7.483199012516353e-05, | |
| "loss": 1.2095, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.5851318944844125, | |
| "grad_norm": 0.548528254032135, | |
| "learning_rate": 7.458706971803055e-05, | |
| "loss": 0.9455, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.5859312549960032, | |
| "grad_norm": 0.5504273176193237, | |
| "learning_rate": 7.434231216949696e-05, | |
| "loss": 0.8916, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.586730615507594, | |
| "grad_norm": 0.6220442652702332, | |
| "learning_rate": 7.409771904808993e-05, | |
| "loss": 1.0541, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.5875299760191847, | |
| "grad_norm": 0.8554894924163818, | |
| "learning_rate": 7.38532919212829e-05, | |
| "loss": 1.0596, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.5883293365307753, | |
| "grad_norm": 0.6036530137062073, | |
| "learning_rate": 7.360903235548553e-05, | |
| "loss": 1.0417, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.5891286970423661, | |
| "grad_norm": 0.5838643312454224, | |
| "learning_rate": 7.336494191603364e-05, | |
| "loss": 0.9401, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.5899280575539568, | |
| "grad_norm": 0.7749225497245789, | |
| "learning_rate": 7.312102216717929e-05, | |
| "loss": 1.0759, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.5907274180655475, | |
| "grad_norm": 0.7013788223266602, | |
| "learning_rate": 7.287727467208056e-05, | |
| "loss": 0.9142, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.5915267785771383, | |
| "grad_norm": 0.7764909863471985, | |
| "learning_rate": 7.263370099279172e-05, | |
| "loss": 1.1373, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.592326139088729, | |
| "grad_norm": 0.6547873616218567, | |
| "learning_rate": 7.239030269025311e-05, | |
| "loss": 1.1794, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.5931254996003198, | |
| "grad_norm": 0.7785171866416931, | |
| "learning_rate": 7.214708132428121e-05, | |
| "loss": 1.0173, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.5939248601119105, | |
| "grad_norm": 0.96076500415802, | |
| "learning_rate": 7.190403845355853e-05, | |
| "loss": 1.2693, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.5947242206235012, | |
| "grad_norm": 0.8062298893928528, | |
| "learning_rate": 7.166117563562375e-05, | |
| "loss": 1.1167, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.595523581135092, | |
| "grad_norm": 1.0789361000061035, | |
| "learning_rate": 7.141849442686168e-05, | |
| "loss": 1.3157, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.5963229416466826, | |
| "grad_norm": 0.9564920663833618, | |
| "learning_rate": 7.117599638249326e-05, | |
| "loss": 1.0115, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.5971223021582733, | |
| "grad_norm": 1.0859782695770264, | |
| "learning_rate": 7.093368305656563e-05, | |
| "loss": 1.1539, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.5979216626698641, | |
| "grad_norm": 1.1587315797805786, | |
| "learning_rate": 7.069155600194223e-05, | |
| "loss": 1.2031, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.5987210231814548, | |
| "grad_norm": 1.4126029014587402, | |
| "learning_rate": 7.044961677029265e-05, | |
| "loss": 1.0768, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.5995203836930456, | |
| "grad_norm": 1.9330767393112183, | |
| "learning_rate": 7.020786691208292e-05, | |
| "loss": 1.2553, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6003197442046363, | |
| "grad_norm": 0.33225733041763306, | |
| "learning_rate": 6.996630797656547e-05, | |
| "loss": 1.0322, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.601119104716227, | |
| "grad_norm": 0.33560457825660706, | |
| "learning_rate": 6.972494151176915e-05, | |
| "loss": 1.2175, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.6019184652278178, | |
| "grad_norm": 0.34604617953300476, | |
| "learning_rate": 6.948376906448942e-05, | |
| "loss": 1.0292, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.6027178257394085, | |
| "grad_norm": 0.3931783139705658, | |
| "learning_rate": 6.924279218027828e-05, | |
| "loss": 1.1754, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.6035171862509991, | |
| "grad_norm": 0.3545131981372833, | |
| "learning_rate": 6.900201240343458e-05, | |
| "loss": 1.0322, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.60431654676259, | |
| "grad_norm": 0.3037162721157074, | |
| "learning_rate": 6.876143127699398e-05, | |
| "loss": 0.9488, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.6051159072741806, | |
| "grad_norm": 0.41264012455940247, | |
| "learning_rate": 6.852105034271902e-05, | |
| "loss": 1.0289, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.6059152677857714, | |
| "grad_norm": 0.4278363287448883, | |
| "learning_rate": 6.82808711410894e-05, | |
| "loss": 1.1346, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.6067146282973621, | |
| "grad_norm": 0.3742501437664032, | |
| "learning_rate": 6.804089521129202e-05, | |
| "loss": 1.0822, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.6075139888089528, | |
| "grad_norm": 0.3738841414451599, | |
| "learning_rate": 6.780112409121099e-05, | |
| "loss": 1.1363, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6083133493205436, | |
| "grad_norm": 0.32057246565818787, | |
| "learning_rate": 6.756155931741802e-05, | |
| "loss": 1.1435, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.6091127098321343, | |
| "grad_norm": 0.3369678556919098, | |
| "learning_rate": 6.732220242516242e-05, | |
| "loss": 1.0887, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.609912070343725, | |
| "grad_norm": 0.34893402457237244, | |
| "learning_rate": 6.708305494836131e-05, | |
| "loss": 1.1949, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.6107114308553158, | |
| "grad_norm": 0.36019623279571533, | |
| "learning_rate": 6.68441184195897e-05, | |
| "loss": 1.1954, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.6115107913669064, | |
| "grad_norm": 0.3615095913410187, | |
| "learning_rate": 6.660539437007081e-05, | |
| "loss": 1.1763, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.6123101518784972, | |
| "grad_norm": 0.40820664167404175, | |
| "learning_rate": 6.636688432966618e-05, | |
| "loss": 1.0974, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.6131095123900879, | |
| "grad_norm": 0.3878170847892761, | |
| "learning_rate": 6.612858982686584e-05, | |
| "loss": 1.2213, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.6139088729016786, | |
| "grad_norm": 0.4106099009513855, | |
| "learning_rate": 6.589051238877858e-05, | |
| "loss": 1.0086, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.6147082334132694, | |
| "grad_norm": 0.37214210629463196, | |
| "learning_rate": 6.565265354112214e-05, | |
| "loss": 0.9487, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.6155075939248601, | |
| "grad_norm": 0.41581258177757263, | |
| "learning_rate": 6.541501480821338e-05, | |
| "loss": 0.9857, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6163069544364509, | |
| "grad_norm": 0.3811487853527069, | |
| "learning_rate": 6.51775977129586e-05, | |
| "loss": 1.0552, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.6171063149480416, | |
| "grad_norm": 0.4689180552959442, | |
| "learning_rate": 6.494040377684372e-05, | |
| "loss": 1.0779, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.6179056754596323, | |
| "grad_norm": 0.3837207853794098, | |
| "learning_rate": 6.470343451992452e-05, | |
| "loss": 1.0286, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.6187050359712231, | |
| "grad_norm": 0.4035284221172333, | |
| "learning_rate": 6.446669146081698e-05, | |
| "loss": 1.0707, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.6195043964828137, | |
| "grad_norm": 0.40140479803085327, | |
| "learning_rate": 6.423017611668745e-05, | |
| "loss": 0.9489, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.6203037569944044, | |
| "grad_norm": 0.39459696412086487, | |
| "learning_rate": 6.3993890003243e-05, | |
| "loss": 0.9949, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.6211031175059952, | |
| "grad_norm": 0.4292733073234558, | |
| "learning_rate": 6.375783463472165e-05, | |
| "loss": 1.0554, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.6219024780175859, | |
| "grad_norm": 0.4641924202442169, | |
| "learning_rate": 6.352201152388268e-05, | |
| "loss": 0.9926, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.6227018385291767, | |
| "grad_norm": 0.527754545211792, | |
| "learning_rate": 6.328642218199702e-05, | |
| "loss": 0.9438, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.6235011990407674, | |
| "grad_norm": 0.4878404140472412, | |
| "learning_rate": 6.30510681188374e-05, | |
| "loss": 0.9029, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6243005595523581, | |
| "grad_norm": 0.5755017995834351, | |
| "learning_rate": 6.281595084266883e-05, | |
| "loss": 0.971, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.6250999200639489, | |
| "grad_norm": 0.5024919509887695, | |
| "learning_rate": 6.258107186023885e-05, | |
| "loss": 1.0335, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.6258992805755396, | |
| "grad_norm": 0.5928468108177185, | |
| "learning_rate": 6.23464326767679e-05, | |
| "loss": 1.0721, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.6266986410871302, | |
| "grad_norm": 0.6729336380958557, | |
| "learning_rate": 6.211203479593968e-05, | |
| "loss": 0.9853, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.627498001598721, | |
| "grad_norm": 0.6404271721839905, | |
| "learning_rate": 6.18778797198915e-05, | |
| "loss": 0.9947, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.6282973621103117, | |
| "grad_norm": 0.7536145448684692, | |
| "learning_rate": 6.164396894920462e-05, | |
| "loss": 1.2851, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.6290967226219025, | |
| "grad_norm": 0.669144332408905, | |
| "learning_rate": 6.141030398289473e-05, | |
| "loss": 1.1976, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.6298960831334932, | |
| "grad_norm": 0.6938853859901428, | |
| "learning_rate": 6.117688631840224e-05, | |
| "loss": 1.0081, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.6306954436450839, | |
| "grad_norm": 0.7957220673561096, | |
| "learning_rate": 6.0943717451582724e-05, | |
| "loss": 1.0587, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.6314948041566747, | |
| "grad_norm": 0.7660002112388611, | |
| "learning_rate": 6.071079887669742e-05, | |
| "loss": 1.0468, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6322941646682654, | |
| "grad_norm": 1.0254074335098267, | |
| "learning_rate": 6.0478132086403426e-05, | |
| "loss": 0.9698, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.6330935251798561, | |
| "grad_norm": 0.902427613735199, | |
| "learning_rate": 6.024571857174443e-05, | |
| "loss": 1.0807, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.6338928856914469, | |
| "grad_norm": 0.872407078742981, | |
| "learning_rate": 6.001355982214092e-05, | |
| "loss": 1.0333, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.6346922462030375, | |
| "grad_norm": 0.7746816277503967, | |
| "learning_rate": 5.978165732538073e-05, | |
| "loss": 1.0817, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.6354916067146283, | |
| "grad_norm": 0.9597741365432739, | |
| "learning_rate": 5.95500125676095e-05, | |
| "loss": 1.0546, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.636290967226219, | |
| "grad_norm": 0.8365216255187988, | |
| "learning_rate": 5.9318627033321204e-05, | |
| "loss": 0.9174, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.6370903277378097, | |
| "grad_norm": 1.751240849494934, | |
| "learning_rate": 5.9087502205348534e-05, | |
| "loss": 1.2519, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.6378896882494005, | |
| "grad_norm": 1.1197962760925293, | |
| "learning_rate": 5.885663956485341e-05, | |
| "loss": 1.084, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.6386890487609912, | |
| "grad_norm": 1.6539058685302734, | |
| "learning_rate": 5.86260405913176e-05, | |
| "loss": 1.1596, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.6394884092725819, | |
| "grad_norm": 1.8245570659637451, | |
| "learning_rate": 5.839570676253316e-05, | |
| "loss": 1.0606, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6402877697841727, | |
| "grad_norm": 0.3631491959095001, | |
| "learning_rate": 5.81656395545929e-05, | |
| "loss": 1.0136, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.6410871302957634, | |
| "grad_norm": 0.34658652544021606, | |
| "learning_rate": 5.7935840441881054e-05, | |
| "loss": 0.9605, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.6418864908073542, | |
| "grad_norm": 0.3638405203819275, | |
| "learning_rate": 5.770631089706373e-05, | |
| "loss": 0.8942, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.6426858513189448, | |
| "grad_norm": 0.42472997307777405, | |
| "learning_rate": 5.747705239107958e-05, | |
| "loss": 1.0283, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.6434852118305355, | |
| "grad_norm": 0.4047344923019409, | |
| "learning_rate": 5.724806639313025e-05, | |
| "loss": 1.0082, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.6442845723421263, | |
| "grad_norm": 0.39204633235931396, | |
| "learning_rate": 5.701935437067096e-05, | |
| "loss": 1.0789, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.645083932853717, | |
| "grad_norm": 0.35815802216529846, | |
| "learning_rate": 5.6790917789401285e-05, | |
| "loss": 1.1287, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.6458832933653078, | |
| "grad_norm": 0.34175869822502136, | |
| "learning_rate": 5.656275811325557e-05, | |
| "loss": 1.0832, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.6466826538768985, | |
| "grad_norm": 0.3721088469028473, | |
| "learning_rate": 5.633487680439361e-05, | |
| "loss": 1.0313, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.6474820143884892, | |
| "grad_norm": 0.399013489484787, | |
| "learning_rate": 5.610727532319131e-05, | |
| "loss": 1.0316, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.64828137490008, | |
| "grad_norm": 0.3578621447086334, | |
| "learning_rate": 5.587995512823133e-05, | |
| "loss": 1.1265, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.6490807354116707, | |
| "grad_norm": 0.3983486294746399, | |
| "learning_rate": 5.565291767629359e-05, | |
| "loss": 1.0861, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.6498800959232613, | |
| "grad_norm": 0.3946421444416046, | |
| "learning_rate": 5.542616442234618e-05, | |
| "loss": 1.0956, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.6506794564348521, | |
| "grad_norm": 0.35159727931022644, | |
| "learning_rate": 5.5199696819535854e-05, | |
| "loss": 1.2644, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.6514788169464428, | |
| "grad_norm": 0.36194556951522827, | |
| "learning_rate": 5.4973516319178774e-05, | |
| "loss": 1.1822, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.6522781774580336, | |
| "grad_norm": 0.3506593108177185, | |
| "learning_rate": 5.474762437075123e-05, | |
| "loss": 1.119, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.6530775379696243, | |
| "grad_norm": 0.4893072545528412, | |
| "learning_rate": 5.4522022421880295e-05, | |
| "loss": 1.0987, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.653876898481215, | |
| "grad_norm": 0.3882545530796051, | |
| "learning_rate": 5.4296711918334644e-05, | |
| "loss": 1.1101, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.6546762589928058, | |
| "grad_norm": 0.4101043939590454, | |
| "learning_rate": 5.407169430401511e-05, | |
| "loss": 1.2256, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.6554756195043965, | |
| "grad_norm": 0.4425513744354248, | |
| "learning_rate": 5.384697102094565e-05, | |
| "loss": 1.1876, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6562749800159872, | |
| "grad_norm": 0.40541303157806396, | |
| "learning_rate": 5.362254350926403e-05, | |
| "loss": 0.9728, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.657074340527578, | |
| "grad_norm": 0.39156901836395264, | |
| "learning_rate": 5.3398413207212396e-05, | |
| "loss": 1.0248, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.6578737010391686, | |
| "grad_norm": 0.3938673734664917, | |
| "learning_rate": 5.317458155112842e-05, | |
| "loss": 1.1646, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.6586730615507594, | |
| "grad_norm": 0.4352782666683197, | |
| "learning_rate": 5.2951049975435786e-05, | |
| "loss": 1.159, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.6594724220623501, | |
| "grad_norm": 0.44448021054267883, | |
| "learning_rate": 5.272781991263516e-05, | |
| "loss": 1.0434, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.6602717825739408, | |
| "grad_norm": 0.415777325630188, | |
| "learning_rate": 5.2504892793295e-05, | |
| "loss": 0.9778, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.6610711430855316, | |
| "grad_norm": 0.4741330146789551, | |
| "learning_rate": 5.228227004604225e-05, | |
| "loss": 1.0315, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.6618705035971223, | |
| "grad_norm": 0.501546323299408, | |
| "learning_rate": 5.205995309755335e-05, | |
| "loss": 1.1244, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.662669864108713, | |
| "grad_norm": 0.5114132165908813, | |
| "learning_rate": 5.183794337254504e-05, | |
| "loss": 1.0154, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.6634692246203038, | |
| "grad_norm": 0.4979310631752014, | |
| "learning_rate": 5.1616242293765226e-05, | |
| "loss": 0.999, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6642685851318945, | |
| "grad_norm": 0.5462687611579895, | |
| "learning_rate": 5.139485128198381e-05, | |
| "loss": 0.997, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.6650679456434853, | |
| "grad_norm": 0.5672511458396912, | |
| "learning_rate": 5.1173771755983724e-05, | |
| "loss": 1.0864, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.6658673061550759, | |
| "grad_norm": 0.5668870210647583, | |
| "learning_rate": 5.095300513255157e-05, | |
| "loss": 1.1027, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.6862683296203613, | |
| "learning_rate": 5.0732552826468896e-05, | |
| "loss": 0.8694, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.6674660271782574, | |
| "grad_norm": 0.7277308106422424, | |
| "learning_rate": 5.051241625050287e-05, | |
| "loss": 1.2427, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.6682653876898481, | |
| "grad_norm": 0.610167384147644, | |
| "learning_rate": 5.029259681539732e-05, | |
| "loss": 1.1544, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.6690647482014388, | |
| "grad_norm": 0.7170546054840088, | |
| "learning_rate": 5.007309592986367e-05, | |
| "loss": 0.8898, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.6698641087130296, | |
| "grad_norm": 0.6002946496009827, | |
| "learning_rate": 4.985391500057192e-05, | |
| "loss": 1.0542, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.6706634692246203, | |
| "grad_norm": 0.7680448889732361, | |
| "learning_rate": 4.963505543214167e-05, | |
| "loss": 0.8763, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.6714628297362111, | |
| "grad_norm": 0.7870104312896729, | |
| "learning_rate": 4.9416518627133e-05, | |
| "loss": 0.9651, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.6722621902478018, | |
| "grad_norm": 0.8171029686927795, | |
| "learning_rate": 4.919830598603763e-05, | |
| "loss": 1.1376, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.6730615507593924, | |
| "grad_norm": 1.0324592590332031, | |
| "learning_rate": 4.8980418907269863e-05, | |
| "loss": 0.9915, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.6738609112709832, | |
| "grad_norm": 0.8155781626701355, | |
| "learning_rate": 4.876285878715764e-05, | |
| "loss": 1.1595, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.6746602717825739, | |
| "grad_norm": 0.9791491627693176, | |
| "learning_rate": 4.8545627019933584e-05, | |
| "loss": 1.2055, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.6754596322941646, | |
| "grad_norm": 1.083111047744751, | |
| "learning_rate": 4.8328724997726095e-05, | |
| "loss": 1.1738, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.6762589928057554, | |
| "grad_norm": 0.8534079194068909, | |
| "learning_rate": 4.811215411055031e-05, | |
| "loss": 1.1604, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.6770583533173461, | |
| "grad_norm": 0.9267480969429016, | |
| "learning_rate": 4.78959157462994e-05, | |
| "loss": 1.0576, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.6778577138289369, | |
| "grad_norm": 1.196876049041748, | |
| "learning_rate": 4.768001129073553e-05, | |
| "loss": 1.3485, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.6786570743405276, | |
| "grad_norm": 1.2235941886901855, | |
| "learning_rate": 4.7464442127481e-05, | |
| "loss": 1.204, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.6794564348521183, | |
| "grad_norm": 1.7451354265213013, | |
| "learning_rate": 4.724920963800939e-05, | |
| "loss": 1.3127, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6802557953637091, | |
| "grad_norm": 0.3947245180606842, | |
| "learning_rate": 4.703431520163675e-05, | |
| "loss": 1.0509, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.6810551558752997, | |
| "grad_norm": 0.3554235100746155, | |
| "learning_rate": 4.6819760195512686e-05, | |
| "loss": 0.8586, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.6818545163868905, | |
| "grad_norm": 0.39175039529800415, | |
| "learning_rate": 4.660554599461151e-05, | |
| "loss": 1.0003, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.6826538768984812, | |
| "grad_norm": 0.38634535670280457, | |
| "learning_rate": 4.639167397172358e-05, | |
| "loss": 1.0581, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.6834532374100719, | |
| "grad_norm": 0.41433098912239075, | |
| "learning_rate": 4.6178145497446404e-05, | |
| "loss": 1.0889, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.6842525979216627, | |
| "grad_norm": 0.3703967034816742, | |
| "learning_rate": 4.596496194017576e-05, | |
| "loss": 1.1005, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.6850519584332534, | |
| "grad_norm": 0.36882877349853516, | |
| "learning_rate": 4.575212466609713e-05, | |
| "loss": 0.9307, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.6858513189448441, | |
| "grad_norm": 0.3600471019744873, | |
| "learning_rate": 4.553963503917681e-05, | |
| "loss": 0.893, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.6866506794564349, | |
| "grad_norm": 0.33053648471832275, | |
| "learning_rate": 4.53274944211532e-05, | |
| "loss": 1.1211, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.6874500399680256, | |
| "grad_norm": 0.37433984875679016, | |
| "learning_rate": 4.5115704171528105e-05, | |
| "loss": 1.0067, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.6882494004796164, | |
| "grad_norm": 0.40663594007492065, | |
| "learning_rate": 4.490426564755792e-05, | |
| "loss": 1.0298, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.689048760991207, | |
| "grad_norm": 0.49298176169395447, | |
| "learning_rate": 4.4693180204245074e-05, | |
| "loss": 1.0433, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.6898481215027977, | |
| "grad_norm": 0.3631519377231598, | |
| "learning_rate": 4.448244919432929e-05, | |
| "loss": 1.1514, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.6906474820143885, | |
| "grad_norm": 0.38605815172195435, | |
| "learning_rate": 4.427207396827886e-05, | |
| "loss": 1.1713, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.6914468425259792, | |
| "grad_norm": 0.341621071100235, | |
| "learning_rate": 4.406205587428206e-05, | |
| "loss": 1.1695, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.6922462030375699, | |
| "grad_norm": 0.3484615981578827, | |
| "learning_rate": 4.3852396258238536e-05, | |
| "loss": 1.0565, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.6930455635491607, | |
| "grad_norm": 0.4451755881309509, | |
| "learning_rate": 4.3643096463750496e-05, | |
| "loss": 1.1425, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.6938449240607514, | |
| "grad_norm": 0.4082989990711212, | |
| "learning_rate": 4.3434157832114373e-05, | |
| "loss": 0.9311, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.6946442845723422, | |
| "grad_norm": 0.3935573101043701, | |
| "learning_rate": 4.3225581702312025e-05, | |
| "loss": 1.1656, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.6954436450839329, | |
| "grad_norm": 0.44610002636909485, | |
| "learning_rate": 4.301736941100223e-05, | |
| "loss": 1.2297, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6962430055955235, | |
| "grad_norm": 0.44538775086402893, | |
| "learning_rate": 4.280952229251212e-05, | |
| "loss": 1.0755, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.6970423661071143, | |
| "grad_norm": 0.4480975568294525, | |
| "learning_rate": 4.26020416788286e-05, | |
| "loss": 1.2264, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.697841726618705, | |
| "grad_norm": 0.5919811129570007, | |
| "learning_rate": 4.239492889958988e-05, | |
| "loss": 1.118, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.6986410871302957, | |
| "grad_norm": 0.4208342730998993, | |
| "learning_rate": 4.21881852820768e-05, | |
| "loss": 0.9921, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.6994404476418865, | |
| "grad_norm": 0.4386043846607208, | |
| "learning_rate": 4.1981812151204513e-05, | |
| "loss": 0.9987, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.7002398081534772, | |
| "grad_norm": 0.5912065505981445, | |
| "learning_rate": 4.177581082951393e-05, | |
| "loss": 1.1662, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.701039168665068, | |
| "grad_norm": 0.5079301595687866, | |
| "learning_rate": 4.1570182637163155e-05, | |
| "loss": 1.0607, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.7018385291766587, | |
| "grad_norm": 0.5225832462310791, | |
| "learning_rate": 4.136492889191914e-05, | |
| "loss": 1.0134, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.7026378896882494, | |
| "grad_norm": 0.49494823813438416, | |
| "learning_rate": 4.1160050909149196e-05, | |
| "loss": 0.9564, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.7034372501998402, | |
| "grad_norm": 0.5825321674346924, | |
| "learning_rate": 4.095555000181257e-05, | |
| "loss": 1.1002, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7042366107114308, | |
| "grad_norm": 0.5384085774421692, | |
| "learning_rate": 4.075142748045194e-05, | |
| "loss": 0.8761, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.7050359712230215, | |
| "grad_norm": 0.5739712119102478, | |
| "learning_rate": 4.05476846531852e-05, | |
| "loss": 1.1312, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.7058353317346123, | |
| "grad_norm": 0.5879523754119873, | |
| "learning_rate": 4.034432282569693e-05, | |
| "loss": 1.1856, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.706634692246203, | |
| "grad_norm": 0.5788174867630005, | |
| "learning_rate": 4.0141343301230116e-05, | |
| "loss": 1.0818, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.7074340527577938, | |
| "grad_norm": 0.6709664463996887, | |
| "learning_rate": 3.99387473805777e-05, | |
| "loss": 1.1062, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.7082334132693845, | |
| "grad_norm": 0.7911237478256226, | |
| "learning_rate": 3.973653636207437e-05, | |
| "loss": 1.1432, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.7090327737809752, | |
| "grad_norm": 0.6831583380699158, | |
| "learning_rate": 3.953471154158808e-05, | |
| "loss": 0.8802, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.709832134292566, | |
| "grad_norm": 0.7837221026420593, | |
| "learning_rate": 3.9333274212511887e-05, | |
| "loss": 0.8944, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.7106314948041567, | |
| "grad_norm": 0.6791643500328064, | |
| "learning_rate": 3.913222566575564e-05, | |
| "loss": 0.8216, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.7114308553157475, | |
| "grad_norm": 0.8798190951347351, | |
| "learning_rate": 3.893156718973767e-05, | |
| "loss": 1.3466, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7122302158273381, | |
| "grad_norm": 0.8324375152587891, | |
| "learning_rate": 3.8731300070376464e-05, | |
| "loss": 1.074, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.7130295763389288, | |
| "grad_norm": 1.132907748222351, | |
| "learning_rate": 3.8531425591082606e-05, | |
| "loss": 1.3694, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.7138289368505196, | |
| "grad_norm": 1.0507283210754395, | |
| "learning_rate": 3.8331945032750374e-05, | |
| "loss": 1.2547, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.7146282973621103, | |
| "grad_norm": 0.9351432919502258, | |
| "learning_rate": 3.813285967374969e-05, | |
| "loss": 1.0931, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.715427657873701, | |
| "grad_norm": 1.1382817029953003, | |
| "learning_rate": 3.7934170789917724e-05, | |
| "loss": 1.2737, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.7162270183852918, | |
| "grad_norm": 0.9496870636940002, | |
| "learning_rate": 3.773587965455093e-05, | |
| "loss": 1.1127, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.7170263788968825, | |
| "grad_norm": 1.0737744569778442, | |
| "learning_rate": 3.753798753839677e-05, | |
| "loss": 1.2043, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.7178257394084733, | |
| "grad_norm": 1.1727651357650757, | |
| "learning_rate": 3.7340495709645585e-05, | |
| "loss": 1.2318, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.718625099920064, | |
| "grad_norm": 1.544258952140808, | |
| "learning_rate": 3.7143405433922485e-05, | |
| "loss": 1.3744, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.7194244604316546, | |
| "grad_norm": 1.887011170387268, | |
| "learning_rate": 3.694671797427924e-05, | |
| "loss": 0.7498, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7202238209432454, | |
| "grad_norm": 0.359557181596756, | |
| "learning_rate": 3.675043459118619e-05, | |
| "loss": 0.9036, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.7210231814548361, | |
| "grad_norm": 0.3796878159046173, | |
| "learning_rate": 3.6554556542524044e-05, | |
| "loss": 1.0923, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.7218225419664268, | |
| "grad_norm": 0.39099475741386414, | |
| "learning_rate": 3.635908508357606e-05, | |
| "loss": 0.987, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.7226219024780176, | |
| "grad_norm": 0.4106023907661438, | |
| "learning_rate": 3.6164021467019814e-05, | |
| "loss": 1.1539, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.7234212629896083, | |
| "grad_norm": 0.391400545835495, | |
| "learning_rate": 3.5969366942919234e-05, | |
| "loss": 1.0677, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.7242206235011991, | |
| "grad_norm": 0.35718274116516113, | |
| "learning_rate": 3.5775122758716585e-05, | |
| "loss": 1.1138, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.7250199840127898, | |
| "grad_norm": 0.3467647135257721, | |
| "learning_rate": 3.558129015922448e-05, | |
| "loss": 1.097, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.7258193445243805, | |
| "grad_norm": 0.37146949768066406, | |
| "learning_rate": 3.5387870386617826e-05, | |
| "loss": 1.0741, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.7266187050359713, | |
| "grad_norm": 0.42259925603866577, | |
| "learning_rate": 3.519486468042602e-05, | |
| "loss": 1.042, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.7274180655475619, | |
| "grad_norm": 0.3869469463825226, | |
| "learning_rate": 3.5002274277524874e-05, | |
| "loss": 0.9691, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7282174260591526, | |
| "grad_norm": 0.40922337770462036, | |
| "learning_rate": 3.4810100412128747e-05, | |
| "loss": 1.2811, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.7290167865707434, | |
| "grad_norm": 0.4367547035217285, | |
| "learning_rate": 3.4618344315782604e-05, | |
| "loss": 1.1267, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.7298161470823341, | |
| "grad_norm": 0.40415459871292114, | |
| "learning_rate": 3.442700721735417e-05, | |
| "loss": 1.118, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.7306155075939249, | |
| "grad_norm": 0.37041985988616943, | |
| "learning_rate": 3.4236090343026004e-05, | |
| "loss": 0.9999, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.7314148681055156, | |
| "grad_norm": 0.3673413395881653, | |
| "learning_rate": 3.4045594916287616e-05, | |
| "loss": 1.0145, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.7322142286171063, | |
| "grad_norm": 0.44406071305274963, | |
| "learning_rate": 3.3855522157927756e-05, | |
| "loss": 1.0797, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.7330135891286971, | |
| "grad_norm": 0.4258328080177307, | |
| "learning_rate": 3.3665873286026436e-05, | |
| "loss": 0.9991, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.7338129496402878, | |
| "grad_norm": 0.381394624710083, | |
| "learning_rate": 3.3476649515947255e-05, | |
| "loss": 0.8769, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.7346123101518784, | |
| "grad_norm": 0.4326583445072174, | |
| "learning_rate": 3.3287852060329514e-05, | |
| "loss": 1.1577, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.7354116706634692, | |
| "grad_norm": 0.40116122364997864, | |
| "learning_rate": 3.309948212908052e-05, | |
| "loss": 0.9067, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7362110311750599, | |
| "grad_norm": 0.418855220079422, | |
| "learning_rate": 3.291154092936768e-05, | |
| "loss": 0.9349, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.7370103916866507, | |
| "grad_norm": 0.3745797276496887, | |
| "learning_rate": 3.2724029665611e-05, | |
| "loss": 0.9903, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.7378097521982414, | |
| "grad_norm": 0.44138798117637634, | |
| "learning_rate": 3.2536949539475193e-05, | |
| "loss": 1.0919, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.7386091127098321, | |
| "grad_norm": 0.4620357155799866, | |
| "learning_rate": 3.235030174986209e-05, | |
| "loss": 1.3411, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.7394084732214229, | |
| "grad_norm": 0.4606659412384033, | |
| "learning_rate": 3.216408749290276e-05, | |
| "loss": 1.0656, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.7402078337330136, | |
| "grad_norm": 0.5463187098503113, | |
| "learning_rate": 3.197830796195011e-05, | |
| "loss": 1.1517, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.7410071942446043, | |
| "grad_norm": 0.511232852935791, | |
| "learning_rate": 3.179296434757105e-05, | |
| "loss": 1.0564, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.7418065547561951, | |
| "grad_norm": 0.5773586630821228, | |
| "learning_rate": 3.160805783753897e-05, | |
| "loss": 0.764, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.7426059152677857, | |
| "grad_norm": 0.5721275806427002, | |
| "learning_rate": 3.1423589616825986e-05, | |
| "loss": 0.8128, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.7434052757793765, | |
| "grad_norm": 0.6489303112030029, | |
| "learning_rate": 3.123956086759553e-05, | |
| "loss": 1.2348, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7442046362909672, | |
| "grad_norm": 0.6034092903137207, | |
| "learning_rate": 3.105597276919465e-05, | |
| "loss": 1.1123, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.7450039968025579, | |
| "grad_norm": 0.6793666481971741, | |
| "learning_rate": 3.087282649814648e-05, | |
| "loss": 1.1772, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.7458033573141487, | |
| "grad_norm": 0.6495110392570496, | |
| "learning_rate": 3.069012322814272e-05, | |
| "loss": 1.0015, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.7466027178257394, | |
| "grad_norm": 0.6330723762512207, | |
| "learning_rate": 3.05078641300361e-05, | |
| "loss": 1.0119, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.7474020783373302, | |
| "grad_norm": 0.8546098470687866, | |
| "learning_rate": 3.0326050371832892e-05, | |
| "loss": 0.9274, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.7482014388489209, | |
| "grad_norm": 0.637100100517273, | |
| "learning_rate": 3.0144683118685335e-05, | |
| "loss": 0.9028, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.7490007993605116, | |
| "grad_norm": 0.7471299767494202, | |
| "learning_rate": 2.9963763532884326e-05, | |
| "loss": 1.0216, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.7498001598721024, | |
| "grad_norm": 0.6812106370925903, | |
| "learning_rate": 2.9783292773851857e-05, | |
| "loss": 1.013, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.750599520383693, | |
| "grad_norm": 0.8330495357513428, | |
| "learning_rate": 2.960327199813362e-05, | |
| "loss": 1.1521, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.750599520383693, | |
| "eval_loss": 1.0665621757507324, | |
| "eval_runtime": 40.4579, | |
| "eval_samples_per_second": 13.026, | |
| "eval_steps_per_second": 6.525, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.7513988808952837, | |
| "grad_norm": 0.926071286201477, | |
| "learning_rate": 2.9423702359391603e-05, | |
| "loss": 0.9088, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.7521982414068745, | |
| "grad_norm": 0.8907631635665894, | |
| "learning_rate": 2.9244585008396695e-05, | |
| "loss": 1.2229, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.7529976019184652, | |
| "grad_norm": 1.0290758609771729, | |
| "learning_rate": 2.9065921093021233e-05, | |
| "loss": 1.0198, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.753796962430056, | |
| "grad_norm": 0.8884785771369934, | |
| "learning_rate": 2.8887711758231783e-05, | |
| "loss": 1.0961, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.7545963229416467, | |
| "grad_norm": 1.128770112991333, | |
| "learning_rate": 2.8709958146081717e-05, | |
| "loss": 1.115, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.7553956834532374, | |
| "grad_norm": 0.9691370725631714, | |
| "learning_rate": 2.853266139570391e-05, | |
| "loss": 1.223, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.7561950439648282, | |
| "grad_norm": 1.039702296257019, | |
| "learning_rate": 2.835582264330343e-05, | |
| "loss": 1.2425, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.7569944044764189, | |
| "grad_norm": 1.0141221284866333, | |
| "learning_rate": 2.81794430221503e-05, | |
| "loss": 0.7924, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.7577937649880095, | |
| "grad_norm": 1.5116673707962036, | |
| "learning_rate": 2.8003523662572194e-05, | |
| "loss": 1.1805, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.7585931254996003, | |
| "grad_norm": 1.3761471509933472, | |
| "learning_rate": 2.7828065691947137e-05, | |
| "loss": 1.0425, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.759392486011191, | |
| "grad_norm": 1.2177839279174805, | |
| "learning_rate": 2.7653070234696434e-05, | |
| "loss": 0.7121, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7601918465227818, | |
| "grad_norm": 0.38295260071754456, | |
| "learning_rate": 2.747853841227732e-05, | |
| "loss": 0.9096, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.7609912070343725, | |
| "grad_norm": 0.35289445519447327, | |
| "learning_rate": 2.7304471343175876e-05, | |
| "loss": 0.9335, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.7617905675459632, | |
| "grad_norm": 0.4038061797618866, | |
| "learning_rate": 2.7130870142899777e-05, | |
| "loss": 1.0322, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.762589928057554, | |
| "grad_norm": 0.34192538261413574, | |
| "learning_rate": 2.6957735923971227e-05, | |
| "loss": 1.0355, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.7633892885691447, | |
| "grad_norm": 0.39813390374183655, | |
| "learning_rate": 2.6785069795919693e-05, | |
| "loss": 1.3233, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.7641886490807354, | |
| "grad_norm": 0.39333102107048035, | |
| "learning_rate": 2.661287286527495e-05, | |
| "loss": 1.075, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.7649880095923262, | |
| "grad_norm": 0.35580936074256897, | |
| "learning_rate": 2.6441146235559933e-05, | |
| "loss": 1.0142, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.7657873701039168, | |
| "grad_norm": 0.35501542687416077, | |
| "learning_rate": 2.6269891007283666e-05, | |
| "loss": 1.1377, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.7665867306155076, | |
| "grad_norm": 0.411457359790802, | |
| "learning_rate": 2.6099108277934103e-05, | |
| "loss": 1.1021, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.7673860911270983, | |
| "grad_norm": 0.3932768702507019, | |
| "learning_rate": 2.5928799141971306e-05, | |
| "loss": 1.187, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.768185451638689, | |
| "grad_norm": 0.3468436598777771, | |
| "learning_rate": 2.5758964690820255e-05, | |
| "loss": 0.9611, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.7689848121502798, | |
| "grad_norm": 0.36244887113571167, | |
| "learning_rate": 2.5589606012863963e-05, | |
| "loss": 1.0625, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.7697841726618705, | |
| "grad_norm": 0.392429918050766, | |
| "learning_rate": 2.5420724193436373e-05, | |
| "loss": 1.0414, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.7705835331734612, | |
| "grad_norm": 0.3554258644580841, | |
| "learning_rate": 2.525232031481557e-05, | |
| "loss": 1.0223, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.771382893685052, | |
| "grad_norm": 0.4706355929374695, | |
| "learning_rate": 2.5084395456216692e-05, | |
| "loss": 0.9994, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.7721822541966427, | |
| "grad_norm": 0.40463125705718994, | |
| "learning_rate": 2.4916950693785136e-05, | |
| "loss": 1.0748, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.7729816147082335, | |
| "grad_norm": 0.37813007831573486, | |
| "learning_rate": 2.4749987100589577e-05, | |
| "loss": 1.0386, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.7737809752198241, | |
| "grad_norm": 0.44974902272224426, | |
| "learning_rate": 2.458350574661511e-05, | |
| "loss": 1.0649, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.7745803357314148, | |
| "grad_norm": 0.4678323566913605, | |
| "learning_rate": 2.441750769875645e-05, | |
| "loss": 1.0105, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.7753796962430056, | |
| "grad_norm": 0.4533700942993164, | |
| "learning_rate": 2.4251994020810954e-05, | |
| "loss": 1.0592, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.7761790567545963, | |
| "grad_norm": 0.45026862621307373, | |
| "learning_rate": 2.408696577347199e-05, | |
| "loss": 1.0666, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.7769784172661871, | |
| "grad_norm": 0.4813355803489685, | |
| "learning_rate": 2.392242401432202e-05, | |
| "loss": 1.1892, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.4178970456123352, | |
| "learning_rate": 2.375836979782582e-05, | |
| "loss": 1.1072, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.7785771382893685, | |
| "grad_norm": 0.4602886736392975, | |
| "learning_rate": 2.3594804175323803e-05, | |
| "loss": 1.0448, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.7793764988009593, | |
| "grad_norm": 0.42374318838119507, | |
| "learning_rate": 2.3431728195025214e-05, | |
| "loss": 0.9626, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.78017585931255, | |
| "grad_norm": 0.4528956413269043, | |
| "learning_rate": 2.3269142902001373e-05, | |
| "loss": 0.8959, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.7809752198241406, | |
| "grad_norm": 0.7776217460632324, | |
| "learning_rate": 2.31070493381791e-05, | |
| "loss": 1.0913, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.7817745803357314, | |
| "grad_norm": 0.5880697965621948, | |
| "learning_rate": 2.2945448542333946e-05, | |
| "loss": 1.1133, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.7825739408473221, | |
| "grad_norm": 0.5221213698387146, | |
| "learning_rate": 2.2784341550083576e-05, | |
| "loss": 0.9721, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.7833733013589129, | |
| "grad_norm": 0.4688611626625061, | |
| "learning_rate": 2.262372939388111e-05, | |
| "loss": 0.9877, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.7841726618705036, | |
| "grad_norm": 0.5054891705513, | |
| "learning_rate": 2.2463613103008495e-05, | |
| "loss": 0.9356, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.7849720223820943, | |
| "grad_norm": 0.6013377904891968, | |
| "learning_rate": 2.230399370356999e-05, | |
| "loss": 0.8171, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.7857713828936851, | |
| "grad_norm": 0.5420698523521423, | |
| "learning_rate": 2.2144872218485436e-05, | |
| "loss": 1.0147, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.7865707434052758, | |
| "grad_norm": 0.6367101669311523, | |
| "learning_rate": 2.198624966748385e-05, | |
| "loss": 1.0311, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.7873701039168665, | |
| "grad_norm": 0.732600748538971, | |
| "learning_rate": 2.1828127067096826e-05, | |
| "loss": 1.1417, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.7881694644284573, | |
| "grad_norm": 0.6787117123603821, | |
| "learning_rate": 2.1670505430652044e-05, | |
| "loss": 0.8856, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.7889688249400479, | |
| "grad_norm": 0.6156232357025146, | |
| "learning_rate": 2.1513385768266715e-05, | |
| "loss": 1.0828, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.7897681854516387, | |
| "grad_norm": 0.6562130451202393, | |
| "learning_rate": 2.1356769086841233e-05, | |
| "loss": 0.8758, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.7905675459632294, | |
| "grad_norm": 0.8883634805679321, | |
| "learning_rate": 2.120065639005252e-05, | |
| "loss": 1.0898, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.7913669064748201, | |
| "grad_norm": 0.717629611492157, | |
| "learning_rate": 2.104504867834781e-05, | |
| "loss": 1.1194, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.7921662669864109, | |
| "grad_norm": 0.8948222994804382, | |
| "learning_rate": 2.0889946948938123e-05, | |
| "loss": 1.206, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.7929656274980016, | |
| "grad_norm": 0.9475370645523071, | |
| "learning_rate": 2.073535219579189e-05, | |
| "loss": 0.9441, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.7937649880095923, | |
| "grad_norm": 1.0661169290542603, | |
| "learning_rate": 2.0581265409628613e-05, | |
| "loss": 1.1246, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.7945643485211831, | |
| "grad_norm": 1.2184830904006958, | |
| "learning_rate": 2.04276875779124e-05, | |
| "loss": 1.202, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.7953637090327738, | |
| "grad_norm": 1.0761971473693848, | |
| "learning_rate": 2.027461968484583e-05, | |
| "loss": 1.2461, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.7961630695443646, | |
| "grad_norm": 1.3752635717391968, | |
| "learning_rate": 2.0122062711363532e-05, | |
| "loss": 1.3903, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.7969624300559552, | |
| "grad_norm": 1.0371414422988892, | |
| "learning_rate": 1.9970017635125836e-05, | |
| "loss": 1.096, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.7977617905675459, | |
| "grad_norm": 1.174883484840393, | |
| "learning_rate": 1.9818485430512646e-05, | |
| "loss": 1.2788, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.7985611510791367, | |
| "grad_norm": 1.1734062433242798, | |
| "learning_rate": 1.9667467068617117e-05, | |
| "loss": 1.1226, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.7993605115907274, | |
| "grad_norm": 1.8263986110687256, | |
| "learning_rate": 1.951696351723945e-05, | |
| "loss": 1.0152, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8001598721023181, | |
| "grad_norm": 0.3164405822753906, | |
| "learning_rate": 1.936697574088068e-05, | |
| "loss": 0.8081, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.8009592326139089, | |
| "grad_norm": 0.36463218927383423, | |
| "learning_rate": 1.9217504700736487e-05, | |
| "loss": 0.9044, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.8017585931254996, | |
| "grad_norm": 0.3530086576938629, | |
| "learning_rate": 1.906855135469112e-05, | |
| "loss": 0.8973, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.8025579536370904, | |
| "grad_norm": 0.35486266016960144, | |
| "learning_rate": 1.892011665731106e-05, | |
| "loss": 1.0328, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.8033573141486811, | |
| "grad_norm": 0.3966931402683258, | |
| "learning_rate": 1.877220155983914e-05, | |
| "loss": 1.1405, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.8041566746602717, | |
| "grad_norm": 0.3827740252017975, | |
| "learning_rate": 1.8624807010188306e-05, | |
| "loss": 1.1148, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.8049560351718625, | |
| "grad_norm": 0.32979387044906616, | |
| "learning_rate": 1.8477933952935567e-05, | |
| "loss": 0.9171, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.8057553956834532, | |
| "grad_norm": 0.3599773049354553, | |
| "learning_rate": 1.833158332931597e-05, | |
| "loss": 1.0972, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.806554756195044, | |
| "grad_norm": 0.32696112990379333, | |
| "learning_rate": 1.8185756077216575e-05, | |
| "loss": 0.9114, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.8073541167066347, | |
| "grad_norm": 0.37722349166870117, | |
| "learning_rate": 1.804045313117031e-05, | |
| "loss": 1.0852, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.8081534772182254, | |
| "grad_norm": 0.3368353247642517, | |
| "learning_rate": 1.78956754223502e-05, | |
| "loss": 0.9988, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.8089528377298162, | |
| "grad_norm": 0.34858715534210205, | |
| "learning_rate": 1.7751423878563268e-05, | |
| "loss": 1.0037, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.8097521982414069, | |
| "grad_norm": 0.3643121123313904, | |
| "learning_rate": 1.7607699424244585e-05, | |
| "loss": 1.0172, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.8105515587529976, | |
| "grad_norm": 0.37529024481773376, | |
| "learning_rate": 1.746450298045138e-05, | |
| "loss": 1.0152, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.8113509192645884, | |
| "grad_norm": 0.3564799726009369, | |
| "learning_rate": 1.732183546485716e-05, | |
| "loss": 0.9851, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.812150279776179, | |
| "grad_norm": 0.3498395085334778, | |
| "learning_rate": 1.7179697791745798e-05, | |
| "loss": 1.1916, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.8129496402877698, | |
| "grad_norm": 0.4197888672351837, | |
| "learning_rate": 1.7038090872005584e-05, | |
| "loss": 1.1798, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.8137490007993605, | |
| "grad_norm": 0.3844601809978485, | |
| "learning_rate": 1.689701561312359e-05, | |
| "loss": 1.0834, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.8145483613109512, | |
| "grad_norm": 0.39758336544036865, | |
| "learning_rate": 1.6756472919179688e-05, | |
| "loss": 1.0161, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.815347721822542, | |
| "grad_norm": 0.37229856848716736, | |
| "learning_rate": 1.6616463690840832e-05, | |
| "loss": 0.9845, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.8161470823341327, | |
| "grad_norm": 0.36239585280418396, | |
| "learning_rate": 1.6476988825355255e-05, | |
| "loss": 0.9263, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.8169464428457234, | |
| "grad_norm": 0.4101642370223999, | |
| "learning_rate": 1.6338049216546723e-05, | |
| "loss": 1.1672, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.8177458033573142, | |
| "grad_norm": 0.41182947158813477, | |
| "learning_rate": 1.6199645754808867e-05, | |
| "loss": 1.0586, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.8185451638689049, | |
| "grad_norm": 0.41741862893104553, | |
| "learning_rate": 1.60617793270993e-05, | |
| "loss": 1.039, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.8193445243804957, | |
| "grad_norm": 0.43758490681648254, | |
| "learning_rate": 1.5924450816934177e-05, | |
| "loss": 0.9945, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.8201438848920863, | |
| "grad_norm": 0.45736342668533325, | |
| "learning_rate": 1.5787661104382355e-05, | |
| "loss": 0.7789, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.820943245403677, | |
| "grad_norm": 0.48528802394866943, | |
| "learning_rate": 1.565141106605985e-05, | |
| "loss": 1.1153, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.8217426059152678, | |
| "grad_norm": 0.4246949255466461, | |
| "learning_rate": 1.551570157512411e-05, | |
| "loss": 1.1034, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.8225419664268585, | |
| "grad_norm": 0.43853768706321716, | |
| "learning_rate": 1.5380533501268545e-05, | |
| "loss": 0.9864, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.8233413269384492, | |
| "grad_norm": 0.4964328408241272, | |
| "learning_rate": 1.5245907710716911e-05, | |
| "loss": 1.1083, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.82414068745004, | |
| "grad_norm": 0.5204165577888489, | |
| "learning_rate": 1.5111825066217666e-05, | |
| "loss": 1.0533, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.8249400479616307, | |
| "grad_norm": 0.5626450777053833, | |
| "learning_rate": 1.4978286427038601e-05, | |
| "loss": 1.0119, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.8257394084732215, | |
| "grad_norm": 0.494026780128479, | |
| "learning_rate": 1.4845292648961218e-05, | |
| "loss": 0.9615, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.8265387689848122, | |
| "grad_norm": 0.5540620684623718, | |
| "learning_rate": 1.4712844584275276e-05, | |
| "loss": 1.0711, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.8273381294964028, | |
| "grad_norm": 0.6289163827896118, | |
| "learning_rate": 1.4580943081773346e-05, | |
| "loss": 1.1879, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.8281374900079936, | |
| "grad_norm": 0.6310924887657166, | |
| "learning_rate": 1.4449588986745365e-05, | |
| "loss": 1.0447, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.8289368505195843, | |
| "grad_norm": 0.6747533679008484, | |
| "learning_rate": 1.4318783140973212e-05, | |
| "loss": 1.1092, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.829736211031175, | |
| "grad_norm": 0.7602023482322693, | |
| "learning_rate": 1.4188526382725253e-05, | |
| "loss": 1.1756, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.8305355715427658, | |
| "grad_norm": 0.6516603827476501, | |
| "learning_rate": 1.4058819546751123e-05, | |
| "loss": 1.0085, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.8313349320543565, | |
| "grad_norm": 0.7728703022003174, | |
| "learning_rate": 1.3929663464276221e-05, | |
| "loss": 1.0542, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8321342925659473, | |
| "grad_norm": 0.8380172848701477, | |
| "learning_rate": 1.3801058962996483e-05, | |
| "loss": 1.1697, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.832933653077538, | |
| "grad_norm": 0.8363956212997437, | |
| "learning_rate": 1.3673006867073012e-05, | |
| "loss": 1.0637, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.8337330135891287, | |
| "grad_norm": 0.9202657341957092, | |
| "learning_rate": 1.354550799712685e-05, | |
| "loss": 0.917, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.8345323741007195, | |
| "grad_norm": 0.9502697587013245, | |
| "learning_rate": 1.341856317023369e-05, | |
| "loss": 1.0422, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.8353317346123101, | |
| "grad_norm": 1.006476640701294, | |
| "learning_rate": 1.3292173199918589e-05, | |
| "loss": 1.1029, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.8361310951239008, | |
| "grad_norm": 1.0370315313339233, | |
| "learning_rate": 1.3166338896150887e-05, | |
| "loss": 1.1268, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.8369304556354916, | |
| "grad_norm": 1.2824859619140625, | |
| "learning_rate": 1.30410610653389e-05, | |
| "loss": 1.3267, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.8377298161470823, | |
| "grad_norm": 1.0529234409332275, | |
| "learning_rate": 1.2916340510324797e-05, | |
| "loss": 1.2708, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.8385291766586731, | |
| "grad_norm": 1.2330455780029297, | |
| "learning_rate": 1.2792178030379454e-05, | |
| "loss": 1.2235, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.8393285371702638, | |
| "grad_norm": 1.4544402360916138, | |
| "learning_rate": 1.2668574421197343e-05, | |
| "loss": 1.324, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8401278976818545, | |
| "grad_norm": 0.4058810770511627, | |
| "learning_rate": 1.2545530474891332e-05, | |
| "loss": 1.1633, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.8409272581934453, | |
| "grad_norm": 0.31929123401641846, | |
| "learning_rate": 1.2423046979987796e-05, | |
| "loss": 0.7378, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.841726618705036, | |
| "grad_norm": 0.39207923412323, | |
| "learning_rate": 1.230112472142142e-05, | |
| "loss": 1.0527, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.8425259792166268, | |
| "grad_norm": 0.3893868625164032, | |
| "learning_rate": 1.2179764480530198e-05, | |
| "loss": 1.0162, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.8433253397282174, | |
| "grad_norm": 0.37037983536720276, | |
| "learning_rate": 1.2058967035050461e-05, | |
| "loss": 1.0753, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.8441247002398081, | |
| "grad_norm": 0.33844029903411865, | |
| "learning_rate": 1.1938733159111848e-05, | |
| "loss": 0.9185, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.8449240607513989, | |
| "grad_norm": 0.3670002520084381, | |
| "learning_rate": 1.1819063623232406e-05, | |
| "loss": 1.0727, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.8457234212629896, | |
| "grad_norm": 0.3631594479084015, | |
| "learning_rate": 1.1699959194313525e-05, | |
| "loss": 1.1735, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.8465227817745803, | |
| "grad_norm": 0.40922659635543823, | |
| "learning_rate": 1.1581420635635198e-05, | |
| "loss": 1.0855, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.8473221422861711, | |
| "grad_norm": 0.4090404808521271, | |
| "learning_rate": 1.1463448706851021e-05, | |
| "loss": 1.109, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.8481215027977618, | |
| "grad_norm": 0.37837451696395874, | |
| "learning_rate": 1.1346044163983327e-05, | |
| "loss": 1.2204, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.8489208633093526, | |
| "grad_norm": 0.3757801353931427, | |
| "learning_rate": 1.1229207759418382e-05, | |
| "loss": 1.1796, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.8497202238209433, | |
| "grad_norm": 0.3450648784637451, | |
| "learning_rate": 1.1112940241901504e-05, | |
| "loss": 1.1792, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.8505195843325339, | |
| "grad_norm": 0.4104471206665039, | |
| "learning_rate": 1.0997242356532334e-05, | |
| "loss": 1.1179, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.8513189448441247, | |
| "grad_norm": 0.36220014095306396, | |
| "learning_rate": 1.088211484476005e-05, | |
| "loss": 0.9711, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.8521183053557154, | |
| "grad_norm": 0.3745596706867218, | |
| "learning_rate": 1.0767558444378534e-05, | |
| "loss": 1.0104, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.8529176658673061, | |
| "grad_norm": 0.3766511380672455, | |
| "learning_rate": 1.065357388952174e-05, | |
| "loss": 1.096, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.8537170263788969, | |
| "grad_norm": 0.3603648245334625, | |
| "learning_rate": 1.0540161910658963e-05, | |
| "loss": 1.1295, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.8545163868904876, | |
| "grad_norm": 0.40518298745155334, | |
| "learning_rate": 1.0427323234590136e-05, | |
| "loss": 1.0624, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.8553157474020784, | |
| "grad_norm": 0.3957744240760803, | |
| "learning_rate": 1.0315058584441184e-05, | |
| "loss": 1.0235, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.8561151079136691, | |
| "grad_norm": 0.4008147418498993, | |
| "learning_rate": 1.0203368679659398e-05, | |
| "loss": 0.9525, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.8569144684252598, | |
| "grad_norm": 0.40113452076911926, | |
| "learning_rate": 1.0092254236008791e-05, | |
| "loss": 0.9687, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.8577138289368506, | |
| "grad_norm": 0.40193572640419006, | |
| "learning_rate": 9.981715965565563e-06, | |
| "loss": 0.9952, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.8585131894484412, | |
| "grad_norm": 0.422347754240036, | |
| "learning_rate": 9.87175457671351e-06, | |
| "loss": 1.0434, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.8593125499600319, | |
| "grad_norm": 0.4626252055168152, | |
| "learning_rate": 9.762370774139485e-06, | |
| "loss": 0.8937, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.8601119104716227, | |
| "grad_norm": 0.519443690776825, | |
| "learning_rate": 9.653565258828879e-06, | |
| "loss": 0.8814, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.8609112709832134, | |
| "grad_norm": 0.44369563460350037, | |
| "learning_rate": 9.545338728061126e-06, | |
| "loss": 1.1225, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.8617106314948042, | |
| "grad_norm": 0.4752413332462311, | |
| "learning_rate": 9.437691875405274e-06, | |
| "loss": 1.0668, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.8625099920063949, | |
| "grad_norm": 0.4670391082763672, | |
| "learning_rate": 9.330625390715441e-06, | |
| "loss": 0.9371, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.8633093525179856, | |
| "grad_norm": 0.6094083786010742, | |
| "learning_rate": 9.224139960126522e-06, | |
| "loss": 0.9715, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.8641087130295764, | |
| "grad_norm": 0.6373944878578186, | |
| "learning_rate": 9.118236266049707e-06, | |
| "loss": 1.0615, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.864908073541167, | |
| "grad_norm": 0.5327449440956116, | |
| "learning_rate": 9.012914987168141e-06, | |
| "loss": 1.1216, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.8657074340527577, | |
| "grad_norm": 0.6360856890678406, | |
| "learning_rate": 8.908176798432555e-06, | |
| "loss": 1.0777, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.8665067945643485, | |
| "grad_norm": 0.6309704780578613, | |
| "learning_rate": 8.804022371056975e-06, | |
| "loss": 1.1369, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.8673061550759392, | |
| "grad_norm": 0.6890113353729248, | |
| "learning_rate": 8.700452372514344e-06, | |
| "loss": 1.0562, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.86810551558753, | |
| "grad_norm": 0.6446107029914856, | |
| "learning_rate": 8.597467466532327e-06, | |
| "loss": 1.0531, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.8689048760991207, | |
| "grad_norm": 0.7217010259628296, | |
| "learning_rate": 8.49506831308905e-06, | |
| "loss": 0.9731, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.8697042366107114, | |
| "grad_norm": 0.8051683902740479, | |
| "learning_rate": 8.393255568408808e-06, | |
| "loss": 1.167, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.8705035971223022, | |
| "grad_norm": 0.73734450340271, | |
| "learning_rate": 8.29202988495793e-06, | |
| "loss": 1.1197, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.8713029576338929, | |
| "grad_norm": 0.7313120365142822, | |
| "learning_rate": 8.19139191144055e-06, | |
| "loss": 0.8269, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.8721023181454837, | |
| "grad_norm": 0.6564885377883911, | |
| "learning_rate": 8.09134229279449e-06, | |
| "loss": 0.9821, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.8729016786570744, | |
| "grad_norm": 0.7633107304573059, | |
| "learning_rate": 7.99188167018704e-06, | |
| "loss": 1.062, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.873701039168665, | |
| "grad_norm": 1.046715497970581, | |
| "learning_rate": 7.893010681010982e-06, | |
| "loss": 1.1348, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.8745003996802558, | |
| "grad_norm": 0.9617335796356201, | |
| "learning_rate": 7.794729958880408e-06, | |
| "loss": 0.9962, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.8752997601918465, | |
| "grad_norm": 0.9675642848014832, | |
| "learning_rate": 7.697040133626698e-06, | |
| "loss": 1.0002, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.8760991207034372, | |
| "grad_norm": 1.3367893695831299, | |
| "learning_rate": 7.599941831294477e-06, | |
| "loss": 1.238, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.876898481215028, | |
| "grad_norm": 1.3670591115951538, | |
| "learning_rate": 7.503435674137549e-06, | |
| "loss": 1.2431, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.8776978417266187, | |
| "grad_norm": 1.4030699729919434, | |
| "learning_rate": 7.40752228061502e-06, | |
| "loss": 1.0153, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.8784972022382095, | |
| "grad_norm": 1.2664506435394287, | |
| "learning_rate": 7.312202265387269e-06, | |
| "loss": 0.8038, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 0.8792965627498002, | |
| "grad_norm": 1.492504358291626, | |
| "learning_rate": 7.2174762393119595e-06, | |
| "loss": 0.9842, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8800959232613909, | |
| "grad_norm": 0.386832594871521, | |
| "learning_rate": 7.1233448094402355e-06, | |
| "loss": 1.0178, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.8808952837729817, | |
| "grad_norm": 0.40789809823036194, | |
| "learning_rate": 7.029808579012765e-06, | |
| "loss": 1.0509, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.8816946442845723, | |
| "grad_norm": 0.34711477160453796, | |
| "learning_rate": 6.936868147455866e-06, | |
| "loss": 0.706, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 0.882494004796163, | |
| "grad_norm": 0.3642677962779999, | |
| "learning_rate": 6.844524110377693e-06, | |
| "loss": 1.0068, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.8832933653077538, | |
| "grad_norm": 0.3810656666755676, | |
| "learning_rate": 6.75277705956443e-06, | |
| "loss": 0.8603, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.8840927258193445, | |
| "grad_norm": 0.4069780111312866, | |
| "learning_rate": 6.661627582976438e-06, | |
| "loss": 1.1551, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.8848920863309353, | |
| "grad_norm": 0.35066914558410645, | |
| "learning_rate": 6.571076264744547e-06, | |
| "loss": 0.9794, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.885691446842526, | |
| "grad_norm": 0.3382989764213562, | |
| "learning_rate": 6.48112368516629e-06, | |
| "loss": 1.0492, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.8864908073541167, | |
| "grad_norm": 0.41927698254585266, | |
| "learning_rate": 6.391770420702214e-06, | |
| "loss": 1.168, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 0.8872901678657075, | |
| "grad_norm": 0.3412163555622101, | |
| "learning_rate": 6.303017043972126e-06, | |
| "loss": 1.1376, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.8880895283772982, | |
| "grad_norm": 0.3779163658618927, | |
| "learning_rate": 6.214864123751463e-06, | |
| "loss": 0.9866, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.33037257194519043, | |
| "learning_rate": 6.127312224967652e-06, | |
| "loss": 1.0248, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.8896882494004796, | |
| "grad_norm": 0.3696483075618744, | |
| "learning_rate": 6.040361908696457e-06, | |
| "loss": 1.1099, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.8904876099120703, | |
| "grad_norm": 0.3396274745464325, | |
| "learning_rate": 5.954013732158437e-06, | |
| "loss": 1.0463, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.8912869704236611, | |
| "grad_norm": 0.37435901165008545, | |
| "learning_rate": 5.868268248715292e-06, | |
| "loss": 1.0425, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.8920863309352518, | |
| "grad_norm": 0.3372223377227783, | |
| "learning_rate": 5.783126007866435e-06, | |
| "loss": 1.1409, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.8928856914468425, | |
| "grad_norm": 0.43907099962234497, | |
| "learning_rate": 5.698587555245339e-06, | |
| "loss": 1.1308, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 0.8936850519584333, | |
| "grad_norm": 0.4471442401409149, | |
| "learning_rate": 5.6146534326161575e-06, | |
| "loss": 0.9418, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.894484412470024, | |
| "grad_norm": 0.39688199758529663, | |
| "learning_rate": 5.531324177870145e-06, | |
| "loss": 1.0384, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.8952837729816147, | |
| "grad_norm": 0.4214858114719391, | |
| "learning_rate": 5.4486003250223016e-06, | |
| "loss": 1.0617, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.8960831334932055, | |
| "grad_norm": 0.45876702666282654, | |
| "learning_rate": 5.366482404207895e-06, | |
| "loss": 1.2123, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 0.8968824940047961, | |
| "grad_norm": 0.4676870107650757, | |
| "learning_rate": 5.284970941679113e-06, | |
| "loss": 1.0053, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.8976818545163869, | |
