QueryAware_0.5 / trainer_state.json
AmeenAli023's picture
Duplicate from AmeenAli023/QueryAware
1f81bdc
Raw
History Blame Contribute Delete
88.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.0029388599574453078,
"eval_steps": 500,
"global_step": 500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 5.877719914890616e-06,
"grad_norm": 42.02911487791772,
"learning_rate": 0.0,
"loss": 2.345,
"step": 1
},
{
"epoch": 1.1755439829781232e-05,
"grad_norm": 50.85266375433337,
"learning_rate": 1.958863858961802e-09,
"loss": 2.6978,
"step": 2
},
{
"epoch": 1.763315974467185e-05,
"grad_norm": 47.291375384339034,
"learning_rate": 3.917727717923604e-09,
"loss": 2.3933,
"step": 3
},
{
"epoch": 2.3510879659562463e-05,
"grad_norm": 46.527545509164675,
"learning_rate": 5.876591576885407e-09,
"loss": 2.44,
"step": 4
},
{
"epoch": 2.938859957445308e-05,
"grad_norm": 46.977570907906056,
"learning_rate": 7.835455435847208e-09,
"loss": 2.5181,
"step": 5
},
{
"epoch": 3.52663194893437e-05,
"grad_norm": 36.871286445588595,
"learning_rate": 9.794319294809012e-09,
"loss": 2.6618,
"step": 6
},
{
"epoch": 4.114403940423431e-05,
"grad_norm": 46.385420313464735,
"learning_rate": 1.1753183153770814e-08,
"loss": 2.6994,
"step": 7
},
{
"epoch": 4.702175931912493e-05,
"grad_norm": 39.33244145538988,
"learning_rate": 1.3712047012732617e-08,
"loss": 2.2604,
"step": 8
},
{
"epoch": 5.289947923401554e-05,
"grad_norm": 42.2604654612667,
"learning_rate": 1.5670910871694417e-08,
"loss": 2.5037,
"step": 9
},
{
"epoch": 5.877719914890616e-05,
"grad_norm": 53.35405279929954,
"learning_rate": 1.7629774730656222e-08,
"loss": 2.7682,
"step": 10
},
{
"epoch": 6.465491906379677e-05,
"grad_norm": 61.096031238374245,
"learning_rate": 1.9588638589618025e-08,
"loss": 2.5123,
"step": 11
},
{
"epoch": 7.05326389786874e-05,
"grad_norm": 34.35305577230433,
"learning_rate": 2.1547502448579827e-08,
"loss": 2.4231,
"step": 12
},
{
"epoch": 7.6410358893578e-05,
"grad_norm": 46.32288420684808,
"learning_rate": 2.350636630754163e-08,
"loss": 2.5667,
"step": 13
},
{
"epoch": 8.228807880846862e-05,
"grad_norm": 39.6869121212424,
"learning_rate": 2.546523016650343e-08,
"loss": 2.6537,
"step": 14
},
{
"epoch": 8.816579872335923e-05,
"grad_norm": 38.4473023992043,
"learning_rate": 2.7424094025465233e-08,
"loss": 2.5401,
"step": 15
},
{
"epoch": 9.404351863824985e-05,
"grad_norm": 35.67137817810734,
"learning_rate": 2.9382957884427032e-08,
"loss": 2.2622,
"step": 16
},
{
"epoch": 9.992123855314047e-05,
"grad_norm": 33.63114184351338,
"learning_rate": 3.1341821743388834e-08,
"loss": 2.5217,
"step": 17
},
{
"epoch": 0.00010579895846803108,
"grad_norm": 39.43620507790292,
"learning_rate": 3.3300685602350636e-08,
"loss": 2.4657,
"step": 18
},
{
"epoch": 0.0001116766783829217,
"grad_norm": 45.917969482317524,
"learning_rate": 3.5259549461312445e-08,
"loss": 2.3468,
"step": 19
},
{
"epoch": 0.00011755439829781231,
"grad_norm": 38.078954655149694,
"learning_rate": 3.721841332027425e-08,
"loss": 2.508,
"step": 20
},
{
"epoch": 0.00012343211821270293,
"grad_norm": 34.493184716086255,
"learning_rate": 3.917727717923605e-08,
"loss": 2.2334,
"step": 21
},
{
"epoch": 0.00012930983812759354,
"grad_norm": 46.491961411990125,
"learning_rate": 4.1136141038197845e-08,
"loss": 2.2112,
"step": 22
},
{
"epoch": 0.00013518755804248415,
"grad_norm": 48.89994737190844,
"learning_rate": 4.3095004897159653e-08,
"loss": 2.4262,
"step": 23
},
{
"epoch": 0.0001410652779573748,
"grad_norm": 30.579301696325047,
"learning_rate": 4.505386875612145e-08,
"loss": 2.7459,
"step": 24
},
{
"epoch": 0.0001469429978722654,
"grad_norm": 39.08437638852243,
"learning_rate": 4.701273261508326e-08,
"loss": 2.5459,
"step": 25
},
{
"epoch": 0.000152820717787156,
"grad_norm": 56.085123066265226,
"learning_rate": 4.897159647404506e-08,
"loss": 2.7213,
"step": 26
},
{
"epoch": 0.00015869843770204662,
"grad_norm": 36.85335937419735,
"learning_rate": 5.093046033300686e-08,
"loss": 2.287,
"step": 27
},
{
"epoch": 0.00016457615761693723,
"grad_norm": 34.87118828311202,
"learning_rate": 5.2889324191968664e-08,
"loss": 2.3702,
"step": 28
},
{
"epoch": 0.00017045387753182785,
"grad_norm": 43.175332148056896,
"learning_rate": 5.4848188050930466e-08,
"loss": 2.2459,
"step": 29
},
{
"epoch": 0.00017633159744671846,
"grad_norm": 32.266335142325374,
"learning_rate": 5.680705190989227e-08,
"loss": 2.5008,
"step": 30
},
{
"epoch": 0.0001822093173616091,
"grad_norm": 43.65661992394771,
"learning_rate": 5.8765915768854064e-08,
"loss": 2.6555,
"step": 31
},
{
"epoch": 0.0001880870372764997,
"grad_norm": 43.84715059263309,
"learning_rate": 6.072477962781587e-08,
"loss": 2.3591,
"step": 32
},
{
"epoch": 0.00019396475719139032,
"grad_norm": 43.406252511374625,
"learning_rate": 6.268364348677767e-08,
"loss": 2.0988,
"step": 33
},
{
"epoch": 0.00019984247710628093,
"grad_norm": 37.561883875948865,
"learning_rate": 6.464250734573948e-08,
"loss": 2.4671,
"step": 34
},
{
"epoch": 0.00020572019702117154,
"grad_norm": 49.1466070616366,
"learning_rate": 6.660137120470127e-08,
"loss": 2.532,
"step": 35
},
{
"epoch": 0.00021159791693606215,
"grad_norm": 39.32458572492562,
"learning_rate": 6.856023506366308e-08,
"loss": 2.5906,
"step": 36
},
{
"epoch": 0.00021747563685095277,
"grad_norm": 45.58644039702116,
"learning_rate": 7.051909892262489e-08,
"loss": 2.7685,
"step": 37
},
{
"epoch": 0.0002233533567658434,
"grad_norm": 34.380760113202136,
"learning_rate": 7.247796278158669e-08,
"loss": 2.2058,
"step": 38
},
{
"epoch": 0.00022923107668073402,
"grad_norm": 33.75738146740774,
"learning_rate": 7.44368266405485e-08,
"loss": 2.1421,
"step": 39
},
{
"epoch": 0.00023510879659562463,
"grad_norm": 58.77325378288631,
"learning_rate": 7.639569049951029e-08,
"loss": 2.173,
"step": 40
},
{
"epoch": 0.00024098651651051524,
"grad_norm": 48.68423052986121,
"learning_rate": 7.83545543584721e-08,
"loss": 2.4029,
"step": 41
},
{
"epoch": 0.00024686423642540585,
"grad_norm": 49.077656798823305,
"learning_rate": 8.031341821743391e-08,
"loss": 2.5765,
"step": 42
},
{
"epoch": 0.00025274195634029646,
"grad_norm": 41.20173204960673,
"learning_rate": 8.227228207639569e-08,
"loss": 2.3414,
"step": 43
},
{
"epoch": 0.0002586196762551871,
"grad_norm": 35.10746373400446,
"learning_rate": 8.42311459353575e-08,
"loss": 2.0016,
"step": 44
},
{
"epoch": 0.0002644973961700777,
"grad_norm": 49.861035229921576,
"learning_rate": 8.619000979431931e-08,
"loss": 2.7158,
"step": 45
},
{
"epoch": 0.0002703751160849683,
"grad_norm": 40.8079607718849,
"learning_rate": 8.814887365328112e-08,
"loss": 2.3449,
"step": 46
},
{
"epoch": 0.0002762528359998589,
"grad_norm": 37.41904404905814,
"learning_rate": 9.01077375122429e-08,
"loss": 2.4975,
"step": 47
},
{
"epoch": 0.0002821305559147496,
"grad_norm": 46.57678726138831,
"learning_rate": 9.20666013712047e-08,
"loss": 2.5687,
"step": 48
},
{
"epoch": 0.0002880082758296402,
"grad_norm": 31.29394678361741,
"learning_rate": 9.402546523016652e-08,
"loss": 2.1173,
"step": 49
},
{
"epoch": 0.0002938859957445308,
"grad_norm": 37.33288236667374,
"learning_rate": 9.598432908912831e-08,
"loss": 2.6534,
"step": 50
},
{
"epoch": 0.0002997637156594214,
"grad_norm": 34.06484936870885,
"learning_rate": 9.794319294809012e-08,
"loss": 2.3917,
"step": 51
},
{
"epoch": 0.000305641435574312,
"grad_norm": 48.254524144077536,
"learning_rate": 9.990205680705191e-08,
"loss": 2.2721,
"step": 52
},
{
"epoch": 0.00031151915548920263,
"grad_norm": 47.77507733499728,
"learning_rate": 1.0186092066601372e-07,
"loss": 2.5648,
"step": 53
},
{
"epoch": 0.00031739687540409324,
"grad_norm": 39.93090562192217,
"learning_rate": 1.0381978452497552e-07,
"loss": 2.7048,
"step": 54
},
{
"epoch": 0.00032327459531898386,
"grad_norm": 45.71987876675893,
"learning_rate": 1.0577864838393733e-07,
"loss": 2.4546,
"step": 55
},
{
"epoch": 0.00032915231523387447,
"grad_norm": 26.902504182785798,
"learning_rate": 1.0773751224289912e-07,
"loss": 2.128,
"step": 56
},
{
"epoch": 0.0003350300351487651,
"grad_norm": 26.86954323404998,
"learning_rate": 1.0969637610186093e-07,
"loss": 2.3557,
"step": 57
},
{
"epoch": 0.0003409077550636557,
"grad_norm": 55.89037077858931,
"learning_rate": 1.1165523996082273e-07,
"loss": 2.8119,
"step": 58
},
{
"epoch": 0.0003467854749785463,
"grad_norm": 42.77040196686266,
"learning_rate": 1.1361410381978454e-07,
"loss": 2.6056,
"step": 59
},
{
"epoch": 0.0003526631948934369,
"grad_norm": 30.77897770617208,
"learning_rate": 1.1557296767874635e-07,
"loss": 2.1565,
"step": 60
},
{
"epoch": 0.0003585409148083276,
"grad_norm": 28.082285297919455,
"learning_rate": 1.1753183153770813e-07,
"loss": 2.3056,
"step": 61
},
{
"epoch": 0.0003644186347232182,
"grad_norm": 51.89300221840249,
"learning_rate": 1.1949069539666995e-07,
"loss": 2.3585,
"step": 62
},
{
"epoch": 0.0003702963546381088,
"grad_norm": 29.104662619846845,
"learning_rate": 1.2144955925563174e-07,
"loss": 2.4636,
"step": 63
},
{
"epoch": 0.0003761740745529994,
"grad_norm": 40.89677913796411,
"learning_rate": 1.2340842311459354e-07,
"loss": 2.3843,
"step": 64
},
{
"epoch": 0.00038205179446789,
"grad_norm": 29.068818361443615,
"learning_rate": 1.2536728697355534e-07,
"loss": 2.3794,
"step": 65
},
{
"epoch": 0.00038792951438278064,
"grad_norm": 27.754009467402703,
"learning_rate": 1.2732615083251716e-07,
"loss": 2.3511,
"step": 66
},
{
"epoch": 0.00039380723429767125,
"grad_norm": 38.91129852443923,
"learning_rate": 1.2928501469147895e-07,
"loss": 2.2822,
"step": 67
},
{
"epoch": 0.00039968495421256186,
"grad_norm": 62.770631236350724,
"learning_rate": 1.3124387855044075e-07,
"loss": 2.5637,
"step": 68
},
{
"epoch": 0.00040556267412745247,
"grad_norm": 43.90327612952796,
"learning_rate": 1.3320274240940254e-07,
"loss": 2.5136,
"step": 69
},
{
"epoch": 0.0004114403940423431,
"grad_norm": 26.674560667801135,
"learning_rate": 1.3516160626836437e-07,
"loss": 1.9693,
"step": 70
},
{
"epoch": 0.0004173181139572337,
"grad_norm": 33.42868793910111,
"learning_rate": 1.3712047012732616e-07,
"loss": 2.4011,
"step": 71
},
{
"epoch": 0.0004231958338721243,
"grad_norm": 35.6133771366324,
"learning_rate": 1.3907933398628796e-07,
"loss": 2.2642,
"step": 72
},
{
"epoch": 0.0004290735537870149,
"grad_norm": 25.976386072062112,
"learning_rate": 1.4103819784524978e-07,
"loss": 2.2197,
"step": 73
},
{
"epoch": 0.00043495127370190553,
"grad_norm": 26.220898436195753,
"learning_rate": 1.4299706170421157e-07,
"loss": 2.3933,
"step": 74
},
{
"epoch": 0.0004408289936167962,
"grad_norm": 31.704081974998505,
"learning_rate": 1.4495592556317337e-07,
"loss": 2.3164,
"step": 75
},
{
"epoch": 0.0004467067135316868,
"grad_norm": 28.191292094592438,
"learning_rate": 1.4691478942213517e-07,
"loss": 2.1898,
"step": 76
},
{
"epoch": 0.0004525844334465774,
"grad_norm": 34.07776057624209,
"learning_rate": 1.48873653281097e-07,
"loss": 2.1479,
"step": 77
},
{
"epoch": 0.00045846215336146803,
"grad_norm": 35.51561977372984,
"learning_rate": 1.5083251714005878e-07,
"loss": 2.1504,
"step": 78
},
{
"epoch": 0.00046433987327635864,
"grad_norm": 32.37780568021879,
"learning_rate": 1.5279138099902058e-07,
"loss": 2.0425,
"step": 79
},
{
"epoch": 0.00047021759319124925,
"grad_norm": 26.28474544112387,
"learning_rate": 1.5475024485798237e-07,
"loss": 2.535,
"step": 80
},
{
"epoch": 0.00047609531310613987,
"grad_norm": 26.94813340817767,
"learning_rate": 1.567091087169442e-07,
"loss": 2.3321,
"step": 81
},
{
"epoch": 0.0004819730330210305,
"grad_norm": 42.30088898391613,
"learning_rate": 1.58667972575906e-07,
"loss": 2.2359,
"step": 82
},
{
"epoch": 0.0004878507529359211,
"grad_norm": 26.92802144279086,
"learning_rate": 1.6062683643486781e-07,
"loss": 2.3239,
"step": 83
},
{
"epoch": 0.0004937284728508117,
"grad_norm": 34.182845875119725,
"learning_rate": 1.6258570029382958e-07,
"loss": 2.2904,
"step": 84
},
{
"epoch": 0.0004996061927657024,
"grad_norm": 33.83477740859824,
"learning_rate": 1.6454456415279138e-07,
"loss": 2.287,
"step": 85
},
{
"epoch": 0.0005054839126805929,
"grad_norm": 31.28814971716291,
"learning_rate": 1.665034280117532e-07,
"loss": 2.3628,
"step": 86
},
{
"epoch": 0.0005113616325954836,
"grad_norm": 24.161461789118633,
"learning_rate": 1.68462291870715e-07,
"loss": 2.3799,
"step": 87
},
{
"epoch": 0.0005172393525103741,
"grad_norm": 38.85070435824102,
"learning_rate": 1.7042115572967682e-07,
"loss": 2.1733,
"step": 88
},
{
"epoch": 0.0005231170724252648,
"grad_norm": 26.583507631290228,
"learning_rate": 1.7238001958863861e-07,
"loss": 2.0755,
"step": 89
},
{
"epoch": 0.0005289947923401554,
"grad_norm": 31.67972636527564,
"learning_rate": 1.7433888344760038e-07,
"loss": 1.9064,
"step": 90
},
{
"epoch": 0.000534872512255046,
"grad_norm": 18.31955982249094,
"learning_rate": 1.7629774730656223e-07,
"loss": 2.3138,
"step": 91
},
{
"epoch": 0.0005407502321699366,
"grad_norm": 22.619392174396776,
"learning_rate": 1.78256611165524e-07,
"loss": 2.0967,
"step": 92
},
{
"epoch": 0.0005466279520848273,
"grad_norm": 28.1552172586705,
"learning_rate": 1.802154750244858e-07,
"loss": 2.0212,
"step": 93
},
{
"epoch": 0.0005525056719997178,
"grad_norm": 26.39585059625926,
"learning_rate": 1.8217433888344762e-07,
"loss": 2.1721,
"step": 94
},
{
"epoch": 0.0005583833919146085,
"grad_norm": 56.55570072945738,
"learning_rate": 1.841332027424094e-07,
"loss": 2.2185,
"step": 95
},
{
"epoch": 0.0005642611118294991,
"grad_norm": 18.255595773563055,
"learning_rate": 1.8609206660137123e-07,
"loss": 1.9785,
"step": 96
},
{
"epoch": 0.0005701388317443897,
"grad_norm": 24.380219043592408,
"learning_rate": 1.8805093046033303e-07,
"loss": 1.9331,
"step": 97
},
{
"epoch": 0.0005760165516592804,
"grad_norm": 23.21420086498007,
"learning_rate": 1.900097943192948e-07,
"loss": 2.1102,
"step": 98
},
{
"epoch": 0.0005818942715741709,
"grad_norm": 25.644314151305732,
"learning_rate": 1.9196865817825662e-07,
"loss": 1.8602,
"step": 99
},
{
"epoch": 0.0005877719914890616,
"grad_norm": 24.678894252503262,
"learning_rate": 1.9392752203721842e-07,
"loss": 1.9086,
"step": 100
},
{
"epoch": 0.0005936497114039522,
"grad_norm": 21.140657194454246,
"learning_rate": 1.9588638589618024e-07,
"loss": 1.8432,
"step": 101
},
{
"epoch": 0.0005995274313188428,
"grad_norm": 22.987569823694244,
"learning_rate": 1.9784524975514203e-07,
"loss": 2.1482,
"step": 102
},
{
"epoch": 0.0006054051512337334,
"grad_norm": 21.456124991404014,
"learning_rate": 1.9980411361410383e-07,
"loss": 1.6048,
"step": 103
},
{
"epoch": 0.000611282871148624,
"grad_norm": 23.506513682703766,
"learning_rate": 2.0176297747306565e-07,
"loss": 2.1504,
"step": 104
},
{
"epoch": 0.0006171605910635146,
"grad_norm": 22.46761557860848,
"learning_rate": 2.0372184133202745e-07,
"loss": 2.0595,
"step": 105
},
{
"epoch": 0.0006230383109784053,
"grad_norm": 22.25174247662017,
"learning_rate": 2.0568070519098922e-07,
"loss": 1.8643,
"step": 106
},
{
"epoch": 0.0006289160308932958,
"grad_norm": 20.762128871169335,
"learning_rate": 2.0763956904995104e-07,
"loss": 1.8806,
"step": 107
},
{
"epoch": 0.0006347937508081865,
"grad_norm": 21.51403808577067,
"learning_rate": 2.0959843290891283e-07,
"loss": 1.737,
"step": 108
},
{
"epoch": 0.0006406714707230772,
"grad_norm": 14.877184572562975,
"learning_rate": 2.1155729676787466e-07,
"loss": 1.8458,
"step": 109
},
{
"epoch": 0.0006465491906379677,
"grad_norm": 17.99619216256105,
"learning_rate": 2.1351616062683645e-07,
"loss": 1.7246,
"step": 110
},
{
"epoch": 0.0006524269105528584,
"grad_norm": 19.28119326563505,
"learning_rate": 2.1547502448579825e-07,
"loss": 1.9538,
"step": 111
},
{
"epoch": 0.0006583046304677489,
"grad_norm": 32.38937989525677,
"learning_rate": 2.1743388834476007e-07,
"loss": 1.9033,
"step": 112
},
{
"epoch": 0.0006641823503826396,
"grad_norm": 17.174288277599942,
"learning_rate": 2.1939275220372186e-07,
"loss": 1.6582,
"step": 113
},
{
"epoch": 0.0006700600702975302,
"grad_norm": 15.380513223594036,
"learning_rate": 2.2135161606268369e-07,
"loss": 1.7601,
"step": 114
},
{
"epoch": 0.0006759377902124208,
"grad_norm": 18.099381133670537,
"learning_rate": 2.2331047992164546e-07,
"loss": 1.7123,
"step": 115
},
{
"epoch": 0.0006818155101273114,
"grad_norm": 15.449085025928484,
"learning_rate": 2.2526934378060725e-07,
"loss": 2.0434,
"step": 116
},
{
"epoch": 0.000687693230042202,
"grad_norm": 18.873494216822824,
"learning_rate": 2.2722820763956907e-07,
"loss": 1.9963,
"step": 117
},
{
"epoch": 0.0006935709499570926,
"grad_norm": 29.52475437227068,
"learning_rate": 2.2918707149853087e-07,
"loss": 1.7232,
"step": 118
},
{
"epoch": 0.0006994486698719833,
"grad_norm": 15.33187463191825,
"learning_rate": 2.311459353574927e-07,
"loss": 1.4605,
"step": 119
},
{
"epoch": 0.0007053263897868738,
"grad_norm": 21.195728129495127,
"learning_rate": 2.3310479921645449e-07,
"loss": 1.6452,
"step": 120
},
{
"epoch": 0.0007112041097017645,
"grad_norm": 12.616990324376612,
"learning_rate": 2.3506366307541625e-07,
"loss": 1.7385,
"step": 121
},
{
"epoch": 0.0007170818296166552,
"grad_norm": 16.59194705973793,
"learning_rate": 2.370225269343781e-07,
"loss": 1.7001,
"step": 122
},
{
"epoch": 0.0007229595495315457,
"grad_norm": 19.98385582713288,
"learning_rate": 2.389813907933399e-07,
"loss": 1.8642,
"step": 123
},
{
"epoch": 0.0007288372694464364,
"grad_norm": 11.984048679052938,
"learning_rate": 2.4094025465230167e-07,
"loss": 1.7128,
"step": 124
},
{
"epoch": 0.0007347149893613269,
"grad_norm": 13.817274924710402,
"learning_rate": 2.428991185112635e-07,
"loss": 1.9165,
"step": 125
},
{
"epoch": 0.0007405927092762176,
"grad_norm": 14.666508072298448,
"learning_rate": 2.4485798237022526e-07,
"loss": 1.3521,
"step": 126
},
{
"epoch": 0.0007464704291911082,
"grad_norm": 15.886204407427591,
"learning_rate": 2.468168462291871e-07,
"loss": 1.6274,
"step": 127
},
{
"epoch": 0.0007523481491059988,
"grad_norm": 27.82376758715373,
"learning_rate": 2.487757100881489e-07,
"loss": 1.6963,
"step": 128
},
{
"epoch": 0.0007582258690208894,
"grad_norm": 32.17667564558956,
"learning_rate": 2.5073457394711067e-07,
"loss": 1.6275,
"step": 129
},
{
"epoch": 0.00076410358893578,
"grad_norm": 16.30806188824753,
"learning_rate": 2.526934378060725e-07,
"loss": 1.5903,
"step": 130
},
{
"epoch": 0.0007699813088506706,
"grad_norm": 17.038993207268526,
"learning_rate": 2.546523016650343e-07,
"loss": 1.6622,
"step": 131
},
{
"epoch": 0.0007758590287655613,
"grad_norm": 14.492825408251898,
"learning_rate": 2.5661116552399614e-07,
"loss": 1.7637,
"step": 132
},
{
"epoch": 0.0007817367486804518,
"grad_norm": 20.71180708089968,
"learning_rate": 2.585700293829579e-07,
"loss": 1.4606,
"step": 133
},
{
"epoch": 0.0007876144685953425,
"grad_norm": 15.231619445935591,
"learning_rate": 2.605288932419197e-07,
"loss": 1.6682,
"step": 134
},
{
"epoch": 0.0007934921885102331,
"grad_norm": 14.170828124279078,
"learning_rate": 2.624877571008815e-07,
"loss": 1.8463,
"step": 135
},
{
"epoch": 0.0007993699084251237,
"grad_norm": 10.59140070763608,
"learning_rate": 2.644466209598433e-07,
"loss": 1.5088,
"step": 136
},
{
"epoch": 0.0008052476283400144,
"grad_norm": 26.814590174464087,
"learning_rate": 2.664054848188051e-07,
"loss": 1.8973,
"step": 137
},
{
"epoch": 0.0008111253482549049,
"grad_norm": 11.60333007084735,
"learning_rate": 2.683643486777669e-07,
"loss": 1.6231,
"step": 138
},
{
"epoch": 0.0008170030681697956,
"grad_norm": 15.760400853465885,
"learning_rate": 2.7032321253672873e-07,
"loss": 1.6369,
"step": 139
},
{
"epoch": 0.0008228807880846862,
"grad_norm": 11.039629773110311,
"learning_rate": 2.7228207639569055e-07,
"loss": 1.7291,
"step": 140
},
{
"epoch": 0.0008287585079995768,
"grad_norm": 13.050093706824535,
"learning_rate": 2.742409402546523e-07,
"loss": 1.3813,
"step": 141
},
{
"epoch": 0.0008346362279144674,
"grad_norm": 13.5488482037351,
"learning_rate": 2.761998041136141e-07,
"loss": 1.8591,
"step": 142
},
{
"epoch": 0.0008405139478293581,
"grad_norm": 11.7620793288059,
"learning_rate": 2.781586679725759e-07,
"loss": 1.721,
"step": 143
},
{
"epoch": 0.0008463916677442486,
"grad_norm": 10.320957727432353,
"learning_rate": 2.8011753183153774e-07,
"loss": 1.5765,
"step": 144
},
{
"epoch": 0.0008522693876591393,
"grad_norm": 13.202134714301499,
"learning_rate": 2.8207639569049956e-07,
"loss": 1.9503,
"step": 145
},
{
"epoch": 0.0008581471075740298,
"grad_norm": 11.162755956662153,
"learning_rate": 2.8403525954946133e-07,
"loss": 1.5982,
"step": 146
},
{
"epoch": 0.0008640248274889205,
"grad_norm": 11.393460380375336,
"learning_rate": 2.8599412340842315e-07,
"loss": 1.6087,
"step": 147
},
{
"epoch": 0.0008699025474038111,
"grad_norm": 14.494889123434431,
"learning_rate": 2.8795298726738497e-07,
"loss": 1.6165,
"step": 148
},
{
"epoch": 0.0008757802673187017,
"grad_norm": 11.174271675177474,
"learning_rate": 2.8991185112634674e-07,
"loss": 1.7131,
"step": 149
},
{
"epoch": 0.0008816579872335924,
"grad_norm": 9.71406247750616,
"learning_rate": 2.918707149853085e-07,
"loss": 1.4839,
"step": 150
},
{
"epoch": 0.000887535707148483,
"grad_norm": 14.012114534660885,
"learning_rate": 2.9382957884427033e-07,
"loss": 1.6935,
"step": 151
},
{
"epoch": 0.0008934134270633736,
"grad_norm": 10.065424608494785,
"learning_rate": 2.9578844270323215e-07,
"loss": 1.5043,
"step": 152
},
{
"epoch": 0.0008992911469782642,
"grad_norm": 10.291357563146,
"learning_rate": 2.97747306562194e-07,
"loss": 1.5935,
"step": 153
},
{
"epoch": 0.0009051688668931548,
"grad_norm": 9.468653200556892,
"learning_rate": 2.9970617042115574e-07,
"loss": 1.437,
"step": 154
},
{
"epoch": 0.0009110465868080454,
"grad_norm": 11.658556544230711,
"learning_rate": 3.0166503428011757e-07,
"loss": 1.8376,
"step": 155
},
{
"epoch": 0.0009169243067229361,
"grad_norm": 9.056893243480333,
"learning_rate": 3.036238981390794e-07,
"loss": 1.2717,
"step": 156
},
{
"epoch": 0.0009228020266378266,
"grad_norm": 9.678352816042532,
"learning_rate": 3.0558276199804116e-07,
"loss": 1.3479,
"step": 157
},
{
"epoch": 0.0009286797465527173,
"grad_norm": 9.018293017844652,
"learning_rate": 3.07541625857003e-07,
"loss": 1.8007,
"step": 158
},
{
"epoch": 0.0009345574664676078,
"grad_norm": 10.85779758090484,
"learning_rate": 3.0950048971596475e-07,
"loss": 1.5511,
"step": 159
},
{
"epoch": 0.0009404351863824985,
"grad_norm": 9.215542538487705,
"learning_rate": 3.1145935357492657e-07,
"loss": 1.5077,
"step": 160
},
{
"epoch": 0.0009463129062973891,
"grad_norm": 10.441875100178423,
"learning_rate": 3.134182174338884e-07,
"loss": 1.2265,
"step": 161
},
{
"epoch": 0.0009521906262122797,
"grad_norm": 11.707651718323877,
"learning_rate": 3.153770812928502e-07,
"loss": 1.6223,
"step": 162
},
{
"epoch": 0.0009580683461271704,
"grad_norm": 11.50471640686927,
"learning_rate": 3.17335945151812e-07,
"loss": 1.5438,
"step": 163
},
{
"epoch": 0.000963946066042061,
"grad_norm": 9.384813095151022,
"learning_rate": 3.192948090107738e-07,
"loss": 1.5544,
"step": 164
},
{
"epoch": 0.0009698237859569516,
"grad_norm": 16.29679950683345,
"learning_rate": 3.2125367286973563e-07,
"loss": 1.9744,
"step": 165
},
{
"epoch": 0.0009757015058718422,
"grad_norm": 9.074030216135546,
"learning_rate": 3.2321253672869734e-07,
"loss": 1.5864,
"step": 166
},
{
"epoch": 0.0009815792257867328,
"grad_norm": 9.283318899361504,
"learning_rate": 3.2517140058765917e-07,
"loss": 1.3906,
"step": 167
},
{
"epoch": 0.0009874569457016234,
"grad_norm": 10.17393481045071,
"learning_rate": 3.27130264446621e-07,
"loss": 1.6783,
"step": 168
},
{
"epoch": 0.000993334665616514,
"grad_norm": 10.387335954943156,
"learning_rate": 3.2908912830558276e-07,
"loss": 1.6027,
"step": 169
},
{
"epoch": 0.0009992123855314047,
"grad_norm": 9.08557174133179,
"learning_rate": 3.310479921645446e-07,
"loss": 1.3654,
"step": 170
},
{
"epoch": 0.0010050901054462953,
"grad_norm": 9.595202107192796,
"learning_rate": 3.330068560235064e-07,
"loss": 1.5666,
"step": 171
},
{
"epoch": 0.0010109678253611858,
"grad_norm": 10.31606733640622,
"learning_rate": 3.3496571988246817e-07,
"loss": 1.6286,
"step": 172
},
{
"epoch": 0.0010168455452760764,
"grad_norm": 9.326281386641854,
"learning_rate": 3.3692458374143e-07,
"loss": 1.9201,
"step": 173
},
{
"epoch": 0.0010227232651909672,
"grad_norm": 8.994312317847095,
"learning_rate": 3.388834476003918e-07,
"loss": 1.5568,
"step": 174
},
{
"epoch": 0.0010286009851058577,
"grad_norm": 9.980025494040948,
"learning_rate": 3.4084231145935364e-07,
"loss": 1.354,
"step": 175
},
{
"epoch": 0.0010344787050207483,
"grad_norm": 11.562220948333023,
"learning_rate": 3.428011753183154e-07,
"loss": 1.5634,
"step": 176
},
{
"epoch": 0.001040356424935639,
"grad_norm": 9.776227430851645,
"learning_rate": 3.4476003917727723e-07,
"loss": 1.4959,
"step": 177
},
{
"epoch": 0.0010462341448505296,
"grad_norm": 11.661631330611714,
"learning_rate": 3.4671890303623905e-07,
"loss": 1.5947,
"step": 178
},
{
"epoch": 0.0010521118647654202,
"grad_norm": 11.68005984757505,
"learning_rate": 3.4867776689520076e-07,
"loss": 1.5006,
"step": 179
},
{
"epoch": 0.0010579895846803107,
"grad_norm": 12.777534176544883,
"learning_rate": 3.506366307541626e-07,
"loss": 1.4467,
"step": 180
},
{
"epoch": 0.0010638673045952015,
"grad_norm": 17.41785652701965,
"learning_rate": 3.5259549461312446e-07,
"loss": 1.9458,
"step": 181
},
{
"epoch": 0.001069745024510092,
"grad_norm": 10.78085942187347,
"learning_rate": 3.545543584720862e-07,
"loss": 1.5085,
"step": 182
},
{
"epoch": 0.0010756227444249826,
"grad_norm": 16.771876557156034,
"learning_rate": 3.56513222331048e-07,
"loss": 1.5092,
"step": 183
},
{
"epoch": 0.0010815004643398732,
"grad_norm": 13.15308515651388,
"learning_rate": 3.584720861900098e-07,
"loss": 1.7019,
"step": 184
},
{
"epoch": 0.001087378184254764,
"grad_norm": 15.191365994444775,
"learning_rate": 3.604309500489716e-07,
"loss": 1.4493,
"step": 185
},
{
"epoch": 0.0010932559041696545,
"grad_norm": 9.667742910814834,
"learning_rate": 3.623898139079334e-07,
"loss": 1.4017,
"step": 186
},
{
"epoch": 0.001099133624084545,
"grad_norm": 8.784865132869584,
"learning_rate": 3.6434867776689523e-07,
"loss": 1.2916,
"step": 187
},
{
"epoch": 0.0011050113439994356,
"grad_norm": 12.035429882395604,
"learning_rate": 3.6630754162585706e-07,
"loss": 1.5701,
"step": 188
},
{
"epoch": 0.0011108890639143264,
"grad_norm": 10.937854560715525,
"learning_rate": 3.682664054848188e-07,
"loss": 1.261,
"step": 189
},
{
"epoch": 0.001116766783829217,
"grad_norm": 13.452124396814133,
"learning_rate": 3.7022526934378065e-07,
"loss": 1.4703,
"step": 190
},
{
"epoch": 0.0011226445037441075,
"grad_norm": 8.894603197974076,
"learning_rate": 3.7218413320274247e-07,
"loss": 1.4685,
"step": 191
},
{
"epoch": 0.0011285222236589983,
"grad_norm": 15.648678530497842,
"learning_rate": 3.7414299706170424e-07,
"loss": 1.8496,
"step": 192
},
{
"epoch": 0.0011343999435738889,
"grad_norm": 11.032596124236516,
"learning_rate": 3.7610186092066606e-07,
"loss": 1.9087,
"step": 193
},
{
"epoch": 0.0011402776634887794,
"grad_norm": 11.308196746519451,
"learning_rate": 3.780607247796279e-07,
"loss": 1.4356,
"step": 194
},
{
"epoch": 0.00114615538340367,
"grad_norm": 10.712918855934676,
"learning_rate": 3.800195886385896e-07,
"loss": 1.5941,
"step": 195
},
{
"epoch": 0.0011520331033185607,
"grad_norm": 10.561945446326497,
"learning_rate": 3.819784524975514e-07,
"loss": 1.1996,
"step": 196
},
{
"epoch": 0.0011579108232334513,
"grad_norm": 9.171518744551488,
"learning_rate": 3.8393731635651324e-07,
"loss": 1.8124,
"step": 197
},
{
"epoch": 0.0011637885431483419,
"grad_norm": 11.090728930779585,
"learning_rate": 3.85896180215475e-07,
"loss": 1.5103,
"step": 198
},
{
"epoch": 0.0011696662630632324,
"grad_norm": 21.075458727240783,
"learning_rate": 3.8785504407443683e-07,
"loss": 1.489,
"step": 199
},
{
"epoch": 0.0011755439829781232,
"grad_norm": 9.785441331082623,
"learning_rate": 3.8981390793339866e-07,
"loss": 1.3845,
"step": 200
},
{
"epoch": 0.0011814217028930137,
"grad_norm": 10.6811684049893,
"learning_rate": 3.917727717923605e-07,
"loss": 1.6402,
"step": 201
},
{
"epoch": 0.0011872994228079043,
"grad_norm": 9.419153945057552,
"learning_rate": 3.9373163565132225e-07,
"loss": 1.4811,
"step": 202
},
{
"epoch": 0.001193177142722795,
"grad_norm": 10.076216378174424,
"learning_rate": 3.9569049951028407e-07,
"loss": 1.2899,
"step": 203
},
{
"epoch": 0.0011990548626376856,
"grad_norm": 9.960886685047456,
"learning_rate": 3.976493633692459e-07,
"loss": 1.4358,
"step": 204
},
{
"epoch": 0.0012049325825525762,
"grad_norm": 8.401960196603293,
"learning_rate": 3.9960822722820766e-07,
"loss": 1.5201,
"step": 205
},
{
"epoch": 0.0012108103024674668,
"grad_norm": 11.58937622232436,
"learning_rate": 4.015670910871695e-07,
"loss": 1.3173,
"step": 206
},
{
"epoch": 0.0012166880223823575,
"grad_norm": 11.03334640239208,
"learning_rate": 4.035259549461313e-07,
"loss": 1.4313,
"step": 207
},
{
"epoch": 0.001222565742297248,
"grad_norm": 10.107621975482537,
"learning_rate": 4.0548481880509307e-07,
"loss": 1.5898,
"step": 208
},
{
"epoch": 0.0012284434622121386,
"grad_norm": 10.061788427468194,
"learning_rate": 4.074436826640549e-07,
"loss": 1.4872,
"step": 209
},
{
"epoch": 0.0012343211821270292,
"grad_norm": 11.047462590018448,
"learning_rate": 4.094025465230167e-07,
"loss": 1.4513,
"step": 210
},
{
"epoch": 0.00124019890204192,
"grad_norm": 9.225859001106716,
"learning_rate": 4.1136141038197843e-07,
"loss": 1.635,
"step": 211
},
{
"epoch": 0.0012460766219568105,
"grad_norm": 9.08800716773603,
"learning_rate": 4.1332027424094025e-07,
"loss": 1.4819,
"step": 212
},
{
"epoch": 0.001251954341871701,
"grad_norm": 8.868603405154412,
"learning_rate": 4.152791380999021e-07,
"loss": 1.2301,
"step": 213
},
{
"epoch": 0.0012578320617865916,
"grad_norm": 14.15909109082643,
"learning_rate": 4.172380019588639e-07,
"loss": 1.9416,
"step": 214
},
{
"epoch": 0.0012637097817014824,
"grad_norm": 9.067239599822472,
"learning_rate": 4.1919686581782567e-07,
"loss": 1.797,
"step": 215
},
{
"epoch": 0.001269587501616373,
"grad_norm": 12.132058750911389,
"learning_rate": 4.211557296767875e-07,
"loss": 1.4109,
"step": 216
},
{
"epoch": 0.0012754652215312635,
"grad_norm": 10.075853653825005,
"learning_rate": 4.231145935357493e-07,
"loss": 1.3572,
"step": 217
},
{
"epoch": 0.0012813429414461543,
"grad_norm": 9.168243399608876,
"learning_rate": 4.250734573947111e-07,
"loss": 1.8158,
"step": 218
},
{
"epoch": 0.0012872206613610449,
"grad_norm": 11.186175864257553,
"learning_rate": 4.270323212536729e-07,
"loss": 1.5282,
"step": 219
},
{
"epoch": 0.0012930983812759354,
"grad_norm": 8.733700945605664,
"learning_rate": 4.289911851126347e-07,
"loss": 1.5974,
"step": 220
},
{
"epoch": 0.001298976101190826,
"grad_norm": 7.788910732857218,
"learning_rate": 4.309500489715965e-07,
"loss": 1.3257,
"step": 221
},
{
"epoch": 0.0013048538211057168,
"grad_norm": 12.030091770377302,
"learning_rate": 4.329089128305583e-07,
"loss": 1.1955,
"step": 222
},
{
"epoch": 0.0013107315410206073,
"grad_norm": 10.080106382940905,
"learning_rate": 4.3486777668952014e-07,
"loss": 1.3482,
"step": 223
},
{
"epoch": 0.0013166092609354979,
"grad_norm": 11.07589048912045,
"learning_rate": 4.3682664054848196e-07,
"loss": 1.3602,
"step": 224
},
{
"epoch": 0.0013224869808503884,
"grad_norm": 15.065493178868694,
"learning_rate": 4.3878550440744373e-07,
"loss": 1.248,
"step": 225
},
{
"epoch": 0.0013283647007652792,
"grad_norm": 9.370347727362022,
"learning_rate": 4.4074436826640555e-07,
"loss": 1.429,
"step": 226
},
{
"epoch": 0.0013342424206801698,
"grad_norm": 9.30558950644716,
"learning_rate": 4.4270323212536737e-07,
"loss": 1.45,
"step": 227
},
{
"epoch": 0.0013401201405950603,
"grad_norm": 10.25603354658129,
"learning_rate": 4.446620959843291e-07,
"loss": 1.5622,
"step": 228
},
{
"epoch": 0.0013459978605099509,
"grad_norm": 12.297542220540503,
"learning_rate": 4.466209598432909e-07,
"loss": 1.2,
"step": 229
},
{
"epoch": 0.0013518755804248416,
"grad_norm": 9.331818179918656,
"learning_rate": 4.4857982370225273e-07,
"loss": 1.3011,
"step": 230
},
{
"epoch": 0.0013577533003397322,
"grad_norm": 10.585992361010224,
"learning_rate": 4.505386875612145e-07,
"loss": 1.6246,
"step": 231
},
{
"epoch": 0.0013636310202546228,
"grad_norm": 9.046465495391782,
"learning_rate": 4.524975514201763e-07,
"loss": 1.6676,
"step": 232
},
{
"epoch": 0.0013695087401695135,
"grad_norm": 8.998925855520136,
"learning_rate": 4.5445641527913815e-07,
"loss": 1.6056,
"step": 233
},
{
"epoch": 0.001375386460084404,
"grad_norm": 9.986723519133724,
"learning_rate": 4.564152791380999e-07,
"loss": 1.377,
"step": 234
},
{
"epoch": 0.0013812641799992947,
"grad_norm": 9.61309475151082,
"learning_rate": 4.5837414299706174e-07,
"loss": 1.4077,
"step": 235
},
{
"epoch": 0.0013871418999141852,
"grad_norm": 12.39156328483041,
"learning_rate": 4.6033300685602356e-07,
"loss": 1.2681,
"step": 236
},
{
"epoch": 0.001393019619829076,
"grad_norm": 11.93909914894201,
"learning_rate": 4.622918707149854e-07,
"loss": 1.5578,
"step": 237
},
{
"epoch": 0.0013988973397439665,
"grad_norm": 9.241067526275764,
"learning_rate": 4.6425073457394715e-07,
"loss": 1.424,
"step": 238
},
{
"epoch": 0.001404775059658857,
"grad_norm": 69.09437420107695,
"learning_rate": 4.6620959843290897e-07,
"loss": 1.7323,
"step": 239
},
{
"epoch": 0.0014106527795737477,
"grad_norm": 9.993143408906265,
"learning_rate": 4.681684622918708e-07,
"loss": 1.5018,
"step": 240
},
{
"epoch": 0.0014165304994886384,
"grad_norm": 10.458045510826041,
"learning_rate": 4.701273261508325e-07,
"loss": 1.3346,
"step": 241
},
{
"epoch": 0.001422408219403529,
"grad_norm": 15.334313398712514,
"learning_rate": 4.720861900097944e-07,
"loss": 1.4986,
"step": 242
},
{
"epoch": 0.0014282859393184195,
"grad_norm": 13.494956398702698,
"learning_rate": 4.740450538687562e-07,
"loss": 1.931,
"step": 243
},
{
"epoch": 0.0014341636592333103,
"grad_norm": 10.152689352885435,
"learning_rate": 4.760039177277179e-07,
"loss": 1.5585,
"step": 244
},
{
"epoch": 0.0014400413791482009,
"grad_norm": 8.834300902788739,
"learning_rate": 4.779627815866798e-07,
"loss": 1.3602,
"step": 245
},
{
"epoch": 0.0014459190990630914,
"grad_norm": 11.381145111843919,
"learning_rate": 4.799216454456416e-07,
"loss": 1.7747,
"step": 246
},
{
"epoch": 0.001451796818977982,
"grad_norm": 12.06148998092397,
"learning_rate": 4.818805093046033e-07,
"loss": 1.5178,
"step": 247
},
{
"epoch": 0.0014576745388928728,
"grad_norm": 8.703106046771826,
"learning_rate": 4.838393731635652e-07,
"loss": 1.5339,
"step": 248
},
{
"epoch": 0.0014635522588077633,
"grad_norm": 11.925440014048084,
"learning_rate": 4.85798237022527e-07,
"loss": 1.3716,
"step": 249
},
{
"epoch": 0.0014694299787226539,
"grad_norm": 9.313391960746207,
"learning_rate": 4.877571008814888e-07,
"loss": 1.5757,
"step": 250
},
{
"epoch": 0.0014753076986375444,
"grad_norm": 8.863788474883224,
"learning_rate": 4.897159647404505e-07,
"loss": 1.3941,
"step": 251
},
{
"epoch": 0.0014811854185524352,
"grad_norm": 10.606405136809217,
"learning_rate": 4.916748285994123e-07,
"loss": 1.3419,
"step": 252
},
{
"epoch": 0.0014870631384673258,
"grad_norm": 12.503980336764862,
"learning_rate": 4.936336924583742e-07,
"loss": 1.4387,
"step": 253
},
{
"epoch": 0.0014929408583822163,
"grad_norm": 11.202398292907109,
"learning_rate": 4.95592556317336e-07,
"loss": 1.3391,
"step": 254
},
{
"epoch": 0.0014988185782971069,
"grad_norm": 8.526431729740759,
"learning_rate": 4.975514201762978e-07,
"loss": 1.2921,
"step": 255
},
{
"epoch": 0.0015046962982119977,
"grad_norm": 8.105752184018986,
"learning_rate": 4.995102840352596e-07,
"loss": 1.6518,
"step": 256
},
{
"epoch": 0.0015105740181268882,
"grad_norm": 8.60563771477929,
"learning_rate": 5.014691478942213e-07,
"loss": 1.4145,
"step": 257
},
{
"epoch": 0.0015164517380417788,
"grad_norm": 9.153779971343228,
"learning_rate": 5.034280117531832e-07,
"loss": 1.3801,
"step": 258
},
{
"epoch": 0.0015223294579566695,
"grad_norm": 8.41877637538237,
