| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5003996802557954, | |
| "eval_steps": 313, | |
| "global_step": 626, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007993605115907274, | |
| "grad_norm": 0.9769428968429565, | |
| "learning_rate": 2e-05, | |
| "loss": 1.579, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0015987210231814548, | |
| "grad_norm": 1.4124329090118408, | |
| "learning_rate": 4e-05, | |
| "loss": 1.6108, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002398081534772182, | |
| "grad_norm": 1.4856560230255127, | |
| "learning_rate": 6e-05, | |
| "loss": 1.5688, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0031974420463629096, | |
| "grad_norm": 1.6137977838516235, | |
| "learning_rate": 8e-05, | |
| "loss": 1.7697, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.003996802557953637, | |
| "grad_norm": 3.157710313796997, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7262, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004796163069544364, | |
| "grad_norm": 1.7610689401626587, | |
| "learning_rate": 0.00012, | |
| "loss": 1.7832, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.005595523581135092, | |
| "grad_norm": 0.8947455883026123, | |
| "learning_rate": 0.00014, | |
| "loss": 1.5354, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.006394884092725819, | |
| "grad_norm": 1.3658140897750854, | |
| "learning_rate": 0.00016, | |
| "loss": 1.4666, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.007194244604316547, | |
| "grad_norm": 1.177503228187561, | |
| "learning_rate": 0.00018, | |
| "loss": 1.5789, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.007993605115907274, | |
| "grad_norm": 0.6313580870628357, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4239, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008792965627498001, | |
| "grad_norm": 0.5577117204666138, | |
| "learning_rate": 0.0001999996795753233, | |
| "loss": 1.3836, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.009592326139088728, | |
| "grad_norm": 0.3578161299228668, | |
| "learning_rate": 0.00019999871830334666, | |
| "loss": 1.4216, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.010391686650679457, | |
| "grad_norm": 0.3458471894264221, | |
| "learning_rate": 0.00019999711619023033, | |
| "loss": 1.3738, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.011191047162270184, | |
| "grad_norm": 0.3871558904647827, | |
| "learning_rate": 0.0001999948732462415, | |
| "loss": 1.6331, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.011990407673860911, | |
| "grad_norm": 0.42943254113197327, | |
| "learning_rate": 0.00019999198948575405, | |
| "loss": 1.4731, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.012789768185451638, | |
| "grad_norm": 0.5344604849815369, | |
| "learning_rate": 0.00019998846492724848, | |
| "loss": 1.4287, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.013589128697042365, | |
| "grad_norm": 0.3863997161388397, | |
| "learning_rate": 0.00019998429959331202, | |
| "loss": 1.4383, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.014388489208633094, | |
| "grad_norm": 0.3050038516521454, | |
| "learning_rate": 0.00019997949351063807, | |
| "loss": 1.397, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.015187849720223821, | |
| "grad_norm": 0.505196213722229, | |
| "learning_rate": 0.00019997404671002645, | |
| "loss": 1.264, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.01598721023181455, | |
| "grad_norm": 0.32306888699531555, | |
| "learning_rate": 0.00019996795922638293, | |
| "loss": 1.2446, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.016786570743405275, | |
| "grad_norm": 0.2961249053478241, | |
| "learning_rate": 0.00019996123109871906, | |
| "loss": 1.2698, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.017585931254996003, | |
| "grad_norm": 0.3262718617916107, | |
| "learning_rate": 0.00019995386237015207, | |
| "loss": 1.3138, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01838529176658673, | |
| "grad_norm": 0.32393717765808105, | |
| "learning_rate": 0.00019994585308790437, | |
| "loss": 1.4055, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.019184652278177457, | |
| "grad_norm": 0.38771146535873413, | |
| "learning_rate": 0.00019993720330330344, | |
| "loss": 1.1768, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.019984012789768184, | |
| "grad_norm": 0.7120429873466492, | |
| "learning_rate": 0.00019992791307178132, | |
| "loss": 1.4573, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.020783373301358914, | |
| "grad_norm": 0.45761018991470337, | |
| "learning_rate": 0.00019991798245287437, | |
| "loss": 1.1941, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.02158273381294964, | |
| "grad_norm": 0.561205267906189, | |
| "learning_rate": 0.00019990741151022301, | |
| "loss": 1.2419, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02238209432454037, | |
| "grad_norm": 0.49985525012016296, | |
| "learning_rate": 0.00019989620031157096, | |
| "loss": 1.2706, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.023181454836131096, | |
| "grad_norm": 0.48108696937561035, | |
| "learning_rate": 0.00019988434892876514, | |
| "loss": 1.2447, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.023980815347721823, | |
| "grad_norm": 0.4763461649417877, | |
| "learning_rate": 0.00019987185743775506, | |
| "loss": 1.3387, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02478017585931255, | |
| "grad_norm": 0.5324568152427673, | |
| "learning_rate": 0.0001998587259185924, | |
| "loss": 1.237, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.025579536370903277, | |
| "grad_norm": 0.5018892884254456, | |
| "learning_rate": 0.00019984495445543035, | |
| "loss": 1.1439, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.026378896882494004, | |
| "grad_norm": 0.6650824546813965, | |
| "learning_rate": 0.00019983054313652327, | |
| "loss": 1.3662, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.02717825739408473, | |
| "grad_norm": 0.5708889365196228, | |
| "learning_rate": 0.000199815492054226, | |
| "loss": 1.257, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.027977617905675458, | |
| "grad_norm": 0.5752307772636414, | |
| "learning_rate": 0.00019979980130499332, | |
| "loss": 1.2674, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02877697841726619, | |
| "grad_norm": 0.6828409433364868, | |
| "learning_rate": 0.0001997834709893793, | |
| "loss": 1.2649, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.029576338928856916, | |
| "grad_norm": 0.7169652581214905, | |
| "learning_rate": 0.00019976650121203662, | |
| "loss": 1.3819, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.030375699440447643, | |
| "grad_norm": 0.818884551525116, | |
| "learning_rate": 0.000199748892081716, | |
| "loss": 1.1481, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03117505995203837, | |
| "grad_norm": 0.7318893074989319, | |
| "learning_rate": 0.00019973064371126548, | |
| "loss": 1.3767, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0319744204636291, | |
| "grad_norm": 0.8119074106216431, | |
| "learning_rate": 0.0001997117562176296, | |
| "loss": 1.2443, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.032773780975219824, | |
| "grad_norm": 0.8081012964248657, | |
| "learning_rate": 0.00019969222972184873, | |
| "loss": 1.3334, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03357314148681055, | |
| "grad_norm": 1.2921013832092285, | |
| "learning_rate": 0.0001996720643490583, | |
| "loss": 1.3498, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03437250199840128, | |
| "grad_norm": 0.8731054067611694, | |
| "learning_rate": 0.00019965126022848795, | |
| "loss": 1.3554, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.035171862509992005, | |
| "grad_norm": 1.6653393507003784, | |
| "learning_rate": 0.00019962981749346078, | |
| "loss": 1.2505, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.03597122302158273, | |
| "grad_norm": 1.1841192245483398, | |
| "learning_rate": 0.0001996077362813924, | |
| "loss": 1.4331, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03677058353317346, | |
| "grad_norm": 1.1024802923202515, | |
| "learning_rate": 0.00019958501673379017, | |
| "loss": 1.3417, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.037569944044764186, | |
| "grad_norm": 1.1995766162872314, | |
| "learning_rate": 0.0001995616589962521, | |
| "loss": 1.3926, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03836930455635491, | |
| "grad_norm": 1.225229024887085, | |
| "learning_rate": 0.0001995376632184661, | |
| "loss": 1.4751, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.03916866506794564, | |
| "grad_norm": 1.2696462869644165, | |
| "learning_rate": 0.000199513029554209, | |
| "loss": 1.5408, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03996802557953637, | |
| "grad_norm": 1.928471326828003, | |
| "learning_rate": 0.00019948775816134545, | |
| "loss": 1.5136, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0407673860911271, | |
| "grad_norm": 0.4314931333065033, | |
| "learning_rate": 0.00019946184920182703, | |
| "loss": 0.9989, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04156674660271783, | |
| "grad_norm": 0.42213746905326843, | |
| "learning_rate": 0.00019943530284169107, | |
| "loss": 1.1271, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.042366107114308556, | |
| "grad_norm": 0.4374404549598694, | |
| "learning_rate": 0.00019940811925105983, | |
| "loss": 1.2358, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04316546762589928, | |
| "grad_norm": 0.4488702714443207, | |
| "learning_rate": 0.00019938029860413914, | |
| "loss": 1.2932, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04396482813749001, | |
| "grad_norm": 0.4582211375236511, | |
| "learning_rate": 0.00019935184107921746, | |
| "loss": 1.3682, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04476418864908074, | |
| "grad_norm": 0.3806026577949524, | |
| "learning_rate": 0.0001993227468586646, | |
| "loss": 1.1767, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.045563549160671464, | |
| "grad_norm": 0.32610875368118286, | |
| "learning_rate": 0.00019929301612893076, | |
| "loss": 1.3175, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.04636290967226219, | |
| "grad_norm": 0.36089614033699036, | |
| "learning_rate": 0.00019926264908054505, | |
| "loss": 1.2795, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.04716227018385292, | |
| "grad_norm": 0.29600343108177185, | |
| "learning_rate": 0.00019923164590811453, | |
| "loss": 1.2695, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.047961630695443645, | |
| "grad_norm": 0.32665324211120605, | |
| "learning_rate": 0.00019920000681032288, | |
| "loss": 1.2684, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04876099120703437, | |
| "grad_norm": 0.3206937611103058, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 1.3253, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0495603517186251, | |
| "grad_norm": 0.341336190700531, | |
| "learning_rate": 0.0001991348216537659, | |
| "loss": 1.3535, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.050359712230215826, | |
| "grad_norm": 0.3452065885066986, | |
| "learning_rate": 0.00019910127601273922, | |
| "loss": 1.2304, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.051159072741806554, | |
| "grad_norm": 0.3307613432407379, | |
| "learning_rate": 0.000199067095281826, | |
| "loss": 1.33, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.05195843325339728, | |
| "grad_norm": 0.45325058698654175, | |
| "learning_rate": 0.00019903227968007326, | |
| "loss": 1.1714, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05275779376498801, | |
| "grad_norm": 0.31866246461868286, | |
| "learning_rate": 0.0001989968294305965, | |
| "loss": 1.1689, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.053557154276578735, | |
| "grad_norm": 0.3664025068283081, | |
| "learning_rate": 0.00019896074476057846, | |
| "loss": 1.1676, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.05435651478816946, | |
| "grad_norm": 0.3401780128479004, | |
| "learning_rate": 0.0001989240259012675, | |
| "loss": 1.1686, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.05515587529976019, | |
| "grad_norm": 0.3274155855178833, | |
| "learning_rate": 0.0001988866730879762, | |
| "loss": 1.2787, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.055955235811350916, | |
| "grad_norm": 0.40872663259506226, | |
| "learning_rate": 0.00019884868656007982, | |
| "loss": 1.1387, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05675459632294165, | |
| "grad_norm": 0.4075217843055725, | |
| "learning_rate": 0.00019881006656101479, | |
| "loss": 1.3203, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.05755395683453238, | |
| "grad_norm": 0.39391955733299255, | |
| "learning_rate": 0.00019877081333827712, | |
| "loss": 1.2622, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.058353317346123104, | |
| "grad_norm": 0.3511233627796173, | |
| "learning_rate": 0.00019873092714342078, | |
| "loss": 1.1735, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.05915267785771383, | |
| "grad_norm": 0.3602863550186157, | |
| "learning_rate": 0.0001986904082320563, | |
| "loss": 1.2766, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.05995203836930456, | |
| "grad_norm": 0.387012243270874, | |
| "learning_rate": 0.00019864925686384873, | |
| "loss": 1.1864, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.060751398880895285, | |
| "grad_norm": 0.40913423895835876, | |
| "learning_rate": 0.00019860747330251646, | |
| "loss": 1.1349, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.06155075939248601, | |
| "grad_norm": 0.4468153715133667, | |
| "learning_rate": 0.0001985650578158291, | |
| "loss": 1.1298, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.06235011990407674, | |
| "grad_norm": 0.39224302768707275, | |
| "learning_rate": 0.00019852201067560606, | |
| "loss": 1.2709, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.06314948041566747, | |
| "grad_norm": 0.5494399070739746, | |
| "learning_rate": 0.00019847833215771467, | |
| "loss": 1.2546, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.0639488409272582, | |
| "grad_norm": 0.41739019751548767, | |
| "learning_rate": 0.0001984340225420684, | |
| "loss": 0.9764, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06474820143884892, | |
| "grad_norm": 0.5113200545310974, | |
| "learning_rate": 0.00019838908211262512, | |
| "loss": 1.1387, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.06554756195043965, | |
| "grad_norm": 0.5975989103317261, | |
| "learning_rate": 0.00019834351115738535, | |
| "loss": 1.1277, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.06634692246203037, | |
| "grad_norm": 0.637183666229248, | |
| "learning_rate": 0.0001982973099683902, | |
| "loss": 1.2837, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.0671462829736211, | |
| "grad_norm": 0.6082845330238342, | |
| "learning_rate": 0.0001982504788417197, | |
| "loss": 1.1614, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.06794564348521183, | |
| "grad_norm": 0.6118457317352295, | |
| "learning_rate": 0.00019820301807749085, | |
| "loss": 1.0804, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.06874500399680256, | |
| "grad_norm": 0.6204389333724976, | |
| "learning_rate": 0.0001981549279798556, | |
| "loss": 1.1872, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.06954436450839328, | |
| "grad_norm": 0.7375397086143494, | |
| "learning_rate": 0.00019810620885699905, | |
| "loss": 1.1549, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.07034372501998401, | |
| "grad_norm": 0.7445656657218933, | |
| "learning_rate": 0.00019805686102113747, | |
| "loss": 1.2021, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.07114308553157474, | |
| "grad_norm": 0.7308616042137146, | |
| "learning_rate": 0.000198006884788516, | |
| "loss": 1.19, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.07194244604316546, | |
| "grad_norm": 0.7465918660163879, | |
| "learning_rate": 0.00019795628047940714, | |
| "loss": 1.206, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07274180655475619, | |
| "grad_norm": 0.7157571911811829, | |
| "learning_rate": 0.00019790504841810818, | |
| "loss": 1.2446, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.07354116706634692, | |
| "grad_norm": 0.7882369756698608, | |
| "learning_rate": 0.0001978531889329395, | |
| "loss": 1.1773, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.07434052757793765, | |
| "grad_norm": 0.8234530091285706, | |
| "learning_rate": 0.00019780070235624221, | |
| "loss": 1.3671, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07513988808952837, | |
| "grad_norm": 0.8139231204986572, | |
| "learning_rate": 0.00019774758902437633, | |
| "loss": 1.0876, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0759392486011191, | |
| "grad_norm": 0.8524796366691589, | |
| "learning_rate": 0.0001976938492777182, | |
| "loss": 1.3702, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07673860911270983, | |
| "grad_norm": 1.0552854537963867, | |
| "learning_rate": 0.00019763948346065863, | |
| "loss": 1.3548, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.07753796962430055, | |
| "grad_norm": 1.154104471206665, | |
| "learning_rate": 0.00019758449192160067, | |
| "loss": 1.3052, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.07833733013589128, | |
| "grad_norm": 1.4590299129486084, | |
| "learning_rate": 0.0001975288750129572, | |
| "loss": 1.5012, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.07913669064748201, | |
| "grad_norm": 1.1523703336715698, | |
| "learning_rate": 0.00019747263309114886, | |
| "loss": 1.2774, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.07993605115907274, | |
| "grad_norm": 1.8166016340255737, | |
| "learning_rate": 0.00019741576651660157, | |
| "loss": 1.5335, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08073541167066348, | |
| "grad_norm": 0.33014848828315735, | |
| "learning_rate": 0.00019735827565374448, | |
| "loss": 0.8385, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.0815347721822542, | |
| "grad_norm": 0.47714775800704956, | |
| "learning_rate": 0.00019730016087100738, | |
| "loss": 1.0898, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.08233413269384493, | |
| "grad_norm": 0.38810276985168457, | |
| "learning_rate": 0.0001972414225408185, | |
| "loss": 1.2567, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.08313349320543566, | |
| "grad_norm": 0.3891359567642212, | |
| "learning_rate": 0.00019718206103960199, | |
| "loss": 1.3001, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.08393285371702638, | |
| "grad_norm": 0.34673169255256653, | |
| "learning_rate": 0.0001971220767477757, | |
| "loss": 0.9833, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.08473221422861711, | |
| "grad_norm": 0.3248400092124939, | |
| "learning_rate": 0.00019706147004974858, | |
| "loss": 1.0155, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.08553157474020784, | |
| "grad_norm": 0.35666921734809875, | |
| "learning_rate": 0.00019700024133391825, | |
| "loss": 1.2418, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.08633093525179857, | |
| "grad_norm": 0.3231503367424011, | |
| "learning_rate": 0.00019693839099266854, | |
| "loss": 1.3613, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.08713029576338929, | |
| "grad_norm": 0.39568766951560974, | |
| "learning_rate": 0.00019687591942236694, | |
| "loss": 1.2904, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.08792965627498002, | |
| "grad_norm": 0.3730657994747162, | |
| "learning_rate": 0.00019681282702336213, | |
| "loss": 1.1627, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08872901678657075, | |
| "grad_norm": 0.325320839881897, | |
| "learning_rate": 0.0001967491141999813, | |
| "loss": 1.2207, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.08952837729816147, | |
| "grad_norm": 0.34427809715270996, | |
| "learning_rate": 0.00019668478136052774, | |
| "loss": 1.1845, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0903277378097522, | |
| "grad_norm": 0.35617393255233765, | |
| "learning_rate": 0.000196619828917278, | |
| "loss": 1.2409, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.09112709832134293, | |
| "grad_norm": 0.27655553817749023, | |
| "learning_rate": 0.0001965542572864794, | |
| "loss": 1.2033, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.09192645883293366, | |
| "grad_norm": 0.3395769000053406, | |
| "learning_rate": 0.00019648806688834727, | |
| "loss": 1.3018, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.09272581934452438, | |
| "grad_norm": 0.3312259018421173, | |
| "learning_rate": 0.0001964212581470624, | |
| "loss": 1.1357, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.09352517985611511, | |
| "grad_norm": 0.3719313144683838, | |
| "learning_rate": 0.00019635383149076816, | |
| "loss": 1.169, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.09432454036770584, | |
| "grad_norm": 0.4446873962879181, | |
| "learning_rate": 0.00019628578735156782, | |
| "loss": 1.1771, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.09512390087929656, | |
| "grad_norm": 0.3680599331855774, | |
| "learning_rate": 0.00019621712616552182, | |
| "loss": 1.1348, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.09592326139088729, | |
| "grad_norm": 0.4281708598136902, | |
| "learning_rate": 0.00019614784837264493, | |
| "loss": 1.281, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09672262190247802, | |
| "grad_norm": 0.47409558296203613, | |
| "learning_rate": 0.00019607795441690343, | |
| "loss": 1.1092, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.09752198241406874, | |
| "grad_norm": 0.42403674125671387, | |
| "learning_rate": 0.00019600744474621232, | |
| "loss": 1.2435, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.09832134292565947, | |
| "grad_norm": 0.3994583487510681, | |
| "learning_rate": 0.0001959363198124323, | |
| "loss": 0.9298, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.0991207034372502, | |
| "grad_norm": 0.40915215015411377, | |
| "learning_rate": 0.00019586458007136713, | |
| "loss": 1.0747, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.09992006394884093, | |
| "grad_norm": 0.48302751779556274, | |
| "learning_rate": 0.00019579222598276042, | |
| "loss": 1.0207, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.10071942446043165, | |
| "grad_norm": 0.4173729121685028, | |
| "learning_rate": 0.0001957192580102929, | |
| "loss": 1.0703, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.10151878497202238, | |
| "grad_norm": 0.4759293496608734, | |
| "learning_rate": 0.0001956456766215793, | |
| "loss": 1.1198, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.10231814548361311, | |
| "grad_norm": 0.482934832572937, | |
| "learning_rate": 0.00019557148228816558, | |
| "loss": 1.0601, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.10311750599520383, | |
| "grad_norm": 0.5856558084487915, | |
| "learning_rate": 0.00019549667548552556, | |
| "loss": 1.0727, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.10391686650679456, | |
| "grad_norm": 0.7809787392616272, | |
| "learning_rate": 0.00019542125669305812, | |
| "loss": 1.3431, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10471622701838529, | |
| "grad_norm": 0.5539796948432922, | |
| "learning_rate": 0.0001953452263940842, | |
| "loss": 1.1235, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.10551558752997602, | |
| "grad_norm": 0.6920113563537598, | |
| "learning_rate": 0.00019526858507584347, | |
| "loss": 1.3767, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.10631494804156674, | |
| "grad_norm": 0.5963138937950134, | |
| "learning_rate": 0.00019519133322949122, | |
| "loss": 1.1087, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.10711430855315747, | |
| "grad_norm": 0.5328074097633362, | |
| "learning_rate": 0.00019511347135009555, | |
| "loss": 1.1354, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1079136690647482, | |
| "grad_norm": 0.6872808337211609, | |
| "learning_rate": 0.0001950349999366337, | |
| "loss": 1.254, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.10871302957633892, | |
| "grad_norm": 0.6161524653434753, | |
| "learning_rate": 0.00019495591949198927, | |
| "loss": 1.1668, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.10951239008792965, | |
| "grad_norm": 0.6678225994110107, | |
| "learning_rate": 0.00019487623052294873, | |
| "loss": 1.2988, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.11031175059952038, | |
| "grad_norm": 0.8969778418540955, | |
| "learning_rate": 0.0001947959335401984, | |
| "loss": 1.2204, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.7081554532051086, | |
| "learning_rate": 0.0001947150290583209, | |
| "loss": 1.3007, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.11191047162270183, | |
| "grad_norm": 0.7982020974159241, | |
| "learning_rate": 0.00019463351759579212, | |
| "loss": 1.2129, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11270983213429256, | |
| "grad_norm": 1.1122978925704956, | |
| "learning_rate": 0.0001945513996749777, | |
| "loss": 1.2605, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.1135091926458833, | |
| "grad_norm": 0.722235381603241, | |
| "learning_rate": 0.00019446867582212986, | |
| "loss": 1.1251, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.11430855315747403, | |
| "grad_norm": 1.081220030784607, | |
| "learning_rate": 0.00019438534656738386, | |
| "loss": 1.2666, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.11510791366906475, | |
| "grad_norm": 0.9172477722167969, | |
| "learning_rate": 0.0001943014124447547, | |
| "loss": 1.3357, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.11590727418065548, | |
| "grad_norm": 1.0490466356277466, | |
| "learning_rate": 0.0001942168739921336, | |
| "loss": 1.5132, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.11670663469224621, | |
| "grad_norm": 0.9120661020278931, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 1.4361, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.11750599520383694, | |
| "grad_norm": 1.3234212398529053, | |
| "learning_rate": 0.0001940459862678416, | |
| "loss": 1.4855, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.11830535571542766, | |
| "grad_norm": 1.1539270877838135, | |
| "learning_rate": 0.00019395963809130354, | |
| "loss": 1.3318, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.11910471622701839, | |
| "grad_norm": 1.735011100769043, | |
| "learning_rate": 0.00019387268777503236, | |
| "loss": 1.1983, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.11990407673860912, | |
| "grad_norm": 1.8227423429489136, | |
| "learning_rate": 0.00019378513587624854, | |
| "loss": 1.3119, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12070343725019984, | |
| "grad_norm": 0.38138511776924133, | |
| "learning_rate": 0.00019369698295602786, | |
| "loss": 0.972, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.12150279776179057, | |
| "grad_norm": 0.43561819195747375, | |
| "learning_rate": 0.0001936082295792978, | |
| "loss": 1.14, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.1223021582733813, | |
| "grad_norm": 0.36653220653533936, | |
| "learning_rate": 0.0001935188763148337, | |
| "loss": 1.0459, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.12310151878497202, | |
| "grad_norm": 0.42525514960289, | |
| "learning_rate": 0.00019342892373525546, | |
| "loss": 1.1425, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.12390087929656275, | |
| "grad_norm": 0.5016980767250061, | |
| "learning_rate": 0.00019333837241702358, | |
| "loss": 1.2008, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.12470023980815348, | |
| "grad_norm": 0.3720166087150574, | |
| "learning_rate": 0.00019324722294043558, | |
| "loss": 1.1646, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.1254996003197442, | |
| "grad_norm": 0.36638587713241577, | |
| "learning_rate": 0.0001931554758896223, | |