| "grad_norm": 0.4472386837005615, | |
| "learning_rate": 5.204066459801615e-06, | |
| "loss": 0.9733, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 0.8984812150279776, | |
| "grad_norm": 0.4509924054145813, | |
| "learning_rate": 5.123769477051255e-06, | |
| "loss": 0.9867, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.8992805755395683, | |
| "grad_norm": 0.5078259110450745, | |
| "learning_rate": 5.044080508010751e-06, | |
| "loss": 1.1323, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.9000799360511591, | |
| "grad_norm": 0.44087496399879456, | |
| "learning_rate": 4.965000063366309e-06, | |
| "loss": 1.0194, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 0.9008792965627498, | |
| "grad_norm": 0.43461111187934875, | |
| "learning_rate": 4.886528649904465e-06, | |
| "loss": 0.9445, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 0.9016786570743405, | |
| "grad_norm": 0.5507033467292786, | |
| "learning_rate": 4.8086667705087515e-06, | |
| "loss": 1.2067, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 0.9024780175859313, | |
| "grad_norm": 0.5718648433685303, | |
| "learning_rate": 4.731414924156541e-06, | |
| "loss": 1.0673, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 0.903277378097522, | |
| "grad_norm": 0.4935750961303711, | |
| "learning_rate": 4.6547736059157855e-06, | |
| "loss": 0.9899, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.9040767386091128, | |
| "grad_norm": 0.531719446182251, | |
| "learning_rate": 4.5787433069418795e-06, | |
| "loss": 1.1598, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 0.9048760991207034, | |
| "grad_norm": 0.663469135761261, | |
| "learning_rate": 4.503324514474483e-06, | |
| "loss": 1.1442, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 0.9056754596322941, | |
| "grad_norm": 0.5921871066093445, | |
| "learning_rate": 4.428517711834468e-06, | |
| "loss": 1.012, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 0.9064748201438849, | |
| "grad_norm": 0.6827823519706726, | |
| "learning_rate": 4.35432337842071e-06, | |
| "loss": 0.9818, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 0.9072741806554756, | |
| "grad_norm": 0.8308103680610657, | |
| "learning_rate": 4.280741989707138e-06, | |
| "loss": 1.1532, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.9080735411670664, | |
| "grad_norm": 0.8198461532592773, | |
| "learning_rate": 4.207774017239608e-06, | |
| "loss": 1.1527, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 0.9088729016786571, | |
| "grad_norm": 0.8963656425476074, | |
| "learning_rate": 4.135419928632889e-06, | |
| "loss": 1.0863, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 0.9096722621902478, | |
| "grad_norm": 0.9033141732215881, | |
| "learning_rate": 4.0636801875676975e-06, | |
| "loss": 1.0033, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 0.9104716227018386, | |
| "grad_norm": 0.6812041997909546, | |
| "learning_rate": 3.992555253787711e-06, | |
| "loss": 0.88, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 0.9112709832134293, | |
| "grad_norm": 0.7249006032943726, | |
| "learning_rate": 3.9220455830965784e-06, | |
| "loss": 1.0334, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.9120703437250199, | |
| "grad_norm": 0.8705779910087585, | |
| "learning_rate": 3.852151627355094e-06, | |
| "loss": 0.9425, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 0.9128697042366107, | |
| "grad_norm": 0.8764076232910156, | |
| "learning_rate": 3.782873834478207e-06, | |
| "loss": 1.1731, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 0.9136690647482014, | |
| "grad_norm": 0.965621829032898, | |
| "learning_rate": 3.7142126484322025e-06, | |
| "loss": 1.0573, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 0.9144684252597922, | |
| "grad_norm": 1.0695438385009766, | |
| "learning_rate": 3.646168509231873e-06, | |
| "loss": 1.2566, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 0.9152677857713829, | |
| "grad_norm": 0.9918877482414246, | |
| "learning_rate": 3.578741852937617e-06, | |
| "loss": 0.9349, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.9160671462829736, | |
| "grad_norm": 1.1387474536895752, | |
| "learning_rate": 3.5119331116527432e-06, | |
| "loss": 1.0953, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 0.9168665067945644, | |
| "grad_norm": 1.5803228616714478, | |
| "learning_rate": 3.4457427135206167e-06, | |
| "loss": 1.1065, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 0.9176658673061551, | |
| "grad_norm": 1.388763666152954, | |
| "learning_rate": 3.3801710827219946e-06, | |
| "loss": 0.98, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 0.9184652278177458, | |
| "grad_norm": 1.2210842370986938, | |
| "learning_rate": 3.3152186394722505e-06, | |
| "loss": 1.1504, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 0.9192645883293366, | |
| "grad_norm": 1.8584920167922974, | |
| "learning_rate": 3.250885800018677e-06, | |
| "loss": 0.8716, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9200639488409272, | |
| "grad_norm": 0.33884188532829285, | |
| "learning_rate": 3.1871729766378865e-06, | |
| "loss": 0.9759, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 0.920863309352518, | |
| "grad_norm": 0.28854814171791077, | |
| "learning_rate": 3.1240805776330818e-06, | |
| "loss": 0.6448, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.9216626698641087, | |
| "grad_norm": 0.4019013047218323, | |
| "learning_rate": 3.0616090073314895e-06, | |
| "loss": 1.0665, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 0.9224620303756994, | |
| "grad_norm": 0.350145548582077, | |
| "learning_rate": 2.9997586660817643e-06, | |
| "loss": 0.9076, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 0.9232613908872902, | |
| "grad_norm": 0.36567509174346924, | |
| "learning_rate": 2.938529950251423e-06, | |
| "loss": 1.0401, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.9240607513988809, | |
| "grad_norm": 0.3485921621322632, | |
| "learning_rate": 2.8779232522243016e-06, | |
| "loss": 1.0657, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 0.9248601119104716, | |
| "grad_norm": 0.37523066997528076, | |
| "learning_rate": 2.8179389603980145e-06, | |
| "loss": 1.0506, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 0.9256594724220624, | |
| "grad_norm": 0.38018035888671875, | |
| "learning_rate": 2.7585774591815215e-06, | |
| "loss": 1.0472, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 0.926458832933653, | |
| "grad_norm": 0.3901759088039398, | |
| "learning_rate": 2.6998391289926316e-06, | |
| "loss": 1.029, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 0.9272581934452439, | |
| "grad_norm": 0.43670159578323364, | |
| "learning_rate": 2.6417243462555364e-06, | |
| "loss": 1.0349, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.9280575539568345, | |
| "grad_norm": 0.3445725440979004, | |
| "learning_rate": 2.5842334833984348e-06, | |
| "loss": 1.0316, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 0.9288569144684252, | |
| "grad_norm": 0.3813253343105316, | |
| "learning_rate": 2.52736690885117e-06, | |
| "loss": 1.1395, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 0.929656274980016, | |
| "grad_norm": 0.37068048119544983, | |
| "learning_rate": 2.4711249870428053e-06, | |
| "loss": 1.1162, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 0.9304556354916067, | |
| "grad_norm": 0.3847525417804718, | |
| "learning_rate": 2.4155080783993424e-06, | |
| "loss": 1.0191, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 0.9312549960031974, | |
| "grad_norm": 0.3729102909564972, | |
| "learning_rate": 2.360516539341384e-06, | |
| "loss": 1.091, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.9320543565147882, | |
| "grad_norm": 0.3424081802368164, | |
| "learning_rate": 2.30615072228183e-06, | |
| "loss": 1.0716, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 0.9328537170263789, | |
| "grad_norm": 0.4647541642189026, | |
| "learning_rate": 2.252410975623698e-06, | |
| "loss": 1.0196, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 0.9336530775379697, | |
| "grad_norm": 0.4275345504283905, | |
| "learning_rate": 2.1992976437577805e-06, | |
| "loss": 0.942, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 0.9344524380495604, | |
| "grad_norm": 0.38524186611175537, | |
| "learning_rate": 2.146811067060539e-06, | |
| "loss": 1.1403, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 0.935251798561151, | |
| "grad_norm": 0.46634790301322937, | |
| "learning_rate": 2.0949515818918464e-06, | |
| "loss": 0.9877, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.9360511590727418, | |
| "grad_norm": 0.43838539719581604, | |
| "learning_rate": 2.0437195205928905e-06, | |
| "loss": 1.0564, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 0.9368505195843325, | |
| "grad_norm": 0.4129061698913574, | |
| "learning_rate": 1.993115211483987e-06, | |
| "loss": 1.0135, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 0.9376498800959233, | |
| "grad_norm": 0.4249756634235382, | |
| "learning_rate": 1.9431389788625576e-06, | |
| "loss": 1.1653, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 0.938449240607514, | |
| "grad_norm": 0.4255049526691437, | |
| "learning_rate": 1.8937911430009336e-06, | |
| "loss": 1.0991, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 0.9392486011191047, | |
| "grad_norm": 0.4508858323097229, | |
| "learning_rate": 1.8450720201444116e-06, | |
| "loss": 1.1633, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.9400479616306955, | |
| "grad_norm": 0.578150749206543, | |
| "learning_rate": 1.7969819225091887e-06, | |
| "loss": 1.0231, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 0.9408473221422862, | |
| "grad_norm": 0.4697416424751282, | |
| "learning_rate": 1.7495211582803205e-06, | |
| "loss": 1.1187, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 0.9416466826538769, | |
| "grad_norm": 0.5787372589111328, | |
| "learning_rate": 1.7026900316098215e-06, | |
| "loss": 0.9238, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 0.9424460431654677, | |
| "grad_norm": 0.47707799077033997, | |
| "learning_rate": 1.6564888426146674e-06, | |
| "loss": 0.8984, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 0.9432454036770583, | |
| "grad_norm": 0.4935532212257385, | |
| "learning_rate": 1.6109178873748854e-06, | |
| "loss": 1.0022, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.9440447641886491, | |
| "grad_norm": 0.4953475892543793, | |
| "learning_rate": 1.5659774579316223e-06, | |
| "loss": 0.8219, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 0.9448441247002398, | |
| "grad_norm": 0.5493616461753845, | |
| "learning_rate": 1.521667842285346e-06, | |
| "loss": 1.0179, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 0.9456434852118305, | |
| "grad_norm": 0.6029142737388611, | |
| "learning_rate": 1.4779893243939359e-06, | |
| "loss": 1.1416, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 0.9464428457234213, | |
| "grad_norm": 0.5970425605773926, | |
| "learning_rate": 1.4349421841709066e-06, | |
| "loss": 1.0044, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 0.947242206235012, | |
| "grad_norm": 0.6485845446586609, | |
| "learning_rate": 1.3925266974835537e-06, | |
| "loss": 1.1542, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.9480415667466027, | |
| "grad_norm": 0.5779854655265808, | |
| "learning_rate": 1.3507431361512668e-06, | |
| "loss": 0.9064, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 0.9488409272581935, | |
| "grad_norm": 0.61083984375, | |
| "learning_rate": 1.3095917679437408e-06, | |
| "loss": 1.0199, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 0.9496402877697842, | |
| "grad_norm": 0.6641895174980164, | |
| "learning_rate": 1.2690728565792232e-06, | |
| "loss": 0.9776, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 0.950439648281375, | |
| "grad_norm": 0.7556144595146179, | |
| "learning_rate": 1.229186661722903e-06, | |
| "loss": 1.1204, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 0.9512390087929656, | |
| "grad_norm": 0.8415480852127075, | |
| "learning_rate": 1.1899334389852134e-06, | |
| "loss": 1.1512, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.9520383693045563, | |
| "grad_norm": 0.8689514398574829, | |
| "learning_rate": 1.1513134399201764e-06, | |
| "loss": 1.1481, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 0.9528377298161471, | |
| "grad_norm": 1.0037050247192383, | |
| "learning_rate": 1.1133269120237932e-06, | |
| "loss": 0.9543, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 0.9536370903277378, | |
| "grad_norm": 0.8194817900657654, | |
| "learning_rate": 1.0759740987325018e-06, | |
| "loss": 0.9775, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 0.9544364508393285, | |
| "grad_norm": 0.9915244579315186, | |
| "learning_rate": 1.0392552394215438e-06, | |
| "loss": 1.0258, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 0.9552358113509193, | |
| "grad_norm": 0.9382218718528748, | |
| "learning_rate": 1.0031705694035109e-06, | |
| "loss": 1.1628, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.95603517186251, | |
| "grad_norm": 1.0913877487182617, | |
| "learning_rate": 9.677203199267682e-07, | |
| "loss": 1.1345, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 0.9568345323741008, | |
| "grad_norm": 1.0560057163238525, | |
| "learning_rate": 9.329047181740103e-07, | |
| "loss": 1.0638, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 0.9576338928856915, | |
| "grad_norm": 1.40395188331604, | |
| "learning_rate": 8.987239872607855e-07, | |
| "loss": 1.164, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 0.9584332533972821, | |
| "grad_norm": 1.05940842628479, | |
| "learning_rate": 8.651783462341078e-07, | |
| "loss": 1.0904, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 0.9592326139088729, | |
| "grad_norm": 1.598750352859497, | |
| "learning_rate": 8.322680100710023e-07, | |
| "loss": 0.94, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9600319744204636, | |
| "grad_norm": 0.32953161001205444, | |
| "learning_rate": 7.999931896771284e-07, | |
| "loss": 0.8256, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 0.9608313349320543, | |
| "grad_norm": 0.367357462644577, | |
| "learning_rate": 7.683540918854593e-07, | |
| "loss": 1.0211, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 0.9616306954436451, | |
| "grad_norm": 0.40370112657546997, | |
| "learning_rate": 7.373509194549599e-07, | |
| "loss": 1.0237, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 0.9624300559552358, | |
| "grad_norm": 0.3742232322692871, | |
| "learning_rate": 7.069838710692667e-07, | |
| "loss": 0.9675, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 0.9632294164668266, | |
| "grad_norm": 0.35942748188972473, | |
| "learning_rate": 6.772531413353988e-07, | |
| "loss": 1.0492, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.9640287769784173, | |
| "grad_norm": 0.35382428765296936, | |
| "learning_rate": 6.4815892078256e-07, | |
| "loss": 1.0462, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 0.964828137490008, | |
| "grad_norm": 0.32497185468673706, | |
| "learning_rate": 6.19701395860861e-07, | |
| "loss": 0.8451, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 0.9656274980015987, | |
| "grad_norm": 0.34937870502471924, | |
| "learning_rate": 5.918807489401768e-07, | |
| "loss": 1.1667, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 0.9664268585131894, | |
| "grad_norm": 0.3648431599140167, | |
| "learning_rate": 5.646971583089355e-07, | |
| "loss": 1.0656, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 0.9672262190247801, | |
| "grad_norm": 0.36765244603157043, | |
| "learning_rate": 5.381507981730094e-07, | |
| "loss": 0.9974, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.9680255795363709, | |
| "grad_norm": 0.3716984987258911, | |
| "learning_rate": 5.122418386545591e-07, | |
| "loss": 1.0341, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 0.9688249400479616, | |
| "grad_norm": 0.3374122679233551, | |
| "learning_rate": 4.869704457910129e-07, | |
| "loss": 1.1617, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 0.9696243005595524, | |
| "grad_norm": 0.37262287735939026, | |
| "learning_rate": 4.623367815339008e-07, | |
| "loss": 1.0739, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 0.9704236610711431, | |
| "grad_norm": 0.3889469802379608, | |
| "learning_rate": 4.3834100374791075e-07, | |
| "loss": 1.0058, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 0.9712230215827338, | |
| "grad_norm": 0.38971421122550964, | |
| "learning_rate": 4.149832662098452e-07, | |
| "loss": 1.1187, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.9720223820943246, | |
| "grad_norm": 0.36595889925956726, | |
| "learning_rate": 3.9226371860758836e-07, | |
| "loss": 0.9833, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 0.9728217426059153, | |
| "grad_norm": 0.37777790427207947, | |
| "learning_rate": 3.701825065392184e-07, | |
| "loss": 1.1064, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 0.973621103117506, | |
| "grad_norm": 0.40676701068878174, | |
| "learning_rate": 3.487397715120633e-07, | |
| "loss": 1.0296, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 0.9744204636290967, | |
| "grad_norm": 0.4983866810798645, | |
| "learning_rate": 3.279356509417242e-07, | |
| "loss": 1.0814, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 0.9752198241406874, | |
| "grad_norm": 0.39095139503479004, | |
| "learning_rate": 3.0777027815128703e-07, | |
| "loss": 0.927, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.9760191846522782, | |
| "grad_norm": 0.48144182562828064, | |
| "learning_rate": 2.882437823704232e-07, | |
| "loss": 1.0725, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 0.9768185451638689, | |
| "grad_norm": 0.48106834292411804, | |
| "learning_rate": 2.6935628873453464e-07, | |
| "loss": 0.9687, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 0.9776179056754596, | |
| "grad_norm": 0.48693814873695374, | |
| "learning_rate": 2.511079182839993e-07, | |
| "loss": 1.1115, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 0.9784172661870504, | |
| "grad_norm": 0.527273952960968, | |
| "learning_rate": 2.334987879634043e-07, | |
| "loss": 1.123, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 0.9792166266986411, | |
| "grad_norm": 0.44314906001091003, | |
| "learning_rate": 2.1652901062072518e-07, | |
| "loss": 1.0677, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.9800159872102319, | |
| "grad_norm": 0.42071110010147095, | |
| "learning_rate": 2.0019869500669252e-07, | |
| "loss": 0.8542, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 0.9808153477218226, | |
| "grad_norm": 0.4927239716053009, | |
| "learning_rate": 1.845079457740151e-07, | |
| "loss": 0.9126, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 0.9816147082334132, | |
| "grad_norm": 0.4800563454627991, | |
| "learning_rate": 1.6945686347674684e-07, | |
| "loss": 1.0087, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 0.982414068745004, | |
| "grad_norm": 0.5370725989341736, | |
| "learning_rate": 1.5504554456966524e-07, | |
| "loss": 0.9965, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 0.9832134292565947, | |
| "grad_norm": 0.5823419094085693, | |
| "learning_rate": 1.4127408140761634e-07, | |
| "loss": 1.0754, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.9840127897681854, | |
| "grad_norm": 0.7353355288505554, | |
| "learning_rate": 1.2814256224493727e-07, | |
| "loss": 1.2792, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 0.9848121502797762, | |
| "grad_norm": 0.6433824896812439, | |
| "learning_rate": 1.1565107123486796e-07, | |
| "loss": 1.0551, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 0.9856115107913669, | |
| "grad_norm": 0.6252878308296204, | |
| "learning_rate": 1.0379968842905153e-07, | |
| "loss": 1.0217, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 0.9864108713029577, | |
| "grad_norm": 0.6532952785491943, | |
| "learning_rate": 9.258848977700129e-08, | |
| "loss": 1.0241, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 0.9872102318145484, | |
| "grad_norm": 0.7605010271072388, | |
| "learning_rate": 8.201754712561238e-08, | |
| "loss": 1.0708, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.988009592326139, | |
| "grad_norm": 0.7377574443817139, | |
| "learning_rate": 7.208692821870644e-08, | |
| "loss": 0.8525, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 0.9888089528377298, | |
| "grad_norm": 0.7121819853782654, | |
| "learning_rate": 6.279669669657651e-08, | |
| "loss": 0.9565, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 0.9896083133493205, | |
| "grad_norm": 0.6941244006156921, | |
| "learning_rate": 5.414691209563172e-08, | |
| "loss": 0.9999, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 0.9904076738609112, | |
| "grad_norm": 0.7420107126235962, | |
| "learning_rate": 4.613762984794212e-08, | |
| "loss": 1.0811, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 0.991207034372502, | |
| "grad_norm": 0.8043553829193115, | |
| "learning_rate": 3.876890128095001e-08, | |
| "loss": 1.1704, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.9920063948840927, | |
| "grad_norm": 1.0237336158752441, | |
| "learning_rate": 3.204077361710356e-08, | |
| "loss": 1.1427, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 0.9928057553956835, | |
| "grad_norm": 1.01031494140625, | |
| "learning_rate": 2.5953289973568162e-08, | |
| "loss": 1.1404, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 0.9936051159072742, | |
| "grad_norm": 0.8859650492668152, | |
| "learning_rate": 2.0506489361937776e-08, | |
| "loss": 1.0392, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 0.9944044764188649, | |
| "grad_norm": 0.845361053943634, | |
| "learning_rate": 1.570040668800177e-08, | |
| "loss": 0.9971, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 0.9952038369304557, | |
| "grad_norm": 0.870662271976471, | |
| "learning_rate": 1.1535072751500676e-08, | |
| "loss": 0.9663, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.9960031974420464, | |
| "grad_norm": 1.0490881204605103, | |
| "learning_rate": 8.010514245970768e-09, | |
| "loss": 1.0316, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 0.996802557953637, | |
| "grad_norm": 1.1360934972763062, | |
| "learning_rate": 5.126753758499803e-09, | |
| "loss": 1.044, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 0.9976019184652278, | |
| "grad_norm": 1.2812937498092651, | |
| "learning_rate": 2.883809769671508e-09, | |
| "loss": 1.1549, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 0.9984012789768185, | |
| "grad_norm": 1.2788760662078857, | |
| "learning_rate": 1.281696653354647e-09, | |
| "loss": 1.0516, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 0.9992006394884093, | |
| "grad_norm": 1.4483753442764282, | |
| "learning_rate": 3.204246767030128e-10, | |
| "loss": 1.1415, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.517924964427948, | |
| "learning_rate": 0.0, | |
| "loss": 1.1982, | |
| "step": 1251 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1251, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 313, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.457769976750408e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |