gsmyrnis's picture
End of training
595ce46 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 1065,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0028169014084507044,
"grad_norm": 6.843351489795488,
"learning_rate": 9.345794392523364e-08,
"loss": 0.9385,
"step": 1
},
{
"epoch": 0.005633802816901409,
"grad_norm": 6.607567814573529,
"learning_rate": 1.8691588785046729e-07,
"loss": 0.9283,
"step": 2
},
{
"epoch": 0.008450704225352112,
"grad_norm": 6.902238922410287,
"learning_rate": 2.8037383177570096e-07,
"loss": 0.9641,
"step": 3
},
{
"epoch": 0.011267605633802818,
"grad_norm": 6.946697059316272,
"learning_rate": 3.7383177570093457e-07,
"loss": 0.9717,
"step": 4
},
{
"epoch": 0.014084507042253521,
"grad_norm": 7.003582107757224,
"learning_rate": 4.6728971962616824e-07,
"loss": 0.9639,
"step": 5
},
{
"epoch": 0.016901408450704224,
"grad_norm": 6.9376047256330144,
"learning_rate": 5.607476635514019e-07,
"loss": 0.9582,
"step": 6
},
{
"epoch": 0.01971830985915493,
"grad_norm": 6.558029428164414,
"learning_rate": 6.542056074766355e-07,
"loss": 0.9168,
"step": 7
},
{
"epoch": 0.022535211267605635,
"grad_norm": 6.345482529471037,
"learning_rate": 7.476635514018691e-07,
"loss": 0.9253,
"step": 8
},
{
"epoch": 0.02535211267605634,
"grad_norm": 6.6326272471163,
"learning_rate": 8.411214953271029e-07,
"loss": 0.9375,
"step": 9
},
{
"epoch": 0.028169014084507043,
"grad_norm": 6.464087540605822,
"learning_rate": 9.345794392523365e-07,
"loss": 0.9562,
"step": 10
},
{
"epoch": 0.030985915492957747,
"grad_norm": 5.166432873083264,
"learning_rate": 1.0280373831775702e-06,
"loss": 0.8792,
"step": 11
},
{
"epoch": 0.03380281690140845,
"grad_norm": 5.079046652537091,
"learning_rate": 1.1214953271028038e-06,
"loss": 0.8914,
"step": 12
},
{
"epoch": 0.036619718309859155,
"grad_norm": 5.037215339408657,
"learning_rate": 1.2149532710280374e-06,
"loss": 0.9172,
"step": 13
},
{
"epoch": 0.03943661971830986,
"grad_norm": 4.6952926229942165,
"learning_rate": 1.308411214953271e-06,
"loss": 0.887,
"step": 14
},
{
"epoch": 0.04225352112676056,
"grad_norm": 2.7137316275253864,
"learning_rate": 1.4018691588785047e-06,
"loss": 0.823,
"step": 15
},
{
"epoch": 0.04507042253521127,
"grad_norm": 2.6632918255705675,
"learning_rate": 1.4953271028037383e-06,
"loss": 0.8242,
"step": 16
},
{
"epoch": 0.04788732394366197,
"grad_norm": 2.451657467696323,
"learning_rate": 1.588785046728972e-06,
"loss": 0.8375,
"step": 17
},
{
"epoch": 0.05070422535211268,
"grad_norm": 2.2343565320461978,
"learning_rate": 1.6822429906542057e-06,
"loss": 0.8349,
"step": 18
},
{
"epoch": 0.05352112676056338,
"grad_norm": 2.1451406757683613,
"learning_rate": 1.7757009345794394e-06,
"loss": 0.8274,
"step": 19
},
{
"epoch": 0.056338028169014086,
"grad_norm": 2.5866576397571164,
"learning_rate": 1.869158878504673e-06,
"loss": 0.8195,
"step": 20
},
{
"epoch": 0.059154929577464786,
"grad_norm": 3.6068541588513856,
"learning_rate": 1.962616822429907e-06,
"loss": 0.8085,
"step": 21
},
{
"epoch": 0.061971830985915494,
"grad_norm": 3.7301124014564735,
"learning_rate": 2.0560747663551404e-06,
"loss": 0.8007,
"step": 22
},
{
"epoch": 0.0647887323943662,
"grad_norm": 3.879846862236254,
"learning_rate": 2.149532710280374e-06,
"loss": 0.8258,
"step": 23
},
{
"epoch": 0.0676056338028169,
"grad_norm": 3.257187025314909,
"learning_rate": 2.2429906542056077e-06,
"loss": 0.7603,
"step": 24
},
{
"epoch": 0.07042253521126761,
"grad_norm": 3.0150439634525137,
"learning_rate": 2.3364485981308413e-06,
"loss": 0.7567,
"step": 25
},
{
"epoch": 0.07323943661971831,
"grad_norm": 2.731962426951936,
"learning_rate": 2.429906542056075e-06,
"loss": 0.7809,
"step": 26
},
{
"epoch": 0.07605633802816901,
"grad_norm": 2.0851170351121966,
"learning_rate": 2.5233644859813085e-06,
"loss": 0.7279,
"step": 27
},
{
"epoch": 0.07887323943661972,
"grad_norm": 1.5871667506827638,
"learning_rate": 2.616822429906542e-06,
"loss": 0.7732,
"step": 28
},
{
"epoch": 0.08169014084507042,
"grad_norm": 1.3057594440862028,
"learning_rate": 2.7102803738317757e-06,
"loss": 0.6913,
"step": 29
},
{
"epoch": 0.08450704225352113,
"grad_norm": 1.3261878434703152,
"learning_rate": 2.8037383177570094e-06,
"loss": 0.7487,
"step": 30
},
{
"epoch": 0.08732394366197183,
"grad_norm": 1.3262343134044554,
"learning_rate": 2.897196261682243e-06,
"loss": 0.7357,
"step": 31
},
{
"epoch": 0.09014084507042254,
"grad_norm": 1.23592234207026,
"learning_rate": 2.9906542056074766e-06,
"loss": 0.718,
"step": 32
},
{
"epoch": 0.09295774647887324,
"grad_norm": 1.1342511290180883,
"learning_rate": 3.08411214953271e-06,
"loss": 0.6962,
"step": 33
},
{
"epoch": 0.09577464788732394,
"grad_norm": 1.0467874998875826,
"learning_rate": 3.177570093457944e-06,
"loss": 0.6934,
"step": 34
},
{
"epoch": 0.09859154929577464,
"grad_norm": 1.0265524735100418,
"learning_rate": 3.2710280373831774e-06,
"loss": 0.6908,
"step": 35
},
{
"epoch": 0.10140845070422536,
"grad_norm": 0.7759183188592782,
"learning_rate": 3.3644859813084115e-06,
"loss": 0.6902,
"step": 36
},
{
"epoch": 0.10422535211267606,
"grad_norm": 0.7307747183179154,
"learning_rate": 3.457943925233645e-06,
"loss": 0.6436,
"step": 37
},
{
"epoch": 0.10704225352112676,
"grad_norm": 0.9476449074596629,
"learning_rate": 3.5514018691588787e-06,
"loss": 0.6674,
"step": 38
},
{
"epoch": 0.10985915492957747,
"grad_norm": 1.023136741079733,
"learning_rate": 3.6448598130841123e-06,
"loss": 0.6525,
"step": 39
},
{
"epoch": 0.11267605633802817,
"grad_norm": 0.9756562165836902,
"learning_rate": 3.738317757009346e-06,
"loss": 0.6604,
"step": 40
},
{
"epoch": 0.11549295774647887,
"grad_norm": 0.6835958250022881,
"learning_rate": 3.8317757009345796e-06,
"loss": 0.6718,
"step": 41
},
{
"epoch": 0.11830985915492957,
"grad_norm": 0.6560085321954412,
"learning_rate": 3.925233644859814e-06,
"loss": 0.6408,
"step": 42
},
{
"epoch": 0.12112676056338029,
"grad_norm": 0.8372121993022718,
"learning_rate": 4.018691588785047e-06,
"loss": 0.6352,
"step": 43
},
{
"epoch": 0.12394366197183099,
"grad_norm": 0.9399706591180332,
"learning_rate": 4.112149532710281e-06,
"loss": 0.6882,
"step": 44
},
{
"epoch": 0.1267605633802817,
"grad_norm": 0.6897032499753717,
"learning_rate": 4.205607476635514e-06,
"loss": 0.6635,
"step": 45
},
{
"epoch": 0.1295774647887324,
"grad_norm": 0.5666941816418474,
"learning_rate": 4.299065420560748e-06,
"loss": 0.6223,
"step": 46
},
{
"epoch": 0.1323943661971831,
"grad_norm": 0.6466666231083906,
"learning_rate": 4.392523364485981e-06,
"loss": 0.6485,
"step": 47
},
{
"epoch": 0.1352112676056338,
"grad_norm": 0.6307749815943701,
"learning_rate": 4.485981308411215e-06,
"loss": 0.6631,
"step": 48
},
{
"epoch": 0.13802816901408452,
"grad_norm": 0.6101350712843577,
"learning_rate": 4.579439252336449e-06,
"loss": 0.6319,
"step": 49
},
{
"epoch": 0.14084507042253522,
"grad_norm": 0.5822616465924019,
"learning_rate": 4.6728971962616825e-06,
"loss": 0.6536,
"step": 50
},
{
"epoch": 0.14366197183098592,
"grad_norm": 0.5720766262956746,
"learning_rate": 4.766355140186917e-06,
"loss": 0.6441,
"step": 51
},
{
"epoch": 0.14647887323943662,
"grad_norm": 0.5821005508097582,
"learning_rate": 4.85981308411215e-06,
"loss": 0.6445,
"step": 52
},
{
"epoch": 0.14929577464788732,
"grad_norm": 0.7326245657069738,
"learning_rate": 4.953271028037384e-06,
"loss": 0.6532,
"step": 53
},
{
"epoch": 0.15211267605633802,
"grad_norm": 0.5055300266839129,
"learning_rate": 5.046728971962617e-06,
"loss": 0.6489,
"step": 54
},
{
"epoch": 0.15492957746478872,
"grad_norm": 0.5340536053639043,
"learning_rate": 5.14018691588785e-06,
"loss": 0.6414,
"step": 55
},
{
"epoch": 0.15774647887323945,
"grad_norm": 0.5829492821591393,
"learning_rate": 5.233644859813084e-06,
"loss": 0.6152,
"step": 56
},
{
"epoch": 0.16056338028169015,
"grad_norm": 0.5453794801583319,
"learning_rate": 5.3271028037383174e-06,
"loss": 0.6039,
"step": 57
},
{
"epoch": 0.16338028169014085,
"grad_norm": 0.5899665130491681,
"learning_rate": 5.4205607476635515e-06,
"loss": 0.6185,
"step": 58
},
{
"epoch": 0.16619718309859155,
"grad_norm": 0.5611513671408634,
"learning_rate": 5.514018691588785e-06,
"loss": 0.6148,
"step": 59
},
{
"epoch": 0.16901408450704225,
"grad_norm": 0.5104752150744826,
"learning_rate": 5.607476635514019e-06,
"loss": 0.6207,
"step": 60
},
{
"epoch": 0.17183098591549295,
"grad_norm": 0.5222500246055776,
"learning_rate": 5.700934579439253e-06,
"loss": 0.6183,
"step": 61
},
{
"epoch": 0.17464788732394365,
"grad_norm": 0.7031567337432721,
"learning_rate": 5.794392523364486e-06,
"loss": 0.6158,
"step": 62
},
{
"epoch": 0.17746478873239438,
"grad_norm": 0.5670544399349872,
"learning_rate": 5.88785046728972e-06,
"loss": 0.6164,
"step": 63
},
{
"epoch": 0.18028169014084508,
"grad_norm": 0.5835195221742685,
"learning_rate": 5.981308411214953e-06,
"loss": 0.6143,
"step": 64
},
{
"epoch": 0.18309859154929578,
"grad_norm": 0.486716857038764,
"learning_rate": 6.074766355140187e-06,
"loss": 0.6218,
"step": 65
},
{
"epoch": 0.18591549295774648,
"grad_norm": 0.6518972141316477,
"learning_rate": 6.16822429906542e-06,
"loss": 0.6158,
"step": 66
},
{
"epoch": 0.18873239436619718,
"grad_norm": 0.5773322831933766,
"learning_rate": 6.2616822429906544e-06,
"loss": 0.6202,
"step": 67
},
{
"epoch": 0.19154929577464788,
"grad_norm": 0.5912349009905348,
"learning_rate": 6.355140186915888e-06,
"loss": 0.6219,
"step": 68
},
{
"epoch": 0.19436619718309858,
"grad_norm": 0.6060790549903095,
"learning_rate": 6.448598130841122e-06,
"loss": 0.5999,
"step": 69
},
{
"epoch": 0.19718309859154928,
"grad_norm": 0.6216646819276722,
"learning_rate": 6.542056074766355e-06,
"loss": 0.6077,
"step": 70
},
{
"epoch": 0.2,
"grad_norm": 0.47329562993465146,
"learning_rate": 6.635514018691589e-06,
"loss": 0.6051,
"step": 71
},
{
"epoch": 0.2028169014084507,
"grad_norm": 0.534157980811324,
"learning_rate": 6.728971962616823e-06,
"loss": 0.6089,
"step": 72
},
{
"epoch": 0.2056338028169014,
"grad_norm": 0.49293802647714885,
"learning_rate": 6.822429906542056e-06,
"loss": 0.5944,
"step": 73
},
{
"epoch": 0.2084507042253521,
"grad_norm": 0.5105802467170905,
"learning_rate": 6.91588785046729e-06,
"loss": 0.6192,
"step": 74
},
{
"epoch": 0.2112676056338028,
"grad_norm": 0.4545810635734364,
"learning_rate": 7.009345794392523e-06,
"loss": 0.5849,
"step": 75
},
{
"epoch": 0.2140845070422535,
"grad_norm": 0.5132015985707264,
"learning_rate": 7.1028037383177574e-06,
"loss": 0.5757,
"step": 76
},
{
"epoch": 0.21690140845070421,
"grad_norm": 0.5240278152519751,
"learning_rate": 7.196261682242991e-06,
"loss": 0.593,
"step": 77
},
{
"epoch": 0.21971830985915494,
"grad_norm": 0.47719974239889557,
"learning_rate": 7.289719626168225e-06,
"loss": 0.578,
"step": 78
},
{
"epoch": 0.22253521126760564,
"grad_norm": 0.5022510515134598,
"learning_rate": 7.383177570093458e-06,
"loss": 0.5789,
"step": 79
},
{
"epoch": 0.22535211267605634,
"grad_norm": 0.5268449362714561,
"learning_rate": 7.476635514018692e-06,
"loss": 0.6014,
"step": 80
},
{
"epoch": 0.22816901408450704,
"grad_norm": 0.50075545195209,
"learning_rate": 7.570093457943926e-06,
"loss": 0.5838,
"step": 81
},
{
"epoch": 0.23098591549295774,
"grad_norm": 0.5519102966922635,
"learning_rate": 7.663551401869159e-06,
"loss": 0.5787,
"step": 82
},
{
"epoch": 0.23380281690140844,
"grad_norm": 0.5305728995724017,
"learning_rate": 7.757009345794392e-06,
"loss": 0.5654,
"step": 83
},
{
"epoch": 0.23661971830985915,
"grad_norm": 0.4984850288710446,
"learning_rate": 7.850467289719627e-06,
"loss": 0.5998,
"step": 84
},
{
"epoch": 0.23943661971830985,
"grad_norm": 0.5870272453921658,
"learning_rate": 7.94392523364486e-06,
"loss": 0.6066,
"step": 85
},
{
"epoch": 0.24225352112676057,
"grad_norm": 0.5113265334277904,
"learning_rate": 8.037383177570094e-06,
"loss": 0.6123,
"step": 86
},
{
"epoch": 0.24507042253521127,
"grad_norm": 0.490831722513275,
"learning_rate": 8.130841121495327e-06,
"loss": 0.5984,
"step": 87
},
{
"epoch": 0.24788732394366197,
"grad_norm": 0.5558569258369573,
"learning_rate": 8.224299065420562e-06,
"loss": 0.6069,
"step": 88
},
{
"epoch": 0.2507042253521127,
"grad_norm": 0.49435562466880034,
"learning_rate": 8.317757009345795e-06,
"loss": 0.6304,
"step": 89
},
{
"epoch": 0.2535211267605634,
"grad_norm": 0.49423876533813854,
"learning_rate": 8.411214953271028e-06,
"loss": 0.6034,
"step": 90
},
{
"epoch": 0.2563380281690141,
"grad_norm": 0.5511212763406935,
"learning_rate": 8.504672897196263e-06,
"loss": 0.614,
"step": 91
},
{
"epoch": 0.2591549295774648,
"grad_norm": 0.4543803002498734,
"learning_rate": 8.598130841121496e-06,
"loss": 0.5713,
"step": 92
},
{
"epoch": 0.2619718309859155,
"grad_norm": 0.5706161712680489,
"learning_rate": 8.69158878504673e-06,
"loss": 0.6034,
"step": 93
},
{
"epoch": 0.2647887323943662,
"grad_norm": 0.505724802570353,
"learning_rate": 8.785046728971963e-06,
"loss": 0.5643,
"step": 94
},
{
"epoch": 0.2676056338028169,
"grad_norm": 0.517868035081914,
"learning_rate": 8.878504672897197e-06,
"loss": 0.598,
"step": 95
},
{
"epoch": 0.2704225352112676,
"grad_norm": 0.5476188621020196,
"learning_rate": 8.97196261682243e-06,
"loss": 0.5783,
"step": 96
},
{
"epoch": 0.27323943661971833,
"grad_norm": 0.536506093724152,
"learning_rate": 9.065420560747664e-06,
"loss": 0.585,
"step": 97
},
{
"epoch": 0.27605633802816903,
"grad_norm": 0.5052391248701771,
"learning_rate": 9.158878504672899e-06,
"loss": 0.5856,
"step": 98
},
{
"epoch": 0.27887323943661974,
"grad_norm": 0.5507911743646905,
"learning_rate": 9.252336448598132e-06,
"loss": 0.5742,
"step": 99
},
{
"epoch": 0.28169014084507044,
"grad_norm": 0.5667507509388369,
"learning_rate": 9.345794392523365e-06,
"loss": 0.5804,
"step": 100
},
{
"epoch": 0.28450704225352114,
"grad_norm": 0.5367761197956139,
"learning_rate": 9.439252336448598e-06,
"loss": 0.6108,
"step": 101
},
{
"epoch": 0.28732394366197184,
"grad_norm": 0.6151261182976093,
"learning_rate": 9.532710280373833e-06,
"loss": 0.5803,
"step": 102
},
{
"epoch": 0.29014084507042254,
"grad_norm": 0.5630295655109795,
"learning_rate": 9.626168224299066e-06,
"loss": 0.569,
"step": 103
},
{
"epoch": 0.29295774647887324,
"grad_norm": 0.595128847891053,
"learning_rate": 9.7196261682243e-06,
"loss": 0.5949,
"step": 104
},
{
"epoch": 0.29577464788732394,
"grad_norm": 0.5932250260888466,
"learning_rate": 9.813084112149533e-06,
"loss": 0.5912,
"step": 105
},
{
"epoch": 0.29859154929577464,
"grad_norm": 0.554296183304856,
"learning_rate": 9.906542056074768e-06,
"loss": 0.5858,
"step": 106
},
{
"epoch": 0.30140845070422534,
"grad_norm": 0.5542586088977368,
"learning_rate": 1e-05,
"loss": 0.5947,
"step": 107
},
{
"epoch": 0.30422535211267604,
"grad_norm": 0.5539014174596538,
"learning_rate": 9.999973115104874e-06,
"loss": 0.564,
"step": 108
},
{
"epoch": 0.30704225352112674,
"grad_norm": 0.5885183535059,
"learning_rate": 9.999892460708615e-06,
"loss": 0.6013,
"step": 109
},
{
"epoch": 0.30985915492957744,
"grad_norm": 0.47510239201375715,
"learning_rate": 9.999758037678576e-06,
"loss": 0.6016,
"step": 110
},
{
"epoch": 0.3126760563380282,
"grad_norm": 0.5191594501253673,
"learning_rate": 9.999569847460335e-06,
"loss": 0.5849,
"step": 111
},
{
"epoch": 0.3154929577464789,
"grad_norm": 0.5208229000776549,
"learning_rate": 9.999327892077683e-06,
"loss": 0.5767,
"step": 112
},
{
"epoch": 0.3183098591549296,
"grad_norm": 0.5371441518969067,
"learning_rate": 9.9990321741326e-06,
"loss": 0.5763,
"step": 113
},
{
"epoch": 0.3211267605633803,
"grad_norm": 0.5928018630279552,
"learning_rate": 9.998682696805224e-06,
"loss": 0.5594,
"step": 114
},
{
"epoch": 0.323943661971831,
"grad_norm": 0.5256757847980414,
"learning_rate": 9.998279463853819e-06,
"loss": 0.5829,
"step": 115
},
{
"epoch": 0.3267605633802817,
"grad_norm": 0.6374633387905916,
"learning_rate": 9.997822479614733e-06,
"loss": 0.5829,
"step": 116
},
{
"epoch": 0.3295774647887324,
"grad_norm": 0.5099003102667371,
"learning_rate": 9.997311749002358e-06,
"loss": 0.5796,
"step": 117
},
{
"epoch": 0.3323943661971831,
"grad_norm": 0.6634412406674159,
"learning_rate": 9.996747277509068e-06,
"loss": 0.5569,
"step": 118
},
{
"epoch": 0.3352112676056338,
"grad_norm": 0.6007705739995757,
"learning_rate": 9.996129071205167e-06,
"loss": 0.5747,
"step": 119
},
{
"epoch": 0.3380281690140845,
"grad_norm": 0.5996405607790697,
"learning_rate": 9.99545713673882e-06,
"loss": 0.5637,
"step": 120
},
{
"epoch": 0.3408450704225352,
"grad_norm": 0.5597246678529963,
"learning_rate": 9.994731481335979e-06,
"loss": 0.582,
"step": 121
},
{
"epoch": 0.3436619718309859,
"grad_norm": 0.5507818897049944,
"learning_rate": 9.993952112800314e-06,
"loss": 0.5796,
"step": 122
},
{
"epoch": 0.3464788732394366,
"grad_norm": 0.5350927421515935,
"learning_rate": 9.993119039513124e-06,
"loss": 0.5715,
"step": 123
},
{
"epoch": 0.3492957746478873,
"grad_norm": 0.5239546056575984,
"learning_rate": 9.992232270433239e-06,
"loss": 0.5771,
"step": 124
},
{
"epoch": 0.352112676056338,
"grad_norm": 0.5126764082447791,
"learning_rate": 9.991291815096941e-06,
"loss": 0.559,
"step": 125
},
{
"epoch": 0.35492957746478876,
"grad_norm": 0.4969718032918457,
"learning_rate": 9.990297683617844e-06,
"loss": 0.5678,
"step": 126
},
{
"epoch": 0.35774647887323946,
"grad_norm": 0.5483487768076548,
"learning_rate": 9.989249886686798e-06,
"loss": 0.5579,
"step": 127
},
{
"epoch": 0.36056338028169016,
"grad_norm": 0.4979281248814945,
"learning_rate": 9.988148435571766e-06,
"loss": 0.527,
"step": 128
},
{
"epoch": 0.36338028169014086,
"grad_norm": 0.5642480799900738,
"learning_rate": 9.98699334211771e-06,
"loss": 0.5692,
"step": 129
},
{
"epoch": 0.36619718309859156,
"grad_norm": 0.6026792858515811,
"learning_rate": 9.985784618746455e-06,
"loss": 0.5801,
"step": 130
},
{
"epoch": 0.36901408450704226,
"grad_norm": 0.5167698331604939,
"learning_rate": 9.984522278456558e-06,
"loss": 0.5539,
"step": 131
},
{
"epoch": 0.37183098591549296,
"grad_norm": 0.5817335775473784,
"learning_rate": 9.983206334823179e-06,
"loss": 0.5606,
"step": 132
},
{
"epoch": 0.37464788732394366,
"grad_norm": 0.5276842927516352,
"learning_rate": 9.981836801997916e-06,
"loss": 0.5722,
"step": 133
},
{
"epoch": 0.37746478873239436,
"grad_norm": 0.5912359974067817,
"learning_rate": 9.98041369470867e-06,
"loss": 0.5614,
"step": 134
},
{
"epoch": 0.38028169014084506,
"grad_norm": 0.5452723193931854,
"learning_rate": 9.978937028259477e-06,
"loss": 0.5515,
"step": 135
},
{
"epoch": 0.38309859154929576,
"grad_norm": 0.5325105100665509,
"learning_rate": 9.977406818530346e-06,
"loss": 0.5466,
"step": 136
},
{
"epoch": 0.38591549295774646,
"grad_norm": 0.5647180917474068,
"learning_rate": 9.975823081977089e-06,
"loss": 0.5624,
"step": 137
},
{
"epoch": 0.38873239436619716,
"grad_norm": 0.5723755319658607,
"learning_rate": 9.974185835631141e-06,
"loss": 0.5626,
"step": 138
},
{
"epoch": 0.39154929577464787,
"grad_norm": 0.5372783352103804,
"learning_rate": 9.972495097099379e-06,
"loss": 0.543,
"step": 139
},
{
"epoch": 0.39436619718309857,
"grad_norm": 0.5397037680212823,
"learning_rate": 9.970750884563938e-06,
"loss": 0.5648,
"step": 140
},
{
"epoch": 0.3971830985915493,
"grad_norm": 0.5552469428542458,
"learning_rate": 9.968953216782005e-06,
"loss": 0.5765,
"step": 141
},
{
"epoch": 0.4,
"grad_norm": 0.5577079940502304,
"learning_rate": 9.96710211308562e-06,
"loss": 0.5582,
"step": 142
},
{
"epoch": 0.4028169014084507,
"grad_norm": 0.5705003650495926,
"learning_rate": 9.965197593381483e-06,
"loss": 0.5666,
"step": 143
},
{
"epoch": 0.4056338028169014,
"grad_norm": 0.4444536833248146,
"learning_rate": 9.963239678150712e-06,
"loss": 0.5609,
"step": 144
},
{
"epoch": 0.4084507042253521,
"grad_norm": 0.6060222914963574,
"learning_rate": 9.961228388448648e-06,
"loss": 0.5603,
"step": 145
},
{
"epoch": 0.4112676056338028,
"grad_norm": 0.49513326472609975,
"learning_rate": 9.959163745904613e-06,
"loss": 0.547,
"step": 146
},
{
"epoch": 0.4140845070422535,
"grad_norm": 0.4629289638803288,
"learning_rate": 9.95704577272169e-06,
"loss": 0.5935,
"step": 147
},
{
"epoch": 0.4169014084507042,
"grad_norm": 0.5136703956308253,
"learning_rate": 9.954874491676474e-06,
"loss": 0.5784,
"step": 148
},
{
"epoch": 0.4197183098591549,
"grad_norm": 0.5669768244151758,
"learning_rate": 9.952649926118827e-06,
"loss": 0.5788,
"step": 149
},
{
"epoch": 0.4225352112676056,
"grad_norm": 0.4669028299546674,
"learning_rate": 9.950372099971635e-06,
"loss": 0.5484,
"step": 150
},
{
"epoch": 0.4253521126760563,
"grad_norm": 0.5798768078016783,
"learning_rate": 9.948041037730546e-06,
"loss": 0.5658,
"step": 151
},
{
"epoch": 0.428169014084507,
"grad_norm": 0.6319591077591975,
"learning_rate": 9.945656764463706e-06,
"loss": 0.5681,
"step": 152
},
{
"epoch": 0.4309859154929577,
"grad_norm": 0.5566876220468727,
"learning_rate": 9.943219305811486e-06,
"loss": 0.586,
"step": 153
},
{
"epoch": 0.43380281690140843,
"grad_norm": 0.6780781695543063,
"learning_rate": 9.940728687986219e-06,
"loss": 0.5678,
"step": 154
},
{
"epoch": 0.43661971830985913,
"grad_norm": 0.5376203276294301,
"learning_rate": 9.938184937771903e-06,
"loss": 0.5576,
"step": 155
},
{
"epoch": 0.4394366197183099,
"grad_norm": 0.589490601515025,
"learning_rate": 9.935588082523917e-06,
"loss": 0.5668,
"step": 156
},
{
"epoch": 0.4422535211267606,
"grad_norm": 0.5240178495748034,
"learning_rate": 9.93293815016874e-06,
"loss": 0.5819,
"step": 157
},
{
"epoch": 0.4450704225352113,
"grad_norm": 0.581734341309261,
"learning_rate": 9.930235169203628e-06,
"loss": 0.5341,
"step": 158
},
{
"epoch": 0.447887323943662,
"grad_norm": 0.4764372205838014,
"learning_rate": 9.927479168696327e-06,
"loss": 0.5507,
"step": 159
},
{
"epoch": 0.4507042253521127,
"grad_norm": 0.5983293375404063,
"learning_rate": 9.924670178284751e-06,
"loss": 0.6017,
"step": 160
},
{
"epoch": 0.4535211267605634,
"grad_norm": 0.4755074939422403,
"learning_rate": 9.921808228176667e-06,
"loss": 0.572,
"step": 161
},
{
"epoch": 0.4563380281690141,
"grad_norm": 0.549427169617322,
"learning_rate": 9.918893349149361e-06,
"loss": 0.5648,
"step": 162
},
{
"epoch": 0.4591549295774648,
"grad_norm": 0.5473994981650376,
"learning_rate": 9.915925572549325e-06,
"loss": 0.5795,
"step": 163
},
{
"epoch": 0.4619718309859155,
"grad_norm": 0.5344229046079484,
"learning_rate": 9.912904930291902e-06,
"loss": 0.5702,
"step": 164
},
{
"epoch": 0.4647887323943662,
"grad_norm": 0.5248831527184084,
"learning_rate": 9.909831454860953e-06,
"loss": 0.544,
"step": 165
},
{
"epoch": 0.4676056338028169,
"grad_norm": 0.5007058816288649,
"learning_rate": 9.906705179308503e-06,
"loss": 0.575,
"step": 166
},
{
"epoch": 0.4704225352112676,
"grad_norm": 0.6076422198886006,
"learning_rate": 9.903526137254386e-06,
"loss": 0.5637,
"step": 167
},
{
"epoch": 0.4732394366197183,
"grad_norm": 0.44954861939381563,
"learning_rate": 9.900294362885892e-06,
"loss": 0.5465,
"step": 168
},
{
"epoch": 0.476056338028169,
"grad_norm": 0.6136256906325439,
"learning_rate": 9.897009890957382e-06,
"loss": 0.5411,
"step": 169
},
{
"epoch": 0.4788732394366197,
"grad_norm": 0.49306626666710585,
"learning_rate": 9.893672756789933e-06,
"loss": 0.5695,
"step": 170
},
{
"epoch": 0.48169014084507045,
"grad_norm": 0.5638207147752434,
"learning_rate": 9.890282996270944e-06,
"loss": 0.562,
"step": 171
},
{
"epoch": 0.48450704225352115,
"grad_norm": 0.5844363211055491,
"learning_rate": 9.886840645853757e-06,
"loss": 0.5713,
"step": 172
},
{
"epoch": 0.48732394366197185,
"grad_norm": 0.4989681859492393,
"learning_rate": 9.883345742557265e-06,
"loss": 0.5184,
"step": 173
},
{
"epoch": 0.49014084507042255,
"grad_norm": 0.6798466666249082,
"learning_rate": 9.879798323965512e-06,
"loss": 0.5482,
"step": 174
},
{
"epoch": 0.49295774647887325,
"grad_norm": 0.5286425298758181,
"learning_rate": 9.876198428227288e-06,
"loss": 0.5738,
"step": 175
},
{
"epoch": 0.49577464788732395,
"grad_norm": 0.5070463361119862,
"learning_rate": 9.872546094055719e-06,
"loss": 0.559,
"step": 176
},
{
"epoch": 0.49859154929577465,
"grad_norm": 0.5974495055987761,
"learning_rate": 9.868841360727856e-06,
"loss": 0.5596,
"step": 177
},
{
"epoch": 0.5014084507042254,
"grad_norm": 0.5018737926842496,
"learning_rate": 9.865084268084247e-06,
"loss": 0.5409,
"step": 178
},
{
"epoch": 0.504225352112676,
"grad_norm": 0.657241380827331,
"learning_rate": 9.861274856528504e-06,
"loss": 0.5534,
"step": 179
},
{
"epoch": 0.5070422535211268,
"grad_norm": 0.5001455789690383,
"learning_rate": 9.857413167026881e-06,
"loss": 0.563,
"step": 180
},
{
"epoch": 0.5098591549295775,
"grad_norm": 0.6635297083955395,
"learning_rate": 9.853499241107827e-06,
"loss": 0.5501,
"step": 181
},
{
"epoch": 0.5126760563380282,
"grad_norm": 0.513999422537837,
"learning_rate": 9.849533120861537e-06,
"loss": 0.5573,
"step": 182
},
{
"epoch": 0.5154929577464789,
"grad_norm": 0.5481718260759659,
"learning_rate": 9.845514848939499e-06,
"loss": 0.5439,
"step": 183
},
{
"epoch": 0.5183098591549296,
"grad_norm": 0.6738586169727009,
"learning_rate": 9.841444468554043e-06,
"loss": 0.5747,
"step": 184
},
{
"epoch": 0.5211267605633803,
"grad_norm": 0.5984162328532678,
"learning_rate": 9.837322023477866e-06,
"loss": 0.5762,
"step": 185
},
{
"epoch": 0.523943661971831,
"grad_norm": 0.5506392735862736,
"learning_rate": 9.833147558043574e-06,
"loss": 0.5295,
"step": 186
},
{
"epoch": 0.5267605633802817,
"grad_norm": 0.6775281123264153,
"learning_rate": 9.82892111714319e-06,
"loss": 0.5666,
"step": 187
},
{
"epoch": 0.5295774647887324,
"grad_norm": 0.5471145431878144,
"learning_rate": 9.824642746227682e-06,
"loss": 0.566,
"step": 188
},
{
"epoch": 0.532394366197183,
"grad_norm": 0.7100217523678872,
"learning_rate": 9.820312491306471e-06,
"loss": 0.5637,
"step": 189
},
{
"epoch": 0.5352112676056338,
"grad_norm": 0.6057794786859599,
"learning_rate": 9.815930398946942e-06,
"loss": 0.5507,
"step": 190
},
{
"epoch": 0.5380281690140845,
"grad_norm": 0.626141535245459,
"learning_rate": 9.811496516273925e-06,
"loss": 0.5716,
"step": 191
},
{
"epoch": 0.5408450704225352,
"grad_norm": 0.5527022890637331,
"learning_rate": 9.807010890969214e-06,
"loss": 0.5532,
"step": 192
},
{
"epoch": 0.543661971830986,
"grad_norm": 0.6020033027406588,
"learning_rate": 9.80247357127103e-06,
"loss": 0.5889,
"step": 193
},
{
"epoch": 0.5464788732394367,
"grad_norm": 0.5199016296325637,
"learning_rate": 9.797884605973523e-06,
"loss": 0.5807,
"step": 194
},
{
"epoch": 0.5492957746478874,
"grad_norm": 0.4523520545534342,
"learning_rate": 9.793244044426233e-06,
"loss": 0.5514,
"step": 195
},
{
"epoch": 0.5521126760563381,
"grad_norm": 0.5379175425838836,
"learning_rate": 9.788551936533561e-06,
"loss": 0.5444,
"step": 196
},
{
"epoch": 0.5549295774647888,
"grad_norm": 0.48732646620471026,
"learning_rate": 9.783808332754242e-06,
"loss": 0.5315,
"step": 197
},
{
"epoch": 0.5577464788732395,
"grad_norm": 0.5557329022733009,
"learning_rate": 9.779013284100791e-06,
"loss": 0.5501,
"step": 198
},
{
"epoch": 0.5605633802816902,
"grad_norm": 0.5373986677295377,
"learning_rate": 9.77416684213896e-06,
"loss": 0.5719,
"step": 199
},
{
"epoch": 0.5633802816901409,
"grad_norm": 0.5919610806064748,
"learning_rate": 9.769269058987181e-06,
"loss": 0.5202,
"step": 200
},
{
"epoch": 0.5661971830985916,
"grad_norm": 0.51134408429211,
"learning_rate": 9.76431998731601e-06,
"loss": 0.5474,
"step": 201
},
{
"epoch": 0.5690140845070423,
"grad_norm": 0.5283277616415307,
"learning_rate": 9.759319680347556e-06,
"loss": 0.5476,
"step": 202
},
{
"epoch": 0.571830985915493,
"grad_norm": 0.5803790463856993,
"learning_rate": 9.75426819185491e-06,
"loss": 0.5542,
"step": 203
},
{
"epoch": 0.5746478873239437,
"grad_norm": 0.5817083296083267,
"learning_rate": 9.74916557616157e-06,
"loss": 0.5365,
"step": 204
},
{
"epoch": 0.5774647887323944,
"grad_norm": 0.4771789889476685,
"learning_rate": 9.744011888140847e-06,
"loss": 0.5335,
"step": 205
},
{
"epoch": 0.5802816901408451,
"grad_norm": 0.59366389105907,
"learning_rate": 9.738807183215288e-06,
"loss": 0.5372,
"step": 206
},
{
"epoch": 0.5830985915492958,
"grad_norm": 0.5577403514252908,
"learning_rate": 9.73355151735607e-06,
"loss": 0.5362,
"step": 207
},
{
"epoch": 0.5859154929577465,
"grad_norm": 0.47605232856572827,
"learning_rate": 9.728244947082405e-06,
"loss": 0.5572,
"step": 208
},
{
"epoch": 0.5887323943661972,
"grad_norm": 0.7070690743472441,
"learning_rate": 9.722887529460928e-06,
"loss": 0.548,
"step": 209
},
{
"epoch": 0.5915492957746479,
"grad_norm": 0.5624040595146416,
"learning_rate": 9.717479322105079e-06,
"loss": 0.5444,
"step": 210
},
{
"epoch": 0.5943661971830986,
"grad_norm": 0.5859424494783367,
"learning_rate": 9.712020383174496e-06,
"loss": 0.5249,
"step": 211
},
{
"epoch": 0.5971830985915493,
"grad_norm": 0.6678507938439178,
"learning_rate": 9.70651077137438e-06,
"loss": 0.533,
"step": 212
},
{
"epoch": 0.6,
"grad_norm": 0.5328470708977834,
"learning_rate": 9.700950545954867e-06,
"loss": 0.5833,
"step": 213
},
{
"epoch": 0.6028169014084507,
"grad_norm": 0.5362441571989329,
"learning_rate": 9.695339766710382e-06,
"loss": 0.5396,
"step": 214
},
{
"epoch": 0.6056338028169014,
"grad_norm": 0.6434345591744017,
"learning_rate": 9.689678493979011e-06,
"loss": 0.5487,
"step": 215
},
{
"epoch": 0.6084507042253521,
"grad_norm": 0.5167603706269648,
"learning_rate": 9.683966788641848e-06,
"loss": 0.5481,
"step": 216
},
{
"epoch": 0.6112676056338028,
"grad_norm": 0.4686781372062126,
"learning_rate": 9.678204712122328e-06,
"loss": 0.5419,
"step": 217
},
{
"epoch": 0.6140845070422535,
"grad_norm": 0.5760752748606316,
"learning_rate": 9.672392326385582e-06,
"loss": 0.557,
"step": 218
},
{
"epoch": 0.6169014084507042,
"grad_norm": 0.6026666041248488,
"learning_rate": 9.666529693937763e-06,
"loss": 0.5529,
"step": 219
},
{
"epoch": 0.6197183098591549,
"grad_norm": 0.5451736184177465,
"learning_rate": 9.660616877825374e-06,
"loss": 0.5361,
"step": 220
},
{
"epoch": 0.6225352112676056,
"grad_norm": 0.5116565555731337,
"learning_rate": 9.65465394163459e-06,
"loss": 0.5087,
"step": 221
},
{
"epoch": 0.6253521126760564,
"grad_norm": 0.6065841927093307,
"learning_rate": 9.64864094949058e-06,
"loss": 0.5465,
"step": 222
},
{
"epoch": 0.6281690140845071,
"grad_norm": 0.5589673127927728,
"learning_rate": 9.642577966056806e-06,
"loss": 0.5148,
"step": 223
},
{
"epoch": 0.6309859154929578,
"grad_norm": 0.5585400710757196,
"learning_rate": 9.636465056534337e-06,
"loss": 0.5267,
"step": 224
},
{
"epoch": 0.6338028169014085,
"grad_norm": 0.574813185908209,
"learning_rate": 9.630302286661149e-06,
"loss": 0.5357,
"step": 225
},
{
"epoch": 0.6366197183098592,
"grad_norm": 0.5191579985554721,
"learning_rate": 9.624089722711409e-06,
"loss": 0.5426,
"step": 226
},
{
"epoch": 0.6394366197183099,
"grad_norm": 0.6688929345165525,
"learning_rate": 9.617827431494769e-06,
"loss": 0.5524,
"step": 227
},
{
"epoch": 0.6422535211267606,
"grad_norm": 0.46805258630886243,
"learning_rate": 9.611515480355644e-06,
"loss": 0.5499,
"step": 228
},
{
"epoch": 0.6450704225352113,
"grad_norm": 0.6144905582687046,
"learning_rate": 9.605153937172495e-06,
"loss": 0.5632,
"step": 229
},
{
"epoch": 0.647887323943662,
"grad_norm": 0.52683034829998,
"learning_rate": 9.598742870357089e-06,
"loss": 0.5555,
"step": 230
},
{
"epoch": 0.6507042253521127,
"grad_norm": 0.603840703034481,
"learning_rate": 9.592282348853772e-06,
"loss": 0.5448,
"step": 231
},
{
"epoch": 0.6535211267605634,
"grad_norm": 0.5800521717117073,
"learning_rate": 9.585772442138717e-06,
"loss": 0.5451,
"step": 232
},
{
"epoch": 0.6563380281690141,
"grad_norm": 0.5500966334452324,
"learning_rate": 9.57921322021919e-06,
"loss": 0.5268,
"step": 233
},
{
"epoch": 0.6591549295774648,
"grad_norm": 0.5570376537579396,
"learning_rate": 9.572604753632789e-06,
"loss": 0.5255,
"step": 234
},
{
"epoch": 0.6619718309859155,
"grad_norm": 0.47661383694647175,
"learning_rate": 9.565947113446685e-06,
"loss": 0.5591,
"step": 235
},
{
"epoch": 0.6647887323943662,
"grad_norm": 0.5445391735613945,
"learning_rate": 9.559240371256861e-06,
"loss": 0.5448,
"step": 236
},
{
"epoch": 0.6676056338028169,
"grad_norm": 0.5194324455486823,
"learning_rate": 9.552484599187344e-06,
"loss": 0.5645,
"step": 237
},
{
"epoch": 0.6704225352112676,
"grad_norm": 0.49906141175236074,
"learning_rate": 9.545679869889422e-06,
"loss": 0.5356,
"step": 238
},
{
"epoch": 0.6732394366197183,
"grad_norm": 0.44801193098795,
"learning_rate": 9.538826256540866e-06,
"loss": 0.5525,
"step": 239
},
{
"epoch": 0.676056338028169,
"grad_norm": 0.4950063463823408,
"learning_rate": 9.53192383284515e-06,
"loss": 0.5251,
"step": 240
},
{
"epoch": 0.6788732394366197,
"grad_norm": 0.4655702750763721,
"learning_rate": 9.524972673030646e-06,
"loss": 0.5532,
"step": 241
},
{
"epoch": 0.6816901408450704,
"grad_norm": 0.5015107720315597,
"learning_rate": 9.517972851849837e-06,
"loss": 0.5363,
"step": 242
},
{
"epoch": 0.6845070422535211,
"grad_norm": 0.4874041712602739,
"learning_rate": 9.510924444578505e-06,
"loss": 0.5138,
"step": 243
},
{
"epoch": 0.6873239436619718,
"grad_norm": 0.5318082437227483,
"learning_rate": 9.503827527014927e-06,
"loss": 0.5568,
"step": 244
},
{
"epoch": 0.6901408450704225,
"grad_norm": 0.5365279955696896,
"learning_rate": 9.496682175479058e-06,
"loss": 0.5568,
"step": 245
},
{
"epoch": 0.6929577464788732,
"grad_norm": 0.4650608267010065,
"learning_rate": 9.489488466811706e-06,
"loss": 0.5294,
"step": 246
},
{
"epoch": 0.6957746478873239,
"grad_norm": 0.502326692197303,
"learning_rate": 9.482246478373713e-06,
"loss": 0.5622,
"step": 247
},
{
"epoch": 0.6985915492957746,
"grad_norm": 0.5139235141970571,
"learning_rate": 9.47495628804512e-06,
"loss": 0.5309,
"step": 248
},
{
"epoch": 0.7014084507042253,
"grad_norm": 0.49082713927721994,
"learning_rate": 9.467617974224326e-06,
"loss": 0.5612,
"step": 249
},
{
"epoch": 0.704225352112676,
"grad_norm": 0.4368824557932469,
"learning_rate": 9.460231615827254e-06,
"loss": 0.544,
"step": 250
},
{
"epoch": 0.7070422535211267,
"grad_norm": 0.4590552453029466,
"learning_rate": 9.452797292286486e-06,
"loss": 0.5403,
"step": 251
},
{
"epoch": 0.7098591549295775,
"grad_norm": 0.4679280855458366,
"learning_rate": 9.445315083550432e-06,
"loss": 0.5213,
"step": 252
},
{
"epoch": 0.7126760563380282,
"grad_norm": 0.45716682819250093,
"learning_rate": 9.437785070082448e-06,
"loss": 0.5486,
"step": 253
},
{
"epoch": 0.7154929577464789,
"grad_norm": 0.5570183710120886,
"learning_rate": 9.430207332859981e-06,
"loss": 0.5501,
"step": 254
},
{
"epoch": 0.7183098591549296,
"grad_norm": 0.4245724220850406,
"learning_rate": 9.422581953373704e-06,
"loss": 0.5356,
"step": 255
},
{
"epoch": 0.7211267605633803,
"grad_norm": 0.48956473422005187,
"learning_rate": 9.414909013626624e-06,
"loss": 0.5313,
"step": 256
},
{
"epoch": 0.723943661971831,
"grad_norm": 0.4582929162313386,
"learning_rate": 9.407188596133212e-06,
"loss": 0.5463,
"step": 257
},
{
"epoch": 0.7267605633802817,
"grad_norm": 0.5115654013081172,
"learning_rate": 9.399420783918518e-06,
"loss": 0.5535,
"step": 258
},
{
"epoch": 0.7295774647887324,
"grad_norm": 0.5269407638629522,
"learning_rate": 9.391605660517268e-06,
"loss": 0.5597,
"step": 259
},
{
"epoch": 0.7323943661971831,
"grad_norm": 0.5438385236535639,
"learning_rate": 9.38374330997297e-06,
"loss": 0.5388,
"step": 260
},
{
"epoch": 0.7352112676056338,
"grad_norm": 0.43926987601895895,
"learning_rate": 9.375833816837012e-06,
"loss": 0.5412,
"step": 261
},
{
"epoch": 0.7380281690140845,
"grad_norm": 0.5836211324705879,
"learning_rate": 9.367877266167752e-06,
"loss": 0.5688,
"step": 262
},
{
"epoch": 0.7408450704225352,
"grad_norm": 0.5045079775368958,
"learning_rate": 9.3598737435296e-06,
"loss": 0.5347,
"step": 263
},
{
"epoch": 0.7436619718309859,
"grad_norm": 0.44338592065262517,
"learning_rate": 9.351823334992105e-06,
"loss": 0.5168,
"step": 264
},
{
"epoch": 0.7464788732394366,
"grad_norm": 0.5235306342556597,
"learning_rate": 9.343726127129023e-06,
"loss": 0.5604,
"step": 265
},
{
"epoch": 0.7492957746478873,
"grad_norm": 0.5172376233558552,
"learning_rate": 9.335582207017385e-06,
"loss": 0.534,
"step": 266
},
{
"epoch": 0.752112676056338,
"grad_norm": 0.5180203870423422,
"learning_rate": 9.327391662236569e-06,
"loss": 0.5379,
"step": 267
},
{
"epoch": 0.7549295774647887,
"grad_norm": 0.4804092941634994,
"learning_rate": 9.31915458086735e-06,
"loss": 0.5052,
"step": 268
},
{
"epoch": 0.7577464788732394,
"grad_norm": 0.46979320817265685,
"learning_rate": 9.310871051490953e-06,
"loss": 0.5295,
"step": 269
},
{
"epoch": 0.7605633802816901,
"grad_norm": 0.6685486725224626,
"learning_rate": 9.302541163188107e-06,
"loss": 0.5358,
"step": 270
},
{
"epoch": 0.7633802816901408,
"grad_norm": 0.4220043036826099,
"learning_rate": 9.294165005538083e-06,
"loss": 0.5597,
"step": 271
},
{
"epoch": 0.7661971830985915,
"grad_norm": 0.5240690708886699,
"learning_rate": 9.285742668617725e-06,
"loss": 0.5429,
"step": 272
},
{
"epoch": 0.7690140845070422,
"grad_norm": 0.5225378668342543,
"learning_rate": 9.277274243000495e-06,
"loss": 0.5314,
"step": 273
},
{
"epoch": 0.7718309859154929,
"grad_norm": 0.5089658349598899,
"learning_rate": 9.268759819755484e-06,
"loss": 0.5404,
"step": 274
},
{
"epoch": 0.7746478873239436,
"grad_norm": 0.5081863591984968,
"learning_rate": 9.260199490446444e-06,
"loss": 0.5436,
"step": 275
},
{
"epoch": 0.7774647887323943,
"grad_norm": 0.5291957024091261,
"learning_rate": 9.251593347130796e-06,
"loss": 0.5459,
"step": 276
},
{
"epoch": 0.780281690140845,
"grad_norm": 0.5546316222295684,
"learning_rate": 9.242941482358646e-06,
"loss": 0.5638,
"step": 277
},
{
"epoch": 0.7830985915492957,
"grad_norm": 0.493107854299303,
"learning_rate": 9.234243989171784e-06,
"loss": 0.5163,
"step": 278
},
{
"epoch": 0.7859154929577464,
"grad_norm": 0.504054739961812,
"learning_rate": 9.225500961102685e-06,
"loss": 0.5403,
"step": 279
},
{
"epoch": 0.7887323943661971,
"grad_norm": 0.46644848784294396,
"learning_rate": 9.216712492173509e-06,
"loss": 0.5319,
"step": 280
},
{
"epoch": 0.7915492957746478,
"grad_norm": 0.48875850819222605,
"learning_rate": 9.20787867689508e-06,
"loss": 0.55,
"step": 281
},
{
"epoch": 0.7943661971830986,
"grad_norm": 0.45374233893805643,
"learning_rate": 9.19899961026588e-06,
"loss": 0.5381,
"step": 282
},
{
"epoch": 0.7971830985915493,
"grad_norm": 0.49081212722421264,
"learning_rate": 9.190075387771014e-06,
"loss": 0.496,
"step": 283
},
{
"epoch": 0.8,
"grad_norm": 0.4898635707497967,
"learning_rate": 9.181106105381201e-06,
"loss": 0.5469,
"step": 284
},
{
"epoch": 0.8028169014084507,
"grad_norm": 0.5742626781139881,
"learning_rate": 9.172091859551727e-06,
"loss": 0.5504,
"step": 285
},
{
"epoch": 0.8056338028169014,
"grad_norm": 0.48424294060741285,
"learning_rate": 9.16303274722141e-06,
"loss": 0.5447,
"step": 286
},
{
"epoch": 0.8084507042253521,
"grad_norm": 0.41774263725909944,
"learning_rate": 9.153928865811567e-06,
"loss": 0.5065,
"step": 287
},
{
"epoch": 0.8112676056338028,
"grad_norm": 0.5366510861069486,
"learning_rate": 9.144780313224955e-06,
"loss": 0.5274,
"step": 288
},
{
"epoch": 0.8140845070422535,
"grad_norm": 0.46697982171080715,
"learning_rate": 9.135587187844727e-06,
"loss": 0.5023,
"step": 289
},
{
"epoch": 0.8169014084507042,
"grad_norm": 0.48443064321010365,
"learning_rate": 9.126349588533367e-06,
"loss": 0.5373,
"step": 290
},
{
"epoch": 0.819718309859155,
"grad_norm": 0.5135389257533832,
"learning_rate": 9.117067614631629e-06,
"loss": 0.5501,
"step": 291
},
{
"epoch": 0.8225352112676056,
"grad_norm": 0.4983544795249396,
"learning_rate": 9.107741365957473e-06,
"loss": 0.52,
"step": 292
},
{
"epoch": 0.8253521126760563,
"grad_norm": 0.474169348982468,
"learning_rate": 9.098370942804984e-06,
"loss": 0.5522,
"step": 293
},
{
"epoch": 0.828169014084507,
"grad_norm": 0.553807859933943,
"learning_rate": 9.0889564459433e-06,
"loss": 0.546,
"step": 294
},
{
"epoch": 0.8309859154929577,
"grad_norm": 0.521288457346182,
"learning_rate": 9.079497976615527e-06,
"loss": 0.5039,
"step": 295
},
{
"epoch": 0.8338028169014085,
"grad_norm": 0.42015807275115125,
"learning_rate": 9.069995636537646e-06,
"loss": 0.5077,
"step": 296
},
{
"epoch": 0.8366197183098592,
"grad_norm": 0.5583730579511402,
"learning_rate": 9.060449527897424e-06,
"loss": 0.5588,
"step": 297
},
{
"epoch": 0.8394366197183099,
"grad_norm": 0.5542124651392272,
"learning_rate": 9.050859753353312e-06,
"loss": 0.5217,
"step": 298
},
{
"epoch": 0.8422535211267606,
"grad_norm": 0.48602364869365977,
"learning_rate": 9.041226416033344e-06,
"loss": 0.54,
"step": 299
},
{
"epoch": 0.8450704225352113,
"grad_norm": 0.653714795345656,
"learning_rate": 9.031549619534025e-06,
"loss": 0.5497,
"step": 300
},
{
"epoch": 0.847887323943662,
"grad_norm": 0.4800359394630375,
"learning_rate": 9.021829467919218e-06,
"loss": 0.5445,
"step": 301
},
{
"epoch": 0.8507042253521127,
"grad_norm": 0.48729987333610464,
"learning_rate": 9.012066065719026e-06,
"loss": 0.5473,
"step": 302
},
{
"epoch": 0.8535211267605634,
"grad_norm": 0.6013760409344134,
"learning_rate": 9.00225951792867e-06,
"loss": 0.5399,
"step": 303
},
{
"epoch": 0.856338028169014,
"grad_norm": 0.4955229451079562,
"learning_rate": 8.99240993000735e-06,
"loss": 0.5373,
"step": 304
},
{
"epoch": 0.8591549295774648,
"grad_norm": 0.5057254756870588,
"learning_rate": 8.98251740787712e-06,
"loss": 0.5308,
"step": 305
},
{
"epoch": 0.8619718309859155,
"grad_norm": 0.5062350967048924,
"learning_rate": 8.972582057921753e-06,
"loss": 0.5379,
"step": 306
},
{
"epoch": 0.8647887323943662,
"grad_norm": 0.5149008679939511,
"learning_rate": 8.962603986985582e-06,
"loss": 0.5447,
"step": 307
},
{
"epoch": 0.8676056338028169,
"grad_norm": 0.5138417832440646,
"learning_rate": 8.952583302372361e-06,
"loss": 0.541,
"step": 308
},
{
"epoch": 0.8704225352112676,
"grad_norm": 0.4988758497493651,
"learning_rate": 8.942520111844117e-06,
"loss": 0.5456,
"step": 309
},
{
"epoch": 0.8732394366197183,
"grad_norm": 0.4908248371860937,
"learning_rate": 8.932414523619979e-06,
"loss": 0.5417,
"step": 310
},
{
"epoch": 0.8760563380281691,
"grad_norm": 0.47467152147405195,
"learning_rate": 8.922266646375012e-06,
"loss": 0.5369,
"step": 311
},
{
"epoch": 0.8788732394366198,
"grad_norm": 0.5674377586617675,
"learning_rate": 8.912076589239071e-06,
"loss": 0.5408,
"step": 312
},
{
"epoch": 0.8816901408450705,
"grad_norm": 0.5006532529649368,
"learning_rate": 8.901844461795597e-06,
"loss": 0.5355,
"step": 313
},
{
"epoch": 0.8845070422535212,
"grad_norm": 0.44995769221377135,
"learning_rate": 8.89157037408046e-06,
"loss": 0.5471,
"step": 314
},
{
"epoch": 0.8873239436619719,
"grad_norm": 0.5082076772096817,
"learning_rate": 8.881254436580771e-06,
"loss": 0.5141,
"step": 315
},
{
"epoch": 0.8901408450704226,
"grad_norm": 0.4979524931446426,
"learning_rate": 8.870896760233687e-06,
"loss": 0.5183,
"step": 316
},
{
"epoch": 0.8929577464788733,
"grad_norm": 0.5033006044158947,
"learning_rate": 8.860497456425226e-06,
"loss": 0.539,
"step": 317
},
{
"epoch": 0.895774647887324,
"grad_norm": 0.5558963290415088,
"learning_rate": 8.850056636989064e-06,
"loss": 0.5214,
"step": 318
},
{
"epoch": 0.8985915492957747,
"grad_norm": 0.528178628352138,
"learning_rate": 8.839574414205335e-06,
"loss": 0.5308,
"step": 319
},
{
"epoch": 0.9014084507042254,
"grad_norm": 0.5988043719853473,
"learning_rate": 8.829050900799426e-06,
"loss": 0.5584,
"step": 320
},
{
"epoch": 0.9042253521126761,
"grad_norm": 0.6178729360079042,
"learning_rate": 8.818486209940754e-06,
"loss": 0.5439,
"step": 321
},
{
"epoch": 0.9070422535211268,
"grad_norm": 0.4727007397282943,
"learning_rate": 8.807880455241562e-06,
"loss": 0.5208,
"step": 322
},
{
"epoch": 0.9098591549295775,
"grad_norm": 0.5865056586686292,
"learning_rate": 8.797233750755696e-06,
"loss": 0.5293,
"step": 323
},
{
"epoch": 0.9126760563380282,
"grad_norm": 0.527614643499507,
"learning_rate": 8.786546210977364e-06,
"loss": 0.51,
"step": 324
},
{
"epoch": 0.9154929577464789,
"grad_norm": 0.5253730216506938,
"learning_rate": 8.775817950839925e-06,
"loss": 0.5464,
"step": 325
},
{
"epoch": 0.9183098591549296,
"grad_norm": 0.5120572114997892,
"learning_rate": 8.765049085714634e-06,
"loss": 0.522,
"step": 326
},
{
"epoch": 0.9211267605633803,
"grad_norm": 0.5579538329061854,
"learning_rate": 8.754239731409418e-06,
"loss": 0.515,
"step": 327
},
{
"epoch": 0.923943661971831,
"grad_norm": 0.4695513820418862,
"learning_rate": 8.743390004167618e-06,
"loss": 0.528,
"step": 328
},
{
"epoch": 0.9267605633802817,
"grad_norm": 0.4744094042874529,
"learning_rate": 8.732500020666748e-06,
"loss": 0.5032,
"step": 329
},
{
"epoch": 0.9295774647887324,
"grad_norm": 0.48310233510167816,
"learning_rate": 8.72156989801723e-06,
"loss": 0.515,
"step": 330
},
{
"epoch": 0.9323943661971831,
"grad_norm": 0.5362590571822845,
"learning_rate": 8.710599753761148e-06,
"loss": 0.5412,
"step": 331
},
{
"epoch": 0.9352112676056338,
"grad_norm": 0.43044803610258253,
"learning_rate": 8.699589705870972e-06,
"loss": 0.5252,
"step": 332
},
{
"epoch": 0.9380281690140845,
"grad_norm": 0.5529255675552455,
"learning_rate": 8.688539872748293e-06,
"loss": 0.554,
"step": 333
},
{
"epoch": 0.9408450704225352,
"grad_norm": 0.4697670392983979,
"learning_rate": 8.677450373222555e-06,
"loss": 0.5249,
"step": 334
},
{
"epoch": 0.9436619718309859,
"grad_norm": 0.466397415688367,
"learning_rate": 8.666321326549771e-06,
"loss": 0.5291,
"step": 335
},
{
"epoch": 0.9464788732394366,
"grad_norm": 0.47431096436255926,
"learning_rate": 8.655152852411242e-06,
"loss": 0.5262,
"step": 336
},
{
"epoch": 0.9492957746478873,
"grad_norm": 0.42290856961673545,
"learning_rate": 8.643945070912269e-06,
"loss": 0.5437,
"step": 337
},
{
"epoch": 0.952112676056338,
"grad_norm": 0.5067980558253683,
"learning_rate": 8.632698102580866e-06,
"loss": 0.5487,
"step": 338
},
{
"epoch": 0.9549295774647887,
"grad_norm": 0.49756235790217707,
"learning_rate": 8.621412068366455e-06,
"loss": 0.5411,
"step": 339
},
{
"epoch": 0.9577464788732394,
"grad_norm": 0.5339747238457127,
"learning_rate": 8.61008708963858e-06,
"loss": 0.5336,
"step": 340
},
{
"epoch": 0.9605633802816902,
"grad_norm": 0.5258263211335326,
"learning_rate": 8.598723288185582e-06,
"loss": 0.5576,
"step": 341
},
{
"epoch": 0.9633802816901409,
"grad_norm": 0.4474885925823666,
"learning_rate": 8.587320786213308e-06,
"loss": 0.5469,
"step": 342
},
{
"epoch": 0.9661971830985916,
"grad_norm": 0.4582164725544087,
"learning_rate": 8.575879706343786e-06,
"loss": 0.5258,
"step": 343
},
{
"epoch": 0.9690140845070423,
"grad_norm": 0.5353459619126049,
"learning_rate": 8.564400171613907e-06,
"loss": 0.5393,
"step": 344
},
{
"epoch": 0.971830985915493,
"grad_norm": 0.5551427263378275,
"learning_rate": 8.552882305474106e-06,
"loss": 0.511,
"step": 345
},
{
"epoch": 0.9746478873239437,
"grad_norm": 0.46371192783441034,
"learning_rate": 8.541326231787036e-06,
"loss": 0.5562,
"step": 346
},
{
"epoch": 0.9774647887323944,
"grad_norm": 0.49791553086908497,
"learning_rate": 8.529732074826225e-06,
"loss": 0.5317,
"step": 347
},
{
"epoch": 0.9802816901408451,
"grad_norm": 0.48902813408872026,
"learning_rate": 8.518099959274751e-06,
"loss": 0.5368,
"step": 348
},
{
"epoch": 0.9830985915492958,
"grad_norm": 0.5196746962677518,
"learning_rate": 8.506430010223899e-06,
"loss": 0.5328,
"step": 349
},
{
"epoch": 0.9859154929577465,
"grad_norm": 0.512561887896862,
"learning_rate": 8.494722353171807e-06,
"loss": 0.5631,
"step": 350
},
{
"epoch": 0.9887323943661972,
"grad_norm": 0.4688614656001333,
"learning_rate": 8.482977114022133e-06,
"loss": 0.5215,
"step": 351
},
{
"epoch": 0.9915492957746479,
"grad_norm": 0.4815835702147859,
"learning_rate": 8.471194419082683e-06,
"loss": 0.5175,
"step": 352
},
{
"epoch": 0.9943661971830986,
"grad_norm": 0.494451086453971,
"learning_rate": 8.459374395064066e-06,
"loss": 0.5347,
"step": 353
},
{
"epoch": 0.9971830985915493,
"grad_norm": 0.5479082871227343,
"learning_rate": 8.447517169078322e-06,
"loss": 0.5342,
"step": 354
},
{
"epoch": 1.0,
"grad_norm": 0.48128318573777745,
"learning_rate": 8.435622868637562e-06,
"loss": 0.5185,
"step": 355
},
{
"epoch": 1.0028169014084507,
"grad_norm": 0.606217938961181,
"learning_rate": 8.423691621652597e-06,
"loss": 0.5392,
"step": 356
},
{
"epoch": 1.0056338028169014,
"grad_norm": 0.5552799943960874,
"learning_rate": 8.411723556431555e-06,
"loss": 0.4923,
"step": 357
},
{
"epoch": 1.008450704225352,
"grad_norm": 0.4737481782588327,
"learning_rate": 8.399718801678507e-06,
"loss": 0.5031,
"step": 358
},
{
"epoch": 1.0112676056338028,
"grad_norm": 0.48881408780555036,
"learning_rate": 8.38767748649208e-06,
"loss": 0.4966,
"step": 359
},
{
"epoch": 1.0140845070422535,
"grad_norm": 0.6036514656317372,
"learning_rate": 8.375599740364076e-06,
"loss": 0.5054,
"step": 360
},
{
"epoch": 1.0169014084507042,
"grad_norm": 0.49847239674819194,
"learning_rate": 8.363485693178068e-06,
"loss": 0.4762,
"step": 361
},
{
"epoch": 1.019718309859155,
"grad_norm": 0.48893728544811266,
"learning_rate": 8.351335475208013e-06,
"loss": 0.4988,
"step": 362
},
{
"epoch": 1.0225352112676056,
"grad_norm": 0.5275361671421374,
"learning_rate": 8.339149217116844e-06,
"loss": 0.5181,
"step": 363
},
{
"epoch": 1.0253521126760563,
"grad_norm": 0.5699229385448366,
"learning_rate": 8.32692704995507e-06,
"loss": 0.4857,
"step": 364
},
{
"epoch": 1.028169014084507,
"grad_norm": 0.5300003505651588,
"learning_rate": 8.314669105159363e-06,
"loss": 0.5019,
"step": 365
},
{
"epoch": 1.0309859154929577,
"grad_norm": 0.47743877846606136,
"learning_rate": 8.302375514551147e-06,
"loss": 0.5031,
"step": 366
},
{
"epoch": 1.0338028169014084,
"grad_norm": 0.46610287314823595,
"learning_rate": 8.29004641033518e-06,
"loss": 0.484,
"step": 367
},
{
"epoch": 1.036619718309859,
"grad_norm": 0.5070608182387246,
"learning_rate": 8.277681925098133e-06,
"loss": 0.4919,
"step": 368
},
{
"epoch": 1.0394366197183098,
"grad_norm": 0.5155487269874641,
"learning_rate": 8.26528219180716e-06,
"loss": 0.4964,
"step": 369
},
{
"epoch": 1.0422535211267605,
"grad_norm": 0.48047795719589964,
"learning_rate": 8.252847343808473e-06,
"loss": 0.51,
"step": 370
},
{
"epoch": 1.0450704225352112,
"grad_norm": 0.5029419422977898,
"learning_rate": 8.240377514825906e-06,
"loss": 0.5096,
"step": 371
},
{
"epoch": 1.047887323943662,
"grad_norm": 0.44300572811738526,
"learning_rate": 8.227872838959478e-06,
"loss": 0.5085,
"step": 372
},
{
"epoch": 1.0507042253521126,
"grad_norm": 0.46826085995665145,
"learning_rate": 8.215333450683945e-06,
"loss": 0.4927,
"step": 373
},
{
"epoch": 1.0535211267605633,
"grad_norm": 0.4992785109969445,
"learning_rate": 8.202759484847366e-06,
"loss": 0.5237,
"step": 374
},
{
"epoch": 1.056338028169014,
"grad_norm": 0.4583536646462298,
"learning_rate": 8.19015107666964e-06,
"loss": 0.4778,
"step": 375
},
{
"epoch": 1.0591549295774647,
"grad_norm": 0.4629467013193756,
"learning_rate": 8.177508361741063e-06,
"loss": 0.5039,
"step": 376
},
{
"epoch": 1.0619718309859154,
"grad_norm": 0.513616475500221,
"learning_rate": 8.164831476020856e-06,
"loss": 0.4902,
"step": 377
},
{
"epoch": 1.064788732394366,
"grad_norm": 0.4564209134623684,
"learning_rate": 8.15212055583572e-06,
"loss": 0.4807,
"step": 378
},
{
"epoch": 1.0676056338028168,
"grad_norm": 0.5148849912441339,
"learning_rate": 8.139375737878356e-06,
"loss": 0.519,
"step": 379
},
{
"epoch": 1.0704225352112675,
"grad_norm": 0.48962484473993256,
"learning_rate": 8.126597159206002e-06,
"loss": 0.4926,
"step": 380
},
{
"epoch": 1.0732394366197182,
"grad_norm": 0.558804039498056,
"learning_rate": 8.113784957238957e-06,
"loss": 0.4954,
"step": 381
},
{
"epoch": 1.076056338028169,
"grad_norm": 0.5328343330334975,
"learning_rate": 8.100939269759103e-06,
"loss": 0.5051,
"step": 382
},
{
"epoch": 1.0788732394366196,
"grad_norm": 0.5293202978458649,
"learning_rate": 8.088060234908425e-06,
"loss": 0.5173,
"step": 383
},
{
"epoch": 1.0816901408450703,
"grad_norm": 0.4837225369393582,
"learning_rate": 8.075147991187521e-06,
"loss": 0.4951,
"step": 384
},
{
"epoch": 1.084507042253521,
"grad_norm": 0.5444025643060015,
"learning_rate": 8.062202677454123e-06,
"loss": 0.4944,
"step": 385
},
{
"epoch": 1.0873239436619717,
"grad_norm": 0.5058337046267952,
"learning_rate": 8.04922443292159e-06,
"loss": 0.4799,
"step": 386
},
{
"epoch": 1.0901408450704226,
"grad_norm": 0.4803075581095113,
"learning_rate": 8.036213397157418e-06,
"loss": 0.4983,
"step": 387
},
{
"epoch": 1.0929577464788733,
"grad_norm": 0.5913188229351856,
"learning_rate": 8.02316971008174e-06,
"loss": 0.493,
"step": 388
},
{
"epoch": 1.095774647887324,
"grad_norm": 0.5178060100858961,
"learning_rate": 8.01009351196582e-06,
"loss": 0.4793,
"step": 389
},
{
"epoch": 1.0985915492957747,
"grad_norm": 0.45957843759530875,
"learning_rate": 7.996984943430544e-06,
"loss": 0.4802,
"step": 390
},
{
"epoch": 1.1014084507042254,
"grad_norm": 0.49107943563289536,
"learning_rate": 7.983844145444908e-06,
"loss": 0.4638,
"step": 391
},
{
"epoch": 1.1042253521126761,
"grad_norm": 0.5865391082035357,
"learning_rate": 7.970671259324502e-06,
"loss": 0.4934,
"step": 392
},
{
"epoch": 1.1070422535211268,
"grad_norm": 0.46382700522033943,
"learning_rate": 7.957466426729995e-06,
"loss": 0.507,
"step": 393
},
{
"epoch": 1.1098591549295775,
"grad_norm": 0.5235903123213329,
"learning_rate": 7.944229789665595e-06,
"loss": 0.5016,
"step": 394
},
{
"epoch": 1.1126760563380282,
"grad_norm": 0.531704325443622,
"learning_rate": 7.930961490477546e-06,
"loss": 0.5434,
"step": 395
},
{
"epoch": 1.115492957746479,
"grad_norm": 0.4854203383807877,
"learning_rate": 7.917661671852582e-06,
"loss": 0.4627,
"step": 396
},
{
"epoch": 1.1183098591549296,
"grad_norm": 0.5733239722826652,
"learning_rate": 7.904330476816391e-06,
"loss": 0.4888,
"step": 397
},
{
"epoch": 1.1211267605633803,
"grad_norm": 0.43773395249537966,
"learning_rate": 7.890968048732091e-06,
"loss": 0.4672,
"step": 398
},
{
"epoch": 1.123943661971831,
"grad_norm": 0.5302196621641707,
"learning_rate": 7.877574531298666e-06,
"loss": 0.505,
"step": 399
},
{
"epoch": 1.1267605633802817,
"grad_norm": 0.46253916299905623,
"learning_rate": 7.864150068549446e-06,
"loss": 0.497,
"step": 400
},
{
"epoch": 1.1295774647887324,
"grad_norm": 0.5434762640908769,
"learning_rate": 7.850694804850538e-06,
"loss": 0.514,
"step": 401
},
{
"epoch": 1.1323943661971831,
"grad_norm": 0.49589560162074636,
"learning_rate": 7.837208884899283e-06,
"loss": 0.4966,
"step": 402
},
{
"epoch": 1.1352112676056338,
"grad_norm": 0.4542407372958917,
"learning_rate": 7.823692453722701e-06,
"loss": 0.4867,
"step": 403
},
{
"epoch": 1.1380281690140845,
"grad_norm": 0.5461139604315557,
"learning_rate": 7.810145656675923e-06,
"loss": 0.5076,
"step": 404
},
{
"epoch": 1.1408450704225352,
"grad_norm": 0.5275707669715958,
"learning_rate": 7.796568639440635e-06,
"loss": 0.496,
"step": 405
},
{
"epoch": 1.143661971830986,
"grad_norm": 0.5140145559296189,
"learning_rate": 7.782961548023515e-06,
"loss": 0.4936,
"step": 406
},
{
"epoch": 1.1464788732394366,
"grad_norm": 0.5073549508475139,
"learning_rate": 7.769324528754653e-06,
"loss": 0.4817,
"step": 407
},
{
"epoch": 1.1492957746478873,
"grad_norm": 0.5143992829310441,
"learning_rate": 7.755657728285979e-06,
"loss": 0.5059,
"step": 408
},
{
"epoch": 1.152112676056338,
"grad_norm": 0.4274674431264774,
"learning_rate": 7.741961293589693e-06,
"loss": 0.5168,
"step": 409
},
{
"epoch": 1.1549295774647887,
"grad_norm": 0.5046190101945648,
"learning_rate": 7.728235371956678e-06,
"loss": 0.5144,
"step": 410
},
{
"epoch": 1.1577464788732394,
"grad_norm": 0.535811089052687,
"learning_rate": 7.714480110994922e-06,
"loss": 0.49,
"step": 411
},
{
"epoch": 1.1605633802816901,
"grad_norm": 0.4873836309283388,
"learning_rate": 7.700695658627924e-06,
"loss": 0.5049,
"step": 412
},
{
"epoch": 1.1633802816901408,
"grad_norm": 0.4727081045199869,
"learning_rate": 7.686882163093106e-06,
"loss": 0.4918,
"step": 413
},
{
"epoch": 1.1661971830985915,
"grad_norm": 0.4826333573880563,
"learning_rate": 7.673039772940218e-06,
"loss": 0.5199,
"step": 414
},
{
"epoch": 1.1690140845070423,
"grad_norm": 0.4529237270053036,
"learning_rate": 7.659168637029746e-06,
"loss": 0.4911,
"step": 415
},
{
"epoch": 1.171830985915493,
"grad_norm": 0.4703668307323843,
"learning_rate": 7.6452689045313e-06,
"loss": 0.4977,
"step": 416
},
{
"epoch": 1.1746478873239437,
"grad_norm": 0.43037435321183687,
"learning_rate": 7.631340724922023e-06,
"loss": 0.4884,
"step": 417
},
{
"epoch": 1.1774647887323944,
"grad_norm": 0.44390781109329,
"learning_rate": 7.617384247984973e-06,
"loss": 0.5007,
"step": 418
},
{
"epoch": 1.180281690140845,
"grad_norm": 0.4303155635436162,
"learning_rate": 7.603399623807519e-06,
"loss": 0.5152,
"step": 419
},
{
"epoch": 1.1830985915492958,
"grad_norm": 0.4472432608010433,
"learning_rate": 7.589387002779722e-06,
"loss": 0.4785,
"step": 420
},
{
"epoch": 1.1859154929577465,
"grad_norm": 0.4418018518887449,
"learning_rate": 7.575346535592721e-06,
"loss": 0.4925,
"step": 421
},
{
"epoch": 1.1887323943661972,
"grad_norm": 0.4358234713383654,
"learning_rate": 7.561278373237108e-06,
"loss": 0.4921,
"step": 422
},
{
"epoch": 1.1915492957746479,
"grad_norm": 0.4758680421804988,
"learning_rate": 7.547182667001316e-06,
"loss": 0.4841,
"step": 423
},
{
"epoch": 1.1943661971830986,
"grad_norm": 0.4632861244285043,
"learning_rate": 7.5330595684699735e-06,
"loss": 0.4705,
"step": 424
},
{
"epoch": 1.1971830985915493,
"grad_norm": 0.5122268438926686,
"learning_rate": 7.5189092295222945e-06,
"loss": 0.5051,
"step": 425
},
{
"epoch": 1.2,
"grad_norm": 0.4542095177777043,
"learning_rate": 7.504731802330427e-06,
"loss": 0.4856,
"step": 426
},
{
"epoch": 1.2028169014084507,
"grad_norm": 0.45093644422410417,
"learning_rate": 7.49052743935783e-06,
"loss": 0.5039,
"step": 427
},
{
"epoch": 1.2056338028169014,
"grad_norm": 0.4698670288362548,
"learning_rate": 7.476296293357626e-06,
"loss": 0.4796,
"step": 428
},
{
"epoch": 1.208450704225352,
"grad_norm": 0.45611072085962195,
"learning_rate": 7.462038517370962e-06,
"loss": 0.5061,
"step": 429
},
{
"epoch": 1.2112676056338028,
"grad_norm": 0.44879399555766897,
"learning_rate": 7.4477542647253645e-06,
"loss": 0.4949,
"step": 430
},
{
"epoch": 1.2140845070422535,
"grad_norm": 0.4568738470061427,
"learning_rate": 7.4334436890330845e-06,
"loss": 0.5085,
"step": 431
},
{
"epoch": 1.2169014084507042,
"grad_norm": 0.4910023364458397,
"learning_rate": 7.4191069441894555e-06,
"loss": 0.5022,
"step": 432
},
{
"epoch": 1.2197183098591549,
"grad_norm": 0.47245665007543136,
"learning_rate": 7.404744184371229e-06,
"loss": 0.4761,
"step": 433
},
{
"epoch": 1.2225352112676056,
"grad_norm": 0.4412540818413418,
"learning_rate": 7.39035556403492e-06,
"loss": 0.4868,
"step": 434
},
{
"epoch": 1.2253521126760563,
"grad_norm": 0.4332333517309346,
"learning_rate": 7.375941237915151e-06,
"loss": 0.4847,
"step": 435
},
{
"epoch": 1.228169014084507,
"grad_norm": 0.42462733830523613,
"learning_rate": 7.361501361022978e-06,
"loss": 0.4801,
"step": 436
},
{
"epoch": 1.2309859154929577,
"grad_norm": 0.41408929063500605,
"learning_rate": 7.347036088644232e-06,
"loss": 0.4633,
"step": 437
},
{
"epoch": 1.2338028169014084,
"grad_norm": 0.4446141488697224,
"learning_rate": 7.3325455763378465e-06,
"loss": 0.4874,
"step": 438
},
{
"epoch": 1.236619718309859,
"grad_norm": 0.39028343567183377,
"learning_rate": 7.318029979934181e-06,
"loss": 0.476,
"step": 439
},
{
"epoch": 1.2394366197183098,
"grad_norm": 0.40546374556808,
"learning_rate": 7.303489455533352e-06,
"loss": 0.4991,
"step": 440
},
{
"epoch": 1.2422535211267607,
"grad_norm": 0.4406508743990959,
"learning_rate": 7.288924159503549e-06,
"loss": 0.5082,
"step": 441
},
{
"epoch": 1.2450704225352114,
"grad_norm": 0.41221274812070907,
"learning_rate": 7.274334248479353e-06,
"loss": 0.4857,
"step": 442
},
{
"epoch": 1.247887323943662,
"grad_norm": 0.4400052134717122,
"learning_rate": 7.259719879360054e-06,
"loss": 0.4856,
"step": 443
},
{
"epoch": 1.2507042253521128,
"grad_norm": 0.4363861455466759,
"learning_rate": 7.2450812093079695e-06,
"loss": 0.4717,
"step": 444
},
{
"epoch": 1.2535211267605635,
"grad_norm": 0.40915950116360555,
"learning_rate": 7.2304183957467385e-06,
"loss": 0.4922,
"step": 445
},
{
"epoch": 1.2563380281690142,
"grad_norm": 0.42971830207213996,
"learning_rate": 7.215731596359645e-06,
"loss": 0.4968,
"step": 446
},
{
"epoch": 1.2591549295774649,
"grad_norm": 0.46604427981545354,
"learning_rate": 7.201020969087913e-06,
"loss": 0.5066,
"step": 447
},
{
"epoch": 1.2619718309859156,
"grad_norm": 0.39614011900749,
"learning_rate": 7.18628667212901e-06,
"loss": 0.4934,
"step": 448
},
{
"epoch": 1.2647887323943663,
"grad_norm": 0.506857483129988,
"learning_rate": 7.17152886393495e-06,
"loss": 0.4836,
"step": 449
},
{
"epoch": 1.267605633802817,
"grad_norm": 0.43280979811122466,
"learning_rate": 7.15674770321058e-06,
"loss": 0.4822,
"step": 450
},
{
"epoch": 1.2704225352112677,
"grad_norm": 0.4291718303065121,
"learning_rate": 7.141943348911885e-06,
"loss": 0.4801,
"step": 451
},
{
"epoch": 1.2732394366197184,
"grad_norm": 0.43238195279590513,
"learning_rate": 7.127115960244269e-06,
"loss": 0.4922,
"step": 452
},
{
"epoch": 1.276056338028169,
"grad_norm": 0.45348003112601215,
"learning_rate": 7.112265696660848e-06,
"loss": 0.4924,
"step": 453
},
{
"epoch": 1.2788732394366198,
"grad_norm": 0.42234280818079467,
"learning_rate": 7.0973927178607335e-06,
"loss": 0.4938,
"step": 454
},
{
"epoch": 1.2816901408450705,
"grad_norm": 0.4396695903135667,
"learning_rate": 7.0824971837873154e-06,
"loss": 0.5022,
"step": 455
},
{
"epoch": 1.2845070422535212,
"grad_norm": 0.433995835377117,
"learning_rate": 7.067579254626543e-06,
"loss": 0.4907,
"step": 456
},
{
"epoch": 1.287323943661972,
"grad_norm": 0.3865805137760006,
"learning_rate": 7.0526390908052e-06,
"loss": 0.4989,
"step": 457
},
{
"epoch": 1.2901408450704226,
"grad_norm": 0.4181083182911281,
"learning_rate": 7.037676852989182e-06,
"loss": 0.4764,
"step": 458
},
{
"epoch": 1.2929577464788733,
"grad_norm": 0.41081992730322214,
"learning_rate": 7.022692702081766e-06,
"loss": 0.4744,
"step": 459
},
{
"epoch": 1.295774647887324,
"grad_norm": 0.40712445897933586,
"learning_rate": 7.007686799221882e-06,
"loss": 0.5068,
"step": 460
},
{
"epoch": 1.2985915492957747,
"grad_norm": 0.4116920369510292,
"learning_rate": 6.992659305782381e-06,
"loss": 0.4652,
"step": 461
},
{
"epoch": 1.3014084507042254,
"grad_norm": 0.4574121168612027,
"learning_rate": 6.977610383368296e-06,
"loss": 0.5119,
"step": 462
},
{
"epoch": 1.304225352112676,
"grad_norm": 0.41865241421774196,
"learning_rate": 6.9625401938151085e-06,
"loss": 0.481,
"step": 463
},
{
"epoch": 1.3070422535211268,
"grad_norm": 0.42466950286656874,
"learning_rate": 6.947448899187004e-06,
"loss": 0.501,
"step": 464
},
{
"epoch": 1.3098591549295775,
"grad_norm": 0.438824118486777,
"learning_rate": 6.932336661775132e-06,
"loss": 0.4961,
"step": 465
},
{
"epoch": 1.3126760563380282,
"grad_norm": 0.4090033280279816,
"learning_rate": 6.91720364409586e-06,
"loss": 0.4741,
"step": 466
},
{
"epoch": 1.315492957746479,
"grad_norm": 0.411571188038661,
"learning_rate": 6.902050008889024e-06,
"loss": 0.4946,
"step": 467
},
{
"epoch": 1.3183098591549296,
"grad_norm": 0.40048050848825795,
"learning_rate": 6.886875919116184e-06,
"loss": 0.4598,
"step": 468
},
{
"epoch": 1.3211267605633803,
"grad_norm": 0.4355207784396914,
"learning_rate": 6.871681537958862e-06,
"loss": 0.5023,
"step": 469
},
{
"epoch": 1.323943661971831,
"grad_norm": 0.40362801280200744,
"learning_rate": 6.856467028816797e-06,
"loss": 0.4994,
"step": 470
},
{
"epoch": 1.3267605633802817,
"grad_norm": 0.46270660160714394,
"learning_rate": 6.841232555306181e-06,
"loss": 0.5157,
"step": 471
},
{
"epoch": 1.3295774647887324,
"grad_norm": 0.42249423161984545,
"learning_rate": 6.825978281257905e-06,
"loss": 0.4798,
"step": 472
},
{
"epoch": 1.332394366197183,
"grad_norm": 0.45487323092224735,
"learning_rate": 6.810704370715791e-06,
"loss": 0.4795,
"step": 473
},
{
"epoch": 1.3352112676056338,
"grad_norm": 0.3919330678327191,
"learning_rate": 6.795410987934834e-06,
"loss": 0.4898,
"step": 474
},
{
"epoch": 1.3380281690140845,
"grad_norm": 0.4353751489003626,
"learning_rate": 6.780098297379427e-06,
"loss": 0.4935,
"step": 475
},
{
"epoch": 1.3408450704225352,
"grad_norm": 0.4629533116217877,
"learning_rate": 6.764766463721605e-06,
"loss": 0.4862,
"step": 476
},
{
"epoch": 1.343661971830986,
"grad_norm": 0.4294849971678444,
"learning_rate": 6.7494156518392625e-06,
"loss": 0.4841,
"step": 477
},
{
"epoch": 1.3464788732394366,
"grad_norm": 0.4469982481591703,
"learning_rate": 6.734046026814388e-06,
"loss": 0.4761,
"step": 478
},
{
"epoch": 1.3492957746478873,
"grad_norm": 0.4431535514689644,
"learning_rate": 6.718657753931284e-06,
"loss": 0.4877,
"step": 479
},
{
"epoch": 1.352112676056338,
"grad_norm": 0.44477269709374434,
"learning_rate": 6.70325099867479e-06,
"loss": 0.492,
"step": 480
},
{
"epoch": 1.3549295774647887,
"grad_norm": 0.44065649629461523,
"learning_rate": 6.687825926728506e-06,
"loss": 0.4779,
"step": 481
},
{
"epoch": 1.3577464788732394,
"grad_norm": 0.46680428200178004,
"learning_rate": 6.672382703973011e-06,
"loss": 0.4979,
"step": 482
},
{
"epoch": 1.36056338028169,
"grad_norm": 0.3753392719693149,
"learning_rate": 6.65692149648407e-06,
"loss": 0.4957,
"step": 483
},
{
"epoch": 1.3633802816901408,
"grad_norm": 0.4464455137831182,
"learning_rate": 6.641442470530866e-06,
"loss": 0.5009,
"step": 484
},
{
"epoch": 1.3661971830985915,
"grad_norm": 0.4558973598726014,
"learning_rate": 6.62594579257419e-06,
"loss": 0.4862,
"step": 485
},
{
"epoch": 1.3690140845070422,
"grad_norm": 0.4244329051781463,
"learning_rate": 6.610431629264669e-06,
"loss": 0.4982,
"step": 486
},
{
"epoch": 1.371830985915493,
"grad_norm": 0.42900553114070356,
"learning_rate": 6.594900147440963e-06,
"loss": 0.5149,
"step": 487
},
{
"epoch": 1.3746478873239436,
"grad_norm": 0.46393543297393147,
"learning_rate": 6.579351514127976e-06,
"loss": 0.504,
"step": 488
},
{
"epoch": 1.3774647887323943,
"grad_norm": 0.5280930975377756,
"learning_rate": 6.56378589653506e-06,
"loss": 0.5027,
"step": 489
},
{
"epoch": 1.380281690140845,
"grad_norm": 0.44081231018896067,
"learning_rate": 6.548203462054211e-06,
"loss": 0.4974,
"step": 490
},
{
"epoch": 1.3830985915492957,
"grad_norm": 0.45229210902770534,
"learning_rate": 6.5326043782582785e-06,
"loss": 0.5099,
"step": 491
},
{
"epoch": 1.3859154929577464,
"grad_norm": 0.4116450533577545,
"learning_rate": 6.516988812899154e-06,
"loss": 0.4901,
"step": 492
},
{
"epoch": 1.388732394366197,
"grad_norm": 0.4941893352727475,
"learning_rate": 6.501356933905973e-06,
"loss": 0.4838,
"step": 493
},
{
"epoch": 1.3915492957746478,
"grad_norm": 0.4473632155182213,
"learning_rate": 6.485708909383306e-06,
"loss": 0.5047,
"step": 494
},
{
"epoch": 1.3943661971830985,
"grad_norm": 0.43832979886064743,
"learning_rate": 6.4700449076093515e-06,
"loss": 0.4723,
"step": 495
},
{
"epoch": 1.3971830985915492,
"grad_norm": 0.44987425974821665,
"learning_rate": 6.454365097034127e-06,
"loss": 0.4766,
"step": 496
},
{
"epoch": 1.4,
"grad_norm": 0.40420889188372705,
"learning_rate": 6.43866964627766e-06,
"loss": 0.4755,
"step": 497
},
{
"epoch": 1.4028169014084506,
"grad_norm": 0.44487770905469926,
"learning_rate": 6.422958724128169e-06,
"loss": 0.4899,
"step": 498
},
{
"epoch": 1.4056338028169013,
"grad_norm": 0.435043357418449,
"learning_rate": 6.40723249954025e-06,
"loss": 0.4972,
"step": 499
},
{
"epoch": 1.408450704225352,
"grad_norm": 0.4627668121771765,
"learning_rate": 6.391491141633064e-06,
"loss": 0.4798,
"step": 500
},
{
"epoch": 1.4112676056338027,
"grad_norm": 0.4070202392162525,
"learning_rate": 6.375734819688514e-06,
"loss": 0.4898,
"step": 501
},
{
"epoch": 1.4140845070422534,
"grad_norm": 0.4213085739571462,
"learning_rate": 6.359963703149424e-06,
"loss": 0.4904,
"step": 502
},
{
"epoch": 1.4169014084507041,
"grad_norm": 0.5558578544437808,
"learning_rate": 6.344177961617719e-06,
"loss": 0.4795,
"step": 503
},
{
"epoch": 1.4197183098591548,
"grad_norm": 0.4638056979361732,
"learning_rate": 6.3283777648526035e-06,
"loss": 0.4944,
"step": 504
},
{
"epoch": 1.4225352112676055,
"grad_norm": 0.44058978200259,
"learning_rate": 6.312563282768729e-06,
"loss": 0.4805,
"step": 505
},
{
"epoch": 1.4253521126760562,
"grad_norm": 0.4712406089314399,
"learning_rate": 6.296734685434373e-06,
"loss": 0.518,
"step": 506
},
{
"epoch": 1.428169014084507,
"grad_norm": 0.50303011695748,
"learning_rate": 6.280892143069607e-06,
"loss": 0.5078,
"step": 507
},
{
"epoch": 1.4309859154929576,
"grad_norm": 0.44401567500787253,
"learning_rate": 6.265035826044467e-06,
"loss": 0.4986,
"step": 508
},
{
"epoch": 1.4338028169014083,
"grad_norm": 0.45802099251252676,
"learning_rate": 6.2491659048771215e-06,
"loss": 0.4962,
"step": 509
},
{
"epoch": 1.436619718309859,
"grad_norm": 0.45900922989672704,
"learning_rate": 6.233282550232036e-06,
"loss": 0.4701,
"step": 510
},
{
"epoch": 1.43943661971831,
"grad_norm": 0.4104785858315547,
"learning_rate": 6.217385932918141e-06,
"loss": 0.4708,
"step": 511
},
{
"epoch": 1.4422535211267606,
"grad_norm": 0.4287021977710135,
"learning_rate": 6.201476223886993e-06,
"loss": 0.4788,
"step": 512
},
{
"epoch": 1.4450704225352113,
"grad_norm": 0.4428659512764066,
"learning_rate": 6.185553594230934e-06,
"loss": 0.5327,
"step": 513
},
{
"epoch": 1.447887323943662,
"grad_norm": 0.39989248720275716,
"learning_rate": 6.169618215181256e-06,
"loss": 0.4913,
"step": 514
},
{
"epoch": 1.4507042253521127,
"grad_norm": 0.4108837509780523,
"learning_rate": 6.153670258106356e-06,
"loss": 0.4905,
"step": 515
},
{
"epoch": 1.4535211267605634,
"grad_norm": 0.41341293933805207,
"learning_rate": 6.137709894509898e-06,
"loss": 0.5087,
"step": 516
},
{
"epoch": 1.4563380281690141,
"grad_norm": 0.42267988314579,
"learning_rate": 6.121737296028959e-06,
"loss": 0.4888,
"step": 517
},
{
"epoch": 1.4591549295774648,
"grad_norm": 0.4511096410354952,
"learning_rate": 6.105752634432195e-06,
"loss": 0.5161,
"step": 518
},
{
"epoch": 1.4619718309859155,
"grad_norm": 0.41852288904862534,
"learning_rate": 6.089756081617987e-06,
"loss": 0.4951,
"step": 519
},
{
"epoch": 1.4647887323943662,
"grad_norm": 0.47057489166038396,
"learning_rate": 6.073747809612591e-06,
"loss": 0.4713,
"step": 520
},
{
"epoch": 1.467605633802817,
"grad_norm": 0.40256689019530845,
"learning_rate": 6.057727990568294e-06,
"loss": 0.4911,
"step": 521
},
{
"epoch": 1.4704225352112676,
"grad_norm": 0.4510853893653407,
"learning_rate": 6.041696796761558e-06,
"loss": 0.5099,
"step": 522
},
{
"epoch": 1.4732394366197183,
"grad_norm": 0.38944003217951545,
"learning_rate": 6.025654400591167e-06,
"loss": 0.4921,
"step": 523
},
{
"epoch": 1.476056338028169,
"grad_norm": 0.3940764581766545,
"learning_rate": 6.009600974576378e-06,
"loss": 0.5037,
"step": 524
},
{
"epoch": 1.4788732394366197,
"grad_norm": 0.4026807626178998,
"learning_rate": 5.9935366913550615e-06,
"loss": 0.4854,
"step": 525
},
{
"epoch": 1.4816901408450704,
"grad_norm": 0.44187668092733434,
"learning_rate": 5.977461723681845e-06,
"loss": 0.4702,
"step": 526
},
{
"epoch": 1.4845070422535211,
"grad_norm": 0.43769176532376947,
"learning_rate": 5.961376244426256e-06,
"loss": 0.5001,
"step": 527
},
{
"epoch": 1.4873239436619718,
"grad_norm": 0.47287330095754926,
"learning_rate": 5.945280426570862e-06,
"loss": 0.4876,
"step": 528
},
{
"epoch": 1.4901408450704225,
"grad_norm": 0.428038527885128,
"learning_rate": 5.929174443209416e-06,
"loss": 0.5191,
"step": 529
},
{
"epoch": 1.4929577464788732,
"grad_norm": 0.4021956491054695,
"learning_rate": 5.913058467544984e-06,
"loss": 0.4842,
"step": 530
},
{
"epoch": 1.495774647887324,
"grad_norm": 0.448727233648992,
"learning_rate": 5.8969326728880964e-06,
"loss": 0.4866,
"step": 531
},
{
"epoch": 1.4985915492957746,
"grad_norm": 0.43755912059789587,
"learning_rate": 5.880797232654869e-06,
"loss": 0.4771,
"step": 532
},
{
"epoch": 1.5014084507042254,
"grad_norm": 0.4425452078492814,
"learning_rate": 5.8646523203651514e-06,
"loss": 0.5125,
"step": 533
},
{
"epoch": 1.504225352112676,
"grad_norm": 0.48313466807852506,
"learning_rate": 5.848498109640652e-06,
"loss": 0.4872,
"step": 534
},
{
"epoch": 1.5070422535211268,
"grad_norm": 0.47985626406062337,
"learning_rate": 5.832334774203076e-06,
"loss": 0.5105,
"step": 535
},
{
"epoch": 1.5098591549295775,
"grad_norm": 0.371740721272159,
"learning_rate": 5.8161624878722545e-06,
"loss": 0.4455,
"step": 536
},
{
"epoch": 1.5126760563380282,
"grad_norm": 0.4869175302843986,
"learning_rate": 5.799981424564275e-06,
"loss": 0.4962,
"step": 537
},
{
"epoch": 1.5154929577464789,
"grad_norm": 0.4814836246010178,
"learning_rate": 5.7837917582896145e-06,
"loss": 0.5113,
"step": 538
},
{
"epoch": 1.5183098591549296,
"grad_norm": 0.41287497701292597,
"learning_rate": 5.767593663151265e-06,
"loss": 0.4781,
"step": 539
},
{
"epoch": 1.5211267605633803,
"grad_norm": 0.5074496740090907,
"learning_rate": 5.751387313342863e-06,
"loss": 0.5008,
"step": 540
},
{
"epoch": 1.523943661971831,
"grad_norm": 0.423624893721716,
"learning_rate": 5.735172883146813e-06,
"loss": 0.5054,
"step": 541
},
{
"epoch": 1.5267605633802817,
"grad_norm": 0.3983612120471499,
"learning_rate": 5.718950546932418e-06,
"loss": 0.5119,
"step": 542
},
{
"epoch": 1.5295774647887324,
"grad_norm": 0.45566126306718424,
"learning_rate": 5.702720479154001e-06,
"loss": 0.4944,
"step": 543
},
{
"epoch": 1.532394366197183,
"grad_norm": 0.45415425821004124,
"learning_rate": 5.686482854349029e-06,
"loss": 0.5087,
"step": 544
},
{
"epoch": 1.5352112676056338,
"grad_norm": 0.418311733394643,
"learning_rate": 5.6702378471362394e-06,
"loss": 0.4881,
"step": 545
},
{
"epoch": 1.5380281690140845,
"grad_norm": 0.4618061508199756,
"learning_rate": 5.653985632213758e-06,
"loss": 0.4851,
"step": 546
},
{
"epoch": 1.5408450704225352,
"grad_norm": 0.4113372184265852,
"learning_rate": 5.637726384357222e-06,
"loss": 0.4928,
"step": 547
},
{
"epoch": 1.543661971830986,
"grad_norm": 0.44563040764122475,
"learning_rate": 5.621460278417901e-06,
"loss": 0.4677,
"step": 548
},
{
"epoch": 1.5464788732394368,
"grad_norm": 0.47662643350622824,
"learning_rate": 5.605187489320815e-06,
"loss": 0.4969,
"step": 549
},
{
"epoch": 1.5492957746478875,
"grad_norm": 0.40408999101715254,
"learning_rate": 5.588908192062858e-06,
"loss": 0.4897,
"step": 550
},
{
"epoch": 1.5521126760563382,
"grad_norm": 0.46478585160154373,
"learning_rate": 5.572622561710906e-06,
"loss": 0.477,
"step": 551
},
{
"epoch": 1.5549295774647889,
"grad_norm": 0.43970878015188386,
"learning_rate": 5.556330773399948e-06,
"loss": 0.4798,
"step": 552
},
{
"epoch": 1.5577464788732396,
"grad_norm": 0.4103181086924883,
"learning_rate": 5.54003300233119e-06,
"loss": 0.4804,
"step": 553
},
{
"epoch": 1.5605633802816903,
"grad_norm": 0.4476411103725844,
"learning_rate": 5.52372942377018e-06,
"loss": 0.5,
"step": 554
},
{
"epoch": 1.563380281690141,
"grad_norm": 0.42059975361762,
"learning_rate": 5.507420213044915e-06,
"loss": 0.4914,
"step": 555
},
{
"epoch": 1.5661971830985917,
"grad_norm": 0.4227459859510044,
"learning_rate": 5.491105545543966e-06,
"loss": 0.479,
"step": 556
},
{
"epoch": 1.5690140845070424,
"grad_norm": 0.44685775608699185,
"learning_rate": 5.474785596714581e-06,
"loss": 0.4847,
"step": 557
},
{
"epoch": 1.571830985915493,
"grad_norm": 0.40998327682835917,
"learning_rate": 5.458460542060807e-06,
"loss": 0.4553,
"step": 558
},
{
"epoch": 1.5746478873239438,
"grad_norm": 0.41568075131048016,
"learning_rate": 5.442130557141595e-06,
"loss": 0.4961,
"step": 559
},
{
"epoch": 1.5774647887323945,
"grad_norm": 0.4130684664753457,
"learning_rate": 5.42579581756892e-06,
"loss": 0.4901,
"step": 560
},
{
"epoch": 1.5802816901408452,
"grad_norm": 0.4265493303005442,
"learning_rate": 5.409456499005883e-06,
"loss": 0.473,
"step": 561
},
{
"epoch": 1.5830985915492959,
"grad_norm": 0.41210586655876397,
"learning_rate": 5.393112777164834e-06,
"loss": 0.4836,
"step": 562
},
{
"epoch": 1.5859154929577466,
"grad_norm": 0.4120627899473988,
"learning_rate": 5.376764827805468e-06,
"loss": 0.4754,
"step": 563
},
{
"epoch": 1.5887323943661973,
"grad_norm": 0.39881714263247103,
"learning_rate": 5.36041282673295e-06,
"loss": 0.4734,
"step": 564
},
{
"epoch": 1.591549295774648,
"grad_norm": 0.4225245003542499,
"learning_rate": 5.3440569497960126e-06,
"loss": 0.5043,
"step": 565
},
{
"epoch": 1.5943661971830987,
"grad_norm": 0.4437638279308617,
"learning_rate": 5.32769737288507e-06,
"loss": 0.4894,
"step": 566
},
{
"epoch": 1.5971830985915494,
"grad_norm": 0.4085471752813184,
"learning_rate": 5.3113342719303266e-06,
"loss": 0.4826,
"step": 567
},
{
"epoch": 1.6,
"grad_norm": 0.44059126729044296,
"learning_rate": 5.294967822899882e-06,
"loss": 0.4842,
"step": 568
},
{
"epoch": 1.6028169014084508,
"grad_norm": 0.4785027561352671,
"learning_rate": 5.278598201797844e-06,
"loss": 0.5029,
"step": 569
},
{
"epoch": 1.6056338028169015,
"grad_norm": 0.39271046798403997,
"learning_rate": 5.262225584662431e-06,
"loss": 0.4926,
"step": 570
},
{
"epoch": 1.6084507042253522,
"grad_norm": 0.42328960022000284,
"learning_rate": 5.245850147564081e-06,
"loss": 0.4865,
"step": 571
},
{
"epoch": 1.611267605633803,
"grad_norm": 0.45771191563988584,
"learning_rate": 5.229472066603558e-06,
"loss": 0.4764,
"step": 572
},
{
"epoch": 1.6140845070422536,
"grad_norm": 0.3867972365885545,
"learning_rate": 5.213091517910056e-06,
"loss": 0.4901,
"step": 573
},
{
"epoch": 1.6169014084507043,
"grad_norm": 0.3873468850604546,
"learning_rate": 5.196708677639311e-06,
"loss": 0.4973,
"step": 574
},
{
"epoch": 1.619718309859155,
"grad_norm": 0.36928905724949573,
"learning_rate": 5.180323721971696e-06,
"loss": 0.4653,
"step": 575
},
{
"epoch": 1.6225352112676057,
"grad_norm": 0.3855607659359868,
"learning_rate": 5.163936827110342e-06,
"loss": 0.4721,
"step": 576
},
{
"epoch": 1.6253521126760564,
"grad_norm": 0.4009987591689016,
"learning_rate": 5.1475481692792235e-06,
"loss": 0.5147,
"step": 577
},
{
"epoch": 1.628169014084507,
"grad_norm": 0.4041174927757299,
"learning_rate": 5.131157924721285e-06,
"loss": 0.4924,
"step": 578
},
{
"epoch": 1.6309859154929578,
"grad_norm": 0.3747685368431732,
"learning_rate": 5.1147662696965254e-06,
"loss": 0.4825,
"step": 579
},
{
"epoch": 1.6338028169014085,
"grad_norm": 0.4088818726448478,
"learning_rate": 5.098373380480114e-06,
"loss": 0.5083,
"step": 580
},
{
"epoch": 1.6366197183098592,
"grad_norm": 0.39419824398459724,
"learning_rate": 5.081979433360498e-06,
"loss": 0.4935,
"step": 581
},
{
"epoch": 1.63943661971831,
"grad_norm": 0.39469243440406887,
"learning_rate": 5.065584604637492e-06,
"loss": 0.4576,
"step": 582
},
{
"epoch": 1.6422535211267606,
"grad_norm": 0.414636635926328,
"learning_rate": 5.0491890706204e-06,
"loss": 0.4842,
"step": 583
},
{
"epoch": 1.6450704225352113,
"grad_norm": 0.43475222695638494,
"learning_rate": 5.0327930076261065e-06,
"loss": 0.4557,
"step": 584
},
{
"epoch": 1.647887323943662,
"grad_norm": 0.3513983093557574,
"learning_rate": 5.0163965919771855e-06,
"loss": 0.4656,
"step": 585
},
{
"epoch": 1.6507042253521127,
"grad_norm": 0.4628906667358179,
"learning_rate": 5e-06,
"loss": 0.5117,
"step": 586
},
{
"epoch": 1.6535211267605634,
"grad_norm": 0.3907044515953839,
"learning_rate": 4.983603408022817e-06,
"loss": 0.4928,
"step": 587
},
{
"epoch": 1.656338028169014,
"grad_norm": 0.4554985455916155,
"learning_rate": 4.967206992373894e-06,
"loss": 0.4957,
"step": 588
},
{
"epoch": 1.6591549295774648,
"grad_norm": 0.40510939357693265,
"learning_rate": 4.9508109293796015e-06,
"loss": 0.4916,
"step": 589
},
{
"epoch": 1.6619718309859155,
"grad_norm": 0.48375130926157284,
"learning_rate": 4.9344153953625095e-06,
"loss": 0.4853,
"step": 590
},
{
"epoch": 1.6647887323943662,
"grad_norm": 0.4937685950817898,
"learning_rate": 4.918020566639505e-06,
"loss": 0.4605,
"step": 591
},
{
"epoch": 1.667605633802817,
"grad_norm": 0.3931913794777372,
"learning_rate": 4.901626619519888e-06,
"loss": 0.4924,
"step": 592
},
{
"epoch": 1.6704225352112676,
"grad_norm": 0.4757949719740506,
"learning_rate": 4.885233730303475e-06,
"loss": 0.479,
"step": 593
},
{
"epoch": 1.6732394366197183,
"grad_norm": 0.46379868898689014,
"learning_rate": 4.868842075278717e-06,
"loss": 0.488,
"step": 594
},
{
"epoch": 1.676056338028169,
"grad_norm": 0.3881777800605308,
"learning_rate": 4.852451830720777e-06,
"loss": 0.4899,
"step": 595
},
{
"epoch": 1.6788732394366197,
"grad_norm": 0.4367216334986508,
"learning_rate": 4.83606317288966e-06,
"loss": 0.4881,
"step": 596
},
{
"epoch": 1.6816901408450704,
"grad_norm": 0.40564739437861447,
"learning_rate": 4.819676278028305e-06,
"loss": 0.4787,
"step": 597
},
{
"epoch": 1.684507042253521,
"grad_norm": 0.4252084939655642,
"learning_rate": 4.803291322360691e-06,
"loss": 0.4763,
"step": 598
},
{
"epoch": 1.6873239436619718,
"grad_norm": 0.41871066030523396,
"learning_rate": 4.7869084820899455e-06,
"loss": 0.4977,
"step": 599
},
{
"epoch": 1.6901408450704225,
"grad_norm": 0.43384887842924486,
"learning_rate": 4.7705279333964435e-06,
"loss": 0.5063,
"step": 600
},
{
"epoch": 1.6929577464788732,
"grad_norm": 0.40350171077067093,
"learning_rate": 4.75414985243592e-06,
"loss": 0.4924,
"step": 601
},
{
"epoch": 1.695774647887324,
"grad_norm": 0.36329329985026526,
"learning_rate": 4.73777441533757e-06,
"loss": 0.476,
"step": 602
},
{
"epoch": 1.6985915492957746,
"grad_norm": 0.37282495932190163,
"learning_rate": 4.721401798202157e-06,
"loss": 0.4959,
"step": 603
},
{
"epoch": 1.7014084507042253,
"grad_norm": 0.393392739621724,
"learning_rate": 4.70503217710012e-06,
"loss": 0.4813,
"step": 604
},
{
"epoch": 1.704225352112676,
"grad_norm": 0.42627654440084173,
"learning_rate": 4.688665728069676e-06,
"loss": 0.5041,
"step": 605
},
{
"epoch": 1.7070422535211267,
"grad_norm": 0.42156592000121074,
"learning_rate": 4.67230262711493e-06,
"loss": 0.4639,
"step": 606
},
{
"epoch": 1.7098591549295774,
"grad_norm": 0.37153582403937235,
"learning_rate": 4.655943050203987e-06,
"loss": 0.4933,
"step": 607
},
{
"epoch": 1.712676056338028,
"grad_norm": 0.3618346879431781,
"learning_rate": 4.63958717326705e-06,
"loss": 0.4731,
"step": 608
},
{
"epoch": 1.7154929577464788,
"grad_norm": 0.4189867275686855,
"learning_rate": 4.623235172194532e-06,
"loss": 0.48,
"step": 609
},
{
"epoch": 1.7183098591549295,
"grad_norm": 0.3544074995168518,
"learning_rate": 4.606887222835168e-06,
"loss": 0.5005,
"step": 610
},
{
"epoch": 1.7211267605633802,
"grad_norm": 0.3582551183885596,
"learning_rate": 4.590543500994118e-06,
"loss": 0.4851,
"step": 611
},
{
"epoch": 1.723943661971831,
"grad_norm": 0.3682415876242219,
"learning_rate": 4.574204182431082e-06,
"loss": 0.4855,
"step": 612
},
{
"epoch": 1.7267605633802816,
"grad_norm": 0.3978386232505586,
"learning_rate": 4.557869442858406e-06,
"loss": 0.5023,
"step": 613
},
{
"epoch": 1.7295774647887323,
"grad_norm": 0.4037363077292521,
"learning_rate": 4.541539457939194e-06,
"loss": 0.4939,
"step": 614
},
{
"epoch": 1.732394366197183,
"grad_norm": 0.4426913520017231,
"learning_rate": 4.525214403285421e-06,
"loss": 0.4967,
"step": 615
},
{
"epoch": 1.7352112676056337,
"grad_norm": 0.41553539084217284,
"learning_rate": 4.5088944544560355e-06,
"loss": 0.4898,
"step": 616
},
{
"epoch": 1.7380281690140844,
"grad_norm": 0.476787686400504,
"learning_rate": 4.4925797869550865e-06,
"loss": 0.4747,
"step": 617
},
{
"epoch": 1.7408450704225351,
"grad_norm": 0.44174190027116295,
"learning_rate": 4.476270576229823e-06,
"loss": 0.5051,
"step": 618
},
{
"epoch": 1.7436619718309858,
"grad_norm": 0.46363672177794424,
"learning_rate": 4.459966997668812e-06,
"loss": 0.4665,
"step": 619
},
{
"epoch": 1.7464788732394365,
"grad_norm": 0.35788247114147254,
"learning_rate": 4.443669226600053e-06,
"loss": 0.4902,
"step": 620
},
{
"epoch": 1.7492957746478872,
"grad_norm": 0.3992859755370345,
"learning_rate": 4.427377438289095e-06,
"loss": 0.4975,
"step": 621
},
{
"epoch": 1.752112676056338,
"grad_norm": 0.4255514294644944,
"learning_rate": 4.411091807937143e-06,
"loss": 0.481,
"step": 622
},
{
"epoch": 1.7549295774647886,
"grad_norm": 0.4210089823530368,
"learning_rate": 4.3948125106791854e-06,
"loss": 0.4771,
"step": 623
},
{
"epoch": 1.7577464788732393,
"grad_norm": 0.3906604723995445,
"learning_rate": 4.3785397215821e-06,
"loss": 0.46,
"step": 624
},
{
"epoch": 1.76056338028169,
"grad_norm": 0.3874823235263679,
"learning_rate": 4.362273615642779e-06,
"loss": 0.5002,
"step": 625
},
{
"epoch": 1.7633802816901407,
"grad_norm": 0.3901434101586165,
"learning_rate": 4.346014367786243e-06,
"loss": 0.4843,
"step": 626
},
{
"epoch": 1.7661971830985914,
"grad_norm": 0.37499873527013877,
"learning_rate": 4.329762152863762e-06,
"loss": 0.4842,
"step": 627
},
{
"epoch": 1.7690140845070421,
"grad_norm": 0.4151492021167118,
"learning_rate": 4.313517145650973e-06,
"loss": 0.4958,
"step": 628
},
{
"epoch": 1.7718309859154928,
"grad_norm": 0.35051417440023813,
"learning_rate": 4.297279520846002e-06,
"loss": 0.5058,
"step": 629
},
{
"epoch": 1.7746478873239435,
"grad_norm": 0.3916747258132481,
"learning_rate": 4.281049453067584e-06,
"loss": 0.4669,
"step": 630
},
{
"epoch": 1.7774647887323942,
"grad_norm": 0.377379436744131,
"learning_rate": 4.264827116853189e-06,
"loss": 0.4962,
"step": 631
},
{
"epoch": 1.780281690140845,
"grad_norm": 0.39444463909483923,
"learning_rate": 4.248612686657139e-06,
"loss": 0.4776,
"step": 632
},
{
"epoch": 1.7830985915492956,
"grad_norm": 0.4224260284589088,
"learning_rate": 4.232406336848734e-06,
"loss": 0.4909,
"step": 633
},
{
"epoch": 1.7859154929577463,
"grad_norm": 0.3831008836136032,
"learning_rate": 4.2162082417103855e-06,
"loss": 0.4927,
"step": 634
},
{
"epoch": 1.788732394366197,
"grad_norm": 0.37541501249859294,
"learning_rate": 4.200018575435726e-06,
"loss": 0.4863,
"step": 635
},
{
"epoch": 1.7915492957746477,
"grad_norm": 0.4326695660097326,
"learning_rate": 4.183837512127747e-06,
"loss": 0.4989,
"step": 636
},
{
"epoch": 1.7943661971830986,
"grad_norm": 0.38404026793531815,
"learning_rate": 4.167665225796925e-06,
"loss": 0.4592,
"step": 637
},
{
"epoch": 1.7971830985915493,
"grad_norm": 0.4162688038323106,
"learning_rate": 4.1515018903593485e-06,
"loss": 0.4807,
"step": 638
},
{
"epoch": 1.8,
"grad_norm": 0.4278400322795686,
"learning_rate": 4.135347679634849e-06,
"loss": 0.4815,
"step": 639
},
{
"epoch": 1.8028169014084507,
"grad_norm": 0.40764842876105817,
"learning_rate": 4.119202767345132e-06,
"loss": 0.5008,
"step": 640
},
{
"epoch": 1.8056338028169014,
"grad_norm": 0.42716211859348846,
"learning_rate": 4.103067327111905e-06,
"loss": 0.4922,
"step": 641
},
{
"epoch": 1.8084507042253521,
"grad_norm": 0.45566066287532137,
"learning_rate": 4.086941532455017e-06,
"loss": 0.4967,
"step": 642
},
{
"epoch": 1.8112676056338028,
"grad_norm": 0.44375813526255004,
"learning_rate": 4.070825556790587e-06,
"loss": 0.497,
"step": 643
},
{
"epoch": 1.8140845070422535,
"grad_norm": 0.38051175030231293,
"learning_rate": 4.05471957342914e-06,
"loss": 0.4742,
"step": 644
},
{
"epoch": 1.8169014084507042,
"grad_norm": 0.42575239690946376,
"learning_rate": 4.0386237555737476e-06,
"loss": 0.5048,
"step": 645
},
{
"epoch": 1.819718309859155,
"grad_norm": 0.37163029748077847,
"learning_rate": 4.022538276318156e-06,
"loss": 0.4837,
"step": 646
},
{
"epoch": 1.8225352112676056,
"grad_norm": 0.45165782259780374,
"learning_rate": 4.006463308644939e-06,
"loss": 0.4921,
"step": 647
},
{
"epoch": 1.8253521126760563,
"grad_norm": 0.455006281945947,
"learning_rate": 3.990399025423622e-06,
"loss": 0.4583,
"step": 648
},
{
"epoch": 1.828169014084507,
"grad_norm": 0.4079794708534016,
"learning_rate": 3.974345599408833e-06,
"loss": 0.5085,
"step": 649
},
{
"epoch": 1.8309859154929577,
"grad_norm": 0.40865373266918004,
"learning_rate": 3.958303203238443e-06,
"loss": 0.476,
"step": 650
},
{
"epoch": 1.8338028169014085,
"grad_norm": 0.4184354451550438,
"learning_rate": 3.942272009431707e-06,
"loss": 0.4589,
"step": 651
},
{
"epoch": 1.8366197183098592,
"grad_norm": 0.49668041654230144,
"learning_rate": 3.92625219038741e-06,
"loss": 0.5024,
"step": 652
},
{
"epoch": 1.8394366197183099,
"grad_norm": 0.3487062852877808,
"learning_rate": 3.910243918382015e-06,
"loss": 0.4815,
"step": 653
},
{
"epoch": 1.8422535211267606,
"grad_norm": 0.4120415306391501,
"learning_rate": 3.894247365567806e-06,
"loss": 0.4795,
"step": 654
},
{
"epoch": 1.8450704225352113,
"grad_norm": 0.4477361963322955,
"learning_rate": 3.878262703971043e-06,
"loss": 0.4957,
"step": 655
},
{
"epoch": 1.847887323943662,
"grad_norm": 0.429401841610224,
"learning_rate": 3.8622901054901045e-06,
"loss": 0.4841,
"step": 656
},
{
"epoch": 1.8507042253521127,
"grad_norm": 0.3942562812288331,
"learning_rate": 3.846329741893646e-06,
"loss": 0.4795,
"step": 657
},
{
"epoch": 1.8535211267605634,
"grad_norm": 0.426170025102285,
"learning_rate": 3.830381784818746e-06,
"loss": 0.5068,
"step": 658
},
{
"epoch": 1.856338028169014,
"grad_norm": 0.38804163901918054,
"learning_rate": 3.814446405769069e-06,
"loss": 0.4835,
"step": 659
},
{
"epoch": 1.8591549295774648,
"grad_norm": 0.3540589772905739,
"learning_rate": 3.7985237761130077e-06,
"loss": 0.4929,
"step": 660
},
{
"epoch": 1.8619718309859155,
"grad_norm": 0.42357776241410305,
"learning_rate": 3.7826140670818597e-06,
"loss": 0.504,
"step": 661
},
{
"epoch": 1.8647887323943662,
"grad_norm": 0.3729025016669601,
"learning_rate": 3.766717449767965e-06,
"loss": 0.4777,
"step": 662
},
{
"epoch": 1.8676056338028169,
"grad_norm": 0.3889169801805458,
"learning_rate": 3.75083409512288e-06,
"loss": 0.515,
"step": 663
},
{
"epoch": 1.8704225352112676,
"grad_norm": 0.38907663234098605,
"learning_rate": 3.7349641739555342e-06,
"loss": 0.4881,
"step": 664
},
{
"epoch": 1.8732394366197183,
"grad_norm": 0.38796929384532813,
"learning_rate": 3.719107856930395e-06,
"loss": 0.4917,
"step": 665
},
{
"epoch": 1.8760563380281692,
"grad_norm": 0.3986305355401847,
"learning_rate": 3.703265314565629e-06,
"loss": 0.5089,
"step": 666
},
{
"epoch": 1.8788732394366199,
"grad_norm": 0.347146307991803,
"learning_rate": 3.687436717231273e-06,
"loss": 0.4766,
"step": 667
},
{
"epoch": 1.8816901408450706,
"grad_norm": 0.40539428665594573,
"learning_rate": 3.6716222351473986e-06,
"loss": 0.5021,
"step": 668
},
{
"epoch": 1.8845070422535213,
"grad_norm": 0.38842403728574076,
"learning_rate": 3.6558220383822824e-06,
"loss": 0.5044,
"step": 669
},
{
"epoch": 1.887323943661972,
"grad_norm": 0.39954575200830245,
"learning_rate": 3.6400362968505776e-06,
"loss": 0.4882,
"step": 670
},
{
"epoch": 1.8901408450704227,
"grad_norm": 0.41445903325977895,
"learning_rate": 3.6242651803114876e-06,
"loss": 0.4994,
"step": 671
},
{
"epoch": 1.8929577464788734,
"grad_norm": 0.38275982826221505,
"learning_rate": 3.6085088583669368e-06,
"loss": 0.4899,
"step": 672
},
{
"epoch": 1.895774647887324,
"grad_norm": 0.3574431084312652,
"learning_rate": 3.59276750045975e-06,
"loss": 0.4829,
"step": 673
},
{
"epoch": 1.8985915492957748,
"grad_norm": 0.3788946557536076,
"learning_rate": 3.5770412758718317e-06,
"loss": 0.4916,
"step": 674
},
{
"epoch": 1.9014084507042255,
"grad_norm": 0.42134326422621154,
"learning_rate": 3.56133035372234e-06,
"loss": 0.4608,
"step": 675
},
{
"epoch": 1.9042253521126762,
"grad_norm": 0.4177754485864574,
"learning_rate": 3.545634902965873e-06,
"loss": 0.4973,
"step": 676
},
{
"epoch": 1.9070422535211269,
"grad_norm": 0.37996812865010277,
"learning_rate": 3.52995509239065e-06,
"loss": 0.4847,
"step": 677
},
{
"epoch": 1.9098591549295776,
"grad_norm": 0.36341888349352885,
"learning_rate": 3.514291090616696e-06,
"loss": 0.4928,
"step": 678
},
{
"epoch": 1.9126760563380283,
"grad_norm": 0.3586723850556533,
"learning_rate": 3.4986430660940283e-06,
"loss": 0.4589,
"step": 679
},
{
"epoch": 1.915492957746479,
"grad_norm": 0.41045984913560996,
"learning_rate": 3.483011187100847e-06,
"loss": 0.4768,
"step": 680
},
{
"epoch": 1.9183098591549297,
"grad_norm": 0.3797315202775918,
"learning_rate": 3.4673956217417228e-06,
"loss": 0.4816,
"step": 681
},
{
"epoch": 1.9211267605633804,
"grad_norm": 0.39011901642864444,
"learning_rate": 3.451796537945791e-06,
"loss": 0.5053,
"step": 682
},
{
"epoch": 1.923943661971831,
"grad_norm": 0.4068699128633064,
"learning_rate": 3.4362141034649434e-06,
"loss": 0.4715,
"step": 683
},
{
"epoch": 1.9267605633802818,
"grad_norm": 0.40734934524843114,
"learning_rate": 3.4206484858720267e-06,
"loss": 0.485,
"step": 684
},
{
"epoch": 1.9295774647887325,
"grad_norm": 0.4177722988348098,
"learning_rate": 3.4050998525590406e-06,
"loss": 0.4897,
"step": 685
},
{
"epoch": 1.9323943661971832,
"grad_norm": 0.35772887554141514,
"learning_rate": 3.389568370735332e-06,
"loss": 0.4658,
"step": 686
},
{
"epoch": 1.935211267605634,
"grad_norm": 0.382070573657635,
"learning_rate": 3.374054207425811e-06,
"loss": 0.452,
"step": 687
},
{
"epoch": 1.9380281690140846,
"grad_norm": 0.37319547891516613,
"learning_rate": 3.3585575294691355e-06,
"loss": 0.4777,
"step": 688
},
{
"epoch": 1.9408450704225353,
"grad_norm": 0.37543346487112356,
"learning_rate": 3.3430785035159297e-06,
"loss": 0.4954,
"step": 689
},
{
"epoch": 1.943661971830986,
"grad_norm": 0.4343137243095733,
"learning_rate": 3.327617296026991e-06,
"loss": 0.4757,
"step": 690
},
{
"epoch": 1.9464788732394367,
"grad_norm": 0.4113551727906176,
"learning_rate": 3.3121740732714953e-06,
"loss": 0.4908,
"step": 691
},
{
"epoch": 1.9492957746478874,
"grad_norm": 0.4150173708033274,
"learning_rate": 3.296749001325212e-06,
"loss": 0.4855,
"step": 692
},
{
"epoch": 1.952112676056338,
"grad_norm": 0.39769439667103085,
"learning_rate": 3.2813422460687176e-06,
"loss": 0.5082,
"step": 693
},
{
"epoch": 1.9549295774647888,
"grad_norm": 0.37230509461533484,
"learning_rate": 3.265953973185613e-06,
"loss": 0.4771,
"step": 694
},
{
"epoch": 1.9577464788732395,
"grad_norm": 0.3921749151984685,
"learning_rate": 3.250584348160738e-06,
"loss": 0.4842,
"step": 695
},
{
"epoch": 1.9605633802816902,
"grad_norm": 0.39815159280717677,
"learning_rate": 3.235233536278396e-06,
"loss": 0.4876,
"step": 696
},
{
"epoch": 1.963380281690141,
"grad_norm": 0.4078684990729484,
"learning_rate": 3.2199017026205744e-06,
"loss": 0.459,
"step": 697
},
{
"epoch": 1.9661971830985916,
"grad_norm": 0.4003910325405714,
"learning_rate": 3.204589012065168e-06,
"loss": 0.4895,
"step": 698
},
{
"epoch": 1.9690140845070423,
"grad_norm": 0.4098497177330705,
"learning_rate": 3.1892956292842103e-06,
"loss": 0.5077,
"step": 699
},
{
"epoch": 1.971830985915493,
"grad_norm": 0.411890736572199,
"learning_rate": 3.1740217187420947e-06,
"loss": 0.4671,
"step": 700
},
{
"epoch": 1.9746478873239437,
"grad_norm": 0.37605262945047896,
"learning_rate": 3.158767444693819e-06,
"loss": 0.4934,
"step": 701
},
{
"epoch": 1.9774647887323944,
"grad_norm": 0.4119794673411224,
"learning_rate": 3.1435329711832042e-06,
"loss": 0.482,
"step": 702
},
{
"epoch": 1.980281690140845,
"grad_norm": 0.38815162205345255,
"learning_rate": 3.1283184620411387e-06,
"loss": 0.482,
"step": 703
},
{
"epoch": 1.9830985915492958,
"grad_norm": 0.3849898691145318,
"learning_rate": 3.1131240808838175e-06,
"loss": 0.4847,
"step": 704
},
{
"epoch": 1.9859154929577465,
"grad_norm": 0.36667596347362735,
"learning_rate": 3.0979499911109768e-06,
"loss": 0.4687,
"step": 705
},
{
"epoch": 1.9887323943661972,
"grad_norm": 0.3778514259931487,
"learning_rate": 3.082796355904142e-06,
"loss": 0.4577,
"step": 706
},
{
"epoch": 1.991549295774648,
"grad_norm": 0.3770331164124935,
"learning_rate": 3.06766333822487e-06,
"loss": 0.5002,
"step": 707
},
{
"epoch": 1.9943661971830986,
"grad_norm": 0.34604698551150276,
"learning_rate": 3.052551100812998e-06,
"loss": 0.4875,
"step": 708
},
{
"epoch": 1.9971830985915493,
"grad_norm": 0.3631995089390127,
"learning_rate": 3.0374598061848936e-06,
"loss": 0.5077,
"step": 709
},
{
"epoch": 2.0,
"grad_norm": 0.37299214959244026,
"learning_rate": 3.022389616631706e-06,
"loss": 0.4918,
"step": 710
},
{
"epoch": 2.0028169014084507,
"grad_norm": 0.4263218715974723,
"learning_rate": 3.0073406942176214e-06,
"loss": 0.4662,
"step": 711
},
{
"epoch": 2.0056338028169014,
"grad_norm": 0.388952450173628,
"learning_rate": 2.9923132007781206e-06,
"loss": 0.4583,
"step": 712
},
{
"epoch": 2.008450704225352,
"grad_norm": 0.3716230490095555,
"learning_rate": 2.977307297918235e-06,
"loss": 0.452,
"step": 713
},
{
"epoch": 2.011267605633803,
"grad_norm": 0.3472437688297708,
"learning_rate": 2.9623231470108194e-06,
"loss": 0.465,
"step": 714
},
{
"epoch": 2.0140845070422535,
"grad_norm": 0.40516520267399664,
"learning_rate": 2.947360909194801e-06,
"loss": 0.4757,
"step": 715
},
{
"epoch": 2.016901408450704,
"grad_norm": 0.4275349577946564,
"learning_rate": 2.9324207453734575e-06,
"loss": 0.4477,
"step": 716
},
{
"epoch": 2.019718309859155,
"grad_norm": 0.4150387078995605,
"learning_rate": 2.917502816212685e-06,
"loss": 0.4516,
"step": 717
},
{
"epoch": 2.0225352112676056,
"grad_norm": 0.4258725320630231,
"learning_rate": 2.902607282139267e-06,
"loss": 0.4485,
"step": 718
},
{
"epoch": 2.0253521126760563,
"grad_norm": 0.37906226457482306,
"learning_rate": 2.8877343033391523e-06,
"loss": 0.4458,
"step": 719
},
{
"epoch": 2.028169014084507,
"grad_norm": 0.4017366735290866,
"learning_rate": 2.8728840397557324e-06,
"loss": 0.4533,
"step": 720
},
{
"epoch": 2.0309859154929577,
"grad_norm": 0.4636710305852195,
"learning_rate": 2.8580566510881158e-06,
"loss": 0.4513,
"step": 721
},
{
"epoch": 2.0338028169014084,
"grad_norm": 0.40403337840200826,
"learning_rate": 2.8432522967894217e-06,
"loss": 0.4466,
"step": 722
},
{
"epoch": 2.036619718309859,
"grad_norm": 0.3839776520892033,
"learning_rate": 2.8284711360650517e-06,
"loss": 0.4582,
"step": 723
},
{
"epoch": 2.03943661971831,
"grad_norm": 0.3957971040790153,
"learning_rate": 2.8137133278709917e-06,
"loss": 0.4673,
"step": 724
},
{
"epoch": 2.0422535211267605,
"grad_norm": 0.4511172081773506,
"learning_rate": 2.7989790309120895e-06,
"loss": 0.4661,
"step": 725
},
{
"epoch": 2.045070422535211,
"grad_norm": 0.42321807954410795,
"learning_rate": 2.7842684036403557e-06,
"loss": 0.4596,
"step": 726
},
{
"epoch": 2.047887323943662,
"grad_norm": 0.4016823973163568,
"learning_rate": 2.769581604253262e-06,
"loss": 0.4572,
"step": 727
},
{
"epoch": 2.0507042253521126,
"grad_norm": 0.35731206857199754,
"learning_rate": 2.754918790692031e-06,
"loss": 0.4496,
"step": 728
},
{
"epoch": 2.0535211267605633,
"grad_norm": 0.3760759988713752,
"learning_rate": 2.7402801206399454e-06,
"loss": 0.4627,
"step": 729
},
{
"epoch": 2.056338028169014,
"grad_norm": 0.36312345984169353,
"learning_rate": 2.7256657515206487e-06,
"loss": 0.4558,
"step": 730
},
{
"epoch": 2.0591549295774647,
"grad_norm": 0.3660535112601631,
"learning_rate": 2.7110758404964534e-06,
"loss": 0.4655,
"step": 731
},
{
"epoch": 2.0619718309859154,
"grad_norm": 0.39210025358373574,
"learning_rate": 2.6965105444666496e-06,
"loss": 0.4609,
"step": 732
},
{
"epoch": 2.064788732394366,
"grad_norm": 0.32546400486520366,
"learning_rate": 2.6819700200658204e-06,
"loss": 0.4469,
"step": 733
},
{
"epoch": 2.067605633802817,
"grad_norm": 0.41216043411606135,
"learning_rate": 2.667454423662156e-06,
"loss": 0.4491,
"step": 734
},
{
"epoch": 2.0704225352112675,
"grad_norm": 0.37826214280359827,
"learning_rate": 2.6529639113557694e-06,
"loss": 0.473,
"step": 735
},
{
"epoch": 2.073239436619718,
"grad_norm": 0.3872061532277833,
"learning_rate": 2.638498638977024e-06,
"loss": 0.4462,
"step": 736
},
{
"epoch": 2.076056338028169,
"grad_norm": 0.35777642541297033,
"learning_rate": 2.6240587620848512e-06,
"loss": 0.4545,
"step": 737
},
{
"epoch": 2.0788732394366196,
"grad_norm": 0.3771646503674181,
"learning_rate": 2.6096444359650817e-06,
"loss": 0.4546,
"step": 738
},
{
"epoch": 2.0816901408450703,
"grad_norm": 0.369712629785824,
"learning_rate": 2.595255815628774e-06,
"loss": 0.426,
"step": 739
},
{
"epoch": 2.084507042253521,
"grad_norm": 0.3948340446779534,
"learning_rate": 2.580893055810545e-06,
"loss": 0.4396,
"step": 740
},
{
"epoch": 2.0873239436619717,
"grad_norm": 0.362237396122223,
"learning_rate": 2.5665563109669155e-06,
"loss": 0.4572,
"step": 741
},
{
"epoch": 2.0901408450704224,
"grad_norm": 0.35377063841196926,
"learning_rate": 2.5522457352746368e-06,
"loss": 0.4484,
"step": 742
},
{
"epoch": 2.092957746478873,
"grad_norm": 0.37686845068586383,
"learning_rate": 2.5379614826290384e-06,
"loss": 0.4659,
"step": 743
},
{
"epoch": 2.095774647887324,
"grad_norm": 0.35203810361621596,
"learning_rate": 2.5237037066423747e-06,
"loss": 0.4473,
"step": 744
},
{
"epoch": 2.0985915492957745,
"grad_norm": 0.3591372556166686,
"learning_rate": 2.509472560642171e-06,
"loss": 0.4597,
"step": 745
},
{
"epoch": 2.101408450704225,
"grad_norm": 0.3708152047071068,
"learning_rate": 2.495268197669573e-06,
"loss": 0.453,
"step": 746
},
{
"epoch": 2.104225352112676,
"grad_norm": 0.3902747213222871,
"learning_rate": 2.481090770477706e-06,
"loss": 0.436,
"step": 747
},
{
"epoch": 2.1070422535211266,
"grad_norm": 0.3887109344687831,
"learning_rate": 2.466940431530026e-06,
"loss": 0.4295,
"step": 748
},
{
"epoch": 2.1098591549295773,
"grad_norm": 0.3958388402882215,
"learning_rate": 2.4528173329986855e-06,
"loss": 0.4481,
"step": 749
},
{
"epoch": 2.112676056338028,
"grad_norm": 0.35749353391449623,
"learning_rate": 2.438721626762892e-06,
"loss": 0.443,
"step": 750
},
{
"epoch": 2.1154929577464787,
"grad_norm": 0.39476997529270275,
"learning_rate": 2.424653464407281e-06,
"loss": 0.4846,
"step": 751
},
{
"epoch": 2.1183098591549294,
"grad_norm": 0.3575214685484145,
"learning_rate": 2.4106129972202793e-06,
"loss": 0.4464,
"step": 752
},
{
"epoch": 2.12112676056338,
"grad_norm": 0.3768049884226859,
"learning_rate": 2.3966003761924816e-06,
"loss": 0.466,
"step": 753
},
{
"epoch": 2.123943661971831,
"grad_norm": 0.4018878366687561,
"learning_rate": 2.382615752015028e-06,
"loss": 0.4601,
"step": 754
},
{
"epoch": 2.1267605633802815,
"grad_norm": 0.40719785808826275,
"learning_rate": 2.3686592750779788e-06,
"loss": 0.4441,
"step": 755
},
{
"epoch": 2.129577464788732,
"grad_norm": 0.3440143155418727,
"learning_rate": 2.3547310954687018e-06,
"loss": 0.477,
"step": 756
},
{
"epoch": 2.132394366197183,
"grad_norm": 0.35333903604899564,
"learning_rate": 2.340831362970257e-06,
"loss": 0.462,
"step": 757
},
{
"epoch": 2.1352112676056336,
"grad_norm": 0.40408511951747045,
"learning_rate": 2.326960227059784e-06,
"loss": 0.4593,
"step": 758
},
{
"epoch": 2.1380281690140843,
"grad_norm": 0.38619122486168744,
"learning_rate": 2.313117836906897e-06,
"loss": 0.4373,
"step": 759
},
{
"epoch": 2.140845070422535,
"grad_norm": 0.411758597632312,
"learning_rate": 2.2993043413720784e-06,
"loss": 0.4655,
"step": 760
},
{
"epoch": 2.1436619718309857,
"grad_norm": 0.3576311304270278,
"learning_rate": 2.28551988900508e-06,
"loss": 0.4465,
"step": 761
},
{
"epoch": 2.1464788732394364,
"grad_norm": 0.3357550677548821,
"learning_rate": 2.271764628043324e-06,
"loss": 0.4769,
"step": 762
},
{
"epoch": 2.149295774647887,
"grad_norm": 0.37905960503822633,
"learning_rate": 2.258038706410311e-06,
"loss": 0.4319,
"step": 763
},
{
"epoch": 2.152112676056338,
"grad_norm": 0.385174993785128,
"learning_rate": 2.2443422717140246e-06,
"loss": 0.4548,
"step": 764
},
{
"epoch": 2.1549295774647885,
"grad_norm": 0.37165354454365357,
"learning_rate": 2.2306754712453504e-06,
"loss": 0.4635,
"step": 765
},
{
"epoch": 2.1577464788732392,
"grad_norm": 0.36455549759311245,
"learning_rate": 2.217038451976485e-06,
"loss": 0.4446,
"step": 766
},
{
"epoch": 2.1605633802816904,
"grad_norm": 0.36114868485444035,
"learning_rate": 2.203431360559365e-06,
"loss": 0.4464,
"step": 767
},
{
"epoch": 2.1633802816901406,
"grad_norm": 0.37195290477502,
"learning_rate": 2.1898543433240787e-06,
"loss": 0.4566,
"step": 768
},
{
"epoch": 2.1661971830985918,
"grad_norm": 0.35611893011878004,
"learning_rate": 2.1763075462773002e-06,
"loss": 0.4405,
"step": 769
},
{
"epoch": 2.169014084507042,
"grad_norm": 0.3568351811276538,
"learning_rate": 2.1627911151007176e-06,
"loss": 0.4468,
"step": 770
},
{
"epoch": 2.171830985915493,
"grad_norm": 0.37439678125672593,
"learning_rate": 2.149305195149463e-06,
"loss": 0.4479,
"step": 771
},
{
"epoch": 2.1746478873239434,
"grad_norm": 0.34449509209968837,
"learning_rate": 2.135849931450555e-06,
"loss": 0.4565,
"step": 772
},
{
"epoch": 2.1774647887323946,
"grad_norm": 0.3525754840068429,
"learning_rate": 2.1224254687013347e-06,
"loss": 0.4648,
"step": 773
},
{
"epoch": 2.1802816901408453,
"grad_norm": 0.3533752171342116,
"learning_rate": 2.1090319512679115e-06,
"loss": 0.4611,
"step": 774
},
{
"epoch": 2.183098591549296,
"grad_norm": 0.37886186027245655,
"learning_rate": 2.0956695231836094e-06,
"loss": 0.4603,
"step": 775
},
{
"epoch": 2.1859154929577467,
"grad_norm": 0.3733834136596295,
"learning_rate": 2.0823383281474202e-06,
"loss": 0.4351,
"step": 776
},
{
"epoch": 2.1887323943661974,
"grad_norm": 0.3728175824778924,
"learning_rate": 2.0690385095224557e-06,
"loss": 0.4788,
"step": 777
},
{
"epoch": 2.191549295774648,
"grad_norm": 0.33738192713218457,
"learning_rate": 2.0557702103344078e-06,
"loss": 0.4627,
"step": 778
},
{
"epoch": 2.1943661971830988,
"grad_norm": 0.3415766698066426,
"learning_rate": 2.0425335732700075e-06,
"loss": 0.4494,
"step": 779
},
{
"epoch": 2.1971830985915495,
"grad_norm": 0.35037808653487684,
"learning_rate": 2.0293287406754976e-06,
"loss": 0.4618,
"step": 780
},
{
"epoch": 2.2,
"grad_norm": 0.3635106636596117,
"learning_rate": 2.0161558545550925e-06,
"loss": 0.458,
"step": 781
},
{
"epoch": 2.202816901408451,
"grad_norm": 0.3305404475827585,
"learning_rate": 2.0030150565694566e-06,
"loss": 0.4566,
"step": 782
},
{
"epoch": 2.2056338028169016,
"grad_norm": 0.36482202121638047,
"learning_rate": 1.989906488034181e-06,
"loss": 0.4296,
"step": 783
},
{
"epoch": 2.2084507042253523,
"grad_norm": 0.38281315419014894,
"learning_rate": 1.976830289918261e-06,
"loss": 0.4497,
"step": 784
},
{
"epoch": 2.211267605633803,
"grad_norm": 0.3401713857213208,
"learning_rate": 1.9637866028425835e-06,
"loss": 0.4579,
"step": 785
},
{
"epoch": 2.2140845070422537,
"grad_norm": 0.38608594850184313,
"learning_rate": 1.950775567078411e-06,
"loss": 0.4632,
"step": 786
},
{
"epoch": 2.2169014084507044,
"grad_norm": 0.377240139541956,
"learning_rate": 1.9377973225458773e-06,
"loss": 0.4448,
"step": 787
},
{
"epoch": 2.219718309859155,
"grad_norm": 0.3486255139361687,
"learning_rate": 1.924852008812479e-06,
"loss": 0.4786,
"step": 788
},
{
"epoch": 2.222535211267606,
"grad_norm": 0.35076148165325083,
"learning_rate": 1.9119397650915774e-06,
"loss": 0.4462,
"step": 789
},
{
"epoch": 2.2253521126760565,
"grad_norm": 0.3531332691720524,
"learning_rate": 1.8990607302408991e-06,
"loss": 0.456,
"step": 790
},
{
"epoch": 2.228169014084507,
"grad_norm": 0.3366858659315378,
"learning_rate": 1.8862150427610453e-06,
"loss": 0.4326,
"step": 791
},
{
"epoch": 2.230985915492958,
"grad_norm": 0.3636085103892985,
"learning_rate": 1.8734028407940003e-06,
"loss": 0.4444,
"step": 792
},
{
"epoch": 2.2338028169014086,
"grad_norm": 0.38948440503258513,
"learning_rate": 1.8606242621216443e-06,
"loss": 0.4569,
"step": 793
},
{
"epoch": 2.2366197183098593,
"grad_norm": 0.3780810617916765,
"learning_rate": 1.84787944416428e-06,
"loss": 0.4694,
"step": 794
},
{
"epoch": 2.23943661971831,
"grad_norm": 0.3603299392400602,
"learning_rate": 1.835168523979144e-06,
"loss": 0.447,
"step": 795
},
{
"epoch": 2.2422535211267607,
"grad_norm": 0.38407466446122274,
"learning_rate": 1.822491638258938e-06,
"loss": 0.4648,
"step": 796
},
{
"epoch": 2.2450704225352114,
"grad_norm": 0.37915726449650966,
"learning_rate": 1.8098489233303595e-06,
"loss": 0.4677,
"step": 797
},
{
"epoch": 2.247887323943662,
"grad_norm": 0.39145196282091305,
"learning_rate": 1.7972405151526339e-06,
"loss": 0.4641,
"step": 798
},
{
"epoch": 2.250704225352113,
"grad_norm": 0.34475277984897656,
"learning_rate": 1.7846665493160548e-06,
"loss": 0.4529,
"step": 799
},
{
"epoch": 2.2535211267605635,
"grad_norm": 0.3704879117075785,
"learning_rate": 1.7721271610405232e-06,
"loss": 0.4305,
"step": 800
},
{
"epoch": 2.256338028169014,
"grad_norm": 0.37161976738203256,
"learning_rate": 1.7596224851740938e-06,
"loss": 0.4825,
"step": 801
},
{
"epoch": 2.259154929577465,
"grad_norm": 0.3524012690682922,
"learning_rate": 1.747152656191527e-06,
"loss": 0.4266,
"step": 802
},
{
"epoch": 2.2619718309859156,
"grad_norm": 0.3753464662024848,
"learning_rate": 1.7347178081928407e-06,
"loss": 0.4544,
"step": 803
},
{
"epoch": 2.2647887323943663,
"grad_norm": 0.34831747174644373,
"learning_rate": 1.7223180749018675e-06,
"loss": 0.4749,
"step": 804
},
{
"epoch": 2.267605633802817,
"grad_norm": 0.3790722784544341,
"learning_rate": 1.7099535896648205e-06,
"loss": 0.4571,
"step": 805
},
{
"epoch": 2.2704225352112677,
"grad_norm": 0.3529967844043124,
"learning_rate": 1.6976244854488545e-06,
"loss": 0.4538,
"step": 806
},
{
"epoch": 2.2732394366197184,
"grad_norm": 0.36580993633176867,
"learning_rate": 1.6853308948406387e-06,
"loss": 0.4559,
"step": 807
},
{
"epoch": 2.276056338028169,
"grad_norm": 0.33773543663604155,
"learning_rate": 1.6730729500449322e-06,
"loss": 0.4491,
"step": 808
},
{
"epoch": 2.27887323943662,
"grad_norm": 0.3818724429980544,
"learning_rate": 1.6608507828831572e-06,
"loss": 0.4471,
"step": 809
},
{
"epoch": 2.2816901408450705,
"grad_norm": 0.3941087744995697,
"learning_rate": 1.648664524791988e-06,
"loss": 0.4477,
"step": 810
},
{
"epoch": 2.284507042253521,
"grad_norm": 0.38090837598627414,
"learning_rate": 1.636514306821933e-06,
"loss": 0.4517,
"step": 811
},
{
"epoch": 2.287323943661972,
"grad_norm": 0.34949670113720865,
"learning_rate": 1.6244002596359255e-06,
"loss": 0.4522,
"step": 812
},
{
"epoch": 2.2901408450704226,
"grad_norm": 0.3547350758000349,
"learning_rate": 1.6123225135079212e-06,
"loss": 0.4562,
"step": 813
},
{
"epoch": 2.2929577464788733,
"grad_norm": 0.38160013725900366,
"learning_rate": 1.6002811983214962e-06,
"loss": 0.4669,
"step": 814
},
{
"epoch": 2.295774647887324,
"grad_norm": 0.35170542214837286,
"learning_rate": 1.5882764435684473e-06,
"loss": 0.4547,
"step": 815
},
{
"epoch": 2.2985915492957747,
"grad_norm": 0.3465195660150116,
"learning_rate": 1.5763083783474048e-06,
"loss": 0.4701,
"step": 816
},
{
"epoch": 2.3014084507042254,
"grad_norm": 0.38491085471157266,
"learning_rate": 1.5643771313624394e-06,
"loss": 0.4272,
"step": 817
},
{
"epoch": 2.304225352112676,
"grad_norm": 0.3759115395560408,
"learning_rate": 1.552482830921681e-06,
"loss": 0.4707,
"step": 818
},
{
"epoch": 2.307042253521127,
"grad_norm": 0.3594305645877611,
"learning_rate": 1.5406256049359359e-06,
"loss": 0.4775,
"step": 819
},
{
"epoch": 2.3098591549295775,
"grad_norm": 0.33521577112471695,
"learning_rate": 1.5288055809173174e-06,
"loss": 0.4516,
"step": 820
},
{
"epoch": 2.312676056338028,
"grad_norm": 0.3674079420964908,
"learning_rate": 1.517022885977868e-06,
"loss": 0.4617,
"step": 821
},
{
"epoch": 2.315492957746479,
"grad_norm": 0.34186215662243075,
"learning_rate": 1.5052776468281933e-06,
"loss": 0.4335,
"step": 822
},
{
"epoch": 2.3183098591549296,
"grad_norm": 0.3958145612608977,
"learning_rate": 1.4935699897761031e-06,
"loss": 0.4662,
"step": 823
},
{
"epoch": 2.3211267605633803,
"grad_norm": 0.3898936062833255,
"learning_rate": 1.4819000407252498e-06,
"loss": 0.4604,
"step": 824
},
{
"epoch": 2.323943661971831,
"grad_norm": 0.3566184175212025,
"learning_rate": 1.4702679251737768e-06,
"loss": 0.4691,
"step": 825
},
{
"epoch": 2.3267605633802817,
"grad_norm": 0.3556640563120501,
"learning_rate": 1.4586737682129653e-06,
"loss": 0.4419,
"step": 826
},
{
"epoch": 2.3295774647887324,
"grad_norm": 0.3482765499624595,
"learning_rate": 1.4471176945258947e-06,
"loss": 0.464,
"step": 827
},
{
"epoch": 2.332394366197183,
"grad_norm": 0.37652126714004625,
"learning_rate": 1.435599828386095e-06,
"loss": 0.4478,
"step": 828
},
{
"epoch": 2.335211267605634,
"grad_norm": 0.34935552553112986,
"learning_rate": 1.4241202936562164e-06,
"loss": 0.4529,
"step": 829
},
{
"epoch": 2.3380281690140845,
"grad_norm": 0.319947879082873,
"learning_rate": 1.412679213786694e-06,
"loss": 0.4521,
"step": 830
},
{
"epoch": 2.340845070422535,
"grad_norm": 0.36668567938711244,
"learning_rate": 1.40127671181442e-06,
"loss": 0.4502,
"step": 831
},
{
"epoch": 2.343661971830986,
"grad_norm": 0.3408592392806045,
"learning_rate": 1.3899129103614229e-06,
"loss": 0.4566,
"step": 832
},
{
"epoch": 2.3464788732394366,
"grad_norm": 0.3670600399372913,
"learning_rate": 1.3785879316335448e-06,
"loss": 0.4362,
"step": 833
},
{
"epoch": 2.3492957746478873,
"grad_norm": 0.3542407917474685,
"learning_rate": 1.3673018974191354e-06,
"loss": 0.4475,
"step": 834
},
{
"epoch": 2.352112676056338,
"grad_norm": 0.3510002567312064,
"learning_rate": 1.3560549290877318e-06,
"loss": 0.4598,
"step": 835
},
{
"epoch": 2.3549295774647887,
"grad_norm": 0.3547228481357092,
"learning_rate": 1.3448471475887587e-06,
"loss": 0.4728,
"step": 836
},
{
"epoch": 2.3577464788732394,
"grad_norm": 0.345967228254802,
"learning_rate": 1.3336786734502294e-06,
"loss": 0.4392,
"step": 837
},
{
"epoch": 2.36056338028169,
"grad_norm": 0.3614460954714215,
"learning_rate": 1.3225496267774452e-06,
"loss": 0.4492,
"step": 838
},
{
"epoch": 2.363380281690141,
"grad_norm": 0.364394650984224,
"learning_rate": 1.311460127251708e-06,
"loss": 0.4733,
"step": 839
},
{
"epoch": 2.3661971830985915,
"grad_norm": 0.35019486310976833,
"learning_rate": 1.3004102941290297e-06,
"loss": 0.4441,
"step": 840
},
{
"epoch": 2.369014084507042,
"grad_norm": 0.3706637778515585,
"learning_rate": 1.2894002462388533e-06,
"loss": 0.4575,
"step": 841
},
{
"epoch": 2.371830985915493,
"grad_norm": 0.32959072572438813,
"learning_rate": 1.2784301019827705e-06,
"loss": 0.4596,
"step": 842
},
{
"epoch": 2.3746478873239436,
"grad_norm": 0.3424607431994825,
"learning_rate": 1.2674999793332539e-06,
"loss": 0.4431,
"step": 843
},
{
"epoch": 2.3774647887323943,
"grad_norm": 0.3676756813732258,
"learning_rate": 1.2566099958323824e-06,
"loss": 0.4356,
"step": 844
},
{
"epoch": 2.380281690140845,
"grad_norm": 0.36058800592394114,
"learning_rate": 1.2457602685905834e-06,
"loss": 0.4316,
"step": 845
},
{
"epoch": 2.3830985915492957,
"grad_norm": 0.3585818367347638,
"learning_rate": 1.2349509142853672e-06,
"loss": 0.4674,
"step": 846
},
{
"epoch": 2.3859154929577464,
"grad_norm": 0.34790992953742333,
"learning_rate": 1.224182049160077e-06,
"loss": 0.4646,
"step": 847
},
{
"epoch": 2.388732394366197,
"grad_norm": 0.36177326432975354,
"learning_rate": 1.2134537890226366e-06,
"loss": 0.4697,
"step": 848
},
{
"epoch": 2.391549295774648,
"grad_norm": 0.3503908592783261,
"learning_rate": 1.202766249244306e-06,
"loss": 0.4678,
"step": 849
},
{
"epoch": 2.3943661971830985,
"grad_norm": 0.34200180240853983,
"learning_rate": 1.1921195447584388e-06,
"loss": 0.4549,
"step": 850
},
{
"epoch": 2.397183098591549,
"grad_norm": 0.3449841650449462,
"learning_rate": 1.1815137900592488e-06,
"loss": 0.4514,
"step": 851
},
{
"epoch": 2.4,
"grad_norm": 0.32359310287845633,
"learning_rate": 1.1709490992005774e-06,
"loss": 0.4587,
"step": 852
},
{
"epoch": 2.4028169014084506,
"grad_norm": 0.362853451889369,
"learning_rate": 1.1604255857946667e-06,
"loss": 0.4528,
"step": 853
},
{
"epoch": 2.4056338028169013,
"grad_norm": 0.3824838433363387,
"learning_rate": 1.149943363010938e-06,
"loss": 0.4692,
"step": 854
},
{
"epoch": 2.408450704225352,
"grad_norm": 0.309820114431787,
"learning_rate": 1.1395025435747759e-06,
"loss": 0.4605,
"step": 855
},
{
"epoch": 2.4112676056338027,
"grad_norm": 0.367780334595774,
"learning_rate": 1.1291032397663142e-06,
"loss": 0.4378,
"step": 856
},
{
"epoch": 2.4140845070422534,
"grad_norm": 0.31307329625243296,
"learning_rate": 1.1187455634192307e-06,
"loss": 0.453,
"step": 857
},
{
"epoch": 2.416901408450704,
"grad_norm": 0.3432753242652767,
"learning_rate": 1.108429625919541e-06,
"loss": 0.464,
"step": 858
},
{
"epoch": 2.419718309859155,
"grad_norm": 0.35797494481992975,
"learning_rate": 1.098155538204404e-06,
"loss": 0.4343,
"step": 859
},
{
"epoch": 2.4225352112676055,
"grad_norm": 0.35317578270799466,
"learning_rate": 1.08792341076093e-06,
"loss": 0.4677,
"step": 860
},
{
"epoch": 2.425352112676056,
"grad_norm": 0.36299248843589665,
"learning_rate": 1.0777333536249873e-06,
"loss": 0.4479,
"step": 861
},
{
"epoch": 2.428169014084507,
"grad_norm": 0.35957601988988697,
"learning_rate": 1.067585476380023e-06,
"loss": 0.4423,
"step": 862
},
{
"epoch": 2.4309859154929576,
"grad_norm": 0.3379973267546356,
"learning_rate": 1.0574798881558834e-06,
"loss": 0.4411,
"step": 863
},
{
"epoch": 2.4338028169014083,
"grad_norm": 0.3471607834308851,
"learning_rate": 1.0474166976276396e-06,
"loss": 0.4478,
"step": 864
},
{
"epoch": 2.436619718309859,
"grad_norm": 0.35472546641648245,
"learning_rate": 1.0373960130144206e-06,
"loss": 0.4522,
"step": 865
},
{
"epoch": 2.4394366197183097,
"grad_norm": 0.3587010132644766,
"learning_rate": 1.0274179420782487e-06,
"loss": 0.4792,
"step": 866
},
{
"epoch": 2.4422535211267604,
"grad_norm": 0.3462299368390657,
"learning_rate": 1.0174825921228803e-06,
"loss": 0.4453,
"step": 867
},
{
"epoch": 2.445070422535211,
"grad_norm": 0.32780133736017886,
"learning_rate": 1.0075900699926523e-06,
"loss": 0.4421,
"step": 868
},
{
"epoch": 2.447887323943662,
"grad_norm": 0.32980232448683194,
"learning_rate": 9.977404820713315e-07,
"loss": 0.4354,
"step": 869
},
{
"epoch": 2.4507042253521125,
"grad_norm": 0.35953716883112774,
"learning_rate": 9.879339342809741e-07,
"loss": 0.4674,
"step": 870
},
{
"epoch": 2.453521126760563,
"grad_norm": 0.3149172395510989,
"learning_rate": 9.781705320807833e-07,
"loss": 0.448,
"step": 871
},
{
"epoch": 2.456338028169014,
"grad_norm": 0.3599954533086056,
"learning_rate": 9.684503804659773e-07,
"loss": 0.4487,
"step": 872
},
{
"epoch": 2.4591549295774646,
"grad_norm": 0.3310011588753474,
"learning_rate": 9.587735839666573e-07,
"loss": 0.431,
"step": 873
},
{
"epoch": 2.4619718309859153,
"grad_norm": 0.3367067646735182,
"learning_rate": 9.491402466466893e-07,
"loss": 0.4454,
"step": 874
},
{
"epoch": 2.464788732394366,
"grad_norm": 0.3403390583206632,
"learning_rate": 9.395504721025773e-07,
"loss": 0.4463,
"step": 875
},
{
"epoch": 2.4676056338028167,
"grad_norm": 0.34010604525094434,
"learning_rate": 9.300043634623546e-07,
"loss": 0.4462,
"step": 876
},
{
"epoch": 2.4704225352112674,
"grad_norm": 0.34940346894141644,
"learning_rate": 9.205020233844736e-07,
"loss": 0.4389,
"step": 877
},
{
"epoch": 2.473239436619718,
"grad_norm": 0.33818530430000493,
"learning_rate": 9.110435540567003e-07,
"loss": 0.4528,
"step": 878
},
{
"epoch": 2.476056338028169,
"grad_norm": 0.32513521696510156,
"learning_rate": 9.016290571950171e-07,
"loss": 0.4475,
"step": 879
},
{
"epoch": 2.4788732394366195,
"grad_norm": 0.348467054421719,
"learning_rate": 8.922586340425288e-07,
"loss": 0.4434,
"step": 880
},
{
"epoch": 2.4816901408450702,
"grad_norm": 0.34598411047159433,
"learning_rate": 8.829323853683719e-07,
"loss": 0.4764,
"step": 881
},
{
"epoch": 2.4845070422535214,
"grad_norm": 0.33736914824699643,
"learning_rate": 8.736504114666345e-07,
"loss": 0.4477,
"step": 882
},
{
"epoch": 2.4873239436619716,
"grad_norm": 0.34008166784157434,
"learning_rate": 8.644128121552742e-07,
"loss": 0.4529,
"step": 883
},
{
"epoch": 2.4901408450704228,
"grad_norm": 0.31419004202982526,
"learning_rate": 8.552196867750462e-07,
"loss": 0.4575,
"step": 884
},
{
"epoch": 2.492957746478873,
"grad_norm": 0.38346622415801657,
"learning_rate": 8.460711341884353e-07,
"loss": 0.461,
"step": 885
},
{
"epoch": 2.495774647887324,
"grad_norm": 0.36160333663713207,
"learning_rate": 8.36967252778591e-07,
"loss": 0.4704,
"step": 886
},
{
"epoch": 2.4985915492957744,
"grad_norm": 0.3426582254986613,
"learning_rate": 8.279081404482748e-07,
"loss": 0.449,
"step": 887
},
{
"epoch": 2.5014084507042256,
"grad_norm": 0.3362056509073023,
"learning_rate": 8.188938946187991e-07,
"loss": 0.4551,
"step": 888
},
{
"epoch": 2.504225352112676,
"grad_norm": 0.35776312404702704,
"learning_rate": 8.099246122289861e-07,
"loss": 0.4616,
"step": 889
},
{
"epoch": 2.507042253521127,
"grad_norm": 0.3516358075048254,
"learning_rate": 8.010003897341212e-07,
"loss": 0.4354,
"step": 890
},
{
"epoch": 2.5098591549295772,
"grad_norm": 0.37697790136846654,
"learning_rate": 7.921213231049196e-07,
"loss": 0.4655,
"step": 891
},
{
"epoch": 2.5126760563380284,
"grad_norm": 0.338057431043845,
"learning_rate": 7.832875078264912e-07,
"loss": 0.4607,
"step": 892
},
{
"epoch": 2.5154929577464786,
"grad_norm": 0.34095350497165633,
"learning_rate": 7.74499038897315e-07,
"loss": 0.4537,
"step": 893
},
{
"epoch": 2.5183098591549298,
"grad_norm": 0.3626292402362458,
"learning_rate": 7.65756010828217e-07,
"loss": 0.4437,
"step": 894
},
{
"epoch": 2.52112676056338,
"grad_norm": 0.33536228545310875,
"learning_rate": 7.570585176413547e-07,
"loss": 0.4506,
"step": 895
},
{
"epoch": 2.523943661971831,
"grad_norm": 0.34757819690062497,
"learning_rate": 7.484066528692041e-07,
"loss": 0.4533,
"step": 896
},
{
"epoch": 2.5267605633802814,
"grad_norm": 0.3544426851266078,
"learning_rate": 7.398005095535565e-07,
"loss": 0.455,
"step": 897
},
{
"epoch": 2.5295774647887326,
"grad_norm": 0.35854954315916865,
"learning_rate": 7.312401802445169e-07,
"loss": 0.4554,
"step": 898
},
{
"epoch": 2.532394366197183,
"grad_norm": 0.34802065964448864,
"learning_rate": 7.227257569995061e-07,
"loss": 0.4418,
"step": 899
},
{
"epoch": 2.535211267605634,
"grad_norm": 0.33929091253717597,
"learning_rate": 7.142573313822754e-07,
"loss": 0.4546,
"step": 900
},
{
"epoch": 2.5380281690140842,
"grad_norm": 0.3455549381014243,
"learning_rate": 7.058349944619186e-07,
"loss": 0.4551,
"step": 901
},
{
"epoch": 2.5408450704225354,
"grad_norm": 0.3730269906106179,
"learning_rate": 6.974588368118934e-07,
"loss": 0.4418,
"step": 902
},
{
"epoch": 2.543661971830986,
"grad_norm": 0.3586546013910396,
"learning_rate": 6.89128948509048e-07,
"loss": 0.4409,
"step": 903
},
{
"epoch": 2.546478873239437,
"grad_norm": 0.3213019984256461,
"learning_rate": 6.808454191326519e-07,
"loss": 0.4395,
"step": 904
},
{
"epoch": 2.5492957746478875,
"grad_norm": 0.3260075624126433,
"learning_rate": 6.726083377634318e-07,
"loss": 0.4524,
"step": 905
},
{
"epoch": 2.552112676056338,
"grad_norm": 0.33357432795345565,
"learning_rate": 6.644177929826162e-07,
"loss": 0.464,
"step": 906
},
{
"epoch": 2.554929577464789,
"grad_norm": 0.36335471129130653,
"learning_rate": 6.562738728709795e-07,
"loss": 0.4508,
"step": 907
},
{
"epoch": 2.5577464788732396,
"grad_norm": 0.3648297045285258,
"learning_rate": 6.481766650078969e-07,
"loss": 0.4448,
"step": 908
},
{
"epoch": 2.5605633802816903,
"grad_norm": 0.3428914158054443,
"learning_rate": 6.401262564704019e-07,
"loss": 0.456,
"step": 909
},
{
"epoch": 2.563380281690141,
"grad_norm": 0.33333366790531865,
"learning_rate": 6.321227338322511e-07,
"loss": 0.4627,
"step": 910
},
{
"epoch": 2.5661971830985917,
"grad_norm": 0.34912392956937827,
"learning_rate": 6.241661831629902e-07,
"loss": 0.4632,
"step": 911
},
{
"epoch": 2.5690140845070424,
"grad_norm": 0.34822142796084354,
"learning_rate": 6.162566900270311e-07,
"loss": 0.462,
"step": 912
},
{
"epoch": 2.571830985915493,
"grad_norm": 0.31962377759459676,
"learning_rate": 6.083943394827329e-07,
"loss": 0.4153,
"step": 913
},
{
"epoch": 2.574647887323944,
"grad_norm": 0.33869239576486176,
"learning_rate": 6.005792160814821e-07,
"loss": 0.4401,
"step": 914
},
{
"epoch": 2.5774647887323945,
"grad_norm": 0.33793034444750547,
"learning_rate": 5.928114038667888e-07,
"loss": 0.4726,
"step": 915
},
{
"epoch": 2.580281690140845,
"grad_norm": 0.3237369136857151,
"learning_rate": 5.850909863733784e-07,
"loss": 0.438,
"step": 916
},
{
"epoch": 2.583098591549296,
"grad_norm": 0.3375720011474295,
"learning_rate": 5.774180466262985e-07,
"loss": 0.4344,
"step": 917
},
{
"epoch": 2.5859154929577466,
"grad_norm": 0.3580068599295479,
"learning_rate": 5.697926671400194e-07,
"loss": 0.4512,
"step": 918
},
{
"epoch": 2.5887323943661973,
"grad_norm": 0.34625637202820847,
"learning_rate": 5.622149299175539e-07,
"loss": 0.4609,
"step": 919
},
{
"epoch": 2.591549295774648,
"grad_norm": 0.34840785097504035,
"learning_rate": 5.546849164495688e-07,
"loss": 0.4423,
"step": 920
},
{
"epoch": 2.5943661971830987,
"grad_norm": 0.3325264584554415,
"learning_rate": 5.472027077135145e-07,
"loss": 0.4358,
"step": 921
},
{
"epoch": 2.5971830985915494,
"grad_norm": 0.3412839911435819,
"learning_rate": 5.397683841727485e-07,
"loss": 0.4217,
"step": 922
},
{
"epoch": 2.6,
"grad_norm": 0.31438716871686184,
"learning_rate": 5.323820257756745e-07,
"loss": 0.4428,
"step": 923
},
{
"epoch": 2.602816901408451,
"grad_norm": 0.335053022759397,
"learning_rate": 5.250437119548817e-07,
"loss": 0.4472,
"step": 924
},
{
"epoch": 2.6056338028169015,
"grad_norm": 0.3386301442199217,
"learning_rate": 5.177535216262885e-07,
"loss": 0.4469,
"step": 925
},
{
"epoch": 2.608450704225352,
"grad_norm": 0.3402630657380998,
"learning_rate": 5.105115331882954e-07,
"loss": 0.4717,
"step": 926
},
{
"epoch": 2.611267605633803,
"grad_norm": 0.33515652533891976,
"learning_rate": 5.033178245209436e-07,
"loss": 0.4606,
"step": 927
},
{
"epoch": 2.6140845070422536,
"grad_norm": 0.37526213127846714,
"learning_rate": 4.961724729850731e-07,
"loss": 0.4673,
"step": 928
},
{
"epoch": 2.6169014084507043,
"grad_norm": 0.3297554414539737,
"learning_rate": 4.890755554214954e-07,
"loss": 0.4298,
"step": 929
},
{
"epoch": 2.619718309859155,
"grad_norm": 0.3757308737779317,
"learning_rate": 4.820271481501642e-07,
"loss": 0.4469,
"step": 930
},
{
"epoch": 2.6225352112676057,
"grad_norm": 0.3560626177653075,
"learning_rate": 4.7502732696935507e-07,
"loss": 0.4478,
"step": 931
},
{
"epoch": 2.6253521126760564,
"grad_norm": 0.3566774647442687,
"learning_rate": 4.680761671548517e-07,
"loss": 0.4594,
"step": 932
},
{
"epoch": 2.628169014084507,
"grad_norm": 0.32636544667152256,
"learning_rate": 4.6117374345913454e-07,
"loss": 0.4491,
"step": 933
},
{
"epoch": 2.630985915492958,
"grad_norm": 0.3297355569759064,
"learning_rate": 4.5432013011057984e-07,
"loss": 0.4562,
"step": 934
},
{
"epoch": 2.6338028169014085,
"grad_norm": 0.33291951317085183,
"learning_rate": 4.4751540081265645e-07,
"loss": 0.4713,
"step": 935
},
{
"epoch": 2.636619718309859,
"grad_norm": 0.3364588123495725,
"learning_rate": 4.407596287431387e-07,
"loss": 0.4701,
"step": 936
},
{
"epoch": 2.63943661971831,
"grad_norm": 0.3569801867527304,
"learning_rate": 4.340528865533161e-07,
"loss": 0.4602,
"step": 937
},
{
"epoch": 2.6422535211267606,
"grad_norm": 0.3457740049414934,
"learning_rate": 4.2739524636721207e-07,
"loss": 0.4523,
"step": 938
},
{
"epoch": 2.6450704225352113,
"grad_norm": 0.3616089928125598,
"learning_rate": 4.207867797808102e-07,
"loss": 0.4178,
"step": 939
},
{
"epoch": 2.647887323943662,
"grad_norm": 0.3153735514351465,
"learning_rate": 4.1422755786128364e-07,
"loss": 0.4484,
"step": 940
},
{
"epoch": 2.6507042253521127,
"grad_norm": 0.32313727348187915,
"learning_rate": 4.0771765114622886e-07,
"loss": 0.4544,
"step": 941
},
{
"epoch": 2.6535211267605634,
"grad_norm": 0.32805844915240456,
"learning_rate": 4.012571296429102e-07,
"loss": 0.4383,
"step": 942
},
{
"epoch": 2.656338028169014,
"grad_norm": 0.31287096142273935,
"learning_rate": 3.948460628275047e-07,
"loss": 0.4502,
"step": 943
},
{
"epoch": 2.659154929577465,
"grad_norm": 0.3371841187040242,
"learning_rate": 3.8848451964435594e-07,
"loss": 0.4557,
"step": 944
},
{
"epoch": 2.6619718309859155,
"grad_norm": 0.32003737481875727,
"learning_rate": 3.8217256850523243e-07,
"loss": 0.4348,
"step": 945
},
{
"epoch": 2.664788732394366,
"grad_norm": 0.33975544248849016,
"learning_rate": 3.759102772885925e-07,
"loss": 0.4421,
"step": 946
},
{
"epoch": 2.667605633802817,
"grad_norm": 0.3266362705011225,
"learning_rate": 3.696977133388524e-07,
"loss": 0.4501,
"step": 947
},
{
"epoch": 2.6704225352112676,
"grad_norm": 0.32481434731258335,
"learning_rate": 3.635349434656638e-07,
"loss": 0.459,
"step": 948
},
{
"epoch": 2.6732394366197183,
"grad_norm": 0.34436172325722814,
"learning_rate": 3.5742203394319606e-07,
"loss": 0.4543,
"step": 949
},
{
"epoch": 2.676056338028169,
"grad_norm": 0.35110488824156655,
"learning_rate": 3.513590505094222e-07,
"loss": 0.4497,
"step": 950
},
{
"epoch": 2.6788732394366197,
"grad_norm": 0.3315333572014592,
"learning_rate": 3.453460583654106e-07,
"loss": 0.4306,
"step": 951
},
{
"epoch": 2.6816901408450704,
"grad_norm": 0.34897677030004204,
"learning_rate": 3.3938312217462686e-07,
"loss": 0.4403,
"step": 952
},
{
"epoch": 2.684507042253521,
"grad_norm": 0.3438403164590787,
"learning_rate": 3.334703060622374e-07,
"loss": 0.4564,
"step": 953
},
{
"epoch": 2.687323943661972,
"grad_norm": 0.3207018561351413,
"learning_rate": 3.2760767361441847e-07,
"loss": 0.475,
"step": 954
},
{
"epoch": 2.6901408450704225,
"grad_norm": 0.3383111012033015,
"learning_rate": 3.2179528787767313e-07,
"loss": 0.452,
"step": 955
},
{
"epoch": 2.692957746478873,
"grad_norm": 0.3341858993813047,
"learning_rate": 3.1603321135815434e-07,
"loss": 0.4584,
"step": 956
},
{
"epoch": 2.695774647887324,
"grad_norm": 0.3257817149894772,
"learning_rate": 3.103215060209902e-07,
"loss": 0.4378,
"step": 957
},
{
"epoch": 2.6985915492957746,
"grad_norm": 0.32341143095549446,
"learning_rate": 3.046602332896209e-07,
"loss": 0.4533,
"step": 958
},
{
"epoch": 2.7014084507042253,
"grad_norm": 0.33437110577140444,
"learning_rate": 2.9904945404513606e-07,
"loss": 0.4423,
"step": 959
},
{
"epoch": 2.704225352112676,
"grad_norm": 0.3280497699271315,
"learning_rate": 2.9348922862562025e-07,
"loss": 0.4344,
"step": 960
},
{
"epoch": 2.7070422535211267,
"grad_norm": 0.3173486025834307,
"learning_rate": 2.879796168255039e-07,
"loss": 0.443,
"step": 961
},
{
"epoch": 2.7098591549295774,
"grad_norm": 0.3249778481256936,
"learning_rate": 2.825206778949219e-07,
"loss": 0.4638,
"step": 962
},
{
"epoch": 2.712676056338028,
"grad_norm": 0.3206975925043833,
"learning_rate": 2.771124705390743e-07,
"loss": 0.4371,
"step": 963
},
{
"epoch": 2.715492957746479,
"grad_norm": 0.3214786347868462,
"learning_rate": 2.717550529175955e-07,
"loss": 0.4563,
"step": 964
},
{
"epoch": 2.7183098591549295,
"grad_norm": 0.3436233501105619,
"learning_rate": 2.664484826439312e-07,
"loss": 0.4552,
"step": 965
},
{
"epoch": 2.72112676056338,
"grad_norm": 0.3474231981101561,
"learning_rate": 2.611928167847133e-07,
"loss": 0.46,
"step": 966
},
{
"epoch": 2.723943661971831,
"grad_norm": 0.34528907019271443,
"learning_rate": 2.5598811185915397e-07,
"loss": 0.4652,
"step": 967
},
{
"epoch": 2.7267605633802816,
"grad_norm": 0.31957962452572586,
"learning_rate": 2.5083442383843136e-07,
"loss": 0.4302,
"step": 968
},
{
"epoch": 2.7295774647887323,
"grad_norm": 0.33868297083611953,
"learning_rate": 2.457318081450899e-07,
"loss": 0.4635,
"step": 969
},
{
"epoch": 2.732394366197183,
"grad_norm": 0.3650237660288557,
"learning_rate": 2.4068031965244486e-07,
"loss": 0.4438,
"step": 970
},
{
"epoch": 2.7352112676056337,
"grad_norm": 0.3422038198247917,
"learning_rate": 2.3568001268399088e-07,
"loss": 0.4496,
"step": 971
},
{
"epoch": 2.7380281690140844,
"grad_norm": 0.3161758059021394,
"learning_rate": 2.3073094101282056e-07,
"loss": 0.4496,
"step": 972
},
{
"epoch": 2.740845070422535,
"grad_norm": 0.3252943511474426,
"learning_rate": 2.258331578610423e-07,
"loss": 0.4431,
"step": 973
},
{
"epoch": 2.743661971830986,
"grad_norm": 0.35599450055875387,
"learning_rate": 2.2098671589921017e-07,
"loss": 0.4461,
"step": 974
},
{
"epoch": 2.7464788732394365,
"grad_norm": 0.3441924744953093,
"learning_rate": 2.1619166724575924e-07,
"loss": 0.4509,
"step": 975
},
{
"epoch": 2.749295774647887,
"grad_norm": 0.3488688599968552,
"learning_rate": 2.114480634664401e-07,
"loss": 0.4507,
"step": 976
},
{
"epoch": 2.752112676056338,
"grad_norm": 0.355239783276596,
"learning_rate": 2.0675595557376916e-07,
"loss": 0.4718,
"step": 977
},
{
"epoch": 2.7549295774647886,
"grad_norm": 0.3437164703895729,
"learning_rate": 2.0211539402647807e-07,
"loss": 0.4654,
"step": 978
},
{
"epoch": 2.7577464788732393,
"grad_norm": 0.31614760184812657,
"learning_rate": 1.9752642872897078e-07,
"loss": 0.4676,
"step": 979
},
{
"epoch": 2.76056338028169,
"grad_norm": 0.321515520308783,
"learning_rate": 1.9298910903078838e-07,
"loss": 0.4305,
"step": 980
},
{
"epoch": 2.7633802816901407,
"grad_norm": 0.34992635800783795,
"learning_rate": 1.8850348372607575e-07,
"loss": 0.4538,
"step": 981
},
{
"epoch": 2.7661971830985914,
"grad_norm": 0.30276880004114215,
"learning_rate": 1.840696010530596e-07,
"loss": 0.4567,
"step": 982
},
{
"epoch": 2.769014084507042,
"grad_norm": 0.34441547278059703,
"learning_rate": 1.7968750869352792e-07,
"loss": 0.4453,
"step": 983
},
{
"epoch": 2.771830985915493,
"grad_norm": 0.30998669583959265,
"learning_rate": 1.753572537723186e-07,
"loss": 0.4425,
"step": 984
},
{
"epoch": 2.7746478873239435,
"grad_norm": 0.3586629209274004,
"learning_rate": 1.7107888285681106e-07,
"loss": 0.4564,
"step": 985
},
{
"epoch": 2.777464788732394,
"grad_norm": 0.32695128234004484,
"learning_rate": 1.6685244195642715e-07,
"loss": 0.4637,
"step": 986
},
{
"epoch": 2.780281690140845,
"grad_norm": 0.34115355713855117,
"learning_rate": 1.6267797652213368e-07,
"loss": 0.4477,
"step": 987
},
{
"epoch": 2.7830985915492956,
"grad_norm": 0.2930530495084058,
"learning_rate": 1.5855553144595858e-07,
"loss": 0.4566,
"step": 988
},
{
"epoch": 2.7859154929577463,
"grad_norm": 0.33022571130555106,
"learning_rate": 1.5448515106050165e-07,
"loss": 0.4308,
"step": 989
},
{
"epoch": 2.788732394366197,
"grad_norm": 0.352771290999933,
"learning_rate": 1.5046687913846392e-07,
"loss": 0.4705,
"step": 990
},
{
"epoch": 2.7915492957746477,
"grad_norm": 0.32707483621349326,
"learning_rate": 1.4650075889217297e-07,
"loss": 0.4602,
"step": 991
},
{
"epoch": 2.7943661971830984,
"grad_norm": 0.32654783950949157,
"learning_rate": 1.4258683297311891e-07,
"loss": 0.4479,
"step": 992
},
{
"epoch": 2.7971830985915496,
"grad_norm": 0.35198467602408556,
"learning_rate": 1.3872514347149756e-07,
"loss": 0.4448,
"step": 993
},
{
"epoch": 2.8,
"grad_norm": 0.342116234775216,
"learning_rate": 1.3491573191575513e-07,
"loss": 0.4582,
"step": 994
},
{
"epoch": 2.802816901408451,
"grad_norm": 0.3507902235732393,
"learning_rate": 1.3115863927214423e-07,
"loss": 0.4546,
"step": 995
},
{
"epoch": 2.8056338028169012,
"grad_norm": 0.3238046742423758,
"learning_rate": 1.2745390594428143e-07,
"loss": 0.4516,
"step": 996
},
{
"epoch": 2.8084507042253524,
"grad_norm": 0.32569158910871976,
"learning_rate": 1.2380157177271369e-07,
"loss": 0.4531,
"step": 997
},
{
"epoch": 2.8112676056338026,
"grad_norm": 0.3072177170540146,
"learning_rate": 1.202016760344893e-07,
"loss": 0.4297,
"step": 998
},
{
"epoch": 2.8140845070422538,
"grad_norm": 0.3158277495077995,
"learning_rate": 1.166542574427354e-07,
"loss": 0.4397,
"step": 999
},
{
"epoch": 2.816901408450704,
"grad_norm": 0.35987034251993116,
"learning_rate": 1.131593541462439e-07,
"loss": 0.4656,
"step": 1000
},
{
"epoch": 2.819718309859155,
"grad_norm": 0.33708080153812514,
"learning_rate": 1.0971700372905736e-07,
"loss": 0.4521,
"step": 1001
},
{
"epoch": 2.8225352112676054,
"grad_norm": 0.3389935828063123,
"learning_rate": 1.0632724321006816e-07,
"loss": 0.4576,
"step": 1002
},
{
"epoch": 2.8253521126760566,
"grad_norm": 0.32671457088330047,
"learning_rate": 1.0299010904261886e-07,
"loss": 0.4413,
"step": 1003
},
{
"epoch": 2.828169014084507,
"grad_norm": 0.3447563387897561,
"learning_rate": 9.970563711410974e-08,
"loss": 0.4534,
"step": 1004
},
{
"epoch": 2.830985915492958,
"grad_norm": 0.3701007432833937,
"learning_rate": 9.647386274561466e-08,
"loss": 0.465,
"step": 1005
},
{
"epoch": 2.8338028169014082,
"grad_norm": 0.31675165370176617,
"learning_rate": 9.329482069149853e-08,
"loss": 0.454,
"step": 1006
},
{
"epoch": 2.8366197183098594,
"grad_norm": 0.3449482405049473,
"learning_rate": 9.016854513904828e-08,
"loss": 0.4599,
"step": 1007
},
{
"epoch": 2.8394366197183096,
"grad_norm": 0.31943993823528516,
"learning_rate": 8.709506970809855e-08,
"loss": 0.4683,
"step": 1008
},
{
"epoch": 2.8422535211267608,
"grad_norm": 0.30682247155872994,
"learning_rate": 8.407442745067552e-08,
"loss": 0.4634,
"step": 1009
},
{
"epoch": 2.845070422535211,
"grad_norm": 0.3068025583439518,
"learning_rate": 8.110665085063918e-08,
"loss": 0.4362,
"step": 1010
},
{
"epoch": 2.847887323943662,
"grad_norm": 0.3326409740995297,
"learning_rate": 7.819177182333493e-08,
"loss": 0.4637,
"step": 1011
},
{
"epoch": 2.8507042253521124,
"grad_norm": 0.37249201472232357,
"learning_rate": 7.532982171524927e-08,
"loss": 0.4492,
"step": 1012
},
{
"epoch": 2.8535211267605636,
"grad_norm": 0.33303906691187746,
"learning_rate": 7.252083130367349e-08,
"loss": 0.4478,
"step": 1013
},
{
"epoch": 2.856338028169014,
"grad_norm": 0.32196502935195925,
"learning_rate": 6.97648307963733e-08,
"loss": 0.4601,
"step": 1014
},
{
"epoch": 2.859154929577465,
"grad_norm": 0.34642386779898743,
"learning_rate": 6.706184983126196e-08,
"loss": 0.43,
"step": 1015
},
{
"epoch": 2.8619718309859152,
"grad_norm": 0.34765468257149207,
"learning_rate": 6.441191747608322e-08,
"loss": 0.4512,
"step": 1016
},
{
"epoch": 2.8647887323943664,
"grad_norm": 0.34037074374729687,
"learning_rate": 6.181506222809885e-08,
"loss": 0.4439,
"step": 1017
},
{
"epoch": 2.8676056338028166,
"grad_norm": 0.3169812664050294,
"learning_rate": 5.9271312013781665e-08,
"loss": 0.443,
"step": 1018
},
{
"epoch": 2.870422535211268,
"grad_norm": 0.32211817240340407,
"learning_rate": 5.678069418851351e-08,
"loss": 0.4676,
"step": 1019
},
{
"epoch": 2.873239436619718,
"grad_norm": 0.3359515873734084,
"learning_rate": 5.4343235536294944e-08,
"loss": 0.4516,
"step": 1020
},
{
"epoch": 2.876056338028169,
"grad_norm": 0.34578515983344105,
"learning_rate": 5.195896226945385e-08,
"loss": 0.4407,
"step": 1021
},
{
"epoch": 2.87887323943662,
"grad_norm": 0.31439962540381255,
"learning_rate": 4.9627900028365036e-08,
"loss": 0.4435,
"step": 1022
},
{
"epoch": 2.8816901408450706,
"grad_norm": 0.34336441391655215,
"learning_rate": 4.735007388117441e-08,
"loss": 0.4386,
"step": 1023
},
{
"epoch": 2.8845070422535213,
"grad_norm": 0.335578371195018,
"learning_rate": 4.5125508323528025e-08,
"loss": 0.4289,
"step": 1024
},
{
"epoch": 2.887323943661972,
"grad_norm": 0.350643103957519,
"learning_rate": 4.2954227278310666e-08,
"loss": 0.4464,
"step": 1025
},
{
"epoch": 2.8901408450704227,
"grad_norm": 0.35068154909138555,
"learning_rate": 4.083625409538772e-08,
"loss": 0.4326,
"step": 1026
},
{
"epoch": 2.8929577464788734,
"grad_norm": 0.34185081917190546,
"learning_rate": 3.877161155135423e-08,
"loss": 0.4593,
"step": 1027
},
{
"epoch": 2.895774647887324,
"grad_norm": 0.3300724585491567,
"learning_rate": 3.676032184928957e-08,
"loss": 0.457,
"step": 1028
},
{
"epoch": 2.898591549295775,
"grad_norm": 0.3411195609156822,
"learning_rate": 3.4802406618518195e-08,
"loss": 0.456,
"step": 1029
},
{
"epoch": 2.9014084507042255,
"grad_norm": 0.3637968816563219,
"learning_rate": 3.289788691437923e-08,
"loss": 0.4394,
"step": 1030
},
{
"epoch": 2.904225352112676,
"grad_norm": 0.3395290765317001,
"learning_rate": 3.10467832179967e-08,
"loss": 0.4506,
"step": 1031
},
{
"epoch": 2.907042253521127,
"grad_norm": 0.3260136250051306,
"learning_rate": 2.9249115436063013e-08,
"loss": 0.4472,
"step": 1032
},
{
"epoch": 2.9098591549295776,
"grad_norm": 0.33498085364365865,
"learning_rate": 2.7504902900621356e-08,
"loss": 0.4434,
"step": 1033
},
{
"epoch": 2.9126760563380283,
"grad_norm": 0.34312265751451554,
"learning_rate": 2.581416436886086e-08,
"loss": 0.4578,
"step": 1034
},
{
"epoch": 2.915492957746479,
"grad_norm": 0.3210591639802571,
"learning_rate": 2.4176918022912333e-08,
"loss": 0.4608,
"step": 1035
},
{
"epoch": 2.9183098591549297,
"grad_norm": 0.3553095286591404,
"learning_rate": 2.2593181469653945e-08,
"loss": 0.4673,
"step": 1036
},
{
"epoch": 2.9211267605633804,
"grad_norm": 0.321850104936369,
"learning_rate": 2.1062971740523076e-08,
"loss": 0.4534,
"step": 1037
},
{
"epoch": 2.923943661971831,
"grad_norm": 0.3229367069170625,
"learning_rate": 1.958630529133032e-08,
"loss": 0.4503,
"step": 1038
},
{
"epoch": 2.926760563380282,
"grad_norm": 0.32793169190757315,
"learning_rate": 1.816319800208466e-08,
"loss": 0.4519,
"step": 1039
},
{
"epoch": 2.9295774647887325,
"grad_norm": 0.3306690320189479,
"learning_rate": 1.679366517682246e-08,
"loss": 0.4484,
"step": 1040
},
{
"epoch": 2.932394366197183,
"grad_norm": 0.35486018214343157,
"learning_rate": 1.547772154344207e-08,
"loss": 0.4564,
"step": 1041
},
{
"epoch": 2.935211267605634,
"grad_norm": 0.3487414018054857,
"learning_rate": 1.4215381253546712e-08,
"loss": 0.4411,
"step": 1042
},
{
"epoch": 2.9380281690140846,
"grad_norm": 0.33851454546424364,
"learning_rate": 1.3006657882290163e-08,
"loss": 0.456,
"step": 1043
},
{
"epoch": 2.9408450704225353,
"grad_norm": 0.33398998229974186,
"learning_rate": 1.1851564428232987e-08,
"loss": 0.4563,
"step": 1044
},
{
"epoch": 2.943661971830986,
"grad_norm": 0.3405604730717625,
"learning_rate": 1.0750113313202082e-08,
"loss": 0.4203,
"step": 1045
},
{
"epoch": 2.9464788732394367,
"grad_norm": 0.3200139617949195,
"learning_rate": 9.70231638215635e-09,
"loss": 0.4472,
"step": 1046
},
{
"epoch": 2.9492957746478874,
"grad_norm": 0.3410890631754851,
"learning_rate": 8.708184903060135e-09,
"loss": 0.4403,
"step": 1047
},
{
"epoch": 2.952112676056338,
"grad_norm": 0.3419082068730841,
"learning_rate": 7.76772956676164e-09,
"loss": 0.4367,
"step": 1048
},
{
"epoch": 2.954929577464789,
"grad_norm": 0.3378106075319102,
"learning_rate": 6.880960486877475e-09,
"loss": 0.4531,
"step": 1049
},
{
"epoch": 2.9577464788732395,
"grad_norm": 0.3278897439419402,
"learning_rate": 6.047887199686075e-09,
"loss": 0.459,
"step": 1050
},
{
"epoch": 2.96056338028169,
"grad_norm": 0.3343112858188528,
"learning_rate": 5.268518664022226e-09,
"loss": 0.459,
"step": 1051
},
{
"epoch": 2.963380281690141,
"grad_norm": 0.32255215872689313,
"learning_rate": 4.54286326118214e-09,
"loss": 0.4336,
"step": 1052
},
{
"epoch": 2.9661971830985916,
"grad_norm": 0.3370218781417737,
"learning_rate": 3.870928794834083e-09,
"loss": 0.4367,
"step": 1053
},
{
"epoch": 2.9690140845070423,
"grad_norm": 0.3323793895278448,
"learning_rate": 3.2527224909328914e-09,
"loss": 0.4625,
"step": 1054
},
{
"epoch": 2.971830985915493,
"grad_norm": 0.33449875147589203,
"learning_rate": 2.6882509976433604e-09,
"loss": 0.4539,
"step": 1055
},
{
"epoch": 2.9746478873239437,
"grad_norm": 0.37530967430639717,
"learning_rate": 2.1775203852680836e-09,
"loss": 0.447,
"step": 1056
},
{
"epoch": 2.9774647887323944,
"grad_norm": 0.32946903451444376,
"learning_rate": 1.7205361461825054e-09,
"loss": 0.443,
"step": 1057
},
{
"epoch": 2.980281690140845,
"grad_norm": 0.3181193426288814,
"learning_rate": 1.317303194776076e-09,
"loss": 0.4281,
"step": 1058
},
{
"epoch": 2.983098591549296,
"grad_norm": 0.34593232841366583,
"learning_rate": 9.678258673995189e-10,
"loss": 0.456,
"step": 1059
},
{
"epoch": 2.9859154929577465,
"grad_norm": 0.3417744937936135,
"learning_rate": 6.721079223165339e-10,
"loss": 0.4374,
"step": 1060
},
{
"epoch": 2.988732394366197,
"grad_norm": 0.3153613761930667,
"learning_rate": 4.301525396666062e-10,
"loss": 0.4499,
"step": 1061
},
{
"epoch": 2.991549295774648,
"grad_norm": 0.31244029690230596,
"learning_rate": 2.4196232142614794e-10,
"loss": 0.4421,
"step": 1062
},
{
"epoch": 2.9943661971830986,
"grad_norm": 0.3084976115932713,
"learning_rate": 1.0753929138629382e-10,
"loss": 0.4275,
"step": 1063
},
{
"epoch": 2.9971830985915493,
"grad_norm": 0.3735924098512123,
"learning_rate": 2.6884895126255695e-11,
"loss": 0.4404,
"step": 1064
},
{
"epoch": 3.0,
"grad_norm": 0.33696360890684507,
"learning_rate": 0.0,
"loss": 0.4535,
"step": 1065
},
{
"epoch": 3.0,
"step": 1065,
"total_flos": 1313980844015616.0,
"train_loss": 0.5105482672861484,
"train_runtime": 17519.9119,
"train_samples_per_second": 5.836,
"train_steps_per_second": 0.061
}
],
"logging_steps": 1,
"max_steps": 1065,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1313980844015616.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}