| "loss": 1.165, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.12629896083133493, | |
| "grad_norm": 0.34399259090423584, | |
| "learning_rate": 0.00019306313185254417, | |
| "loss": 1.1143, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.12709832134292565, | |
| "grad_norm": 0.3347741663455963, | |
| "learning_rate": 0.00019297019142098723, | |
| "loss": 1.3191, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.1278976818545164, | |
| "grad_norm": 0.3507041335105896, | |
| "learning_rate": 0.00019287665519055977, | |
| "loss": 1.1756, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1286970423661071, | |
| "grad_norm": 0.33033809065818787, | |
| "learning_rate": 0.00019278252376068807, | |
| "loss": 1.1667, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.12949640287769784, | |
| "grad_norm": 0.3340887725353241, | |
| "learning_rate": 0.00019268779773461275, | |
| "loss": 1.2313, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.13029576338928858, | |
| "grad_norm": 0.32075756788253784, | |
| "learning_rate": 0.000192592477719385, | |
| "loss": 1.1771, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1310951239008793, | |
| "grad_norm": 0.3003697097301483, | |
| "learning_rate": 0.00019249656432586247, | |
| "loss": 1.1086, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.13189448441247004, | |
| "grad_norm": 0.3765442669391632, | |
| "learning_rate": 0.00019240005816870554, | |
| "loss": 1.1488, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.13269384492406075, | |
| "grad_norm": 0.36099955439567566, | |
| "learning_rate": 0.0001923029598663733, | |
| "loss": 1.0357, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.1334932054356515, | |
| "grad_norm": 0.39176511764526367, | |
| "learning_rate": 0.0001922052700411196, | |
| "loss": 1.1446, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.1342925659472422, | |
| "grad_norm": 0.34224048256874084, | |
| "learning_rate": 0.00019210698931898903, | |
| "loss": 1.1375, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.13509192645883294, | |
| "grad_norm": 0.47467905282974243, | |
| "learning_rate": 0.00019200811832981298, | |
| "loss": 1.1266, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.13589128697042366, | |
| "grad_norm": 0.3941161632537842, | |
| "learning_rate": 0.00019190865770720553, | |
| "loss": 1.0618, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1366906474820144, | |
| "grad_norm": 0.4337694048881531, | |
| "learning_rate": 0.00019180860808855945, | |
| "loss": 1.2431, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.1374900079936051, | |
| "grad_norm": 0.41343578696250916, | |
| "learning_rate": 0.00019170797011504206, | |
| "loss": 1.0933, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.13828936850519585, | |
| "grad_norm": 0.4007498621940613, | |
| "learning_rate": 0.00019160674443159122, | |
| "loss": 1.0775, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.13908872901678657, | |
| "grad_norm": 0.45554855465888977, | |
| "learning_rate": 0.00019150493168691094, | |
| "loss": 1.2463, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.1398880895283773, | |
| "grad_norm": 0.4255776107311249, | |
| "learning_rate": 0.00019140253253346768, | |
| "loss": 1.1902, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.14068745003996802, | |
| "grad_norm": 0.485417902469635, | |
| "learning_rate": 0.00019129954762748568, | |
| "loss": 1.0929, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.14148681055155876, | |
| "grad_norm": 0.4176875650882721, | |
| "learning_rate": 0.00019119597762894304, | |
| "loss": 1.2162, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.14228617106314947, | |
| "grad_norm": 0.630801260471344, | |
| "learning_rate": 0.00019109182320156744, | |
| "loss": 1.1884, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.14308553157474022, | |
| "grad_norm": 0.47190389037132263, | |
| "learning_rate": 0.00019098708501283186, | |
| "loss": 1.1715, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.14388489208633093, | |
| "grad_norm": 0.48340216279029846, | |
| "learning_rate": 0.0001908817637339503, | |
| "loss": 0.9999, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14468425259792167, | |
| "grad_norm": 0.5946831107139587, | |
| "learning_rate": 0.0001907758600398735, | |
| "loss": 1.1012, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.14548361310951238, | |
| "grad_norm": 0.4977732002735138, | |
| "learning_rate": 0.0001906693746092846, | |
| "loss": 1.0698, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.14628297362110312, | |
| "grad_norm": 0.5466732382774353, | |
| "learning_rate": 0.00019056230812459474, | |
| "loss": 1.2019, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.14708233413269384, | |
| "grad_norm": 0.7127442955970764, | |
| "learning_rate": 0.0001904546612719389, | |
| "loss": 1.1404, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.14788169464428458, | |
| "grad_norm": 0.5142548680305481, | |
| "learning_rate": 0.00019034643474117113, | |
| "loss": 1.0968, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1486810551558753, | |
| "grad_norm": 0.7246764898300171, | |
| "learning_rate": 0.00019023762922586052, | |
| "loss": 1.3707, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.14948041566746603, | |
| "grad_norm": 0.6524443030357361, | |
| "learning_rate": 0.00019012824542328648, | |
| "loss": 1.0253, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.15027977617905675, | |
| "grad_norm": 0.7191776633262634, | |
| "learning_rate": 0.00019001828403443445, | |
| "loss": 1.1735, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.1510791366906475, | |
| "grad_norm": 0.7193892598152161, | |
| "learning_rate": 0.00018990774576399123, | |
| "loss": 1.2862, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.1518784972022382, | |
| "grad_norm": 0.6810536980628967, | |
| "learning_rate": 0.00018979663132034062, | |
| "loss": 1.2381, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.15267785771382894, | |
| "grad_norm": 0.8035187125205994, | |
| "learning_rate": 0.0001896849414155588, | |
| "loss": 1.1105, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.15347721822541965, | |
| "grad_norm": 0.8046578764915466, | |
| "learning_rate": 0.00018957267676540987, | |
| "loss": 1.2044, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.1542765787370104, | |
| "grad_norm": 0.8183929920196533, | |
| "learning_rate": 0.00018945983808934104, | |
| "loss": 1.2164, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.1550759392486011, | |
| "grad_norm": 0.8634499907493591, | |
| "learning_rate": 0.00018934642611047827, | |
| "loss": 1.3517, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.15587529976019185, | |
| "grad_norm": 0.8621507287025452, | |
| "learning_rate": 0.00018923244155562147, | |
| "loss": 1.1639, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.15667466027178256, | |
| "grad_norm": 1.0687577724456787, | |
| "learning_rate": 0.00018911788515523997, | |
| "loss": 1.4469, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.1574740207833733, | |
| "grad_norm": 1.067903995513916, | |
| "learning_rate": 0.00018900275764346768, | |
| "loss": 1.4959, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.15827338129496402, | |
| "grad_norm": 1.2181674242019653, | |
| "learning_rate": 0.0001888870597580985, | |
| "loss": 1.3833, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.15907274180655476, | |
| "grad_norm": 1.323157787322998, | |
| "learning_rate": 0.00018877079224058165, | |
| "loss": 1.1393, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.15987210231814547, | |
| "grad_norm": 1.3424099683761597, | |
| "learning_rate": 0.00018865395583601667, | |
| "loss": 1.0496, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1606714628297362, | |
| "grad_norm": 0.347412109375, | |
| "learning_rate": 0.000188536551293149, | |
| "loss": 1.0417, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.16147082334132695, | |
| "grad_norm": 0.41343048214912415, | |
| "learning_rate": 0.0001884185793643648, | |
| "loss": 1.4133, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.16227018385291767, | |
| "grad_norm": 0.3806878626346588, | |
| "learning_rate": 0.0001883000408056865, | |
| "loss": 1.2496, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.1630695443645084, | |
| "grad_norm": 0.43899333477020264, | |
| "learning_rate": 0.0001881809363767676, | |
| "loss": 1.2737, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.16386890487609912, | |
| "grad_norm": 0.34907570481300354, | |
| "learning_rate": 0.00018806126684088815, | |
| "loss": 1.1296, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.16466826538768986, | |
| "grad_norm": 0.4045036733150482, | |
| "learning_rate": 0.00018794103296494955, | |
| "loss": 1.0045, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.16546762589928057, | |
| "grad_norm": 0.34502413868904114, | |
| "learning_rate": 0.0001878202355194698, | |
| "loss": 1.132, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.16626698641087131, | |
| "grad_norm": 0.35705137252807617, | |
| "learning_rate": 0.0001876988752785786, | |
| "loss": 1.098, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.16706634692246203, | |
| "grad_norm": 0.3194409906864166, | |
| "learning_rate": 0.00018757695302001223, | |
| "loss": 1.2312, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.16786570743405277, | |
| "grad_norm": 0.3879018723964691, | |
| "learning_rate": 0.0001874544695251087, | |
| "loss": 1.1976, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.16866506794564348, | |
| "grad_norm": 0.35541602969169617, | |
| "learning_rate": 0.00018733142557880267, | |
| "loss": 1.1774, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.16946442845723422, | |
| "grad_norm": 0.33518531918525696, | |
| "learning_rate": 0.00018720782196962054, | |
| "loss": 1.1635, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.17026378896882494, | |
| "grad_norm": 0.3688577711582184, | |
| "learning_rate": 0.00018708365948967522, | |
| "loss": 1.2767, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.17106314948041568, | |
| "grad_norm": 0.3560887575149536, | |
| "learning_rate": 0.0001869589389346611, | |
| "loss": 1.1967, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.1718625099920064, | |
| "grad_norm": 0.33957144618034363, | |
| "learning_rate": 0.00018683366110384912, | |
| "loss": 1.3085, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.17266187050359713, | |
| "grad_norm": 0.37535560131073, | |
| "learning_rate": 0.00018670782680008143, | |
| "loss": 1.2157, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.17346123101518784, | |
| "grad_norm": 0.32397881150245667, | |
| "learning_rate": 0.00018658143682976635, | |
| "loss": 1.0392, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.17426059152677859, | |
| "grad_norm": 0.3503437638282776, | |
| "learning_rate": 0.00018645449200287314, | |
| "loss": 0.9922, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.1750599520383693, | |
| "grad_norm": 0.42683276534080505, | |
| "learning_rate": 0.000186326993132927, | |
| "loss": 1.1769, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.17585931254996004, | |
| "grad_norm": 0.39922356605529785, | |
| "learning_rate": 0.00018619894103700352, | |
| "loss": 1.1005, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.17665867306155075, | |
| "grad_norm": 0.41728365421295166, | |
| "learning_rate": 0.00018607033653572378, | |
| "loss": 1.0915, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.1774580335731415, | |
| "grad_norm": 0.39291912317276, | |
| "learning_rate": 0.00018594118045324887, | |
| "loss": 1.077, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.1782573940847322, | |
| "grad_norm": 0.4485253095626831, | |
| "learning_rate": 0.00018581147361727477, | |
| "loss": 1.2157, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.17905675459632295, | |
| "grad_norm": 0.4745635986328125, | |
| "learning_rate": 0.00018568121685902682, | |
| "loss": 1.0908, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.17985611510791366, | |
| "grad_norm": 0.44247376918792725, | |
| "learning_rate": 0.00018555041101325466, | |
| "loss": 1.1267, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1806554756195044, | |
| "grad_norm": 0.5008323788642883, | |
| "learning_rate": 0.00018541905691822668, | |
| "loss": 1.2198, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.18145483613109512, | |
| "grad_norm": 0.6248694062232971, | |
| "learning_rate": 0.00018528715541572474, | |
| "loss": 0.9965, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.18225419664268586, | |
| "grad_norm": 0.49490827322006226, | |
| "learning_rate": 0.0001851547073510388, | |
| "loss": 1.1015, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.18305355715427657, | |
| "grad_norm": 0.5184933543205261, | |
| "learning_rate": 0.00018502171357296144, | |
| "loss": 1.0303, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.1838529176658673, | |
| "grad_norm": 0.5275668501853943, | |
| "learning_rate": 0.00018488817493378235, | |
| "loss": 1.1686, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.18465227817745802, | |
| "grad_norm": 0.5399007797241211, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 1.061, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.18545163868904876, | |
| "grad_norm": 0.5668668746948242, | |
| "learning_rate": 0.00018461946649873148, | |
| "loss": 1.016, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.18625099920063948, | |
| "grad_norm": 0.5362419486045837, | |
| "learning_rate": 0.0001844842984248759, | |
| "loss": 1.1065, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.18705035971223022, | |
| "grad_norm": 0.6350337862968445, | |
| "learning_rate": 0.00018434858893394017, | |
| "loss": 1.124, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.18784972022382093, | |
| "grad_norm": 0.6309515237808228, | |
| "learning_rate": 0.00018421233889561765, | |
| "loss": 1.1816, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.18864908073541167, | |
| "grad_norm": 0.7365162372589111, | |
| "learning_rate": 0.00018407554918306582, | |
| "loss": 1.1626, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.18944844124700239, | |
| "grad_norm": 0.6940858960151672, | |
| "learning_rate": 0.0001839382206729007, | |
| "loss": 1.1132, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.19024780175859313, | |
| "grad_norm": 0.8325644731521606, | |
| "learning_rate": 0.00018380035424519115, | |
| "loss": 1.1165, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.19104716227018384, | |
| "grad_norm": 0.6918978095054626, | |
| "learning_rate": 0.00018366195078345327, | |
| "loss": 1.1739, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.19184652278177458, | |
| "grad_norm": 0.6650384068489075, | |
| "learning_rate": 0.00018352301117464473, | |
| "loss": 1.1282, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.1926458832933653, | |
| "grad_norm": 0.8843955397605896, | |
| "learning_rate": 0.00018338353630915915, | |
| "loss": 1.3603, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.19344524380495604, | |
| "grad_norm": 0.8808042407035828, | |
| "learning_rate": 0.00018324352708082032, | |
| "loss": 1.113, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.19424460431654678, | |
| "grad_norm": 1.143660068511963, | |
| "learning_rate": 0.00018310298438687644, | |
| "loss": 1.0883, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.1950439648281375, | |
| "grad_norm": 1.1465717554092407, | |
| "learning_rate": 0.00018296190912799442, | |
| "loss": 1.2652, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.19584332533972823, | |
| "grad_norm": 0.9887382984161377, | |
| "learning_rate": 0.00018282030220825423, | |
| "loss": 1.1181, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.19664268585131894, | |
| "grad_norm": 1.22044038772583, | |
| "learning_rate": 0.00018267816453514284, | |
| "loss": 1.3771, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.19744204636290968, | |
| "grad_norm": 0.9596154093742371, | |
| "learning_rate": 0.00018253549701954862, | |
| "loss": 1.1139, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.1982414068745004, | |
| "grad_norm": 1.6753774881362915, | |
| "learning_rate": 0.00018239230057575542, | |
| "loss": 1.3727, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.19904076738609114, | |
| "grad_norm": 1.4268722534179688, | |
| "learning_rate": 0.00018224857612143675, | |
| "loss": 1.2822, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.19984012789768185, | |
| "grad_norm": 2.222425937652588, | |
| "learning_rate": 0.00018210432457764982, | |
| "loss": 1.1959, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2006394884092726, | |
| "grad_norm": 0.4086918234825134, | |
| "learning_rate": 0.00018195954686882972, | |
| "loss": 1.144, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.2014388489208633, | |
| "grad_norm": 0.41542449593544006, | |
| "learning_rate": 0.00018181424392278345, | |
| "loss": 1.1824, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.20223820943245405, | |
| "grad_norm": 0.4137241542339325, | |
| "learning_rate": 0.00018166841667068405, | |
| "loss": 1.0779, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.20303756994404476, | |
| "grad_norm": 0.4028993546962738, | |
| "learning_rate": 0.00018152206604706445, | |
| "loss": 1.2201, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.2038369304556355, | |
| "grad_norm": 0.4300652742385864, | |
| "learning_rate": 0.00018137519298981172, | |
| "loss": 1.2351, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.20463629096722621, | |
| "grad_norm": 0.3383592963218689, | |
| "learning_rate": 0.00018122779844016086, | |
| "loss": 0.9793, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.20543565147881696, | |
| "grad_norm": 0.39683791995048523, | |
| "learning_rate": 0.00018107988334268896, | |
| "loss": 1.1426, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.20623501199040767, | |
| "grad_norm": 0.42270147800445557, | |
| "learning_rate": 0.0001809314486453089, | |
| "loss": 1.0328, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2070343725019984, | |
| "grad_norm": 0.3438531160354614, | |
| "learning_rate": 0.0001807824952992635, | |
| "loss": 1.1814, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.20783373301358912, | |
| "grad_norm": 0.3860226571559906, | |
| "learning_rate": 0.00018063302425911935, | |
| "loss": 1.2932, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.20863309352517986, | |
| "grad_norm": 0.33305105566978455, | |
| "learning_rate": 0.00018048303648276058, | |
| "loss": 1.2319, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.20943245403677058, | |
| "grad_norm": 0.3034221827983856, | |
| "learning_rate": 0.0001803325329313829, | |
| "loss": 1.1841, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.21023181454836132, | |
| "grad_norm": 0.33632907271385193, | |
| "learning_rate": 0.00018018151456948738, | |
| "loss": 1.0549, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.21103117505995203, | |
| "grad_norm": 0.3618718981742859, | |
| "learning_rate": 0.00018002998236487421, | |
| "loss": 1.132, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.21183053557154277, | |
| "grad_norm": 0.3930991590023041, | |
| "learning_rate": 0.00017987793728863651, | |
| "loss": 1.1224, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.21262989608313348, | |
| "grad_norm": 0.3681231439113617, | |
| "learning_rate": 0.00017972538031515417, | |
| "loss": 1.1517, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.21342925659472423, | |
| "grad_norm": 0.5605345368385315, | |
| "learning_rate": 0.0001795723124220876, | |
| "loss": 1.2004, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.21422861710631494, | |
| "grad_norm": 0.37568894028663635, | |
| "learning_rate": 0.00017941873459037142, | |
| "loss": 1.1675, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.21502797761790568, | |
| "grad_norm": 0.4226769208908081, | |
| "learning_rate": 0.0001792646478042081, | |
| "loss": 1.1093, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.2158273381294964, | |
| "grad_norm": 0.4474649131298065, | |
| "learning_rate": 0.00017911005305106187, | |
| "loss": 1.1787, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.21662669864108713, | |
| "grad_norm": 0.408610999584198, | |
| "learning_rate": 0.0001789549513216522, | |
| "loss": 1.0964, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.21742605915267785, | |
| "grad_norm": 0.4020257890224457, | |
| "learning_rate": 0.0001787993436099475, | |
| "loss": 1.0457, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.2182254196642686, | |
| "grad_norm": 0.479666143655777, | |
| "learning_rate": 0.0001786432309131588, | |
| "loss": 1.1873, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.2190247801758593, | |
| "grad_norm": 0.3813555836677551, | |
| "learning_rate": 0.00017848661423173326, | |
| "loss": 1.2765, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.21982414068745004, | |
| "grad_norm": 0.40289053320884705, | |
| "learning_rate": 0.00017832949456934798, | |
| "loss": 1.1429, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.22062350119904076, | |
| "grad_norm": 0.43944934010505676, | |
| "learning_rate": 0.00017817187293290318, | |
| "loss": 1.226, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.2214228617106315, | |
| "grad_norm": 0.6219374537467957, | |
| "learning_rate": 0.00017801375033251618, | |
| "loss": 1.2194, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.42762303352355957, | |
| "learning_rate": 0.00017785512778151457, | |
| "loss": 0.8679, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.22302158273381295, | |
| "grad_norm": 0.5274598002433777, | |
| "learning_rate": 0.00017769600629643004, | |
| "loss": 1.0911, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.22382094324540366, | |
| "grad_norm": 0.5053704977035522, | |
| "learning_rate": 0.0001775363868969915, | |
| "loss": 1.0298, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2246203037569944, | |
| "grad_norm": 0.45921552181243896, | |
| "learning_rate": 0.0001773762706061189, | |
| "loss": 1.1121, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.22541966426858512, | |
| "grad_norm": 0.6298018097877502, | |
| "learning_rate": 0.00017721565844991643, | |
| "loss": 1.1947, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.22621902478017586, | |
| "grad_norm": 0.5267609357833862, | |
| "learning_rate": 0.00017705455145766607, | |
| "loss": 1.202, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.2270183852917666, | |
| "grad_norm": 0.5780994892120361, | |
| "learning_rate": 0.0001768929506618209, | |
| "loss": 0.9595, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.2278177458033573, | |
| "grad_norm": 0.6282961964607239, | |
| "learning_rate": 0.00017673085709799866, | |
| "loss": 1.1487, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.22861710631494805, | |
| "grad_norm": 0.583211362361908, | |
| "learning_rate": 0.00017656827180497482, | |
| "loss": 1.051, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.22941646682653877, | |
| "grad_norm": 0.5974562168121338, | |
| "learning_rate": 0.0001764051958246762, | |
| "loss": 1.0361, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.2302158273381295, | |
| "grad_norm": 0.7136018872261047, | |
| "learning_rate": 0.0001762416302021742, | |
| "loss": 1.1316, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.23101518784972022, | |
| "grad_norm": 0.6994486451148987, | |
| "learning_rate": 0.000176077575985678, | |
| "loss": 1.2776, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.23181454836131096, | |
| "grad_norm": 0.6960222125053406, | |
| "learning_rate": 0.00017591303422652803, | |
| "loss": 1.3089, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.23261390887290168, | |
| "grad_norm": 0.7990922927856445, | |
| "learning_rate": 0.00017574800597918908, | |
| "loss": 1.2942, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.23341326938449242, | |
| "grad_norm": 0.8984624147415161, | |
| "learning_rate": 0.0001755824923012436, | |
| "loss": 1.0605, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.23421262989608313, | |
| "grad_norm": 0.7995984554290771, | |
| "learning_rate": 0.0001754164942533849, | |
| "loss": 1.3299, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.23501199040767387, | |
| "grad_norm": 0.9280282258987427, | |
| "learning_rate": 0.00017525001289941044, | |
| "loss": 1.3762, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.23581135091926458, | |
| "grad_norm": 0.909788966178894, | |
| "learning_rate": 0.00017508304930621488, | |
| "loss": 1.1487, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.23661071143085532, | |
| "grad_norm": 0.939471423625946, | |
| "learning_rate": 0.00017491560454378332, | |
| "loss": 1.0599, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.23741007194244604, | |
| "grad_norm": 0.9314141869544983, | |
| "learning_rate": 0.00017474767968518446, | |
| "loss": 1.3278, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.23820943245403678, | |
| "grad_norm": 1.094929575920105, | |
| "learning_rate": 0.00017457927580656366, | |
| "loss": 1.2823, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.2390087929656275, | |
| "grad_norm": 1.1693792343139648, | |
| "learning_rate": 0.00017441039398713608, | |
| "loss": 1.3312, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.23980815347721823, | |
| "grad_norm": 1.5119585990905762, | |
| "learning_rate": 0.00017424103530917976, | |
| "loss": 1.4299, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24060751398880895, | |
| "grad_norm": 0.3461761176586151, | |
| "learning_rate": 0.00017407120085802873, | |
| "loss": 1.0737, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.2414068745003997, | |
| "grad_norm": 0.40149208903312683, | |
| "learning_rate": 0.00017390089172206592, | |
| "loss": 1.0599, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.2422062350119904, | |
| "grad_norm": 0.34392961859703064, | |
| "learning_rate": 0.00017373010899271636, | |
| "loss": 1.1225, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.24300559552358114, | |
| "grad_norm": 0.35481372475624084, | |
| "learning_rate": 0.00017355885376444007, | |
| "loss": 1.0724, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.24380495603517185, | |
| "grad_norm": 0.395998477935791, | |
| "learning_rate": 0.00017338712713472508, | |
| "loss": 1.079, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.2446043165467626, | |
| "grad_norm": 0.3567344546318054, | |
| "learning_rate": 0.00017321493020408033, | |
| "loss": 1.2561, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.2454036770583533, | |
| "grad_norm": 0.3671879768371582, | |
| "learning_rate": 0.0001730422640760288, | |
| "loss": 1.2168, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.24620303756994405, | |
| "grad_norm": 0.3822752833366394, | |
| "learning_rate": 0.00017286912985710023, | |
| "loss": 1.1955, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.24700239808153476, | |
| "grad_norm": 0.35947856307029724, | |
| "learning_rate": 0.00017269552865682413, | |
| "loss": 1.1131, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.2478017585931255, | |
| "grad_norm": 0.32692649960517883, | |
| "learning_rate": 0.00017252146158772269, | |
| "loss": 1.1497, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.24860111910471622, | |
| "grad_norm": 0.3335247337818146, | |
| "learning_rate": 0.00017234692976530357, | |
| "loss": 1.2148, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.24940047961630696, | |
| "grad_norm": 0.3448440134525299, | |
| "learning_rate": 0.00017217193430805287, | |
| "loss": 1.1969, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.2501998401278977, | |
| "grad_norm": 0.31313541531562805, | |
| "learning_rate": 0.00017199647633742783, | |
| "loss": 1.1303, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2501998401278977, | |
| "eval_loss": 1.141908884048462, | |
| "eval_runtime": 39.7956, | |
| "eval_samples_per_second": 13.243, | |
| "eval_steps_per_second": 6.634, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2509992006394884, | |
| "grad_norm": 0.3559674322605133, | |
| "learning_rate": 0.0001718205569778497, | |
| "loss": 1.1891, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.2517985611510791, | |
| "grad_norm": 0.40814101696014404, | |
| "learning_rate": 0.0001716441773566966, | |
| "loss": 1.1652, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.25259792166266987, | |
| "grad_norm": 0.3883415162563324, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 1.1556, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2533972821742606, | |
| "grad_norm": 0.37391719222068787, | |
| "learning_rate": 0.0001712900418539183, | |
| "loss": 1.1343, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2541966426858513, | |
| "grad_norm": 0.36951127648353577, | |
| "learning_rate": 0.00017111228824176825, | |
| "loss": 1.1139, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.25499600319744203, | |
| "grad_norm": 0.3808392584323883, | |
| "learning_rate": 0.0001709340789069788, | |
| "loss": 1.1057, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.2557953637090328, | |
| "grad_norm": 0.4488889276981354, | |
| "learning_rate": 0.00017075541499160334, | |
| "loss": 1.1723, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2565947242206235, | |
| "grad_norm": 0.4316102862358093, | |
| "learning_rate": 0.0001705762976406084, | |
| "loss": 1.164, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.2573940847322142, | |
| "grad_norm": 0.4768711030483246, | |
| "learning_rate": 0.00017039672800186638, | |
| "loss": 1.0935, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.25819344524380494, | |
| "grad_norm": 0.4212109446525574, | |
| "learning_rate": 0.00017021670722614817, | |
| "loss": 1.0126, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.2589928057553957, | |
| "grad_norm": 0.4388445019721985, | |
| "learning_rate": 0.0001700362364671157, | |
| "loss": 1.2362, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.2597921662669864, | |
| "grad_norm": 0.4593929350376129, | |
| "learning_rate": 0.0001698553168813147, | |
| "loss": 1.323, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.26059152677857716, | |
| "grad_norm": 0.507211446762085, | |
| "learning_rate": 0.00016967394962816713, | |
| "loss": 1.2032, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.26139088729016785, | |
| "grad_norm": 0.4374110698699951, | |
| "learning_rate": 0.0001694921358699639, | |
| "loss": 0.8798, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.2621902478017586, | |
| "grad_norm": 0.558111310005188, | |
| "learning_rate": 0.00016930987677185727, | |
| "loss": 1.2127, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.26298960831334933, | |
| "grad_norm": 0.4802536070346832, | |
| "learning_rate": 0.00016912717350185355, | |
| "loss": 1.1708, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.2637889688249401, | |
| "grad_norm": 0.572343111038208, | |
| "learning_rate": 0.00016894402723080537, | |
| "loss": 1.2319, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.26458832933653076, | |
| "grad_norm": 0.5261095762252808, | |
| "learning_rate": 0.00016876043913240452, | |
| "loss": 1.1315, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.2653876898481215, | |
| "grad_norm": 0.5496971607208252, | |
| "learning_rate": 0.00016857641038317404, | |
| "loss": 1.1368, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.26618705035971224, | |
| "grad_norm": 0.5394718647003174, | |
| "learning_rate": 0.00016839194216246108, | |
| "loss": 1.0986, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.266986410871303, | |
| "grad_norm": 0.514948844909668, | |
| "learning_rate": 0.00016820703565242898, | |
| "loss": 1.0841, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.26778577138289367, | |
| "grad_norm": 0.6216516494750977, | |
| "learning_rate": 0.0001680216920380499, | |
| "loss": 0.8411, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.2685851318944844, | |
| "grad_norm": 0.6901413798332214, | |
| "learning_rate": 0.00016783591250709728, | |
| "loss": 1.0341, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.26938449240607515, | |
| "grad_norm": 0.6468019485473633, | |
| "learning_rate": 0.00016764969825013793, | |
| "loss": 0.8504, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.2701838529176659, | |
| "grad_norm": 0.8816102743148804, | |
| "learning_rate": 0.0001674630504605248, | |
| "loss": 1.1692, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.2709832134292566, | |
| "grad_norm": 0.677508533000946, | |
| "learning_rate": 0.00016727597033438902, | |
| "loss": 1.0148, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.2717825739408473, | |
| "grad_norm": 0.8252014517784119, | |
| "learning_rate": 0.00016708845907063236, | |
| "loss": 1.1856, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.27258193445243806, | |
| "grad_norm": 0.8761686682701111, | |
| "learning_rate": 0.0001669005178709195, | |
| "loss": 1.168, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.2733812949640288, | |
| "grad_norm": 0.9532282948493958, | |
| "learning_rate": 0.00016671214793967048, | |
| "loss": 1.1328, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.2741806554756195, | |
| "grad_norm": 1.1391276121139526, | |
| "learning_rate": 0.00016652335048405274, | |
| "loss": 1.1273, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.2749800159872102, | |
| "grad_norm": 0.8748697638511658, | |
| "learning_rate": 0.00016633412671397357, | |
| "loss": 0.9519, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.27577937649880097, | |
| "grad_norm": 0.9585081934928894, | |
| "learning_rate": 0.00016614447784207227, | |
| "loss": 1.2655, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.2765787370103917, | |
| "grad_norm": 1.1829018592834473, | |
| "learning_rate": 0.00016595440508371237, | |
| "loss": 1.4466, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.2773780975219824, | |
| "grad_norm": 1.3926594257354736, | |
| "learning_rate": 0.00016576390965697405, | |
| "loss": 0.9497, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.27817745803357313, | |
| "grad_norm": 1.3197882175445557, | |
| "learning_rate": 0.00016557299278264584, | |
| "loss": 1.625, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.2789768185451639, | |
| "grad_norm": 1.5563266277313232, | |
| "learning_rate": 0.0001653816556842174, | |
| "loss": 1.3074, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.2797761790567546, | |
| "grad_norm": 2.229512929916382, | |
| "learning_rate": 0.00016518989958787126, | |
| "loss": 0.8772, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2805755395683453, | |
| "grad_norm": 0.32201531529426575, | |
| "learning_rate": 0.00016499772572247515, | |
| "loss": 0.8794, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.28137490007993604, | |
| "grad_norm": 0.3720848560333252, | |
| "learning_rate": 0.000164805135319574, | |
| "loss": 1.0193, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.2821742605915268, | |
| "grad_norm": 0.44173169136047363, | |
| "learning_rate": 0.0001646121296133822, | |
| "loss": 1.1591, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.2829736211031175, | |
| "grad_norm": 0.3597632646560669, | |
| "learning_rate": 0.00016441870984077554, | |
| "loss": 1.0938, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.2837729816147082, | |
| "grad_norm": 0.4087126851081848, | |
| "learning_rate": 0.00016422487724128342, | |
| "loss": 1.0694, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.28457234212629895, | |
| "grad_norm": 0.528968095779419, | |
| "learning_rate": 0.00016403063305708078, | |
| "loss": 1.024, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.2853717026378897, | |
| "grad_norm": 0.46297335624694824, | |
| "learning_rate": 0.0001638359785329802, | |
| "loss": 1.1045, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.28617106314948043, | |
| "grad_norm": 0.3513520359992981, | |
| "learning_rate": 0.00016364091491642398, | |
| "loss": 1.204, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.2869704236610711, | |
| "grad_norm": 0.3250669240951538, | |
| "learning_rate": 0.000163445443457476, | |
| "loss": 1.0709, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.28776978417266186, | |
| "grad_norm": 0.3269745707511902, | |
| "learning_rate": 0.00016324956540881384, | |
| "loss": 1.0644, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2885691446842526, | |
| "grad_norm": 0.30179765820503235, | |
| "learning_rate": 0.00016305328202572076, | |
| "loss": 1.0858, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.28936850519584334, | |
| "grad_norm": 0.35633420944213867, | |
| "learning_rate": 0.0001628565945660775, | |
| "loss": 1.0879, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.290167865707434, | |
| "grad_norm": 0.3494223356246948, | |
| "learning_rate": 0.00016265950429035443, | |
| "loss": 1.0941, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.29096722621902477, | |
| "grad_norm": 0.3416435420513153, | |
| "learning_rate": 0.00016246201246160327, | |
| "loss": 1.185, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.2917665867306155, | |
| "grad_norm": 0.4749598503112793, | |
| "learning_rate": 0.00016226412034544912, | |
| "loss": 1.1142, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.29256594724220625, | |
| "grad_norm": 0.3291066884994507, | |
| "learning_rate": 0.00016206582921008233, | |
| "loss": 1.0666, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.293365307753797, | |
| "grad_norm": 0.3534018099308014, | |
| "learning_rate": 0.00016186714032625035, | |
| "loss": 1.041, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.2941646682653877, | |
| "grad_norm": 0.44766634702682495, | |
| "learning_rate": 0.00016166805496724965, | |
| "loss": 1.124, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.2949640287769784, | |
| "grad_norm": 0.43692246079444885, | |
| "learning_rate": 0.00016146857440891744, | |
| "loss": 1.1488, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.29576338928856916, | |
| "grad_norm": 0.546944260597229, | |
| "learning_rate": 0.00016126869992962357, | |
| "loss": 1.0826, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2965627498001599, | |
| "grad_norm": 0.3757582902908325, | |
| "learning_rate": 0.00016106843281026237, | |
| "loss": 1.2105, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.2973621103117506, | |
| "grad_norm": 0.4158326983451843, | |
| "learning_rate": 0.00016086777433424435, | |
| "loss": 1.0807, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.2981614708233413, | |
| "grad_norm": 0.41738831996917725, | |
| "learning_rate": 0.00016066672578748814, | |
| "loss": 1.0334, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.29896083133493206, | |
| "grad_norm": 0.4042329490184784, | |
| "learning_rate": 0.00016046528845841196, | |
| "loss": 1.0007, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.2997601918465228, | |
| "grad_norm": 0.43798747658729553, | |
| "learning_rate": 0.00016026346363792567, | |
| "loss": 0.8887, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3005595523581135, | |
| "grad_norm": 0.45503515005111694, | |
| "learning_rate": 0.00016006125261942229, | |
| "loss": 1.002, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.30135891286970423, | |
| "grad_norm": 0.44098415970802307, | |
| "learning_rate": 0.00015985865669876988, | |
| "loss": 0.9226, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.302158273381295, | |
| "grad_norm": 0.5401691198348999, | |
| "learning_rate": 0.00015965567717430304, | |
| "loss": 1.0046, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.3029576338928857, | |
| "grad_norm": 0.4791525602340698, | |
| "learning_rate": 0.00015945231534681483, | |
| "loss": 1.0362, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.3037569944044764, | |
| "grad_norm": 0.5511587858200073, | |
| "learning_rate": 0.00015924857251954806, | |
| "loss": 1.0618, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.30455635491606714, | |
| "grad_norm": 0.5481637716293335, | |
| "learning_rate": 0.00015904444999818745, | |
| "loss": 1.0005, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3053557154276579, | |
| "grad_norm": 0.6237247586250305, | |
| "learning_rate": 0.0001588399490908508, | |
| "loss": 0.885, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.3061550759392486, | |
| "grad_norm": 0.7181023955345154, | |
| "learning_rate": 0.00015863507110808086, | |
| "loss": 1.4557, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3069544364508393, | |
| "grad_norm": 0.5080341100692749, | |
| "learning_rate": 0.00015842981736283686, | |
| "loss": 0.8882, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.30775379696243005, | |
| "grad_norm": 0.6519491672515869, | |
| "learning_rate": 0.0001582241891704861, | |
| "loss": 1.1977, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3085531574740208, | |
| "grad_norm": 0.6563478708267212, | |
| "learning_rate": 0.0001580181878487955, | |
| "loss": 1.0473, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.30935251798561153, | |
| "grad_norm": 0.6795956492424011, | |
| "learning_rate": 0.00015781181471792322, | |
| "loss": 1.1955, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.3101518784972022, | |
| "grad_norm": 0.6601610779762268, | |
| "learning_rate": 0.00015760507110041015, | |
| "loss": 1.0557, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.31095123900879296, | |
| "grad_norm": 0.9046745896339417, | |
| "learning_rate": 0.00015739795832117143, | |
| "loss": 1.1948, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3117505995203837, | |
| "grad_norm": 0.8437576293945312, | |
| "learning_rate": 0.00015719047770748788, | |
| "loss": 0.8749, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.31254996003197444, | |
| "grad_norm": 1.1434983015060425, | |
| "learning_rate": 0.00015698263058899778, | |
| "loss": 1.158, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.3133493205435651, | |
| "grad_norm": 0.8605363965034485, | |
| "learning_rate": 0.000156774418297688, | |
| "loss": 1.2285, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.31414868105515587, | |
| "grad_norm": 0.8296773433685303, | |
| "learning_rate": 0.00015656584216788567, | |
| "loss": 1.2332, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.3149480415667466, | |
| "grad_norm": 1.2430847883224487, | |
| "learning_rate": 0.00015635690353624955, | |
| "loss": 1.1903, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.31574740207833735, | |
| "grad_norm": 1.0443687438964844, | |
| "learning_rate": 0.0001561476037417615, | |
| "loss": 1.311, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.31654676258992803, | |
| "grad_norm": 1.023246169090271, | |
| "learning_rate": 0.00015593794412571796, | |
| "loss": 1.3116, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.3173461231015188, | |
| "grad_norm": 1.2342387437820435, | |
| "learning_rate": 0.00015572792603172115, | |
| "loss": 1.1111, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.3181454836131095, | |
| "grad_norm": 1.0979640483856201, | |
| "learning_rate": 0.00015551755080567075, | |
| "loss": 1.2079, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.31894484412470026, | |
| "grad_norm": 1.2995789051055908, | |
| "learning_rate": 0.00015530681979575496, | |
| "loss": 1.2902, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.31974420463629094, | |
| "grad_norm": 1.4143871068954468, | |
| "learning_rate": 0.00015509573435244214, | |
| "loss": 1.1578, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3205435651478817, | |
| "grad_norm": 0.37532997131347656, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.8548, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.3213429256594724, | |
| "grad_norm": 0.4204126000404358, | |
| "learning_rate": 0.00015467250557884684, | |
| "loss": 1.0921, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.32214228617106316, | |
| "grad_norm": 0.38710692524909973, | |
| "learning_rate": 0.0001544603649608232, | |
| "loss": 1.1638, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.3229416466826539, | |
| "grad_norm": 0.40149056911468506, | |
| "learning_rate": 0.0001542478753339029, | |
| "loss": 1.0853, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.3237410071942446, | |
| "grad_norm": 0.4281335175037384, | |
| "learning_rate": 0.0001540350380598243, | |
| "loss": 0.9679, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.32454036770583533, | |
| "grad_norm": 0.3761771023273468, | |
| "learning_rate": 0.00015382185450255365, | |
| "loss": 1.1288, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.32533972821742607, | |
| "grad_norm": 0.4560992121696472, | |
| "learning_rate": 0.00015360832602827644, | |
| "loss": 1.12, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.3261390887290168, | |
| "grad_norm": 0.35900700092315674, | |
| "learning_rate": 0.00015339445400538852, | |
| "loss": 1.0102, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.3269384492406075, | |
| "grad_norm": 0.3479507267475128, | |
| "learning_rate": 0.00015318023980448735, | |
| "loss": 1.2253, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.32773780975219824, | |
| "grad_norm": 0.3966832458972931, | |
| "learning_rate": 0.00015296568479836325, | |
| "loss": 1.2279, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.328537170263789, | |
| "grad_norm": 0.3351828157901764, | |
| "learning_rate": 0.00015275079036199062, | |
| "loss": 1.161, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.3293365307753797, | |
| "grad_norm": 0.3129248023033142, | |
| "learning_rate": 0.000152535557872519, | |
| "loss": 1.0208, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.3301358912869704, | |
| "grad_norm": 0.3181811273097992, | |
| "learning_rate": 0.00015231998870926446, | |
| "loss": 1.1992, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.33093525179856115, | |
| "grad_norm": 0.32170534133911133, | |
| "learning_rate": 0.0001521040842537006, | |
| "loss": 1.1188, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.3317346123101519, | |
| "grad_norm": 0.348155677318573, | |
| "learning_rate": 0.0001518878458894497, | |
| "loss": 1.1693, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.33253397282174263, | |
| "grad_norm": 0.468597948551178, | |
| "learning_rate": 0.00015167127500227397, | |
| "loss": 1.041, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.35079798102378845, | |
| "learning_rate": 0.0001514543729800664, | |
| "loss": 0.8822, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.33413269384492406, | |
| "grad_norm": 0.36019131541252136, | |
| "learning_rate": 0.0001512371412128424, | |
| "loss": 0.9667, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.3349320543565148, | |
| "grad_norm": 0.37374162673950195, | |
| "learning_rate": 0.00015101958109273015, | |
| "loss": 1.0106, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.33573141486810554, | |
| "grad_norm": 0.39020803570747375, | |
| "learning_rate": 0.0001508016940139624, | |
| "loss": 1.1184, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3365307753796962, | |
| "grad_norm": 0.4587428569793701, | |
| "learning_rate": 0.00015058348137286703, | |
| "loss": 1.3182, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.33733013589128696, | |
| "grad_norm": 0.40048739314079285, | |
| "learning_rate": 0.00015036494456785836, | |
| "loss": 1.1058, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.3381294964028777, | |
| "grad_norm": 0.4465924799442291, | |
| "learning_rate": 0.00015014608499942808, | |
| "loss": 1.0426, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.33892885691446845, | |
| "grad_norm": 0.39607423543930054, | |
| "learning_rate": 0.00014992690407013634, | |
| "loss": 0.9871, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.33972821742605913, | |
| "grad_norm": 0.3925151824951172, | |
| "learning_rate": 0.0001497074031846027, | |
| "loss": 1.0183, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.3405275779376499, | |
| "grad_norm": 0.40714192390441895, | |
| "learning_rate": 0.00014948758374949713, | |
| "loss": 1.1662, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.3413269384492406, | |
| "grad_norm": 0.5418148636817932, | |
| "learning_rate": 0.0001492674471735311, | |
| "loss": 0.9804, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.34212629896083135, | |
| "grad_norm": 0.5795961022377014, | |
| "learning_rate": 0.00014904699486744847, | |
| "loss": 1.2834, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.34292565947242204, | |
| "grad_norm": 0.4905664920806885, | |
| "learning_rate": 0.0001488262282440163, | |
| "loss": 1.0091, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.3437250199840128, | |
| "grad_norm": 0.5001060366630554, | |
| "learning_rate": 0.00014860514871801618, | |
| "loss": 1.1577, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3445243804956035, | |
| "grad_norm": 0.7573508620262146, | |
| "learning_rate": 0.0001483837577062348, | |
| "loss": 0.9795, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.34532374100719426, | |
| "grad_norm": 0.4972291886806488, | |
| "learning_rate": 0.00014816205662745497, | |
| "loss": 1.1244, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.34612310151878495, | |
| "grad_norm": 0.5998781323432922, | |
| "learning_rate": 0.00014794004690244669, | |
| "loss": 1.1667, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.3469224620303757, | |
| "grad_norm": 0.6669907569885254, | |
| "learning_rate": 0.00014771772995395777, | |
| "loss": 1.1025, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.34772182254196643, | |
| "grad_norm": 0.6359307169914246, | |
| "learning_rate": 0.00014749510720670506, | |
| "loss": 0.9878, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.34852118305355717, | |
| "grad_norm": 0.7467820048332214, | |
| "learning_rate": 0.00014727218008736486, | |
| "loss": 1.1126, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.34932054356514786, | |
| "grad_norm": 0.6224162578582764, | |
| "learning_rate": 0.00014704895002456424, | |
| "loss": 1.0024, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.3501199040767386, | |
| "grad_norm": 0.70039963722229, | |
| "learning_rate": 0.00014682541844887164, | |
| "loss": 1.219, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.35091926458832934, | |
| "grad_norm": 0.7462148070335388, | |
| "learning_rate": 0.00014660158679278765, | |
| "loss": 1.106, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.3517186250999201, | |
| "grad_norm": 0.7545698881149292, | |
| "learning_rate": 0.00014637745649073602, | |
| "loss": 1.2432, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.35251798561151076, | |
| "grad_norm": 0.7020461559295654, | |
| "learning_rate": 0.00014615302897905434, | |
| "loss": 1.1659, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.3533173461231015, | |
| "grad_norm": 0.7818939685821533, | |
| "learning_rate": 0.0001459283056959849, | |
| "loss": 0.9831, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.35411670663469225, | |
| "grad_norm": 0.8227357268333435, | |
| "learning_rate": 0.00014570328808166539, | |
| "loss": 1.049, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.354916067146283, | |
| "grad_norm": 0.8649093508720398, | |
| "learning_rate": 0.0001454779775781197, | |
| "loss": 1.3937, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.35571542765787373, | |
| "grad_norm": 0.8531369566917419, | |
| "learning_rate": 0.00014525237562924877, | |
| "loss": 1.2183, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.3565147881694644, | |
| "grad_norm": 1.2276400327682495, | |
| "learning_rate": 0.00014502648368082123, | |
| "loss": 1.4021, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.35731414868105515, | |
| "grad_norm": 1.1620177030563354, | |
| "learning_rate": 0.00014480030318046416, | |
| "loss": 1.3332, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.3581135091926459, | |
| "grad_norm": 1.2599079608917236, | |
| "learning_rate": 0.00014457383557765386, | |
| "loss": 1.4229, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.35891286970423664, | |
| "grad_norm": 1.5748292207717896, | |
| "learning_rate": 0.0001443470823237064, | |
| "loss": 0.9769, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.3597122302158273, | |
| "grad_norm": 1.4127042293548584, | |
| "learning_rate": 0.0001441200448717687, | |
| "loss": 0.8914, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.36051159072741806, | |
| "grad_norm": 0.3052649199962616, | |
| "learning_rate": 0.0001438927246768087, | |
| "loss": 0.8243, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.3613109512390088, | |
| "grad_norm": 0.38650140166282654, | |
| "learning_rate": 0.0001436651231956064, | |
| "loss": 1.1737, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.36211031175059955, | |
| "grad_norm": 0.3861415982246399, | |
| "learning_rate": 0.00014343724188674447, | |
| "loss": 1.1948, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.36290967226219023, | |
| "grad_norm": 0.42397868633270264, | |
| "learning_rate": 0.00014320908221059875, | |
| "loss": 1.0755, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.36370903277378097, | |
| "grad_norm": 0.33729004859924316, | |
| "learning_rate": 0.00014298064562932904, | |
| "loss": 1.0324, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.3645083932853717, | |
| "grad_norm": 0.46672719717025757, | |
| "learning_rate": 0.0001427519336068698, | |
| "loss": 1.1945, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.36530775379696245, | |
| "grad_norm": 0.36421141028404236, | |
| "learning_rate": 0.0001425229476089204, | |
| "loss": 1.1335, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.36610711430855314, | |
| "grad_norm": 0.4340507686138153, | |
| "learning_rate": 0.00014229368910293627, | |
| "loss": 1.0877, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.3669064748201439, | |
| "grad_norm": 0.34189310669898987, | |
| "learning_rate": 0.00014206415955811898, | |
| "loss": 1.0999, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.3677058353317346, | |
| "grad_norm": 0.3289022147655487, | |
| "learning_rate": 0.00014183436044540713, | |
| "loss": 1.055, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.36850519584332536, | |
| "grad_norm": 0.3434530198574066, | |
| "learning_rate": 0.00014160429323746687, | |
| "loss": 1.0693, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.36930455635491605, | |
| "grad_norm": 0.3331538438796997, | |
| "learning_rate": 0.00014137395940868244, | |
| "loss": 1.0318, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.3701039168665068, | |
| "grad_norm": 0.3151646852493286, | |
| "learning_rate": 0.00014114336043514661, | |
| "loss": 1.1513, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.37090327737809753, | |
| "grad_norm": 0.3027523159980774, | |
| "learning_rate": 0.0001409124977946515, | |
| "loss": 1.2295, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.37170263788968827, | |
| "grad_norm": 0.3899821639060974, | |
| "learning_rate": 0.00014068137296667884, | |
| "loss": 1.3589, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.37250199840127896, | |
| "grad_norm": 0.4125242233276367, | |
| "learning_rate": 0.0001404499874323905, | |
| "loss": 1.0456, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.3733013589128697, | |
| "grad_norm": 0.3297238051891327, | |
| "learning_rate": 0.00014021834267461932, | |
| "loss": 1.2101, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.37410071942446044, | |
| "grad_norm": 0.40019673109054565, | |
| "learning_rate": 0.00013998644017785913, | |
| "loss": 1.0543, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.3749000799360512, | |
| "grad_norm": 0.36964860558509827, | |
| "learning_rate": 0.0001397542814282556, | |
| "loss": 1.2206, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.37569944044764186, | |
| "grad_norm": 0.3775829076766968, | |
| "learning_rate": 0.0001395218679135966, | |
| "loss": 0.9938, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.3764988009592326, | |
| "grad_norm": 0.3969877064228058, | |
| "learning_rate": 0.00013928920112330262, | |
| "loss": 1.126, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.37729816147082335, | |
| "grad_norm": 0.43040886521339417, | |
| "learning_rate": 0.00013905628254841729, | |
| "loss": 1.1977, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.3780975219824141, | |
| "grad_norm": 0.35382160544395447, | |
| "learning_rate": 0.00013882311368159782, | |
| "loss": 0.9983, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.37889688249400477, | |
| "grad_norm": 0.362121045589447, | |
| "learning_rate": 0.00013858969601710533, | |
| "loss": 0.9183, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.3796962430055955, | |
| "grad_norm": 0.4062749445438385, | |
| "learning_rate": 0.00013835603105079539, | |
| "loss": 0.9459, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.38049560351718625, | |
| "grad_norm": 0.4215144217014313, | |
| "learning_rate": 0.00013812212028010854, | |
| "loss": 1.0205, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.381294964028777, | |
| "grad_norm": 0.49302709102630615, | |
| "learning_rate": 0.00013788796520406034, | |
| "loss": 1.0908, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.3820943245403677, | |
| "grad_norm": 0.49464038014411926, | |
| "learning_rate": 0.0001376535673232321, | |
| "loss": 1.0684, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.3828936850519584, | |
| "grad_norm": 0.4653661549091339, | |
| "learning_rate": 0.00013741892813976116, | |
| "loss": 1.1386, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.38369304556354916, | |
| "grad_norm": 0.5560212731361389, | |
| "learning_rate": 0.0001371840491573312, | |
| "loss": 1.1828, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.3844924060751399, | |
| "grad_norm": 0.46823370456695557, | |
| "learning_rate": 0.0001369489318811626, | |
| "loss": 1.1321, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.3852917665867306, | |
| "grad_norm": 0.5444937348365784, | |
| "learning_rate": 0.000136713577818003, | |
| "loss": 1.0788, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.38609112709832133, | |
| "grad_norm": 0.5422117710113525, | |
| "learning_rate": 0.0001364779884761173, | |
| "loss": 1.1162, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.38689048760991207, | |
| "grad_norm": 0.5373064875602722, | |
| "learning_rate": 0.0001362421653652784, | |
| "loss": 1.0085, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.3876898481215028, | |
| "grad_norm": 0.6702423095703125, | |
| "learning_rate": 0.00013600610999675702, | |
| "loss": 1.0995, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.38848920863309355, | |
| "grad_norm": 0.5719643235206604, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.9434, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.38928856914468424, | |
| "grad_norm": 0.5578810572624207, | |
| "learning_rate": 0.00013553330853918304, | |
| "loss": 1.1579, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.390087929656275, | |
| "grad_norm": 0.7008306384086609, | |
| "learning_rate": 0.0001352965654800755, | |
| "loss": 1.0893, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.3908872901678657, | |
| "grad_norm": 0.6100425124168396, | |
| "learning_rate": 0.0001350595962231563, | |
| "loss": 1.1744, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.39168665067945646, | |
| "grad_norm": 0.7699117064476013, | |
| "learning_rate": 0.00013482240228704142, | |
| "loss": 1.1863, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.39248601119104715, | |
| "grad_norm": 0.7704407572746277, | |
| "learning_rate": 0.00013458498519178664, | |
| "loss": 1.1027, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.3932853717026379, | |
| "grad_norm": 0.8181988000869751, | |
| "learning_rate": 0.00013434734645887788, | |
| "loss": 0.897, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.39408473221422863, | |
| "grad_norm": 0.8602463603019714, | |
| "learning_rate": 0.00013410948761122143, | |
| "loss": 1.0187, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.39488409272581937, | |
| "grad_norm": 0.8925572037696838, | |
| "learning_rate": 0.00013387141017313418, | |
| "loss": 1.1308, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.39568345323741005, | |
| "grad_norm": 0.8595069050788879, | |
| "learning_rate": 0.00013363311567033382, | |
| "loss": 1.1001, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.3964828137490008, | |
| "grad_norm": 0.7815159559249878, | |
| "learning_rate": 0.0001333946056299292, | |
| "loss": 1.0728, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.39728217426059154, | |
| "grad_norm": 0.9263238310813904, | |
| "learning_rate": 0.00013315588158041032, | |
| "loss": 1.1485, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.3980815347721823, | |
| "grad_norm": 1.121984839439392, | |
| "learning_rate": 0.0001329169450516387, | |
| "loss": 0.9351, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.39888089528377296, | |
| "grad_norm": 1.190191388130188, | |
| "learning_rate": 0.0001326777975748376, | |
| "loss": 1.1409, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.3996802557953637, | |
| "grad_norm": 1.6871143579483032, | |
| "learning_rate": 0.000132438440682582, | |
| "loss": 1.2688, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.40047961630695444, | |
| "grad_norm": 0.3561265468597412, | |
| "learning_rate": 0.00013219887590878903, | |
| "loss": 1.0588, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.4012789768185452, | |
| "grad_norm": 0.37955114245414734, | |
| "learning_rate": 0.00013195910478870803, | |
| "loss": 1.0717, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.40207833733013587, | |
| "grad_norm": 0.3907930850982666, | |
| "learning_rate": 0.00013171912885891063, | |
| "loss": 1.1347, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.4028776978417266, | |
| "grad_norm": 0.3910665214061737, | |
| "learning_rate": 0.000131478949657281, | |
| "loss": 1.2596, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.40367705835331735, | |
| "grad_norm": 0.3303968012332916, | |
| "learning_rate": 0.00013123856872300607, | |
| "loss": 0.8503, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.4044764188649081, | |
| "grad_norm": 0.383739173412323, | |
| "learning_rate": 0.00013099798759656545, | |
| "loss": 1.1542, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.4052757793764988, | |
| "grad_norm": 0.3624424934387207, | |
| "learning_rate": 0.00013075720781972175, | |
| "loss": 1.2591, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.4060751398880895, | |
| "grad_norm": 0.33615046739578247, | |
| "learning_rate": 0.00013051623093551063, | |
| "loss": 1.1455, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.40687450039968026, | |
| "grad_norm": 0.3861737847328186, | |
| "learning_rate": 0.00013027505848823084, | |
| "loss": 1.0889, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.407673860911271, | |
| "grad_norm": 0.43577584624290466, | |
| "learning_rate": 0.00013003369202343453, | |
| "loss": 1.1141, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4084732214228617, | |
| "grad_norm": 0.3891628086566925, | |
| "learning_rate": 0.0001297921330879171, | |
| "loss": 1.2212, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.40927258193445243, | |
| "grad_norm": 0.35330793261528015, | |
| "learning_rate": 0.00012955038322970736, | |
| "loss": 1.1245, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.41007194244604317, | |
| "grad_norm": 0.3083156645298004, | |
| "learning_rate": 0.0001293084439980578, | |
| "loss": 1.1398, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.4108713029576339, | |
| "grad_norm": 0.3527776300907135, | |
| "learning_rate": 0.00012906631694343436, | |
| "loss": 1.0458, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.4116706634692246, | |
| "grad_norm": 0.34124818444252014, | |
| "learning_rate": 0.00012882400361750675, | |
| "loss": 1.2349, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.41247002398081534, | |
| "grad_norm": 0.3764817416667938, | |
| "learning_rate": 0.0001285815055731383, | |
| "loss": 1.1106, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.4132693844924061, | |
| "grad_norm": 0.35788124799728394, | |
| "learning_rate": 0.00012833882436437624, | |
| "loss": 1.0105, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.4140687450039968, | |
| "grad_norm": 0.35195162892341614, | |
| "learning_rate": 0.00012809596154644148, | |
| "loss": 1.034, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.4148681055155875, | |
| "grad_norm": 0.3743147552013397, | |
| "learning_rate": 0.00012785291867571883, | |
| "loss": 1.1551, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.41566746602717825, | |
| "grad_norm": 0.37264588475227356, | |
| "learning_rate": 0.00012760969730974694, | |
| "loss": 1.084, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.416466826538769, | |
| "grad_norm": 0.4359087646007538, | |
| "learning_rate": 0.0001273662990072083, | |
| "loss": 1.1565, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.4172661870503597, | |
| "grad_norm": 0.4252093732357025, | |
| "learning_rate": 0.00012712272532791946, | |
| "loss": 1.1465, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.4180655475619504, | |
| "grad_norm": 0.4450690746307373, | |
| "learning_rate": 0.0001268789778328207, | |
| "loss": 1.051, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.41886490807354115, | |
| "grad_norm": 0.4181789755821228, | |
| "learning_rate": 0.0001266350580839664, | |
| "loss": 0.94, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.4196642685851319, | |
| "grad_norm": 0.43827134370803833, | |
| "learning_rate": 0.0001263909676445145, | |
| "loss": 1.1064, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.42046362909672264, | |
| "grad_norm": 0.43333086371421814, | |
| "learning_rate": 0.0001261467080787171, | |
| "loss": 1.0093, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.4212629896083134, | |
| "grad_norm": 0.4483776092529297, | |
| "learning_rate": 0.0001259022809519101, | |
| "loss": 0.9522, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.42206235011990406, | |
| "grad_norm": 0.42380601167678833, | |
| "learning_rate": 0.00012565768783050305, | |
| "loss": 1.1873, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.4228617106314948, | |
| "grad_norm": 0.46552905440330505, | |
| "learning_rate": 0.00012541293028196944, | |
| "loss": 1.1422, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.42366107114308554, | |
| "grad_norm": 0.4734083414077759, | |
| "learning_rate": 0.0001251680098748365, | |
| "loss": 1.1511, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4244604316546763, | |
| "grad_norm": 0.5062898993492126, | |
| "learning_rate": 0.00012492292817867497, | |
| "loss": 0.8868, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.42525979216626697, | |
| "grad_norm": 0.57962566614151, | |
| "learning_rate": 0.00012467768676408935, | |
| "loss": 1.072, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.4260591526778577, | |
| "grad_norm": 0.5128533244132996, | |
| "learning_rate": 0.00012443228720270767, | |
| "loss": 0.9567, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.42685851318944845, | |
| "grad_norm": 0.5683857202529907, | |
| "learning_rate": 0.00012418673106717143, | |
| "loss": 1.2017, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.4276578737010392, | |
| "grad_norm": 0.5700757503509521, | |
| "learning_rate": 0.00012394101993112557, | |
| "loss": 1.2576, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4284572342126299, | |
| "grad_norm": 0.7162390351295471, | |
| "learning_rate": 0.00012369515536920825, | |
| "loss": 1.0045, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.4292565947242206, | |
| "grad_norm": 0.594262421131134, | |
| "learning_rate": 0.00012344913895704097, | |
| "loss": 0.8915, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.43005595523581136, | |
| "grad_norm": 0.6075459718704224, | |
| "learning_rate": 0.00012320297227121833, | |
| "loss": 1.1498, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.4308553157474021, | |
| "grad_norm": 0.7022448182106018, | |
| "learning_rate": 0.0001229566568892979, | |
| "loss": 1.1091, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.4316546762589928, | |
| "grad_norm": 0.8270899057388306, | |
| "learning_rate": 0.00012271019438979023, | |
| "loss": 1.1842, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4324540367705835, | |
| "grad_norm": 0.7911314368247986, | |
| "learning_rate": 0.00012246358635214868, | |
| "loss": 1.0775, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.43325339728217427, | |
| "grad_norm": 0.6665616035461426, | |
| "learning_rate": 0.0001222168343567592, | |
| "loss": 1.0582, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.434052757793765, | |
| "grad_norm": 0.7318521738052368, | |
| "learning_rate": 0.00012196993998493043, | |
| "loss": 1.1444, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.4348521183053557, | |
| "grad_norm": 0.7249645590782166, | |
| "learning_rate": 0.00012172290481888331, | |
| "loss": 1.086, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.43565147881694644, | |
| "grad_norm": 0.8864445090293884, | |
| "learning_rate": 0.00012147573044174112, | |
| "loss": 1.3008, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.4364508393285372, | |
| "grad_norm": 0.7858752012252808, | |
| "learning_rate": 0.00012122841843751924, | |
| "loss": 1.1622, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.4372501998401279, | |
| "grad_norm": 0.8661579489707947, | |
| "learning_rate": 0.00012098097039111512, | |
| "loss": 1.009, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.4380495603517186, | |
| "grad_norm": 1.0622788667678833, | |
| "learning_rate": 0.00012073338788829787, | |
| "loss": 1.044, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.43884892086330934, | |
| "grad_norm": 1.1994202136993408, | |
| "learning_rate": 0.0001204856725156984, | |
| "loss": 1.1495, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.4396482813749001, | |
| "grad_norm": 1.5216366052627563, | |
| "learning_rate": 0.00012023782586079908, | |
| "loss": 1.1216, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4404476418864908, | |
| "grad_norm": 0.4045037627220154, | |
| "learning_rate": 0.00011998984951192361, | |
| "loss": 1.0062, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.4412470023980815, | |
| "grad_norm": 0.39483803510665894, | |
| "learning_rate": 0.00011974174505822677, | |
| "loss": 0.9856, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.44204636290967225, | |
| "grad_norm": 0.409810870885849, | |
| "learning_rate": 0.00011949351408968439, | |
| "loss": 1.0131, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.442845723421263, | |
| "grad_norm": 0.3824607729911804, | |
| "learning_rate": 0.000119245158197083, | |
| "loss": 1.0047, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.44364508393285373, | |
| "grad_norm": 0.37841466069221497, | |
| "learning_rate": 0.00011899667897200977, | |
| "loss": 1.0711, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.3607119023799896, | |
| "learning_rate": 0.00011874807800684215, | |
| "loss": 1.0795, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.44524380495603516, | |
| "grad_norm": 0.3559788465499878, | |
| "learning_rate": 0.00011849935689473786, | |
| "loss": 1.2118, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.4460431654676259, | |
| "grad_norm": 0.678111732006073, | |
| "learning_rate": 0.00011825051722962453, | |
| "loss": 1.2167, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.44684252597921664, | |
| "grad_norm": 0.3986605107784271, | |
| "learning_rate": 0.00011800156060618955, | |
| "loss": 1.1445, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.44764188649080733, | |
| "grad_norm": 0.36188986897468567, | |
| "learning_rate": 0.00011775248861986978, | |
| "loss": 1.0736, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.44844124700239807, | |
| "grad_norm": 0.33992865681648254, | |
| "learning_rate": 0.00011750330286684153, | |
| "loss": 1.1707, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.4492406075139888, | |
| "grad_norm": 0.35370391607284546, | |
| "learning_rate": 0.00011725400494401, | |
| "loss": 1.1023, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.45003996802557955, | |
| "grad_norm": 0.3561865985393524, | |
| "learning_rate": 0.00011700459644899937, | |
| "loss": 1.2305, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.45083932853717024, | |
| "grad_norm": 0.37104496359825134, | |
| "learning_rate": 0.00011675507898014235, | |
| "loss": 1.1114, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.451638689048761, | |
| "grad_norm": 0.350696861743927, | |
| "learning_rate": 0.00011650545413647, | |
| "loss": 1.1613, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.4524380495603517, | |
| "grad_norm": 0.38277414441108704, | |
| "learning_rate": 0.00011625572351770155, | |
| "loss": 1.1631, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.45323741007194246, | |
| "grad_norm": 0.35308247804641724, | |
| "learning_rate": 0.00011600588872423405, | |
| "loss": 1.1086, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.4540367705835332, | |
| "grad_norm": 0.3877826929092407, | |
| "learning_rate": 0.0001157559513571322, | |
| "loss": 1.1722, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.4548361310951239, | |
| "grad_norm": 0.38781118392944336, | |
| "learning_rate": 0.00011550591301811788, | |
| "loss": 1.013, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.4556354916067146, | |
| "grad_norm": 0.40038326382637024, | |
| "learning_rate": 0.00011525577530956029, | |
| "loss": 1.0764, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.45643485211830537, | |
| "grad_norm": 0.3875632584095001, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 1.1239, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.4572342126298961, | |
| "grad_norm": 0.44367122650146484, | |
| "learning_rate": 0.00011475520819646526, | |
| "loss": 1.0437, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.4580335731414868, | |
| "grad_norm": 0.5386861562728882, | |
| "learning_rate": 0.0001145047819998089, | |
| "loss": 0.9693, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.45883293365307753, | |
| "grad_norm": 0.5589727163314819, | |
| "learning_rate": 0.00011425426284935091, | |
| "loss": 1.0766, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.4596322941646683, | |
| "grad_norm": 0.48679694533348083, | |
| "learning_rate": 0.0001140036523505416, | |
| "loss": 1.1827, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.460431654676259, | |
| "grad_norm": 0.3698740005493164, | |
| "learning_rate": 0.00011375295210941674, | |
| "loss": 1.0565, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.4612310151878497, | |
| "grad_norm": 0.4788266718387604, | |
| "learning_rate": 0.00011350216373258722, | |
| "loss": 1.006, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.46203037569944044, | |
| "grad_norm": 0.6378310322761536, | |
| "learning_rate": 0.00011325128882722875, | |
| "loss": 1.2113, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.4628297362110312, | |
| "grad_norm": 0.5069262981414795, | |
| "learning_rate": 0.00011300032900107147, | |
| "loss": 0.9833, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.4636290967226219, | |
| "grad_norm": 0.44473326206207275, | |
| "learning_rate": 0.00011274928586238986, | |
| "loss": 1.0615, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4644284572342126, | |
| "grad_norm": 0.5024760961532593, | |
| "learning_rate": 0.00011249816101999227, | |
| "loss": 0.9796, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.46522781774580335, | |
| "grad_norm": 0.537169337272644, | |
| "learning_rate": 0.00011224695608321056, | |
| "loss": 1.0713, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.4660271782573941, | |
| "grad_norm": 0.6242894530296326, | |
| "learning_rate": 0.00011199567266188999, | |
| "loss": 1.3315, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.46682653876898483, | |
| "grad_norm": 0.6122602820396423, | |
| "learning_rate": 0.00011174431236637874, | |
| "loss": 1.0745, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.4676258992805755, | |
| "grad_norm": 0.5972760319709778, | |
| "learning_rate": 0.00011149287680751763, | |
| "loss": 1.1155, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.46842525979216626, | |
| "grad_norm": 0.6527268290519714, | |
| "learning_rate": 0.0001112413675966298, | |
| "loss": 1.0451, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.469224620303757, | |
| "grad_norm": 0.662009060382843, | |
| "learning_rate": 0.0001109897863455104, | |
| "loss": 1.1162, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.47002398081534774, | |
| "grad_norm": 0.7552705407142639, | |
| "learning_rate": 0.00011073813466641632, | |
| "loss": 0.9946, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.4708233413269384, | |
| "grad_norm": 0.7745950818061829, | |
| "learning_rate": 0.00011048641417205563, | |
| "loss": 1.1047, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.47162270183852917, | |
| "grad_norm": 0.6825327277183533, | |
| "learning_rate": 0.00011023462647557751, | |
| "loss": 1.0597, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.4724220623501199, | |
| "grad_norm": 0.9221934080123901, | |
| "learning_rate": 0.0001099827731905618, | |
| "loss": 1.0257, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.47322142286171065, | |
| "grad_norm": 0.8638830184936523, | |
| "learning_rate": 0.0001097308559310087, | |
| "loss": 1.0615, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.47402078337330134, | |
| "grad_norm": 0.921091616153717, | |
| "learning_rate": 0.00010947887631132823, | |
| "loss": 1.0958, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.4748201438848921, | |
| "grad_norm": 0.9311054348945618, | |
| "learning_rate": 0.00010922683594633021, | |
| "loss": 1.1697, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.4756195043964828, | |
| "grad_norm": 1.0105490684509277, | |
| "learning_rate": 0.00010897473645121372, | |
| "loss": 1.1954, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.47641886490807356, | |
| "grad_norm": 0.9977965354919434, | |
| "learning_rate": 0.00010872257944155669, | |
| "loss": 1.1052, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.47721822541966424, | |
| "grad_norm": 0.8218235969543457, | |
| "learning_rate": 0.00010847036653330568, | |
| "loss": 1.0016, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.478017585931255, | |
| "grad_norm": 1.238542914390564, | |
| "learning_rate": 0.00010821809934276554, | |
| "loss": 1.1577, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.4788169464428457, | |
| "grad_norm": 1.1989014148712158, | |
| "learning_rate": 0.00010796577948658892, | |
| "loss": 1.0534, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.47961630695443647, | |
| "grad_norm": 1.591269612312317, | |
| "learning_rate": 0.00010771340858176594, | |
| "loss": 1.3851, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.48041566746602715, | |
| "grad_norm": 0.3721429109573364, | |
| "learning_rate": 0.00010746098824561397, | |
| "loss": 0.9508, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.4812150279776179, | |
| "grad_norm": 0.40401971340179443, | |
| "learning_rate": 0.00010720852009576705, | |
| "loss": 0.8829, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.48201438848920863, | |
| "grad_norm": 0.4219840466976166, | |
| "learning_rate": 0.00010695600575016571, | |
| "loss": 0.9494, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.4828137490007994, | |
| "grad_norm": 0.3684583604335785, | |
| "learning_rate": 0.0001067034468270465, | |
| "loss": 0.9306, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.48361310951239006, | |
| "grad_norm": 0.3637100160121918, | |
| "learning_rate": 0.00010645084494493165, | |
| "loss": 1.1087, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.4844124700239808, | |
| "grad_norm": 0.4035799503326416, | |
| "learning_rate": 0.00010619820172261867, | |
| "loss": 1.3074, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.48521183053557154, | |
| "grad_norm": 0.39611637592315674, | |
| "learning_rate": 0.00010594551877917003, | |
| "loss": 0.9484, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.4860111910471623, | |
| "grad_norm": 0.45287951827049255, | |
| "learning_rate": 0.00010569279773390273, | |
| "loss": 1.0403, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.486810551558753, | |
| "grad_norm": 0.3595340847969055, | |
| "learning_rate": 0.00010544004020637799, | |
| "loss": 1.0235, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.4876099120703437, | |
| "grad_norm": 0.3542434275150299, | |
| "learning_rate": 0.00010518724781639073, | |
| "loss": 1.2554, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.48840927258193445, | |
| "grad_norm": 0.35258346796035767, | |
| "learning_rate": 0.00010493442218395936, | |
| "loss": 1.1231, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.4892086330935252, | |
| "grad_norm": 0.3523050844669342, | |
| "learning_rate": 0.0001046815649293154, | |
| "loss": 1.0851, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.49000799360511593, | |
| "grad_norm": 0.336111456155777, | |
| "learning_rate": 0.00010442867767289281, | |
| "loss": 1.0403, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.4908073541167066, | |
| "grad_norm": 0.3554600775241852, | |
| "learning_rate": 0.000104175762035318, | |
| "loss": 1.1551, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.49160671462829736, | |
| "grad_norm": 0.35645225644111633, | |
| "learning_rate": 0.00010392281963739918, | |
| "loss": 1.1704, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.4924060751398881, | |
| "grad_norm": 0.4191342890262604, | |
| "learning_rate": 0.0001036698521001161, | |
| "loss": 1.0728, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.49320543565147884, | |
| "grad_norm": 0.3663305342197418, | |
| "learning_rate": 0.00010341686104460954, | |
| "loss": 1.0764, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.4940047961630695, | |
| "grad_norm": 0.3945816159248352, | |
| "learning_rate": 0.00010316384809217108, | |
| "loss": 1.0224, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.49480415667466027, | |
| "grad_norm": 0.39029353857040405, | |
| "learning_rate": 0.0001029108148642326, | |
| "loss": 1.0606, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.495603517186251, | |
| "grad_norm": 0.3495793342590332, | |
| "learning_rate": 0.00010265776298235586, | |
| "loss": 1.1024, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.49640287769784175, | |
| "grad_norm": 0.3891185224056244, | |
| "learning_rate": 0.00010240469406822221, | |
| "loss": 0.9662, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.49720223820943243, | |
| "grad_norm": 0.41916343569755554, | |
| "learning_rate": 0.00010215160974362223, | |
| "loss": 1.117, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.4980015987210232, | |
| "grad_norm": 0.44954174757003784, | |
| "learning_rate": 0.0001018985116304451, | |
| "loss": 0.9685, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.4988009592326139, | |
| "grad_norm": 0.40312501788139343, | |
| "learning_rate": 0.00010164540135066845, | |
| "loss": 1.022, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.49960031974420466, | |
| "grad_norm": 0.4184950590133667, | |
| "learning_rate": 0.00010139228052634789, | |
| "loss": 1.0082, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.5003996802557954, | |
| "grad_norm": 0.3949190676212311, | |
| "learning_rate": 0.00010113915077960658, | |
| "loss": 1.1159, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.5003996802557954, | |
| "eval_loss": 1.0933643579483032, | |
| "eval_runtime": 39.8289, | |
| "eval_samples_per_second": 13.232, | |
| "eval_steps_per_second": 6.628, | |
| "step": 626 | |
| } | |
| ], | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.738333182674862e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |