LLaMA-O1-Supervised-1129 / trainer_state.json
Di Zhang
Upload folder using huggingface_hub
6e2ada0 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9985559566787003,
"eval_steps": 500,
"global_step": 1730,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0011552346570397113,
"grad_norm": 0.717101514339447,
"learning_rate": 0.0,
"loss": 1.5901,
"step": 1
},
{
"epoch": 0.0023104693140794225,
"grad_norm": 0.7195026874542236,
"learning_rate": 5.017166594399687e-07,
"loss": 1.583,
"step": 2
},
{
"epoch": 0.0034657039711191336,
"grad_norm": 0.7274531722068787,
"learning_rate": 7.952020911994375e-07,
"loss": 1.6153,
"step": 3
},
{
"epoch": 0.004620938628158845,
"grad_norm": 0.7189211249351501,
"learning_rate": 1.0034333188799374e-06,
"loss": 1.5939,
"step": 4
},
{
"epoch": 0.005776173285198556,
"grad_norm": 0.7074140906333923,
"learning_rate": 1.164950007226698e-06,
"loss": 1.6126,
"step": 5
},
{
"epoch": 0.006931407942238267,
"grad_norm": 0.7060695886611938,
"learning_rate": 1.2969187506394062e-06,
"loss": 1.597,
"step": 6
},
{
"epoch": 0.008086642599277978,
"grad_norm": 0.693950891494751,
"learning_rate": 1.4084967333570947e-06,
"loss": 1.584,
"step": 7
},
{
"epoch": 0.00924187725631769,
"grad_norm": 0.6204413771629333,
"learning_rate": 1.5051499783199062e-06,
"loss": 1.5333,
"step": 8
},
{
"epoch": 0.0103971119133574,
"grad_norm": 0.5834112763404846,
"learning_rate": 1.590404182398875e-06,
"loss": 1.5173,
"step": 9
},
{
"epoch": 0.011552346570397111,
"grad_norm": 0.5479596257209778,
"learning_rate": 1.666666666666667e-06,
"loss": 1.4693,
"step": 10
},
{
"epoch": 0.012707581227436824,
"grad_norm": 0.3792664110660553,
"learning_rate": 1.7356544752637086e-06,
"loss": 1.3613,
"step": 11
},
{
"epoch": 0.013862815884476534,
"grad_norm": 0.346934974193573,
"learning_rate": 1.798635410079375e-06,
"loss": 1.3096,
"step": 12
},
{
"epoch": 0.015018050541516245,
"grad_norm": 0.30829983949661255,
"learning_rate": 1.8565722538447281e-06,
"loss": 1.3096,
"step": 13
},
{
"epoch": 0.016173285198555955,
"grad_norm": 0.29539182782173157,
"learning_rate": 1.9102133927970633e-06,
"loss": 1.3067,
"step": 14
},
{
"epoch": 0.017328519855595668,
"grad_norm": 0.17869696021080017,
"learning_rate": 1.960152098426136e-06,
"loss": 1.2191,
"step": 15
},
{
"epoch": 0.01848375451263538,
"grad_norm": 0.3124949634075165,
"learning_rate": 2.0068666377598747e-06,
"loss": 1.2029,
"step": 16
},
{
"epoch": 0.01963898916967509,
"grad_norm": 0.33391040563583374,
"learning_rate": 2.0507482022971233e-06,
"loss": 1.2076,
"step": 17
},
{
"epoch": 0.0207942238267148,
"grad_norm": 0.3112436830997467,
"learning_rate": 2.0921208418388438e-06,
"loss": 1.1862,
"step": 18
},
{
"epoch": 0.021949458483754514,
"grad_norm": 0.29002705216407776,
"learning_rate": 2.1312560015880486e-06,
"loss": 1.1844,
"step": 19
},
{
"epoch": 0.023104693140794223,
"grad_norm": 0.2618213891983032,
"learning_rate": 2.1683833261066357e-06,
"loss": 1.1763,
"step": 20
},
{
"epoch": 0.024259927797833935,
"grad_norm": 0.20633897185325623,
"learning_rate": 2.2036988245565326e-06,
"loss": 1.1683,
"step": 21
},
{
"epoch": 0.025415162454873647,
"grad_norm": 0.17382092773914337,
"learning_rate": 2.2373711347036773e-06,
"loss": 1.1119,
"step": 22
},
{
"epoch": 0.026570397111913356,
"grad_norm": 0.12656044960021973,
"learning_rate": 2.269546393362655e-06,
"loss": 1.1001,
"step": 23
},
{
"epoch": 0.02772563176895307,
"grad_norm": 0.1064457818865776,
"learning_rate": 2.3003520695193436e-06,
"loss": 1.1154,
"step": 24
},
{
"epoch": 0.02888086642599278,
"grad_norm": 0.09630957245826721,
"learning_rate": 2.329900014453396e-06,
"loss": 1.0793,
"step": 25
},
{
"epoch": 0.03003610108303249,
"grad_norm": 0.10916559398174286,
"learning_rate": 2.358288913284697e-06,
"loss": 1.0672,
"step": 26
},
{
"epoch": 0.031191335740072202,
"grad_norm": 0.11219058930873871,
"learning_rate": 2.385606273598312e-06,
"loss": 1.0833,
"step": 27
},
{
"epoch": 0.03234657039711191,
"grad_norm": 0.11489821970462799,
"learning_rate": 2.4119300522370323e-06,
"loss": 1.0795,
"step": 28
},
{
"epoch": 0.03350180505415162,
"grad_norm": 0.11093555390834808,
"learning_rate": 2.4373299964982607e-06,
"loss": 1.0599,
"step": 29
},
{
"epoch": 0.034657039711191336,
"grad_norm": 0.10726247727870941,
"learning_rate": 2.4618687578661045e-06,
"loss": 1.066,
"step": 30
},
{
"epoch": 0.03581227436823105,
"grad_norm": 0.0960482507944107,
"learning_rate": 2.4856028230571215e-06,
"loss": 1.0332,
"step": 31
},
{
"epoch": 0.03696750902527076,
"grad_norm": 0.09282615780830383,
"learning_rate": 2.5085832971998437e-06,
"loss": 1.0566,
"step": 32
},
{
"epoch": 0.03812274368231047,
"grad_norm": 0.0818951278924942,
"learning_rate": 2.530856566463146e-06,
"loss": 1.0351,
"step": 33
},
{
"epoch": 0.03927797833935018,
"grad_norm": 0.07113130390644073,
"learning_rate": 2.5524648617370923e-06,
"loss": 1.0164,
"step": 34
},
{
"epoch": 0.04043321299638989,
"grad_norm": 0.062365710735321045,
"learning_rate": 2.5734467405837933e-06,
"loss": 1.0125,
"step": 35
},
{
"epoch": 0.0415884476534296,
"grad_norm": 0.057508960366249084,
"learning_rate": 2.5938375012788124e-06,
"loss": 1.0392,
"step": 36
},
{
"epoch": 0.042743682310469315,
"grad_norm": 0.0539795346558094,
"learning_rate": 2.6136695401116585e-06,
"loss": 1.0244,
"step": 37
},
{
"epoch": 0.04389891696750903,
"grad_norm": 0.053800784051418304,
"learning_rate": 2.632972661028017e-06,
"loss": 0.9972,
"step": 38
},
{
"epoch": 0.04505415162454874,
"grad_norm": 0.05475891754031181,
"learning_rate": 2.6517743450441657e-06,
"loss": 1.0142,
"step": 39
},
{
"epoch": 0.046209386281588445,
"grad_norm": 0.0571669340133667,
"learning_rate": 2.6700999855466042e-06,
"loss": 0.9995,
"step": 40
},
{
"epoch": 0.04736462093862816,
"grad_norm": 0.05862262472510338,
"learning_rate": 2.687973094532893e-06,
"loss": 0.998,
"step": 41
},
{
"epoch": 0.04851985559566787,
"grad_norm": 0.0590708963572979,
"learning_rate": 2.705415483996501e-06,
"loss": 0.9798,
"step": 42
},
{
"epoch": 0.04967509025270758,
"grad_norm": 0.056892745196819305,
"learning_rate": 2.722447425965978e-06,
"loss": 0.9879,
"step": 43
},
{
"epoch": 0.050830324909747295,
"grad_norm": 0.0541134737432003,
"learning_rate": 2.739087794143646e-06,
"loss": 0.9722,
"step": 44
},
{
"epoch": 0.05198555956678701,
"grad_norm": 0.048609230667352676,
"learning_rate": 2.7553541896255733e-06,
"loss": 0.9491,
"step": 45
},
{
"epoch": 0.05314079422382671,
"grad_norm": 0.04965611547231674,
"learning_rate": 2.771263052802624e-06,
"loss": 1.0037,
"step": 46
},
{
"epoch": 0.054296028880866425,
"grad_norm": 0.043069027364254,
"learning_rate": 2.7868297632261957e-06,
"loss": 0.971,
"step": 47
},
{
"epoch": 0.05545126353790614,
"grad_norm": 0.03820377215743065,
"learning_rate": 2.8020687289593126e-06,
"loss": 1.0084,
"step": 48
},
{
"epoch": 0.05660649819494585,
"grad_norm": 0.03588235378265381,
"learning_rate": 2.8169934667141895e-06,
"loss": 0.9507,
"step": 49
},
{
"epoch": 0.05776173285198556,
"grad_norm": 0.03598296642303467,
"learning_rate": 2.8316166738933647e-06,
"loss": 0.973,
"step": 50
},
{
"epoch": 0.058916967509025274,
"grad_norm": 0.0384756475687027,
"learning_rate": 2.845950293496561e-06,
"loss": 0.9758,
"step": 51
},
{
"epoch": 0.06007220216606498,
"grad_norm": 0.03798473998904228,
"learning_rate": 2.8600055727246655e-06,
"loss": 0.9991,
"step": 52
},
{
"epoch": 0.06122743682310469,
"grad_norm": 0.039415981620550156,
"learning_rate": 2.8737931160013154e-06,
"loss": 0.9661,
"step": 53
},
{
"epoch": 0.062382671480144404,
"grad_norm": 0.04056290537118912,
"learning_rate": 2.887322933038281e-06,
"loss": 0.9773,
"step": 54
},
{
"epoch": 0.06353790613718412,
"grad_norm": 0.03856893256306648,
"learning_rate": 2.900604482490407e-06,
"loss": 0.9644,
"step": 55
},
{
"epoch": 0.06469314079422382,
"grad_norm": 0.03817951679229736,
"learning_rate": 2.9136467116770013e-06,
"loss": 0.9701,
"step": 56
},
{
"epoch": 0.06584837545126354,
"grad_norm": 0.03559865057468414,
"learning_rate": 2.926458092787486e-06,
"loss": 0.9314,
"step": 57
},
{
"epoch": 0.06700361010830325,
"grad_norm": 0.03424916788935661,
"learning_rate": 2.9390466559382293e-06,
"loss": 0.9489,
"step": 58
},
{
"epoch": 0.06815884476534297,
"grad_norm": 0.03320642188191414,
"learning_rate": 2.951420019403574e-06,
"loss": 0.9697,
"step": 59
},
{
"epoch": 0.06931407942238267,
"grad_norm": 0.031773000955581665,
"learning_rate": 2.963585417306073e-06,
"loss": 0.923,
"step": 60
},
{
"epoch": 0.07046931407942238,
"grad_norm": 0.0303028617054224,
"learning_rate": 2.9755497250179457e-06,
"loss": 0.9448,
"step": 61
},
{
"epoch": 0.0716245487364621,
"grad_norm": 0.029582129791378975,
"learning_rate": 2.98731948249709e-06,
"loss": 0.9521,
"step": 62
},
{
"epoch": 0.0727797833935018,
"grad_norm": 0.029476916417479515,
"learning_rate": 2.9989009157559695e-06,
"loss": 0.9429,
"step": 63
},
{
"epoch": 0.07393501805054152,
"grad_norm": 0.02916543185710907,
"learning_rate": 3.0102999566398123e-06,
"loss": 0.9641,
"step": 64
},
{
"epoch": 0.07509025270758123,
"grad_norm": 0.029549594968557358,
"learning_rate": 3.021522261071426e-06,
"loss": 0.9555,
"step": 65
},
{
"epoch": 0.07624548736462095,
"grad_norm": 0.03075719065964222,
"learning_rate": 3.0325732259031143e-06,
"loss": 0.9531,
"step": 66
},
{
"epoch": 0.07740072202166065,
"grad_norm": 0.0289381705224514,
"learning_rate": 3.0434580045013773e-06,
"loss": 0.9358,
"step": 67
},
{
"epoch": 0.07855595667870036,
"grad_norm": 0.03085014969110489,
"learning_rate": 3.054181521177061e-06,
"loss": 0.9496,
"step": 68
},
{
"epoch": 0.07971119133574008,
"grad_norm": 0.02944289892911911,
"learning_rate": 3.064748484562093e-06,
"loss": 0.931,
"step": 69
},
{
"epoch": 0.08086642599277978,
"grad_norm": 0.02958507277071476,
"learning_rate": 3.075163400023762e-06,
"loss": 0.9674,
"step": 70
},
{
"epoch": 0.0820216606498195,
"grad_norm": 0.029631255194544792,
"learning_rate": 3.085430581198459e-06,
"loss": 0.9504,
"step": 71
},
{
"epoch": 0.0831768953068592,
"grad_norm": 0.028514275327324867,
"learning_rate": 3.095554160718781e-06,
"loss": 0.9513,
"step": 72
},
{
"epoch": 0.08433212996389891,
"grad_norm": 0.02768518030643463,
"learning_rate": 3.1055381002007602e-06,
"loss": 0.9152,
"step": 73
},
{
"epoch": 0.08548736462093863,
"grad_norm": 0.028020448982715607,
"learning_rate": 3.1153861995516275e-06,
"loss": 0.9247,
"step": 74
},
{
"epoch": 0.08664259927797834,
"grad_norm": 0.02758488804101944,
"learning_rate": 3.1251021056528336e-06,
"loss": 0.9285,
"step": 75
},
{
"epoch": 0.08779783393501805,
"grad_norm": 0.02713642828166485,
"learning_rate": 3.1346893204679857e-06,
"loss": 0.928,
"step": 76
},
{
"epoch": 0.08895306859205776,
"grad_norm": 0.026601964607834816,
"learning_rate": 3.1441512086208035e-06,
"loss": 0.9135,
"step": 77
},
{
"epoch": 0.09010830324909748,
"grad_norm": 0.02707557938992977,
"learning_rate": 3.1534910044841343e-06,
"loss": 0.9306,
"step": 78
},
{
"epoch": 0.09126353790613718,
"grad_norm": 0.027205491438508034,
"learning_rate": 3.1627118188174026e-06,
"loss": 0.9475,
"step": 79
},
{
"epoch": 0.09241877256317689,
"grad_norm": 0.02767670899629593,
"learning_rate": 3.171816644986573e-06,
"loss": 0.9445,
"step": 80
},
{
"epoch": 0.09357400722021661,
"grad_norm": 0.027132879942655563,
"learning_rate": 3.18080836479775e-06,
"loss": 0.9388,
"step": 81
},
{
"epoch": 0.09472924187725631,
"grad_norm": 0.07792849093675613,
"learning_rate": 3.1896897539728615e-06,
"loss": 0.9097,
"step": 82
},
{
"epoch": 0.09588447653429603,
"grad_norm": 0.026650305837392807,
"learning_rate": 3.1984634872934573e-06,
"loss": 0.9216,
"step": 83
},
{
"epoch": 0.09703971119133574,
"grad_norm": 0.02636835351586342,
"learning_rate": 3.2071321434364693e-06,
"loss": 0.9397,
"step": 84
},
{
"epoch": 0.09819494584837545,
"grad_norm": 0.025988014414906502,
"learning_rate": 3.2156982095238214e-06,
"loss": 0.9166,
"step": 85
},
{
"epoch": 0.09935018050541516,
"grad_norm": 0.02622906304895878,
"learning_rate": 3.2241640854059465e-06,
"loss": 0.9213,
"step": 86
},
{
"epoch": 0.10050541516245487,
"grad_norm": 0.02663242444396019,
"learning_rate": 3.232532087697698e-06,
"loss": 0.9292,
"step": 87
},
{
"epoch": 0.10166064981949459,
"grad_norm": 0.02621094323694706,
"learning_rate": 3.2408044535836154e-06,
"loss": 0.923,
"step": 88
},
{
"epoch": 0.1028158844765343,
"grad_norm": 0.02623113803565502,
"learning_rate": 3.248983344408188e-06,
"loss": 0.9252,
"step": 89
},
{
"epoch": 0.10397111913357401,
"grad_norm": 0.02616356685757637,
"learning_rate": 3.257070849065542e-06,
"loss": 0.9423,
"step": 90
},
{
"epoch": 0.10512635379061372,
"grad_norm": 0.025917142629623413,
"learning_rate": 3.2650689872018227e-06,
"loss": 0.9337,
"step": 91
},
{
"epoch": 0.10628158844765342,
"grad_norm": 0.02607305720448494,
"learning_rate": 3.2729797122425927e-06,
"loss": 0.9134,
"step": 92
},
{
"epoch": 0.10743682310469314,
"grad_norm": 0.02566445618867874,
"learning_rate": 3.280804914256559e-06,
"loss": 0.9182,
"step": 93
},
{
"epoch": 0.10859205776173285,
"grad_norm": 0.025618452578783035,
"learning_rate": 3.2885464226661647e-06,
"loss": 0.883,
"step": 94
},
{
"epoch": 0.10974729241877257,
"grad_norm": 0.02570994757115841,
"learning_rate": 3.2962060088147467e-06,
"loss": 0.9009,
"step": 95
},
{
"epoch": 0.11090252707581227,
"grad_norm": 0.02652270346879959,
"learning_rate": 3.303785388399281e-06,
"loss": 0.9264,
"step": 96
},
{
"epoch": 0.11205776173285198,
"grad_norm": 0.026324449107050896,
"learning_rate": 3.3112862237770753e-06,
"loss": 0.9072,
"step": 97
},
{
"epoch": 0.1132129963898917,
"grad_norm": 0.026178548112511635,
"learning_rate": 3.318710126154159e-06,
"loss": 0.9226,
"step": 98
},
{
"epoch": 0.1143682310469314,
"grad_norm": 0.025392455980181694,
"learning_rate": 3.3260586576625835e-06,
"loss": 0.9002,
"step": 99
},
{
"epoch": 0.11552346570397112,
"grad_norm": 0.02630504220724106,
"learning_rate": 3.333333333333334e-06,
"loss": 0.9335,
"step": 100
},
{
"epoch": 0.11667870036101083,
"grad_norm": 0.026046302169561386,
"learning_rate": 3.340535622971072e-06,
"loss": 0.9079,
"step": 101
},
{
"epoch": 0.11783393501805055,
"grad_norm": 0.025761395692825317,
"learning_rate": 3.3476669529365297e-06,
"loss": 0.9188,
"step": 102
},
{
"epoch": 0.11898916967509025,
"grad_norm": 0.025253284722566605,
"learning_rate": 3.3547287078419544e-06,
"loss": 0.9207,
"step": 103
},
{
"epoch": 0.12014440433212996,
"grad_norm": 0.025493199005723,
"learning_rate": 3.361722232164634e-06,
"loss": 0.8937,
"step": 104
},
{
"epoch": 0.12129963898916968,
"grad_norm": 0.025204647332429886,
"learning_rate": 3.3686488317832306e-06,
"loss": 0.9189,
"step": 105
},
{
"epoch": 0.12245487364620938,
"grad_norm": 0.025601711124181747,
"learning_rate": 3.375509775441284e-06,
"loss": 0.9034,
"step": 106
},
{
"epoch": 0.1236101083032491,
"grad_norm": 0.025162257254123688,
"learning_rate": 3.3823062961420163e-06,
"loss": 0.9008,
"step": 107
},
{
"epoch": 0.12476534296028881,
"grad_norm": 0.025047749280929565,
"learning_rate": 3.3890395924782498e-06,
"loss": 0.8788,
"step": 108
},
{
"epoch": 0.12592057761732853,
"grad_norm": 0.026021016761660576,
"learning_rate": 3.3957108299010395e-06,
"loss": 0.9252,
"step": 109
},
{
"epoch": 0.12707581227436823,
"grad_norm": 0.025183433666825294,
"learning_rate": 3.402321141930376e-06,
"loss": 0.9221,
"step": 110
},
{
"epoch": 0.12823104693140794,
"grad_norm": 0.02558548003435135,
"learning_rate": 3.408871631311096e-06,
"loss": 0.8957,
"step": 111
},
{
"epoch": 0.12938628158844764,
"grad_norm": 0.025612782686948776,
"learning_rate": 3.415363371116969e-06,
"loss": 0.8895,
"step": 112
},
{
"epoch": 0.13054151624548738,
"grad_norm": 0.024990661069750786,
"learning_rate": 3.4217974058057e-06,
"loss": 0.9094,
"step": 113
},
{
"epoch": 0.13169675090252708,
"grad_norm": 0.026293708011507988,
"learning_rate": 3.428174752227455e-06,
"loss": 0.9063,
"step": 114
},
{
"epoch": 0.1328519855595668,
"grad_norm": 0.02521086297929287,
"learning_rate": 3.434496400589353e-06,
"loss": 0.9164,
"step": 115
},
{
"epoch": 0.1340072202166065,
"grad_norm": 0.026099544018507004,
"learning_rate": 3.440763315378198e-06,
"loss": 0.9125,
"step": 116
},
{
"epoch": 0.1351624548736462,
"grad_norm": 0.02469206601381302,
"learning_rate": 3.446976436243603e-06,
"loss": 0.892,
"step": 117
},
{
"epoch": 0.13631768953068593,
"grad_norm": 0.025795504450798035,
"learning_rate": 3.4531366788435426e-06,
"loss": 0.896,
"step": 118
},
{
"epoch": 0.13747292418772564,
"grad_norm": 0.024762745946645737,
"learning_rate": 3.4592449356542185e-06,
"loss": 0.8718,
"step": 119
},
{
"epoch": 0.13862815884476534,
"grad_norm": 0.02494460716843605,
"learning_rate": 3.4653020767460416e-06,
"loss": 0.9059,
"step": 120
},
{
"epoch": 0.13978339350180505,
"grad_norm": 0.02601913921535015,
"learning_rate": 3.471308950527417e-06,
"loss": 0.9188,
"step": 121
},
{
"epoch": 0.14093862815884475,
"grad_norm": 0.025598011910915375,
"learning_rate": 3.4772663844579142e-06,
"loss": 0.9289,
"step": 122
},
{
"epoch": 0.1420938628158845,
"grad_norm": 0.025548091158270836,
"learning_rate": 3.48317518573233e-06,
"loss": 0.9007,
"step": 123
},
{
"epoch": 0.1432490974729242,
"grad_norm": 0.02472161501646042,
"learning_rate": 3.4890361419370587e-06,
"loss": 0.8807,
"step": 124
},
{
"epoch": 0.1444043321299639,
"grad_norm": 0.025738514959812164,
"learning_rate": 3.4948500216800947e-06,
"loss": 0.8969,
"step": 125
},
{
"epoch": 0.1455595667870036,
"grad_norm": 0.025138631463050842,
"learning_rate": 3.5006175751959385e-06,
"loss": 0.9001,
"step": 126
},
{
"epoch": 0.1467148014440433,
"grad_norm": 0.025624800473451614,
"learning_rate": 3.506339534926595e-06,
"loss": 0.9093,
"step": 127
},
{
"epoch": 0.14787003610108304,
"grad_norm": 0.02503197453916073,
"learning_rate": 3.512016616079781e-06,
"loss": 0.8793,
"step": 128
},
{
"epoch": 0.14902527075812275,
"grad_norm": 0.02515401318669319,
"learning_rate": 3.5176495171654153e-06,
"loss": 0.8855,
"step": 129
},
{
"epoch": 0.15018050541516245,
"grad_norm": 0.024870432913303375,
"learning_rate": 3.523238920511395e-06,
"loss": 0.8642,
"step": 130
},
{
"epoch": 0.15133574007220216,
"grad_norm": 0.02646622247993946,
"learning_rate": 3.528785492759607e-06,
"loss": 0.8966,
"step": 131
},
{
"epoch": 0.1524909747292419,
"grad_norm": 0.025933699682354927,
"learning_rate": 3.5342898853430833e-06,
"loss": 0.9036,
"step": 132
},
{
"epoch": 0.1536462093862816,
"grad_norm": 0.025004137307405472,
"learning_rate": 3.5397527349451433e-06,
"loss": 0.8796,
"step": 133
},
{
"epoch": 0.1548014440433213,
"grad_norm": 0.027062473818659782,
"learning_rate": 3.5451746639413463e-06,
"loss": 0.8993,
"step": 134
},
{
"epoch": 0.155956678700361,
"grad_norm": 0.025223543867468834,
"learning_rate": 3.550556280825011e-06,
"loss": 0.8887,
"step": 135
},
{
"epoch": 0.1571119133574007,
"grad_norm": 0.02600521221756935,
"learning_rate": 3.55589818061703e-06,
"loss": 0.8799,
"step": 136
},
{
"epoch": 0.15826714801444045,
"grad_norm": 0.026399623602628708,
"learning_rate": 3.5612009452606784e-06,
"loss": 0.9006,
"step": 137
},
{
"epoch": 0.15942238267148015,
"grad_norm": 0.025787660852074623,
"learning_rate": 3.5664651440020615e-06,
"loss": 0.9254,
"step": 138
},
{
"epoch": 0.16057761732851986,
"grad_norm": 0.025667186826467514,
"learning_rate": 3.5716913337568255e-06,
"loss": 0.8843,
"step": 139
},
{
"epoch": 0.16173285198555956,
"grad_norm": 0.025686215609312057,
"learning_rate": 3.5768800594637304e-06,
"loss": 0.8904,
"step": 140
},
{
"epoch": 0.16288808664259927,
"grad_norm": 0.027956154197454453,
"learning_rate": 3.582031854425634e-06,
"loss": 0.9226,
"step": 141
},
{
"epoch": 0.164043321299639,
"grad_norm": 0.025962915271520615,
"learning_rate": 3.587147240638428e-06,
"loss": 0.9165,
"step": 142
},
{
"epoch": 0.1651985559566787,
"grad_norm": 0.026672059670090675,
"learning_rate": 3.5922267291084367e-06,
"loss": 0.8984,
"step": 143
},
{
"epoch": 0.1663537906137184,
"grad_norm": 0.025454359129071236,
"learning_rate": 3.59727082015875e-06,
"loss": 0.9018,
"step": 144
},
{
"epoch": 0.16750902527075812,
"grad_norm": 0.02756405621767044,
"learning_rate": 3.6022800037249583e-06,
"loss": 0.8928,
"step": 145
},
{
"epoch": 0.16866425992779782,
"grad_norm": 0.025225916877388954,
"learning_rate": 3.607254759640729e-06,
"loss": 0.887,
"step": 146
},
{
"epoch": 0.16981949458483755,
"grad_norm": 0.026238933205604553,
"learning_rate": 3.612195557913627e-06,
"loss": 0.8906,
"step": 147
},
{
"epoch": 0.17097472924187726,
"grad_norm": 0.026623785495758057,
"learning_rate": 3.6171028589915957e-06,
"loss": 0.8866,
"step": 148
},
{
"epoch": 0.17212996389891697,
"grad_norm": 0.02516297437250614,
"learning_rate": 3.6219771140204575e-06,
"loss": 0.875,
"step": 149
},
{
"epoch": 0.17328519855595667,
"grad_norm": 0.026302075013518333,
"learning_rate": 3.626818765092802e-06,
"loss": 0.9158,
"step": 150
},
{
"epoch": 0.17444043321299638,
"grad_norm": 0.02607031911611557,
"learning_rate": 3.631628245488616e-06,
"loss": 0.8595,
"step": 151
},
{
"epoch": 0.1755956678700361,
"grad_norm": 0.025099189952015877,
"learning_rate": 3.6364059799079547e-06,
"loss": 0.8762,
"step": 152
},
{
"epoch": 0.17675090252707581,
"grad_norm": 0.024900630116462708,
"learning_rate": 3.6411523846959985e-06,
"loss": 0.8769,
"step": 153
},
{
"epoch": 0.17790613718411552,
"grad_norm": 0.02502143569290638,
"learning_rate": 3.6458678680607725e-06,
"loss": 0.9109,
"step": 154
},
{
"epoch": 0.17906137184115523,
"grad_norm": 0.025041181594133377,
"learning_rate": 3.6505528302838196e-06,
"loss": 0.8624,
"step": 155
},
{
"epoch": 0.18021660649819496,
"grad_norm": 0.02584444358944893,
"learning_rate": 3.655207663924103e-06,
"loss": 0.8581,
"step": 156
},
{
"epoch": 0.18137184115523466,
"grad_norm": 0.024667399004101753,
"learning_rate": 3.65983275401539e-06,
"loss": 0.8699,
"step": 157
},
{
"epoch": 0.18252707581227437,
"grad_norm": 0.025528263300657272,
"learning_rate": 3.664428478257371e-06,
"loss": 0.8896,
"step": 158
},
{
"epoch": 0.18368231046931408,
"grad_norm": 0.025437112897634506,
"learning_rate": 3.6689952072007528e-06,
"loss": 0.8954,
"step": 159
},
{
"epoch": 0.18483754512635378,
"grad_norm": 0.025140732526779175,
"learning_rate": 3.6735333044265414e-06,
"loss": 0.8794,
"step": 160
},
{
"epoch": 0.1859927797833935,
"grad_norm": 0.026126103475689888,
"learning_rate": 3.6780431267197503e-06,
"loss": 0.9103,
"step": 161
},
{
"epoch": 0.18714801444043322,
"grad_norm": 0.025869259610772133,
"learning_rate": 3.6825250242377186e-06,
"loss": 0.8945,
"step": 162
},
{
"epoch": 0.18830324909747292,
"grad_norm": 0.02550615929067135,
"learning_rate": 3.6869793406732633e-06,
"loss": 0.8884,
"step": 163
},
{
"epoch": 0.18945848375451263,
"grad_norm": 0.02724931389093399,
"learning_rate": 3.69140641341283e-06,
"loss": 0.893,
"step": 164
},
{
"epoch": 0.19061371841155234,
"grad_norm": 0.02519422210752964,
"learning_rate": 3.6958065736898442e-06,
"loss": 0.8739,
"step": 165
},
{
"epoch": 0.19176895306859207,
"grad_norm": 0.025711793452501297,
"learning_rate": 3.700180146733426e-06,
"loss": 0.8832,
"step": 166
},
{
"epoch": 0.19292418772563177,
"grad_norm": 0.02538181096315384,
"learning_rate": 3.7045274519126395e-06,
"loss": 0.8785,
"step": 167
},
{
"epoch": 0.19407942238267148,
"grad_norm": 0.0263076052069664,
"learning_rate": 3.7088488028764387e-06,
"loss": 0.8722,
"step": 168
},
{
"epoch": 0.19523465703971118,
"grad_norm": 0.02613968588411808,
"learning_rate": 3.7131445076894563e-06,
"loss": 0.8833,
"step": 169
},
{
"epoch": 0.1963898916967509,
"grad_norm": 0.024628346785902977,
"learning_rate": 3.717414868963791e-06,
"loss": 0.8795,
"step": 170
},
{
"epoch": 0.19754512635379062,
"grad_norm": 0.026414738968014717,
"learning_rate": 3.721660183986924e-06,
"loss": 0.9025,
"step": 171
},
{
"epoch": 0.19870036101083033,
"grad_norm": 0.025311259552836418,
"learning_rate": 3.725880744845915e-06,
"loss": 0.9047,
"step": 172
},
{
"epoch": 0.19985559566787003,
"grad_norm": 0.025876285508275032,
"learning_rate": 3.7300768385479928e-06,
"loss": 0.8571,
"step": 173
},
{
"epoch": 0.20101083032490974,
"grad_norm": 0.02702619880437851,
"learning_rate": 3.7342487471376667e-06,
"loss": 0.8776,
"step": 174
},
{
"epoch": 0.20216606498194944,
"grad_norm": 0.026685267686843872,
"learning_rate": 3.7383967478104918e-06,
"loss": 0.9018,
"step": 175
},
{
"epoch": 0.20332129963898918,
"grad_norm": 0.026544688269495964,
"learning_rate": 3.7425211130235835e-06,
"loss": 0.8861,
"step": 176
},
{
"epoch": 0.20447653429602888,
"grad_norm": 0.026582278311252594,
"learning_rate": 3.7466221106030114e-06,
"loss": 0.8974,
"step": 177
},
{
"epoch": 0.2056317689530686,
"grad_norm": 0.02530672959983349,
"learning_rate": 3.7507000038481574e-06,
"loss": 0.8904,
"step": 178
},
{
"epoch": 0.2067870036101083,
"grad_norm": 0.02869655378162861,
"learning_rate": 3.7547550516331556e-06,
"loss": 0.8637,
"step": 179
},
{
"epoch": 0.20794223826714803,
"grad_norm": 0.0269013699144125,
"learning_rate": 3.7587875085055104e-06,
"loss": 0.8718,
"step": 180
},
{
"epoch": 0.20909747292418773,
"grad_norm": 0.027417806908488274,
"learning_rate": 3.762797624781975e-06,
"loss": 0.8697,
"step": 181
},
{
"epoch": 0.21025270758122744,
"grad_norm": 0.025989564135670662,
"learning_rate": 3.7667856466417917e-06,
"loss": 0.8838,
"step": 182
},
{
"epoch": 0.21140794223826714,
"grad_norm": 0.026576591655611992,
"learning_rate": 3.7707518162173835e-06,
"loss": 0.8704,
"step": 183
},
{
"epoch": 0.21256317689530685,
"grad_norm": 0.0275897067040205,
"learning_rate": 3.7746963716825613e-06,
"loss": 0.855,
"step": 184
},
{
"epoch": 0.21371841155234658,
"grad_norm": 0.0267089381814003,
"learning_rate": 3.778619547338356e-06,
"loss": 0.8712,
"step": 185
},
{
"epoch": 0.2148736462093863,
"grad_norm": 0.02677794173359871,
"learning_rate": 3.7825215736965283e-06,
"loss": 0.8729,
"step": 186
},
{
"epoch": 0.216028880866426,
"grad_norm": 0.0281060878187418,
"learning_rate": 3.786402677560832e-06,
"loss": 0.8845,
"step": 187
},
{
"epoch": 0.2171841155234657,
"grad_norm": 0.025514084845781326,
"learning_rate": 3.7902630821061337e-06,
"loss": 0.8669,
"step": 188
},
{
"epoch": 0.2183393501805054,
"grad_norm": 0.02548467181622982,
"learning_rate": 3.7941030069554073e-06,
"loss": 0.8897,
"step": 189
},
{
"epoch": 0.21949458483754514,
"grad_norm": 0.026287022978067398,
"learning_rate": 3.7979226682547152e-06,
"loss": 0.8731,
"step": 190
},
{
"epoch": 0.22064981949458484,
"grad_norm": 0.025090090930461884,
"learning_rate": 3.8017222787462132e-06,
"loss": 0.8946,
"step": 191
},
{
"epoch": 0.22180505415162455,
"grad_norm": 0.02495192363858223,
"learning_rate": 3.8055020478392497e-06,
"loss": 0.8567,
"step": 192
},
{
"epoch": 0.22296028880866425,
"grad_norm": 0.025306979194283485,
"learning_rate": 3.8092621816796233e-06,
"loss": 0.8819,
"step": 193
},
{
"epoch": 0.22411552346570396,
"grad_norm": 0.027035973966121674,
"learning_rate": 3.813002883217044e-06,
"loss": 0.8733,
"step": 194
},
{
"epoch": 0.2252707581227437,
"grad_norm": 0.025134000927209854,
"learning_rate": 3.816724352270864e-06,
"loss": 0.8764,
"step": 195
},
{
"epoch": 0.2264259927797834,
"grad_norm": 0.025370297953486443,
"learning_rate": 3.820426785594127e-06,
"loss": 0.8607,
"step": 196
},
{
"epoch": 0.2275812274368231,
"grad_norm": 0.02679363079369068,
"learning_rate": 3.824110376935989e-06,
"loss": 0.8726,
"step": 197
},
{
"epoch": 0.2287364620938628,
"grad_norm": 0.024456890299916267,
"learning_rate": 3.827775317102552e-06,
"loss": 0.858,
"step": 198
},
{
"epoch": 0.2298916967509025,
"grad_norm": 0.026032116264104843,
"learning_rate": 3.831421794016178e-06,
"loss": 0.8606,
"step": 199
},
{
"epoch": 0.23104693140794225,
"grad_norm": 0.026059836149215698,
"learning_rate": 3.835049992773302e-06,
"loss": 0.8613,
"step": 200
},
{
"epoch": 0.23220216606498195,
"grad_norm": 0.02430492639541626,
"learning_rate": 3.8386600957008155e-06,
"loss": 0.866,
"step": 201
},
{
"epoch": 0.23335740072202166,
"grad_norm": 0.026345418766140938,
"learning_rate": 3.8422522824110405e-06,
"loss": 0.8814,
"step": 202
},
{
"epoch": 0.23451263537906136,
"grad_norm": 0.025827715173363686,
"learning_rate": 3.8458267298553554e-06,
"loss": 0.8791,
"step": 203
},
{
"epoch": 0.2356678700361011,
"grad_norm": 0.025157129392027855,
"learning_rate": 3.849383612376498e-06,
"loss": 0.8797,
"step": 204
},
{
"epoch": 0.2368231046931408,
"grad_norm": 0.02571243792772293,
"learning_rate": 3.852923101759591e-06,
"loss": 0.8866,
"step": 205
},
{
"epoch": 0.2379783393501805,
"grad_norm": 0.025041643530130386,
"learning_rate": 3.856445367281923e-06,
"loss": 0.8666,
"step": 206
},
{
"epoch": 0.2391335740072202,
"grad_norm": 0.0259910486638546,
"learning_rate": 3.8599505757615295e-06,
"loss": 0.8727,
"step": 207
},
{
"epoch": 0.24028880866425992,
"grad_norm": 0.02511589229106903,
"learning_rate": 3.863438891604603e-06,
"loss": 0.8801,
"step": 208
},
{
"epoch": 0.24144404332129965,
"grad_norm": 0.024072881788015366,
"learning_rate": 3.866910476851757e-06,
"loss": 0.8203,
"step": 209
},
{
"epoch": 0.24259927797833936,
"grad_norm": 0.025221582502126694,
"learning_rate": 3.870365491223199e-06,
"loss": 0.8384,
"step": 210
},
{
"epoch": 0.24375451263537906,
"grad_norm": 0.02576667070388794,
"learning_rate": 3.873804092162822e-06,
"loss": 0.8423,
"step": 211
},
{
"epoch": 0.24490974729241877,
"grad_norm": 0.02616226114332676,
"learning_rate": 3.877226434881253e-06,
"loss": 0.8521,
"step": 212
},
{
"epoch": 0.24606498194945847,
"grad_norm": 0.024735651910305023,
"learning_rate": 3.880632672397897e-06,
"loss": 0.8847,
"step": 213
},
{
"epoch": 0.2472202166064982,
"grad_norm": 0.02537655271589756,
"learning_rate": 3.884022955581984e-06,
"loss": 0.86,
"step": 214
},
{
"epoch": 0.2483754512635379,
"grad_norm": 0.02478659339249134,
"learning_rate": 3.887397433192676e-06,
"loss": 0.8489,
"step": 215
},
{
"epoch": 0.24953068592057762,
"grad_norm": 0.025592437013983727,
"learning_rate": 3.890756251918219e-06,
"loss": 0.845,
"step": 216
},
{
"epoch": 0.25068592057761735,
"grad_norm": 0.02625833824276924,
"learning_rate": 3.894099556414216e-06,
"loss": 0.8584,
"step": 217
},
{
"epoch": 0.25184115523465705,
"grad_norm": 0.02509414032101631,
"learning_rate": 3.897427489341009e-06,
"loss": 0.8608,
"step": 218
},
{
"epoch": 0.25299638989169676,
"grad_norm": 0.02498655766248703,
"learning_rate": 3.900740191400198e-06,
"loss": 0.8509,
"step": 219
},
{
"epoch": 0.25415162454873647,
"grad_norm": 0.025928007438778877,
"learning_rate": 3.9040378013703444e-06,
"loss": 0.8556,
"step": 220
},
{
"epoch": 0.25530685920577617,
"grad_norm": 0.026060784235596657,
"learning_rate": 3.907320456141851e-06,
"loss": 0.8495,
"step": 221
},
{
"epoch": 0.2564620938628159,
"grad_norm": 0.02592143975198269,
"learning_rate": 3.910588290751064e-06,
"loss": 0.8527,
"step": 222
},
{
"epoch": 0.2576173285198556,
"grad_norm": 0.025333942845463753,
"learning_rate": 3.913841438413601e-06,
"loss": 0.8398,
"step": 223
},
{
"epoch": 0.2587725631768953,
"grad_norm": 0.026303421705961227,
"learning_rate": 3.9170800305569385e-06,
"loss": 0.8575,
"step": 224
},
{
"epoch": 0.259927797833935,
"grad_norm": 0.025790376588702202,
"learning_rate": 3.920304196852272e-06,
"loss": 0.8804,
"step": 225
},
{
"epoch": 0.26108303249097475,
"grad_norm": 0.02600492723286152,
"learning_rate": 3.923514065245669e-06,
"loss": 0.8532,
"step": 226
},
{
"epoch": 0.26223826714801446,
"grad_norm": 0.025459101423621178,
"learning_rate": 3.9267097619885385e-06,
"loss": 0.8597,
"step": 227
},
{
"epoch": 0.26339350180505416,
"grad_norm": 0.025850312784314156,
"learning_rate": 3.9298914116674236e-06,
"loss": 0.8605,
"step": 228
},
{
"epoch": 0.26454873646209387,
"grad_norm": 0.0261206217110157,
"learning_rate": 3.9330591372331475e-06,
"loss": 0.8463,
"step": 229
},
{
"epoch": 0.2657039711191336,
"grad_norm": 0.026850640773773193,
"learning_rate": 3.936213060029322e-06,
"loss": 0.8655,
"step": 230
},
{
"epoch": 0.2668592057761733,
"grad_norm": 0.02615203708410263,
"learning_rate": 3.9393532998202405e-06,
"loss": 0.8729,
"step": 231
},
{
"epoch": 0.268014440433213,
"grad_norm": 0.026082858443260193,
"learning_rate": 3.942479974818166e-06,
"loss": 0.8626,
"step": 232
},
{
"epoch": 0.2691696750902527,
"grad_norm": 0.02707120031118393,
"learning_rate": 3.9455932017100315e-06,
"loss": 0.8665,
"step": 233
},
{
"epoch": 0.2703249097472924,
"grad_norm": 0.02531982958316803,
"learning_rate": 3.9486930956835726e-06,
"loss": 0.841,
"step": 234
},
{
"epoch": 0.27148014440433216,
"grad_norm": 0.02577449567615986,
"learning_rate": 3.951779770452894e-06,
"loss": 0.8528,
"step": 235
},
{
"epoch": 0.27263537906137186,
"grad_norm": 0.025777166709303856,
"learning_rate": 3.954853338283512e-06,
"loss": 0.8636,
"step": 236
},
{
"epoch": 0.27379061371841157,
"grad_norm": 0.027185678482055664,
"learning_rate": 3.95791391001684e-06,
"loss": 0.884,
"step": 237
},
{
"epoch": 0.2749458483754513,
"grad_norm": 0.025784213095903397,
"learning_rate": 3.960961595094187e-06,
"loss": 0.8532,
"step": 238
},
{
"epoch": 0.276101083032491,
"grad_norm": 0.025824446231126785,
"learning_rate": 3.96399650158023e-06,
"loss": 0.8555,
"step": 239
},
{
"epoch": 0.2772563176895307,
"grad_norm": 0.026414718478918076,
"learning_rate": 3.96701873618601e-06,
"loss": 0.8634,
"step": 240
},
{
"epoch": 0.2784115523465704,
"grad_norm": 0.02620028331875801,
"learning_rate": 3.970028404291448e-06,
"loss": 0.8574,
"step": 241
},
{
"epoch": 0.2795667870036101,
"grad_norm": 0.025768935680389404,
"learning_rate": 3.973025609967386e-06,
"loss": 0.8741,
"step": 242
},
{
"epoch": 0.2807220216606498,
"grad_norm": 0.027099501341581345,
"learning_rate": 3.976010455997187e-06,
"loss": 0.8705,
"step": 243
},
{
"epoch": 0.2818772563176895,
"grad_norm": 0.02535291016101837,
"learning_rate": 3.978983043897884e-06,
"loss": 0.853,
"step": 244
},
{
"epoch": 0.28303249097472927,
"grad_norm": 0.02731909230351448,
"learning_rate": 3.981943473940888e-06,
"loss": 0.8776,
"step": 245
},
{
"epoch": 0.284187725631769,
"grad_norm": 0.02630050666630268,
"learning_rate": 3.984891845172299e-06,
"loss": 0.86,
"step": 246
},
{
"epoch": 0.2853429602888087,
"grad_norm": 0.025558151304721832,
"learning_rate": 3.987828255432777e-06,
"loss": 0.8758,
"step": 247
},
{
"epoch": 0.2864981949458484,
"grad_norm": 0.026159491389989853,
"learning_rate": 3.990752801377028e-06,
"loss": 0.8626,
"step": 248
},
{
"epoch": 0.2876534296028881,
"grad_norm": 0.02615942806005478,
"learning_rate": 3.993665578492894e-06,
"loss": 0.858,
"step": 249
},
{
"epoch": 0.2888086642599278,
"grad_norm": 0.025161130353808403,
"learning_rate": 3.996566681120062e-06,
"loss": 0.8429,
"step": 250
},
{
"epoch": 0.2899638989169675,
"grad_norm": 0.02635674737393856,
"learning_rate": 3.999456202468397e-06,
"loss": 0.8686,
"step": 251
},
{
"epoch": 0.2911191335740072,
"grad_norm": 0.025895683094859123,
"learning_rate": 4.0023342346359075e-06,
"loss": 0.8422,
"step": 252
},
{
"epoch": 0.2922743682310469,
"grad_norm": 0.027750222012400627,
"learning_rate": 4.0052008686263635e-06,
"loss": 0.8868,
"step": 253
},
{
"epoch": 0.2934296028880866,
"grad_norm": 0.02600325271487236,
"learning_rate": 4.0080561943665644e-06,
"loss": 0.8639,
"step": 254
},
{
"epoch": 0.2945848375451264,
"grad_norm": 0.02592737227678299,
"learning_rate": 4.010900300723259e-06,
"loss": 0.8453,
"step": 255
},
{
"epoch": 0.2957400722021661,
"grad_norm": 0.025766368955373764,
"learning_rate": 4.0137332755197495e-06,
"loss": 0.8771,
"step": 256
},
{
"epoch": 0.2968953068592058,
"grad_norm": 0.026158476248383522,
"learning_rate": 4.016555205552159e-06,
"loss": 0.8524,
"step": 257
},
{
"epoch": 0.2980505415162455,
"grad_norm": 0.024570690467953682,
"learning_rate": 4.019366176605384e-06,
"loss": 0.8345,
"step": 258
},
{
"epoch": 0.2992057761732852,
"grad_norm": 0.025776837021112442,
"learning_rate": 4.022166273468753e-06,
"loss": 0.8441,
"step": 259
},
{
"epoch": 0.3003610108303249,
"grad_norm": 0.02488660253584385,
"learning_rate": 4.024955579951363e-06,
"loss": 0.8639,
"step": 260
},
{
"epoch": 0.3015162454873646,
"grad_norm": 0.025739185512065887,
"learning_rate": 4.0277341788971355e-06,
"loss": 0.8465,
"step": 261
},
{
"epoch": 0.3026714801444043,
"grad_norm": 0.02594810351729393,
"learning_rate": 4.030502152199576e-06,
"loss": 0.8752,
"step": 262
},
{
"epoch": 0.303826714801444,
"grad_norm": 0.025780096650123596,
"learning_rate": 4.033259580816264e-06,
"loss": 0.8433,
"step": 263
},
{
"epoch": 0.3049819494584838,
"grad_norm": 0.025661000981926918,
"learning_rate": 4.036006544783052e-06,
"loss": 0.8769,
"step": 264
},
{
"epoch": 0.3061371841155235,
"grad_norm": 0.025767603889107704,
"learning_rate": 4.0387431232280135e-06,
"loss": 0.8705,
"step": 265
},
{
"epoch": 0.3072924187725632,
"grad_norm": 0.0245257169008255,
"learning_rate": 4.041469394385113e-06,
"loss": 0.8497,
"step": 266
},
{
"epoch": 0.3084476534296029,
"grad_norm": 0.026073751971125603,
"learning_rate": 4.044185435607626e-06,
"loss": 0.8503,
"step": 267
},
{
"epoch": 0.3096028880866426,
"grad_norm": 0.025399446487426758,
"learning_rate": 4.046891323381315e-06,
"loss": 0.8594,
"step": 268
},
{
"epoch": 0.3107581227436823,
"grad_norm": 0.02536724880337715,
"learning_rate": 4.049587133337347e-06,
"loss": 0.8513,
"step": 269
},
{
"epoch": 0.311913357400722,
"grad_norm": 0.024877896532416344,
"learning_rate": 4.05227294026498e-06,
"loss": 0.8302,
"step": 270
},
{
"epoch": 0.3130685920577617,
"grad_norm": 0.026714155450463295,
"learning_rate": 4.05494881812401e-06,
"loss": 0.8787,
"step": 271
},
{
"epoch": 0.3142238267148014,
"grad_norm": 0.02595258131623268,
"learning_rate": 4.057614840056999e-06,
"loss": 0.8572,
"step": 272
},
{
"epoch": 0.31537906137184113,
"grad_norm": 0.02663499116897583,
"learning_rate": 4.060271078401261e-06,
"loss": 0.8656,
"step": 273
},
{
"epoch": 0.3165342960288809,
"grad_norm": 0.02580447867512703,
"learning_rate": 4.0629176047006474e-06,
"loss": 0.8574,
"step": 274
},
{
"epoch": 0.3176895306859206,
"grad_norm": 0.02633056230843067,
"learning_rate": 4.065554489717105e-06,
"loss": 0.864,
"step": 275
},
{
"epoch": 0.3188447653429603,
"grad_norm": 0.026197999715805054,
"learning_rate": 4.06818180344203e-06,
"loss": 0.8492,
"step": 276
},
{
"epoch": 0.32,
"grad_norm": 0.025230281054973602,
"learning_rate": 4.070799615107415e-06,
"loss": 0.8255,
"step": 277
},
{
"epoch": 0.3211552346570397,
"grad_norm": 0.02663682959973812,
"learning_rate": 4.073407993196794e-06,
"loss": 0.8557,
"step": 278
},
{
"epoch": 0.3223104693140794,
"grad_norm": 0.025046920403838158,
"learning_rate": 4.076007005455996e-06,
"loss": 0.8433,
"step": 279
},
{
"epoch": 0.3234657039711191,
"grad_norm": 0.025554031133651733,
"learning_rate": 4.078596718903699e-06,
"loss": 0.8639,
"step": 280
},
{
"epoch": 0.32462093862815883,
"grad_norm": 0.025266235694289207,
"learning_rate": 4.0811771998418e-06,
"loss": 0.8617,
"step": 281
},
{
"epoch": 0.32577617328519853,
"grad_norm": 0.026751738041639328,
"learning_rate": 4.083748513865602e-06,
"loss": 0.8472,
"step": 282
},
{
"epoch": 0.3269314079422383,
"grad_norm": 0.026282917708158493,
"learning_rate": 4.086310725873818e-06,
"loss": 0.8518,
"step": 283
},
{
"epoch": 0.328086642599278,
"grad_norm": 0.024683093652129173,
"learning_rate": 4.088863900078397e-06,
"loss": 0.856,
"step": 284
},
{
"epoch": 0.3292418772563177,
"grad_norm": 0.025772644206881523,
"learning_rate": 4.091408100014184e-06,
"loss": 0.849,
"step": 285
},
{
"epoch": 0.3303971119133574,
"grad_norm": 0.02520694211125374,
"learning_rate": 4.093943388548406e-06,
"loss": 0.8363,
"step": 286
},
{
"epoch": 0.3315523465703971,
"grad_norm": 0.026992499828338623,
"learning_rate": 4.096469827889988e-06,
"loss": 0.859,
"step": 287
},
{
"epoch": 0.3327075812274368,
"grad_norm": 0.02580207772552967,
"learning_rate": 4.0989874795987185e-06,
"loss": 0.873,
"step": 288
},
{
"epoch": 0.3338628158844765,
"grad_norm": 0.025334378704428673,
"learning_rate": 4.101496404594247e-06,
"loss": 0.848,
"step": 289
},
{
"epoch": 0.33501805054151623,
"grad_norm": 0.027060015127062798,
"learning_rate": 4.103996663164927e-06,
"loss": 0.834,
"step": 290
},
{
"epoch": 0.33617328519855594,
"grad_norm": 0.027299024164676666,
"learning_rate": 4.106488314976513e-06,
"loss": 0.8667,
"step": 291
},
{
"epoch": 0.33732851985559564,
"grad_norm": 0.027205299586057663,
"learning_rate": 4.108971419080698e-06,
"loss": 0.8548,
"step": 292
},
{
"epoch": 0.3384837545126354,
"grad_norm": 0.025082072243094444,
"learning_rate": 4.111446033923516e-06,
"loss": 0.8418,
"step": 293
},
{
"epoch": 0.3396389891696751,
"grad_norm": 0.025831829756498337,
"learning_rate": 4.113912217353596e-06,
"loss": 0.8419,
"step": 294
},
{
"epoch": 0.3407942238267148,
"grad_norm": 0.02547086775302887,
"learning_rate": 4.1163700266302726e-06,
"loss": 0.8708,
"step": 295
},
{
"epoch": 0.3419494584837545,
"grad_norm": 0.02637997455894947,
"learning_rate": 4.118819518431564e-06,
"loss": 0.8492,
"step": 296
},
{
"epoch": 0.3431046931407942,
"grad_norm": 0.02541852556169033,
"learning_rate": 4.121260748862021e-06,
"loss": 0.8607,
"step": 297
},
{
"epoch": 0.34425992779783393,
"grad_norm": 0.026499446481466293,
"learning_rate": 4.123693773460426e-06,
"loss": 0.8543,
"step": 298
},
{
"epoch": 0.34541516245487364,
"grad_norm": 0.025943726301193237,
"learning_rate": 4.126118647207383e-06,
"loss": 0.852,
"step": 299
},
{
"epoch": 0.34657039711191334,
"grad_norm": 0.024514541029930115,
"learning_rate": 4.128535424532771e-06,
"loss": 0.8335,
"step": 300
},
{
"epoch": 0.34772563176895305,
"grad_norm": 0.026036018505692482,
"learning_rate": 4.130944159323072e-06,
"loss": 0.8373,
"step": 301
},
{
"epoch": 0.34888086642599275,
"grad_norm": 0.0261723380535841,
"learning_rate": 4.133344904928585e-06,
"loss": 0.839,
"step": 302
},
{
"epoch": 0.3500361010830325,
"grad_norm": 0.025829095393419266,
"learning_rate": 4.135737714170509e-06,
"loss": 0.8691,
"step": 303
},
{
"epoch": 0.3511913357400722,
"grad_norm": 0.026276499032974243,
"learning_rate": 4.138122639347924e-06,
"loss": 0.839,
"step": 304
},
{
"epoch": 0.3523465703971119,
"grad_norm": 0.026492850854992867,
"learning_rate": 4.140499732244644e-06,
"loss": 0.8585,
"step": 305
},
{
"epoch": 0.35350180505415163,
"grad_norm": 0.027172986418008804,
"learning_rate": 4.142869044135967e-06,
"loss": 0.8369,
"step": 306
},
{
"epoch": 0.35465703971119134,
"grad_norm": 0.026907015591859818,
"learning_rate": 4.145230625795312e-06,
"loss": 0.8578,
"step": 307
},
{
"epoch": 0.35581227436823104,
"grad_norm": 0.02687431313097477,
"learning_rate": 4.147584527500741e-06,
"loss": 0.8727,
"step": 308
},
{
"epoch": 0.35696750902527075,
"grad_norm": 0.026860354468226433,
"learning_rate": 4.149930799041391e-06,
"loss": 0.8026,
"step": 309
},
{
"epoch": 0.35812274368231045,
"grad_norm": 0.02500147931277752,
"learning_rate": 4.152269489723789e-06,
"loss": 0.8309,
"step": 310
},
{
"epoch": 0.35927797833935016,
"grad_norm": 0.026802562177181244,
"learning_rate": 4.154600648378063e-06,
"loss": 0.8524,
"step": 311
},
{
"epoch": 0.3604332129963899,
"grad_norm": 0.026412485167384148,
"learning_rate": 4.156924323364072e-06,
"loss": 0.8485,
"step": 312
},
{
"epoch": 0.3615884476534296,
"grad_norm": 0.027881808578968048,
"learning_rate": 4.159240562577414e-06,
"loss": 0.8654,
"step": 313
},
{
"epoch": 0.36274368231046933,
"grad_norm": 0.026209615170955658,
"learning_rate": 4.1615494134553584e-06,
"loss": 0.8492,
"step": 314
},
{
"epoch": 0.36389891696750903,
"grad_norm": 0.026174476370215416,
"learning_rate": 4.163850922982668e-06,
"loss": 0.818,
"step": 315
},
{
"epoch": 0.36505415162454874,
"grad_norm": 0.02707611583173275,
"learning_rate": 4.166145137697341e-06,
"loss": 0.8586,
"step": 316
},
{
"epoch": 0.36620938628158844,
"grad_norm": 0.02789073996245861,
"learning_rate": 4.1684321036962525e-06,
"loss": 0.8468,
"step": 317
},
{
"epoch": 0.36736462093862815,
"grad_norm": 0.02748328261077404,
"learning_rate": 4.170711866640721e-06,
"loss": 0.8688,
"step": 318
},
{
"epoch": 0.36851985559566786,
"grad_norm": 0.029765894636511803,
"learning_rate": 4.172984471761969e-06,
"loss": 0.8521,
"step": 319
},
{
"epoch": 0.36967509025270756,
"grad_norm": 0.025119414553046227,
"learning_rate": 4.17524996386651e-06,
"loss": 0.829,
"step": 320
},
{
"epoch": 0.37083032490974727,
"grad_norm": 0.028743257746100426,
"learning_rate": 4.177508387341454e-06,
"loss": 0.8676,
"step": 321
},
{
"epoch": 0.371985559566787,
"grad_norm": 0.02732737548649311,
"learning_rate": 4.179759786159719e-06,
"loss": 0.8516,
"step": 322
},
{
"epoch": 0.37314079422382673,
"grad_norm": 0.026275500655174255,
"learning_rate": 4.182004203885172e-06,
"loss": 0.8606,
"step": 323
},
{
"epoch": 0.37429602888086644,
"grad_norm": 0.028016693890094757,
"learning_rate": 4.1842416836776876e-06,
"loss": 0.8272,
"step": 324
},
{
"epoch": 0.37545126353790614,
"grad_norm": 0.026512511074543,
"learning_rate": 4.1864722682981245e-06,
"loss": 0.8438,
"step": 325
},
{
"epoch": 0.37660649819494585,
"grad_norm": 0.026299525052309036,
"learning_rate": 4.188696000113232e-06,
"loss": 0.8485,
"step": 326
},
{
"epoch": 0.37776173285198555,
"grad_norm": 0.02613169699907303,
"learning_rate": 4.190912921100477e-06,
"loss": 0.8184,
"step": 327
},
{
"epoch": 0.37891696750902526,
"grad_norm": 0.02610696479678154,
"learning_rate": 4.193123072852799e-06,
"loss": 0.8609,
"step": 328
},
{
"epoch": 0.38007220216606497,
"grad_norm": 0.026195967569947243,
"learning_rate": 4.1953264965832905e-06,
"loss": 0.8302,
"step": 329
},
{
"epoch": 0.38122743682310467,
"grad_norm": 0.026481349021196365,
"learning_rate": 4.197523233129813e-06,
"loss": 0.8491,
"step": 330
},
{
"epoch": 0.38238267148014443,
"grad_norm": 0.024858448654413223,
"learning_rate": 4.199713322959531e-06,
"loss": 0.8404,
"step": 331
},
{
"epoch": 0.38353790613718414,
"grad_norm": 0.026140978559851646,
"learning_rate": 4.201896806173394e-06,
"loss": 0.8558,
"step": 332
},
{
"epoch": 0.38469314079422384,
"grad_norm": 0.026811236515641212,
"learning_rate": 4.204073722510534e-06,
"loss": 0.8747,
"step": 333
},
{
"epoch": 0.38584837545126355,
"grad_norm": 0.0256299190223217,
"learning_rate": 4.206244111352608e-06,
"loss": 0.8521,
"step": 334
},
{
"epoch": 0.38700361010830325,
"grad_norm": 0.026677099987864494,
"learning_rate": 4.208408011728075e-06,
"loss": 0.8548,
"step": 335
},
{
"epoch": 0.38815884476534296,
"grad_norm": 0.026028303429484367,
"learning_rate": 4.210565462316407e-06,
"loss": 0.8474,
"step": 336
},
{
"epoch": 0.38931407942238266,
"grad_norm": 0.02715076506137848,
"learning_rate": 4.212716501452232e-06,
"loss": 0.8318,
"step": 337
},
{
"epoch": 0.39046931407942237,
"grad_norm": 0.02765047550201416,
"learning_rate": 4.214861167129425e-06,
"loss": 0.8558,
"step": 338
},
{
"epoch": 0.3916245487364621,
"grad_norm": 0.025798538699746132,
"learning_rate": 4.216999497005137e-06,
"loss": 0.8436,
"step": 339
},
{
"epoch": 0.3927797833935018,
"grad_norm": 0.028038429096341133,
"learning_rate": 4.219131528403759e-06,
"loss": 0.845,
"step": 340
},
{
"epoch": 0.39393501805054154,
"grad_norm": 0.026474367827177048,
"learning_rate": 4.22125729832083e-06,
"loss": 0.8411,
"step": 341
},
{
"epoch": 0.39509025270758125,
"grad_norm": 0.025686634704470634,
"learning_rate": 4.223376843426892e-06,
"loss": 0.8485,
"step": 342
},
{
"epoch": 0.39624548736462095,
"grad_norm": 0.026812469586730003,
"learning_rate": 4.225490200071284e-06,
"loss": 0.8571,
"step": 343
},
{
"epoch": 0.39740072202166066,
"grad_norm": 0.02608347311615944,
"learning_rate": 4.227597404285883e-06,
"loss": 0.8426,
"step": 344
},
{
"epoch": 0.39855595667870036,
"grad_norm": 0.02631702460348606,
"learning_rate": 4.229698491788791e-06,
"loss": 0.876,
"step": 345
},
{
"epoch": 0.39971119133574007,
"grad_norm": 0.0272465068846941,
"learning_rate": 4.231793497987962e-06,
"loss": 0.861,
"step": 346
},
{
"epoch": 0.4008664259927798,
"grad_norm": 0.02596902847290039,
"learning_rate": 4.233882457984791e-06,
"loss": 0.8645,
"step": 347
},
{
"epoch": 0.4020216606498195,
"grad_norm": 0.02580340951681137,
"learning_rate": 4.235965406577636e-06,
"loss": 0.8447,
"step": 348
},
{
"epoch": 0.4031768953068592,
"grad_norm": 0.027888286858797073,
"learning_rate": 4.2380423782653e-06,
"loss": 0.8425,
"step": 349
},
{
"epoch": 0.4043321299638989,
"grad_norm": 0.025501729920506477,
"learning_rate": 4.2401134072504595e-06,
"loss": 0.8307,
"step": 350
},
{
"epoch": 0.40548736462093865,
"grad_norm": 0.026491057127714157,
"learning_rate": 4.24217852744304e-06,
"loss": 0.8237,
"step": 351
},
{
"epoch": 0.40664259927797836,
"grad_norm": 0.027327535673975945,
"learning_rate": 4.244237772463552e-06,
"loss": 0.8605,
"step": 352
},
{
"epoch": 0.40779783393501806,
"grad_norm": 0.026969779282808304,
"learning_rate": 4.246291175646372e-06,
"loss": 0.8357,
"step": 353
},
{
"epoch": 0.40895306859205777,
"grad_norm": 0.02584182471036911,
"learning_rate": 4.2483387700429804e-06,
"loss": 0.8348,
"step": 354
},
{
"epoch": 0.41010830324909747,
"grad_norm": 0.025704173371195793,
"learning_rate": 4.2503805884251575e-06,
"loss": 0.8392,
"step": 355
},
{
"epoch": 0.4112635379061372,
"grad_norm": 0.026744280010461807,
"learning_rate": 4.252416663288126e-06,
"loss": 0.8405,
"step": 356
},
{
"epoch": 0.4124187725631769,
"grad_norm": 0.02655099704861641,
"learning_rate": 4.2544470268536555e-06,
"loss": 0.8374,
"step": 357
},
{
"epoch": 0.4135740072202166,
"grad_norm": 0.02521936595439911,
"learning_rate": 4.2564717110731246e-06,
"loss": 0.8442,
"step": 358
},
{
"epoch": 0.4147292418772563,
"grad_norm": 0.02607394941151142,
"learning_rate": 4.258490747630532e-06,
"loss": 0.8273,
"step": 359
},
{
"epoch": 0.41588447653429605,
"grad_norm": 0.027000917121767998,
"learning_rate": 4.26050416794548e-06,
"loss": 0.8664,
"step": 360
},
{
"epoch": 0.41703971119133576,
"grad_norm": 0.026261812075972557,
"learning_rate": 4.262512003176097e-06,
"loss": 0.8351,
"step": 361
},
{
"epoch": 0.41819494584837547,
"grad_norm": 0.025927625596523285,
"learning_rate": 4.264514284221944e-06,
"loss": 0.8419,
"step": 362
},
{
"epoch": 0.41935018050541517,
"grad_norm": 0.02672671154141426,
"learning_rate": 4.2665110417268545e-06,
"loss": 0.84,
"step": 363
},
{
"epoch": 0.4205054151624549,
"grad_norm": 0.02695775218307972,
"learning_rate": 4.26850230608176e-06,
"loss": 0.8483,
"step": 364
},
{
"epoch": 0.4216606498194946,
"grad_norm": 0.026013897731900215,
"learning_rate": 4.270488107427459e-06,
"loss": 0.8371,
"step": 365
},
{
"epoch": 0.4228158844765343,
"grad_norm": 0.026113096624612808,
"learning_rate": 4.272468475657351e-06,
"loss": 0.8448,
"step": 366
},
{
"epoch": 0.423971119133574,
"grad_norm": 0.02611883170902729,
"learning_rate": 4.27444344042015e-06,
"loss": 0.8522,
"step": 367
},
{
"epoch": 0.4251263537906137,
"grad_norm": 0.02588835172355175,
"learning_rate": 4.27641303112253e-06,
"loss": 0.8348,
"step": 368
},
{
"epoch": 0.4262815884476534,
"grad_norm": 0.026020022109150887,
"learning_rate": 4.278377276931768e-06,
"loss": 0.8376,
"step": 369
},
{
"epoch": 0.42743682310469316,
"grad_norm": 0.026923442259430885,
"learning_rate": 4.280336206778326e-06,
"loss": 0.8435,
"step": 370
},
{
"epoch": 0.42859205776173287,
"grad_norm": 0.026913795620203018,
"learning_rate": 4.2822898493584105e-06,
"loss": 0.8374,
"step": 371
},
{
"epoch": 0.4297472924187726,
"grad_norm": 0.025143541395664215,
"learning_rate": 4.2842382331364965e-06,
"loss": 0.8169,
"step": 372
},
{
"epoch": 0.4309025270758123,
"grad_norm": 0.027908481657505035,
"learning_rate": 4.2861813863478135e-06,
"loss": 0.8561,
"step": 373
},
{
"epoch": 0.432057761732852,
"grad_norm": 0.026535294950008392,
"learning_rate": 4.288119337000801e-06,
"loss": 0.8299,
"step": 374
},
{
"epoch": 0.4332129963898917,
"grad_norm": 0.027421629056334496,
"learning_rate": 4.290052112879532e-06,
"loss": 0.8273,
"step": 375
},
{
"epoch": 0.4343682310469314,
"grad_norm": 0.02671075239777565,
"learning_rate": 4.291979741546102e-06,
"loss": 0.8215,
"step": 376
},
{
"epoch": 0.4355234657039711,
"grad_norm": 0.025190116837620735,
"learning_rate": 4.293902250342989e-06,
"loss": 0.871,
"step": 377
},
{
"epoch": 0.4366787003610108,
"grad_norm": 0.026498563587665558,
"learning_rate": 4.295819666395376e-06,
"loss": 0.8582,
"step": 378
},
{
"epoch": 0.43783393501805057,
"grad_norm": 0.02722996659576893,
"learning_rate": 4.297732016613455e-06,
"loss": 0.8288,
"step": 379
},
{
"epoch": 0.4389891696750903,
"grad_norm": 0.028074050322175026,
"learning_rate": 4.299639327694684e-06,
"loss": 0.8574,
"step": 380
},
{
"epoch": 0.44014440433213,
"grad_norm": 0.026562221348285675,
"learning_rate": 4.301541626126033e-06,
"loss": 0.8337,
"step": 381
},
{
"epoch": 0.4412996389891697,
"grad_norm": 0.025497550144791603,
"learning_rate": 4.303438938186182e-06,
"loss": 0.8388,
"step": 382
},
{
"epoch": 0.4424548736462094,
"grad_norm": 0.028150785714387894,
"learning_rate": 4.305331289947705e-06,
"loss": 0.8381,
"step": 383
},
{
"epoch": 0.4436101083032491,
"grad_norm": 0.024889659136533737,
"learning_rate": 4.307218707279219e-06,
"loss": 0.8297,
"step": 384
},
{
"epoch": 0.4447653429602888,
"grad_norm": 0.02572575956583023,
"learning_rate": 4.309101215847502e-06,
"loss": 0.8464,
"step": 385
},
{
"epoch": 0.4459205776173285,
"grad_norm": 0.028156662359833717,
"learning_rate": 4.310978841119592e-06,
"loss": 0.8307,
"step": 386
},
{
"epoch": 0.4470758122743682,
"grad_norm": 0.02584012597799301,
"learning_rate": 4.312851608364853e-06,
"loss": 0.8342,
"step": 387
},
{
"epoch": 0.4482310469314079,
"grad_norm": 0.02681022137403488,
"learning_rate": 4.3147195426570124e-06,
"loss": 0.839,
"step": 388
},
{
"epoch": 0.4493862815884477,
"grad_norm": 0.028267353773117065,
"learning_rate": 4.31658266887618e-06,
"loss": 0.8602,
"step": 389
},
{
"epoch": 0.4505415162454874,
"grad_norm": 0.026196565479040146,
"learning_rate": 4.318441011710832e-06,
"loss": 0.8374,
"step": 390
},
{
"epoch": 0.4516967509025271,
"grad_norm": 0.028164999559521675,
"learning_rate": 4.3202945956597785e-06,
"loss": 0.8458,
"step": 391
},
{
"epoch": 0.4528519855595668,
"grad_norm": 0.026657233014702797,
"learning_rate": 4.322143445034095e-06,
"loss": 0.8238,
"step": 392
},
{
"epoch": 0.4540072202166065,
"grad_norm": 0.025035889819264412,
"learning_rate": 4.323987583959045e-06,
"loss": 0.8371,
"step": 393
},
{
"epoch": 0.4551624548736462,
"grad_norm": 0.02667805179953575,
"learning_rate": 4.325827036375958e-06,
"loss": 0.8201,
"step": 394
},
{
"epoch": 0.4563176895306859,
"grad_norm": 0.028383539989590645,
"learning_rate": 4.327661826044101e-06,
"loss": 0.8385,
"step": 395
},
{
"epoch": 0.4574729241877256,
"grad_norm": 0.025015531107783318,
"learning_rate": 4.329491976542521e-06,
"loss": 0.8013,
"step": 396
},
{
"epoch": 0.4586281588447653,
"grad_norm": 0.0283355675637722,
"learning_rate": 4.3313175112718595e-06,
"loss": 0.8228,
"step": 397
},
{
"epoch": 0.459783393501805,
"grad_norm": 0.026209138333797455,
"learning_rate": 4.333138453456147e-06,
"loss": 0.8336,
"step": 398
},
{
"epoch": 0.4609386281588448,
"grad_norm": 0.024707714095711708,
"learning_rate": 4.334954826144581e-06,
"loss": 0.8417,
"step": 399
},
{
"epoch": 0.4620938628158845,
"grad_norm": 0.02748985029757023,
"learning_rate": 4.336766652213271e-06,
"loss": 0.843,
"step": 400
},
{
"epoch": 0.4632490974729242,
"grad_norm": 0.02513727732002735,
"learning_rate": 4.3385739543669715e-06,
"loss": 0.8301,
"step": 401
},
{
"epoch": 0.4644043321299639,
"grad_norm": 0.028213754296302795,
"learning_rate": 4.340376755140784e-06,
"loss": 0.8497,
"step": 402
},
{
"epoch": 0.4655595667870036,
"grad_norm": 0.02582547254860401,
"learning_rate": 4.34217507690185e-06,
"loss": 0.8252,
"step": 403
},
{
"epoch": 0.4667148014440433,
"grad_norm": 0.028116989880800247,
"learning_rate": 4.343968941851009e-06,
"loss": 0.8312,
"step": 404
},
{
"epoch": 0.467870036101083,
"grad_norm": 0.026428736746311188,
"learning_rate": 4.345758372024448e-06,
"loss": 0.8177,
"step": 405
},
{
"epoch": 0.4690252707581227,
"grad_norm": 0.025875557214021683,
"learning_rate": 4.347543389295324e-06,
"loss": 0.8258,
"step": 406
},
{
"epoch": 0.47018050541516243,
"grad_norm": 0.025924712419509888,
"learning_rate": 4.3493240153753665e-06,
"loss": 0.8379,
"step": 407
},
{
"epoch": 0.4713357400722022,
"grad_norm": 0.025729961693286896,
"learning_rate": 4.3511002718164665e-06,
"loss": 0.8396,
"step": 408
},
{
"epoch": 0.4724909747292419,
"grad_norm": 0.026374271139502525,
"learning_rate": 4.352872180012237e-06,
"loss": 0.8381,
"step": 409
},
{
"epoch": 0.4736462093862816,
"grad_norm": 0.028471384197473526,
"learning_rate": 4.35463976119956e-06,
"loss": 0.8333,
"step": 410
},
{
"epoch": 0.4748014440433213,
"grad_norm": 0.025666672736406326,
"learning_rate": 4.356403036460116e-06,
"loss": 0.8191,
"step": 411
},
{
"epoch": 0.475956678700361,
"grad_norm": 0.026298929005861282,
"learning_rate": 4.358162026721892e-06,
"loss": 0.8225,
"step": 412
},
{
"epoch": 0.4771119133574007,
"grad_norm": 0.027312815189361572,
"learning_rate": 4.359916752760669e-06,
"loss": 0.804,
"step": 413
},
{
"epoch": 0.4782671480144404,
"grad_norm": 0.02659662254154682,
"learning_rate": 4.361667235201499e-06,
"loss": 0.8286,
"step": 414
},
{
"epoch": 0.47942238267148013,
"grad_norm": 0.02585594914853573,
"learning_rate": 4.363413494520154e-06,
"loss": 0.8229,
"step": 415
},
{
"epoch": 0.48057761732851983,
"grad_norm": 0.027580194175243378,
"learning_rate": 4.365155551044572e-06,
"loss": 0.8397,
"step": 416
},
{
"epoch": 0.48173285198555954,
"grad_norm": 0.025151105597615242,
"learning_rate": 4.366893424956263e-06,
"loss": 0.8356,
"step": 417
},
{
"epoch": 0.4828880866425993,
"grad_norm": 0.025705143809318542,
"learning_rate": 4.368627136291726e-06,
"loss": 0.8342,
"step": 418
},
{
"epoch": 0.484043321299639,
"grad_norm": 0.027614813297986984,
"learning_rate": 4.370356704943825e-06,
"loss": 0.8481,
"step": 419
},
{
"epoch": 0.4851985559566787,
"grad_norm": 0.026671946048736572,
"learning_rate": 4.372082150663167e-06,
"loss": 0.8375,
"step": 420
},
{
"epoch": 0.4863537906137184,
"grad_norm": 0.026453910395503044,
"learning_rate": 4.3738034930594475e-06,
"loss": 0.8259,
"step": 421
},
{
"epoch": 0.4875090252707581,
"grad_norm": 0.027380600571632385,
"learning_rate": 4.3755207516027904e-06,
"loss": 0.8394,
"step": 422
},
{
"epoch": 0.48866425992779783,
"grad_norm": 0.02569480426609516,
"learning_rate": 4.3772339456250705e-06,
"loss": 0.82,
"step": 423
},
{
"epoch": 0.48981949458483753,
"grad_norm": 0.028223766013979912,
"learning_rate": 4.3789430943212215e-06,
"loss": 0.8223,
"step": 424
},
{
"epoch": 0.49097472924187724,
"grad_norm": 0.025967473164200783,
"learning_rate": 4.38064821675052e-06,
"loss": 0.8316,
"step": 425
},
{
"epoch": 0.49212996389891694,
"grad_norm": 0.026815980672836304,
"learning_rate": 4.382349331837865e-06,
"loss": 0.8497,
"step": 426
},
{
"epoch": 0.4932851985559567,
"grad_norm": 0.026486823335289955,
"learning_rate": 4.38404645837504e-06,
"loss": 0.8224,
"step": 427
},
{
"epoch": 0.4944404332129964,
"grad_norm": 0.026307394728064537,
"learning_rate": 4.385739615021954e-06,
"loss": 0.8055,
"step": 428
},
{
"epoch": 0.4955956678700361,
"grad_norm": 0.02565447799861431,
"learning_rate": 4.387428820307874e-06,
"loss": 0.85,
"step": 429
},
{
"epoch": 0.4967509025270758,
"grad_norm": 0.026436088606715202,
"learning_rate": 4.389114092632645e-06,
"loss": 0.8343,
"step": 430
},
{
"epoch": 0.4979061371841155,
"grad_norm": 0.028431309387087822,
"learning_rate": 4.390795450267887e-06,
"loss": 0.8457,
"step": 431
},
{
"epoch": 0.49906137184115523,
"grad_norm": 0.026016119867563248,
"learning_rate": 4.392472911358188e-06,
"loss": 0.8307,
"step": 432
},
{
"epoch": 0.500216606498195,
"grad_norm": 0.02676302008330822,
"learning_rate": 4.394146493922276e-06,
"loss": 0.8565,
"step": 433
},
{
"epoch": 0.5013718411552347,
"grad_norm": 0.026580410078167915,
"learning_rate": 4.395816215854185e-06,
"loss": 0.8364,
"step": 434
},
{
"epoch": 0.5025270758122744,
"grad_norm": 0.02553938329219818,
"learning_rate": 4.397482094924396e-06,
"loss": 0.8266,
"step": 435
},
{
"epoch": 0.5036823104693141,
"grad_norm": 0.02656836435198784,
"learning_rate": 4.3991441487809775e-06,
"loss": 0.8181,
"step": 436
},
{
"epoch": 0.5048375451263538,
"grad_norm": 0.025670040398836136,
"learning_rate": 4.400802394950703e-06,
"loss": 0.7895,
"step": 437
},
{
"epoch": 0.5059927797833935,
"grad_norm": 0.02619147300720215,
"learning_rate": 4.402456850840167e-06,
"loss": 0.8364,
"step": 438
},
{
"epoch": 0.5071480144404332,
"grad_norm": 0.025901198387145996,
"learning_rate": 4.4041075337368695e-06,
"loss": 0.8449,
"step": 439
},
{
"epoch": 0.5083032490974729,
"grad_norm": 0.02623150125145912,
"learning_rate": 4.405754460810312e-06,
"loss": 0.8422,
"step": 440
},
{
"epoch": 0.5094584837545126,
"grad_norm": 0.025403983891010284,
"learning_rate": 4.407397649113065e-06,
"loss": 0.851,
"step": 441
},
{
"epoch": 0.5106137184115523,
"grad_norm": 0.027808185666799545,
"learning_rate": 4.40903711558182e-06,
"loss": 0.8238,
"step": 442
},
{
"epoch": 0.511768953068592,
"grad_norm": 0.025581343099474907,
"learning_rate": 4.41067287703845e-06,
"loss": 0.848,
"step": 443
},
{
"epoch": 0.5129241877256318,
"grad_norm": 0.026919787749648094,
"learning_rate": 4.4123049501910335e-06,
"loss": 0.8171,
"step": 444
},
{
"epoch": 0.5140794223826715,
"grad_norm": 0.025836626067757607,
"learning_rate": 4.4139333516348865e-06,
"loss": 0.8338,
"step": 445
},
{
"epoch": 0.5152346570397112,
"grad_norm": 0.026238281279802322,
"learning_rate": 4.415558097853571e-06,
"loss": 0.8164,
"step": 446
},
{
"epoch": 0.5163898916967509,
"grad_norm": 0.02682877704501152,
"learning_rate": 4.4171792052198945e-06,
"loss": 0.8414,
"step": 447
},
{
"epoch": 0.5175451263537906,
"grad_norm": 0.02541991136968136,
"learning_rate": 4.418796689996907e-06,
"loss": 0.8043,
"step": 448
},
{
"epoch": 0.5187003610108303,
"grad_norm": 0.026446549221873283,
"learning_rate": 4.420410568338873e-06,
"loss": 0.8499,
"step": 449
},
{
"epoch": 0.51985559566787,
"grad_norm": 0.02652100659906864,
"learning_rate": 4.42202085629224e-06,
"loss": 0.8331,
"step": 450
},
{
"epoch": 0.5210108303249098,
"grad_norm": 0.025452058762311935,
"learning_rate": 4.423627569796601e-06,
"loss": 0.8265,
"step": 451
},
{
"epoch": 0.5221660649819495,
"grad_norm": 0.025869838893413544,
"learning_rate": 4.425230724685638e-06,
"loss": 0.8276,
"step": 452
},
{
"epoch": 0.5233212996389892,
"grad_norm": 0.026051780208945274,
"learning_rate": 4.426830336688054e-06,
"loss": 0.8575,
"step": 453
},
{
"epoch": 0.5244765342960289,
"grad_norm": 0.02754388563334942,
"learning_rate": 4.428426421428507e-06,
"loss": 0.8255,
"step": 454
},
{
"epoch": 0.5256317689530686,
"grad_norm": 0.02607525885105133,
"learning_rate": 4.430018994428521e-06,
"loss": 0.8397,
"step": 455
},
{
"epoch": 0.5267870036101083,
"grad_norm": 0.026636334136128426,
"learning_rate": 4.431608071107392e-06,
"loss": 0.8377,
"step": 456
},
{
"epoch": 0.527942238267148,
"grad_norm": 0.026849418878555298,
"learning_rate": 4.433193666783084e-06,
"loss": 0.8257,
"step": 457
},
{
"epoch": 0.5290974729241877,
"grad_norm": 0.02537579834461212,
"learning_rate": 4.434775796673116e-06,
"loss": 0.8455,
"step": 458
},
{
"epoch": 0.5302527075812274,
"grad_norm": 0.025962024927139282,
"learning_rate": 4.4363544758954355e-06,
"loss": 0.8263,
"step": 459
},
{
"epoch": 0.5314079422382672,
"grad_norm": 0.027943387627601624,
"learning_rate": 4.437929719469291e-06,
"loss": 0.8546,
"step": 460
},
{
"epoch": 0.5325631768953069,
"grad_norm": 0.02550574764609337,
"learning_rate": 4.439501542316081e-06,
"loss": 0.8367,
"step": 461
},
{
"epoch": 0.5337184115523466,
"grad_norm": 0.030560094863176346,
"learning_rate": 4.44106995926021e-06,
"loss": 0.8443,
"step": 462
},
{
"epoch": 0.5348736462093863,
"grad_norm": 0.025380682200193405,
"learning_rate": 4.442634985029922e-06,
"loss": 0.8341,
"step": 463
},
{
"epoch": 0.536028880866426,
"grad_norm": 0.027356211096048355,
"learning_rate": 4.444196634258136e-06,
"loss": 0.8427,
"step": 464
},
{
"epoch": 0.5371841155234657,
"grad_norm": 0.027884816750884056,
"learning_rate": 4.445754921483257e-06,
"loss": 0.8441,
"step": 465
},
{
"epoch": 0.5383393501805054,
"grad_norm": 0.025398530066013336,
"learning_rate": 4.44730986115e-06,
"loss": 0.8342,
"step": 466
},
{
"epoch": 0.5394945848375451,
"grad_norm": 0.027301618829369545,
"learning_rate": 4.448861467610187e-06,
"loss": 0.8453,
"step": 467
},
{
"epoch": 0.5406498194945848,
"grad_norm": 0.025283178314566612,
"learning_rate": 4.450409755123541e-06,
"loss": 0.8468,
"step": 468
},
{
"epoch": 0.5418050541516245,
"grad_norm": 0.025396686047315598,
"learning_rate": 4.4519547378584725e-06,
"loss": 0.8352,
"step": 469
},
{
"epoch": 0.5429602888086643,
"grad_norm": 0.026122555136680603,
"learning_rate": 4.453496429892863e-06,
"loss": 0.8255,
"step": 470
},
{
"epoch": 0.544115523465704,
"grad_norm": 0.026710988953709602,
"learning_rate": 4.455034845214828e-06,
"loss": 0.8033,
"step": 471
},
{
"epoch": 0.5452707581227437,
"grad_norm": 0.027530129998922348,
"learning_rate": 4.45656999772348e-06,
"loss": 0.827,
"step": 472
},
{
"epoch": 0.5464259927797834,
"grad_norm": 0.02589859999716282,
"learning_rate": 4.458101901229686e-06,
"loss": 0.8585,
"step": 473
},
{
"epoch": 0.5475812274368231,
"grad_norm": 0.026990870013833046,
"learning_rate": 4.459630569456809e-06,
"loss": 0.8228,
"step": 474
},
{
"epoch": 0.5487364620938628,
"grad_norm": 0.026618408039212227,
"learning_rate": 4.461156016041445e-06,
"loss": 0.8308,
"step": 475
},
{
"epoch": 0.5498916967509025,
"grad_norm": 0.026187585666775703,
"learning_rate": 4.4626782545341565e-06,
"loss": 0.841,
"step": 476
},
{
"epoch": 0.5510469314079423,
"grad_norm": 0.024839429184794426,
"learning_rate": 4.4641972984001906e-06,
"loss": 0.8265,
"step": 477
},
{
"epoch": 0.552202166064982,
"grad_norm": 0.02818799950182438,
"learning_rate": 4.465713161020199e-06,
"loss": 0.8231,
"step": 478
},
{
"epoch": 0.5533574007220217,
"grad_norm": 0.02686317451298237,
"learning_rate": 4.46722585569094e-06,
"loss": 0.8266,
"step": 479
},
{
"epoch": 0.5545126353790614,
"grad_norm": 0.02528228797018528,
"learning_rate": 4.468735395625979e-06,
"loss": 0.8296,
"step": 480
},
{
"epoch": 0.5556678700361011,
"grad_norm": 0.02689657174050808,
"learning_rate": 4.470241793956387e-06,
"loss": 0.8345,
"step": 481
},
{
"epoch": 0.5568231046931408,
"grad_norm": 0.028286360204219818,
"learning_rate": 4.471745063731417e-06,
"loss": 0.8371,
"step": 482
},
{
"epoch": 0.5579783393501805,
"grad_norm": 0.0263553187251091,
"learning_rate": 4.473245217919187e-06,
"loss": 0.8183,
"step": 483
},
{
"epoch": 0.5591335740072202,
"grad_norm": 0.027919506654143333,
"learning_rate": 4.474742269407355e-06,
"loss": 0.8317,
"step": 484
},
{
"epoch": 0.5602888086642599,
"grad_norm": 0.02882731519639492,
"learning_rate": 4.476236231003773e-06,
"loss": 0.8414,
"step": 485
},
{
"epoch": 0.5614440433212996,
"grad_norm": 0.02561136521399021,
"learning_rate": 4.477727115437156e-06,
"loss": 0.8004,
"step": 486
},
{
"epoch": 0.5625992779783393,
"grad_norm": 0.02664412185549736,
"learning_rate": 4.479214935357724e-06,
"loss": 0.7967,
"step": 487
},
{
"epoch": 0.563754512635379,
"grad_norm": 0.028001073747873306,
"learning_rate": 4.480699703337852e-06,
"loss": 0.841,
"step": 488
},
{
"epoch": 0.5649097472924187,
"grad_norm": 0.026740211993455887,
"learning_rate": 4.4821814318727016e-06,
"loss": 0.8369,
"step": 489
},
{
"epoch": 0.5660649819494585,
"grad_norm": 0.024850796908140182,
"learning_rate": 4.4836601333808566e-06,
"loss": 0.8405,
"step": 490
},
{
"epoch": 0.5672202166064982,
"grad_norm": 0.025452926754951477,
"learning_rate": 4.485135820204948e-06,
"loss": 0.8106,
"step": 491
},
{
"epoch": 0.568375451263538,
"grad_norm": 0.026447713375091553,
"learning_rate": 4.4866085046122675e-06,
"loss": 0.7992,
"step": 492
},
{
"epoch": 0.5695306859205777,
"grad_norm": 0.027519937604665756,
"learning_rate": 4.488078198795384e-06,
"loss": 0.8227,
"step": 493
},
{
"epoch": 0.5706859205776174,
"grad_norm": 0.026447774842381477,
"learning_rate": 4.4895449148727455e-06,
"loss": 0.8377,
"step": 494
},
{
"epoch": 0.5718411552346571,
"grad_norm": 0.02749788761138916,
"learning_rate": 4.491008664889282e-06,
"loss": 0.789,
"step": 495
},
{
"epoch": 0.5729963898916968,
"grad_norm": 0.025579964742064476,
"learning_rate": 4.492469460816997e-06,
"loss": 0.8288,
"step": 496
},
{
"epoch": 0.5741516245487365,
"grad_norm": 0.027195453643798828,
"learning_rate": 4.493927314555554e-06,
"loss": 0.8194,
"step": 497
},
{
"epoch": 0.5753068592057762,
"grad_norm": 0.027325714007019997,
"learning_rate": 4.495382237932863e-06,
"loss": 0.8136,
"step": 498
},
{
"epoch": 0.5764620938628159,
"grad_norm": 0.026480497792363167,
"learning_rate": 4.496834242705651e-06,
"loss": 0.8239,
"step": 499
},
{
"epoch": 0.5776173285198556,
"grad_norm": 0.026893138885498047,
"learning_rate": 4.498283340560032e-06,
"loss": 0.8234,
"step": 500
},
{
"epoch": 0.5787725631768953,
"grad_norm": 0.02725539728999138,
"learning_rate": 4.4997295431120764e-06,
"loss": 0.8167,
"step": 501
},
{
"epoch": 0.579927797833935,
"grad_norm": 0.027873387560248375,
"learning_rate": 4.5011728619083665e-06,
"loss": 0.8012,
"step": 502
},
{
"epoch": 0.5810830324909747,
"grad_norm": 0.027902130037546158,
"learning_rate": 4.502613308426546e-06,
"loss": 0.8248,
"step": 503
},
{
"epoch": 0.5822382671480144,
"grad_norm": 0.027641048654913902,
"learning_rate": 4.504050894075876e-06,
"loss": 0.8298,
"step": 504
},
{
"epoch": 0.5833935018050541,
"grad_norm": 0.026641003787517548,
"learning_rate": 4.5054856301977696e-06,
"loss": 0.8228,
"step": 505
},
{
"epoch": 0.5845487364620938,
"grad_norm": 0.028835156932473183,
"learning_rate": 4.506917528066332e-06,
"loss": 0.8222,
"step": 506
},
{
"epoch": 0.5857039711191335,
"grad_norm": 0.02770557627081871,
"learning_rate": 4.5083465988888945e-06,
"loss": 0.8455,
"step": 507
},
{
"epoch": 0.5868592057761732,
"grad_norm": 0.0266465675085783,
"learning_rate": 4.509772853806533e-06,
"loss": 0.8269,
"step": 508
},
{
"epoch": 0.588014440433213,
"grad_norm": 0.029550906270742416,
"learning_rate": 4.511196303894598e-06,
"loss": 0.817,
"step": 509
},
{
"epoch": 0.5891696750902528,
"grad_norm": 0.025802530348300934,
"learning_rate": 4.512616960163228e-06,
"loss": 0.8352,
"step": 510
},
{
"epoch": 0.5903249097472925,
"grad_norm": 0.02622455172240734,
"learning_rate": 4.514034833557855e-06,
"loss": 0.8209,
"step": 511
},
{
"epoch": 0.5914801444043322,
"grad_norm": 0.028669588267803192,
"learning_rate": 4.515449934959719e-06,
"loss": 0.8199,
"step": 512
},
{
"epoch": 0.5926353790613719,
"grad_norm": 0.02638787031173706,
"learning_rate": 4.516862275186361e-06,
"loss": 0.8396,
"step": 513
},
{
"epoch": 0.5937906137184116,
"grad_norm": 0.025254135951399803,
"learning_rate": 4.518271864992127e-06,
"loss": 0.8314,
"step": 514
},
{
"epoch": 0.5949458483754513,
"grad_norm": 0.026823000982403755,
"learning_rate": 4.519678715068652e-06,
"loss": 0.8063,
"step": 515
},
{
"epoch": 0.596101083032491,
"grad_norm": 0.029466571286320686,
"learning_rate": 4.521082836045353e-06,
"loss": 0.8279,
"step": 516
},
{
"epoch": 0.5972563176895307,
"grad_norm": 0.02610476315021515,
"learning_rate": 4.5224842384899045e-06,
"loss": 0.8217,
"step": 517
},
{
"epoch": 0.5984115523465704,
"grad_norm": 0.027522264048457146,
"learning_rate": 4.523882932908722e-06,
"loss": 0.8465,
"step": 518
},
{
"epoch": 0.5995667870036101,
"grad_norm": 0.025692617520689964,
"learning_rate": 4.52527892974743e-06,
"loss": 0.8448,
"step": 519
},
{
"epoch": 0.6007220216606498,
"grad_norm": 0.027035709470510483,
"learning_rate": 4.526672239391333e-06,
"loss": 0.8216,
"step": 520
},
{
"epoch": 0.6018772563176895,
"grad_norm": 0.026100315153598785,
"learning_rate": 4.528062872165875e-06,
"loss": 0.8007,
"step": 521
},
{
"epoch": 0.6030324909747292,
"grad_norm": 0.025961345061659813,
"learning_rate": 4.529450838337104e-06,
"loss": 0.8185,
"step": 522
},
{
"epoch": 0.6041877256317689,
"grad_norm": 0.02683369629085064,
"learning_rate": 4.5308361481121244e-06,
"loss": 0.8377,
"step": 523
},
{
"epoch": 0.6053429602888086,
"grad_norm": 0.02755955420434475,
"learning_rate": 4.532218811639545e-06,
"loss": 0.8197,
"step": 524
},
{
"epoch": 0.6064981949458483,
"grad_norm": 0.02744685672223568,
"learning_rate": 4.533598839009929e-06,
"loss": 0.8146,
"step": 525
},
{
"epoch": 0.607653429602888,
"grad_norm": 0.026420993730425835,
"learning_rate": 4.5349762402562325e-06,
"loss": 0.8324,
"step": 526
},
{
"epoch": 0.6088086642599277,
"grad_norm": 0.026225924491882324,
"learning_rate": 4.5363510253542444e-06,
"loss": 0.8299,
"step": 527
},
{
"epoch": 0.6099638989169676,
"grad_norm": 0.025415120646357536,
"learning_rate": 4.537723204223021e-06,
"loss": 0.8373,
"step": 528
},
{
"epoch": 0.6111191335740073,
"grad_norm": 0.026079954579472542,
"learning_rate": 4.53909278672531e-06,
"loss": 0.8064,
"step": 529
},
{
"epoch": 0.612274368231047,
"grad_norm": 0.027824481949210167,
"learning_rate": 4.540459782667983e-06,
"loss": 0.834,
"step": 530
},
{
"epoch": 0.6134296028880867,
"grad_norm": 0.02630164660513401,
"learning_rate": 4.541824201802449e-06,
"loss": 0.8341,
"step": 531
},
{
"epoch": 0.6145848375451264,
"grad_norm": 0.025940794497728348,
"learning_rate": 4.543186053825081e-06,
"loss": 0.8356,
"step": 532
},
{
"epoch": 0.6157400722021661,
"grad_norm": 0.025478770956397057,
"learning_rate": 4.544545348377621e-06,
"loss": 0.8118,
"step": 533
},
{
"epoch": 0.6168953068592058,
"grad_norm": 0.02738497406244278,
"learning_rate": 4.545902095047594e-06,
"loss": 0.8397,
"step": 534
},
{
"epoch": 0.6180505415162455,
"grad_norm": 0.025361735373735428,
"learning_rate": 4.547256303368714e-06,
"loss": 0.8422,
"step": 535
},
{
"epoch": 0.6192057761732852,
"grad_norm": 0.02575971558690071,
"learning_rate": 4.5486079828212835e-06,
"loss": 0.8224,
"step": 536
},
{
"epoch": 0.6203610108303249,
"grad_norm": 0.026920614764094353,
"learning_rate": 4.549957142832593e-06,
"loss": 0.7988,
"step": 537
},
{
"epoch": 0.6215162454873646,
"grad_norm": 0.026717036962509155,
"learning_rate": 4.5513037927773155e-06,
"loss": 0.8397,
"step": 538
},
{
"epoch": 0.6226714801444043,
"grad_norm": 0.028026578947901726,
"learning_rate": 4.552647941977898e-06,
"loss": 0.8421,
"step": 539
},
{
"epoch": 0.623826714801444,
"grad_norm": 0.02701553702354431,
"learning_rate": 4.553989599704948e-06,
"loss": 0.8129,
"step": 540
},
{
"epoch": 0.6249819494584837,
"grad_norm": 0.028897127136588097,
"learning_rate": 4.555328775177616e-06,
"loss": 0.8049,
"step": 541
},
{
"epoch": 0.6261371841155234,
"grad_norm": 0.027024636045098305,
"learning_rate": 4.556665477563979e-06,
"loss": 0.8255,
"step": 542
},
{
"epoch": 0.6272924187725631,
"grad_norm": 0.025836756452918053,
"learning_rate": 4.557999715981412e-06,
"loss": 0.8172,
"step": 543
},
{
"epoch": 0.6284476534296028,
"grad_norm": 0.027048436924815178,
"learning_rate": 4.559331499496967e-06,
"loss": 0.8336,
"step": 544
},
{
"epoch": 0.6296028880866426,
"grad_norm": 0.02669835276901722,
"learning_rate": 4.560660837127738e-06,
"loss": 0.7958,
"step": 545
},
{
"epoch": 0.6307581227436823,
"grad_norm": 0.02850779891014099,
"learning_rate": 4.5619877378412295e-06,
"loss": 0.8313,
"step": 546
},
{
"epoch": 0.6319133574007221,
"grad_norm": 0.027087301015853882,
"learning_rate": 4.563312210555719e-06,
"loss": 0.8257,
"step": 547
},
{
"epoch": 0.6330685920577618,
"grad_norm": 0.029034119099378586,
"learning_rate": 4.564634264140616e-06,
"loss": 0.8292,
"step": 548
},
{
"epoch": 0.6342238267148015,
"grad_norm": 0.02700372040271759,
"learning_rate": 4.5659539074168204e-06,
"loss": 0.828,
"step": 549
},
{
"epoch": 0.6353790613718412,
"grad_norm": 0.027077239006757736,
"learning_rate": 4.5672711491570735e-06,
"loss": 0.829,
"step": 550
},
{
"epoch": 0.6365342960288809,
"grad_norm": 0.027820030227303505,
"learning_rate": 4.568585998086309e-06,
"loss": 0.8323,
"step": 551
},
{
"epoch": 0.6376895306859206,
"grad_norm": 0.02544805407524109,
"learning_rate": 4.569898462881998e-06,
"loss": 0.8424,
"step": 552
},
{
"epoch": 0.6388447653429603,
"grad_norm": 0.025698795914649963,
"learning_rate": 4.571208552174497e-06,
"loss": 0.8471,
"step": 553
},
{
"epoch": 0.64,
"grad_norm": 0.026754576712846756,
"learning_rate": 4.572516274547383e-06,
"loss": 0.8277,
"step": 554
},
{
"epoch": 0.6411552346570397,
"grad_norm": 0.026160219684243202,
"learning_rate": 4.573821638537794e-06,
"loss": 0.8458,
"step": 555
},
{
"epoch": 0.6423104693140794,
"grad_norm": 0.02805655263364315,
"learning_rate": 4.5751246526367635e-06,
"loss": 0.8196,
"step": 556
},
{
"epoch": 0.6434657039711191,
"grad_norm": 0.02578035369515419,
"learning_rate": 4.576425325289549e-06,
"loss": 0.8235,
"step": 557
},
{
"epoch": 0.6446209386281588,
"grad_norm": 0.02676086314022541,
"learning_rate": 4.577723664895965e-06,
"loss": 0.7819,
"step": 558
},
{
"epoch": 0.6457761732851985,
"grad_norm": 0.02770860493183136,
"learning_rate": 4.579019679810706e-06,
"loss": 0.8063,
"step": 559
},
{
"epoch": 0.6469314079422382,
"grad_norm": 0.025705819949507713,
"learning_rate": 4.5803133783436676e-06,
"loss": 0.8145,
"step": 560
},
{
"epoch": 0.648086642599278,
"grad_norm": 0.025856079533696175,
"learning_rate": 4.5816047687602695e-06,
"loss": 0.7849,
"step": 561
},
{
"epoch": 0.6492418772563177,
"grad_norm": 0.026201602071523666,
"learning_rate": 4.582893859281769e-06,
"loss": 0.8262,
"step": 562
},
{
"epoch": 0.6503971119133574,
"grad_norm": 0.026211684569716454,
"learning_rate": 4.584180658085578e-06,
"loss": 0.8371,
"step": 563
},
{
"epoch": 0.6515523465703971,
"grad_norm": 0.02780756726861,
"learning_rate": 4.585465173305571e-06,
"loss": 0.8258,
"step": 564
},
{
"epoch": 0.6527075812274368,
"grad_norm": 0.02710585482418537,
"learning_rate": 4.586747413032399e-06,
"loss": 0.8289,
"step": 565
},
{
"epoch": 0.6538628158844766,
"grad_norm": 0.02654946967959404,
"learning_rate": 4.588027385313786e-06,
"loss": 0.8208,
"step": 566
},
{
"epoch": 0.6550180505415163,
"grad_norm": 0.025670049712061882,
"learning_rate": 4.589305098154845e-06,
"loss": 0.8236,
"step": 567
},
{
"epoch": 0.656173285198556,
"grad_norm": 0.025593513622879982,
"learning_rate": 4.5905805595183656e-06,
"loss": 0.8397,
"step": 568
},
{
"epoch": 0.6573285198555957,
"grad_norm": 0.02713758684694767,
"learning_rate": 4.59185377732512e-06,
"loss": 0.8077,
"step": 569
},
{
"epoch": 0.6584837545126354,
"grad_norm": 0.025683598592877388,
"learning_rate": 4.5931247594541535e-06,
"loss": 0.828,
"step": 570
},
{
"epoch": 0.6596389891696751,
"grad_norm": 0.027867255732417107,
"learning_rate": 4.594393513743081e-06,
"loss": 0.8053,
"step": 571
},
{
"epoch": 0.6607942238267148,
"grad_norm": 0.02871118299663067,
"learning_rate": 4.595660047988374e-06,
"loss": 0.8308,
"step": 572
},
{
"epoch": 0.6619494584837545,
"grad_norm": 0.02591397613286972,
"learning_rate": 4.59692436994565e-06,
"loss": 0.8376,
"step": 573
},
{
"epoch": 0.6631046931407942,
"grad_norm": 0.02970442920923233,
"learning_rate": 4.598186487329957e-06,
"loss": 0.8343,
"step": 574
},
{
"epoch": 0.6642599277978339,
"grad_norm": 0.027048716321587563,
"learning_rate": 4.599446407816052e-06,
"loss": 0.8169,
"step": 575
},
{
"epoch": 0.6654151624548736,
"grad_norm": 0.027099261060357094,
"learning_rate": 4.600704139038687e-06,
"loss": 0.8414,
"step": 576
},
{
"epoch": 0.6665703971119133,
"grad_norm": 0.02764103189110756,
"learning_rate": 4.601959688592886e-06,
"loss": 0.8144,
"step": 577
},
{
"epoch": 0.667725631768953,
"grad_norm": 0.025973627343773842,
"learning_rate": 4.603213064034216e-06,
"loss": 0.8086,
"step": 578
},
{
"epoch": 0.6688808664259928,
"grad_norm": 0.02816241979598999,
"learning_rate": 4.6044642728790615e-06,
"loss": 0.8266,
"step": 579
},
{
"epoch": 0.6700361010830325,
"grad_norm": 0.02614632435142994,
"learning_rate": 4.605713322604896e-06,
"loss": 0.8365,
"step": 580
},
{
"epoch": 0.6711913357400722,
"grad_norm": 0.02658606320619583,
"learning_rate": 4.606960220650551e-06,
"loss": 0.8273,
"step": 581
},
{
"epoch": 0.6723465703971119,
"grad_norm": 0.02635134756565094,
"learning_rate": 4.608204974416482e-06,
"loss": 0.8169,
"step": 582
},
{
"epoch": 0.6735018050541516,
"grad_norm": 0.025622064247727394,
"learning_rate": 4.609447591265024e-06,
"loss": 0.8256,
"step": 583
},
{
"epoch": 0.6746570397111913,
"grad_norm": 0.02630884386599064,
"learning_rate": 4.610688078520667e-06,
"loss": 0.8546,
"step": 584
},
{
"epoch": 0.675812274368231,
"grad_norm": 0.026515616104006767,
"learning_rate": 4.611926443470302e-06,
"loss": 0.829,
"step": 585
},
{
"epoch": 0.6769675090252708,
"grad_norm": 0.025966297835111618,
"learning_rate": 4.613162693363485e-06,
"loss": 0.8206,
"step": 586
},
{
"epoch": 0.6781227436823105,
"grad_norm": 0.025645708665251732,
"learning_rate": 4.614396835412691e-06,
"loss": 0.8214,
"step": 587
},
{
"epoch": 0.6792779783393502,
"grad_norm": 0.026266956701874733,
"learning_rate": 4.6156288767935645e-06,
"loss": 0.8128,
"step": 588
},
{
"epoch": 0.6804332129963899,
"grad_norm": 0.027987468987703323,
"learning_rate": 4.61685882464517e-06,
"loss": 0.8269,
"step": 589
},
{
"epoch": 0.6815884476534296,
"grad_norm": 0.02621353790163994,
"learning_rate": 4.61808668607024e-06,
"loss": 0.817,
"step": 590
},
{
"epoch": 0.6827436823104693,
"grad_norm": 0.026081636548042297,
"learning_rate": 4.619312468135426e-06,
"loss": 0.8168,
"step": 591
},
{
"epoch": 0.683898916967509,
"grad_norm": 0.026946187019348145,
"learning_rate": 4.620536177871534e-06,
"loss": 0.8208,
"step": 592
},
{
"epoch": 0.6850541516245487,
"grad_norm": 0.02429202012717724,
"learning_rate": 4.621757822273772e-06,
"loss": 0.8032,
"step": 593
},
{
"epoch": 0.6862093862815885,
"grad_norm": 0.027351096272468567,
"learning_rate": 4.62297740830199e-06,
"loss": 0.8234,
"step": 594
},
{
"epoch": 0.6873646209386282,
"grad_norm": 0.026615051552653313,
"learning_rate": 4.624194942880917e-06,
"loss": 0.8205,
"step": 595
},
{
"epoch": 0.6885198555956679,
"grad_norm": 0.025758862495422363,
"learning_rate": 4.625410432900395e-06,
"loss": 0.8372,
"step": 596
},
{
"epoch": 0.6896750902527076,
"grad_norm": 0.026305649429559708,
"learning_rate": 4.626623885215616e-06,
"loss": 0.813,
"step": 597
},
{
"epoch": 0.6908303249097473,
"grad_norm": 0.029083121567964554,
"learning_rate": 4.627835306647352e-06,
"loss": 0.801,
"step": 598
},
{
"epoch": 0.691985559566787,
"grad_norm": 0.02733645774424076,
"learning_rate": 4.6290447039821865e-06,
"loss": 0.818,
"step": 599
},
{
"epoch": 0.6931407942238267,
"grad_norm": 0.026958592236042023,
"learning_rate": 4.63025208397274e-06,
"loss": 0.8189,
"step": 600
},
{
"epoch": 0.6942960288808664,
"grad_norm": 0.02624497003853321,
"learning_rate": 4.631457453337899e-06,
"loss": 0.8217,
"step": 601
},
{
"epoch": 0.6954512635379061,
"grad_norm": 0.026590224355459213,
"learning_rate": 4.632660818763041e-06,
"loss": 0.8413,
"step": 602
},
{
"epoch": 0.6966064981949458,
"grad_norm": 0.029455525800585747,
"learning_rate": 4.633862186900253e-06,
"loss": 0.8437,
"step": 603
},
{
"epoch": 0.6977617328519855,
"grad_norm": 0.026593301445245743,
"learning_rate": 4.635061564368554e-06,
"loss": 0.8396,
"step": 604
},
{
"epoch": 0.6989169675090253,
"grad_norm": 0.026324933394789696,
"learning_rate": 4.636258957754115e-06,
"loss": 0.828,
"step": 605
},
{
"epoch": 0.700072202166065,
"grad_norm": 0.027765916660428047,
"learning_rate": 4.6374543736104774e-06,
"loss": 0.8403,
"step": 606
},
{
"epoch": 0.7012274368231047,
"grad_norm": 0.026490546762943268,
"learning_rate": 4.638647818458763e-06,
"loss": 0.8206,
"step": 607
},
{
"epoch": 0.7023826714801444,
"grad_norm": 0.028683004900813103,
"learning_rate": 4.639839298787892e-06,
"loss": 0.8203,
"step": 608
},
{
"epoch": 0.7035379061371841,
"grad_norm": 0.026209495961666107,
"learning_rate": 4.641028821054792e-06,
"loss": 0.8127,
"step": 609
},
{
"epoch": 0.7046931407942238,
"grad_norm": 0.02771477773785591,
"learning_rate": 4.642216391684613e-06,
"loss": 0.8218,
"step": 610
},
{
"epoch": 0.7058483754512636,
"grad_norm": 0.02859386056661606,
"learning_rate": 4.643402017070924e-06,
"loss": 0.7877,
"step": 611
},
{
"epoch": 0.7070036101083033,
"grad_norm": 0.026568232104182243,
"learning_rate": 4.644585703575936e-06,
"loss": 0.8062,
"step": 612
},
{
"epoch": 0.708158844765343,
"grad_norm": 0.030504100024700165,
"learning_rate": 4.645767457530693e-06,
"loss": 0.8068,
"step": 613
},
{
"epoch": 0.7093140794223827,
"grad_norm": 0.026128064841032028,
"learning_rate": 4.64694728523528e-06,
"loss": 0.8185,
"step": 614
},
{
"epoch": 0.7104693140794224,
"grad_norm": 0.026173869147896767,
"learning_rate": 4.648125192959028e-06,
"loss": 0.8305,
"step": 615
},
{
"epoch": 0.7116245487364621,
"grad_norm": 0.02751108445227146,
"learning_rate": 4.649301186940709e-06,
"loss": 0.8176,
"step": 616
},
{
"epoch": 0.7127797833935018,
"grad_norm": 0.026463057845830917,
"learning_rate": 4.650475273388737e-06,
"loss": 0.8011,
"step": 617
},
{
"epoch": 0.7139350180505415,
"grad_norm": 0.026889406144618988,
"learning_rate": 4.65164745848136e-06,
"loss": 0.8319,
"step": 618
},
{
"epoch": 0.7150902527075812,
"grad_norm": 0.02862033247947693,
"learning_rate": 4.652817748366864e-06,
"loss": 0.8394,
"step": 619
},
{
"epoch": 0.7162454873646209,
"grad_norm": 0.025286676362156868,
"learning_rate": 4.653986149163757e-06,
"loss": 0.825,
"step": 620
},
{
"epoch": 0.7174007220216606,
"grad_norm": 0.026653330773115158,
"learning_rate": 4.655152666960967e-06,
"loss": 0.7891,
"step": 621
},
{
"epoch": 0.7185559566787003,
"grad_norm": 0.027324773371219635,
"learning_rate": 4.656317307818032e-06,
"loss": 0.8105,
"step": 622
},
{
"epoch": 0.71971119133574,
"grad_norm": 0.026749735698103905,
"learning_rate": 4.657480077765284e-06,
"loss": 0.8125,
"step": 623
},
{
"epoch": 0.7208664259927798,
"grad_norm": 0.027720510959625244,
"learning_rate": 4.6586409828040405e-06,
"loss": 0.8385,
"step": 624
},
{
"epoch": 0.7220216606498195,
"grad_norm": 0.02868843637406826,
"learning_rate": 4.659800028906792e-06,
"loss": 0.8221,
"step": 625
},
{
"epoch": 0.7231768953068592,
"grad_norm": 0.026634184643626213,
"learning_rate": 4.660957222017384e-06,
"loss": 0.824,
"step": 626
},
{
"epoch": 0.724332129963899,
"grad_norm": 0.02601156197488308,
"learning_rate": 4.662112568051194e-06,
"loss": 0.813,
"step": 627
},
{
"epoch": 0.7254873646209387,
"grad_norm": 0.028468729928135872,
"learning_rate": 4.663266072895327e-06,
"loss": 0.8229,
"step": 628
},
{
"epoch": 0.7266425992779784,
"grad_norm": 0.026884114369750023,
"learning_rate": 4.664417742408782e-06,
"loss": 0.8201,
"step": 629
},
{
"epoch": 0.7277978339350181,
"grad_norm": 0.02819686383008957,
"learning_rate": 4.6655675824226375e-06,
"loss": 0.8048,
"step": 630
},
{
"epoch": 0.7289530685920578,
"grad_norm": 0.02609233930706978,
"learning_rate": 4.666715598740224e-06,
"loss": 0.8145,
"step": 631
},
{
"epoch": 0.7301083032490975,
"grad_norm": 0.025876475498080254,
"learning_rate": 4.667861797137309e-06,
"loss": 0.8236,
"step": 632
},
{
"epoch": 0.7312635379061372,
"grad_norm": 0.025990156456828117,
"learning_rate": 4.669006183362259e-06,
"loss": 0.8416,
"step": 633
},
{
"epoch": 0.7324187725631769,
"grad_norm": 0.026048380881547928,
"learning_rate": 4.670148763136221e-06,
"loss": 0.8004,
"step": 634
},
{
"epoch": 0.7335740072202166,
"grad_norm": 0.0267790500074625,
"learning_rate": 4.6712895421532935e-06,
"loss": 0.8249,
"step": 635
},
{
"epoch": 0.7347292418772563,
"grad_norm": 0.027045415714383125,
"learning_rate": 4.672428526080691e-06,
"loss": 0.8346,
"step": 636
},
{
"epoch": 0.735884476534296,
"grad_norm": 0.026687374338507652,
"learning_rate": 4.673565720558918e-06,
"loss": 0.8206,
"step": 637
},
{
"epoch": 0.7370397111913357,
"grad_norm": 0.025013815611600876,
"learning_rate": 4.674701131201937e-06,
"loss": 0.8278,
"step": 638
},
{
"epoch": 0.7381949458483754,
"grad_norm": 0.02536069042980671,
"learning_rate": 4.675834763597334e-06,
"loss": 0.8051,
"step": 639
},
{
"epoch": 0.7393501805054151,
"grad_norm": 0.025835523381829262,
"learning_rate": 4.676966623306479e-06,
"loss": 0.8131,
"step": 640
},
{
"epoch": 0.7405054151624548,
"grad_norm": 0.026328476145863533,
"learning_rate": 4.678096715864696e-06,
"loss": 0.834,
"step": 641
},
{
"epoch": 0.7416606498194945,
"grad_norm": 0.027017908170819283,
"learning_rate": 4.679225046781422e-06,
"loss": 0.808,
"step": 642
},
{
"epoch": 0.7428158844765343,
"grad_norm": 0.026868514716625214,
"learning_rate": 4.6803516215403704e-06,
"loss": 0.8236,
"step": 643
},
{
"epoch": 0.743971119133574,
"grad_norm": 0.026412272825837135,
"learning_rate": 4.6814764455996875e-06,
"loss": 0.8343,
"step": 644
},
{
"epoch": 0.7451263537906138,
"grad_norm": 0.025585437193512917,
"learning_rate": 4.682599524392114e-06,
"loss": 0.8165,
"step": 645
},
{
"epoch": 0.7462815884476535,
"grad_norm": 0.026194840669631958,
"learning_rate": 4.683720863325141e-06,
"loss": 0.8284,
"step": 646
},
{
"epoch": 0.7474368231046932,
"grad_norm": 0.025945279747247696,
"learning_rate": 4.6848404677811685e-06,
"loss": 0.8213,
"step": 647
},
{
"epoch": 0.7485920577617329,
"grad_norm": 0.02698802575469017,
"learning_rate": 4.685958343117656e-06,
"loss": 0.8399,
"step": 648
},
{
"epoch": 0.7497472924187726,
"grad_norm": 0.02705751173198223,
"learning_rate": 4.687074494667283e-06,
"loss": 0.8254,
"step": 649
},
{
"epoch": 0.7509025270758123,
"grad_norm": 0.02601143904030323,
"learning_rate": 4.688188927738093e-06,
"loss": 0.8283,
"step": 650
},
{
"epoch": 0.752057761732852,
"grad_norm": 0.026112712919712067,
"learning_rate": 4.689301647613653e-06,
"loss": 0.8334,
"step": 651
},
{
"epoch": 0.7532129963898917,
"grad_norm": 0.02763291820883751,
"learning_rate": 4.690412659553201e-06,
"loss": 0.8195,
"step": 652
},
{
"epoch": 0.7543682310469314,
"grad_norm": 0.02655601128935814,
"learning_rate": 4.69152196879179e-06,
"loss": 0.8164,
"step": 653
},
{
"epoch": 0.7555234657039711,
"grad_norm": 0.024583589285612106,
"learning_rate": 4.692629580540446e-06,
"loss": 0.8112,
"step": 654
},
{
"epoch": 0.7566787003610108,
"grad_norm": 0.02777348831295967,
"learning_rate": 4.693735499986305e-06,
"loss": 0.8228,
"step": 655
},
{
"epoch": 0.7578339350180505,
"grad_norm": 0.025867123156785965,
"learning_rate": 4.694839732292768e-06,
"loss": 0.8169,
"step": 656
},
{
"epoch": 0.7589891696750902,
"grad_norm": 0.02569735422730446,
"learning_rate": 4.695942282599635e-06,
"loss": 0.8274,
"step": 657
},
{
"epoch": 0.7601444043321299,
"grad_norm": 0.026997772976756096,
"learning_rate": 4.69704315602326e-06,
"loss": 0.8127,
"step": 658
},
{
"epoch": 0.7612996389891696,
"grad_norm": 0.026205215603113174,
"learning_rate": 4.698142357656684e-06,
"loss": 0.8053,
"step": 659
},
{
"epoch": 0.7624548736462093,
"grad_norm": 0.026929359883069992,
"learning_rate": 4.699239892569782e-06,
"loss": 0.8049,
"step": 660
},
{
"epoch": 0.763610108303249,
"grad_norm": 0.02786414325237274,
"learning_rate": 4.700335765809401e-06,
"loss": 0.8098,
"step": 661
},
{
"epoch": 0.7647653429602889,
"grad_norm": 0.02785409241914749,
"learning_rate": 4.701429982399501e-06,
"loss": 0.8203,
"step": 662
},
{
"epoch": 0.7659205776173286,
"grad_norm": 0.02913859114050865,
"learning_rate": 4.702522547341289e-06,
"loss": 0.8174,
"step": 663
},
{
"epoch": 0.7670758122743683,
"grad_norm": 0.025960449129343033,
"learning_rate": 4.703613465613363e-06,
"loss": 0.8111,
"step": 664
},
{
"epoch": 0.768231046931408,
"grad_norm": 0.027514390647411346,
"learning_rate": 4.704702742171842e-06,
"loss": 0.7999,
"step": 665
},
{
"epoch": 0.7693862815884477,
"grad_norm": 0.026732420548796654,
"learning_rate": 4.705790381950503e-06,
"loss": 0.8332,
"step": 666
},
{
"epoch": 0.7705415162454874,
"grad_norm": 0.025522086769342422,
"learning_rate": 4.706876389860915e-06,
"loss": 0.8179,
"step": 667
},
{
"epoch": 0.7716967509025271,
"grad_norm": 0.026885733008384705,
"learning_rate": 4.707960770792576e-06,
"loss": 0.7875,
"step": 668
},
{
"epoch": 0.7728519855595668,
"grad_norm": 0.026839058846235275,
"learning_rate": 4.709043529613039e-06,
"loss": 0.8029,
"step": 669
},
{
"epoch": 0.7740072202166065,
"grad_norm": 0.02562631480395794,
"learning_rate": 4.710124671168044e-06,
"loss": 0.8184,
"step": 670
},
{
"epoch": 0.7751624548736462,
"grad_norm": 0.026263324543833733,
"learning_rate": 4.7112042002816544e-06,
"loss": 0.8192,
"step": 671
},
{
"epoch": 0.7763176895306859,
"grad_norm": 0.026700858026742935,
"learning_rate": 4.7122821217563755e-06,
"loss": 0.827,
"step": 672
},
{
"epoch": 0.7774729241877256,
"grad_norm": 0.02566208317875862,
"learning_rate": 4.7133584403732955e-06,
"loss": 0.8072,
"step": 673
},
{
"epoch": 0.7786281588447653,
"grad_norm": 0.02590363658964634,
"learning_rate": 4.7144331608922e-06,
"loss": 0.8111,
"step": 674
},
{
"epoch": 0.779783393501805,
"grad_norm": 0.025736572220921516,
"learning_rate": 4.715506288051709e-06,
"loss": 0.8077,
"step": 675
},
{
"epoch": 0.7809386281588447,
"grad_norm": 0.027111530303955078,
"learning_rate": 4.716577826569394e-06,
"loss": 0.8429,
"step": 676
},
{
"epoch": 0.7820938628158844,
"grad_norm": 0.02599494345486164,
"learning_rate": 4.717647781141908e-06,
"loss": 0.8125,
"step": 677
},
{
"epoch": 0.7832490974729241,
"grad_norm": 0.02675134316086769,
"learning_rate": 4.7187161564451065e-06,
"loss": 0.8131,
"step": 678
},
{
"epoch": 0.7844043321299639,
"grad_norm": 0.026113273575901985,
"learning_rate": 4.71978295713417e-06,
"loss": 0.8292,
"step": 679
},
{
"epoch": 0.7855595667870036,
"grad_norm": 0.024617226794362068,
"learning_rate": 4.720848187843727e-06,
"loss": 0.8032,
"step": 680
},
{
"epoch": 0.7867148014440434,
"grad_norm": 0.02645159512758255,
"learning_rate": 4.721911853187975e-06,
"loss": 0.8287,
"step": 681
},
{
"epoch": 0.7878700361010831,
"grad_norm": 0.026287630200386047,
"learning_rate": 4.722973957760799e-06,
"loss": 0.8098,
"step": 682
},
{
"epoch": 0.7890252707581228,
"grad_norm": 0.02640565298497677,
"learning_rate": 4.724034506135888e-06,
"loss": 0.8303,
"step": 683
},
{
"epoch": 0.7901805054151625,
"grad_norm": 0.027293162420392036,
"learning_rate": 4.725093502866861e-06,
"loss": 0.7875,
"step": 684
},
{
"epoch": 0.7913357400722022,
"grad_norm": 0.02671169303357601,
"learning_rate": 4.7261509524873765e-06,
"loss": 0.7986,
"step": 685
},
{
"epoch": 0.7924909747292419,
"grad_norm": 0.025389740243554115,
"learning_rate": 4.727206859511253e-06,
"loss": 0.8168,
"step": 686
},
{
"epoch": 0.7936462093862816,
"grad_norm": 0.02669006958603859,
"learning_rate": 4.7282612284325845e-06,
"loss": 0.8061,
"step": 687
},
{
"epoch": 0.7948014440433213,
"grad_norm": 0.027196478098630905,
"learning_rate": 4.729314063725853e-06,
"loss": 0.8077,
"step": 688
},
{
"epoch": 0.795956678700361,
"grad_norm": 0.026497790589928627,
"learning_rate": 4.730365369846044e-06,
"loss": 0.8161,
"step": 689
},
{
"epoch": 0.7971119133574007,
"grad_norm": 0.025808461010456085,
"learning_rate": 4.73141515122876e-06,
"loss": 0.8042,
"step": 690
},
{
"epoch": 0.7982671480144404,
"grad_norm": 0.02556486800312996,
"learning_rate": 4.732463412290331e-06,
"loss": 0.7871,
"step": 691
},
{
"epoch": 0.7994223826714801,
"grad_norm": 0.02653394266963005,
"learning_rate": 4.73351015742793e-06,
"loss": 0.8163,
"step": 692
},
{
"epoch": 0.8005776173285198,
"grad_norm": 0.027373354882001877,
"learning_rate": 4.734555391019678e-06,
"loss": 0.8073,
"step": 693
},
{
"epoch": 0.8017328519855595,
"grad_norm": 0.02534923329949379,
"learning_rate": 4.735599117424759e-06,
"loss": 0.8186,
"step": 694
},
{
"epoch": 0.8028880866425993,
"grad_norm": 0.025402942672371864,
"learning_rate": 4.736641340983524e-06,
"loss": 0.8198,
"step": 695
},
{
"epoch": 0.804043321299639,
"grad_norm": 0.025877540931105614,
"learning_rate": 4.737682066017604e-06,
"loss": 0.7954,
"step": 696
},
{
"epoch": 0.8051985559566787,
"grad_norm": 0.027829859405755997,
"learning_rate": 4.738721296830016e-06,
"loss": 0.8197,
"step": 697
},
{
"epoch": 0.8063537906137184,
"grad_norm": 0.02695806324481964,
"learning_rate": 4.739759037705269e-06,
"loss": 0.8361,
"step": 698
},
{
"epoch": 0.8075090252707581,
"grad_norm": 0.02655518613755703,
"learning_rate": 4.740795292909469e-06,
"loss": 0.8055,
"step": 699
},
{
"epoch": 0.8086642599277978,
"grad_norm": 0.02709355391561985,
"learning_rate": 4.741830066690428e-06,
"loss": 0.7907,
"step": 700
},
{
"epoch": 0.8098194945848376,
"grad_norm": 0.026433579623699188,
"learning_rate": 4.742863363277765e-06,
"loss": 0.8235,
"step": 701
},
{
"epoch": 0.8109747292418773,
"grad_norm": 0.025645030662417412,
"learning_rate": 4.743895186883009e-06,
"loss": 0.8076,
"step": 702
},
{
"epoch": 0.812129963898917,
"grad_norm": 0.027833297848701477,
"learning_rate": 4.744925541699707e-06,
"loss": 0.8316,
"step": 703
},
{
"epoch": 0.8132851985559567,
"grad_norm": 0.025668229907751083,
"learning_rate": 4.74595443190352e-06,
"loss": 0.8075,
"step": 704
},
{
"epoch": 0.8144404332129964,
"grad_norm": 0.026288649067282677,
"learning_rate": 4.746981861652332e-06,
"loss": 0.8421,
"step": 705
},
{
"epoch": 0.8155956678700361,
"grad_norm": 0.024897335097193718,
"learning_rate": 4.7480078350863404e-06,
"loss": 0.8132,
"step": 706
},
{
"epoch": 0.8167509025270758,
"grad_norm": 0.0259522944688797,
"learning_rate": 4.749032356328167e-06,
"loss": 0.8326,
"step": 707
},
{
"epoch": 0.8179061371841155,
"grad_norm": 0.026102952659130096,
"learning_rate": 4.750055429482949e-06,
"loss": 0.8118,
"step": 708
},
{
"epoch": 0.8190613718411552,
"grad_norm": 0.025872284546494484,
"learning_rate": 4.751077058638445e-06,
"loss": 0.816,
"step": 709
},
{
"epoch": 0.8202166064981949,
"grad_norm": 0.025316176936030388,
"learning_rate": 4.752097247865126e-06,
"loss": 0.8061,
"step": 710
},
{
"epoch": 0.8213718411552346,
"grad_norm": 0.02691754698753357,
"learning_rate": 4.753116001216277e-06,
"loss": 0.7946,
"step": 711
},
{
"epoch": 0.8225270758122744,
"grad_norm": 0.027329301461577415,
"learning_rate": 4.754133322728095e-06,
"loss": 0.8379,
"step": 712
},
{
"epoch": 0.8236823104693141,
"grad_norm": 0.026258062571287155,
"learning_rate": 4.755149216419776e-06,
"loss": 0.8101,
"step": 713
},
{
"epoch": 0.8248375451263538,
"grad_norm": 0.02672073245048523,
"learning_rate": 4.756163686293624e-06,
"loss": 0.8225,
"step": 714
},
{
"epoch": 0.8259927797833935,
"grad_norm": 0.028196029365062714,
"learning_rate": 4.757176736335135e-06,
"loss": 0.848,
"step": 715
},
{
"epoch": 0.8271480144404332,
"grad_norm": 0.02594706416130066,
"learning_rate": 4.758188370513093e-06,
"loss": 0.8426,
"step": 716
},
{
"epoch": 0.8283032490974729,
"grad_norm": 0.028120605275034904,
"learning_rate": 4.759198592779668e-06,
"loss": 0.8306,
"step": 717
},
{
"epoch": 0.8294584837545126,
"grad_norm": 0.027138734236359596,
"learning_rate": 4.760207407070501e-06,
"loss": 0.825,
"step": 718
},
{
"epoch": 0.8306137184115523,
"grad_norm": 0.027197662740945816,
"learning_rate": 4.761214817304805e-06,
"loss": 0.8071,
"step": 719
},
{
"epoch": 0.8317689530685921,
"grad_norm": 0.026295404881238937,
"learning_rate": 4.7622208273854484e-06,
"loss": 0.8158,
"step": 720
},
{
"epoch": 0.8329241877256318,
"grad_norm": 0.02767680026590824,
"learning_rate": 4.763225441199049e-06,
"loss": 0.8226,
"step": 721
},
{
"epoch": 0.8340794223826715,
"grad_norm": 0.025539277121424675,
"learning_rate": 4.764228662616066e-06,
"loss": 0.8025,
"step": 722
},
{
"epoch": 0.8352346570397112,
"grad_norm": 0.02579374797642231,
"learning_rate": 4.765230495490885e-06,
"loss": 0.842,
"step": 723
},
{
"epoch": 0.8363898916967509,
"grad_norm": 0.025891058146953583,
"learning_rate": 4.766230943661912e-06,
"loss": 0.8135,
"step": 724
},
{
"epoch": 0.8375451263537906,
"grad_norm": 0.02527502365410328,
"learning_rate": 4.767230010951657e-06,
"loss": 0.8087,
"step": 725
},
{
"epoch": 0.8387003610108303,
"grad_norm": 0.02618500031530857,
"learning_rate": 4.768227701166823e-06,
"loss": 0.8242,
"step": 726
},
{
"epoch": 0.83985559566787,
"grad_norm": 0.026595618575811386,
"learning_rate": 4.769224018098397e-06,
"loss": 0.8418,
"step": 727
},
{
"epoch": 0.8410108303249098,
"grad_norm": 0.02652081847190857,
"learning_rate": 4.770218965521729e-06,
"loss": 0.8013,
"step": 728
},
{
"epoch": 0.8421660649819495,
"grad_norm": 0.026961620897054672,
"learning_rate": 4.771212547196624e-06,
"loss": 0.7984,
"step": 729
},
{
"epoch": 0.8433212996389892,
"grad_norm": 0.024961085990071297,
"learning_rate": 4.772204766867427e-06,
"loss": 0.809,
"step": 730
},
{
"epoch": 0.8444765342960289,
"grad_norm": 0.026465538889169693,
"learning_rate": 4.773195628263101e-06,
"loss": 0.8186,
"step": 731
},
{
"epoch": 0.8456317689530686,
"grad_norm": 0.025633295997977257,
"learning_rate": 4.77418513509732e-06,
"loss": 0.8111,
"step": 732
},
{
"epoch": 0.8467870036101083,
"grad_norm": 0.026612253859639168,
"learning_rate": 4.775173291068547e-06,
"loss": 0.8218,
"step": 733
},
{
"epoch": 0.847942238267148,
"grad_norm": 0.026292981579899788,
"learning_rate": 4.776160099860117e-06,
"loss": 0.8107,
"step": 734
},
{
"epoch": 0.8490974729241877,
"grad_norm": 0.026987800374627113,
"learning_rate": 4.777145565140325e-06,
"loss": 0.8088,
"step": 735
},
{
"epoch": 0.8502527075812274,
"grad_norm": 0.027124376967549324,
"learning_rate": 4.778129690562499e-06,
"loss": 0.7977,
"step": 736
},
{
"epoch": 0.8514079422382671,
"grad_norm": 0.025853095576167107,
"learning_rate": 4.7791124797650865e-06,
"loss": 0.8105,
"step": 737
},
{
"epoch": 0.8525631768953068,
"grad_norm": 0.027579553425312042,
"learning_rate": 4.780093936371737e-06,
"loss": 0.825,
"step": 738
},
{
"epoch": 0.8537184115523466,
"grad_norm": 0.027602966874837875,
"learning_rate": 4.781074063991377e-06,
"loss": 0.8088,
"step": 739
},
{
"epoch": 0.8548736462093863,
"grad_norm": 0.025124140083789825,
"learning_rate": 4.782052866218294e-06,
"loss": 0.7909,
"step": 740
},
{
"epoch": 0.856028880866426,
"grad_norm": 0.028426257893443108,
"learning_rate": 4.783030346632214e-06,
"loss": 0.8036,
"step": 741
},
{
"epoch": 0.8571841155234657,
"grad_norm": 0.026561887934803963,
"learning_rate": 4.784006508798379e-06,
"loss": 0.8118,
"step": 742
},
{
"epoch": 0.8583393501805054,
"grad_norm": 0.027730902656912804,
"learning_rate": 4.7849813562676265e-06,
"loss": 0.8319,
"step": 743
},
{
"epoch": 0.8594945848375452,
"grad_norm": 0.02669842354953289,
"learning_rate": 4.785954892576465e-06,
"loss": 0.8153,
"step": 744
},
{
"epoch": 0.8606498194945849,
"grad_norm": 0.02621045894920826,
"learning_rate": 4.786927121247155e-06,
"loss": 0.8164,
"step": 745
},
{
"epoch": 0.8618050541516246,
"grad_norm": 0.02661043219268322,
"learning_rate": 4.787898045787781e-06,
"loss": 0.7877,
"step": 746
},
{
"epoch": 0.8629602888086643,
"grad_norm": 0.02501833625137806,
"learning_rate": 4.788867669692332e-06,
"loss": 0.8096,
"step": 747
},
{
"epoch": 0.864115523465704,
"grad_norm": 0.025194406509399414,
"learning_rate": 4.78983599644077e-06,
"loss": 0.791,
"step": 748
},
{
"epoch": 0.8652707581227437,
"grad_norm": 0.026089461520314217,
"learning_rate": 4.790803029499111e-06,
"loss": 0.7906,
"step": 749
},
{
"epoch": 0.8664259927797834,
"grad_norm": 0.025562454015016556,
"learning_rate": 4.7917687723195e-06,
"loss": 0.7749,
"step": 750
},
{
"epoch": 0.8675812274368231,
"grad_norm": 0.025910962373018265,
"learning_rate": 4.792733228340281e-06,
"loss": 0.7991,
"step": 751
},
{
"epoch": 0.8687364620938628,
"grad_norm": 0.026650816202163696,
"learning_rate": 4.793696400986071e-06,
"loss": 0.7999,
"step": 752
},
{
"epoch": 0.8698916967509025,
"grad_norm": 0.025604519993066788,
"learning_rate": 4.794658293667835e-06,
"loss": 0.8337,
"step": 753
},
{
"epoch": 0.8710469314079422,
"grad_norm": 0.02709670551121235,
"learning_rate": 4.795618909782958e-06,
"loss": 0.8128,
"step": 754
},
{
"epoch": 0.8722021660649819,
"grad_norm": 0.027338657528162003,
"learning_rate": 4.7965782527153145e-06,
"loss": 0.8088,
"step": 755
},
{
"epoch": 0.8733574007220216,
"grad_norm": 0.028811268508434296,
"learning_rate": 4.797536325835345e-06,
"loss": 0.8195,
"step": 756
},
{
"epoch": 0.8745126353790613,
"grad_norm": 0.02585168555378914,
"learning_rate": 4.798493132500121e-06,
"loss": 0.8067,
"step": 757
},
{
"epoch": 0.8756678700361011,
"grad_norm": 0.024853166192770004,
"learning_rate": 4.799448676053423e-06,
"loss": 0.8032,
"step": 758
},
{
"epoch": 0.8768231046931408,
"grad_norm": 0.030403772369027138,
"learning_rate": 4.800402959825802e-06,
"loss": 0.8369,
"step": 759
},
{
"epoch": 0.8779783393501805,
"grad_norm": 0.02667844668030739,
"learning_rate": 4.801355987134653e-06,
"loss": 0.8112,
"step": 760
},
{
"epoch": 0.8791335740072203,
"grad_norm": 0.026796355843544006,
"learning_rate": 4.802307761284289e-06,
"loss": 0.819,
"step": 761
},
{
"epoch": 0.88028880866426,
"grad_norm": 0.02874310314655304,
"learning_rate": 4.803258285566001e-06,
"loss": 0.8149,
"step": 762
},
{
"epoch": 0.8814440433212997,
"grad_norm": 0.027145931497216225,
"learning_rate": 4.804207563258135e-06,
"loss": 0.8095,
"step": 763
},
{
"epoch": 0.8825992779783394,
"grad_norm": 0.02635267935693264,
"learning_rate": 4.80515559762615e-06,
"loss": 0.8188,
"step": 764
},
{
"epoch": 0.8837545126353791,
"grad_norm": 0.028352461755275726,
"learning_rate": 4.806102391922697e-06,
"loss": 0.8248,
"step": 765
},
{
"epoch": 0.8849097472924188,
"grad_norm": 0.026468653231859207,
"learning_rate": 4.8070479493876735e-06,
"loss": 0.8083,
"step": 766
},
{
"epoch": 0.8860649819494585,
"grad_norm": 0.02515551634132862,
"learning_rate": 4.8079922732483016e-06,
"loss": 0.8104,
"step": 767
},
{
"epoch": 0.8872202166064982,
"grad_norm": 0.02771628648042679,
"learning_rate": 4.808935366719187e-06,
"loss": 0.8083,
"step": 768
},
{
"epoch": 0.8883754512635379,
"grad_norm": 0.0274192001670599,
"learning_rate": 4.8098772330023855e-06,
"loss": 0.8181,
"step": 769
},
{
"epoch": 0.8895306859205776,
"grad_norm": 0.025738628581166267,
"learning_rate": 4.81081787528747e-06,
"loss": 0.8166,
"step": 770
},
{
"epoch": 0.8906859205776173,
"grad_norm": 0.026444412767887115,
"learning_rate": 4.8117572967515955e-06,
"loss": 0.8061,
"step": 771
},
{
"epoch": 0.891841155234657,
"grad_norm": 0.027295244857668877,
"learning_rate": 4.812695500559561e-06,
"loss": 0.796,
"step": 772
},
{
"epoch": 0.8929963898916967,
"grad_norm": 0.029825204983353615,
"learning_rate": 4.813632489863876e-06,
"loss": 0.8269,
"step": 773
},
{
"epoch": 0.8941516245487364,
"grad_norm": 0.027049973607063293,
"learning_rate": 4.814568267804822e-06,
"loss": 0.8166,
"step": 774
},
{
"epoch": 0.8953068592057761,
"grad_norm": 0.027095604687929153,
"learning_rate": 4.815502837510518e-06,
"loss": 0.8006,
"step": 775
},
{
"epoch": 0.8964620938628158,
"grad_norm": 0.02874009497463703,
"learning_rate": 4.816436202096981e-06,
"loss": 0.7987,
"step": 776
},
{
"epoch": 0.8976173285198557,
"grad_norm": 0.027292495593428612,
"learning_rate": 4.817368364668191e-06,
"loss": 0.8148,
"step": 777
},
{
"epoch": 0.8987725631768954,
"grad_norm": 0.025094101205468178,
"learning_rate": 4.8182993283161485e-06,
"loss": 0.833,
"step": 778
},
{
"epoch": 0.8999277978339351,
"grad_norm": 0.026829306036233902,
"learning_rate": 4.819229096120941e-06,
"loss": 0.8245,
"step": 779
},
{
"epoch": 0.9010830324909748,
"grad_norm": 0.0263043362647295,
"learning_rate": 4.820157671150801e-06,
"loss": 0.826,
"step": 780
},
{
"epoch": 0.9022382671480145,
"grad_norm": 0.029344236478209496,
"learning_rate": 4.821085056462168e-06,
"loss": 0.8322,
"step": 781
},
{
"epoch": 0.9033935018050542,
"grad_norm": 0.02591596357524395,
"learning_rate": 4.822011255099747e-06,
"loss": 0.8305,
"step": 782
},
{
"epoch": 0.9045487364620939,
"grad_norm": 0.02538052201271057,
"learning_rate": 4.8229362700965724e-06,
"loss": 0.8119,
"step": 783
},
{
"epoch": 0.9057039711191336,
"grad_norm": 0.028109556064009666,
"learning_rate": 4.823860104474065e-06,
"loss": 0.8334,
"step": 784
},
{
"epoch": 0.9068592057761733,
"grad_norm": 0.027156807482242584,
"learning_rate": 4.8247827612420875e-06,
"loss": 0.8148,
"step": 785
},
{
"epoch": 0.908014440433213,
"grad_norm": 0.02679731696844101,
"learning_rate": 4.825704243399014e-06,
"loss": 0.8228,
"step": 786
},
{
"epoch": 0.9091696750902527,
"grad_norm": 0.026814332231879234,
"learning_rate": 4.8266245539317745e-06,
"loss": 0.8027,
"step": 787
},
{
"epoch": 0.9103249097472924,
"grad_norm": 0.02507951483130455,
"learning_rate": 4.827543695815926e-06,
"loss": 0.796,
"step": 788
},
{
"epoch": 0.9114801444043321,
"grad_norm": 0.026251574978232384,
"learning_rate": 4.828461672015701e-06,
"loss": 0.8125,
"step": 789
},
{
"epoch": 0.9126353790613718,
"grad_norm": 0.02636023238301277,
"learning_rate": 4.82937848548407e-06,
"loss": 0.7954,
"step": 790
},
{
"epoch": 0.9137906137184115,
"grad_norm": 0.026472898200154305,
"learning_rate": 4.830294139162795e-06,
"loss": 0.8045,
"step": 791
},
{
"epoch": 0.9149458483754512,
"grad_norm": 0.025792265310883522,
"learning_rate": 4.83120863598249e-06,
"loss": 0.8072,
"step": 792
},
{
"epoch": 0.9161010830324909,
"grad_norm": 0.028488213196396828,
"learning_rate": 4.832121978862673e-06,
"loss": 0.7935,
"step": 793
},
{
"epoch": 0.9172563176895306,
"grad_norm": 0.02507089264690876,
"learning_rate": 4.833034170711828e-06,
"loss": 0.7944,
"step": 794
},
{
"epoch": 0.9184115523465703,
"grad_norm": 0.026674294844269753,
"learning_rate": 4.833945214427451e-06,
"loss": 0.8029,
"step": 795
},
{
"epoch": 0.91956678700361,
"grad_norm": 0.02713027596473694,
"learning_rate": 4.834855112896116e-06,
"loss": 0.8342,
"step": 796
},
{
"epoch": 0.9207220216606499,
"grad_norm": 0.026090459898114204,
"learning_rate": 4.835763868993521e-06,
"loss": 0.7986,
"step": 797
},
{
"epoch": 0.9218772563176896,
"grad_norm": 0.02631318010389805,
"learning_rate": 4.83667148558455e-06,
"loss": 0.8021,
"step": 798
},
{
"epoch": 0.9230324909747293,
"grad_norm": 0.027506891638040543,
"learning_rate": 4.8375779655233195e-06,
"loss": 0.8305,
"step": 799
},
{
"epoch": 0.924187725631769,
"grad_norm": 0.02776806242763996,
"learning_rate": 4.83848331165324e-06,
"loss": 0.8141,
"step": 800
},
{
"epoch": 0.9253429602888087,
"grad_norm": 0.027539484202861786,
"learning_rate": 4.839387526807064e-06,
"loss": 0.809,
"step": 801
},
{
"epoch": 0.9264981949458484,
"grad_norm": 0.02546604722738266,
"learning_rate": 4.84029061380694e-06,
"loss": 0.8075,
"step": 802
},
{
"epoch": 0.9276534296028881,
"grad_norm": 0.02793939970433712,
"learning_rate": 4.841192575464469e-06,
"loss": 0.8116,
"step": 803
},
{
"epoch": 0.9288086642599278,
"grad_norm": 0.026220016181468964,
"learning_rate": 4.842093414580753e-06,
"loss": 0.8085,
"step": 804
},
{
"epoch": 0.9299638989169675,
"grad_norm": 0.0267406664788723,
"learning_rate": 4.842993133946448e-06,
"loss": 0.8202,
"step": 805
},
{
"epoch": 0.9311191335740072,
"grad_norm": 0.028274521231651306,
"learning_rate": 4.8438917363418184e-06,
"loss": 0.8222,
"step": 806
},
{
"epoch": 0.9322743682310469,
"grad_norm": 0.025681311264634132,
"learning_rate": 4.844789224536785e-06,
"loss": 0.8009,
"step": 807
},
{
"epoch": 0.9334296028880866,
"grad_norm": 0.02876345068216324,
"learning_rate": 4.845685601290977e-06,
"loss": 0.8129,
"step": 808
},
{
"epoch": 0.9345848375451263,
"grad_norm": 0.027744382619857788,
"learning_rate": 4.846580869353787e-06,
"loss": 0.8084,
"step": 809
},
{
"epoch": 0.935740072202166,
"grad_norm": 0.026011621579527855,
"learning_rate": 4.847475031464417e-06,
"loss": 0.8058,
"step": 810
},
{
"epoch": 0.9368953068592057,
"grad_norm": 0.028499187901616096,
"learning_rate": 4.848368090351928e-06,
"loss": 0.8,
"step": 811
},
{
"epoch": 0.9380505415162455,
"grad_norm": 0.026256656274199486,
"learning_rate": 4.849260048735293e-06,
"loss": 0.8071,
"step": 812
},
{
"epoch": 0.9392057761732852,
"grad_norm": 0.0278424471616745,
"learning_rate": 4.8501509093234474e-06,
"loss": 0.8173,
"step": 813
},
{
"epoch": 0.9403610108303249,
"grad_norm": 0.028124138712882996,
"learning_rate": 4.851040674815336e-06,
"loss": 0.8057,
"step": 814
},
{
"epoch": 0.9415162454873646,
"grad_norm": 0.02599998749792576,
"learning_rate": 4.851929347899962e-06,
"loss": 0.7952,
"step": 815
},
{
"epoch": 0.9426714801444044,
"grad_norm": 0.03042020834982395,
"learning_rate": 4.852816931256436e-06,
"loss": 0.815,
"step": 816
},
{
"epoch": 0.9438267148014441,
"grad_norm": 0.02526993304491043,
"learning_rate": 4.853703427554027e-06,
"loss": 0.8169,
"step": 817
},
{
"epoch": 0.9449819494584838,
"grad_norm": 0.029504787176847458,
"learning_rate": 4.854588839452205e-06,
"loss": 0.8134,
"step": 818
},
{
"epoch": 0.9461371841155235,
"grad_norm": 0.02730252593755722,
"learning_rate": 4.855473169600699e-06,
"loss": 0.8198,
"step": 819
},
{
"epoch": 0.9472924187725632,
"grad_norm": 0.02669229544699192,
"learning_rate": 4.856356420639528e-06,
"loss": 0.8426,
"step": 820
},
{
"epoch": 0.9484476534296029,
"grad_norm": 0.029394106939435005,
"learning_rate": 4.857238595199068e-06,
"loss": 0.8025,
"step": 821
},
{
"epoch": 0.9496028880866426,
"grad_norm": 0.026510139927268028,
"learning_rate": 4.858119695900084e-06,
"loss": 0.8267,
"step": 822
},
{
"epoch": 0.9507581227436823,
"grad_norm": 0.025655683130025864,
"learning_rate": 4.858999725353784e-06,
"loss": 0.7936,
"step": 823
},
{
"epoch": 0.951913357400722,
"grad_norm": 0.029900815337896347,
"learning_rate": 4.859878686161861e-06,
"loss": 0.8245,
"step": 824
},
{
"epoch": 0.9530685920577617,
"grad_norm": 0.02642049826681614,
"learning_rate": 4.860756580916543e-06,
"loss": 0.8036,
"step": 825
},
{
"epoch": 0.9542238267148014,
"grad_norm": 0.027764586731791496,
"learning_rate": 4.861633412200637e-06,
"loss": 0.8132,
"step": 826
},
{
"epoch": 0.9553790613718411,
"grad_norm": 0.02786201983690262,
"learning_rate": 4.862509182587578e-06,
"loss": 0.8189,
"step": 827
},
{
"epoch": 0.9565342960288808,
"grad_norm": 0.02661985345184803,
"learning_rate": 4.8633838946414675e-06,
"loss": 0.8056,
"step": 828
},
{
"epoch": 0.9576895306859206,
"grad_norm": 0.029807835817337036,
"learning_rate": 4.864257550917123e-06,
"loss": 0.8108,
"step": 829
},
{
"epoch": 0.9588447653429603,
"grad_norm": 0.025031132623553276,
"learning_rate": 4.8651301539601235e-06,
"loss": 0.8031,
"step": 830
},
{
"epoch": 0.96,
"grad_norm": 0.028132742270827293,
"learning_rate": 4.866001706306853e-06,
"loss": 0.8199,
"step": 831
},
{
"epoch": 0.9611552346570397,
"grad_norm": 0.02886558324098587,
"learning_rate": 4.866872210484541e-06,
"loss": 0.8175,
"step": 832
},
{
"epoch": 0.9623104693140794,
"grad_norm": 0.026310063898563385,
"learning_rate": 4.867741669011313e-06,
"loss": 0.7998,
"step": 833
},
{
"epoch": 0.9634657039711191,
"grad_norm": 0.02729332074522972,
"learning_rate": 4.868610084396232e-06,
"loss": 0.7809,
"step": 834
},
{
"epoch": 0.9646209386281589,
"grad_norm": 0.027129124850034714,
"learning_rate": 4.869477459139337e-06,
"loss": 0.8219,
"step": 835
},
{
"epoch": 0.9657761732851986,
"grad_norm": 0.027403220534324646,
"learning_rate": 4.8703437957316945e-06,
"loss": 0.807,
"step": 836
},
{
"epoch": 0.9669314079422383,
"grad_norm": 0.028556736186146736,
"learning_rate": 4.871209096655434e-06,
"loss": 0.8213,
"step": 837
},
{
"epoch": 0.968086642599278,
"grad_norm": 0.027089878916740417,
"learning_rate": 4.872073364383795e-06,
"loss": 0.8167,
"step": 838
},
{
"epoch": 0.9692418772563177,
"grad_norm": 0.02591611072421074,
"learning_rate": 4.872936601381167e-06,
"loss": 0.7926,
"step": 839
},
{
"epoch": 0.9703971119133574,
"grad_norm": 0.028653450310230255,
"learning_rate": 4.873798810103137e-06,
"loss": 0.769,
"step": 840
},
{
"epoch": 0.9715523465703971,
"grad_norm": 0.025429286062717438,
"learning_rate": 4.874659992996521e-06,
"loss": 0.8033,
"step": 841
},
{
"epoch": 0.9727075812274368,
"grad_norm": 0.027507685124874115,
"learning_rate": 4.875520152499416e-06,
"loss": 0.7991,
"step": 842
},
{
"epoch": 0.9738628158844765,
"grad_norm": 0.02781766653060913,
"learning_rate": 4.876379291041238e-06,
"loss": 0.811,
"step": 843
},
{
"epoch": 0.9750180505415162,
"grad_norm": 0.02732367254793644,
"learning_rate": 4.87723741104276e-06,
"loss": 0.7795,
"step": 844
},
{
"epoch": 0.976173285198556,
"grad_norm": 0.027977894991636276,
"learning_rate": 4.878094514916154e-06,
"loss": 0.8024,
"step": 845
},
{
"epoch": 0.9773285198555957,
"grad_norm": 0.02706928178668022,
"learning_rate": 4.87895060506504e-06,
"loss": 0.8273,
"step": 846
},
{
"epoch": 0.9784837545126354,
"grad_norm": 0.027973534539341927,
"learning_rate": 4.879805683884512e-06,
"loss": 0.8249,
"step": 847
},
{
"epoch": 0.9796389891696751,
"grad_norm": 0.028234517201781273,
"learning_rate": 4.88065975376119e-06,
"loss": 0.8009,
"step": 848
},
{
"epoch": 0.9807942238267148,
"grad_norm": 0.026143617928028107,
"learning_rate": 4.881512817073255e-06,
"loss": 0.8114,
"step": 849
},
{
"epoch": 0.9819494584837545,
"grad_norm": 0.027196498587727547,
"learning_rate": 4.882364876190489e-06,
"loss": 0.7919,
"step": 850
},
{
"epoch": 0.9831046931407942,
"grad_norm": 0.027922434732317924,
"learning_rate": 4.883215933474314e-06,
"loss": 0.7986,
"step": 851
},
{
"epoch": 0.9842599277978339,
"grad_norm": 0.0268290676176548,
"learning_rate": 4.884065991277834e-06,
"loss": 0.8142,
"step": 852
},
{
"epoch": 0.9854151624548736,
"grad_norm": 0.026363542303442955,
"learning_rate": 4.8849150519458725e-06,
"loss": 0.7958,
"step": 853
},
{
"epoch": 0.9865703971119134,
"grad_norm": 0.026863981038331985,
"learning_rate": 4.88576311781501e-06,
"loss": 0.8132,
"step": 854
},
{
"epoch": 0.9877256317689531,
"grad_norm": 0.026298578828573227,
"learning_rate": 4.886610191213622e-06,
"loss": 0.8156,
"step": 855
},
{
"epoch": 0.9888808664259928,
"grad_norm": 0.026729200035333633,
"learning_rate": 4.8874562744619224e-06,
"loss": 0.8221,
"step": 856
},
{
"epoch": 0.9900361010830325,
"grad_norm": 0.025892335921525955,
"learning_rate": 4.888301369871998e-06,
"loss": 0.8267,
"step": 857
},
{
"epoch": 0.9911913357400722,
"grad_norm": 0.026062045246362686,
"learning_rate": 4.8891454797478435e-06,
"loss": 0.8206,
"step": 858
},
{
"epoch": 0.9923465703971119,
"grad_norm": 0.025573009625077248,
"learning_rate": 4.889988606385404e-06,
"loss": 0.8466,
"step": 859
},
{
"epoch": 0.9935018050541516,
"grad_norm": 0.026895977556705475,
"learning_rate": 4.890830752072613e-06,
"loss": 0.801,
"step": 860
},
{
"epoch": 0.9946570397111913,
"grad_norm": 0.026373952627182007,
"learning_rate": 4.891671919089425e-06,
"loss": 0.7935,
"step": 861
},
{
"epoch": 0.995812274368231,
"grad_norm": 0.025112492963671684,
"learning_rate": 4.892512109707855e-06,
"loss": 0.8104,
"step": 862
},
{
"epoch": 0.9969675090252708,
"grad_norm": 0.025861382484436035,
"learning_rate": 4.893351326192016e-06,
"loss": 0.7755,
"step": 863
},
{
"epoch": 0.9981227436823105,
"grad_norm": 0.025520671159029007,
"learning_rate": 4.8941895707981555e-06,
"loss": 0.783,
"step": 864
},
{
"epoch": 0.9992779783393502,
"grad_norm": 0.02590048499405384,
"learning_rate": 4.895026845774691e-06,
"loss": 0.7952,
"step": 865
},
{
"epoch": 1.00043321299639,
"grad_norm": 0.037370167672634125,
"learning_rate": 4.895863153362244e-06,
"loss": 1.0888,
"step": 866
},
{
"epoch": 1.0015884476534296,
"grad_norm": 0.028089042752981186,
"learning_rate": 4.8966984957936845e-06,
"loss": 0.8026,
"step": 867
},
{
"epoch": 1.0027436823104694,
"grad_norm": 0.02722124569118023,
"learning_rate": 4.897532875294154e-06,
"loss": 0.7892,
"step": 868
},
{
"epoch": 1.003898916967509,
"grad_norm": 0.02784401923418045,
"learning_rate": 4.898366294081111e-06,
"loss": 0.7766,
"step": 869
},
{
"epoch": 1.0050541516245488,
"grad_norm": 0.028576616197824478,
"learning_rate": 4.899198754364365e-06,
"loss": 0.794,
"step": 870
},
{
"epoch": 1.0062093862815884,
"grad_norm": 0.02609376236796379,
"learning_rate": 4.900030258346106e-06,
"loss": 0.787,
"step": 871
},
{
"epoch": 1.0073646209386282,
"grad_norm": 0.026441458612680435,
"learning_rate": 4.900860808220946e-06,
"loss": 0.7726,
"step": 872
},
{
"epoch": 1.0085198555956678,
"grad_norm": 0.026604607701301575,
"learning_rate": 4.90169040617595e-06,
"loss": 0.7672,
"step": 873
},
{
"epoch": 1.0096750902527076,
"grad_norm": 0.025642454624176025,
"learning_rate": 4.902519054390672e-06,
"loss": 0.7804,
"step": 874
},
{
"epoch": 1.0108303249097472,
"grad_norm": 0.027134709060192108,
"learning_rate": 4.9033467550371886e-06,
"loss": 0.7895,
"step": 875
},
{
"epoch": 1.011985559566787,
"grad_norm": 0.02571208029985428,
"learning_rate": 4.904173510280135e-06,
"loss": 0.7916,
"step": 876
},
{
"epoch": 1.0131407942238266,
"grad_norm": 0.026592353358864784,
"learning_rate": 4.904999322276735e-06,
"loss": 0.7975,
"step": 877
},
{
"epoch": 1.0142960288808665,
"grad_norm": 0.02615254372358322,
"learning_rate": 4.905824193176839e-06,
"loss": 0.7881,
"step": 878
},
{
"epoch": 1.015451263537906,
"grad_norm": 0.027889380231499672,
"learning_rate": 4.906648125122953e-06,
"loss": 0.7897,
"step": 879
},
{
"epoch": 1.0166064981949459,
"grad_norm": 0.02607213519513607,
"learning_rate": 4.907471120250281e-06,
"loss": 0.7735,
"step": 880
},
{
"epoch": 1.0177617328519855,
"grad_norm": 0.025891225785017014,
"learning_rate": 4.9082931806867475e-06,
"loss": 0.788,
"step": 881
},
{
"epoch": 1.0189169675090253,
"grad_norm": 0.02776011824607849,
"learning_rate": 4.909114308553034e-06,
"loss": 0.7902,
"step": 882
},
{
"epoch": 1.020072202166065,
"grad_norm": 0.026471663266420364,
"learning_rate": 4.909934505962614e-06,
"loss": 0.8,
"step": 883
},
{
"epoch": 1.0212274368231047,
"grad_norm": 0.026362843811511993,
"learning_rate": 4.9107537750217884e-06,
"loss": 0.7957,
"step": 884
},
{
"epoch": 1.0223826714801445,
"grad_norm": 0.027810046449303627,
"learning_rate": 4.9115721178297095e-06,
"loss": 0.7986,
"step": 885
},
{
"epoch": 1.023537906137184,
"grad_norm": 0.02654910273849964,
"learning_rate": 4.9123895364784185e-06,
"loss": 0.7611,
"step": 886
},
{
"epoch": 1.024693140794224,
"grad_norm": 0.026264095678925514,
"learning_rate": 4.913206033052878e-06,
"loss": 0.808,
"step": 887
},
{
"epoch": 1.0258483754512635,
"grad_norm": 0.02679434046149254,
"learning_rate": 4.914021609631002e-06,
"loss": 0.7432,
"step": 888
},
{
"epoch": 1.0270036101083033,
"grad_norm": 0.026940811425447464,
"learning_rate": 4.914836268283691e-06,
"loss": 0.7872,
"step": 889
},
{
"epoch": 1.028158844765343,
"grad_norm": 0.025167059153318405,
"learning_rate": 4.915650011074855e-06,
"loss": 0.7876,
"step": 890
},
{
"epoch": 1.0293140794223827,
"grad_norm": 0.026694072410464287,
"learning_rate": 4.916462840061458e-06,
"loss": 0.7919,
"step": 891
},
{
"epoch": 1.0304693140794223,
"grad_norm": 0.02730594575405121,
"learning_rate": 4.917274757293539e-06,
"loss": 0.7961,
"step": 892
},
{
"epoch": 1.0316245487364621,
"grad_norm": 0.026232946664094925,
"learning_rate": 4.918085764814244e-06,
"loss": 0.7908,
"step": 893
},
{
"epoch": 1.0327797833935017,
"grad_norm": 0.027884062379598618,
"learning_rate": 4.918895864659863e-06,
"loss": 0.7752,
"step": 894
},
{
"epoch": 1.0339350180505416,
"grad_norm": 0.027524368837475777,
"learning_rate": 4.919705058859854e-06,
"loss": 0.8114,
"step": 895
},
{
"epoch": 1.0350902527075811,
"grad_norm": 0.02609703689813614,
"learning_rate": 4.920513349436876e-06,
"loss": 0.801,
"step": 896
},
{
"epoch": 1.036245487364621,
"grad_norm": 0.027548260986804962,
"learning_rate": 4.921320738406821e-06,
"loss": 0.7736,
"step": 897
},
{
"epoch": 1.0374007220216606,
"grad_norm": 0.028695981949567795,
"learning_rate": 4.922127227778841e-06,
"loss": 0.7968,
"step": 898
},
{
"epoch": 1.0385559566787004,
"grad_norm": 0.02490762434899807,
"learning_rate": 4.922932819555381e-06,
"loss": 0.7695,
"step": 899
},
{
"epoch": 1.03971119133574,
"grad_norm": 0.030678994953632355,
"learning_rate": 4.923737515732209e-06,
"loss": 0.7764,
"step": 900
},
{
"epoch": 1.0408664259927798,
"grad_norm": 0.02829810604453087,
"learning_rate": 4.924541318298439e-06,
"loss": 0.7881,
"step": 901
},
{
"epoch": 1.0420216606498194,
"grad_norm": 0.028521962463855743,
"learning_rate": 4.925344229236571e-06,
"loss": 0.791,
"step": 902
},
{
"epoch": 1.0431768953068592,
"grad_norm": 0.028740065172314644,
"learning_rate": 4.926146250522511e-06,
"loss": 0.778,
"step": 903
},
{
"epoch": 1.044332129963899,
"grad_norm": 0.02634969726204872,
"learning_rate": 4.926947384125606e-06,
"loss": 0.7784,
"step": 904
},
{
"epoch": 1.0454873646209386,
"grad_norm": 0.02609422616660595,
"learning_rate": 4.927747632008672e-06,
"loss": 0.7729,
"step": 905
},
{
"epoch": 1.0466425992779784,
"grad_norm": 0.025464767590165138,
"learning_rate": 4.928546996128022e-06,
"loss": 0.7846,
"step": 906
},
{
"epoch": 1.047797833935018,
"grad_norm": 0.026383783668279648,
"learning_rate": 4.929345478433492e-06,
"loss": 0.7672,
"step": 907
},
{
"epoch": 1.0489530685920578,
"grad_norm": 0.02692020870745182,
"learning_rate": 4.930143080868476e-06,
"loss": 0.777,
"step": 908
},
{
"epoch": 1.0501083032490974,
"grad_norm": 0.026965700089931488,
"learning_rate": 4.930939805369947e-06,
"loss": 0.8104,
"step": 909
},
{
"epoch": 1.0512635379061372,
"grad_norm": 0.0277117807418108,
"learning_rate": 4.931735653868489e-06,
"loss": 0.7729,
"step": 910
},
{
"epoch": 1.0524187725631768,
"grad_norm": 0.027171596884727478,
"learning_rate": 4.932530628288331e-06,
"loss": 0.7957,
"step": 911
},
{
"epoch": 1.0535740072202167,
"grad_norm": 0.0277806855738163,
"learning_rate": 4.933324730547361e-06,
"loss": 0.7988,
"step": 912
},
{
"epoch": 1.0547292418772563,
"grad_norm": 0.02666284516453743,
"learning_rate": 4.934117962557165e-06,
"loss": 0.7637,
"step": 913
},
{
"epoch": 1.055884476534296,
"grad_norm": 0.028444141149520874,
"learning_rate": 4.934910326223052e-06,
"loss": 0.77,
"step": 914
},
{
"epoch": 1.0570397111913357,
"grad_norm": 0.028864728286862373,
"learning_rate": 4.93570182344408e-06,
"loss": 0.7778,
"step": 915
},
{
"epoch": 1.0581949458483755,
"grad_norm": 0.026793263852596283,
"learning_rate": 4.936492456113085e-06,
"loss": 0.7879,
"step": 916
},
{
"epoch": 1.059350180505415,
"grad_norm": 0.027911733835935593,
"learning_rate": 4.937282226116702e-06,
"loss": 0.7698,
"step": 917
},
{
"epoch": 1.060505415162455,
"grad_norm": 0.026382414624094963,
"learning_rate": 4.938071135335405e-06,
"loss": 0.7637,
"step": 918
},
{
"epoch": 1.0616606498194945,
"grad_norm": 0.027525540441274643,
"learning_rate": 4.938859185643519e-06,
"loss": 0.8057,
"step": 919
},
{
"epoch": 1.0628158844765343,
"grad_norm": 0.028536858037114143,
"learning_rate": 4.93964637890926e-06,
"loss": 0.7773,
"step": 920
},
{
"epoch": 1.0639711191335741,
"grad_norm": 0.026540234684944153,
"learning_rate": 4.9404327169947486e-06,
"loss": 0.7912,
"step": 921
},
{
"epoch": 1.0651263537906137,
"grad_norm": 0.026590172201395035,
"learning_rate": 4.941218201756049e-06,
"loss": 0.7873,
"step": 922
},
{
"epoch": 1.0662815884476535,
"grad_norm": 0.026739628985524178,
"learning_rate": 4.942002835043187e-06,
"loss": 0.7853,
"step": 923
},
{
"epoch": 1.0674368231046931,
"grad_norm": 0.027753658592700958,
"learning_rate": 4.9427866187001785e-06,
"loss": 0.7853,
"step": 924
},
{
"epoch": 1.068592057761733,
"grad_norm": 0.0267436932772398,
"learning_rate": 4.943569554565055e-06,
"loss": 0.7851,
"step": 925
},
{
"epoch": 1.0697472924187725,
"grad_norm": 0.02634557895362377,
"learning_rate": 4.9443516444698915e-06,
"loss": 0.7881,
"step": 926
},
{
"epoch": 1.0709025270758123,
"grad_norm": 0.027419744059443474,
"learning_rate": 4.945132890240829e-06,
"loss": 0.813,
"step": 927
},
{
"epoch": 1.072057761732852,
"grad_norm": 0.026219435036182404,
"learning_rate": 4.9459132936981044e-06,
"loss": 0.7998,
"step": 928
},
{
"epoch": 1.0732129963898918,
"grad_norm": 0.026673052459955215,
"learning_rate": 4.94669285665607e-06,
"loss": 0.764,
"step": 929
},
{
"epoch": 1.0743682310469314,
"grad_norm": 0.027227727696299553,
"learning_rate": 4.9474715809232256e-06,
"loss": 0.7734,
"step": 930
},
{
"epoch": 1.0755234657039712,
"grad_norm": 0.025797106325626373,
"learning_rate": 4.948249468302239e-06,
"loss": 0.7959,
"step": 931
},
{
"epoch": 1.0766787003610108,
"grad_norm": 0.026355307549238205,
"learning_rate": 4.9490265205899695e-06,
"loss": 0.7987,
"step": 932
},
{
"epoch": 1.0778339350180506,
"grad_norm": 0.02700088918209076,
"learning_rate": 4.949802739577501e-06,
"loss": 0.8067,
"step": 933
},
{
"epoch": 1.0789891696750902,
"grad_norm": 0.02693251334130764,
"learning_rate": 4.950578127050157e-06,
"loss": 0.7855,
"step": 934
},
{
"epoch": 1.08014440433213,
"grad_norm": 0.02637103572487831,
"learning_rate": 4.95135268478753e-06,
"loss": 0.7821,
"step": 935
},
{
"epoch": 1.0812996389891696,
"grad_norm": 0.02696199156343937,
"learning_rate": 4.952126414563509e-06,
"loss": 0.7954,
"step": 936
},
{
"epoch": 1.0824548736462094,
"grad_norm": 0.025795504450798035,
"learning_rate": 4.952899318146298e-06,
"loss": 0.771,
"step": 937
},
{
"epoch": 1.083610108303249,
"grad_norm": 0.0262451134622097,
"learning_rate": 4.953671397298441e-06,
"loss": 0.7737,
"step": 938
},
{
"epoch": 1.0847653429602888,
"grad_norm": 0.027646278962492943,
"learning_rate": 4.954442653776852e-06,
"loss": 0.7702,
"step": 939
},
{
"epoch": 1.0859205776173284,
"grad_norm": 0.026158776134252548,
"learning_rate": 4.955213089332832e-06,
"loss": 0.772,
"step": 940
},
{
"epoch": 1.0870758122743682,
"grad_norm": 0.02821963280439377,
"learning_rate": 4.955982705712095e-06,
"loss": 0.7828,
"step": 941
},
{
"epoch": 1.088231046931408,
"grad_norm": 0.026813939213752747,
"learning_rate": 4.956751504654796e-06,
"loss": 0.7848,
"step": 942
},
{
"epoch": 1.0893862815884476,
"grad_norm": 0.025459513068199158,
"learning_rate": 4.957519487895548e-06,
"loss": 0.7655,
"step": 943
},
{
"epoch": 1.0905415162454875,
"grad_norm": 0.026770230382680893,
"learning_rate": 4.958286657163448e-06,
"loss": 0.773,
"step": 944
},
{
"epoch": 1.091696750902527,
"grad_norm": 0.02704155258834362,
"learning_rate": 4.959053014182106e-06,
"loss": 0.7755,
"step": 945
},
{
"epoch": 1.0928519855595669,
"grad_norm": 0.02700476534664631,
"learning_rate": 4.959818560669656e-06,
"loss": 0.8009,
"step": 946
},
{
"epoch": 1.0940072202166065,
"grad_norm": 0.028186574578285217,
"learning_rate": 4.96058329833879e-06,
"loss": 0.7705,
"step": 947
},
{
"epoch": 1.0951624548736463,
"grad_norm": 0.027806002646684647,
"learning_rate": 4.9613472288967775e-06,
"loss": 0.7727,
"step": 948
},
{
"epoch": 1.0963176895306859,
"grad_norm": 0.028479190543293953,
"learning_rate": 4.962110354045489e-06,
"loss": 0.7921,
"step": 949
},
{
"epoch": 1.0974729241877257,
"grad_norm": 0.027802351862192154,
"learning_rate": 4.962872675481414e-06,
"loss": 0.7827,
"step": 950
},
{
"epoch": 1.0986281588447653,
"grad_norm": 0.026106078177690506,
"learning_rate": 4.96363419489569e-06,
"loss": 0.7663,
"step": 951
},
{
"epoch": 1.099783393501805,
"grad_norm": 0.027961455285549164,
"learning_rate": 4.964394913974124e-06,
"loss": 0.7735,
"step": 952
},
{
"epoch": 1.1009386281588447,
"grad_norm": 0.02648748643696308,
"learning_rate": 4.965154834397211e-06,
"loss": 0.8092,
"step": 953
},
{
"epoch": 1.1020938628158845,
"grad_norm": 0.026987139135599136,
"learning_rate": 4.965913957840159e-06,
"loss": 0.791,
"step": 954
},
{
"epoch": 1.103249097472924,
"grad_norm": 0.026114538311958313,
"learning_rate": 4.966672285972911e-06,
"loss": 0.7682,
"step": 955
},
{
"epoch": 1.104404332129964,
"grad_norm": 0.027255605906248093,
"learning_rate": 4.967429820460167e-06,
"loss": 0.7934,
"step": 956
},
{
"epoch": 1.1055595667870035,
"grad_norm": 0.026975391432642937,
"learning_rate": 4.968186562961406e-06,
"loss": 0.7758,
"step": 957
},
{
"epoch": 1.1067148014440433,
"grad_norm": 0.026317190378904343,
"learning_rate": 4.968942515130908e-06,
"loss": 0.7949,
"step": 958
},
{
"epoch": 1.1078700361010831,
"grad_norm": 0.027674200013279915,
"learning_rate": 4.969697678617774e-06,
"loss": 0.801,
"step": 959
},
{
"epoch": 1.1090252707581227,
"grad_norm": 0.026397835463285446,
"learning_rate": 4.970452055065948e-06,
"loss": 0.7872,
"step": 960
},
{
"epoch": 1.1101805054151626,
"grad_norm": 0.025970855727791786,
"learning_rate": 4.971205646114243e-06,
"loss": 0.8076,
"step": 961
},
{
"epoch": 1.1113357400722021,
"grad_norm": 0.026946989819407463,
"learning_rate": 4.9719584533963555e-06,
"loss": 0.7646,
"step": 962
},
{
"epoch": 1.112490974729242,
"grad_norm": 0.026165366172790527,
"learning_rate": 4.9727104785408915e-06,
"loss": 0.7704,
"step": 963
},
{
"epoch": 1.1136462093862816,
"grad_norm": 0.026523860171437263,
"learning_rate": 4.973461723171385e-06,
"loss": 0.7787,
"step": 964
},
{
"epoch": 1.1148014440433214,
"grad_norm": 0.02800065465271473,
"learning_rate": 4.974212188906321e-06,
"loss": 0.7824,
"step": 965
},
{
"epoch": 1.115956678700361,
"grad_norm": 0.02660745568573475,
"learning_rate": 4.974961877359156e-06,
"loss": 0.8069,
"step": 966
},
{
"epoch": 1.1171119133574008,
"grad_norm": 0.026590686291456223,
"learning_rate": 4.975710790138337e-06,
"loss": 0.7603,
"step": 967
},
{
"epoch": 1.1182671480144404,
"grad_norm": 0.028299605473876,
"learning_rate": 4.976458928847323e-06,
"loss": 0.7884,
"step": 968
},
{
"epoch": 1.1194223826714802,
"grad_norm": 0.02655804343521595,
"learning_rate": 4.977206295084609e-06,
"loss": 0.7902,
"step": 969
},
{
"epoch": 1.1205776173285198,
"grad_norm": 0.02828875742852688,
"learning_rate": 4.977952890443742e-06,
"loss": 0.7927,
"step": 970
},
{
"epoch": 1.1217328519855596,
"grad_norm": 0.03140028193593025,
"learning_rate": 4.978698716513342e-06,
"loss": 0.8075,
"step": 971
},
{
"epoch": 1.1228880866425992,
"grad_norm": 0.028589541092514992,
"learning_rate": 4.9794437748771245e-06,
"loss": 0.767,
"step": 972
},
{
"epoch": 1.124043321299639,
"grad_norm": 0.028249988332390785,
"learning_rate": 4.98018806711392e-06,
"loss": 0.8103,
"step": 973
},
{
"epoch": 1.1251985559566786,
"grad_norm": 0.027847470715641975,
"learning_rate": 4.980931594797693e-06,
"loss": 0.7939,
"step": 974
},
{
"epoch": 1.1263537906137184,
"grad_norm": 0.02874613367021084,
"learning_rate": 4.981674359497562e-06,
"loss": 0.7986,
"step": 975
},
{
"epoch": 1.1275090252707582,
"grad_norm": 0.02685857191681862,
"learning_rate": 4.98241636277782e-06,
"loss": 0.7765,
"step": 976
},
{
"epoch": 1.1286642599277978,
"grad_norm": 0.026808686554431915,
"learning_rate": 4.9831576061979556e-06,
"loss": 0.7786,
"step": 977
},
{
"epoch": 1.1298194945848374,
"grad_norm": 0.02733052521944046,
"learning_rate": 4.98389809131267e-06,
"loss": 0.7663,
"step": 978
},
{
"epoch": 1.1309747292418773,
"grad_norm": 0.0273138340562582,
"learning_rate": 4.984637819671897e-06,
"loss": 0.8018,
"step": 979
},
{
"epoch": 1.132129963898917,
"grad_norm": 0.02772151120007038,
"learning_rate": 4.985376792820825e-06,
"loss": 0.8052,
"step": 980
},
{
"epoch": 1.1332851985559567,
"grad_norm": 0.0273757204413414,
"learning_rate": 4.986115012299915e-06,
"loss": 0.7608,
"step": 981
},
{
"epoch": 1.1344404332129965,
"grad_norm": 0.026899464428424835,
"learning_rate": 4.986852479644916e-06,
"loss": 0.8069,
"step": 982
},
{
"epoch": 1.135595667870036,
"grad_norm": 0.026683615520596504,
"learning_rate": 4.987589196386893e-06,
"loss": 0.7694,
"step": 983
},
{
"epoch": 1.136750902527076,
"grad_norm": 0.028294900432229042,
"learning_rate": 4.988325164052236e-06,
"loss": 0.7738,
"step": 984
},
{
"epoch": 1.1379061371841155,
"grad_norm": 0.026210768148303032,
"learning_rate": 4.989060384162687e-06,
"loss": 0.7624,
"step": 985
},
{
"epoch": 1.1390613718411553,
"grad_norm": 0.026641108095645905,
"learning_rate": 4.989794858235352e-06,
"loss": 0.7827,
"step": 986
},
{
"epoch": 1.140216606498195,
"grad_norm": 0.027551405131816864,
"learning_rate": 4.990528587782728e-06,
"loss": 0.7849,
"step": 987
},
{
"epoch": 1.1413718411552347,
"grad_norm": 0.026633942499756813,
"learning_rate": 4.991261574312715e-06,
"loss": 0.8019,
"step": 988
},
{
"epoch": 1.1425270758122743,
"grad_norm": 0.027740631252527237,
"learning_rate": 4.991993819328633e-06,
"loss": 0.8075,
"step": 989
},
{
"epoch": 1.1436823104693141,
"grad_norm": 0.027059046551585197,
"learning_rate": 4.992725324329251e-06,
"loss": 0.7916,
"step": 990
},
{
"epoch": 1.1448375451263537,
"grad_norm": 0.027994418516755104,
"learning_rate": 4.993456090808793e-06,
"loss": 0.7943,
"step": 991
},
{
"epoch": 1.1459927797833935,
"grad_norm": 0.026348162442445755,
"learning_rate": 4.994186120256964e-06,
"loss": 0.7993,
"step": 992
},
{
"epoch": 1.1471480144404331,
"grad_norm": 0.027533039450645447,
"learning_rate": 4.99491541415897e-06,
"loss": 0.7806,
"step": 993
},
{
"epoch": 1.148303249097473,
"grad_norm": 0.02699647657573223,
"learning_rate": 4.995643973995523e-06,
"loss": 0.7753,
"step": 994
},
{
"epoch": 1.1494584837545125,
"grad_norm": 0.025739947333931923,
"learning_rate": 4.996371801242876e-06,
"loss": 0.7487,
"step": 995
},
{
"epoch": 1.1506137184115524,
"grad_norm": 0.026720624417066574,
"learning_rate": 4.997098897372832e-06,
"loss": 0.7776,
"step": 996
},
{
"epoch": 1.1517689530685922,
"grad_norm": 0.027732260525226593,
"learning_rate": 4.99782526385276e-06,
"loss": 0.7718,
"step": 997
},
{
"epoch": 1.1529241877256318,
"grad_norm": 0.026213763281702995,
"learning_rate": 4.998550902145619e-06,
"loss": 0.7849,
"step": 998
},
{
"epoch": 1.1540794223826714,
"grad_norm": 0.0285005122423172,
"learning_rate": 4.999275813709971e-06,
"loss": 0.7944,
"step": 999
},
{
"epoch": 1.1552346570397112,
"grad_norm": 0.025959184393286705,
"learning_rate": 5e-06,
"loss": 0.8039,
"step": 1000
},
{
"epoch": 1.156389891696751,
"grad_norm": 0.026393594220280647,
"learning_rate": 4.999976851655541e-06,
"loss": 0.792,
"step": 1001
},
{
"epoch": 1.1575451263537906,
"grad_norm": 0.02741307206451893,
"learning_rate": 4.999907407050885e-06,
"loss": 0.7816,
"step": 1002
},
{
"epoch": 1.1587003610108304,
"grad_norm": 0.02743341587483883,
"learning_rate": 4.999791667472181e-06,
"loss": 0.7473,
"step": 1003
},
{
"epoch": 1.15985559566787,
"grad_norm": 0.025454385206103325,
"learning_rate": 4.999629635062988e-06,
"loss": 0.7964,
"step": 1004
},
{
"epoch": 1.1610108303249098,
"grad_norm": 0.029390884563326836,
"learning_rate": 4.999421312824233e-06,
"loss": 0.7885,
"step": 1005
},
{
"epoch": 1.1621660649819494,
"grad_norm": 0.026846162974834442,
"learning_rate": 4.99916670461415e-06,
"loss": 0.7553,
"step": 1006
},
{
"epoch": 1.1633212996389892,
"grad_norm": 0.027146577835083008,
"learning_rate": 4.998865815148222e-06,
"loss": 0.757,
"step": 1007
},
{
"epoch": 1.1644765342960288,
"grad_norm": 0.028116164728999138,
"learning_rate": 4.998518649999077e-06,
"loss": 0.793,
"step": 1008
},
{
"epoch": 1.1656317689530686,
"grad_norm": 0.027374397963285446,
"learning_rate": 4.998125215596399e-06,
"loss": 0.7684,
"step": 1009
},
{
"epoch": 1.1667870036101082,
"grad_norm": 0.026328133419156075,
"learning_rate": 4.9976855192268e-06,
"loss": 0.794,
"step": 1010
},
{
"epoch": 1.167942238267148,
"grad_norm": 0.028758594766259193,
"learning_rate": 4.997199569033688e-06,
"loss": 0.7872,
"step": 1011
},
{
"epoch": 1.1690974729241876,
"grad_norm": 0.026665737852454185,
"learning_rate": 4.996667374017118e-06,
"loss": 0.797,
"step": 1012
},
{
"epoch": 1.1702527075812275,
"grad_norm": 0.02736995927989483,
"learning_rate": 4.996088944033622e-06,
"loss": 0.7912,
"step": 1013
},
{
"epoch": 1.171407942238267,
"grad_norm": 0.027897171676158905,
"learning_rate": 4.995464289796029e-06,
"loss": 0.7714,
"step": 1014
},
{
"epoch": 1.1725631768953069,
"grad_norm": 0.027566097676753998,
"learning_rate": 4.994793422873265e-06,
"loss": 0.7755,
"step": 1015
},
{
"epoch": 1.1737184115523465,
"grad_norm": 0.026376303285360336,
"learning_rate": 4.994076355690139e-06,
"loss": 0.8052,
"step": 1016
},
{
"epoch": 1.1748736462093863,
"grad_norm": 0.027220716699957848,
"learning_rate": 4.993313101527112e-06,
"loss": 0.7955,
"step": 1017
},
{
"epoch": 1.176028880866426,
"grad_norm": 0.02771547995507717,
"learning_rate": 4.992503674520057e-06,
"loss": 0.7595,
"step": 1018
},
{
"epoch": 1.1771841155234657,
"grad_norm": 0.026792939752340317,
"learning_rate": 4.991648089659987e-06,
"loss": 0.7812,
"step": 1019
},
{
"epoch": 1.1783393501805055,
"grad_norm": 0.026456760242581367,
"learning_rate": 4.990746362792786e-06,
"loss": 0.7884,
"step": 1020
},
{
"epoch": 1.179494584837545,
"grad_norm": 0.028625672683119774,
"learning_rate": 4.989798510618911e-06,
"loss": 0.7999,
"step": 1021
},
{
"epoch": 1.180649819494585,
"grad_norm": 0.026891304180026054,
"learning_rate": 4.988804550693086e-06,
"loss": 0.7949,
"step": 1022
},
{
"epoch": 1.1818050541516245,
"grad_norm": 0.025918442755937576,
"learning_rate": 4.9877645014239686e-06,
"loss": 0.793,
"step": 1023
},
{
"epoch": 1.1829602888086643,
"grad_norm": 0.026719292625784874,
"learning_rate": 4.9866783820738255e-06,
"loss": 0.7915,
"step": 1024
},
{
"epoch": 1.184115523465704,
"grad_norm": 0.027593158185482025,
"learning_rate": 4.985546212758157e-06,
"loss": 0.7678,
"step": 1025
},
{
"epoch": 1.1852707581227437,
"grad_norm": 0.02738940343260765,
"learning_rate": 4.984368014445336e-06,
"loss": 0.79,
"step": 1026
},
{
"epoch": 1.1864259927797833,
"grad_norm": 0.02584446780383587,
"learning_rate": 4.983143808956219e-06,
"loss": 0.7638,
"step": 1027
},
{
"epoch": 1.1875812274368232,
"grad_norm": 0.03016565926373005,
"learning_rate": 4.981873618963737e-06,
"loss": 0.7906,
"step": 1028
},
{
"epoch": 1.1887364620938627,
"grad_norm": 0.026769449934363365,
"learning_rate": 4.980557467992476e-06,
"loss": 0.7743,
"step": 1029
},
{
"epoch": 1.1898916967509026,
"grad_norm": 0.02779136225581169,
"learning_rate": 4.97919538041825e-06,
"loss": 0.7917,
"step": 1030
},
{
"epoch": 1.1910469314079422,
"grad_norm": 0.028257260099053383,
"learning_rate": 4.977787381467638e-06,
"loss": 0.774,
"step": 1031
},
{
"epoch": 1.192202166064982,
"grad_norm": 0.026401590555906296,
"learning_rate": 4.976333497217524e-06,
"loss": 0.7989,
"step": 1032
},
{
"epoch": 1.1933574007220216,
"grad_norm": 0.028177831321954727,
"learning_rate": 4.974833754594611e-06,
"loss": 0.7832,
"step": 1033
},
{
"epoch": 1.1945126353790614,
"grad_norm": 0.028338629752397537,
"learning_rate": 4.9732881813749236e-06,
"loss": 0.7991,
"step": 1034
},
{
"epoch": 1.1956678700361012,
"grad_norm": 0.027354028075933456,
"learning_rate": 4.9716968061832935e-06,
"loss": 0.7917,
"step": 1035
},
{
"epoch": 1.1968231046931408,
"grad_norm": 0.02721875160932541,
"learning_rate": 4.9700596584928275e-06,
"loss": 0.7765,
"step": 1036
},
{
"epoch": 1.1979783393501804,
"grad_norm": 0.026294628158211708,
"learning_rate": 4.968376768624368e-06,
"loss": 0.7871,
"step": 1037
},
{
"epoch": 1.1991335740072202,
"grad_norm": 0.027313537895679474,
"learning_rate": 4.96664816774592e-06,
"loss": 0.7858,
"step": 1038
},
{
"epoch": 1.20028880866426,
"grad_norm": 0.02785896323621273,
"learning_rate": 4.964873887872085e-06,
"loss": 0.7938,
"step": 1039
},
{
"epoch": 1.2014440433212996,
"grad_norm": 0.02687663771212101,
"learning_rate": 4.963053961863463e-06,
"loss": 0.7776,
"step": 1040
},
{
"epoch": 1.2025992779783394,
"grad_norm": 0.026713555678725243,
"learning_rate": 4.961188423426042e-06,
"loss": 0.7741,
"step": 1041
},
{
"epoch": 1.203754512635379,
"grad_norm": 0.02718759886920452,
"learning_rate": 4.959277307110582e-06,
"loss": 0.7667,
"step": 1042
},
{
"epoch": 1.2049097472924188,
"grad_norm": 0.027160272002220154,
"learning_rate": 4.957320648311961e-06,
"loss": 0.7894,
"step": 1043
},
{
"epoch": 1.2060649819494584,
"grad_norm": 0.02789607271552086,
"learning_rate": 4.955318483268535e-06,
"loss": 0.7812,
"step": 1044
},
{
"epoch": 1.2072202166064983,
"grad_norm": 0.02740464359521866,
"learning_rate": 4.953270849061456e-06,
"loss": 0.7585,
"step": 1045
},
{
"epoch": 1.2083754512635378,
"grad_norm": 0.028817661106586456,
"learning_rate": 4.951177783613991e-06,
"loss": 0.8091,
"step": 1046
},
{
"epoch": 1.2095306859205777,
"grad_norm": 0.026185913011431694,
"learning_rate": 4.9490393256908144e-06,
"loss": 0.7956,
"step": 1047
},
{
"epoch": 1.2106859205776173,
"grad_norm": 0.029244182631373405,
"learning_rate": 4.946855514897297e-06,
"loss": 0.7779,
"step": 1048
},
{
"epoch": 1.211841155234657,
"grad_norm": 0.027704982087016106,
"learning_rate": 4.944626391678766e-06,
"loss": 0.7993,
"step": 1049
},
{
"epoch": 1.2129963898916967,
"grad_norm": 0.02663758210837841,
"learning_rate": 4.942351997319761e-06,
"loss": 0.812,
"step": 1050
},
{
"epoch": 1.2141516245487365,
"grad_norm": 0.026222404092550278,
"learning_rate": 4.940032373943262e-06,
"loss": 0.7558,
"step": 1051
},
{
"epoch": 1.215306859205776,
"grad_norm": 0.02578769624233246,
"learning_rate": 4.937667564509923e-06,
"loss": 0.7781,
"step": 1052
},
{
"epoch": 1.216462093862816,
"grad_norm": 0.02732442319393158,
"learning_rate": 4.935257612817259e-06,
"loss": 0.7916,
"step": 1053
},
{
"epoch": 1.2176173285198555,
"grad_norm": 0.029004095122218132,
"learning_rate": 4.932802563498847e-06,
"loss": 0.7979,
"step": 1054
},
{
"epoch": 1.2187725631768953,
"grad_norm": 0.02736952155828476,
"learning_rate": 4.9303024620234966e-06,
"loss": 0.7984,
"step": 1055
},
{
"epoch": 1.2199277978339351,
"grad_norm": 0.026861999183893204,
"learning_rate": 4.927757354694406e-06,
"loss": 0.7986,
"step": 1056
},
{
"epoch": 1.2210830324909747,
"grad_norm": 0.0265817791223526,
"learning_rate": 4.925167288648305e-06,
"loss": 0.7763,
"step": 1057
},
{
"epoch": 1.2222382671480145,
"grad_norm": 0.028195342049002647,
"learning_rate": 4.922532311854585e-06,
"loss": 0.7967,
"step": 1058
},
{
"epoch": 1.2233935018050541,
"grad_norm": 0.028484554961323738,
"learning_rate": 4.919852473114405e-06,
"loss": 0.7926,
"step": 1059
},
{
"epoch": 1.224548736462094,
"grad_norm": 0.025893505662679672,
"learning_rate": 4.917127822059792e-06,
"loss": 0.783,
"step": 1060
},
{
"epoch": 1.2257039711191335,
"grad_norm": 0.028708558529615402,
"learning_rate": 4.9143584091527236e-06,
"loss": 0.7954,
"step": 1061
},
{
"epoch": 1.2268592057761734,
"grad_norm": 0.025731157511472702,
"learning_rate": 4.911544285684186e-06,
"loss": 0.7875,
"step": 1062
},
{
"epoch": 1.228014440433213,
"grad_norm": 0.02737441286444664,
"learning_rate": 4.9086855037732315e-06,
"loss": 0.7618,
"step": 1063
},
{
"epoch": 1.2291696750902528,
"grad_norm": 0.026509573683142662,
"learning_rate": 4.9057821163660085e-06,
"loss": 0.7513,
"step": 1064
},
{
"epoch": 1.2303249097472924,
"grad_norm": 0.027171408757567406,
"learning_rate": 4.9028341772347845e-06,
"loss": 0.7825,
"step": 1065
},
{
"epoch": 1.2314801444043322,
"grad_norm": 0.02830067090690136,
"learning_rate": 4.899841740976947e-06,
"loss": 0.7784,
"step": 1066
},
{
"epoch": 1.2326353790613718,
"grad_norm": 0.02830381691455841,
"learning_rate": 4.8968048630139945e-06,
"loss": 0.8057,
"step": 1067
},
{
"epoch": 1.2337906137184116,
"grad_norm": 0.027096690610051155,
"learning_rate": 4.89372359959051e-06,
"loss": 0.778,
"step": 1068
},
{
"epoch": 1.2349458483754512,
"grad_norm": 0.026616832241415977,
"learning_rate": 4.890598007773116e-06,
"loss": 0.7719,
"step": 1069
},
{
"epoch": 1.236101083032491,
"grad_norm": 0.02842404507100582,
"learning_rate": 4.887428145449425e-06,
"loss": 0.7898,
"step": 1070
},
{
"epoch": 1.2372563176895306,
"grad_norm": 0.029439568519592285,
"learning_rate": 4.884214071326957e-06,
"loss": 0.7586,
"step": 1071
},
{
"epoch": 1.2384115523465704,
"grad_norm": 0.027274589985609055,
"learning_rate": 4.880955844932061e-06,
"loss": 0.7779,
"step": 1072
},
{
"epoch": 1.2395667870036102,
"grad_norm": 0.030161483213305473,
"learning_rate": 4.877653526608811e-06,
"loss": 0.7774,
"step": 1073
},
{
"epoch": 1.2407220216606498,
"grad_norm": 0.026620803400874138,
"learning_rate": 4.874307177517882e-06,
"loss": 0.7714,
"step": 1074
},
{
"epoch": 1.2418772563176894,
"grad_norm": 0.031550221145153046,
"learning_rate": 4.870916859635426e-06,
"loss": 0.7951,
"step": 1075
},
{
"epoch": 1.2430324909747292,
"grad_norm": 0.02677975781261921,
"learning_rate": 4.86748263575192e-06,
"loss": 0.8094,
"step": 1076
},
{
"epoch": 1.244187725631769,
"grad_norm": 0.02815908007323742,
"learning_rate": 4.864004569471e-06,
"loss": 0.7828,
"step": 1077
},
{
"epoch": 1.2453429602888086,
"grad_norm": 0.028923533856868744,
"learning_rate": 4.860482725208291e-06,
"loss": 0.8106,
"step": 1078
},
{
"epoch": 1.2464981949458485,
"grad_norm": 0.026107341051101685,
"learning_rate": 4.856917168190205e-06,
"loss": 0.7893,
"step": 1079
},
{
"epoch": 1.247653429602888,
"grad_norm": 0.029115671291947365,
"learning_rate": 4.853307964452739e-06,
"loss": 0.749,
"step": 1080
},
{
"epoch": 1.2488086642599279,
"grad_norm": 0.02833879366517067,
"learning_rate": 4.849655180840253e-06,
"loss": 0.8041,
"step": 1081
},
{
"epoch": 1.2499638989169675,
"grad_norm": 0.027344336733222008,
"learning_rate": 4.8459588850042255e-06,
"loss": 0.7455,
"step": 1082
},
{
"epoch": 1.2511191335740073,
"grad_norm": 0.02860669605433941,
"learning_rate": 4.8422191454020055e-06,
"loss": 0.7995,
"step": 1083
},
{
"epoch": 1.2522743682310469,
"grad_norm": 0.026610156521201134,
"learning_rate": 4.838436031295543e-06,
"loss": 0.7776,
"step": 1084
},
{
"epoch": 1.2534296028880867,
"grad_norm": 0.029180224984884262,
"learning_rate": 4.834609612750108e-06,
"loss": 0.7755,
"step": 1085
},
{
"epoch": 1.2545848375451263,
"grad_norm": 0.025666316971182823,
"learning_rate": 4.8307399606329906e-06,
"loss": 0.7815,
"step": 1086
},
{
"epoch": 1.255740072202166,
"grad_norm": 0.028815045952796936,
"learning_rate": 4.82682714661219e-06,
"loss": 0.7618,
"step": 1087
},
{
"epoch": 1.2568953068592057,
"grad_norm": 0.02659596875309944,
"learning_rate": 4.822871243155085e-06,
"loss": 0.7906,
"step": 1088
},
{
"epoch": 1.2580505415162455,
"grad_norm": 0.027543647214770317,
"learning_rate": 4.818872323527097e-06,
"loss": 0.7871,
"step": 1089
},
{
"epoch": 1.2592057761732853,
"grad_norm": 0.027373237535357475,
"learning_rate": 4.814830461790327e-06,
"loss": 0.763,
"step": 1090
},
{
"epoch": 1.260361010830325,
"grad_norm": 0.025914175435900688,
"learning_rate": 4.810745732802186e-06,
"loss": 0.7744,
"step": 1091
},
{
"epoch": 1.2615162454873645,
"grad_norm": 0.026502788066864014,
"learning_rate": 4.806618212214009e-06,
"loss": 0.7901,
"step": 1092
},
{
"epoch": 1.2626714801444043,
"grad_norm": 0.026811202988028526,
"learning_rate": 4.802447976469658e-06,
"loss": 0.797,
"step": 1093
},
{
"epoch": 1.2638267148014442,
"grad_norm": 0.027904195711016655,
"learning_rate": 4.7982351028040966e-06,
"loss": 0.7956,
"step": 1094
},
{
"epoch": 1.2649819494584837,
"grad_norm": 0.025807814672589302,
"learning_rate": 4.79397966924197e-06,
"loss": 0.7785,
"step": 1095
},
{
"epoch": 1.2661371841155233,
"grad_norm": 0.02669427916407585,
"learning_rate": 4.789681754596152e-06,
"loss": 0.7882,
"step": 1096
},
{
"epoch": 1.2672924187725632,
"grad_norm": 0.026608362793922424,
"learning_rate": 4.785341438466289e-06,
"loss": 0.7716,
"step": 1097
},
{
"epoch": 1.268447653429603,
"grad_norm": 0.025953490287065506,
"learning_rate": 4.7809588012373255e-06,
"loss": 0.7702,
"step": 1098
},
{
"epoch": 1.2696028880866426,
"grad_norm": 0.02710854634642601,
"learning_rate": 4.776533924078018e-06,
"loss": 0.7734,
"step": 1099
},
{
"epoch": 1.2707581227436824,
"grad_norm": 0.027277441695332527,
"learning_rate": 4.772066888939423e-06,
"loss": 0.7894,
"step": 1100
},
{
"epoch": 1.271913357400722,
"grad_norm": 0.025714052841067314,
"learning_rate": 4.767557778553389e-06,
"loss": 0.7658,
"step": 1101
},
{
"epoch": 1.2730685920577618,
"grad_norm": 0.02643239125609398,
"learning_rate": 4.763006676431021e-06,
"loss": 0.775,
"step": 1102
},
{
"epoch": 1.2742238267148014,
"grad_norm": 0.028819050639867783,
"learning_rate": 4.75841366686113e-06,
"loss": 0.7823,
"step": 1103
},
{
"epoch": 1.2753790613718412,
"grad_norm": 0.026137549430131912,
"learning_rate": 4.753778834908678e-06,
"loss": 0.7607,
"step": 1104
},
{
"epoch": 1.2765342960288808,
"grad_norm": 0.027080422267317772,
"learning_rate": 4.749102266413195e-06,
"loss": 0.7739,
"step": 1105
},
{
"epoch": 1.2776895306859206,
"grad_norm": 0.026252347975969315,
"learning_rate": 4.744384047987202e-06,
"loss": 0.7914,
"step": 1106
},
{
"epoch": 1.2788447653429602,
"grad_norm": 0.02620067074894905,
"learning_rate": 4.739624267014589e-06,
"loss": 0.754,
"step": 1107
},
{
"epoch": 1.28,
"grad_norm": 0.026302233338356018,
"learning_rate": 4.734823011649013e-06,
"loss": 0.7964,
"step": 1108
},
{
"epoch": 1.2811552346570396,
"grad_norm": 0.02616049535572529,
"learning_rate": 4.7299803708122565e-06,
"loss": 0.7831,
"step": 1109
},
{
"epoch": 1.2823104693140794,
"grad_norm": 0.02667761594057083,
"learning_rate": 4.725096434192583e-06,
"loss": 0.7651,
"step": 1110
},
{
"epoch": 1.2834657039711193,
"grad_norm": 0.026840372011065483,
"learning_rate": 4.720171292243076e-06,
"loss": 0.7776,
"step": 1111
},
{
"epoch": 1.2846209386281588,
"grad_norm": 0.027080677449703217,
"learning_rate": 4.7152050361799626e-06,
"loss": 0.7804,
"step": 1112
},
{
"epoch": 1.2857761732851984,
"grad_norm": 0.026866400614380836,
"learning_rate": 4.710197757980923e-06,
"loss": 0.798,
"step": 1113
},
{
"epoch": 1.2869314079422383,
"grad_norm": 0.025592336431145668,
"learning_rate": 4.7051495503833955e-06,
"loss": 0.7717,
"step": 1114
},
{
"epoch": 1.288086642599278,
"grad_norm": 0.02509910985827446,
"learning_rate": 4.700060506882846e-06,
"loss": 0.7494,
"step": 1115
},
{
"epoch": 1.2892418772563177,
"grad_norm": 0.02745307981967926,
"learning_rate": 4.694930721731046e-06,
"loss": 0.7955,
"step": 1116
},
{
"epoch": 1.2903971119133573,
"grad_norm": 0.025139886885881424,
"learning_rate": 4.689760289934322e-06,
"loss": 0.7752,
"step": 1117
},
{
"epoch": 1.291552346570397,
"grad_norm": 0.025314299389719963,
"learning_rate": 4.684549307251799e-06,
"loss": 0.7631,
"step": 1118
},
{
"epoch": 1.292707581227437,
"grad_norm": 0.027501968666911125,
"learning_rate": 4.679297870193628e-06,
"loss": 0.7844,
"step": 1119
},
{
"epoch": 1.2938628158844765,
"grad_norm": 0.026297206059098244,
"learning_rate": 4.6740060760191905e-06,
"loss": 0.74,
"step": 1120
},
{
"epoch": 1.2950180505415163,
"grad_norm": 0.025890007615089417,
"learning_rate": 4.668674022735311e-06,
"loss": 0.7923,
"step": 1121
},
{
"epoch": 1.296173285198556,
"grad_norm": 0.026110520586371422,
"learning_rate": 4.663301809094425e-06,
"loss": 0.7899,
"step": 1122
},
{
"epoch": 1.2973285198555957,
"grad_norm": 0.026979975402355194,
"learning_rate": 4.657889534592767e-06,
"loss": 0.7831,
"step": 1123
},
{
"epoch": 1.2984837545126353,
"grad_norm": 0.026237450540065765,
"learning_rate": 4.652437299468514e-06,
"loss": 0.7728,
"step": 1124
},
{
"epoch": 1.2996389891696751,
"grad_norm": 0.02676897682249546,
"learning_rate": 4.646945204699939e-06,
"loss": 0.7736,
"step": 1125
},
{
"epoch": 1.3007942238267147,
"grad_norm": 0.02639893814921379,
"learning_rate": 4.641413352003534e-06,
"loss": 0.7738,
"step": 1126
},
{
"epoch": 1.3019494584837545,
"grad_norm": 0.026181381195783615,
"learning_rate": 4.63584184383213e-06,
"loss": 0.7766,
"step": 1127
},
{
"epoch": 1.3031046931407944,
"grad_norm": 0.029100872576236725,
"learning_rate": 4.630230783372998e-06,
"loss": 0.7939,
"step": 1128
},
{
"epoch": 1.304259927797834,
"grad_norm": 0.027207808569073677,
"learning_rate": 4.624580274545938e-06,
"loss": 0.7701,
"step": 1129
},
{
"epoch": 1.3054151624548735,
"grad_norm": 0.027454031631350517,
"learning_rate": 4.618890422001355e-06,
"loss": 0.7979,
"step": 1130
},
{
"epoch": 1.3065703971119134,
"grad_norm": 0.02654033899307251,
"learning_rate": 4.61316133111832e-06,
"loss": 0.7473,
"step": 1131
},
{
"epoch": 1.3077256317689532,
"grad_norm": 0.02540537528693676,
"learning_rate": 4.607393108002618e-06,
"loss": 0.7663,
"step": 1132
},
{
"epoch": 1.3088808664259928,
"grad_norm": 0.026679422706365585,
"learning_rate": 4.601585859484786e-06,
"loss": 0.7811,
"step": 1133
},
{
"epoch": 1.3100361010830324,
"grad_norm": 0.02628672868013382,
"learning_rate": 4.595739693118132e-06,
"loss": 0.7836,
"step": 1134
},
{
"epoch": 1.3111913357400722,
"grad_norm": 0.026694132015109062,
"learning_rate": 4.5898547171767375e-06,
"loss": 0.7708,
"step": 1135
},
{
"epoch": 1.312346570397112,
"grad_norm": 0.02691902220249176,
"learning_rate": 4.583931040653466e-06,
"loss": 0.772,
"step": 1136
},
{
"epoch": 1.3135018050541516,
"grad_norm": 0.02754952199757099,
"learning_rate": 4.5779687732579285e-06,
"loss": 0.783,
"step": 1137
},
{
"epoch": 1.3146570397111914,
"grad_norm": 0.025518443435430527,
"learning_rate": 4.571968025414466e-06,
"loss": 0.7736,
"step": 1138
},
{
"epoch": 1.315812274368231,
"grad_norm": 0.025681914761662483,
"learning_rate": 4.565928908260093e-06,
"loss": 0.7823,
"step": 1139
},
{
"epoch": 1.3169675090252708,
"grad_norm": 0.028058268129825592,
"learning_rate": 4.559851533642446e-06,
"loss": 0.7966,
"step": 1140
},
{
"epoch": 1.3181227436823104,
"grad_norm": 0.027302585542201996,
"learning_rate": 4.55373601411771e-06,
"loss": 0.8043,
"step": 1141
},
{
"epoch": 1.3192779783393502,
"grad_norm": 0.027836551889777184,
"learning_rate": 4.547582462948533e-06,
"loss": 0.7903,
"step": 1142
},
{
"epoch": 1.3204332129963898,
"grad_norm": 0.027944888919591904,
"learning_rate": 4.541390994101928e-06,
"loss": 0.7834,
"step": 1143
},
{
"epoch": 1.3215884476534296,
"grad_norm": 0.026407793164253235,
"learning_rate": 4.535161722247168e-06,
"loss": 0.7774,
"step": 1144
},
{
"epoch": 1.3227436823104692,
"grad_norm": 0.026148339733481407,
"learning_rate": 4.528894762753651e-06,
"loss": 0.7909,
"step": 1145
},
{
"epoch": 1.323898916967509,
"grad_norm": 0.027284881100058556,
"learning_rate": 4.522590231688775e-06,
"loss": 0.7885,
"step": 1146
},
{
"epoch": 1.3250541516245486,
"grad_norm": 0.026616675779223442,
"learning_rate": 4.516248245815782e-06,
"loss": 0.7924,
"step": 1147
},
{
"epoch": 1.3262093862815885,
"grad_norm": 0.026135679334402084,
"learning_rate": 4.509868922591595e-06,
"loss": 0.7845,
"step": 1148
},
{
"epoch": 1.3273646209386283,
"grad_norm": 0.027222778648138046,
"learning_rate": 4.503452380164648e-06,
"loss": 0.7983,
"step": 1149
},
{
"epoch": 1.3285198555956679,
"grad_norm": 0.02566412277519703,
"learning_rate": 4.496998737372691e-06,
"loss": 0.7561,
"step": 1150
},
{
"epoch": 1.3296750902527075,
"grad_norm": 0.026768745854496956,
"learning_rate": 4.490508113740594e-06,
"loss": 0.7753,
"step": 1151
},
{
"epoch": 1.3308303249097473,
"grad_norm": 0.027144471183419228,
"learning_rate": 4.483980629478131e-06,
"loss": 0.769,
"step": 1152
},
{
"epoch": 1.331985559566787,
"grad_norm": 0.0260478463023901,
"learning_rate": 4.477416405477754e-06,
"loss": 0.7879,
"step": 1153
},
{
"epoch": 1.3331407942238267,
"grad_norm": 0.025150155648589134,
"learning_rate": 4.470815563312356e-06,
"loss": 0.7795,
"step": 1154
},
{
"epoch": 1.3342960288808663,
"grad_norm": 0.026815764605998993,
"learning_rate": 4.464178225233016e-06,
"loss": 0.7577,
"step": 1155
},
{
"epoch": 1.335451263537906,
"grad_norm": 0.02605642005801201,
"learning_rate": 4.457504514166738e-06,
"loss": 0.7617,
"step": 1156
},
{
"epoch": 1.336606498194946,
"grad_norm": 0.02583806961774826,
"learning_rate": 4.4507945537141714e-06,
"loss": 0.7924,
"step": 1157
},
{
"epoch": 1.3377617328519855,
"grad_norm": 0.027298036962747574,
"learning_rate": 4.444048468147324e-06,
"loss": 0.7756,
"step": 1158
},
{
"epoch": 1.3389169675090253,
"grad_norm": 0.02701791748404503,
"learning_rate": 4.4372663824072595e-06,
"loss": 0.785,
"step": 1159
},
{
"epoch": 1.340072202166065,
"grad_norm": 0.02676539123058319,
"learning_rate": 4.430448422101786e-06,
"loss": 0.7914,
"step": 1160
},
{
"epoch": 1.3412274368231047,
"grad_norm": 0.026288233697414398,
"learning_rate": 4.423594713503124e-06,
"loss": 0.7948,
"step": 1161
},
{
"epoch": 1.3423826714801443,
"grad_norm": 0.02532966062426567,
"learning_rate": 4.416705383545574e-06,
"loss": 0.7571,
"step": 1162
},
{
"epoch": 1.3435379061371842,
"grad_norm": 0.025268996134400368,
"learning_rate": 4.409780559823159e-06,
"loss": 0.7544,
"step": 1163
},
{
"epoch": 1.3446931407942238,
"grad_norm": 0.025901716202497482,
"learning_rate": 4.402820370587272e-06,
"loss": 0.7928,
"step": 1164
},
{
"epoch": 1.3458483754512636,
"grad_norm": 0.026023518294095993,
"learning_rate": 4.395824944744287e-06,
"loss": 0.7821,
"step": 1165
},
{
"epoch": 1.3470036101083034,
"grad_norm": 0.025839975103735924,
"learning_rate": 4.388794411853183e-06,
"loss": 0.7563,
"step": 1166
},
{
"epoch": 1.348158844765343,
"grad_norm": 0.02656048908829689,
"learning_rate": 4.381728902123138e-06,
"loss": 0.7675,
"step": 1167
},
{
"epoch": 1.3493140794223826,
"grad_norm": 0.025841468945145607,
"learning_rate": 4.374628546411121e-06,
"loss": 0.7839,
"step": 1168
},
{
"epoch": 1.3504693140794224,
"grad_norm": 0.02622513473033905,
"learning_rate": 4.367493476219464e-06,
"loss": 0.7764,
"step": 1169
},
{
"epoch": 1.3516245487364622,
"grad_norm": 0.0271604061126709,
"learning_rate": 4.360323823693434e-06,
"loss": 0.7873,
"step": 1170
},
{
"epoch": 1.3527797833935018,
"grad_norm": 0.025581173598766327,
"learning_rate": 4.353119721618778e-06,
"loss": 0.7719,
"step": 1171
},
{
"epoch": 1.3539350180505414,
"grad_norm": 0.027338258922100067,
"learning_rate": 4.345881303419266e-06,
"loss": 0.8023,
"step": 1172
},
{
"epoch": 1.3550902527075812,
"grad_norm": 0.024792378768324852,
"learning_rate": 4.338608703154224e-06,
"loss": 0.7612,
"step": 1173
},
{
"epoch": 1.356245487364621,
"grad_norm": 0.02713456004858017,
"learning_rate": 4.3313020555160445e-06,
"loss": 0.7735,
"step": 1174
},
{
"epoch": 1.3574007220216606,
"grad_norm": 0.026577133685350418,
"learning_rate": 4.323961495827698e-06,
"loss": 0.7942,
"step": 1175
},
{
"epoch": 1.3585559566787004,
"grad_norm": 0.02608906850218773,
"learning_rate": 4.316587160040222e-06,
"loss": 0.7675,
"step": 1176
},
{
"epoch": 1.35971119133574,
"grad_norm": 0.026449469849467278,
"learning_rate": 4.309179184730207e-06,
"loss": 0.7586,
"step": 1177
},
{
"epoch": 1.3608664259927798,
"grad_norm": 0.02639465406537056,
"learning_rate": 4.301737707097264e-06,
"loss": 0.7792,
"step": 1178
},
{
"epoch": 1.3620216606498194,
"grad_norm": 0.026730941608548164,
"learning_rate": 4.294262864961486e-06,
"loss": 0.77,
"step": 1179
},
{
"epoch": 1.3631768953068593,
"grad_norm": 0.02690793387591839,
"learning_rate": 4.286754796760892e-06,
"loss": 0.7898,
"step": 1180
},
{
"epoch": 1.3643321299638989,
"grad_norm": 0.02708018571138382,
"learning_rate": 4.279213641548867e-06,
"loss": 0.7644,
"step": 1181
},
{
"epoch": 1.3654873646209387,
"grad_norm": 0.02672554738819599,
"learning_rate": 4.271639538991584e-06,
"loss": 0.7484,
"step": 1182
},
{
"epoch": 1.3666425992779783,
"grad_norm": 0.028112929314374924,
"learning_rate": 4.264032629365418e-06,
"loss": 0.8274,
"step": 1183
},
{
"epoch": 1.367797833935018,
"grad_norm": 0.026468411087989807,
"learning_rate": 4.256393053554351e-06,
"loss": 0.7438,
"step": 1184
},
{
"epoch": 1.3689530685920577,
"grad_norm": 0.02600760944187641,
"learning_rate": 4.248720953047355e-06,
"loss": 0.764,
"step": 1185
},
{
"epoch": 1.3701083032490975,
"grad_norm": 0.026687582954764366,
"learning_rate": 4.241016469935782e-06,
"loss": 0.764,
"step": 1186
},
{
"epoch": 1.3712635379061373,
"grad_norm": 0.028025714680552483,
"learning_rate": 4.233279746910722e-06,
"loss": 0.7729,
"step": 1187
},
{
"epoch": 1.372418772563177,
"grad_norm": 0.028168534860014915,
"learning_rate": 4.225510927260369e-06,
"loss": 0.7901,
"step": 1188
},
{
"epoch": 1.3735740072202165,
"grad_norm": 0.027783645316958427,
"learning_rate": 4.217710154867362e-06,
"loss": 0.8064,
"step": 1189
},
{
"epoch": 1.3747292418772563,
"grad_norm": 0.025505684316158295,
"learning_rate": 4.209877574206121e-06,
"loss": 0.7248,
"step": 1190
},
{
"epoch": 1.3758844765342961,
"grad_norm": 0.02887566015124321,
"learning_rate": 4.2020133303401714e-06,
"loss": 0.8044,
"step": 1191
},
{
"epoch": 1.3770397111913357,
"grad_norm": 0.02724134922027588,
"learning_rate": 4.1941175689194615e-06,
"loss": 0.771,
"step": 1192
},
{
"epoch": 1.3781949458483753,
"grad_norm": 0.026623480021953583,
"learning_rate": 4.186190436177655e-06,
"loss": 0.7926,
"step": 1193
},
{
"epoch": 1.3793501805054151,
"grad_norm": 0.026381775736808777,
"learning_rate": 4.178232078929437e-06,
"loss": 0.7658,
"step": 1194
},
{
"epoch": 1.380505415162455,
"grad_norm": 0.027352899312973022,
"learning_rate": 4.170242644567781e-06,
"loss": 0.7598,
"step": 1195
},
{
"epoch": 1.3816606498194945,
"grad_norm": 0.027212858200073242,
"learning_rate": 4.162222281061226e-06,
"loss": 0.8047,
"step": 1196
},
{
"epoch": 1.3828158844765344,
"grad_norm": 0.02757844887673855,
"learning_rate": 4.154171136951136e-06,
"loss": 0.7707,
"step": 1197
},
{
"epoch": 1.383971119133574,
"grad_norm": 0.02665034867823124,
"learning_rate": 4.146089361348948e-06,
"loss": 0.7659,
"step": 1198
},
{
"epoch": 1.3851263537906138,
"grad_norm": 0.02670939266681671,
"learning_rate": 4.137977103933411e-06,
"loss": 0.7898,
"step": 1199
},
{
"epoch": 1.3862815884476534,
"grad_norm": 0.02680130861699581,
"learning_rate": 4.1298345149478115e-06,
"loss": 0.8019,
"step": 1200
},
{
"epoch": 1.3874368231046932,
"grad_norm": 0.02670557238161564,
"learning_rate": 4.121661745197195e-06,
"loss": 0.7749,
"step": 1201
},
{
"epoch": 1.3885920577617328,
"grad_norm": 0.02557177096605301,
"learning_rate": 4.113458946045569e-06,
"loss": 0.7807,
"step": 1202
},
{
"epoch": 1.3897472924187726,
"grad_norm": 0.027189351618289948,
"learning_rate": 4.1052262694131e-06,
"loss": 0.8127,
"step": 1203
},
{
"epoch": 1.3909025270758124,
"grad_norm": 0.026801222935318947,
"learning_rate": 4.096963867773306e-06,
"loss": 0.7887,
"step": 1204
},
{
"epoch": 1.392057761732852,
"grad_norm": 0.027125921100378036,
"learning_rate": 4.0886718941502215e-06,
"loss": 0.7705,
"step": 1205
},
{
"epoch": 1.3932129963898916,
"grad_norm": 0.027330022305250168,
"learning_rate": 4.080350502115575e-06,
"loss": 0.7655,
"step": 1206
},
{
"epoch": 1.3943682310469314,
"grad_norm": 0.025932665914297104,
"learning_rate": 4.071999845785938e-06,
"loss": 0.7839,
"step": 1207
},
{
"epoch": 1.3955234657039712,
"grad_norm": 0.02584053948521614,
"learning_rate": 4.06362007981987e-06,
"loss": 0.7676,
"step": 1208
},
{
"epoch": 1.3966787003610108,
"grad_norm": 0.025614172220230103,
"learning_rate": 4.055211359415058e-06,
"loss": 0.7685,
"step": 1209
},
{
"epoch": 1.3978339350180504,
"grad_norm": 0.026585347950458527,
"learning_rate": 4.0467738403054425e-06,
"loss": 0.7716,
"step": 1210
},
{
"epoch": 1.3989891696750902,
"grad_norm": 0.02616872265934944,
"learning_rate": 4.038307678758328e-06,
"loss": 0.8029,
"step": 1211
},
{
"epoch": 1.40014440433213,
"grad_norm": 0.02624761499464512,
"learning_rate": 4.0298130315714934e-06,
"loss": 0.766,
"step": 1212
},
{
"epoch": 1.4012996389891696,
"grad_norm": 0.025757934898138046,
"learning_rate": 4.021290056070287e-06,
"loss": 0.7724,
"step": 1213
},
{
"epoch": 1.4024548736462095,
"grad_norm": 0.027395786717534065,
"learning_rate": 4.012738910104711e-06,
"loss": 0.7987,
"step": 1214
},
{
"epoch": 1.403610108303249,
"grad_norm": 0.0256296806037426,
"learning_rate": 4.004159752046501e-06,
"loss": 0.7516,
"step": 1215
},
{
"epoch": 1.4047653429602889,
"grad_norm": 0.027848385274410248,
"learning_rate": 3.99555274078619e-06,
"loss": 0.7778,
"step": 1216
},
{
"epoch": 1.4059205776173285,
"grad_norm": 0.02629752829670906,
"learning_rate": 3.986918035730169e-06,
"loss": 0.7689,
"step": 1217
},
{
"epoch": 1.4070758122743683,
"grad_norm": 0.027596496045589447,
"learning_rate": 3.9782557967977295e-06,
"loss": 0.7976,
"step": 1218
},
{
"epoch": 1.4082310469314079,
"grad_norm": 0.02730746567249298,
"learning_rate": 3.969566184418109e-06,
"loss": 0.7775,
"step": 1219
},
{
"epoch": 1.4093862815884477,
"grad_norm": 0.025179557502269745,
"learning_rate": 3.960849359527515e-06,
"loss": 0.7639,
"step": 1220
},
{
"epoch": 1.4105415162454873,
"grad_norm": 0.026616785675287247,
"learning_rate": 3.952105483566139e-06,
"loss": 0.7583,
"step": 1221
},
{
"epoch": 1.411696750902527,
"grad_norm": 0.027852557599544525,
"learning_rate": 3.94333471847518e-06,
"loss": 0.7936,
"step": 1222
},
{
"epoch": 1.4128519855595667,
"grad_norm": 0.026742927730083466,
"learning_rate": 3.934537226693836e-06,
"loss": 0.761,
"step": 1223
},
{
"epoch": 1.4140072202166065,
"grad_norm": 0.025545494630932808,
"learning_rate": 3.925713171156295e-06,
"loss": 0.7952,
"step": 1224
},
{
"epoch": 1.4151624548736463,
"grad_norm": 0.02676277793943882,
"learning_rate": 3.9168627152887214e-06,
"loss": 0.7623,
"step": 1225
},
{
"epoch": 1.416317689530686,
"grad_norm": 0.024539409205317497,
"learning_rate": 3.9079860230062286e-06,
"loss": 0.7451,
"step": 1226
},
{
"epoch": 1.4174729241877255,
"grad_norm": 0.026942811906337738,
"learning_rate": 3.8990832587098396e-06,
"loss": 0.7825,
"step": 1227
},
{
"epoch": 1.4186281588447653,
"grad_norm": 0.026073571294546127,
"learning_rate": 3.890154587283447e-06,
"loss": 0.7788,
"step": 1228
},
{
"epoch": 1.4197833935018052,
"grad_norm": 0.02608587220311165,
"learning_rate": 3.881200174090758e-06,
"loss": 0.7752,
"step": 1229
},
{
"epoch": 1.4209386281588448,
"grad_norm": 0.02780032530426979,
"learning_rate": 3.8722201849722264e-06,
"loss": 0.7834,
"step": 1230
},
{
"epoch": 1.4220938628158843,
"grad_norm": 0.025982137769460678,
"learning_rate": 3.863214786241993e-06,
"loss": 0.7633,
"step": 1231
},
{
"epoch": 1.4232490974729242,
"grad_norm": 0.026244191452860832,
"learning_rate": 3.85418414468479e-06,
"loss": 0.7901,
"step": 1232
},
{
"epoch": 1.424404332129964,
"grad_norm": 0.028074799105525017,
"learning_rate": 3.845128427552865e-06,
"loss": 0.7579,
"step": 1233
},
{
"epoch": 1.4255595667870036,
"grad_norm": 0.027848297730088234,
"learning_rate": 3.836047802562878e-06,
"loss": 0.7633,
"step": 1234
},
{
"epoch": 1.4267148014440434,
"grad_norm": 0.026308046653866768,
"learning_rate": 3.8269424378927925e-06,
"loss": 0.7959,
"step": 1235
},
{
"epoch": 1.427870036101083,
"grad_norm": 0.02657618559896946,
"learning_rate": 3.81781250217877e-06,
"loss": 0.782,
"step": 1236
},
{
"epoch": 1.4290252707581228,
"grad_norm": 0.02592160366475582,
"learning_rate": 3.8086581645120347e-06,
"loss": 0.7875,
"step": 1237
},
{
"epoch": 1.4301805054151624,
"grad_norm": 0.02750803716480732,
"learning_rate": 3.799479594435751e-06,
"loss": 0.7712,
"step": 1238
},
{
"epoch": 1.4313357400722022,
"grad_norm": 0.02752981334924698,
"learning_rate": 3.7902769619418783e-06,
"loss": 0.8094,
"step": 1239
},
{
"epoch": 1.4324909747292418,
"grad_norm": 0.025965960696339607,
"learning_rate": 3.7810504374680274e-06,
"loss": 0.7687,
"step": 1240
},
{
"epoch": 1.4336462093862816,
"grad_norm": 0.025762738659977913,
"learning_rate": 3.7718001918942982e-06,
"loss": 0.7943,
"step": 1241
},
{
"epoch": 1.4348014440433212,
"grad_norm": 0.026879120618104935,
"learning_rate": 3.762526396540121e-06,
"loss": 0.7735,
"step": 1242
},
{
"epoch": 1.435956678700361,
"grad_norm": 0.02714950405061245,
"learning_rate": 3.7532292231610774e-06,
"loss": 0.7641,
"step": 1243
},
{
"epoch": 1.4371119133574006,
"grad_norm": 0.02679312974214554,
"learning_rate": 3.7439088439457257e-06,
"loss": 0.7627,
"step": 1244
},
{
"epoch": 1.4382671480144404,
"grad_norm": 0.027805205434560776,
"learning_rate": 3.7345654315124035e-06,
"loss": 0.7811,
"step": 1245
},
{
"epoch": 1.4394223826714803,
"grad_norm": 0.02763986401259899,
"learning_rate": 3.725199158906041e-06,
"loss": 0.7924,
"step": 1246
},
{
"epoch": 1.4405776173285199,
"grad_norm": 0.026331216096878052,
"learning_rate": 3.7158101995949486e-06,
"loss": 0.7601,
"step": 1247
},
{
"epoch": 1.4417328519855594,
"grad_norm": 0.025142505764961243,
"learning_rate": 3.706398727467606e-06,
"loss": 0.7624,
"step": 1248
},
{
"epoch": 1.4428880866425993,
"grad_norm": 0.02704186551272869,
"learning_rate": 3.696964916829443e-06,
"loss": 0.7867,
"step": 1249
},
{
"epoch": 1.444043321299639,
"grad_norm": 0.027327412739396095,
"learning_rate": 3.6875089423996097e-06,
"loss": 0.7861,
"step": 1250
},
{
"epoch": 1.4451985559566787,
"grad_norm": 0.027434786781668663,
"learning_rate": 3.678030979307741e-06,
"loss": 0.7869,
"step": 1251
},
{
"epoch": 1.4463537906137185,
"grad_norm": 0.02410939894616604,
"learning_rate": 3.6685312030907167e-06,
"loss": 0.7568,
"step": 1252
},
{
"epoch": 1.447509025270758,
"grad_norm": 0.02866499125957489,
"learning_rate": 3.6590097896894017e-06,
"loss": 0.8089,
"step": 1253
},
{
"epoch": 1.448664259927798,
"grad_norm": 0.025909971445798874,
"learning_rate": 3.6494669154453978e-06,
"loss": 0.794,
"step": 1254
},
{
"epoch": 1.4498194945848375,
"grad_norm": 0.026149652898311615,
"learning_rate": 3.6399027570977703e-06,
"loss": 0.7657,
"step": 1255
},
{
"epoch": 1.4509747292418773,
"grad_norm": 0.027486305683851242,
"learning_rate": 3.6303174917797783e-06,
"loss": 0.7573,
"step": 1256
},
{
"epoch": 1.452129963898917,
"grad_norm": 0.026965491473674774,
"learning_rate": 3.6207112970155915e-06,
"loss": 0.7739,
"step": 1257
},
{
"epoch": 1.4532851985559567,
"grad_norm": 0.026813145726919174,
"learning_rate": 3.611084350717008e-06,
"loss": 0.7612,
"step": 1258
},
{
"epoch": 1.4544404332129963,
"grad_norm": 0.027455372735857964,
"learning_rate": 3.60143683118015e-06,
"loss": 0.7892,
"step": 1259
},
{
"epoch": 1.4555956678700361,
"grad_norm": 0.027145925909280777,
"learning_rate": 3.5917689170821714e-06,
"loss": 0.7806,
"step": 1260
},
{
"epoch": 1.4567509025270757,
"grad_norm": 0.02541586011648178,
"learning_rate": 3.582080787477941e-06,
"loss": 0.7772,
"step": 1261
},
{
"epoch": 1.4579061371841155,
"grad_norm": 0.028484217822551727,
"learning_rate": 3.5723726217967315e-06,
"loss": 0.7926,
"step": 1262
},
{
"epoch": 1.4590613718411554,
"grad_norm": 0.027112239971756935,
"learning_rate": 3.5626445998388926e-06,
"loss": 0.7819,
"step": 1263
},
{
"epoch": 1.460216606498195,
"grad_norm": 0.027995990589261055,
"learning_rate": 3.552896901772525e-06,
"loss": 0.7845,
"step": 1264
},
{
"epoch": 1.4613718411552346,
"grad_norm": 0.026287071406841278,
"learning_rate": 3.543129708130139e-06,
"loss": 0.7979,
"step": 1265
},
{
"epoch": 1.4625270758122744,
"grad_norm": 0.026646843180060387,
"learning_rate": 3.533343199805315e-06,
"loss": 0.7869,
"step": 1266
},
{
"epoch": 1.4636823104693142,
"grad_norm": 0.025323543697595596,
"learning_rate": 3.52353755804935e-06,
"loss": 0.7765,
"step": 1267
},
{
"epoch": 1.4648375451263538,
"grad_norm": 0.025138625875115395,
"learning_rate": 3.5137129644679035e-06,
"loss": 0.7606,
"step": 1268
},
{
"epoch": 1.4659927797833934,
"grad_norm": 0.026612093672156334,
"learning_rate": 3.5038696010176316e-06,
"loss": 0.7624,
"step": 1269
},
{
"epoch": 1.4671480144404332,
"grad_norm": 0.025180159136652946,
"learning_rate": 3.4940076500028193e-06,
"loss": 0.7704,
"step": 1270
},
{
"epoch": 1.468303249097473,
"grad_norm": 0.0247277170419693,
"learning_rate": 3.484127294072003e-06,
"loss": 0.7799,
"step": 1271
},
{
"epoch": 1.4694584837545126,
"grad_norm": 0.025542214512825012,
"learning_rate": 3.474228716214588e-06,
"loss": 0.7859,
"step": 1272
},
{
"epoch": 1.4706137184115524,
"grad_norm": 0.02490636333823204,
"learning_rate": 3.464312099757457e-06,
"loss": 0.7553,
"step": 1273
},
{
"epoch": 1.471768953068592,
"grad_norm": 0.027338897809386253,
"learning_rate": 3.4543776283615815e-06,
"loss": 0.7869,
"step": 1274
},
{
"epoch": 1.4729241877256318,
"grad_norm": 0.0249380711466074,
"learning_rate": 3.444425486018611e-06,
"loss": 0.7734,
"step": 1275
},
{
"epoch": 1.4740794223826714,
"grad_norm": 0.02633252553641796,
"learning_rate": 3.434455857047475e-06,
"loss": 0.7714,
"step": 1276
},
{
"epoch": 1.4752346570397112,
"grad_norm": 0.026209495961666107,
"learning_rate": 3.424468926090961e-06,
"loss": 0.7912,
"step": 1277
},
{
"epoch": 1.4763898916967508,
"grad_norm": 0.02743818424642086,
"learning_rate": 3.4144648781122987e-06,
"loss": 0.7725,
"step": 1278
},
{
"epoch": 1.4775451263537906,
"grad_norm": 0.02479531429708004,
"learning_rate": 3.404443898391737e-06,
"loss": 0.7548,
"step": 1279
},
{
"epoch": 1.4787003610108302,
"grad_norm": 0.026155853644013405,
"learning_rate": 3.3944061725231055e-06,
"loss": 0.7882,
"step": 1280
},
{
"epoch": 1.47985559566787,
"grad_norm": 0.02570260874927044,
"learning_rate": 3.3843518864103857e-06,
"loss": 0.7617,
"step": 1281
},
{
"epoch": 1.4810108303249097,
"grad_norm": 0.026218149811029434,
"learning_rate": 3.374281226264259e-06,
"loss": 0.7798,
"step": 1282
},
{
"epoch": 1.4821660649819495,
"grad_norm": 0.025268124416470528,
"learning_rate": 3.364194378598666e-06,
"loss": 0.7647,
"step": 1283
},
{
"epoch": 1.4833212996389893,
"grad_norm": 0.028134865686297417,
"learning_rate": 3.3540915302273476e-06,
"loss": 0.777,
"step": 1284
},
{
"epoch": 1.4844765342960289,
"grad_norm": 0.026391401886940002,
"learning_rate": 3.343972868260385e-06,
"loss": 0.7704,
"step": 1285
},
{
"epoch": 1.4856317689530685,
"grad_norm": 0.025334736332297325,
"learning_rate": 3.333838580100737e-06,
"loss": 0.7465,
"step": 1286
},
{
"epoch": 1.4867870036101083,
"grad_norm": 0.025972386822104454,
"learning_rate": 3.32368885344077e-06,
"loss": 0.798,
"step": 1287
},
{
"epoch": 1.487942238267148,
"grad_norm": 0.02676592580974102,
"learning_rate": 3.3135238762587723e-06,
"loss": 0.7946,
"step": 1288
},
{
"epoch": 1.4890974729241877,
"grad_norm": 0.026947803795337677,
"learning_rate": 3.3033438368154886e-06,
"loss": 0.7762,
"step": 1289
},
{
"epoch": 1.4902527075812273,
"grad_norm": 0.02568644843995571,
"learning_rate": 3.2931489236506183e-06,
"loss": 0.7977,
"step": 1290
},
{
"epoch": 1.4914079422382671,
"grad_norm": 0.02489590272307396,
"learning_rate": 3.282939325579333e-06,
"loss": 0.7634,
"step": 1291
},
{
"epoch": 1.492563176895307,
"grad_norm": 0.02679099142551422,
"learning_rate": 3.272715231688775e-06,
"loss": 0.7917,
"step": 1292
},
{
"epoch": 1.4937184115523465,
"grad_norm": 0.025316089391708374,
"learning_rate": 3.2624768313345594e-06,
"loss": 0.7507,
"step": 1293
},
{
"epoch": 1.4948736462093863,
"grad_norm": 0.024965766817331314,
"learning_rate": 3.25222431413726e-06,
"loss": 0.7751,
"step": 1294
},
{
"epoch": 1.496028880866426,
"grad_norm": 0.026891566812992096,
"learning_rate": 3.241957869978907e-06,
"loss": 0.7697,
"step": 1295
},
{
"epoch": 1.4971841155234658,
"grad_norm": 0.02724074199795723,
"learning_rate": 3.2316776889994627e-06,
"loss": 0.8087,
"step": 1296
},
{
"epoch": 1.4983393501805053,
"grad_norm": 0.025544699281454086,
"learning_rate": 3.221383961593303e-06,
"loss": 0.7511,
"step": 1297
},
{
"epoch": 1.4994945848375452,
"grad_norm": 0.02615639939904213,
"learning_rate": 3.2110768784056884e-06,
"loss": 0.7741,
"step": 1298
},
{
"epoch": 1.5006498194945848,
"grad_norm": 0.026993228122591972,
"learning_rate": 3.2007566303292416e-06,
"loss": 0.7631,
"step": 1299
},
{
"epoch": 1.5018050541516246,
"grad_norm": 0.02476736344397068,
"learning_rate": 3.1904234085003976e-06,
"loss": 0.7574,
"step": 1300
},
{
"epoch": 1.5029602888086644,
"grad_norm": 0.025930294767022133,
"learning_rate": 3.180077404295881e-06,
"loss": 0.8047,
"step": 1301
},
{
"epoch": 1.504115523465704,
"grad_norm": 0.025386760011315346,
"learning_rate": 3.1697188093291454e-06,
"loss": 0.7663,
"step": 1302
},
{
"epoch": 1.5052707581227436,
"grad_norm": 0.025852493941783905,
"learning_rate": 3.1593478154468364e-06,
"loss": 0.8022,
"step": 1303
},
{
"epoch": 1.5064259927797834,
"grad_norm": 0.024930665269494057,
"learning_rate": 3.148964614725232e-06,
"loss": 0.779,
"step": 1304
},
{
"epoch": 1.5075812274368232,
"grad_norm": 0.025383364409208298,
"learning_rate": 3.138569399466689e-06,
"loss": 0.7833,
"step": 1305
},
{
"epoch": 1.5087364620938628,
"grad_norm": 0.026384184136986732,
"learning_rate": 3.128162362196077e-06,
"loss": 0.7753,
"step": 1306
},
{
"epoch": 1.5098916967509024,
"grad_norm": 0.02558097243309021,
"learning_rate": 3.117743695657219e-06,
"loss": 0.7648,
"step": 1307
},
{
"epoch": 1.5110469314079422,
"grad_norm": 0.02592851221561432,
"learning_rate": 3.107313592809315e-06,
"loss": 0.7819,
"step": 1308
},
{
"epoch": 1.512202166064982,
"grad_norm": 0.02552459016442299,
"learning_rate": 3.0968722468233743e-06,
"loss": 0.7831,
"step": 1309
},
{
"epoch": 1.5133574007220216,
"grad_norm": 0.026385486125946045,
"learning_rate": 3.086419851078632e-06,
"loss": 0.7874,
"step": 1310
},
{
"epoch": 1.5145126353790612,
"grad_norm": 0.025948703289031982,
"learning_rate": 3.0759565991589733e-06,
"loss": 0.7687,
"step": 1311
},
{
"epoch": 1.515667870036101,
"grad_norm": 0.02731943503022194,
"learning_rate": 3.0654826848493436e-06,
"loss": 0.7722,
"step": 1312
},
{
"epoch": 1.5168231046931409,
"grad_norm": 0.02646474353969097,
"learning_rate": 3.0549983021321634e-06,
"loss": 0.7731,
"step": 1313
},
{
"epoch": 1.5179783393501805,
"grad_norm": 0.0249773059040308,
"learning_rate": 3.044503645183731e-06,
"loss": 0.7773,
"step": 1314
},
{
"epoch": 1.5191335740072203,
"grad_norm": 0.0276559516787529,
"learning_rate": 3.0339989083706306e-06,
"loss": 0.7911,
"step": 1315
},
{
"epoch": 1.5202888086642599,
"grad_norm": 0.026679757982492447,
"learning_rate": 3.0234842862461307e-06,
"loss": 0.7383,
"step": 1316
},
{
"epoch": 1.5214440433212997,
"grad_norm": 0.02656986005604267,
"learning_rate": 3.0129599735465816e-06,
"loss": 0.766,
"step": 1317
},
{
"epoch": 1.5225992779783395,
"grad_norm": 0.024834152311086655,
"learning_rate": 3.0024261651878056e-06,
"loss": 0.7661,
"step": 1318
},
{
"epoch": 1.523754512635379,
"grad_norm": 0.026497021317481995,
"learning_rate": 2.9918830562614927e-06,
"loss": 0.785,
"step": 1319
},
{
"epoch": 1.5249097472924187,
"grad_norm": 0.02679789625108242,
"learning_rate": 2.9813308420315825e-06,
"loss": 0.7814,
"step": 1320
},
{
"epoch": 1.5260649819494585,
"grad_norm": 0.02663526125252247,
"learning_rate": 2.9707697179306505e-06,
"loss": 0.7875,
"step": 1321
},
{
"epoch": 1.5272202166064983,
"grad_norm": 0.02751866728067398,
"learning_rate": 2.9601998795562865e-06,
"loss": 0.7683,
"step": 1322
},
{
"epoch": 1.528375451263538,
"grad_norm": 0.026420842856168747,
"learning_rate": 2.9496215226674747e-06,
"loss": 0.7671,
"step": 1323
},
{
"epoch": 1.5295306859205775,
"grad_norm": 0.026854515075683594,
"learning_rate": 2.9390348431809656e-06,
"loss": 0.7705,
"step": 1324
},
{
"epoch": 1.5306859205776173,
"grad_norm": 0.025556296110153198,
"learning_rate": 2.9284400371676486e-06,
"loss": 0.7771,
"step": 1325
},
{
"epoch": 1.5318411552346571,
"grad_norm": 0.025579258799552917,
"learning_rate": 2.917837300848923e-06,
"loss": 0.7683,
"step": 1326
},
{
"epoch": 1.5329963898916967,
"grad_norm": 0.027229176834225655,
"learning_rate": 2.9072268305930578e-06,
"loss": 0.789,
"step": 1327
},
{
"epoch": 1.5341516245487363,
"grad_norm": 0.025784116238355637,
"learning_rate": 2.896608822911561e-06,
"loss": 0.7699,
"step": 1328
},
{
"epoch": 1.5353068592057761,
"grad_norm": 0.02648346498608589,
"learning_rate": 2.8859834744555377e-06,
"loss": 0.773,
"step": 1329
},
{
"epoch": 1.536462093862816,
"grad_norm": 0.025939345359802246,
"learning_rate": 2.8753509820120464e-06,
"loss": 0.791,
"step": 1330
},
{
"epoch": 1.5376173285198556,
"grad_norm": 0.025402942672371864,
"learning_rate": 2.864711542500457e-06,
"loss": 0.7643,
"step": 1331
},
{
"epoch": 1.5387725631768951,
"grad_norm": 0.026020778343081474,
"learning_rate": 2.8540653529688033e-06,
"loss": 0.7813,
"step": 1332
},
{
"epoch": 1.539927797833935,
"grad_norm": 0.026241673156619072,
"learning_rate": 2.8434126105901333e-06,
"loss": 0.7905,
"step": 1333
},
{
"epoch": 1.5410830324909748,
"grad_norm": 0.02648642659187317,
"learning_rate": 2.8327535126588563e-06,
"loss": 0.771,
"step": 1334
},
{
"epoch": 1.5422382671480146,
"grad_norm": 0.02610144577920437,
"learning_rate": 2.8220882565870893e-06,
"loss": 0.792,
"step": 1335
},
{
"epoch": 1.5433935018050542,
"grad_norm": 0.025254933163523674,
"learning_rate": 2.811417039901001e-06,
"loss": 0.7674,
"step": 1336
},
{
"epoch": 1.5445487364620938,
"grad_norm": 0.025538668036460876,
"learning_rate": 2.800740060237157e-06,
"loss": 0.7745,
"step": 1337
},
{
"epoch": 1.5457039711191336,
"grad_norm": 0.025685755535960197,
"learning_rate": 2.7900575153388525e-06,
"loss": 0.7645,
"step": 1338
},
{
"epoch": 1.5468592057761734,
"grad_norm": 0.026221225038170815,
"learning_rate": 2.779369603052456e-06,
"loss": 0.8149,
"step": 1339
},
{
"epoch": 1.548014440433213,
"grad_norm": 0.02525060623884201,
"learning_rate": 2.768676521323742e-06,
"loss": 0.7552,
"step": 1340
},
{
"epoch": 1.5491696750902526,
"grad_norm": 0.026158245280385017,
"learning_rate": 2.757978468194226e-06,
"loss": 0.7712,
"step": 1341
},
{
"epoch": 1.5503249097472924,
"grad_norm": 0.02666587382555008,
"learning_rate": 2.747275641797496e-06,
"loss": 0.7511,
"step": 1342
},
{
"epoch": 1.5514801444043322,
"grad_norm": 0.026437923312187195,
"learning_rate": 2.736568240355544e-06,
"loss": 0.7659,
"step": 1343
},
{
"epoch": 1.5526353790613718,
"grad_norm": 0.0248862411826849,
"learning_rate": 2.725856462175095e-06,
"loss": 0.7547,
"step": 1344
},
{
"epoch": 1.5537906137184114,
"grad_norm": 0.026547906920313835,
"learning_rate": 2.715140505643932e-06,
"loss": 0.7996,
"step": 1345
},
{
"epoch": 1.5549458483754512,
"grad_norm": 0.02599526010453701,
"learning_rate": 2.704420569227224e-06,
"loss": 0.7641,
"step": 1346
},
{
"epoch": 1.556101083032491,
"grad_norm": 0.02453439123928547,
"learning_rate": 2.6936968514638495e-06,
"loss": 0.7466,
"step": 1347
},
{
"epoch": 1.5572563176895307,
"grad_norm": 0.027833128347992897,
"learning_rate": 2.68296955096272e-06,
"loss": 0.7826,
"step": 1348
},
{
"epoch": 1.5584115523465703,
"grad_norm": 0.026869479566812515,
"learning_rate": 2.672238866399099e-06,
"loss": 0.789,
"step": 1349
},
{
"epoch": 1.55956678700361,
"grad_norm": 0.025666510686278343,
"learning_rate": 2.6615049965109296e-06,
"loss": 0.784,
"step": 1350
},
{
"epoch": 1.5607220216606499,
"grad_norm": 0.02734527923166752,
"learning_rate": 2.6507681400951414e-06,
"loss": 0.7934,
"step": 1351
},
{
"epoch": 1.5618772563176895,
"grad_norm": 0.025378704071044922,
"learning_rate": 2.640028496003984e-06,
"loss": 0.7554,
"step": 1352
},
{
"epoch": 1.5630324909747293,
"grad_norm": 0.02528882399201393,
"learning_rate": 2.629286263141329e-06,
"loss": 0.7652,
"step": 1353
},
{
"epoch": 1.5641877256317689,
"grad_norm": 0.0248930174857378,
"learning_rate": 2.6185416404589985e-06,
"loss": 0.7515,
"step": 1354
},
{
"epoch": 1.5653429602888087,
"grad_norm": 0.025928398594260216,
"learning_rate": 2.607794826953075e-06,
"loss": 0.7839,
"step": 1355
},
{
"epoch": 1.5664981949458485,
"grad_norm": 0.025345591828227043,
"learning_rate": 2.5970460216602133e-06,
"loss": 0.773,
"step": 1356
},
{
"epoch": 1.5676534296028881,
"grad_norm": 0.02638527937233448,
"learning_rate": 2.5862954236539587e-06,
"loss": 0.7641,
"step": 1357
},
{
"epoch": 1.5688086642599277,
"grad_norm": 0.025557557120919228,
"learning_rate": 2.5755432320410594e-06,
"loss": 0.7869,
"step": 1358
},
{
"epoch": 1.5699638989169675,
"grad_norm": 0.025332162156701088,
"learning_rate": 2.5647896459577757e-06,
"loss": 0.7873,
"step": 1359
},
{
"epoch": 1.5711191335740073,
"grad_norm": 0.025012118741869926,
"learning_rate": 2.5540348645661965e-06,
"loss": 0.7805,
"step": 1360
},
{
"epoch": 1.572274368231047,
"grad_norm": 0.02651703916490078,
"learning_rate": 2.5432790870505487e-06,
"loss": 0.791,
"step": 1361
},
{
"epoch": 1.5734296028880865,
"grad_norm": 0.025666911154985428,
"learning_rate": 2.5325225126135045e-06,
"loss": 0.7814,
"step": 1362
},
{
"epoch": 1.5745848375451263,
"grad_norm": 0.02593865618109703,
"learning_rate": 2.521765340472499e-06,
"loss": 0.8046,
"step": 1363
},
{
"epoch": 1.5757400722021662,
"grad_norm": 0.025667283684015274,
"learning_rate": 2.5110077698560364e-06,
"loss": 0.757,
"step": 1364
},
{
"epoch": 1.5768953068592058,
"grad_norm": 0.025065291672945023,
"learning_rate": 2.5002500000000003e-06,
"loss": 0.7729,
"step": 1365
},
{
"epoch": 1.5780505415162454,
"grad_norm": 0.025896169245243073,
"learning_rate": 2.4894922301439643e-06,
"loss": 0.7825,
"step": 1366
},
{
"epoch": 1.5792057761732852,
"grad_norm": 0.02581976167857647,
"learning_rate": 2.478734659527501e-06,
"loss": 0.7783,
"step": 1367
},
{
"epoch": 1.580361010830325,
"grad_norm": 0.025588102638721466,
"learning_rate": 2.467977487386497e-06,
"loss": 0.7617,
"step": 1368
},
{
"epoch": 1.5815162454873646,
"grad_norm": 0.02477039210498333,
"learning_rate": 2.4572209129494524e-06,
"loss": 0.7494,
"step": 1369
},
{
"epoch": 1.5826714801444042,
"grad_norm": 0.027211980894207954,
"learning_rate": 2.4464651354338037e-06,
"loss": 0.8013,
"step": 1370
},
{
"epoch": 1.583826714801444,
"grad_norm": 0.02574116736650467,
"learning_rate": 2.435710354042225e-06,
"loss": 0.7706,
"step": 1371
},
{
"epoch": 1.5849819494584838,
"grad_norm": 0.025494717061519623,
"learning_rate": 2.4249567679589417e-06,
"loss": 0.8006,
"step": 1372
},
{
"epoch": 1.5861371841155236,
"grad_norm": 0.024281838908791542,
"learning_rate": 2.4142045763460415e-06,
"loss": 0.7609,
"step": 1373
},
{
"epoch": 1.5872924187725632,
"grad_norm": 0.025773445144295692,
"learning_rate": 2.403453978339787e-06,
"loss": 0.8048,
"step": 1374
},
{
"epoch": 1.5884476534296028,
"grad_norm": 0.02534416690468788,
"learning_rate": 2.392705173046925e-06,
"loss": 0.7731,
"step": 1375
},
{
"epoch": 1.5896028880866426,
"grad_norm": 0.02497626096010208,
"learning_rate": 2.3819583595410013e-06,
"loss": 0.7858,
"step": 1376
},
{
"epoch": 1.5907581227436824,
"grad_norm": 0.02502652071416378,
"learning_rate": 2.3712137368586717e-06,
"loss": 0.7438,
"step": 1377
},
{
"epoch": 1.591913357400722,
"grad_norm": 0.024966862052679062,
"learning_rate": 2.3604715039960173e-06,
"loss": 0.7835,
"step": 1378
},
{
"epoch": 1.5930685920577616,
"grad_norm": 0.025360535830259323,
"learning_rate": 2.3497318599048592e-06,
"loss": 0.7716,
"step": 1379
},
{
"epoch": 1.5942238267148015,
"grad_norm": 0.026140306144952774,
"learning_rate": 2.338995003489072e-06,
"loss": 0.7556,
"step": 1380
},
{
"epoch": 1.5953790613718413,
"grad_norm": 0.02430606074631214,
"learning_rate": 2.328261133600902e-06,
"loss": 0.7708,
"step": 1381
},
{
"epoch": 1.5965342960288809,
"grad_norm": 0.02544374018907547,
"learning_rate": 2.317530449037281e-06,
"loss": 0.7837,
"step": 1382
},
{
"epoch": 1.5976895306859205,
"grad_norm": 0.025840749964118004,
"learning_rate": 2.3068031485361516e-06,
"loss": 0.7575,
"step": 1383
},
{
"epoch": 1.5988447653429603,
"grad_norm": 0.027307111769914627,
"learning_rate": 2.296079430772777e-06,
"loss": 0.8002,
"step": 1384
},
{
"epoch": 1.6,
"grad_norm": 0.026851478964090347,
"learning_rate": 2.285359494356068e-06,
"loss": 0.8076,
"step": 1385
},
{
"epoch": 1.6011552346570397,
"grad_norm": 0.025269558653235435,
"learning_rate": 2.274643537824905e-06,
"loss": 0.7601,
"step": 1386
},
{
"epoch": 1.6023104693140793,
"grad_norm": 0.024502936750650406,
"learning_rate": 2.263931759644456e-06,
"loss": 0.7777,
"step": 1387
},
{
"epoch": 1.603465703971119,
"grad_norm": 0.025695212185382843,
"learning_rate": 2.2532243582025048e-06,
"loss": 0.7779,
"step": 1388
},
{
"epoch": 1.604620938628159,
"grad_norm": 0.025510141626000404,
"learning_rate": 2.242521531805775e-06,
"loss": 0.7623,
"step": 1389
},
{
"epoch": 1.6057761732851985,
"grad_norm": 0.025397466495633125,
"learning_rate": 2.2318234786762586e-06,
"loss": 0.769,
"step": 1390
},
{
"epoch": 1.6069314079422383,
"grad_norm": 0.02594076469540596,
"learning_rate": 2.2211303969475444e-06,
"loss": 0.8051,
"step": 1391
},
{
"epoch": 1.608086642599278,
"grad_norm": 0.024837691336870193,
"learning_rate": 2.210442484661148e-06,
"loss": 0.8117,
"step": 1392
},
{
"epoch": 1.6092418772563177,
"grad_norm": 0.026649747043848038,
"learning_rate": 2.199759939762843e-06,
"loss": 0.7768,
"step": 1393
},
{
"epoch": 1.6103971119133575,
"grad_norm": 0.026813820004463196,
"learning_rate": 2.1890829600989995e-06,
"loss": 0.7575,
"step": 1394
},
{
"epoch": 1.6115523465703971,
"grad_norm": 0.026133766397833824,
"learning_rate": 2.1784117434129113e-06,
"loss": 0.7599,
"step": 1395
},
{
"epoch": 1.6127075812274367,
"grad_norm": 0.026183927431702614,
"learning_rate": 2.167746487341144e-06,
"loss": 0.7661,
"step": 1396
},
{
"epoch": 1.6138628158844766,
"grad_norm": 0.02729174494743347,
"learning_rate": 2.157087389409867e-06,
"loss": 0.7698,
"step": 1397
},
{
"epoch": 1.6150180505415164,
"grad_norm": 0.026295259594917297,
"learning_rate": 2.1464346470311965e-06,
"loss": 0.7695,
"step": 1398
},
{
"epoch": 1.616173285198556,
"grad_norm": 0.026081614196300507,
"learning_rate": 2.135788457499544e-06,
"loss": 0.7622,
"step": 1399
},
{
"epoch": 1.6173285198555956,
"grad_norm": 0.025280749425292015,
"learning_rate": 2.1251490179879547e-06,
"loss": 0.7639,
"step": 1400
},
{
"epoch": 1.6184837545126354,
"grad_norm": 0.02538752183318138,
"learning_rate": 2.1145165255444643e-06,
"loss": 0.7893,
"step": 1401
},
{
"epoch": 1.6196389891696752,
"grad_norm": 0.026337897405028343,
"learning_rate": 2.1038911770884395e-06,
"loss": 0.7936,
"step": 1402
},
{
"epoch": 1.6207942238267148,
"grad_norm": 0.027165520936250687,
"learning_rate": 2.0932731694069438e-06,
"loss": 0.7506,
"step": 1403
},
{
"epoch": 1.6219494584837544,
"grad_norm": 0.02547120675444603,
"learning_rate": 2.0826626991510775e-06,
"loss": 0.7623,
"step": 1404
},
{
"epoch": 1.6231046931407942,
"grad_norm": 0.023889468982815742,
"learning_rate": 2.072059962832352e-06,
"loss": 0.7444,
"step": 1405
},
{
"epoch": 1.624259927797834,
"grad_norm": 0.025629183277487755,
"learning_rate": 2.0614651568190354e-06,
"loss": 0.7869,
"step": 1406
},
{
"epoch": 1.6254151624548736,
"grad_norm": 0.02670186199247837,
"learning_rate": 2.050878477332526e-06,
"loss": 0.8075,
"step": 1407
},
{
"epoch": 1.6265703971119132,
"grad_norm": 0.025403102859854698,
"learning_rate": 2.040300120443714e-06,
"loss": 0.7647,
"step": 1408
},
{
"epoch": 1.627725631768953,
"grad_norm": 0.025784939527511597,
"learning_rate": 2.02973028206935e-06,
"loss": 0.7594,
"step": 1409
},
{
"epoch": 1.6288808664259928,
"grad_norm": 0.024970732629299164,
"learning_rate": 2.019169157968418e-06,
"loss": 0.7664,
"step": 1410
},
{
"epoch": 1.6300361010830327,
"grad_norm": 0.025724420323967934,
"learning_rate": 2.008616943738508e-06,
"loss": 0.7637,
"step": 1411
},
{
"epoch": 1.6311913357400722,
"grad_norm": 0.02508534863591194,
"learning_rate": 1.9980738348121955e-06,
"loss": 0.7674,
"step": 1412
},
{
"epoch": 1.6323465703971118,
"grad_norm": 0.024077199399471283,
"learning_rate": 1.9875400264534194e-06,
"loss": 0.7703,
"step": 1413
},
{
"epoch": 1.6335018050541517,
"grad_norm": 0.02483062632381916,
"learning_rate": 1.97701571375387e-06,
"loss": 0.7825,
"step": 1414
},
{
"epoch": 1.6346570397111915,
"grad_norm": 0.025150980800390244,
"learning_rate": 1.9665010916293704e-06,
"loss": 0.7875,
"step": 1415
},
{
"epoch": 1.635812274368231,
"grad_norm": 0.02459767460823059,
"learning_rate": 1.95599635481627e-06,
"loss": 0.7745,
"step": 1416
},
{
"epoch": 1.6369675090252707,
"grad_norm": 0.024894608184695244,
"learning_rate": 1.9455016978678377e-06,
"loss": 0.7825,
"step": 1417
},
{
"epoch": 1.6381227436823105,
"grad_norm": 0.02483339235186577,
"learning_rate": 1.935017315150656e-06,
"loss": 0.7654,
"step": 1418
},
{
"epoch": 1.6392779783393503,
"grad_norm": 0.025194387882947922,
"learning_rate": 1.9245434008410273e-06,
"loss": 0.7513,
"step": 1419
},
{
"epoch": 1.6404332129963899,
"grad_norm": 0.025890490040183067,
"learning_rate": 1.914080148921368e-06,
"loss": 0.7756,
"step": 1420
},
{
"epoch": 1.6415884476534295,
"grad_norm": 0.025699496269226074,
"learning_rate": 1.903627753176627e-06,
"loss": 0.7702,
"step": 1421
},
{
"epoch": 1.6427436823104693,
"grad_norm": 0.02435440570116043,
"learning_rate": 1.893186407190685e-06,
"loss": 0.7727,
"step": 1422
},
{
"epoch": 1.6438989169675091,
"grad_norm": 0.024738416075706482,
"learning_rate": 1.8827563043427818e-06,
"loss": 0.7856,
"step": 1423
},
{
"epoch": 1.6450541516245487,
"grad_norm": 0.025477735325694084,
"learning_rate": 1.8723376378039229e-06,
"loss": 0.775,
"step": 1424
},
{
"epoch": 1.6462093862815883,
"grad_norm": 0.02498653531074524,
"learning_rate": 1.8619306005333118e-06,
"loss": 0.7464,
"step": 1425
},
{
"epoch": 1.6473646209386281,
"grad_norm": 0.025513045489788055,
"learning_rate": 1.851535385274768e-06,
"loss": 0.7944,
"step": 1426
},
{
"epoch": 1.648519855595668,
"grad_norm": 0.024236127734184265,
"learning_rate": 1.8411521845531638e-06,
"loss": 0.7646,
"step": 1427
},
{
"epoch": 1.6496750902527075,
"grad_norm": 0.025221243500709534,
"learning_rate": 1.8307811906708552e-06,
"loss": 0.7802,
"step": 1428
},
{
"epoch": 1.6508303249097471,
"grad_norm": 0.024581240490078926,
"learning_rate": 1.8204225957041197e-06,
"loss": 0.7686,
"step": 1429
},
{
"epoch": 1.651985559566787,
"grad_norm": 0.025168737396597862,
"learning_rate": 1.8100765914996026e-06,
"loss": 0.79,
"step": 1430
},
{
"epoch": 1.6531407942238268,
"grad_norm": 0.024541551247239113,
"learning_rate": 1.7997433696707597e-06,
"loss": 0.766,
"step": 1431
},
{
"epoch": 1.6542960288808666,
"grad_norm": 0.02476644702255726,
"learning_rate": 1.789423121594312e-06,
"loss": 0.7718,
"step": 1432
},
{
"epoch": 1.6554512635379062,
"grad_norm": 0.027230584993958473,
"learning_rate": 1.779116038406698e-06,
"loss": 0.7813,
"step": 1433
},
{
"epoch": 1.6566064981949458,
"grad_norm": 0.025931481271982193,
"learning_rate": 1.768822311000538e-06,
"loss": 0.7772,
"step": 1434
},
{
"epoch": 1.6577617328519856,
"grad_norm": 0.025082925334572792,
"learning_rate": 1.7585421300210932e-06,
"loss": 0.7958,
"step": 1435
},
{
"epoch": 1.6589169675090254,
"grad_norm": 0.02672106772661209,
"learning_rate": 1.7482756858627406e-06,
"loss": 0.774,
"step": 1436
},
{
"epoch": 1.660072202166065,
"grad_norm": 0.026696957647800446,
"learning_rate": 1.7380231686654419e-06,
"loss": 0.7763,
"step": 1437
},
{
"epoch": 1.6612274368231046,
"grad_norm": 0.025145625695586205,
"learning_rate": 1.7277847683112251e-06,
"loss": 0.7602,
"step": 1438
},
{
"epoch": 1.6623826714801444,
"grad_norm": 0.025679778307676315,
"learning_rate": 1.7175606744206683e-06,
"loss": 0.7772,
"step": 1439
},
{
"epoch": 1.6635379061371842,
"grad_norm": 0.02643396146595478,
"learning_rate": 1.707351076349382e-06,
"loss": 0.7776,
"step": 1440
},
{
"epoch": 1.6646931407942238,
"grad_norm": 0.025024591013789177,
"learning_rate": 1.697156163184512e-06,
"loss": 0.769,
"step": 1441
},
{
"epoch": 1.6658483754512634,
"grad_norm": 0.02451016753911972,
"learning_rate": 1.6869761237412275e-06,
"loss": 0.7614,
"step": 1442
},
{
"epoch": 1.6670036101083032,
"grad_norm": 0.02593870460987091,
"learning_rate": 1.6768111465592312e-06,
"loss": 0.7828,
"step": 1443
},
{
"epoch": 1.668158844765343,
"grad_norm": 0.024905387312173843,
"learning_rate": 1.6666614198992625e-06,
"loss": 0.7847,
"step": 1444
},
{
"epoch": 1.6693140794223826,
"grad_norm": 0.025363383814692497,
"learning_rate": 1.6565271317396156e-06,
"loss": 0.7757,
"step": 1445
},
{
"epoch": 1.6704693140794222,
"grad_norm": 0.02425825409591198,
"learning_rate": 1.6464084697726532e-06,
"loss": 0.756,
"step": 1446
},
{
"epoch": 1.671624548736462,
"grad_norm": 0.025201529264450073,
"learning_rate": 1.636305621401335e-06,
"loss": 0.7701,
"step": 1447
},
{
"epoch": 1.6727797833935019,
"grad_norm": 0.02498772367835045,
"learning_rate": 1.626218773735742e-06,
"loss": 0.8012,
"step": 1448
},
{
"epoch": 1.6739350180505417,
"grad_norm": 0.02500982955098152,
"learning_rate": 1.6161481135896154e-06,
"loss": 0.7796,
"step": 1449
},
{
"epoch": 1.6750902527075813,
"grad_norm": 0.025785304605960846,
"learning_rate": 1.606093827476895e-06,
"loss": 0.7584,
"step": 1450
},
{
"epoch": 1.6762454873646209,
"grad_norm": 0.024516083300113678,
"learning_rate": 1.5960561016082638e-06,
"loss": 0.7812,
"step": 1451
},
{
"epoch": 1.6774007220216607,
"grad_norm": 0.025357436388731003,
"learning_rate": 1.5860351218877021e-06,
"loss": 0.7642,
"step": 1452
},
{
"epoch": 1.6785559566787005,
"grad_norm": 0.025423161685466766,
"learning_rate": 1.5760310739090402e-06,
"loss": 0.7838,
"step": 1453
},
{
"epoch": 1.67971119133574,
"grad_norm": 0.025020305067300797,
"learning_rate": 1.5660441429525258e-06,
"loss": 0.7713,
"step": 1454
},
{
"epoch": 1.6808664259927797,
"grad_norm": 0.02521519362926483,
"learning_rate": 1.556074513981389e-06,
"loss": 0.8022,
"step": 1455
},
{
"epoch": 1.6820216606498195,
"grad_norm": 0.02515571005642414,
"learning_rate": 1.5461223716384196e-06,
"loss": 0.7794,
"step": 1456
},
{
"epoch": 1.6831768953068593,
"grad_norm": 0.025751549750566483,
"learning_rate": 1.5361879002425432e-06,
"loss": 0.8067,
"step": 1457
},
{
"epoch": 1.684332129963899,
"grad_norm": 0.024167675524950027,
"learning_rate": 1.5262712837854139e-06,
"loss": 0.7497,
"step": 1458
},
{
"epoch": 1.6854873646209385,
"grad_norm": 0.025504915043711662,
"learning_rate": 1.5163727059279972e-06,
"loss": 0.7645,
"step": 1459
},
{
"epoch": 1.6866425992779783,
"grad_norm": 0.026490593329072,
"learning_rate": 1.506492349997181e-06,
"loss": 0.7698,
"step": 1460
},
{
"epoch": 1.6877978339350181,
"grad_norm": 0.024468280375003815,
"learning_rate": 1.4966303989823688e-06,
"loss": 0.7541,
"step": 1461
},
{
"epoch": 1.6889530685920577,
"grad_norm": 0.02487323433160782,
"learning_rate": 1.4867870355320972e-06,
"loss": 0.7611,
"step": 1462
},
{
"epoch": 1.6901083032490973,
"grad_norm": 0.024844063445925713,
"learning_rate": 1.4769624419506507e-06,
"loss": 0.7598,
"step": 1463
},
{
"epoch": 1.6912635379061371,
"grad_norm": 0.024609016254544258,
"learning_rate": 1.4671568001946851e-06,
"loss": 0.7863,
"step": 1464
},
{
"epoch": 1.692418772563177,
"grad_norm": 0.025440316647291183,
"learning_rate": 1.4573702918698615e-06,
"loss": 0.7955,
"step": 1465
},
{
"epoch": 1.6935740072202166,
"grad_norm": 0.024768877774477005,
"learning_rate": 1.4476030982274746e-06,
"loss": 0.7666,
"step": 1466
},
{
"epoch": 1.6947292418772562,
"grad_norm": 0.024406442418694496,
"learning_rate": 1.4378554001611087e-06,
"loss": 0.747,
"step": 1467
},
{
"epoch": 1.695884476534296,
"grad_norm": 0.02518375776708126,
"learning_rate": 1.4281273782032696e-06,
"loss": 0.7914,
"step": 1468
},
{
"epoch": 1.6970397111913358,
"grad_norm": 0.025701193138957024,
"learning_rate": 1.41841921252206e-06,
"loss": 0.7879,
"step": 1469
},
{
"epoch": 1.6981949458483756,
"grad_norm": 0.025167284533381462,
"learning_rate": 1.4087310829178295e-06,
"loss": 0.7707,
"step": 1470
},
{
"epoch": 1.6993501805054152,
"grad_norm": 0.024784136563539505,
"learning_rate": 1.3990631688198505e-06,
"loss": 0.7753,
"step": 1471
},
{
"epoch": 1.7005054151624548,
"grad_norm": 0.024584993720054626,
"learning_rate": 1.3894156492829932e-06,
"loss": 0.7579,
"step": 1472
},
{
"epoch": 1.7016606498194946,
"grad_norm": 0.025017768144607544,
"learning_rate": 1.3797887029844083e-06,
"loss": 0.7512,
"step": 1473
},
{
"epoch": 1.7028158844765344,
"grad_norm": 0.02527061477303505,
"learning_rate": 1.3701825082202227e-06,
"loss": 0.772,
"step": 1474
},
{
"epoch": 1.703971119133574,
"grad_norm": 0.02521699294447899,
"learning_rate": 1.3605972429022308e-06,
"loss": 0.766,
"step": 1475
},
{
"epoch": 1.7051263537906136,
"grad_norm": 0.024668460711836815,
"learning_rate": 1.3510330845546029e-06,
"loss": 0.7638,
"step": 1476
},
{
"epoch": 1.7062815884476534,
"grad_norm": 0.02491624839603901,
"learning_rate": 1.3414902103105984e-06,
"loss": 0.7572,
"step": 1477
},
{
"epoch": 1.7074368231046932,
"grad_norm": 0.02490483969449997,
"learning_rate": 1.3319687969092844e-06,
"loss": 0.7763,
"step": 1478
},
{
"epoch": 1.7085920577617328,
"grad_norm": 0.02502390742301941,
"learning_rate": 1.3224690206922594e-06,
"loss": 0.7859,
"step": 1479
},
{
"epoch": 1.7097472924187724,
"grad_norm": 0.024547290056943893,
"learning_rate": 1.3129910576003918e-06,
"loss": 0.76,
"step": 1480
},
{
"epoch": 1.7109025270758123,
"grad_norm": 0.024496039375662804,
"learning_rate": 1.3035350831705581e-06,
"loss": 0.7541,
"step": 1481
},
{
"epoch": 1.712057761732852,
"grad_norm": 0.0256884153932333,
"learning_rate": 1.2941012725323943e-06,
"loss": 0.7794,
"step": 1482
},
{
"epoch": 1.7132129963898917,
"grad_norm": 0.025446278974413872,
"learning_rate": 1.2846898004050527e-06,
"loss": 0.789,
"step": 1483
},
{
"epoch": 1.7143682310469313,
"grad_norm": 0.02467747963964939,
"learning_rate": 1.2753008410939593e-06,
"loss": 0.7704,
"step": 1484
},
{
"epoch": 1.715523465703971,
"grad_norm": 0.024466682225465775,
"learning_rate": 1.2659345684875971e-06,
"loss": 0.7751,
"step": 1485
},
{
"epoch": 1.716678700361011,
"grad_norm": 0.02474912814795971,
"learning_rate": 1.2565911560542752e-06,
"loss": 0.7721,
"step": 1486
},
{
"epoch": 1.7178339350180507,
"grad_norm": 0.024523675441741943,
"learning_rate": 1.2472707768389226e-06,
"loss": 0.7638,
"step": 1487
},
{
"epoch": 1.7189891696750903,
"grad_norm": 0.02448326162993908,
"learning_rate": 1.2379736034598793e-06,
"loss": 0.7898,
"step": 1488
},
{
"epoch": 1.72014440433213,
"grad_norm": 0.024665791541337967,
"learning_rate": 1.2286998081057024e-06,
"loss": 0.757,
"step": 1489
},
{
"epoch": 1.7212996389891697,
"grad_norm": 0.024392489343881607,
"learning_rate": 1.219449562531973e-06,
"loss": 0.7756,
"step": 1490
},
{
"epoch": 1.7224548736462095,
"grad_norm": 0.024781666696071625,
"learning_rate": 1.2102230380581221e-06,
"loss": 0.7851,
"step": 1491
},
{
"epoch": 1.7236101083032491,
"grad_norm": 0.024531826376914978,
"learning_rate": 1.2010204055642504e-06,
"loss": 0.7786,
"step": 1492
},
{
"epoch": 1.7247653429602887,
"grad_norm": 0.024737264961004257,
"learning_rate": 1.1918418354879655e-06,
"loss": 0.7918,
"step": 1493
},
{
"epoch": 1.7259205776173285,
"grad_norm": 0.024697288870811462,
"learning_rate": 1.1826874978212304e-06,
"loss": 0.7797,
"step": 1494
},
{
"epoch": 1.7270758122743683,
"grad_norm": 0.023727286607027054,
"learning_rate": 1.1735575621072076e-06,
"loss": 0.7622,
"step": 1495
},
{
"epoch": 1.728231046931408,
"grad_norm": 0.024227799847722054,
"learning_rate": 1.1644521974371236e-06,
"loss": 0.7538,
"step": 1496
},
{
"epoch": 1.7293862815884475,
"grad_norm": 0.02443331852555275,
"learning_rate": 1.1553715724471356e-06,
"loss": 0.7567,
"step": 1497
},
{
"epoch": 1.7305415162454874,
"grad_norm": 0.02441231720149517,
"learning_rate": 1.146315855315211e-06,
"loss": 0.763,
"step": 1498
},
{
"epoch": 1.7316967509025272,
"grad_norm": 0.02491583675146103,
"learning_rate": 1.1372852137580082e-06,
"loss": 0.7519,
"step": 1499
},
{
"epoch": 1.7328519855595668,
"grad_norm": 0.024958504363894463,
"learning_rate": 1.1282798150277738e-06,
"loss": 0.7704,
"step": 1500
},
{
"epoch": 1.7340072202166064,
"grad_norm": 0.02489922009408474,
"learning_rate": 1.1192998259092428e-06,
"loss": 0.8001,
"step": 1501
},
{
"epoch": 1.7351624548736462,
"grad_norm": 0.024233028292655945,
"learning_rate": 1.1103454127165526e-06,
"loss": 0.7725,
"step": 1502
},
{
"epoch": 1.736317689530686,
"grad_norm": 0.02405642904341221,
"learning_rate": 1.1014167412901609e-06,
"loss": 0.7227,
"step": 1503
},
{
"epoch": 1.7374729241877256,
"grad_norm": 0.025087906047701836,
"learning_rate": 1.092513976993772e-06,
"loss": 0.7852,
"step": 1504
},
{
"epoch": 1.7386281588447652,
"grad_norm": 0.025061985477805138,
"learning_rate": 1.083637284711279e-06,
"loss": 0.7648,
"step": 1505
},
{
"epoch": 1.739783393501805,
"grad_norm": 0.024631349369883537,
"learning_rate": 1.074786828843705e-06,
"loss": 0.741,
"step": 1506
},
{
"epoch": 1.7409386281588448,
"grad_norm": 0.024879056960344315,
"learning_rate": 1.0659627733061643e-06,
"loss": 0.7355,
"step": 1507
},
{
"epoch": 1.7420938628158846,
"grad_norm": 0.024683699011802673,
"learning_rate": 1.0571652815248199e-06,
"loss": 0.8,
"step": 1508
},
{
"epoch": 1.7432490974729242,
"grad_norm": 0.02492067962884903,
"learning_rate": 1.0483945164338617e-06,
"loss": 0.7701,
"step": 1509
},
{
"epoch": 1.7444043321299638,
"grad_norm": 0.02489222027361393,
"learning_rate": 1.039650640472486e-06,
"loss": 0.7831,
"step": 1510
},
{
"epoch": 1.7455595667870036,
"grad_norm": 0.024434035643935204,
"learning_rate": 1.0309338155818906e-06,
"loss": 0.7896,
"step": 1511
},
{
"epoch": 1.7467148014440435,
"grad_norm": 0.025025788694620132,
"learning_rate": 1.0222442032022705e-06,
"loss": 0.7806,
"step": 1512
},
{
"epoch": 1.747870036101083,
"grad_norm": 0.024875549599528313,
"learning_rate": 1.0135819642698318e-06,
"loss": 0.7844,
"step": 1513
},
{
"epoch": 1.7490252707581226,
"grad_norm": 0.024595865979790688,
"learning_rate": 1.0049472592138105e-06,
"loss": 0.7766,
"step": 1514
},
{
"epoch": 1.7501805054151625,
"grad_norm": 0.02502664178609848,
"learning_rate": 9.963402479535004e-07,
"loss": 0.7813,
"step": 1515
},
{
"epoch": 1.7513357400722023,
"grad_norm": 0.024631381034851074,
"learning_rate": 9.877610898952902e-07,
"loss": 0.79,
"step": 1516
},
{
"epoch": 1.7524909747292419,
"grad_norm": 0.023990454152226448,
"learning_rate": 9.792099439297134e-07,
"loss": 0.7875,
"step": 1517
},
{
"epoch": 1.7536462093862815,
"grad_norm": 0.0241263285279274,
"learning_rate": 9.706869684285066e-07,
"loss": 0.7605,
"step": 1518
},
{
"epoch": 1.7548014440433213,
"grad_norm": 0.02398371510207653,
"learning_rate": 9.621923212416713e-07,
"loss": 0.7778,
"step": 1519
},
{
"epoch": 1.755956678700361,
"grad_norm": 0.025420811027288437,
"learning_rate": 9.53726159694558e-07,
"loss": 0.7881,
"step": 1520
},
{
"epoch": 1.7571119133574007,
"grad_norm": 0.024340439587831497,
"learning_rate": 9.452886405849418e-07,
"loss": 0.7725,
"step": 1521
},
{
"epoch": 1.7582671480144403,
"grad_norm": 0.02469663880765438,
"learning_rate": 9.36879920180131e-07,
"loss": 0.771,
"step": 1522
},
{
"epoch": 1.75942238267148,
"grad_norm": 0.024449417367577553,
"learning_rate": 9.285001542140625e-07,
"loss": 0.7624,
"step": 1523
},
{
"epoch": 1.76057761732852,
"grad_norm": 0.024617070332169533,
"learning_rate": 9.201494978844251e-07,
"loss": 0.7763,
"step": 1524
},
{
"epoch": 1.7617328519855595,
"grad_norm": 0.02444697730243206,
"learning_rate": 9.118281058497792e-07,
"loss": 0.7634,
"step": 1525
},
{
"epoch": 1.7628880866425993,
"grad_norm": 0.02412438951432705,
"learning_rate": 9.03536132226695e-07,
"loss": 0.7501,
"step": 1526
},
{
"epoch": 1.764043321299639,
"grad_norm": 0.024980876594781876,
"learning_rate": 8.952737305869002e-07,
"loss": 0.802,
"step": 1527
},
{
"epoch": 1.7651985559566787,
"grad_norm": 0.02424163930118084,
"learning_rate": 8.870410539544322e-07,
"loss": 0.7846,
"step": 1528
},
{
"epoch": 1.7663537906137186,
"grad_norm": 0.025128301233053207,
"learning_rate": 8.788382548028056e-07,
"loss": 0.7551,
"step": 1529
},
{
"epoch": 1.7675090252707581,
"grad_norm": 0.02520165592432022,
"learning_rate": 8.706654850521883e-07,
"loss": 0.7736,
"step": 1530
},
{
"epoch": 1.7686642599277977,
"grad_norm": 0.02502501755952835,
"learning_rate": 8.625228960665893e-07,
"loss": 0.7632,
"step": 1531
},
{
"epoch": 1.7698194945848376,
"grad_norm": 0.02456534653902054,
"learning_rate": 8.544106386510526e-07,
"loss": 0.7514,
"step": 1532
},
{
"epoch": 1.7709747292418774,
"grad_norm": 0.02477426454424858,
"learning_rate": 8.463288630488653e-07,
"loss": 0.7741,
"step": 1533
},
{
"epoch": 1.772129963898917,
"grad_norm": 0.02434214949607849,
"learning_rate": 8.382777189387751e-07,
"loss": 0.7721,
"step": 1534
},
{
"epoch": 1.7732851985559566,
"grad_norm": 0.023917306214571,
"learning_rate": 8.302573554322192e-07,
"loss": 0.7507,
"step": 1535
},
{
"epoch": 1.7744404332129964,
"grad_norm": 0.024669747799634933,
"learning_rate": 8.222679210705636e-07,
"loss": 0.7993,
"step": 1536
},
{
"epoch": 1.7755956678700362,
"grad_norm": 0.024916965514421463,
"learning_rate": 8.143095638223446e-07,
"loss": 0.7536,
"step": 1537
},
{
"epoch": 1.7767509025270758,
"grad_norm": 0.02413412742316723,
"learning_rate": 8.063824310805396e-07,
"loss": 0.7552,
"step": 1538
},
{
"epoch": 1.7779061371841154,
"grad_norm": 0.025234825909137726,
"learning_rate": 7.984866696598284e-07,
"loss": 0.7845,
"step": 1539
},
{
"epoch": 1.7790613718411552,
"grad_norm": 0.024565840139985085,
"learning_rate": 7.906224257938796e-07,
"loss": 0.7674,
"step": 1540
},
{
"epoch": 1.780216606498195,
"grad_norm": 0.024029076099395752,
"learning_rate": 7.827898451326386e-07,
"loss": 0.7637,
"step": 1541
},
{
"epoch": 1.7813718411552346,
"grad_norm": 0.0241931714117527,
"learning_rate": 7.749890727396312e-07,
"loss": 0.7754,
"step": 1542
},
{
"epoch": 1.7825270758122742,
"grad_norm": 0.0242807324975729,
"learning_rate": 7.672202530892778e-07,
"loss": 0.7751,
"step": 1543
},
{
"epoch": 1.783682310469314,
"grad_norm": 0.024539347738027573,
"learning_rate": 7.594835300642188e-07,
"loss": 0.7955,
"step": 1544
},
{
"epoch": 1.7848375451263538,
"grad_norm": 0.024527210742235184,
"learning_rate": 7.517790469526456e-07,
"loss": 0.7713,
"step": 1545
},
{
"epoch": 1.7859927797833937,
"grad_norm": 0.024129964411258698,
"learning_rate": 7.441069464456497e-07,
"loss": 0.7888,
"step": 1546
},
{
"epoch": 1.7871480144404333,
"grad_norm": 0.024345900863409042,
"learning_rate": 7.364673706345822e-07,
"loss": 0.7548,
"step": 1547
},
{
"epoch": 1.7883032490974728,
"grad_norm": 0.024231646209955215,
"learning_rate": 7.288604610084172e-07,
"loss": 0.7645,
"step": 1548
},
{
"epoch": 1.7894584837545127,
"grad_norm": 0.023612959310412407,
"learning_rate": 7.212863584511341e-07,
"loss": 0.7565,
"step": 1549
},
{
"epoch": 1.7906137184115525,
"grad_norm": 0.02500954270362854,
"learning_rate": 7.137452032391084e-07,
"loss": 0.7646,
"step": 1550
},
{
"epoch": 1.791768953068592,
"grad_norm": 0.024092212319374084,
"learning_rate": 7.062371350385143e-07,
"loss": 0.7667,
"step": 1551
},
{
"epoch": 1.7929241877256317,
"grad_norm": 0.023889616131782532,
"learning_rate": 6.987622929027362e-07,
"loss": 0.7848,
"step": 1552
},
{
"epoch": 1.7940794223826715,
"grad_norm": 0.024128958582878113,
"learning_rate": 6.913208152697935e-07,
"loss": 0.7667,
"step": 1553
},
{
"epoch": 1.7952346570397113,
"grad_norm": 0.024358974769711494,
"learning_rate": 6.839128399597784e-07,
"loss": 0.7742,
"step": 1554
},
{
"epoch": 1.796389891696751,
"grad_norm": 0.024110984057188034,
"learning_rate": 6.765385041723024e-07,
"loss": 0.7783,
"step": 1555
},
{
"epoch": 1.7975451263537905,
"grad_norm": 0.02420348860323429,
"learning_rate": 6.691979444839561e-07,
"loss": 0.7737,
"step": 1556
},
{
"epoch": 1.7987003610108303,
"grad_norm": 0.024441605433821678,
"learning_rate": 6.618912968457771e-07,
"loss": 0.7767,
"step": 1557
},
{
"epoch": 1.7998555956678701,
"grad_norm": 0.024723242968320847,
"learning_rate": 6.546186965807347e-07,
"loss": 0.7607,
"step": 1558
},
{
"epoch": 1.8010108303249097,
"grad_norm": 0.025666577741503716,
"learning_rate": 6.473802783812225e-07,
"loss": 0.8028,
"step": 1559
},
{
"epoch": 1.8021660649819493,
"grad_norm": 0.024900216609239578,
"learning_rate": 6.401761763065661e-07,
"loss": 0.7452,
"step": 1560
},
{
"epoch": 1.8033212996389891,
"grad_norm": 0.024593856185674667,
"learning_rate": 6.330065237805361e-07,
"loss": 0.7879,
"step": 1561
},
{
"epoch": 1.804476534296029,
"grad_norm": 0.024126332253217697,
"learning_rate": 6.258714535888803e-07,
"loss": 0.7687,
"step": 1562
},
{
"epoch": 1.8056317689530685,
"grad_norm": 0.02449883334338665,
"learning_rate": 6.187710978768624e-07,
"loss": 0.7622,
"step": 1563
},
{
"epoch": 1.8067870036101084,
"grad_norm": 0.024287080392241478,
"learning_rate": 6.117055881468175e-07,
"loss": 0.7863,
"step": 1564
},
{
"epoch": 1.807942238267148,
"grad_norm": 0.024112451821565628,
"learning_rate": 6.046750552557139e-07,
"loss": 0.7887,
"step": 1565
},
{
"epoch": 1.8090974729241878,
"grad_norm": 0.02391749620437622,
"learning_rate": 5.976796294127286e-07,
"loss": 0.7687,
"step": 1566
},
{
"epoch": 1.8102527075812276,
"grad_norm": 0.023689934983849525,
"learning_rate": 5.907194401768411e-07,
"loss": 0.7705,
"step": 1567
},
{
"epoch": 1.8114079422382672,
"grad_norm": 0.025095578283071518,
"learning_rate": 5.837946164544276e-07,
"loss": 0.805,
"step": 1568
},
{
"epoch": 1.8125631768953068,
"grad_norm": 0.02462933026254177,
"learning_rate": 5.769052864968768e-07,
"loss": 0.7699,
"step": 1569
},
{
"epoch": 1.8137184115523466,
"grad_norm": 0.023868851363658905,
"learning_rate": 5.700515778982142e-07,
"loss": 0.7506,
"step": 1570
},
{
"epoch": 1.8148736462093864,
"grad_norm": 0.024339932948350906,
"learning_rate": 5.632336175927404e-07,
"loss": 0.7663,
"step": 1571
},
{
"epoch": 1.816028880866426,
"grad_norm": 0.023983048275113106,
"learning_rate": 5.564515318526759e-07,
"loss": 0.7555,
"step": 1572
},
{
"epoch": 1.8171841155234656,
"grad_norm": 0.024354344233870506,
"learning_rate": 5.497054462858296e-07,
"loss": 0.7856,
"step": 1573
},
{
"epoch": 1.8183393501805054,
"grad_norm": 0.02329828403890133,
"learning_rate": 5.429954858332623e-07,
"loss": 0.7755,
"step": 1574
},
{
"epoch": 1.8194945848375452,
"grad_norm": 0.024324821308255196,
"learning_rate": 5.363217747669843e-07,
"loss": 0.7871,
"step": 1575
},
{
"epoch": 1.8206498194945848,
"grad_norm": 0.02310972288250923,
"learning_rate": 5.296844366876441e-07,
"loss": 0.7499,
"step": 1576
},
{
"epoch": 1.8218050541516244,
"grad_norm": 0.02391042560338974,
"learning_rate": 5.230835945222463e-07,
"loss": 0.7941,
"step": 1577
},
{
"epoch": 1.8229602888086642,
"grad_norm": 0.024979131296277046,
"learning_rate": 5.165193705218699e-07,
"loss": 0.7719,
"step": 1578
},
{
"epoch": 1.824115523465704,
"grad_norm": 0.024278726428747177,
"learning_rate": 5.099918862594065e-07,
"loss": 0.7815,
"step": 1579
},
{
"epoch": 1.8252707581227436,
"grad_norm": 0.024197446182370186,
"learning_rate": 5.035012626273094e-07,
"loss": 0.7538,
"step": 1580
},
{
"epoch": 1.8264259927797832,
"grad_norm": 0.02464877814054489,
"learning_rate": 4.970476198353524e-07,
"loss": 0.764,
"step": 1581
},
{
"epoch": 1.827581227436823,
"grad_norm": 0.02464544028043747,
"learning_rate": 4.906310774084055e-07,
"loss": 0.787,
"step": 1582
},
{
"epoch": 1.8287364620938629,
"grad_norm": 0.023831835016608238,
"learning_rate": 4.842517541842186e-07,
"loss": 0.7527,
"step": 1583
},
{
"epoch": 1.8298916967509027,
"grad_norm": 0.02423412911593914,
"learning_rate": 4.779097683112254e-07,
"loss": 0.7639,
"step": 1584
},
{
"epoch": 1.8310469314079423,
"grad_norm": 0.023963099345564842,
"learning_rate": 4.7160523724634964e-07,
"loss": 0.7676,
"step": 1585
},
{
"epoch": 1.8322021660649819,
"grad_norm": 0.023802466690540314,
"learning_rate": 4.653382777528332e-07,
"loss": 0.7805,
"step": 1586
},
{
"epoch": 1.8333574007220217,
"grad_norm": 0.024113498628139496,
"learning_rate": 4.5910900589807164e-07,
"loss": 0.7578,
"step": 1587
},
{
"epoch": 1.8345126353790615,
"grad_norm": 0.024253182113170624,
"learning_rate": 4.5291753705146685e-07,
"loss": 0.7816,
"step": 1588
},
{
"epoch": 1.835667870036101,
"grad_norm": 0.024671630933880806,
"learning_rate": 4.4676398588229044e-07,
"loss": 0.7712,
"step": 1589
},
{
"epoch": 1.8368231046931407,
"grad_norm": 0.02365119196474552,
"learning_rate": 4.40648466357554e-07,
"loss": 0.7673,
"step": 1590
},
{
"epoch": 1.8379783393501805,
"grad_norm": 0.024656254798173904,
"learning_rate": 4.3457109173990736e-07,
"loss": 0.7627,
"step": 1591
},
{
"epoch": 1.8391335740072203,
"grad_norm": 0.02385905385017395,
"learning_rate": 4.285319745855341e-07,
"loss": 0.7758,
"step": 1592
},
{
"epoch": 1.84028880866426,
"grad_norm": 0.02395222708582878,
"learning_rate": 4.2253122674207165e-07,
"loss": 0.7763,
"step": 1593
},
{
"epoch": 1.8414440433212995,
"grad_norm": 0.02432720549404621,
"learning_rate": 4.165689593465352e-07,
"loss": 0.7873,
"step": 1594
},
{
"epoch": 1.8425992779783393,
"grad_norm": 0.02468527853488922,
"learning_rate": 4.1064528282326263e-07,
"loss": 0.7517,
"step": 1595
},
{
"epoch": 1.8437545126353792,
"grad_norm": 0.02376159094274044,
"learning_rate": 4.047603068818685e-07,
"loss": 0.7715,
"step": 1596
},
{
"epoch": 1.8449097472924187,
"grad_norm": 0.02394728921353817,
"learning_rate": 3.9891414051521345e-07,
"loss": 0.7697,
"step": 1597
},
{
"epoch": 1.8460649819494583,
"grad_norm": 0.024393731728196144,
"learning_rate": 3.93106891997382e-07,
"loss": 0.7651,
"step": 1598
},
{
"epoch": 1.8472202166064982,
"grad_norm": 0.02447488158941269,
"learning_rate": 3.8733866888168026e-07,
"loss": 0.7862,
"step": 1599
},
{
"epoch": 1.848375451263538,
"grad_norm": 0.023694733157753944,
"learning_rate": 3.816095779986455e-07,
"loss": 0.7704,
"step": 1600
},
{
"epoch": 1.8495306859205776,
"grad_norm": 0.023735985159873962,
"learning_rate": 3.759197254540624e-07,
"loss": 0.7712,
"step": 1601
},
{
"epoch": 1.8506859205776174,
"grad_norm": 0.02489444985985756,
"learning_rate": 3.702692166270027e-07,
"loss": 0.7918,
"step": 1602
},
{
"epoch": 1.851841155234657,
"grad_norm": 0.02421208843588829,
"learning_rate": 3.6465815616787e-07,
"loss": 0.7888,
"step": 1603
},
{
"epoch": 1.8529963898916968,
"grad_norm": 0.0240098275244236,
"learning_rate": 3.590866479964664e-07,
"loss": 0.747,
"step": 1604
},
{
"epoch": 1.8541516245487366,
"grad_norm": 0.025053462013602257,
"learning_rate": 3.535547953000618e-07,
"loss": 0.7872,
"step": 1605
},
{
"epoch": 1.8553068592057762,
"grad_norm": 0.023561321198940277,
"learning_rate": 3.4806270053148703e-07,
"loss": 0.7748,
"step": 1606
},
{
"epoch": 1.8564620938628158,
"grad_norm": 0.023926684632897377,
"learning_rate": 3.4261046540723424e-07,
"loss": 0.7576,
"step": 1607
},
{
"epoch": 1.8576173285198556,
"grad_norm": 0.023767825216054916,
"learning_rate": 3.371981909055759e-07,
"loss": 0.7577,
"step": 1608
},
{
"epoch": 1.8587725631768954,
"grad_norm": 0.0232445877045393,
"learning_rate": 3.318259772646901e-07,
"loss": 0.7442,
"step": 1609
},
{
"epoch": 1.859927797833935,
"grad_norm": 0.024164149537682533,
"learning_rate": 3.264939239808092e-07,
"loss": 0.7589,
"step": 1610
},
{
"epoch": 1.8610830324909746,
"grad_norm": 0.023966316133737564,
"learning_rate": 3.2120212980637286e-07,
"loss": 0.768,
"step": 1611
},
{
"epoch": 1.8622382671480144,
"grad_norm": 0.023847166448831558,
"learning_rate": 3.1595069274820076e-07,
"loss": 0.7603,
"step": 1612
},
{
"epoch": 1.8633935018050543,
"grad_norm": 0.02418961189687252,
"learning_rate": 3.1073971006567867e-07,
"loss": 0.7814,
"step": 1613
},
{
"epoch": 1.8645487364620938,
"grad_norm": 0.023753009736537933,
"learning_rate": 3.055692782689551e-07,
"loss": 0.7667,
"step": 1614
},
{
"epoch": 1.8657039711191334,
"grad_norm": 0.02413536049425602,
"learning_rate": 3.0043949311715463e-07,
"loss": 0.7826,
"step": 1615
},
{
"epoch": 1.8668592057761733,
"grad_norm": 0.024381062015891075,
"learning_rate": 2.953504496166049e-07,
"loss": 0.7678,
"step": 1616
},
{
"epoch": 1.868014440433213,
"grad_norm": 0.02448740415275097,
"learning_rate": 2.903022420190769e-07,
"loss": 0.7751,
"step": 1617
},
{
"epoch": 1.8691696750902527,
"grad_norm": 0.024086831137537956,
"learning_rate": 2.8529496382003895e-07,
"loss": 0.7799,
"step": 1618
},
{
"epoch": 1.8703249097472923,
"grad_norm": 0.02341514825820923,
"learning_rate": 2.8032870775692436e-07,
"loss": 0.7723,
"step": 1619
},
{
"epoch": 1.871480144404332,
"grad_norm": 0.023932676762342453,
"learning_rate": 2.75403565807417e-07,
"loss": 0.7635,
"step": 1620
},
{
"epoch": 1.872635379061372,
"grad_norm": 0.023613903671503067,
"learning_rate": 2.705196291877439e-07,
"loss": 0.7605,
"step": 1621
},
{
"epoch": 1.8737906137184117,
"grad_norm": 0.024287506937980652,
"learning_rate": 2.6567698835098765e-07,
"loss": 0.7815,
"step": 1622
},
{
"epoch": 1.8749458483754513,
"grad_norm": 0.023454533889889717,
"learning_rate": 2.608757329854116e-07,
"loss": 0.759,
"step": 1623
},
{
"epoch": 1.876101083032491,
"grad_norm": 0.024026375263929367,
"learning_rate": 2.561159520127993e-07,
"loss": 0.7847,
"step": 1624
},
{
"epoch": 1.8772563176895307,
"grad_norm": 0.024074561893939972,
"learning_rate": 2.5139773358680497e-07,
"loss": 0.7734,
"step": 1625
},
{
"epoch": 1.8784115523465705,
"grad_norm": 0.024127675220370293,
"learning_rate": 2.467211650913235e-07,
"loss": 0.7627,
"step": 1626
},
{
"epoch": 1.8795667870036101,
"grad_norm": 0.024527592584490776,
"learning_rate": 2.4208633313887026e-07,
"loss": 0.8107,
"step": 1627
},
{
"epoch": 1.8807220216606497,
"grad_norm": 0.02324732393026352,
"learning_rate": 2.3749332356897935e-07,
"loss": 0.7739,
"step": 1628
},
{
"epoch": 1.8818772563176895,
"grad_norm": 0.02475915476679802,
"learning_rate": 2.3294222144661086e-07,
"loss": 0.778,
"step": 1629
},
{
"epoch": 1.8830324909747294,
"grad_norm": 0.024105004966259003,
"learning_rate": 2.2843311106057757e-07,
"loss": 0.7772,
"step": 1630
},
{
"epoch": 1.884187725631769,
"grad_norm": 0.02386721596121788,
"learning_rate": 2.2396607592198304e-07,
"loss": 0.7631,
"step": 1631
},
{
"epoch": 1.8853429602888085,
"grad_norm": 0.024259351193904877,
"learning_rate": 2.1954119876267445e-07,
"loss": 0.7966,
"step": 1632
},
{
"epoch": 1.8864981949458484,
"grad_norm": 0.023345720022916794,
"learning_rate": 2.1515856153371186e-07,
"loss": 0.7685,
"step": 1633
},
{
"epoch": 1.8876534296028882,
"grad_norm": 0.024387696757912636,
"learning_rate": 2.108182454038492e-07,
"loss": 0.7807,
"step": 1634
},
{
"epoch": 1.8888086642599278,
"grad_norm": 0.024348227307200432,
"learning_rate": 2.065203307580309e-07,
"loss": 0.7761,
"step": 1635
},
{
"epoch": 1.8899638989169674,
"grad_norm": 0.023634381592273712,
"learning_rate": 2.0226489719590362e-07,
"loss": 0.7619,
"step": 1636
},
{
"epoch": 1.8911191335740072,
"grad_norm": 0.023854384198784828,
"learning_rate": 1.9805202353034296e-07,
"loss": 0.7587,
"step": 1637
},
{
"epoch": 1.892274368231047,
"grad_norm": 0.023444773629307747,
"learning_rate": 1.9388178778599145e-07,
"loss": 0.7739,
"step": 1638
},
{
"epoch": 1.8934296028880866,
"grad_norm": 0.023775937035679817,
"learning_rate": 1.8975426719781569e-07,
"loss": 0.7671,
"step": 1639
},
{
"epoch": 1.8945848375451262,
"grad_norm": 0.023820441216230392,
"learning_rate": 1.85669538209674e-07,
"loss": 0.7873,
"step": 1640
},
{
"epoch": 1.895740072202166,
"grad_norm": 0.023935571312904358,
"learning_rate": 1.816276764729035e-07,
"loss": 0.7698,
"step": 1641
},
{
"epoch": 1.8968953068592058,
"grad_norm": 0.02413167431950569,
"learning_rate": 1.7762875684491536e-07,
"loss": 0.7851,
"step": 1642
},
{
"epoch": 1.8980505415162456,
"grad_norm": 0.02435237169265747,
"learning_rate": 1.7367285338781114e-07,
"loss": 0.7699,
"step": 1643
},
{
"epoch": 1.8992057761732852,
"grad_norm": 0.024776320904493332,
"learning_rate": 1.6976003936701042e-07,
"loss": 0.7768,
"step": 1644
},
{
"epoch": 1.9003610108303248,
"grad_norm": 0.02364964410662651,
"learning_rate": 1.6589038724989278e-07,
"loss": 0.7931,
"step": 1645
},
{
"epoch": 1.9015162454873646,
"grad_norm": 0.023821519687771797,
"learning_rate": 1.6206396870445796e-07,
"loss": 0.7533,
"step": 1646
},
{
"epoch": 1.9026714801444045,
"grad_norm": 0.02380574308335781,
"learning_rate": 1.582808545979954e-07,
"loss": 0.7875,
"step": 1647
},
{
"epoch": 1.903826714801444,
"grad_norm": 0.024249248206615448,
"learning_rate": 1.5454111499577497e-07,
"loss": 0.7561,
"step": 1648
},
{
"epoch": 1.9049819494584836,
"grad_norm": 0.02454947866499424,
"learning_rate": 1.5084481915974666e-07,
"loss": 0.7695,
"step": 1649
},
{
"epoch": 1.9061371841155235,
"grad_norm": 0.023830397054553032,
"learning_rate": 1.471920355472605e-07,
"loss": 0.7765,
"step": 1650
},
{
"epoch": 1.9072924187725633,
"grad_norm": 0.023539626970887184,
"learning_rate": 1.4358283180979607e-07,
"loss": 0.8193,
"step": 1651
},
{
"epoch": 1.9084476534296029,
"grad_norm": 0.02411411888897419,
"learning_rate": 1.4001727479171022e-07,
"loss": 0.8001,
"step": 1652
},
{
"epoch": 1.9096028880866425,
"grad_norm": 0.024499407038092613,
"learning_rate": 1.3649543052900035e-07,
"loss": 0.7714,
"step": 1653
},
{
"epoch": 1.9107581227436823,
"grad_norm": 0.023902500048279762,
"learning_rate": 1.3301736424808105e-07,
"loss": 0.7906,
"step": 1654
},
{
"epoch": 1.911913357400722,
"grad_norm": 0.024184612557291985,
"learning_rate": 1.295831403645745e-07,
"loss": 0.7643,
"step": 1655
},
{
"epoch": 1.9130685920577617,
"grad_norm": 0.024717319756746292,
"learning_rate": 1.2619282248211896e-07,
"loss": 0.8194,
"step": 1656
},
{
"epoch": 1.9142238267148013,
"grad_norm": 0.023884067311882973,
"learning_rate": 1.2284647339118997e-07,
"loss": 0.753,
"step": 1657
},
{
"epoch": 1.915379061371841,
"grad_norm": 0.023947538807988167,
"learning_rate": 1.1954415506793914e-07,
"loss": 0.7659,
"step": 1658
},
{
"epoch": 1.916534296028881,
"grad_norm": 0.02345140650868416,
"learning_rate": 1.1628592867304393e-07,
"loss": 0.7673,
"step": 1659
},
{
"epoch": 1.9176895306859207,
"grad_norm": 0.02360740303993225,
"learning_rate": 1.1307185455057616e-07,
"loss": 0.7703,
"step": 1660
},
{
"epoch": 1.9188447653429603,
"grad_norm": 0.023865753784775734,
"learning_rate": 1.0990199222688366e-07,
"loss": 0.7839,
"step": 1661
},
{
"epoch": 1.92,
"grad_norm": 0.02310926280915737,
"learning_rate": 1.0677640040949089e-07,
"loss": 0.7723,
"step": 1662
},
{
"epoch": 1.9211552346570397,
"grad_norm": 0.023065300658345222,
"learning_rate": 1.0369513698600607e-07,
"loss": 0.7699,
"step": 1663
},
{
"epoch": 1.9223104693140796,
"grad_norm": 0.023747362196445465,
"learning_rate": 1.0065825902305331e-07,
"loss": 0.7736,
"step": 1664
},
{
"epoch": 1.9234657039711192,
"grad_norm": 0.02400401420891285,
"learning_rate": 9.766582276521622e-08,
"loss": 0.7661,
"step": 1665
},
{
"epoch": 1.9246209386281588,
"grad_norm": 0.02401837147772312,
"learning_rate": 9.471788363399199e-08,
"loss": 0.7723,
"step": 1666
},
{
"epoch": 1.9257761732851986,
"grad_norm": 0.024011146277189255,
"learning_rate": 9.181449622676956e-08,
"loss": 0.7702,
"step": 1667
},
{
"epoch": 1.9269314079422384,
"grad_norm": 0.023694701492786407,
"learning_rate": 8.8955714315814e-08,
"loss": 0.7538,
"step": 1668
},
{
"epoch": 1.928086642599278,
"grad_norm": 0.02426433376967907,
"learning_rate": 8.614159084727651e-08,
"loss": 0.7792,
"step": 1669
},
{
"epoch": 1.9292418772563176,
"grad_norm": 0.024202091619372368,
"learning_rate": 8.337217794020758e-08,
"loss": 0.7423,
"step": 1670
},
{
"epoch": 1.9303971119133574,
"grad_norm": 0.024061063304543495,
"learning_rate": 8.064752688559553e-08,
"loss": 0.7549,
"step": 1671
},
{
"epoch": 1.9315523465703972,
"grad_norm": 0.024105606600642204,
"learning_rate": 7.796768814541562e-08,
"loss": 0.7845,
"step": 1672
},
{
"epoch": 1.9327075812274368,
"grad_norm": 0.024117371067404747,
"learning_rate": 7.533271135169531e-08,
"loss": 0.7706,
"step": 1673
},
{
"epoch": 1.9338628158844764,
"grad_norm": 0.023541245609521866,
"learning_rate": 7.274264530559488e-08,
"loss": 0.7794,
"step": 1674
},
{
"epoch": 1.9350180505415162,
"grad_norm": 0.023918094113469124,
"learning_rate": 7.019753797650377e-08,
"loss": 0.7515,
"step": 1675
},
{
"epoch": 1.936173285198556,
"grad_norm": 0.02431383728981018,
"learning_rate": 6.769743650115355e-08,
"loss": 0.7852,
"step": 1676
},
{
"epoch": 1.9373285198555956,
"grad_norm": 0.022908180952072144,
"learning_rate": 6.524238718274186e-08,
"loss": 0.7451,
"step": 1677
},
{
"epoch": 1.9384837545126352,
"grad_norm": 0.023673677816987038,
"learning_rate": 6.283243549007756e-08,
"loss": 0.7679,
"step": 1678
},
{
"epoch": 1.939638989169675,
"grad_norm": 0.02336922660470009,
"learning_rate": 6.046762605673734e-08,
"loss": 0.7619,
"step": 1679
},
{
"epoch": 1.9407942238267148,
"grad_norm": 0.023723188787698746,
"learning_rate": 5.814800268024005e-08,
"loss": 0.7792,
"step": 1680
},
{
"epoch": 1.9419494584837547,
"grad_norm": 0.02384623885154724,
"learning_rate": 5.58736083212344e-08,
"loss": 0.7648,
"step": 1681
},
{
"epoch": 1.9431046931407943,
"grad_norm": 0.024481408298015594,
"learning_rate": 5.364448510270387e-08,
"loss": 0.7907,
"step": 1682
},
{
"epoch": 1.9442599277978339,
"grad_norm": 0.023579951375722885,
"learning_rate": 5.146067430918623e-08,
"loss": 0.7737,
"step": 1683
},
{
"epoch": 1.9454151624548737,
"grad_norm": 0.02421787939965725,
"learning_rate": 4.932221638601039e-08,
"loss": 0.7746,
"step": 1684
},
{
"epoch": 1.9465703971119135,
"grad_norm": 0.02382575534284115,
"learning_rate": 4.722915093854457e-08,
"loss": 0.7547,
"step": 1685
},
{
"epoch": 1.947725631768953,
"grad_norm": 0.024289904162287712,
"learning_rate": 4.518151673146557e-08,
"loss": 0.7786,
"step": 1686
},
{
"epoch": 1.9488808664259927,
"grad_norm": 0.0236971415579319,
"learning_rate": 4.3179351688039386e-08,
"loss": 0.7785,
"step": 1687
},
{
"epoch": 1.9500361010830325,
"grad_norm": 0.023959195241332054,
"learning_rate": 4.122269288941911e-08,
"loss": 0.7665,
"step": 1688
},
{
"epoch": 1.9511913357400723,
"grad_norm": 0.02382025495171547,
"learning_rate": 3.931157657395747e-08,
"loss": 0.7539,
"step": 1689
},
{
"epoch": 1.952346570397112,
"grad_norm": 0.024375753477215767,
"learning_rate": 3.744603813653741e-08,
"loss": 0.7962,
"step": 1690
},
{
"epoch": 1.9535018050541515,
"grad_norm": 0.023500755429267883,
"learning_rate": 3.562611212791524e-08,
"loss": 0.7373,
"step": 1691
},
{
"epoch": 1.9546570397111913,
"grad_norm": 0.024235745891928673,
"learning_rate": 3.3851832254080314e-08,
"loss": 0.7971,
"step": 1692
},
{
"epoch": 1.9558122743682311,
"grad_norm": 0.02372434362769127,
"learning_rate": 3.212323137563259e-08,
"loss": 0.746,
"step": 1693
},
{
"epoch": 1.9569675090252707,
"grad_norm": 0.02388044260442257,
"learning_rate": 3.044034150717202e-08,
"loss": 0.7903,
"step": 1694
},
{
"epoch": 1.9581227436823103,
"grad_norm": 0.02452683448791504,
"learning_rate": 2.880319381670718e-08,
"loss": 0.7752,
"step": 1695
},
{
"epoch": 1.9592779783393501,
"grad_norm": 0.023390717804431915,
"learning_rate": 2.721181862507687e-08,
"loss": 0.7453,
"step": 1696
},
{
"epoch": 1.96043321299639,
"grad_norm": 0.023405345156788826,
"learning_rate": 2.566624540538925e-08,
"loss": 0.7601,
"step": 1697
},
{
"epoch": 1.9615884476534298,
"grad_norm": 0.023698341101408005,
"learning_rate": 2.4166502782475928e-08,
"loss": 0.7577,
"step": 1698
},
{
"epoch": 1.9627436823104694,
"grad_norm": 0.024021916091442108,
"learning_rate": 2.2712618532361902e-08,
"loss": 0.7723,
"step": 1699
},
{
"epoch": 1.963898916967509,
"grad_norm": 0.024296529591083527,
"learning_rate": 2.130461958174988e-08,
"loss": 0.7896,
"step": 1700
},
{
"epoch": 1.9650541516245488,
"grad_norm": 0.02386784926056862,
"learning_rate": 1.9942532007523544e-08,
"loss": 0.7587,
"step": 1701
},
{
"epoch": 1.9662093862815886,
"grad_norm": 0.023428186774253845,
"learning_rate": 1.8626381036264058e-08,
"loss": 0.7714,
"step": 1702
},
{
"epoch": 1.9673646209386282,
"grad_norm": 0.024159930646419525,
"learning_rate": 1.735619104378108e-08,
"loss": 0.7937,
"step": 1703
},
{
"epoch": 1.9685198555956678,
"grad_norm": 0.024223441258072853,
"learning_rate": 1.6131985554663975e-08,
"loss": 0.761,
"step": 1704
},
{
"epoch": 1.9696750902527076,
"grad_norm": 0.02402389794588089,
"learning_rate": 1.4953787241843607e-08,
"loss": 0.7742,
"step": 1705
},
{
"epoch": 1.9708303249097474,
"grad_norm": 0.023779474198818207,
"learning_rate": 1.382161792617522e-08,
"loss": 0.7489,
"step": 1706
},
{
"epoch": 1.971985559566787,
"grad_norm": 0.02408597618341446,
"learning_rate": 1.273549857603129e-08,
"loss": 0.7927,
"step": 1707
},
{
"epoch": 1.9731407942238266,
"grad_norm": 0.023695794865489006,
"learning_rate": 1.16954493069155e-08,
"loss": 0.7603,
"step": 1708
},
{
"epoch": 1.9742960288808664,
"grad_norm": 0.025053711608052254,
"learning_rate": 1.0701489381089175e-08,
"loss": 0.7886,
"step": 1709
},
{
"epoch": 1.9754512635379062,
"grad_norm": 0.023919204249978065,
"learning_rate": 9.753637207214111e-09,
"loss": 0.7837,
"step": 1710
},
{
"epoch": 1.9766064981949458,
"grad_norm": 0.023150555789470673,
"learning_rate": 8.85191034001343e-09,
"loss": 0.7449,
"step": 1711
},
{
"epoch": 1.9777617328519854,
"grad_norm": 0.023474812507629395,
"learning_rate": 7.996325479943542e-09,
"loss": 0.7825,
"step": 1712
},
{
"epoch": 1.9789169675090252,
"grad_norm": 0.02393323741853237,
"learning_rate": 7.186898472888319e-09,
"loss": 0.7702,
"step": 1713
},
{
"epoch": 1.980072202166065,
"grad_norm": 0.02350715547800064,
"learning_rate": 6.423644309862407e-09,
"loss": 0.7662,
"step": 1714
},
{
"epoch": 1.9812274368231046,
"grad_norm": 0.023993050679564476,
"learning_rate": 5.706577126735919e-09,
"loss": 0.794,
"step": 1715
},
{
"epoch": 1.9823826714801442,
"grad_norm": 0.024228619411587715,
"learning_rate": 5.035710203971343e-09,
"loss": 0.7874,
"step": 1716
},
{
"epoch": 1.983537906137184,
"grad_norm": 0.024353953078389168,
"learning_rate": 4.411055966378209e-09,
"loss": 0.7666,
"step": 1717
},
{
"epoch": 1.9846931407942239,
"grad_norm": 0.023848267272114754,
"learning_rate": 3.832625982882452e-09,
"loss": 0.7824,
"step": 1718
},
{
"epoch": 1.9858483754512637,
"grad_norm": 0.02303747460246086,
"learning_rate": 3.300430966312174e-09,
"loss": 0.7451,
"step": 1719
},
{
"epoch": 1.9870036101083033,
"grad_norm": 0.02363884262740612,
"learning_rate": 2.8144807732005904e-09,
"loss": 0.7673,
"step": 1720
},
{
"epoch": 1.9881588447653429,
"grad_norm": 0.024231605231761932,
"learning_rate": 2.374784403601199e-09,
"loss": 0.7823,
"step": 1721
},
{
"epoch": 1.9893140794223827,
"grad_norm": 0.02352130599319935,
"learning_rate": 1.9813500009229292e-09,
"loss": 0.7676,
"step": 1722
},
{
"epoch": 1.9904693140794225,
"grad_norm": 0.02459130994975567,
"learning_rate": 1.634184851778611e-09,
"loss": 0.7937,
"step": 1723
},
{
"epoch": 1.991624548736462,
"grad_norm": 0.023892242461442947,
"learning_rate": 1.3332953858495395e-09,
"loss": 0.7666,
"step": 1724
},
{
"epoch": 1.9927797833935017,
"grad_norm": 0.0234362855553627,
"learning_rate": 1.0786871757678059e-09,
"loss": 0.7796,
"step": 1725
},
{
"epoch": 1.9939350180505415,
"grad_norm": 0.024267012253403664,
"learning_rate": 8.703649370116684e-10,
"loss": 0.738,
"step": 1726
},
{
"epoch": 1.9950902527075813,
"grad_norm": 0.023687295615673065,
"learning_rate": 7.083325278189627e-10,
"loss": 0.7633,
"step": 1727
},
{
"epoch": 1.996245487364621,
"grad_norm": 0.02364824153482914,
"learning_rate": 5.925929491152231e-10,
"loss": 0.769,
"step": 1728
},
{
"epoch": 1.9974007220216605,
"grad_norm": 0.024108268320560455,
"learning_rate": 5.231483444587321e-10,
"loss": 0.7676,
"step": 1729
},
{
"epoch": 1.9985559566787003,
"grad_norm": 0.025233233347535133,
"learning_rate": 5e-10,
"loss": 0.7958,
"step": 1730
},
{
"epoch": 1.9985559566787003,
"step": 1730,
"total_flos": 2.356928687040889e+19,
"train_loss": 0.8199663722446199,
"train_runtime": 64534.4414,
"train_samples_per_second": 10.301,
"train_steps_per_second": 0.027
}
],
"logging_steps": 1.0,
"max_steps": 1730,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 250,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.356928687040889e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}