b1_science_top_1 / trainer_state.json
ryanmarten's picture
End of training
4abc457 verified
{
"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": 7.607806892343653,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.2119,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 7.393543715315901,
"learning_rate": 6.504065040650407e-07,
"loss": 1.1697,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 7.587891111647724,
"learning_rate": 9.75609756097561e-07,
"loss": 1.2016,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 7.389606447410321,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.2039,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 6.901935377218557,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.1595,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 6.553709239876541,
"learning_rate": 1.951219512195122e-06,
"loss": 1.1535,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 5.432204378526824,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.1082,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 5.18858556280531,
"learning_rate": 2.601626016260163e-06,
"loss": 1.157,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.5021502676871745,
"learning_rate": 2.926829268292683e-06,
"loss": 1.048,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.5751873347974916,
"learning_rate": 3.2520325203252037e-06,
"loss": 1.1021,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.380686380761333,
"learning_rate": 3.577235772357724e-06,
"loss": 1.0547,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 5.06251749142768,
"learning_rate": 3.902439024390244e-06,
"loss": 1.0381,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 5.365912186823246,
"learning_rate": 4.227642276422765e-06,
"loss": 1.024,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.961808309885242,
"learning_rate": 4.552845528455285e-06,
"loss": 0.986,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 4.763192943048146,
"learning_rate": 4.8780487804878055e-06,
"loss": 1.0194,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 3.2337528945129703,
"learning_rate": 5.203252032520326e-06,
"loss": 0.9848,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 3.1218456230066876,
"learning_rate": 5.528455284552846e-06,
"loss": 0.9809,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.7808334397964187,
"learning_rate": 5.853658536585366e-06,
"loss": 0.972,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 2.4052550456847213,
"learning_rate": 6.178861788617887e-06,
"loss": 0.9493,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 2.0960063428632636,
"learning_rate": 6.504065040650407e-06,
"loss": 0.9183,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.6866414453397307,
"learning_rate": 6.829268292682928e-06,
"loss": 0.9165,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.7142698944958636,
"learning_rate": 7.154471544715448e-06,
"loss": 0.9025,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.7324690669511538,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.8978,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.5636804891209075,
"learning_rate": 7.804878048780489e-06,
"loss": 0.867,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.348788169862765,
"learning_rate": 8.130081300813009e-06,
"loss": 0.8974,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.2527988436241864,
"learning_rate": 8.45528455284553e-06,
"loss": 0.8962,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.3296803332107812,
"learning_rate": 8.78048780487805e-06,
"loss": 0.8813,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 1.0618332830950192,
"learning_rate": 9.10569105691057e-06,
"loss": 0.8846,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 1.0255128456223102,
"learning_rate": 9.43089430894309e-06,
"loss": 0.8681,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 1.2194901932497608,
"learning_rate": 9.756097560975611e-06,
"loss": 0.8967,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 1.1779625300925984,
"learning_rate": 1.008130081300813e-05,
"loss": 0.8688,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.8574709670903634,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.8332,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 1.0114970594932358,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.8489,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.8189156342818956,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.856,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.9216848319604047,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.8642,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.7916486933738868,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.8773,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.7918548184801053,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.863,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.8422942835626791,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.8626,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.722002220846205,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.8305,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.8630654399393634,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.8481,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.9912374528520872,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.8443,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.7576315020438189,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.846,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.7946990350491588,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.8263,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 1.0212033622459902,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.8618,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.7180732270783728,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.8383,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.9062328804123592,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.8266,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.7728467999618894,
"learning_rate": 1.528455284552846e-05,
"loss": 0.8129,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.7682707390706401,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.8158,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.7296930879876335,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.8082,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.8690192285277243,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.8253,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.7328410805504642,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.8021,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.7437904549841122,
"learning_rate": 1.691056910569106e-05,
"loss": 0.8285,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.796799274011133,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.8263,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.8662069296351417,
"learning_rate": 1.75609756097561e-05,
"loss": 0.8174,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.9363846171371568,
"learning_rate": 1.788617886178862e-05,
"loss": 0.8253,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.9283631659372158,
"learning_rate": 1.821138211382114e-05,
"loss": 0.8202,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.8313759837555533,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.8144,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.9772384087578951,
"learning_rate": 1.886178861788618e-05,
"loss": 0.8151,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.8909185330662883,
"learning_rate": 1.91869918699187e-05,
"loss": 0.84,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.8495784413130499,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.7895,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.7848621689435769,
"learning_rate": 1.983739837398374e-05,
"loss": 0.8023,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.7893157015176571,
"learning_rate": 2.016260162601626e-05,
"loss": 0.8291,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.8438921349506423,
"learning_rate": 2.048780487804878e-05,
"loss": 0.8145,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.7431217671106334,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.8021,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.7562210986316193,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.8173,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.8089594303803846,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.8096,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.7895651354292681,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.7993,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.9730987533682705,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.8216,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.8336929454237045,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.8348,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 1.1724631278797377,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.831,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.8479225439601672,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.7987,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.8726140504048198,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.8211,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.7309980429125531,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.8137,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.7348319436490693,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.8186,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.9064099894180973,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.8169,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.826557092721544,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.8001,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.9429991419587874,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.8042,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.831651803916315,
"learning_rate": 2.536585365853659e-05,
"loss": 0.7985,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.8646236204601424,
"learning_rate": 2.569105691056911e-05,
"loss": 0.7886,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.7706310175888754,
"learning_rate": 2.601626016260163e-05,
"loss": 0.7662,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.9385485021675102,
"learning_rate": 2.634146341463415e-05,
"loss": 0.791,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.8365604593603766,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.8159,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.8611814831287041,
"learning_rate": 2.699186991869919e-05,
"loss": 0.8255,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 0.7495260901222208,
"learning_rate": 2.731707317073171e-05,
"loss": 0.7984,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.7367610537913875,
"learning_rate": 2.764227642276423e-05,
"loss": 0.7811,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.7712256807586277,
"learning_rate": 2.796747967479675e-05,
"loss": 0.7905,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.7653863776465746,
"learning_rate": 2.829268292682927e-05,
"loss": 0.8,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 1.2162930668371825,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.808,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.9594815165673303,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.8212,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.8785989690027431,
"learning_rate": 2.926829268292683e-05,
"loss": 0.8148,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.8345097119413978,
"learning_rate": 2.959349593495935e-05,
"loss": 0.8082,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.8253619074422055,
"learning_rate": 2.991869918699187e-05,
"loss": 0.7869,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.8278528745369086,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.7943,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.8320943678998864,
"learning_rate": 3.056910569105692e-05,
"loss": 0.808,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.8408486615898506,
"learning_rate": 3.089430894308943e-05,
"loss": 0.816,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.974823492146757,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.794,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.9387077659600153,
"learning_rate": 3.154471544715447e-05,
"loss": 0.808,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.7581409117795597,
"learning_rate": 3.186991869918699e-05,
"loss": 0.813,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.9093371674508962,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.8066,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.826622841384607,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.7949,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.9422116050180303,
"learning_rate": 3.284552845528456e-05,
"loss": 0.7658,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.8624545604048585,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.8074,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 1.0890087339016399,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.8134,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.0716840798014613,
"learning_rate": 3.382113821138212e-05,
"loss": 0.7994,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.9309737140550247,
"learning_rate": 3.414634146341463e-05,
"loss": 0.8057,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.9854229468697615,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.7932,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.99951158992149,
"learning_rate": 3.479674796747968e-05,
"loss": 0.7933,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.7666478753053647,
"learning_rate": 3.51219512195122e-05,
"loss": 0.7949,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.8738437097697433,
"learning_rate": 3.544715447154472e-05,
"loss": 0.8273,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.0399250721030842,
"learning_rate": 3.577235772357724e-05,
"loss": 0.8317,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 0.9948943611983216,
"learning_rate": 3.609756097560976e-05,
"loss": 0.8149,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.921095745746958,
"learning_rate": 3.642276422764228e-05,
"loss": 0.7895,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.054168149082262,
"learning_rate": 3.67479674796748e-05,
"loss": 0.8116,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.8077667863467105,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.7715,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.822370401539034,
"learning_rate": 3.739837398373984e-05,
"loss": 0.7922,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.990919311818821,
"learning_rate": 3.772357723577236e-05,
"loss": 0.793,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.9102881202961368,
"learning_rate": 3.804878048780488e-05,
"loss": 0.7863,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.6688205449236377,
"learning_rate": 3.83739837398374e-05,
"loss": 0.7849,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.8584797942738227,
"learning_rate": 3.869918699186992e-05,
"loss": 0.7913,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.0090082248742935,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.7834,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 1.0264338358925533,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.8353,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 0.897054743915044,
"learning_rate": 3.967479674796748e-05,
"loss": 0.7684,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 0.8287228061874787,
"learning_rate": 4e-05,
"loss": 0.8102,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 1.072352432193976,
"learning_rate": 3.999991946137476e-05,
"loss": 0.8073,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.1270082739549225,
"learning_rate": 3.999967784614766e-05,
"loss": 0.8043,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.7626540968654951,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.8169,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.758574364014147,
"learning_rate": 3.999871139496895e-05,
"loss": 0.7967,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.8855783523258045,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.8067,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.9050681189602536,
"learning_rate": 3.999710067759846e-05,
"loss": 0.8074,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.8367433012513539,
"learning_rate": 3.999605373449617e-05,
"loss": 0.813,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.8685169187038129,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.8051,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.8068020556703477,
"learning_rate": 3.999347672161713e-05,
"loss": 0.7775,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.7650541096604251,
"learning_rate": 3.999194667259528e-05,
"loss": 0.7878,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.9400276810088601,
"learning_rate": 3.999025561118334e-05,
"loss": 0.7794,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.8329942350955717,
"learning_rate": 3.998840355100086e-05,
"loss": 0.813,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.8513128640302231,
"learning_rate": 3.998639050696409e-05,
"loss": 0.8015,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.8327135958991501,
"learning_rate": 3.998421649528582e-05,
"loss": 0.7934,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.7927845032898265,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.8024,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.7896201108199237,
"learning_rate": 3.997938564033779e-05,
"loss": 0.8015,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 1.0360900458262565,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.7897,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.8101750975605219,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.7947,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.7618665287216916,
"learning_rate": 3.997093258045973e-05,
"loss": 0.7907,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.8990146350170286,
"learning_rate": 3.996779317598936e-05,
"loss": 0.779,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.8208388141174423,
"learning_rate": 3.996449295365782e-05,
"loss": 0.782,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.7522646799284676,
"learning_rate": 3.996103194004467e-05,
"loss": 0.8008,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.960518537504372,
"learning_rate": 3.995741016302441e-05,
"loss": 0.7918,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.6279404752189512,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.7861,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.9016414782319416,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.7744,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.7621676913631221,
"learning_rate": 3.994558054968643e-05,
"loss": 0.8,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.9283398561918387,
"learning_rate": 3.994131602367481e-05,
"loss": 0.8024,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.781589767960697,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.7886,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.9400147515940361,
"learning_rate": 3.993230519343752e-05,
"loss": 0.7995,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.7717543864351121,
"learning_rate": 3.992755896178383e-05,
"loss": 0.7821,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.6606895873449645,
"learning_rate": 3.992265223630981e-05,
"loss": 0.81,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.7641823467438348,
"learning_rate": 3.991758505653355e-05,
"loss": 0.8155,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.7355840749358501,
"learning_rate": 3.991235746326543e-05,
"loss": 0.7971,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.6823531565541342,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.7937,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.7904818895211045,
"learning_rate": 3.990142120595444e-05,
"loss": 0.7963,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.6321554768426836,
"learning_rate": 3.98957126299907e-05,
"loss": 0.7793,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.6334771059732883,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.7801,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.652003603034157,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.7669,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.739960831449817,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.7896,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.6613945549236503,
"learning_rate": 3.98712764319087e-05,
"loss": 0.7805,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.6612509939735441,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.766,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.6266662240879284,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.7733,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.6600159593477919,
"learning_rate": 3.985126869010129e-05,
"loss": 0.793,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.6085488604072714,
"learning_rate": 3.984427961114833e-05,
"loss": 0.7918,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.6938794028628833,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.8041,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.624898563659077,
"learning_rate": 3.982982204151901e-05,
"loss": 0.7833,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.7156325646715105,
"learning_rate": 3.982235366728193e-05,
"loss": 0.7947,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.7254621651650967,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.8046,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.7670123686202683,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.7934,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.986381372956543,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.8031,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.7410862707427704,
"learning_rate": 3.979088432599627e-05,
"loss": 0.7968,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.7097946368044431,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.8011,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.6300621214323768,
"learning_rate": 3.977419304032111e-05,
"loss": 0.7802,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.8258169132546216,
"learning_rate": 3.976560847560697e-05,
"loss": 0.7876,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.6441953776199969,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.7856,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.9155285525686293,
"learning_rate": 3.974796184811956e-05,
"loss": 0.7758,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.7370643541854416,
"learning_rate": 3.973889992746979e-05,
"loss": 0.7635,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.7298728400613076,
"learning_rate": 3.972967903243361e-05,
"loss": 0.7715,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.7846936221045062,
"learning_rate": 3.972029923727486e-05,
"loss": 0.7815,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.6388143502266158,
"learning_rate": 3.971076061753709e-05,
"loss": 0.7833,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.772964740359074,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.7667,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.6725064730093347,
"learning_rate": 3.969120721289402e-05,
"loss": 0.8042,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.6536397034870945,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.784,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.6754775456540817,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.7861,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.7474496232561747,
"learning_rate": 3.966068788369422e-05,
"loss": 0.8096,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.8092006108863046,
"learning_rate": 3.965019797448622e-05,
"loss": 0.785,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.6727092249044911,
"learning_rate": 3.963954980528515e-05,
"loss": 0.7798,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.6784459700768211,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.7739,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.7374008708932581,
"learning_rate": 3.961777903121329e-05,
"loss": 0.7832,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.6610823044280854,
"learning_rate": 3.960665660168131e-05,
"loss": 0.7954,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.7803097605833994,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.8031,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.7160498799218283,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.7881,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.7845430299968533,
"learning_rate": 3.957234222185657e-05,
"loss": 0.7947,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.6066336574477401,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.78,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.803647154454646,
"learning_rate": 3.954867765033605e-05,
"loss": 0.7896,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.5909052221563678,
"learning_rate": 3.953660915306728e-05,
"loss": 0.7778,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.6887943564327977,
"learning_rate": 3.952438331063419e-05,
"loss": 0.8025,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.6842023594578742,
"learning_rate": 3.951200022150205e-05,
"loss": 0.8072,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.654965102375549,
"learning_rate": 3.949945998540253e-05,
"loss": 0.7829,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.6866649421206683,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.7707,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.7971618168002252,
"learning_rate": 3.947390847755559e-05,
"loss": 0.7629,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.8485756947244271,
"learning_rate": 3.946089741159648e-05,
"loss": 0.79,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.7617834222040127,
"learning_rate": 3.944772961024501e-05,
"loss": 0.7947,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.7137818310572526,
"learning_rate": 3.943440517955285e-05,
"loss": 0.7665,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.7706666735329565,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.7476,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.6646262318901943,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.7771,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.8147032424032501,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.7827,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.8010693189068027,
"learning_rate": 3.937954332854371e-05,
"loss": 0.784,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.6595145961054906,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.7359,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.8145877201641332,
"learning_rate": 3.935117547697387e-05,
"loss": 0.7774,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.6351140873809643,
"learning_rate": 3.933675771619675e-05,
"loss": 0.7581,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.7285734077076174,
"learning_rate": 3.932218421983131e-05,
"loss": 0.7919,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.8736518889798665,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.7864,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.6453116666707841,
"learning_rate": 3.929257049108054e-05,
"loss": 0.7849,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.9330422696536262,
"learning_rate": 3.927753049720011e-05,
"loss": 0.7905,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.6380711950900357,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.7847,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.6962700847679576,
"learning_rate": 3.92469848560784e-05,
"loss": 0.8198,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.6178586617798486,
"learning_rate": 3.923147945484751e-05,
"loss": 0.7683,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.6198607797517478,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.7786,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.59399145948892,
"learning_rate": 3.920000411543654e-05,
"loss": 0.7786,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.6402736551472451,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.7748,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.7161971208598362,
"learning_rate": 3.916791024049648e-05,
"loss": 0.7873,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.5812942347545466,
"learning_rate": 3.91516316745245e-05,
"loss": 0.7906,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.7145295346306079,
"learning_rate": 3.913519886394389e-05,
"loss": 0.7766,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.6757050535705096,
"learning_rate": 3.911861194110225e-05,
"loss": 0.7779,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.6479486242669653,
"learning_rate": 3.910187103958837e-05,
"loss": 0.7557,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.7554995617709788,
"learning_rate": 3.908497629423117e-05,
"loss": 0.7967,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.6564436050874437,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.7667,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.610700362445129,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.7742,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.6057236169590109,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.7635,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.657656885725812,
"learning_rate": 3.901586161429081e-05,
"loss": 0.8011,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.6053463267382987,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.7694,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.5683621384193809,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.7665,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.6039033220268641,
"learning_rate": 3.896241703466667e-05,
"loss": 0.7886,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.587780866086855,
"learning_rate": 3.894429654085585e-05,
"loss": 0.7609,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.8193094752257639,
"learning_rate": 3.892602347228505e-05,
"loss": 0.7899,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.7072631063689974,
"learning_rate": 3.890759797612307e-05,
"loss": 0.798,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.6839201171834809,
"learning_rate": 3.888902020076632e-05,
"loss": 0.7487,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.7732593277287587,
"learning_rate": 3.887029029583764e-05,
"loss": 0.7513,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.5907507322292838,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.7748,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.5682099860914966,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.7827,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.5942542634747586,
"learning_rate": 3.881318931821972e-05,
"loss": 0.7585,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.6547716899012801,
"learning_rate": 3.879385241571817e-05,
"loss": 0.7408,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 0.581156411601929,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.7338,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.8732065557669838,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.6138,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.5774703976681932,
"learning_rate": 3.873493415863256e-05,
"loss": 0.6002,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.8016609635376105,
"learning_rate": 3.871499275032111e-05,
"loss": 0.5956,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.6348202165509207,
"learning_rate": 3.869490061403091e-05,
"loss": 0.5924,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 0.7397198748572259,
"learning_rate": 3.867465791158124e-05,
"loss": 0.6043,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.7438903125186032,
"learning_rate": 3.865426480600407e-05,
"loss": 0.5804,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 0.6517426575355707,
"learning_rate": 3.863372146154264e-05,
"loss": 0.6044,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.7355656116261422,
"learning_rate": 3.861302804365024e-05,
"loss": 0.6226,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.7050590661652748,
"learning_rate": 3.85921847189888e-05,
"loss": 0.5917,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.726115140083047,
"learning_rate": 3.85711916554276e-05,
"loss": 0.6022,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.697942757679148,
"learning_rate": 3.85500490220419e-05,
"loss": 0.5858,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.7014091480357524,
"learning_rate": 3.852875698911154e-05,
"loss": 0.6105,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.6757257985792863,
"learning_rate": 3.850731572811963e-05,
"loss": 0.5968,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.6732737427652161,
"learning_rate": 3.848572541175116e-05,
"loss": 0.6054,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.6697570849937825,
"learning_rate": 3.846398621389154e-05,
"loss": 0.6032,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.6250904869014002,
"learning_rate": 3.84420983096253e-05,
"loss": 0.5693,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.6501995874346592,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.6059,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.6997982500731121,
"learning_rate": 3.839787708819787e-05,
"loss": 0.5979,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.6588315715003042,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.5962,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.6538291768594633,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.6017,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.7725449216986721,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.5979,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.7871015710325765,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.5731,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.6942143850446544,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.5844,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.6811479513269205,
"learning_rate": 3.826166304956594e-05,
"loss": 0.5864,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.7078691059176183,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.5929,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.7871680040623192,
"learning_rate": 3.821507978767479e-05,
"loss": 0.5865,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.6520418545547645,
"learning_rate": 3.819156801001693e-05,
"loss": 0.593,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.7976141507010054,
"learning_rate": 3.816790971997121e-05,
"loss": 0.5708,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.7225248324961825,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.5803,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.6258312256752028,
"learning_rate": 3.81201543660571e-05,
"loss": 0.5655,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.7285970572344127,
"learning_rate": 3.809605768680377e-05,
"loss": 0.5848,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.7022041310420769,
"learning_rate": 3.807181526438958e-05,
"loss": 0.5942,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.7701689958270314,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.5949,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.8364420443448863,
"learning_rate": 3.802289397223145e-05,
"loss": 0.5941,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.8614208850279532,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.6116,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.9694211980915449,
"learning_rate": 3.797339206560096e-05,
"loss": 0.6062,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.907302588754404,
"learning_rate": 3.794842387948027e-05,
"loss": 0.5767,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.7851151457110381,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.5839,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.7880071764185652,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.5963,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.7020333688361554,
"learning_rate": 3.787265280646825e-05,
"loss": 0.5959,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.7112725990773352,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.5973,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.7927761785383373,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.6089,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.73178322775999,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.5921,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.6021175280856262,
"learning_rate": 3.776961046830306e-05,
"loss": 0.588,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.7733664793111539,
"learning_rate": 3.774349157719215e-05,
"loss": 0.5973,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.5828190116047194,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.6191,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.6703343760901636,
"learning_rate": 3.769082529555359e-05,
"loss": 0.587,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.623527635454609,
"learning_rate": 3.766427832919294e-05,
"loss": 0.6024,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.6027957757651938,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.6009,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.6056397600516513,
"learning_rate": 3.761075781441526e-05,
"loss": 0.6075,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 0.7126314400849378,
"learning_rate": 3.75837846970451e-05,
"loss": 0.5928,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.5694261769746022,
"learning_rate": 3.755666996229032e-05,
"loss": 0.5685,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.7171688004721847,
"learning_rate": 3.752941382852927e-05,
"loss": 0.6006,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.5749550840667167,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.613,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.609479660279044,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.6167,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.5845260044338126,
"learning_rate": 3.744679923406351e-05,
"loss": 0.6262,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.5689841248845234,
"learning_rate": 3.741897971081043e-05,
"loss": 0.5999,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.6428705182766801,
"learning_rate": 3.739101989748946e-05,
"loss": 0.5859,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.5695064613036215,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.5843,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.594552071437002,
"learning_rate": 3.73346803025098e-05,
"loss": 0.5921,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 0.5677430523790219,
"learning_rate": 3.730630097460247e-05,
"loss": 0.5758,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.6055829831651833,
"learning_rate": 3.727778226412628e-05,
"loss": 0.6051,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 0.5783659968042157,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.5964,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.595589515647746,
"learning_rate": 3.722032761533114e-05,
"loss": 0.6047,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.5933336549788346,
"learning_rate": 3.719139213974403e-05,
"loss": 0.6056,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 0.5618398489134566,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.6034,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.609848527566902,
"learning_rate": 3.713310605140055e-05,
"loss": 0.5888,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.5964679475561165,
"learning_rate": 3.710375590807233e-05,
"loss": 0.5944,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.5638806901286798,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.5839,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.5730949129599185,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.5802,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.6129081356880398,
"learning_rate": 3.701487992136854e-05,
"loss": 0.61,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 0.5601064200718908,
"learning_rate": 3.69849802022222e-05,
"loss": 0.6148,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.6373439994259741,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.6199,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.6137363790033474,
"learning_rate": 3.692477062175289e-05,
"loss": 0.6053,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.615858376887754,
"learning_rate": 3.689446124534958e-05,
"loss": 0.5886,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.6455499429383174,
"learning_rate": 3.686401580327799e-05,
"loss": 0.609,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.6601873711864136,
"learning_rate": 3.683343454074149e-05,
"loss": 0.6099,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.6000204139441397,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.6078,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.6394403547379943,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.607,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.6830999698076322,
"learning_rate": 3.674087829877297e-05,
"loss": 0.5884,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.5797417199434945,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.597,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.7044823771416117,
"learning_rate": 3.667849957966501e-05,
"loss": 0.587,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.6509131176040178,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.6014,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.6707920762521203,
"learning_rate": 3.661558355626795e-05,
"loss": 0.603,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.6384027614901036,
"learning_rate": 3.658392468818163e-05,
"loss": 0.6072,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.5213511011083612,
"learning_rate": 3.655213225544574e-05,
"loss": 0.5883,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.6866646293619093,
"learning_rate": 3.652020651411218e-05,
"loss": 0.6015,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.5864474546453821,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.5847,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.6064684703392841,
"learning_rate": 3.645595613522574e-05,
"loss": 0.6034,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.6498110486010376,
"learning_rate": 3.642363201513657e-05,
"loss": 0.5951,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.5823117085443584,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.5894,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.6704881687182457,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.5758,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.6636901133433927,
"learning_rate": 3.632586705948318e-05,
"loss": 0.5702,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.6199089397899836,
"learning_rate": 3.629301541734311e-05,
"loss": 0.6113,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.7132000868090463,
"learning_rate": 3.626003255349676e-05,
"loss": 0.6166,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.5721926601615668,
"learning_rate": 3.622691873358357e-05,
"loss": 0.5838,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.7421125397109054,
"learning_rate": 3.61936742242977e-05,
"loss": 0.5986,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.6074658758003356,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.6087,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.6665224849022102,
"learning_rate": 3.612679420964516e-05,
"loss": 0.6164,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.5660552070395921,
"learning_rate": 3.609315924292092e-05,
"loss": 0.5909,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.6196862223179042,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.606,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.6144771878752318,
"learning_rate": 3.602550074513133e-05,
"loss": 0.6021,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.590864743627934,
"learning_rate": 3.599147775897822e-05,
"loss": 0.6043,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.6323663822080761,
"learning_rate": 3.595732597966167e-05,
"loss": 0.6176,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.5195163142270381,
"learning_rate": 3.592304568223542e-05,
"loss": 0.5863,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.5767979503471956,
"learning_rate": 3.588863714278826e-05,
"loss": 0.5917,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.5558931535535298,
"learning_rate": 3.585410063844186e-05,
"loss": 0.603,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.5782598032497832,
"learning_rate": 3.581943644734846e-05,
"loss": 0.5921,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.5598695195195617,
"learning_rate": 3.578464484868869e-05,
"loss": 0.5967,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.5304126686403173,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.5943,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.5787660186859815,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.597,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.5838854541336683,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.6029,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.5503297574726093,
"learning_rate": 3.564420999786543e-05,
"loss": 0.6105,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.5973268027529769,
"learning_rate": 3.560878558491842e-05,
"loss": 0.6017,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.509645033734509,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.5849,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.5912689591958639,
"learning_rate": 3.553755991230039e-05,
"loss": 0.6549,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.578785990671761,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.5953,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.5202436690951565,
"learning_rate": 3.546583369120419e-05,
"loss": 0.5775,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.5631963000792698,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.5787,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.5810435693184101,
"learning_rate": 3.539360923231766e-05,
"loss": 0.6089,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.5928218727332761,
"learning_rate": 3.535731089018394e-05,
"loss": 0.6221,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.5665055947608707,
"learning_rate": 3.532088886237956e-05,
"loss": 0.6062,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.6407140597155299,
"learning_rate": 3.528434344224253e-05,
"loss": 0.6201,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.6578658862440514,
"learning_rate": 3.524767492410464e-05,
"loss": 0.5986,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.6169335460660622,
"learning_rate": 3.521088360328908e-05,
"loss": 0.5992,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.7663578839328806,
"learning_rate": 3.517396977610811e-05,
"loss": 0.5973,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.6206567785222147,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.6065,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.6035589150886722,
"learning_rate": 3.509977579282971e-05,
"loss": 0.5948,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.607577811152864,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.6206,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.6569519494833488,
"learning_rate": 3.502509536445719e-05,
"loss": 0.6246,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.6655281113012875,
"learning_rate": 3.498757348458147e-05,
"loss": 0.6095,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.6033028441689147,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.617,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.5958430901818426,
"learning_rate": 3.491216790442899e-05,
"loss": 0.6219,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.5949658285766714,
"learning_rate": 3.487428481145839e-05,
"loss": 0.5916,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.6169269062140614,
"learning_rate": 3.483628192304278e-05,
"loss": 0.5868,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.6932823779136169,
"learning_rate": 3.479815954525219e-05,
"loss": 0.6095,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.545593135402959,
"learning_rate": 3.475991798511899e-05,
"loss": 0.6163,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.7937430562599953,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.6117,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.5843516374763659,
"learning_rate": 3.468307855075128e-05,
"loss": 0.6046,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.7453029580726032,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.5926,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.589238316369403,
"learning_rate": 3.460576609535163e-05,
"loss": 0.6163,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.7347538384810385,
"learning_rate": 3.456693326250006e-05,
"loss": 0.592,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.6253054160170282,
"learning_rate": 3.452798310957058e-05,
"loss": 0.6222,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.5757829351701256,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.5959,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.5738061418955158,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.6239,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.6268443722430391,
"learning_rate": 3.441043187201574e-05,
"loss": 0.6028,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.5840238562822351,
"learning_rate": 3.437101558517728e-05,
"loss": 0.6178,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.6468881376538268,
"learning_rate": 3.433148355615496e-05,
"loss": 0.6107,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.6147281657696353,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.6354,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.5905289409730526,
"learning_rate": 3.425207354603043e-05,
"loss": 0.6119,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.6274703435350056,
"learning_rate": 3.421219620448553e-05,
"loss": 0.5966,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.5698921773177216,
"learning_rate": 3.417220439986623e-05,
"loss": 0.6003,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.6043144463580495,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.5869,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.6057641155696274,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.6245,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.5967549419828996,
"learning_rate": 3.405154543304065e-05,
"loss": 0.5828,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.6101111670762375,
"learning_rate": 3.401109900618843e-05,
"loss": 0.6029,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.568169035566485,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.6312,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.649153847001911,
"learning_rate": 3.392986794874714e-05,
"loss": 0.6054,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.5090140119049416,
"learning_rate": 3.388908397238184e-05,
"loss": 0.5923,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.6345680032501896,
"learning_rate": 3.384818813524362e-05,
"loss": 0.5836,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.6474102576084231,
"learning_rate": 3.380718076670195e-05,
"loss": 0.5978,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.5918165014043868,
"learning_rate": 3.376606219702454e-05,
"loss": 0.6253,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.6700691739590833,
"learning_rate": 3.372483275737468e-05,
"loss": 0.6017,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.5776426981285693,
"learning_rate": 3.368349277980861e-05,
"loss": 0.6053,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.5968460352269188,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.6012,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.6458230224705035,
"learning_rate": 3.360048254360144e-05,
"loss": 0.5864,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.5707303038692311,
"learning_rate": 3.355881295351336e-05,
"loss": 0.5786,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.6064969483023611,
"learning_rate": 3.351703416260975e-05,
"loss": 0.6057,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.6048745921835993,
"learning_rate": 3.347514650737126e-05,
"loss": 0.6239,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.557182661157355,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.5917,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.664161619706652,
"learning_rate": 3.339104595419334e-05,
"loss": 0.6132,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.5404261803849054,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.5996,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.6382519841047599,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.5982,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.5058115383558751,
"learning_rate": 3.326408710419996e-05,
"loss": 0.6028,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.6480562719670722,
"learning_rate": 3.322155337795454e-05,
"loss": 0.6085,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.5193408522045253,
"learning_rate": 3.317891316713587e-05,
"loss": 0.5904,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.6827434260430911,
"learning_rate": 3.313616681516231e-05,
"loss": 0.6052,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.7392649383363337,
"learning_rate": 3.309331466630713e-05,
"loss": 0.6119,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.6446805939996907,
"learning_rate": 3.305035706569563e-05,
"loss": 0.6145,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.7433554156777252,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.6067,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.5830874018290577,
"learning_rate": 3.296412689394864e-05,
"loss": 0.5985,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.6245719703526939,
"learning_rate": 3.292085501729909e-05,
"loss": 0.5914,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.5445311607567979,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.5871,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.5724599793638191,
"learning_rate": 3.283399942497381e-05,
"loss": 0.6013,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.5902559747859615,
"learning_rate": 3.279041640882108e-05,
"loss": 0.6224,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.5935695780785047,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.5878,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.6164113506454896,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.6049,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.5659283589309865,
"learning_rate": 3.265905069502244e-05,
"loss": 0.5876,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.5455705321629816,
"learning_rate": 3.261505774420016e-05,
"loss": 0.586,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.5267654752834655,
"learning_rate": 3.257096319343707e-05,
"loss": 0.5957,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.609862547457684,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.6249,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.5829314360706592,
"learning_rate": 3.248247071342966e-05,
"loss": 0.6028,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.5753968597382567,
"learning_rate": 3.243807349689161e-05,
"loss": 0.6096,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.6032875381259633,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.6085,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.5685861347664593,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.6139,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.6637714743157007,
"learning_rate": 3.230428223438075e-05,
"loss": 0.5955,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.5339397854457546,
"learning_rate": 3.225948647317459e-05,
"loss": 0.5852,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.6265293279817646,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.5989,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.6236367806553621,
"learning_rate": 3.216959910368034e-05,
"loss": 0.6249,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.5524115907530536,
"learning_rate": 3.212450821933277e-05,
"loss": 0.606,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.5756534369949164,
"learning_rate": 3.207931968586281e-05,
"loss": 0.6076,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.5567528041833827,
"learning_rate": 3.203403386721272e-05,
"loss": 0.5953,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.5592501563916369,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.6042,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.5577523678996232,
"learning_rate": 3.194317183405573e-05,
"loss": 0.5754,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.6017297261920005,
"learning_rate": 3.189759635133914e-05,
"loss": 0.6127,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.569750620788789,
"learning_rate": 3.185192504701718e-05,
"loss": 0.594,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.6098661247438405,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.6147,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.6148239810242565,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.6034,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.602183245016038,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.5803,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.5629973289232163,
"learning_rate": 3.166828898179731e-05,
"loss": 0.6061,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.5839011686378607,
"learning_rate": 3.162214410223536e-05,
"loss": 0.607,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.5836343492251241,
"learning_rate": 3.157590561952257e-05,
"loss": 0.5975,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.5913351121817788,
"learning_rate": 3.152957390605732e-05,
"loss": 0.6257,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.554560924830124,
"learning_rate": 3.148314933498886e-05,
"loss": 0.6192,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.6683614760088459,
"learning_rate": 3.143663228021431e-05,
"loss": 0.6134,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.5584687882256828,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.6191,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.6889606760288862,
"learning_rate": 3.134332221885661e-05,
"loss": 0.6198,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.5509971994942906,
"learning_rate": 3.129652996377987e-05,
"loss": 0.6283,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.5830151312377695,
"learning_rate": 3.12496467280038e-05,
"loss": 0.5877,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.5917958832755541,
"learning_rate": 3.120267288911952e-05,
"loss": 0.6066,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.5725074630418118,
"learning_rate": 3.11556088254479e-05,
"loss": 0.611,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.6406195250751695,
"learning_rate": 3.11084549160364e-05,
"loss": 0.5838,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.594351054044271,
"learning_rate": 3.106121154065615e-05,
"loss": 0.6056,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.5927561118688978,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.5965,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.6651532986627803,
"learning_rate": 3.096645791467348e-05,
"loss": 0.5916,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.6448098754027665,
"learning_rate": 3.091894842720373e-05,
"loss": 0.6058,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.6602863420368841,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.6357,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.5946482552297947,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.6072,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.5760585457460217,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.6037,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.637895925465041,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.6012,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.5953799141380951,
"learning_rate": 3.06800895716334e-05,
"loss": 0.6194,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.5778396459522998,
"learning_rate": 3.063205821011029e-05,
"loss": 0.5984,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.6237654672677111,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.6099,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.628064337061949,
"learning_rate": 3.053573898718618e-05,
"loss": 0.6177,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.670412819902393,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.6187,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.584011490209677,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.5972,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.6143770472121562,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.5845,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.537344952257066,
"learning_rate": 3.034208541656255e-05,
"loss": 0.5943,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.7066145923751287,
"learning_rate": 3.029346281309218e-05,
"loss": 0.6265,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.5190272968858315,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.5907,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.6494226313410707,
"learning_rate": 3.019596929201569e-05,
"loss": 0.6115,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.5334868587206975,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.6143,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.5993096259860139,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.6032,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.5276651564933268,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.591,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.5480856745913394,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6045,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.5889910894626287,
"learning_rate": 2.995080534235264e-05,
"loss": 0.5969,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.5907113684687959,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.5692,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.8963160234827804,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.3897,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 1.1864436209189444,
"learning_rate": 2.980274210295326e-05,
"loss": 0.4029,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.8870054032835404,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.385,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.7273957276562338,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.4088,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.7168504787405506,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3927,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.5929369315239049,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.3911,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.6662649325022685,
"learning_rate": 2.955439637024526e-05,
"loss": 0.3826,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.6624091676431377,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.382,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.6920407226375248,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.3945,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.6230479062109411,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3835,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.6411188543079344,
"learning_rate": 2.93543322968538e-05,
"loss": 0.3704,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.6429664811617349,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.3707,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.6235251678976552,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.3946,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.6139078694694637,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3906,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.6262568760393064,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.387,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.6074405773013294,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.3898,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.5475358640344165,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.3687,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.5806241490062988,
"learning_rate": 2.900133526131588e-05,
"loss": 0.3869,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.5817043261181017,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.3754,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.6356996492084521,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3733,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.5696426565740269,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3795,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.607165134499063,
"learning_rate": 2.879801882631847e-05,
"loss": 0.4003,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.5206623838753841,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.3891,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.5428293320350098,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.3859,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.5525870147816259,
"learning_rate": 2.864478605989494e-05,
"loss": 0.3984,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.588495439422463,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.4033,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.5200452562347823,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.3822,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.5828180444273979,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3911,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.5541778095223793,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.4016,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.5226436371072429,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.4083,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.5438870393057756,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.3955,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.5254106129795638,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.3939,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.5654781029836254,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.394,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.4914208410102574,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.3745,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.6020512986924352,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3687,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.5414491342266459,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.3794,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.5502120030625025,
"learning_rate": 2.802570062143278e-05,
"loss": 0.3905,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.5535387258341357,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.3817,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.5502130874334074,
"learning_rate": 2.792159532078314e-05,
"loss": 0.3966,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.49424650021261046,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.3723,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.5720434982838802,
"learning_rate": 2.781723482288857e-05,
"loss": 0.3954,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.4702824262942927,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.387,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.5770883090150516,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3911,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.47605345697449547,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.3851,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.5167038461991028,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3632,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.5008850085464035,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.371,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.5007058711940726,
"learning_rate": 2.7502655806328e-05,
"loss": 0.4056,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.48583468378581907,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.4087,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.5158963746028612,
"learning_rate": 2.739730822017673e-05,
"loss": 0.3762,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.5026809860390608,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.3802,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.5379281716560401,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.3861,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.539583736896801,
"learning_rate": 2.723884107774236e-05,
"loss": 0.3965,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.5640914787139998,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3989,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.5200072242520184,
"learning_rate": 2.71329041039194e-05,
"loss": 0.3903,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.503523270099919,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3881,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.5555009584591764,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.3735,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.5338877652433439,
"learning_rate": 2.697356885686189e-05,
"loss": 0.3965,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.5907311880056733,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3856,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.527372946564449,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.3812,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.5382968370719524,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.4192,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.5407254540952011,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.4078,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.5688005486396707,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.382,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.575377697029208,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3783,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.5458374789708114,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.3888,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.5208608839774398,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.371,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.5112652307611913,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.3933,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.5144047836057486,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.3956,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.5314257289060231,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3943,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.5071183482669611,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.3667,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.5773797085353732,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.3936,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.5524706355389346,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.3893,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.4731505239544197,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.3849,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.5168959525732323,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3959,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.5390771027262704,
"learning_rate": 2.606146264569603e-05,
"loss": 0.4063,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.5172877859231843,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.3776,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.569154100114174,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.3765,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.5185013453308233,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.3843,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.5349164094498392,
"learning_rate": 2.584472010606015e-05,
"loss": 0.4032,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.525826416510778,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.3752,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.5251902669608044,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3898,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.5414122678106538,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3856,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.5083698569237726,
"learning_rate": 2.56272244128509e-05,
"loss": 0.4207,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.5826810287016296,
"learning_rate": 2.55727360915895e-05,
"loss": 0.3948,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.5169671513859312,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.3968,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.5139569970541039,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3977,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.494780237262699,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3846,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.5281998769416739,
"learning_rate": 2.535433837985559e-05,
"loss": 0.4015,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.49009842697368566,
"learning_rate": 2.529963004393324e-05,
"loss": 0.4019,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.5173169329423096,
"learning_rate": 2.524487902551908e-05,
"loss": 0.374,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.4893410467411231,
"learning_rate": 2.519008576557029e-05,
"loss": 0.3608,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.49478868820036737,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.4058,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.5235882012611537,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.3796,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.49353565755780077,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.401,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.5074921808683411,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.3761,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.5015989466482824,
"learning_rate": 2.491550129992345e-05,
"loss": 0.3892,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.5177804748217657,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3721,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.49411839469498015,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.3957,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.5536047631083485,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.393,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.5030284154980249,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.4003,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.5164278468659577,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.4005,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.5007651530903415,
"learning_rate": 2.458469841202444e-05,
"loss": 0.3818,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.5213268277356008,
"learning_rate": 2.452943276859226e-05,
"loss": 0.3754,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.5067244399562729,
"learning_rate": 2.447413064573125e-05,
"loss": 0.3707,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.5645241539086214,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.3995,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.5141417925971842,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.4234,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.5235105011307918,
"learning_rate": 2.430800985597963e-05,
"loss": 0.4032,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.5112689794555305,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.3902,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.5035710466853918,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3937,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.4981967698878523,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.3927,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.5064107542630472,
"learning_rate": 2.408603181104957e-05,
"loss": 0.4058,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.4927968279835035,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.3987,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.4801833409292795,
"learning_rate": 2.397484355008521e-05,
"loss": 0.3917,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.4900292199189647,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.3917,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.49993248054101447,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.4092,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.5108172937974874,
"learning_rate": 2.380782218329337e-05,
"loss": 0.3904,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.4551304918758009,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.4044,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.486382948774806,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.3911,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.5044645784201129,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.394,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.5057384394763091,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.3998,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.5089957664683604,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3791,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.49262405239198626,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.377,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.5238168849074862,
"learning_rate": 2.341705325992901e-05,
"loss": 0.3856,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.5028842119200384,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3769,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.4735957893219881,
"learning_rate": 2.33051505621937e-05,
"loss": 0.4051,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.49349694915524556,
"learning_rate": 2.324915905911693e-05,
"loss": 0.4012,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.45607252359901074,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3823,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.534403865824684,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.3722,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.5213311678936735,
"learning_rate": 2.308102934504593e-05,
"loss": 0.425,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.4520540248671105,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.3713,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.5255619752851586,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.3722,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.46571868937002026,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3942,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.5029478699626156,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.3701,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.47031672795107055,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3735,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.4739198840275075,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.4061,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.531708301120094,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.3901,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.5044611391350264,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3982,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.47092620032536536,
"learning_rate": 2.257534907110328e-05,
"loss": 0.3844,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.5029316353244448,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.3725,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.48708443841946764,
"learning_rate": 2.246273589557294e-05,
"loss": 0.3957,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.5072104223247064,
"learning_rate": 2.240639932921966e-05,
"loss": 0.3919,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.4740826770764929,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.4078,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.5355330515135893,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.4013,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.48743504483822914,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.401,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.49750363591741265,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.3781,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.523673090637859,
"learning_rate": 2.212443486523819e-05,
"loss": 0.378,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.49008991239767036,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.3861,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.5058568030806204,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.397,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.5179997967160612,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.3921,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.5173402338963075,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3859,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.4848468783564759,
"learning_rate": 2.184204266048803e-05,
"loss": 0.3825,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.4931264485195604,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.4002,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.5239659523188732,
"learning_rate": 2.17289787446841e-05,
"loss": 0.3837,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.49553560569631744,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.4013,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.5497641414629512,
"learning_rate": 2.161585912916385e-05,
"loss": 0.3935,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.4722438036606276,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.3831,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.5202039275645676,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3937,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.5126659481348737,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.3943,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.4797623776694997,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.4018,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.5134588525716584,
"learning_rate": 2.133284032321232e-05,
"loss": 0.3897,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.5325631809636909,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.3835,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.47580475980110565,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3976,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.5097187254603867,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.3721,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.5406968200156804,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.3916,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.4607240071330362,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3655,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.507937148433873,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3808,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.5415521983209989,
"learning_rate": 2.093617592756158e-05,
"loss": 0.4028,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.4685174930896097,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.3866,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.49013116665664325,
"learning_rate": 2.082276853708978e-05,
"loss": 0.3903,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.5222785860065743,
"learning_rate": 2.076605467380071e-05,
"loss": 0.3731,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.4680784164757683,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.4034,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.5347346292027513,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3817,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.5146153562564106,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3908,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.5126305405573354,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.3731,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.5076066045879966,
"learning_rate": 2.048240194884496e-05,
"loss": 0.3935,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.48689201304585306,
"learning_rate": 2.042565792045695e-05,
"loss": 0.4112,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.47106967376589726,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3761,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.4681483682571758,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.3832,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.47918147180293086,
"learning_rate": 2.025540709431948e-05,
"loss": 0.3795,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.4730815400042616,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3922,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.4716943641195888,
"learning_rate": 2.014189549672245e-05,
"loss": 0.3766,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.4786182726260851,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.4046,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.4790814244628887,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.3915,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.4894735529176241,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.3744,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.4767976619775886,
"learning_rate": 1.991486224484123e-05,
"loss": 0.3818,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.4704910670723517,
"learning_rate": 1.985810450327756e-05,
"loss": 0.3877,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.4911480271568773,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.3753,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.4888722067282446,
"learning_rate": 1.974459290568053e-05,
"loss": 0.3807,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.49836571802170593,
"learning_rate": 1.968783996385397e-05,
"loss": 0.3814,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.47226717805061924,
"learning_rate": 1.963108953612143e-05,
"loss": 0.4079,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.4735755397614252,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3783,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.4817356541203408,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3796,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.48190856511030916,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.4007,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.48011013156204807,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.3878,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.5186987647552895,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.4148,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.472251514896484,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3806,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.4855327563034292,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.404,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.4758074777741581,
"learning_rate": 1.917723146291022e-05,
"loss": 0.3715,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.45972673265274283,
"learning_rate": 1.912052422608584e-05,
"loss": 0.3926,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.47367580076305593,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.402,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.4747008692066484,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.3677,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.4766879237628887,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3936,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.46075643462091204,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3933,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.4576649732613267,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.3841,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.5226473376426551,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.388,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.4658200441450426,
"learning_rate": 1.872379746550743e-05,
"loss": 0.3739,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.5233262754071876,
"learning_rate": 1.866715967678769e-05,
"loss": 0.3698,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.512365745796851,
"learning_rate": 1.861053262258067e-05,
"loss": 0.407,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.4781829674237248,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.3722,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.5024007424075414,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.3673,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.49504412562133876,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.3939,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.45576487730934045,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.3693,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.5089456381034941,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.4141,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.46897024532643905,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.3928,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.4556387234081726,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3756,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.511942503704913,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.3906,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.4870627517084135,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.3754,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.48339723678860774,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.3633,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.48488306916941315,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.375,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.46996767886591456,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.3767,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.4659735069971277,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.3935,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.4922445499429909,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.372,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.48086439577955137,
"learning_rate": 1.776272483901182e-05,
"loss": 0.3933,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.4635965718908389,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.3947,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.4600444344852932,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3907,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.46087187624953085,
"learning_rate": 1.759360067078035e-05,
"loss": 0.3945,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.47737162373193837,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.3818,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.4735445820170233,
"learning_rate": 1.748094737261012e-05,
"loss": 0.3739,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.4792067053384025,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.3651,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.47575950272882656,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.4043,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.46911365725419396,
"learning_rate": 1.731212071922883e-05,
"loss": 0.3932,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.47405124738485055,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.392,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.48678852213173895,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.3926,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.4744107370211298,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3915,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.4655037155657012,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.3907,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.47550657421001713,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3864,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.5368769868637894,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.3613,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.4799117040456909,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.3664,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.5393040399248687,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3966,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.4908418444671761,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3943,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.5187623355387534,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3839,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.5610854186914066,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.3865,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.48431923741052046,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3764,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.5230567228539424,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3816,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.49260982756140215,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3751,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.5179227527312028,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.3904,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.46094860869347865,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.3848,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.548662004678164,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.337,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.6483811986714486,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2372,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.4309924617093014,
"learning_rate": 1.624791353909428e-05,
"loss": 0.2387,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.7237145800792029,
"learning_rate": 1.619217781670663e-05,
"loss": 0.2185,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.8614651417073202,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.2513,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.4941541484421965,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2361,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.5785474277338208,
"learning_rate": 1.60251564499148e-05,
"loss": 0.2079,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.6480726367007109,
"learning_rate": 1.596954608907176e-05,
"loss": 0.2651,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.5819579392729073,
"learning_rate": 1.591396818895043e-05,
"loss": 0.2594,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.4825654189043782,
"learning_rate": 1.585842319716759e-05,
"loss": 0.2155,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.5192023148230461,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.2636,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.604608574913149,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.2445,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.5202829757989416,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.2515,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.4735574341724489,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.2345,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.5181595655613881,
"learning_rate": 1.558120751116291e-05,
"loss": 0.2533,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.49326345910791214,
"learning_rate": 1.552586935426876e-05,
"loss": 0.2149,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.47203382050808756,
"learning_rate": 1.547056723140774e-05,
"loss": 0.2313,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.5064148455553177,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.2019,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.4526336256783694,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2433,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.4727717663380634,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.2291,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.5143035493658922,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.2318,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.4353658353106565,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.23,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.43610983360690864,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2278,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.44400799094140336,
"learning_rate": 1.508449870007656e-05,
"loss": 0.2303,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.4095673239961345,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.219,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.4290000468627575,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.238,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.4290337498131177,
"learning_rate": 1.4919625713405e-05,
"loss": 0.2378,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.4076345576339971,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.2537,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.41771526397351816,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.2416,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.43753033076931813,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.2305,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.4095083147185756,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.2321,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.39666363263701687,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.24,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.4191338501863226,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.2416,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.409242913865038,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2442,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.4133569489364237,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.2296,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.4060864917216162,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2492,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.4115982227020448,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.2215,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.4170152217448272,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.2516,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.43626410032689855,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.2486,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.4046598644319759,
"learning_rate": 1.420958430225303e-05,
"loss": 0.1992,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.3928520900654167,
"learning_rate": 1.415527989393985e-05,
"loss": 0.24,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.40721507377871363,
"learning_rate": 1.410102255819891e-05,
"loss": 0.219,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.43152220202864056,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2255,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.39094717525308564,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2406,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.41058635662793475,
"learning_rate": 1.393853735430398e-05,
"loss": 0.2414,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.396465258329573,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.2367,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.3950635641173411,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2392,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.41969369545311597,
"learning_rate": 1.377649151173163e-05,
"loss": 0.2245,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.4146565844730358,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.2049,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.3961675075525315,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.2414,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.4012315225660514,
"learning_rate": 1.361489677627324e-05,
"loss": 0.2119,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.40741168023368807,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.2338,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.3962528561702981,
"learning_rate": 1.350742335331241e-05,
"loss": 0.2225,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.39663348384752367,
"learning_rate": 1.345376486102198e-05,
"loss": 0.2543,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.40727936246478996,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.2021,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.40219237072658637,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.2395,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.41729084835398855,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.2206,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.3877895201353035,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.2302,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.4051380643182279,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2275,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.43637546572885333,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.2136,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.4000357816223864,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.2099,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.3986432922076072,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.2165,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.4038314521701066,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.2211,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.4124991062800909,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.2181,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.3993321260092609,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.221,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.3915481506460887,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.2243,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.40470831726156453,
"learning_rate": 1.276115892225764e-05,
"loss": 0.2455,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.3984503955448292,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.2366,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.3898740034632375,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.2139,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.38333491986807533,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.2277,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.4063114681707547,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.222,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.4049456393240289,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.2292,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.3878712296663139,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.2307,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.40027045777263504,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.2244,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.3996957504408001,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.2358,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.400025804607359,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.2361,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.3992598617328097,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2319,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.3991538804106501,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.2295,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.39500175064098353,
"learning_rate": 1.213055323844297e-05,
"loss": 0.2038,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.39798082678481456,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.2349,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.3792680730531119,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.239,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.37861272987139843,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.235,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.40512043396249764,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2131,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.38904401661911664,
"learning_rate": 1.187045262895488e-05,
"loss": 0.2201,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.38648543921003126,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.2496,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.39250154689581146,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.2285,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.397312183849785,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.2382,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.3931171946283858,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.2376,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.39286079123288803,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2364,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.3986386616007412,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2272,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.39587585683408316,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.222,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.4094438512075004,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2271,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.412055746671488,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.2227,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.39924324881469553,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2479,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.38529993810749813,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.2289,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.3773263982382473,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2456,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.39958916822345464,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.2072,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.4030880936922618,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.213,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.3949382730023591,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.2444,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.3978502603230516,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.2164,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.38598833549720724,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.2105,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.39627791098220283,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2247,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.38934347628989013,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.2338,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.40652839885232916,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.2368,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.39397666328869196,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.1929,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.3884411580667285,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2369,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.387781725836,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.2366,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.4005655271160201,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.2448,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.38157894749056254,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.2364,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.4050646301085264,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.2341,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.3937978760397373,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.2316,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.4048219836552997,
"learning_rate": 1.044560362975474e-05,
"loss": 0.2314,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.3994669334170764,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.2302,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.37331589219311123,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.2072,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.3949015406746707,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.2127,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.39780879379518475,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.2389,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.3865570832292721,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.2237,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.40131374950348664,
"learning_rate": 1.014782400334433e-05,
"loss": 0.2387,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.3852558352488051,
"learning_rate": 1.009846945771296e-05,
"loss": 0.2531,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.3762309521200769,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.2515,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.3917950493963973,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2141,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.3983594722626298,
"learning_rate": 9.950885880977891e-06,
"loss": 0.244,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.376858978988554,
"learning_rate": 9.901852696139382e-06,
"loss": 0.2368,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.4009032297910105,
"learning_rate": 9.852900840391027e-06,
"loss": 0.229,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.4014782239398572,
"learning_rate": 9.804030707984313e-06,
"loss": 0.2414,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.38500863773988614,
"learning_rate": 9.755242692512599e-06,
"loss": 0.2319,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.4003950721790685,
"learning_rate": 9.70653718690782e-06,
"loss": 0.2352,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.3919159705520726,
"learning_rate": 9.657914583437454e-06,
"loss": 0.2367,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.3859233287587241,
"learning_rate": 9.609375273701246e-06,
"loss": 0.2293,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.38430697910653266,
"learning_rate": 9.560919648628133e-06,
"loss": 0.212,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.3836415939625435,
"learning_rate": 9.512548098473047e-06,
"loss": 0.2036,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.3895105564026446,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2405,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.39384075412875885,
"learning_rate": 9.416058780547987e-06,
"loss": 0.2242,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.37294082452506705,
"learning_rate": 9.367941789889714e-06,
"loss": 0.2162,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.37793520274498177,
"learning_rate": 9.319910428366607e-06,
"loss": 0.2543,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.39610253790606026,
"learning_rate": 9.271965082816667e-06,
"loss": 0.1979,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.37766286475123423,
"learning_rate": 9.224106139385111e-06,
"loss": 0.219,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.38004669556445503,
"learning_rate": 9.176333983521291e-06,
"loss": 0.2094,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.3988759861735903,
"learning_rate": 9.12864899997558e-06,
"loss": 0.2205,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.4056585136933967,
"learning_rate": 9.08105157279628e-06,
"loss": 0.2076,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.372638836997894,
"learning_rate": 9.03354208532653e-06,
"loss": 0.2424,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.41376490674960215,
"learning_rate": 8.986120920201205e-06,
"loss": 0.2324,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.4000504142613534,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2252,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.38662456121851135,
"learning_rate": 8.8915450839636e-06,
"loss": 0.2184,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.3703450635587868,
"learning_rate": 8.844391174552116e-06,
"loss": 0.2493,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.3913556013059037,
"learning_rate": 8.797327110880479e-06,
"loss": 0.233,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.3798998398684034,
"learning_rate": 8.750353271996206e-06,
"loss": 0.2236,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.38032330931595315,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2096,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.3845465258041697,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2264,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.4117777114229305,
"learning_rate": 8.609976883624377e-06,
"loss": 0.2299,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.39125914369096,
"learning_rate": 8.563367719785698e-06,
"loss": 0.2228,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.4018324827341087,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2192,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.3915674641725832,
"learning_rate": 8.47042609394269e-06,
"loss": 0.2246,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.39356521811966394,
"learning_rate": 8.424094380477432e-06,
"loss": 0.2318,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.404316061792575,
"learning_rate": 8.37785589776465e-06,
"loss": 0.2375,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.4027018527148028,
"learning_rate": 8.331711018202694e-06,
"loss": 0.2289,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.3945251119973975,
"learning_rate": 8.285660113436104e-06,
"loss": 0.2195,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.39118621281885196,
"learning_rate": 8.239703554352527e-06,
"loss": 0.2196,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.410880986584116,
"learning_rate": 8.193841711079775e-06,
"loss": 0.2128,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.40209225956115296,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2385,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.3978704757333634,
"learning_rate": 8.102403648660859e-06,
"loss": 0.234,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.3974605326739747,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2296,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.3956532282241566,
"learning_rate": 8.011348871891762e-06,
"loss": 0.2515,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.37374509270018436,
"learning_rate": 7.965966132787287e-06,
"loss": 0.2483,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.390063533445569,
"learning_rate": 7.920680314137189e-06,
"loss": 0.2371,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.38369564269562095,
"learning_rate": 7.875491780667246e-06,
"loss": 0.2163,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.37422257607798026,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2265,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.38878777035333845,
"learning_rate": 7.785408024250259e-06,
"loss": 0.2316,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.3725597861007591,
"learning_rate": 7.74051352682542e-06,
"loss": 0.2334,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.39726334128088314,
"learning_rate": 7.695717765619257e-06,
"loss": 0.19,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.41259808009994,
"learning_rate": 7.651021101410673e-06,
"loss": 0.2397,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.3675545142509866,
"learning_rate": 7.606423894180464e-06,
"loss": 0.2694,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.36876964038668925,
"learning_rate": 7.56192650310839e-06,
"loss": 0.2324,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.3863631123644292,
"learning_rate": 7.517529286570349e-06,
"loss": 0.218,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.3866670856844285,
"learning_rate": 7.473232602135387e-06,
"loss": 0.2044,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.37749828827225024,
"learning_rate": 7.429036806562935e-06,
"loss": 0.236,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.3747642949941145,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.2419,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.380535802638636,
"learning_rate": 7.340949304977567e-06,
"loss": 0.2278,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.37562305175722444,
"learning_rate": 7.297058308409282e-06,
"loss": 0.2551,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.384108606527611,
"learning_rate": 7.25326961958704e-06,
"loss": 0.2401,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.38041654936693486,
"learning_rate": 7.209583591178921e-06,
"loss": 0.2552,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.37232632881275773,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.2444,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.37850337447773763,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.2448,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.391325107412597,
"learning_rate": 7.079144982700909e-06,
"loss": 0.255,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.3799432439697686,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.2203,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.3804883676988822,
"learning_rate": 6.99270564069757e-06,
"loss": 0.2224,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.37608410172333295,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2236,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.3729885940812076,
"learning_rate": 6.906685333692871e-06,
"loss": 0.2591,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.3931298589611444,
"learning_rate": 6.86383318483769e-06,
"loss": 0.2174,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.3870721591014519,
"learning_rate": 6.821086832864139e-06,
"loss": 0.2221,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.39010826876922294,
"learning_rate": 6.77844662204546e-06,
"loss": 0.2306,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.3799891927501264,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2337,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.3952523395673949,
"learning_rate": 6.693485996688695e-06,
"loss": 0.236,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.38258080082725726,
"learning_rate": 6.651166266411801e-06,
"loss": 0.2257,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.3757640286651047,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.2373,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.37540569573791566,
"learning_rate": 6.566849674844711e-06,
"loss": 0.2067,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.3737445717226183,
"learning_rate": 6.524853492628747e-06,
"loss": 0.2098,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.38927051972198295,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.1836,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.38155404678178345,
"learning_rate": 6.441187046486648e-06,
"loss": 0.2368,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.3752853013968466,
"learning_rate": 6.399517456398567e-06,
"loss": 0.2473,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.37477153348959946,
"learning_rate": 6.357957402727164e-06,
"loss": 0.243,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.38256296440280374,
"learning_rate": 6.316507220191395e-06,
"loss": 0.2435,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.3855520395131836,
"learning_rate": 6.275167242625331e-06,
"loss": 0.2162,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.3609700900963953,
"learning_rate": 6.233937802975471e-06,
"loss": 0.2345,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.3730870175399244,
"learning_rate": 6.192819233298046e-06,
"loss": 0.2366,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.37913854976350114,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2337,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.3750441206042768,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.2214,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.37717637802582904,
"learning_rate": 6.070132051252868e-06,
"loss": 0.201,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.3770799518758209,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2233,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.3797112438127224,
"learning_rate": 5.988900993811575e-06,
"loss": 0.226,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.38496368183438284,
"learning_rate": 5.948454566959363e-06,
"loss": 0.222,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.3678787492344046,
"learning_rate": 5.908121309322328e-06,
"loss": 0.2083,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.3760941569012882,
"learning_rate": 5.867901545738976e-06,
"loss": 0.2277,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.3808674901989816,
"learning_rate": 5.827795600133774e-06,
"loss": 0.2548,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.3740202483346168,
"learning_rate": 5.787803795514466e-06,
"loss": 0.253,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.37668319827086766,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2319,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.370266663530591,
"learning_rate": 5.708163896665708e-06,
"loss": 0.205,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.37661888644834446,
"learning_rate": 5.668516443845047e-06,
"loss": 0.2418,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.3677652164821992,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.2401,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.37828346925264916,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.2202,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.36617790410833606,
"learning_rate": 5.550267900783019e-06,
"loss": 0.2094,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.3856115796929128,
"learning_rate": 5.511084049737623e-06,
"loss": 0.2154,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.37302376321223657,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.2302,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.3879091106957532,
"learning_rate": 5.433066737499948e-06,
"loss": 0.2341,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.39047732676326363,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2708,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.3733463519707308,
"learning_rate": 5.355518704628997e-06,
"loss": 0.2379,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.3743682415140321,
"learning_rate": 5.316921449248731e-06,
"loss": 0.2189,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.38775184310650573,
"learning_rate": 5.278442449364538e-06,
"loss": 0.2137,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.3678468803514688,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2424,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.36814922859089155,
"learning_rate": 5.201840454747822e-06,
"loss": 0.2395,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.37188589450570153,
"learning_rate": 5.163718076957223e-06,
"loss": 0.2271,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.37021087055948304,
"learning_rate": 5.125715188541609e-06,
"loss": 0.2363,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.3902624365692607,
"learning_rate": 5.087832095571021e-06,
"loss": 0.2119,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.38347967864482463,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.2328,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.3802675470508117,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2436,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.3630835121095611,
"learning_rate": 4.974904635542815e-06,
"loss": 0.2246,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.37681658604475776,
"learning_rate": 4.937503765719582e-06,
"loss": 0.2168,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.3786619874405179,
"learning_rate": 4.900224207170299e-06,
"loss": 0.1908,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.3657577253594321,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2222,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.3702977274633754,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.2066,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.3637640808207908,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2322,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.37710682610058205,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2297,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.3682587499317269,
"learning_rate": 4.715656557757473e-06,
"loss": 0.2623,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.387592687495728,
"learning_rate": 4.679111137620442e-06,
"loss": 0.2315,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.3782555395406791,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2475,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.3691708116966061,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2438,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.3723627931260361,
"learning_rate": 4.570216403561141e-06,
"loss": 0.2213,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.38286495821381944,
"learning_rate": 4.534166308795815e-06,
"loss": 0.2038,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.5180430869262554,
"learning_rate": 4.498240773728859e-06,
"loss": 0.1515,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.40366328215044833,
"learning_rate": 4.462440087699609e-06,
"loss": 0.1336,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.35196846816186844,
"learning_rate": 4.426764539041861e-06,
"loss": 0.1454,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 0.3044896437005124,
"learning_rate": 4.391214415081582e-06,
"loss": 0.1637,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.34082951569339537,
"learning_rate": 4.355790002134579e-06,
"loss": 0.1415,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.4529069410082483,
"learning_rate": 4.320491585504207e-06,
"loss": 0.1558,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.4211384990309533,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.1405,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.5155467525589661,
"learning_rate": 4.250273877330691e-06,
"loss": 0.1737,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.39721449995270236,
"learning_rate": 4.215355151311313e-06,
"loss": 0.1526,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.3534967535684442,
"learning_rate": 4.180563552651542e-06,
"loss": 0.1825,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.3624572325510896,
"learning_rate": 4.145899361558147e-06,
"loss": 0.1347,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.37073092673857033,
"learning_rate": 4.111362857211738e-06,
"loss": 0.1704,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.3642182178754355,
"learning_rate": 4.076954317764592e-06,
"loss": 0.1681,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.35556666713618146,
"learning_rate": 4.042674020338335e-06,
"loss": 0.1579,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.3161910160384897,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.1684,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.3110075508291893,
"learning_rate": 3.974499254868674e-06,
"loss": 0.1408,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.3101611804990918,
"learning_rate": 3.940605335895451e-06,
"loss": 0.1618,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.3384838086351906,
"learning_rate": 3.90684075707908e-06,
"loss": 0.1753,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.32196688067041396,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.1533,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.3556999783351077,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.1645,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.33019640323074856,
"learning_rate": 3.806325775702304e-06,
"loss": 0.1753,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.3014021619958938,
"learning_rate": 3.773081266416434e-06,
"loss": 0.1472,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.3054010608625938,
"learning_rate": 3.739967446503245e-06,
"loss": 0.1533,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.2956763009258815,
"learning_rate": 3.706984582656894e-06,
"loss": 0.1415,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.281524636170751,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.1266,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.2961327063948867,
"learning_rate": 3.641412784665648e-06,
"loss": 0.138,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.29550898230473027,
"learning_rate": 3.608824378627005e-06,
"loss": 0.1774,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.3153118047948625,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.15,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.2979954768738364,
"learning_rate": 3.544043864774269e-06,
"loss": 0.1581,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.2949773090367727,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.1518,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.2938848687071148,
"learning_rate": 3.479793485887819e-06,
"loss": 0.1456,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.29304570307364947,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.1491,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.2932826644045669,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.158,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.29139949598518705,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.163,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.3127055796381909,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.1599,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.2834434718399097,
"learning_rate": 3.321500420335e-06,
"loss": 0.1458,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.28267278609023555,
"learning_rate": 3.290243771741275e-06,
"loss": 0.1485,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.2811871684407441,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.1522,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.27598125316675376,
"learning_rate": 3.228134459445149e-06,
"loss": 0.1445,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.29011857363968413,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.1552,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.2958015021641299,
"learning_rate": 3.166565459258513e-06,
"loss": 0.1438,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.2961158144954576,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.1656,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.29015954135378486,
"learning_rate": 3.105538754650419e-06,
"loss": 0.1576,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.2956897779183912,
"learning_rate": 3.07522937824712e-06,
"loss": 0.178,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.27998146109847627,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.1399,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.30357676591468935,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.1425,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.28433302617697304,
"learning_rate": 2.985120078631465e-06,
"loss": 0.1346,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.28955150011468717,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.1554,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.29813350779234543,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.1712,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.2986916915415432,
"learning_rate": 2.896244091927678e-06,
"loss": 0.1654,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.29568914938578084,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.1751,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.2824040982448574,
"learning_rate": 2.837681792951994e-06,
"loss": 0.1574,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.29359413461297285,
"learning_rate": 2.808607860255981e-06,
"loss": 0.1502,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.2895708933789207,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.147,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.2854616198703325,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.1457,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.288405757376502,
"learning_rate": 2.722217735873718e-06,
"loss": 0.1559,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.29421556358364787,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.1411,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.28581022364656355,
"learning_rate": 2.665319697490205e-06,
"loss": 0.1501,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.2897650858608695,
"learning_rate": 2.637079980714945e-06,
"loss": 0.1686,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.2994023856856442,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.1688,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.28687141255200443,
"learning_rate": 2.581020289189571e-06,
"loss": 0.1679,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.2869071758624578,
"learning_rate": 2.553200765936501e-06,
"loss": 0.1621,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.2954702194643741,
"learning_rate": 2.525521756805962e-06,
"loss": 0.1599,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.2890963186912492,
"learning_rate": 2.497983484720885e-06,
"loss": 0.1667,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.2876224841650776,
"learning_rate": 2.470586171470728e-06,
"loss": 0.1424,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.2994653417684325,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.1566,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.2963042452350008,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.1646,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.29047260307012934,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.1406,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.29853552474137296,
"learning_rate": 2.362410902836978e-06,
"loss": 0.1314,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.2977875868056949,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.1566,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.2847228156523625,
"learning_rate": 2.309174704446411e-06,
"loss": 0.1584,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.2831162819316391,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.18,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.2940841880580904,
"learning_rate": 2.256508422807855e-06,
"loss": 0.1558,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.2953300340497041,
"learning_rate": 2.230389531696946e-06,
"loss": 0.1463,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.28785584140134624,
"learning_rate": 2.204413754585857e-06,
"loss": 0.1783,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.28751418922083033,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.1562,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.2945524265113578,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.1555,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.28546153062908636,
"learning_rate": 2.127347193531757e-06,
"loss": 0.1601,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.280832151526901,
"learning_rate": 2.101945952921942e-06,
"loss": 0.1301,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.2834987998901822,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.1436,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.2879134268917202,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.1228,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.2851934939825675,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.1552,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.27916477782476423,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.1616,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.28510792885851194,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.1573,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.28283908386918144,
"learning_rate": 1.95257270594031e-06,
"loss": 0.1432,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.2897599961669186,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.1632,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.27658654050392184,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.1321,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.2867534257747529,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.1685,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.29891197996490665,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.1833,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.2965012453272018,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.1657,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.2873798518825974,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.1391,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.28407818068866164,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.1596,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.2910937521408632,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.1561,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.2774596901246709,
"learning_rate": 1.738336950434061e-06,
"loss": 0.1421,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.2897714015189224,
"learning_rate": 1.715265841375957e-06,
"loss": 0.1574,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.2926199664998968,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.151,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.2980763789342174,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.147,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.2977066380807668,
"learning_rate": 1.646936827927441e-06,
"loss": 0.1526,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.28236188757082564,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.1536,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.27302909337740716,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.1531,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.2898985064238512,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.1847,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.27516550512865157,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.1585,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.27116587953229127,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.1391,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.2949477902687687,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.155,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.26972153406273885,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.1405,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.2862597056774692,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.1428,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.2940771343920121,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.1862,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.28568971941600224,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.1556,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.27013942301826327,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.1358,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.29508824168532083,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.1565,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.29095081336014117,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.1882,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.28828729882315424,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.1627,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.2856023929127915,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.155,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.2900033978375863,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.1542,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.2883861565526437,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.1619,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.2710673638822653,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.1574,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.28224527616415734,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.1603,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.28987584754749857,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.1533,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.28899208943064697,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.1691,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.287969638453569,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.1704,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.28686964307820373,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.1563,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.29159242892313514,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.148,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.28255625394324696,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.1447,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.2797204503584998,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.1616,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.2732559593326253,
"learning_rate": 1.092402023876933e-06,
"loss": 0.1638,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.2843697255616171,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.161,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.2724947081361211,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.1661,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.2964807139748485,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.1735,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.2992634602471327,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.1564,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.275141313902956,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.1557,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.2958256857068839,
"learning_rate": 9.841383857091947e-07,
"loss": 0.1359,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.2829648213370622,
"learning_rate": 9.666296380321616e-07,
"loss": 0.1554,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.26880359795182296,
"learning_rate": 9.492741825034124e-07,
"loss": 0.1574,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.29169658159677114,
"learning_rate": 9.320721589013892e-07,
"loss": 0.1634,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.2788637263504898,
"learning_rate": 9.150237057688339e-07,
"loss": 0.1682,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.2800195691214698,
"learning_rate": 8.981289604116328e-07,
"loss": 0.1402,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.2785993574832048,
"learning_rate": 8.813880588977542e-07,
"loss": 0.1588,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.28621500952032075,
"learning_rate": 8.648011360561126e-07,
"loss": 0.1707,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.2823617763683345,
"learning_rate": 8.483683254755037e-07,
"loss": 0.138,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.2908675221821683,
"learning_rate": 8.320897595035227e-07,
"loss": 0.1675,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.2824463751116879,
"learning_rate": 8.159655692455093e-07,
"loss": 0.1628,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.2771133399727016,
"learning_rate": 7.999958845634648e-07,
"loss": 0.1526,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.2908967732453443,
"learning_rate": 7.841808340750478e-07,
"loss": 0.1582,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.2850517496302957,
"learning_rate": 7.685205451524869e-07,
"loss": 0.1394,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.2783222397206319,
"learning_rate": 7.530151439216027e-07,
"loss": 0.1499,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.2944343682770104,
"learning_rate": 7.376647552607675e-07,
"loss": 0.1179,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.2921881297819173,
"learning_rate": 7.224695027998963e-07,
"loss": 0.1325,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.28604121298471474,
"learning_rate": 7.07429508919466e-07,
"loss": 0.1498,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.28436352914521634,
"learning_rate": 6.925448947495206e-07,
"loss": 0.1524,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.2916159344042771,
"learning_rate": 6.778157801686936e-07,
"loss": 0.1489,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.29564154260270886,
"learning_rate": 6.632422838032515e-07,
"loss": 0.1643,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.28241503253911093,
"learning_rate": 6.488245230261281e-07,
"loss": 0.1356,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.29023921019741594,
"learning_rate": 6.345626139559868e-07,
"loss": 0.1614,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.28829350567858514,
"learning_rate": 6.204566714562866e-07,
"loss": 0.1444,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.29816909669372405,
"learning_rate": 6.06506809134344e-07,
"loss": 0.1494,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.29286390722254,
"learning_rate": 5.927131393404373e-07,
"loss": 0.1686,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.2834853626589851,
"learning_rate": 5.790757731668817e-07,
"loss": 0.1407,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.2932472431610219,
"learning_rate": 5.655948204471507e-07,
"loss": 0.1674,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.28536981988022997,
"learning_rate": 5.522703897549875e-07,
"loss": 0.1682,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.2854086996635656,
"learning_rate": 5.391025884035239e-07,
"loss": 0.1597,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.28551342985828077,
"learning_rate": 5.260915224444207e-07,
"loss": 0.1358,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.2957755715867365,
"learning_rate": 5.132372966670129e-07,
"loss": 0.1795,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.2921886048774336,
"learning_rate": 5.005400145974704e-07,
"loss": 0.1613,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.2858494930470142,
"learning_rate": 4.879997784979562e-07,
"loss": 0.1322,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.2844363307042408,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.1622,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.2706948931080056,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.1397,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.2842214212555906,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.1326,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.30797064227163196,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.1866,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.2999804547770024,
"learning_rate": 4.27657778143431e-07,
"loss": 0.1443,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.2962682786890784,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.1419,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.29512521886813825,
"learning_rate": 4.046237371675177e-07,
"loss": 0.1532,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.2769998247171845,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.1431,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.2820204689112018,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.1362,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.27745176800130433,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.1414,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.28218671761446923,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.1647,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.29605229983037123,
"learning_rate": 3.498020255137813e-07,
"loss": 0.1549,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.30289023427088835,
"learning_rate": 3.393121163057811e-07,
"loss": 0.1787,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.282034983590021,
"learning_rate": 3.289805515751399e-07,
"loss": 0.1522,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.29367449540238144,
"learning_rate": 3.188074145308573e-07,
"loss": 0.1479,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.2871629190662411,
"learning_rate": 3.087927871059804e-07,
"loss": 0.1499,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.27250487514846666,
"learning_rate": 2.989367499569418e-07,
"loss": 0.154,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.26995800560197575,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.1393,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.2845708768076084,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.1706,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.29100280475753854,
"learning_rate": 2.703209675663887e-07,
"loss": 0.151,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.290989301374153,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.1461,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.2736977949136364,
"learning_rate": 2.520381518804471e-07,
"loss": 0.1798,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.2825825736565179,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.1479,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.2975499166749772,
"learning_rate": 2.343915243930317e-07,
"loss": 0.1577,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.28527264059547475,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.167,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.28330685076381884,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.1472,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.28446741107646434,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.1571,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.26653358668276067,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.1594,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.28114077020166733,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.1576,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.270496113180834,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.1434,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.28317898468501423,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.1517,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.2847973705387425,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.1785,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.28811956122493604,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.1482,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.28651288305948563,
"learning_rate": 1.557203888516745e-07,
"loss": 0.1659,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.27136151703514727,
"learning_rate": 1.487313098987131e-07,
"loss": 0.1197,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.2813963774191875,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.1631,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.28909761180315485,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.1428,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.2891228820638488,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.1469,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.27645228387867016,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.162,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.27352097574627876,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.1495,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.28559702608147,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.1449,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.2904318407777788,
"learning_rate": 1.042873700093061e-07,
"loss": 0.1575,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.29421031627930744,
"learning_rate": 9.857879404556291e-08,
"loss": 0.1582,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.279740883621928,
"learning_rate": 9.303050139225722e-08,
"loss": 0.141,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.27929292598504185,
"learning_rate": 8.76425367345779e-08,
"loss": 0.1653,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.27176161779472385,
"learning_rate": 8.241494346644897e-08,
"loss": 0.174,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.2906680126526785,
"learning_rate": 7.734776369019204e-08,
"loss": 0.1594,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.2852148071130496,
"learning_rate": 7.244103821617332e-08,
"loss": 0.1549,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.28956039096900515,
"learning_rate": 6.769480656248606e-08,
"loss": 0.1605,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.2845968476199527,
"learning_rate": 6.310910695462635e-08,
"loss": 0.1335,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.27999103416893634,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.1482,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.2850341190561167,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.1616,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.28240725128519917,
"learning_rate": 5.031556326567488e-08,
"loss": 0.1474,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.26912246305438003,
"learning_rate": 4.637234823364312e-08,
"loss": 0.17,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.2839208105070727,
"learning_rate": 4.258983697558838e-08,
"loss": 0.1412,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.2973541354938876,
"learning_rate": 3.896805995533548e-08,
"loss": 0.157,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.2875567752053633,
"learning_rate": 3.550704634218028e-08,
"loss": 0.1525,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.28891565256157936,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.1734,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.28331480324435826,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.1423,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.2829512199917889,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.172,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.28767920880032904,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.1613,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.28643972079931307,
"learning_rate": 2.061435966221881e-08,
"loss": 0.131,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.3045210959607141,
"learning_rate": 1.811846652477245e-08,
"loss": 0.171,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.27924089247811257,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.1542,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.2834975795275094,
"learning_rate": 1.360949303591097e-08,
"loss": 0.1537,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.2923497745322423,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.1651,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.28577678204065243,
"learning_rate": 9.744388816668172e-09,
"loss": 0.1533,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.2811918362814767,
"learning_rate": 8.05332740472009e-09,
"loss": 0.138,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.2796249314107885,
"learning_rate": 6.523278382872811e-09,
"loss": 0.1714,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.27281012861408516,
"learning_rate": 5.15425407393133e-09,
"loss": 0.1576,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.27678360933957147,
"learning_rate": 3.94626550383137e-09,
"loss": 0.1517,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.30369133918609154,
"learning_rate": 2.899322401546112e-09,
"loss": 0.1636,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.28460776016975164,
"learning_rate": 2.013433199010706e-09,
"loss": 0.145,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.27393567290878423,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.1612,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.26582694318598615,
"learning_rate": 7.248437353468695e-10,
"loss": 0.1231,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.2733834870135608,
"learning_rate": 3.221538523412449e-10,
"loss": 0.1415,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.27884812777416856,
"learning_rate": 8.053862524670663e-11,
"loss": 0.1466,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.29586491216484684,
"learning_rate": 0.0,
"loss": 0.1498,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 1.7399549575514767e+18,
"train_loss": 0.440576109496074,
"train_runtime": 29537.6018,
"train_samples_per_second": 5.349,
"train_steps_per_second": 0.042
}
],
"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": 1.7399549575514767e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}