{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.982278481012658, "eval_steps": 500, "global_step": 1230, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.004050632911392405, "grad_norm": 6.366441440266172, "learning_rate": 3.2520325203252037e-07, "loss": 0.821, "step": 1 }, { "epoch": 0.00810126582278481, "grad_norm": 6.545198015340827, "learning_rate": 6.504065040650407e-07, "loss": 0.8314, "step": 2 }, { "epoch": 0.012151898734177215, "grad_norm": 6.593447305395962, "learning_rate": 9.75609756097561e-07, "loss": 0.846, "step": 3 }, { "epoch": 0.01620253164556962, "grad_norm": 6.243632182509292, "learning_rate": 1.3008130081300815e-06, "loss": 0.8259, "step": 4 }, { "epoch": 0.020253164556962026, "grad_norm": 5.647554768410843, "learning_rate": 1.6260162601626018e-06, "loss": 0.7923, "step": 5 }, { "epoch": 0.02430379746835443, "grad_norm": 5.679414762915671, "learning_rate": 1.951219512195122e-06, "loss": 0.7957, "step": 6 }, { "epoch": 0.028354430379746835, "grad_norm": 4.359931679539345, "learning_rate": 2.2764227642276426e-06, "loss": 0.7756, "step": 7 }, { "epoch": 0.03240506329113924, "grad_norm": 4.090092545752834, "learning_rate": 2.601626016260163e-06, "loss": 0.7696, "step": 8 }, { "epoch": 0.03645569620253165, "grad_norm": 2.406973407843762, "learning_rate": 2.926829268292683e-06, "loss": 0.7396, "step": 9 }, { "epoch": 0.04050632911392405, "grad_norm": 2.0606566657250003, "learning_rate": 3.2520325203252037e-06, "loss": 0.7426, "step": 10 }, { "epoch": 0.044556962025316456, "grad_norm": 1.9438707002394675, "learning_rate": 3.577235772357724e-06, "loss": 0.7267, "step": 11 }, { "epoch": 0.04860759493670886, "grad_norm": 3.524284979788005, "learning_rate": 3.902439024390244e-06, "loss": 0.6893, "step": 12 }, { "epoch": 0.052658227848101265, "grad_norm": 4.114948108617655, "learning_rate": 4.227642276422765e-06, "loss": 0.7147, "step": 13 }, { "epoch": 0.05670886075949367, "grad_norm": 4.055082261140213, "learning_rate": 4.552845528455285e-06, "loss": 0.7122, "step": 14 }, { "epoch": 0.060759493670886074, "grad_norm": 3.9634353558808284, "learning_rate": 4.8780487804878055e-06, "loss": 0.6941, "step": 15 }, { "epoch": 0.06481012658227848, "grad_norm": 3.2766960646757024, "learning_rate": 5.203252032520326e-06, "loss": 0.684, "step": 16 }, { "epoch": 0.06886075949367089, "grad_norm": 2.8654477774579803, "learning_rate": 5.528455284552846e-06, "loss": 0.6704, "step": 17 }, { "epoch": 0.0729113924050633, "grad_norm": 2.2130829165928496, "learning_rate": 5.853658536585366e-06, "loss": 0.6545, "step": 18 }, { "epoch": 0.0769620253164557, "grad_norm": 1.3672367554143627, "learning_rate": 6.178861788617887e-06, "loss": 0.6319, "step": 19 }, { "epoch": 0.0810126582278481, "grad_norm": 1.2356600786501815, "learning_rate": 6.504065040650407e-06, "loss": 0.6326, "step": 20 }, { "epoch": 0.08506329113924051, "grad_norm": 1.269364302772384, "learning_rate": 6.829268292682928e-06, "loss": 0.6061, "step": 21 }, { "epoch": 0.08911392405063291, "grad_norm": 1.3486435448786163, "learning_rate": 7.154471544715448e-06, "loss": 0.6079, "step": 22 }, { "epoch": 0.09316455696202532, "grad_norm": 1.100257835991344, "learning_rate": 7.4796747967479676e-06, "loss": 0.6209, "step": 23 }, { "epoch": 0.09721518987341772, "grad_norm": 0.8360825943209261, "learning_rate": 7.804878048780489e-06, "loss": 0.5907, "step": 24 }, { "epoch": 0.10126582278481013, "grad_norm": 1.4804549689806694, "learning_rate": 8.130081300813009e-06, "loss": 0.6293, "step": 25 }, { "epoch": 0.10531645569620253, "grad_norm": 1.0136577916808593, "learning_rate": 8.45528455284553e-06, "loss": 0.5834, "step": 26 }, { "epoch": 0.10936708860759493, "grad_norm": 0.8200281782730398, "learning_rate": 8.78048780487805e-06, "loss": 0.5887, "step": 27 }, { "epoch": 0.11341772151898734, "grad_norm": 0.7727161614581294, "learning_rate": 9.10569105691057e-06, "loss": 0.5766, "step": 28 }, { "epoch": 0.11746835443037974, "grad_norm": 0.8066947915166085, "learning_rate": 9.43089430894309e-06, "loss": 0.5719, "step": 29 }, { "epoch": 0.12151898734177215, "grad_norm": 0.7117595795844477, "learning_rate": 9.756097560975611e-06, "loss": 0.5606, "step": 30 }, { "epoch": 0.12556962025316457, "grad_norm": 0.6122696616691725, "learning_rate": 1.008130081300813e-05, "loss": 0.5485, "step": 31 }, { "epoch": 0.12962025316455697, "grad_norm": 0.7791422050469652, "learning_rate": 1.0406504065040652e-05, "loss": 0.558, "step": 32 }, { "epoch": 0.13367088607594937, "grad_norm": 0.6545150482063794, "learning_rate": 1.0731707317073172e-05, "loss": 0.5411, "step": 33 }, { "epoch": 0.13772151898734178, "grad_norm": 0.4409882649391751, "learning_rate": 1.1056910569105692e-05, "loss": 0.5423, "step": 34 }, { "epoch": 0.14177215189873418, "grad_norm": 0.5729369589767671, "learning_rate": 1.1382113821138213e-05, "loss": 0.5633, "step": 35 }, { "epoch": 0.1458227848101266, "grad_norm": 0.5872452125614008, "learning_rate": 1.1707317073170731e-05, "loss": 0.5457, "step": 36 }, { "epoch": 0.149873417721519, "grad_norm": 0.484872339479273, "learning_rate": 1.2032520325203254e-05, "loss": 0.5472, "step": 37 }, { "epoch": 0.1539240506329114, "grad_norm": 0.4873632508333773, "learning_rate": 1.2357723577235774e-05, "loss": 0.5303, "step": 38 }, { "epoch": 0.1579746835443038, "grad_norm": 0.5306590359837672, "learning_rate": 1.2682926829268294e-05, "loss": 0.5431, "step": 39 }, { "epoch": 0.1620253164556962, "grad_norm": 0.42316684122668263, "learning_rate": 1.3008130081300815e-05, "loss": 0.5291, "step": 40 }, { "epoch": 0.1660759493670886, "grad_norm": 0.5977149658826423, "learning_rate": 1.3333333333333333e-05, "loss": 0.5368, "step": 41 }, { "epoch": 0.17012658227848101, "grad_norm": 0.5066466576095802, "learning_rate": 1.3658536585365855e-05, "loss": 0.5188, "step": 42 }, { "epoch": 0.17417721518987342, "grad_norm": 0.4649625107507997, "learning_rate": 1.3983739837398376e-05, "loss": 0.5281, "step": 43 }, { "epoch": 0.17822784810126582, "grad_norm": 0.5385547888117231, "learning_rate": 1.4308943089430896e-05, "loss": 0.5142, "step": 44 }, { "epoch": 0.18227848101265823, "grad_norm": 0.4651183389879011, "learning_rate": 1.4634146341463415e-05, "loss": 0.5258, "step": 45 }, { "epoch": 0.18632911392405063, "grad_norm": 0.6770787341892783, "learning_rate": 1.4959349593495935e-05, "loss": 0.504, "step": 46 }, { "epoch": 0.19037974683544304, "grad_norm": 0.47808868693953954, "learning_rate": 1.528455284552846e-05, "loss": 0.5132, "step": 47 }, { "epoch": 0.19443037974683544, "grad_norm": 0.3859467704668216, "learning_rate": 1.5609756097560978e-05, "loss": 0.5059, "step": 48 }, { "epoch": 0.19848101265822785, "grad_norm": 0.521681415686226, "learning_rate": 1.5934959349593496e-05, "loss": 0.5189, "step": 49 }, { "epoch": 0.20253164556962025, "grad_norm": 0.4320969141357702, "learning_rate": 1.6260162601626018e-05, "loss": 0.5042, "step": 50 }, { "epoch": 0.20658227848101265, "grad_norm": 0.39767325297597317, "learning_rate": 1.6585365853658537e-05, "loss": 0.5128, "step": 51 }, { "epoch": 0.21063291139240506, "grad_norm": 0.4734419270712812, "learning_rate": 1.691056910569106e-05, "loss": 0.5051, "step": 52 }, { "epoch": 0.21468354430379746, "grad_norm": 0.4363703877768567, "learning_rate": 1.7235772357723578e-05, "loss": 0.5178, "step": 53 }, { "epoch": 0.21873417721518987, "grad_norm": 0.42402427944977317, "learning_rate": 1.75609756097561e-05, "loss": 0.4936, "step": 54 }, { "epoch": 0.22278481012658227, "grad_norm": 0.47090932896367166, "learning_rate": 1.788617886178862e-05, "loss": 0.4986, "step": 55 }, { "epoch": 0.22683544303797468, "grad_norm": 0.38945310503114267, "learning_rate": 1.821138211382114e-05, "loss": 0.5034, "step": 56 }, { "epoch": 0.23088607594936708, "grad_norm": 0.49078736654092026, "learning_rate": 1.8536585365853663e-05, "loss": 0.4939, "step": 57 }, { "epoch": 0.23493670886075949, "grad_norm": 0.43418061174289796, "learning_rate": 1.886178861788618e-05, "loss": 0.4916, "step": 58 }, { "epoch": 0.2389873417721519, "grad_norm": 0.3963981471214845, "learning_rate": 1.91869918699187e-05, "loss": 0.4946, "step": 59 }, { "epoch": 0.2430379746835443, "grad_norm": 0.491781945152039, "learning_rate": 1.9512195121951222e-05, "loss": 0.4862, "step": 60 }, { "epoch": 0.2470886075949367, "grad_norm": 0.5424965969922666, "learning_rate": 1.983739837398374e-05, "loss": 0.4773, "step": 61 }, { "epoch": 0.25113924050632913, "grad_norm": 0.7129918960468913, "learning_rate": 2.016260162601626e-05, "loss": 0.492, "step": 62 }, { "epoch": 0.25518987341772154, "grad_norm": 0.8385339964707252, "learning_rate": 2.048780487804878e-05, "loss": 0.512, "step": 63 }, { "epoch": 0.25924050632911394, "grad_norm": 0.7155233535976742, "learning_rate": 2.0813008130081303e-05, "loss": 0.4837, "step": 64 }, { "epoch": 0.26329113924050634, "grad_norm": 0.43959274616752836, "learning_rate": 2.1138211382113822e-05, "loss": 0.4673, "step": 65 }, { "epoch": 0.26734177215189875, "grad_norm": 0.6893810785438176, "learning_rate": 2.1463414634146344e-05, "loss": 0.5004, "step": 66 }, { "epoch": 0.27139240506329115, "grad_norm": 0.8055210832307125, "learning_rate": 2.1788617886178863e-05, "loss": 0.4879, "step": 67 }, { "epoch": 0.27544303797468356, "grad_norm": 0.6710137189102823, "learning_rate": 2.2113821138211385e-05, "loss": 0.4888, "step": 68 }, { "epoch": 0.27949367088607596, "grad_norm": 0.707042844525267, "learning_rate": 2.2439024390243907e-05, "loss": 0.5051, "step": 69 }, { "epoch": 0.28354430379746837, "grad_norm": 0.7007895289824153, "learning_rate": 2.2764227642276426e-05, "loss": 0.5075, "step": 70 }, { "epoch": 0.28759493670886077, "grad_norm": 0.6620953982378566, "learning_rate": 2.3089430894308948e-05, "loss": 0.487, "step": 71 }, { "epoch": 0.2916455696202532, "grad_norm": 0.5914359432555613, "learning_rate": 2.3414634146341463e-05, "loss": 0.4972, "step": 72 }, { "epoch": 0.2956962025316456, "grad_norm": 0.6906027650575874, "learning_rate": 2.3739837398373985e-05, "loss": 0.4712, "step": 73 }, { "epoch": 0.299746835443038, "grad_norm": 0.6228319662029798, "learning_rate": 2.4065040650406507e-05, "loss": 0.4707, "step": 74 }, { "epoch": 0.3037974683544304, "grad_norm": 0.6027303054048482, "learning_rate": 2.4390243902439026e-05, "loss": 0.4757, "step": 75 }, { "epoch": 0.3078481012658228, "grad_norm": 0.5623183741162822, "learning_rate": 2.4715447154471548e-05, "loss": 0.4755, "step": 76 }, { "epoch": 0.3118987341772152, "grad_norm": 0.6196870355774003, "learning_rate": 2.5040650406504066e-05, "loss": 0.4836, "step": 77 }, { "epoch": 0.3159493670886076, "grad_norm": 0.585808195959988, "learning_rate": 2.536585365853659e-05, "loss": 0.4912, "step": 78 }, { "epoch": 0.32, "grad_norm": 0.6499168466363382, "learning_rate": 2.569105691056911e-05, "loss": 0.4856, "step": 79 }, { "epoch": 0.3240506329113924, "grad_norm": 0.640722520871733, "learning_rate": 2.601626016260163e-05, "loss": 0.4698, "step": 80 }, { "epoch": 0.3281012658227848, "grad_norm": 0.7862368621873075, "learning_rate": 2.634146341463415e-05, "loss": 0.4875, "step": 81 }, { "epoch": 0.3321518987341772, "grad_norm": 0.9788852183483757, "learning_rate": 2.6666666666666667e-05, "loss": 0.4823, "step": 82 }, { "epoch": 0.3362025316455696, "grad_norm": 0.9850442568440896, "learning_rate": 2.699186991869919e-05, "loss": 0.475, "step": 83 }, { "epoch": 0.34025316455696203, "grad_norm": 1.15027445745085, "learning_rate": 2.731707317073171e-05, "loss": 0.4871, "step": 84 }, { "epoch": 0.34430379746835443, "grad_norm": 1.0561772634901054, "learning_rate": 2.764227642276423e-05, "loss": 0.4744, "step": 85 }, { "epoch": 0.34835443037974684, "grad_norm": 1.6959884143038297, "learning_rate": 2.796747967479675e-05, "loss": 0.4794, "step": 86 }, { "epoch": 0.35240506329113924, "grad_norm": 1.059593020888486, "learning_rate": 2.829268292682927e-05, "loss": 0.4884, "step": 87 }, { "epoch": 0.35645569620253165, "grad_norm": 1.415149407908605, "learning_rate": 2.8617886178861792e-05, "loss": 0.4804, "step": 88 }, { "epoch": 0.36050632911392405, "grad_norm": 0.8148204568794046, "learning_rate": 2.8943089430894314e-05, "loss": 0.4903, "step": 89 }, { "epoch": 0.36455696202531646, "grad_norm": 1.5996990796387476, "learning_rate": 2.926829268292683e-05, "loss": 0.4951, "step": 90 }, { "epoch": 0.36860759493670886, "grad_norm": 0.8822334355274946, "learning_rate": 2.959349593495935e-05, "loss": 0.4649, "step": 91 }, { "epoch": 0.37265822784810126, "grad_norm": 1.4700712322907636, "learning_rate": 2.991869918699187e-05, "loss": 0.487, "step": 92 }, { "epoch": 0.37670886075949367, "grad_norm": 0.9376022645405021, "learning_rate": 3.0243902439024392e-05, "loss": 0.4917, "step": 93 }, { "epoch": 0.3807594936708861, "grad_norm": 1.197351608233562, "learning_rate": 3.056910569105692e-05, "loss": 0.4718, "step": 94 }, { "epoch": 0.3848101265822785, "grad_norm": 0.9867121419360544, "learning_rate": 3.089430894308943e-05, "loss": 0.4987, "step": 95 }, { "epoch": 0.3888607594936709, "grad_norm": 1.0954296227800238, "learning_rate": 3.1219512195121955e-05, "loss": 0.488, "step": 96 }, { "epoch": 0.3929113924050633, "grad_norm": 0.8527787675662697, "learning_rate": 3.154471544715447e-05, "loss": 0.482, "step": 97 }, { "epoch": 0.3969620253164557, "grad_norm": 0.902515561340026, "learning_rate": 3.186991869918699e-05, "loss": 0.4762, "step": 98 }, { "epoch": 0.4010126582278481, "grad_norm": 0.7073462135060716, "learning_rate": 3.2195121951219514e-05, "loss": 0.4688, "step": 99 }, { "epoch": 0.4050632911392405, "grad_norm": 0.8455745635478479, "learning_rate": 3.2520325203252037e-05, "loss": 0.463, "step": 100 }, { "epoch": 0.4091139240506329, "grad_norm": 0.8993103322511871, "learning_rate": 3.284552845528456e-05, "loss": 0.485, "step": 101 }, { "epoch": 0.4131645569620253, "grad_norm": 1.0664632469728645, "learning_rate": 3.3170731707317074e-05, "loss": 0.4795, "step": 102 }, { "epoch": 0.4172151898734177, "grad_norm": 0.8503052911995067, "learning_rate": 3.3495934959349596e-05, "loss": 0.4758, "step": 103 }, { "epoch": 0.4212658227848101, "grad_norm": 0.9775800924797176, "learning_rate": 3.382113821138212e-05, "loss": 0.4951, "step": 104 }, { "epoch": 0.4253164556962025, "grad_norm": 0.8049617972995645, "learning_rate": 3.414634146341463e-05, "loss": 0.4597, "step": 105 }, { "epoch": 0.4293670886075949, "grad_norm": 0.7900797136838957, "learning_rate": 3.4471544715447155e-05, "loss": 0.4838, "step": 106 }, { "epoch": 0.43341772151898733, "grad_norm": 0.7880645188095051, "learning_rate": 3.479674796747968e-05, "loss": 0.482, "step": 107 }, { "epoch": 0.43746835443037974, "grad_norm": 0.7258304537248991, "learning_rate": 3.51219512195122e-05, "loss": 0.4613, "step": 108 }, { "epoch": 0.44151898734177214, "grad_norm": 0.619891381840048, "learning_rate": 3.544715447154472e-05, "loss": 0.4852, "step": 109 }, { "epoch": 0.44556962025316454, "grad_norm": 0.8069916108810657, "learning_rate": 3.577235772357724e-05, "loss": 0.4898, "step": 110 }, { "epoch": 0.44962025316455695, "grad_norm": 0.7163028652186608, "learning_rate": 3.609756097560976e-05, "loss": 0.4762, "step": 111 }, { "epoch": 0.45367088607594935, "grad_norm": 0.6695133766259735, "learning_rate": 3.642276422764228e-05, "loss": 0.4881, "step": 112 }, { "epoch": 0.45772151898734176, "grad_norm": 45.34593792355903, "learning_rate": 3.67479674796748e-05, "loss": 0.4908, "step": 113 }, { "epoch": 0.46177215189873416, "grad_norm": 1.2332748445746946, "learning_rate": 3.7073170731707325e-05, "loss": 0.4856, "step": 114 }, { "epoch": 0.46582278481012657, "grad_norm": 1.5308061880596793, "learning_rate": 3.739837398373984e-05, "loss": 0.4847, "step": 115 }, { "epoch": 0.46987341772151897, "grad_norm": 1.0846008623758148, "learning_rate": 3.772357723577236e-05, "loss": 0.4862, "step": 116 }, { "epoch": 0.4739240506329114, "grad_norm": 1.802203368376355, "learning_rate": 3.804878048780488e-05, "loss": 0.4649, "step": 117 }, { "epoch": 0.4779746835443038, "grad_norm": 1.1795553399845844, "learning_rate": 3.83739837398374e-05, "loss": 0.461, "step": 118 }, { "epoch": 0.4820253164556962, "grad_norm": 1.8414870173610898, "learning_rate": 3.869918699186992e-05, "loss": 0.4825, "step": 119 }, { "epoch": 0.4860759493670886, "grad_norm": 1.869409147792413, "learning_rate": 3.9024390243902444e-05, "loss": 0.5074, "step": 120 }, { "epoch": 0.490126582278481, "grad_norm": 1.2587939981813374, "learning_rate": 3.9349593495934966e-05, "loss": 0.484, "step": 121 }, { "epoch": 0.4941772151898734, "grad_norm": 1.1925572601165648, "learning_rate": 3.967479674796748e-05, "loss": 0.4657, "step": 122 }, { "epoch": 0.4982278481012658, "grad_norm": 1.0506632426848639, "learning_rate": 4e-05, "loss": 0.4823, "step": 123 }, { "epoch": 0.5022784810126583, "grad_norm": 1.117820980499303, "learning_rate": 3.999991946137476e-05, "loss": 0.4705, "step": 124 }, { "epoch": 0.5063291139240507, "grad_norm": 0.8042164147432055, "learning_rate": 3.999967784614766e-05, "loss": 0.4763, "step": 125 }, { "epoch": 0.5103797468354431, "grad_norm": 0.9497144313833796, "learning_rate": 3.9999275156264656e-05, "loss": 0.4649, "step": 126 }, { "epoch": 0.5144303797468355, "grad_norm": 0.7821106786355714, "learning_rate": 3.999871139496895e-05, "loss": 0.4741, "step": 127 }, { "epoch": 0.5184810126582279, "grad_norm": 0.7750007729736481, "learning_rate": 3.9997986566800995e-05, "loss": 0.4621, "step": 128 }, { "epoch": 0.5225316455696203, "grad_norm": 0.8754827610574405, "learning_rate": 3.999710067759846e-05, "loss": 0.4501, "step": 129 }, { "epoch": 0.5265822784810127, "grad_norm": 0.5960096735721176, "learning_rate": 3.999605373449617e-05, "loss": 0.4588, "step": 130 }, { "epoch": 0.5306329113924051, "grad_norm": 0.7760303749358244, "learning_rate": 3.9994845745926075e-05, "loss": 0.4839, "step": 131 }, { "epoch": 0.5346835443037975, "grad_norm": 0.731621104248255, "learning_rate": 3.999347672161713e-05, "loss": 0.4594, "step": 132 }, { "epoch": 0.5387341772151899, "grad_norm": 0.7626960430720853, "learning_rate": 3.999194667259528e-05, "loss": 0.4794, "step": 133 }, { "epoch": 0.5427848101265823, "grad_norm": 0.6866213432819137, "learning_rate": 3.999025561118334e-05, "loss": 0.4554, "step": 134 }, { "epoch": 0.5468354430379747, "grad_norm": 0.6353463940084648, "learning_rate": 3.998840355100086e-05, "loss": 0.479, "step": 135 }, { "epoch": 0.5508860759493671, "grad_norm": 0.9845967580472496, "learning_rate": 3.998639050696409e-05, "loss": 0.4621, "step": 136 }, { "epoch": 0.5549367088607595, "grad_norm": 0.5845404969197919, "learning_rate": 3.998421649528582e-05, "loss": 0.4529, "step": 137 }, { "epoch": 0.5589873417721519, "grad_norm": 0.5507574777072013, "learning_rate": 3.9981881533475234e-05, "loss": 0.4759, "step": 138 }, { "epoch": 0.5630379746835443, "grad_norm": 0.5788995081902416, "learning_rate": 3.997938564033779e-05, "loss": 0.4694, "step": 139 }, { "epoch": 0.5670886075949367, "grad_norm": 0.47306766541902345, "learning_rate": 3.9976728835975064e-05, "loss": 0.4523, "step": 140 }, { "epoch": 0.5711392405063291, "grad_norm": 0.5578094314798875, "learning_rate": 3.9973911141784605e-05, "loss": 0.4737, "step": 141 }, { "epoch": 0.5751898734177215, "grad_norm": 0.587613924284837, "learning_rate": 3.997093258045973e-05, "loss": 0.4676, "step": 142 }, { "epoch": 0.579240506329114, "grad_norm": 0.6675556450393575, "learning_rate": 3.996779317598936e-05, "loss": 0.486, "step": 143 }, { "epoch": 0.5832911392405064, "grad_norm": 0.5429977507423774, "learning_rate": 3.996449295365782e-05, "loss": 0.4591, "step": 144 }, { "epoch": 0.5873417721518988, "grad_norm": 0.5551091908547249, "learning_rate": 3.996103194004467e-05, "loss": 0.4604, "step": 145 }, { "epoch": 0.5913924050632912, "grad_norm": 0.5236884504029311, "learning_rate": 3.995741016302441e-05, "loss": 0.4436, "step": 146 }, { "epoch": 0.5954430379746836, "grad_norm": 0.5932152799666758, "learning_rate": 3.9953627651766364e-05, "loss": 0.4694, "step": 147 }, { "epoch": 0.599493670886076, "grad_norm": 0.42940159404433464, "learning_rate": 3.9949684436734325e-05, "loss": 0.4524, "step": 148 }, { "epoch": 0.6035443037974684, "grad_norm": 0.4729916270112219, "learning_rate": 3.994558054968643e-05, "loss": 0.4542, "step": 149 }, { "epoch": 0.6075949367088608, "grad_norm": 0.6452380119123143, "learning_rate": 3.994131602367481e-05, "loss": 0.4668, "step": 150 }, { "epoch": 0.6116455696202532, "grad_norm": 0.5694765373888674, "learning_rate": 3.9936890893045376e-05, "loss": 0.4571, "step": 151 }, { "epoch": 0.6156962025316456, "grad_norm": 0.5289939776189743, "learning_rate": 3.993230519343752e-05, "loss": 0.4673, "step": 152 }, { "epoch": 0.619746835443038, "grad_norm": 0.6706922568648612, "learning_rate": 3.992755896178383e-05, "loss": 0.4622, "step": 153 }, { "epoch": 0.6237974683544304, "grad_norm": 0.4911565253842146, "learning_rate": 3.992265223630981e-05, "loss": 0.4515, "step": 154 }, { "epoch": 0.6278481012658228, "grad_norm": 0.6319619776012189, "learning_rate": 3.991758505653355e-05, "loss": 0.4592, "step": 155 }, { "epoch": 0.6318987341772152, "grad_norm": 0.5541760259967372, "learning_rate": 3.991235746326543e-05, "loss": 0.4558, "step": 156 }, { "epoch": 0.6359493670886076, "grad_norm": 1.20808974791757, "learning_rate": 3.9906969498607745e-05, "loss": 0.4528, "step": 157 }, { "epoch": 0.64, "grad_norm": 0.4550861588839403, "learning_rate": 3.990142120595444e-05, "loss": 0.4523, "step": 158 }, { "epoch": 0.6440506329113924, "grad_norm": 0.513967742807425, "learning_rate": 3.98957126299907e-05, "loss": 0.4707, "step": 159 }, { "epoch": 0.6481012658227848, "grad_norm": 0.5646641537877993, "learning_rate": 3.9889843816692596e-05, "loss": 0.4515, "step": 160 }, { "epoch": 0.6521518987341772, "grad_norm": 0.6227473985109685, "learning_rate": 3.9883814813326766e-05, "loss": 0.4656, "step": 161 }, { "epoch": 0.6562025316455696, "grad_norm": 0.4035712931223199, "learning_rate": 3.9877625668449956e-05, "loss": 0.4462, "step": 162 }, { "epoch": 0.660253164556962, "grad_norm": 0.5978179114014814, "learning_rate": 3.98712764319087e-05, "loss": 0.4524, "step": 163 }, { "epoch": 0.6643037974683544, "grad_norm": 0.5844591629699496, "learning_rate": 3.9864767154838864e-05, "loss": 0.4706, "step": 164 }, { "epoch": 0.6683544303797468, "grad_norm": 0.5653575014055141, "learning_rate": 3.9858097889665277e-05, "loss": 0.4607, "step": 165 }, { "epoch": 0.6724050632911392, "grad_norm": 0.9867601362820566, "learning_rate": 3.985126869010129e-05, "loss": 0.4688, "step": 166 }, { "epoch": 0.6764556962025317, "grad_norm": 0.7026010347230134, "learning_rate": 3.984427961114833e-05, "loss": 0.4596, "step": 167 }, { "epoch": 0.6805063291139241, "grad_norm": 0.5055972697923043, "learning_rate": 3.9837130709095475e-05, "loss": 0.4671, "step": 168 }, { "epoch": 0.6845569620253165, "grad_norm": 0.5405975434223228, "learning_rate": 3.982982204151901e-05, "loss": 0.4554, "step": 169 }, { "epoch": 0.6886075949367089, "grad_norm": 0.4858036637584922, "learning_rate": 3.982235366728193e-05, "loss": 0.4638, "step": 170 }, { "epoch": 0.6926582278481013, "grad_norm": 0.7347743790948159, "learning_rate": 3.9814725646533505e-05, "loss": 0.452, "step": 171 }, { "epoch": 0.6967088607594937, "grad_norm": 0.45038811310466714, "learning_rate": 3.9806938040708746e-05, "loss": 0.4522, "step": 172 }, { "epoch": 0.7007594936708861, "grad_norm": 0.5228083070140135, "learning_rate": 3.9798990912527976e-05, "loss": 0.456, "step": 173 }, { "epoch": 0.7048101265822785, "grad_norm": 0.49408084025324794, "learning_rate": 3.979088432599627e-05, "loss": 0.4464, "step": 174 }, { "epoch": 0.7088607594936709, "grad_norm": 0.4062178360142027, "learning_rate": 3.9782618346402964e-05, "loss": 0.4638, "step": 175 }, { "epoch": 0.7129113924050633, "grad_norm": 0.47889379638923885, "learning_rate": 3.977419304032111e-05, "loss": 0.4534, "step": 176 }, { "epoch": 0.7169620253164557, "grad_norm": 0.4679040532263957, "learning_rate": 3.976560847560697e-05, "loss": 0.4471, "step": 177 }, { "epoch": 0.7210126582278481, "grad_norm": 0.4360749701300504, "learning_rate": 3.9756864721399456e-05, "loss": 0.4463, "step": 178 }, { "epoch": 0.7250632911392405, "grad_norm": 1.2668442761082717, "learning_rate": 3.974796184811956e-05, "loss": 0.4452, "step": 179 }, { "epoch": 0.7291139240506329, "grad_norm": 0.4043256217595065, "learning_rate": 3.973889992746979e-05, "loss": 0.453, "step": 180 }, { "epoch": 0.7331645569620253, "grad_norm": 0.3929741617263602, "learning_rate": 3.972967903243361e-05, "loss": 0.4484, "step": 181 }, { "epoch": 0.7372151898734177, "grad_norm": 0.5726755377129074, "learning_rate": 3.972029923727486e-05, "loss": 0.4519, "step": 182 }, { "epoch": 0.7412658227848101, "grad_norm": 0.5322168484793741, "learning_rate": 3.971076061753709e-05, "loss": 0.46, "step": 183 }, { "epoch": 0.7453164556962025, "grad_norm": 0.4022328605040161, "learning_rate": 3.9701063250043066e-05, "loss": 0.4552, "step": 184 }, { "epoch": 0.7493670886075949, "grad_norm": 0.46657749121389286, "learning_rate": 3.969120721289402e-05, "loss": 0.4617, "step": 185 }, { "epoch": 0.7534177215189873, "grad_norm": 0.5054315882793944, "learning_rate": 3.9681192585469146e-05, "loss": 0.4558, "step": 186 }, { "epoch": 0.7574683544303797, "grad_norm": 0.525269323159095, "learning_rate": 3.9671019448424865e-05, "loss": 0.4597, "step": 187 }, { "epoch": 0.7615189873417721, "grad_norm": 0.638640585763432, "learning_rate": 3.966068788369422e-05, "loss": 0.4538, "step": 188 }, { "epoch": 0.7655696202531646, "grad_norm": 0.6901134410284889, "learning_rate": 3.965019797448622e-05, "loss": 0.4447, "step": 189 }, { "epoch": 0.769620253164557, "grad_norm": 0.5690778412789963, "learning_rate": 3.963954980528515e-05, "loss": 0.4548, "step": 190 }, { "epoch": 0.7736708860759494, "grad_norm": 0.4136360502062, "learning_rate": 3.9628743461849905e-05, "loss": 0.4679, "step": 191 }, { "epoch": 0.7777215189873418, "grad_norm": 1.2077145496908277, "learning_rate": 3.961777903121329e-05, "loss": 0.467, "step": 192 }, { "epoch": 0.7817721518987342, "grad_norm": 0.5759780966544537, "learning_rate": 3.960665660168131e-05, "loss": 0.4574, "step": 193 }, { "epoch": 0.7858227848101266, "grad_norm": 0.5064005324711073, "learning_rate": 3.9595376262832485e-05, "loss": 0.4471, "step": 194 }, { "epoch": 0.789873417721519, "grad_norm": 0.457545974013236, "learning_rate": 3.9583938105517127e-05, "loss": 0.4514, "step": 195 }, { "epoch": 0.7939240506329114, "grad_norm": 0.5682310318840603, "learning_rate": 3.957234222185657e-05, "loss": 0.4441, "step": 196 }, { "epoch": 0.7979746835443038, "grad_norm": 0.4875990674152402, "learning_rate": 3.9560588705242474e-05, "loss": 0.4522, "step": 197 }, { "epoch": 0.8020253164556962, "grad_norm": 0.547576500978827, "learning_rate": 3.954867765033605e-05, "loss": 0.4589, "step": 198 }, { "epoch": 0.8060759493670886, "grad_norm": 0.5637853570343997, "learning_rate": 3.953660915306728e-05, "loss": 0.4664, "step": 199 }, { "epoch": 0.810126582278481, "grad_norm": 0.5880564804758105, "learning_rate": 3.952438331063419e-05, "loss": 0.4369, "step": 200 }, { "epoch": 0.8141772151898734, "grad_norm": 0.45042086316730334, "learning_rate": 3.951200022150205e-05, "loss": 0.4553, "step": 201 }, { "epoch": 0.8182278481012658, "grad_norm": 0.6070273905890764, "learning_rate": 3.949945998540253e-05, "loss": 0.4519, "step": 202 }, { "epoch": 0.8222784810126582, "grad_norm": 0.6403122383488195, "learning_rate": 3.9486762703332993e-05, "loss": 0.4617, "step": 203 }, { "epoch": 0.8263291139240506, "grad_norm": 0.4802260287826234, "learning_rate": 3.947390847755559e-05, "loss": 0.4615, "step": 204 }, { "epoch": 0.830379746835443, "grad_norm": 0.4130121924164318, "learning_rate": 3.946089741159648e-05, "loss": 0.4452, "step": 205 }, { "epoch": 0.8344303797468354, "grad_norm": 0.4389147806595703, "learning_rate": 3.944772961024501e-05, "loss": 0.4518, "step": 206 }, { "epoch": 0.8384810126582278, "grad_norm": 0.42946750007511586, "learning_rate": 3.943440517955285e-05, "loss": 0.4634, "step": 207 }, { "epoch": 0.8425316455696202, "grad_norm": 0.43434742395040316, "learning_rate": 3.9420924226833126e-05, "loss": 0.4461, "step": 208 }, { "epoch": 0.8465822784810126, "grad_norm": 0.5494606177268588, "learning_rate": 3.9407286860659566e-05, "loss": 0.4635, "step": 209 }, { "epoch": 0.850632911392405, "grad_norm": 0.5639218738297073, "learning_rate": 3.9393493190865657e-05, "loss": 0.4566, "step": 210 }, { "epoch": 0.8546835443037974, "grad_norm": 0.4944063738098886, "learning_rate": 3.937954332854371e-05, "loss": 0.4481, "step": 211 }, { "epoch": 0.8587341772151899, "grad_norm": 0.627403551711974, "learning_rate": 3.9365437386044016e-05, "loss": 0.4656, "step": 212 }, { "epoch": 0.8627848101265823, "grad_norm": 0.9534155727838589, "learning_rate": 3.935117547697387e-05, "loss": 0.4558, "step": 213 }, { "epoch": 0.8668354430379747, "grad_norm": 0.8115189388164498, "learning_rate": 3.933675771619675e-05, "loss": 0.4615, "step": 214 }, { "epoch": 0.8708860759493671, "grad_norm": 0.5963477776517585, "learning_rate": 3.932218421983131e-05, "loss": 0.4519, "step": 215 }, { "epoch": 0.8749367088607595, "grad_norm": 0.6902480258752931, "learning_rate": 3.9307455105250484e-05, "loss": 0.4581, "step": 216 }, { "epoch": 0.8789873417721519, "grad_norm": 0.5253440234892756, "learning_rate": 3.929257049108054e-05, "loss": 0.4326, "step": 217 }, { "epoch": 0.8830379746835443, "grad_norm": 0.48631465143975744, "learning_rate": 3.927753049720011e-05, "loss": 0.4369, "step": 218 }, { "epoch": 0.8870886075949367, "grad_norm": 0.5447424569107017, "learning_rate": 3.9262335244739234e-05, "loss": 0.4526, "step": 219 }, { "epoch": 0.8911392405063291, "grad_norm": 0.528615873990411, "learning_rate": 3.92469848560784e-05, "loss": 0.4433, "step": 220 }, { "epoch": 0.8951898734177215, "grad_norm": 0.5844521165072634, "learning_rate": 3.923147945484751e-05, "loss": 0.4397, "step": 221 }, { "epoch": 0.8992405063291139, "grad_norm": 0.5098741092232181, "learning_rate": 3.9215819165924956e-05, "loss": 0.4384, "step": 222 }, { "epoch": 0.9032911392405063, "grad_norm": 0.6294553075872699, "learning_rate": 3.920000411543654e-05, "loss": 0.4538, "step": 223 }, { "epoch": 0.9073417721518987, "grad_norm": 0.6110386987392465, "learning_rate": 3.9184034430754495e-05, "loss": 0.4509, "step": 224 }, { "epoch": 0.9113924050632911, "grad_norm": 0.4846978316684442, "learning_rate": 3.916791024049648e-05, "loss": 0.4335, "step": 225 }, { "epoch": 0.9154430379746835, "grad_norm": 0.619992657381423, "learning_rate": 3.91516316745245e-05, "loss": 0.4599, "step": 226 }, { "epoch": 0.9194936708860759, "grad_norm": 0.43137838555735036, "learning_rate": 3.913519886394389e-05, "loss": 0.4506, "step": 227 }, { "epoch": 0.9235443037974683, "grad_norm": 0.5408005488726805, "learning_rate": 3.911861194110225e-05, "loss": 0.444, "step": 228 }, { "epoch": 0.9275949367088607, "grad_norm": 0.4323557926296712, "learning_rate": 3.910187103958837e-05, "loss": 0.4326, "step": 229 }, { "epoch": 0.9316455696202531, "grad_norm": 0.6499192933088839, "learning_rate": 3.908497629423117e-05, "loss": 0.4467, "step": 230 }, { "epoch": 0.9356962025316455, "grad_norm": 0.4101551675320743, "learning_rate": 3.9067927841098614e-05, "loss": 0.4405, "step": 231 }, { "epoch": 0.9397468354430379, "grad_norm": 0.6080366554972899, "learning_rate": 3.9050725817496594e-05, "loss": 0.4412, "step": 232 }, { "epoch": 0.9437974683544303, "grad_norm": 0.4954675940109334, "learning_rate": 3.9033370361967844e-05, "loss": 0.451, "step": 233 }, { "epoch": 0.9478481012658228, "grad_norm": 0.4757295575886923, "learning_rate": 3.901586161429081e-05, "loss": 0.4459, "step": 234 }, { "epoch": 0.9518987341772152, "grad_norm": 0.5454484768758371, "learning_rate": 3.8998199715478545e-05, "loss": 0.4574, "step": 235 }, { "epoch": 0.9559493670886076, "grad_norm": 0.520437725745447, "learning_rate": 3.8980384807777564e-05, "loss": 0.4453, "step": 236 }, { "epoch": 0.96, "grad_norm": 0.502503557322242, "learning_rate": 3.896241703466667e-05, "loss": 0.4581, "step": 237 }, { "epoch": 0.9640506329113924, "grad_norm": 0.3804479750907366, "learning_rate": 3.894429654085585e-05, "loss": 0.45, "step": 238 }, { "epoch": 0.9681012658227848, "grad_norm": 0.42843029817624884, "learning_rate": 3.892602347228505e-05, "loss": 0.4511, "step": 239 }, { "epoch": 0.9721518987341772, "grad_norm": 0.3450892771695689, "learning_rate": 3.890759797612307e-05, "loss": 0.4432, "step": 240 }, { "epoch": 0.9762025316455696, "grad_norm": 0.4988305832373456, "learning_rate": 3.888902020076632e-05, "loss": 0.4456, "step": 241 }, { "epoch": 0.980253164556962, "grad_norm": 0.440636215685347, "learning_rate": 3.887029029583764e-05, "loss": 0.451, "step": 242 }, { "epoch": 0.9843037974683544, "grad_norm": 0.3770093961850818, "learning_rate": 3.8851408412185125e-05, "loss": 0.4423, "step": 243 }, { "epoch": 0.9883544303797468, "grad_norm": 0.49408254743874336, "learning_rate": 3.8832374701880855e-05, "loss": 0.4496, "step": 244 }, { "epoch": 0.9924050632911392, "grad_norm": 0.3562574275838168, "learning_rate": 3.881318931821972e-05, "loss": 0.4339, "step": 245 }, { "epoch": 0.9964556962025316, "grad_norm": 0.393700107062393, "learning_rate": 3.879385241571817e-05, "loss": 0.4323, "step": 246 }, { "epoch": 1.000506329113924, "grad_norm": 0.4293997147407447, "learning_rate": 3.8774364150112955e-05, "loss": 0.4396, "step": 247 }, { "epoch": 1.0045569620253165, "grad_norm": 0.40437836407453115, "learning_rate": 3.8754724678359884e-05, "loss": 0.4141, "step": 248 }, { "epoch": 1.0086075949367088, "grad_norm": 0.3633809774581597, "learning_rate": 3.873493415863256e-05, "loss": 0.406, "step": 249 }, { "epoch": 1.0126582278481013, "grad_norm": 0.4049749511455976, "learning_rate": 3.871499275032111e-05, "loss": 0.4049, "step": 250 }, { "epoch": 1.0167088607594936, "grad_norm": 0.4890247954087134, "learning_rate": 3.869490061403091e-05, "loss": 0.4071, "step": 251 }, { "epoch": 1.0207594936708861, "grad_norm": 0.5410497673079426, "learning_rate": 3.867465791158124e-05, "loss": 0.4169, "step": 252 }, { "epoch": 1.0248101265822784, "grad_norm": 0.5647130483111472, "learning_rate": 3.865426480600407e-05, "loss": 0.4197, "step": 253 }, { "epoch": 1.028860759493671, "grad_norm": 0.5823043777872356, "learning_rate": 3.863372146154264e-05, "loss": 0.4038, "step": 254 }, { "epoch": 1.0329113924050632, "grad_norm": 0.37166334665081135, "learning_rate": 3.861302804365024e-05, "loss": 0.4046, "step": 255 }, { "epoch": 1.0369620253164558, "grad_norm": 0.651847558325141, "learning_rate": 3.85921847189888e-05, "loss": 0.4282, "step": 256 }, { "epoch": 1.041012658227848, "grad_norm": 0.6322552092538848, "learning_rate": 3.85711916554276e-05, "loss": 0.4097, "step": 257 }, { "epoch": 1.0450632911392406, "grad_norm": 0.4849356610309448, "learning_rate": 3.85500490220419e-05, "loss": 0.4247, "step": 258 }, { "epoch": 1.0491139240506329, "grad_norm": 0.5564981378269173, "learning_rate": 3.852875698911154e-05, "loss": 0.4063, "step": 259 }, { "epoch": 1.0531645569620254, "grad_norm": 0.43876176866315497, "learning_rate": 3.850731572811963e-05, "loss": 0.4012, "step": 260 }, { "epoch": 1.0572151898734177, "grad_norm": 0.9079066128222953, "learning_rate": 3.848572541175116e-05, "loss": 0.4081, "step": 261 }, { "epoch": 1.0612658227848102, "grad_norm": 0.5307506717176992, "learning_rate": 3.846398621389154e-05, "loss": 0.4217, "step": 262 }, { "epoch": 1.0653164556962025, "grad_norm": 0.45318268346474966, "learning_rate": 3.84420983096253e-05, "loss": 0.4069, "step": 263 }, { "epoch": 1.069367088607595, "grad_norm": 0.628381888826879, "learning_rate": 3.8420061875234606e-05, "loss": 0.4168, "step": 264 }, { "epoch": 1.0734177215189873, "grad_norm": 0.5544584688734254, "learning_rate": 3.839787708819787e-05, "loss": 0.4058, "step": 265 }, { "epoch": 1.0774683544303798, "grad_norm": 0.48951314044054384, "learning_rate": 3.8375544127188325e-05, "loss": 0.4047, "step": 266 }, { "epoch": 1.081518987341772, "grad_norm": 0.4385220657102478, "learning_rate": 3.8353063172072564e-05, "loss": 0.4056, "step": 267 }, { "epoch": 1.0855696202531646, "grad_norm": 0.6010794367110275, "learning_rate": 3.8330434403909105e-05, "loss": 0.4122, "step": 268 }, { "epoch": 1.089620253164557, "grad_norm": 0.7787564555233977, "learning_rate": 3.8307658004946934e-05, "loss": 0.4094, "step": 269 }, { "epoch": 1.0936708860759494, "grad_norm": 0.38696257240713183, "learning_rate": 3.8284734158624046e-05, "loss": 0.3998, "step": 270 }, { "epoch": 1.0977215189873417, "grad_norm": 0.46673171634744814, "learning_rate": 3.826166304956594e-05, "loss": 0.407, "step": 271 }, { "epoch": 1.1017721518987342, "grad_norm": 0.5108988927425662, "learning_rate": 3.8238444863584164e-05, "loss": 0.4051, "step": 272 }, { "epoch": 1.1058227848101265, "grad_norm": 0.4092455885328882, "learning_rate": 3.821507978767479e-05, "loss": 0.4124, "step": 273 }, { "epoch": 1.109873417721519, "grad_norm": 0.4786895805902475, "learning_rate": 3.819156801001693e-05, "loss": 0.4199, "step": 274 }, { "epoch": 1.1139240506329113, "grad_norm": 0.47786839657684677, "learning_rate": 3.816790971997121e-05, "loss": 0.4042, "step": 275 }, { "epoch": 1.1179746835443038, "grad_norm": 0.3285332849970864, "learning_rate": 3.8144105108078246e-05, "loss": 0.4078, "step": 276 }, { "epoch": 1.1220253164556961, "grad_norm": 0.418369153216123, "learning_rate": 3.81201543660571e-05, "loss": 0.4098, "step": 277 }, { "epoch": 1.1260759493670887, "grad_norm": 0.5178001843359239, "learning_rate": 3.809605768680377e-05, "loss": 0.4097, "step": 278 }, { "epoch": 1.130126582278481, "grad_norm": 0.3779023970570908, "learning_rate": 3.807181526438958e-05, "loss": 0.4108, "step": 279 }, { "epoch": 1.1341772151898735, "grad_norm": 0.3812562550580138, "learning_rate": 3.8047427294059697e-05, "loss": 0.398, "step": 280 }, { "epoch": 1.1382278481012658, "grad_norm": 0.3927532548714664, "learning_rate": 3.802289397223145e-05, "loss": 0.4133, "step": 281 }, { "epoch": 1.1422784810126583, "grad_norm": 0.4472387403436717, "learning_rate": 3.7998215496492854e-05, "loss": 0.4119, "step": 282 }, { "epoch": 1.1463291139240506, "grad_norm": 0.37276808279268686, "learning_rate": 3.797339206560096e-05, "loss": 0.402, "step": 283 }, { "epoch": 1.150379746835443, "grad_norm": 0.378082102795398, "learning_rate": 3.794842387948027e-05, "loss": 0.4083, "step": 284 }, { "epoch": 1.1544303797468354, "grad_norm": 0.4687945687236878, "learning_rate": 3.7923311139221114e-05, "loss": 0.4234, "step": 285 }, { "epoch": 1.158481012658228, "grad_norm": 0.48331785336384586, "learning_rate": 3.7898054047078054e-05, "loss": 0.4234, "step": 286 }, { "epoch": 1.1625316455696202, "grad_norm": 0.3728948930524858, "learning_rate": 3.787265280646825e-05, "loss": 0.4217, "step": 287 }, { "epoch": 1.1665822784810127, "grad_norm": 0.35803880378389674, "learning_rate": 3.7847107621969786e-05, "loss": 0.4096, "step": 288 }, { "epoch": 1.170632911392405, "grad_norm": 0.49977159155717144, "learning_rate": 3.7821418699320064e-05, "loss": 0.4034, "step": 289 }, { "epoch": 1.1746835443037975, "grad_norm": 0.4395444103149848, "learning_rate": 3.7795586245414145e-05, "loss": 0.4059, "step": 290 }, { "epoch": 1.1787341772151898, "grad_norm": 0.36459313241932667, "learning_rate": 3.776961046830306e-05, "loss": 0.4188, "step": 291 }, { "epoch": 1.1827848101265823, "grad_norm": 0.43536618933987103, "learning_rate": 3.774349157719215e-05, "loss": 0.4142, "step": 292 }, { "epoch": 1.1868354430379746, "grad_norm": 1.213319924394842, "learning_rate": 3.7717229782439365e-05, "loss": 0.4611, "step": 293 }, { "epoch": 1.1908860759493671, "grad_norm": 0.366669307775165, "learning_rate": 3.769082529555359e-05, "loss": 0.4075, "step": 294 }, { "epoch": 1.1949367088607594, "grad_norm": 0.46310817721043224, "learning_rate": 3.766427832919294e-05, "loss": 0.4099, "step": 295 }, { "epoch": 1.198987341772152, "grad_norm": 0.41573861192497497, "learning_rate": 3.7637589097163024e-05, "loss": 0.3936, "step": 296 }, { "epoch": 1.2030379746835442, "grad_norm": 0.4445164678352878, "learning_rate": 3.761075781441526e-05, "loss": 0.4115, "step": 297 }, { "epoch": 1.2070886075949367, "grad_norm": 0.40955379531202113, "learning_rate": 3.75837846970451e-05, "loss": 0.414, "step": 298 }, { "epoch": 1.211139240506329, "grad_norm": 0.5218488745785649, "learning_rate": 3.755666996229032e-05, "loss": 0.4043, "step": 299 }, { "epoch": 1.2151898734177216, "grad_norm": 0.42347994822612833, "learning_rate": 3.752941382852927e-05, "loss": 0.4166, "step": 300 }, { "epoch": 1.2192405063291138, "grad_norm": 0.38293741652918867, "learning_rate": 3.7502016515279115e-05, "loss": 0.4025, "step": 301 }, { "epoch": 1.2232911392405064, "grad_norm": 0.463237828891802, "learning_rate": 3.7474478243194043e-05, "loss": 0.4169, "step": 302 }, { "epoch": 1.2273417721518987, "grad_norm": 0.4071988083626525, "learning_rate": 3.744679923406351e-05, "loss": 0.4166, "step": 303 }, { "epoch": 1.2313924050632912, "grad_norm": 0.3569818322292202, "learning_rate": 3.741897971081043e-05, "loss": 0.4152, "step": 304 }, { "epoch": 1.2354430379746835, "grad_norm": 0.3523833599087922, "learning_rate": 3.739101989748946e-05, "loss": 0.4128, "step": 305 }, { "epoch": 1.239493670886076, "grad_norm": 0.37760175754585296, "learning_rate": 3.7362920019285066e-05, "loss": 0.4276, "step": 306 }, { "epoch": 1.2435443037974683, "grad_norm": 0.3888698418420509, "learning_rate": 3.73346803025098e-05, "loss": 0.4117, "step": 307 }, { "epoch": 1.2475949367088608, "grad_norm": 0.3590205348799031, "learning_rate": 3.730630097460247e-05, "loss": 0.407, "step": 308 }, { "epoch": 1.251645569620253, "grad_norm": 0.3967754178644449, "learning_rate": 3.727778226412628e-05, "loss": 0.4193, "step": 309 }, { "epoch": 1.2556962025316456, "grad_norm": 0.40298461522477697, "learning_rate": 3.7249124400767006e-05, "loss": 0.4077, "step": 310 }, { "epoch": 1.259746835443038, "grad_norm": 0.37485308837944686, "learning_rate": 3.722032761533114e-05, "loss": 0.405, "step": 311 }, { "epoch": 1.2637974683544304, "grad_norm": 0.4081569500487906, "learning_rate": 3.719139213974403e-05, "loss": 0.4008, "step": 312 }, { "epoch": 1.2678481012658227, "grad_norm": 0.40142419840136745, "learning_rate": 3.7162318207048006e-05, "loss": 0.4072, "step": 313 }, { "epoch": 1.2718987341772152, "grad_norm": 0.3727397197774807, "learning_rate": 3.713310605140055e-05, "loss": 0.401, "step": 314 }, { "epoch": 1.2759493670886077, "grad_norm": 0.42756129620489264, "learning_rate": 3.710375590807233e-05, "loss": 0.4179, "step": 315 }, { "epoch": 1.28, "grad_norm": 0.36129696247139487, "learning_rate": 3.7074268013445365e-05, "loss": 0.4136, "step": 316 }, { "epoch": 1.2840506329113923, "grad_norm": 0.5027776250966887, "learning_rate": 3.7044642605011114e-05, "loss": 0.4205, "step": 317 }, { "epoch": 1.2881012658227848, "grad_norm": 0.4267302876239031, "learning_rate": 3.701487992136854e-05, "loss": 0.4004, "step": 318 }, { "epoch": 1.2921518987341774, "grad_norm": 0.3504270636564766, "learning_rate": 3.69849802022222e-05, "loss": 0.4058, "step": 319 }, { "epoch": 1.2962025316455696, "grad_norm": 0.45106181908648374, "learning_rate": 3.6954943688380334e-05, "loss": 0.4145, "step": 320 }, { "epoch": 1.300253164556962, "grad_norm": 0.483900593892956, "learning_rate": 3.692477062175289e-05, "loss": 0.3944, "step": 321 }, { "epoch": 1.3043037974683545, "grad_norm": 0.38431547791686654, "learning_rate": 3.689446124534958e-05, "loss": 0.4167, "step": 322 }, { "epoch": 1.308354430379747, "grad_norm": 0.4073645485924419, "learning_rate": 3.686401580327799e-05, "loss": 0.4036, "step": 323 }, { "epoch": 1.3124050632911393, "grad_norm": 0.3916980138986248, "learning_rate": 3.683343454074149e-05, "loss": 0.4095, "step": 324 }, { "epoch": 1.3164556962025316, "grad_norm": 0.3724229171074615, "learning_rate": 3.6802717704037386e-05, "loss": 0.4019, "step": 325 }, { "epoch": 1.320506329113924, "grad_norm": 0.4662449337720794, "learning_rate": 3.6771865540554855e-05, "loss": 0.4062, "step": 326 }, { "epoch": 1.3245569620253166, "grad_norm": 0.36877742197640206, "learning_rate": 3.674087829877297e-05, "loss": 0.3877, "step": 327 }, { "epoch": 1.3286075949367089, "grad_norm": 0.5132746702018466, "learning_rate": 3.6709756228258735e-05, "loss": 0.4107, "step": 328 }, { "epoch": 1.3326582278481012, "grad_norm": 0.4144946617731807, "learning_rate": 3.667849957966501e-05, "loss": 0.4063, "step": 329 }, { "epoch": 1.3367088607594937, "grad_norm": 0.4716406852636758, "learning_rate": 3.6647108604728546e-05, "loss": 0.4065, "step": 330 }, { "epoch": 1.3407594936708862, "grad_norm": 0.47617446964437954, "learning_rate": 3.661558355626795e-05, "loss": 0.4185, "step": 331 }, { "epoch": 1.3448101265822785, "grad_norm": 0.35565215094664676, "learning_rate": 3.658392468818163e-05, "loss": 0.4092, "step": 332 }, { "epoch": 1.3488607594936708, "grad_norm": 9.366565233337584, "learning_rate": 3.655213225544574e-05, "loss": 0.4427, "step": 333 }, { "epoch": 1.3529113924050633, "grad_norm": 0.7012922815128182, "learning_rate": 3.652020651411218e-05, "loss": 0.4005, "step": 334 }, { "epoch": 1.3569620253164558, "grad_norm": 0.7559871505053798, "learning_rate": 3.6488147721306474e-05, "loss": 0.4098, "step": 335 }, { "epoch": 1.3610126582278481, "grad_norm": 0.49883330283620053, "learning_rate": 3.645595613522574e-05, "loss": 0.4153, "step": 336 }, { "epoch": 1.3650632911392404, "grad_norm": 0.6030132201000052, "learning_rate": 3.642363201513657e-05, "loss": 0.4104, "step": 337 }, { "epoch": 1.369113924050633, "grad_norm": 0.4688844824117302, "learning_rate": 3.6391175621373006e-05, "loss": 0.4203, "step": 338 }, { "epoch": 1.3731645569620254, "grad_norm": 0.628658167514011, "learning_rate": 3.6358587215334355e-05, "loss": 0.404, "step": 339 }, { "epoch": 1.3772151898734177, "grad_norm": 0.928985419313828, "learning_rate": 3.632586705948318e-05, "loss": 0.4298, "step": 340 }, { "epoch": 1.38126582278481, "grad_norm": 0.5874852640678295, "learning_rate": 3.629301541734311e-05, "loss": 0.4051, "step": 341 }, { "epoch": 1.3853164556962025, "grad_norm": 0.5981008062377415, "learning_rate": 3.626003255349676e-05, "loss": 0.4186, "step": 342 }, { "epoch": 1.389367088607595, "grad_norm": 0.5125470889155176, "learning_rate": 3.622691873358357e-05, "loss": 0.401, "step": 343 }, { "epoch": 1.3934177215189874, "grad_norm": 0.6487254290982464, "learning_rate": 3.61936742242977e-05, "loss": 0.404, "step": 344 }, { "epoch": 1.3974683544303796, "grad_norm": 0.4190299124209669, "learning_rate": 3.6160299293385864e-05, "loss": 0.4061, "step": 345 }, { "epoch": 1.4015189873417722, "grad_norm": 0.5568734008265345, "learning_rate": 3.612679420964516e-05, "loss": 0.4091, "step": 346 }, { "epoch": 1.4055696202531647, "grad_norm": 0.3479861717295567, "learning_rate": 3.609315924292092e-05, "loss": 0.4187, "step": 347 }, { "epoch": 1.409620253164557, "grad_norm": 0.5219845086426631, "learning_rate": 3.6059394664104554e-05, "loss": 0.414, "step": 348 }, { "epoch": 1.4136708860759493, "grad_norm": 0.3829340113917253, "learning_rate": 3.602550074513133e-05, "loss": 0.401, "step": 349 }, { "epoch": 1.4177215189873418, "grad_norm": 0.45256380867462437, "learning_rate": 3.599147775897822e-05, "loss": 0.4047, "step": 350 }, { "epoch": 1.4217721518987343, "grad_norm": 0.3377350434731355, "learning_rate": 3.595732597966167e-05, "loss": 0.4046, "step": 351 }, { "epoch": 1.4258227848101266, "grad_norm": 0.45534360107352684, "learning_rate": 3.592304568223542e-05, "loss": 0.4105, "step": 352 }, { "epoch": 1.4298734177215189, "grad_norm": 0.3382184296261859, "learning_rate": 3.588863714278826e-05, "loss": 0.4173, "step": 353 }, { "epoch": 1.4339240506329114, "grad_norm": 0.34454587479545573, "learning_rate": 3.585410063844186e-05, "loss": 0.394, "step": 354 }, { "epoch": 1.437974683544304, "grad_norm": 0.3629241477692016, "learning_rate": 3.581943644734846e-05, "loss": 0.4003, "step": 355 }, { "epoch": 1.4420253164556962, "grad_norm": 0.35794596020864655, "learning_rate": 3.578464484868869e-05, "loss": 0.4097, "step": 356 }, { "epoch": 1.4460759493670885, "grad_norm": 0.6867670650859381, "learning_rate": 3.5749726122669316e-05, "loss": 0.4473, "step": 357 }, { "epoch": 1.450126582278481, "grad_norm": 0.36619110600282573, "learning_rate": 3.5714680550520943e-05, "loss": 0.4064, "step": 358 }, { "epoch": 1.4541772151898735, "grad_norm": 0.3502601645753421, "learning_rate": 3.5679508414495794e-05, "loss": 0.4189, "step": 359 }, { "epoch": 1.4582278481012658, "grad_norm": 0.3524033703039227, "learning_rate": 3.564420999786543e-05, "loss": 0.4108, "step": 360 }, { "epoch": 1.4622784810126581, "grad_norm": 0.3452711877499318, "learning_rate": 3.560878558491842e-05, "loss": 0.4076, "step": 361 }, { "epoch": 1.4663291139240506, "grad_norm": 0.3369056810256558, "learning_rate": 3.5573235460958145e-05, "loss": 0.386, "step": 362 }, { "epoch": 1.4703797468354431, "grad_norm": 0.3076398852203883, "learning_rate": 3.553755991230039e-05, "loss": 0.4056, "step": 363 }, { "epoch": 1.4744303797468354, "grad_norm": 0.4188531533786415, "learning_rate": 3.5501759226271144e-05, "loss": 0.4142, "step": 364 }, { "epoch": 1.4784810126582277, "grad_norm": 0.3390823863031162, "learning_rate": 3.546583369120419e-05, "loss": 0.4106, "step": 365 }, { "epoch": 1.4825316455696202, "grad_norm": 0.36836724958794437, "learning_rate": 3.5429783596438864e-05, "loss": 0.4062, "step": 366 }, { "epoch": 1.4865822784810128, "grad_norm": 0.402447393167961, "learning_rate": 3.539360923231766e-05, "loss": 0.4062, "step": 367 }, { "epoch": 1.490632911392405, "grad_norm": 0.35066327037440526, "learning_rate": 3.535731089018394e-05, "loss": 0.4048, "step": 368 }, { "epoch": 1.4946835443037974, "grad_norm": 0.37851180259196354, "learning_rate": 3.532088886237956e-05, "loss": 0.4093, "step": 369 }, { "epoch": 1.4987341772151899, "grad_norm": 0.42126721755723956, "learning_rate": 3.528434344224253e-05, "loss": 0.4132, "step": 370 }, { "epoch": 1.5027848101265824, "grad_norm": 0.3840047301880263, "learning_rate": 3.524767492410464e-05, "loss": 0.4102, "step": 371 }, { "epoch": 1.5068354430379747, "grad_norm": 0.4123243851256507, "learning_rate": 3.521088360328908e-05, "loss": 0.4191, "step": 372 }, { "epoch": 1.510886075949367, "grad_norm": 0.45510450504068983, "learning_rate": 3.517396977610811e-05, "loss": 0.4076, "step": 373 }, { "epoch": 1.5149367088607595, "grad_norm": 0.43633437671170994, "learning_rate": 3.5136933739860595e-05, "loss": 0.3863, "step": 374 }, { "epoch": 1.518987341772152, "grad_norm": 0.46343317992146377, "learning_rate": 3.509977579282971e-05, "loss": 0.4141, "step": 375 }, { "epoch": 1.5230379746835443, "grad_norm": 0.3700760888529492, "learning_rate": 3.5062496234280424e-05, "loss": 0.407, "step": 376 }, { "epoch": 1.5270886075949366, "grad_norm": 0.4233647515093545, "learning_rate": 3.502509536445719e-05, "loss": 0.4179, "step": 377 }, { "epoch": 1.531139240506329, "grad_norm": 0.31398221287498335, "learning_rate": 3.498757348458147e-05, "loss": 0.4026, "step": 378 }, { "epoch": 1.5351898734177216, "grad_norm": 0.32646975494840935, "learning_rate": 3.4949930896849324e-05, "loss": 0.3915, "step": 379 }, { "epoch": 1.539240506329114, "grad_norm": 0.3183561704423831, "learning_rate": 3.491216790442899e-05, "loss": 0.4043, "step": 380 }, { "epoch": 1.5432911392405062, "grad_norm": 0.32277330894841516, "learning_rate": 3.487428481145839e-05, "loss": 0.3974, "step": 381 }, { "epoch": 1.5473417721518987, "grad_norm": 0.36746734071755915, "learning_rate": 3.483628192304278e-05, "loss": 0.4146, "step": 382 }, { "epoch": 1.5513924050632912, "grad_norm": 0.35860735993253423, "learning_rate": 3.479815954525219e-05, "loss": 0.4106, "step": 383 }, { "epoch": 1.5554430379746835, "grad_norm": 0.2988897319128841, "learning_rate": 3.475991798511899e-05, "loss": 0.3958, "step": 384 }, { "epoch": 1.5594936708860758, "grad_norm": 0.33082627431904343, "learning_rate": 3.4721557550635464e-05, "loss": 0.4025, "step": 385 }, { "epoch": 1.5635443037974683, "grad_norm": 0.29912496295981306, "learning_rate": 3.468307855075128e-05, "loss": 0.4006, "step": 386 }, { "epoch": 1.5675949367088609, "grad_norm": 0.34291861421461384, "learning_rate": 3.4644481295371005e-05, "loss": 0.4061, "step": 387 }, { "epoch": 1.5716455696202531, "grad_norm": 0.30192340743260465, "learning_rate": 3.460576609535163e-05, "loss": 0.4048, "step": 388 }, { "epoch": 1.5756962025316454, "grad_norm": 0.3334676571542644, "learning_rate": 3.456693326250006e-05, "loss": 0.4075, "step": 389 }, { "epoch": 1.579746835443038, "grad_norm": 0.34571184436202584, "learning_rate": 3.452798310957058e-05, "loss": 0.4096, "step": 390 }, { "epoch": 1.5837974683544305, "grad_norm": 0.40844207161622526, "learning_rate": 3.4488915950262386e-05, "loss": 0.4, "step": 391 }, { "epoch": 1.5878481012658228, "grad_norm": 0.34216979899936756, "learning_rate": 3.4449732099216985e-05, "loss": 0.3973, "step": 392 }, { "epoch": 1.591898734177215, "grad_norm": 0.3687451039909502, "learning_rate": 3.441043187201574e-05, "loss": 0.4047, "step": 393 }, { "epoch": 1.5959493670886076, "grad_norm": 0.389606378925937, "learning_rate": 3.437101558517728e-05, "loss": 0.4012, "step": 394 }, { "epoch": 1.6, "grad_norm": 0.35516005273661355, "learning_rate": 3.433148355615496e-05, "loss": 0.41, "step": 395 }, { "epoch": 1.6040506329113924, "grad_norm": 0.4795883491747531, "learning_rate": 3.4291836103334294e-05, "loss": 0.4188, "step": 396 }, { "epoch": 1.6081012658227847, "grad_norm": 0.4257179074014646, "learning_rate": 3.425207354603043e-05, "loss": 0.3972, "step": 397 }, { "epoch": 1.6121518987341772, "grad_norm": 0.40697677168816465, "learning_rate": 3.421219620448553e-05, "loss": 0.393, "step": 398 }, { "epoch": 1.6162025316455697, "grad_norm": 0.4960354530499013, "learning_rate": 3.417220439986623e-05, "loss": 0.3997, "step": 399 }, { "epoch": 1.620253164556962, "grad_norm": 0.37226707808412557, "learning_rate": 3.4132098454261024e-05, "loss": 0.416, "step": 400 }, { "epoch": 1.6243037974683543, "grad_norm": 0.33006726357472277, "learning_rate": 3.4091878690677676e-05, "loss": 0.406, "step": 401 }, { "epoch": 1.6283544303797468, "grad_norm": 0.446117795027989, "learning_rate": 3.405154543304065e-05, "loss": 0.4113, "step": 402 }, { "epoch": 1.6324050632911393, "grad_norm": 0.32263745080738027, "learning_rate": 3.401109900618843e-05, "loss": 0.4096, "step": 403 }, { "epoch": 1.6364556962025316, "grad_norm": 0.466169685089556, "learning_rate": 3.3970539735870996e-05, "loss": 0.4108, "step": 404 }, { "epoch": 1.640506329113924, "grad_norm": 0.3735439675142462, "learning_rate": 3.392986794874714e-05, "loss": 0.4096, "step": 405 }, { "epoch": 1.6445569620253164, "grad_norm": 0.3968688253507354, "learning_rate": 3.388908397238184e-05, "loss": 0.3988, "step": 406 }, { "epoch": 1.648607594936709, "grad_norm": 0.39619234162770506, "learning_rate": 3.384818813524362e-05, "loss": 0.3918, "step": 407 }, { "epoch": 1.6526582278481012, "grad_norm": 0.3957534525306337, "learning_rate": 3.380718076670195e-05, "loss": 0.4025, "step": 408 }, { "epoch": 1.6567088607594935, "grad_norm": 0.35528659430603127, "learning_rate": 3.376606219702454e-05, "loss": 0.4059, "step": 409 }, { "epoch": 1.660759493670886, "grad_norm": 0.32797680010473895, "learning_rate": 3.372483275737468e-05, "loss": 0.4024, "step": 410 }, { "epoch": 1.6648101265822786, "grad_norm": 0.3954949425211598, "learning_rate": 3.368349277980861e-05, "loss": 0.4139, "step": 411 }, { "epoch": 1.6688607594936709, "grad_norm": 0.30444247852572903, "learning_rate": 3.3642042597272844e-05, "loss": 0.401, "step": 412 }, { "epoch": 1.6729113924050631, "grad_norm": 0.3769095834893029, "learning_rate": 3.360048254360144e-05, "loss": 0.3822, "step": 413 }, { "epoch": 1.6769620253164557, "grad_norm": 0.34046246556161974, "learning_rate": 3.355881295351336e-05, "loss": 0.3995, "step": 414 }, { "epoch": 1.6810126582278482, "grad_norm": 0.49275405864849703, "learning_rate": 3.351703416260975e-05, "loss": 0.3908, "step": 415 }, { "epoch": 1.6850632911392405, "grad_norm": 0.3174355147836213, "learning_rate": 3.347514650737126e-05, "loss": 0.404, "step": 416 }, { "epoch": 1.6891139240506328, "grad_norm": 0.4385144438611822, "learning_rate": 3.3433150325155295e-05, "loss": 0.4001, "step": 417 }, { "epoch": 1.6931645569620253, "grad_norm": 0.32716488896839924, "learning_rate": 3.339104595419334e-05, "loss": 0.4067, "step": 418 }, { "epoch": 1.6972151898734178, "grad_norm": 0.42832621964659173, "learning_rate": 3.3348833733588204e-05, "loss": 0.4046, "step": 419 }, { "epoch": 1.70126582278481, "grad_norm": 0.33906912081689927, "learning_rate": 3.3306514003311305e-05, "loss": 0.3987, "step": 420 }, { "epoch": 1.7053164556962024, "grad_norm": 0.31755895253065974, "learning_rate": 3.326408710419996e-05, "loss": 0.4019, "step": 421 }, { "epoch": 1.709367088607595, "grad_norm": 0.3501237729220953, "learning_rate": 3.322155337795454e-05, "loss": 0.3982, "step": 422 }, { "epoch": 1.7134177215189874, "grad_norm": 0.3107519035734609, "learning_rate": 3.317891316713587e-05, "loss": 0.4042, "step": 423 }, { "epoch": 1.7174683544303797, "grad_norm": 0.35350052810500393, "learning_rate": 3.313616681516231e-05, "loss": 0.3969, "step": 424 }, { "epoch": 1.721518987341772, "grad_norm": 0.3592199739096979, "learning_rate": 3.309331466630713e-05, "loss": 0.4081, "step": 425 }, { "epoch": 1.7255696202531645, "grad_norm": 0.37548086029598593, "learning_rate": 3.305035706569563e-05, "loss": 0.4164, "step": 426 }, { "epoch": 1.729620253164557, "grad_norm": 0.3051727152787476, "learning_rate": 3.3007294359302433e-05, "loss": 0.3898, "step": 427 }, { "epoch": 1.7336708860759493, "grad_norm": 0.3415600034225169, "learning_rate": 3.296412689394864e-05, "loss": 0.4125, "step": 428 }, { "epoch": 1.7377215189873416, "grad_norm": 0.33414296104853713, "learning_rate": 3.292085501729909e-05, "loss": 0.4092, "step": 429 }, { "epoch": 1.7417721518987341, "grad_norm": 0.3343775568961719, "learning_rate": 3.2877479077859534e-05, "loss": 0.3929, "step": 430 }, { "epoch": 1.7458227848101266, "grad_norm": 0.31329042961700254, "learning_rate": 3.283399942497381e-05, "loss": 0.411, "step": 431 }, { "epoch": 1.749873417721519, "grad_norm": 0.3505324601481064, "learning_rate": 3.279041640882108e-05, "loss": 0.3937, "step": 432 }, { "epoch": 1.7539240506329112, "grad_norm": 0.31755086481202827, "learning_rate": 3.2746730380412964e-05, "loss": 0.3893, "step": 433 }, { "epoch": 1.7579746835443038, "grad_norm": 0.36287871389166015, "learning_rate": 3.2702941691590726e-05, "loss": 0.4122, "step": 434 }, { "epoch": 1.7620253164556963, "grad_norm": 0.37387896125947573, "learning_rate": 3.265905069502244e-05, "loss": 0.395, "step": 435 }, { "epoch": 1.7660759493670886, "grad_norm": 0.3367694207327517, "learning_rate": 3.261505774420016e-05, "loss": 0.3986, "step": 436 }, { "epoch": 1.7701265822784809, "grad_norm": 0.36805379718530284, "learning_rate": 3.257096319343707e-05, "loss": 0.4038, "step": 437 }, { "epoch": 1.7741772151898734, "grad_norm": 0.39374503590279686, "learning_rate": 3.2526767397864614e-05, "loss": 0.4094, "step": 438 }, { "epoch": 1.7782278481012659, "grad_norm": 0.3484979221341852, "learning_rate": 3.248247071342966e-05, "loss": 0.3945, "step": 439 }, { "epoch": 1.7822784810126582, "grad_norm": 0.3059415577383054, "learning_rate": 3.243807349689161e-05, "loss": 0.3815, "step": 440 }, { "epoch": 1.7863291139240505, "grad_norm": 0.33864760795519117, "learning_rate": 3.2393576105819544e-05, "loss": 0.4058, "step": 441 }, { "epoch": 1.790379746835443, "grad_norm": 0.38941563901452825, "learning_rate": 3.2348978898589333e-05, "loss": 0.3959, "step": 442 }, { "epoch": 1.7944303797468355, "grad_norm": 0.2859466411309488, "learning_rate": 3.230428223438075e-05, "loss": 0.3909, "step": 443 }, { "epoch": 1.7984810126582278, "grad_norm": 0.42003770937884305, "learning_rate": 3.225948647317459e-05, "loss": 0.39, "step": 444 }, { "epoch": 1.80253164556962, "grad_norm": 0.3407248061161464, "learning_rate": 3.2214591975749745e-05, "loss": 0.4022, "step": 445 }, { "epoch": 1.8065822784810126, "grad_norm": 0.34053553131835085, "learning_rate": 3.216959910368034e-05, "loss": 0.3941, "step": 446 }, { "epoch": 1.8106329113924051, "grad_norm": 0.3541973027573286, "learning_rate": 3.212450821933277e-05, "loss": 0.3857, "step": 447 }, { "epoch": 1.8146835443037974, "grad_norm": 0.35252496113798526, "learning_rate": 3.207931968586281e-05, "loss": 0.4069, "step": 448 }, { "epoch": 1.81873417721519, "grad_norm": 0.3022456084049639, "learning_rate": 3.203403386721272e-05, "loss": 0.4104, "step": 449 }, { "epoch": 1.8227848101265822, "grad_norm": 0.3648744786121604, "learning_rate": 3.1988651128108245e-05, "loss": 0.3979, "step": 450 }, { "epoch": 1.8268354430379747, "grad_norm": 0.3241771405919617, "learning_rate": 3.194317183405573e-05, "loss": 0.4155, "step": 451 }, { "epoch": 1.8308860759493673, "grad_norm": 0.34814445355026885, "learning_rate": 3.189759635133914e-05, "loss": 0.4063, "step": 452 }, { "epoch": 1.8349367088607595, "grad_norm": 0.33363875686942174, "learning_rate": 3.185192504701718e-05, "loss": 0.3907, "step": 453 }, { "epoch": 1.8389873417721518, "grad_norm": 0.5930265463591506, "learning_rate": 3.1806158288920234e-05, "loss": 0.4078, "step": 454 }, { "epoch": 1.8430379746835444, "grad_norm": 0.37357469526822096, "learning_rate": 3.1760296445647477e-05, "loss": 0.4138, "step": 455 }, { "epoch": 1.8470886075949369, "grad_norm": 0.33241238288238634, "learning_rate": 3.1714339886563896e-05, "loss": 0.3964, "step": 456 }, { "epoch": 1.8511392405063292, "grad_norm": 0.4084605471583437, "learning_rate": 3.166828898179731e-05, "loss": 0.4153, "step": 457 }, { "epoch": 1.8551898734177215, "grad_norm": 0.29612168038754294, "learning_rate": 3.162214410223536e-05, "loss": 0.4063, "step": 458 }, { "epoch": 1.859240506329114, "grad_norm": 0.3540587361757342, "learning_rate": 3.157590561952257e-05, "loss": 0.4046, "step": 459 }, { "epoch": 1.8632911392405065, "grad_norm": 0.3177115642750938, "learning_rate": 3.152957390605732e-05, "loss": 0.406, "step": 460 }, { "epoch": 1.8673417721518988, "grad_norm": 0.40521909249477284, "learning_rate": 3.148314933498886e-05, "loss": 0.419, "step": 461 }, { "epoch": 1.871392405063291, "grad_norm": 0.33096346327700726, "learning_rate": 3.143663228021431e-05, "loss": 0.4033, "step": 462 }, { "epoch": 1.8754430379746836, "grad_norm": 0.2891673128870199, "learning_rate": 3.1390023116375624e-05, "loss": 0.3928, "step": 463 }, { "epoch": 1.879493670886076, "grad_norm": 0.3360876935548682, "learning_rate": 3.134332221885661e-05, "loss": 0.4019, "step": 464 }, { "epoch": 1.8835443037974684, "grad_norm": 0.34062735851341924, "learning_rate": 3.129652996377987e-05, "loss": 0.3952, "step": 465 }, { "epoch": 1.8875949367088607, "grad_norm": 0.34664944913629697, "learning_rate": 3.12496467280038e-05, "loss": 0.3983, "step": 466 }, { "epoch": 1.8916455696202532, "grad_norm": 0.3132927529740008, "learning_rate": 3.120267288911952e-05, "loss": 0.4044, "step": 467 }, { "epoch": 1.8956962025316457, "grad_norm": 0.3889673510168891, "learning_rate": 3.11556088254479e-05, "loss": 0.402, "step": 468 }, { "epoch": 1.899746835443038, "grad_norm": 0.3064597902987973, "learning_rate": 3.11084549160364e-05, "loss": 0.3987, "step": 469 }, { "epoch": 1.9037974683544303, "grad_norm": 0.3695874015773935, "learning_rate": 3.106121154065615e-05, "loss": 0.3999, "step": 470 }, { "epoch": 1.9078481012658228, "grad_norm": 0.38805943200644516, "learning_rate": 3.1013879079798805e-05, "loss": 0.404, "step": 471 }, { "epoch": 1.9118987341772153, "grad_norm": 0.305971151663065, "learning_rate": 3.096645791467348e-05, "loss": 0.4054, "step": 472 }, { "epoch": 1.9159493670886076, "grad_norm": 0.3962761602009394, "learning_rate": 3.091894842720373e-05, "loss": 0.4142, "step": 473 }, { "epoch": 1.92, "grad_norm": 0.33140563445044074, "learning_rate": 3.0871351000024425e-05, "loss": 0.387, "step": 474 }, { "epoch": 1.9240506329113924, "grad_norm": 0.3324738049961148, "learning_rate": 3.0823666016478716e-05, "loss": 0.4011, "step": 475 }, { "epoch": 1.928101265822785, "grad_norm": 0.3080440984749449, "learning_rate": 3.0775893860614896e-05, "loss": 0.4028, "step": 476 }, { "epoch": 1.9321518987341773, "grad_norm": 0.4135673201926859, "learning_rate": 3.0728034917183336e-05, "loss": 0.4083, "step": 477 }, { "epoch": 1.9362025316455695, "grad_norm": 0.2795187346816477, "learning_rate": 3.06800895716334e-05, "loss": 0.4016, "step": 478 }, { "epoch": 1.940253164556962, "grad_norm": 0.3668665699854496, "learning_rate": 3.063205821011029e-05, "loss": 0.3991, "step": 479 }, { "epoch": 1.9443037974683546, "grad_norm": 0.3769500509881642, "learning_rate": 3.0583941219452016e-05, "loss": 0.3917, "step": 480 }, { "epoch": 1.9483544303797469, "grad_norm": 0.3749114194455825, "learning_rate": 3.053573898718618e-05, "loss": 0.4007, "step": 481 }, { "epoch": 1.9524050632911392, "grad_norm": 0.37287009399041693, "learning_rate": 3.0487451901526956e-05, "loss": 0.3983, "step": 482 }, { "epoch": 1.9564556962025317, "grad_norm": 0.3705616618580174, "learning_rate": 3.0439080351371875e-05, "loss": 0.406, "step": 483 }, { "epoch": 1.9605063291139242, "grad_norm": 0.4626811130838967, "learning_rate": 3.0390624726298764e-05, "loss": 0.3962, "step": 484 }, { "epoch": 1.9645569620253165, "grad_norm": 0.27111895074002373, "learning_rate": 3.034208541656255e-05, "loss": 0.3939, "step": 485 }, { "epoch": 1.9686075949367088, "grad_norm": 0.38092522312409227, "learning_rate": 3.029346281309218e-05, "loss": 0.4066, "step": 486 }, { "epoch": 1.9726582278481013, "grad_norm": 0.3026413298785291, "learning_rate": 3.0244757307487415e-05, "loss": 0.3953, "step": 487 }, { "epoch": 1.9767088607594938, "grad_norm": 0.30448035240484556, "learning_rate": 3.019596929201569e-05, "loss": 0.3958, "step": 488 }, { "epoch": 1.980759493670886, "grad_norm": 0.2806332133195858, "learning_rate": 3.0147099159608985e-05, "loss": 0.3945, "step": 489 }, { "epoch": 1.9848101265822784, "grad_norm": 0.27839364365680125, "learning_rate": 3.0098147303860616e-05, "loss": 0.3862, "step": 490 }, { "epoch": 1.988860759493671, "grad_norm": 0.3062903383533262, "learning_rate": 3.0049114119022117e-05, "loss": 0.411, "step": 491 }, { "epoch": 1.9929113924050634, "grad_norm": 0.3312550526235843, "learning_rate": 3.0000000000000004e-05, "loss": 0.4069, "step": 492 }, { "epoch": 1.9969620253164557, "grad_norm": 0.29975468409675016, "learning_rate": 2.995080534235264e-05, "loss": 0.4037, "step": 493 }, { "epoch": 2.001012658227848, "grad_norm": 0.33250712317106523, "learning_rate": 2.9901530542287044e-05, "loss": 0.391, "step": 494 }, { "epoch": 2.0050632911392405, "grad_norm": 0.41194537905015965, "learning_rate": 2.9852175996655676e-05, "loss": 0.3555, "step": 495 }, { "epoch": 2.009113924050633, "grad_norm": 0.561820245444463, "learning_rate": 2.980274210295326e-05, "loss": 0.3588, "step": 496 }, { "epoch": 2.013164556962025, "grad_norm": 0.4708076974682147, "learning_rate": 2.9753229259313578e-05, "loss": 0.3479, "step": 497 }, { "epoch": 2.0172151898734176, "grad_norm": 0.6215109715443167, "learning_rate": 2.9703637864506274e-05, "loss": 0.3654, "step": 498 }, { "epoch": 2.02126582278481, "grad_norm": 0.6381952012079997, "learning_rate": 2.965396831793362e-05, "loss": 0.3479, "step": 499 }, { "epoch": 2.0253164556962027, "grad_norm": 0.9384162077053636, "learning_rate": 2.9604221019627316e-05, "loss": 0.3626, "step": 500 }, { "epoch": 2.0293670886075947, "grad_norm": 0.6445005552300003, "learning_rate": 2.955439637024526e-05, "loss": 0.3623, "step": 501 }, { "epoch": 2.0334177215189873, "grad_norm": 0.6220957447418921, "learning_rate": 2.9504494771068334e-05, "loss": 0.3577, "step": 502 }, { "epoch": 2.0374683544303798, "grad_norm": 0.7139985974006574, "learning_rate": 2.9454516623997156e-05, "loss": 0.3494, "step": 503 }, { "epoch": 2.0415189873417723, "grad_norm": 0.4507109490671395, "learning_rate": 2.9404462331548847e-05, "loss": 0.3581, "step": 504 }, { "epoch": 2.0455696202531644, "grad_norm": 0.5717364865940981, "learning_rate": 2.93543322968538e-05, "loss": 0.3605, "step": 505 }, { "epoch": 2.049620253164557, "grad_norm": 0.5523533813967964, "learning_rate": 2.9304126923652428e-05, "loss": 0.3486, "step": 506 }, { "epoch": 2.0536708860759494, "grad_norm": 0.4617267021880591, "learning_rate": 2.9253846616291896e-05, "loss": 0.3622, "step": 507 }, { "epoch": 2.057721518987342, "grad_norm": 0.6013304597613198, "learning_rate": 2.9203491779722896e-05, "loss": 0.3397, "step": 508 }, { "epoch": 2.061772151898734, "grad_norm": 0.41344645030648214, "learning_rate": 2.9153062819496357e-05, "loss": 0.3524, "step": 509 }, { "epoch": 2.0658227848101265, "grad_norm": 0.550387971291183, "learning_rate": 2.9102560141760178e-05, "loss": 0.3423, "step": 510 }, { "epoch": 2.069873417721519, "grad_norm": 0.4615952690021445, "learning_rate": 2.9051984153256004e-05, "loss": 0.3575, "step": 511 }, { "epoch": 2.0739240506329115, "grad_norm": 0.45247057113837613, "learning_rate": 2.900133526131588e-05, "loss": 0.3592, "step": 512 }, { "epoch": 2.0779746835443036, "grad_norm": 0.5230292572777449, "learning_rate": 2.8950613873859025e-05, "loss": 0.3601, "step": 513 }, { "epoch": 2.082025316455696, "grad_norm": 0.4405640869296072, "learning_rate": 2.8899820399388515e-05, "loss": 0.3579, "step": 514 }, { "epoch": 2.0860759493670886, "grad_norm": 0.5350530503158747, "learning_rate": 2.8848955246988012e-05, "loss": 0.3584, "step": 515 }, { "epoch": 2.090126582278481, "grad_norm": 0.4249184984036266, "learning_rate": 2.879801882631847e-05, "loss": 0.359, "step": 516 }, { "epoch": 2.094177215189873, "grad_norm": 0.47269772213061834, "learning_rate": 2.8747011547614808e-05, "loss": 0.352, "step": 517 }, { "epoch": 2.0982278481012657, "grad_norm": 0.3771879019852577, "learning_rate": 2.8695933821682635e-05, "loss": 0.3558, "step": 518 }, { "epoch": 2.1022784810126582, "grad_norm": 0.42983289269102265, "learning_rate": 2.864478605989494e-05, "loss": 0.3458, "step": 519 }, { "epoch": 2.1063291139240508, "grad_norm": 0.2873439616673898, "learning_rate": 2.8593568674188765e-05, "loss": 0.3449, "step": 520 }, { "epoch": 2.110379746835443, "grad_norm": 0.38758747542224586, "learning_rate": 2.8542282077061892e-05, "loss": 0.3551, "step": 521 }, { "epoch": 2.1144303797468353, "grad_norm": 0.31388745880255237, "learning_rate": 2.8490926681569523e-05, "loss": 0.3503, "step": 522 }, { "epoch": 2.118481012658228, "grad_norm": 0.3261534447933216, "learning_rate": 2.8439502901320956e-05, "loss": 0.3584, "step": 523 }, { "epoch": 2.1225316455696204, "grad_norm": 0.32153484024310713, "learning_rate": 2.8388011150476237e-05, "loss": 0.3503, "step": 524 }, { "epoch": 2.1265822784810124, "grad_norm": 0.3190256319627238, "learning_rate": 2.8336451843742866e-05, "loss": 0.3493, "step": 525 }, { "epoch": 2.130632911392405, "grad_norm": 0.35110118336795776, "learning_rate": 2.8284825396372387e-05, "loss": 0.349, "step": 526 }, { "epoch": 2.1346835443037975, "grad_norm": 0.29507478390602604, "learning_rate": 2.8233132224157132e-05, "loss": 0.3576, "step": 527 }, { "epoch": 2.13873417721519, "grad_norm": 0.3910659949700422, "learning_rate": 2.8181372743426805e-05, "loss": 0.3611, "step": 528 }, { "epoch": 2.1427848101265825, "grad_norm": 0.40146409891808194, "learning_rate": 2.8129547371045128e-05, "loss": 0.3425, "step": 529 }, { "epoch": 2.1468354430379746, "grad_norm": 0.4376920480202133, "learning_rate": 2.8077656524406534e-05, "loss": 0.3548, "step": 530 }, { "epoch": 2.150886075949367, "grad_norm": 0.35216757052211245, "learning_rate": 2.802570062143278e-05, "loss": 0.3472, "step": 531 }, { "epoch": 2.1549367088607596, "grad_norm": 0.350045185908929, "learning_rate": 2.7973680080569555e-05, "loss": 0.3557, "step": 532 }, { "epoch": 2.1589873417721517, "grad_norm": 0.3640399325806392, "learning_rate": 2.792159532078314e-05, "loss": 0.3515, "step": 533 }, { "epoch": 2.163037974683544, "grad_norm": 0.2628433295710214, "learning_rate": 2.7869446761557033e-05, "loss": 0.3567, "step": 534 }, { "epoch": 2.1670886075949367, "grad_norm": 0.34306225858462375, "learning_rate": 2.781723482288857e-05, "loss": 0.3534, "step": 535 }, { "epoch": 2.1711392405063292, "grad_norm": 0.33087479379881396, "learning_rate": 2.7764959925285517e-05, "loss": 0.3499, "step": 536 }, { "epoch": 2.1751898734177217, "grad_norm": 0.3339419341640183, "learning_rate": 2.771262248976272e-05, "loss": 0.3503, "step": 537 }, { "epoch": 2.179240506329114, "grad_norm": 0.37300108638916285, "learning_rate": 2.7660222937838677e-05, "loss": 0.3574, "step": 538 }, { "epoch": 2.1832911392405063, "grad_norm": 0.36967642436563375, "learning_rate": 2.7607761691532186e-05, "loss": 0.3434, "step": 539 }, { "epoch": 2.187341772151899, "grad_norm": 0.3094941495323904, "learning_rate": 2.7555239173358916e-05, "loss": 0.3348, "step": 540 }, { "epoch": 2.191392405063291, "grad_norm": 0.34566551861732914, "learning_rate": 2.7502655806328e-05, "loss": 0.3425, "step": 541 }, { "epoch": 2.1954430379746834, "grad_norm": 0.2836166587118865, "learning_rate": 2.7450012013938648e-05, "loss": 0.3356, "step": 542 }, { "epoch": 2.199493670886076, "grad_norm": 0.328188208585498, "learning_rate": 2.739730822017673e-05, "loss": 0.36, "step": 543 }, { "epoch": 2.2035443037974685, "grad_norm": 0.3289476551499328, "learning_rate": 2.7344544849511355e-05, "loss": 0.3626, "step": 544 }, { "epoch": 2.207594936708861, "grad_norm": 0.28413381745350036, "learning_rate": 2.7291722326891456e-05, "loss": 0.3549, "step": 545 }, { "epoch": 2.211645569620253, "grad_norm": 0.33532950614515306, "learning_rate": 2.723884107774236e-05, "loss": 0.3523, "step": 546 }, { "epoch": 2.2156962025316456, "grad_norm": 0.3410523524718243, "learning_rate": 2.718590152796239e-05, "loss": 0.3578, "step": 547 }, { "epoch": 2.219746835443038, "grad_norm": 0.3281400239657301, "learning_rate": 2.71329041039194e-05, "loss": 0.3541, "step": 548 }, { "epoch": 2.22379746835443, "grad_norm": 0.31780361479043795, "learning_rate": 2.7079849232447357e-05, "loss": 0.3485, "step": 549 }, { "epoch": 2.2278481012658227, "grad_norm": 0.3213025770976673, "learning_rate": 2.7026737340842895e-05, "loss": 0.346, "step": 550 }, { "epoch": 2.231898734177215, "grad_norm": 0.3121711196675289, "learning_rate": 2.697356885686189e-05, "loss": 0.3706, "step": 551 }, { "epoch": 2.2359493670886077, "grad_norm": 0.3110292444356443, "learning_rate": 2.6920344208716014e-05, "loss": 0.3481, "step": 552 }, { "epoch": 2.24, "grad_norm": 0.2978485752259437, "learning_rate": 2.6867063825069252e-05, "loss": 0.3409, "step": 553 }, { "epoch": 2.2440506329113923, "grad_norm": 0.2885374678935957, "learning_rate": 2.6813728135034494e-05, "loss": 0.3534, "step": 554 }, { "epoch": 2.248101265822785, "grad_norm": 0.30255851001933226, "learning_rate": 2.6760337568170056e-05, "loss": 0.3591, "step": 555 }, { "epoch": 2.2521518987341773, "grad_norm": 0.2773599553458637, "learning_rate": 2.6706892554476226e-05, "loss": 0.3575, "step": 556 }, { "epoch": 2.2562025316455694, "grad_norm": 0.28124337878960876, "learning_rate": 2.6653393524391795e-05, "loss": 0.3584, "step": 557 }, { "epoch": 2.260253164556962, "grad_norm": 0.30282761399336344, "learning_rate": 2.6599840908790592e-05, "loss": 0.3499, "step": 558 }, { "epoch": 2.2643037974683544, "grad_norm": 0.29833513322265276, "learning_rate": 2.6546235138978028e-05, "loss": 0.3481, "step": 559 }, { "epoch": 2.268354430379747, "grad_norm": 0.3334601749549423, "learning_rate": 2.6492576646687597e-05, "loss": 0.354, "step": 560 }, { "epoch": 2.2724050632911394, "grad_norm": 0.35036086382887344, "learning_rate": 2.6438865864077425e-05, "loss": 0.3509, "step": 561 }, { "epoch": 2.2764556962025315, "grad_norm": 0.5485838801366879, "learning_rate": 2.6385103223726766e-05, "loss": 0.3563, "step": 562 }, { "epoch": 2.280506329113924, "grad_norm": 0.3577837653950019, "learning_rate": 2.6331289158632537e-05, "loss": 0.3472, "step": 563 }, { "epoch": 2.2845569620253166, "grad_norm": 0.37651705483821934, "learning_rate": 2.6277424102205817e-05, "loss": 0.3454, "step": 564 }, { "epoch": 2.2886075949367086, "grad_norm": 0.3262485026105536, "learning_rate": 2.6223508488268374e-05, "loss": 0.3655, "step": 565 }, { "epoch": 2.292658227848101, "grad_norm": 0.4192202123598278, "learning_rate": 2.6169542751049148e-05, "loss": 0.3567, "step": 566 }, { "epoch": 2.2967088607594937, "grad_norm": 0.2970931613457095, "learning_rate": 2.6115527325180754e-05, "loss": 0.351, "step": 567 }, { "epoch": 2.300759493670886, "grad_norm": 0.4019284611991859, "learning_rate": 2.606146264569603e-05, "loss": 0.3559, "step": 568 }, { "epoch": 2.3048101265822787, "grad_norm": 0.3007673820418208, "learning_rate": 2.6007349148024447e-05, "loss": 0.3509, "step": 569 }, { "epoch": 2.3088607594936708, "grad_norm": 0.4042024506930742, "learning_rate": 2.5953187267988694e-05, "loss": 0.3594, "step": 570 }, { "epoch": 2.3129113924050633, "grad_norm": 0.3080473346353238, "learning_rate": 2.5898977441801097e-05, "loss": 0.3488, "step": 571 }, { "epoch": 2.316962025316456, "grad_norm": 0.32116846858649006, "learning_rate": 2.584472010606015e-05, "loss": 0.3554, "step": 572 }, { "epoch": 2.321012658227848, "grad_norm": 0.28312175365914083, "learning_rate": 2.5790415697746976e-05, "loss": 0.3665, "step": 573 }, { "epoch": 2.3250632911392404, "grad_norm": 0.33087563883698484, "learning_rate": 2.5736064654221808e-05, "loss": 0.3483, "step": 574 }, { "epoch": 2.329113924050633, "grad_norm": 0.2913539115323236, "learning_rate": 2.568166741322048e-05, "loss": 0.3522, "step": 575 }, { "epoch": 2.3331645569620254, "grad_norm": 0.2807798249515716, "learning_rate": 2.56272244128509e-05, "loss": 0.3667, "step": 576 }, { "epoch": 2.337215189873418, "grad_norm": 0.31222273402550205, "learning_rate": 2.55727360915895e-05, "loss": 0.3605, "step": 577 }, { "epoch": 2.34126582278481, "grad_norm": 0.2883664321464256, "learning_rate": 2.5518202888277734e-05, "loss": 0.3549, "step": 578 }, { "epoch": 2.3453164556962025, "grad_norm": 0.3215723678352346, "learning_rate": 2.5463625242118523e-05, "loss": 0.3427, "step": 579 }, { "epoch": 2.349367088607595, "grad_norm": 0.2809937174066703, "learning_rate": 2.5409003592672723e-05, "loss": 0.3659, "step": 580 }, { "epoch": 2.353417721518987, "grad_norm": 0.3128412885275924, "learning_rate": 2.535433837985559e-05, "loss": 0.3564, "step": 581 }, { "epoch": 2.3574683544303796, "grad_norm": 0.2870285464167094, "learning_rate": 2.529963004393324e-05, "loss": 0.3479, "step": 582 }, { "epoch": 2.361518987341772, "grad_norm": 0.28641614247920893, "learning_rate": 2.524487902551908e-05, "loss": 0.348, "step": 583 }, { "epoch": 2.3655696202531646, "grad_norm": 0.3283338316253794, "learning_rate": 2.519008576557029e-05, "loss": 0.3436, "step": 584 }, { "epoch": 2.369620253164557, "grad_norm": 0.30502024596699523, "learning_rate": 2.5135250705384254e-05, "loss": 0.3444, "step": 585 }, { "epoch": 2.3736708860759492, "grad_norm": 0.3457459217337302, "learning_rate": 2.5080374286595007e-05, "loss": 0.3393, "step": 586 }, { "epoch": 2.3777215189873417, "grad_norm": 0.3191137877023231, "learning_rate": 2.5025456951169677e-05, "loss": 0.3555, "step": 587 }, { "epoch": 2.3817721518987343, "grad_norm": 0.35698406638200886, "learning_rate": 2.4970499141404942e-05, "loss": 0.3665, "step": 588 }, { "epoch": 2.3858227848101268, "grad_norm": 0.29380144235676336, "learning_rate": 2.491550129992345e-05, "loss": 0.3585, "step": 589 }, { "epoch": 2.389873417721519, "grad_norm": 0.33617330378500315, "learning_rate": 2.486046386967024e-05, "loss": 0.3617, "step": 590 }, { "epoch": 2.3939240506329114, "grad_norm": 0.2884594249483237, "learning_rate": 2.4805387293909214e-05, "loss": 0.3501, "step": 591 }, { "epoch": 2.397974683544304, "grad_norm": 0.3007429049688558, "learning_rate": 2.4750272016219552e-05, "loss": 0.3455, "step": 592 }, { "epoch": 2.4020253164556964, "grad_norm": 0.2916305449669507, "learning_rate": 2.4695118480492114e-05, "loss": 0.3717, "step": 593 }, { "epoch": 2.4060759493670885, "grad_norm": 0.2612268958362754, "learning_rate": 2.4639927130925898e-05, "loss": 0.3478, "step": 594 }, { "epoch": 2.410126582278481, "grad_norm": 0.2925832900758547, "learning_rate": 2.458469841202444e-05, "loss": 0.3576, "step": 595 }, { "epoch": 2.4141772151898735, "grad_norm": 0.2543450562351551, "learning_rate": 2.452943276859226e-05, "loss": 0.3468, "step": 596 }, { "epoch": 2.418227848101266, "grad_norm": 0.31196250362556405, "learning_rate": 2.447413064573125e-05, "loss": 0.359, "step": 597 }, { "epoch": 2.422278481012658, "grad_norm": 0.2899703913819664, "learning_rate": 2.4418792488837095e-05, "loss": 0.3532, "step": 598 }, { "epoch": 2.4263291139240506, "grad_norm": 0.3077893901308835, "learning_rate": 2.4363418743595713e-05, "loss": 0.3562, "step": 599 }, { "epoch": 2.430379746835443, "grad_norm": 0.2985796644206488, "learning_rate": 2.430800985597963e-05, "loss": 0.3613, "step": 600 }, { "epoch": 2.4344303797468356, "grad_norm": 0.32103299809637986, "learning_rate": 2.4252566272244415e-05, "loss": 0.3614, "step": 601 }, { "epoch": 2.4384810126582277, "grad_norm": 0.27972283542949733, "learning_rate": 2.4197088438925063e-05, "loss": 0.3488, "step": 602 }, { "epoch": 2.44253164556962, "grad_norm": 0.3330380083884755, "learning_rate": 2.4141576802832417e-05, "loss": 0.3529, "step": 603 }, { "epoch": 2.4465822784810127, "grad_norm": 0.2821798919134842, "learning_rate": 2.408603181104957e-05, "loss": 0.3637, "step": 604 }, { "epoch": 2.4506329113924052, "grad_norm": 0.3361549470081877, "learning_rate": 2.4030453910928245e-05, "loss": 0.3576, "step": 605 }, { "epoch": 2.4546835443037973, "grad_norm": 0.2814742368549593, "learning_rate": 2.397484355008521e-05, "loss": 0.3494, "step": 606 }, { "epoch": 2.45873417721519, "grad_norm": 0.3795343754779306, "learning_rate": 2.3919201176398662e-05, "loss": 0.3565, "step": 607 }, { "epoch": 2.4627848101265823, "grad_norm": 0.40756327167632045, "learning_rate": 2.3863527238004633e-05, "loss": 0.3556, "step": 608 }, { "epoch": 2.466835443037975, "grad_norm": 0.3313623608412181, "learning_rate": 2.380782218329337e-05, "loss": 0.3408, "step": 609 }, { "epoch": 2.470886075949367, "grad_norm": 0.4103918508004759, "learning_rate": 2.3752086460905725e-05, "loss": 0.3586, "step": 610 }, { "epoch": 2.4749367088607594, "grad_norm": 0.280053180103094, "learning_rate": 2.3696320519729544e-05, "loss": 0.3512, "step": 611 }, { "epoch": 2.478987341772152, "grad_norm": 0.3835807722106156, "learning_rate": 2.3640524808896045e-05, "loss": 0.3546, "step": 612 }, { "epoch": 2.4830379746835445, "grad_norm": 0.2930305873637649, "learning_rate": 2.3584699777776222e-05, "loss": 0.3524, "step": 613 }, { "epoch": 2.4870886075949366, "grad_norm": 0.32111310401650134, "learning_rate": 2.3528845875977195e-05, "loss": 0.3423, "step": 614 }, { "epoch": 2.491139240506329, "grad_norm": 0.32097596191119515, "learning_rate": 2.3472963553338614e-05, "loss": 0.3584, "step": 615 }, { "epoch": 2.4951898734177216, "grad_norm": 0.2819465257458148, "learning_rate": 2.341705325992901e-05, "loss": 0.3564, "step": 616 }, { "epoch": 2.499240506329114, "grad_norm": 0.27908756903700815, "learning_rate": 2.336111544604222e-05, "loss": 0.3438, "step": 617 }, { "epoch": 2.503291139240506, "grad_norm": 0.2823505385879975, "learning_rate": 2.33051505621937e-05, "loss": 0.3577, "step": 618 }, { "epoch": 2.5073417721518987, "grad_norm": 0.2803659193669749, "learning_rate": 2.324915905911693e-05, "loss": 0.3458, "step": 619 }, { "epoch": 2.511392405063291, "grad_norm": 0.25746135935955605, "learning_rate": 2.319314138775977e-05, "loss": 0.3614, "step": 620 }, { "epoch": 2.5154430379746833, "grad_norm": 0.3179116479307971, "learning_rate": 2.3137097999280856e-05, "loss": 0.3578, "step": 621 }, { "epoch": 2.519493670886076, "grad_norm": 0.26228101233436046, "learning_rate": 2.308102934504593e-05, "loss": 0.3498, "step": 622 }, { "epoch": 2.5235443037974683, "grad_norm": 0.26662834092716375, "learning_rate": 2.3024935876624222e-05, "loss": 0.3481, "step": 623 }, { "epoch": 2.527594936708861, "grad_norm": 0.24054522443047538, "learning_rate": 2.2968818045784813e-05, "loss": 0.3502, "step": 624 }, { "epoch": 2.5316455696202533, "grad_norm": 0.2752235021452881, "learning_rate": 2.2912676304493006e-05, "loss": 0.3425, "step": 625 }, { "epoch": 2.5356962025316454, "grad_norm": 0.24117455332358698, "learning_rate": 2.2856511104906668e-05, "loss": 0.3546, "step": 626 }, { "epoch": 2.539746835443038, "grad_norm": 0.32060939306061154, "learning_rate": 2.2800322899372586e-05, "loss": 0.355, "step": 627 }, { "epoch": 2.5437974683544304, "grad_norm": 0.2546984641435136, "learning_rate": 2.2744112140422844e-05, "loss": 0.3543, "step": 628 }, { "epoch": 2.547848101265823, "grad_norm": 0.3318202712865923, "learning_rate": 2.2687879280771177e-05, "loss": 0.351, "step": 629 }, { "epoch": 2.5518987341772155, "grad_norm": 0.27791308037003803, "learning_rate": 2.26316247733093e-05, "loss": 0.346, "step": 630 }, { "epoch": 2.5559493670886075, "grad_norm": 0.25240516205710434, "learning_rate": 2.257534907110328e-05, "loss": 0.3487, "step": 631 }, { "epoch": 2.56, "grad_norm": 0.2792597377692073, "learning_rate": 2.2519052627389882e-05, "loss": 0.3498, "step": 632 }, { "epoch": 2.5640506329113926, "grad_norm": 0.2568065187315663, "learning_rate": 2.246273589557294e-05, "loss": 0.3445, "step": 633 }, { "epoch": 2.5681012658227846, "grad_norm": 0.26201318211411107, "learning_rate": 2.240639932921966e-05, "loss": 0.3529, "step": 634 }, { "epoch": 2.572151898734177, "grad_norm": 0.27771810962144694, "learning_rate": 2.2350043382056995e-05, "loss": 0.3577, "step": 635 }, { "epoch": 2.5762025316455697, "grad_norm": 0.2609125401221827, "learning_rate": 2.2293668507968015e-05, "loss": 0.3598, "step": 636 }, { "epoch": 2.580253164556962, "grad_norm": 0.28353080957253146, "learning_rate": 2.2237275160988186e-05, "loss": 0.3445, "step": 637 }, { "epoch": 2.5843037974683547, "grad_norm": 0.3033844062776866, "learning_rate": 2.2180863795301787e-05, "loss": 0.3468, "step": 638 }, { "epoch": 2.5883544303797468, "grad_norm": 0.28688173460368416, "learning_rate": 2.212443486523819e-05, "loss": 0.3575, "step": 639 }, { "epoch": 2.5924050632911393, "grad_norm": 0.3167786333818367, "learning_rate": 2.2067988825268243e-05, "loss": 0.351, "step": 640 }, { "epoch": 2.596455696202532, "grad_norm": 0.2958899554341526, "learning_rate": 2.2011526130000596e-05, "loss": 0.3604, "step": 641 }, { "epoch": 2.600506329113924, "grad_norm": 0.32553016047356076, "learning_rate": 2.1955047234178038e-05, "loss": 0.3549, "step": 642 }, { "epoch": 2.6045569620253164, "grad_norm": 0.2752569779255134, "learning_rate": 2.1898552592673825e-05, "loss": 0.3364, "step": 643 }, { "epoch": 2.608607594936709, "grad_norm": 0.31479612483127634, "learning_rate": 2.184204266048803e-05, "loss": 0.3568, "step": 644 }, { "epoch": 2.6126582278481014, "grad_norm": 0.30600301391025686, "learning_rate": 2.1785517892743887e-05, "loss": 0.3657, "step": 645 }, { "epoch": 2.616708860759494, "grad_norm": 0.29777047709580107, "learning_rate": 2.17289787446841e-05, "loss": 0.341, "step": 646 }, { "epoch": 2.620759493670886, "grad_norm": 0.29360079001717454, "learning_rate": 2.1672425671667198e-05, "loss": 0.3504, "step": 647 }, { "epoch": 2.6248101265822785, "grad_norm": 0.29590338348071293, "learning_rate": 2.161585912916385e-05, "loss": 0.3482, "step": 648 }, { "epoch": 2.628860759493671, "grad_norm": 0.29789097855496394, "learning_rate": 2.1559279572753214e-05, "loss": 0.3526, "step": 649 }, { "epoch": 2.632911392405063, "grad_norm": 0.2729447801868544, "learning_rate": 2.1502687458119268e-05, "loss": 0.3492, "step": 650 }, { "epoch": 2.6369620253164556, "grad_norm": 0.27424415664271845, "learning_rate": 2.1446083241047116e-05, "loss": 0.3608, "step": 651 }, { "epoch": 2.641012658227848, "grad_norm": 0.2608365249026713, "learning_rate": 2.1389467377419333e-05, "loss": 0.3586, "step": 652 }, { "epoch": 2.6450632911392407, "grad_norm": 0.2957742549416258, "learning_rate": 2.133284032321232e-05, "loss": 0.3635, "step": 653 }, { "epoch": 2.649113924050633, "grad_norm": 0.25329108504599707, "learning_rate": 2.1276202534492566e-05, "loss": 0.3528, "step": 654 }, { "epoch": 2.6531645569620252, "grad_norm": 0.2810039217771068, "learning_rate": 2.121955446741306e-05, "loss": 0.3566, "step": 655 }, { "epoch": 2.6572151898734178, "grad_norm": 0.25032916015904594, "learning_rate": 2.1162896578209517e-05, "loss": 0.3546, "step": 656 }, { "epoch": 2.6612658227848103, "grad_norm": 0.3821252366627388, "learning_rate": 2.1106229323196813e-05, "loss": 0.3549, "step": 657 }, { "epoch": 2.6653164556962023, "grad_norm": 0.32435305370240525, "learning_rate": 2.1049553158765214e-05, "loss": 0.3615, "step": 658 }, { "epoch": 2.669367088607595, "grad_norm": 0.2552776101044122, "learning_rate": 2.0992868541376764e-05, "loss": 0.3583, "step": 659 }, { "epoch": 2.6734177215189874, "grad_norm": 0.26958689615974374, "learning_rate": 2.093617592756158e-05, "loss": 0.3555, "step": 660 }, { "epoch": 2.67746835443038, "grad_norm": 0.2459136831393981, "learning_rate": 2.0879475773914167e-05, "loss": 0.3452, "step": 661 }, { "epoch": 2.6815189873417724, "grad_norm": 0.29822139421748733, "learning_rate": 2.082276853708978e-05, "loss": 0.3555, "step": 662 }, { "epoch": 2.6855696202531645, "grad_norm": 0.25268590592527707, "learning_rate": 2.076605467380071e-05, "loss": 0.3479, "step": 663 }, { "epoch": 2.689620253164557, "grad_norm": 0.2866034494707634, "learning_rate": 2.0709334640812613e-05, "loss": 0.3494, "step": 664 }, { "epoch": 2.6936708860759495, "grad_norm": 0.2650172023668211, "learning_rate": 2.0652608894940824e-05, "loss": 0.3453, "step": 665 }, { "epoch": 2.6977215189873416, "grad_norm": 0.26773942917567817, "learning_rate": 2.0595877893046722e-05, "loss": 0.3536, "step": 666 }, { "epoch": 2.701772151898734, "grad_norm": 0.2788962904733358, "learning_rate": 2.0539142092033985e-05, "loss": 0.3561, "step": 667 }, { "epoch": 2.7058227848101266, "grad_norm": 0.3586568578660402, "learning_rate": 2.048240194884496e-05, "loss": 0.3557, "step": 668 }, { "epoch": 2.709873417721519, "grad_norm": 0.2959530527241258, "learning_rate": 2.042565792045695e-05, "loss": 0.3599, "step": 669 }, { "epoch": 2.7139240506329116, "grad_norm": 0.2523104279244123, "learning_rate": 2.036891046387857e-05, "loss": 0.3499, "step": 670 }, { "epoch": 2.7179746835443037, "grad_norm": 0.3056476954335154, "learning_rate": 2.0312160036146036e-05, "loss": 0.3647, "step": 671 }, { "epoch": 2.7220253164556962, "grad_norm": 0.22413712371038882, "learning_rate": 2.025540709431948e-05, "loss": 0.3655, "step": 672 }, { "epoch": 2.7260759493670887, "grad_norm": 0.2486402436643554, "learning_rate": 2.0198652095479298e-05, "loss": 0.351, "step": 673 }, { "epoch": 2.730126582278481, "grad_norm": 0.25564214496345433, "learning_rate": 2.014189549672245e-05, "loss": 0.3649, "step": 674 }, { "epoch": 2.7341772151898733, "grad_norm": 0.24105685540980332, "learning_rate": 2.0085137755158776e-05, "loss": 0.347, "step": 675 }, { "epoch": 2.738227848101266, "grad_norm": 0.2458787477485068, "learning_rate": 2.0028379327907327e-05, "loss": 0.3554, "step": 676 }, { "epoch": 2.7422784810126584, "grad_norm": 0.2678788536629327, "learning_rate": 1.9971620672092676e-05, "loss": 0.3646, "step": 677 }, { "epoch": 2.746329113924051, "grad_norm": 0.22830599566740112, "learning_rate": 1.991486224484123e-05, "loss": 0.349, "step": 678 }, { "epoch": 2.750379746835443, "grad_norm": 0.2750807725450295, "learning_rate": 1.985810450327756e-05, "loss": 0.3529, "step": 679 }, { "epoch": 2.7544303797468355, "grad_norm": 0.2517392665335761, "learning_rate": 1.9801347904520706e-05, "loss": 0.3521, "step": 680 }, { "epoch": 2.758481012658228, "grad_norm": 0.26375852286129237, "learning_rate": 1.974459290568053e-05, "loss": 0.3471, "step": 681 }, { "epoch": 2.76253164556962, "grad_norm": 0.28357624897590883, "learning_rate": 1.968783996385397e-05, "loss": 0.3527, "step": 682 }, { "epoch": 2.7665822784810126, "grad_norm": 0.2734191057993035, "learning_rate": 1.963108953612143e-05, "loss": 0.3547, "step": 683 }, { "epoch": 2.770632911392405, "grad_norm": 0.2779287486540737, "learning_rate": 1.9574342079543056e-05, "loss": 0.3623, "step": 684 }, { "epoch": 2.7746835443037976, "grad_norm": 0.25662237299135066, "learning_rate": 1.9517598051155046e-05, "loss": 0.3588, "step": 685 }, { "epoch": 2.77873417721519, "grad_norm": 0.2615258182743059, "learning_rate": 1.9460857907966025e-05, "loss": 0.3485, "step": 686 }, { "epoch": 2.782784810126582, "grad_norm": 0.260436013558495, "learning_rate": 1.9404122106953285e-05, "loss": 0.3506, "step": 687 }, { "epoch": 2.7868354430379747, "grad_norm": 0.26048633463384413, "learning_rate": 1.9347391105059176e-05, "loss": 0.3668, "step": 688 }, { "epoch": 2.790886075949367, "grad_norm": 0.24340059438930503, "learning_rate": 1.92906653591874e-05, "loss": 0.3536, "step": 689 }, { "epoch": 2.7949367088607593, "grad_norm": 0.2874532911030448, "learning_rate": 1.9233945326199295e-05, "loss": 0.3503, "step": 690 }, { "epoch": 2.798987341772152, "grad_norm": 0.2331474965421145, "learning_rate": 1.917723146291022e-05, "loss": 0.3584, "step": 691 }, { "epoch": 2.8030379746835443, "grad_norm": 0.31016734603018065, "learning_rate": 1.912052422608584e-05, "loss": 0.33, "step": 692 }, { "epoch": 2.807088607594937, "grad_norm": 0.23504848508069606, "learning_rate": 1.9063824072438428e-05, "loss": 0.3517, "step": 693 }, { "epoch": 2.8111392405063294, "grad_norm": 0.2846736273021319, "learning_rate": 1.9007131458623246e-05, "loss": 0.3623, "step": 694 }, { "epoch": 2.8151898734177214, "grad_norm": 0.22944987997395846, "learning_rate": 1.895044684123479e-05, "loss": 0.343, "step": 695 }, { "epoch": 2.819240506329114, "grad_norm": 0.2920422510250883, "learning_rate": 1.8893770676803194e-05, "loss": 0.3621, "step": 696 }, { "epoch": 2.8232911392405065, "grad_norm": 0.2643287508765151, "learning_rate": 1.8837103421790486e-05, "loss": 0.359, "step": 697 }, { "epoch": 2.8273417721518985, "grad_norm": 0.2921147271486843, "learning_rate": 1.8780445532586952e-05, "loss": 0.3593, "step": 698 }, { "epoch": 2.831392405063291, "grad_norm": 0.24408170305080495, "learning_rate": 1.872379746550743e-05, "loss": 0.3465, "step": 699 }, { "epoch": 2.8354430379746836, "grad_norm": 0.2775940969123264, "learning_rate": 1.866715967678769e-05, "loss": 0.3379, "step": 700 }, { "epoch": 2.839493670886076, "grad_norm": 0.24916221865331875, "learning_rate": 1.861053262258067e-05, "loss": 0.3441, "step": 701 }, { "epoch": 2.8435443037974686, "grad_norm": 0.2762123898728658, "learning_rate": 1.8553916758952897e-05, "loss": 0.3418, "step": 702 }, { "epoch": 2.8475949367088607, "grad_norm": 0.2334895600713022, "learning_rate": 1.8497312541880735e-05, "loss": 0.3432, "step": 703 }, { "epoch": 2.851645569620253, "grad_norm": 0.27654903435179645, "learning_rate": 1.8440720427246786e-05, "loss": 0.3493, "step": 704 }, { "epoch": 2.8556962025316457, "grad_norm": 0.23713160814790554, "learning_rate": 1.8384140870836157e-05, "loss": 0.3501, "step": 705 }, { "epoch": 2.8597468354430378, "grad_norm": 0.24944642789700247, "learning_rate": 1.8327574328332806e-05, "loss": 0.3454, "step": 706 }, { "epoch": 2.8637974683544303, "grad_norm": 0.24662962942693806, "learning_rate": 1.8271021255315906e-05, "loss": 0.347, "step": 707 }, { "epoch": 2.867848101265823, "grad_norm": 0.2735839682604408, "learning_rate": 1.8214482107256117e-05, "loss": 0.3644, "step": 708 }, { "epoch": 2.8718987341772153, "grad_norm": 0.2637130306516853, "learning_rate": 1.8157957339511968e-05, "loss": 0.3557, "step": 709 }, { "epoch": 2.875949367088608, "grad_norm": 0.24474019586465587, "learning_rate": 1.8101447407326182e-05, "loss": 0.3379, "step": 710 }, { "epoch": 2.88, "grad_norm": 0.25572083985633537, "learning_rate": 1.8044952765821966e-05, "loss": 0.3494, "step": 711 }, { "epoch": 2.8840506329113924, "grad_norm": 0.2672182082785668, "learning_rate": 1.7988473869999407e-05, "loss": 0.3569, "step": 712 }, { "epoch": 2.888101265822785, "grad_norm": 0.2460503603667764, "learning_rate": 1.7932011174731764e-05, "loss": 0.3526, "step": 713 }, { "epoch": 2.892151898734177, "grad_norm": 0.276479306172656, "learning_rate": 1.7875565134761817e-05, "loss": 0.3519, "step": 714 }, { "epoch": 2.8962025316455695, "grad_norm": 0.2534490713840529, "learning_rate": 1.7819136204698226e-05, "loss": 0.3579, "step": 715 }, { "epoch": 2.900253164556962, "grad_norm": 0.27259740854301656, "learning_rate": 1.776272483901182e-05, "loss": 0.3426, "step": 716 }, { "epoch": 2.9043037974683545, "grad_norm": 0.2520859612071462, "learning_rate": 1.7706331492031995e-05, "loss": 0.3597, "step": 717 }, { "epoch": 2.908354430379747, "grad_norm": 0.25911253160128145, "learning_rate": 1.764995661794301e-05, "loss": 0.3569, "step": 718 }, { "epoch": 2.912405063291139, "grad_norm": 0.2538089518823638, "learning_rate": 1.759360067078035e-05, "loss": 0.3358, "step": 719 }, { "epoch": 2.9164556962025316, "grad_norm": 0.2501045026094366, "learning_rate": 1.7537264104427064e-05, "loss": 0.3573, "step": 720 }, { "epoch": 2.920506329113924, "grad_norm": 0.25518943673772637, "learning_rate": 1.748094737261012e-05, "loss": 0.3534, "step": 721 }, { "epoch": 2.9245569620253162, "grad_norm": 0.25741870247217385, "learning_rate": 1.7424650928896726e-05, "loss": 0.3511, "step": 722 }, { "epoch": 2.9286075949367087, "grad_norm": 0.2575085991180075, "learning_rate": 1.7368375226690712e-05, "loss": 0.3668, "step": 723 }, { "epoch": 2.9326582278481013, "grad_norm": 0.27178763460259814, "learning_rate": 1.731212071922883e-05, "loss": 0.351, "step": 724 }, { "epoch": 2.9367088607594938, "grad_norm": 0.2592253817180511, "learning_rate": 1.7255887859577156e-05, "loss": 0.3375, "step": 725 }, { "epoch": 2.9407594936708863, "grad_norm": 0.24225201124527052, "learning_rate": 1.7199677100627427e-05, "loss": 0.3548, "step": 726 }, { "epoch": 2.9448101265822784, "grad_norm": 0.27014277590737923, "learning_rate": 1.7143488895093343e-05, "loss": 0.358, "step": 727 }, { "epoch": 2.948860759493671, "grad_norm": 0.23760734592849483, "learning_rate": 1.7087323695506994e-05, "loss": 0.355, "step": 728 }, { "epoch": 2.9529113924050634, "grad_norm": 0.24311694953762616, "learning_rate": 1.7031181954215194e-05, "loss": 0.3385, "step": 729 }, { "epoch": 2.9569620253164555, "grad_norm": 0.24071839208302714, "learning_rate": 1.6975064123375788e-05, "loss": 0.3568, "step": 730 }, { "epoch": 2.961012658227848, "grad_norm": 0.22123017932218406, "learning_rate": 1.6918970654954084e-05, "loss": 0.3457, "step": 731 }, { "epoch": 2.9650632911392405, "grad_norm": 0.2166294721817881, "learning_rate": 1.686290200071915e-05, "loss": 0.3681, "step": 732 }, { "epoch": 2.969113924050633, "grad_norm": 0.2290988546616795, "learning_rate": 1.6806858612240234e-05, "loss": 0.3553, "step": 733 }, { "epoch": 2.9731645569620255, "grad_norm": 0.23202529097783417, "learning_rate": 1.6750840940883078e-05, "loss": 0.3558, "step": 734 }, { "epoch": 2.9772151898734176, "grad_norm": 0.23798449551526937, "learning_rate": 1.6694849437806305e-05, "loss": 0.361, "step": 735 }, { "epoch": 2.98126582278481, "grad_norm": 0.24259785857008317, "learning_rate": 1.663888455395778e-05, "loss": 0.3515, "step": 736 }, { "epoch": 2.9853164556962026, "grad_norm": 0.23944971388445735, "learning_rate": 1.6582946740070995e-05, "loss": 0.3576, "step": 737 }, { "epoch": 2.9893670886075947, "grad_norm": 0.24056041873701758, "learning_rate": 1.6527036446661396e-05, "loss": 0.3377, "step": 738 }, { "epoch": 2.993417721518987, "grad_norm": 0.2554661531848799, "learning_rate": 1.6471154124022818e-05, "loss": 0.3651, "step": 739 }, { "epoch": 2.9974683544303797, "grad_norm": 0.26722268320978365, "learning_rate": 1.6415300222223788e-05, "loss": 0.354, "step": 740 }, { "epoch": 3.0015189873417722, "grad_norm": 0.3425888270150753, "learning_rate": 1.6359475191103958e-05, "loss": 0.3381, "step": 741 }, { "epoch": 3.0055696202531648, "grad_norm": 0.3313150470815187, "learning_rate": 1.6303679480270466e-05, "loss": 0.3079, "step": 742 }, { "epoch": 3.009620253164557, "grad_norm": 0.4564998543641076, "learning_rate": 1.624791353909428e-05, "loss": 0.3052, "step": 743 }, { "epoch": 3.0136708860759494, "grad_norm": 0.3100666578431888, "learning_rate": 1.619217781670663e-05, "loss": 0.302, "step": 744 }, { "epoch": 3.017721518987342, "grad_norm": 0.3594276958917497, "learning_rate": 1.6136472761995373e-05, "loss": 0.2915, "step": 745 }, { "epoch": 3.0217721518987344, "grad_norm": 0.3017414687748239, "learning_rate": 1.608079882360134e-05, "loss": 0.3066, "step": 746 }, { "epoch": 3.0258227848101265, "grad_norm": 0.31306391861880023, "learning_rate": 1.60251564499148e-05, "loss": 0.3068, "step": 747 }, { "epoch": 3.029873417721519, "grad_norm": 0.3069473916155731, "learning_rate": 1.596954608907176e-05, "loss": 0.2979, "step": 748 }, { "epoch": 3.0339240506329115, "grad_norm": 0.2673190795441396, "learning_rate": 1.591396818895043e-05, "loss": 0.3117, "step": 749 }, { "epoch": 3.037974683544304, "grad_norm": 0.2937916146565775, "learning_rate": 1.585842319716759e-05, "loss": 0.3059, "step": 750 }, { "epoch": 3.042025316455696, "grad_norm": 0.28955007694944807, "learning_rate": 1.5802911561074944e-05, "loss": 0.3067, "step": 751 }, { "epoch": 3.0460759493670886, "grad_norm": 0.33091918347626825, "learning_rate": 1.5747433727755595e-05, "loss": 0.3018, "step": 752 }, { "epoch": 3.050126582278481, "grad_norm": 0.262240618019802, "learning_rate": 1.5691990144020376e-05, "loss": 0.3159, "step": 753 }, { "epoch": 3.0541772151898736, "grad_norm": 0.27863387099635717, "learning_rate": 1.5636581256404297e-05, "loss": 0.3184, "step": 754 }, { "epoch": 3.0582278481012657, "grad_norm": 0.24511156082981866, "learning_rate": 1.558120751116291e-05, "loss": 0.3053, "step": 755 }, { "epoch": 3.062278481012658, "grad_norm": 0.29418124420998093, "learning_rate": 1.552586935426876e-05, "loss": 0.296, "step": 756 }, { "epoch": 3.0663291139240507, "grad_norm": 0.2544588284908696, "learning_rate": 1.547056723140774e-05, "loss": 0.2992, "step": 757 }, { "epoch": 3.0703797468354432, "grad_norm": 0.26741538744132, "learning_rate": 1.5415301587975565e-05, "loss": 0.3038, "step": 758 }, { "epoch": 3.0744303797468353, "grad_norm": 0.26791415254832257, "learning_rate": 1.536007286907411e-05, "loss": 0.3088, "step": 759 }, { "epoch": 3.078481012658228, "grad_norm": 0.2392599550154332, "learning_rate": 1.5304881519507896e-05, "loss": 0.3117, "step": 760 }, { "epoch": 3.0825316455696203, "grad_norm": 0.2746493096548342, "learning_rate": 1.5249727983780453e-05, "loss": 0.3005, "step": 761 }, { "epoch": 3.086582278481013, "grad_norm": 0.24163185665862896, "learning_rate": 1.5194612706090786e-05, "loss": 0.3067, "step": 762 }, { "epoch": 3.090632911392405, "grad_norm": 0.2700010199555972, "learning_rate": 1.5139536130329771e-05, "loss": 0.3062, "step": 763 }, { "epoch": 3.0946835443037974, "grad_norm": 0.24618310461398976, "learning_rate": 1.508449870007656e-05, "loss": 0.3039, "step": 764 }, { "epoch": 3.09873417721519, "grad_norm": 0.2652656273936138, "learning_rate": 1.5029500858595056e-05, "loss": 0.3001, "step": 765 }, { "epoch": 3.1027848101265825, "grad_norm": 0.2380837269983529, "learning_rate": 1.4974543048830328e-05, "loss": 0.2933, "step": 766 }, { "epoch": 3.1068354430379745, "grad_norm": 0.2960387431215136, "learning_rate": 1.4919625713405e-05, "loss": 0.3066, "step": 767 }, { "epoch": 3.110886075949367, "grad_norm": 0.2602171918553114, "learning_rate": 1.4864749294615756e-05, "loss": 0.2997, "step": 768 }, { "epoch": 3.1149367088607596, "grad_norm": 0.28854620126908503, "learning_rate": 1.4809914234429716e-05, "loss": 0.2952, "step": 769 }, { "epoch": 3.118987341772152, "grad_norm": 0.22953770583137745, "learning_rate": 1.4755120974480923e-05, "loss": 0.3084, "step": 770 }, { "epoch": 3.123037974683544, "grad_norm": 0.24316956783078447, "learning_rate": 1.4700369956066771e-05, "loss": 0.3057, "step": 771 }, { "epoch": 3.1270886075949367, "grad_norm": 0.22096892852115443, "learning_rate": 1.4645661620144413e-05, "loss": 0.3094, "step": 772 }, { "epoch": 3.131139240506329, "grad_norm": 0.24778827458028146, "learning_rate": 1.4590996407327284e-05, "loss": 0.299, "step": 773 }, { "epoch": 3.1351898734177217, "grad_norm": 0.24275376987918432, "learning_rate": 1.4536374757881487e-05, "loss": 0.3103, "step": 774 }, { "epoch": 3.1392405063291138, "grad_norm": 0.25528133034628997, "learning_rate": 1.4481797111722271e-05, "loss": 0.3048, "step": 775 }, { "epoch": 3.1432911392405063, "grad_norm": 0.25471383027081573, "learning_rate": 1.4427263908410507e-05, "loss": 0.3014, "step": 776 }, { "epoch": 3.147341772151899, "grad_norm": 0.25880066648315286, "learning_rate": 1.4372775587149108e-05, "loss": 0.3026, "step": 777 }, { "epoch": 3.1513924050632913, "grad_norm": 0.26549545874758546, "learning_rate": 1.4318332586779522e-05, "loss": 0.3048, "step": 778 }, { "epoch": 3.1554430379746834, "grad_norm": 0.256910116833301, "learning_rate": 1.4263935345778202e-05, "loss": 0.3117, "step": 779 }, { "epoch": 3.159493670886076, "grad_norm": 0.31149773495010513, "learning_rate": 1.420958430225303e-05, "loss": 0.3082, "step": 780 }, { "epoch": 3.1635443037974684, "grad_norm": 0.2232255784239736, "learning_rate": 1.415527989393985e-05, "loss": 0.2955, "step": 781 }, { "epoch": 3.167594936708861, "grad_norm": 0.27573836179026207, "learning_rate": 1.410102255819891e-05, "loss": 0.309, "step": 782 }, { "epoch": 3.171645569620253, "grad_norm": 0.23660011624279467, "learning_rate": 1.404681273201131e-05, "loss": 0.307, "step": 783 }, { "epoch": 3.1756962025316455, "grad_norm": 0.27898009413434033, "learning_rate": 1.399265085197556e-05, "loss": 0.298, "step": 784 }, { "epoch": 3.179746835443038, "grad_norm": 0.23760771112750403, "learning_rate": 1.393853735430398e-05, "loss": 0.2957, "step": 785 }, { "epoch": 3.1837974683544306, "grad_norm": 0.253581176271666, "learning_rate": 1.3884472674819246e-05, "loss": 0.3078, "step": 786 }, { "epoch": 3.1878481012658226, "grad_norm": 0.23786882860798483, "learning_rate": 1.3830457248950864e-05, "loss": 0.2966, "step": 787 }, { "epoch": 3.191898734177215, "grad_norm": 0.248809614601311, "learning_rate": 1.377649151173163e-05, "loss": 0.3003, "step": 788 }, { "epoch": 3.1959493670886077, "grad_norm": 0.2790110777580637, "learning_rate": 1.3722575897794181e-05, "loss": 0.296, "step": 789 }, { "epoch": 3.2, "grad_norm": 0.22126108451814072, "learning_rate": 1.3668710841367472e-05, "loss": 0.3049, "step": 790 }, { "epoch": 3.2040506329113922, "grad_norm": 0.3108180121509644, "learning_rate": 1.361489677627324e-05, "loss": 0.2972, "step": 791 }, { "epoch": 3.2081012658227848, "grad_norm": 0.24557822237760302, "learning_rate": 1.3561134135922585e-05, "loss": 0.3114, "step": 792 }, { "epoch": 3.2121518987341773, "grad_norm": 0.35033565861911536, "learning_rate": 1.350742335331241e-05, "loss": 0.3079, "step": 793 }, { "epoch": 3.21620253164557, "grad_norm": 0.2919444276073894, "learning_rate": 1.345376486102198e-05, "loss": 0.3185, "step": 794 }, { "epoch": 3.220253164556962, "grad_norm": 0.262364895978286, "learning_rate": 1.3400159091209414e-05, "loss": 0.3035, "step": 795 }, { "epoch": 3.2243037974683544, "grad_norm": 0.2884229777494557, "learning_rate": 1.3346606475608216e-05, "loss": 0.3145, "step": 796 }, { "epoch": 3.228354430379747, "grad_norm": 0.2647752327566661, "learning_rate": 1.3293107445523781e-05, "loss": 0.2997, "step": 797 }, { "epoch": 3.2324050632911394, "grad_norm": 0.24614849577660988, "learning_rate": 1.3239662431829949e-05, "loss": 0.3016, "step": 798 }, { "epoch": 3.2364556962025315, "grad_norm": 0.26249896937101325, "learning_rate": 1.3186271864965509e-05, "loss": 0.2991, "step": 799 }, { "epoch": 3.240506329113924, "grad_norm": 0.23100577941289274, "learning_rate": 1.3132936174930756e-05, "loss": 0.313, "step": 800 }, { "epoch": 3.2445569620253165, "grad_norm": 0.26237786208120456, "learning_rate": 1.3079655791283995e-05, "loss": 0.3011, "step": 801 }, { "epoch": 3.248607594936709, "grad_norm": 0.24441502314890692, "learning_rate": 1.3026431143138108e-05, "loss": 0.3061, "step": 802 }, { "epoch": 3.252658227848101, "grad_norm": 0.21486835614793737, "learning_rate": 1.2973262659157114e-05, "loss": 0.2979, "step": 803 }, { "epoch": 3.2567088607594936, "grad_norm": 0.27494499032971986, "learning_rate": 1.2920150767552651e-05, "loss": 0.3153, "step": 804 }, { "epoch": 3.260759493670886, "grad_norm": 0.23550615238099812, "learning_rate": 1.2867095896080607e-05, "loss": 0.3035, "step": 805 }, { "epoch": 3.2648101265822786, "grad_norm": 0.22788995367488474, "learning_rate": 1.2814098472037612e-05, "loss": 0.3093, "step": 806 }, { "epoch": 3.2688607594936707, "grad_norm": 0.27385910186177964, "learning_rate": 1.276115892225764e-05, "loss": 0.2874, "step": 807 }, { "epoch": 3.2729113924050632, "grad_norm": 0.2585000723324782, "learning_rate": 1.2708277673108555e-05, "loss": 0.2984, "step": 808 }, { "epoch": 3.2769620253164558, "grad_norm": 0.27321692653989915, "learning_rate": 1.2655455150488649e-05, "loss": 0.3046, "step": 809 }, { "epoch": 3.2810126582278483, "grad_norm": 0.24020101150076276, "learning_rate": 1.2602691779823272e-05, "loss": 0.3015, "step": 810 }, { "epoch": 3.2850632911392403, "grad_norm": 0.2335971544607564, "learning_rate": 1.2549987986061355e-05, "loss": 0.3101, "step": 811 }, { "epoch": 3.289113924050633, "grad_norm": 0.2553361725393974, "learning_rate": 1.2497344193672005e-05, "loss": 0.3046, "step": 812 }, { "epoch": 3.2931645569620254, "grad_norm": 0.31343100999336954, "learning_rate": 1.2444760826641092e-05, "loss": 0.3079, "step": 813 }, { "epoch": 3.297215189873418, "grad_norm": 0.24559367252067774, "learning_rate": 1.2392238308467817e-05, "loss": 0.3034, "step": 814 }, { "epoch": 3.30126582278481, "grad_norm": 0.2622530879012044, "learning_rate": 1.2339777062161326e-05, "loss": 0.3065, "step": 815 }, { "epoch": 3.3053164556962025, "grad_norm": 0.25760598915174143, "learning_rate": 1.2287377510237293e-05, "loss": 0.3075, "step": 816 }, { "epoch": 3.309367088607595, "grad_norm": 0.25183926506767373, "learning_rate": 1.2235040074714488e-05, "loss": 0.3049, "step": 817 }, { "epoch": 3.3134177215189875, "grad_norm": 0.2668896128086339, "learning_rate": 1.2182765177111434e-05, "loss": 0.3114, "step": 818 }, { "epoch": 3.3174683544303796, "grad_norm": 0.23262242732779898, "learning_rate": 1.213055323844297e-05, "loss": 0.2981, "step": 819 }, { "epoch": 3.321518987341772, "grad_norm": 0.25223965161039613, "learning_rate": 1.2078404679216864e-05, "loss": 0.3037, "step": 820 }, { "epoch": 3.3255696202531646, "grad_norm": 0.24605485137219998, "learning_rate": 1.2026319919430458e-05, "loss": 0.3009, "step": 821 }, { "epoch": 3.329620253164557, "grad_norm": 0.2133798845078682, "learning_rate": 1.1974299378567227e-05, "loss": 0.3011, "step": 822 }, { "epoch": 3.333670886075949, "grad_norm": 0.23525279882503228, "learning_rate": 1.1922343475593462e-05, "loss": 0.3077, "step": 823 }, { "epoch": 3.3377215189873417, "grad_norm": 0.23093480687476123, "learning_rate": 1.187045262895488e-05, "loss": 0.3091, "step": 824 }, { "epoch": 3.3417721518987342, "grad_norm": 0.2348147374879687, "learning_rate": 1.1818627256573203e-05, "loss": 0.3042, "step": 825 }, { "epoch": 3.3458227848101267, "grad_norm": 0.2474178249067499, "learning_rate": 1.1766867775842864e-05, "loss": 0.3058, "step": 826 }, { "epoch": 3.349873417721519, "grad_norm": 0.22763985854488628, "learning_rate": 1.1715174603627615e-05, "loss": 0.2985, "step": 827 }, { "epoch": 3.3539240506329113, "grad_norm": 0.22313188650939966, "learning_rate": 1.1663548156257147e-05, "loss": 0.2969, "step": 828 }, { "epoch": 3.357974683544304, "grad_norm": 0.22761781729640918, "learning_rate": 1.161198884952377e-05, "loss": 0.3022, "step": 829 }, { "epoch": 3.3620253164556964, "grad_norm": 0.23279093680225235, "learning_rate": 1.1560497098679056e-05, "loss": 0.3148, "step": 830 }, { "epoch": 3.3660759493670884, "grad_norm": 0.2154358810036318, "learning_rate": 1.1509073318430479e-05, "loss": 0.3038, "step": 831 }, { "epoch": 3.370126582278481, "grad_norm": 0.23099366720299935, "learning_rate": 1.1457717922938116e-05, "loss": 0.3139, "step": 832 }, { "epoch": 3.3741772151898735, "grad_norm": 0.20667386728595938, "learning_rate": 1.1406431325811233e-05, "loss": 0.293, "step": 833 }, { "epoch": 3.378227848101266, "grad_norm": 0.21463591317790967, "learning_rate": 1.135521394010506e-05, "loss": 0.3043, "step": 834 }, { "epoch": 3.382278481012658, "grad_norm": 0.21600476062709234, "learning_rate": 1.1304066178317367e-05, "loss": 0.2914, "step": 835 }, { "epoch": 3.3863291139240506, "grad_norm": 0.22034905283342088, "learning_rate": 1.1252988452385199e-05, "loss": 0.2921, "step": 836 }, { "epoch": 3.390379746835443, "grad_norm": 0.22474078510449785, "learning_rate": 1.1201981173681536e-05, "loss": 0.3069, "step": 837 }, { "epoch": 3.3944303797468356, "grad_norm": 0.21656466014051962, "learning_rate": 1.1151044753011991e-05, "loss": 0.288, "step": 838 }, { "epoch": 3.3984810126582277, "grad_norm": 0.22754128297993984, "learning_rate": 1.1100179600611491e-05, "loss": 0.3031, "step": 839 }, { "epoch": 3.40253164556962, "grad_norm": 0.23418579598218983, "learning_rate": 1.1049386126140985e-05, "loss": 0.3126, "step": 840 }, { "epoch": 3.4065822784810127, "grad_norm": 0.1996155647132429, "learning_rate": 1.0998664738684128e-05, "loss": 0.2955, "step": 841 }, { "epoch": 3.410632911392405, "grad_norm": 0.2322487966316973, "learning_rate": 1.0948015846744e-05, "loss": 0.3011, "step": 842 }, { "epoch": 3.4146835443037973, "grad_norm": 0.23223468499584005, "learning_rate": 1.0897439858239832e-05, "loss": 0.3023, "step": 843 }, { "epoch": 3.41873417721519, "grad_norm": 0.20814764527010607, "learning_rate": 1.0846937180503652e-05, "loss": 0.3031, "step": 844 }, { "epoch": 3.4227848101265823, "grad_norm": 0.2377855922849024, "learning_rate": 1.0796508220277117e-05, "loss": 0.2971, "step": 845 }, { "epoch": 3.426835443037975, "grad_norm": 0.22293373346875606, "learning_rate": 1.0746153383708107e-05, "loss": 0.2889, "step": 846 }, { "epoch": 3.430886075949367, "grad_norm": 0.2199371250483576, "learning_rate": 1.0695873076347579e-05, "loss": 0.2964, "step": 847 }, { "epoch": 3.4349367088607594, "grad_norm": 0.2288457494084828, "learning_rate": 1.0645667703146205e-05, "loss": 0.3147, "step": 848 }, { "epoch": 3.438987341772152, "grad_norm": 0.22444973671638674, "learning_rate": 1.0595537668451161e-05, "loss": 0.306, "step": 849 }, { "epoch": 3.4430379746835444, "grad_norm": 0.22833339000959885, "learning_rate": 1.0545483376002854e-05, "loss": 0.2984, "step": 850 }, { "epoch": 3.4470886075949365, "grad_norm": 0.22350879835220694, "learning_rate": 1.0495505228931676e-05, "loss": 0.3051, "step": 851 }, { "epoch": 3.451139240506329, "grad_norm": 0.225219437960772, "learning_rate": 1.044560362975474e-05, "loss": 0.3075, "step": 852 }, { "epoch": 3.4551898734177215, "grad_norm": 0.2358035934408475, "learning_rate": 1.0395778980372695e-05, "loss": 0.3027, "step": 853 }, { "epoch": 3.459240506329114, "grad_norm": 0.210102453032059, "learning_rate": 1.0346031682066381e-05, "loss": 0.3066, "step": 854 }, { "epoch": 3.463291139240506, "grad_norm": 0.22545257742260238, "learning_rate": 1.0296362135493724e-05, "loss": 0.3124, "step": 855 }, { "epoch": 3.4673417721518986, "grad_norm": 0.2008276738416522, "learning_rate": 1.0246770740686422e-05, "loss": 0.293, "step": 856 }, { "epoch": 3.471392405063291, "grad_norm": 0.21091461375323128, "learning_rate": 1.0197257897046743e-05, "loss": 0.299, "step": 857 }, { "epoch": 3.4754430379746837, "grad_norm": 0.20668484614630603, "learning_rate": 1.014782400334433e-05, "loss": 0.3084, "step": 858 }, { "epoch": 3.479493670886076, "grad_norm": 0.26794665728286965, "learning_rate": 1.009846945771296e-05, "loss": 0.305, "step": 859 }, { "epoch": 3.4835443037974683, "grad_norm": 0.2282645342948944, "learning_rate": 1.0049194657647363e-05, "loss": 0.3105, "step": 860 }, { "epoch": 3.487594936708861, "grad_norm": 0.2101591611674635, "learning_rate": 1.0000000000000006e-05, "loss": 0.303, "step": 861 }, { "epoch": 3.4916455696202533, "grad_norm": 0.2230752578939505, "learning_rate": 9.950885880977891e-06, "loss": 0.3073, "step": 862 }, { "epoch": 3.4956962025316454, "grad_norm": 0.2045566965079551, "learning_rate": 9.901852696139382e-06, "loss": 0.2911, "step": 863 }, { "epoch": 3.499746835443038, "grad_norm": 0.21264971413355024, "learning_rate": 9.852900840391027e-06, "loss": 0.3094, "step": 864 }, { "epoch": 3.5037974683544304, "grad_norm": 0.21078866778909822, "learning_rate": 9.804030707984313e-06, "loss": 0.2994, "step": 865 }, { "epoch": 3.507848101265823, "grad_norm": 0.21970452364777016, "learning_rate": 9.755242692512599e-06, "loss": 0.2964, "step": 866 }, { "epoch": 3.5118987341772154, "grad_norm": 0.20304156473258475, "learning_rate": 9.70653718690782e-06, "loss": 0.2941, "step": 867 }, { "epoch": 3.5159493670886075, "grad_norm": 0.2008403699900234, "learning_rate": 9.657914583437454e-06, "loss": 0.2985, "step": 868 }, { "epoch": 3.52, "grad_norm": 0.2297875491951306, "learning_rate": 9.609375273701246e-06, "loss": 0.2992, "step": 869 }, { "epoch": 3.5240506329113925, "grad_norm": 0.23261031007670416, "learning_rate": 9.560919648628133e-06, "loss": 0.2931, "step": 870 }, { "epoch": 3.5281012658227846, "grad_norm": 0.19813893881465705, "learning_rate": 9.512548098473047e-06, "loss": 0.2956, "step": 871 }, { "epoch": 3.532151898734177, "grad_norm": 0.20522417544224172, "learning_rate": 9.464261012813825e-06, "loss": 0.3056, "step": 872 }, { "epoch": 3.5362025316455696, "grad_norm": 0.20688688294437843, "learning_rate": 9.416058780547987e-06, "loss": 0.303, "step": 873 }, { "epoch": 3.540253164556962, "grad_norm": 0.2066907023431564, "learning_rate": 9.367941789889714e-06, "loss": 0.3089, "step": 874 }, { "epoch": 3.5443037974683547, "grad_norm": 0.21259558877288492, "learning_rate": 9.319910428366607e-06, "loss": 0.3112, "step": 875 }, { "epoch": 3.5483544303797467, "grad_norm": 0.19356567678633826, "learning_rate": 9.271965082816667e-06, "loss": 0.3032, "step": 876 }, { "epoch": 3.5524050632911393, "grad_norm": 0.21344259724427736, "learning_rate": 9.224106139385111e-06, "loss": 0.3044, "step": 877 }, { "epoch": 3.5564556962025318, "grad_norm": 0.20997789521105398, "learning_rate": 9.176333983521291e-06, "loss": 0.3055, "step": 878 }, { "epoch": 3.560506329113924, "grad_norm": 0.2055775755327742, "learning_rate": 9.12864899997558e-06, "loss": 0.2955, "step": 879 }, { "epoch": 3.5645569620253164, "grad_norm": 0.2032013317605556, "learning_rate": 9.08105157279628e-06, "loss": 0.2981, "step": 880 }, { "epoch": 3.568607594936709, "grad_norm": 0.19828853643666902, "learning_rate": 9.03354208532653e-06, "loss": 0.3047, "step": 881 }, { "epoch": 3.5726582278481014, "grad_norm": 0.21454217836690237, "learning_rate": 8.986120920201205e-06, "loss": 0.3062, "step": 882 }, { "epoch": 3.576708860759494, "grad_norm": 0.20056265689027003, "learning_rate": 8.938788459343852e-06, "loss": 0.3087, "step": 883 }, { "epoch": 3.580759493670886, "grad_norm": 0.2184782761501924, "learning_rate": 8.8915450839636e-06, "loss": 0.2991, "step": 884 }, { "epoch": 3.5848101265822785, "grad_norm": 0.22241797371289915, "learning_rate": 8.844391174552116e-06, "loss": 0.2984, "step": 885 }, { "epoch": 3.588860759493671, "grad_norm": 0.2018275469206433, "learning_rate": 8.797327110880479e-06, "loss": 0.3032, "step": 886 }, { "epoch": 3.592911392405063, "grad_norm": 0.2191670868650659, "learning_rate": 8.750353271996206e-06, "loss": 0.311, "step": 887 }, { "epoch": 3.5969620253164556, "grad_norm": 0.2471043281244445, "learning_rate": 8.703470036220132e-06, "loss": 0.297, "step": 888 }, { "epoch": 3.601012658227848, "grad_norm": 0.22071254274626972, "learning_rate": 8.656677781143394e-06, "loss": 0.3024, "step": 889 }, { "epoch": 3.6050632911392406, "grad_norm": 0.20597153822613434, "learning_rate": 8.609976883624377e-06, "loss": 0.3099, "step": 890 }, { "epoch": 3.609113924050633, "grad_norm": 0.22935105939085257, "learning_rate": 8.563367719785698e-06, "loss": 0.3044, "step": 891 }, { "epoch": 3.613164556962025, "grad_norm": 0.23473550102203986, "learning_rate": 8.516850665011138e-06, "loss": 0.2961, "step": 892 }, { "epoch": 3.6172151898734177, "grad_norm": 0.2263309709701646, "learning_rate": 8.47042609394269e-06, "loss": 0.3021, "step": 893 }, { "epoch": 3.6212658227848102, "grad_norm": 0.20956128233930382, "learning_rate": 8.424094380477432e-06, "loss": 0.3086, "step": 894 }, { "epoch": 3.6253164556962023, "grad_norm": 0.240471806478542, "learning_rate": 8.37785589776465e-06, "loss": 0.3093, "step": 895 }, { "epoch": 3.629367088607595, "grad_norm": 0.26684623465319424, "learning_rate": 8.331711018202694e-06, "loss": 0.2961, "step": 896 }, { "epoch": 3.6334177215189873, "grad_norm": 0.21853920188539874, "learning_rate": 8.285660113436104e-06, "loss": 0.3015, "step": 897 }, { "epoch": 3.63746835443038, "grad_norm": 0.21309870944851686, "learning_rate": 8.239703554352527e-06, "loss": 0.3101, "step": 898 }, { "epoch": 3.6415189873417724, "grad_norm": 0.24886917991842616, "learning_rate": 8.193841711079775e-06, "loss": 0.3011, "step": 899 }, { "epoch": 3.6455696202531644, "grad_norm": 0.23035321889000995, "learning_rate": 8.148074952982828e-06, "loss": 0.3064, "step": 900 }, { "epoch": 3.649620253164557, "grad_norm": 0.21606634563803992, "learning_rate": 8.102403648660859e-06, "loss": 0.309, "step": 901 }, { "epoch": 3.6536708860759495, "grad_norm": 0.229375507385167, "learning_rate": 8.056828165944282e-06, "loss": 0.3079, "step": 902 }, { "epoch": 3.6577215189873415, "grad_norm": 0.21602996796868992, "learning_rate": 8.011348871891762e-06, "loss": 0.3132, "step": 903 }, { "epoch": 3.661772151898734, "grad_norm": 0.22240155079351762, "learning_rate": 7.965966132787287e-06, "loss": 0.3067, "step": 904 }, { "epoch": 3.6658227848101266, "grad_norm": 0.23200433222384975, "learning_rate": 7.920680314137189e-06, "loss": 0.3012, "step": 905 }, { "epoch": 3.669873417721519, "grad_norm": 0.20941231896517232, "learning_rate": 7.875491780667246e-06, "loss": 0.3093, "step": 906 }, { "epoch": 3.6739240506329116, "grad_norm": 0.22534454822532676, "learning_rate": 7.830400896319667e-06, "loss": 0.3008, "step": 907 }, { "epoch": 3.6779746835443037, "grad_norm": 0.22605804156919124, "learning_rate": 7.785408024250259e-06, "loss": 0.2972, "step": 908 }, { "epoch": 3.682025316455696, "grad_norm": 0.21752248563606183, "learning_rate": 7.74051352682542e-06, "loss": 0.3002, "step": 909 }, { "epoch": 3.6860759493670887, "grad_norm": 0.23756241941341258, "learning_rate": 7.695717765619257e-06, "loss": 0.3059, "step": 910 }, { "epoch": 3.690126582278481, "grad_norm": 0.21733242551406146, "learning_rate": 7.651021101410673e-06, "loss": 0.2995, "step": 911 }, { "epoch": 3.6941772151898733, "grad_norm": 0.2355772741694523, "learning_rate": 7.606423894180464e-06, "loss": 0.3007, "step": 912 }, { "epoch": 3.698227848101266, "grad_norm": 0.21437204719323288, "learning_rate": 7.56192650310839e-06, "loss": 0.2929, "step": 913 }, { "epoch": 3.7022784810126583, "grad_norm": 0.2282136781974808, "learning_rate": 7.517529286570349e-06, "loss": 0.3104, "step": 914 }, { "epoch": 3.706329113924051, "grad_norm": 0.20596535781036066, "learning_rate": 7.473232602135387e-06, "loss": 0.3046, "step": 915 }, { "epoch": 3.710379746835443, "grad_norm": 0.19150963664163637, "learning_rate": 7.429036806562935e-06, "loss": 0.3057, "step": 916 }, { "epoch": 3.7144303797468354, "grad_norm": 0.21156305579388207, "learning_rate": 7.3849422557998455e-06, "loss": 0.3064, "step": 917 }, { "epoch": 3.718481012658228, "grad_norm": 0.2058629081387625, "learning_rate": 7.340949304977567e-06, "loss": 0.313, "step": 918 }, { "epoch": 3.72253164556962, "grad_norm": 0.20040439900090334, "learning_rate": 7.297058308409282e-06, "loss": 0.2966, "step": 919 }, { "epoch": 3.7265822784810125, "grad_norm": 0.2185105513643437, "learning_rate": 7.25326961958704e-06, "loss": 0.3112, "step": 920 }, { "epoch": 3.730632911392405, "grad_norm": 0.24930124253480204, "learning_rate": 7.209583591178921e-06, "loss": 0.3063, "step": 921 }, { "epoch": 3.7346835443037976, "grad_norm": 0.20487196082271142, "learning_rate": 7.1660005750261925e-06, "loss": 0.3044, "step": 922 }, { "epoch": 3.73873417721519, "grad_norm": 0.19873260022428135, "learning_rate": 7.1225209221404765e-06, "loss": 0.3027, "step": 923 }, { "epoch": 3.742784810126582, "grad_norm": 0.2184533460555972, "learning_rate": 7.079144982700909e-06, "loss": 0.3001, "step": 924 }, { "epoch": 3.7468354430379747, "grad_norm": 0.22880275340494977, "learning_rate": 7.0358731060513695e-06, "loss": 0.2922, "step": 925 }, { "epoch": 3.750886075949367, "grad_norm": 0.21801039500112102, "learning_rate": 6.99270564069757e-06, "loss": 0.3014, "step": 926 }, { "epoch": 3.7549367088607593, "grad_norm": 0.21972679707039325, "learning_rate": 6.949642934304375e-06, "loss": 0.3034, "step": 927 }, { "epoch": 3.7589873417721518, "grad_norm": 0.22321058199059743, "learning_rate": 6.906685333692871e-06, "loss": 0.3102, "step": 928 }, { "epoch": 3.7630379746835443, "grad_norm": 0.21841782133284252, "learning_rate": 6.86383318483769e-06, "loss": 0.3064, "step": 929 }, { "epoch": 3.767088607594937, "grad_norm": 0.19778283684922807, "learning_rate": 6.821086832864139e-06, "loss": 0.3106, "step": 930 }, { "epoch": 3.7711392405063293, "grad_norm": 0.20211865997239342, "learning_rate": 6.77844662204546e-06, "loss": 0.2919, "step": 931 }, { "epoch": 3.7751898734177214, "grad_norm": 0.21795876133232991, "learning_rate": 6.7359128958000455e-06, "loss": 0.3076, "step": 932 }, { "epoch": 3.779240506329114, "grad_norm": 0.20552385294982045, "learning_rate": 6.693485996688695e-06, "loss": 0.2981, "step": 933 }, { "epoch": 3.7832911392405064, "grad_norm": 0.1949182851527134, "learning_rate": 6.651166266411801e-06, "loss": 0.3106, "step": 934 }, { "epoch": 3.7873417721518985, "grad_norm": 0.18902384503221442, "learning_rate": 6.6089540458066725e-06, "loss": 0.2932, "step": 935 }, { "epoch": 3.791392405063291, "grad_norm": 0.21375125925391122, "learning_rate": 6.566849674844711e-06, "loss": 0.2993, "step": 936 }, { "epoch": 3.7954430379746835, "grad_norm": 0.21097825790393973, "learning_rate": 6.524853492628747e-06, "loss": 0.3029, "step": 937 }, { "epoch": 3.799493670886076, "grad_norm": 0.19260585660485505, "learning_rate": 6.4829658373902536e-06, "loss": 0.2983, "step": 938 }, { "epoch": 3.8035443037974686, "grad_norm": 0.21915807405861926, "learning_rate": 6.441187046486648e-06, "loss": 0.3073, "step": 939 }, { "epoch": 3.8075949367088606, "grad_norm": 0.19633072461245998, "learning_rate": 6.399517456398567e-06, "loss": 0.3078, "step": 940 }, { "epoch": 3.811645569620253, "grad_norm": 0.23527506528125705, "learning_rate": 6.357957402727164e-06, "loss": 0.2945, "step": 941 }, { "epoch": 3.8156962025316457, "grad_norm": 0.22505986087943358, "learning_rate": 6.316507220191395e-06, "loss": 0.3057, "step": 942 }, { "epoch": 3.8197468354430377, "grad_norm": 0.20685631735845467, "learning_rate": 6.275167242625331e-06, "loss": 0.3082, "step": 943 }, { "epoch": 3.8237974683544302, "grad_norm": 0.20455126356972106, "learning_rate": 6.233937802975471e-06, "loss": 0.3133, "step": 944 }, { "epoch": 3.8278481012658228, "grad_norm": 0.22199719929629194, "learning_rate": 6.192819233298046e-06, "loss": 0.3147, "step": 945 }, { "epoch": 3.8318987341772153, "grad_norm": 0.1982998887680543, "learning_rate": 6.151811864756383e-06, "loss": 0.3005, "step": 946 }, { "epoch": 3.835949367088608, "grad_norm": 0.19601081069367987, "learning_rate": 6.1109160276181655e-06, "loss": 0.2986, "step": 947 }, { "epoch": 3.84, "grad_norm": 0.20515868449587227, "learning_rate": 6.070132051252868e-06, "loss": 0.2949, "step": 948 }, { "epoch": 3.8440506329113924, "grad_norm": 0.20443849413660453, "learning_rate": 6.0294602641290034e-06, "loss": 0.3065, "step": 949 }, { "epoch": 3.848101265822785, "grad_norm": 0.19209706345870914, "learning_rate": 5.988900993811575e-06, "loss": 0.2979, "step": 950 }, { "epoch": 3.852151898734177, "grad_norm": 0.19452748591910599, "learning_rate": 5.948454566959363e-06, "loss": 0.3131, "step": 951 }, { "epoch": 3.8562025316455695, "grad_norm": 0.19295403324977134, "learning_rate": 5.908121309322328e-06, "loss": 0.3087, "step": 952 }, { "epoch": 3.860253164556962, "grad_norm": 0.21433534113209154, "learning_rate": 5.867901545738976e-06, "loss": 0.3074, "step": 953 }, { "epoch": 3.8643037974683545, "grad_norm": 0.1906561037017946, "learning_rate": 5.827795600133774e-06, "loss": 0.3052, "step": 954 }, { "epoch": 3.868354430379747, "grad_norm": 0.21151707090844765, "learning_rate": 5.787803795514466e-06, "loss": 0.2991, "step": 955 }, { "epoch": 3.872405063291139, "grad_norm": 0.20711582275778048, "learning_rate": 5.747926453969576e-06, "loss": 0.3255, "step": 956 }, { "epoch": 3.8764556962025316, "grad_norm": 0.2050136137816677, "learning_rate": 5.708163896665708e-06, "loss": 0.2832, "step": 957 }, { "epoch": 3.880506329113924, "grad_norm": 0.1871022983268096, "learning_rate": 5.668516443845047e-06, "loss": 0.2995, "step": 958 }, { "epoch": 3.884556962025316, "grad_norm": 0.2109800793885071, "learning_rate": 5.6289844148227225e-06, "loss": 0.3042, "step": 959 }, { "epoch": 3.8886075949367087, "grad_norm": 0.19689400273015953, "learning_rate": 5.5895681279842615e-06, "loss": 0.3106, "step": 960 }, { "epoch": 3.8926582278481012, "grad_norm": 0.1908835943025496, "learning_rate": 5.550267900783019e-06, "loss": 0.2995, "step": 961 }, { "epoch": 3.8967088607594937, "grad_norm": 0.19776550381119962, "learning_rate": 5.511084049737623e-06, "loss": 0.3068, "step": 962 }, { "epoch": 3.9007594936708863, "grad_norm": 0.20506311097529226, "learning_rate": 5.4720168904294215e-06, "loss": 0.303, "step": 963 }, { "epoch": 3.9048101265822783, "grad_norm": 0.19378178907396437, "learning_rate": 5.433066737499948e-06, "loss": 0.2984, "step": 964 }, { "epoch": 3.908860759493671, "grad_norm": 0.19722020458034104, "learning_rate": 5.394233904648376e-06, "loss": 0.3064, "step": 965 }, { "epoch": 3.9129113924050634, "grad_norm": 0.2282982777755024, "learning_rate": 5.355518704628997e-06, "loss": 0.3159, "step": 966 }, { "epoch": 3.9169620253164554, "grad_norm": 0.19691040291020473, "learning_rate": 5.316921449248731e-06, "loss": 0.3071, "step": 967 }, { "epoch": 3.921012658227848, "grad_norm": 0.1982472658811993, "learning_rate": 5.278442449364538e-06, "loss": 0.3021, "step": 968 }, { "epoch": 3.9250632911392405, "grad_norm": 0.1956617647873405, "learning_rate": 5.240082014881016e-06, "loss": 0.3159, "step": 969 }, { "epoch": 3.929113924050633, "grad_norm": 0.1985865590783985, "learning_rate": 5.201840454747822e-06, "loss": 0.3064, "step": 970 }, { "epoch": 3.9331645569620255, "grad_norm": 0.1939051934066514, "learning_rate": 5.163718076957223e-06, "loss": 0.3059, "step": 971 }, { "epoch": 3.9372151898734176, "grad_norm": 0.1860258955026145, "learning_rate": 5.125715188541609e-06, "loss": 0.2932, "step": 972 }, { "epoch": 3.94126582278481, "grad_norm": 0.18774053818752298, "learning_rate": 5.087832095571021e-06, "loss": 0.3094, "step": 973 }, { "epoch": 3.9453164556962026, "grad_norm": 0.1945895812065499, "learning_rate": 5.0500691031506766e-06, "loss": 0.3052, "step": 974 }, { "epoch": 3.9493670886075947, "grad_norm": 0.20105351631871118, "learning_rate": 5.01242651541854e-06, "loss": 0.3166, "step": 975 }, { "epoch": 3.953417721518987, "grad_norm": 0.18869225172110773, "learning_rate": 4.974904635542815e-06, "loss": 0.3059, "step": 976 }, { "epoch": 3.9574683544303797, "grad_norm": 0.20525330735960662, "learning_rate": 4.937503765719582e-06, "loss": 0.3085, "step": 977 }, { "epoch": 3.961518987341772, "grad_norm": 0.1775555412831799, "learning_rate": 4.900224207170299e-06, "loss": 0.3031, "step": 978 }, { "epoch": 3.9655696202531647, "grad_norm": 0.1908043818108333, "learning_rate": 4.8630662601394065e-06, "loss": 0.3041, "step": 979 }, { "epoch": 3.969620253164557, "grad_norm": 0.19752299022837683, "learning_rate": 4.8260302238918995e-06, "loss": 0.2974, "step": 980 }, { "epoch": 3.9736708860759493, "grad_norm": 0.19595534529050687, "learning_rate": 4.789116396710924e-06, "loss": 0.3026, "step": 981 }, { "epoch": 3.977721518987342, "grad_norm": 0.18874155607487086, "learning_rate": 4.752325075895368e-06, "loss": 0.3108, "step": 982 }, { "epoch": 3.981772151898734, "grad_norm": 0.1977186258251516, "learning_rate": 4.715656557757473e-06, "loss": 0.3026, "step": 983 }, { "epoch": 3.9858227848101264, "grad_norm": 0.2032417475350747, "learning_rate": 4.679111137620442e-06, "loss": 0.3087, "step": 984 }, { "epoch": 3.989873417721519, "grad_norm": 0.18790509753521983, "learning_rate": 4.6426891098160585e-06, "loss": 0.3024, "step": 985 }, { "epoch": 3.9939240506329114, "grad_norm": 0.18242128978579458, "learning_rate": 4.6063907676823474e-06, "loss": 0.3075, "step": 986 }, { "epoch": 3.997974683544304, "grad_norm": 0.20487634558723675, "learning_rate": 4.570216403561141e-06, "loss": 0.3028, "step": 987 }, { "epoch": 4.002025316455696, "grad_norm": 0.2197135874024959, "learning_rate": 4.534166308795815e-06, "loss": 0.2918, "step": 988 }, { "epoch": 4.006075949367089, "grad_norm": 0.26998119764996853, "learning_rate": 4.498240773728859e-06, "loss": 0.2651, "step": 989 }, { "epoch": 4.010126582278481, "grad_norm": 0.23551450634177484, "learning_rate": 4.462440087699609e-06, "loss": 0.2664, "step": 990 }, { "epoch": 4.014177215189873, "grad_norm": 0.26430197206930167, "learning_rate": 4.426764539041861e-06, "loss": 0.2713, "step": 991 }, { "epoch": 4.018227848101266, "grad_norm": 0.30164114386189744, "learning_rate": 4.391214415081582e-06, "loss": 0.2727, "step": 992 }, { "epoch": 4.022278481012658, "grad_norm": 0.21979376861451252, "learning_rate": 4.355790002134579e-06, "loss": 0.2577, "step": 993 }, { "epoch": 4.02632911392405, "grad_norm": 0.23981956422761774, "learning_rate": 4.320491585504207e-06, "loss": 0.2776, "step": 994 }, { "epoch": 4.030379746835443, "grad_norm": 0.27686312887635195, "learning_rate": 4.2853194494790615e-06, "loss": 0.265, "step": 995 }, { "epoch": 4.034430379746835, "grad_norm": 0.23868756638020783, "learning_rate": 4.250273877330691e-06, "loss": 0.2591, "step": 996 }, { "epoch": 4.038481012658228, "grad_norm": 0.2152454855180847, "learning_rate": 4.215355151311313e-06, "loss": 0.267, "step": 997 }, { "epoch": 4.04253164556962, "grad_norm": 0.2318708255190135, "learning_rate": 4.180563552651542e-06, "loss": 0.259, "step": 998 }, { "epoch": 4.046582278481012, "grad_norm": 0.2538831113761317, "learning_rate": 4.145899361558147e-06, "loss": 0.2672, "step": 999 }, { "epoch": 4.050632911392405, "grad_norm": 0.21269381211923385, "learning_rate": 4.111362857211738e-06, "loss": 0.27, "step": 1000 }, { "epoch": 4.054683544303797, "grad_norm": 0.19218579443964878, "learning_rate": 4.076954317764592e-06, "loss": 0.2578, "step": 1001 }, { "epoch": 4.0587341772151895, "grad_norm": 0.20662028640326913, "learning_rate": 4.042674020338335e-06, "loss": 0.2588, "step": 1002 }, { "epoch": 4.062784810126582, "grad_norm": 0.22018943901794322, "learning_rate": 4.0085222410217835e-06, "loss": 0.2699, "step": 1003 }, { "epoch": 4.0668354430379745, "grad_norm": 0.18543646655043716, "learning_rate": 3.974499254868674e-06, "loss": 0.265, "step": 1004 }, { "epoch": 4.0708860759493675, "grad_norm": 0.19625353570086015, "learning_rate": 3.940605335895451e-06, "loss": 0.2772, "step": 1005 }, { "epoch": 4.0749367088607595, "grad_norm": 0.2050249677050791, "learning_rate": 3.90684075707908e-06, "loss": 0.2605, "step": 1006 }, { "epoch": 4.078987341772152, "grad_norm": 0.1939216533181128, "learning_rate": 3.8732057903548505e-06, "loss": 0.2601, "step": 1007 }, { "epoch": 4.083037974683545, "grad_norm": 0.20239926202789593, "learning_rate": 3.8397007066141375e-06, "loss": 0.2667, "step": 1008 }, { "epoch": 4.087088607594937, "grad_norm": 0.19638027829224136, "learning_rate": 3.806325775702304e-06, "loss": 0.2671, "step": 1009 }, { "epoch": 4.091139240506329, "grad_norm": 0.1920873952842713, "learning_rate": 3.773081266416434e-06, "loss": 0.2613, "step": 1010 }, { "epoch": 4.095189873417722, "grad_norm": 0.19703806702388463, "learning_rate": 3.739967446503245e-06, "loss": 0.273, "step": 1011 }, { "epoch": 4.099240506329114, "grad_norm": 0.1959734007562541, "learning_rate": 3.706984582656894e-06, "loss": 0.2589, "step": 1012 }, { "epoch": 4.103291139240507, "grad_norm": 0.20149788294634022, "learning_rate": 3.6741329405168237e-06, "loss": 0.2688, "step": 1013 }, { "epoch": 4.107341772151899, "grad_norm": 0.18645967097467214, "learning_rate": 3.641412784665648e-06, "loss": 0.2702, "step": 1014 }, { "epoch": 4.111392405063291, "grad_norm": 0.18738173354318136, "learning_rate": 3.608824378627005e-06, "loss": 0.2721, "step": 1015 }, { "epoch": 4.115443037974684, "grad_norm": 0.1818975820474809, "learning_rate": 3.5763679848634337e-06, "loss": 0.268, "step": 1016 }, { "epoch": 4.119493670886076, "grad_norm": 0.1816464093813645, "learning_rate": 3.544043864774269e-06, "loss": 0.2747, "step": 1017 }, { "epoch": 4.123544303797468, "grad_norm": 0.19065557249183065, "learning_rate": 3.5118522786935282e-06, "loss": 0.2758, "step": 1018 }, { "epoch": 4.127594936708861, "grad_norm": 0.18705401075135755, "learning_rate": 3.479793485887819e-06, "loss": 0.2689, "step": 1019 }, { "epoch": 4.131645569620253, "grad_norm": 0.18913089810198772, "learning_rate": 3.4478677445542653e-06, "loss": 0.2803, "step": 1020 }, { "epoch": 4.135696202531646, "grad_norm": 0.1838776962158285, "learning_rate": 3.4160753118183767e-06, "loss": 0.2708, "step": 1021 }, { "epoch": 4.139746835443038, "grad_norm": 0.19746188159931838, "learning_rate": 3.3844164437320527e-06, "loss": 0.2736, "step": 1022 }, { "epoch": 4.14379746835443, "grad_norm": 0.1883160418671491, "learning_rate": 3.3528913952714558e-06, "loss": 0.2719, "step": 1023 }, { "epoch": 4.147848101265823, "grad_norm": 0.1800624797525394, "learning_rate": 3.321500420335e-06, "loss": 0.274, "step": 1024 }, { "epoch": 4.151898734177215, "grad_norm": 0.19326954377535563, "learning_rate": 3.290243771741275e-06, "loss": 0.2769, "step": 1025 }, { "epoch": 4.155949367088607, "grad_norm": 0.18831625009697306, "learning_rate": 3.2591217012270325e-06, "loss": 0.2646, "step": 1026 }, { "epoch": 4.16, "grad_norm": 0.1771271081425216, "learning_rate": 3.228134459445149e-06, "loss": 0.2787, "step": 1027 }, { "epoch": 4.164050632911392, "grad_norm": 0.18121441483186942, "learning_rate": 3.1972822959626205e-06, "loss": 0.2659, "step": 1028 }, { "epoch": 4.168101265822785, "grad_norm": 0.18082044349191392, "learning_rate": 3.166565459258513e-06, "loss": 0.2695, "step": 1029 }, { "epoch": 4.172151898734177, "grad_norm": 0.1774983090237879, "learning_rate": 3.1359841967220193e-06, "loss": 0.2689, "step": 1030 }, { "epoch": 4.176202531645569, "grad_norm": 0.18011718461616597, "learning_rate": 3.105538754650419e-06, "loss": 0.2716, "step": 1031 }, { "epoch": 4.180253164556962, "grad_norm": 0.18629865944541546, "learning_rate": 3.07522937824712e-06, "loss": 0.2458, "step": 1032 }, { "epoch": 4.184303797468354, "grad_norm": 0.18591050647341809, "learning_rate": 3.0450563116196697e-06, "loss": 0.2842, "step": 1033 }, { "epoch": 4.188354430379746, "grad_norm": 0.18480037167727084, "learning_rate": 3.0150197977778008e-06, "loss": 0.2781, "step": 1034 }, { "epoch": 4.192405063291139, "grad_norm": 0.16941481123951876, "learning_rate": 2.985120078631465e-06, "loss": 0.2787, "step": 1035 }, { "epoch": 4.1964556962025314, "grad_norm": 0.1830219626905376, "learning_rate": 2.9553573949888893e-06, "loss": 0.2734, "step": 1036 }, { "epoch": 4.200506329113924, "grad_norm": 0.19113288970688128, "learning_rate": 2.9257319865546384e-06, "loss": 0.281, "step": 1037 }, { "epoch": 4.2045569620253165, "grad_norm": 0.18632079184832898, "learning_rate": 2.896244091927678e-06, "loss": 0.2749, "step": 1038 }, { "epoch": 4.2086075949367086, "grad_norm": 0.1793382501156949, "learning_rate": 2.8668939485994584e-06, "loss": 0.2678, "step": 1039 }, { "epoch": 4.2126582278481015, "grad_norm": 0.17510347093281162, "learning_rate": 2.837681792951994e-06, "loss": 0.2687, "step": 1040 }, { "epoch": 4.216708860759494, "grad_norm": 0.1680930563168772, "learning_rate": 2.808607860255981e-06, "loss": 0.2684, "step": 1041 }, { "epoch": 4.220759493670886, "grad_norm": 0.1889396771441588, "learning_rate": 2.7796723846688634e-06, "loss": 0.2653, "step": 1042 }, { "epoch": 4.224810126582279, "grad_norm": 0.18255915928768243, "learning_rate": 2.7508755992329937e-06, "loss": 0.2752, "step": 1043 }, { "epoch": 4.228860759493671, "grad_norm": 0.1775785008335017, "learning_rate": 2.722217735873718e-06, "loss": 0.2805, "step": 1044 }, { "epoch": 4.232911392405064, "grad_norm": 0.171331647736653, "learning_rate": 2.6936990253975315e-06, "loss": 0.2704, "step": 1045 }, { "epoch": 4.236962025316456, "grad_norm": 0.18664641811747368, "learning_rate": 2.665319697490205e-06, "loss": 0.2681, "step": 1046 }, { "epoch": 4.241012658227848, "grad_norm": 0.17032052998807604, "learning_rate": 2.637079980714945e-06, "loss": 0.2753, "step": 1047 }, { "epoch": 4.245063291139241, "grad_norm": 0.17567117645062066, "learning_rate": 2.6089801025105453e-06, "loss": 0.2752, "step": 1048 }, { "epoch": 4.249113924050633, "grad_norm": 0.17540641559015155, "learning_rate": 2.581020289189571e-06, "loss": 0.2617, "step": 1049 }, { "epoch": 4.253164556962025, "grad_norm": 0.22306399337504093, "learning_rate": 2.553200765936501e-06, "loss": 0.2766, "step": 1050 }, { "epoch": 4.257215189873418, "grad_norm": 0.17653268712657524, "learning_rate": 2.525521756805962e-06, "loss": 0.269, "step": 1051 }, { "epoch": 4.26126582278481, "grad_norm": 0.17921862585381307, "learning_rate": 2.497983484720885e-06, "loss": 0.2562, "step": 1052 }, { "epoch": 4.265316455696203, "grad_norm": 0.17689358521449663, "learning_rate": 2.470586171470728e-06, "loss": 0.2727, "step": 1053 }, { "epoch": 4.269367088607595, "grad_norm": 0.1705856609607582, "learning_rate": 2.4433300377096836e-06, "loss": 0.271, "step": 1054 }, { "epoch": 4.273417721518987, "grad_norm": 0.17217523362217174, "learning_rate": 2.4162153029549073e-06, "loss": 0.2736, "step": 1055 }, { "epoch": 4.27746835443038, "grad_norm": 0.18129482579513126, "learning_rate": 2.3892421855847458e-06, "loss": 0.2617, "step": 1056 }, { "epoch": 4.281518987341772, "grad_norm": 0.20437930384659564, "learning_rate": 2.362410902836978e-06, "loss": 0.267, "step": 1057 }, { "epoch": 4.285569620253165, "grad_norm": 0.19049021126245683, "learning_rate": 2.3357216708070653e-06, "loss": 0.2676, "step": 1058 }, { "epoch": 4.289620253164557, "grad_norm": 0.16897439623162117, "learning_rate": 2.309174704446411e-06, "loss": 0.2791, "step": 1059 }, { "epoch": 4.293670886075949, "grad_norm": 0.17750778279985455, "learning_rate": 2.2827702175606437e-06, "loss": 0.2831, "step": 1060 }, { "epoch": 4.297721518987342, "grad_norm": 0.1817245040483692, "learning_rate": 2.256508422807855e-06, "loss": 0.278, "step": 1061 }, { "epoch": 4.301772151898734, "grad_norm": 0.17393606347811322, "learning_rate": 2.230389531696946e-06, "loss": 0.2616, "step": 1062 }, { "epoch": 4.305822784810126, "grad_norm": 0.18126370110124387, "learning_rate": 2.204413754585857e-06, "loss": 0.2678, "step": 1063 }, { "epoch": 4.309873417721519, "grad_norm": 0.17404353905691528, "learning_rate": 2.1785813006799406e-06, "loss": 0.2695, "step": 1064 }, { "epoch": 4.313924050632911, "grad_norm": 0.18467699264211893, "learning_rate": 2.1528923780302224e-06, "loss": 0.2596, "step": 1065 }, { "epoch": 4.317974683544303, "grad_norm": 0.18376513342447745, "learning_rate": 2.127347193531757e-06, "loss": 0.2641, "step": 1066 }, { "epoch": 4.322025316455696, "grad_norm": 0.20265079788686177, "learning_rate": 2.101945952921942e-06, "loss": 0.2616, "step": 1067 }, { "epoch": 4.326075949367088, "grad_norm": 0.1816440441360374, "learning_rate": 2.0766888607788906e-06, "loss": 0.2603, "step": 1068 }, { "epoch": 4.330126582278481, "grad_norm": 0.1778358289297519, "learning_rate": 2.0515761205197337e-06, "loss": 0.2725, "step": 1069 }, { "epoch": 4.334177215189873, "grad_norm": 0.17325034752138313, "learning_rate": 2.0266079343990453e-06, "loss": 0.2648, "step": 1070 }, { "epoch": 4.3382278481012655, "grad_norm": 0.1782602905506633, "learning_rate": 2.0017845035071494e-06, "loss": 0.2652, "step": 1071 }, { "epoch": 4.3422784810126585, "grad_norm": 0.18748739663053945, "learning_rate": 1.9771060277685537e-06, "loss": 0.2541, "step": 1072 }, { "epoch": 4.3463291139240505, "grad_norm": 0.18225395134088956, "learning_rate": 1.95257270594031e-06, "loss": 0.2672, "step": 1073 }, { "epoch": 4.3503797468354435, "grad_norm": 0.17686481073871677, "learning_rate": 1.9281847356104188e-06, "loss": 0.266, "step": 1074 }, { "epoch": 4.3544303797468356, "grad_norm": 0.17314855217946823, "learning_rate": 1.9039423131962365e-06, "loss": 0.2761, "step": 1075 }, { "epoch": 4.358481012658228, "grad_norm": 0.17212231886866583, "learning_rate": 1.8798456339429027e-06, "loss": 0.259, "step": 1076 }, { "epoch": 4.362531645569621, "grad_norm": 0.18221529715048582, "learning_rate": 1.8558948919217612e-06, "loss": 0.2749, "step": 1077 }, { "epoch": 4.366582278481013, "grad_norm": 0.17313675591972028, "learning_rate": 1.8320902800287954e-06, "loss": 0.2668, "step": 1078 }, { "epoch": 4.370632911392405, "grad_norm": 0.17916016836156506, "learning_rate": 1.8084319899830726e-06, "loss": 0.2666, "step": 1079 }, { "epoch": 4.374683544303798, "grad_norm": 0.18637237540104587, "learning_rate": 1.7849202123252097e-06, "loss": 0.2615, "step": 1080 }, { "epoch": 4.37873417721519, "grad_norm": 0.18676073021266595, "learning_rate": 1.7615551364158401e-06, "loss": 0.2742, "step": 1081 }, { "epoch": 4.382784810126582, "grad_norm": 0.18415476368031175, "learning_rate": 1.738336950434061e-06, "loss": 0.2712, "step": 1082 }, { "epoch": 4.386835443037975, "grad_norm": 0.1729468617830621, "learning_rate": 1.715265841375957e-06, "loss": 0.2559, "step": 1083 }, { "epoch": 4.390886075949367, "grad_norm": 0.2115998900853037, "learning_rate": 1.6923419950530684e-06, "loss": 0.258, "step": 1084 }, { "epoch": 4.39493670886076, "grad_norm": 0.18232698983592716, "learning_rate": 1.6695655960909008e-06, "loss": 0.2711, "step": 1085 }, { "epoch": 4.398987341772152, "grad_norm": 0.18243767596877183, "learning_rate": 1.646936827927441e-06, "loss": 0.2736, "step": 1086 }, { "epoch": 4.403037974683544, "grad_norm": 0.17107199300582066, "learning_rate": 1.6244558728116766e-06, "loss": 0.2664, "step": 1087 }, { "epoch": 4.407088607594937, "grad_norm": 0.1683506039094082, "learning_rate": 1.6021229118021265e-06, "loss": 0.2632, "step": 1088 }, { "epoch": 4.411139240506329, "grad_norm": 0.18094564358258364, "learning_rate": 1.5799381247653967e-06, "loss": 0.2689, "step": 1089 }, { "epoch": 4.415189873417722, "grad_norm": 0.18742339742548336, "learning_rate": 1.5579016903747013e-06, "loss": 0.2684, "step": 1090 }, { "epoch": 4.419240506329114, "grad_norm": 0.17860853742745825, "learning_rate": 1.5360137861084656e-06, "loss": 0.2653, "step": 1091 }, { "epoch": 4.423291139240506, "grad_norm": 0.17387045587740899, "learning_rate": 1.5142745882488475e-06, "loss": 0.2666, "step": 1092 }, { "epoch": 4.427341772151899, "grad_norm": 0.1761974578764661, "learning_rate": 1.4926842718803691e-06, "loss": 0.2732, "step": 1093 }, { "epoch": 4.431392405063291, "grad_norm": 0.17010419800986654, "learning_rate": 1.4712430108884657e-06, "loss": 0.2672, "step": 1094 }, { "epoch": 4.435443037974683, "grad_norm": 0.17321156488473297, "learning_rate": 1.4499509779581078e-06, "loss": 0.2759, "step": 1095 }, { "epoch": 4.439493670886076, "grad_norm": 0.16830274877271403, "learning_rate": 1.4288083445723988e-06, "loss": 0.2649, "step": 1096 }, { "epoch": 4.443544303797468, "grad_norm": 0.18360245285024313, "learning_rate": 1.4078152810112045e-06, "loss": 0.2698, "step": 1097 }, { "epoch": 4.44759493670886, "grad_norm": 0.16796318075200037, "learning_rate": 1.3869719563497697e-06, "loss": 0.2751, "step": 1098 }, { "epoch": 4.451645569620253, "grad_norm": 0.19242135315578027, "learning_rate": 1.3662785384573663e-06, "loss": 0.2805, "step": 1099 }, { "epoch": 4.455696202531645, "grad_norm": 0.18041591391070408, "learning_rate": 1.3457351939959383e-06, "loss": 0.2714, "step": 1100 }, { "epoch": 4.459746835443038, "grad_norm": 0.17112323974191648, "learning_rate": 1.3253420884187551e-06, "loss": 0.2776, "step": 1101 }, { "epoch": 4.46379746835443, "grad_norm": 0.17639459436033353, "learning_rate": 1.3050993859690953e-06, "loss": 0.2719, "step": 1102 }, { "epoch": 4.467848101265822, "grad_norm": 0.1781091703607025, "learning_rate": 1.2850072496788869e-06, "loss": 0.2723, "step": 1103 }, { "epoch": 4.471898734177215, "grad_norm": 0.1777506356023915, "learning_rate": 1.2650658413674434e-06, "loss": 0.2672, "step": 1104 }, { "epoch": 4.4759493670886075, "grad_norm": 0.16720171781597362, "learning_rate": 1.2452753216401226e-06, "loss": 0.2763, "step": 1105 }, { "epoch": 4.48, "grad_norm": 0.1869414075459926, "learning_rate": 1.2256358498870503e-06, "loss": 0.2626, "step": 1106 }, { "epoch": 4.4840506329113925, "grad_norm": 0.20551535783092276, "learning_rate": 1.2061475842818337e-06, "loss": 0.266, "step": 1107 }, { "epoch": 4.488101265822785, "grad_norm": 0.17613463899857787, "learning_rate": 1.1868106817802816e-06, "loss": 0.2707, "step": 1108 }, { "epoch": 4.4921518987341775, "grad_norm": 0.17265242741481296, "learning_rate": 1.1676252981191482e-06, "loss": 0.2683, "step": 1109 }, { "epoch": 4.49620253164557, "grad_norm": 0.17569542163548574, "learning_rate": 1.1485915878148823e-06, "loss": 0.2709, "step": 1110 }, { "epoch": 4.500253164556962, "grad_norm": 0.17450677915794194, "learning_rate": 1.1297097041623584e-06, "loss": 0.2723, "step": 1111 }, { "epoch": 4.504303797468355, "grad_norm": 0.17078902794938555, "learning_rate": 1.1109797992336847e-06, "loss": 0.267, "step": 1112 }, { "epoch": 4.508354430379747, "grad_norm": 0.1748465638753489, "learning_rate": 1.092402023876933e-06, "loss": 0.2601, "step": 1113 }, { "epoch": 4.512405063291139, "grad_norm": 0.17772988633345374, "learning_rate": 1.0739765277149527e-06, "loss": 0.2667, "step": 1114 }, { "epoch": 4.516455696202532, "grad_norm": 0.17211181616752075, "learning_rate": 1.0557034591441596e-06, "loss": 0.2742, "step": 1115 }, { "epoch": 4.520506329113924, "grad_norm": 0.171296723824673, "learning_rate": 1.0375829653333324e-06, "loss": 0.2717, "step": 1116 }, { "epoch": 4.524556962025317, "grad_norm": 0.1743397508848111, "learning_rate": 1.0196151922224385e-06, "loss": 0.2678, "step": 1117 }, { "epoch": 4.528607594936709, "grad_norm": 0.18974519986586924, "learning_rate": 1.0018002845214526e-06, "loss": 0.2692, "step": 1118 }, { "epoch": 4.532658227848101, "grad_norm": 0.18068062622518818, "learning_rate": 9.841383857091947e-07, "loss": 0.2712, "step": 1119 }, { "epoch": 4.536708860759494, "grad_norm": 0.18926123459564256, "learning_rate": 9.666296380321616e-07, "loss": 0.2776, "step": 1120 }, { "epoch": 4.540759493670886, "grad_norm": 0.1671148401311281, "learning_rate": 9.492741825034124e-07, "loss": 0.2682, "step": 1121 }, { "epoch": 4.544810126582279, "grad_norm": 0.1746402532951112, "learning_rate": 9.320721589013892e-07, "loss": 0.277, "step": 1122 }, { "epoch": 4.548860759493671, "grad_norm": 0.17295673521205726, "learning_rate": 9.150237057688339e-07, "loss": 0.2765, "step": 1123 }, { "epoch": 4.552911392405063, "grad_norm": 0.17295658181095241, "learning_rate": 8.981289604116328e-07, "loss": 0.2707, "step": 1124 }, { "epoch": 4.556962025316456, "grad_norm": 0.17082165364927318, "learning_rate": 8.813880588977542e-07, "loss": 0.2737, "step": 1125 }, { "epoch": 4.561012658227848, "grad_norm": 0.1709990658088844, "learning_rate": 8.648011360561126e-07, "loss": 0.276, "step": 1126 }, { "epoch": 4.56506329113924, "grad_norm": 0.17647854522071516, "learning_rate": 8.483683254755037e-07, "loss": 0.271, "step": 1127 }, { "epoch": 4.569113924050633, "grad_norm": 0.17318279560922056, "learning_rate": 8.320897595035227e-07, "loss": 0.2745, "step": 1128 }, { "epoch": 4.573164556962025, "grad_norm": 0.17236089659467393, "learning_rate": 8.159655692455093e-07, "loss": 0.2764, "step": 1129 }, { "epoch": 4.577215189873417, "grad_norm": 0.17434029558979885, "learning_rate": 7.999958845634648e-07, "loss": 0.2706, "step": 1130 }, { "epoch": 4.58126582278481, "grad_norm": 0.1680638260464098, "learning_rate": 7.841808340750478e-07, "loss": 0.265, "step": 1131 }, { "epoch": 4.585316455696202, "grad_norm": 0.16802969111783295, "learning_rate": 7.685205451524869e-07, "loss": 0.257, "step": 1132 }, { "epoch": 4.589367088607595, "grad_norm": 0.17864899129789008, "learning_rate": 7.530151439216027e-07, "loss": 0.2706, "step": 1133 }, { "epoch": 4.593417721518987, "grad_norm": 0.17019090868637074, "learning_rate": 7.376647552607675e-07, "loss": 0.2682, "step": 1134 }, { "epoch": 4.597468354430379, "grad_norm": 0.17299393114803316, "learning_rate": 7.224695027998963e-07, "loss": 0.263, "step": 1135 }, { "epoch": 4.601518987341772, "grad_norm": 0.17493009249588512, "learning_rate": 7.07429508919466e-07, "loss": 0.2688, "step": 1136 }, { "epoch": 4.605569620253164, "grad_norm": 0.18398141680083396, "learning_rate": 6.925448947495206e-07, "loss": 0.2736, "step": 1137 }, { "epoch": 4.609620253164557, "grad_norm": 0.16931723397699974, "learning_rate": 6.778157801686936e-07, "loss": 0.2774, "step": 1138 }, { "epoch": 4.613670886075949, "grad_norm": 0.16561521007118757, "learning_rate": 6.632422838032515e-07, "loss": 0.2643, "step": 1139 }, { "epoch": 4.6177215189873415, "grad_norm": 0.18038177308902992, "learning_rate": 6.488245230261281e-07, "loss": 0.2652, "step": 1140 }, { "epoch": 4.6217721518987345, "grad_norm": 0.17339381548766777, "learning_rate": 6.345626139559868e-07, "loss": 0.2645, "step": 1141 }, { "epoch": 4.6258227848101265, "grad_norm": 0.1733745775939773, "learning_rate": 6.204566714562866e-07, "loss": 0.276, "step": 1142 }, { "epoch": 4.629873417721519, "grad_norm": 0.16848400731086355, "learning_rate": 6.06506809134344e-07, "loss": 0.2728, "step": 1143 }, { "epoch": 4.633924050632912, "grad_norm": 0.18452488260942448, "learning_rate": 5.927131393404373e-07, "loss": 0.2564, "step": 1144 }, { "epoch": 4.637974683544304, "grad_norm": 0.17617705979615994, "learning_rate": 5.790757731668817e-07, "loss": 0.2628, "step": 1145 }, { "epoch": 4.642025316455696, "grad_norm": 0.16580358563126654, "learning_rate": 5.655948204471507e-07, "loss": 0.2754, "step": 1146 }, { "epoch": 4.646075949367089, "grad_norm": 0.1702224904697736, "learning_rate": 5.522703897549875e-07, "loss": 0.2744, "step": 1147 }, { "epoch": 4.650126582278481, "grad_norm": 0.1762561885029394, "learning_rate": 5.391025884035239e-07, "loss": 0.2665, "step": 1148 }, { "epoch": 4.654177215189874, "grad_norm": 0.1681683689370741, "learning_rate": 5.260915224444207e-07, "loss": 0.2703, "step": 1149 }, { "epoch": 4.658227848101266, "grad_norm": 0.1711832096798835, "learning_rate": 5.132372966670129e-07, "loss": 0.275, "step": 1150 }, { "epoch": 4.662278481012658, "grad_norm": 0.1657260991072199, "learning_rate": 5.005400145974704e-07, "loss": 0.2702, "step": 1151 }, { "epoch": 4.666329113924051, "grad_norm": 0.1760058474291978, "learning_rate": 4.879997784979562e-07, "loss": 0.2695, "step": 1152 }, { "epoch": 4.670379746835443, "grad_norm": 0.17670824458736725, "learning_rate": 4.7561668936580984e-07, "loss": 0.2675, "step": 1153 }, { "epoch": 4.674430379746836, "grad_norm": 0.1654578311709247, "learning_rate": 4.6339084693272306e-07, "loss": 0.2653, "step": 1154 }, { "epoch": 4.678481012658228, "grad_norm": 0.16825726457804732, "learning_rate": 4.5132234966395847e-07, "loss": 0.2737, "step": 1155 }, { "epoch": 4.68253164556962, "grad_norm": 0.19694164045532347, "learning_rate": 4.3941129475752795e-07, "loss": 0.2677, "step": 1156 }, { "epoch": 4.686582278481013, "grad_norm": 0.16612390193338897, "learning_rate": 4.27657778143431e-07, "loss": 0.2683, "step": 1157 }, { "epoch": 4.690632911392405, "grad_norm": 0.17444067271128483, "learning_rate": 4.1606189448287757e-07, "loss": 0.2676, "step": 1158 }, { "epoch": 4.694683544303797, "grad_norm": 0.17541869943524732, "learning_rate": 4.046237371675177e-07, "loss": 0.2587, "step": 1159 }, { "epoch": 4.69873417721519, "grad_norm": 0.1802668188171414, "learning_rate": 3.9334339831869963e-07, "loss": 0.2574, "step": 1160 }, { "epoch": 4.702784810126582, "grad_norm": 0.17312974701967004, "learning_rate": 3.8222096878671955e-07, "loss": 0.2606, "step": 1161 }, { "epoch": 4.706835443037974, "grad_norm": 0.17571450596199187, "learning_rate": 3.7125653815009545e-07, "loss": 0.2622, "step": 1162 }, { "epoch": 4.710886075949367, "grad_norm": 0.1700879634168538, "learning_rate": 3.6045019471484974e-07, "loss": 0.2744, "step": 1163 }, { "epoch": 4.714936708860759, "grad_norm": 0.1703242473890918, "learning_rate": 3.498020255137813e-07, "loss": 0.2739, "step": 1164 }, { "epoch": 4.718987341772152, "grad_norm": 0.17323957538340023, "learning_rate": 3.393121163057811e-07, "loss": 0.267, "step": 1165 }, { "epoch": 4.723037974683544, "grad_norm": 0.17104266531073326, "learning_rate": 3.289805515751399e-07, "loss": 0.273, "step": 1166 }, { "epoch": 4.727088607594936, "grad_norm": 0.16538956372501468, "learning_rate": 3.188074145308573e-07, "loss": 0.2689, "step": 1167 }, { "epoch": 4.731139240506329, "grad_norm": 0.168150120646448, "learning_rate": 3.087927871059804e-07, "loss": 0.2629, "step": 1168 }, { "epoch": 4.735189873417721, "grad_norm": 0.17369081704693906, "learning_rate": 2.989367499569418e-07, "loss": 0.276, "step": 1169 }, { "epoch": 4.739240506329114, "grad_norm": 0.1655742539967124, "learning_rate": 2.8923938246290917e-07, "loss": 0.2594, "step": 1170 }, { "epoch": 4.743291139240506, "grad_norm": 0.16427788181180378, "learning_rate": 2.7970076272514804e-07, "loss": 0.2755, "step": 1171 }, { "epoch": 4.7473417721518985, "grad_norm": 0.16818559476036674, "learning_rate": 2.703209675663887e-07, "loss": 0.2721, "step": 1172 }, { "epoch": 4.751392405063291, "grad_norm": 0.1673132809447521, "learning_rate": 2.6110007253021374e-07, "loss": 0.264, "step": 1173 }, { "epoch": 4.7554430379746835, "grad_norm": 0.16942499831388239, "learning_rate": 2.520381518804471e-07, "loss": 0.2641, "step": 1174 }, { "epoch": 4.759493670886076, "grad_norm": 0.16408817515892846, "learning_rate": 2.4313527860054585e-07, "loss": 0.2549, "step": 1175 }, { "epoch": 4.7635443037974685, "grad_norm": 0.17740851587354625, "learning_rate": 2.343915243930317e-07, "loss": 0.2622, "step": 1176 }, { "epoch": 4.767594936708861, "grad_norm": 0.16296928266439767, "learning_rate": 2.2580695967889367e-07, "loss": 0.2478, "step": 1177 }, { "epoch": 4.7716455696202535, "grad_norm": 0.16958026661624437, "learning_rate": 2.1738165359704189e-07, "loss": 0.264, "step": 1178 }, { "epoch": 4.775696202531646, "grad_norm": 0.16627589208770494, "learning_rate": 2.0911567400373257e-07, "loss": 0.2747, "step": 1179 }, { "epoch": 4.779746835443038, "grad_norm": 0.1705230947248207, "learning_rate": 2.0100908747202607e-07, "loss": 0.265, "step": 1180 }, { "epoch": 4.783797468354431, "grad_norm": 0.16427781564825272, "learning_rate": 1.9306195929125638e-07, "loss": 0.2828, "step": 1181 }, { "epoch": 4.787848101265823, "grad_norm": 0.16521711093846808, "learning_rate": 1.8527435346650247e-07, "loss": 0.2613, "step": 1182 }, { "epoch": 4.791898734177215, "grad_norm": 0.17014496786505212, "learning_rate": 1.7764633271807108e-07, "loss": 0.2793, "step": 1183 }, { "epoch": 4.795949367088608, "grad_norm": 0.16590831522178537, "learning_rate": 1.7017795848099262e-07, "loss": 0.265, "step": 1184 }, { "epoch": 4.8, "grad_norm": 0.1719773875067652, "learning_rate": 1.6286929090452596e-07, "loss": 0.2673, "step": 1185 }, { "epoch": 4.804050632911393, "grad_norm": 0.17835494945141514, "learning_rate": 1.557203888516745e-07, "loss": 0.2761, "step": 1186 }, { "epoch": 4.808101265822785, "grad_norm": 0.17021082134538537, "learning_rate": 1.487313098987131e-07, "loss": 0.2563, "step": 1187 }, { "epoch": 4.812151898734177, "grad_norm": 0.16522446494350745, "learning_rate": 1.4190211033472402e-07, "loss": 0.2748, "step": 1188 }, { "epoch": 4.81620253164557, "grad_norm": 0.16878322831006776, "learning_rate": 1.3523284516113955e-07, "loss": 0.273, "step": 1189 }, { "epoch": 4.820253164556962, "grad_norm": 0.1695396914987955, "learning_rate": 1.2872356809130682e-07, "loss": 0.2685, "step": 1190 }, { "epoch": 4.824303797468355, "grad_norm": 0.1678245985277798, "learning_rate": 1.2237433155004807e-07, "loss": 0.2641, "step": 1191 }, { "epoch": 4.828354430379747, "grad_norm": 0.16932655422450385, "learning_rate": 1.1618518667323886e-07, "loss": 0.2739, "step": 1192 }, { "epoch": 4.832405063291139, "grad_norm": 0.17330214974855607, "learning_rate": 1.1015618330740385e-07, "loss": 0.2746, "step": 1193 }, { "epoch": 4.836455696202532, "grad_norm": 0.16507633998769908, "learning_rate": 1.042873700093061e-07, "loss": 0.2803, "step": 1194 }, { "epoch": 4.840506329113924, "grad_norm": 0.18588628460557186, "learning_rate": 9.857879404556291e-08, "loss": 0.2736, "step": 1195 }, { "epoch": 4.844556962025316, "grad_norm": 0.1811062981177184, "learning_rate": 9.303050139225722e-08, "loss": 0.2584, "step": 1196 }, { "epoch": 4.848607594936709, "grad_norm": 0.17159252026821128, "learning_rate": 8.76425367345779e-08, "loss": 0.2675, "step": 1197 }, { "epoch": 4.852658227848101, "grad_norm": 0.17164950367894455, "learning_rate": 8.241494346644897e-08, "loss": 0.2706, "step": 1198 }, { "epoch": 4.856708860759493, "grad_norm": 0.17261624068721043, "learning_rate": 7.734776369019204e-08, "loss": 0.2564, "step": 1199 }, { "epoch": 4.860759493670886, "grad_norm": 0.18184296020151883, "learning_rate": 7.244103821617332e-08, "loss": 0.2632, "step": 1200 }, { "epoch": 4.864810126582278, "grad_norm": 0.1672072885267681, "learning_rate": 6.769480656248606e-08, "loss": 0.2636, "step": 1201 }, { "epoch": 4.868860759493671, "grad_norm": 0.16751013925340177, "learning_rate": 6.310910695462635e-08, "loss": 0.2816, "step": 1202 }, { "epoch": 4.872911392405063, "grad_norm": 0.20348247433962832, "learning_rate": 5.8683976325191185e-08, "loss": 0.2611, "step": 1203 }, { "epoch": 4.876962025316455, "grad_norm": 0.16456470633305129, "learning_rate": 5.4419450313571984e-08, "loss": 0.2699, "step": 1204 }, { "epoch": 4.881012658227848, "grad_norm": 0.16827065879924596, "learning_rate": 5.031556326567488e-08, "loss": 0.2704, "step": 1205 }, { "epoch": 4.88506329113924, "grad_norm": 0.17438225122228174, "learning_rate": 4.637234823364312e-08, "loss": 0.2778, "step": 1206 }, { "epoch": 4.889113924050633, "grad_norm": 0.17105327267112502, "learning_rate": 4.258983697558838e-08, "loss": 0.2566, "step": 1207 }, { "epoch": 4.8931645569620255, "grad_norm": 0.16997965068395682, "learning_rate": 3.896805995533548e-08, "loss": 0.2775, "step": 1208 }, { "epoch": 4.8972151898734175, "grad_norm": 0.17200095430423018, "learning_rate": 3.550704634218028e-08, "loss": 0.2756, "step": 1209 }, { "epoch": 4.9012658227848105, "grad_norm": 0.16958938351519343, "learning_rate": 3.2206824010647676e-08, "loss": 0.2654, "step": 1210 }, { "epoch": 4.905316455696203, "grad_norm": 0.16871044167499089, "learning_rate": 2.9067419540278476e-08, "loss": 0.2774, "step": 1211 }, { "epoch": 4.909367088607595, "grad_norm": 0.1733790324684269, "learning_rate": 2.6088858215400638e-08, "loss": 0.2744, "step": 1212 }, { "epoch": 4.913417721518988, "grad_norm": 0.17138957301660562, "learning_rate": 2.3271164024940564e-08, "loss": 0.2712, "step": 1213 }, { "epoch": 4.91746835443038, "grad_norm": 0.17486708269093773, "learning_rate": 2.061435966221881e-08, "loss": 0.2701, "step": 1214 }, { "epoch": 4.921518987341772, "grad_norm": 0.16904815017266434, "learning_rate": 1.811846652477245e-08, "loss": 0.2705, "step": 1215 }, { "epoch": 4.925569620253165, "grad_norm": 0.17002487951286704, "learning_rate": 1.5783504714184106e-08, "loss": 0.2696, "step": 1216 }, { "epoch": 4.929620253164557, "grad_norm": 0.17239626099303695, "learning_rate": 1.360949303591097e-08, "loss": 0.2583, "step": 1217 }, { "epoch": 4.93367088607595, "grad_norm": 0.16653796590121814, "learning_rate": 1.1596448999144916e-08, "loss": 0.2783, "step": 1218 }, { "epoch": 4.937721518987342, "grad_norm": 0.17000859479623293, "learning_rate": 9.744388816668172e-09, "loss": 0.2644, "step": 1219 }, { "epoch": 4.941772151898734, "grad_norm": 0.16557646520952873, "learning_rate": 8.05332740472009e-09, "loss": 0.2805, "step": 1220 }, { "epoch": 4.945822784810127, "grad_norm": 0.1638077479862759, "learning_rate": 6.523278382872811e-09, "loss": 0.2738, "step": 1221 }, { "epoch": 4.949873417721519, "grad_norm": 0.17264230236657646, "learning_rate": 5.15425407393133e-09, "loss": 0.2698, "step": 1222 }, { "epoch": 4.953924050632912, "grad_norm": 0.16458612136634912, "learning_rate": 3.94626550383137e-09, "loss": 0.2734, "step": 1223 }, { "epoch": 4.957974683544304, "grad_norm": 0.16539355268938205, "learning_rate": 2.899322401546112e-09, "loss": 0.2667, "step": 1224 }, { "epoch": 4.962025316455696, "grad_norm": 0.16846458465575848, "learning_rate": 2.013433199010706e-09, "loss": 0.2673, "step": 1225 }, { "epoch": 4.966075949367089, "grad_norm": 0.17153120539828676, "learning_rate": 1.2886050310556563e-09, "loss": 0.2825, "step": 1226 }, { "epoch": 4.970126582278481, "grad_norm": 0.16802106931242636, "learning_rate": 7.248437353468695e-10, "loss": 0.264, "step": 1227 }, { "epoch": 4.974177215189873, "grad_norm": 0.16968282691221612, "learning_rate": 3.221538523412449e-10, "loss": 0.2699, "step": 1228 }, { "epoch": 4.978227848101266, "grad_norm": 0.1689982641945008, "learning_rate": 8.053862524670663e-11, "loss": 0.2763, "step": 1229 }, { "epoch": 4.982278481012658, "grad_norm": 0.16438629550097103, "learning_rate": 0.0, "loss": 0.2736, "step": 1230 }, { "epoch": 4.982278481012658, "step": 1230, "total_flos": 5.925003403780555e+18, "train_loss": 0.36658277566113123, "train_runtime": 54896.0887, "train_samples_per_second": 2.878, "train_steps_per_second": 0.022 } ], "logging_steps": 1, "max_steps": 1230, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.925003403780555e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }