{ "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": 6.354339283759368, "learning_rate": 3.2520325203252037e-07, "loss": 0.8211, "step": 1 }, { "epoch": 0.00810126582278481, "grad_norm": 6.416933318891079, "learning_rate": 6.504065040650407e-07, "loss": 0.8312, "step": 2 }, { "epoch": 0.012151898734177215, "grad_norm": 6.414899667254344, "learning_rate": 9.75609756097561e-07, "loss": 0.8301, "step": 3 }, { "epoch": 0.01620253164556962, "grad_norm": 6.386716978142179, "learning_rate": 1.3008130081300815e-06, "loss": 0.8277, "step": 4 }, { "epoch": 0.020253164556962026, "grad_norm": 5.882568438970497, "learning_rate": 1.6260162601626018e-06, "loss": 0.8217, "step": 5 }, { "epoch": 0.02430379746835443, "grad_norm": 5.700661342109287, "learning_rate": 1.951219512195122e-06, "loss": 0.8144, "step": 6 }, { "epoch": 0.028354430379746835, "grad_norm": 4.245040148656788, "learning_rate": 2.2764227642276426e-06, "loss": 0.7527, "step": 7 }, { "epoch": 0.03240506329113924, "grad_norm": 4.185507022272931, "learning_rate": 2.601626016260163e-06, "loss": 0.787, "step": 8 }, { "epoch": 0.03645569620253165, "grad_norm": 2.3856892893269976, "learning_rate": 2.926829268292683e-06, "loss": 0.7291, "step": 9 }, { "epoch": 0.04050632911392405, "grad_norm": 2.006222010523658, "learning_rate": 3.2520325203252037e-06, "loss": 0.7238, "step": 10 }, { "epoch": 0.044556962025316456, "grad_norm": 1.9407620284017522, "learning_rate": 3.577235772357724e-06, "loss": 0.7099, "step": 11 }, { "epoch": 0.04860759493670886, "grad_norm": 3.890693265339749, "learning_rate": 3.902439024390244e-06, "loss": 0.7331, "step": 12 }, { "epoch": 0.052658227848101265, "grad_norm": 4.089114470342537, "learning_rate": 4.227642276422765e-06, "loss": 0.7168, "step": 13 }, { "epoch": 0.05670886075949367, "grad_norm": 4.249698672063286, "learning_rate": 4.552845528455285e-06, "loss": 0.7312, "step": 14 }, { "epoch": 0.060759493670886074, "grad_norm": 3.9153466313215484, "learning_rate": 4.8780487804878055e-06, "loss": 0.7086, "step": 15 }, { "epoch": 0.06481012658227848, "grad_norm": 3.086501625177861, "learning_rate": 5.203252032520326e-06, "loss": 0.6619, "step": 16 }, { "epoch": 0.06886075949367089, "grad_norm": 2.766050907522193, "learning_rate": 5.528455284552846e-06, "loss": 0.6476, "step": 17 }, { "epoch": 0.0729113924050633, "grad_norm": 2.1241977559135177, "learning_rate": 5.853658536585366e-06, "loss": 0.6574, "step": 18 }, { "epoch": 0.0769620253164557, "grad_norm": 1.3918860141255291, "learning_rate": 6.178861788617887e-06, "loss": 0.6392, "step": 19 }, { "epoch": 0.0810126582278481, "grad_norm": 1.2229135131097035, "learning_rate": 6.504065040650407e-06, "loss": 0.6153, "step": 20 }, { "epoch": 0.08506329113924051, "grad_norm": 1.361133823719395, "learning_rate": 6.829268292682928e-06, "loss": 0.6086, "step": 21 }, { "epoch": 0.08911392405063291, "grad_norm": 1.314597722612901, "learning_rate": 7.154471544715448e-06, "loss": 0.5917, "step": 22 }, { "epoch": 0.09316455696202532, "grad_norm": 1.0306401833610233, "learning_rate": 7.4796747967479676e-06, "loss": 0.5928, "step": 23 }, { "epoch": 0.09721518987341772, "grad_norm": 0.7918658311128399, "learning_rate": 7.804878048780489e-06, "loss": 0.5755, "step": 24 }, { "epoch": 0.10126582278481013, "grad_norm": 0.9035420932503686, "learning_rate": 8.130081300813009e-06, "loss": 0.5668, "step": 25 }, { "epoch": 0.10531645569620253, "grad_norm": 0.8539027614400364, "learning_rate": 8.45528455284553e-06, "loss": 0.5805, "step": 26 }, { "epoch": 0.10936708860759493, "grad_norm": 0.6956954568185441, "learning_rate": 8.78048780487805e-06, "loss": 0.575, "step": 27 }, { "epoch": 0.11341772151898734, "grad_norm": 0.8249387486863041, "learning_rate": 9.10569105691057e-06, "loss": 0.5505, "step": 28 }, { "epoch": 0.11746835443037974, "grad_norm": 0.8059038375097219, "learning_rate": 9.43089430894309e-06, "loss": 0.5548, "step": 29 }, { "epoch": 0.12151898734177215, "grad_norm": 0.5685910835691669, "learning_rate": 9.756097560975611e-06, "loss": 0.5606, "step": 30 }, { "epoch": 0.12556962025316457, "grad_norm": 1.4195054388512505, "learning_rate": 1.008130081300813e-05, "loss": 0.5739, "step": 31 }, { "epoch": 0.12962025316455697, "grad_norm": 0.8341390501711973, "learning_rate": 1.0406504065040652e-05, "loss": 0.551, "step": 32 }, { "epoch": 0.13367088607594937, "grad_norm": 0.7033749668936686, "learning_rate": 1.0731707317073172e-05, "loss": 0.5359, "step": 33 }, { "epoch": 0.13772151898734178, "grad_norm": 0.499755411515897, "learning_rate": 1.1056910569105692e-05, "loss": 0.54, "step": 34 }, { "epoch": 0.14177215189873418, "grad_norm": 0.615038048315872, "learning_rate": 1.1382113821138213e-05, "loss": 0.5317, "step": 35 }, { "epoch": 0.1458227848101266, "grad_norm": 0.5677176435627704, "learning_rate": 1.1707317073170731e-05, "loss": 0.5296, "step": 36 }, { "epoch": 0.149873417721519, "grad_norm": 0.5280313373726304, "learning_rate": 1.2032520325203254e-05, "loss": 0.5271, "step": 37 }, { "epoch": 0.1539240506329114, "grad_norm": 0.6045979051130415, "learning_rate": 1.2357723577235774e-05, "loss": 0.524, "step": 38 }, { "epoch": 0.1579746835443038, "grad_norm": 0.4455931760778868, "learning_rate": 1.2682926829268294e-05, "loss": 0.5369, "step": 39 }, { "epoch": 0.1620253164556962, "grad_norm": 0.47548647577871567, "learning_rate": 1.3008130081300815e-05, "loss": 0.5129, "step": 40 }, { "epoch": 0.1660759493670886, "grad_norm": 0.6041637265440302, "learning_rate": 1.3333333333333333e-05, "loss": 0.5203, "step": 41 }, { "epoch": 0.17012658227848101, "grad_norm": 0.48018603401281734, "learning_rate": 1.3658536585365855e-05, "loss": 0.5189, "step": 42 }, { "epoch": 0.17417721518987342, "grad_norm": 0.582040911217193, "learning_rate": 1.3983739837398376e-05, "loss": 0.5041, "step": 43 }, { "epoch": 0.17822784810126582, "grad_norm": 0.4928531138664464, "learning_rate": 1.4308943089430896e-05, "loss": 0.5235, "step": 44 }, { "epoch": 0.18227848101265823, "grad_norm": 0.5128546631869223, "learning_rate": 1.4634146341463415e-05, "loss": 0.5068, "step": 45 }, { "epoch": 0.18632911392405063, "grad_norm": 0.42968644443840476, "learning_rate": 1.4959349593495935e-05, "loss": 0.5118, "step": 46 }, { "epoch": 0.19037974683544304, "grad_norm": 0.36290155707009164, "learning_rate": 1.528455284552846e-05, "loss": 0.5134, "step": 47 }, { "epoch": 0.19443037974683544, "grad_norm": 0.44540933210165046, "learning_rate": 1.5609756097560978e-05, "loss": 0.5057, "step": 48 }, { "epoch": 0.19848101265822785, "grad_norm": 0.40206667256209594, "learning_rate": 1.5934959349593496e-05, "loss": 0.5061, "step": 49 }, { "epoch": 0.20253164556962025, "grad_norm": 0.36307642119771744, "learning_rate": 1.6260162601626018e-05, "loss": 0.5084, "step": 50 }, { "epoch": 0.20658227848101265, "grad_norm": 0.3555490932593833, "learning_rate": 1.6585365853658537e-05, "loss": 0.5099, "step": 51 }, { "epoch": 0.21063291139240506, "grad_norm": 0.38069885167871387, "learning_rate": 1.691056910569106e-05, "loss": 0.5118, "step": 52 }, { "epoch": 0.21468354430379746, "grad_norm": 0.4502291186073661, "learning_rate": 1.7235772357723578e-05, "loss": 0.4909, "step": 53 }, { "epoch": 0.21873417721518987, "grad_norm": 0.4245697851373596, "learning_rate": 1.75609756097561e-05, "loss": 0.5112, "step": 54 }, { "epoch": 0.22278481012658227, "grad_norm": 0.4006940641592711, "learning_rate": 1.788617886178862e-05, "loss": 0.4945, "step": 55 }, { "epoch": 0.22683544303797468, "grad_norm": 0.4076793692068587, "learning_rate": 1.821138211382114e-05, "loss": 0.506, "step": 56 }, { "epoch": 0.23088607594936708, "grad_norm": 0.4839433743547984, "learning_rate": 1.8536585365853663e-05, "loss": 0.5045, "step": 57 }, { "epoch": 0.23493670886075949, "grad_norm": 0.4349275076765884, "learning_rate": 1.886178861788618e-05, "loss": 0.4905, "step": 58 }, { "epoch": 0.2389873417721519, "grad_norm": 0.4123546561818302, "learning_rate": 1.91869918699187e-05, "loss": 0.4963, "step": 59 }, { "epoch": 0.2430379746835443, "grad_norm": 0.44351452789016754, "learning_rate": 1.9512195121951222e-05, "loss": 0.4773, "step": 60 }, { "epoch": 0.2470886075949367, "grad_norm": 0.4069170337344362, "learning_rate": 1.983739837398374e-05, "loss": 0.4933, "step": 61 }, { "epoch": 0.25113924050632913, "grad_norm": 0.45913313317495524, "learning_rate": 2.016260162601626e-05, "loss": 0.499, "step": 62 }, { "epoch": 0.25518987341772154, "grad_norm": 0.5605099670122329, "learning_rate": 2.048780487804878e-05, "loss": 0.4915, "step": 63 }, { "epoch": 0.25924050632911394, "grad_norm": 0.6798553175445468, "learning_rate": 2.0813008130081303e-05, "loss": 0.4886, "step": 64 }, { "epoch": 0.26329113924050634, "grad_norm": 0.7667721311358375, "learning_rate": 2.1138211382113822e-05, "loss": 0.4829, "step": 65 }, { "epoch": 0.26734177215189875, "grad_norm": 0.8948928159414264, "learning_rate": 2.1463414634146344e-05, "loss": 0.49, "step": 66 }, { "epoch": 0.27139240506329115, "grad_norm": 0.4801041330681864, "learning_rate": 2.1788617886178863e-05, "loss": 0.5068, "step": 67 }, { "epoch": 0.27544303797468356, "grad_norm": 1.0052628816570772, "learning_rate": 2.2113821138211385e-05, "loss": 0.4836, "step": 68 }, { "epoch": 0.27949367088607596, "grad_norm": 0.9103904842990755, "learning_rate": 2.2439024390243907e-05, "loss": 0.4824, "step": 69 }, { "epoch": 0.28354430379746837, "grad_norm": 0.8721525601642346, "learning_rate": 2.2764227642276426e-05, "loss": 0.5101, "step": 70 }, { "epoch": 0.28759493670886077, "grad_norm": 1.01260490389457, "learning_rate": 2.3089430894308948e-05, "loss": 0.4782, "step": 71 }, { "epoch": 0.2916455696202532, "grad_norm": 0.9538616028145454, "learning_rate": 2.3414634146341463e-05, "loss": 0.4901, "step": 72 }, { "epoch": 0.2956962025316456, "grad_norm": 0.9589353698179944, "learning_rate": 2.3739837398373985e-05, "loss": 0.4868, "step": 73 }, { "epoch": 0.299746835443038, "grad_norm": 0.6149225903344657, "learning_rate": 2.4065040650406507e-05, "loss": 0.4818, "step": 74 }, { "epoch": 0.3037974683544304, "grad_norm": 0.8154204379658267, "learning_rate": 2.4390243902439026e-05, "loss": 0.4887, "step": 75 }, { "epoch": 0.3078481012658228, "grad_norm": 0.6934805200236519, "learning_rate": 2.4715447154471548e-05, "loss": 0.4673, "step": 76 }, { "epoch": 0.3118987341772152, "grad_norm": 0.77792420736469, "learning_rate": 2.5040650406504066e-05, "loss": 0.4808, "step": 77 }, { "epoch": 0.3159493670886076, "grad_norm": 0.7033610222949429, "learning_rate": 2.536585365853659e-05, "loss": 0.4911, "step": 78 }, { "epoch": 0.32, "grad_norm": 0.7406495133059409, "learning_rate": 2.569105691056911e-05, "loss": 0.4797, "step": 79 }, { "epoch": 0.3240506329113924, "grad_norm": 1.069522682669949, "learning_rate": 2.601626016260163e-05, "loss": 0.4904, "step": 80 }, { "epoch": 0.3281012658227848, "grad_norm": 0.5376326991237551, "learning_rate": 2.634146341463415e-05, "loss": 0.4865, "step": 81 }, { "epoch": 0.3321518987341772, "grad_norm": 1.0762009594241047, "learning_rate": 2.6666666666666667e-05, "loss": 0.4747, "step": 82 }, { "epoch": 0.3362025316455696, "grad_norm": 0.6869275430116534, "learning_rate": 2.699186991869919e-05, "loss": 0.4932, "step": 83 }, { "epoch": 0.34025316455696203, "grad_norm": 0.9505170826761062, "learning_rate": 2.731707317073171e-05, "loss": 0.4835, "step": 84 }, { "epoch": 0.34430379746835443, "grad_norm": 0.8725555359747081, "learning_rate": 2.764227642276423e-05, "loss": 0.4815, "step": 85 }, { "epoch": 0.34835443037974684, "grad_norm": 0.771112234427663, "learning_rate": 2.796747967479675e-05, "loss": 0.4781, "step": 86 }, { "epoch": 0.35240506329113924, "grad_norm": 0.8387892067753853, "learning_rate": 2.829268292682927e-05, "loss": 0.4642, "step": 87 }, { "epoch": 0.35645569620253165, "grad_norm": 1.251829897380383, "learning_rate": 2.8617886178861792e-05, "loss": 0.4587, "step": 88 }, { "epoch": 0.36050632911392405, "grad_norm": 0.6684073716866425, "learning_rate": 2.8943089430894314e-05, "loss": 0.4795, "step": 89 }, { "epoch": 0.36455696202531646, "grad_norm": 1.025887241698641, "learning_rate": 2.926829268292683e-05, "loss": 0.4727, "step": 90 }, { "epoch": 0.36860759493670886, "grad_norm": 1.415915375333013, "learning_rate": 2.959349593495935e-05, "loss": 0.4897, "step": 91 }, { "epoch": 0.37265822784810126, "grad_norm": 0.5808027799721855, "learning_rate": 2.991869918699187e-05, "loss": 0.4783, "step": 92 }, { "epoch": 0.37670886075949367, "grad_norm": 1.465661752730975, "learning_rate": 3.0243902439024392e-05, "loss": 0.478, "step": 93 }, { "epoch": 0.3807594936708861, "grad_norm": 0.5827329193408555, "learning_rate": 3.056910569105692e-05, "loss": 0.4825, "step": 94 }, { "epoch": 0.3848101265822785, "grad_norm": 1.1958266810308655, "learning_rate": 3.089430894308943e-05, "loss": 0.4816, "step": 95 }, { "epoch": 0.3888607594936709, "grad_norm": 0.748393040020456, "learning_rate": 3.1219512195121955e-05, "loss": 0.4802, "step": 96 }, { "epoch": 0.3929113924050633, "grad_norm": 0.8403727277938129, "learning_rate": 3.154471544715447e-05, "loss": 0.4932, "step": 97 }, { "epoch": 0.3969620253164557, "grad_norm": 0.9722271141815407, "learning_rate": 3.186991869918699e-05, "loss": 0.4802, "step": 98 }, { "epoch": 0.4010126582278481, "grad_norm": 0.960455603312288, "learning_rate": 3.2195121951219514e-05, "loss": 0.498, "step": 99 }, { "epoch": 0.4050632911392405, "grad_norm": 1.2852787590102466, "learning_rate": 3.2520325203252037e-05, "loss": 0.4814, "step": 100 }, { "epoch": 0.4091139240506329, "grad_norm": 0.9351782011006846, "learning_rate": 3.284552845528456e-05, "loss": 0.4764, "step": 101 }, { "epoch": 0.4131645569620253, "grad_norm": 0.7248752164471148, "learning_rate": 3.3170731707317074e-05, "loss": 0.4581, "step": 102 }, { "epoch": 0.4172151898734177, "grad_norm": 0.8610803704781315, "learning_rate": 3.3495934959349596e-05, "loss": 0.4573, "step": 103 }, { "epoch": 0.4212658227848101, "grad_norm": 1.141244786313121, "learning_rate": 3.382113821138212e-05, "loss": 0.4799, "step": 104 }, { "epoch": 0.4253164556962025, "grad_norm": 0.7048368982757099, "learning_rate": 3.414634146341463e-05, "loss": 0.4728, "step": 105 }, { "epoch": 0.4293670886075949, "grad_norm": 1.00859922076752, "learning_rate": 3.4471544715447155e-05, "loss": 0.4708, "step": 106 }, { "epoch": 0.43341772151898733, "grad_norm": 0.9536182714053715, "learning_rate": 3.479674796747968e-05, "loss": 0.4786, "step": 107 }, { "epoch": 0.43746835443037974, "grad_norm": 0.8042343619886441, "learning_rate": 3.51219512195122e-05, "loss": 0.4612, "step": 108 }, { "epoch": 0.44151898734177214, "grad_norm": 0.855567285023661, "learning_rate": 3.544715447154472e-05, "loss": 0.454, "step": 109 }, { "epoch": 0.44556962025316454, "grad_norm": 0.7724251682120098, "learning_rate": 3.577235772357724e-05, "loss": 0.4651, "step": 110 }, { "epoch": 0.44962025316455695, "grad_norm": 0.7057852639653921, "learning_rate": 3.609756097560976e-05, "loss": 0.4646, "step": 111 }, { "epoch": 0.45367088607594935, "grad_norm": 0.6441186667322315, "learning_rate": 3.642276422764228e-05, "loss": 0.4729, "step": 112 }, { "epoch": 0.45772151898734176, "grad_norm": 0.713022719489293, "learning_rate": 3.67479674796748e-05, "loss": 0.465, "step": 113 }, { "epoch": 0.46177215189873416, "grad_norm": 0.7021623285505268, "learning_rate": 3.7073170731707325e-05, "loss": 0.468, "step": 114 }, { "epoch": 0.46582278481012657, "grad_norm": 0.8237095008726756, "learning_rate": 3.739837398373984e-05, "loss": 0.473, "step": 115 }, { "epoch": 0.46987341772151897, "grad_norm": 0.6592858074699595, "learning_rate": 3.772357723577236e-05, "loss": 0.4538, "step": 116 }, { "epoch": 0.4739240506329114, "grad_norm": 0.7270118338443609, "learning_rate": 3.804878048780488e-05, "loss": 0.4678, "step": 117 }, { "epoch": 0.4779746835443038, "grad_norm": 0.5880423395141625, "learning_rate": 3.83739837398374e-05, "loss": 0.46, "step": 118 }, { "epoch": 0.4820253164556962, "grad_norm": 0.8625321269740713, "learning_rate": 3.869918699186992e-05, "loss": 0.467, "step": 119 }, { "epoch": 0.4860759493670886, "grad_norm": 0.9516728819178885, "learning_rate": 3.9024390243902444e-05, "loss": 0.4668, "step": 120 }, { "epoch": 0.490126582278481, "grad_norm": 1.1353698223958442, "learning_rate": 3.9349593495934966e-05, "loss": 0.4662, "step": 121 }, { "epoch": 0.4941772151898734, "grad_norm": 0.7293733611261571, "learning_rate": 3.967479674796748e-05, "loss": 0.4506, "step": 122 }, { "epoch": 0.4982278481012658, "grad_norm": 0.7684341607887686, "learning_rate": 4e-05, "loss": 0.4663, "step": 123 }, { "epoch": 0.5022784810126583, "grad_norm": 0.8932030964443223, "learning_rate": 3.999991946137476e-05, "loss": 0.4627, "step": 124 }, { "epoch": 0.5063291139240507, "grad_norm": 1.2542509331290435, "learning_rate": 3.999967784614766e-05, "loss": 0.4732, "step": 125 }, { "epoch": 0.5103797468354431, "grad_norm": 0.8755860529109312, "learning_rate": 3.9999275156264656e-05, "loss": 0.4657, "step": 126 }, { "epoch": 0.5144303797468355, "grad_norm": 0.761161786248994, "learning_rate": 3.999871139496895e-05, "loss": 0.453, "step": 127 }, { "epoch": 0.5184810126582279, "grad_norm": 1.1440725847152042, "learning_rate": 3.9997986566800995e-05, "loss": 0.4588, "step": 128 }, { "epoch": 0.5225316455696203, "grad_norm": 0.753057665449726, "learning_rate": 3.999710067759846e-05, "loss": 0.4681, "step": 129 }, { "epoch": 0.5265822784810127, "grad_norm": 0.7916043010203132, "learning_rate": 3.999605373449617e-05, "loss": 0.4622, "step": 130 }, { "epoch": 0.5306329113924051, "grad_norm": 0.7867488878034677, "learning_rate": 3.9994845745926075e-05, "loss": 0.4551, "step": 131 }, { "epoch": 0.5346835443037975, "grad_norm": 0.7447091709425686, "learning_rate": 3.999347672161713e-05, "loss": 0.4655, "step": 132 }, { "epoch": 0.5387341772151899, "grad_norm": 0.7318781992172791, "learning_rate": 3.999194667259528e-05, "loss": 0.4716, "step": 133 }, { "epoch": 0.5427848101265823, "grad_norm": 0.8004642439480708, "learning_rate": 3.999025561118334e-05, "loss": 0.4565, "step": 134 }, { "epoch": 0.5468354430379747, "grad_norm": 0.7270839740086383, "learning_rate": 3.998840355100086e-05, "loss": 0.4667, "step": 135 }, { "epoch": 0.5508860759493671, "grad_norm": 0.6772535519816474, "learning_rate": 3.998639050696409e-05, "loss": 0.4592, "step": 136 }, { "epoch": 0.5549367088607595, "grad_norm": 0.6489868253392055, "learning_rate": 3.998421649528582e-05, "loss": 0.4567, "step": 137 }, { "epoch": 0.5589873417721519, "grad_norm": 0.5295933965451042, "learning_rate": 3.9981881533475234e-05, "loss": 0.4575, "step": 138 }, { "epoch": 0.5630379746835443, "grad_norm": 0.6508584701668786, "learning_rate": 3.997938564033779e-05, "loss": 0.4611, "step": 139 }, { "epoch": 0.5670886075949367, "grad_norm": 0.5629919545717997, "learning_rate": 3.9976728835975064e-05, "loss": 0.4524, "step": 140 }, { "epoch": 0.5711392405063291, "grad_norm": 0.585829005414746, "learning_rate": 3.9973911141784605e-05, "loss": 0.4491, "step": 141 }, { "epoch": 0.5751898734177215, "grad_norm": 0.587433886517133, "learning_rate": 3.997093258045973e-05, "loss": 0.4562, "step": 142 }, { "epoch": 0.579240506329114, "grad_norm": 0.5540779674946045, "learning_rate": 3.996779317598936e-05, "loss": 0.4677, "step": 143 }, { "epoch": 0.5832911392405064, "grad_norm": 0.61683391542742, "learning_rate": 3.996449295365782e-05, "loss": 0.4674, "step": 144 }, { "epoch": 0.5873417721518988, "grad_norm": 0.6453789462130225, "learning_rate": 3.996103194004467e-05, "loss": 0.4482, "step": 145 }, { "epoch": 0.5913924050632912, "grad_norm": 0.6369683549090321, "learning_rate": 3.995741016302441e-05, "loss": 0.4553, "step": 146 }, { "epoch": 0.5954430379746836, "grad_norm": 0.6091954268882337, "learning_rate": 3.9953627651766364e-05, "loss": 0.4616, "step": 147 }, { "epoch": 0.599493670886076, "grad_norm": 0.5967311607173645, "learning_rate": 3.9949684436734325e-05, "loss": 0.4559, "step": 148 }, { "epoch": 0.6035443037974684, "grad_norm": 0.5139302556885356, "learning_rate": 3.994558054968643e-05, "loss": 0.4578, "step": 149 }, { "epoch": 0.6075949367088608, "grad_norm": 0.5655567589375522, "learning_rate": 3.994131602367481e-05, "loss": 0.458, "step": 150 }, { "epoch": 0.6116455696202532, "grad_norm": 0.5198205808869507, "learning_rate": 3.9936890893045376e-05, "loss": 0.46, "step": 151 }, { "epoch": 0.6156962025316456, "grad_norm": 0.4904023380149228, "learning_rate": 3.993230519343752e-05, "loss": 0.4612, "step": 152 }, { "epoch": 0.619746835443038, "grad_norm": 0.6446799796290733, "learning_rate": 3.992755896178383e-05, "loss": 0.4629, "step": 153 }, { "epoch": 0.6237974683544304, "grad_norm": 0.7601401072448184, "learning_rate": 3.992265223630981e-05, "loss": 0.4636, "step": 154 }, { "epoch": 0.6278481012658228, "grad_norm": 0.6946672738307469, "learning_rate": 3.991758505653355e-05, "loss": 0.4717, "step": 155 }, { "epoch": 0.6318987341772152, "grad_norm": 0.6085651022290662, "learning_rate": 3.991235746326543e-05, "loss": 0.4598, "step": 156 }, { "epoch": 0.6359493670886076, "grad_norm": 0.6789908868684463, "learning_rate": 3.9906969498607745e-05, "loss": 0.4459, "step": 157 }, { "epoch": 0.64, "grad_norm": 0.8119705097169638, "learning_rate": 3.990142120595444e-05, "loss": 0.4525, "step": 158 }, { "epoch": 0.6440506329113924, "grad_norm": 0.7492516198671819, "learning_rate": 3.98957126299907e-05, "loss": 0.4703, "step": 159 }, { "epoch": 0.6481012658227848, "grad_norm": 0.6950345042165725, "learning_rate": 3.9889843816692596e-05, "loss": 0.4494, "step": 160 }, { "epoch": 0.6521518987341772, "grad_norm": 0.7265130937591903, "learning_rate": 3.9883814813326766e-05, "loss": 0.4424, "step": 161 }, { "epoch": 0.6562025316455696, "grad_norm": 0.6540836577560566, "learning_rate": 3.9877625668449956e-05, "loss": 0.4393, "step": 162 }, { "epoch": 0.660253164556962, "grad_norm": 0.38537859991629697, "learning_rate": 3.98712764319087e-05, "loss": 0.4616, "step": 163 }, { "epoch": 0.6643037974683544, "grad_norm": 0.5818881212301503, "learning_rate": 3.9864767154838864e-05, "loss": 0.4544, "step": 164 }, { "epoch": 0.6683544303797468, "grad_norm": 0.6862334656444792, "learning_rate": 3.9858097889665277e-05, "loss": 0.4551, "step": 165 }, { "epoch": 0.6724050632911392, "grad_norm": 0.5961688871759786, "learning_rate": 3.985126869010129e-05, "loss": 0.4359, "step": 166 }, { "epoch": 0.6764556962025317, "grad_norm": 0.4921960039429861, "learning_rate": 3.984427961114833e-05, "loss": 0.4633, "step": 167 }, { "epoch": 0.6805063291139241, "grad_norm": 0.570490180866991, "learning_rate": 3.9837130709095475e-05, "loss": 0.4565, "step": 168 }, { "epoch": 0.6845569620253165, "grad_norm": 0.578660312464601, "learning_rate": 3.982982204151901e-05, "loss": 0.4486, "step": 169 }, { "epoch": 0.6886075949367089, "grad_norm": 0.518664517535379, "learning_rate": 3.982235366728193e-05, "loss": 0.4544, "step": 170 }, { "epoch": 0.6926582278481013, "grad_norm": 0.6255199055274627, "learning_rate": 3.9814725646533505e-05, "loss": 0.4587, "step": 171 }, { "epoch": 0.6967088607594937, "grad_norm": 0.5112793576976395, "learning_rate": 3.9806938040708746e-05, "loss": 0.4516, "step": 172 }, { "epoch": 0.7007594936708861, "grad_norm": 0.606173855455474, "learning_rate": 3.9798990912527976e-05, "loss": 0.4514, "step": 173 }, { "epoch": 0.7048101265822785, "grad_norm": 0.5323550695490584, "learning_rate": 3.979088432599627e-05, "loss": 0.4463, "step": 174 }, { "epoch": 0.7088607594936709, "grad_norm": 0.5020515091743822, "learning_rate": 3.9782618346402964e-05, "loss": 0.4388, "step": 175 }, { "epoch": 0.7129113924050633, "grad_norm": 0.6349141305000346, "learning_rate": 3.977419304032111e-05, "loss": 0.4511, "step": 176 }, { "epoch": 0.7169620253164557, "grad_norm": 0.6287554423475711, "learning_rate": 3.976560847560697e-05, "loss": 0.4486, "step": 177 }, { "epoch": 0.7210126582278481, "grad_norm": 0.5641203501236475, "learning_rate": 3.9756864721399456e-05, "loss": 0.4571, "step": 178 }, { "epoch": 0.7250632911392405, "grad_norm": 0.5385048280618474, "learning_rate": 3.974796184811956e-05, "loss": 0.4486, "step": 179 }, { "epoch": 0.7291139240506329, "grad_norm": 0.4999683995956364, "learning_rate": 3.973889992746979e-05, "loss": 0.4499, "step": 180 }, { "epoch": 0.7331645569620253, "grad_norm": 0.45505477272563766, "learning_rate": 3.972967903243361e-05, "loss": 0.4434, "step": 181 }, { "epoch": 0.7372151898734177, "grad_norm": 0.5345516521212655, "learning_rate": 3.972029923727486e-05, "loss": 0.44, "step": 182 }, { "epoch": 0.7412658227848101, "grad_norm": 0.5591366640756519, "learning_rate": 3.971076061753709e-05, "loss": 0.4478, "step": 183 }, { "epoch": 0.7453164556962025, "grad_norm": 0.443098495924804, "learning_rate": 3.9701063250043066e-05, "loss": 0.4522, "step": 184 }, { "epoch": 0.7493670886075949, "grad_norm": 0.5785778234734176, "learning_rate": 3.969120721289402e-05, "loss": 0.4423, "step": 185 }, { "epoch": 0.7534177215189873, "grad_norm": 0.4623166517409209, "learning_rate": 3.9681192585469146e-05, "loss": 0.4288, "step": 186 }, { "epoch": 0.7574683544303797, "grad_norm": 0.5258613997741649, "learning_rate": 3.9671019448424865e-05, "loss": 0.4527, "step": 187 }, { "epoch": 0.7615189873417721, "grad_norm": 0.5533252145454644, "learning_rate": 3.966068788369422e-05, "loss": 0.4487, "step": 188 }, { "epoch": 0.7655696202531646, "grad_norm": 0.5192259314417941, "learning_rate": 3.965019797448622e-05, "loss": 0.4413, "step": 189 }, { "epoch": 0.769620253164557, "grad_norm": 0.5253996700099032, "learning_rate": 3.963954980528515e-05, "loss": 0.4344, "step": 190 }, { "epoch": 0.7736708860759494, "grad_norm": 0.4913628113143923, "learning_rate": 3.9628743461849905e-05, "loss": 0.456, "step": 191 }, { "epoch": 0.7777215189873418, "grad_norm": 0.5777199199006973, "learning_rate": 3.961777903121329e-05, "loss": 0.4624, "step": 192 }, { "epoch": 0.7817721518987342, "grad_norm": 0.506221472517052, "learning_rate": 3.960665660168131e-05, "loss": 0.4535, "step": 193 }, { "epoch": 0.7858227848101266, "grad_norm": 0.6252024456421241, "learning_rate": 3.9595376262832485e-05, "loss": 0.4475, "step": 194 }, { "epoch": 0.789873417721519, "grad_norm": 0.5033671822151695, "learning_rate": 3.9583938105517127e-05, "loss": 0.4456, "step": 195 }, { "epoch": 0.7939240506329114, "grad_norm": 0.5345223786098035, "learning_rate": 3.957234222185657e-05, "loss": 0.4592, "step": 196 }, { "epoch": 0.7979746835443038, "grad_norm": 0.6540640303259668, "learning_rate": 3.9560588705242474e-05, "loss": 0.4572, "step": 197 }, { "epoch": 0.8020253164556962, "grad_norm": 0.459383262604561, "learning_rate": 3.954867765033605e-05, "loss": 0.4471, "step": 198 }, { "epoch": 0.8060759493670886, "grad_norm": 0.7474927240394595, "learning_rate": 3.953660915306728e-05, "loss": 0.438, "step": 199 }, { "epoch": 0.810126582278481, "grad_norm": 0.6614175363474404, "learning_rate": 3.952438331063419e-05, "loss": 0.4555, "step": 200 }, { "epoch": 0.8141772151898734, "grad_norm": 0.4985555518128984, "learning_rate": 3.951200022150205e-05, "loss": 0.4388, "step": 201 }, { "epoch": 0.8182278481012658, "grad_norm": 0.4523834388317368, "learning_rate": 3.949945998540253e-05, "loss": 0.4486, "step": 202 }, { "epoch": 0.8222784810126582, "grad_norm": 0.48589729482807326, "learning_rate": 3.9486762703332993e-05, "loss": 0.434, "step": 203 }, { "epoch": 0.8263291139240506, "grad_norm": 0.522581208427055, "learning_rate": 3.947390847755559e-05, "loss": 0.4456, "step": 204 }, { "epoch": 0.830379746835443, "grad_norm": 0.4527973612657358, "learning_rate": 3.946089741159648e-05, "loss": 0.4442, "step": 205 }, { "epoch": 0.8344303797468354, "grad_norm": 0.882393638623316, "learning_rate": 3.944772961024501e-05, "loss": 0.4584, "step": 206 }, { "epoch": 0.8384810126582278, "grad_norm": 0.40800459102619274, "learning_rate": 3.943440517955285e-05, "loss": 0.4525, "step": 207 }, { "epoch": 0.8425316455696202, "grad_norm": 0.44924184236585407, "learning_rate": 3.9420924226833126e-05, "loss": 0.4337, "step": 208 }, { "epoch": 0.8465822784810126, "grad_norm": 0.4041719312567507, "learning_rate": 3.9407286860659566e-05, "loss": 0.4324, "step": 209 }, { "epoch": 0.850632911392405, "grad_norm": 0.45771843315154365, "learning_rate": 3.9393493190865657e-05, "loss": 0.4393, "step": 210 }, { "epoch": 0.8546835443037974, "grad_norm": 0.500005956493709, "learning_rate": 3.937954332854371e-05, "loss": 0.4492, "step": 211 }, { "epoch": 0.8587341772151899, "grad_norm": 0.3936507130274401, "learning_rate": 3.9365437386044016e-05, "loss": 0.4452, "step": 212 }, { "epoch": 0.8627848101265823, "grad_norm": 0.6418437945379682, "learning_rate": 3.935117547697387e-05, "loss": 0.453, "step": 213 }, { "epoch": 0.8668354430379747, "grad_norm": 1.4064233051059871, "learning_rate": 3.933675771619675e-05, "loss": 0.4498, "step": 214 }, { "epoch": 0.8708860759493671, "grad_norm": 1.059396596568302, "learning_rate": 3.932218421983131e-05, "loss": 0.4457, "step": 215 }, { "epoch": 0.8749367088607595, "grad_norm": 0.555035640642864, "learning_rate": 3.9307455105250484e-05, "loss": 0.4487, "step": 216 }, { "epoch": 0.8789873417721519, "grad_norm": 0.6388083604475806, "learning_rate": 3.929257049108054e-05, "loss": 0.4541, "step": 217 }, { "epoch": 0.8830379746835443, "grad_norm": 0.6337000967345201, "learning_rate": 3.927753049720011e-05, "loss": 0.44, "step": 218 }, { "epoch": 0.8870886075949367, "grad_norm": 0.5916352754296074, "learning_rate": 3.9262335244739234e-05, "loss": 0.4587, "step": 219 }, { "epoch": 0.8911392405063291, "grad_norm": 0.6450166137949507, "learning_rate": 3.92469848560784e-05, "loss": 0.4398, "step": 220 }, { "epoch": 0.8951898734177215, "grad_norm": 0.48682015338986173, "learning_rate": 3.923147945484751e-05, "loss": 0.4384, "step": 221 }, { "epoch": 0.8992405063291139, "grad_norm": 0.6318496594026971, "learning_rate": 3.9215819165924956e-05, "loss": 0.4564, "step": 222 }, { "epoch": 0.9032911392405063, "grad_norm": 0.5598431057155867, "learning_rate": 3.920000411543654e-05, "loss": 0.4475, "step": 223 }, { "epoch": 0.9073417721518987, "grad_norm": 0.5626795955741168, "learning_rate": 3.9184034430754495e-05, "loss": 0.4492, "step": 224 }, { "epoch": 0.9113924050632911, "grad_norm": 0.5319723995485589, "learning_rate": 3.916791024049648e-05, "loss": 0.4337, "step": 225 }, { "epoch": 0.9154430379746835, "grad_norm": 0.38074104484453375, "learning_rate": 3.91516316745245e-05, "loss": 0.4503, "step": 226 }, { "epoch": 0.9194936708860759, "grad_norm": 0.590608863254065, "learning_rate": 3.913519886394389e-05, "loss": 0.438, "step": 227 }, { "epoch": 0.9235443037974683, "grad_norm": 0.45580102393622285, "learning_rate": 3.911861194110225e-05, "loss": 0.4459, "step": 228 }, { "epoch": 0.9275949367088607, "grad_norm": 0.42145605309516404, "learning_rate": 3.910187103958837e-05, "loss": 0.4508, "step": 229 }, { "epoch": 0.9316455696202531, "grad_norm": 0.5372502606389523, "learning_rate": 3.908497629423117e-05, "loss": 0.45, "step": 230 }, { "epoch": 0.9356962025316455, "grad_norm": 0.41715445722165223, "learning_rate": 3.9067927841098614e-05, "loss": 0.4291, "step": 231 }, { "epoch": 0.9397468354430379, "grad_norm": 0.48500532340692554, "learning_rate": 3.9050725817496594e-05, "loss": 0.441, "step": 232 }, { "epoch": 0.9437974683544303, "grad_norm": 0.4372441700131902, "learning_rate": 3.9033370361967844e-05, "loss": 0.4505, "step": 233 }, { "epoch": 0.9478481012658228, "grad_norm": 0.44714546257575394, "learning_rate": 3.901586161429081e-05, "loss": 0.4468, "step": 234 }, { "epoch": 0.9518987341772152, "grad_norm": 0.4253462819281477, "learning_rate": 3.8998199715478545e-05, "loss": 0.4297, "step": 235 }, { "epoch": 0.9559493670886076, "grad_norm": 0.35777335331416255, "learning_rate": 3.8980384807777564e-05, "loss": 0.4446, "step": 236 }, { "epoch": 0.96, "grad_norm": 0.39173494368261785, "learning_rate": 3.896241703466667e-05, "loss": 0.4386, "step": 237 }, { "epoch": 0.9640506329113924, "grad_norm": 0.45479835900007387, "learning_rate": 3.894429654085585e-05, "loss": 0.4421, "step": 238 }, { "epoch": 0.9681012658227848, "grad_norm": 0.420865539671571, "learning_rate": 3.892602347228505e-05, "loss": 0.4434, "step": 239 }, { "epoch": 0.9721518987341772, "grad_norm": 0.36782245053631857, "learning_rate": 3.890759797612307e-05, "loss": 0.4343, "step": 240 }, { "epoch": 0.9762025316455696, "grad_norm": 0.3970838396235413, "learning_rate": 3.888902020076632e-05, "loss": 0.4419, "step": 241 }, { "epoch": 0.980253164556962, "grad_norm": 6.281224121487392, "learning_rate": 3.887029029583764e-05, "loss": 0.4583, "step": 242 }, { "epoch": 0.9843037974683544, "grad_norm": 48.019826996879395, "learning_rate": 3.8851408412185125e-05, "loss": 0.5306, "step": 243 }, { "epoch": 0.9883544303797468, "grad_norm": 0.8630253185074803, "learning_rate": 3.8832374701880855e-05, "loss": 0.4414, "step": 244 }, { "epoch": 0.9924050632911392, "grad_norm": 1.0716511218164966, "learning_rate": 3.881318931821972e-05, "loss": 0.4344, "step": 245 }, { "epoch": 0.9964556962025316, "grad_norm": 1.287513217381779, "learning_rate": 3.879385241571817e-05, "loss": 0.4482, "step": 246 }, { "epoch": 1.000506329113924, "grad_norm": 0.7180479023729714, "learning_rate": 3.8774364150112955e-05, "loss": 0.4416, "step": 247 }, { "epoch": 1.0045569620253165, "grad_norm": 0.6475750896996986, "learning_rate": 3.8754724678359884e-05, "loss": 0.3954, "step": 248 }, { "epoch": 1.0086075949367088, "grad_norm": 0.5907860004840504, "learning_rate": 3.873493415863256e-05, "loss": 0.4178, "step": 249 }, { "epoch": 1.0126582278481013, "grad_norm": 0.6915294145793293, "learning_rate": 3.871499275032111e-05, "loss": 0.4119, "step": 250 }, { "epoch": 1.0167088607594936, "grad_norm": 0.6561274446432093, "learning_rate": 3.869490061403091e-05, "loss": 0.4212, "step": 251 }, { "epoch": 1.0207594936708861, "grad_norm": 0.6889007295963919, "learning_rate": 3.867465791158124e-05, "loss": 0.4173, "step": 252 }, { "epoch": 1.0248101265822784, "grad_norm": 0.6710399351716738, "learning_rate": 3.865426480600407e-05, "loss": 0.4044, "step": 253 }, { "epoch": 1.028860759493671, "grad_norm": 0.9978360645019638, "learning_rate": 3.863372146154264e-05, "loss": 0.4137, "step": 254 }, { "epoch": 1.0329113924050632, "grad_norm": 1.0223943888247593, "learning_rate": 3.861302804365024e-05, "loss": 0.4146, "step": 255 }, { "epoch": 1.0369620253164558, "grad_norm": 0.6160403470338607, "learning_rate": 3.85921847189888e-05, "loss": 0.3999, "step": 256 }, { "epoch": 1.041012658227848, "grad_norm": 0.8692948257404531, "learning_rate": 3.85711916554276e-05, "loss": 0.4099, "step": 257 }, { "epoch": 1.0450632911392406, "grad_norm": 0.6243057520657582, "learning_rate": 3.85500490220419e-05, "loss": 0.4044, "step": 258 }, { "epoch": 1.0491139240506329, "grad_norm": 0.7059237110996236, "learning_rate": 3.852875698911154e-05, "loss": 0.4242, "step": 259 }, { "epoch": 1.0531645569620254, "grad_norm": 0.7291797562735614, "learning_rate": 3.850731572811963e-05, "loss": 0.423, "step": 260 }, { "epoch": 1.0572151898734177, "grad_norm": 0.5117129891699297, "learning_rate": 3.848572541175116e-05, "loss": 0.3997, "step": 261 }, { "epoch": 1.0612658227848102, "grad_norm": 0.6252979825905661, "learning_rate": 3.846398621389154e-05, "loss": 0.4212, "step": 262 }, { "epoch": 1.0653164556962025, "grad_norm": 0.5530314730524939, "learning_rate": 3.84420983096253e-05, "loss": 0.4077, "step": 263 }, { "epoch": 1.069367088607595, "grad_norm": 0.6411859714827605, "learning_rate": 3.8420061875234606e-05, "loss": 0.4, "step": 264 }, { "epoch": 1.0734177215189873, "grad_norm": 0.44585757210756016, "learning_rate": 3.839787708819787e-05, "loss": 0.4016, "step": 265 }, { "epoch": 1.0774683544303798, "grad_norm": 0.5548668704042862, "learning_rate": 3.8375544127188325e-05, "loss": 0.4076, "step": 266 }, { "epoch": 1.081518987341772, "grad_norm": 0.5114373883487318, "learning_rate": 3.8353063172072564e-05, "loss": 0.4027, "step": 267 }, { "epoch": 1.0855696202531646, "grad_norm": 0.4664108835888721, "learning_rate": 3.8330434403909105e-05, "loss": 0.4202, "step": 268 }, { "epoch": 1.089620253164557, "grad_norm": 0.5723032304018557, "learning_rate": 3.8307658004946934e-05, "loss": 0.424, "step": 269 }, { "epoch": 1.0936708860759494, "grad_norm": 0.47140734868681783, "learning_rate": 3.8284734158624046e-05, "loss": 0.4017, "step": 270 }, { "epoch": 1.0977215189873417, "grad_norm": 0.49927271940689755, "learning_rate": 3.826166304956594e-05, "loss": 0.4012, "step": 271 }, { "epoch": 1.1017721518987342, "grad_norm": 0.5045937861849679, "learning_rate": 3.8238444863584164e-05, "loss": 0.3967, "step": 272 }, { "epoch": 1.1058227848101265, "grad_norm": 0.7776718646659695, "learning_rate": 3.821507978767479e-05, "loss": 0.421, "step": 273 }, { "epoch": 1.109873417721519, "grad_norm": 0.47346571610055793, "learning_rate": 3.819156801001693e-05, "loss": 0.3899, "step": 274 }, { "epoch": 1.1139240506329113, "grad_norm": 1.2302898056625906, "learning_rate": 3.816790971997121e-05, "loss": 0.4088, "step": 275 }, { "epoch": 1.1179746835443038, "grad_norm": 0.4751909630516516, "learning_rate": 3.8144105108078246e-05, "loss": 0.4086, "step": 276 }, { "epoch": 1.1220253164556961, "grad_norm": 0.426367814419382, "learning_rate": 3.81201543660571e-05, "loss": 0.3994, "step": 277 }, { "epoch": 1.1260759493670887, "grad_norm": 0.41105811409476395, "learning_rate": 3.809605768680377e-05, "loss": 0.4004, "step": 278 }, { "epoch": 1.130126582278481, "grad_norm": 0.40341870950913256, "learning_rate": 3.807181526438958e-05, "loss": 0.4006, "step": 279 }, { "epoch": 1.1341772151898735, "grad_norm": 0.38475966340291284, "learning_rate": 3.8047427294059697e-05, "loss": 0.3976, "step": 280 }, { "epoch": 1.1382278481012658, "grad_norm": 0.3863771209882561, "learning_rate": 3.802289397223145e-05, "loss": 0.4073, "step": 281 }, { "epoch": 1.1422784810126583, "grad_norm": 0.385748010355889, "learning_rate": 3.7998215496492854e-05, "loss": 0.4054, "step": 282 }, { "epoch": 1.1463291139240506, "grad_norm": 0.3710441041158405, "learning_rate": 3.797339206560096e-05, "loss": 0.4093, "step": 283 }, { "epoch": 1.150379746835443, "grad_norm": 0.39062851823045397, "learning_rate": 3.794842387948027e-05, "loss": 0.3999, "step": 284 }, { "epoch": 1.1544303797468354, "grad_norm": 0.3956504938398933, "learning_rate": 3.7923311139221114e-05, "loss": 0.4021, "step": 285 }, { "epoch": 1.158481012658228, "grad_norm": 0.3774213474995563, "learning_rate": 3.7898054047078054e-05, "loss": 0.4, "step": 286 }, { "epoch": 1.1625316455696202, "grad_norm": 0.4105015084568266, "learning_rate": 3.787265280646825e-05, "loss": 0.4183, "step": 287 }, { "epoch": 1.1665822784810127, "grad_norm": 0.38571994154792705, "learning_rate": 3.7847107621969786e-05, "loss": 0.3958, "step": 288 }, { "epoch": 1.170632911392405, "grad_norm": 0.362492096292577, "learning_rate": 3.7821418699320064e-05, "loss": 0.4042, "step": 289 }, { "epoch": 1.1746835443037975, "grad_norm": 0.48463635279530576, "learning_rate": 3.7795586245414145e-05, "loss": 0.4058, "step": 290 }, { "epoch": 1.1787341772151898, "grad_norm": 0.45193404197306297, "learning_rate": 3.776961046830306e-05, "loss": 0.4075, "step": 291 }, { "epoch": 1.1827848101265823, "grad_norm": 0.5239432545256942, "learning_rate": 3.774349157719215e-05, "loss": 0.4033, "step": 292 }, { "epoch": 1.1868354430379746, "grad_norm": 0.5612616002747873, "learning_rate": 3.7717229782439365e-05, "loss": 0.4018, "step": 293 }, { "epoch": 1.1908860759493671, "grad_norm": 0.35569119553987294, "learning_rate": 3.769082529555359e-05, "loss": 0.3887, "step": 294 }, { "epoch": 1.1949367088607594, "grad_norm": 0.5709223020171331, "learning_rate": 3.766427832919294e-05, "loss": 0.4087, "step": 295 }, { "epoch": 1.198987341772152, "grad_norm": 0.43323250946661257, "learning_rate": 3.7637589097163024e-05, "loss": 0.3946, "step": 296 }, { "epoch": 1.2030379746835442, "grad_norm": 0.37706321102492235, "learning_rate": 3.761075781441526e-05, "loss": 0.3947, "step": 297 }, { "epoch": 1.2070886075949367, "grad_norm": 0.48735966158751476, "learning_rate": 3.75837846970451e-05, "loss": 0.4039, "step": 298 }, { "epoch": 1.211139240506329, "grad_norm": 0.3534461066959703, "learning_rate": 3.755666996229032e-05, "loss": 0.3926, "step": 299 }, { "epoch": 1.2151898734177216, "grad_norm": 0.46876386333034714, "learning_rate": 3.752941382852927e-05, "loss": 0.4015, "step": 300 }, { "epoch": 1.2192405063291138, "grad_norm": 0.4053131071391026, "learning_rate": 3.7502016515279115e-05, "loss": 0.3976, "step": 301 }, { "epoch": 1.2232911392405064, "grad_norm": 0.399776164333771, "learning_rate": 3.7474478243194043e-05, "loss": 0.3968, "step": 302 }, { "epoch": 1.2273417721518987, "grad_norm": 0.37878995819634403, "learning_rate": 3.744679923406351e-05, "loss": 0.4049, "step": 303 }, { "epoch": 1.2313924050632912, "grad_norm": 0.4893428334383951, "learning_rate": 3.741897971081043e-05, "loss": 0.4009, "step": 304 }, { "epoch": 1.2354430379746835, "grad_norm": 0.40181787639041633, "learning_rate": 3.739101989748946e-05, "loss": 0.4028, "step": 305 }, { "epoch": 1.239493670886076, "grad_norm": 0.4685183092668485, "learning_rate": 3.7362920019285066e-05, "loss": 0.4014, "step": 306 }, { "epoch": 1.2435443037974683, "grad_norm": 0.38654218085500935, "learning_rate": 3.73346803025098e-05, "loss": 0.3844, "step": 307 }, { "epoch": 1.2475949367088608, "grad_norm": 0.3804708199830811, "learning_rate": 3.730630097460247e-05, "loss": 0.3962, "step": 308 }, { "epoch": 1.251645569620253, "grad_norm": 0.423654504877119, "learning_rate": 3.727778226412628e-05, "loss": 0.3975, "step": 309 }, { "epoch": 1.2556962025316456, "grad_norm": 0.3699512932992789, "learning_rate": 3.7249124400767006e-05, "loss": 0.3982, "step": 310 }, { "epoch": 1.259746835443038, "grad_norm": 0.4902110361625691, "learning_rate": 3.722032761533114e-05, "loss": 0.3939, "step": 311 }, { "epoch": 1.2637974683544304, "grad_norm": 0.33014371690224825, "learning_rate": 3.719139213974403e-05, "loss": 0.4054, "step": 312 }, { "epoch": 1.2678481012658227, "grad_norm": 0.3617830490606708, "learning_rate": 3.7162318207048006e-05, "loss": 0.4095, "step": 313 }, { "epoch": 1.2718987341772152, "grad_norm": 0.367396037019516, "learning_rate": 3.713310605140055e-05, "loss": 0.4043, "step": 314 }, { "epoch": 1.2759493670886077, "grad_norm": 0.449055525191707, "learning_rate": 3.710375590807233e-05, "loss": 0.3981, "step": 315 }, { "epoch": 1.28, "grad_norm": 0.35934230998369937, "learning_rate": 3.7074268013445365e-05, "loss": 0.426, "step": 316 }, { "epoch": 1.2840506329113923, "grad_norm": 0.3285985429050412, "learning_rate": 3.7044642605011114e-05, "loss": 0.418, "step": 317 }, { "epoch": 1.2881012658227848, "grad_norm": 0.4075766190903428, "learning_rate": 3.701487992136854e-05, "loss": 0.3885, "step": 318 }, { "epoch": 1.2921518987341774, "grad_norm": 0.3228080902487238, "learning_rate": 3.69849802022222e-05, "loss": 0.3875, "step": 319 }, { "epoch": 1.2962025316455696, "grad_norm": 0.4617479658477641, "learning_rate": 3.6954943688380334e-05, "loss": 0.4087, "step": 320 }, { "epoch": 1.300253164556962, "grad_norm": 0.5152769568169459, "learning_rate": 3.692477062175289e-05, "loss": 0.4047, "step": 321 }, { "epoch": 1.3043037974683545, "grad_norm": 0.3596724687124773, "learning_rate": 3.689446124534958e-05, "loss": 0.4048, "step": 322 }, { "epoch": 1.308354430379747, "grad_norm": 0.4661112443538026, "learning_rate": 3.686401580327799e-05, "loss": 0.3843, "step": 323 }, { "epoch": 1.3124050632911393, "grad_norm": 0.42953373346043006, "learning_rate": 3.683343454074149e-05, "loss": 0.4063, "step": 324 }, { "epoch": 1.3164556962025316, "grad_norm": 0.312169172262583, "learning_rate": 3.6802717704037386e-05, "loss": 0.3876, "step": 325 }, { "epoch": 1.320506329113924, "grad_norm": 0.5845639978844808, "learning_rate": 3.6771865540554855e-05, "loss": 0.3909, "step": 326 }, { "epoch": 1.3245569620253166, "grad_norm": 0.40098260097522576, "learning_rate": 3.674087829877297e-05, "loss": 0.3991, "step": 327 }, { "epoch": 1.3286075949367089, "grad_norm": 0.3226693964297329, "learning_rate": 3.6709756228258735e-05, "loss": 0.4002, "step": 328 }, { "epoch": 1.3326582278481012, "grad_norm": 0.3430280923061136, "learning_rate": 3.667849957966501e-05, "loss": 0.3984, "step": 329 }, { "epoch": 1.3367088607594937, "grad_norm": 0.36950444858899223, "learning_rate": 3.6647108604728546e-05, "loss": 0.4077, "step": 330 }, { "epoch": 1.3407594936708862, "grad_norm": 0.30526593561624904, "learning_rate": 3.661558355626795e-05, "loss": 0.3936, "step": 331 }, { "epoch": 1.3448101265822785, "grad_norm": 0.3488228326234552, "learning_rate": 3.658392468818163e-05, "loss": 0.3972, "step": 332 }, { "epoch": 1.3488607594936708, "grad_norm": 0.29923236579709517, "learning_rate": 3.655213225544574e-05, "loss": 0.4029, "step": 333 }, { "epoch": 1.3529113924050633, "grad_norm": 0.31947244810281444, "learning_rate": 3.652020651411218e-05, "loss": 0.3956, "step": 334 }, { "epoch": 1.3569620253164558, "grad_norm": 0.357625164742112, "learning_rate": 3.6488147721306474e-05, "loss": 0.4056, "step": 335 }, { "epoch": 1.3610126582278481, "grad_norm": 0.31420201852156276, "learning_rate": 3.645595613522574e-05, "loss": 0.4103, "step": 336 }, { "epoch": 1.3650632911392404, "grad_norm": 0.3394911778299976, "learning_rate": 3.642363201513657e-05, "loss": 0.4016, "step": 337 }, { "epoch": 1.369113924050633, "grad_norm": 0.3713009208797141, "learning_rate": 3.6391175621373006e-05, "loss": 0.3994, "step": 338 }, { "epoch": 1.3731645569620254, "grad_norm": 0.3881532630088444, "learning_rate": 3.6358587215334355e-05, "loss": 0.4077, "step": 339 }, { "epoch": 1.3772151898734177, "grad_norm": 0.35667030988814036, "learning_rate": 3.632586705948318e-05, "loss": 0.4058, "step": 340 }, { "epoch": 1.38126582278481, "grad_norm": 0.35959694462238456, "learning_rate": 3.629301541734311e-05, "loss": 0.393, "step": 341 }, { "epoch": 1.3853164556962025, "grad_norm": 0.36042838075129613, "learning_rate": 3.626003255349676e-05, "loss": 0.3985, "step": 342 }, { "epoch": 1.389367088607595, "grad_norm": 0.4538843315746254, "learning_rate": 3.622691873358357e-05, "loss": 0.4098, "step": 343 }, { "epoch": 1.3934177215189874, "grad_norm": 0.3741891371894677, "learning_rate": 3.61936742242977e-05, "loss": 0.3791, "step": 344 }, { "epoch": 1.3974683544303796, "grad_norm": 0.4769482421062264, "learning_rate": 3.6160299293385864e-05, "loss": 0.3912, "step": 345 }, { "epoch": 1.4015189873417722, "grad_norm": 0.4256288024034492, "learning_rate": 3.612679420964516e-05, "loss": 0.4109, "step": 346 }, { "epoch": 1.4055696202531647, "grad_norm": 0.3796606000564335, "learning_rate": 3.609315924292092e-05, "loss": 0.3937, "step": 347 }, { "epoch": 1.409620253164557, "grad_norm": 0.38202762516394695, "learning_rate": 3.6059394664104554e-05, "loss": 0.4036, "step": 348 }, { "epoch": 1.4136708860759493, "grad_norm": 0.33342642370793873, "learning_rate": 3.602550074513133e-05, "loss": 0.401, "step": 349 }, { "epoch": 1.4177215189873418, "grad_norm": 0.3531635605737397, "learning_rate": 3.599147775897822e-05, "loss": 0.3894, "step": 350 }, { "epoch": 1.4217721518987343, "grad_norm": 0.404393962963754, "learning_rate": 3.595732597966167e-05, "loss": 0.3974, "step": 351 }, { "epoch": 1.4258227848101266, "grad_norm": 0.7680224051837081, "learning_rate": 3.592304568223542e-05, "loss": 0.4099, "step": 352 }, { "epoch": 1.4298734177215189, "grad_norm": 0.37212599637154004, "learning_rate": 3.588863714278826e-05, "loss": 0.4056, "step": 353 }, { "epoch": 1.4339240506329114, "grad_norm": 0.433469387775846, "learning_rate": 3.585410063844186e-05, "loss": 0.4012, "step": 354 }, { "epoch": 1.437974683544304, "grad_norm": 0.355325135156294, "learning_rate": 3.581943644734846e-05, "loss": 0.4017, "step": 355 }, { "epoch": 1.4420253164556962, "grad_norm": 0.40912309772774946, "learning_rate": 3.578464484868869e-05, "loss": 0.3868, "step": 356 }, { "epoch": 1.4460759493670885, "grad_norm": 0.33366801612400815, "learning_rate": 3.5749726122669316e-05, "loss": 0.4016, "step": 357 }, { "epoch": 1.450126582278481, "grad_norm": 0.5426564013997984, "learning_rate": 3.5714680550520943e-05, "loss": 0.4011, "step": 358 }, { "epoch": 1.4541772151898735, "grad_norm": 0.3124580324456678, "learning_rate": 3.5679508414495794e-05, "loss": 0.4022, "step": 359 }, { "epoch": 1.4582278481012658, "grad_norm": 0.4215643155481631, "learning_rate": 3.564420999786543e-05, "loss": 0.3964, "step": 360 }, { "epoch": 1.4622784810126581, "grad_norm": 0.3967196547438108, "learning_rate": 3.560878558491842e-05, "loss": 0.3816, "step": 361 }, { "epoch": 1.4663291139240506, "grad_norm": 0.37069462426263416, "learning_rate": 3.5573235460958145e-05, "loss": 0.4311, "step": 362 }, { "epoch": 1.4703797468354431, "grad_norm": 0.41378574198727897, "learning_rate": 3.553755991230039e-05, "loss": 0.3967, "step": 363 }, { "epoch": 1.4744303797468354, "grad_norm": 0.3501793786965155, "learning_rate": 3.5501759226271144e-05, "loss": 0.4003, "step": 364 }, { "epoch": 1.4784810126582277, "grad_norm": 0.7501392558626657, "learning_rate": 3.546583369120419e-05, "loss": 0.4083, "step": 365 }, { "epoch": 1.4825316455696202, "grad_norm": 0.438442855161263, "learning_rate": 3.5429783596438864e-05, "loss": 0.408, "step": 366 }, { "epoch": 1.4865822784810128, "grad_norm": 0.4997610764088511, "learning_rate": 3.539360923231766e-05, "loss": 0.4126, "step": 367 }, { "epoch": 1.490632911392405, "grad_norm": 0.6237275014262749, "learning_rate": 3.535731089018394e-05, "loss": 0.3877, "step": 368 }, { "epoch": 1.4946835443037974, "grad_norm": 0.39755713878080085, "learning_rate": 3.532088886237956e-05, "loss": 0.4028, "step": 369 }, { "epoch": 1.4987341772151899, "grad_norm": 0.403621153243375, "learning_rate": 3.528434344224253e-05, "loss": 0.4085, "step": 370 }, { "epoch": 1.5027848101265824, "grad_norm": 0.3531301184542502, "learning_rate": 3.524767492410464e-05, "loss": 0.3969, "step": 371 }, { "epoch": 1.5068354430379747, "grad_norm": 0.4506477954354387, "learning_rate": 3.521088360328908e-05, "loss": 0.4071, "step": 372 }, { "epoch": 1.510886075949367, "grad_norm": 0.4065446330772298, "learning_rate": 3.517396977610811e-05, "loss": 0.3975, "step": 373 }, { "epoch": 1.5149367088607595, "grad_norm": 0.36831991514046797, "learning_rate": 3.5136933739860595e-05, "loss": 0.3909, "step": 374 }, { "epoch": 1.518987341772152, "grad_norm": 0.4166318966183373, "learning_rate": 3.509977579282971e-05, "loss": 0.3951, "step": 375 }, { "epoch": 1.5230379746835443, "grad_norm": 0.37454501137857504, "learning_rate": 3.5062496234280424e-05, "loss": 0.3976, "step": 376 }, { "epoch": 1.5270886075949366, "grad_norm": 0.431805506409059, "learning_rate": 3.502509536445719e-05, "loss": 0.41, "step": 377 }, { "epoch": 1.531139240506329, "grad_norm": 0.35917886076413286, "learning_rate": 3.498757348458147e-05, "loss": 0.4026, "step": 378 }, { "epoch": 1.5351898734177216, "grad_norm": 0.4236357622698421, "learning_rate": 3.4949930896849324e-05, "loss": 0.3986, "step": 379 }, { "epoch": 1.539240506329114, "grad_norm": 0.4234974146795041, "learning_rate": 3.491216790442899e-05, "loss": 0.3905, "step": 380 }, { "epoch": 1.5432911392405062, "grad_norm": 0.38549415918883256, "learning_rate": 3.487428481145839e-05, "loss": 0.3983, "step": 381 }, { "epoch": 1.5473417721518987, "grad_norm": 0.40013758701752217, "learning_rate": 3.483628192304278e-05, "loss": 0.404, "step": 382 }, { "epoch": 1.5513924050632912, "grad_norm": 0.5473185783889324, "learning_rate": 3.479815954525219e-05, "loss": 0.4088, "step": 383 }, { "epoch": 1.5554430379746835, "grad_norm": 0.4324732635708064, "learning_rate": 3.475991798511899e-05, "loss": 0.3958, "step": 384 }, { "epoch": 1.5594936708860758, "grad_norm": 0.4359035762646964, "learning_rate": 3.4721557550635464e-05, "loss": 0.4118, "step": 385 }, { "epoch": 1.5635443037974683, "grad_norm": 0.526432519493724, "learning_rate": 3.468307855075128e-05, "loss": 0.4019, "step": 386 }, { "epoch": 1.5675949367088609, "grad_norm": 0.49680873044895546, "learning_rate": 3.4644481295371005e-05, "loss": 0.4062, "step": 387 }, { "epoch": 1.5716455696202531, "grad_norm": 0.40636094054544974, "learning_rate": 3.460576609535163e-05, "loss": 0.4072, "step": 388 }, { "epoch": 1.5756962025316454, "grad_norm": 0.4581392649215959, "learning_rate": 3.456693326250006e-05, "loss": 0.4058, "step": 389 }, { "epoch": 1.579746835443038, "grad_norm": 0.45054545932296286, "learning_rate": 3.452798310957058e-05, "loss": 0.3967, "step": 390 }, { "epoch": 1.5837974683544305, "grad_norm": 1.213395788947817, "learning_rate": 3.4488915950262386e-05, "loss": 0.4099, "step": 391 }, { "epoch": 1.5878481012658228, "grad_norm": 0.4466310123177808, "learning_rate": 3.4449732099216985e-05, "loss": 0.4043, "step": 392 }, { "epoch": 1.591898734177215, "grad_norm": 0.3480542669889557, "learning_rate": 3.441043187201574e-05, "loss": 0.4047, "step": 393 }, { "epoch": 1.5959493670886076, "grad_norm": 0.3100746361160821, "learning_rate": 3.437101558517728e-05, "loss": 0.4006, "step": 394 }, { "epoch": 1.6, "grad_norm": 0.3417533735017672, "learning_rate": 3.433148355615496e-05, "loss": 0.3958, "step": 395 }, { "epoch": 1.6040506329113924, "grad_norm": 0.3002694833144843, "learning_rate": 3.4291836103334294e-05, "loss": 0.4039, "step": 396 }, { "epoch": 1.6081012658227847, "grad_norm": 0.3435190043967952, "learning_rate": 3.425207354603043e-05, "loss": 0.4145, "step": 397 }, { "epoch": 1.6121518987341772, "grad_norm": 0.3686988847746781, "learning_rate": 3.421219620448553e-05, "loss": 0.4059, "step": 398 }, { "epoch": 1.6162025316455697, "grad_norm": 0.40094660723346986, "learning_rate": 3.417220439986623e-05, "loss": 0.4048, "step": 399 }, { "epoch": 1.620253164556962, "grad_norm": 0.2957181699497376, "learning_rate": 3.4132098454261024e-05, "loss": 0.39, "step": 400 }, { "epoch": 1.6243037974683543, "grad_norm": 0.46668111514791805, "learning_rate": 3.4091878690677676e-05, "loss": 0.409, "step": 401 }, { "epoch": 1.6283544303797468, "grad_norm": 0.35834298275586185, "learning_rate": 3.405154543304065e-05, "loss": 0.4056, "step": 402 }, { "epoch": 1.6324050632911393, "grad_norm": 0.47542677714447734, "learning_rate": 3.401109900618843e-05, "loss": 0.3932, "step": 403 }, { "epoch": 1.6364556962025316, "grad_norm": 0.5201224553862324, "learning_rate": 3.3970539735870996e-05, "loss": 0.4037, "step": 404 }, { "epoch": 1.640506329113924, "grad_norm": 0.36365690579108495, "learning_rate": 3.392986794874714e-05, "loss": 0.396, "step": 405 }, { "epoch": 1.6445569620253164, "grad_norm": 0.5184169380486937, "learning_rate": 3.388908397238184e-05, "loss": 0.3792, "step": 406 }, { "epoch": 1.648607594936709, "grad_norm": 0.3597581262111428, "learning_rate": 3.384818813524362e-05, "loss": 0.3995, "step": 407 }, { "epoch": 1.6526582278481012, "grad_norm": 0.5076987299078384, "learning_rate": 3.380718076670195e-05, "loss": 0.3973, "step": 408 }, { "epoch": 1.6567088607594935, "grad_norm": 0.42413434164724434, "learning_rate": 3.376606219702454e-05, "loss": 0.393, "step": 409 }, { "epoch": 1.660759493670886, "grad_norm": 0.40416385891770895, "learning_rate": 3.372483275737468e-05, "loss": 0.3928, "step": 410 }, { "epoch": 1.6648101265822786, "grad_norm": 0.3789074289242101, "learning_rate": 3.368349277980861e-05, "loss": 0.4058, "step": 411 }, { "epoch": 1.6688607594936709, "grad_norm": 0.4179003030171004, "learning_rate": 3.3642042597272844e-05, "loss": 0.4082, "step": 412 }, { "epoch": 1.6729113924050631, "grad_norm": 0.31196166197507835, "learning_rate": 3.360048254360144e-05, "loss": 0.3959, "step": 413 }, { "epoch": 1.6769620253164557, "grad_norm": 0.4202764396925827, "learning_rate": 3.355881295351336e-05, "loss": 0.4027, "step": 414 }, { "epoch": 1.6810126582278482, "grad_norm": 0.3775516596831335, "learning_rate": 3.351703416260975e-05, "loss": 0.3975, "step": 415 }, { "epoch": 1.6850632911392405, "grad_norm": 0.3882379835961418, "learning_rate": 3.347514650737126e-05, "loss": 0.395, "step": 416 }, { "epoch": 1.6891139240506328, "grad_norm": 0.4009864677372137, "learning_rate": 3.3433150325155295e-05, "loss": 0.3952, "step": 417 }, { "epoch": 1.6931645569620253, "grad_norm": 0.36808041267418495, "learning_rate": 3.339104595419334e-05, "loss": 0.392, "step": 418 }, { "epoch": 1.6972151898734178, "grad_norm": 0.3723788781474722, "learning_rate": 3.3348833733588204e-05, "loss": 0.404, "step": 419 }, { "epoch": 1.70126582278481, "grad_norm": 0.359919096192835, "learning_rate": 3.3306514003311305e-05, "loss": 0.3723, "step": 420 }, { "epoch": 1.7053164556962024, "grad_norm": 0.31561320651194413, "learning_rate": 3.326408710419996e-05, "loss": 0.3983, "step": 421 }, { "epoch": 1.709367088607595, "grad_norm": 0.3101146476076507, "learning_rate": 3.322155337795454e-05, "loss": 0.3915, "step": 422 }, { "epoch": 1.7134177215189874, "grad_norm": 0.36550332245227574, "learning_rate": 3.317891316713587e-05, "loss": 0.403, "step": 423 }, { "epoch": 1.7174683544303797, "grad_norm": 0.3538721871393044, "learning_rate": 3.313616681516231e-05, "loss": 0.3895, "step": 424 }, { "epoch": 1.721518987341772, "grad_norm": 0.3047310996112344, "learning_rate": 3.309331466630713e-05, "loss": 0.3976, "step": 425 }, { "epoch": 1.7255696202531645, "grad_norm": 0.3718213850792479, "learning_rate": 3.305035706569563e-05, "loss": 0.3888, "step": 426 }, { "epoch": 1.729620253164557, "grad_norm": 0.3579961243640536, "learning_rate": 3.3007294359302433e-05, "loss": 0.3945, "step": 427 }, { "epoch": 1.7336708860759493, "grad_norm": 0.37346659681753896, "learning_rate": 3.296412689394864e-05, "loss": 0.3866, "step": 428 }, { "epoch": 1.7377215189873416, "grad_norm": 0.3693690017196737, "learning_rate": 3.292085501729909e-05, "loss": 0.4035, "step": 429 }, { "epoch": 1.7417721518987341, "grad_norm": 0.33883595369478886, "learning_rate": 3.2877479077859534e-05, "loss": 0.4109, "step": 430 }, { "epoch": 1.7458227848101266, "grad_norm": 0.3806458727165953, "learning_rate": 3.283399942497381e-05, "loss": 0.3991, "step": 431 }, { "epoch": 1.749873417721519, "grad_norm": 0.3692009842774455, "learning_rate": 3.279041640882108e-05, "loss": 0.3913, "step": 432 }, { "epoch": 1.7539240506329112, "grad_norm": 0.3024191514953693, "learning_rate": 3.2746730380412964e-05, "loss": 0.3856, "step": 433 }, { "epoch": 1.7579746835443038, "grad_norm": 0.35701472801271095, "learning_rate": 3.2702941691590726e-05, "loss": 0.3894, "step": 434 }, { "epoch": 1.7620253164556963, "grad_norm": 0.29106071042396975, "learning_rate": 3.265905069502244e-05, "loss": 0.405, "step": 435 }, { "epoch": 1.7660759493670886, "grad_norm": 0.326475626831347, "learning_rate": 3.261505774420016e-05, "loss": 0.386, "step": 436 }, { "epoch": 1.7701265822784809, "grad_norm": 0.28014879093130124, "learning_rate": 3.257096319343707e-05, "loss": 0.3991, "step": 437 }, { "epoch": 1.7741772151898734, "grad_norm": 0.307343430783666, "learning_rate": 3.2526767397864614e-05, "loss": 0.4019, "step": 438 }, { "epoch": 1.7782278481012659, "grad_norm": 0.31203063708670065, "learning_rate": 3.248247071342966e-05, "loss": 0.4034, "step": 439 }, { "epoch": 1.7822784810126582, "grad_norm": 0.3140796384874758, "learning_rate": 3.243807349689161e-05, "loss": 0.4005, "step": 440 }, { "epoch": 1.7863291139240505, "grad_norm": 0.3831238092948371, "learning_rate": 3.2393576105819544e-05, "loss": 0.3957, "step": 441 }, { "epoch": 1.790379746835443, "grad_norm": 0.31511280841053807, "learning_rate": 3.2348978898589333e-05, "loss": 0.4009, "step": 442 }, { "epoch": 1.7944303797468355, "grad_norm": 0.33844434775774024, "learning_rate": 3.230428223438075e-05, "loss": 0.3958, "step": 443 }, { "epoch": 1.7984810126582278, "grad_norm": 0.33638565981328356, "learning_rate": 3.225948647317459e-05, "loss": 0.3996, "step": 444 }, { "epoch": 1.80253164556962, "grad_norm": 0.3190502223132732, "learning_rate": 3.2214591975749745e-05, "loss": 0.402, "step": 445 }, { "epoch": 1.8065822784810126, "grad_norm": 0.36576101855834015, "learning_rate": 3.216959910368034e-05, "loss": 0.4057, "step": 446 }, { "epoch": 1.8106329113924051, "grad_norm": 0.2917335017625684, "learning_rate": 3.212450821933277e-05, "loss": 0.384, "step": 447 }, { "epoch": 1.8146835443037974, "grad_norm": 0.3537330112056853, "learning_rate": 3.207931968586281e-05, "loss": 0.4026, "step": 448 }, { "epoch": 1.81873417721519, "grad_norm": 0.3026430961938655, "learning_rate": 3.203403386721272e-05, "loss": 0.393, "step": 449 }, { "epoch": 1.8227848101265822, "grad_norm": 0.2981917031010366, "learning_rate": 3.1988651128108245e-05, "loss": 0.3946, "step": 450 }, { "epoch": 1.8268354430379747, "grad_norm": 0.3028319398902521, "learning_rate": 3.194317183405573e-05, "loss": 0.3914, "step": 451 }, { "epoch": 1.8308860759493673, "grad_norm": 0.372431253574424, "learning_rate": 3.189759635133914e-05, "loss": 0.397, "step": 452 }, { "epoch": 1.8349367088607595, "grad_norm": 0.33739547513875423, "learning_rate": 3.185192504701718e-05, "loss": 0.3928, "step": 453 }, { "epoch": 1.8389873417721518, "grad_norm": 0.32026945962913894, "learning_rate": 3.1806158288920234e-05, "loss": 0.401, "step": 454 }, { "epoch": 1.8430379746835444, "grad_norm": 0.32986549276732113, "learning_rate": 3.1760296445647477e-05, "loss": 0.3977, "step": 455 }, { "epoch": 1.8470886075949369, "grad_norm": 0.31087317700082884, "learning_rate": 3.1714339886563896e-05, "loss": 0.3887, "step": 456 }, { "epoch": 1.8511392405063292, "grad_norm": 0.3422253674714748, "learning_rate": 3.166828898179731e-05, "loss": 0.3931, "step": 457 }, { "epoch": 1.8551898734177215, "grad_norm": 0.3922914430207445, "learning_rate": 3.162214410223536e-05, "loss": 0.3992, "step": 458 }, { "epoch": 1.859240506329114, "grad_norm": 0.3382760335485976, "learning_rate": 3.157590561952257e-05, "loss": 0.3817, "step": 459 }, { "epoch": 1.8632911392405065, "grad_norm": 0.34607767415461604, "learning_rate": 3.152957390605732e-05, "loss": 0.3947, "step": 460 }, { "epoch": 1.8673417721518988, "grad_norm": 0.42755023854204455, "learning_rate": 3.148314933498886e-05, "loss": 0.3933, "step": 461 }, { "epoch": 1.871392405063291, "grad_norm": 0.3881214990359343, "learning_rate": 3.143663228021431e-05, "loss": 0.4051, "step": 462 }, { "epoch": 1.8754430379746836, "grad_norm": 0.4426829426196495, "learning_rate": 3.1390023116375624e-05, "loss": 0.3929, "step": 463 }, { "epoch": 1.879493670886076, "grad_norm": 0.30776224037318045, "learning_rate": 3.134332221885661e-05, "loss": 0.3985, "step": 464 }, { "epoch": 1.8835443037974684, "grad_norm": 0.4255776447564245, "learning_rate": 3.129652996377987e-05, "loss": 0.398, "step": 465 }, { "epoch": 1.8875949367088607, "grad_norm": 0.36966355226981457, "learning_rate": 3.12496467280038e-05, "loss": 0.3815, "step": 466 }, { "epoch": 1.8916455696202532, "grad_norm": 0.3869871476109576, "learning_rate": 3.120267288911952e-05, "loss": 0.3806, "step": 467 }, { "epoch": 1.8956962025316457, "grad_norm": 0.32095283464785657, "learning_rate": 3.11556088254479e-05, "loss": 0.4055, "step": 468 }, { "epoch": 1.899746835443038, "grad_norm": 0.39722743537442784, "learning_rate": 3.11084549160364e-05, "loss": 0.3843, "step": 469 }, { "epoch": 1.9037974683544303, "grad_norm": 0.32758666816333826, "learning_rate": 3.106121154065615e-05, "loss": 0.3986, "step": 470 }, { "epoch": 1.9078481012658228, "grad_norm": 0.3348042654957808, "learning_rate": 3.1013879079798805e-05, "loss": 0.3894, "step": 471 }, { "epoch": 1.9118987341772153, "grad_norm": 0.3325078939909019, "learning_rate": 3.096645791467348e-05, "loss": 0.3957, "step": 472 }, { "epoch": 1.9159493670886076, "grad_norm": 0.29764741655951993, "learning_rate": 3.091894842720373e-05, "loss": 0.4031, "step": 473 }, { "epoch": 1.92, "grad_norm": 0.36861541304325757, "learning_rate": 3.0871351000024425e-05, "loss": 0.3932, "step": 474 }, { "epoch": 1.9240506329113924, "grad_norm": 0.33513920054283197, "learning_rate": 3.0823666016478716e-05, "loss": 0.4061, "step": 475 }, { "epoch": 1.928101265822785, "grad_norm": 0.3794245467250986, "learning_rate": 3.0775893860614896e-05, "loss": 0.3898, "step": 476 }, { "epoch": 1.9321518987341773, "grad_norm": 0.3584848454063876, "learning_rate": 3.0728034917183336e-05, "loss": 0.4066, "step": 477 }, { "epoch": 1.9362025316455695, "grad_norm": 0.3463222015856723, "learning_rate": 3.06800895716334e-05, "loss": 0.3924, "step": 478 }, { "epoch": 1.940253164556962, "grad_norm": 0.3477137350686893, "learning_rate": 3.063205821011029e-05, "loss": 0.3958, "step": 479 }, { "epoch": 1.9443037974683546, "grad_norm": 0.32035678498243914, "learning_rate": 3.0583941219452016e-05, "loss": 0.3944, "step": 480 }, { "epoch": 1.9483544303797469, "grad_norm": 0.3612000350176266, "learning_rate": 3.053573898718618e-05, "loss": 0.3904, "step": 481 }, { "epoch": 1.9524050632911392, "grad_norm": 0.29134764833259225, "learning_rate": 3.0487451901526956e-05, "loss": 0.4088, "step": 482 }, { "epoch": 1.9564556962025317, "grad_norm": 0.4872036540012547, "learning_rate": 3.0439080351371875e-05, "loss": 0.4046, "step": 483 }, { "epoch": 1.9605063291139242, "grad_norm": 0.29279666081722633, "learning_rate": 3.0390624726298764e-05, "loss": 0.3816, "step": 484 }, { "epoch": 1.9645569620253165, "grad_norm": 0.46273676339174363, "learning_rate": 3.034208541656255e-05, "loss": 0.4015, "step": 485 }, { "epoch": 1.9686075949367088, "grad_norm": 0.3252953860319193, "learning_rate": 3.029346281309218e-05, "loss": 0.3814, "step": 486 }, { "epoch": 1.9726582278481013, "grad_norm": 0.40561104257245206, "learning_rate": 3.0244757307487415e-05, "loss": 0.3979, "step": 487 }, { "epoch": 1.9767088607594938, "grad_norm": 0.32681292879942125, "learning_rate": 3.019596929201569e-05, "loss": 0.3996, "step": 488 }, { "epoch": 1.980759493670886, "grad_norm": 0.3999220643968569, "learning_rate": 3.0147099159608985e-05, "loss": 0.3824, "step": 489 }, { "epoch": 1.9848101265822784, "grad_norm": 0.3103885548003107, "learning_rate": 3.0098147303860616e-05, "loss": 0.3926, "step": 490 }, { "epoch": 1.988860759493671, "grad_norm": 0.36053251843900147, "learning_rate": 3.0049114119022117e-05, "loss": 0.3961, "step": 491 }, { "epoch": 1.9929113924050634, "grad_norm": 0.36199818297638436, "learning_rate": 3.0000000000000004e-05, "loss": 0.3958, "step": 492 }, { "epoch": 1.9969620253164557, "grad_norm": 0.3318964329452759, "learning_rate": 2.995080534235264e-05, "loss": 0.3839, "step": 493 }, { "epoch": 2.001012658227848, "grad_norm": 0.3572508270703928, "learning_rate": 2.9901530542287044e-05, "loss": 0.3893, "step": 494 }, { "epoch": 2.0050632911392405, "grad_norm": 0.4052574997186816, "learning_rate": 2.9852175996655676e-05, "loss": 0.355, "step": 495 }, { "epoch": 2.009113924050633, "grad_norm": 0.6613035363747818, "learning_rate": 2.980274210295326e-05, "loss": 0.3494, "step": 496 }, { "epoch": 2.013164556962025, "grad_norm": 0.5666188874056033, "learning_rate": 2.9753229259313578e-05, "loss": 0.3547, "step": 497 }, { "epoch": 2.0172151898734176, "grad_norm": 0.49382183594069123, "learning_rate": 2.9703637864506274e-05, "loss": 0.3509, "step": 498 }, { "epoch": 2.02126582278481, "grad_norm": 0.4773302791296366, "learning_rate": 2.965396831793362e-05, "loss": 0.3444, "step": 499 }, { "epoch": 2.0253164556962027, "grad_norm": 0.46951189142809846, "learning_rate": 2.9604221019627316e-05, "loss": 0.3504, "step": 500 }, { "epoch": 2.0293670886075947, "grad_norm": 0.3992025093760167, "learning_rate": 2.955439637024526e-05, "loss": 0.3395, "step": 501 }, { "epoch": 2.0334177215189873, "grad_norm": 0.38500799263858704, "learning_rate": 2.9504494771068334e-05, "loss": 0.3351, "step": 502 }, { "epoch": 2.0374683544303798, "grad_norm": 0.47505061616408917, "learning_rate": 2.9454516623997156e-05, "loss": 0.3319, "step": 503 }, { "epoch": 2.0415189873417723, "grad_norm": 0.3488703905774492, "learning_rate": 2.9404462331548847e-05, "loss": 0.3521, "step": 504 }, { "epoch": 2.0455696202531644, "grad_norm": 0.4726281788351472, "learning_rate": 2.93543322968538e-05, "loss": 0.3331, "step": 505 }, { "epoch": 2.049620253164557, "grad_norm": 0.38812629186707953, "learning_rate": 2.9304126923652428e-05, "loss": 0.3375, "step": 506 }, { "epoch": 2.0536708860759494, "grad_norm": 0.4441837575800042, "learning_rate": 2.9253846616291896e-05, "loss": 0.3421, "step": 507 }, { "epoch": 2.057721518987342, "grad_norm": 0.3928678221625555, "learning_rate": 2.9203491779722896e-05, "loss": 0.3259, "step": 508 }, { "epoch": 2.061772151898734, "grad_norm": 0.41456788529074284, "learning_rate": 2.9153062819496357e-05, "loss": 0.3427, "step": 509 }, { "epoch": 2.0658227848101265, "grad_norm": 0.35320860539363624, "learning_rate": 2.9102560141760178e-05, "loss": 0.3423, "step": 510 }, { "epoch": 2.069873417721519, "grad_norm": 0.37640853184549883, "learning_rate": 2.9051984153256004e-05, "loss": 0.3478, "step": 511 }, { "epoch": 2.0739240506329115, "grad_norm": 0.34780731296543754, "learning_rate": 2.900133526131588e-05, "loss": 0.3484, "step": 512 }, { "epoch": 2.0779746835443036, "grad_norm": 0.34001659363465436, "learning_rate": 2.8950613873859025e-05, "loss": 0.3529, "step": 513 }, { "epoch": 2.082025316455696, "grad_norm": 0.2917613829850685, "learning_rate": 2.8899820399388515e-05, "loss": 0.335, "step": 514 }, { "epoch": 2.0860759493670886, "grad_norm": 0.3409178054467281, "learning_rate": 2.8848955246988012e-05, "loss": 0.3268, "step": 515 }, { "epoch": 2.090126582278481, "grad_norm": 0.2866135660026131, "learning_rate": 2.879801882631847e-05, "loss": 0.3488, "step": 516 }, { "epoch": 2.094177215189873, "grad_norm": 0.36722798926710076, "learning_rate": 2.8747011547614808e-05, "loss": 0.3529, "step": 517 }, { "epoch": 2.0982278481012657, "grad_norm": 0.3236516176273472, "learning_rate": 2.8695933821682635e-05, "loss": 0.367, "step": 518 }, { "epoch": 2.1022784810126582, "grad_norm": 0.37526656911772444, "learning_rate": 2.864478605989494e-05, "loss": 0.352, "step": 519 }, { "epoch": 2.1063291139240508, "grad_norm": 0.3236872258042403, "learning_rate": 2.8593568674188765e-05, "loss": 0.3417, "step": 520 }, { "epoch": 2.110379746835443, "grad_norm": 0.2865912915216018, "learning_rate": 2.8542282077061892e-05, "loss": 0.3428, "step": 521 }, { "epoch": 2.1144303797468353, "grad_norm": 0.295460415291979, "learning_rate": 2.8490926681569523e-05, "loss": 0.34, "step": 522 }, { "epoch": 2.118481012658228, "grad_norm": 0.3095730541943112, "learning_rate": 2.8439502901320956e-05, "loss": 0.3416, "step": 523 }, { "epoch": 2.1225316455696204, "grad_norm": 0.31894161159403156, "learning_rate": 2.8388011150476237e-05, "loss": 0.3479, "step": 524 }, { "epoch": 2.1265822784810124, "grad_norm": 0.26968410917274976, "learning_rate": 2.8336451843742866e-05, "loss": 0.3397, "step": 525 }, { "epoch": 2.130632911392405, "grad_norm": 0.318400524258485, "learning_rate": 2.8284825396372387e-05, "loss": 0.3415, "step": 526 }, { "epoch": 2.1346835443037975, "grad_norm": 0.28404173496093293, "learning_rate": 2.8233132224157132e-05, "loss": 0.3447, "step": 527 }, { "epoch": 2.13873417721519, "grad_norm": 0.37585954245967457, "learning_rate": 2.8181372743426805e-05, "loss": 0.3309, "step": 528 }, { "epoch": 2.1427848101265825, "grad_norm": 0.27945916198372567, "learning_rate": 2.8129547371045128e-05, "loss": 0.3368, "step": 529 }, { "epoch": 2.1468354430379746, "grad_norm": 0.3446829315872565, "learning_rate": 2.8077656524406534e-05, "loss": 0.3487, "step": 530 }, { "epoch": 2.150886075949367, "grad_norm": 0.30617831300356785, "learning_rate": 2.802570062143278e-05, "loss": 0.3444, "step": 531 }, { "epoch": 2.1549367088607596, "grad_norm": 0.3204649029971748, "learning_rate": 2.7973680080569555e-05, "loss": 0.3522, "step": 532 }, { "epoch": 2.1589873417721517, "grad_norm": 0.3021823865413493, "learning_rate": 2.792159532078314e-05, "loss": 0.3397, "step": 533 }, { "epoch": 2.163037974683544, "grad_norm": 0.30371247604501617, "learning_rate": 2.7869446761557033e-05, "loss": 0.357, "step": 534 }, { "epoch": 2.1670886075949367, "grad_norm": 0.32741924272820105, "learning_rate": 2.781723482288857e-05, "loss": 0.345, "step": 535 }, { "epoch": 2.1711392405063292, "grad_norm": 0.2935587323623186, "learning_rate": 2.7764959925285517e-05, "loss": 0.3495, "step": 536 }, { "epoch": 2.1751898734177217, "grad_norm": 0.26395047860240317, "learning_rate": 2.771262248976272e-05, "loss": 0.3423, "step": 537 }, { "epoch": 2.179240506329114, "grad_norm": 0.3373561106320075, "learning_rate": 2.7660222937838677e-05, "loss": 0.3529, "step": 538 }, { "epoch": 2.1832911392405063, "grad_norm": 0.26628245472614404, "learning_rate": 2.7607761691532186e-05, "loss": 0.3448, "step": 539 }, { "epoch": 2.187341772151899, "grad_norm": 0.345088379313994, "learning_rate": 2.7555239173358916e-05, "loss": 0.3507, "step": 540 }, { "epoch": 2.191392405063291, "grad_norm": 0.30038056146184966, "learning_rate": 2.7502655806328e-05, "loss": 0.3401, "step": 541 }, { "epoch": 2.1954430379746834, "grad_norm": 0.30877590433288504, "learning_rate": 2.7450012013938648e-05, "loss": 0.3459, "step": 542 }, { "epoch": 2.199493670886076, "grad_norm": 0.34816882887005496, "learning_rate": 2.739730822017673e-05, "loss": 0.3493, "step": 543 }, { "epoch": 2.2035443037974685, "grad_norm": 0.29473816929502683, "learning_rate": 2.7344544849511355e-05, "loss": 0.3571, "step": 544 }, { "epoch": 2.207594936708861, "grad_norm": 0.3365597316065496, "learning_rate": 2.7291722326891456e-05, "loss": 0.3435, "step": 545 }, { "epoch": 2.211645569620253, "grad_norm": 0.26370074471953026, "learning_rate": 2.723884107774236e-05, "loss": 0.3406, "step": 546 }, { "epoch": 2.2156962025316456, "grad_norm": 0.3235363587973158, "learning_rate": 2.718590152796239e-05, "loss": 0.3495, "step": 547 }, { "epoch": 2.219746835443038, "grad_norm": 0.27737507798765904, "learning_rate": 2.71329041039194e-05, "loss": 0.3402, "step": 548 }, { "epoch": 2.22379746835443, "grad_norm": 0.3015520557631859, "learning_rate": 2.7079849232447357e-05, "loss": 0.3488, "step": 549 }, { "epoch": 2.2278481012658227, "grad_norm": 0.28373286303876527, "learning_rate": 2.7026737340842895e-05, "loss": 0.3388, "step": 550 }, { "epoch": 2.231898734177215, "grad_norm": 0.32995169061768687, "learning_rate": 2.697356885686189e-05, "loss": 0.3457, "step": 551 }, { "epoch": 2.2359493670886077, "grad_norm": 0.3003085184608397, "learning_rate": 2.6920344208716014e-05, "loss": 0.3334, "step": 552 }, { "epoch": 2.24, "grad_norm": 0.3040236863651046, "learning_rate": 2.6867063825069252e-05, "loss": 0.3405, "step": 553 }, { "epoch": 2.2440506329113923, "grad_norm": 0.285984633091963, "learning_rate": 2.6813728135034494e-05, "loss": 0.339, "step": 554 }, { "epoch": 2.248101265822785, "grad_norm": 0.31971632053610094, "learning_rate": 2.6760337568170056e-05, "loss": 0.3522, "step": 555 }, { "epoch": 2.2521518987341773, "grad_norm": 0.4109832153977163, "learning_rate": 2.6706892554476226e-05, "loss": 0.3522, "step": 556 }, { "epoch": 2.2562025316455694, "grad_norm": 0.3422206018385529, "learning_rate": 2.6653393524391795e-05, "loss": 0.3519, "step": 557 }, { "epoch": 2.260253164556962, "grad_norm": 0.3539786509320766, "learning_rate": 2.6599840908790592e-05, "loss": 0.3472, "step": 558 }, { "epoch": 2.2643037974683544, "grad_norm": 0.3416000828801056, "learning_rate": 2.6546235138978028e-05, "loss": 0.34, "step": 559 }, { "epoch": 2.268354430379747, "grad_norm": 0.34418850419088554, "learning_rate": 2.6492576646687597e-05, "loss": 0.3448, "step": 560 }, { "epoch": 2.2724050632911394, "grad_norm": 0.30822217513518685, "learning_rate": 2.6438865864077425e-05, "loss": 0.3455, "step": 561 }, { "epoch": 2.2764556962025315, "grad_norm": 0.34213763918984796, "learning_rate": 2.6385103223726766e-05, "loss": 0.3553, "step": 562 }, { "epoch": 2.280506329113924, "grad_norm": 0.32322428265614167, "learning_rate": 2.6331289158632537e-05, "loss": 0.3591, "step": 563 }, { "epoch": 2.2845569620253166, "grad_norm": 0.3607360296681508, "learning_rate": 2.6277424102205817e-05, "loss": 0.3401, "step": 564 }, { "epoch": 2.2886075949367086, "grad_norm": 0.2959552370540033, "learning_rate": 2.6223508488268374e-05, "loss": 0.3468, "step": 565 }, { "epoch": 2.292658227848101, "grad_norm": 0.3303075201206515, "learning_rate": 2.6169542751049148e-05, "loss": 0.3464, "step": 566 }, { "epoch": 2.2967088607594937, "grad_norm": 0.2963894722796868, "learning_rate": 2.6115527325180754e-05, "loss": 0.3484, "step": 567 }, { "epoch": 2.300759493670886, "grad_norm": 0.3494622872420442, "learning_rate": 2.606146264569603e-05, "loss": 0.3533, "step": 568 }, { "epoch": 2.3048101265822787, "grad_norm": 0.2937261652750821, "learning_rate": 2.6007349148024447e-05, "loss": 0.345, "step": 569 }, { "epoch": 2.3088607594936708, "grad_norm": 0.34337656540229644, "learning_rate": 2.5953187267988694e-05, "loss": 0.3445, "step": 570 }, { "epoch": 2.3129113924050633, "grad_norm": 0.2967203252863564, "learning_rate": 2.5898977441801097e-05, "loss": 0.3506, "step": 571 }, { "epoch": 2.316962025316456, "grad_norm": 0.403431012395974, "learning_rate": 2.584472010606015e-05, "loss": 0.3506, "step": 572 }, { "epoch": 2.321012658227848, "grad_norm": 0.3059399331880436, "learning_rate": 2.5790415697746976e-05, "loss": 0.3511, "step": 573 }, { "epoch": 2.3250632911392404, "grad_norm": 0.30881978194521004, "learning_rate": 2.5736064654221808e-05, "loss": 0.3265, "step": 574 }, { "epoch": 2.329113924050633, "grad_norm": 0.33972147234562927, "learning_rate": 2.568166741322048e-05, "loss": 0.3521, "step": 575 }, { "epoch": 2.3331645569620254, "grad_norm": 0.31147806224714136, "learning_rate": 2.56272244128509e-05, "loss": 0.3592, "step": 576 }, { "epoch": 2.337215189873418, "grad_norm": 0.30495445193579995, "learning_rate": 2.55727360915895e-05, "loss": 0.3439, "step": 577 }, { "epoch": 2.34126582278481, "grad_norm": 0.26209331992238594, "learning_rate": 2.5518202888277734e-05, "loss": 0.3408, "step": 578 }, { "epoch": 2.3453164556962025, "grad_norm": 0.28675635757200535, "learning_rate": 2.5463625242118523e-05, "loss": 0.3482, "step": 579 }, { "epoch": 2.349367088607595, "grad_norm": 0.3125659298248449, "learning_rate": 2.5409003592672723e-05, "loss": 0.3578, "step": 580 }, { "epoch": 2.353417721518987, "grad_norm": 0.2772674945237436, "learning_rate": 2.535433837985559e-05, "loss": 0.3495, "step": 581 }, { "epoch": 2.3574683544303796, "grad_norm": 0.2625286034944694, "learning_rate": 2.529963004393324e-05, "loss": 0.3557, "step": 582 }, { "epoch": 2.361518987341772, "grad_norm": 0.2911391174488446, "learning_rate": 2.524487902551908e-05, "loss": 0.3512, "step": 583 }, { "epoch": 2.3655696202531646, "grad_norm": 0.24119958076285855, "learning_rate": 2.519008576557029e-05, "loss": 0.3521, "step": 584 }, { "epoch": 2.369620253164557, "grad_norm": 0.26515298379405144, "learning_rate": 2.5135250705384254e-05, "loss": 0.3369, "step": 585 }, { "epoch": 2.3736708860759492, "grad_norm": 0.2846736661965167, "learning_rate": 2.5080374286595007e-05, "loss": 0.3428, "step": 586 }, { "epoch": 2.3777215189873417, "grad_norm": 0.2772721835212197, "learning_rate": 2.5025456951169677e-05, "loss": 0.3539, "step": 587 }, { "epoch": 2.3817721518987343, "grad_norm": 0.3092943926788682, "learning_rate": 2.4970499141404942e-05, "loss": 0.3361, "step": 588 }, { "epoch": 2.3858227848101268, "grad_norm": 0.2610319260139128, "learning_rate": 2.491550129992345e-05, "loss": 0.3543, "step": 589 }, { "epoch": 2.389873417721519, "grad_norm": 0.27276365705248945, "learning_rate": 2.486046386967024e-05, "loss": 0.3471, "step": 590 }, { "epoch": 2.3939240506329114, "grad_norm": 0.27089537376724787, "learning_rate": 2.4805387293909214e-05, "loss": 0.3497, "step": 591 }, { "epoch": 2.397974683544304, "grad_norm": 0.30121144645653036, "learning_rate": 2.4750272016219552e-05, "loss": 0.3571, "step": 592 }, { "epoch": 2.4020253164556964, "grad_norm": 0.2675612478322901, "learning_rate": 2.4695118480492114e-05, "loss": 0.3524, "step": 593 }, { "epoch": 2.4060759493670885, "grad_norm": 0.29028343333990486, "learning_rate": 2.4639927130925898e-05, "loss": 0.3555, "step": 594 }, { "epoch": 2.410126582278481, "grad_norm": 0.2710716036940737, "learning_rate": 2.458469841202444e-05, "loss": 0.3359, "step": 595 }, { "epoch": 2.4141772151898735, "grad_norm": 0.3181025636399892, "learning_rate": 2.452943276859226e-05, "loss": 0.3512, "step": 596 }, { "epoch": 2.418227848101266, "grad_norm": 0.28666500945275164, "learning_rate": 2.447413064573125e-05, "loss": 0.3424, "step": 597 }, { "epoch": 2.422278481012658, "grad_norm": 0.3430963206036322, "learning_rate": 2.4418792488837095e-05, "loss": 0.3522, "step": 598 }, { "epoch": 2.4263291139240506, "grad_norm": 0.2784564641882, "learning_rate": 2.4363418743595713e-05, "loss": 0.3454, "step": 599 }, { "epoch": 2.430379746835443, "grad_norm": 0.2831639550767728, "learning_rate": 2.430800985597963e-05, "loss": 0.3403, "step": 600 }, { "epoch": 2.4344303797468356, "grad_norm": 0.27547758382325876, "learning_rate": 2.4252566272244415e-05, "loss": 0.3367, "step": 601 }, { "epoch": 2.4384810126582277, "grad_norm": 0.2853393218153495, "learning_rate": 2.4197088438925063e-05, "loss": 0.3489, "step": 602 }, { "epoch": 2.44253164556962, "grad_norm": 0.28370974808830307, "learning_rate": 2.4141576802832417e-05, "loss": 0.3498, "step": 603 }, { "epoch": 2.4465822784810127, "grad_norm": 0.25403115500606416, "learning_rate": 2.408603181104957e-05, "loss": 0.3465, "step": 604 }, { "epoch": 2.4506329113924052, "grad_norm": 0.3008967022063414, "learning_rate": 2.4030453910928245e-05, "loss": 0.3552, "step": 605 }, { "epoch": 2.4546835443037973, "grad_norm": 0.3057235976081265, "learning_rate": 2.397484355008521e-05, "loss": 0.3459, "step": 606 }, { "epoch": 2.45873417721519, "grad_norm": 0.3027442724890845, "learning_rate": 2.3919201176398662e-05, "loss": 0.3394, "step": 607 }, { "epoch": 2.4627848101265823, "grad_norm": 0.29076806045131814, "learning_rate": 2.3863527238004633e-05, "loss": 0.3321, "step": 608 }, { "epoch": 2.466835443037975, "grad_norm": 0.28933585392201383, "learning_rate": 2.380782218329337e-05, "loss": 0.3471, "step": 609 }, { "epoch": 2.470886075949367, "grad_norm": 0.2813147195749776, "learning_rate": 2.3752086460905725e-05, "loss": 0.3482, "step": 610 }, { "epoch": 2.4749367088607594, "grad_norm": 0.26562160315451117, "learning_rate": 2.3696320519729544e-05, "loss": 0.3467, "step": 611 }, { "epoch": 2.478987341772152, "grad_norm": 0.31195123881244863, "learning_rate": 2.3640524808896045e-05, "loss": 0.3511, "step": 612 }, { "epoch": 2.4830379746835445, "grad_norm": 0.299435688241796, "learning_rate": 2.3584699777776222e-05, "loss": 0.3617, "step": 613 }, { "epoch": 2.4870886075949366, "grad_norm": 0.3084963480660079, "learning_rate": 2.3528845875977195e-05, "loss": 0.3475, "step": 614 }, { "epoch": 2.491139240506329, "grad_norm": 0.27758636412844423, "learning_rate": 2.3472963553338614e-05, "loss": 0.3542, "step": 615 }, { "epoch": 2.4951898734177216, "grad_norm": 0.2866618988847706, "learning_rate": 2.341705325992901e-05, "loss": 0.3626, "step": 616 }, { "epoch": 2.499240506329114, "grad_norm": 0.33613128948056825, "learning_rate": 2.336111544604222e-05, "loss": 0.3556, "step": 617 }, { "epoch": 2.503291139240506, "grad_norm": 0.277582025279553, "learning_rate": 2.33051505621937e-05, "loss": 0.3497, "step": 618 }, { "epoch": 2.5073417721518987, "grad_norm": 0.321009492224204, "learning_rate": 2.324915905911693e-05, "loss": 0.3399, "step": 619 }, { "epoch": 2.511392405063291, "grad_norm": 0.28342702519812735, "learning_rate": 2.319314138775977e-05, "loss": 0.351, "step": 620 }, { "epoch": 2.5154430379746833, "grad_norm": 0.27802440336964634, "learning_rate": 2.3137097999280856e-05, "loss": 0.342, "step": 621 }, { "epoch": 2.519493670886076, "grad_norm": 0.23976142998324954, "learning_rate": 2.308102934504593e-05, "loss": 0.3454, "step": 622 }, { "epoch": 2.5235443037974683, "grad_norm": 0.32095862992380264, "learning_rate": 2.3024935876624222e-05, "loss": 0.3416, "step": 623 }, { "epoch": 2.527594936708861, "grad_norm": 0.2780927230015669, "learning_rate": 2.2968818045784813e-05, "loss": 0.3546, "step": 624 }, { "epoch": 2.5316455696202533, "grad_norm": 0.26812730256749023, "learning_rate": 2.2912676304493006e-05, "loss": 0.342, "step": 625 }, { "epoch": 2.5356962025316454, "grad_norm": 0.2627702215673269, "learning_rate": 2.2856511104906668e-05, "loss": 0.3482, "step": 626 }, { "epoch": 2.539746835443038, "grad_norm": 0.257111990841057, "learning_rate": 2.2800322899372586e-05, "loss": 0.341, "step": 627 }, { "epoch": 2.5437974683544304, "grad_norm": 0.2490457083439576, "learning_rate": 2.2744112140422844e-05, "loss": 0.3493, "step": 628 }, { "epoch": 2.547848101265823, "grad_norm": 0.29599491761173313, "learning_rate": 2.2687879280771177e-05, "loss": 0.3523, "step": 629 }, { "epoch": 2.5518987341772155, "grad_norm": 0.2667371103129047, "learning_rate": 2.26316247733093e-05, "loss": 0.3471, "step": 630 }, { "epoch": 2.5559493670886075, "grad_norm": 0.28370173163437806, "learning_rate": 2.257534907110328e-05, "loss": 0.3527, "step": 631 }, { "epoch": 2.56, "grad_norm": 0.27994900219680485, "learning_rate": 2.2519052627389882e-05, "loss": 0.3454, "step": 632 }, { "epoch": 2.5640506329113926, "grad_norm": 0.2561654599383484, "learning_rate": 2.246273589557294e-05, "loss": 0.3481, "step": 633 }, { "epoch": 2.5681012658227846, "grad_norm": 0.2949206916660058, "learning_rate": 2.240639932921966e-05, "loss": 0.3305, "step": 634 }, { "epoch": 2.572151898734177, "grad_norm": 0.23444775980830773, "learning_rate": 2.2350043382056995e-05, "loss": 0.3495, "step": 635 }, { "epoch": 2.5762025316455697, "grad_norm": 0.31618427211741496, "learning_rate": 2.2293668507968015e-05, "loss": 0.3446, "step": 636 }, { "epoch": 2.580253164556962, "grad_norm": 0.2585703212493614, "learning_rate": 2.2237275160988186e-05, "loss": 0.3407, "step": 637 }, { "epoch": 2.5843037974683547, "grad_norm": 0.26282779381094423, "learning_rate": 2.2180863795301787e-05, "loss": 0.3385, "step": 638 }, { "epoch": 2.5883544303797468, "grad_norm": 0.27391076281603527, "learning_rate": 2.212443486523819e-05, "loss": 0.3514, "step": 639 }, { "epoch": 2.5924050632911393, "grad_norm": 0.2383952712806843, "learning_rate": 2.2067988825268243e-05, "loss": 0.3518, "step": 640 }, { "epoch": 2.596455696202532, "grad_norm": 0.23341056035917437, "learning_rate": 2.2011526130000596e-05, "loss": 0.345, "step": 641 }, { "epoch": 2.600506329113924, "grad_norm": 0.26125249931613165, "learning_rate": 2.1955047234178038e-05, "loss": 0.3503, "step": 642 }, { "epoch": 2.6045569620253164, "grad_norm": 0.2190845601964268, "learning_rate": 2.1898552592673825e-05, "loss": 0.3401, "step": 643 }, { "epoch": 2.608607594936709, "grad_norm": 0.2708661564258868, "learning_rate": 2.184204266048803e-05, "loss": 0.3406, "step": 644 }, { "epoch": 2.6126582278481014, "grad_norm": 0.25464971232127714, "learning_rate": 2.1785517892743887e-05, "loss": 0.3367, "step": 645 }, { "epoch": 2.616708860759494, "grad_norm": 0.256704449940484, "learning_rate": 2.17289787446841e-05, "loss": 0.3532, "step": 646 }, { "epoch": 2.620759493670886, "grad_norm": 0.2650136342401565, "learning_rate": 2.1672425671667198e-05, "loss": 0.3461, "step": 647 }, { "epoch": 2.6248101265822785, "grad_norm": 0.5660313590571984, "learning_rate": 2.161585912916385e-05, "loss": 0.3522, "step": 648 }, { "epoch": 2.628860759493671, "grad_norm": 0.26443309725612724, "learning_rate": 2.1559279572753214e-05, "loss": 0.3483, "step": 649 }, { "epoch": 2.632911392405063, "grad_norm": 0.25386434125552393, "learning_rate": 2.1502687458119268e-05, "loss": 0.3465, "step": 650 }, { "epoch": 2.6369620253164556, "grad_norm": 0.2887130961767785, "learning_rate": 2.1446083241047116e-05, "loss": 0.3407, "step": 651 }, { "epoch": 2.641012658227848, "grad_norm": 0.2732730607176014, "learning_rate": 2.1389467377419333e-05, "loss": 0.3471, "step": 652 }, { "epoch": 2.6450632911392407, "grad_norm": 0.2976138089129315, "learning_rate": 2.133284032321232e-05, "loss": 0.3477, "step": 653 }, { "epoch": 2.649113924050633, "grad_norm": 0.2720935382822592, "learning_rate": 2.1276202534492566e-05, "loss": 0.3355, "step": 654 }, { "epoch": 2.6531645569620252, "grad_norm": 0.2959578441352938, "learning_rate": 2.121955446741306e-05, "loss": 0.3282, "step": 655 }, { "epoch": 2.6572151898734178, "grad_norm": 0.27484986244438814, "learning_rate": 2.1162896578209517e-05, "loss": 0.3431, "step": 656 }, { "epoch": 2.6612658227848103, "grad_norm": 0.27081914622027703, "learning_rate": 2.1106229323196813e-05, "loss": 0.3507, "step": 657 }, { "epoch": 2.6653164556962023, "grad_norm": 0.30567102010367336, "learning_rate": 2.1049553158765214e-05, "loss": 0.3398, "step": 658 }, { "epoch": 2.669367088607595, "grad_norm": 0.2495280637525675, "learning_rate": 2.0992868541376764e-05, "loss": 0.3446, "step": 659 }, { "epoch": 2.6734177215189874, "grad_norm": 0.3203995508223298, "learning_rate": 2.093617592756158e-05, "loss": 0.3506, "step": 660 }, { "epoch": 2.67746835443038, "grad_norm": 0.2635994767339217, "learning_rate": 2.0879475773914167e-05, "loss": 0.3522, "step": 661 }, { "epoch": 2.6815189873417724, "grad_norm": 0.30184799864173084, "learning_rate": 2.082276853708978e-05, "loss": 0.3568, "step": 662 }, { "epoch": 2.6855696202531645, "grad_norm": 0.26914735311573884, "learning_rate": 2.076605467380071e-05, "loss": 0.3431, "step": 663 }, { "epoch": 2.689620253164557, "grad_norm": 0.28229218874253365, "learning_rate": 2.0709334640812613e-05, "loss": 0.3536, "step": 664 }, { "epoch": 2.6936708860759495, "grad_norm": 0.2798376321903946, "learning_rate": 2.0652608894940824e-05, "loss": 0.3359, "step": 665 }, { "epoch": 2.6977215189873416, "grad_norm": 0.24573350245610467, "learning_rate": 2.0595877893046722e-05, "loss": 0.347, "step": 666 }, { "epoch": 2.701772151898734, "grad_norm": 0.2688325653283201, "learning_rate": 2.0539142092033985e-05, "loss": 0.3487, "step": 667 }, { "epoch": 2.7058227848101266, "grad_norm": 0.26974545349520823, "learning_rate": 2.048240194884496e-05, "loss": 0.3403, "step": 668 }, { "epoch": 2.709873417721519, "grad_norm": 0.256463958121147, "learning_rate": 2.042565792045695e-05, "loss": 0.3402, "step": 669 }, { "epoch": 2.7139240506329116, "grad_norm": 0.28498369936706064, "learning_rate": 2.036891046387857e-05, "loss": 0.3384, "step": 670 }, { "epoch": 2.7179746835443037, "grad_norm": 0.24873225333430415, "learning_rate": 2.0312160036146036e-05, "loss": 0.3527, "step": 671 }, { "epoch": 2.7220253164556962, "grad_norm": 0.30670848370655934, "learning_rate": 2.025540709431948e-05, "loss": 0.3538, "step": 672 }, { "epoch": 2.7260759493670887, "grad_norm": 0.23914451034472464, "learning_rate": 2.0198652095479298e-05, "loss": 0.3409, "step": 673 }, { "epoch": 2.730126582278481, "grad_norm": 0.2803352911091695, "learning_rate": 2.014189549672245e-05, "loss": 0.3476, "step": 674 }, { "epoch": 2.7341772151898733, "grad_norm": 0.2695162330827431, "learning_rate": 2.0085137755158776e-05, "loss": 0.3455, "step": 675 }, { "epoch": 2.738227848101266, "grad_norm": 0.254415938256953, "learning_rate": 2.0028379327907327e-05, "loss": 0.3539, "step": 676 }, { "epoch": 2.7422784810126584, "grad_norm": 0.27122765565436124, "learning_rate": 1.9971620672092676e-05, "loss": 0.3517, "step": 677 }, { "epoch": 2.746329113924051, "grad_norm": 0.2550076413303751, "learning_rate": 1.991486224484123e-05, "loss": 0.3465, "step": 678 }, { "epoch": 2.750379746835443, "grad_norm": 0.2716667898862722, "learning_rate": 1.985810450327756e-05, "loss": 0.3604, "step": 679 }, { "epoch": 2.7544303797468355, "grad_norm": 0.24285647121202317, "learning_rate": 1.9801347904520706e-05, "loss": 0.3444, "step": 680 }, { "epoch": 2.758481012658228, "grad_norm": 0.27673317451458657, "learning_rate": 1.974459290568053e-05, "loss": 0.3561, "step": 681 }, { "epoch": 2.76253164556962, "grad_norm": 0.25991513175279024, "learning_rate": 1.968783996385397e-05, "loss": 0.3484, "step": 682 }, { "epoch": 2.7665822784810126, "grad_norm": 0.2774196513222486, "learning_rate": 1.963108953612143e-05, "loss": 0.3523, "step": 683 }, { "epoch": 2.770632911392405, "grad_norm": 0.28174311520221007, "learning_rate": 1.9574342079543056e-05, "loss": 0.3468, "step": 684 }, { "epoch": 2.7746835443037976, "grad_norm": 0.2749813135026754, "learning_rate": 1.9517598051155046e-05, "loss": 0.3523, "step": 685 }, { "epoch": 2.77873417721519, "grad_norm": 0.30526702777005194, "learning_rate": 1.9460857907966025e-05, "loss": 0.3538, "step": 686 }, { "epoch": 2.782784810126582, "grad_norm": 0.255746994323296, "learning_rate": 1.9404122106953285e-05, "loss": 0.3538, "step": 687 }, { "epoch": 2.7868354430379747, "grad_norm": 0.2790237880459053, "learning_rate": 1.9347391105059176e-05, "loss": 0.3703, "step": 688 }, { "epoch": 2.790886075949367, "grad_norm": 0.2869334479603526, "learning_rate": 1.92906653591874e-05, "loss": 0.3356, "step": 689 }, { "epoch": 2.7949367088607593, "grad_norm": 0.25626000006924715, "learning_rate": 1.9233945326199295e-05, "loss": 0.3552, "step": 690 }, { "epoch": 2.798987341772152, "grad_norm": 0.2338290024590152, "learning_rate": 1.917723146291022e-05, "loss": 0.3493, "step": 691 }, { "epoch": 2.8030379746835443, "grad_norm": 0.24926548605190074, "learning_rate": 1.912052422608584e-05, "loss": 0.3452, "step": 692 }, { "epoch": 2.807088607594937, "grad_norm": 0.25316335629446785, "learning_rate": 1.9063824072438428e-05, "loss": 0.3491, "step": 693 }, { "epoch": 2.8111392405063294, "grad_norm": 0.29921176203620337, "learning_rate": 1.9007131458623246e-05, "loss": 0.3532, "step": 694 }, { "epoch": 2.8151898734177214, "grad_norm": 0.2573282474189339, "learning_rate": 1.895044684123479e-05, "loss": 0.3443, "step": 695 }, { "epoch": 2.819240506329114, "grad_norm": 0.2664136150796945, "learning_rate": 1.8893770676803194e-05, "loss": 0.3548, "step": 696 }, { "epoch": 2.8232911392405065, "grad_norm": 0.23285406248750126, "learning_rate": 1.8837103421790486e-05, "loss": 0.3444, "step": 697 }, { "epoch": 2.8273417721518985, "grad_norm": 0.250075781174467, "learning_rate": 1.8780445532586952e-05, "loss": 0.3526, "step": 698 }, { "epoch": 2.831392405063291, "grad_norm": 0.2597797071997067, "learning_rate": 1.872379746550743e-05, "loss": 0.3472, "step": 699 }, { "epoch": 2.8354430379746836, "grad_norm": 0.24846736427908309, "learning_rate": 1.866715967678769e-05, "loss": 0.3371, "step": 700 }, { "epoch": 2.839493670886076, "grad_norm": 0.24182564206001136, "learning_rate": 1.861053262258067e-05, "loss": 0.3443, "step": 701 }, { "epoch": 2.8435443037974686, "grad_norm": 0.234182694367731, "learning_rate": 1.8553916758952897e-05, "loss": 0.3444, "step": 702 }, { "epoch": 2.8475949367088607, "grad_norm": 0.2605939426722998, "learning_rate": 1.8497312541880735e-05, "loss": 0.346, "step": 703 }, { "epoch": 2.851645569620253, "grad_norm": 0.2431531213686959, "learning_rate": 1.8440720427246786e-05, "loss": 0.3405, "step": 704 }, { "epoch": 2.8556962025316457, "grad_norm": 0.23162372800836972, "learning_rate": 1.8384140870836157e-05, "loss": 0.349, "step": 705 }, { "epoch": 2.8597468354430378, "grad_norm": 0.23523989072155538, "learning_rate": 1.8327574328332806e-05, "loss": 0.3315, "step": 706 }, { "epoch": 2.8637974683544303, "grad_norm": 0.2263787816884725, "learning_rate": 1.8271021255315906e-05, "loss": 0.3508, "step": 707 }, { "epoch": 2.867848101265823, "grad_norm": 0.23678156360492286, "learning_rate": 1.8214482107256117e-05, "loss": 0.3482, "step": 708 }, { "epoch": 2.8718987341772153, "grad_norm": 0.2402268097237413, "learning_rate": 1.8157957339511968e-05, "loss": 0.3514, "step": 709 }, { "epoch": 2.875949367088608, "grad_norm": 0.2599542821727442, "learning_rate": 1.8101447407326182e-05, "loss": 0.3412, "step": 710 }, { "epoch": 2.88, "grad_norm": 0.22932628708356462, "learning_rate": 1.8044952765821966e-05, "loss": 0.3398, "step": 711 }, { "epoch": 2.8840506329113924, "grad_norm": 0.2876920049073328, "learning_rate": 1.7988473869999407e-05, "loss": 0.343, "step": 712 }, { "epoch": 2.888101265822785, "grad_norm": 0.27566306203028096, "learning_rate": 1.7932011174731764e-05, "loss": 0.3569, "step": 713 }, { "epoch": 2.892151898734177, "grad_norm": 0.2349246844113701, "learning_rate": 1.7875565134761817e-05, "loss": 0.3481, "step": 714 }, { "epoch": 2.8962025316455695, "grad_norm": 0.2785386689697172, "learning_rate": 1.7819136204698226e-05, "loss": 0.347, "step": 715 }, { "epoch": 2.900253164556962, "grad_norm": 0.20782842493130585, "learning_rate": 1.776272483901182e-05, "loss": 0.3392, "step": 716 }, { "epoch": 2.9043037974683545, "grad_norm": 0.262575770766965, "learning_rate": 1.7706331492031995e-05, "loss": 0.3514, "step": 717 }, { "epoch": 2.908354430379747, "grad_norm": 0.2588167957454398, "learning_rate": 1.764995661794301e-05, "loss": 0.3359, "step": 718 }, { "epoch": 2.912405063291139, "grad_norm": 0.2641219621406834, "learning_rate": 1.759360067078035e-05, "loss": 0.3444, "step": 719 }, { "epoch": 2.9164556962025316, "grad_norm": 0.23627271333557395, "learning_rate": 1.7537264104427064e-05, "loss": 0.3492, "step": 720 }, { "epoch": 2.920506329113924, "grad_norm": 0.21883515452655877, "learning_rate": 1.748094737261012e-05, "loss": 0.3538, "step": 721 }, { "epoch": 2.9245569620253162, "grad_norm": 0.2513171698519925, "learning_rate": 1.7424650928896726e-05, "loss": 0.347, "step": 722 }, { "epoch": 2.9286075949367087, "grad_norm": 0.26589307034999266, "learning_rate": 1.7368375226690712e-05, "loss": 0.3498, "step": 723 }, { "epoch": 2.9326582278481013, "grad_norm": 0.2099036560150294, "learning_rate": 1.731212071922883e-05, "loss": 0.3511, "step": 724 }, { "epoch": 2.9367088607594938, "grad_norm": 0.27453568523517474, "learning_rate": 1.7255887859577156e-05, "loss": 0.3527, "step": 725 }, { "epoch": 2.9407594936708863, "grad_norm": 0.22231860571430492, "learning_rate": 1.7199677100627427e-05, "loss": 0.3435, "step": 726 }, { "epoch": 2.9448101265822784, "grad_norm": 0.24687328550594798, "learning_rate": 1.7143488895093343e-05, "loss": 0.34, "step": 727 }, { "epoch": 2.948860759493671, "grad_norm": 0.2490491235976839, "learning_rate": 1.7087323695506994e-05, "loss": 0.3422, "step": 728 }, { "epoch": 2.9529113924050634, "grad_norm": 0.22705479416572094, "learning_rate": 1.7031181954215194e-05, "loss": 0.3478, "step": 729 }, { "epoch": 2.9569620253164555, "grad_norm": 0.24670363175595897, "learning_rate": 1.6975064123375788e-05, "loss": 0.3504, "step": 730 }, { "epoch": 2.961012658227848, "grad_norm": 0.25278172393357906, "learning_rate": 1.6918970654954084e-05, "loss": 0.338, "step": 731 }, { "epoch": 2.9650632911392405, "grad_norm": 0.2312580172898857, "learning_rate": 1.686290200071915e-05, "loss": 0.3507, "step": 732 }, { "epoch": 2.969113924050633, "grad_norm": 0.25206479926673736, "learning_rate": 1.6806858612240234e-05, "loss": 0.3514, "step": 733 }, { "epoch": 2.9731645569620255, "grad_norm": 0.22741679869117506, "learning_rate": 1.6750840940883078e-05, "loss": 0.3417, "step": 734 }, { "epoch": 2.9772151898734176, "grad_norm": 0.26439580276783076, "learning_rate": 1.6694849437806305e-05, "loss": 0.3369, "step": 735 }, { "epoch": 2.98126582278481, "grad_norm": 0.25052362068384293, "learning_rate": 1.663888455395778e-05, "loss": 0.3517, "step": 736 }, { "epoch": 2.9853164556962026, "grad_norm": 0.26485426676435514, "learning_rate": 1.6582946740070995e-05, "loss": 0.3462, "step": 737 }, { "epoch": 2.9893670886075947, "grad_norm": 0.22445802555485958, "learning_rate": 1.6527036446661396e-05, "loss": 0.3386, "step": 738 }, { "epoch": 2.993417721518987, "grad_norm": 0.27426179007885826, "learning_rate": 1.6471154124022818e-05, "loss": 0.3524, "step": 739 }, { "epoch": 2.9974683544303797, "grad_norm": 0.23866057794073944, "learning_rate": 1.6415300222223788e-05, "loss": 0.3514, "step": 740 }, { "epoch": 3.0015189873417722, "grad_norm": 0.2671127821217465, "learning_rate": 1.6359475191103958e-05, "loss": 0.3308, "step": 741 }, { "epoch": 3.0055696202531648, "grad_norm": 0.32700967615108856, "learning_rate": 1.6303679480270466e-05, "loss": 0.2963, "step": 742 }, { "epoch": 3.009620253164557, "grad_norm": 0.5662692745243918, "learning_rate": 1.624791353909428e-05, "loss": 0.2952, "step": 743 }, { "epoch": 3.0136708860759494, "grad_norm": 0.33888863680404896, "learning_rate": 1.619217781670663e-05, "loss": 0.2999, "step": 744 }, { "epoch": 3.017721518987342, "grad_norm": 0.3762118745724246, "learning_rate": 1.6136472761995373e-05, "loss": 0.2915, "step": 745 }, { "epoch": 3.0217721518987344, "grad_norm": 0.34069928918549003, "learning_rate": 1.608079882360134e-05, "loss": 0.3005, "step": 746 }, { "epoch": 3.0258227848101265, "grad_norm": 0.33993319123720245, "learning_rate": 1.60251564499148e-05, "loss": 0.2959, "step": 747 }, { "epoch": 3.029873417721519, "grad_norm": 0.31806279146366007, "learning_rate": 1.596954608907176e-05, "loss": 0.282, "step": 748 }, { "epoch": 3.0339240506329115, "grad_norm": 0.28190215285642156, "learning_rate": 1.591396818895043e-05, "loss": 0.3002, "step": 749 }, { "epoch": 3.037974683544304, "grad_norm": 0.33086885621119155, "learning_rate": 1.585842319716759e-05, "loss": 0.3009, "step": 750 }, { "epoch": 3.042025316455696, "grad_norm": 0.27795830656524134, "learning_rate": 1.5802911561074944e-05, "loss": 0.2942, "step": 751 }, { "epoch": 3.0460759493670886, "grad_norm": 0.31405997457573676, "learning_rate": 1.5747433727755595e-05, "loss": 0.2967, "step": 752 }, { "epoch": 3.050126582278481, "grad_norm": 0.3141831706260811, "learning_rate": 1.5691990144020376e-05, "loss": 0.2943, "step": 753 }, { "epoch": 3.0541772151898736, "grad_norm": 0.23633930981050494, "learning_rate": 1.5636581256404297e-05, "loss": 0.3009, "step": 754 }, { "epoch": 3.0582278481012657, "grad_norm": 0.30343911412960617, "learning_rate": 1.558120751116291e-05, "loss": 0.2974, "step": 755 }, { "epoch": 3.062278481012658, "grad_norm": 0.26743262338317836, "learning_rate": 1.552586935426876e-05, "loss": 0.2892, "step": 756 }, { "epoch": 3.0663291139240507, "grad_norm": 0.27690042496156125, "learning_rate": 1.547056723140774e-05, "loss": 0.2972, "step": 757 }, { "epoch": 3.0703797468354432, "grad_norm": 0.29457000949342665, "learning_rate": 1.5415301587975565e-05, "loss": 0.2869, "step": 758 }, { "epoch": 3.0744303797468353, "grad_norm": 0.25657091581286007, "learning_rate": 1.536007286907411e-05, "loss": 0.2933, "step": 759 }, { "epoch": 3.078481012658228, "grad_norm": 0.26728196659766046, "learning_rate": 1.5304881519507896e-05, "loss": 0.2962, "step": 760 }, { "epoch": 3.0825316455696203, "grad_norm": 0.26707716534486153, "learning_rate": 1.5249727983780453e-05, "loss": 0.2944, "step": 761 }, { "epoch": 3.086582278481013, "grad_norm": 0.23794017715521562, "learning_rate": 1.5194612706090786e-05, "loss": 0.2921, "step": 762 }, { "epoch": 3.090632911392405, "grad_norm": 0.23974448677112947, "learning_rate": 1.5139536130329771e-05, "loss": 0.293, "step": 763 }, { "epoch": 3.0946835443037974, "grad_norm": 0.23052112299951022, "learning_rate": 1.508449870007656e-05, "loss": 0.2974, "step": 764 }, { "epoch": 3.09873417721519, "grad_norm": 0.2249938316150708, "learning_rate": 1.5029500858595056e-05, "loss": 0.3018, "step": 765 }, { "epoch": 3.1027848101265825, "grad_norm": 0.2300711662360829, "learning_rate": 1.4974543048830328e-05, "loss": 0.2917, "step": 766 }, { "epoch": 3.1068354430379745, "grad_norm": 0.2301734582896938, "learning_rate": 1.4919625713405e-05, "loss": 0.2894, "step": 767 }, { "epoch": 3.110886075949367, "grad_norm": 0.22627936183445624, "learning_rate": 1.4864749294615756e-05, "loss": 0.2915, "step": 768 }, { "epoch": 3.1149367088607596, "grad_norm": 0.2317710166909585, "learning_rate": 1.4809914234429716e-05, "loss": 0.3048, "step": 769 }, { "epoch": 3.118987341772152, "grad_norm": 0.24580811924149384, "learning_rate": 1.4755120974480923e-05, "loss": 0.2916, "step": 770 }, { "epoch": 3.123037974683544, "grad_norm": 0.23867105139181782, "learning_rate": 1.4700369956066771e-05, "loss": 0.2965, "step": 771 }, { "epoch": 3.1270886075949367, "grad_norm": 0.20660898292460342, "learning_rate": 1.4645661620144413e-05, "loss": 0.3017, "step": 772 }, { "epoch": 3.131139240506329, "grad_norm": 0.22537932847173378, "learning_rate": 1.4590996407327284e-05, "loss": 0.2947, "step": 773 }, { "epoch": 3.1351898734177217, "grad_norm": 0.24055523414234725, "learning_rate": 1.4536374757881487e-05, "loss": 0.3021, "step": 774 }, { "epoch": 3.1392405063291138, "grad_norm": 0.21127127345985164, "learning_rate": 1.4481797111722271e-05, "loss": 0.2898, "step": 775 }, { "epoch": 3.1432911392405063, "grad_norm": 0.23895144494033452, "learning_rate": 1.4427263908410507e-05, "loss": 0.2958, "step": 776 }, { "epoch": 3.147341772151899, "grad_norm": 0.22883815106298042, "learning_rate": 1.4372775587149108e-05, "loss": 0.3036, "step": 777 }, { "epoch": 3.1513924050632913, "grad_norm": 0.2296331632350077, "learning_rate": 1.4318332586779522e-05, "loss": 0.297, "step": 778 }, { "epoch": 3.1554430379746834, "grad_norm": 0.23792960167633892, "learning_rate": 1.4263935345778202e-05, "loss": 0.313, "step": 779 }, { "epoch": 3.159493670886076, "grad_norm": 0.2224999915063212, "learning_rate": 1.420958430225303e-05, "loss": 0.2921, "step": 780 }, { "epoch": 3.1635443037974684, "grad_norm": 0.22325431146685243, "learning_rate": 1.415527989393985e-05, "loss": 0.2981, "step": 781 }, { "epoch": 3.167594936708861, "grad_norm": 0.21518867170290634, "learning_rate": 1.410102255819891e-05, "loss": 0.2927, "step": 782 }, { "epoch": 3.171645569620253, "grad_norm": 0.21918444297144596, "learning_rate": 1.404681273201131e-05, "loss": 0.2943, "step": 783 }, { "epoch": 3.1756962025316455, "grad_norm": 0.22975770367037254, "learning_rate": 1.399265085197556e-05, "loss": 0.3081, "step": 784 }, { "epoch": 3.179746835443038, "grad_norm": 0.2204340584044393, "learning_rate": 1.393853735430398e-05, "loss": 0.3092, "step": 785 }, { "epoch": 3.1837974683544306, "grad_norm": 0.2344029851678538, "learning_rate": 1.3884472674819246e-05, "loss": 0.2977, "step": 786 }, { "epoch": 3.1878481012658226, "grad_norm": 0.22071994673613368, "learning_rate": 1.3830457248950864e-05, "loss": 0.2994, "step": 787 }, { "epoch": 3.191898734177215, "grad_norm": 0.2418564411806224, "learning_rate": 1.377649151173163e-05, "loss": 0.2964, "step": 788 }, { "epoch": 3.1959493670886077, "grad_norm": 0.24805695480864387, "learning_rate": 1.3722575897794181e-05, "loss": 0.3082, "step": 789 }, { "epoch": 3.2, "grad_norm": 0.22397059680755932, "learning_rate": 1.3668710841367472e-05, "loss": 0.2989, "step": 790 }, { "epoch": 3.2040506329113922, "grad_norm": 0.23465326904401348, "learning_rate": 1.361489677627324e-05, "loss": 0.2889, "step": 791 }, { "epoch": 3.2081012658227848, "grad_norm": 0.2295047487610294, "learning_rate": 1.3561134135922585e-05, "loss": 0.3, "step": 792 }, { "epoch": 3.2121518987341773, "grad_norm": 0.23511997158264816, "learning_rate": 1.350742335331241e-05, "loss": 0.294, "step": 793 }, { "epoch": 3.21620253164557, "grad_norm": 0.23042518162834938, "learning_rate": 1.345376486102198e-05, "loss": 0.3003, "step": 794 }, { "epoch": 3.220253164556962, "grad_norm": 0.22496363853003426, "learning_rate": 1.3400159091209414e-05, "loss": 0.2926, "step": 795 }, { "epoch": 3.2243037974683544, "grad_norm": 0.22664376009597723, "learning_rate": 1.3346606475608216e-05, "loss": 0.3026, "step": 796 }, { "epoch": 3.228354430379747, "grad_norm": 0.23864766390661799, "learning_rate": 1.3293107445523781e-05, "loss": 0.312, "step": 797 }, { "epoch": 3.2324050632911394, "grad_norm": 0.2054157409002744, "learning_rate": 1.3239662431829949e-05, "loss": 0.304, "step": 798 }, { "epoch": 3.2364556962025315, "grad_norm": 0.24153541260185526, "learning_rate": 1.3186271864965509e-05, "loss": 0.2998, "step": 799 }, { "epoch": 3.240506329113924, "grad_norm": 0.2213708067924794, "learning_rate": 1.3132936174930756e-05, "loss": 0.3038, "step": 800 }, { "epoch": 3.2445569620253165, "grad_norm": 0.24210530142451153, "learning_rate": 1.3079655791283995e-05, "loss": 0.3028, "step": 801 }, { "epoch": 3.248607594936709, "grad_norm": 0.2476953175253225, "learning_rate": 1.3026431143138108e-05, "loss": 0.2929, "step": 802 }, { "epoch": 3.252658227848101, "grad_norm": 0.26546649106439807, "learning_rate": 1.2973262659157114e-05, "loss": 0.2937, "step": 803 }, { "epoch": 3.2567088607594936, "grad_norm": 0.24202065921870378, "learning_rate": 1.2920150767552651e-05, "loss": 0.2993, "step": 804 }, { "epoch": 3.260759493670886, "grad_norm": 0.2191441200090244, "learning_rate": 1.2867095896080607e-05, "loss": 0.2942, "step": 805 }, { "epoch": 3.2648101265822786, "grad_norm": 0.23671083243882643, "learning_rate": 1.2814098472037612e-05, "loss": 0.2856, "step": 806 }, { "epoch": 3.2688607594936707, "grad_norm": 0.2402314935665623, "learning_rate": 1.276115892225764e-05, "loss": 0.3077, "step": 807 }, { "epoch": 3.2729113924050632, "grad_norm": 0.22546701238687886, "learning_rate": 1.2708277673108555e-05, "loss": 0.2986, "step": 808 }, { "epoch": 3.2769620253164558, "grad_norm": 0.23699519925180149, "learning_rate": 1.2655455150488649e-05, "loss": 0.2981, "step": 809 }, { "epoch": 3.2810126582278483, "grad_norm": 0.23948872706687718, "learning_rate": 1.2602691779823272e-05, "loss": 0.3016, "step": 810 }, { "epoch": 3.2850632911392403, "grad_norm": 0.21201631839836652, "learning_rate": 1.2549987986061355e-05, "loss": 0.3028, "step": 811 }, { "epoch": 3.289113924050633, "grad_norm": 0.24596553905356522, "learning_rate": 1.2497344193672005e-05, "loss": 0.2838, "step": 812 }, { "epoch": 3.2931645569620254, "grad_norm": 0.2289484399580497, "learning_rate": 1.2444760826641092e-05, "loss": 0.2907, "step": 813 }, { "epoch": 3.297215189873418, "grad_norm": 0.2422624534983616, "learning_rate": 1.2392238308467817e-05, "loss": 0.3002, "step": 814 }, { "epoch": 3.30126582278481, "grad_norm": 0.23228124322777952, "learning_rate": 1.2339777062161326e-05, "loss": 0.2918, "step": 815 }, { "epoch": 3.3053164556962025, "grad_norm": 0.24272831963652697, "learning_rate": 1.2287377510237293e-05, "loss": 0.3032, "step": 816 }, { "epoch": 3.309367088607595, "grad_norm": 0.2637128675192321, "learning_rate": 1.2235040074714488e-05, "loss": 0.3018, "step": 817 }, { "epoch": 3.3134177215189875, "grad_norm": 0.22140406443748895, "learning_rate": 1.2182765177111434e-05, "loss": 0.3017, "step": 818 }, { "epoch": 3.3174683544303796, "grad_norm": 0.22916979672748397, "learning_rate": 1.213055323844297e-05, "loss": 0.3018, "step": 819 }, { "epoch": 3.321518987341772, "grad_norm": 0.2311047132728669, "learning_rate": 1.2078404679216864e-05, "loss": 0.2978, "step": 820 }, { "epoch": 3.3255696202531646, "grad_norm": 0.2109040793527875, "learning_rate": 1.2026319919430458e-05, "loss": 0.2999, "step": 821 }, { "epoch": 3.329620253164557, "grad_norm": 0.2266480094658967, "learning_rate": 1.1974299378567227e-05, "loss": 0.2921, "step": 822 }, { "epoch": 3.333670886075949, "grad_norm": 0.24134289706675108, "learning_rate": 1.1922343475593462e-05, "loss": 0.2997, "step": 823 }, { "epoch": 3.3377215189873417, "grad_norm": 0.22914474234831292, "learning_rate": 1.187045262895488e-05, "loss": 0.291, "step": 824 }, { "epoch": 3.3417721518987342, "grad_norm": 0.21973370936342074, "learning_rate": 1.1818627256573203e-05, "loss": 0.3017, "step": 825 }, { "epoch": 3.3458227848101267, "grad_norm": 0.21836439470053728, "learning_rate": 1.1766867775842864e-05, "loss": 0.2993, "step": 826 }, { "epoch": 3.349873417721519, "grad_norm": 0.2191120219261651, "learning_rate": 1.1715174603627615e-05, "loss": 0.3022, "step": 827 }, { "epoch": 3.3539240506329113, "grad_norm": 0.2337172616945094, "learning_rate": 1.1663548156257147e-05, "loss": 0.3055, "step": 828 }, { "epoch": 3.357974683544304, "grad_norm": 0.2115235722231807, "learning_rate": 1.161198884952377e-05, "loss": 0.3001, "step": 829 }, { "epoch": 3.3620253164556964, "grad_norm": 0.2144870725596607, "learning_rate": 1.1560497098679056e-05, "loss": 0.2839, "step": 830 }, { "epoch": 3.3660759493670884, "grad_norm": 0.21544644252530254, "learning_rate": 1.1509073318430479e-05, "loss": 0.2968, "step": 831 }, { "epoch": 3.370126582278481, "grad_norm": 0.2019537258119688, "learning_rate": 1.1457717922938116e-05, "loss": 0.2933, "step": 832 }, { "epoch": 3.3741772151898735, "grad_norm": 0.21993250498652023, "learning_rate": 1.1406431325811233e-05, "loss": 0.2877, "step": 833 }, { "epoch": 3.378227848101266, "grad_norm": 0.21999338163601306, "learning_rate": 1.135521394010506e-05, "loss": 0.2973, "step": 834 }, { "epoch": 3.382278481012658, "grad_norm": 0.19521501480136622, "learning_rate": 1.1304066178317367e-05, "loss": 0.3073, "step": 835 }, { "epoch": 3.3863291139240506, "grad_norm": 0.22667839126141817, "learning_rate": 1.1252988452385199e-05, "loss": 0.3065, "step": 836 }, { "epoch": 3.390379746835443, "grad_norm": 0.21233179263051133, "learning_rate": 1.1201981173681536e-05, "loss": 0.2945, "step": 837 }, { "epoch": 3.3944303797468356, "grad_norm": 0.21214010497047148, "learning_rate": 1.1151044753011991e-05, "loss": 0.3028, "step": 838 }, { "epoch": 3.3984810126582277, "grad_norm": 0.2759025034253367, "learning_rate": 1.1100179600611491e-05, "loss": 0.2867, "step": 839 }, { "epoch": 3.40253164556962, "grad_norm": 0.19624978929566464, "learning_rate": 1.1049386126140985e-05, "loss": 0.2958, "step": 840 }, { "epoch": 3.4065822784810127, "grad_norm": 0.22427727753569934, "learning_rate": 1.0998664738684128e-05, "loss": 0.2865, "step": 841 }, { "epoch": 3.410632911392405, "grad_norm": 0.24246622914798643, "learning_rate": 1.0948015846744e-05, "loss": 0.2957, "step": 842 }, { "epoch": 3.4146835443037973, "grad_norm": 0.23009388027867625, "learning_rate": 1.0897439858239832e-05, "loss": 0.2862, "step": 843 }, { "epoch": 3.41873417721519, "grad_norm": 0.24832634532977813, "learning_rate": 1.0846937180503652e-05, "loss": 0.294, "step": 844 }, { "epoch": 3.4227848101265823, "grad_norm": 0.25034744617129345, "learning_rate": 1.0796508220277117e-05, "loss": 0.3108, "step": 845 }, { "epoch": 3.426835443037975, "grad_norm": 0.2712695287529746, "learning_rate": 1.0746153383708107e-05, "loss": 0.2982, "step": 846 }, { "epoch": 3.430886075949367, "grad_norm": 0.2613824881042837, "learning_rate": 1.0695873076347579e-05, "loss": 0.2937, "step": 847 }, { "epoch": 3.4349367088607594, "grad_norm": 0.23678916360206037, "learning_rate": 1.0645667703146205e-05, "loss": 0.3046, "step": 848 }, { "epoch": 3.438987341772152, "grad_norm": 0.2140129948958066, "learning_rate": 1.0595537668451161e-05, "loss": 0.296, "step": 849 }, { "epoch": 3.4430379746835444, "grad_norm": 0.24512447684203084, "learning_rate": 1.0545483376002854e-05, "loss": 0.2934, "step": 850 }, { "epoch": 3.4470886075949365, "grad_norm": 0.2437536013444335, "learning_rate": 1.0495505228931676e-05, "loss": 0.3048, "step": 851 }, { "epoch": 3.451139240506329, "grad_norm": 0.22588537856372023, "learning_rate": 1.044560362975474e-05, "loss": 0.3016, "step": 852 }, { "epoch": 3.4551898734177215, "grad_norm": 0.23354786408136272, "learning_rate": 1.0395778980372695e-05, "loss": 0.2892, "step": 853 }, { "epoch": 3.459240506329114, "grad_norm": 0.21937717744953136, "learning_rate": 1.0346031682066381e-05, "loss": 0.2967, "step": 854 }, { "epoch": 3.463291139240506, "grad_norm": 0.24543436992630852, "learning_rate": 1.0296362135493724e-05, "loss": 0.2935, "step": 855 }, { "epoch": 3.4673417721518986, "grad_norm": 0.22720034984779852, "learning_rate": 1.0246770740686422e-05, "loss": 0.3039, "step": 856 }, { "epoch": 3.471392405063291, "grad_norm": 0.21288531000886463, "learning_rate": 1.0197257897046743e-05, "loss": 0.2982, "step": 857 }, { "epoch": 3.4754430379746837, "grad_norm": 0.23300957115299184, "learning_rate": 1.014782400334433e-05, "loss": 0.3007, "step": 858 }, { "epoch": 3.479493670886076, "grad_norm": 0.234975849107781, "learning_rate": 1.009846945771296e-05, "loss": 0.3022, "step": 859 }, { "epoch": 3.4835443037974683, "grad_norm": 0.2111378928739831, "learning_rate": 1.0049194657647363e-05, "loss": 0.3039, "step": 860 }, { "epoch": 3.487594936708861, "grad_norm": 0.24427368499130855, "learning_rate": 1.0000000000000006e-05, "loss": 0.2912, "step": 861 }, { "epoch": 3.4916455696202533, "grad_norm": 0.20263786884388144, "learning_rate": 9.950885880977891e-06, "loss": 0.3009, "step": 862 }, { "epoch": 3.4956962025316454, "grad_norm": 0.22316240292191958, "learning_rate": 9.901852696139382e-06, "loss": 0.2916, "step": 863 }, { "epoch": 3.499746835443038, "grad_norm": 0.22848696299903676, "learning_rate": 9.852900840391027e-06, "loss": 0.2981, "step": 864 }, { "epoch": 3.5037974683544304, "grad_norm": 0.22154330668847896, "learning_rate": 9.804030707984313e-06, "loss": 0.2872, "step": 865 }, { "epoch": 3.507848101265823, "grad_norm": 0.20099964284643462, "learning_rate": 9.755242692512599e-06, "loss": 0.2901, "step": 866 }, { "epoch": 3.5118987341772154, "grad_norm": 0.22180494001928072, "learning_rate": 9.70653718690782e-06, "loss": 0.2981, "step": 867 }, { "epoch": 3.5159493670886075, "grad_norm": 0.22078802427011943, "learning_rate": 9.657914583437454e-06, "loss": 0.3001, "step": 868 }, { "epoch": 3.52, "grad_norm": 0.20827304714113387, "learning_rate": 9.609375273701246e-06, "loss": 0.3005, "step": 869 }, { "epoch": 3.5240506329113925, "grad_norm": 0.19086472409933208, "learning_rate": 9.560919648628133e-06, "loss": 0.3096, "step": 870 }, { "epoch": 3.5281012658227846, "grad_norm": 0.20531739218954453, "learning_rate": 9.512548098473047e-06, "loss": 0.3159, "step": 871 }, { "epoch": 3.532151898734177, "grad_norm": 0.21118019576366864, "learning_rate": 9.464261012813825e-06, "loss": 0.2916, "step": 872 }, { "epoch": 3.5362025316455696, "grad_norm": 0.2066485581305117, "learning_rate": 9.416058780547987e-06, "loss": 0.3042, "step": 873 }, { "epoch": 3.540253164556962, "grad_norm": 0.21019122945575386, "learning_rate": 9.367941789889714e-06, "loss": 0.2951, "step": 874 }, { "epoch": 3.5443037974683547, "grad_norm": 0.2276028902451978, "learning_rate": 9.319910428366607e-06, "loss": 0.3018, "step": 875 }, { "epoch": 3.5483544303797467, "grad_norm": 0.2270990811598102, "learning_rate": 9.271965082816667e-06, "loss": 0.302, "step": 876 }, { "epoch": 3.5524050632911393, "grad_norm": 0.20688464895237427, "learning_rate": 9.224106139385111e-06, "loss": 0.2827, "step": 877 }, { "epoch": 3.5564556962025318, "grad_norm": 0.22043039474018092, "learning_rate": 9.176333983521291e-06, "loss": 0.3062, "step": 878 }, { "epoch": 3.560506329113924, "grad_norm": 0.20088903304866879, "learning_rate": 9.12864899997558e-06, "loss": 0.3037, "step": 879 }, { "epoch": 3.5645569620253164, "grad_norm": 0.2050928490484553, "learning_rate": 9.08105157279628e-06, "loss": 0.2826, "step": 880 }, { "epoch": 3.568607594936709, "grad_norm": 0.21404425137526678, "learning_rate": 9.03354208532653e-06, "loss": 0.2968, "step": 881 }, { "epoch": 3.5726582278481014, "grad_norm": 0.20587687299023408, "learning_rate": 8.986120920201205e-06, "loss": 0.3073, "step": 882 }, { "epoch": 3.576708860759494, "grad_norm": 0.23241786087364977, "learning_rate": 8.938788459343852e-06, "loss": 0.296, "step": 883 }, { "epoch": 3.580759493670886, "grad_norm": 0.2136475905720522, "learning_rate": 8.8915450839636e-06, "loss": 0.3002, "step": 884 }, { "epoch": 3.5848101265822785, "grad_norm": 0.22661695700100606, "learning_rate": 8.844391174552116e-06, "loss": 0.2989, "step": 885 }, { "epoch": 3.588860759493671, "grad_norm": 0.20915327969026234, "learning_rate": 8.797327110880479e-06, "loss": 0.2939, "step": 886 }, { "epoch": 3.592911392405063, "grad_norm": 0.2292587150126473, "learning_rate": 8.750353271996206e-06, "loss": 0.2986, "step": 887 }, { "epoch": 3.5969620253164556, "grad_norm": 0.20723365831564844, "learning_rate": 8.703470036220132e-06, "loss": 0.2823, "step": 888 }, { "epoch": 3.601012658227848, "grad_norm": 0.21764655000691358, "learning_rate": 8.656677781143394e-06, "loss": 0.2872, "step": 889 }, { "epoch": 3.6050632911392406, "grad_norm": 0.22235577971618928, "learning_rate": 8.609976883624377e-06, "loss": 0.297, "step": 890 }, { "epoch": 3.609113924050633, "grad_norm": 0.24936734784788883, "learning_rate": 8.563367719785698e-06, "loss": 0.2969, "step": 891 }, { "epoch": 3.613164556962025, "grad_norm": 0.2224793214722332, "learning_rate": 8.516850665011138e-06, "loss": 0.2944, "step": 892 }, { "epoch": 3.6172151898734177, "grad_norm": 0.22395363204292007, "learning_rate": 8.47042609394269e-06, "loss": 0.2891, "step": 893 }, { "epoch": 3.6212658227848102, "grad_norm": 0.24889100864725458, "learning_rate": 8.424094380477432e-06, "loss": 0.2919, "step": 894 }, { "epoch": 3.6253164556962023, "grad_norm": 0.21154785788524383, "learning_rate": 8.37785589776465e-06, "loss": 0.299, "step": 895 }, { "epoch": 3.629367088607595, "grad_norm": 0.23484313040986554, "learning_rate": 8.331711018202694e-06, "loss": 0.2838, "step": 896 }, { "epoch": 3.6334177215189873, "grad_norm": 0.23493911543729917, "learning_rate": 8.285660113436104e-06, "loss": 0.2899, "step": 897 }, { "epoch": 3.63746835443038, "grad_norm": 0.22804564830635235, "learning_rate": 8.239703554352527e-06, "loss": 0.3008, "step": 898 }, { "epoch": 3.6415189873417724, "grad_norm": 0.23350854307572472, "learning_rate": 8.193841711079775e-06, "loss": 0.2977, "step": 899 }, { "epoch": 3.6455696202531644, "grad_norm": 0.23164260604699313, "learning_rate": 8.148074952982828e-06, "loss": 0.2987, "step": 900 }, { "epoch": 3.649620253164557, "grad_norm": 0.20277049887434007, "learning_rate": 8.102403648660859e-06, "loss": 0.2921, "step": 901 }, { "epoch": 3.6536708860759495, "grad_norm": 0.2273545575703857, "learning_rate": 8.056828165944282e-06, "loss": 0.2983, "step": 902 }, { "epoch": 3.6577215189873415, "grad_norm": 0.2262207395176346, "learning_rate": 8.011348871891762e-06, "loss": 0.2933, "step": 903 }, { "epoch": 3.661772151898734, "grad_norm": 0.19222082126151843, "learning_rate": 7.965966132787287e-06, "loss": 0.2961, "step": 904 }, { "epoch": 3.6658227848101266, "grad_norm": 0.2204626608691333, "learning_rate": 7.920680314137189e-06, "loss": 0.3017, "step": 905 }, { "epoch": 3.669873417721519, "grad_norm": 0.23797978441926712, "learning_rate": 7.875491780667246e-06, "loss": 0.305, "step": 906 }, { "epoch": 3.6739240506329116, "grad_norm": 0.212562143068109, "learning_rate": 7.830400896319667e-06, "loss": 0.3022, "step": 907 }, { "epoch": 3.6779746835443037, "grad_norm": 0.23065692184149705, "learning_rate": 7.785408024250259e-06, "loss": 0.293, "step": 908 }, { "epoch": 3.682025316455696, "grad_norm": 0.23803939300218976, "learning_rate": 7.74051352682542e-06, "loss": 0.2994, "step": 909 }, { "epoch": 3.6860759493670887, "grad_norm": 0.20700538246943803, "learning_rate": 7.695717765619257e-06, "loss": 0.2935, "step": 910 }, { "epoch": 3.690126582278481, "grad_norm": 0.2250172057751531, "learning_rate": 7.651021101410673e-06, "loss": 0.3015, "step": 911 }, { "epoch": 3.6941772151898733, "grad_norm": 0.24106025157002867, "learning_rate": 7.606423894180464e-06, "loss": 0.2948, "step": 912 }, { "epoch": 3.698227848101266, "grad_norm": 0.2037398684162737, "learning_rate": 7.56192650310839e-06, "loss": 0.2916, "step": 913 }, { "epoch": 3.7022784810126583, "grad_norm": 0.20131982256707956, "learning_rate": 7.517529286570349e-06, "loss": 0.2969, "step": 914 }, { "epoch": 3.706329113924051, "grad_norm": 0.24164886496346524, "learning_rate": 7.473232602135387e-06, "loss": 0.3058, "step": 915 }, { "epoch": 3.710379746835443, "grad_norm": 0.20073328927806172, "learning_rate": 7.429036806562935e-06, "loss": 0.2944, "step": 916 }, { "epoch": 3.7144303797468354, "grad_norm": 0.19878158769191506, "learning_rate": 7.3849422557998455e-06, "loss": 0.2871, "step": 917 }, { "epoch": 3.718481012658228, "grad_norm": 0.2077804379419975, "learning_rate": 7.340949304977567e-06, "loss": 0.2979, "step": 918 }, { "epoch": 3.72253164556962, "grad_norm": 0.1999145580348887, "learning_rate": 7.297058308409282e-06, "loss": 0.2911, "step": 919 }, { "epoch": 3.7265822784810125, "grad_norm": 0.20119435847230693, "learning_rate": 7.25326961958704e-06, "loss": 0.297, "step": 920 }, { "epoch": 3.730632911392405, "grad_norm": 0.20020390302542743, "learning_rate": 7.209583591178921e-06, "loss": 0.2913, "step": 921 }, { "epoch": 3.7346835443037976, "grad_norm": 0.2079071594628412, "learning_rate": 7.1660005750261925e-06, "loss": 0.3149, "step": 922 }, { "epoch": 3.73873417721519, "grad_norm": 0.19874407300228764, "learning_rate": 7.1225209221404765e-06, "loss": 0.3005, "step": 923 }, { "epoch": 3.742784810126582, "grad_norm": 0.2242417675599834, "learning_rate": 7.079144982700909e-06, "loss": 0.3048, "step": 924 }, { "epoch": 3.7468354430379747, "grad_norm": 0.20723585024139052, "learning_rate": 7.0358731060513695e-06, "loss": 0.2965, "step": 925 }, { "epoch": 3.750886075949367, "grad_norm": 0.21298966965568525, "learning_rate": 6.99270564069757e-06, "loss": 0.3007, "step": 926 }, { "epoch": 3.7549367088607593, "grad_norm": 0.1981757842524555, "learning_rate": 6.949642934304375e-06, "loss": 0.3009, "step": 927 }, { "epoch": 3.7589873417721518, "grad_norm": 0.20690639839672656, "learning_rate": 6.906685333692871e-06, "loss": 0.2963, "step": 928 }, { "epoch": 3.7630379746835443, "grad_norm": 0.21072837649242063, "learning_rate": 6.86383318483769e-06, "loss": 0.2984, "step": 929 }, { "epoch": 3.767088607594937, "grad_norm": 0.20489328660753858, "learning_rate": 6.821086832864139e-06, "loss": 0.2984, "step": 930 }, { "epoch": 3.7711392405063293, "grad_norm": 0.19482918239199107, "learning_rate": 6.77844662204546e-06, "loss": 0.3007, "step": 931 }, { "epoch": 3.7751898734177214, "grad_norm": 0.21291107658265734, "learning_rate": 6.7359128958000455e-06, "loss": 0.2929, "step": 932 }, { "epoch": 3.779240506329114, "grad_norm": 0.20252220519519135, "learning_rate": 6.693485996688695e-06, "loss": 0.301, "step": 933 }, { "epoch": 3.7832911392405064, "grad_norm": 0.1965155573380087, "learning_rate": 6.651166266411801e-06, "loss": 0.2956, "step": 934 }, { "epoch": 3.7873417721518985, "grad_norm": 0.20478820380135002, "learning_rate": 6.6089540458066725e-06, "loss": 0.2922, "step": 935 }, { "epoch": 3.791392405063291, "grad_norm": 0.19762254243352342, "learning_rate": 6.566849674844711e-06, "loss": 0.2914, "step": 936 }, { "epoch": 3.7954430379746835, "grad_norm": 0.2029627033042507, "learning_rate": 6.524853492628747e-06, "loss": 0.3039, "step": 937 }, { "epoch": 3.799493670886076, "grad_norm": 0.19529815690903715, "learning_rate": 6.4829658373902536e-06, "loss": 0.2875, "step": 938 }, { "epoch": 3.8035443037974686, "grad_norm": 0.20403692070948126, "learning_rate": 6.441187046486648e-06, "loss": 0.2916, "step": 939 }, { "epoch": 3.8075949367088606, "grad_norm": 0.20452089463951903, "learning_rate": 6.399517456398567e-06, "loss": 0.3049, "step": 940 }, { "epoch": 3.811645569620253, "grad_norm": 0.20732360516590187, "learning_rate": 6.357957402727164e-06, "loss": 0.2991, "step": 941 }, { "epoch": 3.8156962025316457, "grad_norm": 0.2086062170924424, "learning_rate": 6.316507220191395e-06, "loss": 0.301, "step": 942 }, { "epoch": 3.8197468354430377, "grad_norm": 0.18974890227998575, "learning_rate": 6.275167242625331e-06, "loss": 0.2938, "step": 943 }, { "epoch": 3.8237974683544302, "grad_norm": 0.19520277422066137, "learning_rate": 6.233937802975471e-06, "loss": 0.2932, "step": 944 }, { "epoch": 3.8278481012658228, "grad_norm": 0.21564298470894072, "learning_rate": 6.192819233298046e-06, "loss": 0.2914, "step": 945 }, { "epoch": 3.8318987341772153, "grad_norm": 0.22424412848112762, "learning_rate": 6.151811864756383e-06, "loss": 0.2986, "step": 946 }, { "epoch": 3.835949367088608, "grad_norm": 0.19644651908733132, "learning_rate": 6.1109160276181655e-06, "loss": 0.2945, "step": 947 }, { "epoch": 3.84, "grad_norm": 0.21268182538767896, "learning_rate": 6.070132051252868e-06, "loss": 0.3015, "step": 948 }, { "epoch": 3.8440506329113924, "grad_norm": 0.21358906501482977, "learning_rate": 6.0294602641290034e-06, "loss": 0.2934, "step": 949 }, { "epoch": 3.848101265822785, "grad_norm": 0.2118549789334887, "learning_rate": 5.988900993811575e-06, "loss": 0.2944, "step": 950 }, { "epoch": 3.852151898734177, "grad_norm": 0.19293018711901333, "learning_rate": 5.948454566959363e-06, "loss": 0.2983, "step": 951 }, { "epoch": 3.8562025316455695, "grad_norm": 0.20421387212787998, "learning_rate": 5.908121309322328e-06, "loss": 0.2968, "step": 952 }, { "epoch": 3.860253164556962, "grad_norm": 0.19941691321842164, "learning_rate": 5.867901545738976e-06, "loss": 0.3127, "step": 953 }, { "epoch": 3.8643037974683545, "grad_norm": 0.19906169377118452, "learning_rate": 5.827795600133774e-06, "loss": 0.2994, "step": 954 }, { "epoch": 3.868354430379747, "grad_norm": 0.19186071421772735, "learning_rate": 5.787803795514466e-06, "loss": 0.293, "step": 955 }, { "epoch": 3.872405063291139, "grad_norm": 0.19755593861479698, "learning_rate": 5.747926453969576e-06, "loss": 0.2944, "step": 956 }, { "epoch": 3.8764556962025316, "grad_norm": 0.19655151357974573, "learning_rate": 5.708163896665708e-06, "loss": 0.2978, "step": 957 }, { "epoch": 3.880506329113924, "grad_norm": 0.18897213239067184, "learning_rate": 5.668516443845047e-06, "loss": 0.2956, "step": 958 }, { "epoch": 3.884556962025316, "grad_norm": 0.19315259229331935, "learning_rate": 5.6289844148227225e-06, "loss": 0.2937, "step": 959 }, { "epoch": 3.8886075949367087, "grad_norm": 0.19871637758535193, "learning_rate": 5.5895681279842615e-06, "loss": 0.2995, "step": 960 }, { "epoch": 3.8926582278481012, "grad_norm": 0.18722295209578066, "learning_rate": 5.550267900783019e-06, "loss": 0.3069, "step": 961 }, { "epoch": 3.8967088607594937, "grad_norm": 0.19744570712395706, "learning_rate": 5.511084049737623e-06, "loss": 0.3058, "step": 962 }, { "epoch": 3.9007594936708863, "grad_norm": 0.18383770538934693, "learning_rate": 5.4720168904294215e-06, "loss": 0.2915, "step": 963 }, { "epoch": 3.9048101265822783, "grad_norm": 0.19228767857907167, "learning_rate": 5.433066737499948e-06, "loss": 0.2919, "step": 964 }, { "epoch": 3.908860759493671, "grad_norm": 0.17786635081639676, "learning_rate": 5.394233904648376e-06, "loss": 0.2859, "step": 965 }, { "epoch": 3.9129113924050634, "grad_norm": 0.19617606388263256, "learning_rate": 5.355518704628997e-06, "loss": 0.3016, "step": 966 }, { "epoch": 3.9169620253164554, "grad_norm": 0.18911855093816834, "learning_rate": 5.316921449248731e-06, "loss": 0.2978, "step": 967 }, { "epoch": 3.921012658227848, "grad_norm": 0.18915906429618912, "learning_rate": 5.278442449364538e-06, "loss": 0.2894, "step": 968 }, { "epoch": 3.9250632911392405, "grad_norm": 0.19281337117658862, "learning_rate": 5.240082014881016e-06, "loss": 0.2985, "step": 969 }, { "epoch": 3.929113924050633, "grad_norm": 0.3237093413532034, "learning_rate": 5.201840454747822e-06, "loss": 0.303, "step": 970 }, { "epoch": 3.9331645569620255, "grad_norm": 0.18885926736036582, "learning_rate": 5.163718076957223e-06, "loss": 0.2944, "step": 971 }, { "epoch": 3.9372151898734176, "grad_norm": 0.1906957371789111, "learning_rate": 5.125715188541609e-06, "loss": 0.2935, "step": 972 }, { "epoch": 3.94126582278481, "grad_norm": 0.18144051509436201, "learning_rate": 5.087832095571021e-06, "loss": 0.2915, "step": 973 }, { "epoch": 3.9453164556962026, "grad_norm": 0.18867265519206908, "learning_rate": 5.0500691031506766e-06, "loss": 0.2846, "step": 974 }, { "epoch": 3.9493670886075947, "grad_norm": 0.19812622126284624, "learning_rate": 5.01242651541854e-06, "loss": 0.2972, "step": 975 }, { "epoch": 3.953417721518987, "grad_norm": 0.19931523633742118, "learning_rate": 4.974904635542815e-06, "loss": 0.2876, "step": 976 }, { "epoch": 3.9574683544303797, "grad_norm": 0.17844064483796118, "learning_rate": 4.937503765719582e-06, "loss": 0.2973, "step": 977 }, { "epoch": 3.961518987341772, "grad_norm": 0.18874265117744374, "learning_rate": 4.900224207170299e-06, "loss": 0.3014, "step": 978 }, { "epoch": 3.9655696202531647, "grad_norm": 0.19351637889377732, "learning_rate": 4.8630662601394065e-06, "loss": 0.2976, "step": 979 }, { "epoch": 3.969620253164557, "grad_norm": 0.19672784637423588, "learning_rate": 4.8260302238918995e-06, "loss": 0.3056, "step": 980 }, { "epoch": 3.9736708860759493, "grad_norm": 0.1907869153772834, "learning_rate": 4.789116396710924e-06, "loss": 0.3095, "step": 981 }, { "epoch": 3.977721518987342, "grad_norm": 0.1849155518049543, "learning_rate": 4.752325075895368e-06, "loss": 0.2899, "step": 982 }, { "epoch": 3.981772151898734, "grad_norm": 0.18379128361031877, "learning_rate": 4.715656557757473e-06, "loss": 0.3055, "step": 983 }, { "epoch": 3.9858227848101264, "grad_norm": 0.200229406417035, "learning_rate": 4.679111137620442e-06, "loss": 0.2962, "step": 984 }, { "epoch": 3.989873417721519, "grad_norm": 0.18451930914554604, "learning_rate": 4.6426891098160585e-06, "loss": 0.2956, "step": 985 }, { "epoch": 3.9939240506329114, "grad_norm": 0.23524232045369814, "learning_rate": 4.6063907676823474e-06, "loss": 0.3004, "step": 986 }, { "epoch": 3.997974683544304, "grad_norm": 0.18857421238805425, "learning_rate": 4.570216403561141e-06, "loss": 0.3089, "step": 987 }, { "epoch": 4.002025316455696, "grad_norm": 0.23806815593390587, "learning_rate": 4.534166308795815e-06, "loss": 0.2789, "step": 988 }, { "epoch": 4.006075949367089, "grad_norm": 0.2906065808399264, "learning_rate": 4.498240773728859e-06, "loss": 0.2697, "step": 989 }, { "epoch": 4.010126582278481, "grad_norm": 0.1942777529665251, "learning_rate": 4.462440087699609e-06, "loss": 0.2706, "step": 990 }, { "epoch": 4.014177215189873, "grad_norm": 0.32888421581328126, "learning_rate": 4.426764539041861e-06, "loss": 0.2685, "step": 991 }, { "epoch": 4.018227848101266, "grad_norm": 0.2986968246966496, "learning_rate": 4.391214415081582e-06, "loss": 0.2512, "step": 992 }, { "epoch": 4.022278481012658, "grad_norm": 0.2175468456731634, "learning_rate": 4.355790002134579e-06, "loss": 0.2671, "step": 993 }, { "epoch": 4.02632911392405, "grad_norm": 0.2655713388614624, "learning_rate": 4.320491585504207e-06, "loss": 0.2602, "step": 994 }, { "epoch": 4.030379746835443, "grad_norm": 0.31778811908446913, "learning_rate": 4.2853194494790615e-06, "loss": 0.2603, "step": 995 }, { "epoch": 4.034430379746835, "grad_norm": 0.2515793110502782, "learning_rate": 4.250273877330691e-06, "loss": 0.2547, "step": 996 }, { "epoch": 4.038481012658228, "grad_norm": 0.19995784819070922, "learning_rate": 4.215355151311313e-06, "loss": 0.2621, "step": 997 }, { "epoch": 4.04253164556962, "grad_norm": 0.2586452311742753, "learning_rate": 4.180563552651542e-06, "loss": 0.2639, "step": 998 }, { "epoch": 4.046582278481012, "grad_norm": 0.2718861770632561, "learning_rate": 4.145899361558147e-06, "loss": 0.2675, "step": 999 }, { "epoch": 4.050632911392405, "grad_norm": 0.21056241709670676, "learning_rate": 4.111362857211738e-06, "loss": 0.2639, "step": 1000 }, { "epoch": 4.054683544303797, "grad_norm": 0.1991360270728414, "learning_rate": 4.076954317764592e-06, "loss": 0.2644, "step": 1001 }, { "epoch": 4.0587341772151895, "grad_norm": 0.23210828524992688, "learning_rate": 4.042674020338335e-06, "loss": 0.2674, "step": 1002 }, { "epoch": 4.062784810126582, "grad_norm": 0.22376685551287578, "learning_rate": 4.0085222410217835e-06, "loss": 0.2636, "step": 1003 }, { "epoch": 4.0668354430379745, "grad_norm": 0.19091894520215077, "learning_rate": 3.974499254868674e-06, "loss": 0.2644, "step": 1004 }, { "epoch": 4.0708860759493675, "grad_norm": 0.19161730574760916, "learning_rate": 3.940605335895451e-06, "loss": 0.2659, "step": 1005 }, { "epoch": 4.0749367088607595, "grad_norm": 0.22670797194833556, "learning_rate": 3.90684075707908e-06, "loss": 0.2661, "step": 1006 }, { "epoch": 4.078987341772152, "grad_norm": 0.21697093164630576, "learning_rate": 3.8732057903548505e-06, "loss": 0.264, "step": 1007 }, { "epoch": 4.083037974683545, "grad_norm": 0.1875726831530171, "learning_rate": 3.8397007066141375e-06, "loss": 0.2616, "step": 1008 }, { "epoch": 4.087088607594937, "grad_norm": 0.20131154894915798, "learning_rate": 3.806325775702304e-06, "loss": 0.2638, "step": 1009 }, { "epoch": 4.091139240506329, "grad_norm": 0.19558111730187658, "learning_rate": 3.773081266416434e-06, "loss": 0.2606, "step": 1010 }, { "epoch": 4.095189873417722, "grad_norm": 0.1867493741695061, "learning_rate": 3.739967446503245e-06, "loss": 0.2653, "step": 1011 }, { "epoch": 4.099240506329114, "grad_norm": 0.18408659550536433, "learning_rate": 3.706984582656894e-06, "loss": 0.264, "step": 1012 }, { "epoch": 4.103291139240507, "grad_norm": 0.1870673206346807, "learning_rate": 3.6741329405168237e-06, "loss": 0.2536, "step": 1013 }, { "epoch": 4.107341772151899, "grad_norm": 0.19250519703715271, "learning_rate": 3.641412784665648e-06, "loss": 0.2556, "step": 1014 }, { "epoch": 4.111392405063291, "grad_norm": 0.19064696707826026, "learning_rate": 3.608824378627005e-06, "loss": 0.2502, "step": 1015 }, { "epoch": 4.115443037974684, "grad_norm": 0.19507138908426205, "learning_rate": 3.5763679848634337e-06, "loss": 0.2623, "step": 1016 }, { "epoch": 4.119493670886076, "grad_norm": 0.1845611288188549, "learning_rate": 3.544043864774269e-06, "loss": 0.255, "step": 1017 }, { "epoch": 4.123544303797468, "grad_norm": 0.18362461976498917, "learning_rate": 3.5118522786935282e-06, "loss": 0.2613, "step": 1018 }, { "epoch": 4.127594936708861, "grad_norm": 0.18289735156418846, "learning_rate": 3.479793485887819e-06, "loss": 0.2683, "step": 1019 }, { "epoch": 4.131645569620253, "grad_norm": 0.1881178360965204, "learning_rate": 3.4478677445542653e-06, "loss": 0.2633, "step": 1020 }, { "epoch": 4.135696202531646, "grad_norm": 0.1818853921495059, "learning_rate": 3.4160753118183767e-06, "loss": 0.2675, "step": 1021 }, { "epoch": 4.139746835443038, "grad_norm": 0.19198378234006924, "learning_rate": 3.3844164437320527e-06, "loss": 0.253, "step": 1022 }, { "epoch": 4.14379746835443, "grad_norm": 0.18394771831142784, "learning_rate": 3.3528913952714558e-06, "loss": 0.2688, "step": 1023 }, { "epoch": 4.147848101265823, "grad_norm": 0.1912628379730283, "learning_rate": 3.321500420335e-06, "loss": 0.2503, "step": 1024 }, { "epoch": 4.151898734177215, "grad_norm": 0.19458437942799844, "learning_rate": 3.290243771741275e-06, "loss": 0.2656, "step": 1025 }, { "epoch": 4.155949367088607, "grad_norm": 0.17890893149303397, "learning_rate": 3.2591217012270325e-06, "loss": 0.2704, "step": 1026 }, { "epoch": 4.16, "grad_norm": 0.18402565942343857, "learning_rate": 3.228134459445149e-06, "loss": 0.2698, "step": 1027 }, { "epoch": 4.164050632911392, "grad_norm": 0.1987470325632958, "learning_rate": 3.1972822959626205e-06, "loss": 0.268, "step": 1028 }, { "epoch": 4.168101265822785, "grad_norm": 0.1889957996257092, "learning_rate": 3.166565459258513e-06, "loss": 0.2551, "step": 1029 }, { "epoch": 4.172151898734177, "grad_norm": 0.18075641860482247, "learning_rate": 3.1359841967220193e-06, "loss": 0.2731, "step": 1030 }, { "epoch": 4.176202531645569, "grad_norm": 0.17876625164770493, "learning_rate": 3.105538754650419e-06, "loss": 0.2534, "step": 1031 }, { "epoch": 4.180253164556962, "grad_norm": 0.18018581098755787, "learning_rate": 3.07522937824712e-06, "loss": 0.2538, "step": 1032 }, { "epoch": 4.184303797468354, "grad_norm": 0.18188083017057405, "learning_rate": 3.0450563116196697e-06, "loss": 0.2813, "step": 1033 }, { "epoch": 4.188354430379746, "grad_norm": 0.1855595782773524, "learning_rate": 3.0150197977778008e-06, "loss": 0.2726, "step": 1034 }, { "epoch": 4.192405063291139, "grad_norm": 0.17337017549762215, "learning_rate": 2.985120078631465e-06, "loss": 0.262, "step": 1035 }, { "epoch": 4.1964556962025314, "grad_norm": 0.17679294437726914, "learning_rate": 2.9553573949888893e-06, "loss": 0.2482, "step": 1036 }, { "epoch": 4.200506329113924, "grad_norm": 0.18260931512248557, "learning_rate": 2.9257319865546384e-06, "loss": 0.268, "step": 1037 }, { "epoch": 4.2045569620253165, "grad_norm": 0.17487397419923512, "learning_rate": 2.896244091927678e-06, "loss": 0.2683, "step": 1038 }, { "epoch": 4.2086075949367086, "grad_norm": 0.19036463880758253, "learning_rate": 2.8668939485994584e-06, "loss": 0.2582, "step": 1039 }, { "epoch": 4.2126582278481015, "grad_norm": 0.1827577346801575, "learning_rate": 2.837681792951994e-06, "loss": 0.2617, "step": 1040 }, { "epoch": 4.216708860759494, "grad_norm": 0.18165449852279816, "learning_rate": 2.808607860255981e-06, "loss": 0.269, "step": 1041 }, { "epoch": 4.220759493670886, "grad_norm": 0.18719311596462815, "learning_rate": 2.7796723846688634e-06, "loss": 0.2501, "step": 1042 }, { "epoch": 4.224810126582279, "grad_norm": 0.17580376827069216, "learning_rate": 2.7508755992329937e-06, "loss": 0.2619, "step": 1043 }, { "epoch": 4.228860759493671, "grad_norm": 0.1771894211333577, "learning_rate": 2.722217735873718e-06, "loss": 0.2622, "step": 1044 }, { "epoch": 4.232911392405064, "grad_norm": 0.17820447654834146, "learning_rate": 2.6936990253975315e-06, "loss": 0.2632, "step": 1045 }, { "epoch": 4.236962025316456, "grad_norm": 0.18138170026149694, "learning_rate": 2.665319697490205e-06, "loss": 0.2852, "step": 1046 }, { "epoch": 4.241012658227848, "grad_norm": 0.18813451832693628, "learning_rate": 2.637079980714945e-06, "loss": 0.2656, "step": 1047 }, { "epoch": 4.245063291139241, "grad_norm": 0.17796112904087688, "learning_rate": 2.6089801025105453e-06, "loss": 0.2522, "step": 1048 }, { "epoch": 4.249113924050633, "grad_norm": 0.17704152419901772, "learning_rate": 2.581020289189571e-06, "loss": 0.2584, "step": 1049 }, { "epoch": 4.253164556962025, "grad_norm": 0.1740878725959348, "learning_rate": 2.553200765936501e-06, "loss": 0.2586, "step": 1050 }, { "epoch": 4.257215189873418, "grad_norm": 0.17455349995849795, "learning_rate": 2.525521756805962e-06, "loss": 0.2659, "step": 1051 }, { "epoch": 4.26126582278481, "grad_norm": 0.17899183398206314, "learning_rate": 2.497983484720885e-06, "loss": 0.2646, "step": 1052 }, { "epoch": 4.265316455696203, "grad_norm": 0.18924275902926224, "learning_rate": 2.470586171470728e-06, "loss": 0.2536, "step": 1053 }, { "epoch": 4.269367088607595, "grad_norm": 0.18060290160170783, "learning_rate": 2.4433300377096836e-06, "loss": 0.265, "step": 1054 }, { "epoch": 4.273417721518987, "grad_norm": 0.1826537450626836, "learning_rate": 2.4162153029549073e-06, "loss": 0.2664, "step": 1055 }, { "epoch": 4.27746835443038, "grad_norm": 0.1832544930919502, "learning_rate": 2.3892421855847458e-06, "loss": 0.2567, "step": 1056 }, { "epoch": 4.281518987341772, "grad_norm": 0.18065500657179628, "learning_rate": 2.362410902836978e-06, "loss": 0.255, "step": 1057 }, { "epoch": 4.285569620253165, "grad_norm": 0.19087569644625257, "learning_rate": 2.3357216708070653e-06, "loss": 0.2516, "step": 1058 }, { "epoch": 4.289620253164557, "grad_norm": 0.23823623291056742, "learning_rate": 2.309174704446411e-06, "loss": 0.2657, "step": 1059 }, { "epoch": 4.293670886075949, "grad_norm": 0.17335476726164542, "learning_rate": 2.2827702175606437e-06, "loss": 0.2679, "step": 1060 }, { "epoch": 4.297721518987342, "grad_norm": 0.22188394998193287, "learning_rate": 2.256508422807855e-06, "loss": 0.2679, "step": 1061 }, { "epoch": 4.301772151898734, "grad_norm": 0.18567042125215708, "learning_rate": 2.230389531696946e-06, "loss": 0.2622, "step": 1062 }, { "epoch": 4.305822784810126, "grad_norm": 0.1772909982360832, "learning_rate": 2.204413754585857e-06, "loss": 0.2573, "step": 1063 }, { "epoch": 4.309873417721519, "grad_norm": 0.17831969939504996, "learning_rate": 2.1785813006799406e-06, "loss": 0.249, "step": 1064 }, { "epoch": 4.313924050632911, "grad_norm": 0.18192199955630795, "learning_rate": 2.1528923780302224e-06, "loss": 0.2584, "step": 1065 }, { "epoch": 4.317974683544303, "grad_norm": 0.1858507703864459, "learning_rate": 2.127347193531757e-06, "loss": 0.2578, "step": 1066 }, { "epoch": 4.322025316455696, "grad_norm": 0.1732408907884258, "learning_rate": 2.101945952921942e-06, "loss": 0.2709, "step": 1067 }, { "epoch": 4.326075949367088, "grad_norm": 0.1716996951330163, "learning_rate": 2.0766888607788906e-06, "loss": 0.2588, "step": 1068 }, { "epoch": 4.330126582278481, "grad_norm": 0.17184410659981955, "learning_rate": 2.0515761205197337e-06, "loss": 0.2736, "step": 1069 }, { "epoch": 4.334177215189873, "grad_norm": 0.1699426110071505, "learning_rate": 2.0266079343990453e-06, "loss": 0.2657, "step": 1070 }, { "epoch": 4.3382278481012655, "grad_norm": 0.16860337153783694, "learning_rate": 2.0017845035071494e-06, "loss": 0.2571, "step": 1071 }, { "epoch": 4.3422784810126585, "grad_norm": 0.17363117883404766, "learning_rate": 1.9771060277685537e-06, "loss": 0.2624, "step": 1072 }, { "epoch": 4.3463291139240505, "grad_norm": 0.17407961841824235, "learning_rate": 1.95257270594031e-06, "loss": 0.2706, "step": 1073 }, { "epoch": 4.3503797468354435, "grad_norm": 0.1719172795806497, "learning_rate": 1.9281847356104188e-06, "loss": 0.2553, "step": 1074 }, { "epoch": 4.3544303797468356, "grad_norm": 0.20607035727810488, "learning_rate": 1.9039423131962365e-06, "loss": 0.2593, "step": 1075 }, { "epoch": 4.358481012658228, "grad_norm": 0.169223852540592, "learning_rate": 1.8798456339429027e-06, "loss": 0.2621, "step": 1076 }, { "epoch": 4.362531645569621, "grad_norm": 0.21241319545913523, "learning_rate": 1.8558948919217612e-06, "loss": 0.2633, "step": 1077 }, { "epoch": 4.366582278481013, "grad_norm": 0.1734312845798049, "learning_rate": 1.8320902800287954e-06, "loss": 0.2631, "step": 1078 }, { "epoch": 4.370632911392405, "grad_norm": 0.17304753760392294, "learning_rate": 1.8084319899830726e-06, "loss": 0.2656, "step": 1079 }, { "epoch": 4.374683544303798, "grad_norm": 0.17081992943466603, "learning_rate": 1.7849202123252097e-06, "loss": 0.2575, "step": 1080 }, { "epoch": 4.37873417721519, "grad_norm": 0.17708054840438936, "learning_rate": 1.7615551364158401e-06, "loss": 0.2551, "step": 1081 }, { "epoch": 4.382784810126582, "grad_norm": 0.16677253918936719, "learning_rate": 1.738336950434061e-06, "loss": 0.2697, "step": 1082 }, { "epoch": 4.386835443037975, "grad_norm": 0.17591888419356735, "learning_rate": 1.715265841375957e-06, "loss": 0.2617, "step": 1083 }, { "epoch": 4.390886075949367, "grad_norm": 0.18144850223753167, "learning_rate": 1.6923419950530684e-06, "loss": 0.2678, "step": 1084 }, { "epoch": 4.39493670886076, "grad_norm": 0.1745107581378683, "learning_rate": 1.6695655960909008e-06, "loss": 0.257, "step": 1085 }, { "epoch": 4.398987341772152, "grad_norm": 0.17637367960649716, "learning_rate": 1.646936827927441e-06, "loss": 0.2573, "step": 1086 }, { "epoch": 4.403037974683544, "grad_norm": 0.1744419237406387, "learning_rate": 1.6244558728116766e-06, "loss": 0.2519, "step": 1087 }, { "epoch": 4.407088607594937, "grad_norm": 0.22166614944932445, "learning_rate": 1.6021229118021265e-06, "loss": 0.2732, "step": 1088 }, { "epoch": 4.411139240506329, "grad_norm": 0.17884441653446817, "learning_rate": 1.5799381247653967e-06, "loss": 0.2588, "step": 1089 }, { "epoch": 4.415189873417722, "grad_norm": 0.17514082517249915, "learning_rate": 1.5579016903747013e-06, "loss": 0.2568, "step": 1090 }, { "epoch": 4.419240506329114, "grad_norm": 0.17973717404476688, "learning_rate": 1.5360137861084656e-06, "loss": 0.2722, "step": 1091 }, { "epoch": 4.423291139240506, "grad_norm": 0.17656125911461457, "learning_rate": 1.5142745882488475e-06, "loss": 0.2503, "step": 1092 }, { "epoch": 4.427341772151899, "grad_norm": 0.17717832524473356, "learning_rate": 1.4926842718803691e-06, "loss": 0.255, "step": 1093 }, { "epoch": 4.431392405063291, "grad_norm": 0.17556570157851464, "learning_rate": 1.4712430108884657e-06, "loss": 0.2674, "step": 1094 }, { "epoch": 4.435443037974683, "grad_norm": 0.1734663949833259, "learning_rate": 1.4499509779581078e-06, "loss": 0.2661, "step": 1095 }, { "epoch": 4.439493670886076, "grad_norm": 0.17290488757291522, "learning_rate": 1.4288083445723988e-06, "loss": 0.2633, "step": 1096 }, { "epoch": 4.443544303797468, "grad_norm": 0.17509546071027032, "learning_rate": 1.4078152810112045e-06, "loss": 0.2602, "step": 1097 }, { "epoch": 4.44759493670886, "grad_norm": 0.171152697515967, "learning_rate": 1.3869719563497697e-06, "loss": 0.2607, "step": 1098 }, { "epoch": 4.451645569620253, "grad_norm": 0.1812658473916176, "learning_rate": 1.3662785384573663e-06, "loss": 0.2667, "step": 1099 }, { "epoch": 4.455696202531645, "grad_norm": 0.20496886906898065, "learning_rate": 1.3457351939959383e-06, "loss": 0.2663, "step": 1100 }, { "epoch": 4.459746835443038, "grad_norm": 0.17070154600989423, "learning_rate": 1.3253420884187551e-06, "loss": 0.2572, "step": 1101 }, { "epoch": 4.46379746835443, "grad_norm": 0.17571662733326188, "learning_rate": 1.3050993859690953e-06, "loss": 0.2699, "step": 1102 }, { "epoch": 4.467848101265822, "grad_norm": 0.17239957800367287, "learning_rate": 1.2850072496788869e-06, "loss": 0.269, "step": 1103 }, { "epoch": 4.471898734177215, "grad_norm": 0.17172641836345212, "learning_rate": 1.2650658413674434e-06, "loss": 0.2558, "step": 1104 }, { "epoch": 4.4759493670886075, "grad_norm": 0.16641144112295692, "learning_rate": 1.2452753216401226e-06, "loss": 0.2517, "step": 1105 }, { "epoch": 4.48, "grad_norm": 0.16989553258420625, "learning_rate": 1.2256358498870503e-06, "loss": 0.2642, "step": 1106 }, { "epoch": 4.4840506329113925, "grad_norm": 0.17683292562836392, "learning_rate": 1.2061475842818337e-06, "loss": 0.2629, "step": 1107 }, { "epoch": 4.488101265822785, "grad_norm": 0.16858202485394183, "learning_rate": 1.1868106817802816e-06, "loss": 0.273, "step": 1108 }, { "epoch": 4.4921518987341775, "grad_norm": 0.17686903014429978, "learning_rate": 1.1676252981191482e-06, "loss": 0.2701, "step": 1109 }, { "epoch": 4.49620253164557, "grad_norm": 0.17364858129258454, "learning_rate": 1.1485915878148823e-06, "loss": 0.2642, "step": 1110 }, { "epoch": 4.500253164556962, "grad_norm": 0.1805671581944283, "learning_rate": 1.1297097041623584e-06, "loss": 0.2695, "step": 1111 }, { "epoch": 4.504303797468355, "grad_norm": 0.17523384988756346, "learning_rate": 1.1109797992336847e-06, "loss": 0.2587, "step": 1112 }, { "epoch": 4.508354430379747, "grad_norm": 0.16880575855612254, "learning_rate": 1.092402023876933e-06, "loss": 0.2655, "step": 1113 }, { "epoch": 4.512405063291139, "grad_norm": 0.17779705988986513, "learning_rate": 1.0739765277149527e-06, "loss": 0.2556, "step": 1114 }, { "epoch": 4.516455696202532, "grad_norm": 0.18507369176859664, "learning_rate": 1.0557034591441596e-06, "loss": 0.2463, "step": 1115 }, { "epoch": 4.520506329113924, "grad_norm": 0.1699657852597437, "learning_rate": 1.0375829653333324e-06, "loss": 0.2712, "step": 1116 }, { "epoch": 4.524556962025317, "grad_norm": 0.17537014369405374, "learning_rate": 1.0196151922224385e-06, "loss": 0.2711, "step": 1117 }, { "epoch": 4.528607594936709, "grad_norm": 0.20079492167294405, "learning_rate": 1.0018002845214526e-06, "loss": 0.2671, "step": 1118 }, { "epoch": 4.532658227848101, "grad_norm": 0.17242904240957016, "learning_rate": 9.841383857091947e-07, "loss": 0.2687, "step": 1119 }, { "epoch": 4.536708860759494, "grad_norm": 0.16155165532320995, "learning_rate": 9.666296380321616e-07, "loss": 0.2594, "step": 1120 }, { "epoch": 4.540759493670886, "grad_norm": 0.16602705206336374, "learning_rate": 9.492741825034124e-07, "loss": 0.2566, "step": 1121 }, { "epoch": 4.544810126582279, "grad_norm": 0.1761553506595148, "learning_rate": 9.320721589013892e-07, "loss": 0.265, "step": 1122 }, { "epoch": 4.548860759493671, "grad_norm": 0.16895240737436515, "learning_rate": 9.150237057688339e-07, "loss": 0.2676, "step": 1123 }, { "epoch": 4.552911392405063, "grad_norm": 0.1670968272122602, "learning_rate": 8.981289604116328e-07, "loss": 0.2677, "step": 1124 }, { "epoch": 4.556962025316456, "grad_norm": 0.17175053986959368, "learning_rate": 8.813880588977542e-07, "loss": 0.2692, "step": 1125 }, { "epoch": 4.561012658227848, "grad_norm": 0.16797885564114112, "learning_rate": 8.648011360561126e-07, "loss": 0.278, "step": 1126 }, { "epoch": 4.56506329113924, "grad_norm": 0.17835443853202926, "learning_rate": 8.483683254755037e-07, "loss": 0.2725, "step": 1127 }, { "epoch": 4.569113924050633, "grad_norm": 0.17364709667738087, "learning_rate": 8.320897595035227e-07, "loss": 0.2615, "step": 1128 }, { "epoch": 4.573164556962025, "grad_norm": 0.16545328628336678, "learning_rate": 8.159655692455093e-07, "loss": 0.265, "step": 1129 }, { "epoch": 4.577215189873417, "grad_norm": 0.17396119959111145, "learning_rate": 7.999958845634648e-07, "loss": 0.2666, "step": 1130 }, { "epoch": 4.58126582278481, "grad_norm": 0.16846419453989353, "learning_rate": 7.841808340750478e-07, "loss": 0.247, "step": 1131 }, { "epoch": 4.585316455696202, "grad_norm": 0.16739845172510215, "learning_rate": 7.685205451524869e-07, "loss": 0.2551, "step": 1132 }, { "epoch": 4.589367088607595, "grad_norm": 0.1760782524347757, "learning_rate": 7.530151439216027e-07, "loss": 0.2652, "step": 1133 }, { "epoch": 4.593417721518987, "grad_norm": 0.16891134033164168, "learning_rate": 7.376647552607675e-07, "loss": 0.2546, "step": 1134 }, { "epoch": 4.597468354430379, "grad_norm": 0.17818065413227635, "learning_rate": 7.224695027998963e-07, "loss": 0.2647, "step": 1135 }, { "epoch": 4.601518987341772, "grad_norm": 0.1638385940723332, "learning_rate": 7.07429508919466e-07, "loss": 0.2695, "step": 1136 }, { "epoch": 4.605569620253164, "grad_norm": 0.17065748402797515, "learning_rate": 6.925448947495206e-07, "loss": 0.2551, "step": 1137 }, { "epoch": 4.609620253164557, "grad_norm": 0.176089892700094, "learning_rate": 6.778157801686936e-07, "loss": 0.2643, "step": 1138 }, { "epoch": 4.613670886075949, "grad_norm": 0.1761189102801655, "learning_rate": 6.632422838032515e-07, "loss": 0.2752, "step": 1139 }, { "epoch": 4.6177215189873415, "grad_norm": 0.16489662760018264, "learning_rate": 6.488245230261281e-07, "loss": 0.2585, "step": 1140 }, { "epoch": 4.6217721518987345, "grad_norm": 0.21146479140148242, "learning_rate": 6.345626139559868e-07, "loss": 0.2575, "step": 1141 }, { "epoch": 4.6258227848101265, "grad_norm": 0.1843251571392768, "learning_rate": 6.204566714562866e-07, "loss": 0.2488, "step": 1142 }, { "epoch": 4.629873417721519, "grad_norm": 0.17302985071701743, "learning_rate": 6.06506809134344e-07, "loss": 0.2493, "step": 1143 }, { "epoch": 4.633924050632912, "grad_norm": 0.1663029242783467, "learning_rate": 5.927131393404373e-07, "loss": 0.2674, "step": 1144 }, { "epoch": 4.637974683544304, "grad_norm": 0.18321307592028582, "learning_rate": 5.790757731668817e-07, "loss": 0.2484, "step": 1145 }, { "epoch": 4.642025316455696, "grad_norm": 0.1730121084126217, "learning_rate": 5.655948204471507e-07, "loss": 0.2683, "step": 1146 }, { "epoch": 4.646075949367089, "grad_norm": 0.17159015598252067, "learning_rate": 5.522703897549875e-07, "loss": 0.261, "step": 1147 }, { "epoch": 4.650126582278481, "grad_norm": 0.17292228124257294, "learning_rate": 5.391025884035239e-07, "loss": 0.2615, "step": 1148 }, { "epoch": 4.654177215189874, "grad_norm": 0.17342590933189148, "learning_rate": 5.260915224444207e-07, "loss": 0.2774, "step": 1149 }, { "epoch": 4.658227848101266, "grad_norm": 0.16759139718021196, "learning_rate": 5.132372966670129e-07, "loss": 0.262, "step": 1150 }, { "epoch": 4.662278481012658, "grad_norm": 0.1793017349928622, "learning_rate": 5.005400145974704e-07, "loss": 0.2448, "step": 1151 }, { "epoch": 4.666329113924051, "grad_norm": 0.1795280148418071, "learning_rate": 4.879997784979562e-07, "loss": 0.2557, "step": 1152 }, { "epoch": 4.670379746835443, "grad_norm": 0.1690257358813071, "learning_rate": 4.7561668936580984e-07, "loss": 0.2585, "step": 1153 }, { "epoch": 4.674430379746836, "grad_norm": 0.160996333307283, "learning_rate": 4.6339084693272306e-07, "loss": 0.266, "step": 1154 }, { "epoch": 4.678481012658228, "grad_norm": 0.16835908599094015, "learning_rate": 4.5132234966395847e-07, "loss": 0.2688, "step": 1155 }, { "epoch": 4.68253164556962, "grad_norm": 0.17240198160752407, "learning_rate": 4.3941129475752795e-07, "loss": 0.2739, "step": 1156 }, { "epoch": 4.686582278481013, "grad_norm": 0.17345647069935338, "learning_rate": 4.27657778143431e-07, "loss": 0.2695, "step": 1157 }, { "epoch": 4.690632911392405, "grad_norm": 0.16560486207954836, "learning_rate": 4.1606189448287757e-07, "loss": 0.2614, "step": 1158 }, { "epoch": 4.694683544303797, "grad_norm": 0.17003710927059454, "learning_rate": 4.046237371675177e-07, "loss": 0.2695, "step": 1159 }, { "epoch": 4.69873417721519, "grad_norm": 0.16075012175171147, "learning_rate": 3.9334339831869963e-07, "loss": 0.2594, "step": 1160 }, { "epoch": 4.702784810126582, "grad_norm": 0.16455733557489655, "learning_rate": 3.8222096878671955e-07, "loss": 0.2667, "step": 1161 }, { "epoch": 4.706835443037974, "grad_norm": 0.17592000414915118, "learning_rate": 3.7125653815009545e-07, "loss": 0.2634, "step": 1162 }, { "epoch": 4.710886075949367, "grad_norm": 0.1650124187357302, "learning_rate": 3.6045019471484974e-07, "loss": 0.2553, "step": 1163 }, { "epoch": 4.714936708860759, "grad_norm": 0.17913740288764143, "learning_rate": 3.498020255137813e-07, "loss": 0.2612, "step": 1164 }, { "epoch": 4.718987341772152, "grad_norm": 0.17191594861388748, "learning_rate": 3.393121163057811e-07, "loss": 0.2691, "step": 1165 }, { "epoch": 4.723037974683544, "grad_norm": 0.16936218781142345, "learning_rate": 3.289805515751399e-07, "loss": 0.2657, "step": 1166 }, { "epoch": 4.727088607594936, "grad_norm": 0.1764064095354331, "learning_rate": 3.188074145308573e-07, "loss": 0.2689, "step": 1167 }, { "epoch": 4.731139240506329, "grad_norm": 0.9363333716987495, "learning_rate": 3.087927871059804e-07, "loss": 0.2771, "step": 1168 }, { "epoch": 4.735189873417721, "grad_norm": 0.17246813520979618, "learning_rate": 2.989367499569418e-07, "loss": 0.2629, "step": 1169 }, { "epoch": 4.739240506329114, "grad_norm": 0.17132566416677933, "learning_rate": 2.8923938246290917e-07, "loss": 0.2571, "step": 1170 }, { "epoch": 4.743291139240506, "grad_norm": 0.17077322000826106, "learning_rate": 2.7970076272514804e-07, "loss": 0.2653, "step": 1171 }, { "epoch": 4.7473417721518985, "grad_norm": 0.17496297747564346, "learning_rate": 2.703209675663887e-07, "loss": 0.2513, "step": 1172 }, { "epoch": 4.751392405063291, "grad_norm": 0.16568321734963254, "learning_rate": 2.6110007253021374e-07, "loss": 0.2565, "step": 1173 }, { "epoch": 4.7554430379746835, "grad_norm": 0.1676968624984896, "learning_rate": 2.520381518804471e-07, "loss": 0.2707, "step": 1174 }, { "epoch": 4.759493670886076, "grad_norm": 0.16255724915581296, "learning_rate": 2.4313527860054585e-07, "loss": 0.2659, "step": 1175 }, { "epoch": 4.7635443037974685, "grad_norm": 0.1733946118632074, "learning_rate": 2.343915243930317e-07, "loss": 0.2632, "step": 1176 }, { "epoch": 4.767594936708861, "grad_norm": 0.1690650834061682, "learning_rate": 2.2580695967889367e-07, "loss": 0.2522, "step": 1177 }, { "epoch": 4.7716455696202535, "grad_norm": 0.17199295704018536, "learning_rate": 2.1738165359704189e-07, "loss": 0.2717, "step": 1178 }, { "epoch": 4.775696202531646, "grad_norm": 0.17485054499719305, "learning_rate": 2.0911567400373257e-07, "loss": 0.2643, "step": 1179 }, { "epoch": 4.779746835443038, "grad_norm": 0.17055954184753258, "learning_rate": 2.0100908747202607e-07, "loss": 0.2607, "step": 1180 }, { "epoch": 4.783797468354431, "grad_norm": 0.18374415324199886, "learning_rate": 1.9306195929125638e-07, "loss": 0.2587, "step": 1181 }, { "epoch": 4.787848101265823, "grad_norm": 0.16701879190609126, "learning_rate": 1.8527435346650247e-07, "loss": 0.2687, "step": 1182 }, { "epoch": 4.791898734177215, "grad_norm": 0.16527639355456647, "learning_rate": 1.7764633271807108e-07, "loss": 0.2691, "step": 1183 }, { "epoch": 4.795949367088608, "grad_norm": 0.1675679540992759, "learning_rate": 1.7017795848099262e-07, "loss": 0.268, "step": 1184 }, { "epoch": 4.8, "grad_norm": 0.18173024857810147, "learning_rate": 1.6286929090452596e-07, "loss": 0.2701, "step": 1185 }, { "epoch": 4.804050632911393, "grad_norm": 0.1798825053645834, "learning_rate": 1.557203888516745e-07, "loss": 0.2539, "step": 1186 }, { "epoch": 4.808101265822785, "grad_norm": 0.16527916818124852, "learning_rate": 1.487313098987131e-07, "loss": 0.2449, "step": 1187 }, { "epoch": 4.812151898734177, "grad_norm": 0.16676869441480985, "learning_rate": 1.4190211033472402e-07, "loss": 0.257, "step": 1188 }, { "epoch": 4.81620253164557, "grad_norm": 0.16552872848250347, "learning_rate": 1.3523284516113955e-07, "loss": 0.264, "step": 1189 }, { "epoch": 4.820253164556962, "grad_norm": 0.17225861194941297, "learning_rate": 1.2872356809130682e-07, "loss": 0.2572, "step": 1190 }, { "epoch": 4.824303797468355, "grad_norm": 0.16123282787037901, "learning_rate": 1.2237433155004807e-07, "loss": 0.2687, "step": 1191 }, { "epoch": 4.828354430379747, "grad_norm": 0.17438378954050002, "learning_rate": 1.1618518667323886e-07, "loss": 0.2665, "step": 1192 }, { "epoch": 4.832405063291139, "grad_norm": 0.17309197273927135, "learning_rate": 1.1015618330740385e-07, "loss": 0.2663, "step": 1193 }, { "epoch": 4.836455696202532, "grad_norm": 0.172561793829551, "learning_rate": 1.042873700093061e-07, "loss": 0.2538, "step": 1194 }, { "epoch": 4.840506329113924, "grad_norm": 0.1656202293883688, "learning_rate": 9.857879404556291e-08, "loss": 0.2566, "step": 1195 }, { "epoch": 4.844556962025316, "grad_norm": 0.1651287147451129, "learning_rate": 9.303050139225722e-08, "loss": 0.2672, "step": 1196 }, { "epoch": 4.848607594936709, "grad_norm": 0.17325516530247936, "learning_rate": 8.76425367345779e-08, "loss": 0.2696, "step": 1197 }, { "epoch": 4.852658227848101, "grad_norm": 0.16879900919128904, "learning_rate": 8.241494346644897e-08, "loss": 0.2598, "step": 1198 }, { "epoch": 4.856708860759493, "grad_norm": 0.16937786663500234, "learning_rate": 7.734776369019204e-08, "loss": 0.2713, "step": 1199 }, { "epoch": 4.860759493670886, "grad_norm": 0.17038900223822956, "learning_rate": 7.244103821617332e-08, "loss": 0.2487, "step": 1200 }, { "epoch": 4.864810126582278, "grad_norm": 0.17001938336816527, "learning_rate": 6.769480656248606e-08, "loss": 0.2665, "step": 1201 }, { "epoch": 4.868860759493671, "grad_norm": 0.169575347295222, "learning_rate": 6.310910695462635e-08, "loss": 0.2697, "step": 1202 }, { "epoch": 4.872911392405063, "grad_norm": 0.17522744442479415, "learning_rate": 5.8683976325191185e-08, "loss": 0.2636, "step": 1203 }, { "epoch": 4.876962025316455, "grad_norm": 0.16272763788613978, "learning_rate": 5.4419450313571984e-08, "loss": 0.2665, "step": 1204 }, { "epoch": 4.881012658227848, "grad_norm": 0.16372313177264763, "learning_rate": 5.031556326567488e-08, "loss": 0.262, "step": 1205 }, { "epoch": 4.88506329113924, "grad_norm": 0.16981802093767961, "learning_rate": 4.637234823364312e-08, "loss": 0.2605, "step": 1206 }, { "epoch": 4.889113924050633, "grad_norm": 0.17028259410766333, "learning_rate": 4.258983697558838e-08, "loss": 0.2606, "step": 1207 }, { "epoch": 4.8931645569620255, "grad_norm": 0.19347694062346182, "learning_rate": 3.896805995533548e-08, "loss": 0.2661, "step": 1208 }, { "epoch": 4.8972151898734175, "grad_norm": 0.1685401305411941, "learning_rate": 3.550704634218028e-08, "loss": 0.2713, "step": 1209 }, { "epoch": 4.9012658227848105, "grad_norm": 0.16514842152962206, "learning_rate": 3.2206824010647676e-08, "loss": 0.2561, "step": 1210 }, { "epoch": 4.905316455696203, "grad_norm": 0.16392194317171258, "learning_rate": 2.9067419540278476e-08, "loss": 0.2504, "step": 1211 }, { "epoch": 4.909367088607595, "grad_norm": 0.16553368025932194, "learning_rate": 2.6088858215400638e-08, "loss": 0.2622, "step": 1212 }, { "epoch": 4.913417721518988, "grad_norm": 0.1695200269581906, "learning_rate": 2.3271164024940564e-08, "loss": 0.2608, "step": 1213 }, { "epoch": 4.91746835443038, "grad_norm": 0.17183791166821652, "learning_rate": 2.061435966221881e-08, "loss": 0.2666, "step": 1214 }, { "epoch": 4.921518987341772, "grad_norm": 0.15926760285649086, "learning_rate": 1.811846652477245e-08, "loss": 0.2708, "step": 1215 }, { "epoch": 4.925569620253165, "grad_norm": 0.1725060289548194, "learning_rate": 1.5783504714184106e-08, "loss": 0.2639, "step": 1216 }, { "epoch": 4.929620253164557, "grad_norm": 0.16994741566934754, "learning_rate": 1.360949303591097e-08, "loss": 0.265, "step": 1217 }, { "epoch": 4.93367088607595, "grad_norm": 0.16577399087139824, "learning_rate": 1.1596448999144916e-08, "loss": 0.2539, "step": 1218 }, { "epoch": 4.937721518987342, "grad_norm": 0.15870869719554823, "learning_rate": 9.744388816668172e-09, "loss": 0.2741, "step": 1219 }, { "epoch": 4.941772151898734, "grad_norm": 0.16502803894429138, "learning_rate": 8.05332740472009e-09, "loss": 0.2581, "step": 1220 }, { "epoch": 4.945822784810127, "grad_norm": 0.15995744754929736, "learning_rate": 6.523278382872811e-09, "loss": 0.261, "step": 1221 }, { "epoch": 4.949873417721519, "grad_norm": 0.16595890633790222, "learning_rate": 5.15425407393133e-09, "loss": 0.2563, "step": 1222 }, { "epoch": 4.953924050632912, "grad_norm": 0.16891391179416618, "learning_rate": 3.94626550383137e-09, "loss": 0.2728, "step": 1223 }, { "epoch": 4.957974683544304, "grad_norm": 0.17802236845879255, "learning_rate": 2.899322401546112e-09, "loss": 0.2523, "step": 1224 }, { "epoch": 4.962025316455696, "grad_norm": 0.1692196302560271, "learning_rate": 2.013433199010706e-09, "loss": 0.2436, "step": 1225 }, { "epoch": 4.966075949367089, "grad_norm": 0.16715890652828533, "learning_rate": 1.2886050310556563e-09, "loss": 0.2701, "step": 1226 }, { "epoch": 4.970126582278481, "grad_norm": 0.1696329611631096, "learning_rate": 7.248437353468695e-10, "loss": 0.2615, "step": 1227 }, { "epoch": 4.974177215189873, "grad_norm": 0.1669740656556205, "learning_rate": 3.221538523412449e-10, "loss": 0.2575, "step": 1228 }, { "epoch": 4.978227848101266, "grad_norm": 0.1625962091729828, "learning_rate": 8.053862524670663e-11, "loss": 0.2646, "step": 1229 }, { "epoch": 4.982278481012658, "grad_norm": 0.17423698449881814, "learning_rate": 0.0, "loss": 0.2532, "step": 1230 }, { "epoch": 4.982278481012658, "step": 1230, "total_flos": 5.906598825203728e+18, "train_loss": 0.3604403211939626, "train_runtime": 54728.4997, "train_samples_per_second": 2.887, "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": 5.906598825203728e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }