random_9glD99SwISCcOl0I / trainer_state.json
cutelemonlili's picture
Add files using upload-large-folder tool
da7a5e1 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 1022,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0019569471624266144,
"grad_norm": 6.49816938838842,
"learning_rate": 9.99997637686018e-06,
"loss": 0.3831,
"step": 1
},
{
"epoch": 0.003913894324853229,
"grad_norm": 5.380693356203711,
"learning_rate": 9.999905507663936e-06,
"loss": 0.3581,
"step": 2
},
{
"epoch": 0.005870841487279843,
"grad_norm": 4.854102607288882,
"learning_rate": 9.999787393080931e-06,
"loss": 0.3323,
"step": 3
},
{
"epoch": 0.007827788649706457,
"grad_norm": 3.3339489731022236,
"learning_rate": 9.99962203422726e-06,
"loss": 0.2206,
"step": 4
},
{
"epoch": 0.009784735812133072,
"grad_norm": 3.089407392178408,
"learning_rate": 9.999409432665442e-06,
"loss": 0.2369,
"step": 5
},
{
"epoch": 0.011741682974559686,
"grad_norm": 2.6560729801276,
"learning_rate": 9.9991495904044e-06,
"loss": 0.2,
"step": 6
},
{
"epoch": 0.0136986301369863,
"grad_norm": 2.5800447510898294,
"learning_rate": 9.998842509899456e-06,
"loss": 0.1678,
"step": 7
},
{
"epoch": 0.015655577299412915,
"grad_norm": 2.823342371763127,
"learning_rate": 9.998488194052287e-06,
"loss": 0.2169,
"step": 8
},
{
"epoch": 0.01761252446183953,
"grad_norm": 3.4565865518835786,
"learning_rate": 9.998086646210916e-06,
"loss": 0.2734,
"step": 9
},
{
"epoch": 0.019569471624266144,
"grad_norm": 3.155959188591151,
"learning_rate": 9.997637870169673e-06,
"loss": 0.2101,
"step": 10
},
{
"epoch": 0.021526418786692758,
"grad_norm": 2.5892748160647394,
"learning_rate": 9.997141870169154e-06,
"loss": 0.2341,
"step": 11
},
{
"epoch": 0.023483365949119372,
"grad_norm": 2.2770534492930223,
"learning_rate": 9.996598650896191e-06,
"loss": 0.2109,
"step": 12
},
{
"epoch": 0.025440313111545987,
"grad_norm": 2.6766408472493586,
"learning_rate": 9.996008217483806e-06,
"loss": 0.252,
"step": 13
},
{
"epoch": 0.0273972602739726,
"grad_norm": 2.42877025676844,
"learning_rate": 9.995370575511151e-06,
"loss": 0.2012,
"step": 14
},
{
"epoch": 0.029354207436399216,
"grad_norm": 2.3106434819204464,
"learning_rate": 9.994685731003469e-06,
"loss": 0.1809,
"step": 15
},
{
"epoch": 0.03131115459882583,
"grad_norm": 2.7138963384238206,
"learning_rate": 9.993953690432032e-06,
"loss": 0.2227,
"step": 16
},
{
"epoch": 0.033268101761252444,
"grad_norm": 1.9560459455673258,
"learning_rate": 9.99317446071408e-06,
"loss": 0.1861,
"step": 17
},
{
"epoch": 0.03522504892367906,
"grad_norm": 2.2259934753526567,
"learning_rate": 9.99234804921275e-06,
"loss": 0.2005,
"step": 18
},
{
"epoch": 0.03718199608610567,
"grad_norm": 2.771920003009874,
"learning_rate": 9.991474463737018e-06,
"loss": 0.2211,
"step": 19
},
{
"epoch": 0.03913894324853229,
"grad_norm": 3.4898151511234126,
"learning_rate": 9.990553712541617e-06,
"loss": 0.2587,
"step": 20
},
{
"epoch": 0.0410958904109589,
"grad_norm": 2.82405436363526,
"learning_rate": 9.989585804326963e-06,
"loss": 0.2295,
"step": 21
},
{
"epoch": 0.043052837573385516,
"grad_norm": 2.932912521576679,
"learning_rate": 9.988570748239062e-06,
"loss": 0.2594,
"step": 22
},
{
"epoch": 0.04500978473581213,
"grad_norm": 2.495304772832742,
"learning_rate": 9.987508553869444e-06,
"loss": 0.1973,
"step": 23
},
{
"epoch": 0.046966731898238745,
"grad_norm": 2.601956252074884,
"learning_rate": 9.986399231255057e-06,
"loss": 0.1905,
"step": 24
},
{
"epoch": 0.04892367906066536,
"grad_norm": 2.529030852639673,
"learning_rate": 9.985242790878168e-06,
"loss": 0.1691,
"step": 25
},
{
"epoch": 0.050880626223091974,
"grad_norm": 2.4882841479475317,
"learning_rate": 9.984039243666284e-06,
"loss": 0.1979,
"step": 26
},
{
"epoch": 0.05283757338551859,
"grad_norm": 2.9257563764546406,
"learning_rate": 9.982788600992027e-06,
"loss": 0.2072,
"step": 27
},
{
"epoch": 0.0547945205479452,
"grad_norm": 2.8838722311323255,
"learning_rate": 9.98149087467304e-06,
"loss": 0.2328,
"step": 28
},
{
"epoch": 0.05675146771037182,
"grad_norm": 2.181327262811659,
"learning_rate": 9.980146076971873e-06,
"loss": 0.2007,
"step": 29
},
{
"epoch": 0.05870841487279843,
"grad_norm": 2.8572486821313197,
"learning_rate": 9.978754220595861e-06,
"loss": 0.2135,
"step": 30
},
{
"epoch": 0.060665362035225046,
"grad_norm": 2.5374509542410886,
"learning_rate": 9.977315318697013e-06,
"loss": 0.2073,
"step": 31
},
{
"epoch": 0.06262230919765166,
"grad_norm": 2.3130227856443275,
"learning_rate": 9.975829384871884e-06,
"loss": 0.1879,
"step": 32
},
{
"epoch": 0.06457925636007827,
"grad_norm": 2.4793830732334765,
"learning_rate": 9.974296433161437e-06,
"loss": 0.2215,
"step": 33
},
{
"epoch": 0.06653620352250489,
"grad_norm": 2.6918960743148457,
"learning_rate": 9.97271647805093e-06,
"loss": 0.2037,
"step": 34
},
{
"epoch": 0.0684931506849315,
"grad_norm": 2.656918035841296,
"learning_rate": 9.97108953446976e-06,
"loss": 0.2072,
"step": 35
},
{
"epoch": 0.07045009784735812,
"grad_norm": 2.2989268048651597,
"learning_rate": 9.969415617791336e-06,
"loss": 0.2243,
"step": 36
},
{
"epoch": 0.07240704500978473,
"grad_norm": 2.2220346400231508,
"learning_rate": 9.967694743832923e-06,
"loss": 0.1945,
"step": 37
},
{
"epoch": 0.07436399217221135,
"grad_norm": 2.0711067608688665,
"learning_rate": 9.965926928855498e-06,
"loss": 0.1712,
"step": 38
},
{
"epoch": 0.07632093933463796,
"grad_norm": 2.5322489103260537,
"learning_rate": 9.964112189563601e-06,
"loss": 0.2341,
"step": 39
},
{
"epoch": 0.07827788649706457,
"grad_norm": 2.1485732934115074,
"learning_rate": 9.962250543105167e-06,
"loss": 0.1974,
"step": 40
},
{
"epoch": 0.08023483365949119,
"grad_norm": 2.0625357973409857,
"learning_rate": 9.960342007071368e-06,
"loss": 0.1452,
"step": 41
},
{
"epoch": 0.0821917808219178,
"grad_norm": 2.7180375898424654,
"learning_rate": 9.95838659949645e-06,
"loss": 0.2038,
"step": 42
},
{
"epoch": 0.08414872798434442,
"grad_norm": 2.0868047221494765,
"learning_rate": 9.956384338857561e-06,
"loss": 0.1651,
"step": 43
},
{
"epoch": 0.08610567514677103,
"grad_norm": 2.8213948689249593,
"learning_rate": 9.954335244074575e-06,
"loss": 0.2619,
"step": 44
},
{
"epoch": 0.08806262230919765,
"grad_norm": 1.871771628868505,
"learning_rate": 9.952239334509909e-06,
"loss": 0.1617,
"step": 45
},
{
"epoch": 0.09001956947162426,
"grad_norm": 2.731855359483773,
"learning_rate": 9.950096629968353e-06,
"loss": 0.2518,
"step": 46
},
{
"epoch": 0.09197651663405088,
"grad_norm": 2.6371657115421763,
"learning_rate": 9.947907150696868e-06,
"loss": 0.1958,
"step": 47
},
{
"epoch": 0.09393346379647749,
"grad_norm": 2.6444310296811615,
"learning_rate": 9.945670917384404e-06,
"loss": 0.1583,
"step": 48
},
{
"epoch": 0.0958904109589041,
"grad_norm": 2.4826045149475946,
"learning_rate": 9.943387951161702e-06,
"loss": 0.2238,
"step": 49
},
{
"epoch": 0.09784735812133072,
"grad_norm": 2.5082949960285084,
"learning_rate": 9.941058273601097e-06,
"loss": 0.1619,
"step": 50
},
{
"epoch": 0.09980430528375733,
"grad_norm": 2.3729389287378164,
"learning_rate": 9.938681906716305e-06,
"loss": 0.1744,
"step": 51
},
{
"epoch": 0.10176125244618395,
"grad_norm": 2.648730419799791,
"learning_rate": 9.936258872962229e-06,
"loss": 0.2122,
"step": 52
},
{
"epoch": 0.10371819960861056,
"grad_norm": 2.486658025546424,
"learning_rate": 9.93378919523473e-06,
"loss": 0.1649,
"step": 53
},
{
"epoch": 0.10567514677103718,
"grad_norm": 2.2724210250046513,
"learning_rate": 9.931272896870427e-06,
"loss": 0.2044,
"step": 54
},
{
"epoch": 0.10763209393346379,
"grad_norm": 1.9775758430433867,
"learning_rate": 9.928710001646467e-06,
"loss": 0.1763,
"step": 55
},
{
"epoch": 0.1095890410958904,
"grad_norm": 2.4370142170794913,
"learning_rate": 9.926100533780304e-06,
"loss": 0.2233,
"step": 56
},
{
"epoch": 0.11154598825831702,
"grad_norm": 2.5141917988151343,
"learning_rate": 9.923444517929467e-06,
"loss": 0.2492,
"step": 57
},
{
"epoch": 0.11350293542074363,
"grad_norm": 2.752368590034408,
"learning_rate": 9.92074197919133e-06,
"loss": 0.1967,
"step": 58
},
{
"epoch": 0.11545988258317025,
"grad_norm": 2.765803987257559,
"learning_rate": 9.917992943102872e-06,
"loss": 0.2097,
"step": 59
},
{
"epoch": 0.11741682974559686,
"grad_norm": 2.5147132924542674,
"learning_rate": 9.91519743564044e-06,
"loss": 0.2212,
"step": 60
},
{
"epoch": 0.11937377690802348,
"grad_norm": 2.0155053032184087,
"learning_rate": 9.912355483219498e-06,
"loss": 0.1638,
"step": 61
},
{
"epoch": 0.12133072407045009,
"grad_norm": 2.021296244758309,
"learning_rate": 9.909467112694385e-06,
"loss": 0.1672,
"step": 62
},
{
"epoch": 0.1232876712328767,
"grad_norm": 2.2173585967079745,
"learning_rate": 9.906532351358047e-06,
"loss": 0.174,
"step": 63
},
{
"epoch": 0.12524461839530332,
"grad_norm": 2.899798726270873,
"learning_rate": 9.903551226941801e-06,
"loss": 0.253,
"step": 64
},
{
"epoch": 0.12720156555772993,
"grad_norm": 2.503996867252731,
"learning_rate": 9.900523767615052e-06,
"loss": 0.2198,
"step": 65
},
{
"epoch": 0.12915851272015655,
"grad_norm": 2.3952499898965094,
"learning_rate": 9.897450001985038e-06,
"loss": 0.2071,
"step": 66
},
{
"epoch": 0.13111545988258316,
"grad_norm": 2.399907837161599,
"learning_rate": 9.894329959096559e-06,
"loss": 0.254,
"step": 67
},
{
"epoch": 0.13307240704500978,
"grad_norm": 2.7441327698125404,
"learning_rate": 9.891163668431696e-06,
"loss": 0.186,
"step": 68
},
{
"epoch": 0.1350293542074364,
"grad_norm": 2.2747673192909934,
"learning_rate": 9.887951159909541e-06,
"loss": 0.2202,
"step": 69
},
{
"epoch": 0.136986301369863,
"grad_norm": 2.702240638553894,
"learning_rate": 9.88469246388591e-06,
"loss": 0.1936,
"step": 70
},
{
"epoch": 0.13894324853228962,
"grad_norm": 2.39315583619449,
"learning_rate": 9.881387611153053e-06,
"loss": 0.1679,
"step": 71
},
{
"epoch": 0.14090019569471623,
"grad_norm": 2.7394787651459644,
"learning_rate": 9.878036632939374e-06,
"loss": 0.2118,
"step": 72
},
{
"epoch": 0.14285714285714285,
"grad_norm": 2.7400844254551364,
"learning_rate": 9.874639560909118e-06,
"loss": 0.2248,
"step": 73
},
{
"epoch": 0.14481409001956946,
"grad_norm": 3.387587218567297,
"learning_rate": 9.871196427162094e-06,
"loss": 0.2027,
"step": 74
},
{
"epoch": 0.14677103718199608,
"grad_norm": 2.8157583006918454,
"learning_rate": 9.867707264233349e-06,
"loss": 0.2308,
"step": 75
},
{
"epoch": 0.1487279843444227,
"grad_norm": 2.1042150721161326,
"learning_rate": 9.86417210509288e-06,
"loss": 0.1666,
"step": 76
},
{
"epoch": 0.1506849315068493,
"grad_norm": 2.133705709859829,
"learning_rate": 9.860590983145307e-06,
"loss": 0.2,
"step": 77
},
{
"epoch": 0.15264187866927592,
"grad_norm": 2.1633564482046803,
"learning_rate": 9.85696393222957e-06,
"loss": 0.1774,
"step": 78
},
{
"epoch": 0.15459882583170254,
"grad_norm": 2.4915588839950273,
"learning_rate": 9.853290986618604e-06,
"loss": 0.2003,
"step": 79
},
{
"epoch": 0.15655577299412915,
"grad_norm": 2.210488926028733,
"learning_rate": 9.849572181019008e-06,
"loss": 0.1895,
"step": 80
},
{
"epoch": 0.15851272015655576,
"grad_norm": 2.902575002928562,
"learning_rate": 9.845807550570727e-06,
"loss": 0.2201,
"step": 81
},
{
"epoch": 0.16046966731898238,
"grad_norm": 2.460701914976216,
"learning_rate": 9.84199713084672e-06,
"loss": 0.169,
"step": 82
},
{
"epoch": 0.162426614481409,
"grad_norm": 2.291537819889769,
"learning_rate": 9.83814095785262e-06,
"loss": 0.2016,
"step": 83
},
{
"epoch": 0.1643835616438356,
"grad_norm": 2.4535458882383785,
"learning_rate": 9.834239068026388e-06,
"loss": 0.1808,
"step": 84
},
{
"epoch": 0.16634050880626222,
"grad_norm": 2.4284249353537364,
"learning_rate": 9.830291498237983e-06,
"loss": 0.2074,
"step": 85
},
{
"epoch": 0.16829745596868884,
"grad_norm": 2.181210964875374,
"learning_rate": 9.826298285789002e-06,
"loss": 0.2071,
"step": 86
},
{
"epoch": 0.17025440313111545,
"grad_norm": 2.193375948249739,
"learning_rate": 9.822259468412329e-06,
"loss": 0.1642,
"step": 87
},
{
"epoch": 0.17221135029354206,
"grad_norm": 2.7426121907059016,
"learning_rate": 9.818175084271786e-06,
"loss": 0.1825,
"step": 88
},
{
"epoch": 0.17416829745596868,
"grad_norm": 2.2015774825289895,
"learning_rate": 9.814045171961762e-06,
"loss": 0.2004,
"step": 89
},
{
"epoch": 0.1761252446183953,
"grad_norm": 2.0524771992508195,
"learning_rate": 9.809869770506855e-06,
"loss": 0.163,
"step": 90
},
{
"epoch": 0.1780821917808219,
"grad_norm": 2.3527235519778147,
"learning_rate": 9.805648919361505e-06,
"loss": 0.1897,
"step": 91
},
{
"epoch": 0.18003913894324852,
"grad_norm": 2.1754448826743444,
"learning_rate": 9.801382658409611e-06,
"loss": 0.209,
"step": 92
},
{
"epoch": 0.18199608610567514,
"grad_norm": 2.0361688732737337,
"learning_rate": 9.797071027964166e-06,
"loss": 0.1833,
"step": 93
},
{
"epoch": 0.18395303326810175,
"grad_norm": 2.5653082795621986,
"learning_rate": 9.792714068766872e-06,
"loss": 0.2326,
"step": 94
},
{
"epoch": 0.18590998043052837,
"grad_norm": 1.826467304079506,
"learning_rate": 9.788311821987748e-06,
"loss": 0.1548,
"step": 95
},
{
"epoch": 0.18786692759295498,
"grad_norm": 1.6993615861848013,
"learning_rate": 9.783864329224752e-06,
"loss": 0.1342,
"step": 96
},
{
"epoch": 0.1898238747553816,
"grad_norm": 2.624746390682379,
"learning_rate": 9.779371632503382e-06,
"loss": 0.2394,
"step": 97
},
{
"epoch": 0.1917808219178082,
"grad_norm": 2.3275235518937722,
"learning_rate": 9.774833774276278e-06,
"loss": 0.2222,
"step": 98
},
{
"epoch": 0.19373776908023482,
"grad_norm": 2.86146034805017,
"learning_rate": 9.770250797422822e-06,
"loss": 0.2029,
"step": 99
},
{
"epoch": 0.19569471624266144,
"grad_norm": 2.741265179665307,
"learning_rate": 9.765622745248739e-06,
"loss": 0.1934,
"step": 100
},
{
"epoch": 0.19765166340508805,
"grad_norm": 2.5823472746443557,
"learning_rate": 9.760949661485676e-06,
"loss": 0.1962,
"step": 101
},
{
"epoch": 0.19960861056751467,
"grad_norm": 2.1759848293570743,
"learning_rate": 9.7562315902908e-06,
"loss": 0.1897,
"step": 102
},
{
"epoch": 0.20156555772994128,
"grad_norm": 2.3771044101959595,
"learning_rate": 9.75146857624637e-06,
"loss": 0.1987,
"step": 103
},
{
"epoch": 0.2035225048923679,
"grad_norm": 1.8921875081895094,
"learning_rate": 9.746660664359326e-06,
"loss": 0.181,
"step": 104
},
{
"epoch": 0.2054794520547945,
"grad_norm": 2.515657563734455,
"learning_rate": 9.741807900060858e-06,
"loss": 0.2228,
"step": 105
},
{
"epoch": 0.20743639921722112,
"grad_norm": 2.8310381056783567,
"learning_rate": 9.73691032920598e-06,
"loss": 0.2107,
"step": 106
},
{
"epoch": 0.20939334637964774,
"grad_norm": 2.5486385399926843,
"learning_rate": 9.731967998073087e-06,
"loss": 0.1505,
"step": 107
},
{
"epoch": 0.21135029354207435,
"grad_norm": 2.1347198320178227,
"learning_rate": 9.726980953363536e-06,
"loss": 0.1803,
"step": 108
},
{
"epoch": 0.21330724070450097,
"grad_norm": 2.641580369943664,
"learning_rate": 9.721949242201185e-06,
"loss": 0.2239,
"step": 109
},
{
"epoch": 0.21526418786692758,
"grad_norm": 2.5211140766736495,
"learning_rate": 9.716872912131964e-06,
"loss": 0.1749,
"step": 110
},
{
"epoch": 0.2172211350293542,
"grad_norm": 2.087904485693696,
"learning_rate": 9.711752011123412e-06,
"loss": 0.1743,
"step": 111
},
{
"epoch": 0.2191780821917808,
"grad_norm": 2.2983871399852367,
"learning_rate": 9.706586587564236e-06,
"loss": 0.1643,
"step": 112
},
{
"epoch": 0.22113502935420742,
"grad_norm": 2.780628581073493,
"learning_rate": 9.701376690263845e-06,
"loss": 0.2444,
"step": 113
},
{
"epoch": 0.22309197651663404,
"grad_norm": 2.326905056397722,
"learning_rate": 9.696122368451887e-06,
"loss": 0.1673,
"step": 114
},
{
"epoch": 0.22504892367906065,
"grad_norm": 2.519042300696672,
"learning_rate": 9.690823671777799e-06,
"loss": 0.1694,
"step": 115
},
{
"epoch": 0.22700587084148727,
"grad_norm": 1.7378931792828742,
"learning_rate": 9.685480650310319e-06,
"loss": 0.1391,
"step": 116
},
{
"epoch": 0.22896281800391388,
"grad_norm": 2.034879982322088,
"learning_rate": 9.680093354537027e-06,
"loss": 0.1715,
"step": 117
},
{
"epoch": 0.2309197651663405,
"grad_norm": 2.994384283826758,
"learning_rate": 9.67466183536386e-06,
"loss": 0.2517,
"step": 118
},
{
"epoch": 0.2328767123287671,
"grad_norm": 1.8992995415631702,
"learning_rate": 9.669186144114627e-06,
"loss": 0.1415,
"step": 119
},
{
"epoch": 0.23483365949119372,
"grad_norm": 2.269855574311111,
"learning_rate": 9.663666332530541e-06,
"loss": 0.2372,
"step": 120
},
{
"epoch": 0.23679060665362034,
"grad_norm": 2.3574308339486465,
"learning_rate": 9.658102452769712e-06,
"loss": 0.2276,
"step": 121
},
{
"epoch": 0.23874755381604695,
"grad_norm": 2.1149611303024494,
"learning_rate": 9.652494557406666e-06,
"loss": 0.1977,
"step": 122
},
{
"epoch": 0.24070450097847357,
"grad_norm": 2.121396498728366,
"learning_rate": 9.646842699431842e-06,
"loss": 0.2404,
"step": 123
},
{
"epoch": 0.24266144814090018,
"grad_norm": 2.275643415388438,
"learning_rate": 9.641146932251088e-06,
"loss": 0.1869,
"step": 124
},
{
"epoch": 0.2446183953033268,
"grad_norm": 2.047919519630945,
"learning_rate": 9.635407309685171e-06,
"loss": 0.1825,
"step": 125
},
{
"epoch": 0.2465753424657534,
"grad_norm": 2.681245894602953,
"learning_rate": 9.62962388596925e-06,
"loss": 0.2251,
"step": 126
},
{
"epoch": 0.24853228962818003,
"grad_norm": 2.419599821005517,
"learning_rate": 9.623796715752378e-06,
"loss": 0.1732,
"step": 127
},
{
"epoch": 0.25048923679060664,
"grad_norm": 2.1938648465449826,
"learning_rate": 9.617925854096975e-06,
"loss": 0.2004,
"step": 128
},
{
"epoch": 0.25244618395303325,
"grad_norm": 2.267142684845981,
"learning_rate": 9.61201135647832e-06,
"loss": 0.2056,
"step": 129
},
{
"epoch": 0.25440313111545987,
"grad_norm": 2.41015151814145,
"learning_rate": 9.606053278784009e-06,
"loss": 0.2091,
"step": 130
},
{
"epoch": 0.2563600782778865,
"grad_norm": 2.1753469191770667,
"learning_rate": 9.600051677313446e-06,
"loss": 0.2118,
"step": 131
},
{
"epoch": 0.2583170254403131,
"grad_norm": 2.2023732989411866,
"learning_rate": 9.5940066087773e-06,
"loss": 0.2141,
"step": 132
},
{
"epoch": 0.2602739726027397,
"grad_norm": 2.5292796304039022,
"learning_rate": 9.587918130296969e-06,
"loss": 0.2303,
"step": 133
},
{
"epoch": 0.2622309197651663,
"grad_norm": 2.867701225542371,
"learning_rate": 9.581786299404046e-06,
"loss": 0.1757,
"step": 134
},
{
"epoch": 0.26418786692759294,
"grad_norm": 2.0953787309643257,
"learning_rate": 9.57561117403977e-06,
"loss": 0.2003,
"step": 135
},
{
"epoch": 0.26614481409001955,
"grad_norm": 2.4141335556858063,
"learning_rate": 9.56939281255448e-06,
"loss": 0.2104,
"step": 136
},
{
"epoch": 0.26810176125244617,
"grad_norm": 2.620656202730439,
"learning_rate": 9.563131273707065e-06,
"loss": 0.233,
"step": 137
},
{
"epoch": 0.2700587084148728,
"grad_norm": 2.440317368365604,
"learning_rate": 9.556826616664408e-06,
"loss": 0.2622,
"step": 138
},
{
"epoch": 0.2720156555772994,
"grad_norm": 2.2609850867758037,
"learning_rate": 9.550478901000827e-06,
"loss": 0.1784,
"step": 139
},
{
"epoch": 0.273972602739726,
"grad_norm": 3.2204287542922185,
"learning_rate": 9.544088186697515e-06,
"loss": 0.2581,
"step": 140
},
{
"epoch": 0.2759295499021526,
"grad_norm": 2.1488795287046236,
"learning_rate": 9.537654534141964e-06,
"loss": 0.1823,
"step": 141
},
{
"epoch": 0.27788649706457924,
"grad_norm": 2.5173257456737974,
"learning_rate": 9.531178004127404e-06,
"loss": 0.2183,
"step": 142
},
{
"epoch": 0.27984344422700586,
"grad_norm": 2.150436741808308,
"learning_rate": 9.524658657852223e-06,
"loss": 0.1989,
"step": 143
},
{
"epoch": 0.28180039138943247,
"grad_norm": 2.5222707702692153,
"learning_rate": 9.518096556919396e-06,
"loss": 0.2068,
"step": 144
},
{
"epoch": 0.2837573385518591,
"grad_norm": 1.9379853268562073,
"learning_rate": 9.511491763335893e-06,
"loss": 0.1559,
"step": 145
},
{
"epoch": 0.2857142857142857,
"grad_norm": 2.0238963605764977,
"learning_rate": 9.504844339512096e-06,
"loss": 0.1886,
"step": 146
},
{
"epoch": 0.2876712328767123,
"grad_norm": 2.7078248088939834,
"learning_rate": 9.498154348261217e-06,
"loss": 0.2552,
"step": 147
},
{
"epoch": 0.2896281800391389,
"grad_norm": 1.833615803597142,
"learning_rate": 9.491421852798695e-06,
"loss": 0.164,
"step": 148
},
{
"epoch": 0.29158512720156554,
"grad_norm": 2.0789055569223307,
"learning_rate": 9.484646916741602e-06,
"loss": 0.1642,
"step": 149
},
{
"epoch": 0.29354207436399216,
"grad_norm": 2.4365222052159745,
"learning_rate": 9.477829604108044e-06,
"loss": 0.2071,
"step": 150
},
{
"epoch": 0.29549902152641877,
"grad_norm": 2.7626517916406317,
"learning_rate": 9.47096997931655e-06,
"loss": 0.2399,
"step": 151
},
{
"epoch": 0.2974559686888454,
"grad_norm": 1.756442646434427,
"learning_rate": 9.464068107185476e-06,
"loss": 0.1683,
"step": 152
},
{
"epoch": 0.299412915851272,
"grad_norm": 2.3951752279986156,
"learning_rate": 9.45712405293237e-06,
"loss": 0.2283,
"step": 153
},
{
"epoch": 0.3013698630136986,
"grad_norm": 2.7991273888999504,
"learning_rate": 9.450137882173385e-06,
"loss": 0.2873,
"step": 154
},
{
"epoch": 0.30332681017612523,
"grad_norm": 2.665588066905155,
"learning_rate": 9.443109660922632e-06,
"loss": 0.1771,
"step": 155
},
{
"epoch": 0.30528375733855184,
"grad_norm": 2.358906853004939,
"learning_rate": 9.436039455591574e-06,
"loss": 0.1669,
"step": 156
},
{
"epoch": 0.30724070450097846,
"grad_norm": 2.5899056219491925,
"learning_rate": 9.42892733298839e-06,
"loss": 0.2035,
"step": 157
},
{
"epoch": 0.30919765166340507,
"grad_norm": 2.4047599442576737,
"learning_rate": 9.421773360317348e-06,
"loss": 0.1816,
"step": 158
},
{
"epoch": 0.3111545988258317,
"grad_norm": 2.4544158682406594,
"learning_rate": 9.414577605178166e-06,
"loss": 0.2101,
"step": 159
},
{
"epoch": 0.3131115459882583,
"grad_norm": 2.609981440605299,
"learning_rate": 9.407340135565375e-06,
"loss": 0.2205,
"step": 160
},
{
"epoch": 0.3150684931506849,
"grad_norm": 2.092482917971009,
"learning_rate": 9.40006101986768e-06,
"loss": 0.1806,
"step": 161
},
{
"epoch": 0.31702544031311153,
"grad_norm": 2.2336246760665666,
"learning_rate": 9.392740326867304e-06,
"loss": 0.1999,
"step": 162
},
{
"epoch": 0.31898238747553814,
"grad_norm": 2.2633274486118298,
"learning_rate": 9.385378125739352e-06,
"loss": 0.2153,
"step": 163
},
{
"epoch": 0.32093933463796476,
"grad_norm": 2.4013991115498308,
"learning_rate": 9.377974486051149e-06,
"loss": 0.2377,
"step": 164
},
{
"epoch": 0.32289628180039137,
"grad_norm": 2.4717198255171433,
"learning_rate": 9.370529477761574e-06,
"loss": 0.1903,
"step": 165
},
{
"epoch": 0.324853228962818,
"grad_norm": 2.1896594597201218,
"learning_rate": 9.363043171220423e-06,
"loss": 0.193,
"step": 166
},
{
"epoch": 0.3268101761252446,
"grad_norm": 2.261922478295587,
"learning_rate": 9.355515637167719e-06,
"loss": 0.1875,
"step": 167
},
{
"epoch": 0.3287671232876712,
"grad_norm": 2.549880582238543,
"learning_rate": 9.347946946733055e-06,
"loss": 0.2321,
"step": 168
},
{
"epoch": 0.33072407045009783,
"grad_norm": 2.0563506953858695,
"learning_rate": 9.34033717143493e-06,
"loss": 0.1962,
"step": 169
},
{
"epoch": 0.33268101761252444,
"grad_norm": 2.1707802126832507,
"learning_rate": 9.332686383180055e-06,
"loss": 0.193,
"step": 170
},
{
"epoch": 0.33463796477495106,
"grad_norm": 2.9408127854748667,
"learning_rate": 9.324994654262687e-06,
"loss": 0.2347,
"step": 171
},
{
"epoch": 0.33659491193737767,
"grad_norm": 2.4906740775132334,
"learning_rate": 9.31726205736394e-06,
"loss": 0.217,
"step": 172
},
{
"epoch": 0.3385518590998043,
"grad_norm": 2.498445091973163,
"learning_rate": 9.309488665551104e-06,
"loss": 0.2161,
"step": 173
},
{
"epoch": 0.3405088062622309,
"grad_norm": 2.790520427002426,
"learning_rate": 9.301674552276942e-06,
"loss": 0.231,
"step": 174
},
{
"epoch": 0.3424657534246575,
"grad_norm": 1.9895496968942854,
"learning_rate": 9.293819791379016e-06,
"loss": 0.1649,
"step": 175
},
{
"epoch": 0.34442270058708413,
"grad_norm": 1.833766391446337,
"learning_rate": 9.28592445707897e-06,
"loss": 0.1585,
"step": 176
},
{
"epoch": 0.34637964774951074,
"grad_norm": 2.391738037807345,
"learning_rate": 9.277988623981837e-06,
"loss": 0.1699,
"step": 177
},
{
"epoch": 0.34833659491193736,
"grad_norm": 2.396251438784599,
"learning_rate": 9.270012367075337e-06,
"loss": 0.2266,
"step": 178
},
{
"epoch": 0.350293542074364,
"grad_norm": 2.5986052542988705,
"learning_rate": 9.26199576172916e-06,
"loss": 0.2442,
"step": 179
},
{
"epoch": 0.3522504892367906,
"grad_norm": 2.406433783735256,
"learning_rate": 9.253938883694266e-06,
"loss": 0.2242,
"step": 180
},
{
"epoch": 0.3542074363992172,
"grad_norm": 1.8409370969769863,
"learning_rate": 9.245841809102156e-06,
"loss": 0.2212,
"step": 181
},
{
"epoch": 0.3561643835616438,
"grad_norm": 2.229640332374793,
"learning_rate": 9.237704614464157e-06,
"loss": 0.1877,
"step": 182
},
{
"epoch": 0.35812133072407043,
"grad_norm": 1.748208942535522,
"learning_rate": 9.229527376670706e-06,
"loss": 0.1646,
"step": 183
},
{
"epoch": 0.36007827788649704,
"grad_norm": 2.1833227413069864,
"learning_rate": 9.221310172990616e-06,
"loss": 0.2057,
"step": 184
},
{
"epoch": 0.36203522504892366,
"grad_norm": 1.8945740870610521,
"learning_rate": 9.213053081070346e-06,
"loss": 0.1807,
"step": 185
},
{
"epoch": 0.3639921722113503,
"grad_norm": 2.216741563710101,
"learning_rate": 9.204756178933274e-06,
"loss": 0.1885,
"step": 186
},
{
"epoch": 0.3659491193737769,
"grad_norm": 2.256774055316552,
"learning_rate": 9.19641954497895e-06,
"loss": 0.1902,
"step": 187
},
{
"epoch": 0.3679060665362035,
"grad_norm": 2.0194869754940763,
"learning_rate": 9.18804325798236e-06,
"loss": 0.1781,
"step": 188
},
{
"epoch": 0.3698630136986301,
"grad_norm": 1.9980319322395583,
"learning_rate": 9.179627397093184e-06,
"loss": 0.1749,
"step": 189
},
{
"epoch": 0.37181996086105673,
"grad_norm": 2.3626866121126247,
"learning_rate": 9.171172041835048e-06,
"loss": 0.194,
"step": 190
},
{
"epoch": 0.37377690802348335,
"grad_norm": 2.2438869339005274,
"learning_rate": 9.162677272104767e-06,
"loss": 0.1958,
"step": 191
},
{
"epoch": 0.37573385518590996,
"grad_norm": 2.106976935821723,
"learning_rate": 9.154143168171594e-06,
"loss": 0.1622,
"step": 192
},
{
"epoch": 0.3776908023483366,
"grad_norm": 2.2068971862392877,
"learning_rate": 9.145569810676458e-06,
"loss": 0.167,
"step": 193
},
{
"epoch": 0.3796477495107632,
"grad_norm": 2.39645414205477,
"learning_rate": 9.136957280631212e-06,
"loss": 0.1793,
"step": 194
},
{
"epoch": 0.3816046966731898,
"grad_norm": 2.5229709065064077,
"learning_rate": 9.128305659417856e-06,
"loss": 0.1865,
"step": 195
},
{
"epoch": 0.3835616438356164,
"grad_norm": 2.182990698297152,
"learning_rate": 9.119615028787771e-06,
"loss": 0.1881,
"step": 196
},
{
"epoch": 0.38551859099804303,
"grad_norm": 2.274532942330284,
"learning_rate": 9.110885470860953e-06,
"loss": 0.1951,
"step": 197
},
{
"epoch": 0.38747553816046965,
"grad_norm": 2.445554265575562,
"learning_rate": 9.102117068125227e-06,
"loss": 0.1956,
"step": 198
},
{
"epoch": 0.38943248532289626,
"grad_norm": 2.4358676162272515,
"learning_rate": 9.093309903435475e-06,
"loss": 0.2076,
"step": 199
},
{
"epoch": 0.3913894324853229,
"grad_norm": 2.1878514933367916,
"learning_rate": 9.084464060012849e-06,
"loss": 0.1673,
"step": 200
},
{
"epoch": 0.3933463796477495,
"grad_norm": 1.9958089790089841,
"learning_rate": 9.075579621443991e-06,
"loss": 0.1473,
"step": 201
},
{
"epoch": 0.3953033268101761,
"grad_norm": 2.115877596984378,
"learning_rate": 9.066656671680231e-06,
"loss": 0.157,
"step": 202
},
{
"epoch": 0.3972602739726027,
"grad_norm": 1.8022014598062654,
"learning_rate": 9.057695295036806e-06,
"loss": 0.1598,
"step": 203
},
{
"epoch": 0.39921722113502933,
"grad_norm": 3.064431282962799,
"learning_rate": 9.048695576192058e-06,
"loss": 0.2295,
"step": 204
},
{
"epoch": 0.40117416829745595,
"grad_norm": 2.288011058306396,
"learning_rate": 9.039657600186633e-06,
"loss": 0.1733,
"step": 205
},
{
"epoch": 0.40313111545988256,
"grad_norm": 2.2885142765205835,
"learning_rate": 9.03058145242268e-06,
"loss": 0.2037,
"step": 206
},
{
"epoch": 0.4050880626223092,
"grad_norm": 2.018202070053573,
"learning_rate": 9.021467218663042e-06,
"loss": 0.1923,
"step": 207
},
{
"epoch": 0.4070450097847358,
"grad_norm": 2.697309772059589,
"learning_rate": 9.012314985030445e-06,
"loss": 0.1838,
"step": 208
},
{
"epoch": 0.4090019569471624,
"grad_norm": 2.1136959142478187,
"learning_rate": 9.003124838006689e-06,
"loss": 0.1919,
"step": 209
},
{
"epoch": 0.410958904109589,
"grad_norm": 4.860487229796089,
"learning_rate": 8.993896864431825e-06,
"loss": 0.1891,
"step": 210
},
{
"epoch": 0.41291585127201563,
"grad_norm": 2.144576023613455,
"learning_rate": 8.984631151503335e-06,
"loss": 0.1327,
"step": 211
},
{
"epoch": 0.41487279843444225,
"grad_norm": 1.906283506983683,
"learning_rate": 8.975327786775316e-06,
"loss": 0.1505,
"step": 212
},
{
"epoch": 0.41682974559686886,
"grad_norm": 4.628602425835523,
"learning_rate": 8.965986858157636e-06,
"loss": 0.1955,
"step": 213
},
{
"epoch": 0.4187866927592955,
"grad_norm": 2.8764301880578182,
"learning_rate": 8.956608453915126e-06,
"loss": 0.2246,
"step": 214
},
{
"epoch": 0.4207436399217221,
"grad_norm": 2.7919477716369254,
"learning_rate": 8.947192662666724e-06,
"loss": 0.2058,
"step": 215
},
{
"epoch": 0.4227005870841487,
"grad_norm": 6.407872718797146,
"learning_rate": 8.937739573384653e-06,
"loss": 0.188,
"step": 216
},
{
"epoch": 0.4246575342465753,
"grad_norm": 2.7809615856986905,
"learning_rate": 8.928249275393572e-06,
"loss": 0.2357,
"step": 217
},
{
"epoch": 0.42661448140900193,
"grad_norm": 2.695289142220112,
"learning_rate": 8.918721858369738e-06,
"loss": 0.156,
"step": 218
},
{
"epoch": 0.42857142857142855,
"grad_norm": 2.176281626881343,
"learning_rate": 8.90915741234015e-06,
"loss": 0.2453,
"step": 219
},
{
"epoch": 0.43052837573385516,
"grad_norm": 3.8445700444424515,
"learning_rate": 8.899556027681708e-06,
"loss": 0.2147,
"step": 220
},
{
"epoch": 0.4324853228962818,
"grad_norm": 3.2286157538199673,
"learning_rate": 8.88991779512035e-06,
"loss": 0.1934,
"step": 221
},
{
"epoch": 0.4344422700587084,
"grad_norm": 2.3347055437322064,
"learning_rate": 8.880242805730208e-06,
"loss": 0.1725,
"step": 222
},
{
"epoch": 0.436399217221135,
"grad_norm": 2.0637964590518663,
"learning_rate": 8.870531150932727e-06,
"loss": 0.1882,
"step": 223
},
{
"epoch": 0.4383561643835616,
"grad_norm": 2.674428942823747,
"learning_rate": 8.860782922495821e-06,
"loss": 0.1747,
"step": 224
},
{
"epoch": 0.44031311154598823,
"grad_norm": 2.2742296106240807,
"learning_rate": 8.850998212532998e-06,
"loss": 0.2379,
"step": 225
},
{
"epoch": 0.44227005870841485,
"grad_norm": 1.8587039860876922,
"learning_rate": 8.84117711350248e-06,
"loss": 0.1805,
"step": 226
},
{
"epoch": 0.44422700587084146,
"grad_norm": 2.4996506588033345,
"learning_rate": 8.831319718206353e-06,
"loss": 0.1926,
"step": 227
},
{
"epoch": 0.4461839530332681,
"grad_norm": 2.4201920192241353,
"learning_rate": 8.821426119789662e-06,
"loss": 0.2049,
"step": 228
},
{
"epoch": 0.4481409001956947,
"grad_norm": 2.3947181002335793,
"learning_rate": 8.811496411739552e-06,
"loss": 0.2055,
"step": 229
},
{
"epoch": 0.4500978473581213,
"grad_norm": 2.8927660475394585,
"learning_rate": 8.801530687884378e-06,
"loss": 0.2204,
"step": 230
},
{
"epoch": 0.4520547945205479,
"grad_norm": 4.2812142457221,
"learning_rate": 8.791529042392813e-06,
"loss": 0.2235,
"step": 231
},
{
"epoch": 0.45401174168297453,
"grad_norm": 2.5990468504023165,
"learning_rate": 8.781491569772966e-06,
"loss": 0.2153,
"step": 232
},
{
"epoch": 0.45596868884540115,
"grad_norm": 2.303420913649967,
"learning_rate": 8.771418364871483e-06,
"loss": 0.1853,
"step": 233
},
{
"epoch": 0.45792563600782776,
"grad_norm": 2.587544539691289,
"learning_rate": 8.761309522872657e-06,
"loss": 0.2136,
"step": 234
},
{
"epoch": 0.4598825831702544,
"grad_norm": 2.4243317462016774,
"learning_rate": 8.751165139297522e-06,
"loss": 0.2163,
"step": 235
},
{
"epoch": 0.461839530332681,
"grad_norm": 2.480230009126469,
"learning_rate": 8.740985310002956e-06,
"loss": 0.1733,
"step": 236
},
{
"epoch": 0.4637964774951076,
"grad_norm": 2.0406551042359484,
"learning_rate": 8.730770131180771e-06,
"loss": 0.1622,
"step": 237
},
{
"epoch": 0.4657534246575342,
"grad_norm": 2.2972584301769676,
"learning_rate": 8.720519699356804e-06,
"loss": 0.2411,
"step": 238
},
{
"epoch": 0.46771037181996084,
"grad_norm": 2.143621529692524,
"learning_rate": 8.71023411139001e-06,
"loss": 0.1565,
"step": 239
},
{
"epoch": 0.46966731898238745,
"grad_norm": 2.6594355800009986,
"learning_rate": 8.699913464471543e-06,
"loss": 0.2167,
"step": 240
},
{
"epoch": 0.47162426614481406,
"grad_norm": 1.9364714197948296,
"learning_rate": 8.689557856123838e-06,
"loss": 0.1917,
"step": 241
},
{
"epoch": 0.4735812133072407,
"grad_norm": 2.594548452571705,
"learning_rate": 8.679167384199686e-06,
"loss": 0.2053,
"step": 242
},
{
"epoch": 0.4755381604696673,
"grad_norm": 2.683685283122139,
"learning_rate": 8.668742146881316e-06,
"loss": 0.2513,
"step": 243
},
{
"epoch": 0.4774951076320939,
"grad_norm": 1.9502017134281109,
"learning_rate": 8.658282242679461e-06,
"loss": 0.1752,
"step": 244
},
{
"epoch": 0.4794520547945205,
"grad_norm": 2.1703811289170525,
"learning_rate": 8.647787770432439e-06,
"loss": 0.1793,
"step": 245
},
{
"epoch": 0.48140900195694714,
"grad_norm": 2.2411637178177726,
"learning_rate": 8.6372588293052e-06,
"loss": 0.1989,
"step": 246
},
{
"epoch": 0.48336594911937375,
"grad_norm": 1.9662273933653867,
"learning_rate": 8.626695518788403e-06,
"loss": 0.1727,
"step": 247
},
{
"epoch": 0.48532289628180036,
"grad_norm": 2.5135384160700016,
"learning_rate": 8.616097938697476e-06,
"loss": 0.2073,
"step": 248
},
{
"epoch": 0.487279843444227,
"grad_norm": 2.3034756521167576,
"learning_rate": 8.60546618917166e-06,
"loss": 0.2145,
"step": 249
},
{
"epoch": 0.4892367906066536,
"grad_norm": 2.15481510860185,
"learning_rate": 8.594800370673083e-06,
"loss": 0.2035,
"step": 250
},
{
"epoch": 0.4911937377690802,
"grad_norm": 2.5333361180263023,
"learning_rate": 8.584100583985791e-06,
"loss": 0.1949,
"step": 251
},
{
"epoch": 0.4931506849315068,
"grad_norm": 1.9994598084504018,
"learning_rate": 8.573366930214807e-06,
"loss": 0.1685,
"step": 252
},
{
"epoch": 0.49510763209393344,
"grad_norm": 2.376019589572924,
"learning_rate": 8.562599510785171e-06,
"loss": 0.1956,
"step": 253
},
{
"epoch": 0.49706457925636005,
"grad_norm": 2.2120050988896884,
"learning_rate": 8.551798427440985e-06,
"loss": 0.1671,
"step": 254
},
{
"epoch": 0.49902152641878667,
"grad_norm": 2.4559111510151426,
"learning_rate": 8.540963782244455e-06,
"loss": 0.2117,
"step": 255
},
{
"epoch": 0.5009784735812133,
"grad_norm": 2.5103983294113963,
"learning_rate": 8.53009567757491e-06,
"loss": 0.1918,
"step": 256
},
{
"epoch": 0.50293542074364,
"grad_norm": 2.0942954436320953,
"learning_rate": 8.519194216127854e-06,
"loss": 0.1814,
"step": 257
},
{
"epoch": 0.5048923679060665,
"grad_norm": 2.1088644604940607,
"learning_rate": 8.50825950091399e-06,
"loss": 0.1766,
"step": 258
},
{
"epoch": 0.5068493150684932,
"grad_norm": 2.706846628068543,
"learning_rate": 8.497291635258235e-06,
"loss": 0.2975,
"step": 259
},
{
"epoch": 0.5088062622309197,
"grad_norm": 2.759614134739062,
"learning_rate": 8.486290722798765e-06,
"loss": 0.2473,
"step": 260
},
{
"epoch": 0.5107632093933464,
"grad_norm": 2.055003880255024,
"learning_rate": 8.475256867486011e-06,
"loss": 0.171,
"step": 261
},
{
"epoch": 0.512720156555773,
"grad_norm": 1.948751515908191,
"learning_rate": 8.464190173581698e-06,
"loss": 0.1799,
"step": 262
},
{
"epoch": 0.5146771037181996,
"grad_norm": 1.8704298582706453,
"learning_rate": 8.45309074565785e-06,
"loss": 0.1854,
"step": 263
},
{
"epoch": 0.5166340508806262,
"grad_norm": 1.9403290225766887,
"learning_rate": 8.441958688595802e-06,
"loss": 0.1603,
"step": 264
},
{
"epoch": 0.5185909980430529,
"grad_norm": 2.0627854794372107,
"learning_rate": 8.430794107585208e-06,
"loss": 0.1998,
"step": 265
},
{
"epoch": 0.5205479452054794,
"grad_norm": 3.134871484922293,
"learning_rate": 8.419597108123054e-06,
"loss": 0.28,
"step": 266
},
{
"epoch": 0.5225048923679061,
"grad_norm": 2.1328599665868437,
"learning_rate": 8.408367796012652e-06,
"loss": 0.1762,
"step": 267
},
{
"epoch": 0.5244618395303327,
"grad_norm": 1.9861081126962148,
"learning_rate": 8.397106277362647e-06,
"loss": 0.1785,
"step": 268
},
{
"epoch": 0.5264187866927593,
"grad_norm": 2.570735848683898,
"learning_rate": 8.38581265858601e-06,
"loss": 0.1896,
"step": 269
},
{
"epoch": 0.5283757338551859,
"grad_norm": 2.3585440623771348,
"learning_rate": 8.374487046399035e-06,
"loss": 0.1905,
"step": 270
},
{
"epoch": 0.5303326810176126,
"grad_norm": 2.0282507335721736,
"learning_rate": 8.363129547820333e-06,
"loss": 0.1839,
"step": 271
},
{
"epoch": 0.5322896281800391,
"grad_norm": 2.076096289424388,
"learning_rate": 8.35174027016981e-06,
"loss": 0.1723,
"step": 272
},
{
"epoch": 0.5342465753424658,
"grad_norm": 2.2002147091456914,
"learning_rate": 8.340319321067668e-06,
"loss": 0.1939,
"step": 273
},
{
"epoch": 0.5362035225048923,
"grad_norm": 2.2187956327644955,
"learning_rate": 8.328866808433378e-06,
"loss": 0.1886,
"step": 274
},
{
"epoch": 0.538160469667319,
"grad_norm": 1.948043222652721,
"learning_rate": 8.317382840484663e-06,
"loss": 0.1956,
"step": 275
},
{
"epoch": 0.5401174168297456,
"grad_norm": 2.2498449881181353,
"learning_rate": 8.305867525736475e-06,
"loss": 0.2214,
"step": 276
},
{
"epoch": 0.5420743639921722,
"grad_norm": 2.2583633223543855,
"learning_rate": 8.29432097299997e-06,
"loss": 0.2291,
"step": 277
},
{
"epoch": 0.5440313111545988,
"grad_norm": 2.172742832768641,
"learning_rate": 8.28274329138148e-06,
"loss": 0.1946,
"step": 278
},
{
"epoch": 0.5459882583170255,
"grad_norm": 2.8694946256241907,
"learning_rate": 8.271134590281482e-06,
"loss": 0.2084,
"step": 279
},
{
"epoch": 0.547945205479452,
"grad_norm": 2.409144386542836,
"learning_rate": 8.259494979393563e-06,
"loss": 0.2286,
"step": 280
},
{
"epoch": 0.5499021526418787,
"grad_norm": 2.1209827677563395,
"learning_rate": 8.247824568703386e-06,
"loss": 0.1681,
"step": 281
},
{
"epoch": 0.5518590998043053,
"grad_norm": 2.1367613658790154,
"learning_rate": 8.236123468487649e-06,
"loss": 0.1765,
"step": 282
},
{
"epoch": 0.5538160469667319,
"grad_norm": 2.150473396723976,
"learning_rate": 8.224391789313038e-06,
"loss": 0.2243,
"step": 283
},
{
"epoch": 0.5557729941291585,
"grad_norm": 1.9172230370495653,
"learning_rate": 8.2126296420352e-06,
"loss": 0.1764,
"step": 284
},
{
"epoch": 0.5577299412915852,
"grad_norm": 2.383363524883776,
"learning_rate": 8.200837137797665e-06,
"loss": 0.2118,
"step": 285
},
{
"epoch": 0.5596868884540117,
"grad_norm": 2.153619634620154,
"learning_rate": 8.189014388030834e-06,
"loss": 0.2059,
"step": 286
},
{
"epoch": 0.5616438356164384,
"grad_norm": 2.1441994457955467,
"learning_rate": 8.177161504450887e-06,
"loss": 0.1775,
"step": 287
},
{
"epoch": 0.5636007827788649,
"grad_norm": 1.6038648606526436,
"learning_rate": 8.16527859905876e-06,
"loss": 0.1778,
"step": 288
},
{
"epoch": 0.5655577299412916,
"grad_norm": 1.7633769339721965,
"learning_rate": 8.153365784139065e-06,
"loss": 0.1541,
"step": 289
},
{
"epoch": 0.5675146771037182,
"grad_norm": 1.8018002573379492,
"learning_rate": 8.141423172259038e-06,
"loss": 0.1447,
"step": 290
},
{
"epoch": 0.5694716242661448,
"grad_norm": 2.7466777060039056,
"learning_rate": 8.129450876267475e-06,
"loss": 0.245,
"step": 291
},
{
"epoch": 0.5714285714285714,
"grad_norm": 2.833360917876741,
"learning_rate": 8.117449009293668e-06,
"loss": 0.2029,
"step": 292
},
{
"epoch": 0.5733855185909981,
"grad_norm": 2.10574467332215,
"learning_rate": 8.105417684746327e-06,
"loss": 0.1509,
"step": 293
},
{
"epoch": 0.5753424657534246,
"grad_norm": 1.8282908744111137,
"learning_rate": 8.093357016312518e-06,
"loss": 0.1617,
"step": 294
},
{
"epoch": 0.5772994129158513,
"grad_norm": 2.0000453924465886,
"learning_rate": 8.081267117956581e-06,
"loss": 0.1522,
"step": 295
},
{
"epoch": 0.5792563600782779,
"grad_norm": 2.009827552095151,
"learning_rate": 8.069148103919064e-06,
"loss": 0.1436,
"step": 296
},
{
"epoch": 0.5812133072407045,
"grad_norm": 2.4734575920532875,
"learning_rate": 8.057000088715631e-06,
"loss": 0.2021,
"step": 297
},
{
"epoch": 0.5831702544031311,
"grad_norm": 1.611541409791449,
"learning_rate": 8.044823187135984e-06,
"loss": 0.128,
"step": 298
},
{
"epoch": 0.5851272015655578,
"grad_norm": 2.272604965276249,
"learning_rate": 8.032617514242784e-06,
"loss": 0.2376,
"step": 299
},
{
"epoch": 0.5870841487279843,
"grad_norm": 1.7826119333125128,
"learning_rate": 8.020383185370559e-06,
"loss": 0.1271,
"step": 300
},
{
"epoch": 0.589041095890411,
"grad_norm": 2.332798169783043,
"learning_rate": 8.008120316124612e-06,
"loss": 0.1933,
"step": 301
},
{
"epoch": 0.5909980430528375,
"grad_norm": 2.198482018882503,
"learning_rate": 7.995829022379937e-06,
"loss": 0.1782,
"step": 302
},
{
"epoch": 0.5929549902152642,
"grad_norm": 1.8388860798415916,
"learning_rate": 7.983509420280107e-06,
"loss": 0.1486,
"step": 303
},
{
"epoch": 0.5949119373776908,
"grad_norm": 1.9960324775951768,
"learning_rate": 7.9711616262362e-06,
"loss": 0.1903,
"step": 304
},
{
"epoch": 0.5968688845401174,
"grad_norm": 2.115797877624949,
"learning_rate": 7.958785756925684e-06,
"loss": 0.1922,
"step": 305
},
{
"epoch": 0.598825831702544,
"grad_norm": 1.8193570816264995,
"learning_rate": 7.94638192929131e-06,
"loss": 0.1772,
"step": 306
},
{
"epoch": 0.6007827788649707,
"grad_norm": 2.3493170468607256,
"learning_rate": 7.933950260540023e-06,
"loss": 0.2133,
"step": 307
},
{
"epoch": 0.6027397260273972,
"grad_norm": 1.879332176666921,
"learning_rate": 7.921490868141843e-06,
"loss": 0.1766,
"step": 308
},
{
"epoch": 0.6046966731898239,
"grad_norm": 1.7341708223195131,
"learning_rate": 7.909003869828757e-06,
"loss": 0.1677,
"step": 309
},
{
"epoch": 0.6066536203522505,
"grad_norm": 2.2507292732004007,
"learning_rate": 7.896489383593606e-06,
"loss": 0.1601,
"step": 310
},
{
"epoch": 0.6086105675146771,
"grad_norm": 2.0594027791268923,
"learning_rate": 7.883947527688976e-06,
"loss": 0.1532,
"step": 311
},
{
"epoch": 0.6105675146771037,
"grad_norm": 2.408708926474526,
"learning_rate": 7.871378420626072e-06,
"loss": 0.2021,
"step": 312
},
{
"epoch": 0.6125244618395304,
"grad_norm": 2.2830641099498035,
"learning_rate": 7.858782181173602e-06,
"loss": 0.1723,
"step": 313
},
{
"epoch": 0.6144814090019569,
"grad_norm": 1.9982758429450047,
"learning_rate": 7.84615892835666e-06,
"loss": 0.16,
"step": 314
},
{
"epoch": 0.6164383561643836,
"grad_norm": 2.634699473199227,
"learning_rate": 7.83350878145559e-06,
"loss": 0.2396,
"step": 315
},
{
"epoch": 0.6183953033268101,
"grad_norm": 2.439100354318057,
"learning_rate": 7.820831860004867e-06,
"loss": 0.135,
"step": 316
},
{
"epoch": 0.6203522504892368,
"grad_norm": 2.293955055772235,
"learning_rate": 7.808128283791967e-06,
"loss": 0.2077,
"step": 317
},
{
"epoch": 0.6223091976516634,
"grad_norm": 1.5197542859653332,
"learning_rate": 7.795398172856234e-06,
"loss": 0.1289,
"step": 318
},
{
"epoch": 0.62426614481409,
"grad_norm": 1.8961618220022762,
"learning_rate": 7.782641647487741e-06,
"loss": 0.1331,
"step": 319
},
{
"epoch": 0.6262230919765166,
"grad_norm": 2.0491282710401677,
"learning_rate": 7.769858828226165e-06,
"loss": 0.1697,
"step": 320
},
{
"epoch": 0.6281800391389433,
"grad_norm": 1.9617574024954327,
"learning_rate": 7.757049835859635e-06,
"loss": 0.1639,
"step": 321
},
{
"epoch": 0.6301369863013698,
"grad_norm": 2.71550580279651,
"learning_rate": 7.744214791423597e-06,
"loss": 0.2303,
"step": 322
},
{
"epoch": 0.6320939334637965,
"grad_norm": 2.433519685679652,
"learning_rate": 7.731353816199672e-06,
"loss": 0.2156,
"step": 323
},
{
"epoch": 0.6340508806262231,
"grad_norm": 2.664762358240488,
"learning_rate": 7.718467031714506e-06,
"loss": 0.1864,
"step": 324
},
{
"epoch": 0.6360078277886497,
"grad_norm": 2.340396402101541,
"learning_rate": 7.705554559738623e-06,
"loss": 0.24,
"step": 325
},
{
"epoch": 0.6379647749510763,
"grad_norm": 2.648004004757832,
"learning_rate": 7.692616522285278e-06,
"loss": 0.1616,
"step": 326
},
{
"epoch": 0.639921722113503,
"grad_norm": 2.1379122455101895,
"learning_rate": 7.679653041609296e-06,
"loss": 0.1608,
"step": 327
},
{
"epoch": 0.6418786692759295,
"grad_norm": 2.016213472489541,
"learning_rate": 7.666664240205922e-06,
"loss": 0.1862,
"step": 328
},
{
"epoch": 0.6438356164383562,
"grad_norm": 2.023888144392631,
"learning_rate": 7.653650240809667e-06,
"loss": 0.1689,
"step": 329
},
{
"epoch": 0.6457925636007827,
"grad_norm": 1.9091564689003773,
"learning_rate": 7.640611166393142e-06,
"loss": 0.1462,
"step": 330
},
{
"epoch": 0.6477495107632094,
"grad_norm": 1.8185912288115094,
"learning_rate": 7.627547140165899e-06,
"loss": 0.1647,
"step": 331
},
{
"epoch": 0.649706457925636,
"grad_norm": 1.9563867765498288,
"learning_rate": 7.614458285573262e-06,
"loss": 0.1722,
"step": 332
},
{
"epoch": 0.6516634050880626,
"grad_norm": 1.8173428796019335,
"learning_rate": 7.60134472629517e-06,
"loss": 0.1437,
"step": 333
},
{
"epoch": 0.6536203522504892,
"grad_norm": 2.3721957835019847,
"learning_rate": 7.588206586245001e-06,
"loss": 0.1989,
"step": 334
},
{
"epoch": 0.6555772994129159,
"grad_norm": 2.0578841275944266,
"learning_rate": 7.575043989568401e-06,
"loss": 0.1802,
"step": 335
},
{
"epoch": 0.6575342465753424,
"grad_norm": 2.33956924434927,
"learning_rate": 7.56185706064212e-06,
"loss": 0.1776,
"step": 336
},
{
"epoch": 0.6594911937377691,
"grad_norm": 2.6488764141970744,
"learning_rate": 7.548645924072816e-06,
"loss": 0.1789,
"step": 337
},
{
"epoch": 0.6614481409001957,
"grad_norm": 2.0121629729257657,
"learning_rate": 7.535410704695907e-06,
"loss": 0.1784,
"step": 338
},
{
"epoch": 0.6634050880626223,
"grad_norm": 2.5686994589463423,
"learning_rate": 7.522151527574362e-06,
"loss": 0.2442,
"step": 339
},
{
"epoch": 0.6653620352250489,
"grad_norm": 2.836071043986302,
"learning_rate": 7.508868517997544e-06,
"loss": 0.2319,
"step": 340
},
{
"epoch": 0.6673189823874756,
"grad_norm": 2.227411079361732,
"learning_rate": 7.495561801480009e-06,
"loss": 0.1767,
"step": 341
},
{
"epoch": 0.6692759295499021,
"grad_norm": 2.1374935257419923,
"learning_rate": 7.4822315037603245e-06,
"loss": 0.1884,
"step": 342
},
{
"epoch": 0.6712328767123288,
"grad_norm": 1.5918580893658882,
"learning_rate": 7.468877750799887e-06,
"loss": 0.125,
"step": 343
},
{
"epoch": 0.6731898238747553,
"grad_norm": 2.1374541789525177,
"learning_rate": 7.455500668781725e-06,
"loss": 0.1634,
"step": 344
},
{
"epoch": 0.675146771037182,
"grad_norm": 1.9794879471155153,
"learning_rate": 7.442100384109313e-06,
"loss": 0.1868,
"step": 345
},
{
"epoch": 0.6771037181996086,
"grad_norm": 2.148898308745395,
"learning_rate": 7.428677023405366e-06,
"loss": 0.1833,
"step": 346
},
{
"epoch": 0.6790606653620352,
"grad_norm": 2.0421994641633905,
"learning_rate": 7.415230713510657e-06,
"loss": 0.1697,
"step": 347
},
{
"epoch": 0.6810176125244618,
"grad_norm": 2.1165597448492535,
"learning_rate": 7.40176158148281e-06,
"loss": 0.1849,
"step": 348
},
{
"epoch": 0.6829745596868885,
"grad_norm": 2.761270328760098,
"learning_rate": 7.3882697545951e-06,
"loss": 0.2161,
"step": 349
},
{
"epoch": 0.684931506849315,
"grad_norm": 2.179460448802176,
"learning_rate": 7.374755360335253e-06,
"loss": 0.199,
"step": 350
},
{
"epoch": 0.6868884540117417,
"grad_norm": 3.005401317413042,
"learning_rate": 7.361218526404237e-06,
"loss": 0.1952,
"step": 351
},
{
"epoch": 0.6888454011741683,
"grad_norm": 2.2722310438981412,
"learning_rate": 7.3476593807150606e-06,
"loss": 0.1965,
"step": 352
},
{
"epoch": 0.6908023483365949,
"grad_norm": 2.287920474174499,
"learning_rate": 7.334078051391563e-06,
"loss": 0.1676,
"step": 353
},
{
"epoch": 0.6927592954990215,
"grad_norm": 2.001353989212772,
"learning_rate": 7.320474666767201e-06,
"loss": 0.13,
"step": 354
},
{
"epoch": 0.6947162426614482,
"grad_norm": 2.241629320644228,
"learning_rate": 7.3068493553838386e-06,
"loss": 0.1819,
"step": 355
},
{
"epoch": 0.6966731898238747,
"grad_norm": 2.293034396349547,
"learning_rate": 7.293202245990526e-06,
"loss": 0.2079,
"step": 356
},
{
"epoch": 0.6986301369863014,
"grad_norm": 2.0722814594404517,
"learning_rate": 7.279533467542295e-06,
"loss": 0.1691,
"step": 357
},
{
"epoch": 0.700587084148728,
"grad_norm": 2.013824330796792,
"learning_rate": 7.265843149198931e-06,
"loss": 0.1483,
"step": 358
},
{
"epoch": 0.7025440313111546,
"grad_norm": 1.8293971221527265,
"learning_rate": 7.252131420323757e-06,
"loss": 0.1502,
"step": 359
},
{
"epoch": 0.7045009784735812,
"grad_norm": 2.2775240476862098,
"learning_rate": 7.238398410482408e-06,
"loss": 0.1886,
"step": 360
},
{
"epoch": 0.7064579256360078,
"grad_norm": 2.3849802054367077,
"learning_rate": 7.224644249441606e-06,
"loss": 0.1495,
"step": 361
},
{
"epoch": 0.7084148727984344,
"grad_norm": 2.3168751357812907,
"learning_rate": 7.210869067167942e-06,
"loss": 0.1873,
"step": 362
},
{
"epoch": 0.7103718199608611,
"grad_norm": 2.046664547966605,
"learning_rate": 7.197072993826637e-06,
"loss": 0.1612,
"step": 363
},
{
"epoch": 0.7123287671232876,
"grad_norm": 2.5318621007477273,
"learning_rate": 7.183256159780321e-06,
"loss": 0.2566,
"step": 364
},
{
"epoch": 0.7142857142857143,
"grad_norm": 2.576097997051626,
"learning_rate": 7.169418695587791e-06,
"loss": 0.2234,
"step": 365
},
{
"epoch": 0.7162426614481409,
"grad_norm": 2.0978024481980633,
"learning_rate": 7.155560732002792e-06,
"loss": 0.1662,
"step": 366
},
{
"epoch": 0.7181996086105675,
"grad_norm": 2.363993610816794,
"learning_rate": 7.141682399972765e-06,
"loss": 0.1935,
"step": 367
},
{
"epoch": 0.7201565557729941,
"grad_norm": 3.068135522185098,
"learning_rate": 7.127783830637625e-06,
"loss": 0.2496,
"step": 368
},
{
"epoch": 0.7221135029354208,
"grad_norm": 2.093219630568525,
"learning_rate": 7.113865155328506e-06,
"loss": 0.1906,
"step": 369
},
{
"epoch": 0.7240704500978473,
"grad_norm": 2.5084679208006384,
"learning_rate": 7.099926505566537e-06,
"loss": 0.2078,
"step": 370
},
{
"epoch": 0.726027397260274,
"grad_norm": 2.1721439053705387,
"learning_rate": 7.085968013061585e-06,
"loss": 0.1625,
"step": 371
},
{
"epoch": 0.7279843444227005,
"grad_norm": 2.644362370137884,
"learning_rate": 7.071989809711018e-06,
"loss": 0.263,
"step": 372
},
{
"epoch": 0.7299412915851272,
"grad_norm": 2.0980923615020814,
"learning_rate": 7.057992027598459e-06,
"loss": 0.1756,
"step": 373
},
{
"epoch": 0.7318982387475538,
"grad_norm": 2.1688025512369813,
"learning_rate": 7.043974798992532e-06,
"loss": 0.182,
"step": 374
},
{
"epoch": 0.7338551859099804,
"grad_norm": 2.249817038860846,
"learning_rate": 7.029938256345617e-06,
"loss": 0.1578,
"step": 375
},
{
"epoch": 0.735812133072407,
"grad_norm": 2.531776097231375,
"learning_rate": 7.015882532292598e-06,
"loss": 0.1886,
"step": 376
},
{
"epoch": 0.7377690802348337,
"grad_norm": 2.367720121263562,
"learning_rate": 7.001807759649609e-06,
"loss": 0.1973,
"step": 377
},
{
"epoch": 0.7397260273972602,
"grad_norm": 2.315462953608418,
"learning_rate": 6.987714071412781e-06,
"loss": 0.1946,
"step": 378
},
{
"epoch": 0.7416829745596869,
"grad_norm": 2.04094554695451,
"learning_rate": 6.973601600756978e-06,
"loss": 0.1998,
"step": 379
},
{
"epoch": 0.7436399217221135,
"grad_norm": 1.8432650259541994,
"learning_rate": 6.959470481034547e-06,
"loss": 0.1796,
"step": 380
},
{
"epoch": 0.7455968688845401,
"grad_norm": 2.112901859688788,
"learning_rate": 6.945320845774057e-06,
"loss": 0.1735,
"step": 381
},
{
"epoch": 0.7475538160469667,
"grad_norm": 1.7613330270494343,
"learning_rate": 6.931152828679033e-06,
"loss": 0.1502,
"step": 382
},
{
"epoch": 0.7495107632093934,
"grad_norm": 2.674901871712495,
"learning_rate": 6.9169665636266925e-06,
"loss": 0.2437,
"step": 383
},
{
"epoch": 0.7514677103718199,
"grad_norm": 2.5422669549986128,
"learning_rate": 6.902762184666687e-06,
"loss": 0.1897,
"step": 384
},
{
"epoch": 0.7534246575342466,
"grad_norm": 1.9666127900842278,
"learning_rate": 6.888539826019824e-06,
"loss": 0.1539,
"step": 385
},
{
"epoch": 0.7553816046966731,
"grad_norm": 2.406136825101305,
"learning_rate": 6.874299622076817e-06,
"loss": 0.2033,
"step": 386
},
{
"epoch": 0.7573385518590998,
"grad_norm": 1.7232207896617806,
"learning_rate": 6.860041707396993e-06,
"loss": 0.1476,
"step": 387
},
{
"epoch": 0.7592954990215264,
"grad_norm": 2.0480956588223336,
"learning_rate": 6.845766216707037e-06,
"loss": 0.1438,
"step": 388
},
{
"epoch": 0.761252446183953,
"grad_norm": 1.9065836666555238,
"learning_rate": 6.831473284899715e-06,
"loss": 0.153,
"step": 389
},
{
"epoch": 0.7632093933463796,
"grad_norm": 2.544910841115613,
"learning_rate": 6.817163047032598e-06,
"loss": 0.217,
"step": 390
},
{
"epoch": 0.7651663405088063,
"grad_norm": 2.373086577098212,
"learning_rate": 6.802835638326784e-06,
"loss": 0.2042,
"step": 391
},
{
"epoch": 0.7671232876712328,
"grad_norm": 2.188989414027513,
"learning_rate": 6.788491194165629e-06,
"loss": 0.1821,
"step": 392
},
{
"epoch": 0.7690802348336595,
"grad_norm": 1.882204780094551,
"learning_rate": 6.774129850093451e-06,
"loss": 0.1207,
"step": 393
},
{
"epoch": 0.7710371819960861,
"grad_norm": 2.0143426714345862,
"learning_rate": 6.759751741814271e-06,
"loss": 0.1703,
"step": 394
},
{
"epoch": 0.7729941291585127,
"grad_norm": 1.8577228941871937,
"learning_rate": 6.74535700519051e-06,
"loss": 0.1265,
"step": 395
},
{
"epoch": 0.7749510763209393,
"grad_norm": 1.9986366972767378,
"learning_rate": 6.7309457762417214e-06,
"loss": 0.1706,
"step": 396
},
{
"epoch": 0.776908023483366,
"grad_norm": 2.55488139449399,
"learning_rate": 6.7165181911432955e-06,
"loss": 0.2356,
"step": 397
},
{
"epoch": 0.7788649706457925,
"grad_norm": 2.023963402324868,
"learning_rate": 6.702074386225175e-06,
"loss": 0.2164,
"step": 398
},
{
"epoch": 0.7808219178082192,
"grad_norm": 2.0119370135736605,
"learning_rate": 6.687614497970567e-06,
"loss": 0.1775,
"step": 399
},
{
"epoch": 0.7827788649706457,
"grad_norm": 2.2378042919112935,
"learning_rate": 6.67313866301466e-06,
"loss": 0.1607,
"step": 400
},
{
"epoch": 0.7847358121330724,
"grad_norm": 2.4473506044542184,
"learning_rate": 6.658647018143322e-06,
"loss": 0.1968,
"step": 401
},
{
"epoch": 0.786692759295499,
"grad_norm": 2.065741540272696,
"learning_rate": 6.6441397002918165e-06,
"loss": 0.1627,
"step": 402
},
{
"epoch": 0.7886497064579256,
"grad_norm": 2.1800644343624764,
"learning_rate": 6.629616846543497e-06,
"loss": 0.2436,
"step": 403
},
{
"epoch": 0.7906066536203522,
"grad_norm": 2.0766431734944812,
"learning_rate": 6.615078594128531e-06,
"loss": 0.1624,
"step": 404
},
{
"epoch": 0.7925636007827789,
"grad_norm": 1.959807511254783,
"learning_rate": 6.600525080422583e-06,
"loss": 0.1766,
"step": 405
},
{
"epoch": 0.7945205479452054,
"grad_norm": 1.94596835349326,
"learning_rate": 6.585956442945531e-06,
"loss": 0.1634,
"step": 406
},
{
"epoch": 0.7964774951076321,
"grad_norm": 1.957611360720237,
"learning_rate": 6.571372819360158e-06,
"loss": 0.1602,
"step": 407
},
{
"epoch": 0.7984344422700587,
"grad_norm": 2.157155308661062,
"learning_rate": 6.5567743474708545e-06,
"loss": 0.1846,
"step": 408
},
{
"epoch": 0.8003913894324853,
"grad_norm": 2.306982809696333,
"learning_rate": 6.5421611652223185e-06,
"loss": 0.2247,
"step": 409
},
{
"epoch": 0.8023483365949119,
"grad_norm": 1.82765638257364,
"learning_rate": 6.52753341069825e-06,
"loss": 0.1623,
"step": 410
},
{
"epoch": 0.8043052837573386,
"grad_norm": 2.2805728167927413,
"learning_rate": 6.512891222120044e-06,
"loss": 0.1893,
"step": 411
},
{
"epoch": 0.8062622309197651,
"grad_norm": 2.0168151394995406,
"learning_rate": 6.498234737845488e-06,
"loss": 0.1693,
"step": 412
},
{
"epoch": 0.8082191780821918,
"grad_norm": 2.198351064216878,
"learning_rate": 6.483564096367452e-06,
"loss": 0.1635,
"step": 413
},
{
"epoch": 0.8101761252446184,
"grad_norm": 2.081853940644963,
"learning_rate": 6.468879436312584e-06,
"loss": 0.2099,
"step": 414
},
{
"epoch": 0.812133072407045,
"grad_norm": 1.890784041976538,
"learning_rate": 6.454180896439993e-06,
"loss": 0.1817,
"step": 415
},
{
"epoch": 0.8140900195694716,
"grad_norm": 2.4301334004946082,
"learning_rate": 6.439468615639947e-06,
"loss": 0.2134,
"step": 416
},
{
"epoch": 0.8160469667318982,
"grad_norm": 2.2710552894492797,
"learning_rate": 6.424742732932547e-06,
"loss": 0.1803,
"step": 417
},
{
"epoch": 0.8180039138943248,
"grad_norm": 1.8802636849356569,
"learning_rate": 6.410003387466433e-06,
"loss": 0.1701,
"step": 418
},
{
"epoch": 0.8199608610567515,
"grad_norm": 1.8785987026944129,
"learning_rate": 6.3952507185174505e-06,
"loss": 0.1515,
"step": 419
},
{
"epoch": 0.821917808219178,
"grad_norm": 2.448476109773896,
"learning_rate": 6.380484865487346e-06,
"loss": 0.186,
"step": 420
},
{
"epoch": 0.8238747553816047,
"grad_norm": 2.464947399234228,
"learning_rate": 6.365705967902439e-06,
"loss": 0.1628,
"step": 421
},
{
"epoch": 0.8258317025440313,
"grad_norm": 1.5135706949117274,
"learning_rate": 6.35091416541232e-06,
"loss": 0.1552,
"step": 422
},
{
"epoch": 0.8277886497064579,
"grad_norm": 2.0339945027320625,
"learning_rate": 6.3361095977885125e-06,
"loss": 0.1922,
"step": 423
},
{
"epoch": 0.8297455968688845,
"grad_norm": 2.4839169687346723,
"learning_rate": 6.3212924049231676e-06,
"loss": 0.2332,
"step": 424
},
{
"epoch": 0.8317025440313112,
"grad_norm": 1.9431387370715525,
"learning_rate": 6.306462726827733e-06,
"loss": 0.1869,
"step": 425
},
{
"epoch": 0.8336594911937377,
"grad_norm": 2.1199143348931306,
"learning_rate": 6.29162070363163e-06,
"loss": 0.1476,
"step": 426
},
{
"epoch": 0.8356164383561644,
"grad_norm": 1.8748486888730402,
"learning_rate": 6.276766475580935e-06,
"loss": 0.1663,
"step": 427
},
{
"epoch": 0.837573385518591,
"grad_norm": 2.1588105700286415,
"learning_rate": 6.261900183037053e-06,
"loss": 0.238,
"step": 428
},
{
"epoch": 0.8395303326810176,
"grad_norm": 2.189612463579074,
"learning_rate": 6.247021966475385e-06,
"loss": 0.179,
"step": 429
},
{
"epoch": 0.8414872798434442,
"grad_norm": 2.1892679059620543,
"learning_rate": 6.232131966484007e-06,
"loss": 0.1269,
"step": 430
},
{
"epoch": 0.8434442270058709,
"grad_norm": 2.6317084779940534,
"learning_rate": 6.217230323762338e-06,
"loss": 0.1994,
"step": 431
},
{
"epoch": 0.8454011741682974,
"grad_norm": 1.7973110272935775,
"learning_rate": 6.202317179119817e-06,
"loss": 0.1276,
"step": 432
},
{
"epoch": 0.8473581213307241,
"grad_norm": 2.3523541173540274,
"learning_rate": 6.187392673474562e-06,
"loss": 0.19,
"step": 433
},
{
"epoch": 0.8493150684931506,
"grad_norm": 1.9466611158615799,
"learning_rate": 6.1724569478520495e-06,
"loss": 0.1354,
"step": 434
},
{
"epoch": 0.8512720156555773,
"grad_norm": 2.030324983182261,
"learning_rate": 6.15751014338377e-06,
"loss": 0.1634,
"step": 435
},
{
"epoch": 0.8532289628180039,
"grad_norm": 2.535851970665704,
"learning_rate": 6.142552401305907e-06,
"loss": 0.2085,
"step": 436
},
{
"epoch": 0.8551859099804305,
"grad_norm": 2.776722870414143,
"learning_rate": 6.1275838629579906e-06,
"loss": 0.2938,
"step": 437
},
{
"epoch": 0.8571428571428571,
"grad_norm": 2.5286257239586565,
"learning_rate": 6.112604669781572e-06,
"loss": 0.2056,
"step": 438
},
{
"epoch": 0.8590998043052838,
"grad_norm": 2.2206772657404343,
"learning_rate": 6.097614963318883e-06,
"loss": 0.1711,
"step": 439
},
{
"epoch": 0.8610567514677103,
"grad_norm": 2.151811845313509,
"learning_rate": 6.0826148852114945e-06,
"loss": 0.175,
"step": 440
},
{
"epoch": 0.863013698630137,
"grad_norm": 2.56573554662175,
"learning_rate": 6.067604577198981e-06,
"loss": 0.1751,
"step": 441
},
{
"epoch": 0.8649706457925636,
"grad_norm": 1.8051537157371025,
"learning_rate": 6.052584181117589e-06,
"loss": 0.1292,
"step": 442
},
{
"epoch": 0.8669275929549902,
"grad_norm": 2.058692378129427,
"learning_rate": 6.037553838898882e-06,
"loss": 0.1923,
"step": 443
},
{
"epoch": 0.8688845401174168,
"grad_norm": 2.3162309090883504,
"learning_rate": 6.022513692568412e-06,
"loss": 0.1804,
"step": 444
},
{
"epoch": 0.8708414872798435,
"grad_norm": 1.8005749395750652,
"learning_rate": 6.00746388424437e-06,
"loss": 0.1809,
"step": 445
},
{
"epoch": 0.87279843444227,
"grad_norm": 2.426858839571719,
"learning_rate": 5.992404556136247e-06,
"loss": 0.1817,
"step": 446
},
{
"epoch": 0.8747553816046967,
"grad_norm": 2.5119474822070105,
"learning_rate": 5.977335850543488e-06,
"loss": 0.2053,
"step": 447
},
{
"epoch": 0.8767123287671232,
"grad_norm": 2.0283698043023364,
"learning_rate": 5.96225790985415e-06,
"loss": 0.1259,
"step": 448
},
{
"epoch": 0.8786692759295499,
"grad_norm": 2.7604953637866414,
"learning_rate": 5.947170876543553e-06,
"loss": 0.1657,
"step": 449
},
{
"epoch": 0.8806262230919765,
"grad_norm": 2.128906444124902,
"learning_rate": 5.932074893172934e-06,
"loss": 0.1932,
"step": 450
},
{
"epoch": 0.8825831702544031,
"grad_norm": 1.983151604590541,
"learning_rate": 5.916970102388104e-06,
"loss": 0.1976,
"step": 451
},
{
"epoch": 0.8845401174168297,
"grad_norm": 2.145851420074306,
"learning_rate": 5.9018566469181e-06,
"loss": 0.1423,
"step": 452
},
{
"epoch": 0.8864970645792564,
"grad_norm": 2.1241035461670497,
"learning_rate": 5.886734669573827e-06,
"loss": 0.1771,
"step": 453
},
{
"epoch": 0.8884540117416829,
"grad_norm": 1.9479268148718478,
"learning_rate": 5.87160431324672e-06,
"loss": 0.1328,
"step": 454
},
{
"epoch": 0.8904109589041096,
"grad_norm": 2.1174977715657293,
"learning_rate": 5.856465720907388e-06,
"loss": 0.211,
"step": 455
},
{
"epoch": 0.8923679060665362,
"grad_norm": 2.4205453845240323,
"learning_rate": 5.841319035604267e-06,
"loss": 0.1666,
"step": 456
},
{
"epoch": 0.8943248532289628,
"grad_norm": 2.297680335093687,
"learning_rate": 5.826164400462259e-06,
"loss": 0.2323,
"step": 457
},
{
"epoch": 0.8962818003913894,
"grad_norm": 2.493174486491035,
"learning_rate": 5.8110019586813944e-06,
"loss": 0.1973,
"step": 458
},
{
"epoch": 0.898238747553816,
"grad_norm": 2.2644834447659963,
"learning_rate": 5.795831853535461e-06,
"loss": 0.1763,
"step": 459
},
{
"epoch": 0.9001956947162426,
"grad_norm": 2.3845938800970656,
"learning_rate": 5.780654228370669e-06,
"loss": 0.1862,
"step": 460
},
{
"epoch": 0.9021526418786693,
"grad_norm": 2.344583110265581,
"learning_rate": 5.76546922660428e-06,
"loss": 0.1807,
"step": 461
},
{
"epoch": 0.9041095890410958,
"grad_norm": 2.1398071686563975,
"learning_rate": 5.7502769917232635e-06,
"loss": 0.1589,
"step": 462
},
{
"epoch": 0.9060665362035225,
"grad_norm": 1.9714418531756854,
"learning_rate": 5.735077667282935e-06,
"loss": 0.129,
"step": 463
},
{
"epoch": 0.9080234833659491,
"grad_norm": 2.212318091375408,
"learning_rate": 5.719871396905603e-06,
"loss": 0.1943,
"step": 464
},
{
"epoch": 0.9099804305283757,
"grad_norm": 1.5451056855048066,
"learning_rate": 5.704658324279203e-06,
"loss": 0.1701,
"step": 465
},
{
"epoch": 0.9119373776908023,
"grad_norm": 2.139023612246954,
"learning_rate": 5.6894385931559555e-06,
"loss": 0.1888,
"step": 466
},
{
"epoch": 0.913894324853229,
"grad_norm": 2.1793452689245143,
"learning_rate": 5.674212347350996e-06,
"loss": 0.1588,
"step": 467
},
{
"epoch": 0.9158512720156555,
"grad_norm": 1.9974126080003265,
"learning_rate": 5.658979730741014e-06,
"loss": 0.1713,
"step": 468
},
{
"epoch": 0.9178082191780822,
"grad_norm": 2.380983147927289,
"learning_rate": 5.643740887262905e-06,
"loss": 0.1815,
"step": 469
},
{
"epoch": 0.9197651663405088,
"grad_norm": 1.9336954495370033,
"learning_rate": 5.6284959609124e-06,
"loss": 0.1507,
"step": 470
},
{
"epoch": 0.9217221135029354,
"grad_norm": 2.23889600104245,
"learning_rate": 5.613245095742711e-06,
"loss": 0.2161,
"step": 471
},
{
"epoch": 0.923679060665362,
"grad_norm": 1.934229378085192,
"learning_rate": 5.5979884358631665e-06,
"loss": 0.1526,
"step": 472
},
{
"epoch": 0.9256360078277887,
"grad_norm": 1.9889854036080192,
"learning_rate": 5.582726125437846e-06,
"loss": 0.1731,
"step": 473
},
{
"epoch": 0.9275929549902152,
"grad_norm": 2.360199878036082,
"learning_rate": 5.567458308684233e-06,
"loss": 0.1513,
"step": 474
},
{
"epoch": 0.9295499021526419,
"grad_norm": 2.3158352876921926,
"learning_rate": 5.5521851298718295e-06,
"loss": 0.2106,
"step": 475
},
{
"epoch": 0.9315068493150684,
"grad_norm": 2.44754781105346,
"learning_rate": 5.536906733320816e-06,
"loss": 0.1977,
"step": 476
},
{
"epoch": 0.9334637964774951,
"grad_norm": 2.720718018235558,
"learning_rate": 5.5216232634006695e-06,
"loss": 0.2127,
"step": 477
},
{
"epoch": 0.9354207436399217,
"grad_norm": 2.3186341405245767,
"learning_rate": 5.506334864528808e-06,
"loss": 0.1702,
"step": 478
},
{
"epoch": 0.9373776908023483,
"grad_norm": 2.120993296789949,
"learning_rate": 5.491041681169226e-06,
"loss": 0.1885,
"step": 479
},
{
"epoch": 0.9393346379647749,
"grad_norm": 2.151691597269135,
"learning_rate": 5.4757438578311275e-06,
"loss": 0.2019,
"step": 480
},
{
"epoch": 0.9412915851272016,
"grad_norm": 2.3001760321495555,
"learning_rate": 5.460441539067559e-06,
"loss": 0.2007,
"step": 481
},
{
"epoch": 0.9432485322896281,
"grad_norm": 2.047203799197217,
"learning_rate": 5.4451348694740495e-06,
"loss": 0.1669,
"step": 482
},
{
"epoch": 0.9452054794520548,
"grad_norm": 2.1777995685000997,
"learning_rate": 5.429823993687234e-06,
"loss": 0.1768,
"step": 483
},
{
"epoch": 0.9471624266144814,
"grad_norm": 2.3015372444552438,
"learning_rate": 5.414509056383498e-06,
"loss": 0.1855,
"step": 484
},
{
"epoch": 0.949119373776908,
"grad_norm": 2.1104267668286134,
"learning_rate": 5.399190202277602e-06,
"loss": 0.1881,
"step": 485
},
{
"epoch": 0.9510763209393346,
"grad_norm": 2.6686581513854652,
"learning_rate": 5.383867576121324e-06,
"loss": 0.2014,
"step": 486
},
{
"epoch": 0.9530332681017613,
"grad_norm": 1.9621455477131546,
"learning_rate": 5.368541322702074e-06,
"loss": 0.1773,
"step": 487
},
{
"epoch": 0.9549902152641878,
"grad_norm": 2.4455326885270376,
"learning_rate": 5.353211586841547e-06,
"loss": 0.184,
"step": 488
},
{
"epoch": 0.9569471624266145,
"grad_norm": 2.2250938934280087,
"learning_rate": 5.337878513394337e-06,
"loss": 0.172,
"step": 489
},
{
"epoch": 0.958904109589041,
"grad_norm": 2.4311155979257957,
"learning_rate": 5.322542247246583e-06,
"loss": 0.1935,
"step": 490
},
{
"epoch": 0.9608610567514677,
"grad_norm": 2.072682084849118,
"learning_rate": 5.307202933314586e-06,
"loss": 0.1575,
"step": 491
},
{
"epoch": 0.9628180039138943,
"grad_norm": 2.0063370920638124,
"learning_rate": 5.29186071654345e-06,
"loss": 0.1973,
"step": 492
},
{
"epoch": 0.9647749510763209,
"grad_norm": 1.9638393601170117,
"learning_rate": 5.276515741905708e-06,
"loss": 0.1659,
"step": 493
},
{
"epoch": 0.9667318982387475,
"grad_norm": 2.37792580273935,
"learning_rate": 5.261168154399953e-06,
"loss": 0.2104,
"step": 494
},
{
"epoch": 0.9686888454011742,
"grad_norm": 2.2859496040978913,
"learning_rate": 5.245818099049467e-06,
"loss": 0.1724,
"step": 495
},
{
"epoch": 0.9706457925636007,
"grad_norm": 1.9929497963870733,
"learning_rate": 5.23046572090085e-06,
"loss": 0.1866,
"step": 496
},
{
"epoch": 0.9726027397260274,
"grad_norm": 2.5725173955202902,
"learning_rate": 5.215111165022653e-06,
"loss": 0.2702,
"step": 497
},
{
"epoch": 0.974559686888454,
"grad_norm": 1.900349030846749,
"learning_rate": 5.199754576504006e-06,
"loss": 0.1553,
"step": 498
},
{
"epoch": 0.9765166340508806,
"grad_norm": 2.009465322944539,
"learning_rate": 5.184396100453242e-06,
"loss": 0.1979,
"step": 499
},
{
"epoch": 0.9784735812133072,
"grad_norm": 2.510565501697165,
"learning_rate": 5.1690358819965334e-06,
"loss": 0.1794,
"step": 500
},
{
"epoch": 0.9784735812133072,
"eval_loss": 0.1730533093214035,
"eval_runtime": 1.7673,
"eval_samples_per_second": 23.765,
"eval_steps_per_second": 6.224,
"step": 500
},
{
"epoch": 0.9804305283757339,
"grad_norm": 1.9913594200136842,
"learning_rate": 5.153674066276513e-06,
"loss": 0.1724,
"step": 501
},
{
"epoch": 0.9823874755381604,
"grad_norm": 2.092031396013016,
"learning_rate": 5.138310798450912e-06,
"loss": 0.1524,
"step": 502
},
{
"epoch": 0.9843444227005871,
"grad_norm": 2.4719622291429366,
"learning_rate": 5.122946223691177e-06,
"loss": 0.26,
"step": 503
},
{
"epoch": 0.9863013698630136,
"grad_norm": 1.9265073115138192,
"learning_rate": 5.107580487181112e-06,
"loss": 0.1427,
"step": 504
},
{
"epoch": 0.9882583170254403,
"grad_norm": 2.293444490699572,
"learning_rate": 5.092213734115489e-06,
"loss": 0.1742,
"step": 505
},
{
"epoch": 0.9902152641878669,
"grad_norm": 1.998119128505986,
"learning_rate": 5.0768461096986935e-06,
"loss": 0.1824,
"step": 506
},
{
"epoch": 0.9921722113502935,
"grad_norm": 2.4031471707620216,
"learning_rate": 5.0614777591433375e-06,
"loss": 0.2025,
"step": 507
},
{
"epoch": 0.9941291585127201,
"grad_norm": 2.2343620512025697,
"learning_rate": 5.0461088276689026e-06,
"loss": 0.1577,
"step": 508
},
{
"epoch": 0.9960861056751468,
"grad_norm": 1.6683261552912807,
"learning_rate": 5.0307394605003554e-06,
"loss": 0.1689,
"step": 509
},
{
"epoch": 0.9980430528375733,
"grad_norm": 1.460659346326579,
"learning_rate": 5.01536980286678e-06,
"loss": 0.1062,
"step": 510
},
{
"epoch": 1.0,
"grad_norm": 1.6860617130527176,
"learning_rate": 5e-06,
"loss": 0.1376,
"step": 511
},
{
"epoch": 1.0019569471624266,
"grad_norm": 1.4364604535711802,
"learning_rate": 4.984630197133224e-06,
"loss": 0.0943,
"step": 512
},
{
"epoch": 1.0039138943248531,
"grad_norm": 1.5725953632244196,
"learning_rate": 4.969260539499645e-06,
"loss": 0.0807,
"step": 513
},
{
"epoch": 1.00587084148728,
"grad_norm": 1.4497646797478543,
"learning_rate": 4.953891172331098e-06,
"loss": 0.0819,
"step": 514
},
{
"epoch": 1.0078277886497065,
"grad_norm": 1.6849830170920463,
"learning_rate": 4.938522240856664e-06,
"loss": 0.0927,
"step": 515
},
{
"epoch": 1.009784735812133,
"grad_norm": 1.5300862825979846,
"learning_rate": 4.92315389030131e-06,
"loss": 0.1013,
"step": 516
},
{
"epoch": 1.0117416829745598,
"grad_norm": 1.406057897590303,
"learning_rate": 4.907786265884511e-06,
"loss": 0.0828,
"step": 517
},
{
"epoch": 1.0136986301369864,
"grad_norm": 1.5621100292746029,
"learning_rate": 4.89241951281889e-06,
"loss": 0.0741,
"step": 518
},
{
"epoch": 1.015655577299413,
"grad_norm": 1.4601764822523038,
"learning_rate": 4.8770537763088234e-06,
"loss": 0.0857,
"step": 519
},
{
"epoch": 1.0176125244618395,
"grad_norm": 1.4678940809229446,
"learning_rate": 4.8616892015490905e-06,
"loss": 0.0976,
"step": 520
},
{
"epoch": 1.019569471624266,
"grad_norm": 1.6757635833037305,
"learning_rate": 4.846325933723487e-06,
"loss": 0.0754,
"step": 521
},
{
"epoch": 1.0215264187866928,
"grad_norm": 1.6595726843007137,
"learning_rate": 4.830964118003468e-06,
"loss": 0.0826,
"step": 522
},
{
"epoch": 1.0234833659491194,
"grad_norm": 1.928983509127001,
"learning_rate": 4.8156038995467606e-06,
"loss": 0.1359,
"step": 523
},
{
"epoch": 1.025440313111546,
"grad_norm": 1.5732507111682255,
"learning_rate": 4.800245423495997e-06,
"loss": 0.0747,
"step": 524
},
{
"epoch": 1.0273972602739727,
"grad_norm": 1.383530254503628,
"learning_rate": 4.784888834977347e-06,
"loss": 0.076,
"step": 525
},
{
"epoch": 1.0293542074363993,
"grad_norm": 1.899212911308038,
"learning_rate": 4.769534279099151e-06,
"loss": 0.0884,
"step": 526
},
{
"epoch": 1.0313111545988258,
"grad_norm": 3.569009618772651,
"learning_rate": 4.7541819009505354e-06,
"loss": 0.0917,
"step": 527
},
{
"epoch": 1.0332681017612524,
"grad_norm": 1.6455734820411143,
"learning_rate": 4.73883184560005e-06,
"loss": 0.0812,
"step": 528
},
{
"epoch": 1.035225048923679,
"grad_norm": 2.483396518537711,
"learning_rate": 4.723484258094293e-06,
"loss": 0.099,
"step": 529
},
{
"epoch": 1.0371819960861057,
"grad_norm": 1.8626635991158047,
"learning_rate": 4.708139283456551e-06,
"loss": 0.0744,
"step": 530
},
{
"epoch": 1.0391389432485323,
"grad_norm": 1.6382008334734672,
"learning_rate": 4.692797066685415e-06,
"loss": 0.0557,
"step": 531
},
{
"epoch": 1.0410958904109588,
"grad_norm": 2.3189584109701107,
"learning_rate": 4.6774577527534195e-06,
"loss": 0.0822,
"step": 532
},
{
"epoch": 1.0430528375733856,
"grad_norm": 1.9716766753796242,
"learning_rate": 4.662121486605663e-06,
"loss": 0.0679,
"step": 533
},
{
"epoch": 1.0450097847358122,
"grad_norm": 2.4490741026154295,
"learning_rate": 4.646788413158455e-06,
"loss": 0.067,
"step": 534
},
{
"epoch": 1.0469667318982387,
"grad_norm": 2.1309358698955267,
"learning_rate": 4.631458677297927e-06,
"loss": 0.0766,
"step": 535
},
{
"epoch": 1.0489236790606653,
"grad_norm": 1.8030435026312444,
"learning_rate": 4.616132423878679e-06,
"loss": 0.056,
"step": 536
},
{
"epoch": 1.0508806262230919,
"grad_norm": 1.380389372236134,
"learning_rate": 4.600809797722398e-06,
"loss": 0.0576,
"step": 537
},
{
"epoch": 1.0528375733855186,
"grad_norm": 2.068579626120395,
"learning_rate": 4.585490943616504e-06,
"loss": 0.0669,
"step": 538
},
{
"epoch": 1.0547945205479452,
"grad_norm": 1.8160828277729508,
"learning_rate": 4.570176006312769e-06,
"loss": 0.0809,
"step": 539
},
{
"epoch": 1.0567514677103718,
"grad_norm": 2.181532539041997,
"learning_rate": 4.554865130525953e-06,
"loss": 0.0794,
"step": 540
},
{
"epoch": 1.0587084148727985,
"grad_norm": 2.481238669867583,
"learning_rate": 4.539558460932442e-06,
"loss": 0.0984,
"step": 541
},
{
"epoch": 1.060665362035225,
"grad_norm": 2.2043986315536244,
"learning_rate": 4.524256142168874e-06,
"loss": 0.102,
"step": 542
},
{
"epoch": 1.0626223091976517,
"grad_norm": 2.2712373413405222,
"learning_rate": 4.508958318830776e-06,
"loss": 0.1251,
"step": 543
},
{
"epoch": 1.0645792563600782,
"grad_norm": 2.207718955938066,
"learning_rate": 4.493665135471194e-06,
"loss": 0.1086,
"step": 544
},
{
"epoch": 1.0665362035225048,
"grad_norm": 1.8486169243660096,
"learning_rate": 4.478376736599331e-06,
"loss": 0.0781,
"step": 545
},
{
"epoch": 1.0684931506849316,
"grad_norm": 2.006416167891028,
"learning_rate": 4.463093266679185e-06,
"loss": 0.0743,
"step": 546
},
{
"epoch": 1.0704500978473581,
"grad_norm": 1.889566490327463,
"learning_rate": 4.447814870128172e-06,
"loss": 0.079,
"step": 547
},
{
"epoch": 1.0724070450097847,
"grad_norm": 2.0918219113878824,
"learning_rate": 4.4325416913157706e-06,
"loss": 0.0829,
"step": 548
},
{
"epoch": 1.0743639921722115,
"grad_norm": 2.100853326565339,
"learning_rate": 4.417273874562154e-06,
"loss": 0.0701,
"step": 549
},
{
"epoch": 1.076320939334638,
"grad_norm": 1.9967564516049237,
"learning_rate": 4.402011564136835e-06,
"loss": 0.0973,
"step": 550
},
{
"epoch": 1.0782778864970646,
"grad_norm": 2.2959094406493703,
"learning_rate": 4.386754904257291e-06,
"loss": 0.0744,
"step": 551
},
{
"epoch": 1.0802348336594911,
"grad_norm": 2.6420602167568177,
"learning_rate": 4.371504039087602e-06,
"loss": 0.1216,
"step": 552
},
{
"epoch": 1.0821917808219177,
"grad_norm": 2.109242856524176,
"learning_rate": 4.356259112737096e-06,
"loss": 0.0844,
"step": 553
},
{
"epoch": 1.0841487279843445,
"grad_norm": 2.384937837425204,
"learning_rate": 4.341020269258987e-06,
"loss": 0.0666,
"step": 554
},
{
"epoch": 1.086105675146771,
"grad_norm": 2.157971248151676,
"learning_rate": 4.325787652649006e-06,
"loss": 0.0695,
"step": 555
},
{
"epoch": 1.0880626223091976,
"grad_norm": 1.8104462500826197,
"learning_rate": 4.310561406844045e-06,
"loss": 0.0855,
"step": 556
},
{
"epoch": 1.0900195694716244,
"grad_norm": 1.570124801382527,
"learning_rate": 4.295341675720797e-06,
"loss": 0.0667,
"step": 557
},
{
"epoch": 1.091976516634051,
"grad_norm": 1.6419424872581734,
"learning_rate": 4.280128603094399e-06,
"loss": 0.0673,
"step": 558
},
{
"epoch": 1.0939334637964775,
"grad_norm": 1.7049297271694,
"learning_rate": 4.264922332717066e-06,
"loss": 0.0947,
"step": 559
},
{
"epoch": 1.095890410958904,
"grad_norm": 1.429444451255871,
"learning_rate": 4.249723008276737e-06,
"loss": 0.0464,
"step": 560
},
{
"epoch": 1.0978473581213306,
"grad_norm": 1.875457161300417,
"learning_rate": 4.234530773395721e-06,
"loss": 0.0797,
"step": 561
},
{
"epoch": 1.0998043052837574,
"grad_norm": 2.4251321421083034,
"learning_rate": 4.219345771629333e-06,
"loss": 0.0768,
"step": 562
},
{
"epoch": 1.101761252446184,
"grad_norm": 1.4484743583801212,
"learning_rate": 4.20416814646454e-06,
"loss": 0.0538,
"step": 563
},
{
"epoch": 1.1037181996086105,
"grad_norm": 1.6140116924282306,
"learning_rate": 4.188998041318608e-06,
"loss": 0.0618,
"step": 564
},
{
"epoch": 1.1056751467710373,
"grad_norm": 1.491637132489362,
"learning_rate": 4.173835599537741e-06,
"loss": 0.0602,
"step": 565
},
{
"epoch": 1.1076320939334638,
"grad_norm": 2.258769844171861,
"learning_rate": 4.158680964395734e-06,
"loss": 0.105,
"step": 566
},
{
"epoch": 1.1095890410958904,
"grad_norm": 2.353160174308116,
"learning_rate": 4.143534279092613e-06,
"loss": 0.1023,
"step": 567
},
{
"epoch": 1.111545988258317,
"grad_norm": 1.6529832773429178,
"learning_rate": 4.1283956867532825e-06,
"loss": 0.0629,
"step": 568
},
{
"epoch": 1.1135029354207435,
"grad_norm": 1.6040135598410907,
"learning_rate": 4.113265330426175e-06,
"loss": 0.0671,
"step": 569
},
{
"epoch": 1.1154598825831703,
"grad_norm": 1.7624997243953948,
"learning_rate": 4.098143353081902e-06,
"loss": 0.0617,
"step": 570
},
{
"epoch": 1.1174168297455969,
"grad_norm": 1.665207214554899,
"learning_rate": 4.083029897611897e-06,
"loss": 0.0606,
"step": 571
},
{
"epoch": 1.1193737769080234,
"grad_norm": 2.0723758505494008,
"learning_rate": 4.067925106827068e-06,
"loss": 0.0885,
"step": 572
},
{
"epoch": 1.1213307240704502,
"grad_norm": 1.695932395186667,
"learning_rate": 4.052829123456448e-06,
"loss": 0.0733,
"step": 573
},
{
"epoch": 1.1232876712328768,
"grad_norm": 1.9748916137627088,
"learning_rate": 4.037742090145851e-06,
"loss": 0.0931,
"step": 574
},
{
"epoch": 1.1252446183953033,
"grad_norm": 2.3317215321689226,
"learning_rate": 4.0226641494565125e-06,
"loss": 0.1095,
"step": 575
},
{
"epoch": 1.1272015655577299,
"grad_norm": 1.4616848832641363,
"learning_rate": 4.007595443863755e-06,
"loss": 0.0677,
"step": 576
},
{
"epoch": 1.1291585127201564,
"grad_norm": 1.6330963920798616,
"learning_rate": 3.992536115755631e-06,
"loss": 0.0703,
"step": 577
},
{
"epoch": 1.1311154598825832,
"grad_norm": 1.692042486424253,
"learning_rate": 3.977486307431589e-06,
"loss": 0.0451,
"step": 578
},
{
"epoch": 1.1330724070450098,
"grad_norm": 1.6093687911763785,
"learning_rate": 3.962446161101119e-06,
"loss": 0.0694,
"step": 579
},
{
"epoch": 1.1350293542074363,
"grad_norm": 1.5833758240377953,
"learning_rate": 3.9474158188824145e-06,
"loss": 0.0834,
"step": 580
},
{
"epoch": 1.1369863013698631,
"grad_norm": 1.4774406501732416,
"learning_rate": 3.93239542280102e-06,
"loss": 0.0609,
"step": 581
},
{
"epoch": 1.1389432485322897,
"grad_norm": 1.7667781026660516,
"learning_rate": 3.917385114788508e-06,
"loss": 0.0631,
"step": 582
},
{
"epoch": 1.1409001956947162,
"grad_norm": 1.5212996511473915,
"learning_rate": 3.902385036681118e-06,
"loss": 0.0592,
"step": 583
},
{
"epoch": 1.1428571428571428,
"grad_norm": 1.8205593753047988,
"learning_rate": 3.887395330218429e-06,
"loss": 0.0704,
"step": 584
},
{
"epoch": 1.1448140900195694,
"grad_norm": 2.178134092464723,
"learning_rate": 3.87241613704201e-06,
"loss": 0.0733,
"step": 585
},
{
"epoch": 1.1467710371819961,
"grad_norm": 2.054050330693599,
"learning_rate": 3.857447598694095e-06,
"loss": 0.0696,
"step": 586
},
{
"epoch": 1.1487279843444227,
"grad_norm": 2.5812369887030413,
"learning_rate": 3.842489856616231e-06,
"loss": 0.11,
"step": 587
},
{
"epoch": 1.1506849315068493,
"grad_norm": 2.6523077137659854,
"learning_rate": 3.827543052147952e-06,
"loss": 0.1018,
"step": 588
},
{
"epoch": 1.152641878669276,
"grad_norm": 1.8132608675165665,
"learning_rate": 3.8126073265254377e-06,
"loss": 0.0578,
"step": 589
},
{
"epoch": 1.1545988258317026,
"grad_norm": 1.8873048557902383,
"learning_rate": 3.797682820880184e-06,
"loss": 0.0773,
"step": 590
},
{
"epoch": 1.1565557729941291,
"grad_norm": 1.7843790872280314,
"learning_rate": 3.782769676237663e-06,
"loss": 0.0695,
"step": 591
},
{
"epoch": 1.1585127201565557,
"grad_norm": 1.4344205996994106,
"learning_rate": 3.7678680335159955e-06,
"loss": 0.0754,
"step": 592
},
{
"epoch": 1.1604696673189823,
"grad_norm": 2.128916228668013,
"learning_rate": 3.7529780335246164e-06,
"loss": 0.0912,
"step": 593
},
{
"epoch": 1.162426614481409,
"grad_norm": 1.8648494638612307,
"learning_rate": 3.7380998169629477e-06,
"loss": 0.0756,
"step": 594
},
{
"epoch": 1.1643835616438356,
"grad_norm": 2.194430280271575,
"learning_rate": 3.7232335244190656e-06,
"loss": 0.0945,
"step": 595
},
{
"epoch": 1.1663405088062622,
"grad_norm": 1.7225857798012905,
"learning_rate": 3.708379296368372e-06,
"loss": 0.0614,
"step": 596
},
{
"epoch": 1.168297455968689,
"grad_norm": 1.8868709347648658,
"learning_rate": 3.6935372731722686e-06,
"loss": 0.0709,
"step": 597
},
{
"epoch": 1.1702544031311155,
"grad_norm": 1.8641881344679696,
"learning_rate": 3.6787075950768337e-06,
"loss": 0.0966,
"step": 598
},
{
"epoch": 1.172211350293542,
"grad_norm": 2.1874579509990886,
"learning_rate": 3.663890402211489e-06,
"loss": 0.0848,
"step": 599
},
{
"epoch": 1.1741682974559686,
"grad_norm": 2.175980645817372,
"learning_rate": 3.649085834587683e-06,
"loss": 0.1247,
"step": 600
},
{
"epoch": 1.1761252446183952,
"grad_norm": 2.273811598238986,
"learning_rate": 3.6342940320975613e-06,
"loss": 0.0952,
"step": 601
},
{
"epoch": 1.178082191780822,
"grad_norm": 1.5514710678019257,
"learning_rate": 3.6195151345126556e-06,
"loss": 0.059,
"step": 602
},
{
"epoch": 1.1800391389432485,
"grad_norm": 1.9792612346178367,
"learning_rate": 3.6047492814825508e-06,
"loss": 0.0806,
"step": 603
},
{
"epoch": 1.181996086105675,
"grad_norm": 2.116168997974382,
"learning_rate": 3.5899966125335684e-06,
"loss": 0.0887,
"step": 604
},
{
"epoch": 1.1839530332681019,
"grad_norm": 1.5968956529678024,
"learning_rate": 3.575257267067453e-06,
"loss": 0.0681,
"step": 605
},
{
"epoch": 1.1859099804305284,
"grad_norm": 1.5628253080473218,
"learning_rate": 3.5605313843600555e-06,
"loss": 0.0786,
"step": 606
},
{
"epoch": 1.187866927592955,
"grad_norm": 1.8502056594609926,
"learning_rate": 3.5458191035600077e-06,
"loss": 0.0713,
"step": 607
},
{
"epoch": 1.1898238747553815,
"grad_norm": 1.7829714144983377,
"learning_rate": 3.531120563687419e-06,
"loss": 0.0791,
"step": 608
},
{
"epoch": 1.191780821917808,
"grad_norm": 2.0630863633398775,
"learning_rate": 3.5164359036325483e-06,
"loss": 0.082,
"step": 609
},
{
"epoch": 1.1937377690802349,
"grad_norm": 1.6702377825969237,
"learning_rate": 3.5017652621545133e-06,
"loss": 0.0761,
"step": 610
},
{
"epoch": 1.1956947162426614,
"grad_norm": 1.9955178647736718,
"learning_rate": 3.487108777879957e-06,
"loss": 0.0574,
"step": 611
},
{
"epoch": 1.197651663405088,
"grad_norm": 1.6405584478809219,
"learning_rate": 3.4724665893017517e-06,
"loss": 0.068,
"step": 612
},
{
"epoch": 1.1996086105675148,
"grad_norm": 1.4159797202771183,
"learning_rate": 3.457838834777682e-06,
"loss": 0.0539,
"step": 613
},
{
"epoch": 1.2015655577299413,
"grad_norm": 1.8425970070227418,
"learning_rate": 3.4432256525291468e-06,
"loss": 0.0494,
"step": 614
},
{
"epoch": 1.203522504892368,
"grad_norm": 2.1511155581331285,
"learning_rate": 3.428627180639844e-06,
"loss": 0.0743,
"step": 615
},
{
"epoch": 1.2054794520547945,
"grad_norm": 2.96920797708028,
"learning_rate": 3.4140435570544708e-06,
"loss": 0.0979,
"step": 616
},
{
"epoch": 1.207436399217221,
"grad_norm": 2.231144073150809,
"learning_rate": 3.3994749195774173e-06,
"loss": 0.0754,
"step": 617
},
{
"epoch": 1.2093933463796478,
"grad_norm": 1.678087237123076,
"learning_rate": 3.3849214058714707e-06,
"loss": 0.0541,
"step": 618
},
{
"epoch": 1.2113502935420744,
"grad_norm": 2.1255313109001728,
"learning_rate": 3.370383153456504e-06,
"loss": 0.0907,
"step": 619
},
{
"epoch": 1.213307240704501,
"grad_norm": 1.8026174387405463,
"learning_rate": 3.355860299708187e-06,
"loss": 0.091,
"step": 620
},
{
"epoch": 1.2152641878669277,
"grad_norm": 1.6620341835571215,
"learning_rate": 3.341352981856678e-06,
"loss": 0.067,
"step": 621
},
{
"epoch": 1.2172211350293543,
"grad_norm": 1.7928164787526168,
"learning_rate": 3.326861336985341e-06,
"loss": 0.0585,
"step": 622
},
{
"epoch": 1.2191780821917808,
"grad_norm": 1.9225348551493409,
"learning_rate": 3.3123855020294344e-06,
"loss": 0.0619,
"step": 623
},
{
"epoch": 1.2211350293542074,
"grad_norm": 1.7052958212898235,
"learning_rate": 3.2979256137748283e-06,
"loss": 0.0608,
"step": 624
},
{
"epoch": 1.223091976516634,
"grad_norm": 2.1053259750518487,
"learning_rate": 3.283481808856706e-06,
"loss": 0.076,
"step": 625
},
{
"epoch": 1.2250489236790607,
"grad_norm": 2.138631150625641,
"learning_rate": 3.269054223758279e-06,
"loss": 0.0628,
"step": 626
},
{
"epoch": 1.2270058708414873,
"grad_norm": 1.9492993526884388,
"learning_rate": 3.2546429948094904e-06,
"loss": 0.0812,
"step": 627
},
{
"epoch": 1.2289628180039138,
"grad_norm": 2.206847852850266,
"learning_rate": 3.240248258185731e-06,
"loss": 0.0756,
"step": 628
},
{
"epoch": 1.2309197651663406,
"grad_norm": 1.9337659120686344,
"learning_rate": 3.225870149906549e-06,
"loss": 0.0808,
"step": 629
},
{
"epoch": 1.2328767123287672,
"grad_norm": 2.0132409152376307,
"learning_rate": 3.2115088058343725e-06,
"loss": 0.0702,
"step": 630
},
{
"epoch": 1.2348336594911937,
"grad_norm": 1.9874258173187966,
"learning_rate": 3.1971643616732174e-06,
"loss": 0.0684,
"step": 631
},
{
"epoch": 1.2367906066536203,
"grad_norm": 2.1824915965699567,
"learning_rate": 3.182836952967405e-06,
"loss": 0.0824,
"step": 632
},
{
"epoch": 1.2387475538160468,
"grad_norm": 1.3990548059614947,
"learning_rate": 3.1685267151002856e-06,
"loss": 0.037,
"step": 633
},
{
"epoch": 1.2407045009784736,
"grad_norm": 1.7547699264870562,
"learning_rate": 3.154233783292964e-06,
"loss": 0.0726,
"step": 634
},
{
"epoch": 1.2426614481409002,
"grad_norm": 2.0574421319620315,
"learning_rate": 3.139958292603009e-06,
"loss": 0.0621,
"step": 635
},
{
"epoch": 1.2446183953033267,
"grad_norm": 2.03808288504389,
"learning_rate": 3.125700377923186e-06,
"loss": 0.0649,
"step": 636
},
{
"epoch": 1.2465753424657535,
"grad_norm": 1.664664975582277,
"learning_rate": 3.111460173980175e-06,
"loss": 0.062,
"step": 637
},
{
"epoch": 1.24853228962818,
"grad_norm": 1.9350624481077692,
"learning_rate": 3.097237815333315e-06,
"loss": 0.0865,
"step": 638
},
{
"epoch": 1.2504892367906066,
"grad_norm": 2.161111111656984,
"learning_rate": 3.0830334363733083e-06,
"loss": 0.0642,
"step": 639
},
{
"epoch": 1.2524461839530332,
"grad_norm": 1.7014923942244589,
"learning_rate": 3.068847171320969e-06,
"loss": 0.0751,
"step": 640
},
{
"epoch": 1.2544031311154598,
"grad_norm": 1.8034868786489808,
"learning_rate": 3.054679154225943e-06,
"loss": 0.0804,
"step": 641
},
{
"epoch": 1.2563600782778865,
"grad_norm": 2.0184120611736787,
"learning_rate": 3.0405295189654537e-06,
"loss": 0.0674,
"step": 642
},
{
"epoch": 1.258317025440313,
"grad_norm": 1.616484343181486,
"learning_rate": 3.0263983992430245e-06,
"loss": 0.0707,
"step": 643
},
{
"epoch": 1.2602739726027397,
"grad_norm": 1.9054811123440305,
"learning_rate": 3.0122859285872214e-06,
"loss": 0.0448,
"step": 644
},
{
"epoch": 1.2622309197651664,
"grad_norm": 1.9502074583748314,
"learning_rate": 2.9981922403503914e-06,
"loss": 0.104,
"step": 645
},
{
"epoch": 1.264187866927593,
"grad_norm": 1.3929361049229838,
"learning_rate": 2.9841174677074037e-06,
"loss": 0.0499,
"step": 646
},
{
"epoch": 1.2661448140900196,
"grad_norm": 2.015090065912094,
"learning_rate": 2.9700617436543854e-06,
"loss": 0.0736,
"step": 647
},
{
"epoch": 1.2681017612524461,
"grad_norm": 2.283355451205927,
"learning_rate": 2.956025201007471e-06,
"loss": 0.0943,
"step": 648
},
{
"epoch": 1.2700587084148727,
"grad_norm": 1.5205045455930266,
"learning_rate": 2.942007972401542e-06,
"loss": 0.0391,
"step": 649
},
{
"epoch": 1.2720156555772995,
"grad_norm": 2.076497401164596,
"learning_rate": 2.9280101902889823e-06,
"loss": 0.0958,
"step": 650
},
{
"epoch": 1.273972602739726,
"grad_norm": 1.8724323612868614,
"learning_rate": 2.914031986938417e-06,
"loss": 0.0456,
"step": 651
},
{
"epoch": 1.2759295499021526,
"grad_norm": 2.1900139731732544,
"learning_rate": 2.900073494433464e-06,
"loss": 0.0647,
"step": 652
},
{
"epoch": 1.2778864970645794,
"grad_norm": 1.7668764564888597,
"learning_rate": 2.8861348446714942e-06,
"loss": 0.0772,
"step": 653
},
{
"epoch": 1.279843444227006,
"grad_norm": 2.055426213301513,
"learning_rate": 2.8722161693623775e-06,
"loss": 0.0598,
"step": 654
},
{
"epoch": 1.2818003913894325,
"grad_norm": 1.7950726294135408,
"learning_rate": 2.858317600027235e-06,
"loss": 0.0585,
"step": 655
},
{
"epoch": 1.283757338551859,
"grad_norm": 1.6907221551089933,
"learning_rate": 2.8444392679972107e-06,
"loss": 0.082,
"step": 656
},
{
"epoch": 1.2857142857142856,
"grad_norm": 1.94067562510593,
"learning_rate": 2.83058130441221e-06,
"loss": 0.0904,
"step": 657
},
{
"epoch": 1.2876712328767124,
"grad_norm": 1.9445258745628324,
"learning_rate": 2.816743840219681e-06,
"loss": 0.0607,
"step": 658
},
{
"epoch": 1.289628180039139,
"grad_norm": 2.201875980183344,
"learning_rate": 2.802927006173365e-06,
"loss": 0.0906,
"step": 659
},
{
"epoch": 1.2915851272015655,
"grad_norm": 2.0052366821231127,
"learning_rate": 2.789130932832059e-06,
"loss": 0.0556,
"step": 660
},
{
"epoch": 1.2935420743639923,
"grad_norm": 2.399732508582058,
"learning_rate": 2.7753557505583945e-06,
"loss": 0.0893,
"step": 661
},
{
"epoch": 1.2954990215264188,
"grad_norm": 1.577728985135844,
"learning_rate": 2.7616015895175953e-06,
"loss": 0.054,
"step": 662
},
{
"epoch": 1.2974559686888454,
"grad_norm": 1.8668715387815782,
"learning_rate": 2.747868579676244e-06,
"loss": 0.08,
"step": 663
},
{
"epoch": 1.299412915851272,
"grad_norm": 2.0460778407301263,
"learning_rate": 2.734156850801071e-06,
"loss": 0.07,
"step": 664
},
{
"epoch": 1.3013698630136985,
"grad_norm": 1.6605738012339846,
"learning_rate": 2.720466532457707e-06,
"loss": 0.0658,
"step": 665
},
{
"epoch": 1.3033268101761253,
"grad_norm": 2.1548936437704387,
"learning_rate": 2.706797754009476e-06,
"loss": 0.0834,
"step": 666
},
{
"epoch": 1.3052837573385518,
"grad_norm": 2.177951932952928,
"learning_rate": 2.693150644616165e-06,
"loss": 0.0869,
"step": 667
},
{
"epoch": 1.3072407045009784,
"grad_norm": 3.0852971035539807,
"learning_rate": 2.6795253332327996e-06,
"loss": 0.0774,
"step": 668
},
{
"epoch": 1.3091976516634052,
"grad_norm": 1.621723168698709,
"learning_rate": 2.6659219486084366e-06,
"loss": 0.058,
"step": 669
},
{
"epoch": 1.3111545988258317,
"grad_norm": 1.8052594279486867,
"learning_rate": 2.652340619284939e-06,
"loss": 0.0744,
"step": 670
},
{
"epoch": 1.3131115459882583,
"grad_norm": 1.674212030789698,
"learning_rate": 2.638781473595765e-06,
"loss": 0.0572,
"step": 671
},
{
"epoch": 1.3150684931506849,
"grad_norm": 2.043418047769638,
"learning_rate": 2.6252446396647503e-06,
"loss": 0.1077,
"step": 672
},
{
"epoch": 1.3170254403131114,
"grad_norm": 1.8877359312902404,
"learning_rate": 2.6117302454049e-06,
"loss": 0.058,
"step": 673
},
{
"epoch": 1.3189823874755382,
"grad_norm": 1.863879266501068,
"learning_rate": 2.5982384185171905e-06,
"loss": 0.0718,
"step": 674
},
{
"epoch": 1.3209393346379648,
"grad_norm": 1.8105314172937954,
"learning_rate": 2.5847692864893423e-06,
"loss": 0.073,
"step": 675
},
{
"epoch": 1.3228962818003913,
"grad_norm": 1.9600381215931317,
"learning_rate": 2.5713229765946355e-06,
"loss": 0.0706,
"step": 676
},
{
"epoch": 1.324853228962818,
"grad_norm": 2.120293640627276,
"learning_rate": 2.5578996158906887e-06,
"loss": 0.1015,
"step": 677
},
{
"epoch": 1.3268101761252447,
"grad_norm": 1.5896445265701151,
"learning_rate": 2.544499331218274e-06,
"loss": 0.0677,
"step": 678
},
{
"epoch": 1.3287671232876712,
"grad_norm": 1.8591938043813965,
"learning_rate": 2.531122249200114e-06,
"loss": 0.0644,
"step": 679
},
{
"epoch": 1.3307240704500978,
"grad_norm": 2.0210099379232416,
"learning_rate": 2.5177684962396764e-06,
"loss": 0.0813,
"step": 680
},
{
"epoch": 1.3326810176125243,
"grad_norm": 2.202997245026879,
"learning_rate": 2.504438198519992e-06,
"loss": 0.0723,
"step": 681
},
{
"epoch": 1.3346379647749511,
"grad_norm": 2.6249294578090714,
"learning_rate": 2.4911314820024566e-06,
"loss": 0.0722,
"step": 682
},
{
"epoch": 1.3365949119373777,
"grad_norm": 1.9431999055505595,
"learning_rate": 2.477848472425638e-06,
"loss": 0.0951,
"step": 683
},
{
"epoch": 1.3385518590998042,
"grad_norm": 1.9215215707488533,
"learning_rate": 2.464589295304096e-06,
"loss": 0.0875,
"step": 684
},
{
"epoch": 1.340508806262231,
"grad_norm": 1.6960829537877524,
"learning_rate": 2.4513540759271855e-06,
"loss": 0.0539,
"step": 685
},
{
"epoch": 1.3424657534246576,
"grad_norm": 1.6202828640137272,
"learning_rate": 2.438142939357882e-06,
"loss": 0.0512,
"step": 686
},
{
"epoch": 1.3444227005870841,
"grad_norm": 1.5345872398540568,
"learning_rate": 2.4249560104315992e-06,
"loss": 0.0651,
"step": 687
},
{
"epoch": 1.3463796477495107,
"grad_norm": 1.6927372641026215,
"learning_rate": 2.4117934137550003e-06,
"loss": 0.0612,
"step": 688
},
{
"epoch": 1.3483365949119372,
"grad_norm": 1.5871614922314186,
"learning_rate": 2.398655273704831e-06,
"loss": 0.0705,
"step": 689
},
{
"epoch": 1.350293542074364,
"grad_norm": 1.8292578591146453,
"learning_rate": 2.38554171442674e-06,
"loss": 0.0808,
"step": 690
},
{
"epoch": 1.3522504892367906,
"grad_norm": 2.3792236017494077,
"learning_rate": 2.3724528598341033e-06,
"loss": 0.0932,
"step": 691
},
{
"epoch": 1.3542074363992171,
"grad_norm": 1.546699861880311,
"learning_rate": 2.35938883360686e-06,
"loss": 0.0641,
"step": 692
},
{
"epoch": 1.356164383561644,
"grad_norm": 2.0358263165883943,
"learning_rate": 2.346349759190332e-06,
"loss": 0.066,
"step": 693
},
{
"epoch": 1.3581213307240705,
"grad_norm": 1.490842276288226,
"learning_rate": 2.3333357597940794e-06,
"loss": 0.05,
"step": 694
},
{
"epoch": 1.360078277886497,
"grad_norm": 1.791146355805155,
"learning_rate": 2.3203469583907074e-06,
"loss": 0.0705,
"step": 695
},
{
"epoch": 1.3620352250489236,
"grad_norm": 2.2593808299860005,
"learning_rate": 2.3073834777147237e-06,
"loss": 0.0677,
"step": 696
},
{
"epoch": 1.3639921722113502,
"grad_norm": 2.1732183959015954,
"learning_rate": 2.294445440261377e-06,
"loss": 0.0885,
"step": 697
},
{
"epoch": 1.365949119373777,
"grad_norm": 1.5312213730289557,
"learning_rate": 2.2815329682854944e-06,
"loss": 0.0688,
"step": 698
},
{
"epoch": 1.3679060665362035,
"grad_norm": 1.8904929752087385,
"learning_rate": 2.2686461838003294e-06,
"loss": 0.0763,
"step": 699
},
{
"epoch": 1.36986301369863,
"grad_norm": 2.06876254381848,
"learning_rate": 2.2557852085764053e-06,
"loss": 0.0811,
"step": 700
},
{
"epoch": 1.3718199608610568,
"grad_norm": 1.9637438532435185,
"learning_rate": 2.2429501641403657e-06,
"loss": 0.0774,
"step": 701
},
{
"epoch": 1.3737769080234834,
"grad_norm": 1.5903631475108062,
"learning_rate": 2.230141171773836e-06,
"loss": 0.0436,
"step": 702
},
{
"epoch": 1.37573385518591,
"grad_norm": 2.024028265421277,
"learning_rate": 2.2173583525122595e-06,
"loss": 0.0861,
"step": 703
},
{
"epoch": 1.3776908023483365,
"grad_norm": 2.2452926161627826,
"learning_rate": 2.204601827143769e-06,
"loss": 0.0864,
"step": 704
},
{
"epoch": 1.379647749510763,
"grad_norm": 1.774376807897319,
"learning_rate": 2.1918717162080343e-06,
"loss": 0.067,
"step": 705
},
{
"epoch": 1.3816046966731899,
"grad_norm": 2.125239212808248,
"learning_rate": 2.179168139995134e-06,
"loss": 0.0687,
"step": 706
},
{
"epoch": 1.3835616438356164,
"grad_norm": 1.9386214287212566,
"learning_rate": 2.1664912185444127e-06,
"loss": 0.0719,
"step": 707
},
{
"epoch": 1.385518590998043,
"grad_norm": 1.7387292264200185,
"learning_rate": 2.1538410716433415e-06,
"loss": 0.0686,
"step": 708
},
{
"epoch": 1.3874755381604698,
"grad_norm": 1.848864992503259,
"learning_rate": 2.1412178188263975e-06,
"loss": 0.0624,
"step": 709
},
{
"epoch": 1.3894324853228963,
"grad_norm": 1.5697045992324241,
"learning_rate": 2.12862157937393e-06,
"loss": 0.0523,
"step": 710
},
{
"epoch": 1.3913894324853229,
"grad_norm": 1.9824122518786331,
"learning_rate": 2.1160524723110253e-06,
"loss": 0.114,
"step": 711
},
{
"epoch": 1.3933463796477494,
"grad_norm": 2.855313761219161,
"learning_rate": 2.103510616406396e-06,
"loss": 0.1014,
"step": 712
},
{
"epoch": 1.395303326810176,
"grad_norm": 1.7406203751163734,
"learning_rate": 2.090996130171245e-06,
"loss": 0.0549,
"step": 713
},
{
"epoch": 1.3972602739726028,
"grad_norm": 1.71306351635305,
"learning_rate": 2.0785091318581577e-06,
"loss": 0.061,
"step": 714
},
{
"epoch": 1.3992172211350293,
"grad_norm": 1.6706721747054911,
"learning_rate": 2.0660497394599786e-06,
"loss": 0.0863,
"step": 715
},
{
"epoch": 1.401174168297456,
"grad_norm": 1.7914000516782471,
"learning_rate": 2.053618070708691e-06,
"loss": 0.0586,
"step": 716
},
{
"epoch": 1.4031311154598827,
"grad_norm": 2.1371743888191195,
"learning_rate": 2.0412142430743175e-06,
"loss": 0.0616,
"step": 717
},
{
"epoch": 1.4050880626223092,
"grad_norm": 2.075993465348145,
"learning_rate": 2.0288383737638007e-06,
"loss": 0.0693,
"step": 718
},
{
"epoch": 1.4070450097847358,
"grad_norm": 1.2437743177865122,
"learning_rate": 2.0164905797198936e-06,
"loss": 0.0488,
"step": 719
},
{
"epoch": 1.4090019569471623,
"grad_norm": 2.087784278527308,
"learning_rate": 2.004170977620066e-06,
"loss": 0.0484,
"step": 720
},
{
"epoch": 1.410958904109589,
"grad_norm": 1.913903365922053,
"learning_rate": 1.991879683875386e-06,
"loss": 0.0815,
"step": 721
},
{
"epoch": 1.4129158512720157,
"grad_norm": 1.7286447187971836,
"learning_rate": 1.9796168146294415e-06,
"loss": 0.077,
"step": 722
},
{
"epoch": 1.4148727984344422,
"grad_norm": 1.763246064167294,
"learning_rate": 1.9673824857572173e-06,
"loss": 0.0637,
"step": 723
},
{
"epoch": 1.4168297455968688,
"grad_norm": 2.247186262832292,
"learning_rate": 1.9551768128640173e-06,
"loss": 0.0735,
"step": 724
},
{
"epoch": 1.4187866927592956,
"grad_norm": 2.000376581105247,
"learning_rate": 1.94299991128437e-06,
"loss": 0.0858,
"step": 725
},
{
"epoch": 1.4207436399217221,
"grad_norm": 1.749021015554715,
"learning_rate": 1.9308518960809353e-06,
"loss": 0.0777,
"step": 726
},
{
"epoch": 1.4227005870841487,
"grad_norm": 1.7947719557204047,
"learning_rate": 1.9187328820434185e-06,
"loss": 0.0613,
"step": 727
},
{
"epoch": 1.4246575342465753,
"grad_norm": 2.141085404619341,
"learning_rate": 1.9066429836874844e-06,
"loss": 0.07,
"step": 728
},
{
"epoch": 1.4266144814090018,
"grad_norm": 2.23270739670154,
"learning_rate": 1.894582315253673e-06,
"loss": 0.1098,
"step": 729
},
{
"epoch": 1.4285714285714286,
"grad_norm": 1.5198522541763244,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.0554,
"step": 730
},
{
"epoch": 1.4305283757338552,
"grad_norm": 1.8101495775510452,
"learning_rate": 1.8705491237325246e-06,
"loss": 0.0738,
"step": 731
},
{
"epoch": 1.4324853228962817,
"grad_norm": 1.5320855667688547,
"learning_rate": 1.8585768277409639e-06,
"loss": 0.0556,
"step": 732
},
{
"epoch": 1.4344422700587085,
"grad_norm": 1.9703876452576585,
"learning_rate": 1.8466342158609367e-06,
"loss": 0.0764,
"step": 733
},
{
"epoch": 1.436399217221135,
"grad_norm": 1.8602853311790735,
"learning_rate": 1.83472140094124e-06,
"loss": 0.0821,
"step": 734
},
{
"epoch": 1.4383561643835616,
"grad_norm": 1.965044526084634,
"learning_rate": 1.8228384955491136e-06,
"loss": 0.0653,
"step": 735
},
{
"epoch": 1.4403131115459882,
"grad_norm": 2.2333824222337713,
"learning_rate": 1.8109856119691672e-06,
"loss": 0.073,
"step": 736
},
{
"epoch": 1.4422700587084147,
"grad_norm": 2.774306757176038,
"learning_rate": 1.7991628622023338e-06,
"loss": 0.1171,
"step": 737
},
{
"epoch": 1.4442270058708415,
"grad_norm": 1.9679288460755902,
"learning_rate": 1.7873703579648033e-06,
"loss": 0.0781,
"step": 738
},
{
"epoch": 1.446183953033268,
"grad_norm": 1.7538055228481852,
"learning_rate": 1.775608210686962e-06,
"loss": 0.0553,
"step": 739
},
{
"epoch": 1.4481409001956946,
"grad_norm": 1.6322103669371992,
"learning_rate": 1.763876531512354e-06,
"loss": 0.0789,
"step": 740
},
{
"epoch": 1.4500978473581214,
"grad_norm": 2.434369958490434,
"learning_rate": 1.7521754312966155e-06,
"loss": 0.0989,
"step": 741
},
{
"epoch": 1.452054794520548,
"grad_norm": 1.73301653362213,
"learning_rate": 1.7405050206064372e-06,
"loss": 0.0769,
"step": 742
},
{
"epoch": 1.4540117416829745,
"grad_norm": 2.441179651598841,
"learning_rate": 1.7288654097185193e-06,
"loss": 0.0913,
"step": 743
},
{
"epoch": 1.455968688845401,
"grad_norm": 1.8343276866479539,
"learning_rate": 1.7172567086185205e-06,
"loss": 0.0823,
"step": 744
},
{
"epoch": 1.4579256360078277,
"grad_norm": 1.9318527185380996,
"learning_rate": 1.7056790270000302e-06,
"loss": 0.06,
"step": 745
},
{
"epoch": 1.4598825831702544,
"grad_norm": 1.8346825955364,
"learning_rate": 1.6941324742635262e-06,
"loss": 0.0783,
"step": 746
},
{
"epoch": 1.461839530332681,
"grad_norm": 1.715657417837221,
"learning_rate": 1.682617159515338e-06,
"loss": 0.0624,
"step": 747
},
{
"epoch": 1.4637964774951076,
"grad_norm": 1.971842956839842,
"learning_rate": 1.671133191566624e-06,
"loss": 0.1111,
"step": 748
},
{
"epoch": 1.4657534246575343,
"grad_norm": 1.4999361160456088,
"learning_rate": 1.6596806789323317e-06,
"loss": 0.0511,
"step": 749
},
{
"epoch": 1.467710371819961,
"grad_norm": 1.620700043026342,
"learning_rate": 1.6482597298301917e-06,
"loss": 0.0693,
"step": 750
},
{
"epoch": 1.4696673189823874,
"grad_norm": 1.6236918289991626,
"learning_rate": 1.6368704521796702e-06,
"loss": 0.0457,
"step": 751
},
{
"epoch": 1.471624266144814,
"grad_norm": 2.088344992354519,
"learning_rate": 1.625512953600966e-06,
"loss": 0.0733,
"step": 752
},
{
"epoch": 1.4735812133072406,
"grad_norm": 1.468629838665795,
"learning_rate": 1.6141873414139914e-06,
"loss": 0.0433,
"step": 753
},
{
"epoch": 1.4755381604696673,
"grad_norm": 2.012450159214436,
"learning_rate": 1.6028937226373536e-06,
"loss": 0.0686,
"step": 754
},
{
"epoch": 1.477495107632094,
"grad_norm": 2.473842629693045,
"learning_rate": 1.5916322039873494e-06,
"loss": 0.0934,
"step": 755
},
{
"epoch": 1.4794520547945205,
"grad_norm": 1.6619221245291644,
"learning_rate": 1.5804028918769488e-06,
"loss": 0.0505,
"step": 756
},
{
"epoch": 1.4814090019569472,
"grad_norm": 2.6427238058753466,
"learning_rate": 1.5692058924147924e-06,
"loss": 0.1059,
"step": 757
},
{
"epoch": 1.4833659491193738,
"grad_norm": 1.6317849738060646,
"learning_rate": 1.5580413114042003e-06,
"loss": 0.0614,
"step": 758
},
{
"epoch": 1.4853228962818004,
"grad_norm": 1.2690105925041866,
"learning_rate": 1.546909254342151e-06,
"loss": 0.0466,
"step": 759
},
{
"epoch": 1.487279843444227,
"grad_norm": 1.553475043184994,
"learning_rate": 1.535809826418304e-06,
"loss": 0.0677,
"step": 760
},
{
"epoch": 1.4892367906066535,
"grad_norm": 1.5284923085319786,
"learning_rate": 1.524743132513991e-06,
"loss": 0.0593,
"step": 761
},
{
"epoch": 1.4911937377690803,
"grad_norm": 1.3124264891394795,
"learning_rate": 1.513709277201237e-06,
"loss": 0.0567,
"step": 762
},
{
"epoch": 1.4931506849315068,
"grad_norm": 1.755704359284656,
"learning_rate": 1.5027083647417657e-06,
"loss": 0.0601,
"step": 763
},
{
"epoch": 1.4951076320939334,
"grad_norm": 2.2830828761320743,
"learning_rate": 1.4917404990860112e-06,
"loss": 0.0874,
"step": 764
},
{
"epoch": 1.4970645792563602,
"grad_norm": 1.773173966136802,
"learning_rate": 1.4808057838721451e-06,
"loss": 0.0724,
"step": 765
},
{
"epoch": 1.4990215264187867,
"grad_norm": 1.5195227151205504,
"learning_rate": 1.4699043224250919e-06,
"loss": 0.0513,
"step": 766
},
{
"epoch": 1.5009784735812133,
"grad_norm": 1.6714993787934345,
"learning_rate": 1.4590362177555467e-06,
"loss": 0.0582,
"step": 767
},
{
"epoch": 1.5029354207436398,
"grad_norm": 1.9594711967405802,
"learning_rate": 1.4482015725590159e-06,
"loss": 0.0917,
"step": 768
},
{
"epoch": 1.5048923679060664,
"grad_norm": 1.909198400825252,
"learning_rate": 1.4374004892148312e-06,
"loss": 0.0593,
"step": 769
},
{
"epoch": 1.5068493150684932,
"grad_norm": 2.116365097993436,
"learning_rate": 1.4266330697851955e-06,
"loss": 0.0873,
"step": 770
},
{
"epoch": 1.5088062622309197,
"grad_norm": 1.6381746475196974,
"learning_rate": 1.4158994160142114e-06,
"loss": 0.0536,
"step": 771
},
{
"epoch": 1.5107632093933465,
"grad_norm": 1.4806980311898863,
"learning_rate": 1.4051996293269182e-06,
"loss": 0.0456,
"step": 772
},
{
"epoch": 1.512720156555773,
"grad_norm": 1.6647220544897314,
"learning_rate": 1.3945338108283402e-06,
"loss": 0.0613,
"step": 773
},
{
"epoch": 1.5146771037181996,
"grad_norm": 2.09656630511193,
"learning_rate": 1.3839020613025267e-06,
"loss": 0.0684,
"step": 774
},
{
"epoch": 1.5166340508806262,
"grad_norm": 2.256173645282195,
"learning_rate": 1.373304481211598e-06,
"loss": 0.0583,
"step": 775
},
{
"epoch": 1.5185909980430528,
"grad_norm": 1.6610885904954371,
"learning_rate": 1.3627411706948029e-06,
"loss": 0.0636,
"step": 776
},
{
"epoch": 1.5205479452054793,
"grad_norm": 2.3270370609147313,
"learning_rate": 1.3522122295675616e-06,
"loss": 0.095,
"step": 777
},
{
"epoch": 1.522504892367906,
"grad_norm": 2.7066972456569074,
"learning_rate": 1.3417177573205399e-06,
"loss": 0.1073,
"step": 778
},
{
"epoch": 1.5244618395303327,
"grad_norm": 1.718220736308844,
"learning_rate": 1.3312578531186881e-06,
"loss": 0.0585,
"step": 779
},
{
"epoch": 1.5264187866927594,
"grad_norm": 2.137712382730657,
"learning_rate": 1.3208326158003171e-06,
"loss": 0.0592,
"step": 780
},
{
"epoch": 1.528375733855186,
"grad_norm": 1.691010288213673,
"learning_rate": 1.310442143876164e-06,
"loss": 0.0704,
"step": 781
},
{
"epoch": 1.5303326810176126,
"grad_norm": 1.6085542841570966,
"learning_rate": 1.3000865355284565e-06,
"loss": 0.0812,
"step": 782
},
{
"epoch": 1.532289628180039,
"grad_norm": 1.8358654366245675,
"learning_rate": 1.2897658886099906e-06,
"loss": 0.0765,
"step": 783
},
{
"epoch": 1.5342465753424657,
"grad_norm": 1.6146523843083154,
"learning_rate": 1.2794803006431984e-06,
"loss": 0.0517,
"step": 784
},
{
"epoch": 1.5362035225048922,
"grad_norm": 2.233762078628394,
"learning_rate": 1.2692298688192301e-06,
"loss": 0.0785,
"step": 785
},
{
"epoch": 1.538160469667319,
"grad_norm": 1.9244175768687313,
"learning_rate": 1.2590146899970446e-06,
"loss": 0.0587,
"step": 786
},
{
"epoch": 1.5401174168297456,
"grad_norm": 2.1949381026114483,
"learning_rate": 1.2488348607024781e-06,
"loss": 0.057,
"step": 787
},
{
"epoch": 1.5420743639921723,
"grad_norm": 1.590461422127919,
"learning_rate": 1.2386904771273444e-06,
"loss": 0.0601,
"step": 788
},
{
"epoch": 1.544031311154599,
"grad_norm": 1.7830886323972521,
"learning_rate": 1.228581635128518e-06,
"loss": 0.0655,
"step": 789
},
{
"epoch": 1.5459882583170255,
"grad_norm": 2.202097036310967,
"learning_rate": 1.218508430227035e-06,
"loss": 0.0669,
"step": 790
},
{
"epoch": 1.547945205479452,
"grad_norm": 1.7240485850799274,
"learning_rate": 1.2084709576071885e-06,
"loss": 0.0581,
"step": 791
},
{
"epoch": 1.5499021526418786,
"grad_norm": 1.8690540221728638,
"learning_rate": 1.1984693121156226e-06,
"loss": 0.0735,
"step": 792
},
{
"epoch": 1.5518590998043051,
"grad_norm": 1.9985603296477679,
"learning_rate": 1.1885035882604478e-06,
"loss": 0.0599,
"step": 793
},
{
"epoch": 1.553816046966732,
"grad_norm": 1.9632432326128284,
"learning_rate": 1.1785738802103396e-06,
"loss": 0.0659,
"step": 794
},
{
"epoch": 1.5557729941291585,
"grad_norm": 2.1076162425919613,
"learning_rate": 1.1686802817936488e-06,
"loss": 0.0983,
"step": 795
},
{
"epoch": 1.5577299412915853,
"grad_norm": 2.508888145682337,
"learning_rate": 1.158822886497521e-06,
"loss": 0.0711,
"step": 796
},
{
"epoch": 1.5596868884540118,
"grad_norm": 2.6025480727277945,
"learning_rate": 1.1490017874670046e-06,
"loss": 0.0944,
"step": 797
},
{
"epoch": 1.5616438356164384,
"grad_norm": 1.9451696196039674,
"learning_rate": 1.1392170775041788e-06,
"loss": 0.0629,
"step": 798
},
{
"epoch": 1.563600782778865,
"grad_norm": 3.2653566365483795,
"learning_rate": 1.1294688490672744e-06,
"loss": 0.0687,
"step": 799
},
{
"epoch": 1.5655577299412915,
"grad_norm": 1.5491988761961808,
"learning_rate": 1.1197571942697938e-06,
"loss": 0.054,
"step": 800
},
{
"epoch": 1.567514677103718,
"grad_norm": 1.824384057199256,
"learning_rate": 1.1100822048796494e-06,
"loss": 0.0589,
"step": 801
},
{
"epoch": 1.5694716242661448,
"grad_norm": 2.201153569998338,
"learning_rate": 1.1004439723182942e-06,
"loss": 0.0764,
"step": 802
},
{
"epoch": 1.5714285714285714,
"grad_norm": 1.9648971451834376,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.053,
"step": 803
},
{
"epoch": 1.5733855185909982,
"grad_norm": 1.517310898398437,
"learning_rate": 1.0812781416302642e-06,
"loss": 0.0467,
"step": 804
},
{
"epoch": 1.5753424657534247,
"grad_norm": 1.6308992310290062,
"learning_rate": 1.0717507246064273e-06,
"loss": 0.0571,
"step": 805
},
{
"epoch": 1.5772994129158513,
"grad_norm": 2.1510156600364736,
"learning_rate": 1.0622604266153485e-06,
"loss": 0.0758,
"step": 806
},
{
"epoch": 1.5792563600782779,
"grad_norm": 2.5066959662419666,
"learning_rate": 1.0528073373332782e-06,
"loss": 0.098,
"step": 807
},
{
"epoch": 1.5812133072407044,
"grad_norm": 1.7950784494365253,
"learning_rate": 1.0433915460848764e-06,
"loss": 0.0808,
"step": 808
},
{
"epoch": 1.583170254403131,
"grad_norm": 1.945706064195573,
"learning_rate": 1.0340131418423648e-06,
"loss": 0.0723,
"step": 809
},
{
"epoch": 1.5851272015655578,
"grad_norm": 1.8535920553916685,
"learning_rate": 1.0246722132246857e-06,
"loss": 0.0699,
"step": 810
},
{
"epoch": 1.5870841487279843,
"grad_norm": 2.0127340595113705,
"learning_rate": 1.015368848496665e-06,
"loss": 0.0994,
"step": 811
},
{
"epoch": 1.589041095890411,
"grad_norm": 2.4337962673275975,
"learning_rate": 1.0061031355681766e-06,
"loss": 0.0996,
"step": 812
},
{
"epoch": 1.5909980430528377,
"grad_norm": 2.181307109964522,
"learning_rate": 9.968751619933104e-07,
"loss": 0.0961,
"step": 813
},
{
"epoch": 1.5929549902152642,
"grad_norm": 1.9296584753054122,
"learning_rate": 9.876850149695556e-07,
"loss": 0.0541,
"step": 814
},
{
"epoch": 1.5949119373776908,
"grad_norm": 1.994731492571611,
"learning_rate": 9.785327813369589e-07,
"loss": 0.0734,
"step": 815
},
{
"epoch": 1.5968688845401173,
"grad_norm": 1.421639543623951,
"learning_rate": 9.694185475773216e-07,
"loss": 0.05,
"step": 816
},
{
"epoch": 1.5988258317025439,
"grad_norm": 2.1869164132735643,
"learning_rate": 9.603423998133682e-07,
"loss": 0.0995,
"step": 817
},
{
"epoch": 1.6007827788649707,
"grad_norm": 1.7405090775323613,
"learning_rate": 9.513044238079428e-07,
"loss": 0.0622,
"step": 818
},
{
"epoch": 1.6027397260273972,
"grad_norm": 1.8370517766711922,
"learning_rate": 9.423047049631956e-07,
"loss": 0.0551,
"step": 819
},
{
"epoch": 1.604696673189824,
"grad_norm": 2.2762116573304554,
"learning_rate": 9.333433283197702e-07,
"loss": 0.0891,
"step": 820
},
{
"epoch": 1.6066536203522506,
"grad_norm": 1.7411192220053084,
"learning_rate": 9.244203785560096e-07,
"loss": 0.0567,
"step": 821
},
{
"epoch": 1.6086105675146771,
"grad_norm": 1.7736189889474319,
"learning_rate": 9.15535939987151e-07,
"loss": 0.0684,
"step": 822
},
{
"epoch": 1.6105675146771037,
"grad_norm": 1.8719690213825595,
"learning_rate": 9.066900965645259e-07,
"loss": 0.0921,
"step": 823
},
{
"epoch": 1.6125244618395302,
"grad_norm": 1.840570752780444,
"learning_rate": 8.978829318747745e-07,
"loss": 0.0792,
"step": 824
},
{
"epoch": 1.6144814090019568,
"grad_norm": 1.6807047675287288,
"learning_rate": 8.891145291390479e-07,
"loss": 0.0552,
"step": 825
},
{
"epoch": 1.6164383561643836,
"grad_norm": 1.773823746963521,
"learning_rate": 8.803849712122292e-07,
"loss": 0.0545,
"step": 826
},
{
"epoch": 1.6183953033268101,
"grad_norm": 1.8565310167524063,
"learning_rate": 8.71694340582146e-07,
"loss": 0.0854,
"step": 827
},
{
"epoch": 1.620352250489237,
"grad_norm": 1.5958695872112978,
"learning_rate": 8.63042719368789e-07,
"loss": 0.0481,
"step": 828
},
{
"epoch": 1.6223091976516635,
"grad_norm": 1.992274507166537,
"learning_rate": 8.544301893235424e-07,
"loss": 0.0703,
"step": 829
},
{
"epoch": 1.62426614481409,
"grad_norm": 2.39304611409934,
"learning_rate": 8.458568318284088e-07,
"loss": 0.0958,
"step": 830
},
{
"epoch": 1.6262230919765166,
"grad_norm": 2.0011959902157046,
"learning_rate": 8.37322727895234e-07,
"loss": 0.0598,
"step": 831
},
{
"epoch": 1.6281800391389432,
"grad_norm": 2.1186286042119447,
"learning_rate": 8.288279581649528e-07,
"loss": 0.0672,
"step": 832
},
{
"epoch": 1.6301369863013697,
"grad_norm": 2.297749015581445,
"learning_rate": 8.203726029068149e-07,
"loss": 0.0683,
"step": 833
},
{
"epoch": 1.6320939334637965,
"grad_norm": 1.8706622090948464,
"learning_rate": 8.119567420176411e-07,
"loss": 0.0571,
"step": 834
},
{
"epoch": 1.634050880626223,
"grad_norm": 2.2156132547479257,
"learning_rate": 8.035804550210524e-07,
"loss": 0.0981,
"step": 835
},
{
"epoch": 1.6360078277886498,
"grad_norm": 2.034032246744709,
"learning_rate": 7.952438210667268e-07,
"loss": 0.0757,
"step": 836
},
{
"epoch": 1.6379647749510764,
"grad_norm": 1.8414115101866129,
"learning_rate": 7.869469189296536e-07,
"loss": 0.0656,
"step": 837
},
{
"epoch": 1.639921722113503,
"grad_norm": 1.8540327590161976,
"learning_rate": 7.786898270093846e-07,
"loss": 0.0659,
"step": 838
},
{
"epoch": 1.6418786692759295,
"grad_norm": 1.5558739414609088,
"learning_rate": 7.704726233292953e-07,
"loss": 0.0534,
"step": 839
},
{
"epoch": 1.643835616438356,
"grad_norm": 2.387602918988076,
"learning_rate": 7.622953855358456e-07,
"loss": 0.0935,
"step": 840
},
{
"epoch": 1.6457925636007826,
"grad_norm": 2.197185671479319,
"learning_rate": 7.541581908978457e-07,
"loss": 0.0818,
"step": 841
},
{
"epoch": 1.6477495107632094,
"grad_norm": 1.7591340994370461,
"learning_rate": 7.460611163057347e-07,
"loss": 0.0626,
"step": 842
},
{
"epoch": 1.649706457925636,
"grad_norm": 1.559856507626264,
"learning_rate": 7.380042382708396e-07,
"loss": 0.0453,
"step": 843
},
{
"epoch": 1.6516634050880628,
"grad_norm": 2.442581643653225,
"learning_rate": 7.299876329246652e-07,
"loss": 0.0944,
"step": 844
},
{
"epoch": 1.6536203522504893,
"grad_norm": 1.7541253181991128,
"learning_rate": 7.220113760181641e-07,
"loss": 0.094,
"step": 845
},
{
"epoch": 1.6555772994129159,
"grad_norm": 1.6420776983566667,
"learning_rate": 7.140755429210316e-07,
"loss": 0.068,
"step": 846
},
{
"epoch": 1.6575342465753424,
"grad_norm": 1.7498731702981907,
"learning_rate": 7.061802086209857e-07,
"loss": 0.0759,
"step": 847
},
{
"epoch": 1.659491193737769,
"grad_norm": 2.4339774387308943,
"learning_rate": 6.983254477230589e-07,
"loss": 0.0908,
"step": 848
},
{
"epoch": 1.6614481409001955,
"grad_norm": 1.846420165794024,
"learning_rate": 6.905113344488984e-07,
"loss": 0.0737,
"step": 849
},
{
"epoch": 1.6634050880626223,
"grad_norm": 1.9242564438313337,
"learning_rate": 6.827379426360614e-07,
"loss": 0.0625,
"step": 850
},
{
"epoch": 1.6653620352250489,
"grad_norm": 2.197282654783137,
"learning_rate": 6.750053457373146e-07,
"loss": 0.0894,
"step": 851
},
{
"epoch": 1.6673189823874757,
"grad_norm": 2.6576429147771354,
"learning_rate": 6.673136168199467e-07,
"loss": 0.0899,
"step": 852
},
{
"epoch": 1.6692759295499022,
"grad_norm": 1.9630535523395625,
"learning_rate": 6.596628285650713e-07,
"loss": 0.0719,
"step": 853
},
{
"epoch": 1.6712328767123288,
"grad_norm": 2.206645522504302,
"learning_rate": 6.52053053266945e-07,
"loss": 0.0759,
"step": 854
},
{
"epoch": 1.6731898238747553,
"grad_norm": 1.9264488543590772,
"learning_rate": 6.444843628322845e-07,
"loss": 0.0538,
"step": 855
},
{
"epoch": 1.675146771037182,
"grad_norm": 1.6842348090533155,
"learning_rate": 6.36956828779579e-07,
"loss": 0.0528,
"step": 856
},
{
"epoch": 1.6771037181996085,
"grad_norm": 1.792602770814028,
"learning_rate": 6.294705222384262e-07,
"loss": 0.0665,
"step": 857
},
{
"epoch": 1.6790606653620352,
"grad_norm": 1.8160457436985673,
"learning_rate": 6.22025513948854e-07,
"loss": 0.0702,
"step": 858
},
{
"epoch": 1.6810176125244618,
"grad_norm": 2.108720491261194,
"learning_rate": 6.14621874260648e-07,
"loss": 0.0641,
"step": 859
},
{
"epoch": 1.6829745596868886,
"grad_norm": 2.4917813762183716,
"learning_rate": 6.072596731326974e-07,
"loss": 0.06,
"step": 860
},
{
"epoch": 1.6849315068493151,
"grad_norm": 2.316984338756159,
"learning_rate": 5.999389801323219e-07,
"loss": 0.0999,
"step": 861
},
{
"epoch": 1.6868884540117417,
"grad_norm": 1.7702101966699655,
"learning_rate": 5.92659864434626e-07,
"loss": 0.0447,
"step": 862
},
{
"epoch": 1.6888454011741683,
"grad_norm": 1.6704028640578121,
"learning_rate": 5.854223948218362e-07,
"loss": 0.0666,
"step": 863
},
{
"epoch": 1.6908023483365948,
"grad_norm": 1.9296826996203102,
"learning_rate": 5.782266396826536e-07,
"loss": 0.0869,
"step": 864
},
{
"epoch": 1.6927592954990214,
"grad_norm": 1.8331245595640584,
"learning_rate": 5.710726670116107e-07,
"loss": 0.0611,
"step": 865
},
{
"epoch": 1.6947162426614482,
"grad_norm": 1.759987621884817,
"learning_rate": 5.639605444084273e-07,
"loss": 0.0643,
"step": 866
},
{
"epoch": 1.6966731898238747,
"grad_norm": 2.2535944167073176,
"learning_rate": 5.5689033907737e-07,
"loss": 0.0594,
"step": 867
},
{
"epoch": 1.6986301369863015,
"grad_norm": 2.0582335813644987,
"learning_rate": 5.498621178266167e-07,
"loss": 0.0667,
"step": 868
},
{
"epoch": 1.700587084148728,
"grad_norm": 1.4734945674425672,
"learning_rate": 5.428759470676304e-07,
"loss": 0.0692,
"step": 869
},
{
"epoch": 1.7025440313111546,
"grad_norm": 1.438880082445018,
"learning_rate": 5.359318928145262e-07,
"loss": 0.0433,
"step": 870
},
{
"epoch": 1.7045009784735812,
"grad_norm": 1.8006475656064704,
"learning_rate": 5.290300206834492e-07,
"loss": 0.0638,
"step": 871
},
{
"epoch": 1.7064579256360077,
"grad_norm": 1.8182606673468897,
"learning_rate": 5.221703958919572e-07,
"loss": 0.0464,
"step": 872
},
{
"epoch": 1.7084148727984343,
"grad_norm": 1.9332821863374574,
"learning_rate": 5.15353083258398e-07,
"loss": 0.0611,
"step": 873
},
{
"epoch": 1.710371819960861,
"grad_norm": 1.9296910916381425,
"learning_rate": 5.085781472013051e-07,
"loss": 0.0871,
"step": 874
},
{
"epoch": 1.7123287671232876,
"grad_norm": 1.7286752168515525,
"learning_rate": 5.018456517387837e-07,
"loss": 0.0586,
"step": 875
},
{
"epoch": 1.7142857142857144,
"grad_norm": 2.226057499528436,
"learning_rate": 4.951556604879049e-07,
"loss": 0.0692,
"step": 876
},
{
"epoch": 1.716242661448141,
"grad_norm": 1.9745290525165387,
"learning_rate": 4.885082366641081e-07,
"loss": 0.0777,
"step": 877
},
{
"epoch": 1.7181996086105675,
"grad_norm": 1.9875308866429826,
"learning_rate": 4.819034430806047e-07,
"loss": 0.0522,
"step": 878
},
{
"epoch": 1.720156555772994,
"grad_norm": 1.99767624142291,
"learning_rate": 4.7534134214777704e-07,
"loss": 0.0696,
"step": 879
},
{
"epoch": 1.7221135029354206,
"grad_norm": 2.391649153451773,
"learning_rate": 4.688219958725981e-07,
"loss": 0.0847,
"step": 880
},
{
"epoch": 1.7240704500978472,
"grad_norm": 1.6757770521682804,
"learning_rate": 4.623454658580373e-07,
"loss": 0.0494,
"step": 881
},
{
"epoch": 1.726027397260274,
"grad_norm": 1.8764151160416955,
"learning_rate": 4.5591181330248534e-07,
"loss": 0.0645,
"step": 882
},
{
"epoch": 1.7279843444227005,
"grad_norm": 2.4269974217154644,
"learning_rate": 4.495210989991733e-07,
"loss": 0.0809,
"step": 883
},
{
"epoch": 1.7299412915851273,
"grad_norm": 1.5352329290549542,
"learning_rate": 4.4317338333559335e-07,
"loss": 0.0514,
"step": 884
},
{
"epoch": 1.7318982387475539,
"grad_norm": 1.8934163926622125,
"learning_rate": 4.368687262929372e-07,
"loss": 0.0876,
"step": 885
},
{
"epoch": 1.7338551859099804,
"grad_norm": 1.5647468775643874,
"learning_rate": 4.306071874455225e-07,
"loss": 0.0446,
"step": 886
},
{
"epoch": 1.735812133072407,
"grad_norm": 1.4759074865949746,
"learning_rate": 4.243888259602319e-07,
"loss": 0.0438,
"step": 887
},
{
"epoch": 1.7377690802348336,
"grad_norm": 1.9116678154650808,
"learning_rate": 4.1821370059595577e-07,
"loss": 0.063,
"step": 888
},
{
"epoch": 1.7397260273972601,
"grad_norm": 1.9472488295782138,
"learning_rate": 4.1208186970303097e-07,
"loss": 0.0676,
"step": 889
},
{
"epoch": 1.741682974559687,
"grad_norm": 1.7358050997057057,
"learning_rate": 4.0599339122270097e-07,
"loss": 0.0742,
"step": 890
},
{
"epoch": 1.7436399217221135,
"grad_norm": 3.410224019277629,
"learning_rate": 3.999483226865547e-07,
"loss": 0.0928,
"step": 891
},
{
"epoch": 1.7455968688845402,
"grad_norm": 1.494341352032203,
"learning_rate": 3.939467212159925e-07,
"loss": 0.0485,
"step": 892
},
{
"epoch": 1.7475538160469668,
"grad_norm": 1.9668783500632143,
"learning_rate": 3.879886435216823e-07,
"loss": 0.0801,
"step": 893
},
{
"epoch": 1.7495107632093934,
"grad_norm": 1.4835079443041854,
"learning_rate": 3.8207414590302525e-07,
"loss": 0.0526,
"step": 894
},
{
"epoch": 1.75146771037182,
"grad_norm": 1.9624247621725115,
"learning_rate": 3.762032842476238e-07,
"loss": 0.0797,
"step": 895
},
{
"epoch": 1.7534246575342465,
"grad_norm": 2.2713864588963,
"learning_rate": 3.70376114030751e-07,
"loss": 0.0528,
"step": 896
},
{
"epoch": 1.755381604696673,
"grad_norm": 1.8607715836319558,
"learning_rate": 3.6459269031483e-07,
"loss": 0.0604,
"step": 897
},
{
"epoch": 1.7573385518590998,
"grad_norm": 2.157314365026463,
"learning_rate": 3.5885306774891214e-07,
"loss": 0.0639,
"step": 898
},
{
"epoch": 1.7592954990215264,
"grad_norm": 1.8590858131336914,
"learning_rate": 3.5315730056815933e-07,
"loss": 0.0634,
"step": 899
},
{
"epoch": 1.7612524461839532,
"grad_norm": 2.53844017163972,
"learning_rate": 3.475054425933344e-07,
"loss": 0.0808,
"step": 900
},
{
"epoch": 1.7632093933463797,
"grad_norm": 1.5469904171521347,
"learning_rate": 3.4189754723028825e-07,
"loss": 0.0716,
"step": 901
},
{
"epoch": 1.7651663405088063,
"grad_norm": 1.8092626706927226,
"learning_rate": 3.363336674694601e-07,
"loss": 0.0558,
"step": 902
},
{
"epoch": 1.7671232876712328,
"grad_norm": 1.9266647400975379,
"learning_rate": 3.308138558853746e-07,
"loss": 0.0666,
"step": 903
},
{
"epoch": 1.7690802348336594,
"grad_norm": 2.01907407855337,
"learning_rate": 3.2533816463614253e-07,
"loss": 0.071,
"step": 904
},
{
"epoch": 1.771037181996086,
"grad_norm": 1.578190203315228,
"learning_rate": 3.1990664546297287e-07,
"loss": 0.0499,
"step": 905
},
{
"epoch": 1.7729941291585127,
"grad_norm": 1.7808877152612739,
"learning_rate": 3.145193496896809e-07,
"loss": 0.0569,
"step": 906
},
{
"epoch": 1.7749510763209393,
"grad_norm": 2.0494636284256655,
"learning_rate": 3.0917632822220214e-07,
"loss": 0.0823,
"step": 907
},
{
"epoch": 1.776908023483366,
"grad_norm": 2.366853153663543,
"learning_rate": 3.0387763154811457e-07,
"loss": 0.0776,
"step": 908
},
{
"epoch": 1.7788649706457926,
"grad_norm": 2.058236493087252,
"learning_rate": 2.986233097361579e-07,
"loss": 0.0569,
"step": 909
},
{
"epoch": 1.7808219178082192,
"grad_norm": 1.5657496542565184,
"learning_rate": 2.934134124357646e-07,
"loss": 0.0566,
"step": 910
},
{
"epoch": 1.7827788649706457,
"grad_norm": 1.7752713769858823,
"learning_rate": 2.882479888765888e-07,
"loss": 0.0692,
"step": 911
},
{
"epoch": 1.7847358121330723,
"grad_norm": 1.6094995240857757,
"learning_rate": 2.83127087868037e-07,
"loss": 0.0459,
"step": 912
},
{
"epoch": 1.7866927592954989,
"grad_norm": 2.0324434872220465,
"learning_rate": 2.780507577988156e-07,
"loss": 0.0786,
"step": 913
},
{
"epoch": 1.7886497064579256,
"grad_norm": 1.7949748652031825,
"learning_rate": 2.730190466364652e-07,
"loss": 0.0625,
"step": 914
},
{
"epoch": 1.7906066536203522,
"grad_norm": 2.010465715781938,
"learning_rate": 2.6803200192691393e-07,
"loss": 0.0879,
"step": 915
},
{
"epoch": 1.792563600782779,
"grad_norm": 1.5607669434339806,
"learning_rate": 2.630896707940228e-07,
"loss": 0.0515,
"step": 916
},
{
"epoch": 1.7945205479452055,
"grad_norm": 2.0845137127362903,
"learning_rate": 2.5819209993914185e-07,
"loss": 0.068,
"step": 917
},
{
"epoch": 1.796477495107632,
"grad_norm": 2.30313347246288,
"learning_rate": 2.533393356406749e-07,
"loss": 0.0736,
"step": 918
},
{
"epoch": 1.7984344422700587,
"grad_norm": 2.057945814247378,
"learning_rate": 2.485314237536307e-07,
"loss": 0.0611,
"step": 919
},
{
"epoch": 1.8003913894324852,
"grad_norm": 1.8708164497629671,
"learning_rate": 2.4376840970920133e-07,
"loss": 0.0624,
"step": 920
},
{
"epoch": 1.8023483365949118,
"grad_norm": 1.8708508443320144,
"learning_rate": 2.3905033851432393e-07,
"loss": 0.059,
"step": 921
},
{
"epoch": 1.8043052837573386,
"grad_norm": 2.4262504636872437,
"learning_rate": 2.3437725475126126e-07,
"loss": 0.102,
"step": 922
},
{
"epoch": 1.8062622309197651,
"grad_norm": 1.932768763902872,
"learning_rate": 2.297492025771786e-07,
"loss": 0.0684,
"step": 923
},
{
"epoch": 1.808219178082192,
"grad_norm": 1.9841509005780171,
"learning_rate": 2.2516622572372416e-07,
"loss": 0.0532,
"step": 924
},
{
"epoch": 1.8101761252446185,
"grad_norm": 1.875366096325679,
"learning_rate": 2.206283674966192e-07,
"loss": 0.0644,
"step": 925
},
{
"epoch": 1.812133072407045,
"grad_norm": 1.7750663158412083,
"learning_rate": 2.1613567077524878e-07,
"loss": 0.0625,
"step": 926
},
{
"epoch": 1.8140900195694716,
"grad_norm": 2.6808620966636174,
"learning_rate": 2.1168817801225262e-07,
"loss": 0.0628,
"step": 927
},
{
"epoch": 1.8160469667318981,
"grad_norm": 1.6735846693480432,
"learning_rate": 2.0728593123312934e-07,
"loss": 0.0549,
"step": 928
},
{
"epoch": 1.8180039138943247,
"grad_norm": 2.0635673798678127,
"learning_rate": 2.0292897203583405e-07,
"loss": 0.0703,
"step": 929
},
{
"epoch": 1.8199608610567515,
"grad_norm": 2.0259852061877854,
"learning_rate": 1.986173415903897e-07,
"loss": 0.0592,
"step": 930
},
{
"epoch": 1.821917808219178,
"grad_norm": 2.1390807393548488,
"learning_rate": 1.9435108063849684e-07,
"loss": 0.0834,
"step": 931
},
{
"epoch": 1.8238747553816048,
"grad_norm": 1.9498409008032929,
"learning_rate": 1.9013022949314563e-07,
"loss": 0.0671,
"step": 932
},
{
"epoch": 1.8258317025440314,
"grad_norm": 1.8431007375401527,
"learning_rate": 1.8595482803823894e-07,
"loss": 0.0692,
"step": 933
},
{
"epoch": 1.827788649706458,
"grad_norm": 2.1037553628231023,
"learning_rate": 1.8182491572821536e-07,
"loss": 0.0692,
"step": 934
},
{
"epoch": 1.8297455968688845,
"grad_norm": 2.262656095578377,
"learning_rate": 1.777405315876718e-07,
"loss": 0.0573,
"step": 935
},
{
"epoch": 1.831702544031311,
"grad_norm": 2.1548954201305963,
"learning_rate": 1.737017142109998e-07,
"loss": 0.0701,
"step": 936
},
{
"epoch": 1.8336594911937376,
"grad_norm": 1.687332897598958,
"learning_rate": 1.6970850176201704e-07,
"loss": 0.0662,
"step": 937
},
{
"epoch": 1.8356164383561644,
"grad_norm": 1.8324219399727675,
"learning_rate": 1.6576093197361253e-07,
"loss": 0.0663,
"step": 938
},
{
"epoch": 1.837573385518591,
"grad_norm": 1.778099967692163,
"learning_rate": 1.6185904214738203e-07,
"loss": 0.0769,
"step": 939
},
{
"epoch": 1.8395303326810177,
"grad_norm": 1.8386100239909018,
"learning_rate": 1.5800286915328034e-07,
"loss": 0.0502,
"step": 940
},
{
"epoch": 1.8414872798434443,
"grad_norm": 2.0741422332829784,
"learning_rate": 1.5419244942927402e-07,
"loss": 0.0768,
"step": 941
},
{
"epoch": 1.8434442270058709,
"grad_norm": 1.4998109917237057,
"learning_rate": 1.5042781898099433e-07,
"loss": 0.0608,
"step": 942
},
{
"epoch": 1.8454011741682974,
"grad_norm": 1.7383769482825193,
"learning_rate": 1.4670901338139754e-07,
"loss": 0.057,
"step": 943
},
{
"epoch": 1.847358121330724,
"grad_norm": 1.8954651684696426,
"learning_rate": 1.430360677704301e-07,
"loss": 0.0572,
"step": 944
},
{
"epoch": 1.8493150684931505,
"grad_norm": 1.653226606022547,
"learning_rate": 1.39409016854693e-07,
"loss": 0.0626,
"step": 945
},
{
"epoch": 1.8512720156555773,
"grad_norm": 2.272238448379762,
"learning_rate": 1.3582789490712179e-07,
"loss": 0.0764,
"step": 946
},
{
"epoch": 1.8532289628180039,
"grad_norm": 1.4662218534342983,
"learning_rate": 1.3229273576665148e-07,
"loss": 0.0432,
"step": 947
},
{
"epoch": 1.8551859099804306,
"grad_norm": 1.5357933122998098,
"learning_rate": 1.2880357283790778e-07,
"loss": 0.0579,
"step": 948
},
{
"epoch": 1.8571428571428572,
"grad_norm": 1.3146244294690186,
"learning_rate": 1.253604390908819e-07,
"loss": 0.0327,
"step": 949
},
{
"epoch": 1.8590998043052838,
"grad_norm": 1.8993226307286248,
"learning_rate": 1.219633670606274e-07,
"loss": 0.0698,
"step": 950
},
{
"epoch": 1.8610567514677103,
"grad_norm": 2.104947974431727,
"learning_rate": 1.1861238884694759e-07,
"loss": 0.0791,
"step": 951
},
{
"epoch": 1.8630136986301369,
"grad_norm": 2.2062112293699183,
"learning_rate": 1.1530753611409151e-07,
"loss": 0.0629,
"step": 952
},
{
"epoch": 1.8649706457925634,
"grad_norm": 1.4946985884492225,
"learning_rate": 1.1204884009046013e-07,
"loss": 0.0577,
"step": 953
},
{
"epoch": 1.8669275929549902,
"grad_norm": 2.0735034634231053,
"learning_rate": 1.0883633156830553e-07,
"loss": 0.0665,
"step": 954
},
{
"epoch": 1.8688845401174168,
"grad_norm": 1.98280703889663,
"learning_rate": 1.0567004090344224e-07,
"loss": 0.0884,
"step": 955
},
{
"epoch": 1.8708414872798436,
"grad_norm": 1.3864192140071434,
"learning_rate": 1.0254999801496246e-07,
"loss": 0.0559,
"step": 956
},
{
"epoch": 1.8727984344422701,
"grad_norm": 2.7182462480969485,
"learning_rate": 9.947623238494853e-08,
"loss": 0.0763,
"step": 957
},
{
"epoch": 1.8747553816046967,
"grad_norm": 1.63927921140115,
"learning_rate": 9.644877305819977e-08,
"loss": 0.0522,
"step": 958
},
{
"epoch": 1.8767123287671232,
"grad_norm": 1.9363451489211527,
"learning_rate": 9.346764864195335e-08,
"loss": 0.0834,
"step": 959
},
{
"epoch": 1.8786692759295498,
"grad_norm": 3.465960323766893,
"learning_rate": 9.053288730561716e-08,
"loss": 0.0911,
"step": 960
},
{
"epoch": 1.8806262230919764,
"grad_norm": 1.8539797672965617,
"learning_rate": 8.764451678050234e-08,
"loss": 0.0733,
"step": 961
},
{
"epoch": 1.8825831702544031,
"grad_norm": 1.517581847538663,
"learning_rate": 8.480256435956124e-08,
"loss": 0.0641,
"step": 962
},
{
"epoch": 1.8845401174168297,
"grad_norm": 1.7781989287288233,
"learning_rate": 8.200705689712873e-08,
"loss": 0.0502,
"step": 963
},
{
"epoch": 1.8864970645792565,
"grad_norm": 1.9133766890551818,
"learning_rate": 7.925802080867129e-08,
"loss": 0.0729,
"step": 964
},
{
"epoch": 1.888454011741683,
"grad_norm": 2.152636514583198,
"learning_rate": 7.655548207053332e-08,
"loss": 0.0745,
"step": 965
},
{
"epoch": 1.8904109589041096,
"grad_norm": 2.1252250197952534,
"learning_rate": 7.389946621969679e-08,
"loss": 0.0818,
"step": 966
},
{
"epoch": 1.8923679060665362,
"grad_norm": 1.6061641427162516,
"learning_rate": 7.128999835353423e-08,
"loss": 0.0561,
"step": 967
},
{
"epoch": 1.8943248532289627,
"grad_norm": 1.8297838429795426,
"learning_rate": 6.872710312957498e-08,
"loss": 0.0658,
"step": 968
},
{
"epoch": 1.8962818003913893,
"grad_norm": 2.0574209357464546,
"learning_rate": 6.621080476527209e-08,
"loss": 0.0561,
"step": 969
},
{
"epoch": 1.898238747553816,
"grad_norm": 1.8748345374595397,
"learning_rate": 6.374112703777302e-08,
"loss": 0.0493,
"step": 970
},
{
"epoch": 1.9001956947162426,
"grad_norm": 1.7467928636031131,
"learning_rate": 6.131809328369542e-08,
"loss": 0.0699,
"step": 971
},
{
"epoch": 1.9021526418786694,
"grad_norm": 2.0868922280803495,
"learning_rate": 5.8941726398903944e-08,
"loss": 0.0673,
"step": 972
},
{
"epoch": 1.904109589041096,
"grad_norm": 2.1883066037691288,
"learning_rate": 5.661204883829763e-08,
"loss": 0.0768,
"step": 973
},
{
"epoch": 1.9060665362035225,
"grad_norm": 2.212775319654743,
"learning_rate": 5.432908261559733e-08,
"loss": 0.0935,
"step": 974
},
{
"epoch": 1.908023483365949,
"grad_norm": 1.5359910765464901,
"learning_rate": 5.209284930313418e-08,
"loss": 0.0595,
"step": 975
},
{
"epoch": 1.9099804305283756,
"grad_norm": 1.6101646885927374,
"learning_rate": 4.9903370031649246e-08,
"loss": 0.0662,
"step": 976
},
{
"epoch": 1.9119373776908022,
"grad_norm": 2.3218869497095995,
"learning_rate": 4.776066549009195e-08,
"loss": 0.0621,
"step": 977
},
{
"epoch": 1.913894324853229,
"grad_norm": 1.8174608522656108,
"learning_rate": 4.5664755925426406e-08,
"loss": 0.0629,
"step": 978
},
{
"epoch": 1.9158512720156555,
"grad_norm": 1.9482990709823016,
"learning_rate": 4.3615661142439316e-08,
"loss": 0.0512,
"step": 979
},
{
"epoch": 1.9178082191780823,
"grad_norm": 1.5556668948512693,
"learning_rate": 4.1613400503550114e-08,
"loss": 0.0494,
"step": 980
},
{
"epoch": 1.9197651663405089,
"grad_norm": 1.8631717493061775,
"learning_rate": 3.9657992928632813e-08,
"loss": 0.0672,
"step": 981
},
{
"epoch": 1.9217221135029354,
"grad_norm": 1.6992409719622914,
"learning_rate": 3.774945689483445e-08,
"loss": 0.0658,
"step": 982
},
{
"epoch": 1.923679060665362,
"grad_norm": 1.4963212768398808,
"learning_rate": 3.5887810436399106e-08,
"loss": 0.0571,
"step": 983
},
{
"epoch": 1.9256360078277885,
"grad_norm": 2.350928057455318,
"learning_rate": 3.407307114450253e-08,
"loss": 0.0833,
"step": 984
},
{
"epoch": 1.927592954990215,
"grad_norm": 1.9342743047902613,
"learning_rate": 3.2305256167078916e-08,
"loss": 0.0818,
"step": 985
},
{
"epoch": 1.9295499021526419,
"grad_norm": 1.8028486027561332,
"learning_rate": 3.058438220866544e-08,
"loss": 0.0532,
"step": 986
},
{
"epoch": 1.9315068493150684,
"grad_norm": 2.6236519448970492,
"learning_rate": 2.8910465530240793e-08,
"loss": 0.0722,
"step": 987
},
{
"epoch": 1.9334637964774952,
"grad_norm": 1.6965572650458751,
"learning_rate": 2.7283521949070802e-08,
"loss": 0.0746,
"step": 988
},
{
"epoch": 1.9354207436399218,
"grad_norm": 1.8074958593917527,
"learning_rate": 2.570356683856301e-08,
"loss": 0.0592,
"step": 989
},
{
"epoch": 1.9373776908023483,
"grad_norm": 2.0951472727513485,
"learning_rate": 2.4170615128117358e-08,
"loss": 0.0828,
"step": 990
},
{
"epoch": 1.939334637964775,
"grad_norm": 2.1318537581694104,
"learning_rate": 2.2684681302986288e-08,
"loss": 0.0672,
"step": 991
},
{
"epoch": 1.9412915851272015,
"grad_norm": 1.9079226553485233,
"learning_rate": 2.12457794041393e-08,
"loss": 0.0581,
"step": 992
},
{
"epoch": 1.943248532289628,
"grad_norm": 1.5478411354004666,
"learning_rate": 1.9853923028128607e-08,
"loss": 0.0593,
"step": 993
},
{
"epoch": 1.9452054794520548,
"grad_norm": 1.9272575394874891,
"learning_rate": 1.850912532696092e-08,
"loss": 0.0703,
"step": 994
},
{
"epoch": 1.9471624266144814,
"grad_norm": 1.9052324223439343,
"learning_rate": 1.7211399007974748e-08,
"loss": 0.0738,
"step": 995
},
{
"epoch": 1.9491193737769081,
"grad_norm": 1.9102787610536185,
"learning_rate": 1.596075633371774e-08,
"loss": 0.0578,
"step": 996
},
{
"epoch": 1.9510763209393347,
"grad_norm": 2.0392105696294602,
"learning_rate": 1.4757209121832317e-08,
"loss": 0.0493,
"step": 997
},
{
"epoch": 1.9530332681017613,
"grad_norm": 2.5071601664620573,
"learning_rate": 1.3600768744944648e-08,
"loss": 0.0937,
"step": 998
},
{
"epoch": 1.9549902152641878,
"grad_norm": 2.0900696663375586,
"learning_rate": 1.2491446130555306e-08,
"loss": 0.0839,
"step": 999
},
{
"epoch": 1.9569471624266144,
"grad_norm": 1.9294282508872134,
"learning_rate": 1.1429251760938232e-08,
"loss": 0.071,
"step": 1000
},
{
"epoch": 1.9569471624266144,
"eval_loss": 0.18407420814037323,
"eval_runtime": 1.7325,
"eval_samples_per_second": 24.242,
"eval_steps_per_second": 6.349,
"step": 1000
},
{
"epoch": 1.958904109589041,
"grad_norm": 1.8112041643323307,
"learning_rate": 1.041419567303914e-08,
"loss": 0.0546,
"step": 1001
},
{
"epoch": 1.9608610567514677,
"grad_norm": 1.7361095395145336,
"learning_rate": 9.446287458383385e-09,
"loss": 0.0606,
"step": 1002
},
{
"epoch": 1.9628180039138943,
"grad_norm": 1.642561448442732,
"learning_rate": 8.525536262982692e-09,
"loss": 0.0619,
"step": 1003
},
{
"epoch": 1.964774951076321,
"grad_norm": 1.7278412829751106,
"learning_rate": 7.651950787251339e-09,
"loss": 0.0565,
"step": 1004
},
{
"epoch": 1.9667318982387476,
"grad_norm": 2.055949077214264,
"learning_rate": 6.825539285922334e-09,
"loss": 0.0905,
"step": 1005
},
{
"epoch": 1.9686888454011742,
"grad_norm": 2.084485320232909,
"learning_rate": 6.046309567968589e-09,
"loss": 0.0637,
"step": 1006
},
{
"epoch": 1.9706457925636007,
"grad_norm": 2.021062002959845,
"learning_rate": 5.3142689965318685e-09,
"loss": 0.0709,
"step": 1007
},
{
"epoch": 1.9726027397260273,
"grad_norm": 2.0873794075707646,
"learning_rate": 4.629424488850065e-09,
"loss": 0.0754,
"step": 1008
},
{
"epoch": 1.9745596868884538,
"grad_norm": 2.3480923316543394,
"learning_rate": 3.9917825161950305e-09,
"loss": 0.0674,
"step": 1009
},
{
"epoch": 1.9765166340508806,
"grad_norm": 1.2943772394277153,
"learning_rate": 3.4013491038087376e-09,
"loss": 0.0484,
"step": 1010
},
{
"epoch": 1.9784735812133072,
"grad_norm": 3.179898815857171,
"learning_rate": 2.858129830847767e-09,
"loss": 0.0672,
"step": 1011
},
{
"epoch": 1.980430528375734,
"grad_norm": 2.157241986375847,
"learning_rate": 2.3621298303294626e-09,
"loss": 0.0714,
"step": 1012
},
{
"epoch": 1.9823874755381605,
"grad_norm": 2.571040416488115,
"learning_rate": 1.9133537890853036e-09,
"loss": 0.0598,
"step": 1013
},
{
"epoch": 1.984344422700587,
"grad_norm": 1.9537109603426126,
"learning_rate": 1.5118059477142732e-09,
"loss": 0.0559,
"step": 1014
},
{
"epoch": 1.9863013698630136,
"grad_norm": 1.5101317189873702,
"learning_rate": 1.1574901005456662e-09,
"loss": 0.0729,
"step": 1015
},
{
"epoch": 1.9882583170254402,
"grad_norm": 2.2356169846937997,
"learning_rate": 8.504095956002323e-10,
"loss": 0.068,
"step": 1016
},
{
"epoch": 1.9902152641878668,
"grad_norm": 1.9729640574520855,
"learning_rate": 5.905673345601992e-10,
"loss": 0.0582,
"step": 1017
},
{
"epoch": 1.9921722113502935,
"grad_norm": 1.5930362156361813,
"learning_rate": 3.7796577274096244e-10,
"loss": 0.0467,
"step": 1018
},
{
"epoch": 1.99412915851272,
"grad_norm": 2.587174880869091,
"learning_rate": 2.1260691906999087e-10,
"loss": 0.0897,
"step": 1019
},
{
"epoch": 1.9960861056751469,
"grad_norm": 1.9009713158806287,
"learning_rate": 9.449233606573238e-11,
"loss": 0.0626,
"step": 1020
},
{
"epoch": 1.9980430528375734,
"grad_norm": 1.745787129347406,
"learning_rate": 2.3623139821515874e-11,
"loss": 0.0678,
"step": 1021
},
{
"epoch": 2.0,
"grad_norm": 1.4273950099579886,
"learning_rate": 0.0,
"loss": 0.0411,
"step": 1022
},
{
"epoch": 2.0,
"step": 1022,
"total_flos": 3162915078144.0,
"train_loss": 0.13159879108930986,
"train_runtime": 773.0704,
"train_samples_per_second": 10.566,
"train_steps_per_second": 1.322
}
],
"logging_steps": 1,
"max_steps": 1022,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 50000000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3162915078144.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}