haoningwu's picture
Upload folder using huggingface_hub
2c11a33 verified
Raw
History Blame Contribute Delete
129 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 739,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013540961408259986,
"grad_norm": 58.75459324449355,
"learning_rate": 0.0,
"loss": 2.2042,
"step": 1
},
{
"epoch": 0.002708192281651997,
"grad_norm": 94.22912279043327,
"learning_rate": 4.347826086956522e-07,
"loss": 3.197,
"step": 2
},
{
"epoch": 0.004062288422477996,
"grad_norm": 65.72485559821217,
"learning_rate": 8.695652173913044e-07,
"loss": 2.5046,
"step": 3
},
{
"epoch": 0.005416384563303994,
"grad_norm": 51.99343130799949,
"learning_rate": 1.3043478260869566e-06,
"loss": 2.0259,
"step": 4
},
{
"epoch": 0.006770480704129994,
"grad_norm": 69.29867918648851,
"learning_rate": 1.7391304347826088e-06,
"loss": 2.4339,
"step": 5
},
{
"epoch": 0.008124576844955992,
"grad_norm": 54.59400957825116,
"learning_rate": 2.173913043478261e-06,
"loss": 1.8228,
"step": 6
},
{
"epoch": 0.009478672985781991,
"grad_norm": 45.00438853136862,
"learning_rate": 2.6086956521739132e-06,
"loss": 1.6494,
"step": 7
},
{
"epoch": 0.010832769126607989,
"grad_norm": 26.35830318559937,
"learning_rate": 3.043478260869566e-06,
"loss": 1.186,
"step": 8
},
{
"epoch": 0.012186865267433988,
"grad_norm": 27.341674365743902,
"learning_rate": 3.4782608695652175e-06,
"loss": 1.2635,
"step": 9
},
{
"epoch": 0.013540961408259987,
"grad_norm": 22.739524149963685,
"learning_rate": 3.91304347826087e-06,
"loss": 0.9027,
"step": 10
},
{
"epoch": 0.014895057549085985,
"grad_norm": 12.005297918322773,
"learning_rate": 4.347826086956522e-06,
"loss": 0.5808,
"step": 11
},
{
"epoch": 0.016249153689911984,
"grad_norm": 12.959063807909851,
"learning_rate": 4.782608695652174e-06,
"loss": 0.5787,
"step": 12
},
{
"epoch": 0.017603249830737983,
"grad_norm": 9.459524282042983,
"learning_rate": 5.2173913043478265e-06,
"loss": 0.4613,
"step": 13
},
{
"epoch": 0.018957345971563982,
"grad_norm": 9.324441123280721,
"learning_rate": 5.652173913043479e-06,
"loss": 0.5642,
"step": 14
},
{
"epoch": 0.020311442112389978,
"grad_norm": 8.77116211873085,
"learning_rate": 6.086956521739132e-06,
"loss": 0.543,
"step": 15
},
{
"epoch": 0.021665538253215978,
"grad_norm": 7.754905587455139,
"learning_rate": 6.521739130434783e-06,
"loss": 0.5051,
"step": 16
},
{
"epoch": 0.023019634394041977,
"grad_norm": 4.425304568033925,
"learning_rate": 6.956521739130435e-06,
"loss": 0.4089,
"step": 17
},
{
"epoch": 0.024373730534867976,
"grad_norm": 3.864807643294916,
"learning_rate": 7.391304347826087e-06,
"loss": 0.4284,
"step": 18
},
{
"epoch": 0.025727826675693975,
"grad_norm": 2.921422676713869,
"learning_rate": 7.82608695652174e-06,
"loss": 0.3232,
"step": 19
},
{
"epoch": 0.027081922816519974,
"grad_norm": 2.401849031514139,
"learning_rate": 8.260869565217392e-06,
"loss": 0.4365,
"step": 20
},
{
"epoch": 0.02843601895734597,
"grad_norm": 2.1731580563273822,
"learning_rate": 8.695652173913044e-06,
"loss": 0.3214,
"step": 21
},
{
"epoch": 0.02979011509817197,
"grad_norm": 3.2444865831166796,
"learning_rate": 9.130434782608697e-06,
"loss": 0.4032,
"step": 22
},
{
"epoch": 0.03114421123899797,
"grad_norm": 2.8135461309910297,
"learning_rate": 9.565217391304349e-06,
"loss": 0.3738,
"step": 23
},
{
"epoch": 0.03249830737982397,
"grad_norm": 1.9691389402557273,
"learning_rate": 1e-05,
"loss": 0.3164,
"step": 24
},
{
"epoch": 0.033852403520649964,
"grad_norm": 2.4633214348390333,
"learning_rate": 9.999951870315733e-06,
"loss": 0.3204,
"step": 25
},
{
"epoch": 0.035206499661475966,
"grad_norm": 1.8643763015434776,
"learning_rate": 9.999807482189519e-06,
"loss": 0.2867,
"step": 26
},
{
"epoch": 0.03656059580230196,
"grad_norm": 1.322958721056099,
"learning_rate": 9.999566838401099e-06,
"loss": 0.2484,
"step": 27
},
{
"epoch": 0.037914691943127965,
"grad_norm": 1.6645222918784135,
"learning_rate": 9.999229943583318e-06,
"loss": 0.3396,
"step": 28
},
{
"epoch": 0.03926878808395396,
"grad_norm": 1.5285989329971796,
"learning_rate": 9.99879680422203e-06,
"loss": 0.3029,
"step": 29
},
{
"epoch": 0.040622884224779957,
"grad_norm": 1.3131301707824983,
"learning_rate": 9.998267428655983e-06,
"loss": 0.3211,
"step": 30
},
{
"epoch": 0.04197698036560596,
"grad_norm": 1.5658380222131454,
"learning_rate": 9.997641827076645e-06,
"loss": 0.2779,
"step": 31
},
{
"epoch": 0.043331076506431955,
"grad_norm": 1.5701857701736117,
"learning_rate": 9.996920011528022e-06,
"loss": 0.3446,
"step": 32
},
{
"epoch": 0.04468517264725796,
"grad_norm": 1.705765841199752,
"learning_rate": 9.996101995906411e-06,
"loss": 0.263,
"step": 33
},
{
"epoch": 0.046039268788083954,
"grad_norm": 1.420065009706357,
"learning_rate": 9.995187795960152e-06,
"loss": 0.2998,
"step": 34
},
{
"epoch": 0.04739336492890995,
"grad_norm": 1.2846639754720788,
"learning_rate": 9.994177429289302e-06,
"loss": 0.235,
"step": 35
},
{
"epoch": 0.04874746106973595,
"grad_norm": 1.1768495485407515,
"learning_rate": 9.993070915345313e-06,
"loss": 0.2709,
"step": 36
},
{
"epoch": 0.05010155721056195,
"grad_norm": 0.9432569476682318,
"learning_rate": 9.991868275430653e-06,
"loss": 0.2419,
"step": 37
},
{
"epoch": 0.05145565335138795,
"grad_norm": 1.3363246413476688,
"learning_rate": 9.990569532698393e-06,
"loss": 0.2833,
"step": 38
},
{
"epoch": 0.052809749492213946,
"grad_norm": 0.992899612944998,
"learning_rate": 9.989174712151766e-06,
"loss": 0.2698,
"step": 39
},
{
"epoch": 0.05416384563303995,
"grad_norm": 1.5040236358854777,
"learning_rate": 9.987683840643679e-06,
"loss": 0.3216,
"step": 40
},
{
"epoch": 0.055517941773865945,
"grad_norm": 1.0805017840834301,
"learning_rate": 9.9860969468762e-06,
"loss": 0.2789,
"step": 41
},
{
"epoch": 0.05687203791469194,
"grad_norm": 1.0478619610186182,
"learning_rate": 9.984414061400011e-06,
"loss": 0.2631,
"step": 42
},
{
"epoch": 0.05822613405551794,
"grad_norm": 1.5143408134324512,
"learning_rate": 9.982635216613812e-06,
"loss": 0.2255,
"step": 43
},
{
"epoch": 0.05958023019634394,
"grad_norm": 0.9084082196669798,
"learning_rate": 9.980760446763693e-06,
"loss": 0.2396,
"step": 44
},
{
"epoch": 0.06093432633716994,
"grad_norm": 1.461957476582148,
"learning_rate": 9.978789787942489e-06,
"loss": 0.2728,
"step": 45
},
{
"epoch": 0.06228842247799594,
"grad_norm": 0.9785095117004499,
"learning_rate": 9.976723278089075e-06,
"loss": 0.2481,
"step": 46
},
{
"epoch": 0.06364251861882193,
"grad_norm": 1.180230494885314,
"learning_rate": 9.974560956987637e-06,
"loss": 0.2365,
"step": 47
},
{
"epoch": 0.06499661475964794,
"grad_norm": 1.2765663421283717,
"learning_rate": 9.97230286626691e-06,
"loss": 0.2748,
"step": 48
},
{
"epoch": 0.06635071090047394,
"grad_norm": 1.1010775499004417,
"learning_rate": 9.969949049399368e-06,
"loss": 0.2618,
"step": 49
},
{
"epoch": 0.06770480704129993,
"grad_norm": 1.1732237911684615,
"learning_rate": 9.967499551700398e-06,
"loss": 0.1844,
"step": 50
},
{
"epoch": 0.06905890318212593,
"grad_norm": 1.1194659690438453,
"learning_rate": 9.96495442032742e-06,
"loss": 0.2171,
"step": 51
},
{
"epoch": 0.07041299932295193,
"grad_norm": 1.0236299756973815,
"learning_rate": 9.962313704278981e-06,
"loss": 0.2353,
"step": 52
},
{
"epoch": 0.07176709546377792,
"grad_norm": 1.2744862550898641,
"learning_rate": 9.959577454393815e-06,
"loss": 0.2501,
"step": 53
},
{
"epoch": 0.07312119160460392,
"grad_norm": 1.1254396952923758,
"learning_rate": 9.956745723349857e-06,
"loss": 0.2578,
"step": 54
},
{
"epoch": 0.07447528774542993,
"grad_norm": 1.1926581392388238,
"learning_rate": 9.953818565663238e-06,
"loss": 0.2865,
"step": 55
},
{
"epoch": 0.07582938388625593,
"grad_norm": 0.767478906113829,
"learning_rate": 9.950796037687224e-06,
"loss": 0.1897,
"step": 56
},
{
"epoch": 0.07718348002708192,
"grad_norm": 1.5014624602044033,
"learning_rate": 9.947678197611145e-06,
"loss": 0.2713,
"step": 57
},
{
"epoch": 0.07853757616790792,
"grad_norm": 1.1195821932762393,
"learning_rate": 9.944465105459265e-06,
"loss": 0.2652,
"step": 58
},
{
"epoch": 0.07989167230873392,
"grad_norm": 0.9661794451732615,
"learning_rate": 9.941156823089623e-06,
"loss": 0.2055,
"step": 59
},
{
"epoch": 0.08124576844955991,
"grad_norm": 1.048803546094244,
"learning_rate": 9.937753414192862e-06,
"loss": 0.2401,
"step": 60
},
{
"epoch": 0.08259986459038592,
"grad_norm": 1.0034754783751856,
"learning_rate": 9.934254944290973e-06,
"loss": 0.2243,
"step": 61
},
{
"epoch": 0.08395396073121192,
"grad_norm": 0.8344114565045573,
"learning_rate": 9.93066148073606e-06,
"loss": 0.221,
"step": 62
},
{
"epoch": 0.08530805687203792,
"grad_norm": 1.1450289169392145,
"learning_rate": 9.926973092709028e-06,
"loss": 0.2605,
"step": 63
},
{
"epoch": 0.08666215301286391,
"grad_norm": 1.0255599694668958,
"learning_rate": 9.923189851218259e-06,
"loss": 0.2399,
"step": 64
},
{
"epoch": 0.08801624915368991,
"grad_norm": 1.0068478592486565,
"learning_rate": 9.91931182909824e-06,
"loss": 0.2191,
"step": 65
},
{
"epoch": 0.08937034529451592,
"grad_norm": 0.7608680193843377,
"learning_rate": 9.915339101008162e-06,
"loss": 0.2296,
"step": 66
},
{
"epoch": 0.0907244414353419,
"grad_norm": 0.940854255896407,
"learning_rate": 9.911271743430486e-06,
"loss": 0.2312,
"step": 67
},
{
"epoch": 0.09207853757616791,
"grad_norm": 0.9244076934941405,
"learning_rate": 9.907109834669465e-06,
"loss": 0.2682,
"step": 68
},
{
"epoch": 0.09343263371699391,
"grad_norm": 0.821804771536457,
"learning_rate": 9.902853454849641e-06,
"loss": 0.2348,
"step": 69
},
{
"epoch": 0.0947867298578199,
"grad_norm": 0.8680451643351618,
"learning_rate": 9.898502685914301e-06,
"loss": 0.1197,
"step": 70
},
{
"epoch": 0.0961408259986459,
"grad_norm": 0.9994956759958264,
"learning_rate": 9.8940576116239e-06,
"loss": 0.2304,
"step": 71
},
{
"epoch": 0.0974949221394719,
"grad_norm": 1.4541767351892134,
"learning_rate": 9.889518317554446e-06,
"loss": 0.182,
"step": 72
},
{
"epoch": 0.0988490182802979,
"grad_norm": 1.2921251364583113,
"learning_rate": 9.884884891095856e-06,
"loss": 0.2223,
"step": 73
},
{
"epoch": 0.1002031144211239,
"grad_norm": 0.9332864413336698,
"learning_rate": 9.880157421450269e-06,
"loss": 0.1786,
"step": 74
},
{
"epoch": 0.1015572105619499,
"grad_norm": 1.3454577550906328,
"learning_rate": 9.875335999630335e-06,
"loss": 0.2029,
"step": 75
},
{
"epoch": 0.1029113067027759,
"grad_norm": 1.176662037418281,
"learning_rate": 9.870420718457458e-06,
"loss": 0.2602,
"step": 76
},
{
"epoch": 0.10426540284360189,
"grad_norm": 0.8553980297846864,
"learning_rate": 9.865411672560012e-06,
"loss": 0.1509,
"step": 77
},
{
"epoch": 0.10561949898442789,
"grad_norm": 1.1273861441350155,
"learning_rate": 9.860308958371511e-06,
"loss": 0.2474,
"step": 78
},
{
"epoch": 0.1069735951252539,
"grad_norm": 1.0482442634564217,
"learning_rate": 9.85511267412877e-06,
"loss": 0.252,
"step": 79
},
{
"epoch": 0.1083276912660799,
"grad_norm": 1.0810809817775093,
"learning_rate": 9.849822919869993e-06,
"loss": 0.2632,
"step": 80
},
{
"epoch": 0.10968178740690589,
"grad_norm": 1.1748724609500094,
"learning_rate": 9.844439797432863e-06,
"loss": 0.2531,
"step": 81
},
{
"epoch": 0.11103588354773189,
"grad_norm": 1.0244575328596122,
"learning_rate": 9.838963410452572e-06,
"loss": 0.1986,
"step": 82
},
{
"epoch": 0.11238997968855789,
"grad_norm": 1.3509903275828423,
"learning_rate": 9.833393864359832e-06,
"loss": 0.1916,
"step": 83
},
{
"epoch": 0.11374407582938388,
"grad_norm": 1.092912573226465,
"learning_rate": 9.827731266378839e-06,
"loss": 0.2166,
"step": 84
},
{
"epoch": 0.11509817197020988,
"grad_norm": 0.8737686118883372,
"learning_rate": 9.821975725525214e-06,
"loss": 0.1849,
"step": 85
},
{
"epoch": 0.11645226811103589,
"grad_norm": 0.8588608179296646,
"learning_rate": 9.816127352603907e-06,
"loss": 0.2124,
"step": 86
},
{
"epoch": 0.11780636425186188,
"grad_norm": 1.0101312434872192,
"learning_rate": 9.810186260207049e-06,
"loss": 0.2312,
"step": 87
},
{
"epoch": 0.11916046039268788,
"grad_norm": 0.927357041776064,
"learning_rate": 9.804152562711804e-06,
"loss": 0.2467,
"step": 88
},
{
"epoch": 0.12051455653351388,
"grad_norm": 1.1213558139288597,
"learning_rate": 9.798026376278154e-06,
"loss": 0.1274,
"step": 89
},
{
"epoch": 0.12186865267433988,
"grad_norm": 0.9239992217843024,
"learning_rate": 9.791807818846665e-06,
"loss": 0.19,
"step": 90
},
{
"epoch": 0.12322274881516587,
"grad_norm": 1.2284079929947043,
"learning_rate": 9.78549701013622e-06,
"loss": 0.2428,
"step": 91
},
{
"epoch": 0.12457684495599188,
"grad_norm": 0.9879883907634985,
"learning_rate": 9.779094071641712e-06,
"loss": 0.1994,
"step": 92
},
{
"epoch": 0.12593094109681788,
"grad_norm": 1.2290184677087364,
"learning_rate": 9.772599126631704e-06,
"loss": 0.2489,
"step": 93
},
{
"epoch": 0.12728503723764387,
"grad_norm": 1.1916290717292688,
"learning_rate": 9.766012300146058e-06,
"loss": 0.2153,
"step": 94
},
{
"epoch": 0.12863913337846988,
"grad_norm": 1.2236322538337256,
"learning_rate": 9.759333718993523e-06,
"loss": 0.1991,
"step": 95
},
{
"epoch": 0.12999322951929587,
"grad_norm": 1.168201726456461,
"learning_rate": 9.752563511749301e-06,
"loss": 0.2061,
"step": 96
},
{
"epoch": 0.13134732566012186,
"grad_norm": 1.0930895786336507,
"learning_rate": 9.745701808752567e-06,
"loss": 0.1987,
"step": 97
},
{
"epoch": 0.13270142180094788,
"grad_norm": 1.0550881111573123,
"learning_rate": 9.738748742103961e-06,
"loss": 0.2089,
"step": 98
},
{
"epoch": 0.13405551794177387,
"grad_norm": 1.053208397756083,
"learning_rate": 9.731704445663044e-06,
"loss": 0.1808,
"step": 99
},
{
"epoch": 0.13540961408259986,
"grad_norm": 1.0437742742280756,
"learning_rate": 9.724569055045722e-06,
"loss": 0.1938,
"step": 100
},
{
"epoch": 0.13676371022342587,
"grad_norm": 0.8675035188305833,
"learning_rate": 9.717342707621633e-06,
"loss": 0.2356,
"step": 101
},
{
"epoch": 0.13811780636425186,
"grad_norm": 0.9882014823227931,
"learning_rate": 9.710025542511503e-06,
"loss": 0.1701,
"step": 102
},
{
"epoch": 0.13947190250507785,
"grad_norm": 0.7120751752819733,
"learning_rate": 9.702617700584475e-06,
"loss": 0.1778,
"step": 103
},
{
"epoch": 0.14082599864590387,
"grad_norm": 1.022363650123575,
"learning_rate": 9.695119324455383e-06,
"loss": 0.229,
"step": 104
},
{
"epoch": 0.14218009478672985,
"grad_norm": 1.2710855942816435,
"learning_rate": 9.687530558482016e-06,
"loss": 0.2448,
"step": 105
},
{
"epoch": 0.14353419092755584,
"grad_norm": 1.3972517693315574,
"learning_rate": 9.679851548762344e-06,
"loss": 0.2267,
"step": 106
},
{
"epoch": 0.14488828706838186,
"grad_norm": 0.9515942767685095,
"learning_rate": 9.672082443131687e-06,
"loss": 0.2011,
"step": 107
},
{
"epoch": 0.14624238320920785,
"grad_norm": 0.9260581673007142,
"learning_rate": 9.664223391159885e-06,
"loss": 0.1701,
"step": 108
},
{
"epoch": 0.14759647935003387,
"grad_norm": 1.0426375277169637,
"learning_rate": 9.656274544148419e-06,
"loss": 0.1977,
"step": 109
},
{
"epoch": 0.14895057549085985,
"grad_norm": 0.9378857012288075,
"learning_rate": 9.648236055127485e-06,
"loss": 0.1984,
"step": 110
},
{
"epoch": 0.15030467163168584,
"grad_norm": 1.1050430959912767,
"learning_rate": 9.640108078853056e-06,
"loss": 0.2436,
"step": 111
},
{
"epoch": 0.15165876777251186,
"grad_norm": 1.0263090733524953,
"learning_rate": 9.631890771803909e-06,
"loss": 0.1591,
"step": 112
},
{
"epoch": 0.15301286391333785,
"grad_norm": 0.9939978817405343,
"learning_rate": 9.623584292178598e-06,
"loss": 0.1964,
"step": 113
},
{
"epoch": 0.15436696005416384,
"grad_norm": 0.9735197738420288,
"learning_rate": 9.61518879989242e-06,
"loss": 0.1511,
"step": 114
},
{
"epoch": 0.15572105619498985,
"grad_norm": 1.3423630153145567,
"learning_rate": 9.606704456574335e-06,
"loss": 0.162,
"step": 115
},
{
"epoch": 0.15707515233581584,
"grad_norm": 1.0904085060296527,
"learning_rate": 9.598131425563847e-06,
"loss": 0.1748,
"step": 116
},
{
"epoch": 0.15842924847664183,
"grad_norm": 1.1011865365338906,
"learning_rate": 9.589469871907868e-06,
"loss": 0.2672,
"step": 117
},
{
"epoch": 0.15978334461746785,
"grad_norm": 1.1035518932847574,
"learning_rate": 9.580719962357532e-06,
"loss": 0.1632,
"step": 118
},
{
"epoch": 0.16113744075829384,
"grad_norm": 0.8799291363501105,
"learning_rate": 9.571881865364994e-06,
"loss": 0.2087,
"step": 119
},
{
"epoch": 0.16249153689911983,
"grad_norm": 1.339799499848271,
"learning_rate": 9.562955751080183e-06,
"loss": 0.2999,
"step": 120
},
{
"epoch": 0.16384563303994584,
"grad_norm": 0.9095336124045952,
"learning_rate": 9.553941791347521e-06,
"loss": 0.1641,
"step": 121
},
{
"epoch": 0.16519972918077183,
"grad_norm": 0.9453021772977844,
"learning_rate": 9.544840159702626e-06,
"loss": 0.2139,
"step": 122
},
{
"epoch": 0.16655382532159782,
"grad_norm": 1.0259601500595346,
"learning_rate": 9.535651031368957e-06,
"loss": 0.2375,
"step": 123
},
{
"epoch": 0.16790792146242384,
"grad_norm": 1.1534944073504834,
"learning_rate": 9.526374583254454e-06,
"loss": 0.2295,
"step": 124
},
{
"epoch": 0.16926201760324983,
"grad_norm": 0.8938413776954387,
"learning_rate": 9.517010993948125e-06,
"loss": 0.1698,
"step": 125
},
{
"epoch": 0.17061611374407584,
"grad_norm": 0.976906813471617,
"learning_rate": 9.507560443716608e-06,
"loss": 0.1735,
"step": 126
},
{
"epoch": 0.17197020988490183,
"grad_norm": 0.9685431449341421,
"learning_rate": 9.498023114500703e-06,
"loss": 0.2085,
"step": 127
},
{
"epoch": 0.17332430602572782,
"grad_norm": 0.9959204101378731,
"learning_rate": 9.488399189911866e-06,
"loss": 0.1912,
"step": 128
},
{
"epoch": 0.17467840216655384,
"grad_norm": 1.034136062151613,
"learning_rate": 9.47868885522868e-06,
"loss": 0.1626,
"step": 129
},
{
"epoch": 0.17603249830737983,
"grad_norm": 0.7691856191665672,
"learning_rate": 9.468892297393281e-06,
"loss": 0.1761,
"step": 130
},
{
"epoch": 0.17738659444820581,
"grad_norm": 0.8776384463846831,
"learning_rate": 9.459009705007763e-06,
"loss": 0.1937,
"step": 131
},
{
"epoch": 0.17874069058903183,
"grad_norm": 0.9118755090140794,
"learning_rate": 9.449041268330549e-06,
"loss": 0.1679,
"step": 132
},
{
"epoch": 0.18009478672985782,
"grad_norm": 1.154475718929298,
"learning_rate": 9.438987179272716e-06,
"loss": 0.2637,
"step": 133
},
{
"epoch": 0.1814488828706838,
"grad_norm": 1.0042096648891632,
"learning_rate": 9.428847631394324e-06,
"loss": 0.197,
"step": 134
},
{
"epoch": 0.18280297901150983,
"grad_norm": 1.1633877283787264,
"learning_rate": 9.418622819900666e-06,
"loss": 0.1915,
"step": 135
},
{
"epoch": 0.18415707515233581,
"grad_norm": 0.8789749091396356,
"learning_rate": 9.408312941638522e-06,
"loss": 0.1996,
"step": 136
},
{
"epoch": 0.1855111712931618,
"grad_norm": 0.9572325603942851,
"learning_rate": 9.397918195092363e-06,
"loss": 0.2093,
"step": 137
},
{
"epoch": 0.18686526743398782,
"grad_norm": 1.0036010820156096,
"learning_rate": 9.387438780380541e-06,
"loss": 0.1874,
"step": 138
},
{
"epoch": 0.1882193635748138,
"grad_norm": 0.8483358547594898,
"learning_rate": 9.376874899251423e-06,
"loss": 0.1832,
"step": 139
},
{
"epoch": 0.1895734597156398,
"grad_norm": 0.8814475602494724,
"learning_rate": 9.366226755079513e-06,
"loss": 0.1893,
"step": 140
},
{
"epoch": 0.1909275558564658,
"grad_norm": 0.8984206487803382,
"learning_rate": 9.35549455286154e-06,
"loss": 0.2376,
"step": 141
},
{
"epoch": 0.1922816519972918,
"grad_norm": 1.2602645227178595,
"learning_rate": 9.344678499212503e-06,
"loss": 0.1846,
"step": 142
},
{
"epoch": 0.19363574813811782,
"grad_norm": 1.0754028623283816,
"learning_rate": 9.333778802361702e-06,
"loss": 0.208,
"step": 143
},
{
"epoch": 0.1949898442789438,
"grad_norm": 1.2494706287979813,
"learning_rate": 9.322795672148726e-06,
"loss": 0.1875,
"step": 144
},
{
"epoch": 0.1963439404197698,
"grad_norm": 1.1174298813034655,
"learning_rate": 9.311729320019409e-06,
"loss": 0.1603,
"step": 145
},
{
"epoch": 0.1976980365605958,
"grad_norm": 1.0516070554506416,
"learning_rate": 9.300579959021766e-06,
"loss": 0.1797,
"step": 146
},
{
"epoch": 0.1990521327014218,
"grad_norm": 0.7777435565032388,
"learning_rate": 9.289347803801884e-06,
"loss": 0.1582,
"step": 147
},
{
"epoch": 0.2004062288422478,
"grad_norm": 0.9150038459987193,
"learning_rate": 9.2780330705998e-06,
"loss": 0.1422,
"step": 148
},
{
"epoch": 0.2017603249830738,
"grad_norm": 0.9763500410727148,
"learning_rate": 9.266635977245328e-06,
"loss": 0.2153,
"step": 149
},
{
"epoch": 0.2031144211238998,
"grad_norm": 0.9140601904245541,
"learning_rate": 9.255156743153867e-06,
"loss": 0.1434,
"step": 150
},
{
"epoch": 0.20446851726472579,
"grad_norm": 0.9610037864274474,
"learning_rate": 9.243595589322186e-06,
"loss": 0.1822,
"step": 151
},
{
"epoch": 0.2058226134055518,
"grad_norm": 1.1478861873798138,
"learning_rate": 9.231952738324155e-06,
"loss": 0.1643,
"step": 152
},
{
"epoch": 0.2071767095463778,
"grad_norm": 0.865018836238558,
"learning_rate": 9.220228414306473e-06,
"loss": 0.1893,
"step": 153
},
{
"epoch": 0.20853080568720378,
"grad_norm": 1.1216257094499948,
"learning_rate": 9.208422842984345e-06,
"loss": 0.2286,
"step": 154
},
{
"epoch": 0.2098849018280298,
"grad_norm": 0.9111389732012027,
"learning_rate": 9.196536251637138e-06,
"loss": 0.1813,
"step": 155
},
{
"epoch": 0.21123899796885579,
"grad_norm": 0.8607864872521915,
"learning_rate": 9.18456886910401e-06,
"loss": 0.1732,
"step": 156
},
{
"epoch": 0.21259309410968177,
"grad_norm": 1.0271739500313661,
"learning_rate": 9.172520925779495e-06,
"loss": 0.2313,
"step": 157
},
{
"epoch": 0.2139471902505078,
"grad_norm": 0.9483899365554289,
"learning_rate": 9.16039265360908e-06,
"loss": 0.219,
"step": 158
},
{
"epoch": 0.21530128639133378,
"grad_norm": 0.7105580429924478,
"learning_rate": 9.148184286084725e-06,
"loss": 0.1729,
"step": 159
},
{
"epoch": 0.2166553825321598,
"grad_norm": 0.821239929837702,
"learning_rate": 9.135896058240384e-06,
"loss": 0.2121,
"step": 160
},
{
"epoch": 0.21800947867298578,
"grad_norm": 0.830497460866812,
"learning_rate": 9.123528206647464e-06,
"loss": 0.1662,
"step": 161
},
{
"epoch": 0.21936357481381177,
"grad_norm": 0.79208744987268,
"learning_rate": 9.111080969410282e-06,
"loss": 0.171,
"step": 162
},
{
"epoch": 0.2207176709546378,
"grad_norm": 0.9062547941929919,
"learning_rate": 9.098554586161479e-06,
"loss": 0.1725,
"step": 163
},
{
"epoch": 0.22207176709546378,
"grad_norm": 0.8940620101622849,
"learning_rate": 9.085949298057402e-06,
"loss": 0.1506,
"step": 164
},
{
"epoch": 0.22342586323628977,
"grad_norm": 0.9641149069998645,
"learning_rate": 9.073265347773467e-06,
"loss": 0.2045,
"step": 165
},
{
"epoch": 0.22477995937711578,
"grad_norm": 0.9826087106032385,
"learning_rate": 9.060502979499484e-06,
"loss": 0.2055,
"step": 166
},
{
"epoch": 0.22613405551794177,
"grad_norm": 1.0667338061689005,
"learning_rate": 9.047662438934951e-06,
"loss": 0.2091,
"step": 167
},
{
"epoch": 0.22748815165876776,
"grad_norm": 1.0029199137600147,
"learning_rate": 9.034743973284337e-06,
"loss": 0.1896,
"step": 168
},
{
"epoch": 0.22884224779959378,
"grad_norm": 0.9433703092856318,
"learning_rate": 9.02174783125231e-06,
"loss": 0.1742,
"step": 169
},
{
"epoch": 0.23019634394041977,
"grad_norm": 0.8339090612516012,
"learning_rate": 9.008674263038954e-06,
"loss": 0.2025,
"step": 170
},
{
"epoch": 0.23155044008124576,
"grad_norm": 1.1566636515576076,
"learning_rate": 8.995523520334955e-06,
"loss": 0.1397,
"step": 171
},
{
"epoch": 0.23290453622207177,
"grad_norm": 0.8912412273863639,
"learning_rate": 8.98229585631675e-06,
"loss": 0.2204,
"step": 172
},
{
"epoch": 0.23425863236289776,
"grad_norm": 0.7876655957681313,
"learning_rate": 8.968991525641656e-06,
"loss": 0.1355,
"step": 173
},
{
"epoch": 0.23561272850372375,
"grad_norm": 0.895254404794835,
"learning_rate": 8.955610784442967e-06,
"loss": 0.2152,
"step": 174
},
{
"epoch": 0.23696682464454977,
"grad_norm": 0.972221207475063,
"learning_rate": 8.942153890325023e-06,
"loss": 0.18,
"step": 175
},
{
"epoch": 0.23832092078537576,
"grad_norm": 1.3432280871101725,
"learning_rate": 8.928621102358248e-06,
"loss": 0.2153,
"step": 176
},
{
"epoch": 0.23967501692620177,
"grad_norm": 1.0756319494574786,
"learning_rate": 8.91501268107417e-06,
"loss": 0.1761,
"step": 177
},
{
"epoch": 0.24102911306702776,
"grad_norm": 0.9434522004041053,
"learning_rate": 8.901328888460394e-06,
"loss": 0.1454,
"step": 178
},
{
"epoch": 0.24238320920785375,
"grad_norm": 0.907841413603741,
"learning_rate": 8.887569987955572e-06,
"loss": 0.1808,
"step": 179
},
{
"epoch": 0.24373730534867977,
"grad_norm": 0.8041723496745592,
"learning_rate": 8.873736244444311e-06,
"loss": 0.1562,
"step": 180
},
{
"epoch": 0.24509140148950576,
"grad_norm": 0.8870781334889946,
"learning_rate": 8.8598279242521e-06,
"loss": 0.2273,
"step": 181
},
{
"epoch": 0.24644549763033174,
"grad_norm": 0.9446919815152058,
"learning_rate": 8.845845295140163e-06,
"loss": 0.2153,
"step": 182
},
{
"epoch": 0.24779959377115776,
"grad_norm": 1.1720978022902202,
"learning_rate": 8.831788626300306e-06,
"loss": 0.1878,
"step": 183
},
{
"epoch": 0.24915368991198375,
"grad_norm": 1.1483691522298392,
"learning_rate": 8.817658188349745e-06,
"loss": 0.2075,
"step": 184
},
{
"epoch": 0.25050778605280977,
"grad_norm": 1.2425142755743255,
"learning_rate": 8.803454253325885e-06,
"loss": 0.1934,
"step": 185
},
{
"epoch": 0.25186188219363576,
"grad_norm": 1.4983624161516467,
"learning_rate": 8.789177094681091e-06,
"loss": 0.204,
"step": 186
},
{
"epoch": 0.25321597833446174,
"grad_norm": 1.1404994869762957,
"learning_rate": 8.774826987277418e-06,
"loss": 0.1981,
"step": 187
},
{
"epoch": 0.25457007447528773,
"grad_norm": 0.9276473640561954,
"learning_rate": 8.76040420738132e-06,
"loss": 0.1729,
"step": 188
},
{
"epoch": 0.2559241706161137,
"grad_norm": 1.0167294916404863,
"learning_rate": 8.745909032658338e-06,
"loss": 0.1726,
"step": 189
},
{
"epoch": 0.25727826675693977,
"grad_norm": 0.9556133613938518,
"learning_rate": 8.731341742167739e-06,
"loss": 0.1599,
"step": 190
},
{
"epoch": 0.25863236289776576,
"grad_norm": 1.0707254403056727,
"learning_rate": 8.716702616357163e-06,
"loss": 0.2124,
"step": 191
},
{
"epoch": 0.25998645903859174,
"grad_norm": 0.8660742749708209,
"learning_rate": 8.701991937057211e-06,
"loss": 0.1734,
"step": 192
},
{
"epoch": 0.26134055517941773,
"grad_norm": 1.091624593932146,
"learning_rate": 8.687209987476023e-06,
"loss": 0.1935,
"step": 193
},
{
"epoch": 0.2626946513202437,
"grad_norm": 0.9567638947510108,
"learning_rate": 8.672357052193826e-06,
"loss": 0.2246,
"step": 194
},
{
"epoch": 0.2640487474610697,
"grad_norm": 0.8538639528585871,
"learning_rate": 8.657433417157455e-06,
"loss": 0.1719,
"step": 195
},
{
"epoch": 0.26540284360189575,
"grad_norm": 0.8348742120140251,
"learning_rate": 8.642439369674845e-06,
"loss": 0.1808,
"step": 196
},
{
"epoch": 0.26675693974272174,
"grad_norm": 0.9506526907199666,
"learning_rate": 8.627375198409504e-06,
"loss": 0.1133,
"step": 197
},
{
"epoch": 0.26811103588354773,
"grad_norm": 1.0025252148263613,
"learning_rate": 8.612241193374958e-06,
"loss": 0.1884,
"step": 198
},
{
"epoch": 0.2694651320243737,
"grad_norm": 1.0145341568269548,
"learning_rate": 8.597037645929155e-06,
"loss": 0.2421,
"step": 199
},
{
"epoch": 0.2708192281651997,
"grad_norm": 0.9097200647294701,
"learning_rate": 8.581764848768878e-06,
"loss": 0.1738,
"step": 200
},
{
"epoch": 0.27217332430602575,
"grad_norm": 1.09132067294482,
"learning_rate": 8.566423095924084e-06,
"loss": 0.1384,
"step": 201
},
{
"epoch": 0.27352742044685174,
"grad_norm": 0.9664210000994968,
"learning_rate": 8.551012682752262e-06,
"loss": 0.2122,
"step": 202
},
{
"epoch": 0.27488151658767773,
"grad_norm": 1.6619821129300765,
"learning_rate": 8.535533905932739e-06,
"loss": 0.1479,
"step": 203
},
{
"epoch": 0.2762356127285037,
"grad_norm": 0.8646981883906724,
"learning_rate": 8.519987063460973e-06,
"loss": 0.1268,
"step": 204
},
{
"epoch": 0.2775897088693297,
"grad_norm": 0.8475054046630563,
"learning_rate": 8.504372454642813e-06,
"loss": 0.1899,
"step": 205
},
{
"epoch": 0.2789438050101557,
"grad_norm": 1.0092938733068637,
"learning_rate": 8.488690380088733e-06,
"loss": 0.1667,
"step": 206
},
{
"epoch": 0.28029790115098174,
"grad_norm": 0.9262032116715689,
"learning_rate": 8.472941141708055e-06,
"loss": 0.2281,
"step": 207
},
{
"epoch": 0.28165199729180773,
"grad_norm": 1.3040750226891595,
"learning_rate": 8.457125042703124e-06,
"loss": 0.2357,
"step": 208
},
{
"epoch": 0.2830060934326337,
"grad_norm": 0.8851084535470077,
"learning_rate": 8.441242387563484e-06,
"loss": 0.1407,
"step": 209
},
{
"epoch": 0.2843601895734597,
"grad_norm": 1.0177068536946232,
"learning_rate": 8.425293482060002e-06,
"loss": 0.2058,
"step": 210
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.6175630890536948,
"learning_rate": 8.409278633238996e-06,
"loss": 0.1036,
"step": 211
},
{
"epoch": 0.2870683818551117,
"grad_norm": 0.85876871183371,
"learning_rate": 8.393198149416311e-06,
"loss": 0.1458,
"step": 212
},
{
"epoch": 0.28842247799593773,
"grad_norm": 0.8298097096659509,
"learning_rate": 8.377052340171391e-06,
"loss": 0.1649,
"step": 213
},
{
"epoch": 0.2897765741367637,
"grad_norm": 0.7744205117359284,
"learning_rate": 8.360841516341319e-06,
"loss": 0.1871,
"step": 214
},
{
"epoch": 0.2911306702775897,
"grad_norm": 0.7971622822754814,
"learning_rate": 8.344565990014823e-06,
"loss": 0.1766,
"step": 215
},
{
"epoch": 0.2924847664184157,
"grad_norm": 0.898641724790306,
"learning_rate": 8.328226074526284e-06,
"loss": 0.1601,
"step": 216
},
{
"epoch": 0.2938388625592417,
"grad_norm": 0.9363104121305316,
"learning_rate": 8.31182208444969e-06,
"loss": 0.1805,
"step": 217
},
{
"epoch": 0.29519295870006773,
"grad_norm": 1.0671382277757204,
"learning_rate": 8.295354335592588e-06,
"loss": 0.1358,
"step": 218
},
{
"epoch": 0.2965470548408937,
"grad_norm": 0.8097781049952084,
"learning_rate": 8.278823144989998e-06,
"loss": 0.1601,
"step": 219
},
{
"epoch": 0.2979011509817197,
"grad_norm": 0.9201468007946983,
"learning_rate": 8.262228830898313e-06,
"loss": 0.1631,
"step": 220
},
{
"epoch": 0.2992552471225457,
"grad_norm": 0.9844751974893571,
"learning_rate": 8.245571712789172e-06,
"loss": 0.2207,
"step": 221
},
{
"epoch": 0.3006093432633717,
"grad_norm": 0.9173780834498073,
"learning_rate": 8.22885211134331e-06,
"loss": 0.2173,
"step": 222
},
{
"epoch": 0.3019634394041977,
"grad_norm": 0.7994027807629596,
"learning_rate": 8.212070348444382e-06,
"loss": 0.171,
"step": 223
},
{
"epoch": 0.3033175355450237,
"grad_norm": 0.8789502143454573,
"learning_rate": 8.19522674717277e-06,
"loss": 0.1747,
"step": 224
},
{
"epoch": 0.3046716316858497,
"grad_norm": 0.9981490737218109,
"learning_rate": 8.178321631799358e-06,
"loss": 0.238,
"step": 225
},
{
"epoch": 0.3060257278266757,
"grad_norm": 1.0249216907747705,
"learning_rate": 8.161355327779289e-06,
"loss": 0.1446,
"step": 226
},
{
"epoch": 0.3073798239675017,
"grad_norm": 0.8769945528949843,
"learning_rate": 8.144328161745708e-06,
"loss": 0.1596,
"step": 227
},
{
"epoch": 0.3087339201083277,
"grad_norm": 0.8328103559380179,
"learning_rate": 8.127240461503462e-06,
"loss": 0.1407,
"step": 228
},
{
"epoch": 0.31008801624915366,
"grad_norm": 0.788222773346233,
"learning_rate": 8.110092556022803e-06,
"loss": 0.1671,
"step": 229
},
{
"epoch": 0.3114421123899797,
"grad_norm": 0.9812041646222222,
"learning_rate": 8.092884775433037e-06,
"loss": 0.2449,
"step": 230
},
{
"epoch": 0.3127962085308057,
"grad_norm": 0.9060762242185922,
"learning_rate": 8.075617451016187e-06,
"loss": 0.1481,
"step": 231
},
{
"epoch": 0.3141503046716317,
"grad_norm": 0.8242528715629991,
"learning_rate": 8.058290915200597e-06,
"loss": 0.1626,
"step": 232
},
{
"epoch": 0.3155044008124577,
"grad_norm": 0.7734381426678048,
"learning_rate": 8.04090550155455e-06,
"loss": 0.1661,
"step": 233
},
{
"epoch": 0.31685849695328366,
"grad_norm": 0.9253057348506322,
"learning_rate": 8.023461544779833e-06,
"loss": 0.1519,
"step": 234
},
{
"epoch": 0.3182125930941097,
"grad_norm": 0.8468301461753109,
"learning_rate": 8.005959380705298e-06,
"loss": 0.167,
"step": 235
},
{
"epoch": 0.3195666892349357,
"grad_norm": 1.048267366132464,
"learning_rate": 7.988399346280398e-06,
"loss": 0.1969,
"step": 236
},
{
"epoch": 0.3209207853757617,
"grad_norm": 0.9368622391453264,
"learning_rate": 7.970781779568699e-06,
"loss": 0.2483,
"step": 237
},
{
"epoch": 0.3222748815165877,
"grad_norm": 0.9084222255162543,
"learning_rate": 7.953107019741366e-06,
"loss": 0.1979,
"step": 238
},
{
"epoch": 0.32362897765741366,
"grad_norm": 0.8506001609560738,
"learning_rate": 7.935375407070648e-06,
"loss": 0.1868,
"step": 239
},
{
"epoch": 0.32498307379823965,
"grad_norm": 0.9362984711106395,
"learning_rate": 7.917587282923312e-06,
"loss": 0.1927,
"step": 240
},
{
"epoch": 0.3263371699390657,
"grad_norm": 0.8031982474765188,
"learning_rate": 7.899742989754075e-06,
"loss": 0.1293,
"step": 241
},
{
"epoch": 0.3276912660798917,
"grad_norm": 0.8714612871618181,
"learning_rate": 7.88184287109902e-06,
"loss": 0.157,
"step": 242
},
{
"epoch": 0.3290453622207177,
"grad_norm": 0.8465253088676411,
"learning_rate": 7.863887271568963e-06,
"loss": 0.1827,
"step": 243
},
{
"epoch": 0.33039945836154366,
"grad_norm": 0.8170187193691545,
"learning_rate": 7.845876536842846e-06,
"loss": 0.1955,
"step": 244
},
{
"epoch": 0.33175355450236965,
"grad_norm": 0.82710317567014,
"learning_rate": 7.827811013661056e-06,
"loss": 0.1929,
"step": 245
},
{
"epoch": 0.33310765064319564,
"grad_norm": 0.9680376246257761,
"learning_rate": 7.809691049818766e-06,
"loss": 0.1567,
"step": 246
},
{
"epoch": 0.3344617467840217,
"grad_norm": 0.8596611616027606,
"learning_rate": 7.791516994159226e-06,
"loss": 0.2252,
"step": 247
},
{
"epoch": 0.3358158429248477,
"grad_norm": 0.9273469717702996,
"learning_rate": 7.773289196567066e-06,
"loss": 0.1934,
"step": 248
},
{
"epoch": 0.33716993906567366,
"grad_norm": 0.7503085680062017,
"learning_rate": 7.755008007961542e-06,
"loss": 0.1646,
"step": 249
},
{
"epoch": 0.33852403520649965,
"grad_norm": 0.6653161546952755,
"learning_rate": 7.736673780289788e-06,
"loss": 0.1243,
"step": 250
},
{
"epoch": 0.33987813134732564,
"grad_norm": 0.7916280862974827,
"learning_rate": 7.718286866520037e-06,
"loss": 0.185,
"step": 251
},
{
"epoch": 0.3412322274881517,
"grad_norm": 1.0039542842854647,
"learning_rate": 7.699847620634834e-06,
"loss": 0.1593,
"step": 252
},
{
"epoch": 0.3425863236289777,
"grad_norm": 0.8921215532786686,
"learning_rate": 7.681356397624212e-06,
"loss": 0.1802,
"step": 253
},
{
"epoch": 0.34394041976980366,
"grad_norm": 0.6880619313855061,
"learning_rate": 7.662813553478857e-06,
"loss": 0.1567,
"step": 254
},
{
"epoch": 0.34529451591062965,
"grad_norm": 0.9662186570231487,
"learning_rate": 7.644219445183267e-06,
"loss": 0.1693,
"step": 255
},
{
"epoch": 0.34664861205145564,
"grad_norm": 0.8207802837695596,
"learning_rate": 7.625574430708867e-06,
"loss": 0.1819,
"step": 256
},
{
"epoch": 0.34800270819228163,
"grad_norm": 1.0177226117646463,
"learning_rate": 7.606878869007117e-06,
"loss": 0.2308,
"step": 257
},
{
"epoch": 0.3493568043331077,
"grad_norm": 0.8484528807429219,
"learning_rate": 7.588133120002612e-06,
"loss": 0.1576,
"step": 258
},
{
"epoch": 0.35071090047393366,
"grad_norm": 0.897583876701313,
"learning_rate": 7.569337544586145e-06,
"loss": 0.146,
"step": 259
},
{
"epoch": 0.35206499661475965,
"grad_norm": 0.8910099512360655,
"learning_rate": 7.5504925046077596e-06,
"loss": 0.1689,
"step": 260
},
{
"epoch": 0.35341909275558564,
"grad_norm": 0.8901159708602862,
"learning_rate": 7.531598362869784e-06,
"loss": 0.1524,
"step": 261
},
{
"epoch": 0.35477318889641163,
"grad_norm": 1.660762562072437,
"learning_rate": 7.5126554831198506e-06,
"loss": 0.1485,
"step": 262
},
{
"epoch": 0.3561272850372376,
"grad_norm": 0.8457275292404339,
"learning_rate": 7.493664230043889e-06,
"loss": 0.2006,
"step": 263
},
{
"epoch": 0.35748138117806366,
"grad_norm": 1.0641705532624839,
"learning_rate": 7.474624969259101e-06,
"loss": 0.2046,
"step": 264
},
{
"epoch": 0.35883547731888965,
"grad_norm": 0.6643471526907597,
"learning_rate": 7.455538067306935e-06,
"loss": 0.1473,
"step": 265
},
{
"epoch": 0.36018957345971564,
"grad_norm": 0.8984130182781388,
"learning_rate": 7.436403891646014e-06,
"loss": 0.2102,
"step": 266
},
{
"epoch": 0.36154366960054163,
"grad_norm": 0.8651905796462697,
"learning_rate": 7.417222810645072e-06,
"loss": 0.1844,
"step": 267
},
{
"epoch": 0.3628977657413676,
"grad_norm": 0.7464787685017356,
"learning_rate": 7.3979951935758596e-06,
"loss": 0.1783,
"step": 268
},
{
"epoch": 0.36425186188219366,
"grad_norm": 0.6194945596623783,
"learning_rate": 7.378721410606029e-06,
"loss": 0.1405,
"step": 269
},
{
"epoch": 0.36560595802301965,
"grad_norm": 1.0052453539387527,
"learning_rate": 7.359401832792019e-06,
"loss": 0.1775,
"step": 270
},
{
"epoch": 0.36696005416384564,
"grad_norm": 0.94187429309211,
"learning_rate": 7.340036832071899e-06,
"loss": 0.1594,
"step": 271
},
{
"epoch": 0.36831415030467163,
"grad_norm": 0.8879159531033977,
"learning_rate": 7.32062678125822e-06,
"loss": 0.1541,
"step": 272
},
{
"epoch": 0.3696682464454976,
"grad_norm": 0.8561355343272015,
"learning_rate": 7.301172054030829e-06,
"loss": 0.1568,
"step": 273
},
{
"epoch": 0.3710223425863236,
"grad_norm": 0.687826025987938,
"learning_rate": 7.281673024929674e-06,
"loss": 0.1764,
"step": 274
},
{
"epoch": 0.37237643872714965,
"grad_norm": 1.1494624950206707,
"learning_rate": 7.262130069347603e-06,
"loss": 0.206,
"step": 275
},
{
"epoch": 0.37373053486797564,
"grad_norm": 1.0473138916163023,
"learning_rate": 7.242543563523128e-06,
"loss": 0.1918,
"step": 276
},
{
"epoch": 0.37508463100880163,
"grad_norm": 1.186224872212997,
"learning_rate": 7.222913884533187e-06,
"loss": 0.1644,
"step": 277
},
{
"epoch": 0.3764387271496276,
"grad_norm": 0.9080200192584014,
"learning_rate": 7.2032414102858795e-06,
"loss": 0.1674,
"step": 278
},
{
"epoch": 0.3777928232904536,
"grad_norm": 0.8937111694347315,
"learning_rate": 7.183526519513196e-06,
"loss": 0.2005,
"step": 279
},
{
"epoch": 0.3791469194312796,
"grad_norm": 0.9687175608521055,
"learning_rate": 7.163769591763723e-06,
"loss": 0.165,
"step": 280
},
{
"epoch": 0.38050101557210564,
"grad_norm": 0.9345466042023967,
"learning_rate": 7.143971007395338e-06,
"loss": 0.1681,
"step": 281
},
{
"epoch": 0.3818551117129316,
"grad_norm": 1.0056884078563586,
"learning_rate": 7.1241311475678896e-06,
"loss": 0.1991,
"step": 282
},
{
"epoch": 0.3832092078537576,
"grad_norm": 1.1727033512490486,
"learning_rate": 7.10425039423585e-06,
"loss": 0.1378,
"step": 283
},
{
"epoch": 0.3845633039945836,
"grad_norm": 1.0359435064455782,
"learning_rate": 7.084329130140972e-06,
"loss": 0.1721,
"step": 284
},
{
"epoch": 0.3859174001354096,
"grad_norm": 0.8534569543272673,
"learning_rate": 7.0643677388049195e-06,
"loss": 0.1553,
"step": 285
},
{
"epoch": 0.38727149627623564,
"grad_norm": 1.10949728912015,
"learning_rate": 7.044366604521874e-06,
"loss": 0.2151,
"step": 286
},
{
"epoch": 0.3886255924170616,
"grad_norm": 1.0482046221958576,
"learning_rate": 7.024326112351147e-06,
"loss": 0.1647,
"step": 287
},
{
"epoch": 0.3899796885578876,
"grad_norm": 1.0334351957745223,
"learning_rate": 7.004246648109765e-06,
"loss": 0.2101,
"step": 288
},
{
"epoch": 0.3913337846987136,
"grad_norm": 1.059711322689107,
"learning_rate": 6.984128598365034e-06,
"loss": 0.1836,
"step": 289
},
{
"epoch": 0.3926878808395396,
"grad_norm": 1.0500778508321833,
"learning_rate": 6.963972350427112e-06,
"loss": 0.1568,
"step": 290
},
{
"epoch": 0.3940419769803656,
"grad_norm": 1.0066770854298477,
"learning_rate": 6.943778292341535e-06,
"loss": 0.1527,
"step": 291
},
{
"epoch": 0.3953960731211916,
"grad_norm": 0.80089632811313,
"learning_rate": 6.923546812881759e-06,
"loss": 0.1411,
"step": 292
},
{
"epoch": 0.3967501692620176,
"grad_norm": 0.9715891823560623,
"learning_rate": 6.903278301541673e-06,
"loss": 0.2066,
"step": 293
},
{
"epoch": 0.3981042654028436,
"grad_norm": 0.9485977558655331,
"learning_rate": 6.882973148528096e-06,
"loss": 0.1679,
"step": 294
},
{
"epoch": 0.3994583615436696,
"grad_norm": 1.0893018711589424,
"learning_rate": 6.862631744753271e-06,
"loss": 0.2311,
"step": 295
},
{
"epoch": 0.4008124576844956,
"grad_norm": 0.5920054417789278,
"learning_rate": 6.8422544818273336e-06,
"loss": 0.1286,
"step": 296
},
{
"epoch": 0.40216655382532157,
"grad_norm": 0.8060571295468734,
"learning_rate": 6.821841752050775e-06,
"loss": 0.186,
"step": 297
},
{
"epoch": 0.4035206499661476,
"grad_norm": 0.9545489844306732,
"learning_rate": 6.801393948406894e-06,
"loss": 0.1574,
"step": 298
},
{
"epoch": 0.4048747461069736,
"grad_norm": 0.8503347251875324,
"learning_rate": 6.78091146455422e-06,
"loss": 0.1511,
"step": 299
},
{
"epoch": 0.4062288422477996,
"grad_norm": 0.9008908971915316,
"learning_rate": 6.760394694818949e-06,
"loss": 0.1983,
"step": 300
},
{
"epoch": 0.4075829383886256,
"grad_norm": 0.9531852644889576,
"learning_rate": 6.73984403418734e-06,
"loss": 0.1382,
"step": 301
},
{
"epoch": 0.40893703452945157,
"grad_norm": 0.7752309245303294,
"learning_rate": 6.719259878298113e-06,
"loss": 0.1735,
"step": 302
},
{
"epoch": 0.4102911306702776,
"grad_norm": 0.9698820961720023,
"learning_rate": 6.69864262343484e-06,
"loss": 0.2193,
"step": 303
},
{
"epoch": 0.4116452268111036,
"grad_norm": 0.9071520692579097,
"learning_rate": 6.67799266651831e-06,
"loss": 0.1955,
"step": 304
},
{
"epoch": 0.4129993229519296,
"grad_norm": 0.7963117853971042,
"learning_rate": 6.65731040509888e-06,
"loss": 0.1601,
"step": 305
},
{
"epoch": 0.4143534190927556,
"grad_norm": 0.6866600321908848,
"learning_rate": 6.63659623734884e-06,
"loss": 0.1557,
"step": 306
},
{
"epoch": 0.41570751523358157,
"grad_norm": 0.9218704833633397,
"learning_rate": 6.61585056205473e-06,
"loss": 0.1841,
"step": 307
},
{
"epoch": 0.41706161137440756,
"grad_norm": 0.8921876197901185,
"learning_rate": 6.59507377860967e-06,
"loss": 0.1637,
"step": 308
},
{
"epoch": 0.4184157075152336,
"grad_norm": 0.7419286800605366,
"learning_rate": 6.574266287005671e-06,
"loss": 0.1942,
"step": 309
},
{
"epoch": 0.4197698036560596,
"grad_norm": 0.7162168821216858,
"learning_rate": 6.553428487825933e-06,
"loss": 0.1611,
"step": 310
},
{
"epoch": 0.4211238997968856,
"grad_norm": 0.9678841945505542,
"learning_rate": 6.532560782237134e-06,
"loss": 0.2552,
"step": 311
},
{
"epoch": 0.42247799593771157,
"grad_norm": 0.7070147171401922,
"learning_rate": 6.511663571981708e-06,
"loss": 0.135,
"step": 312
},
{
"epoch": 0.42383209207853756,
"grad_norm": 0.9277260998490678,
"learning_rate": 6.490737259370109e-06,
"loss": 0.1507,
"step": 313
},
{
"epoch": 0.42518618821936355,
"grad_norm": 0.9475671400048534,
"learning_rate": 6.46978224727306e-06,
"loss": 0.1741,
"step": 314
},
{
"epoch": 0.4265402843601896,
"grad_norm": 1.0298508784862794,
"learning_rate": 6.448798939113812e-06,
"loss": 0.1671,
"step": 315
},
{
"epoch": 0.4278943805010156,
"grad_norm": 0.7652129031035252,
"learning_rate": 6.42778773886036e-06,
"loss": 0.095,
"step": 316
},
{
"epoch": 0.42924847664184157,
"grad_norm": 1.1221102672776428,
"learning_rate": 6.406749051017679e-06,
"loss": 0.1634,
"step": 317
},
{
"epoch": 0.43060257278266756,
"grad_norm": 0.9970063697928587,
"learning_rate": 6.3856832806199324e-06,
"loss": 0.1849,
"step": 318
},
{
"epoch": 0.43195666892349355,
"grad_norm": 0.8237393384268415,
"learning_rate": 6.364590833222669e-06,
"loss": 0.1426,
"step": 319
},
{
"epoch": 0.4333107650643196,
"grad_norm": 0.9183348693596578,
"learning_rate": 6.343472114895022e-06,
"loss": 0.1596,
"step": 320
},
{
"epoch": 0.4346648612051456,
"grad_norm": 0.9407001435883583,
"learning_rate": 6.32232753221189e-06,
"loss": 0.2142,
"step": 321
},
{
"epoch": 0.43601895734597157,
"grad_norm": 0.9166483547213038,
"learning_rate": 6.301157492246111e-06,
"loss": 0.1919,
"step": 322
},
{
"epoch": 0.43737305348679756,
"grad_norm": 0.940612186841125,
"learning_rate": 6.279962402560618e-06,
"loss": 0.1784,
"step": 323
},
{
"epoch": 0.43872714962762355,
"grad_norm": 1.08449020749176,
"learning_rate": 6.2587426712006005e-06,
"loss": 0.2249,
"step": 324
},
{
"epoch": 0.44008124576844954,
"grad_norm": 0.7054507878576708,
"learning_rate": 6.237498706685647e-06,
"loss": 0.1506,
"step": 325
},
{
"epoch": 0.4414353419092756,
"grad_norm": 0.8192376829806236,
"learning_rate": 6.216230918001881e-06,
"loss": 0.201,
"step": 326
},
{
"epoch": 0.44278943805010157,
"grad_norm": 0.93471951121777,
"learning_rate": 6.194939714594085e-06,
"loss": 0.1457,
"step": 327
},
{
"epoch": 0.44414353419092756,
"grad_norm": 1.0087133713551801,
"learning_rate": 6.173625506357814e-06,
"loss": 0.1671,
"step": 328
},
{
"epoch": 0.44549763033175355,
"grad_norm": 0.8095347337691632,
"learning_rate": 6.152288703631516e-06,
"loss": 0.1452,
"step": 329
},
{
"epoch": 0.44685172647257954,
"grad_norm": 0.715220216221791,
"learning_rate": 6.1309297171886225e-06,
"loss": 0.1595,
"step": 330
},
{
"epoch": 0.4482058226134055,
"grad_norm": 0.966168514208031,
"learning_rate": 6.109548958229642e-06,
"loss": 0.1863,
"step": 331
},
{
"epoch": 0.44955991875423157,
"grad_norm": 1.1283252500415997,
"learning_rate": 6.088146838374247e-06,
"loss": 0.1834,
"step": 332
},
{
"epoch": 0.45091401489505756,
"grad_norm": 0.6855172744728713,
"learning_rate": 6.0667237696533445e-06,
"loss": 0.1377,
"step": 333
},
{
"epoch": 0.45226811103588355,
"grad_norm": 0.8118145348083093,
"learning_rate": 6.045280164501147e-06,
"loss": 0.1865,
"step": 334
},
{
"epoch": 0.45362220717670954,
"grad_norm": 1.0733023519681528,
"learning_rate": 6.023816435747236e-06,
"loss": 0.1859,
"step": 335
},
{
"epoch": 0.4549763033175355,
"grad_norm": 0.9509551501670634,
"learning_rate": 6.002332996608605e-06,
"loss": 0.1874,
"step": 336
},
{
"epoch": 0.45633039945836157,
"grad_norm": 0.7992733428671231,
"learning_rate": 5.98083026068171e-06,
"loss": 0.181,
"step": 337
},
{
"epoch": 0.45768449559918756,
"grad_norm": 0.7825955330561137,
"learning_rate": 5.9593086419345104e-06,
"loss": 0.1281,
"step": 338
},
{
"epoch": 0.45903859174001355,
"grad_norm": 0.8154627381438526,
"learning_rate": 5.937768554698489e-06,
"loss": 0.1752,
"step": 339
},
{
"epoch": 0.46039268788083954,
"grad_norm": 0.9492427857414811,
"learning_rate": 5.916210413660687e-06,
"loss": 0.1716,
"step": 340
},
{
"epoch": 0.4617467840216655,
"grad_norm": 0.7557552905128929,
"learning_rate": 5.8946346338557125e-06,
"loss": 0.1529,
"step": 341
},
{
"epoch": 0.4631008801624915,
"grad_norm": 0.836085178403307,
"learning_rate": 5.873041630657752e-06,
"loss": 0.1782,
"step": 342
},
{
"epoch": 0.46445497630331756,
"grad_norm": 1.0332932470938718,
"learning_rate": 5.851431819772579e-06,
"loss": 0.1449,
"step": 343
},
{
"epoch": 0.46580907244414355,
"grad_norm": 0.8152998001997726,
"learning_rate": 5.82980561722954e-06,
"loss": 0.1712,
"step": 344
},
{
"epoch": 0.46716316858496953,
"grad_norm": 0.7094776992512005,
"learning_rate": 5.808163439373557e-06,
"loss": 0.1547,
"step": 345
},
{
"epoch": 0.4685172647257955,
"grad_norm": 0.6221862542819474,
"learning_rate": 5.786505702857107e-06,
"loss": 0.1365,
"step": 346
},
{
"epoch": 0.4698713608666215,
"grad_norm": 0.9572200776402384,
"learning_rate": 5.764832824632194e-06,
"loss": 0.1863,
"step": 347
},
{
"epoch": 0.4712254570074475,
"grad_norm": 0.7205234742021339,
"learning_rate": 5.743145221942333e-06,
"loss": 0.1417,
"step": 348
},
{
"epoch": 0.47257955314827355,
"grad_norm": 0.5876542860033283,
"learning_rate": 5.721443312314517e-06,
"loss": 0.164,
"step": 349
},
{
"epoch": 0.47393364928909953,
"grad_norm": 0.8345495578505313,
"learning_rate": 5.6997275135511635e-06,
"loss": 0.193,
"step": 350
},
{
"epoch": 0.4752877454299255,
"grad_norm": 0.6774752368569341,
"learning_rate": 5.677998243722089e-06,
"loss": 0.1463,
"step": 351
},
{
"epoch": 0.4766418415707515,
"grad_norm": 0.726429222521974,
"learning_rate": 5.656255921156455e-06,
"loss": 0.155,
"step": 352
},
{
"epoch": 0.4779959377115775,
"grad_norm": 0.9662923542491118,
"learning_rate": 5.634500964434707e-06,
"loss": 0.1902,
"step": 353
},
{
"epoch": 0.47935003385240355,
"grad_norm": 0.9471787277812976,
"learning_rate": 5.612733792380525e-06,
"loss": 0.2186,
"step": 354
},
{
"epoch": 0.48070412999322953,
"grad_norm": 0.7464409121548943,
"learning_rate": 5.590954824052756e-06,
"loss": 0.1799,
"step": 355
},
{
"epoch": 0.4820582261340555,
"grad_norm": 1.0258190434055041,
"learning_rate": 5.569164478737346e-06,
"loss": 0.2128,
"step": 356
},
{
"epoch": 0.4834123222748815,
"grad_norm": 0.7888106152195081,
"learning_rate": 5.547363175939274e-06,
"loss": 0.1752,
"step": 357
},
{
"epoch": 0.4847664184157075,
"grad_norm": 1.0014329871966479,
"learning_rate": 5.525551335374469e-06,
"loss": 0.2052,
"step": 358
},
{
"epoch": 0.4861205145565335,
"grad_norm": 0.8062259305401849,
"learning_rate": 5.503729376961727e-06,
"loss": 0.2014,
"step": 359
},
{
"epoch": 0.48747461069735953,
"grad_norm": 0.8479210573125648,
"learning_rate": 5.4818977208146375e-06,
"loss": 0.1499,
"step": 360
},
{
"epoch": 0.4888287068381855,
"grad_norm": 0.7281211701979133,
"learning_rate": 5.460056787233488e-06,
"loss": 0.1519,
"step": 361
},
{
"epoch": 0.4901828029790115,
"grad_norm": 0.8807049900826209,
"learning_rate": 5.438206996697173e-06,
"loss": 0.1713,
"step": 362
},
{
"epoch": 0.4915368991198375,
"grad_norm": 1.1636439821378188,
"learning_rate": 5.416348769855099e-06,
"loss": 0.2012,
"step": 363
},
{
"epoch": 0.4928909952606635,
"grad_norm": 0.799118068498799,
"learning_rate": 5.394482527519089e-06,
"loss": 0.1856,
"step": 364
},
{
"epoch": 0.4942450914014895,
"grad_norm": 0.7560905749606406,
"learning_rate": 5.372608690655281e-06,
"loss": 0.154,
"step": 365
},
{
"epoch": 0.4955991875423155,
"grad_norm": 0.8398501589092475,
"learning_rate": 5.350727680376019e-06,
"loss": 0.1572,
"step": 366
},
{
"epoch": 0.4969532836831415,
"grad_norm": 1.0320929285283185,
"learning_rate": 5.3288399179317485e-06,
"loss": 0.1889,
"step": 367
},
{
"epoch": 0.4983073798239675,
"grad_norm": 0.8009871361146,
"learning_rate": 5.306945824702908e-06,
"loss": 0.1775,
"step": 368
},
{
"epoch": 0.4996614759647935,
"grad_norm": 0.8799149541646992,
"learning_rate": 5.2850458221918155e-06,
"loss": 0.1529,
"step": 369
},
{
"epoch": 0.5010155721056195,
"grad_norm": 0.85124023187912,
"learning_rate": 5.263140332014553e-06,
"loss": 0.1273,
"step": 370
},
{
"epoch": 0.5023696682464455,
"grad_norm": 0.8676029255003597,
"learning_rate": 5.241229775892852e-06,
"loss": 0.1619,
"step": 371
},
{
"epoch": 0.5037237643872715,
"grad_norm": 0.8988586584091905,
"learning_rate": 5.219314575645972e-06,
"loss": 0.1803,
"step": 372
},
{
"epoch": 0.5050778605280974,
"grad_norm": 0.8705134986176656,
"learning_rate": 5.197395153182577e-06,
"loss": 0.175,
"step": 373
},
{
"epoch": 0.5064319566689235,
"grad_norm": 1.2438609511972156,
"learning_rate": 5.1754719304926245e-06,
"loss": 0.1355,
"step": 374
},
{
"epoch": 0.5077860528097495,
"grad_norm": 0.9665941670457021,
"learning_rate": 5.153545329639226e-06,
"loss": 0.1846,
"step": 375
},
{
"epoch": 0.5091401489505755,
"grad_norm": 1.0083680990308304,
"learning_rate": 5.131615772750534e-06,
"loss": 0.1508,
"step": 376
},
{
"epoch": 0.5104942450914015,
"grad_norm": 0.835658637521355,
"learning_rate": 5.109683682011608e-06,
"loss": 0.1757,
"step": 377
},
{
"epoch": 0.5118483412322274,
"grad_norm": 0.7954729122512308,
"learning_rate": 5.087749479656286e-06,
"loss": 0.1736,
"step": 378
},
{
"epoch": 0.5132024373730535,
"grad_norm": 0.8445866330407172,
"learning_rate": 5.065813587959065e-06,
"loss": 0.1586,
"step": 379
},
{
"epoch": 0.5145565335138795,
"grad_norm": 0.7101496990744409,
"learning_rate": 5.043876429226962e-06,
"loss": 0.2093,
"step": 380
},
{
"epoch": 0.5159106296547055,
"grad_norm": 0.9731725655936514,
"learning_rate": 5.021938425791381e-06,
"loss": 0.1718,
"step": 381
},
{
"epoch": 0.5172647257955315,
"grad_norm": 1.0315457677005235,
"learning_rate": 5e-06,
"loss": 0.1924,
"step": 382
},
{
"epoch": 0.5186188219363574,
"grad_norm": 1.0031406841155923,
"learning_rate": 4.9780615742086195e-06,
"loss": 0.1944,
"step": 383
},
{
"epoch": 0.5199729180771835,
"grad_norm": 0.8276747667590124,
"learning_rate": 4.95612357077304e-06,
"loss": 0.1819,
"step": 384
},
{
"epoch": 0.5213270142180095,
"grad_norm": 0.7136904332435295,
"learning_rate": 4.934186412040935e-06,
"loss": 0.1337,
"step": 385
},
{
"epoch": 0.5226811103588355,
"grad_norm": 0.826465773743981,
"learning_rate": 4.912250520343715e-06,
"loss": 0.1712,
"step": 386
},
{
"epoch": 0.5240352064996615,
"grad_norm": 0.9818165784203778,
"learning_rate": 4.890316317988395e-06,
"loss": 0.224,
"step": 387
},
{
"epoch": 0.5253893026404874,
"grad_norm": 0.8195786046166481,
"learning_rate": 4.868384227249468e-06,
"loss": 0.1951,
"step": 388
},
{
"epoch": 0.5267433987813135,
"grad_norm": 0.9055091305049465,
"learning_rate": 4.846454670360774e-06,
"loss": 0.1771,
"step": 389
},
{
"epoch": 0.5280974949221394,
"grad_norm": 0.8051183618479892,
"learning_rate": 4.824528069507377e-06,
"loss": 0.162,
"step": 390
},
{
"epoch": 0.5294515910629655,
"grad_norm": 0.9464547078433297,
"learning_rate": 4.8026048468174245e-06,
"loss": 0.2001,
"step": 391
},
{
"epoch": 0.5308056872037915,
"grad_norm": 0.9682696888060136,
"learning_rate": 4.7806854243540304e-06,
"loss": 0.1873,
"step": 392
},
{
"epoch": 0.5321597833446174,
"grad_norm": 0.7596809353365747,
"learning_rate": 4.758770224107149e-06,
"loss": 0.1738,
"step": 393
},
{
"epoch": 0.5335138794854435,
"grad_norm": 0.8491525767524211,
"learning_rate": 4.736859667985448e-06,
"loss": 0.1645,
"step": 394
},
{
"epoch": 0.5348679756262694,
"grad_norm": 0.7912156480903655,
"learning_rate": 4.714954177808186e-06,
"loss": 0.1757,
"step": 395
},
{
"epoch": 0.5362220717670955,
"grad_norm": 0.8665850758921498,
"learning_rate": 4.693054175297095e-06,
"loss": 0.137,
"step": 396
},
{
"epoch": 0.5375761679079215,
"grad_norm": 0.9114269137215836,
"learning_rate": 4.671160082068253e-06,
"loss": 0.1935,
"step": 397
},
{
"epoch": 0.5389302640487474,
"grad_norm": 0.8874667670758649,
"learning_rate": 4.649272319623983e-06,
"loss": 0.2052,
"step": 398
},
{
"epoch": 0.5402843601895735,
"grad_norm": 0.918735013107402,
"learning_rate": 4.62739130934472e-06,
"loss": 0.2007,
"step": 399
},
{
"epoch": 0.5416384563303994,
"grad_norm": 1.033246888514312,
"learning_rate": 4.605517472480912e-06,
"loss": 0.1898,
"step": 400
},
{
"epoch": 0.5429925524712255,
"grad_norm": 0.7644612043411352,
"learning_rate": 4.583651230144902e-06,
"loss": 0.1175,
"step": 401
},
{
"epoch": 0.5443466486120515,
"grad_norm": 1.3031910956124564,
"learning_rate": 4.561793003302829e-06,
"loss": 0.2048,
"step": 402
},
{
"epoch": 0.5457007447528774,
"grad_norm": 0.849000966226295,
"learning_rate": 4.539943212766513e-06,
"loss": 0.1381,
"step": 403
},
{
"epoch": 0.5470548408937035,
"grad_norm": 0.8991590513631682,
"learning_rate": 4.5181022791853625e-06,
"loss": 0.1571,
"step": 404
},
{
"epoch": 0.5484089370345294,
"grad_norm": 0.8751940696739857,
"learning_rate": 4.496270623038276e-06,
"loss": 0.1551,
"step": 405
},
{
"epoch": 0.5497630331753555,
"grad_norm": 0.6899911330169567,
"learning_rate": 4.474448664625533e-06,
"loss": 0.1378,
"step": 406
},
{
"epoch": 0.5511171293161814,
"grad_norm": 0.9025384407109227,
"learning_rate": 4.452636824060727e-06,
"loss": 0.1629,
"step": 407
},
{
"epoch": 0.5524712254570074,
"grad_norm": 0.7865489765683099,
"learning_rate": 4.430835521262656e-06,
"loss": 0.1578,
"step": 408
},
{
"epoch": 0.5538253215978335,
"grad_norm": 0.8545156484076634,
"learning_rate": 4.409045175947246e-06,
"loss": 0.1678,
"step": 409
},
{
"epoch": 0.5551794177386594,
"grad_norm": 1.1762434024484008,
"learning_rate": 4.387266207619477e-06,
"loss": 0.1869,
"step": 410
},
{
"epoch": 0.5565335138794855,
"grad_norm": 0.7005722277233816,
"learning_rate": 4.3654990355652945e-06,
"loss": 0.1853,
"step": 411
},
{
"epoch": 0.5578876100203114,
"grad_norm": 1.0964938250471914,
"learning_rate": 4.343744078843545e-06,
"loss": 0.1695,
"step": 412
},
{
"epoch": 0.5592417061611374,
"grad_norm": 0.8337790321425751,
"learning_rate": 4.32200175627791e-06,
"loss": 0.1732,
"step": 413
},
{
"epoch": 0.5605958023019635,
"grad_norm": 0.8878392588601526,
"learning_rate": 4.300272486448839e-06,
"loss": 0.1602,
"step": 414
},
{
"epoch": 0.5619498984427894,
"grad_norm": 1.0606024225314115,
"learning_rate": 4.278556687685485e-06,
"loss": 0.1984,
"step": 415
},
{
"epoch": 0.5633039945836155,
"grad_norm": 0.824001955189102,
"learning_rate": 4.256854778057668e-06,
"loss": 0.1272,
"step": 416
},
{
"epoch": 0.5646580907244414,
"grad_norm": 0.9144341825717467,
"learning_rate": 4.235167175367808e-06,
"loss": 0.1629,
"step": 417
},
{
"epoch": 0.5660121868652674,
"grad_norm": 1.0138382252246012,
"learning_rate": 4.213494297142895e-06,
"loss": 0.1837,
"step": 418
},
{
"epoch": 0.5673662830060935,
"grad_norm": 0.9904393788643632,
"learning_rate": 4.1918365606264434e-06,
"loss": 0.1577,
"step": 419
},
{
"epoch": 0.5687203791469194,
"grad_norm": 0.6434826413413879,
"learning_rate": 4.170194382770462e-06,
"loss": 0.1054,
"step": 420
},
{
"epoch": 0.5700744752877455,
"grad_norm": 0.9270851331377845,
"learning_rate": 4.148568180227422e-06,
"loss": 0.1644,
"step": 421
},
{
"epoch": 0.5714285714285714,
"grad_norm": 1.1098669419754115,
"learning_rate": 4.1269583693422484e-06,
"loss": 0.1984,
"step": 422
},
{
"epoch": 0.5727826675693974,
"grad_norm": 1.1039317291612822,
"learning_rate": 4.10536536614429e-06,
"loss": 0.1846,
"step": 423
},
{
"epoch": 0.5741367637102234,
"grad_norm": 0.8683532682645492,
"learning_rate": 4.083789586339315e-06,
"loss": 0.194,
"step": 424
},
{
"epoch": 0.5754908598510494,
"grad_norm": 0.9924994612157984,
"learning_rate": 4.0622314453015135e-06,
"loss": 0.1856,
"step": 425
},
{
"epoch": 0.5768449559918755,
"grad_norm": 1.1094733293815595,
"learning_rate": 4.040691358065491e-06,
"loss": 0.198,
"step": 426
},
{
"epoch": 0.5781990521327014,
"grad_norm": 0.767361509742422,
"learning_rate": 4.0191697393182904e-06,
"loss": 0.1583,
"step": 427
},
{
"epoch": 0.5795531482735274,
"grad_norm": 1.114387942936837,
"learning_rate": 3.997667003391397e-06,
"loss": 0.1454,
"step": 428
},
{
"epoch": 0.5809072444143534,
"grad_norm": 0.868911375630595,
"learning_rate": 3.976183564252765e-06,
"loss": 0.1774,
"step": 429
},
{
"epoch": 0.5822613405551794,
"grad_norm": 0.8688457161898233,
"learning_rate": 3.954719835498853e-06,
"loss": 0.1959,
"step": 430
},
{
"epoch": 0.5836154366960055,
"grad_norm": 1.0279290659792153,
"learning_rate": 3.933276230346657e-06,
"loss": 0.2451,
"step": 431
},
{
"epoch": 0.5849695328368314,
"grad_norm": 0.758061369534775,
"learning_rate": 3.911853161625756e-06,
"loss": 0.1793,
"step": 432
},
{
"epoch": 0.5863236289776574,
"grad_norm": 0.8289344689406578,
"learning_rate": 3.89045104177036e-06,
"loss": 0.188,
"step": 433
},
{
"epoch": 0.5876777251184834,
"grad_norm": 0.7472232494528747,
"learning_rate": 3.869070282811379e-06,
"loss": 0.1983,
"step": 434
},
{
"epoch": 0.5890318212593094,
"grad_norm": 1.020582238837582,
"learning_rate": 3.847711296368485e-06,
"loss": 0.1702,
"step": 435
},
{
"epoch": 0.5903859174001355,
"grad_norm": 0.8802132225405479,
"learning_rate": 3.826374493642187e-06,
"loss": 0.1766,
"step": 436
},
{
"epoch": 0.5917400135409614,
"grad_norm": 0.9704169964321497,
"learning_rate": 3.8050602854059164e-06,
"loss": 0.1637,
"step": 437
},
{
"epoch": 0.5930941096817874,
"grad_norm": 0.7836273051542119,
"learning_rate": 3.7837690819981187e-06,
"loss": 0.155,
"step": 438
},
{
"epoch": 0.5944482058226134,
"grad_norm": 0.8796850878274921,
"learning_rate": 3.7625012933143533e-06,
"loss": 0.1305,
"step": 439
},
{
"epoch": 0.5958023019634394,
"grad_norm": 0.9191392037904091,
"learning_rate": 3.7412573287994e-06,
"loss": 0.1601,
"step": 440
},
{
"epoch": 0.5971563981042654,
"grad_norm": 0.951625687969982,
"learning_rate": 3.7200375974393847e-06,
"loss": 0.2069,
"step": 441
},
{
"epoch": 0.5985104942450914,
"grad_norm": 0.7631480251748166,
"learning_rate": 3.6988425077538904e-06,
"loss": 0.1484,
"step": 442
},
{
"epoch": 0.5998645903859174,
"grad_norm": 0.8863237160810076,
"learning_rate": 3.6776724677881105e-06,
"loss": 0.1264,
"step": 443
},
{
"epoch": 0.6012186865267434,
"grad_norm": 0.8853537316373861,
"learning_rate": 3.6565278851049803e-06,
"loss": 0.1746,
"step": 444
},
{
"epoch": 0.6025727826675694,
"grad_norm": 0.7897245672969606,
"learning_rate": 3.6354091667773327e-06,
"loss": 0.1207,
"step": 445
},
{
"epoch": 0.6039268788083954,
"grad_norm": 0.7897033598013464,
"learning_rate": 3.614316719380069e-06,
"loss": 0.1453,
"step": 446
},
{
"epoch": 0.6052809749492214,
"grad_norm": 0.9395735699864135,
"learning_rate": 3.593250948982322e-06,
"loss": 0.1497,
"step": 447
},
{
"epoch": 0.6066350710900474,
"grad_norm": 0.7843529078550943,
"learning_rate": 3.5722122611396416e-06,
"loss": 0.1301,
"step": 448
},
{
"epoch": 0.6079891672308734,
"grad_norm": 0.7403196083567885,
"learning_rate": 3.551201060886189e-06,
"loss": 0.1184,
"step": 449
},
{
"epoch": 0.6093432633716994,
"grad_norm": 0.8531649403019163,
"learning_rate": 3.530217752726942e-06,
"loss": 0.1752,
"step": 450
},
{
"epoch": 0.6106973595125254,
"grad_norm": 0.9747077006239969,
"learning_rate": 3.5092627406298935e-06,
"loss": 0.1714,
"step": 451
},
{
"epoch": 0.6120514556533514,
"grad_norm": 0.9646009841505984,
"learning_rate": 3.488336428018293e-06,
"loss": 0.1686,
"step": 452
},
{
"epoch": 0.6134055517941774,
"grad_norm": 1.1374673045848798,
"learning_rate": 3.467439217762868e-06,
"loss": 0.1697,
"step": 453
},
{
"epoch": 0.6147596479350034,
"grad_norm": 0.9791286404845176,
"learning_rate": 3.4465715121740684e-06,
"loss": 0.1371,
"step": 454
},
{
"epoch": 0.6161137440758294,
"grad_norm": 0.9353453895769026,
"learning_rate": 3.4257337129943315e-06,
"loss": 0.1747,
"step": 455
},
{
"epoch": 0.6174678402166554,
"grad_norm": 0.9931105163272578,
"learning_rate": 3.404926221390332e-06,
"loss": 0.152,
"step": 456
},
{
"epoch": 0.6188219363574814,
"grad_norm": 0.7850621788003889,
"learning_rate": 3.3841494379452706e-06,
"loss": 0.1664,
"step": 457
},
{
"epoch": 0.6201760324983073,
"grad_norm": 0.8310172264571184,
"learning_rate": 3.3634037626511597e-06,
"loss": 0.1698,
"step": 458
},
{
"epoch": 0.6215301286391334,
"grad_norm": 0.7875734707943924,
"learning_rate": 3.3426895949011206e-06,
"loss": 0.1994,
"step": 459
},
{
"epoch": 0.6228842247799594,
"grad_norm": 0.910308226862476,
"learning_rate": 3.3220073334816928e-06,
"loss": 0.1626,
"step": 460
},
{
"epoch": 0.6242383209207854,
"grad_norm": 0.9987547301753712,
"learning_rate": 3.301357376565161e-06,
"loss": 0.1633,
"step": 461
},
{
"epoch": 0.6255924170616114,
"grad_norm": 0.9770148435339038,
"learning_rate": 3.280740121701888e-06,
"loss": 0.2137,
"step": 462
},
{
"epoch": 0.6269465132024373,
"grad_norm": 0.8463709983475592,
"learning_rate": 3.2601559658126624e-06,
"loss": 0.2045,
"step": 463
},
{
"epoch": 0.6283006093432634,
"grad_norm": 1.0760045338751967,
"learning_rate": 3.2396053051810515e-06,
"loss": 0.2418,
"step": 464
},
{
"epoch": 0.6296547054840894,
"grad_norm": 0.8596285475209885,
"learning_rate": 3.2190885354457797e-06,
"loss": 0.1625,
"step": 465
},
{
"epoch": 0.6310088016249153,
"grad_norm": 0.8281786331847009,
"learning_rate": 3.198606051593107e-06,
"loss": 0.1829,
"step": 466
},
{
"epoch": 0.6323628977657414,
"grad_norm": 0.900026158209928,
"learning_rate": 3.178158247949225e-06,
"loss": 0.1866,
"step": 467
},
{
"epoch": 0.6337169939065673,
"grad_norm": 0.8501792251206072,
"learning_rate": 3.157745518172669e-06,
"loss": 0.1588,
"step": 468
},
{
"epoch": 0.6350710900473934,
"grad_norm": 0.9328206797666032,
"learning_rate": 3.137368255246731e-06,
"loss": 0.1398,
"step": 469
},
{
"epoch": 0.6364251861882194,
"grad_norm": 0.8577144585087838,
"learning_rate": 3.117026851471905e-06,
"loss": 0.1764,
"step": 470
},
{
"epoch": 0.6377792823290453,
"grad_norm": 0.9065671481463856,
"learning_rate": 3.0967216984583288e-06,
"loss": 0.1801,
"step": 471
},
{
"epoch": 0.6391333784698714,
"grad_norm": 0.9085706722717214,
"learning_rate": 3.0764531871182422e-06,
"loss": 0.1452,
"step": 472
},
{
"epoch": 0.6404874746106973,
"grad_norm": 0.7246381935588333,
"learning_rate": 3.056221707658467e-06,
"loss": 0.1911,
"step": 473
},
{
"epoch": 0.6418415707515234,
"grad_norm": 0.8989547170826399,
"learning_rate": 3.036027649572889e-06,
"loss": 0.2099,
"step": 474
},
{
"epoch": 0.6431956668923493,
"grad_norm": 0.8313019913299642,
"learning_rate": 3.0158714016349657e-06,
"loss": 0.1752,
"step": 475
},
{
"epoch": 0.6445497630331753,
"grad_norm": 1.0117818889704953,
"learning_rate": 2.9957533518902376e-06,
"loss": 0.1421,
"step": 476
},
{
"epoch": 0.6459038591740014,
"grad_norm": 0.9517444237518378,
"learning_rate": 2.975673887648855e-06,
"loss": 0.1831,
"step": 477
},
{
"epoch": 0.6472579553148273,
"grad_norm": 1.0452749034776285,
"learning_rate": 2.955633395478129e-06,
"loss": 0.2005,
"step": 478
},
{
"epoch": 0.6486120514556534,
"grad_norm": 0.8233004574922531,
"learning_rate": 2.9356322611950826e-06,
"loss": 0.1584,
"step": 479
},
{
"epoch": 0.6499661475964793,
"grad_norm": 0.9703081283556748,
"learning_rate": 2.9156708698590273e-06,
"loss": 0.1257,
"step": 480
},
{
"epoch": 0.6513202437373053,
"grad_norm": 0.8571230869759529,
"learning_rate": 2.895749605764151e-06,
"loss": 0.1992,
"step": 481
},
{
"epoch": 0.6526743398781314,
"grad_norm": 0.8674721293982799,
"learning_rate": 2.875868852432112e-06,
"loss": 0.1976,
"step": 482
},
{
"epoch": 0.6540284360189573,
"grad_norm": 0.9875377769226037,
"learning_rate": 2.8560289926046625e-06,
"loss": 0.1504,
"step": 483
},
{
"epoch": 0.6553825321597834,
"grad_norm": 1.0777480990045858,
"learning_rate": 2.836230408236278e-06,
"loss": 0.1739,
"step": 484
},
{
"epoch": 0.6567366283006093,
"grad_norm": 0.9310552880618396,
"learning_rate": 2.8164734804868055e-06,
"loss": 0.1262,
"step": 485
},
{
"epoch": 0.6580907244414353,
"grad_norm": 1.0717331111337627,
"learning_rate": 2.7967585897141213e-06,
"loss": 0.1903,
"step": 486
},
{
"epoch": 0.6594448205822614,
"grad_norm": 1.008597081125778,
"learning_rate": 2.777086115466814e-06,
"loss": 0.143,
"step": 487
},
{
"epoch": 0.6607989167230873,
"grad_norm": 0.7396532461802736,
"learning_rate": 2.757456436476873e-06,
"loss": 0.1311,
"step": 488
},
{
"epoch": 0.6621530128639134,
"grad_norm": 0.8827988160796161,
"learning_rate": 2.7378699306523992e-06,
"loss": 0.1783,
"step": 489
},
{
"epoch": 0.6635071090047393,
"grad_norm": 1.0077964943136801,
"learning_rate": 2.7183269750703283e-06,
"loss": 0.1699,
"step": 490
},
{
"epoch": 0.6648612051455653,
"grad_norm": 0.895765133945653,
"learning_rate": 2.698827945969174e-06,
"loss": 0.1467,
"step": 491
},
{
"epoch": 0.6662153012863913,
"grad_norm": 0.7160118252865467,
"learning_rate": 2.67937321874178e-06,
"loss": 0.1356,
"step": 492
},
{
"epoch": 0.6675693974272173,
"grad_norm": 0.8898777393194023,
"learning_rate": 2.6599631679281e-06,
"loss": 0.1328,
"step": 493
},
{
"epoch": 0.6689234935680434,
"grad_norm": 0.7362694146835967,
"learning_rate": 2.640598167207984e-06,
"loss": 0.1359,
"step": 494
},
{
"epoch": 0.6702775897088693,
"grad_norm": 0.8334391707823784,
"learning_rate": 2.6212785893939728e-06,
"loss": 0.1485,
"step": 495
},
{
"epoch": 0.6716316858496953,
"grad_norm": 0.765528650625843,
"learning_rate": 2.602004806424144e-06,
"loss": 0.1352,
"step": 496
},
{
"epoch": 0.6729857819905213,
"grad_norm": 0.9240242492295755,
"learning_rate": 2.5827771893549286e-06,
"loss": 0.1271,
"step": 497
},
{
"epoch": 0.6743398781313473,
"grad_norm": 0.978722389203178,
"learning_rate": 2.563596108353987e-06,
"loss": 0.1766,
"step": 498
},
{
"epoch": 0.6756939742721734,
"grad_norm": 0.7988094774959305,
"learning_rate": 2.5444619326930664e-06,
"loss": 0.2104,
"step": 499
},
{
"epoch": 0.6770480704129993,
"grad_norm": 0.9859987789965777,
"learning_rate": 2.5253750307408996e-06,
"loss": 0.198,
"step": 500
},
{
"epoch": 0.6784021665538253,
"grad_norm": 0.6429567487285057,
"learning_rate": 2.506335769956113e-06,
"loss": 0.1241,
"step": 501
},
{
"epoch": 0.6797562626946513,
"grad_norm": 1.0583387999145726,
"learning_rate": 2.4873445168801502e-06,
"loss": 0.1795,
"step": 502
},
{
"epoch": 0.6811103588354773,
"grad_norm": 0.8712652403942699,
"learning_rate": 2.468401637130217e-06,
"loss": 0.1642,
"step": 503
},
{
"epoch": 0.6824644549763034,
"grad_norm": 0.714900582719695,
"learning_rate": 2.4495074953922425e-06,
"loss": 0.1656,
"step": 504
},
{
"epoch": 0.6838185511171293,
"grad_norm": 0.8548515473047144,
"learning_rate": 2.430662455413857e-06,
"loss": 0.1833,
"step": 505
},
{
"epoch": 0.6851726472579553,
"grad_norm": 0.9813156385175185,
"learning_rate": 2.4118668799973897e-06,
"loss": 0.1733,
"step": 506
},
{
"epoch": 0.6865267433987813,
"grad_norm": 0.780900861609635,
"learning_rate": 2.393121130992886e-06,
"loss": 0.1616,
"step": 507
},
{
"epoch": 0.6878808395396073,
"grad_norm": 1.0766405883907129,
"learning_rate": 2.3744255692911345e-06,
"loss": 0.1594,
"step": 508
},
{
"epoch": 0.6892349356804333,
"grad_norm": 0.831835188215008,
"learning_rate": 2.355780554816733e-06,
"loss": 0.15,
"step": 509
},
{
"epoch": 0.6905890318212593,
"grad_norm": 0.8792871797147107,
"learning_rate": 2.3371864465211434e-06,
"loss": 0.1086,
"step": 510
},
{
"epoch": 0.6919431279620853,
"grad_norm": 0.865372105591103,
"learning_rate": 2.31864360237579e-06,
"loss": 0.16,
"step": 511
},
{
"epoch": 0.6932972241029113,
"grad_norm": 0.8759256620541797,
"learning_rate": 2.3001523793651688e-06,
"loss": 0.1672,
"step": 512
},
{
"epoch": 0.6946513202437373,
"grad_norm": 0.7611576506431957,
"learning_rate": 2.281713133479966e-06,
"loss": 0.1571,
"step": 513
},
{
"epoch": 0.6960054163845633,
"grad_norm": 0.7480218314376521,
"learning_rate": 2.2633262197102133e-06,
"loss": 0.1487,
"step": 514
},
{
"epoch": 0.6973595125253893,
"grad_norm": 0.9751237391913985,
"learning_rate": 2.244991992038459e-06,
"loss": 0.158,
"step": 515
},
{
"epoch": 0.6987136086662153,
"grad_norm": 0.8420391985331888,
"learning_rate": 2.2267108034329343e-06,
"loss": 0.149,
"step": 516
},
{
"epoch": 0.7000677048070413,
"grad_norm": 0.9592644136610753,
"learning_rate": 2.208483005840774e-06,
"loss": 0.1745,
"step": 517
},
{
"epoch": 0.7014218009478673,
"grad_norm": 0.8192959742680409,
"learning_rate": 2.1903089501812363e-06,
"loss": 0.1563,
"step": 518
},
{
"epoch": 0.7027758970886933,
"grad_norm": 0.9556520507012928,
"learning_rate": 2.172188986338945e-06,
"loss": 0.16,
"step": 519
},
{
"epoch": 0.7041299932295193,
"grad_norm": 1.0144657135230208,
"learning_rate": 2.1541234631571533e-06,
"loss": 0.2052,
"step": 520
},
{
"epoch": 0.7054840893703453,
"grad_norm": 0.826674013679155,
"learning_rate": 2.136112728431038e-06,
"loss": 0.16,
"step": 521
},
{
"epoch": 0.7068381855111713,
"grad_norm": 0.8359658974700702,
"learning_rate": 2.118157128900984e-06,
"loss": 0.1653,
"step": 522
},
{
"epoch": 0.7081922816519973,
"grad_norm": 0.934884772166819,
"learning_rate": 2.100257010245927e-06,
"loss": 0.1461,
"step": 523
},
{
"epoch": 0.7095463777928233,
"grad_norm": 0.8065347707337489,
"learning_rate": 2.0824127170766904e-06,
"loss": 0.1649,
"step": 524
},
{
"epoch": 0.7109004739336493,
"grad_norm": 0.7332654341301846,
"learning_rate": 2.064624592929352e-06,
"loss": 0.1641,
"step": 525
},
{
"epoch": 0.7122545700744752,
"grad_norm": 0.7709818205915376,
"learning_rate": 2.0468929802586345e-06,
"loss": 0.1429,
"step": 526
},
{
"epoch": 0.7136086662153013,
"grad_norm": 0.7811238184755382,
"learning_rate": 2.0292182204313033e-06,
"loss": 0.1425,
"step": 527
},
{
"epoch": 0.7149627623561273,
"grad_norm": 0.8424111245040724,
"learning_rate": 2.0116006537196033e-06,
"loss": 0.1821,
"step": 528
},
{
"epoch": 0.7163168584969533,
"grad_norm": 1.0289276927583162,
"learning_rate": 1.994040619294703e-06,
"loss": 0.1651,
"step": 529
},
{
"epoch": 0.7176709546377793,
"grad_norm": 0.7844541917484446,
"learning_rate": 1.9765384552201683e-06,
"loss": 0.1673,
"step": 530
},
{
"epoch": 0.7190250507786052,
"grad_norm": 0.8529191584179351,
"learning_rate": 1.9590944984454514e-06,
"loss": 0.1821,
"step": 531
},
{
"epoch": 0.7203791469194313,
"grad_norm": 0.8838527354314403,
"learning_rate": 1.941709084799404e-06,
"loss": 0.199,
"step": 532
},
{
"epoch": 0.7217332430602573,
"grad_norm": 0.7817751665056095,
"learning_rate": 1.924382548983815e-06,
"loss": 0.168,
"step": 533
},
{
"epoch": 0.7230873392010833,
"grad_norm": 0.9100711551277519,
"learning_rate": 1.907115224566964e-06,
"loss": 0.1566,
"step": 534
},
{
"epoch": 0.7244414353419093,
"grad_norm": 1.1593625020685057,
"learning_rate": 1.8899074439771991e-06,
"loss": 0.2133,
"step": 535
},
{
"epoch": 0.7257955314827352,
"grad_norm": 1.203640883281713,
"learning_rate": 1.872759538496539e-06,
"loss": 0.164,
"step": 536
},
{
"epoch": 0.7271496276235613,
"grad_norm": 1.0402884488627961,
"learning_rate": 1.8556718382542933e-06,
"loss": 0.1905,
"step": 537
},
{
"epoch": 0.7285037237643873,
"grad_norm": 0.8238443338731649,
"learning_rate": 1.8386446722207113e-06,
"loss": 0.1877,
"step": 538
},
{
"epoch": 0.7298578199052133,
"grad_norm": 0.6733401188436547,
"learning_rate": 1.8216783682006445e-06,
"loss": 0.1312,
"step": 539
},
{
"epoch": 0.7312119160460393,
"grad_norm": 0.874450628869087,
"learning_rate": 1.804773252827231e-06,
"loss": 0.1477,
"step": 540
},
{
"epoch": 0.7325660121868652,
"grad_norm": 1.0742708606697025,
"learning_rate": 1.7879296515556182e-06,
"loss": 0.1423,
"step": 541
},
{
"epoch": 0.7339201083276913,
"grad_norm": 0.88936511888949,
"learning_rate": 1.7711478886566902e-06,
"loss": 0.1435,
"step": 542
},
{
"epoch": 0.7352742044685172,
"grad_norm": 0.8478534218140374,
"learning_rate": 1.7544282872108287e-06,
"loss": 0.1586,
"step": 543
},
{
"epoch": 0.7366283006093433,
"grad_norm": 1.03601989275073,
"learning_rate": 1.7377711691016885e-06,
"loss": 0.1529,
"step": 544
},
{
"epoch": 0.7379823967501693,
"grad_norm": 0.8036643443027366,
"learning_rate": 1.7211768550100034e-06,
"loss": 0.1225,
"step": 545
},
{
"epoch": 0.7393364928909952,
"grad_norm": 0.9125791465535871,
"learning_rate": 1.7046456644074127e-06,
"loss": 0.1989,
"step": 546
},
{
"epoch": 0.7406905890318213,
"grad_norm": 0.8175945530084422,
"learning_rate": 1.6881779155503103e-06,
"loss": 0.1918,
"step": 547
},
{
"epoch": 0.7420446851726472,
"grad_norm": 0.8924556720434464,
"learning_rate": 1.671773925473717e-06,
"loss": 0.1988,
"step": 548
},
{
"epoch": 0.7433987813134733,
"grad_norm": 1.009812657174965,
"learning_rate": 1.655434009985178e-06,
"loss": 0.167,
"step": 549
},
{
"epoch": 0.7447528774542993,
"grad_norm": 0.7649192955900349,
"learning_rate": 1.6391584836586833e-06,
"loss": 0.155,
"step": 550
},
{
"epoch": 0.7461069735951252,
"grad_norm": 0.7864619182208713,
"learning_rate": 1.6229476598286097e-06,
"loss": 0.1519,
"step": 551
},
{
"epoch": 0.7474610697359513,
"grad_norm": 0.9523554540226679,
"learning_rate": 1.6068018505836901e-06,
"loss": 0.1544,
"step": 552
},
{
"epoch": 0.7488151658767772,
"grad_norm": 0.9308166565300809,
"learning_rate": 1.590721366761004e-06,
"loss": 0.1709,
"step": 553
},
{
"epoch": 0.7501692620176033,
"grad_norm": 0.9859569961233892,
"learning_rate": 1.574706517939998e-06,
"loss": 0.1809,
"step": 554
},
{
"epoch": 0.7515233581584293,
"grad_norm": 0.9098445378362635,
"learning_rate": 1.5587576124365171e-06,
"loss": 0.1825,
"step": 555
},
{
"epoch": 0.7528774542992552,
"grad_norm": 0.8498371512718124,
"learning_rate": 1.542874957296876e-06,
"loss": 0.1433,
"step": 556
},
{
"epoch": 0.7542315504400813,
"grad_norm": 0.6560999269526281,
"learning_rate": 1.5270588582919477e-06,
"loss": 0.1324,
"step": 557
},
{
"epoch": 0.7555856465809072,
"grad_norm": 1.0483646360567538,
"learning_rate": 1.5113096199112676e-06,
"loss": 0.1958,
"step": 558
},
{
"epoch": 0.7569397427217333,
"grad_norm": 0.9474474720473122,
"learning_rate": 1.4956275453571883e-06,
"loss": 0.1571,
"step": 559
},
{
"epoch": 0.7582938388625592,
"grad_norm": 0.8773738362649157,
"learning_rate": 1.4800129365390282e-06,
"loss": 0.1714,
"step": 560
},
{
"epoch": 0.7596479350033852,
"grad_norm": 0.8286787799865905,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.1426,
"step": 561
},
{
"epoch": 0.7610020311442113,
"grad_norm": 1.0447162185133374,
"learning_rate": 1.448987317247741e-06,
"loss": 0.2036,
"step": 562
},
{
"epoch": 0.7623561272850372,
"grad_norm": 0.8804077593437535,
"learning_rate": 1.4335769040759185e-06,
"loss": 0.1807,
"step": 563
},
{
"epoch": 0.7637102234258633,
"grad_norm": 0.7979970967985389,
"learning_rate": 1.4182351512311237e-06,
"loss": 0.1592,
"step": 564
},
{
"epoch": 0.7650643195666892,
"grad_norm": 0.9944623851550066,
"learning_rate": 1.4029623540708443e-06,
"loss": 0.1396,
"step": 565
},
{
"epoch": 0.7664184157075152,
"grad_norm": 0.8069253672364819,
"learning_rate": 1.3877588066250452e-06,
"loss": 0.1418,
"step": 566
},
{
"epoch": 0.7677725118483413,
"grad_norm": 0.9803767727910536,
"learning_rate": 1.372624801590498e-06,
"loss": 0.1759,
"step": 567
},
{
"epoch": 0.7691266079891672,
"grad_norm": 1.0148509273862558,
"learning_rate": 1.3575606303251582e-06,
"loss": 0.1572,
"step": 568
},
{
"epoch": 0.7704807041299933,
"grad_norm": 0.8298319078396565,
"learning_rate": 1.3425665828425483e-06,
"loss": 0.2074,
"step": 569
},
{
"epoch": 0.7718348002708192,
"grad_norm": 0.7471240928046166,
"learning_rate": 1.3276429478061742e-06,
"loss": 0.1494,
"step": 570
},
{
"epoch": 0.7731888964116452,
"grad_norm": 0.7456007063522406,
"learning_rate": 1.312790012523978e-06,
"loss": 0.1873,
"step": 571
},
{
"epoch": 0.7745429925524713,
"grad_norm": 0.6884698164752326,
"learning_rate": 1.2980080629427904e-06,
"loss": 0.1293,
"step": 572
},
{
"epoch": 0.7758970886932972,
"grad_norm": 1.0105622262739957,
"learning_rate": 1.2832973836428387e-06,
"loss": 0.15,
"step": 573
},
{
"epoch": 0.7772511848341233,
"grad_norm": 0.8780268317440114,
"learning_rate": 1.2686582578322631e-06,
"loss": 0.1612,
"step": 574
},
{
"epoch": 0.7786052809749492,
"grad_norm": 1.0316670485881008,
"learning_rate": 1.2540909673416647e-06,
"loss": 0.1544,
"step": 575
},
{
"epoch": 0.7799593771157752,
"grad_norm": 0.9656272467166316,
"learning_rate": 1.2395957926186802e-06,
"loss": 0.1745,
"step": 576
},
{
"epoch": 0.7813134732566012,
"grad_norm": 0.9142011193873664,
"learning_rate": 1.2251730127225825e-06,
"loss": 0.1524,
"step": 577
},
{
"epoch": 0.7826675693974272,
"grad_norm": 0.8858819163335048,
"learning_rate": 1.2108229053189097e-06,
"loss": 0.1592,
"step": 578
},
{
"epoch": 0.7840216655382533,
"grad_norm": 1.153101739411035,
"learning_rate": 1.1965457466741159e-06,
"loss": 0.1798,
"step": 579
},
{
"epoch": 0.7853757616790792,
"grad_norm": 1.1132573753926343,
"learning_rate": 1.1823418116502566e-06,
"loss": 0.1565,
"step": 580
},
{
"epoch": 0.7867298578199052,
"grad_norm": 1.1913758521559024,
"learning_rate": 1.1682113736996943e-06,
"loss": 0.1518,
"step": 581
},
{
"epoch": 0.7880839539607312,
"grad_norm": 0.9470522999130591,
"learning_rate": 1.1541547048598383e-06,
"loss": 0.1644,
"step": 582
},
{
"epoch": 0.7894380501015572,
"grad_norm": 0.9632724018139464,
"learning_rate": 1.1401720757478995e-06,
"loss": 0.1729,
"step": 583
},
{
"epoch": 0.7907921462423833,
"grad_norm": 0.8864746264835872,
"learning_rate": 1.1262637555556905e-06,
"loss": 0.1439,
"step": 584
},
{
"epoch": 0.7921462423832092,
"grad_norm": 1.057421481242857,
"learning_rate": 1.1124300120444315e-06,
"loss": 0.126,
"step": 585
},
{
"epoch": 0.7935003385240352,
"grad_norm": 0.7025678365256196,
"learning_rate": 1.098671111539606e-06,
"loss": 0.1342,
"step": 586
},
{
"epoch": 0.7948544346648612,
"grad_norm": 0.8465000452800346,
"learning_rate": 1.0849873189258309e-06,
"loss": 0.1932,
"step": 587
},
{
"epoch": 0.7962085308056872,
"grad_norm": 1.0558235926053003,
"learning_rate": 1.0713788976417522e-06,
"loss": 0.1646,
"step": 588
},
{
"epoch": 0.7975626269465133,
"grad_norm": 0.7910168388174552,
"learning_rate": 1.0578461096749787e-06,
"loss": 0.1475,
"step": 589
},
{
"epoch": 0.7989167230873392,
"grad_norm": 0.8024590115170623,
"learning_rate": 1.044389215557034e-06,
"loss": 0.1496,
"step": 590
},
{
"epoch": 0.8002708192281652,
"grad_norm": 1.2680943598783014,
"learning_rate": 1.031008474358345e-06,
"loss": 0.1759,
"step": 591
},
{
"epoch": 0.8016249153689912,
"grad_norm": 0.9494469535373047,
"learning_rate": 1.0177041436832508e-06,
"loss": 0.1567,
"step": 592
},
{
"epoch": 0.8029790115098172,
"grad_norm": 0.9017524358791295,
"learning_rate": 1.0044764796650458e-06,
"loss": 0.1662,
"step": 593
},
{
"epoch": 0.8043331076506431,
"grad_norm": 0.8109800233977214,
"learning_rate": 9.913257369610473e-07,
"loss": 0.1812,
"step": 594
},
{
"epoch": 0.8056872037914692,
"grad_norm": 0.9085262058404419,
"learning_rate": 9.78252168747692e-07,
"loss": 0.1765,
"step": 595
},
{
"epoch": 0.8070412999322952,
"grad_norm": 0.8509703171867035,
"learning_rate": 9.652560267156647e-07,
"loss": 0.1917,
"step": 596
},
{
"epoch": 0.8083953960731212,
"grad_norm": 0.923128999787286,
"learning_rate": 9.523375610650509e-07,
"loss": 0.235,
"step": 597
},
{
"epoch": 0.8097494922139472,
"grad_norm": 0.8197761990522552,
"learning_rate": 9.394970205005178e-07,
"loss": 0.1447,
"step": 598
},
{
"epoch": 0.8111035883547731,
"grad_norm": 1.0190657855810292,
"learning_rate": 9.26734652226533e-07,
"loss": 0.1703,
"step": 599
},
{
"epoch": 0.8124576844955992,
"grad_norm": 0.8301957356124218,
"learning_rate": 9.140507019425981e-07,
"loss": 0.1321,
"step": 600
},
{
"epoch": 0.8138117806364252,
"grad_norm": 1.102789314784226,
"learning_rate": 9.014454138385225e-07,
"loss": 0.1362,
"step": 601
},
{
"epoch": 0.8151658767772512,
"grad_norm": 0.7292530336633652,
"learning_rate": 8.88919030589721e-07,
"loss": 0.1774,
"step": 602
},
{
"epoch": 0.8165199729180772,
"grad_norm": 0.7649185353132691,
"learning_rate": 8.764717933525386e-07,
"loss": 0.108,
"step": 603
},
{
"epoch": 0.8178740690589031,
"grad_norm": 1.0758784075311514,
"learning_rate": 8.641039417596181e-07,
"loss": 0.1446,
"step": 604
},
{
"epoch": 0.8192281651997292,
"grad_norm": 0.8676990449422515,
"learning_rate": 8.518157139152755e-07,
"loss": 0.1482,
"step": 605
},
{
"epoch": 0.8205822613405552,
"grad_norm": 1.086119982711249,
"learning_rate": 8.39607346390921e-07,
"loss": 0.1974,
"step": 606
},
{
"epoch": 0.8219363574813812,
"grad_norm": 0.6977664028669072,
"learning_rate": 8.274790742205053e-07,
"loss": 0.1219,
"step": 607
},
{
"epoch": 0.8232904536222072,
"grad_norm": 0.9909467453210977,
"learning_rate": 8.154311308959911e-07,
"loss": 0.1373,
"step": 608
},
{
"epoch": 0.8246445497630331,
"grad_norm": 0.8237390733802635,
"learning_rate": 8.03463748362861e-07,
"loss": 0.1426,
"step": 609
},
{
"epoch": 0.8259986459038592,
"grad_norm": 0.7908534273954079,
"learning_rate": 7.915771570156555e-07,
"loss": 0.1654,
"step": 610
},
{
"epoch": 0.8273527420446851,
"grad_norm": 1.1434632665159954,
"learning_rate": 7.797715856935284e-07,
"loss": 0.2302,
"step": 611
},
{
"epoch": 0.8287068381855112,
"grad_norm": 0.7946881280367531,
"learning_rate": 7.680472616758467e-07,
"loss": 0.1746,
"step": 612
},
{
"epoch": 0.8300609343263372,
"grad_norm": 0.847682725716897,
"learning_rate": 7.564044106778162e-07,
"loss": 0.2033,
"step": 613
},
{
"epoch": 0.8314150304671631,
"grad_norm": 0.9418692837025284,
"learning_rate": 7.448432568461345e-07,
"loss": 0.1591,
"step": 614
},
{
"epoch": 0.8327691266079892,
"grad_norm": 0.957079428755692,
"learning_rate": 7.333640227546734e-07,
"loss": 0.1065,
"step": 615
},
{
"epoch": 0.8341232227488151,
"grad_norm": 1.062583687237453,
"learning_rate": 7.219669294002002e-07,
"loss": 0.2117,
"step": 616
},
{
"epoch": 0.8354773188896412,
"grad_norm": 0.9983157916993027,
"learning_rate": 7.106521961981161e-07,
"loss": 0.1477,
"step": 617
},
{
"epoch": 0.8368314150304672,
"grad_norm": 1.0043585174128125,
"learning_rate": 6.994200409782354e-07,
"loss": 0.1722,
"step": 618
},
{
"epoch": 0.8381855111712931,
"grad_norm": 0.992847791234847,
"learning_rate": 6.882706799805916e-07,
"loss": 0.1449,
"step": 619
},
{
"epoch": 0.8395396073121192,
"grad_norm": 0.8968979820235704,
"learning_rate": 6.772043278512747e-07,
"loss": 0.1541,
"step": 620
},
{
"epoch": 0.8408937034529451,
"grad_norm": 0.7659571414300276,
"learning_rate": 6.662211976382987e-07,
"loss": 0.1569,
"step": 621
},
{
"epoch": 0.8422477995937712,
"grad_norm": 0.9530184421321656,
"learning_rate": 6.553215007874986e-07,
"loss": 0.2171,
"step": 622
},
{
"epoch": 0.8436018957345972,
"grad_norm": 0.8085681048716992,
"learning_rate": 6.445054471384627e-07,
"loss": 0.1627,
"step": 623
},
{
"epoch": 0.8449559918754231,
"grad_norm": 1.1513346523607006,
"learning_rate": 6.337732449204886e-07,
"loss": 0.1971,
"step": 624
},
{
"epoch": 0.8463100880162492,
"grad_norm": 0.9290862956073291,
"learning_rate": 6.231251007485789e-07,
"loss": 0.1735,
"step": 625
},
{
"epoch": 0.8476641841570751,
"grad_norm": 0.8230644430785699,
"learning_rate": 6.125612196194591e-07,
"loss": 0.1462,
"step": 626
},
{
"epoch": 0.8490182802979012,
"grad_norm": 0.7332437563473756,
"learning_rate": 6.020818049076371e-07,
"loss": 0.1314,
"step": 627
},
{
"epoch": 0.8503723764387271,
"grad_norm": 0.896133426059931,
"learning_rate": 5.916870583614792e-07,
"loss": 0.1659,
"step": 628
},
{
"epoch": 0.8517264725795531,
"grad_norm": 0.9844465279941125,
"learning_rate": 5.813771800993351e-07,
"loss": 0.1285,
"step": 629
},
{
"epoch": 0.8530805687203792,
"grad_norm": 1.0793385360508658,
"learning_rate": 5.711523686056769e-07,
"loss": 0.1795,
"step": 630
},
{
"epoch": 0.8544346648612051,
"grad_norm": 0.7964413760869411,
"learning_rate": 5.610128207272841e-07,
"loss": 0.1713,
"step": 631
},
{
"epoch": 0.8557887610020312,
"grad_norm": 1.0567491635722657,
"learning_rate": 5.509587316694537e-07,
"loss": 0.1847,
"step": 632
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.9508923229360245,
"learning_rate": 5.409902949922374e-07,
"loss": 0.1289,
"step": 633
},
{
"epoch": 0.8584969532836831,
"grad_norm": 0.8773375736525768,
"learning_rate": 5.311077026067196e-07,
"loss": 0.1891,
"step": 634
},
{
"epoch": 0.8598510494245092,
"grad_norm": 0.7712902759070395,
"learning_rate": 5.21311144771321e-07,
"loss": 0.1504,
"step": 635
},
{
"epoch": 0.8612051455653351,
"grad_norm": 0.8721730621406611,
"learning_rate": 5.116008100881348e-07,
"loss": 0.1205,
"step": 636
},
{
"epoch": 0.8625592417061612,
"grad_norm": 0.7911360388279146,
"learning_rate": 5.019768854992984e-07,
"loss": 0.1191,
"step": 637
},
{
"epoch": 0.8639133378469871,
"grad_norm": 0.8471120557866042,
"learning_rate": 4.924395562833934e-07,
"loss": 0.1689,
"step": 638
},
{
"epoch": 0.8652674339878131,
"grad_norm": 0.8809005387688132,
"learning_rate": 4.829890060518766e-07,
"loss": 0.1176,
"step": 639
},
{
"epoch": 0.8666215301286392,
"grad_norm": 0.8480821404391728,
"learning_rate": 4.736254167455473e-07,
"loss": 0.1439,
"step": 640
},
{
"epoch": 0.8679756262694651,
"grad_norm": 0.8367277555953317,
"learning_rate": 4.6434896863104427e-07,
"loss": 0.1332,
"step": 641
},
{
"epoch": 0.8693297224102912,
"grad_norm": 0.8135772302364183,
"learning_rate": 4.551598402973761e-07,
"loss": 0.1379,
"step": 642
},
{
"epoch": 0.8706838185511171,
"grad_norm": 1.0170206424906376,
"learning_rate": 4.46058208652479e-07,
"loss": 0.1593,
"step": 643
},
{
"epoch": 0.8720379146919431,
"grad_norm": 0.8976897096216934,
"learning_rate": 4.370442489198179e-07,
"loss": 0.1609,
"step": 644
},
{
"epoch": 0.8733920108327691,
"grad_norm": 0.9586643409595541,
"learning_rate": 4.2811813463500605e-07,
"loss": 0.1573,
"step": 645
},
{
"epoch": 0.8747461069735951,
"grad_norm": 0.9262839537695102,
"learning_rate": 4.1928003764246936e-07,
"loss": 0.1286,
"step": 646
},
{
"epoch": 0.8761002031144212,
"grad_norm": 0.9640859981007688,
"learning_rate": 4.105301280921348e-07,
"loss": 0.1439,
"step": 647
},
{
"epoch": 0.8774542992552471,
"grad_norm": 1.0290960863591725,
"learning_rate": 4.018685744361539e-07,
"loss": 0.1325,
"step": 648
},
{
"epoch": 0.8788083953960731,
"grad_norm": 0.7401998087864601,
"learning_rate": 3.932955434256658e-07,
"loss": 0.0975,
"step": 649
},
{
"epoch": 0.8801624915368991,
"grad_norm": 0.7194519181258913,
"learning_rate": 3.84811200107581e-07,
"loss": 0.074,
"step": 650
},
{
"epoch": 0.8815165876777251,
"grad_norm": 0.9925206012728999,
"learning_rate": 3.7641570782140424e-07,
"loss": 0.204,
"step": 651
},
{
"epoch": 0.8828706838185512,
"grad_norm": 0.9154916121563816,
"learning_rate": 3.681092281960935e-07,
"loss": 0.1983,
"step": 652
},
{
"epoch": 0.8842247799593771,
"grad_norm": 0.9527621958167523,
"learning_rate": 3.598919211469454e-07,
"loss": 0.152,
"step": 653
},
{
"epoch": 0.8855788761002031,
"grad_norm": 1.109411296451597,
"learning_rate": 3.517639448725163e-07,
"loss": 0.1871,
"step": 654
},
{
"epoch": 0.8869329722410291,
"grad_norm": 0.7632172195548418,
"learning_rate": 3.437254558515818e-07,
"loss": 0.1304,
"step": 655
},
{
"epoch": 0.8882870683818551,
"grad_norm": 0.733767599608247,
"learning_rate": 3.357766088401149e-07,
"loss": 0.1508,
"step": 656
},
{
"epoch": 0.8896411645226812,
"grad_norm": 0.7997429308289891,
"learning_rate": 3.279175568683146e-07,
"loss": 0.1885,
"step": 657
},
{
"epoch": 0.8909952606635071,
"grad_norm": 0.858789719675283,
"learning_rate": 3.2014845123765734e-07,
"loss": 0.1527,
"step": 658
},
{
"epoch": 0.8923493568043331,
"grad_norm": 0.9190640557237344,
"learning_rate": 3.1246944151798296e-07,
"loss": 0.1633,
"step": 659
},
{
"epoch": 0.8937034529451591,
"grad_norm": 0.9823062126026297,
"learning_rate": 3.048806755446182e-07,
"loss": 0.1596,
"step": 660
},
{
"epoch": 0.8950575490859851,
"grad_norm": 0.8111976351650877,
"learning_rate": 2.973822994155262e-07,
"loss": 0.1446,
"step": 661
},
{
"epoch": 0.896411645226811,
"grad_norm": 0.9577556936376426,
"learning_rate": 2.8997445748849715e-07,
"loss": 0.1887,
"step": 662
},
{
"epoch": 0.8977657413676371,
"grad_norm": 0.7654559373406858,
"learning_rate": 2.826572923783688e-07,
"loss": 0.161,
"step": 663
},
{
"epoch": 0.8991198375084631,
"grad_norm": 0.9668484225993964,
"learning_rate": 2.7543094495427913e-07,
"loss": 0.1787,
"step": 664
},
{
"epoch": 0.9004739336492891,
"grad_norm": 0.7407556180154452,
"learning_rate": 2.6829555433695585e-07,
"loss": 0.1355,
"step": 665
},
{
"epoch": 0.9018280297901151,
"grad_norm": 1.0038786188102993,
"learning_rate": 2.612512578960391e-07,
"loss": 0.1809,
"step": 666
},
{
"epoch": 0.903182125930941,
"grad_norm": 0.953827571995608,
"learning_rate": 2.542981912474335e-07,
"loss": 0.1395,
"step": 667
},
{
"epoch": 0.9045362220717671,
"grad_norm": 0.9298090401188408,
"learning_rate": 2.474364882507002e-07,
"loss": 0.149,
"step": 668
},
{
"epoch": 0.9058903182125931,
"grad_norm": 0.9977184030296391,
"learning_rate": 2.406662810064786e-07,
"loss": 0.1397,
"step": 669
},
{
"epoch": 0.9072444143534191,
"grad_norm": 0.8234901813464861,
"learning_rate": 2.3398769985394422e-07,
"loss": 0.155,
"step": 670
},
{
"epoch": 0.9085985104942451,
"grad_norm": 0.7792494874881529,
"learning_rate": 2.2740087336829586e-07,
"loss": 0.1459,
"step": 671
},
{
"epoch": 0.909952606635071,
"grad_norm": 1.0184529018694468,
"learning_rate": 2.2090592835828817e-07,
"loss": 0.1572,
"step": 672
},
{
"epoch": 0.9113067027758971,
"grad_norm": 0.8068482199106803,
"learning_rate": 2.145029898637807e-07,
"loss": 0.155,
"step": 673
},
{
"epoch": 0.9126607989167231,
"grad_norm": 0.9862107222155843,
"learning_rate": 2.081921811533366e-07,
"loss": 0.2064,
"step": 674
},
{
"epoch": 0.9140148950575491,
"grad_norm": 0.7265750219835868,
"learning_rate": 2.0197362372184782e-07,
"loss": 0.1442,
"step": 675
},
{
"epoch": 0.9153689911983751,
"grad_norm": 0.7559974069531963,
"learning_rate": 1.958474372881969e-07,
"loss": 0.1258,
"step": 676
},
{
"epoch": 0.916723087339201,
"grad_norm": 0.7408372985614482,
"learning_rate": 1.8981373979295226e-07,
"loss": 0.142,
"step": 677
},
{
"epoch": 0.9180771834800271,
"grad_norm": 0.9608350066309304,
"learning_rate": 1.83872647396095e-07,
"loss": 0.2019,
"step": 678
},
{
"epoch": 0.919431279620853,
"grad_norm": 0.9011191290839695,
"learning_rate": 1.780242744747862e-07,
"loss": 0.1577,
"step": 679
},
{
"epoch": 0.9207853757616791,
"grad_norm": 0.8039010974882403,
"learning_rate": 1.722687336211626e-07,
"loss": 0.1422,
"step": 680
},
{
"epoch": 0.9221394719025051,
"grad_norm": 0.7568610226010065,
"learning_rate": 1.6660613564016992e-07,
"loss": 0.1379,
"step": 681
},
{
"epoch": 0.923493568043331,
"grad_norm": 1.090089294973423,
"learning_rate": 1.6103658954742917e-07,
"loss": 0.1846,
"step": 682
},
{
"epoch": 0.9248476641841571,
"grad_norm": 1.0551239948466493,
"learning_rate": 1.5556020256713823e-07,
"loss": 0.1313,
"step": 683
},
{
"epoch": 0.926201760324983,
"grad_norm": 0.8620277827708585,
"learning_rate": 1.5017708013000787e-07,
"loss": 0.17,
"step": 684
},
{
"epoch": 0.9275558564658091,
"grad_norm": 0.98129855677002,
"learning_rate": 1.4488732587123078e-07,
"loss": 0.1646,
"step": 685
},
{
"epoch": 0.9289099526066351,
"grad_norm": 0.9854464091612312,
"learning_rate": 1.3969104162848912e-07,
"loss": 0.1686,
"step": 686
},
{
"epoch": 0.930264048747461,
"grad_norm": 0.9045285861475807,
"learning_rate": 1.3458832743998894e-07,
"loss": 0.1702,
"step": 687
},
{
"epoch": 0.9316181448882871,
"grad_norm": 0.7723383286218737,
"learning_rate": 1.2957928154254174e-07,
"loss": 0.1803,
"step": 688
},
{
"epoch": 0.932972241029113,
"grad_norm": 0.7827682969554892,
"learning_rate": 1.2466400036966475e-07,
"loss": 0.1412,
"step": 689
},
{
"epoch": 0.9343263371699391,
"grad_norm": 0.8448319348134118,
"learning_rate": 1.1984257854973148e-07,
"loss": 0.1643,
"step": 690
},
{
"epoch": 0.9356804333107651,
"grad_norm": 0.7422214238572734,
"learning_rate": 1.1511510890414579e-07,
"loss": 0.1612,
"step": 691
},
{
"epoch": 0.937034529451591,
"grad_norm": 0.740675842792783,
"learning_rate": 1.1048168244555513e-07,
"loss": 0.1666,
"step": 692
},
{
"epoch": 0.9383886255924171,
"grad_norm": 0.9025732659231845,
"learning_rate": 1.0594238837610071e-07,
"loss": 0.1721,
"step": 693
},
{
"epoch": 0.939742721733243,
"grad_norm": 0.8073371610377824,
"learning_rate": 1.0149731408569951e-07,
"loss": 0.1421,
"step": 694
},
{
"epoch": 0.9410968178740691,
"grad_norm": 0.913842718835237,
"learning_rate": 9.71465451503606e-08,
"loss": 0.1582,
"step": 695
},
{
"epoch": 0.942450914014895,
"grad_norm": 0.8337199028366609,
"learning_rate": 9.289016533053696e-08,
"loss": 0.1486,
"step": 696
},
{
"epoch": 0.943805010155721,
"grad_norm": 0.75256023589629,
"learning_rate": 8.872825656951579e-08,
"loss": 0.1747,
"step": 697
},
{
"epoch": 0.9451591062965471,
"grad_norm": 0.936827364092228,
"learning_rate": 8.46608989918396e-08,
"loss": 0.1417,
"step": 698
},
{
"epoch": 0.946513202437373,
"grad_norm": 0.8949365811281319,
"learning_rate": 8.068817090176152e-08,
"loss": 0.153,
"step": 699
},
{
"epoch": 0.9478672985781991,
"grad_norm": 1.0437596642065827,
"learning_rate": 7.681014878174187e-08,
"loss": 0.1693,
"step": 700
},
{
"epoch": 0.949221394719025,
"grad_norm": 0.7789595445602694,
"learning_rate": 7.302690729097284e-08,
"loss": 0.1522,
"step": 701
},
{
"epoch": 0.950575490859851,
"grad_norm": 0.8940623265854092,
"learning_rate": 6.933851926394175e-08,
"loss": 0.1286,
"step": 702
},
{
"epoch": 0.9519295870006771,
"grad_norm": 1.142446406589418,
"learning_rate": 6.574505570902834e-08,
"loss": 0.1395,
"step": 703
},
{
"epoch": 0.953283683141503,
"grad_norm": 1.0612372846936757,
"learning_rate": 6.224658580713971e-08,
"loss": 0.1506,
"step": 704
},
{
"epoch": 0.9546377792823291,
"grad_norm": 1.013425855017681,
"learning_rate": 5.884317691037644e-08,
"loss": 0.1871,
"step": 705
},
{
"epoch": 0.955991875423155,
"grad_norm": 0.7394374943850645,
"learning_rate": 5.5534894540736907e-08,
"loss": 0.1281,
"step": 706
},
{
"epoch": 0.957345971563981,
"grad_norm": 0.9540536984878906,
"learning_rate": 5.232180238885609e-08,
"loss": 0.143,
"step": 707
},
{
"epoch": 0.9587000677048071,
"grad_norm": 0.779577635592253,
"learning_rate": 4.920396231277713e-08,
"loss": 0.1413,
"step": 708
},
{
"epoch": 0.960054163845633,
"grad_norm": 0.9400896572907169,
"learning_rate": 4.618143433676392e-08,
"loss": 0.1838,
"step": 709
},
{
"epoch": 0.9614082599864591,
"grad_norm": 0.7855375755976217,
"learning_rate": 4.325427665014314e-08,
"loss": 0.1393,
"step": 710
},
{
"epoch": 0.962762356127285,
"grad_norm": 1.0950947385510803,
"learning_rate": 4.042254560618575e-08,
"loss": 0.1837,
"step": 711
},
{
"epoch": 0.964116452268111,
"grad_norm": 0.8654573487033121,
"learning_rate": 3.76862957210189e-08,
"loss": 0.1904,
"step": 712
},
{
"epoch": 0.965470548408937,
"grad_norm": 1.050450057508448,
"learning_rate": 3.504557967258127e-08,
"loss": 0.1755,
"step": 713
},
{
"epoch": 0.966824644549763,
"grad_norm": 1.023119496158692,
"learning_rate": 3.2500448299603306e-08,
"loss": 0.2075,
"step": 714
},
{
"epoch": 0.9681787406905891,
"grad_norm": 0.8314547854411589,
"learning_rate": 3.005095060063301e-08,
"loss": 0.1861,
"step": 715
},
{
"epoch": 0.969532836831415,
"grad_norm": 0.8079908672851998,
"learning_rate": 2.7697133733091686e-08,
"loss": 0.1679,
"step": 716
},
{
"epoch": 0.970886932972241,
"grad_norm": 0.8968978076423817,
"learning_rate": 2.5439043012363552e-08,
"loss": 0.192,
"step": 717
},
{
"epoch": 0.972241029113067,
"grad_norm": 0.849822458333106,
"learning_rate": 2.3276721910926447e-08,
"loss": 0.1568,
"step": 718
},
{
"epoch": 0.973595125253893,
"grad_norm": 0.9849472347334293,
"learning_rate": 2.1210212057512502e-08,
"loss": 0.1682,
"step": 719
},
{
"epoch": 0.9749492213947191,
"grad_norm": 0.9886969907696579,
"learning_rate": 1.923955323630877e-08,
"loss": 0.1981,
"step": 720
},
{
"epoch": 0.976303317535545,
"grad_norm": 0.8549929136152121,
"learning_rate": 1.7364783386190075e-08,
"loss": 0.1615,
"step": 721
},
{
"epoch": 0.977657413676371,
"grad_norm": 1.027574343041217,
"learning_rate": 1.5585938599989027e-08,
"loss": 0.1795,
"step": 722
},
{
"epoch": 0.979011509817197,
"grad_norm": 0.8370405769917955,
"learning_rate": 1.390305312380047e-08,
"loss": 0.1915,
"step": 723
},
{
"epoch": 0.980365605958023,
"grad_norm": 0.8202833272442734,
"learning_rate": 1.2316159356323132e-08,
"loss": 0.1664,
"step": 724
},
{
"epoch": 0.9817197020988491,
"grad_norm": 1.0365081702442294,
"learning_rate": 1.0825287848235666e-08,
"loss": 0.1743,
"step": 725
},
{
"epoch": 0.983073798239675,
"grad_norm": 0.9499125012394468,
"learning_rate": 9.430467301607683e-09,
"loss": 0.1379,
"step": 726
},
{
"epoch": 0.984427894380501,
"grad_norm": 0.9836812744718438,
"learning_rate": 8.131724569347965e-09,
"loss": 0.1493,
"step": 727
},
{
"epoch": 0.985781990521327,
"grad_norm": 1.2302423950361594,
"learning_rate": 6.929084654688223e-09,
"loss": 0.1494,
"step": 728
},
{
"epoch": 0.987136086662153,
"grad_norm": 0.9716945962250909,
"learning_rate": 5.822570710699582e-09,
"loss": 0.1327,
"step": 729
},
{
"epoch": 0.988490182802979,
"grad_norm": 1.0808054852232092,
"learning_rate": 4.8122040398496105e-09,
"loss": 0.201,
"step": 730
},
{
"epoch": 0.989844278943805,
"grad_norm": 0.910060114520982,
"learning_rate": 3.898004093588759e-09,
"loss": 0.144,
"step": 731
},
{
"epoch": 0.991198375084631,
"grad_norm": 0.8261598315155959,
"learning_rate": 3.0799884719795448e-09,
"loss": 0.1861,
"step": 732
},
{
"epoch": 0.992552471225457,
"grad_norm": 0.7275162003167082,
"learning_rate": 2.358172923355717e-09,
"loss": 0.1287,
"step": 733
},
{
"epoch": 0.993906567366283,
"grad_norm": 0.7793579401930688,
"learning_rate": 1.7325713440180525e-09,
"loss": 0.1456,
"step": 734
},
{
"epoch": 0.995260663507109,
"grad_norm": 0.7762445980687355,
"learning_rate": 1.2031957779701231e-09,
"loss": 0.151,
"step": 735
},
{
"epoch": 0.996614759647935,
"grad_norm": 0.8667813119888734,
"learning_rate": 7.700564166834845e-10,
"loss": 0.17,
"step": 736
},
{
"epoch": 0.997968855788761,
"grad_norm": 0.8597669318682817,
"learning_rate": 4.3316159890172126e-10,
"loss": 0.1977,
"step": 737
},
{
"epoch": 0.999322951929587,
"grad_norm": 0.7899670827381629,
"learning_rate": 1.9251781048168494e-10,
"loss": 0.1363,
"step": 738
},
{
"epoch": 1.0,
"grad_norm": 1.10272243697496,
"learning_rate": 4.8129684266928765e-11,
"loss": 0.128,
"step": 739
}
],
"logging_steps": 1.0,
"max_steps": 739,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 708686067433472.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}