namansmishaps's picture
Upload folder using huggingface_hub
c9d330c verified
Raw
History Blame Contribute Delete
126 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9907502569373072,
"eval_steps": 500,
"global_step": 729,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0041109969167523125,
"grad_norm": 10.707864137961929,
"learning_rate": 9.09090909090909e-09,
"loss": 1.653,
"step": 1
},
{
"epoch": 0.008221993833504625,
"grad_norm": 10.079889894789929,
"learning_rate": 1.818181818181818e-08,
"loss": 1.611,
"step": 2
},
{
"epoch": 0.012332990750256937,
"grad_norm": 9.419370762311278,
"learning_rate": 2.727272727272727e-08,
"loss": 1.5125,
"step": 3
},
{
"epoch": 0.01644398766700925,
"grad_norm": 11.08145357513862,
"learning_rate": 3.636363636363636e-08,
"loss": 1.7065,
"step": 4
},
{
"epoch": 0.020554984583761562,
"grad_norm": 13.693470206701171,
"learning_rate": 4.545454545454545e-08,
"loss": 1.6996,
"step": 5
},
{
"epoch": 0.024665981500513873,
"grad_norm": 10.792826728200351,
"learning_rate": 5.454545454545454e-08,
"loss": 1.6366,
"step": 6
},
{
"epoch": 0.02877697841726619,
"grad_norm": 9.259967603342691,
"learning_rate": 6.363636363636363e-08,
"loss": 1.4384,
"step": 7
},
{
"epoch": 0.0328879753340185,
"grad_norm": 11.634024710191847,
"learning_rate": 7.272727272727273e-08,
"loss": 1.8226,
"step": 8
},
{
"epoch": 0.03699897225077081,
"grad_norm": 12.43829169869022,
"learning_rate": 8.181818181818182e-08,
"loss": 1.8798,
"step": 9
},
{
"epoch": 0.041109969167523124,
"grad_norm": 14.122605378135763,
"learning_rate": 9.09090909090909e-08,
"loss": 1.9207,
"step": 10
},
{
"epoch": 0.045220966084275435,
"grad_norm": 10.436429425580595,
"learning_rate": 1e-07,
"loss": 1.4087,
"step": 11
},
{
"epoch": 0.04933196300102775,
"grad_norm": 13.354320080172663,
"learning_rate": 1.0909090909090908e-07,
"loss": 1.8416,
"step": 12
},
{
"epoch": 0.05344295991778006,
"grad_norm": 11.046025387388442,
"learning_rate": 1.1818181818181818e-07,
"loss": 1.6022,
"step": 13
},
{
"epoch": 0.05755395683453238,
"grad_norm": 12.871188554801648,
"learning_rate": 1.2727272727272726e-07,
"loss": 1.932,
"step": 14
},
{
"epoch": 0.06166495375128469,
"grad_norm": 11.103321688706771,
"learning_rate": 1.3636363636363635e-07,
"loss": 1.6527,
"step": 15
},
{
"epoch": 0.065775950668037,
"grad_norm": 11.262727784803356,
"learning_rate": 1.4545454545454545e-07,
"loss": 1.6742,
"step": 16
},
{
"epoch": 0.0698869475847893,
"grad_norm": 11.782474338729536,
"learning_rate": 1.5454545454545453e-07,
"loss": 1.7641,
"step": 17
},
{
"epoch": 0.07399794450154162,
"grad_norm": 11.272950463022264,
"learning_rate": 1.6363636363636364e-07,
"loss": 1.7117,
"step": 18
},
{
"epoch": 0.07810894141829394,
"grad_norm": 11.59120293181086,
"learning_rate": 1.7272727272727272e-07,
"loss": 1.7428,
"step": 19
},
{
"epoch": 0.08221993833504625,
"grad_norm": 10.57519693614169,
"learning_rate": 1.818181818181818e-07,
"loss": 1.6703,
"step": 20
},
{
"epoch": 0.08633093525179857,
"grad_norm": 11.453228551846658,
"learning_rate": 1.909090909090909e-07,
"loss": 1.7573,
"step": 21
},
{
"epoch": 0.09044193216855087,
"grad_norm": 11.29402839117027,
"learning_rate": 2e-07,
"loss": 1.6777,
"step": 22
},
{
"epoch": 0.09455292908530319,
"grad_norm": 11.891258741251171,
"learning_rate": 1.9999901274303226e-07,
"loss": 1.7828,
"step": 23
},
{
"epoch": 0.0986639260020555,
"grad_norm": 9.417998938008715,
"learning_rate": 1.9999605099162257e-07,
"loss": 1.5301,
"step": 24
},
{
"epoch": 0.10277492291880781,
"grad_norm": 12.434612682087748,
"learning_rate": 1.999911148042511e-07,
"loss": 1.9377,
"step": 25
},
{
"epoch": 0.10688591983556012,
"grad_norm": 9.43561201481426,
"learning_rate": 1.999842042783836e-07,
"loss": 1.5347,
"step": 26
},
{
"epoch": 0.11099691675231244,
"grad_norm": 10.413570520205068,
"learning_rate": 1.9997531955046938e-07,
"loss": 1.6379,
"step": 27
},
{
"epoch": 0.11510791366906475,
"grad_norm": 9.618497593862777,
"learning_rate": 1.9996446079593855e-07,
"loss": 1.583,
"step": 28
},
{
"epoch": 0.11921891058581706,
"grad_norm": 12.173893065489027,
"learning_rate": 1.999516282291988e-07,
"loss": 1.9173,
"step": 29
},
{
"epoch": 0.12332990750256938,
"grad_norm": 11.178809895169248,
"learning_rate": 1.9993682210363092e-07,
"loss": 1.7341,
"step": 30
},
{
"epoch": 0.12744090441932168,
"grad_norm": 10.364692752106809,
"learning_rate": 1.9992004271158393e-07,
"loss": 1.6695,
"step": 31
},
{
"epoch": 0.131551901336074,
"grad_norm": 25.766594072812342,
"learning_rate": 1.9990129038436926e-07,
"loss": 1.9172,
"step": 32
},
{
"epoch": 0.13566289825282632,
"grad_norm": 10.248873523838787,
"learning_rate": 1.9988056549225425e-07,
"loss": 1.6077,
"step": 33
},
{
"epoch": 0.1397738951695786,
"grad_norm": 9.657467999208903,
"learning_rate": 1.9985786844445474e-07,
"loss": 1.6677,
"step": 34
},
{
"epoch": 0.14388489208633093,
"grad_norm": 10.179618115404484,
"learning_rate": 1.9983319968912713e-07,
"loss": 1.7537,
"step": 35
},
{
"epoch": 0.14799588900308325,
"grad_norm": 9.216582739378328,
"learning_rate": 1.9980655971335942e-07,
"loss": 1.6093,
"step": 36
},
{
"epoch": 0.15210688591983557,
"grad_norm": 10.03947910756885,
"learning_rate": 1.9977794904316165e-07,
"loss": 1.3945,
"step": 37
},
{
"epoch": 0.15621788283658788,
"grad_norm": 9.153514820925254,
"learning_rate": 1.9974736824345551e-07,
"loss": 1.5444,
"step": 38
},
{
"epoch": 0.16032887975334018,
"grad_norm": 10.588030905814275,
"learning_rate": 1.997148179180631e-07,
"loss": 1.7547,
"step": 39
},
{
"epoch": 0.1644398766700925,
"grad_norm": 8.635529189480858,
"learning_rate": 1.9968029870969518e-07,
"loss": 1.5311,
"step": 40
},
{
"epoch": 0.1685508735868448,
"grad_norm": 8.796854932608584,
"learning_rate": 1.996438112999383e-07,
"loss": 1.5621,
"step": 41
},
{
"epoch": 0.17266187050359713,
"grad_norm": 9.844561069662058,
"learning_rate": 1.9960535640924147e-07,
"loss": 1.7654,
"step": 42
},
{
"epoch": 0.17677286742034942,
"grad_norm": 8.169528609598125,
"learning_rate": 1.9956493479690186e-07,
"loss": 1.565,
"step": 43
},
{
"epoch": 0.18088386433710174,
"grad_norm": 9.177103070862636,
"learning_rate": 1.9952254726104981e-07,
"loss": 1.6504,
"step": 44
},
{
"epoch": 0.18499486125385406,
"grad_norm": 8.135361612565449,
"learning_rate": 1.9947819463863317e-07,
"loss": 1.5706,
"step": 45
},
{
"epoch": 0.18910585817060638,
"grad_norm": 11.213544735693581,
"learning_rate": 1.9943187780540063e-07,
"loss": 1.6303,
"step": 46
},
{
"epoch": 0.1932168550873587,
"grad_norm": 9.732781400753217,
"learning_rate": 1.9938359767588452e-07,
"loss": 1.6898,
"step": 47
},
{
"epoch": 0.197327852004111,
"grad_norm": 9.269604213022049,
"learning_rate": 1.9933335520338272e-07,
"loss": 1.6342,
"step": 48
},
{
"epoch": 0.2014388489208633,
"grad_norm": 9.640024814932346,
"learning_rate": 1.9928115137993982e-07,
"loss": 1.645,
"step": 49
},
{
"epoch": 0.20554984583761562,
"grad_norm": 10.042562364639352,
"learning_rate": 1.9922698723632764e-07,
"loss": 1.7879,
"step": 50
},
{
"epoch": 0.20966084275436794,
"grad_norm": 10.677611513314368,
"learning_rate": 1.9917086384202472e-07,
"loss": 1.7366,
"step": 51
},
{
"epoch": 0.21377183967112023,
"grad_norm": 11.12670398027465,
"learning_rate": 1.9911278230519534e-07,
"loss": 1.7761,
"step": 52
},
{
"epoch": 0.21788283658787255,
"grad_norm": 9.17809047840844,
"learning_rate": 1.9905274377266745e-07,
"loss": 1.719,
"step": 53
},
{
"epoch": 0.22199383350462487,
"grad_norm": 9.66213410187294,
"learning_rate": 1.9899074942991033e-07,
"loss": 1.6941,
"step": 54
},
{
"epoch": 0.2261048304213772,
"grad_norm": 9.316901012218453,
"learning_rate": 1.9892680050101086e-07,
"loss": 1.8349,
"step": 55
},
{
"epoch": 0.2302158273381295,
"grad_norm": 8.386332787168893,
"learning_rate": 1.988608982486496e-07,
"loss": 1.6031,
"step": 56
},
{
"epoch": 0.2343268242548818,
"grad_norm": 9.201728730915688,
"learning_rate": 1.9879304397407566e-07,
"loss": 1.6979,
"step": 57
},
{
"epoch": 0.23843782117163412,
"grad_norm": 9.19524420119839,
"learning_rate": 1.9872323901708118e-07,
"loss": 1.8266,
"step": 58
},
{
"epoch": 0.24254881808838644,
"grad_norm": 8.30925669939803,
"learning_rate": 1.9865148475597474e-07,
"loss": 1.4002,
"step": 59
},
{
"epoch": 0.24665981500513876,
"grad_norm": 8.708995475073078,
"learning_rate": 1.9857778260755426e-07,
"loss": 1.7404,
"step": 60
},
{
"epoch": 0.25077081192189105,
"grad_norm": 7.027030274599196,
"learning_rate": 1.985021340270789e-07,
"loss": 1.5258,
"step": 61
},
{
"epoch": 0.25488180883864336,
"grad_norm": 8.012080200049331,
"learning_rate": 1.9842454050824042e-07,
"loss": 1.5637,
"step": 62
},
{
"epoch": 0.2589928057553957,
"grad_norm": 6.3981910130624176,
"learning_rate": 1.983450035831337e-07,
"loss": 1.3657,
"step": 63
},
{
"epoch": 0.263103802672148,
"grad_norm": 7.478235385879245,
"learning_rate": 1.9826352482222637e-07,
"loss": 1.5445,
"step": 64
},
{
"epoch": 0.2672147995889003,
"grad_norm": 8.440469191638593,
"learning_rate": 1.9818010583432793e-07,
"loss": 1.376,
"step": 65
},
{
"epoch": 0.27132579650565264,
"grad_norm": 7.748465926383861,
"learning_rate": 1.9809474826655792e-07,
"loss": 1.5774,
"step": 66
},
{
"epoch": 0.27543679342240496,
"grad_norm": 8.57381708963539,
"learning_rate": 1.980074538043134e-07,
"loss": 1.5867,
"step": 67
},
{
"epoch": 0.2795477903391572,
"grad_norm": 7.603992819406835,
"learning_rate": 1.9791822417123574e-07,
"loss": 1.5607,
"step": 68
},
{
"epoch": 0.28365878725590954,
"grad_norm": 7.372496720744107,
"learning_rate": 1.9782706112917644e-07,
"loss": 1.5389,
"step": 69
},
{
"epoch": 0.28776978417266186,
"grad_norm": 7.379234155109177,
"learning_rate": 1.9773396647816248e-07,
"loss": 1.498,
"step": 70
},
{
"epoch": 0.2918807810894142,
"grad_norm": 11.113450259449483,
"learning_rate": 1.976389420563607e-07,
"loss": 1.6178,
"step": 71
},
{
"epoch": 0.2959917780061665,
"grad_norm": 8.367445274605876,
"learning_rate": 1.9754198974004156e-07,
"loss": 1.3662,
"step": 72
},
{
"epoch": 0.3001027749229188,
"grad_norm": 7.174507123313505,
"learning_rate": 1.9744311144354206e-07,
"loss": 1.5961,
"step": 73
},
{
"epoch": 0.30421377183967113,
"grad_norm": 7.835771561867673,
"learning_rate": 1.9734230911922796e-07,
"loss": 1.569,
"step": 74
},
{
"epoch": 0.30832476875642345,
"grad_norm": 12.836201576697274,
"learning_rate": 1.972395847574552e-07,
"loss": 1.6943,
"step": 75
},
{
"epoch": 0.31243576567317577,
"grad_norm": 7.396564528406367,
"learning_rate": 1.9713494038653056e-07,
"loss": 1.6064,
"step": 76
},
{
"epoch": 0.31654676258992803,
"grad_norm": 8.513071826753036,
"learning_rate": 1.9702837807267178e-07,
"loss": 1.4589,
"step": 77
},
{
"epoch": 0.32065775950668035,
"grad_norm": 6.82558705878682,
"learning_rate": 1.969198999199666e-07,
"loss": 1.4491,
"step": 78
},
{
"epoch": 0.32476875642343267,
"grad_norm": 5.930099566107373,
"learning_rate": 1.9680950807033125e-07,
"loss": 1.3579,
"step": 79
},
{
"epoch": 0.328879753340185,
"grad_norm": 6.956499321420972,
"learning_rate": 1.9669720470346817e-07,
"loss": 1.5192,
"step": 80
},
{
"epoch": 0.3329907502569373,
"grad_norm": 9.205199818219354,
"learning_rate": 1.9658299203682302e-07,
"loss": 1.7676,
"step": 81
},
{
"epoch": 0.3371017471736896,
"grad_norm": 7.974414130409151,
"learning_rate": 1.964668723255408e-07,
"loss": 1.5004,
"step": 82
},
{
"epoch": 0.34121274409044194,
"grad_norm": 10.46228879637872,
"learning_rate": 1.9634884786242137e-07,
"loss": 1.4274,
"step": 83
},
{
"epoch": 0.34532374100719426,
"grad_norm": 7.006447147726963,
"learning_rate": 1.9622892097787424e-07,
"loss": 1.4474,
"step": 84
},
{
"epoch": 0.3494347379239466,
"grad_norm": 7.304736439924326,
"learning_rate": 1.9610709403987246e-07,
"loss": 1.4933,
"step": 85
},
{
"epoch": 0.35354573484069884,
"grad_norm": 8.572376141029476,
"learning_rate": 1.9598336945390582e-07,
"loss": 1.4764,
"step": 86
},
{
"epoch": 0.35765673175745116,
"grad_norm": 6.62015523991803,
"learning_rate": 1.9585774966293361e-07,
"loss": 1.3693,
"step": 87
},
{
"epoch": 0.3617677286742035,
"grad_norm": 11.713543438005354,
"learning_rate": 1.957302371473361e-07,
"loss": 1.692,
"step": 88
},
{
"epoch": 0.3658787255909558,
"grad_norm": 6.261294458504216,
"learning_rate": 1.9560083442486562e-07,
"loss": 1.4308,
"step": 89
},
{
"epoch": 0.3699897225077081,
"grad_norm": 6.400229031166525,
"learning_rate": 1.9546954405059699e-07,
"loss": 1.3351,
"step": 90
},
{
"epoch": 0.37410071942446044,
"grad_norm": 6.347971290847363,
"learning_rate": 1.9533636861687695e-07,
"loss": 1.3089,
"step": 91
},
{
"epoch": 0.37821171634121276,
"grad_norm": 8.1022669923147,
"learning_rate": 1.9520131075327298e-07,
"loss": 1.6091,
"step": 92
},
{
"epoch": 0.3823227132579651,
"grad_norm": 8.724654001862117,
"learning_rate": 1.9506437312652145e-07,
"loss": 1.5754,
"step": 93
},
{
"epoch": 0.3864337101747174,
"grad_norm": 6.286621286224387,
"learning_rate": 1.9492555844047483e-07,
"loss": 1.3631,
"step": 94
},
{
"epoch": 0.39054470709146966,
"grad_norm": 7.327425938420709,
"learning_rate": 1.947848694360485e-07,
"loss": 1.4446,
"step": 95
},
{
"epoch": 0.394655704008222,
"grad_norm": 7.261174570313381,
"learning_rate": 1.946423088911664e-07,
"loss": 1.4527,
"step": 96
},
{
"epoch": 0.3987667009249743,
"grad_norm": 8.334586000579163,
"learning_rate": 1.9449787962070642e-07,
"loss": 1.5765,
"step": 97
},
{
"epoch": 0.4028776978417266,
"grad_norm": 6.019711509620174,
"learning_rate": 1.9435158447644456e-07,
"loss": 1.2973,
"step": 98
},
{
"epoch": 0.40698869475847893,
"grad_norm": 6.268742524449732,
"learning_rate": 1.942034263469989e-07,
"loss": 1.3864,
"step": 99
},
{
"epoch": 0.41109969167523125,
"grad_norm": 6.459311232973291,
"learning_rate": 1.9405340815777232e-07,
"loss": 1.328,
"step": 100
},
{
"epoch": 0.41521068859198357,
"grad_norm": 6.685489852709574,
"learning_rate": 1.9390153287089485e-07,
"loss": 1.3571,
"step": 101
},
{
"epoch": 0.4193216855087359,
"grad_norm": 5.799450095444668,
"learning_rate": 1.9374780348516525e-07,
"loss": 1.2997,
"step": 102
},
{
"epoch": 0.4234326824254882,
"grad_norm": 6.966273782296385,
"learning_rate": 1.9359222303599161e-07,
"loss": 1.5034,
"step": 103
},
{
"epoch": 0.42754367934224047,
"grad_norm": 6.667050936120606,
"learning_rate": 1.9343479459533158e-07,
"loss": 1.532,
"step": 104
},
{
"epoch": 0.4316546762589928,
"grad_norm": 6.5093558351926175,
"learning_rate": 1.932755212716317e-07,
"loss": 1.3541,
"step": 105
},
{
"epoch": 0.4357656731757451,
"grad_norm": 5.967905266361118,
"learning_rate": 1.9311440620976594e-07,
"loss": 1.2638,
"step": 106
},
{
"epoch": 0.4398766700924974,
"grad_norm": 11.025029669401425,
"learning_rate": 1.9295145259097363e-07,
"loss": 1.6295,
"step": 107
},
{
"epoch": 0.44398766700924974,
"grad_norm": 6.804163379870639,
"learning_rate": 1.9278666363279665e-07,
"loss": 1.3899,
"step": 108
},
{
"epoch": 0.44809866392600206,
"grad_norm": 5.570635795736665,
"learning_rate": 1.92620042589016e-07,
"loss": 1.343,
"step": 109
},
{
"epoch": 0.4522096608427544,
"grad_norm": 7.9712461371351075,
"learning_rate": 1.9245159274958738e-07,
"loss": 1.1742,
"step": 110
},
{
"epoch": 0.4563206577595067,
"grad_norm": 5.847438622442037,
"learning_rate": 1.9228131744057634e-07,
"loss": 1.2814,
"step": 111
},
{
"epoch": 0.460431654676259,
"grad_norm": 5.415902230738819,
"learning_rate": 1.921092200240926e-07,
"loss": 1.249,
"step": 112
},
{
"epoch": 0.4645426515930113,
"grad_norm": 5.559511138928407,
"learning_rate": 1.9193530389822361e-07,
"loss": 1.234,
"step": 113
},
{
"epoch": 0.4686536485097636,
"grad_norm": 5.6472024695669605,
"learning_rate": 1.9175957249696757e-07,
"loss": 1.3378,
"step": 114
},
{
"epoch": 0.4727646454265159,
"grad_norm": 9.835736386430504,
"learning_rate": 1.915820292901654e-07,
"loss": 1.223,
"step": 115
},
{
"epoch": 0.47687564234326824,
"grad_norm": 5.001772612228533,
"learning_rate": 1.914026777834325e-07,
"loss": 1.2777,
"step": 116
},
{
"epoch": 0.48098663926002055,
"grad_norm": 5.747620038850084,
"learning_rate": 1.912215215180894e-07,
"loss": 1.3384,
"step": 117
},
{
"epoch": 0.4850976361767729,
"grad_norm": 5.434192975971742,
"learning_rate": 1.9103856407109174e-07,
"loss": 1.331,
"step": 118
},
{
"epoch": 0.4892086330935252,
"grad_norm": 8.822807056425992,
"learning_rate": 1.9085380905495986e-07,
"loss": 1.3666,
"step": 119
},
{
"epoch": 0.4933196300102775,
"grad_norm": 5.195072980892559,
"learning_rate": 1.9066726011770723e-07,
"loss": 1.3101,
"step": 120
},
{
"epoch": 0.49743062692702983,
"grad_norm": 6.061820219659554,
"learning_rate": 1.904789209427687e-07,
"loss": 1.3194,
"step": 121
},
{
"epoch": 0.5015416238437821,
"grad_norm": 5.683541772683969,
"learning_rate": 1.902887952489275e-07,
"loss": 1.3017,
"step": 122
},
{
"epoch": 0.5056526207605344,
"grad_norm": 7.003289691664117,
"learning_rate": 1.900968867902419e-07,
"loss": 1.5633,
"step": 123
},
{
"epoch": 0.5097636176772867,
"grad_norm": 5.3307566117226,
"learning_rate": 1.8990319935597122e-07,
"loss": 1.2872,
"step": 124
},
{
"epoch": 0.513874614594039,
"grad_norm": 4.637510901432101,
"learning_rate": 1.897077367705008e-07,
"loss": 1.2278,
"step": 125
},
{
"epoch": 0.5179856115107914,
"grad_norm": 7.725170952109564,
"learning_rate": 1.8951050289326665e-07,
"loss": 1.3531,
"step": 126
},
{
"epoch": 0.5220966084275437,
"grad_norm": 6.5432794886992935,
"learning_rate": 1.8931150161867915e-07,
"loss": 1.4712,
"step": 127
},
{
"epoch": 0.526207605344296,
"grad_norm": 5.099034062070253,
"learning_rate": 1.891107368760462e-07,
"loss": 1.3124,
"step": 128
},
{
"epoch": 0.5303186022610483,
"grad_norm": 7.34830932448487,
"learning_rate": 1.8890821262949563e-07,
"loss": 1.3615,
"step": 129
},
{
"epoch": 0.5344295991778006,
"grad_norm": 4.974499367955476,
"learning_rate": 1.8870393287789693e-07,
"loss": 1.275,
"step": 130
},
{
"epoch": 0.538540596094553,
"grad_norm": 6.277837399258011,
"learning_rate": 1.884979016547822e-07,
"loss": 1.4911,
"step": 131
},
{
"epoch": 0.5426515930113053,
"grad_norm": 6.300414640449073,
"learning_rate": 1.8829012302826676e-07,
"loss": 1.3662,
"step": 132
},
{
"epoch": 0.5467625899280576,
"grad_norm": 6.897111795676912,
"learning_rate": 1.880806011009684e-07,
"loss": 1.3777,
"step": 133
},
{
"epoch": 0.5508735868448099,
"grad_norm": 6.365236910899277,
"learning_rate": 1.8786934000992687e-07,
"loss": 1.2849,
"step": 134
},
{
"epoch": 0.5549845837615622,
"grad_norm": 4.95321554450534,
"learning_rate": 1.8765634392652182e-07,
"loss": 1.2829,
"step": 135
},
{
"epoch": 0.5590955806783144,
"grad_norm": 5.4461660071709606,
"learning_rate": 1.8744161705639067e-07,
"loss": 1.2983,
"step": 136
},
{
"epoch": 0.5632065775950668,
"grad_norm": 5.540349675940396,
"learning_rate": 1.8722516363934528e-07,
"loss": 1.2619,
"step": 137
},
{
"epoch": 0.5673175745118191,
"grad_norm": 6.964049263220438,
"learning_rate": 1.870069879492886e-07,
"loss": 1.4972,
"step": 138
},
{
"epoch": 0.5714285714285714,
"grad_norm": 5.716437085523343,
"learning_rate": 1.8678709429413e-07,
"loss": 1.3306,
"step": 139
},
{
"epoch": 0.5755395683453237,
"grad_norm": 9.277392356133683,
"learning_rate": 1.865654870157004e-07,
"loss": 1.3488,
"step": 140
},
{
"epoch": 0.579650565262076,
"grad_norm": 6.1551196516884445,
"learning_rate": 1.8634217048966633e-07,
"loss": 1.38,
"step": 141
},
{
"epoch": 0.5837615621788284,
"grad_norm": 4.976550280547202,
"learning_rate": 1.8611714912544376e-07,
"loss": 1.2261,
"step": 142
},
{
"epoch": 0.5878725590955807,
"grad_norm": 5.234504589611595,
"learning_rate": 1.8589042736611093e-07,
"loss": 1.3224,
"step": 143
},
{
"epoch": 0.591983556012333,
"grad_norm": 18.72759020571204,
"learning_rate": 1.8566200968832043e-07,
"loss": 1.1957,
"step": 144
},
{
"epoch": 0.5960945529290853,
"grad_norm": 4.360003556078857,
"learning_rate": 1.8543190060221125e-07,
"loss": 1.264,
"step": 145
},
{
"epoch": 0.6002055498458376,
"grad_norm": 5.886825199475184,
"learning_rate": 1.8520010465131934e-07,
"loss": 1.3873,
"step": 146
},
{
"epoch": 0.60431654676259,
"grad_norm": 4.792369934205899,
"learning_rate": 1.8496662641248806e-07,
"loss": 1.2899,
"step": 147
},
{
"epoch": 0.6084275436793423,
"grad_norm": 5.908144217721895,
"learning_rate": 1.8473147049577773e-07,
"loss": 1.3477,
"step": 148
},
{
"epoch": 0.6125385405960946,
"grad_norm": 5.624489648110953,
"learning_rate": 1.8449464154437474e-07,
"loss": 1.344,
"step": 149
},
{
"epoch": 0.6166495375128469,
"grad_norm": 5.718772099976671,
"learning_rate": 1.8425614423449972e-07,
"loss": 1.2532,
"step": 150
},
{
"epoch": 0.6207605344295992,
"grad_norm": 6.393465254386759,
"learning_rate": 1.8401598327531534e-07,
"loss": 1.2969,
"step": 151
},
{
"epoch": 0.6248715313463515,
"grad_norm": 5.647205221965829,
"learning_rate": 1.8377416340883313e-07,
"loss": 1.3353,
"step": 152
},
{
"epoch": 0.6289825282631039,
"grad_norm": 5.936967173792457,
"learning_rate": 1.8353068940982007e-07,
"loss": 1.2068,
"step": 153
},
{
"epoch": 0.6330935251798561,
"grad_norm": 6.864964609522403,
"learning_rate": 1.832855660857042e-07,
"loss": 1.2872,
"step": 154
},
{
"epoch": 0.6372045220966084,
"grad_norm": 5.135047720122144,
"learning_rate": 1.8303879827647974e-07,
"loss": 1.2048,
"step": 155
},
{
"epoch": 0.6413155190133607,
"grad_norm": 5.346765155903461,
"learning_rate": 1.8279039085461148e-07,
"loss": 1.3305,
"step": 156
},
{
"epoch": 0.645426515930113,
"grad_norm": 5.040099277567391,
"learning_rate": 1.8254034872493853e-07,
"loss": 1.2649,
"step": 157
},
{
"epoch": 0.6495375128468653,
"grad_norm": 5.329511339157392,
"learning_rate": 1.8228867682457762e-07,
"loss": 1.2711,
"step": 158
},
{
"epoch": 0.6536485097636177,
"grad_norm": 5.6417382542593275,
"learning_rate": 1.820353801228255e-07,
"loss": 1.3227,
"step": 159
},
{
"epoch": 0.65775950668037,
"grad_norm": 5.0402605192268615,
"learning_rate": 1.8178046362106083e-07,
"loss": 1.1433,
"step": 160
},
{
"epoch": 0.6618705035971223,
"grad_norm": 5.290791064280284,
"learning_rate": 1.8152393235264543e-07,
"loss": 1.1927,
"step": 161
},
{
"epoch": 0.6659815005138746,
"grad_norm": 5.723238591665759,
"learning_rate": 1.81265791382825e-07,
"loss": 1.3449,
"step": 162
},
{
"epoch": 0.6700924974306269,
"grad_norm": 6.894894192664154,
"learning_rate": 1.8100604580862898e-07,
"loss": 1.3515,
"step": 163
},
{
"epoch": 0.6742034943473793,
"grad_norm": 5.292876203921864,
"learning_rate": 1.8074470075876983e-07,
"loss": 1.2181,
"step": 164
},
{
"epoch": 0.6783144912641316,
"grad_norm": 6.016200507600216,
"learning_rate": 1.8048176139354206e-07,
"loss": 1.339,
"step": 165
},
{
"epoch": 0.6824254881808839,
"grad_norm": 10.65140323955686,
"learning_rate": 1.8021723290472007e-07,
"loss": 1.3989,
"step": 166
},
{
"epoch": 0.6865364850976362,
"grad_norm": 6.835299309886434,
"learning_rate": 1.799511205154557e-07,
"loss": 1.1971,
"step": 167
},
{
"epoch": 0.6906474820143885,
"grad_norm": 4.879232611898712,
"learning_rate": 1.796834294801752e-07,
"loss": 1.275,
"step": 168
},
{
"epoch": 0.6947584789311408,
"grad_norm": 9.236322964731619,
"learning_rate": 1.7941416508447532e-07,
"loss": 1.1331,
"step": 169
},
{
"epoch": 0.6988694758478932,
"grad_norm": 6.628893482135026,
"learning_rate": 1.7914333264501913e-07,
"loss": 1.2046,
"step": 170
},
{
"epoch": 0.7029804727646455,
"grad_norm": 5.173325379014087,
"learning_rate": 1.7887093750943086e-07,
"loss": 1.2001,
"step": 171
},
{
"epoch": 0.7070914696813977,
"grad_norm": 4.640674342071268,
"learning_rate": 1.7859698505619044e-07,
"loss": 1.1967,
"step": 172
},
{
"epoch": 0.71120246659815,
"grad_norm": 5.480644072332889,
"learning_rate": 1.783214806945272e-07,
"loss": 1.3128,
"step": 173
},
{
"epoch": 0.7153134635149023,
"grad_norm": 5.07951059573175,
"learning_rate": 1.7804442986431316e-07,
"loss": 1.2685,
"step": 174
},
{
"epoch": 0.7194244604316546,
"grad_norm": 5.447927413738588,
"learning_rate": 1.777658380359556e-07,
"loss": 1.1988,
"step": 175
},
{
"epoch": 0.723535457348407,
"grad_norm": 4.079539267934105,
"learning_rate": 1.7748571071028897e-07,
"loss": 1.1171,
"step": 176
},
{
"epoch": 0.7276464542651593,
"grad_norm": 10.428836094857438,
"learning_rate": 1.7720405341846637e-07,
"loss": 1.445,
"step": 177
},
{
"epoch": 0.7317574511819116,
"grad_norm": 4.994200087027091,
"learning_rate": 1.7692087172185027e-07,
"loss": 1.1787,
"step": 178
},
{
"epoch": 0.7358684480986639,
"grad_norm": 8.329601013646606,
"learning_rate": 1.7663617121190271e-07,
"loss": 1.2562,
"step": 179
},
{
"epoch": 0.7399794450154162,
"grad_norm": 5.611087131141067,
"learning_rate": 1.76349957510075e-07,
"loss": 1.2815,
"step": 180
},
{
"epoch": 0.7440904419321686,
"grad_norm": 4.097510869957281,
"learning_rate": 1.760622362676965e-07,
"loss": 1.1629,
"step": 181
},
{
"epoch": 0.7482014388489209,
"grad_norm": 5.552004707795244,
"learning_rate": 1.7577301316586323e-07,
"loss": 1.2387,
"step": 182
},
{
"epoch": 0.7523124357656732,
"grad_norm": 4.8462890072104425,
"learning_rate": 1.754822939153257e-07,
"loss": 1.1054,
"step": 183
},
{
"epoch": 0.7564234326824255,
"grad_norm": 6.536235722142915,
"learning_rate": 1.7519008425637597e-07,
"loss": 1.3187,
"step": 184
},
{
"epoch": 0.7605344295991778,
"grad_norm": 4.951878169170016,
"learning_rate": 1.7489638995873453e-07,
"loss": 1.2254,
"step": 185
},
{
"epoch": 0.7646454265159301,
"grad_norm": 3.958765970797126,
"learning_rate": 1.7460121682143615e-07,
"loss": 1.072,
"step": 186
},
{
"epoch": 0.7687564234326825,
"grad_norm": 6.022137357961096,
"learning_rate": 1.7430457067271563e-07,
"loss": 1.2429,
"step": 187
},
{
"epoch": 0.7728674203494348,
"grad_norm": 8.466655888789814,
"learning_rate": 1.7400645736989245e-07,
"loss": 1.2505,
"step": 188
},
{
"epoch": 0.7769784172661871,
"grad_norm": 5.222771499693532,
"learning_rate": 1.737068827992554e-07,
"loss": 1.2429,
"step": 189
},
{
"epoch": 0.7810894141829393,
"grad_norm": 6.355910286345292,
"learning_rate": 1.7340585287594602e-07,
"loss": 1.3068,
"step": 190
},
{
"epoch": 0.7852004110996916,
"grad_norm": 5.126600401356926,
"learning_rate": 1.7310337354384215e-07,
"loss": 1.2144,
"step": 191
},
{
"epoch": 0.789311408016444,
"grad_norm": 8.352368899667336,
"learning_rate": 1.7279945077544035e-07,
"loss": 1.1347,
"step": 192
},
{
"epoch": 0.7934224049331963,
"grad_norm": 5.268839329383105,
"learning_rate": 1.7249409057173807e-07,
"loss": 1.1184,
"step": 193
},
{
"epoch": 0.7975334018499486,
"grad_norm": 3.8916065501136843,
"learning_rate": 1.7218729896211503e-07,
"loss": 1.0728,
"step": 194
},
{
"epoch": 0.8016443987667009,
"grad_norm": 4.80958946575301,
"learning_rate": 1.7187908200421434e-07,
"loss": 1.2634,
"step": 195
},
{
"epoch": 0.8057553956834532,
"grad_norm": 4.584370732817813,
"learning_rate": 1.7156944578382276e-07,
"loss": 1.1399,
"step": 196
},
{
"epoch": 0.8098663926002055,
"grad_norm": 6.130616255679746,
"learning_rate": 1.712583964147507e-07,
"loss": 1.1173,
"step": 197
},
{
"epoch": 0.8139773895169579,
"grad_norm": 13.319035759981675,
"learning_rate": 1.7094594003871117e-07,
"loss": 1.1742,
"step": 198
},
{
"epoch": 0.8180883864337102,
"grad_norm": 3.779192478211289,
"learning_rate": 1.7063208282519895e-07,
"loss": 1.0236,
"step": 199
},
{
"epoch": 0.8221993833504625,
"grad_norm": 4.109209036990689,
"learning_rate": 1.7031683097136843e-07,
"loss": 1.1069,
"step": 200
},
{
"epoch": 0.8263103802672148,
"grad_norm": 4.0762226712183605,
"learning_rate": 1.700001907019114e-07,
"loss": 1.0031,
"step": 201
},
{
"epoch": 0.8304213771839671,
"grad_norm": 3.8654259991990916,
"learning_rate": 1.6968216826893404e-07,
"loss": 0.9732,
"step": 202
},
{
"epoch": 0.8345323741007195,
"grad_norm": 3.869952065623661,
"learning_rate": 1.6936276995183373e-07,
"loss": 1.1063,
"step": 203
},
{
"epoch": 0.8386433710174718,
"grad_norm": 4.704453370534266,
"learning_rate": 1.6904200205717466e-07,
"loss": 1.0475,
"step": 204
},
{
"epoch": 0.8427543679342241,
"grad_norm": 5.026304033396767,
"learning_rate": 1.6871987091856366e-07,
"loss": 1.1108,
"step": 205
},
{
"epoch": 0.8468653648509764,
"grad_norm": 5.035225900625049,
"learning_rate": 1.683963828965249e-07,
"loss": 1.043,
"step": 206
},
{
"epoch": 0.8509763617677287,
"grad_norm": 3.662612352820592,
"learning_rate": 1.6807154437837453e-07,
"loss": 1.0623,
"step": 207
},
{
"epoch": 0.8550873586844809,
"grad_norm": 3.8410544152768655,
"learning_rate": 1.6774536177809427e-07,
"loss": 0.9906,
"step": 208
},
{
"epoch": 0.8591983556012333,
"grad_norm": 4.129293605901065,
"learning_rate": 1.6741784153620508e-07,
"loss": 1.1204,
"step": 209
},
{
"epoch": 0.8633093525179856,
"grad_norm": 4.935467560854357,
"learning_rate": 1.6708899011963976e-07,
"loss": 1.1119,
"step": 210
},
{
"epoch": 0.8674203494347379,
"grad_norm": 4.015734926775564,
"learning_rate": 1.6675881402161536e-07,
"loss": 1.1626,
"step": 211
},
{
"epoch": 0.8715313463514902,
"grad_norm": 5.446078766699763,
"learning_rate": 1.6642731976150494e-07,
"loss": 1.131,
"step": 212
},
{
"epoch": 0.8756423432682425,
"grad_norm": 4.566173880082412,
"learning_rate": 1.6609451388470885e-07,
"loss": 1.0112,
"step": 213
},
{
"epoch": 0.8797533401849948,
"grad_norm": 3.276523865697787,
"learning_rate": 1.6576040296252554e-07,
"loss": 0.9645,
"step": 214
},
{
"epoch": 0.8838643371017472,
"grad_norm": 3.6911629992951047,
"learning_rate": 1.6542499359202168e-07,
"loss": 1.0944,
"step": 215
},
{
"epoch": 0.8879753340184995,
"grad_norm": 4.183238776507183,
"learning_rate": 1.650882923959021e-07,
"loss": 1.1043,
"step": 216
},
{
"epoch": 0.8920863309352518,
"grad_norm": 3.192052677057548,
"learning_rate": 1.6475030602237876e-07,
"loss": 0.9394,
"step": 217
},
{
"epoch": 0.8961973278520041,
"grad_norm": 3.2005085455189444,
"learning_rate": 1.6441104114503977e-07,
"loss": 0.9451,
"step": 218
},
{
"epoch": 0.9003083247687564,
"grad_norm": 5.048492570878951,
"learning_rate": 1.6407050446271738e-07,
"loss": 1.1157,
"step": 219
},
{
"epoch": 0.9044193216855088,
"grad_norm": 3.341145430999988,
"learning_rate": 1.6372870269935582e-07,
"loss": 0.9953,
"step": 220
},
{
"epoch": 0.9085303186022611,
"grad_norm": 4.14587154305856,
"learning_rate": 1.6338564260387862e-07,
"loss": 0.9303,
"step": 221
},
{
"epoch": 0.9126413155190134,
"grad_norm": 4.260201583701495,
"learning_rate": 1.6304133095005507e-07,
"loss": 1.109,
"step": 222
},
{
"epoch": 0.9167523124357657,
"grad_norm": 4.092619824286429,
"learning_rate": 1.626957745363668e-07,
"loss": 1.1044,
"step": 223
},
{
"epoch": 0.920863309352518,
"grad_norm": 4.0502484125320475,
"learning_rate": 1.6234898018587336e-07,
"loss": 1.0063,
"step": 224
},
{
"epoch": 0.9249743062692704,
"grad_norm": 3.5506086131246124,
"learning_rate": 1.6200095474607751e-07,
"loss": 0.9798,
"step": 225
},
{
"epoch": 0.9290853031860226,
"grad_norm": 12.459826992818279,
"learning_rate": 1.6165170508879008e-07,
"loss": 1.0563,
"step": 226
},
{
"epoch": 0.9331963001027749,
"grad_norm": 4.234594438916428,
"learning_rate": 1.613012381099942e-07,
"loss": 1.0542,
"step": 227
},
{
"epoch": 0.9373072970195272,
"grad_norm": 5.581281099863202,
"learning_rate": 1.6094956072970924e-07,
"loss": 0.9864,
"step": 228
},
{
"epoch": 0.9414182939362795,
"grad_norm": 5.482341245037625,
"learning_rate": 1.6059667989185406e-07,
"loss": 1.0739,
"step": 229
},
{
"epoch": 0.9455292908530318,
"grad_norm": 4.8712204495422124,
"learning_rate": 1.6024260256410994e-07,
"loss": 1.0664,
"step": 230
},
{
"epoch": 0.9496402877697842,
"grad_norm": 3.910681290271946,
"learning_rate": 1.5988733573778315e-07,
"loss": 1.0877,
"step": 231
},
{
"epoch": 0.9537512846865365,
"grad_norm": 3.6293772691939954,
"learning_rate": 1.595308864276666e-07,
"loss": 1.0067,
"step": 232
},
{
"epoch": 0.9578622816032888,
"grad_norm": 3.3442637604223013,
"learning_rate": 1.5917326167190165e-07,
"loss": 1.0216,
"step": 233
},
{
"epoch": 0.9619732785200411,
"grad_norm": 4.399760739803475,
"learning_rate": 1.588144685318389e-07,
"loss": 1.1903,
"step": 234
},
{
"epoch": 0.9660842754367934,
"grad_norm": 5.12258557268555,
"learning_rate": 1.5845451409189888e-07,
"loss": 0.9688,
"step": 235
},
{
"epoch": 0.9701952723535457,
"grad_norm": 3.674152497012296,
"learning_rate": 1.580934054594322e-07,
"loss": 1.0486,
"step": 236
},
{
"epoch": 0.9743062692702981,
"grad_norm": 4.238866161199344,
"learning_rate": 1.5773114976457914e-07,
"loss": 1.0302,
"step": 237
},
{
"epoch": 0.9784172661870504,
"grad_norm": 4.367082292611695,
"learning_rate": 1.573677541601289e-07,
"loss": 1.0433,
"step": 238
},
{
"epoch": 0.9825282631038027,
"grad_norm": 4.273812234099169,
"learning_rate": 1.5700322582137827e-07,
"loss": 1.0055,
"step": 239
},
{
"epoch": 0.986639260020555,
"grad_norm": 4.491229260617013,
"learning_rate": 1.5663757194599012e-07,
"loss": 1.0768,
"step": 240
},
{
"epoch": 0.9907502569373073,
"grad_norm": 3.8347744908641483,
"learning_rate": 1.562707997538512e-07,
"loss": 0.9993,
"step": 241
},
{
"epoch": 0.9948612538540597,
"grad_norm": 3.27938600158038,
"learning_rate": 1.5590291648692952e-07,
"loss": 0.9517,
"step": 242
},
{
"epoch": 0.998972250770812,
"grad_norm": 7.3523047468326075,
"learning_rate": 1.5553392940913147e-07,
"loss": 1.0323,
"step": 243
},
{
"epoch": 1.0,
"grad_norm": 7.3523047468326075,
"learning_rate": 1.551638458061583e-07,
"loss": 0.2618,
"step": 244
},
{
"epoch": 1.0041109969167523,
"grad_norm": 4.587527924546633,
"learning_rate": 1.5479267298536238e-07,
"loss": 1.0414,
"step": 245
},
{
"epoch": 1.0082219938335046,
"grad_norm": 3.2655329122499066,
"learning_rate": 1.5442041827560272e-07,
"loss": 0.922,
"step": 246
},
{
"epoch": 1.012332990750257,
"grad_norm": 3.6435788817559556,
"learning_rate": 1.5404708902710048e-07,
"loss": 1.0512,
"step": 247
},
{
"epoch": 1.0164439876670093,
"grad_norm": 3.532035670122312,
"learning_rate": 1.5367269261129367e-07,
"loss": 1.1142,
"step": 248
},
{
"epoch": 1.0205549845837616,
"grad_norm": 3.1281688701442976,
"learning_rate": 1.532972364206917e-07,
"loss": 0.8561,
"step": 249
},
{
"epoch": 1.024665981500514,
"grad_norm": 7.538457318745296,
"learning_rate": 1.529207278687294e-07,
"loss": 0.9928,
"step": 250
},
{
"epoch": 1.0287769784172662,
"grad_norm": 4.9764365177078425,
"learning_rate": 1.5254317438962053e-07,
"loss": 1.0124,
"step": 251
},
{
"epoch": 1.0328879753340185,
"grad_norm": 4.3471137210647335,
"learning_rate": 1.521645834382112e-07,
"loss": 0.9735,
"step": 252
},
{
"epoch": 1.0369989722507709,
"grad_norm": 3.767687792184999,
"learning_rate": 1.5178496248983254e-07,
"loss": 0.9494,
"step": 253
},
{
"epoch": 1.0411099691675232,
"grad_norm": 3.922983144917104,
"learning_rate": 1.51404319040153e-07,
"loss": 1.0714,
"step": 254
},
{
"epoch": 1.0452209660842755,
"grad_norm": 3.9564590894983325,
"learning_rate": 1.510226606050306e-07,
"loss": 1.0723,
"step": 255
},
{
"epoch": 1.0493319630010278,
"grad_norm": 5.090704470034626,
"learning_rate": 1.5063999472036431e-07,
"loss": 0.8843,
"step": 256
},
{
"epoch": 1.0534429599177801,
"grad_norm": 3.791932714047322,
"learning_rate": 1.502563289419453e-07,
"loss": 1.0045,
"step": 257
},
{
"epoch": 1.0575539568345325,
"grad_norm": 4.205770790261918,
"learning_rate": 1.4987167084530789e-07,
"loss": 0.9499,
"step": 258
},
{
"epoch": 1.0616649537512848,
"grad_norm": 3.61064474688858,
"learning_rate": 1.494860280255798e-07,
"loss": 1.0903,
"step": 259
},
{
"epoch": 1.065775950668037,
"grad_norm": 4.220812615971972,
"learning_rate": 1.4909940809733221e-07,
"loss": 1.0432,
"step": 260
},
{
"epoch": 1.0698869475847892,
"grad_norm": 4.1564691945837176,
"learning_rate": 1.4871181869442952e-07,
"loss": 1.0334,
"step": 261
},
{
"epoch": 1.0739979445015417,
"grad_norm": 3.4456733669634705,
"learning_rate": 1.4832326746987846e-07,
"loss": 0.8977,
"step": 262
},
{
"epoch": 1.0781089414182938,
"grad_norm": 8.867540475958243,
"learning_rate": 1.4793376209567715e-07,
"loss": 0.9876,
"step": 263
},
{
"epoch": 1.0822199383350461,
"grad_norm": 3.758481717795243,
"learning_rate": 1.4754331026266343e-07,
"loss": 1.139,
"step": 264
},
{
"epoch": 1.0863309352517985,
"grad_norm": 3.423033101790078,
"learning_rate": 1.4715191968036324e-07,
"loss": 0.9716,
"step": 265
},
{
"epoch": 1.0904419321685508,
"grad_norm": 5.334887865735263,
"learning_rate": 1.467595980768381e-07,
"loss": 1.0051,
"step": 266
},
{
"epoch": 1.094552929085303,
"grad_norm": 6.59849216352483,
"learning_rate": 1.4636635319853273e-07,
"loss": 1.0045,
"step": 267
},
{
"epoch": 1.0986639260020554,
"grad_norm": 4.986823873514777,
"learning_rate": 1.4597219281012207e-07,
"loss": 0.9797,
"step": 268
},
{
"epoch": 1.1027749229188077,
"grad_norm": 3.312438263885284,
"learning_rate": 1.4557712469435797e-07,
"loss": 0.9593,
"step": 269
},
{
"epoch": 1.10688591983556,
"grad_norm": 3.6481380380427852,
"learning_rate": 1.451811566519154e-07,
"loss": 0.9023,
"step": 270
},
{
"epoch": 1.1109969167523124,
"grad_norm": 3.5037083932955766,
"learning_rate": 1.4478429650123852e-07,
"loss": 1.0164,
"step": 271
},
{
"epoch": 1.1151079136690647,
"grad_norm": 5.373660229379728,
"learning_rate": 1.4438655207838628e-07,
"loss": 0.9792,
"step": 272
},
{
"epoch": 1.119218910585817,
"grad_norm": 3.256466523727759,
"learning_rate": 1.4398793123687779e-07,
"loss": 0.8795,
"step": 273
},
{
"epoch": 1.1233299075025693,
"grad_norm": 2.9632076523747375,
"learning_rate": 1.435884418475371e-07,
"loss": 1.0495,
"step": 274
},
{
"epoch": 1.1274409044193217,
"grad_norm": 3.3022340284078835,
"learning_rate": 1.431880917983379e-07,
"loss": 0.8945,
"step": 275
},
{
"epoch": 1.131551901336074,
"grad_norm": 3.5932638648288857,
"learning_rate": 1.4278688899424765e-07,
"loss": 0.9901,
"step": 276
},
{
"epoch": 1.1356628982528263,
"grad_norm": 3.919024819518497,
"learning_rate": 1.4238484135707163e-07,
"loss": 0.9543,
"step": 277
},
{
"epoch": 1.1397738951695786,
"grad_norm": 3.7589281660297296,
"learning_rate": 1.419819568252965e-07,
"loss": 0.8906,
"step": 278
},
{
"epoch": 1.143884892086331,
"grad_norm": 3.5161227693050727,
"learning_rate": 1.415782433539335e-07,
"loss": 0.909,
"step": 279
},
{
"epoch": 1.1479958890030832,
"grad_norm": 4.358301844170438,
"learning_rate": 1.4117370891436134e-07,
"loss": 1.0172,
"step": 280
},
{
"epoch": 1.1521068859198356,
"grad_norm": 3.3000172459440296,
"learning_rate": 1.4076836149416886e-07,
"loss": 0.9679,
"step": 281
},
{
"epoch": 1.1562178828365879,
"grad_norm": 3.8632482737448584,
"learning_rate": 1.4036220909699746e-07,
"loss": 0.9702,
"step": 282
},
{
"epoch": 1.1603288797533402,
"grad_norm": 3.909839901638121,
"learning_rate": 1.399552597423828e-07,
"loss": 1.0682,
"step": 283
},
{
"epoch": 1.1644398766700925,
"grad_norm": 2.7714797573184256,
"learning_rate": 1.3954752146559652e-07,
"loss": 0.8751,
"step": 284
},
{
"epoch": 1.1685508735868448,
"grad_norm": 3.2678409235203723,
"learning_rate": 1.3913900231748777e-07,
"loss": 0.8514,
"step": 285
},
{
"epoch": 1.1726618705035972,
"grad_norm": 4.628365909485524,
"learning_rate": 1.3872971036432407e-07,
"loss": 0.8754,
"step": 286
},
{
"epoch": 1.1767728674203495,
"grad_norm": 3.1810861375665778,
"learning_rate": 1.3831965368763204e-07,
"loss": 0.9274,
"step": 287
},
{
"epoch": 1.1808838643371018,
"grad_norm": 2.676778659249158,
"learning_rate": 1.3790884038403793e-07,
"loss": 0.9122,
"step": 288
},
{
"epoch": 1.1849948612538541,
"grad_norm": 4.50938245514188,
"learning_rate": 1.3749727856510767e-07,
"loss": 0.966,
"step": 289
},
{
"epoch": 1.1891058581706064,
"grad_norm": 4.5085156293745685,
"learning_rate": 1.3708497635718673e-07,
"loss": 0.9738,
"step": 290
},
{
"epoch": 1.1932168550873588,
"grad_norm": 3.840888130011014,
"learning_rate": 1.3667194190123958e-07,
"loss": 0.9885,
"step": 291
},
{
"epoch": 1.197327852004111,
"grad_norm": 5.03067762124161,
"learning_rate": 1.3625818335268925e-07,
"loss": 0.9908,
"step": 292
},
{
"epoch": 1.2014388489208634,
"grad_norm": 3.012942460130905,
"learning_rate": 1.358437088812558e-07,
"loss": 0.9007,
"step": 293
},
{
"epoch": 1.2055498458376157,
"grad_norm": 3.9729600234803795,
"learning_rate": 1.3542852667079556e-07,
"loss": 0.9564,
"step": 294
},
{
"epoch": 1.209660842754368,
"grad_norm": 3.4923568698520047,
"learning_rate": 1.3501264491913906e-07,
"loss": 0.9348,
"step": 295
},
{
"epoch": 1.2137718396711201,
"grad_norm": 13.54234756183888,
"learning_rate": 1.3459607183792945e-07,
"loss": 1.0596,
"step": 296
},
{
"epoch": 1.2178828365878727,
"grad_norm": 3.3853216831853126,
"learning_rate": 1.3417881565246028e-07,
"loss": 1.0132,
"step": 297
},
{
"epoch": 1.2219938335046248,
"grad_norm": 3.076392845182414,
"learning_rate": 1.3376088460151304e-07,
"loss": 0.9107,
"step": 298
},
{
"epoch": 1.2261048304213773,
"grad_norm": 2.8152450155019855,
"learning_rate": 1.3334228693719464e-07,
"loss": 0.865,
"step": 299
},
{
"epoch": 1.2302158273381294,
"grad_norm": 3.598739419126404,
"learning_rate": 1.3292303092477425e-07,
"loss": 0.9782,
"step": 300
},
{
"epoch": 1.2343268242548817,
"grad_norm": 3.511571183489361,
"learning_rate": 1.3250312484252023e-07,
"loss": 0.9511,
"step": 301
},
{
"epoch": 1.238437821171634,
"grad_norm": 3.074787788547139,
"learning_rate": 1.3208257698153675e-07,
"loss": 0.8803,
"step": 302
},
{
"epoch": 1.2425488180883864,
"grad_norm": 2.746368624615067,
"learning_rate": 1.3166139564559992e-07,
"loss": 0.8559,
"step": 303
},
{
"epoch": 1.2466598150051387,
"grad_norm": 3.4881838173477284,
"learning_rate": 1.3123958915099391e-07,
"loss": 0.9375,
"step": 304
},
{
"epoch": 1.250770811921891,
"grad_norm": 4.412925782674072,
"learning_rate": 1.3081716582634672e-07,
"loss": 1.0466,
"step": 305
},
{
"epoch": 1.2548818088386433,
"grad_norm": 5.046254104640164,
"learning_rate": 1.3039413401246576e-07,
"loss": 1.0148,
"step": 306
},
{
"epoch": 1.2589928057553956,
"grad_norm": 3.8480322391567277,
"learning_rate": 1.2997050206217316e-07,
"loss": 0.9868,
"step": 307
},
{
"epoch": 1.263103802672148,
"grad_norm": 3.514561158345834,
"learning_rate": 1.2954627834014081e-07,
"loss": 1.0085,
"step": 308
},
{
"epoch": 1.2672147995889003,
"grad_norm": 4.351642190234845,
"learning_rate": 1.2912147122272522e-07,
"loss": 0.9864,
"step": 309
},
{
"epoch": 1.2713257965056526,
"grad_norm": 3.5989585736661835,
"learning_rate": 1.286960890978021e-07,
"loss": 0.9998,
"step": 310
},
{
"epoch": 1.275436793422405,
"grad_norm": 3.4737199129099645,
"learning_rate": 1.2827014036460083e-07,
"loss": 0.8694,
"step": 311
},
{
"epoch": 1.2795477903391572,
"grad_norm": 2.539715373657141,
"learning_rate": 1.2784363343353847e-07,
"loss": 0.9282,
"step": 312
},
{
"epoch": 1.2836587872559095,
"grad_norm": 3.2040501701385504,
"learning_rate": 1.2741657672605383e-07,
"loss": 0.8948,
"step": 313
},
{
"epoch": 1.2877697841726619,
"grad_norm": 4.476229955091483,
"learning_rate": 1.2698897867444112e-07,
"loss": 0.9465,
"step": 314
},
{
"epoch": 1.2918807810894142,
"grad_norm": 2.735316244383116,
"learning_rate": 1.2656084772168342e-07,
"loss": 0.8718,
"step": 315
},
{
"epoch": 1.2959917780061665,
"grad_norm": 15.132216877168823,
"learning_rate": 1.2613219232128607e-07,
"loss": 0.9205,
"step": 316
},
{
"epoch": 1.3001027749229188,
"grad_norm": 2.852225281113358,
"learning_rate": 1.257030209371097e-07,
"loss": 0.9716,
"step": 317
},
{
"epoch": 1.3042137718396711,
"grad_norm": 3.472991419069752,
"learning_rate": 1.2527334204320306e-07,
"loss": 0.9016,
"step": 318
},
{
"epoch": 1.3083247687564235,
"grad_norm": 6.268518446736897,
"learning_rate": 1.2484316412363585e-07,
"loss": 1.0182,
"step": 319
},
{
"epoch": 1.3124357656731758,
"grad_norm": 4.343662117964228,
"learning_rate": 1.2441249567233098e-07,
"loss": 0.8995,
"step": 320
},
{
"epoch": 1.316546762589928,
"grad_norm": 5.4253442976393815,
"learning_rate": 1.239813451928971e-07,
"loss": 1.0105,
"step": 321
},
{
"epoch": 1.3206577595066804,
"grad_norm": 3.1886335139175865,
"learning_rate": 1.2354972119846047e-07,
"loss": 0.894,
"step": 322
},
{
"epoch": 1.3247687564234327,
"grad_norm": 6.633026435651971,
"learning_rate": 1.2311763221149697e-07,
"loss": 0.9594,
"step": 323
},
{
"epoch": 1.328879753340185,
"grad_norm": 3.200014759412536,
"learning_rate": 1.2268508676366393e-07,
"loss": 0.7737,
"step": 324
},
{
"epoch": 1.3329907502569374,
"grad_norm": 2.9522876335978556,
"learning_rate": 1.2225209339563143e-07,
"loss": 0.8886,
"step": 325
},
{
"epoch": 1.3371017471736897,
"grad_norm": 2.9896642725764515,
"learning_rate": 1.218186606569139e-07,
"loss": 0.9793,
"step": 326
},
{
"epoch": 1.341212744090442,
"grad_norm": 6.08730869790015,
"learning_rate": 1.2138479710570122e-07,
"loss": 1.0776,
"step": 327
},
{
"epoch": 1.3453237410071943,
"grad_norm": 4.452907765904937,
"learning_rate": 1.2095051130868957e-07,
"loss": 0.9614,
"step": 328
},
{
"epoch": 1.3494347379239466,
"grad_norm": 3.108116007100476,
"learning_rate": 1.2051581184091263e-07,
"loss": 0.8032,
"step": 329
},
{
"epoch": 1.353545734840699,
"grad_norm": 2.905780323154698,
"learning_rate": 1.2008070728557184e-07,
"loss": 0.8724,
"step": 330
},
{
"epoch": 1.357656731757451,
"grad_norm": 3.7786375892411477,
"learning_rate": 1.196452062338674e-07,
"loss": 0.9391,
"step": 331
},
{
"epoch": 1.3617677286742036,
"grad_norm": 3.9772000342593525,
"learning_rate": 1.1920931728482823e-07,
"loss": 0.9264,
"step": 332
},
{
"epoch": 1.3658787255909557,
"grad_norm": 3.817880764626275,
"learning_rate": 1.1877304904514231e-07,
"loss": 0.9745,
"step": 333
},
{
"epoch": 1.3699897225077082,
"grad_norm": 3.7294607952564442,
"learning_rate": 1.1833641012898689e-07,
"loss": 0.8814,
"step": 334
},
{
"epoch": 1.3741007194244603,
"grad_norm": 3.894043685988304,
"learning_rate": 1.1789940915785823e-07,
"loss": 0.8693,
"step": 335
},
{
"epoch": 1.3782117163412129,
"grad_norm": 3.2573926116380725,
"learning_rate": 1.1746205476040136e-07,
"loss": 0.9465,
"step": 336
},
{
"epoch": 1.382322713257965,
"grad_norm": 4.12092698927328,
"learning_rate": 1.1702435557223986e-07,
"loss": 1.0532,
"step": 337
},
{
"epoch": 1.3864337101747175,
"grad_norm": 5.26471469117091,
"learning_rate": 1.1658632023580515e-07,
"loss": 0.9369,
"step": 338
},
{
"epoch": 1.3905447070914696,
"grad_norm": 3.223766697269215,
"learning_rate": 1.1614795740016599e-07,
"loss": 1.0186,
"step": 339
},
{
"epoch": 1.394655704008222,
"grad_norm": 3.4553250886935554,
"learning_rate": 1.1570927572085765e-07,
"loss": 0.8929,
"step": 340
},
{
"epoch": 1.3987667009249742,
"grad_norm": 4.722435348796641,
"learning_rate": 1.1527028385971107e-07,
"loss": 0.809,
"step": 341
},
{
"epoch": 1.4028776978417266,
"grad_norm": 3.9226576826541426,
"learning_rate": 1.1483099048468168e-07,
"loss": 1.0225,
"step": 342
},
{
"epoch": 1.4069886947584789,
"grad_norm": 7.093144884392558,
"learning_rate": 1.143914042696784e-07,
"loss": 0.9861,
"step": 343
},
{
"epoch": 1.4110996916752312,
"grad_norm": 3.1456777723164695,
"learning_rate": 1.1395153389439232e-07,
"loss": 0.8725,
"step": 344
},
{
"epoch": 1.4152106885919835,
"grad_norm": 2.9260651476179875,
"learning_rate": 1.1351138804412524e-07,
"loss": 0.8668,
"step": 345
},
{
"epoch": 1.4193216855087358,
"grad_norm": 2.4690552378582686,
"learning_rate": 1.1307097540961838e-07,
"loss": 0.8523,
"step": 346
},
{
"epoch": 1.4234326824254881,
"grad_norm": 4.573741288832168,
"learning_rate": 1.1263030468688058e-07,
"loss": 0.9331,
"step": 347
},
{
"epoch": 1.4275436793422405,
"grad_norm": 3.109286001494723,
"learning_rate": 1.1218938457701661e-07,
"loss": 0.864,
"step": 348
},
{
"epoch": 1.4316546762589928,
"grad_norm": 6.457283012285657,
"learning_rate": 1.1174822378605551e-07,
"loss": 1.0116,
"step": 349
},
{
"epoch": 1.435765673175745,
"grad_norm": 3.0895002017704183,
"learning_rate": 1.1130683102477863e-07,
"loss": 0.8629,
"step": 350
},
{
"epoch": 1.4398766700924974,
"grad_norm": 3.6763052961248133,
"learning_rate": 1.1086521500854744e-07,
"loss": 0.8757,
"step": 351
},
{
"epoch": 1.4439876670092497,
"grad_norm": 3.1496716800740985,
"learning_rate": 1.1042338445713184e-07,
"loss": 0.817,
"step": 352
},
{
"epoch": 1.448098663926002,
"grad_norm": 5.1946658489020825,
"learning_rate": 1.0998134809453756e-07,
"loss": 0.732,
"step": 353
},
{
"epoch": 1.4522096608427544,
"grad_norm": 3.829241413844801,
"learning_rate": 1.095391146488342e-07,
"loss": 0.9864,
"step": 354
},
{
"epoch": 1.4563206577595067,
"grad_norm": 3.7899885166138696,
"learning_rate": 1.0909669285198279e-07,
"loss": 0.9475,
"step": 355
},
{
"epoch": 1.460431654676259,
"grad_norm": 3.5441163883202242,
"learning_rate": 1.0865409143966337e-07,
"loss": 0.9404,
"step": 356
},
{
"epoch": 1.4645426515930113,
"grad_norm": 5.127809830712663,
"learning_rate": 1.0821131915110247e-07,
"loss": 1.0162,
"step": 357
},
{
"epoch": 1.4686536485097637,
"grad_norm": 3.8210553301046595,
"learning_rate": 1.0776838472890064e-07,
"loss": 0.904,
"step": 358
},
{
"epoch": 1.472764645426516,
"grad_norm": 3.423276389121683,
"learning_rate": 1.0732529691885977e-07,
"loss": 0.9258,
"step": 359
},
{
"epoch": 1.4768756423432683,
"grad_norm": 4.818251672123113,
"learning_rate": 1.0688206446981041e-07,
"loss": 0.9906,
"step": 360
},
{
"epoch": 1.4809866392600206,
"grad_norm": 4.163954679546723,
"learning_rate": 1.0643869613343906e-07,
"loss": 0.9437,
"step": 361
},
{
"epoch": 1.485097636176773,
"grad_norm": 3.9891570717000064,
"learning_rate": 1.059952006641153e-07,
"loss": 0.9786,
"step": 362
},
{
"epoch": 1.4892086330935252,
"grad_norm": 3.55257226968209,
"learning_rate": 1.0555158681871896e-07,
"loss": 0.9195,
"step": 363
},
{
"epoch": 1.4933196300102776,
"grad_norm": 3.241093467267381,
"learning_rate": 1.0510786335646723e-07,
"loss": 0.9582,
"step": 364
},
{
"epoch": 1.4974306269270299,
"grad_norm": 3.242022186577696,
"learning_rate": 1.0466403903874175e-07,
"loss": 0.8839,
"step": 365
},
{
"epoch": 1.501541623843782,
"grad_norm": 2.983071257935304,
"learning_rate": 1.0422012262891549e-07,
"loss": 0.8955,
"step": 366
},
{
"epoch": 1.5056526207605345,
"grad_norm": 4.190418089862188,
"learning_rate": 1.0377612289217982e-07,
"loss": 0.9547,
"step": 367
},
{
"epoch": 1.5097636176772866,
"grad_norm": 3.0118207209284056,
"learning_rate": 1.0333204859537142e-07,
"loss": 0.9189,
"step": 368
},
{
"epoch": 1.5138746145940392,
"grad_norm": 4.4088932253245385,
"learning_rate": 1.0288790850679916e-07,
"loss": 0.9642,
"step": 369
},
{
"epoch": 1.5179856115107913,
"grad_norm": 4.256377473558271,
"learning_rate": 1.02443711396071e-07,
"loss": 0.9967,
"step": 370
},
{
"epoch": 1.5220966084275438,
"grad_norm": 3.5285920421327877,
"learning_rate": 1.0199946603392078e-07,
"loss": 0.9941,
"step": 371
},
{
"epoch": 1.526207605344296,
"grad_norm": 3.228747946556862,
"learning_rate": 1.015551811920351e-07,
"loss": 0.8876,
"step": 372
},
{
"epoch": 1.5303186022610484,
"grad_norm": 4.690078191994573,
"learning_rate": 1.0111086564288003e-07,
"loss": 0.977,
"step": 373
},
{
"epoch": 1.5344295991778005,
"grad_norm": 3.4054812092989932,
"learning_rate": 1.0066652815952805e-07,
"loss": 0.8812,
"step": 374
},
{
"epoch": 1.538540596094553,
"grad_norm": 4.145441599920812,
"learning_rate": 1.0022217751548469e-07,
"loss": 0.8706,
"step": 375
},
{
"epoch": 1.5426515930113052,
"grad_norm": 2.8932427154784195,
"learning_rate": 9.977782248451532e-08,
"loss": 0.9032,
"step": 376
},
{
"epoch": 1.5467625899280577,
"grad_norm": 3.963915959396549,
"learning_rate": 9.933347184047194e-08,
"loss": 0.9419,
"step": 377
},
{
"epoch": 1.5508735868448098,
"grad_norm": 8.043380806796971,
"learning_rate": 9.888913435711997e-08,
"loss": 0.9767,
"step": 378
},
{
"epoch": 1.5549845837615623,
"grad_norm": 3.5125234270241474,
"learning_rate": 9.84448188079649e-08,
"loss": 0.9214,
"step": 379
},
{
"epoch": 1.5590955806783144,
"grad_norm": 4.003469090505716,
"learning_rate": 9.80005339660792e-08,
"loss": 0.9552,
"step": 380
},
{
"epoch": 1.5632065775950668,
"grad_norm": 4.640026799945273,
"learning_rate": 9.7556288603929e-08,
"loss": 0.9121,
"step": 381
},
{
"epoch": 1.567317574511819,
"grad_norm": 5.0645678331872785,
"learning_rate": 9.711209149320083e-08,
"loss": 0.9166,
"step": 382
},
{
"epoch": 1.5714285714285714,
"grad_norm": 2.7634743082527797,
"learning_rate": 9.666795140462859e-08,
"loss": 0.8692,
"step": 383
},
{
"epoch": 1.5755395683453237,
"grad_norm": 4.1199673904860195,
"learning_rate": 9.622387710782017e-08,
"loss": 0.8751,
"step": 384
},
{
"epoch": 1.579650565262076,
"grad_norm": 2.8379235943085903,
"learning_rate": 9.577987737108453e-08,
"loss": 0.8888,
"step": 385
},
{
"epoch": 1.5837615621788284,
"grad_norm": 5.882295965122743,
"learning_rate": 9.533596096125824e-08,
"loss": 0.9051,
"step": 386
},
{
"epoch": 1.5878725590955807,
"grad_norm": 3.420446189681305,
"learning_rate": 9.489213664353276e-08,
"loss": 0.9566,
"step": 387
},
{
"epoch": 1.591983556012333,
"grad_norm": 5.305798545532549,
"learning_rate": 9.444841318128103e-08,
"loss": 1.0193,
"step": 388
},
{
"epoch": 1.5960945529290853,
"grad_norm": 4.247055711315682,
"learning_rate": 9.400479933588468e-08,
"loss": 0.9384,
"step": 389
},
{
"epoch": 1.6002055498458376,
"grad_norm": 8.733285816075485,
"learning_rate": 9.356130386656093e-08,
"loss": 0.8369,
"step": 390
},
{
"epoch": 1.60431654676259,
"grad_norm": 2.8786782614816198,
"learning_rate": 9.311793553018958e-08,
"loss": 0.8824,
"step": 391
},
{
"epoch": 1.6084275436793423,
"grad_norm": 4.363120915051614,
"learning_rate": 9.267470308114025e-08,
"loss": 0.9449,
"step": 392
},
{
"epoch": 1.6125385405960946,
"grad_norm": 2.825067613906357,
"learning_rate": 9.223161527109936e-08,
"loss": 0.8717,
"step": 393
},
{
"epoch": 1.616649537512847,
"grad_norm": 3.4773329919642504,
"learning_rate": 9.178868084889756e-08,
"loss": 0.9516,
"step": 394
},
{
"epoch": 1.6207605344295992,
"grad_norm": 3.346888644174058,
"learning_rate": 9.134590856033663e-08,
"loss": 0.8735,
"step": 395
},
{
"epoch": 1.6248715313463515,
"grad_norm": 3.9626979838092122,
"learning_rate": 9.090330714801723e-08,
"loss": 0.9276,
"step": 396
},
{
"epoch": 1.6289825282631039,
"grad_norm": 4.276330156677023,
"learning_rate": 9.04608853511658e-08,
"loss": 0.8676,
"step": 397
},
{
"epoch": 1.6330935251798562,
"grad_norm": 3.1592785717660177,
"learning_rate": 9.001865190546244e-08,
"loss": 0.8915,
"step": 398
},
{
"epoch": 1.6372045220966083,
"grad_norm": 3.4798718911610456,
"learning_rate": 8.957661554286817e-08,
"loss": 0.8507,
"step": 399
},
{
"epoch": 1.6413155190133608,
"grad_norm": 5.913042910096891,
"learning_rate": 8.913478499145254e-08,
"loss": 0.9228,
"step": 400
},
{
"epoch": 1.645426515930113,
"grad_norm": 3.392731962857513,
"learning_rate": 8.86931689752214e-08,
"loss": 1.0059,
"step": 401
},
{
"epoch": 1.6495375128468655,
"grad_norm": 4.487647832133066,
"learning_rate": 8.825177621394448e-08,
"loss": 0.9056,
"step": 402
},
{
"epoch": 1.6536485097636175,
"grad_norm": 5.511907110296882,
"learning_rate": 8.781061542298341e-08,
"loss": 0.8975,
"step": 403
},
{
"epoch": 1.65775950668037,
"grad_norm": 3.329792413319853,
"learning_rate": 8.736969531311943e-08,
"loss": 0.8379,
"step": 404
},
{
"epoch": 1.6618705035971222,
"grad_norm": 3.2987510339919752,
"learning_rate": 8.69290245903816e-08,
"loss": 0.8905,
"step": 405
},
{
"epoch": 1.6659815005138747,
"grad_norm": 2.62381585897142,
"learning_rate": 8.648861195587475e-08,
"loss": 0.8473,
"step": 406
},
{
"epoch": 1.6700924974306268,
"grad_norm": 6.587278654350038,
"learning_rate": 8.60484661056077e-08,
"loss": 0.8982,
"step": 407
},
{
"epoch": 1.6742034943473794,
"grad_norm": 4.234639340313541,
"learning_rate": 8.560859573032161e-08,
"loss": 0.9484,
"step": 408
},
{
"epoch": 1.6783144912641315,
"grad_norm": 4.421720531070511,
"learning_rate": 8.516900951531831e-08,
"loss": 0.8253,
"step": 409
},
{
"epoch": 1.682425488180884,
"grad_norm": 2.8081070483958,
"learning_rate": 8.472971614028895e-08,
"loss": 0.8682,
"step": 410
},
{
"epoch": 1.686536485097636,
"grad_norm": 4.068593293283322,
"learning_rate": 8.429072427914234e-08,
"loss": 0.8271,
"step": 411
},
{
"epoch": 1.6906474820143886,
"grad_norm": 3.67545913643879,
"learning_rate": 8.385204259983403e-08,
"loss": 0.9262,
"step": 412
},
{
"epoch": 1.6947584789311407,
"grad_norm": 4.307597160621971,
"learning_rate": 8.341367976419485e-08,
"loss": 0.9034,
"step": 413
},
{
"epoch": 1.6988694758478933,
"grad_norm": 3.447324847960237,
"learning_rate": 8.297564442776013e-08,
"loss": 0.8751,
"step": 414
},
{
"epoch": 1.7029804727646454,
"grad_norm": 10.600275954525147,
"learning_rate": 8.253794523959863e-08,
"loss": 0.8835,
"step": 415
},
{
"epoch": 1.7070914696813977,
"grad_norm": 3.7164747813173538,
"learning_rate": 8.210059084214177e-08,
"loss": 0.9566,
"step": 416
},
{
"epoch": 1.71120246659815,
"grad_norm": 3.7287482542685337,
"learning_rate": 8.16635898710131e-08,
"loss": 0.8051,
"step": 417
},
{
"epoch": 1.7153134635149023,
"grad_norm": 4.997785314839548,
"learning_rate": 8.122695095485768e-08,
"loss": 0.9255,
"step": 418
},
{
"epoch": 1.7194244604316546,
"grad_norm": 3.7793087996899355,
"learning_rate": 8.079068271517182e-08,
"loss": 0.8596,
"step": 419
},
{
"epoch": 1.723535457348407,
"grad_norm": 2.719092083725673,
"learning_rate": 8.035479376613262e-08,
"loss": 0.8287,
"step": 420
},
{
"epoch": 1.7276464542651593,
"grad_norm": 3.617941539110509,
"learning_rate": 7.991929271442817e-08,
"loss": 0.9264,
"step": 421
},
{
"epoch": 1.7317574511819116,
"grad_norm": 2.7010258220031087,
"learning_rate": 7.948418815908739e-08,
"loss": 0.8017,
"step": 422
},
{
"epoch": 1.735868448098664,
"grad_norm": 2.928539904181109,
"learning_rate": 7.904948869131039e-08,
"loss": 0.7854,
"step": 423
},
{
"epoch": 1.7399794450154162,
"grad_norm": 4.509777337415114,
"learning_rate": 7.86152028942988e-08,
"loss": 0.9151,
"step": 424
},
{
"epoch": 1.7440904419321686,
"grad_norm": 3.114608066755098,
"learning_rate": 7.818133934308607e-08,
"loss": 0.8351,
"step": 425
},
{
"epoch": 1.7482014388489209,
"grad_norm": 3.3727557427516848,
"learning_rate": 7.774790660436856e-08,
"loss": 0.9537,
"step": 426
},
{
"epoch": 1.7523124357656732,
"grad_norm": 6.530624051161062,
"learning_rate": 7.731491323633608e-08,
"loss": 0.9381,
"step": 427
},
{
"epoch": 1.7564234326824255,
"grad_norm": 3.9314754514276076,
"learning_rate": 7.688236778850306e-08,
"loss": 0.9122,
"step": 428
},
{
"epoch": 1.7605344295991778,
"grad_norm": 3.547169223933416,
"learning_rate": 7.645027880153956e-08,
"loss": 0.8632,
"step": 429
},
{
"epoch": 1.7646454265159301,
"grad_norm": 3.213275075980098,
"learning_rate": 7.601865480710289e-08,
"loss": 0.8503,
"step": 430
},
{
"epoch": 1.7687564234326825,
"grad_norm": 4.934271588319233,
"learning_rate": 7.558750432766902e-08,
"loss": 0.8459,
"step": 431
},
{
"epoch": 1.7728674203494348,
"grad_norm": 3.029832690714125,
"learning_rate": 7.515683587636412e-08,
"loss": 0.9439,
"step": 432
},
{
"epoch": 1.776978417266187,
"grad_norm": 2.9552031221774473,
"learning_rate": 7.472665795679693e-08,
"loss": 0.8648,
"step": 433
},
{
"epoch": 1.7810894141829392,
"grad_norm": 3.227122116361125,
"learning_rate": 7.42969790628903e-08,
"loss": 0.8632,
"step": 434
},
{
"epoch": 1.7852004110996917,
"grad_norm": 3.4231798668824993,
"learning_rate": 7.386780767871396e-08,
"loss": 0.8415,
"step": 435
},
{
"epoch": 1.7893114080164438,
"grad_norm": 3.033262471123595,
"learning_rate": 7.343915227831661e-08,
"loss": 0.8922,
"step": 436
},
{
"epoch": 1.7934224049331964,
"grad_norm": 3.1916355473778473,
"learning_rate": 7.301102132555891e-08,
"loss": 0.8329,
"step": 437
},
{
"epoch": 1.7975334018499485,
"grad_norm": 3.8996217289980373,
"learning_rate": 7.258342327394616e-08,
"loss": 0.9226,
"step": 438
},
{
"epoch": 1.801644398766701,
"grad_norm": 2.9500360635886103,
"learning_rate": 7.215636656646151e-08,
"loss": 0.8526,
"step": 439
},
{
"epoch": 1.8057553956834531,
"grad_norm": 3.7180248722745795,
"learning_rate": 7.17298596353992e-08,
"loss": 0.8821,
"step": 440
},
{
"epoch": 1.8098663926002057,
"grad_norm": 3.077649806347787,
"learning_rate": 7.130391090219789e-08,
"loss": 0.8878,
"step": 441
},
{
"epoch": 1.8139773895169578,
"grad_norm": 4.71809796222561,
"learning_rate": 7.08785287772748e-08,
"loss": 0.8094,
"step": 442
},
{
"epoch": 1.8180883864337103,
"grad_norm": 3.2490450105344957,
"learning_rate": 7.045372165985919e-08,
"loss": 0.8366,
"step": 443
},
{
"epoch": 1.8221993833504624,
"grad_norm": 2.8069728461143417,
"learning_rate": 7.002949793782686e-08,
"loss": 0.8419,
"step": 444
},
{
"epoch": 1.826310380267215,
"grad_norm": 4.051243587736159,
"learning_rate": 6.960586598753427e-08,
"loss": 0.9234,
"step": 445
},
{
"epoch": 1.830421377183967,
"grad_norm": 3.474969546813908,
"learning_rate": 6.91828341736533e-08,
"loss": 0.9398,
"step": 446
},
{
"epoch": 1.8345323741007196,
"grad_norm": 3.767887619664162,
"learning_rate": 6.87604108490061e-08,
"loss": 0.8652,
"step": 447
},
{
"epoch": 1.8386433710174717,
"grad_norm": 4.108851946013631,
"learning_rate": 6.833860435440006e-08,
"loss": 1.0247,
"step": 448
},
{
"epoch": 1.8427543679342242,
"grad_norm": 3.4217065691964823,
"learning_rate": 6.791742301846325e-08,
"loss": 0.9332,
"step": 449
},
{
"epoch": 1.8468653648509763,
"grad_norm": 4.003287335513478,
"learning_rate": 6.749687515747977e-08,
"loss": 0.7957,
"step": 450
},
{
"epoch": 1.8509763617677288,
"grad_norm": 3.725894755816065,
"learning_rate": 6.707696907522577e-08,
"loss": 1.0451,
"step": 451
},
{
"epoch": 1.855087358684481,
"grad_norm": 4.443696729854911,
"learning_rate": 6.665771306280537e-08,
"loss": 0.8194,
"step": 452
},
{
"epoch": 1.8591983556012333,
"grad_norm": 4.132556354444622,
"learning_rate": 6.623911539848696e-08,
"loss": 0.8215,
"step": 453
},
{
"epoch": 1.8633093525179856,
"grad_norm": 7.461273938081753,
"learning_rate": 6.582118434753973e-08,
"loss": 0.9291,
"step": 454
},
{
"epoch": 1.867420349434738,
"grad_norm": 3.455918543455807,
"learning_rate": 6.540392816207054e-08,
"loss": 0.9164,
"step": 455
},
{
"epoch": 1.8715313463514902,
"grad_norm": 3.487975074813655,
"learning_rate": 6.498735508086093e-08,
"loss": 0.8576,
"step": 456
},
{
"epoch": 1.8756423432682425,
"grad_norm": 5.4993696031792085,
"learning_rate": 6.457147332920441e-08,
"loss": 0.8603,
"step": 457
},
{
"epoch": 1.8797533401849948,
"grad_norm": 5.194499862457218,
"learning_rate": 6.415629111874418e-08,
"loss": 0.9161,
"step": 458
},
{
"epoch": 1.8838643371017472,
"grad_norm": 6.9097273980176555,
"learning_rate": 6.374181664731075e-08,
"loss": 0.8747,
"step": 459
},
{
"epoch": 1.8879753340184995,
"grad_norm": 2.9565953135358893,
"learning_rate": 6.332805809876041e-08,
"loss": 0.8312,
"step": 460
},
{
"epoch": 1.8920863309352518,
"grad_norm": 4.208691232847417,
"learning_rate": 6.291502364281329e-08,
"loss": 0.8237,
"step": 461
},
{
"epoch": 1.8961973278520041,
"grad_norm": 3.4133001376901433,
"learning_rate": 6.250272143489236e-08,
"loss": 0.8498,
"step": 462
},
{
"epoch": 1.9003083247687564,
"grad_norm": 2.6542452726369117,
"learning_rate": 6.209115961596207e-08,
"loss": 0.8447,
"step": 463
},
{
"epoch": 1.9044193216855088,
"grad_norm": 2.6996240555205233,
"learning_rate": 6.168034631236794e-08,
"loss": 0.7443,
"step": 464
},
{
"epoch": 1.908530318602261,
"grad_norm": 2.9679087112357503,
"learning_rate": 6.127028963567593e-08,
"loss": 0.8142,
"step": 465
},
{
"epoch": 1.9126413155190134,
"grad_norm": 2.970029406337129,
"learning_rate": 6.086099768251221e-08,
"loss": 0.8812,
"step": 466
},
{
"epoch": 1.9167523124357657,
"grad_norm": 4.687799481829812,
"learning_rate": 6.04524785344035e-08,
"loss": 0.8872,
"step": 467
},
{
"epoch": 1.920863309352518,
"grad_norm": 4.867826422752404,
"learning_rate": 6.004474025761723e-08,
"loss": 0.9387,
"step": 468
},
{
"epoch": 1.9249743062692704,
"grad_norm": 3.0577878862885512,
"learning_rate": 5.963779090300256e-08,
"loss": 0.8316,
"step": 469
},
{
"epoch": 1.9290853031860227,
"grad_norm": 3.0514746459758264,
"learning_rate": 5.923163850583113e-08,
"loss": 0.7958,
"step": 470
},
{
"epoch": 1.9331963001027748,
"grad_norm": 4.368093714560947,
"learning_rate": 5.88262910856387e-08,
"loss": 0.8024,
"step": 471
},
{
"epoch": 1.9373072970195273,
"grad_norm": 2.8368973994018627,
"learning_rate": 5.84217566460665e-08,
"loss": 0.8502,
"step": 472
},
{
"epoch": 1.9414182939362794,
"grad_norm": 3.721426547887253,
"learning_rate": 5.801804317470349e-08,
"loss": 0.9421,
"step": 473
},
{
"epoch": 1.945529290853032,
"grad_norm": 2.8531515325709464,
"learning_rate": 5.761515864292835e-08,
"loss": 0.8858,
"step": 474
},
{
"epoch": 1.949640287769784,
"grad_norm": 4.49045301081062,
"learning_rate": 5.721311100575234e-08,
"loss": 0.9739,
"step": 475
},
{
"epoch": 1.9537512846865366,
"grad_norm": 4.314117767346784,
"learning_rate": 5.681190820166213e-08,
"loss": 0.8303,
"step": 476
},
{
"epoch": 1.9578622816032887,
"grad_norm": 4.1257319802992205,
"learning_rate": 5.641155815246289e-08,
"loss": 0.8144,
"step": 477
},
{
"epoch": 1.9619732785200412,
"grad_norm": 2.691841289359226,
"learning_rate": 5.601206876312223e-08,
"loss": 0.7583,
"step": 478
},
{
"epoch": 1.9660842754367933,
"grad_norm": 7.270643954599217,
"learning_rate": 5.561344792161373e-08,
"loss": 0.9272,
"step": 479
},
{
"epoch": 1.9701952723535459,
"grad_norm": 3.5160400139097066,
"learning_rate": 5.52157034987615e-08,
"loss": 0.8659,
"step": 480
},
{
"epoch": 1.974306269270298,
"grad_norm": 3.7017434660767075,
"learning_rate": 5.481884334808463e-08,
"loss": 0.9278,
"step": 481
},
{
"epoch": 1.9784172661870505,
"grad_norm": 2.591584218932893,
"learning_rate": 5.4422875305642026e-08,
"loss": 0.8878,
"step": 482
},
{
"epoch": 1.9825282631038026,
"grad_norm": 3.4697793428223456,
"learning_rate": 5.4027807189877894e-08,
"loss": 0.8272,
"step": 483
},
{
"epoch": 1.9866392600205551,
"grad_norm": 2.7032797326677453,
"learning_rate": 5.363364680146725e-08,
"loss": 0.783,
"step": 484
},
{
"epoch": 1.9907502569373072,
"grad_norm": 3.3338940021602412,
"learning_rate": 5.3240401923161924e-08,
"loss": 0.8049,
"step": 485
},
{
"epoch": 1.9948612538540598,
"grad_norm": 5.992113758655148,
"learning_rate": 5.2848080319636754e-08,
"loss": 0.8352,
"step": 486
},
{
"epoch": 1.9989722507708119,
"grad_norm": 3.29825757609477,
"learning_rate": 5.245668973733657e-08,
"loss": 0.8789,
"step": 487
},
{
"epoch": 2.0,
"grad_norm": 3.29825757609477,
"learning_rate": 5.206623790432285e-08,
"loss": 0.2231,
"step": 488
},
{
"epoch": 2.004110996916752,
"grad_norm": 3.788125568384661,
"learning_rate": 5.1676732530121526e-08,
"loss": 0.8911,
"step": 489
},
{
"epoch": 2.0082219938335046,
"grad_norm": 3.1409210087540695,
"learning_rate": 5.12881813055705e-08,
"loss": 0.8666,
"step": 490
},
{
"epoch": 2.0123329907502567,
"grad_norm": 2.936287868990542,
"learning_rate": 5.090059190266779e-08,
"loss": 0.7776,
"step": 491
},
{
"epoch": 2.0164439876670093,
"grad_norm": 3.4695711872528885,
"learning_rate": 5.0513971974420196e-08,
"loss": 0.9269,
"step": 492
},
{
"epoch": 2.0205549845837614,
"grad_norm": 3.2288752885643293,
"learning_rate": 5.012832915469207e-08,
"loss": 0.8954,
"step": 493
},
{
"epoch": 2.024665981500514,
"grad_norm": 3.2951959602692367,
"learning_rate": 4.97436710580547e-08,
"loss": 0.8826,
"step": 494
},
{
"epoch": 2.028776978417266,
"grad_norm": 2.917240255101616,
"learning_rate": 4.93600052796357e-08,
"loss": 0.8632,
"step": 495
},
{
"epoch": 2.0328879753340185,
"grad_norm": 5.187844927004196,
"learning_rate": 4.897733939496942e-08,
"loss": 0.9416,
"step": 496
},
{
"epoch": 2.0369989722507706,
"grad_norm": 3.328555248428224,
"learning_rate": 4.8595680959846993e-08,
"loss": 0.8429,
"step": 497
},
{
"epoch": 2.041109969167523,
"grad_norm": 3.3795750665305295,
"learning_rate": 4.8215037510167455e-08,
"loss": 0.9195,
"step": 498
},
{
"epoch": 2.0452209660842753,
"grad_norm": 3.8481339463830206,
"learning_rate": 4.78354165617888e-08,
"loss": 0.8822,
"step": 499
},
{
"epoch": 2.049331963001028,
"grad_norm": 4.595587711491347,
"learning_rate": 4.745682561037947e-08,
"loss": 0.826,
"step": 500
},
{
"epoch": 2.05344295991778,
"grad_norm": 3.604573911368687,
"learning_rate": 4.707927213127062e-08,
"loss": 0.9112,
"step": 501
},
{
"epoch": 2.0575539568345325,
"grad_norm": 4.386570815030496,
"learning_rate": 4.670276357930828e-08,
"loss": 0.7542,
"step": 502
},
{
"epoch": 2.0616649537512846,
"grad_norm": 2.808587931273854,
"learning_rate": 4.6327307388706335e-08,
"loss": 0.874,
"step": 503
},
{
"epoch": 2.065775950668037,
"grad_norm": 5.420247112836083,
"learning_rate": 4.5952910972899514e-08,
"loss": 0.8795,
"step": 504
},
{
"epoch": 2.069886947584789,
"grad_norm": 3.453879289279688,
"learning_rate": 4.557958172439725e-08,
"loss": 0.8367,
"step": 505
},
{
"epoch": 2.0739979445015417,
"grad_norm": 3.0676038054022476,
"learning_rate": 4.520732701463762e-08,
"loss": 0.8598,
"step": 506
},
{
"epoch": 2.078108941418294,
"grad_norm": 7.756550710403295,
"learning_rate": 4.483615419384167e-08,
"loss": 0.8316,
"step": 507
},
{
"epoch": 2.0822199383350464,
"grad_norm": 3.7300401780041756,
"learning_rate": 4.4466070590868543e-08,
"loss": 0.8424,
"step": 508
},
{
"epoch": 2.0863309352517985,
"grad_norm": 2.5645367588246355,
"learning_rate": 4.4097083513070486e-08,
"loss": 0.6788,
"step": 509
},
{
"epoch": 2.090441932168551,
"grad_norm": 2.5093860960425434,
"learning_rate": 4.372920024614883e-08,
"loss": 0.8171,
"step": 510
},
{
"epoch": 2.094552929085303,
"grad_norm": 3.4968063083940306,
"learning_rate": 4.336242805400989e-08,
"loss": 0.8415,
"step": 511
},
{
"epoch": 2.0986639260020556,
"grad_norm": 3.3707767027220195,
"learning_rate": 4.299677417862173e-08,
"loss": 0.864,
"step": 512
},
{
"epoch": 2.1027749229188077,
"grad_norm": 3.331745614115561,
"learning_rate": 4.2632245839871095e-08,
"loss": 0.9011,
"step": 513
},
{
"epoch": 2.1068859198355603,
"grad_norm": 3.2473204408643226,
"learning_rate": 4.226885023542083e-08,
"loss": 0.9682,
"step": 514
},
{
"epoch": 2.1109969167523124,
"grad_norm": 4.193878520121946,
"learning_rate": 4.19065945405678e-08,
"loss": 0.8534,
"step": 515
},
{
"epoch": 2.115107913669065,
"grad_norm": 2.8706632043507443,
"learning_rate": 4.154548590810113e-08,
"loss": 0.8982,
"step": 516
},
{
"epoch": 2.119218910585817,
"grad_norm": 4.48257638858821,
"learning_rate": 4.118553146816115e-08,
"loss": 0.8248,
"step": 517
},
{
"epoch": 2.1233299075025696,
"grad_norm": 3.0041615871924523,
"learning_rate": 4.082673832809838e-08,
"loss": 0.855,
"step": 518
},
{
"epoch": 2.1274409044193217,
"grad_norm": 3.1022048007172414,
"learning_rate": 4.046911357233342e-08,
"loss": 0.8076,
"step": 519
},
{
"epoch": 2.131551901336074,
"grad_norm": 3.2451093336201846,
"learning_rate": 4.0112664262216865e-08,
"loss": 0.8012,
"step": 520
},
{
"epoch": 2.1356628982528263,
"grad_norm": 3.149969064323508,
"learning_rate": 3.975739743589004e-08,
"loss": 0.7997,
"step": 521
},
{
"epoch": 2.1397738951695784,
"grad_norm": 3.179256311222312,
"learning_rate": 3.9403320108145946e-08,
"loss": 0.8268,
"step": 522
},
{
"epoch": 2.143884892086331,
"grad_norm": 3.151437821639254,
"learning_rate": 3.9050439270290735e-08,
"loss": 0.806,
"step": 523
},
{
"epoch": 2.1479958890030835,
"grad_norm": 3.338862297804263,
"learning_rate": 3.8698761890005795e-08,
"loss": 0.9055,
"step": 524
},
{
"epoch": 2.1521068859198356,
"grad_norm": 4.031692479052263,
"learning_rate": 3.834829491120991e-08,
"loss": 0.8099,
"step": 525
},
{
"epoch": 2.1562178828365877,
"grad_norm": 3.330009061639912,
"learning_rate": 3.7999045253922504e-08,
"loss": 0.8204,
"step": 526
},
{
"epoch": 2.16032887975334,
"grad_norm": 3.3480890276183253,
"learning_rate": 3.765101981412665e-08,
"loss": 0.8667,
"step": 527
},
{
"epoch": 2.1644398766700923,
"grad_norm": 3.0538487683948308,
"learning_rate": 3.7304225463633234e-08,
"loss": 0.8266,
"step": 528
},
{
"epoch": 2.168550873586845,
"grad_norm": 3.246745268650053,
"learning_rate": 3.695866904994496e-08,
"loss": 0.8202,
"step": 529
},
{
"epoch": 2.172661870503597,
"grad_norm": 3.5140206385964494,
"learning_rate": 3.66143573961214e-08,
"loss": 0.8063,
"step": 530
},
{
"epoch": 2.1767728674203495,
"grad_norm": 5.111702162587348,
"learning_rate": 3.6271297300644153e-08,
"loss": 0.8617,
"step": 531
},
{
"epoch": 2.1808838643371016,
"grad_norm": 3.0115422681262007,
"learning_rate": 3.592949553728259e-08,
"loss": 0.7375,
"step": 532
},
{
"epoch": 2.184994861253854,
"grad_norm": 2.666365414113028,
"learning_rate": 3.5588958854960224e-08,
"loss": 0.8883,
"step": 533
},
{
"epoch": 2.189105858170606,
"grad_norm": 2.53476271631746,
"learning_rate": 3.524969397762122e-08,
"loss": 0.7137,
"step": 534
},
{
"epoch": 2.1932168550873588,
"grad_norm": 3.760906900900401,
"learning_rate": 3.491170760409792e-08,
"loss": 0.931,
"step": 535
},
{
"epoch": 2.197327852004111,
"grad_norm": 4.831171143278271,
"learning_rate": 3.45750064079783e-08,
"loss": 0.8715,
"step": 536
},
{
"epoch": 2.2014388489208634,
"grad_norm": 2.976809093825922,
"learning_rate": 3.4239597037474486e-08,
"loss": 0.7694,
"step": 537
},
{
"epoch": 2.2055498458376155,
"grad_norm": 3.0028148210421626,
"learning_rate": 3.390548611529116e-08,
"loss": 0.8088,
"step": 538
},
{
"epoch": 2.209660842754368,
"grad_norm": 3.4315940934966522,
"learning_rate": 3.3572680238495066e-08,
"loss": 0.9181,
"step": 539
},
{
"epoch": 2.21377183967112,
"grad_norm": 4.705649298484479,
"learning_rate": 3.324118597838463e-08,
"loss": 0.8721,
"step": 540
},
{
"epoch": 2.2178828365878727,
"grad_norm": 9.135213871248446,
"learning_rate": 3.291100988036022e-08,
"loss": 0.7863,
"step": 541
},
{
"epoch": 2.2219938335046248,
"grad_norm": 2.862245170449049,
"learning_rate": 3.258215846379492e-08,
"loss": 0.8541,
"step": 542
},
{
"epoch": 2.2261048304213773,
"grad_norm": 68.0212080766877,
"learning_rate": 3.2254638221905715e-08,
"loss": 0.7869,
"step": 543
},
{
"epoch": 2.2302158273381294,
"grad_norm": 3.7564261616277226,
"learning_rate": 3.192845562162549e-08,
"loss": 0.8967,
"step": 544
},
{
"epoch": 2.234326824254882,
"grad_norm": 3.69398211708455,
"learning_rate": 3.160361710347508e-08,
"loss": 0.8964,
"step": 545
},
{
"epoch": 2.238437821171634,
"grad_norm": 4.48268222512928,
"learning_rate": 3.128012908143636e-08,
"loss": 0.9815,
"step": 546
},
{
"epoch": 2.2425488180883866,
"grad_norm": 4.045335878354018,
"learning_rate": 3.0957997942825333e-08,
"loss": 0.887,
"step": 547
},
{
"epoch": 2.2466598150051387,
"grad_norm": 3.6274208839986586,
"learning_rate": 3.063723004816626e-08,
"loss": 0.9293,
"step": 548
},
{
"epoch": 2.250770811921891,
"grad_norm": 3.083802546670731,
"learning_rate": 3.0317831731065954e-08,
"loss": 0.7954,
"step": 549
},
{
"epoch": 2.2548818088386433,
"grad_norm": 2.5117889993907374,
"learning_rate": 2.999980929808863e-08,
"loss": 0.8399,
"step": 550
},
{
"epoch": 2.258992805755396,
"grad_norm": 3.6475431912649334,
"learning_rate": 2.9683169028631573e-08,
"loss": 0.8361,
"step": 551
},
{
"epoch": 2.263103802672148,
"grad_norm": 3.008144832101907,
"learning_rate": 2.9367917174801037e-08,
"loss": 0.8918,
"step": 552
},
{
"epoch": 2.2672147995889005,
"grad_norm": 3.082675833585662,
"learning_rate": 2.9054059961288823e-08,
"loss": 0.7753,
"step": 553
},
{
"epoch": 2.2713257965056526,
"grad_norm": 3.712032750324046,
"learning_rate": 2.8741603585249308e-08,
"loss": 0.8329,
"step": 554
},
{
"epoch": 2.275436793422405,
"grad_norm": 5.148982914157342,
"learning_rate": 2.8430554216177206e-08,
"loss": 0.8071,
"step": 555
},
{
"epoch": 2.279547790339157,
"grad_norm": 2.6445849682408182,
"learning_rate": 2.8120917995785665e-08,
"loss": 0.8853,
"step": 556
},
{
"epoch": 2.2836587872559093,
"grad_norm": 3.3217979182178707,
"learning_rate": 2.7812701037884968e-08,
"loss": 0.7882,
"step": 557
},
{
"epoch": 2.287769784172662,
"grad_norm": 2.774730734505095,
"learning_rate": 2.750590942826195e-08,
"loss": 0.8469,
"step": 558
},
{
"epoch": 2.2918807810894144,
"grad_norm": 3.5447043537286533,
"learning_rate": 2.7200549224559643e-08,
"loss": 0.8577,
"step": 559
},
{
"epoch": 2.2959917780061665,
"grad_norm": 4.022819506448639,
"learning_rate": 2.6896626456157844e-08,
"loss": 0.9231,
"step": 560
},
{
"epoch": 2.3001027749229186,
"grad_norm": 3.5997919371836393,
"learning_rate": 2.6594147124053977e-08,
"loss": 0.8417,
"step": 561
},
{
"epoch": 2.304213771839671,
"grad_norm": 3.8744215623947786,
"learning_rate": 2.629311720074464e-08,
"loss": 0.8707,
"step": 562
},
{
"epoch": 2.3083247687564237,
"grad_norm": 7.031602732118638,
"learning_rate": 2.5993542630107535e-08,
"loss": 0.8512,
"step": 563
},
{
"epoch": 2.3124357656731758,
"grad_norm": 8.88042690396173,
"learning_rate": 2.5695429327284356e-08,
"loss": 0.7847,
"step": 564
},
{
"epoch": 2.316546762589928,
"grad_norm": 3.708864526434595,
"learning_rate": 2.5398783178563844e-08,
"loss": 0.8185,
"step": 565
},
{
"epoch": 2.3206577595066804,
"grad_norm": 15.573104340264974,
"learning_rate": 2.5103610041265476e-08,
"loss": 0.9161,
"step": 566
},
{
"epoch": 2.3247687564234325,
"grad_norm": 7.781988013139914,
"learning_rate": 2.4809915743624033e-08,
"loss": 0.8637,
"step": 567
},
{
"epoch": 2.328879753340185,
"grad_norm": 7.17219485886084,
"learning_rate": 2.4517706084674317e-08,
"loss": 0.8198,
"step": 568
},
{
"epoch": 2.332990750256937,
"grad_norm": 4.1048918408793345,
"learning_rate": 2.4226986834136766e-08,
"loss": 0.903,
"step": 569
},
{
"epoch": 2.3371017471736897,
"grad_norm": 2.9736841676267933,
"learning_rate": 2.3937763732303505e-08,
"loss": 0.8344,
"step": 570
},
{
"epoch": 2.3412127440904418,
"grad_norm": 3.017208388644247,
"learning_rate": 2.3650042489924992e-08,
"loss": 0.8497,
"step": 571
},
{
"epoch": 2.3453237410071943,
"grad_norm": 37.33575099459087,
"learning_rate": 2.3363828788097273e-08,
"loss": 0.9639,
"step": 572
},
{
"epoch": 2.3494347379239464,
"grad_norm": 4.310058275385428,
"learning_rate": 2.3079128278149717e-08,
"loss": 0.8119,
"step": 573
},
{
"epoch": 2.353545734840699,
"grad_norm": 10.04395663981547,
"learning_rate": 2.279594658153363e-08,
"loss": 0.9621,
"step": 574
},
{
"epoch": 2.357656731757451,
"grad_norm": 6.035004365620923,
"learning_rate": 2.2514289289711018e-08,
"loss": 0.8278,
"step": 575
},
{
"epoch": 2.3617677286742036,
"grad_norm": 3.9886942484442556,
"learning_rate": 2.223416196404442e-08,
"loss": 0.8021,
"step": 576
},
{
"epoch": 2.3658787255909557,
"grad_norm": 3.479308541969328,
"learning_rate": 2.195557013568684e-08,
"loss": 0.8671,
"step": 577
},
{
"epoch": 2.3699897225077082,
"grad_norm": 3.153045230800823,
"learning_rate": 2.167851930547283e-08,
"loss": 0.9059,
"step": 578
},
{
"epoch": 2.3741007194244603,
"grad_norm": 4.341122896062692,
"learning_rate": 2.140301494380956e-08,
"loss": 0.8108,
"step": 579
},
{
"epoch": 2.378211716341213,
"grad_norm": 6.067412801857111,
"learning_rate": 2.1129062490569104e-08,
"loss": 0.8742,
"step": 580
},
{
"epoch": 2.382322713257965,
"grad_norm": 3.719527238514641,
"learning_rate": 2.085666735498085e-08,
"loss": 0.9089,
"step": 581
},
{
"epoch": 2.3864337101747175,
"grad_norm": 4.4092615707885665,
"learning_rate": 2.0585834915524647e-08,
"loss": 0.9143,
"step": 582
},
{
"epoch": 2.3905447070914696,
"grad_norm": 3.850699314352751,
"learning_rate": 2.0316570519824806e-08,
"loss": 0.8269,
"step": 583
},
{
"epoch": 2.394655704008222,
"grad_norm": 3.106803915842057,
"learning_rate": 2.004887948454428e-08,
"loss": 0.849,
"step": 584
},
{
"epoch": 2.3987667009249742,
"grad_norm": 3.3528527556785632,
"learning_rate": 1.9782767095279938e-08,
"loss": 0.8775,
"step": 585
},
{
"epoch": 2.402877697841727,
"grad_norm": 3.569646163136982,
"learning_rate": 1.9518238606457928e-08,
"loss": 0.8509,
"step": 586
},
{
"epoch": 2.406988694758479,
"grad_norm": 4.765189710112887,
"learning_rate": 1.9255299241230184e-08,
"loss": 0.8963,
"step": 587
},
{
"epoch": 2.4110996916752314,
"grad_norm": 3.444243144599029,
"learning_rate": 1.8993954191371044e-08,
"loss": 0.9227,
"step": 588
},
{
"epoch": 2.4152106885919835,
"grad_norm": 4.8921534659605355,
"learning_rate": 1.8734208617174984e-08,
"loss": 0.7786,
"step": 589
},
{
"epoch": 2.419321685508736,
"grad_norm": 2.923651684956854,
"learning_rate": 1.8476067647354553e-08,
"loss": 0.883,
"step": 590
},
{
"epoch": 2.423432682425488,
"grad_norm": 3.5194748675039333,
"learning_rate": 1.8219536378939172e-08,
"loss": 0.8822,
"step": 591
},
{
"epoch": 2.4275436793422402,
"grad_norm": 3.4954859859634366,
"learning_rate": 1.7964619877174515e-08,
"loss": 0.8292,
"step": 592
},
{
"epoch": 2.431654676258993,
"grad_norm": 2.7558605688331683,
"learning_rate": 1.7711323175422376e-08,
"loss": 0.8403,
"step": 593
},
{
"epoch": 2.4357656731757453,
"grad_norm": 3.6546323506771503,
"learning_rate": 1.7459651275061482e-08,
"loss": 0.8484,
"step": 594
},
{
"epoch": 2.4398766700924974,
"grad_norm": 2.8602332467102554,
"learning_rate": 1.7209609145388538e-08,
"loss": 0.8007,
"step": 595
},
{
"epoch": 2.4439876670092495,
"grad_norm": 3.395869287848844,
"learning_rate": 1.6961201723520247e-08,
"loss": 0.8711,
"step": 596
},
{
"epoch": 2.448098663926002,
"grad_norm": 3.8655664953844147,
"learning_rate": 1.6714433914295813e-08,
"loss": 0.8997,
"step": 597
},
{
"epoch": 2.4522096608427546,
"grad_norm": 3.082882184138082,
"learning_rate": 1.646931059017994e-08,
"loss": 0.8074,
"step": 598
},
{
"epoch": 2.4563206577595067,
"grad_norm": 3.8019349707346293,
"learning_rate": 1.6225836591166886e-08,
"loss": 0.8959,
"step": 599
},
{
"epoch": 2.460431654676259,
"grad_norm": 5.711695849364997,
"learning_rate": 1.5984016724684657e-08,
"loss": 0.8462,
"step": 600
},
{
"epoch": 2.4645426515930113,
"grad_norm": 3.0357669324430883,
"learning_rate": 1.5743855765500258e-08,
"loss": 0.8364,
"step": 601
},
{
"epoch": 2.4686536485097634,
"grad_norm": 3.5696190593271924,
"learning_rate": 1.5505358455625227e-08,
"loss": 0.7571,
"step": 602
},
{
"epoch": 2.472764645426516,
"grad_norm": 2.940498976987975,
"learning_rate": 1.526852950422226e-08,
"loss": 0.8663,
"step": 603
},
{
"epoch": 2.476875642343268,
"grad_norm": 4.367892519028362,
"learning_rate": 1.5033373587511945e-08,
"loss": 0.849,
"step": 604
},
{
"epoch": 2.4809866392600206,
"grad_norm": 2.717195294882748,
"learning_rate": 1.4799895348680646e-08,
"loss": 0.8264,
"step": 605
},
{
"epoch": 2.4850976361767727,
"grad_norm": 2.9304814296246287,
"learning_rate": 1.4568099397788746e-08,
"loss": 0.8071,
"step": 606
},
{
"epoch": 2.4892086330935252,
"grad_norm": 2.9661720190431304,
"learning_rate": 1.4337990311679571e-08,
"loss": 0.8461,
"step": 607
},
{
"epoch": 2.4933196300102773,
"grad_norm": 2.9674241787582365,
"learning_rate": 1.4109572633889088e-08,
"loss": 0.8322,
"step": 608
},
{
"epoch": 2.49743062692703,
"grad_norm": 3.5877409004366085,
"learning_rate": 1.3882850874556207e-08,
"loss": 0.786,
"step": 609
},
{
"epoch": 2.501541623843782,
"grad_norm": 3.1505885217575327,
"learning_rate": 1.3657829510333652e-08,
"loss": 0.8131,
"step": 610
},
{
"epoch": 2.5056526207605345,
"grad_norm": 2.520354798383407,
"learning_rate": 1.3434512984299595e-08,
"loss": 0.8327,
"step": 611
},
{
"epoch": 2.5097636176772866,
"grad_norm": 3.6434837180016624,
"learning_rate": 1.3212905705869992e-08,
"loss": 0.8327,
"step": 612
},
{
"epoch": 2.513874614594039,
"grad_norm": 2.677761452843666,
"learning_rate": 1.2993012050711406e-08,
"loss": 0.8096,
"step": 613
},
{
"epoch": 2.5179856115107913,
"grad_norm": 3.315810263396286,
"learning_rate": 1.2774836360654717e-08,
"loss": 0.8357,
"step": 614
},
{
"epoch": 2.522096608427544,
"grad_norm": 3.3376459149001128,
"learning_rate": 1.2558382943609358e-08,
"loss": 0.7958,
"step": 615
},
{
"epoch": 2.526207605344296,
"grad_norm": 5.332602967938458,
"learning_rate": 1.2343656073478158e-08,
"loss": 0.845,
"step": 616
},
{
"epoch": 2.5303186022610484,
"grad_norm": 3.328379844626599,
"learning_rate": 1.2130659990073145e-08,
"loss": 0.8106,
"step": 617
},
{
"epoch": 2.5344295991778005,
"grad_norm": 4.427066231420292,
"learning_rate": 1.1919398899031585e-08,
"loss": 0.829,
"step": 618
},
{
"epoch": 2.538540596094553,
"grad_norm": 4.201813235555996,
"learning_rate": 1.1709876971733267e-08,
"loss": 0.7907,
"step": 619
},
{
"epoch": 2.542651593011305,
"grad_norm": 2.7822781712834836,
"learning_rate": 1.150209834521777e-08,
"loss": 0.8899,
"step": 620
},
{
"epoch": 2.5467625899280577,
"grad_norm": 2.6297739304892267,
"learning_rate": 1.1296067122103059e-08,
"loss": 0.8182,
"step": 621
},
{
"epoch": 2.55087358684481,
"grad_norm": 3.3878130722233433,
"learning_rate": 1.1091787370504347e-08,
"loss": 0.9358,
"step": 622
},
{
"epoch": 2.5549845837615623,
"grad_norm": 6.328833340578769,
"learning_rate": 1.0889263123953774e-08,
"loss": 0.866,
"step": 623
},
{
"epoch": 2.5590955806783144,
"grad_norm": 4.854945622680054,
"learning_rate": 1.0688498381320853e-08,
"loss": 0.8313,
"step": 624
},
{
"epoch": 2.563206577595067,
"grad_norm": 3.1753110601435846,
"learning_rate": 1.0489497106733346e-08,
"loss": 0.867,
"step": 625
},
{
"epoch": 2.567317574511819,
"grad_norm": 3.7564805253370763,
"learning_rate": 1.029226322949921e-08,
"loss": 0.7942,
"step": 626
},
{
"epoch": 2.571428571428571,
"grad_norm": 7.681715445645189,
"learning_rate": 1.0096800644028791e-08,
"loss": 0.84,
"step": 627
},
{
"epoch": 2.5755395683453237,
"grad_norm": 3.651375559118566,
"learning_rate": 9.903113209758097e-09,
"loss": 0.808,
"step": 628
},
{
"epoch": 2.5796505652620763,
"grad_norm": 3.0372621659311325,
"learning_rate": 9.7112047510725e-09,
"loss": 0.9234,
"step": 629
},
{
"epoch": 2.5837615621788284,
"grad_norm": 3.921253331472848,
"learning_rate": 9.521079057231273e-09,
"loss": 0.9036,
"step": 630
},
{
"epoch": 2.5878725590955804,
"grad_norm": 3.2169002274136336,
"learning_rate": 9.332739882292751e-09,
"loss": 0.8106,
"step": 631
},
{
"epoch": 2.591983556012333,
"grad_norm": 6.452579795168197,
"learning_rate": 9.146190945040144e-09,
"loss": 0.7761,
"step": 632
},
{
"epoch": 2.5960945529290855,
"grad_norm": 3.3837697189410623,
"learning_rate": 8.961435928908268e-09,
"loss": 0.8426,
"step": 633
},
{
"epoch": 2.6002055498458376,
"grad_norm": 3.4349518498334155,
"learning_rate": 8.778478481910612e-09,
"loss": 0.8111,
"step": 634
},
{
"epoch": 2.6043165467625897,
"grad_norm": 3.925148969414287,
"learning_rate": 8.597322216567492e-09,
"loss": 0.9536,
"step": 635
},
{
"epoch": 2.6084275436793423,
"grad_norm": 2.9922197995668323,
"learning_rate": 8.41797070983461e-09,
"loss": 0.872,
"step": 636
},
{
"epoch": 2.612538540596095,
"grad_norm": 3.310137154991517,
"learning_rate": 8.240427503032443e-09,
"loss": 0.8188,
"step": 637
},
{
"epoch": 2.616649537512847,
"grad_norm": 3.1101161666153225,
"learning_rate": 8.064696101776358e-09,
"loss": 0.8388,
"step": 638
},
{
"epoch": 2.620760534429599,
"grad_norm": 2.574123766024681,
"learning_rate": 7.89077997590738e-09,
"loss": 0.8894,
"step": 639
},
{
"epoch": 2.6248715313463515,
"grad_norm": 3.050084057911728,
"learning_rate": 7.71868255942365e-09,
"loss": 0.7846,
"step": 640
},
{
"epoch": 2.628982528263104,
"grad_norm": 3.719049663953326,
"learning_rate": 7.548407250412613e-09,
"loss": 0.8126,
"step": 641
},
{
"epoch": 2.633093525179856,
"grad_norm": 3.237959977090943,
"learning_rate": 7.379957410983994e-09,
"loss": 0.78,
"step": 642
},
{
"epoch": 2.6372045220966083,
"grad_norm": 3.0769679112962627,
"learning_rate": 7.213336367203338e-09,
"loss": 0.8138,
"step": 643
},
{
"epoch": 2.641315519013361,
"grad_norm": 4.749266169904812,
"learning_rate": 7.048547409026384e-09,
"loss": 0.9123,
"step": 644
},
{
"epoch": 2.645426515930113,
"grad_norm": 7.742779907813039,
"learning_rate": 6.885593790234057e-09,
"loss": 0.806,
"step": 645
},
{
"epoch": 2.6495375128468655,
"grad_norm": 4.508430942876999,
"learning_rate": 6.7244787283682765e-09,
"loss": 0.925,
"step": 646
},
{
"epoch": 2.6536485097636175,
"grad_norm": 6.815785595171212,
"learning_rate": 6.5652054046683945e-09,
"loss": 0.8828,
"step": 647
},
{
"epoch": 2.65775950668037,
"grad_norm": 3.8021810178386164,
"learning_rate": 6.4077769640083825e-09,
"loss": 0.8803,
"step": 648
},
{
"epoch": 2.661870503597122,
"grad_norm": 3.957459657139488,
"learning_rate": 6.252196514834751e-09,
"loss": 0.8844,
"step": 649
},
{
"epoch": 2.6659815005138747,
"grad_norm": 3.068224928717054,
"learning_rate": 6.098467129105123e-09,
"loss": 0.9214,
"step": 650
},
{
"epoch": 2.670092497430627,
"grad_norm": 4.263631671447282,
"learning_rate": 5.946591842227677e-09,
"loss": 0.9347,
"step": 651
},
{
"epoch": 2.6742034943473794,
"grad_norm": 3.3185240835345566,
"learning_rate": 5.796573653001091e-09,
"loss": 0.7484,
"step": 652
},
{
"epoch": 2.6783144912641315,
"grad_norm": 4.16664895900444,
"learning_rate": 5.6484155235554275e-09,
"loss": 1.0615,
"step": 653
},
{
"epoch": 2.682425488180884,
"grad_norm": 3.647067682268931,
"learning_rate": 5.5021203792935845e-09,
"loss": 0.8857,
"step": 654
},
{
"epoch": 2.686536485097636,
"grad_norm": 4.364802496783611,
"learning_rate": 5.357691108833584e-09,
"loss": 0.8893,
"step": 655
},
{
"epoch": 2.6906474820143886,
"grad_norm": 2.624845495273688,
"learning_rate": 5.2151305639515195e-09,
"loss": 0.7829,
"step": 656
},
{
"epoch": 2.6947584789311407,
"grad_norm": 4.64252053639452,
"learning_rate": 5.0744415595251665e-09,
"loss": 0.8786,
"step": 657
},
{
"epoch": 2.6988694758478933,
"grad_norm": 5.876633976573797,
"learning_rate": 4.93562687347856e-09,
"loss": 0.8088,
"step": 658
},
{
"epoch": 2.7029804727646454,
"grad_norm": 3.8195676526281654,
"learning_rate": 4.798689246727006e-09,
"loss": 0.8378,
"step": 659
},
{
"epoch": 2.707091469681398,
"grad_norm": 4.397865887265361,
"learning_rate": 4.663631383123057e-09,
"loss": 0.7731,
"step": 660
},
{
"epoch": 2.71120246659815,
"grad_norm": 3.3686107705218005,
"learning_rate": 4.530455949403e-09,
"loss": 0.8434,
"step": 661
},
{
"epoch": 2.715313463514902,
"grad_norm": 3.683379272534775,
"learning_rate": 4.3991655751343784e-09,
"loss": 0.8673,
"step": 662
},
{
"epoch": 2.7194244604316546,
"grad_norm": 12.36137313275503,
"learning_rate": 4.269762852663894e-09,
"loss": 0.8432,
"step": 663
},
{
"epoch": 2.723535457348407,
"grad_norm": 3.09541559967358,
"learning_rate": 4.142250337066356e-09,
"loss": 0.9148,
"step": 664
},
{
"epoch": 2.7276464542651593,
"grad_norm": 3.5939723980137233,
"learning_rate": 4.0166305460941576e-09,
"loss": 0.9074,
"step": 665
},
{
"epoch": 2.7317574511819114,
"grad_norm": 2.714058041033449,
"learning_rate": 3.892905960127546e-09,
"loss": 0.709,
"step": 666
},
{
"epoch": 2.735868448098664,
"grad_norm": 2.982802966954589,
"learning_rate": 3.7710790221257445e-09,
"loss": 0.7859,
"step": 667
},
{
"epoch": 2.7399794450154165,
"grad_norm": 3.442761462953877,
"learning_rate": 3.651152137578617e-09,
"loss": 0.804,
"step": 668
},
{
"epoch": 2.7440904419321686,
"grad_norm": 2.372601936691603,
"learning_rate": 3.533127674459202e-09,
"loss": 0.8447,
"step": 669
},
{
"epoch": 2.7482014388489207,
"grad_norm": 3.655236664258741,
"learning_rate": 3.4170079631769765e-09,
"loss": 0.8877,
"step": 670
},
{
"epoch": 2.752312435765673,
"grad_norm": 4.144320095609748,
"learning_rate": 3.302795296531813e-09,
"loss": 0.8645,
"step": 671
},
{
"epoch": 2.7564234326824257,
"grad_norm": 3.552089367818775,
"learning_rate": 3.1904919296687484e-09,
"loss": 0.87,
"step": 672
},
{
"epoch": 2.760534429599178,
"grad_norm": 3.4223600911422913,
"learning_rate": 3.0801000800333876e-09,
"loss": 0.8475,
"step": 673
},
{
"epoch": 2.76464542651593,
"grad_norm": 3.289725575981045,
"learning_rate": 2.971621927328216e-09,
"loss": 0.9519,
"step": 674
},
{
"epoch": 2.7687564234326825,
"grad_norm": 3.3889285846976263,
"learning_rate": 2.865059613469434e-09,
"loss": 0.8779,
"step": 675
},
{
"epoch": 2.772867420349435,
"grad_norm": 3.5198142186730155,
"learning_rate": 2.7604152425448114e-09,
"loss": 0.7689,
"step": 676
},
{
"epoch": 2.776978417266187,
"grad_norm": 2.7823724178901217,
"learning_rate": 2.6576908807720233e-09,
"loss": 0.867,
"step": 677
},
{
"epoch": 2.781089414182939,
"grad_norm": 3.930670697134359,
"learning_rate": 2.556888556457926e-09,
"loss": 0.8453,
"step": 678
},
{
"epoch": 2.7852004110996917,
"grad_norm": 3.0308327243392372,
"learning_rate": 2.458010259958432e-09,
"loss": 0.8289,
"step": 679
},
{
"epoch": 2.789311408016444,
"grad_norm": 2.7477560997301174,
"learning_rate": 2.3610579436393e-09,
"loss": 0.7384,
"step": 680
},
{
"epoch": 2.7934224049331964,
"grad_norm": 2.791404195204494,
"learning_rate": 2.2660335218375295e-09,
"loss": 0.7649,
"step": 681
},
{
"epoch": 2.7975334018499485,
"grad_norm": 3.6681649953988327,
"learning_rate": 2.1729388708235487e-09,
"loss": 0.8459,
"step": 682
},
{
"epoch": 2.801644398766701,
"grad_norm": 2.6921912128783125,
"learning_rate": 2.081775828764254e-09,
"loss": 0.8259,
"step": 683
},
{
"epoch": 2.805755395683453,
"grad_norm": 3.391989217239566,
"learning_rate": 1.992546195686573e-09,
"loss": 0.821,
"step": 684
},
{
"epoch": 2.8098663926002057,
"grad_norm": 3.441211894452777,
"learning_rate": 1.9052517334420704e-09,
"loss": 0.8306,
"step": 685
},
{
"epoch": 2.8139773895169578,
"grad_norm": 3.972146357484525,
"learning_rate": 1.8198941656720645e-09,
"loss": 0.8597,
"step": 686
},
{
"epoch": 2.8180883864337103,
"grad_norm": 8.140870841074713,
"learning_rate": 1.736475177773633e-09,
"loss": 0.7867,
"step": 687
},
{
"epoch": 2.8221993833504624,
"grad_norm": 3.211799028443169,
"learning_rate": 1.6549964168663056e-09,
"loss": 0.8649,
"step": 688
},
{
"epoch": 2.826310380267215,
"grad_norm": 4.393260872097472,
"learning_rate": 1.5754594917595565e-09,
"loss": 0.8108,
"step": 689
},
{
"epoch": 2.830421377183967,
"grad_norm": 6.3597430610397,
"learning_rate": 1.4978659729210973e-09,
"loss": 0.9452,
"step": 690
},
{
"epoch": 2.8345323741007196,
"grad_norm": 3.488514923184495,
"learning_rate": 1.4222173924457349e-09,
"loss": 0.9172,
"step": 691
},
{
"epoch": 2.8386433710174717,
"grad_norm": 4.1426558033306256,
"learning_rate": 1.348515244025239e-09,
"loss": 0.8387,
"step": 692
},
{
"epoch": 2.842754367934224,
"grad_norm": 2.7871439777605422,
"learning_rate": 1.276760982918812e-09,
"loss": 0.7337,
"step": 693
},
{
"epoch": 2.8468653648509763,
"grad_norm": 2.484656601422799,
"learning_rate": 1.2069560259243327e-09,
"loss": 0.8153,
"step": 694
},
{
"epoch": 2.850976361767729,
"grad_norm": 7.363069212965746,
"learning_rate": 1.139101751350402e-09,
"loss": 0.9584,
"step": 695
},
{
"epoch": 2.855087358684481,
"grad_norm": 3.609856862436703,
"learning_rate": 1.0731994989891303e-09,
"loss": 0.7602,
"step": 696
},
{
"epoch": 2.859198355601233,
"grad_norm": 2.9639245106226486,
"learning_rate": 1.0092505700896704e-09,
"loss": 0.8467,
"step": 697
},
{
"epoch": 2.8633093525179856,
"grad_norm": 3.7204302459502117,
"learning_rate": 9.47256227332538e-10,
"loss": 0.8915,
"step": 698
},
{
"epoch": 2.867420349434738,
"grad_norm": 3.2380239528856936,
"learning_rate": 8.872176948046761e-10,
"loss": 0.8157,
"step": 699
},
{
"epoch": 2.87153134635149,
"grad_norm": 3.138285957539406,
"learning_rate": 8.291361579752631e-10,
"loss": 0.8503,
"step": 700
},
{
"epoch": 2.8756423432682423,
"grad_norm": 2.8626684047971502,
"learning_rate": 7.730127636723538e-10,
"loss": 0.8021,
"step": 701
},
{
"epoch": 2.879753340184995,
"grad_norm": 2.512856374229095,
"learning_rate": 7.188486200601751e-10,
"loss": 0.8225,
"step": 702
},
{
"epoch": 2.8838643371017474,
"grad_norm": 4.2783651150549815,
"learning_rate": 6.666447966172994e-10,
"loss": 0.9079,
"step": 703
},
{
"epoch": 2.8879753340184995,
"grad_norm": 4.896981732743199,
"learning_rate": 6.164023241154837e-10,
"loss": 0.8157,
"step": 704
},
{
"epoch": 2.8920863309352516,
"grad_norm": 4.883469998735079,
"learning_rate": 5.681221945993631e-10,
"loss": 0.9009,
"step": 705
},
{
"epoch": 2.896197327852004,
"grad_norm": 2.6595323338130017,
"learning_rate": 5.218053613668116e-10,
"loss": 0.7755,
"step": 706
},
{
"epoch": 2.9003083247687567,
"grad_norm": 3.5046044020047264,
"learning_rate": 4.774527389501682e-10,
"loss": 0.8365,
"step": 707
},
{
"epoch": 2.9044193216855088,
"grad_norm": 9.110076846985548,
"learning_rate": 4.3506520309813943e-10,
"loss": 0.8721,
"step": 708
},
{
"epoch": 2.908530318602261,
"grad_norm": 3.602979018567322,
"learning_rate": 3.946435907585255e-10,
"loss": 0.8837,
"step": 709
},
{
"epoch": 2.9126413155190134,
"grad_norm": 6.139265281175144,
"learning_rate": 3.561887000616881e-10,
"loss": 0.7613,
"step": 710
},
{
"epoch": 2.916752312435766,
"grad_norm": 3.265415149532557,
"learning_rate": 3.1970129030481905e-10,
"loss": 0.9008,
"step": 711
},
{
"epoch": 2.920863309352518,
"grad_norm": 3.3186659191410435,
"learning_rate": 2.8518208193689666e-10,
"loss": 0.7702,
"step": 712
},
{
"epoch": 2.92497430626927,
"grad_norm": 2.7682912491468197,
"learning_rate": 2.5263175654449685e-10,
"loss": 0.8004,
"step": 713
},
{
"epoch": 2.9290853031860227,
"grad_norm": 3.372557832698698,
"learning_rate": 2.2205095683833774e-10,
"loss": 0.8595,
"step": 714
},
{
"epoch": 2.9331963001027748,
"grad_norm": 3.5066168294870512,
"learning_rate": 1.9344028664056712e-10,
"loss": 0.8714,
"step": 715
},
{
"epoch": 2.9373072970195273,
"grad_norm": 4.172353240228205,
"learning_rate": 1.6680031087286106e-10,
"loss": 0.7844,
"step": 716
},
{
"epoch": 2.9414182939362794,
"grad_norm": 2.705807912981015,
"learning_rate": 1.4213155554525513e-10,
"loss": 0.795,
"step": 717
},
{
"epoch": 2.945529290853032,
"grad_norm": 10.17019453316457,
"learning_rate": 1.194345077457415e-10,
"loss": 0.8667,
"step": 718
},
{
"epoch": 2.949640287769784,
"grad_norm": 3.619800197834828,
"learning_rate": 9.8709615630721e-11,
"loss": 0.8294,
"step": 719
},
{
"epoch": 2.9537512846865366,
"grad_norm": 24.951408876544917,
"learning_rate": 7.995728841605486e-11,
"loss": 0.8178,
"step": 720
},
{
"epoch": 2.9578622816032887,
"grad_norm": 2.759507591055827,
"learning_rate": 6.3177896369071e-11,
"loss": 0.8172,
"step": 721
},
{
"epoch": 2.961973278520041,
"grad_norm": 4.253217027820731,
"learning_rate": 4.8371770801192146e-11,
"loss": 0.8004,
"step": 722
},
{
"epoch": 2.9660842754367933,
"grad_norm": 2.5583171524197503,
"learning_rate": 3.553920406144106e-11,
"loss": 0.8248,
"step": 723
},
{
"epoch": 2.970195272353546,
"grad_norm": 3.028860987474932,
"learning_rate": 2.4680449530622982e-11,
"loss": 0.8425,
"step": 724
},
{
"epoch": 2.974306269270298,
"grad_norm": 2.9287708287232497,
"learning_rate": 1.579572161637399e-11,
"loss": 0.8734,
"step": 725
},
{
"epoch": 2.9784172661870505,
"grad_norm": 3.6945258866601276,
"learning_rate": 8.885195748875585e-12,
"loss": 0.7734,
"step": 726
},
{
"epoch": 2.9825282631038026,
"grad_norm": 12.81977369580803,
"learning_rate": 3.9490083774240676e-12,
"loss": 0.8421,
"step": 727
},
{
"epoch": 2.986639260020555,
"grad_norm": 7.495618249033629,
"learning_rate": 9.872569677438215e-13,
"loss": 0.8617,
"step": 728
},
{
"epoch": 2.9907502569373072,
"grad_norm": 5.005440264811633,
"learning_rate": 0.0,
"loss": 0.8836,
"step": 729
}
],
"logging_steps": 1.0,
"max_steps": 729,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 729,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 438094649032704.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}