"learning_rate": 5.05386875612145e-07,
"loss": 1.3788,
"step": 259
},
{
"epoch": 0.00152820717787156,
"grad_norm": 8.94067688880904,
"learning_rate": 5.073457394711068e-07,
"loss": 1.27,
"step": 260
},
{
"epoch": 0.0015340848977864507,
"grad_norm": 10.981582990635001,
"learning_rate": 5.093046033300686e-07,
"loss": 1.4033,
"step": 261
},
{
"epoch": 0.0015399626177013412,
"grad_norm": 10.577569060024489,
"learning_rate": 5.112634671890305e-07,
"loss": 1.2214,
"step": 262
},
{
"epoch": 0.001545840337616232,
"grad_norm": 12.805732199689936,
"learning_rate": 5.132223310479923e-07,
"loss": 1.303,
"step": 263
},
{
"epoch": 0.0015517180575311226,
"grad_norm": 11.328061019168393,
"learning_rate": 5.15181194906954e-07,
"loss": 1.239,
"step": 264
},
{
"epoch": 0.0015575957774460131,
"grad_norm": 8.779910655482228,
"learning_rate": 5.171400587659158e-07,
"loss": 1.5309,
"step": 265
},
{
"epoch": 0.0015634734973609037,
"grad_norm": 12.48858841285161,
"learning_rate": 5.190989226248776e-07,
"loss": 1.3235,
"step": 266
},
{
"epoch": 0.0015693512172757944,
"grad_norm": 9.529030183678163,
"learning_rate": 5.210577864838394e-07,
"loss": 1.7585,
"step": 267
},
{
"epoch": 0.001575228937190685,
"grad_norm": 8.60745560803701,
"learning_rate": 5.230166503428012e-07,
"loss": 1.3312,
"step": 268
},
{
"epoch": 0.0015811066571055756,
"grad_norm": 11.884696881425304,
"learning_rate": 5.24975514201763e-07,
"loss": 1.6793,
"step": 269
},
{
"epoch": 0.0015869843770204661,
"grad_norm": 8.967117197526099,
"learning_rate": 5.269343780607248e-07,
"loss": 1.4354,
"step": 270
},
{
"epoch": 0.0015928620969353569,
"grad_norm": 10.164939243499552,
"learning_rate": 5.288932419196866e-07,
"loss": 1.3092,
"step": 271
},
{
"epoch": 0.0015987398168502474,
"grad_norm": 10.854358248443695,
"learning_rate": 5.308521057786485e-07,
"loss": 1.4924,
"step": 272
},
{
"epoch": 0.001604617536765138,
"grad_norm": 8.960399631709869,
"learning_rate": 5.328109696376102e-07,
"loss": 1.5164,
"step": 273
},
{
"epoch": 0.0016104952566800288,
"grad_norm": 31.472372986870518,
"learning_rate": 5.34769833496572e-07,
"loss": 1.8094,
"step": 274
},
{
"epoch": 0.0016163729765949193,
"grad_norm": 8.473406841108375,
"learning_rate": 5.367286973555338e-07,
"loss": 1.2746,
"step": 275
},
{
"epoch": 0.0016222506965098099,
"grad_norm": 9.601218622750052,
"learning_rate": 5.386875612144956e-07,
"loss": 1.5087,
"step": 276
},
{
"epoch": 0.0016281284164247004,
"grad_norm": 9.649058387787647,
"learning_rate": 5.406464250734575e-07,
"loss": 1.4106,
"step": 277
},
{
"epoch": 0.0016340061363395912,
"grad_norm": 12.337790226855372,
"learning_rate": 5.426052889324193e-07,
"loss": 1.1803,
"step": 278
},
{
"epoch": 0.0016398838562544818,
"grad_norm": 10.953084253896563,
"learning_rate": 5.445641527913811e-07,
"loss": 1.5586,
"step": 279
},
{
"epoch": 0.0016457615761693723,
"grad_norm": 10.642696358664933,
"learning_rate": 5.465230166503428e-07,
"loss": 1.5048,
"step": 280
},
{
"epoch": 0.001651639296084263,
"grad_norm": 10.736683560071118,
"learning_rate": 5.484818805093046e-07,
"loss": 1.6119,
"step": 281
},
{
"epoch": 0.0016575170159991537,
"grad_norm": 10.990806499370061,
"learning_rate": 5.504407443682665e-07,
"loss": 1.5904,
"step": 282
},
{
"epoch": 0.0016633947359140442,
"grad_norm": 8.163597300046485,
"learning_rate": 5.523996082272282e-07,
"loss": 1.4993,
"step": 283
},
{
"epoch": 0.0016692724558289348,
"grad_norm": 8.147151905533738,
"learning_rate": 5.5435847208619e-07,
"loss": 1.4757,
"step": 284
},
{
"epoch": 0.0016751501757438256,
"grad_norm": 9.599530081078425,
"learning_rate": 5.563173359451518e-07,
"loss": 1.5775,
"step": 285
},
{
"epoch": 0.0016810278956587161,
"grad_norm": 10.21224899778634,
"learning_rate": 5.582761998041137e-07,
"loss": 1.6991,
"step": 286
},
{
"epoch": 0.0016869056155736067,
"grad_norm": 9.785513490047899,
"learning_rate": 5.602350636630755e-07,
"loss": 1.4311,
"step": 287
},
{
"epoch": 0.0016927833354884972,
"grad_norm": 7.874953791900144,
"learning_rate": 5.621939275220373e-07,
"loss": 1.3383,
"step": 288
},
{
"epoch": 0.001698661055403388,
"grad_norm": 9.336147650202287,
"learning_rate": 5.641527913809991e-07,
"loss": 1.3211,
"step": 289
},
{
"epoch": 0.0017045387753182786,
"grad_norm": 9.542961142594972,
"learning_rate": 5.661116552399608e-07,
"loss": 1.2768,
"step": 290
},
{
"epoch": 0.0017104164952331691,
"grad_norm": 9.590134538210235,
"learning_rate": 5.680705190989227e-07,
"loss": 1.6988,
"step": 291
},
{
"epoch": 0.0017162942151480597,
"grad_norm": 10.325514884799286,
"learning_rate": 5.700293829578845e-07,
"loss": 1.4005,
"step": 292
},
{
"epoch": 0.0017221719350629505,
"grad_norm": 10.585596643296642,
"learning_rate": 5.719882468168463e-07,
"loss": 1.4279,
"step": 293
},
{
"epoch": 0.001728049654977841,
"grad_norm": 10.587814188427249,
"learning_rate": 5.739471106758081e-07,
"loss": 1.1622,
"step": 294
},
{
"epoch": 0.0017339273748927316,
"grad_norm": 8.114925702550204,
"learning_rate": 5.759059745347699e-07,
"loss": 1.2287,
"step": 295
},
{
"epoch": 0.0017398050948076221,
"grad_norm": 10.570600276337354,
"learning_rate": 5.778648383937317e-07,
"loss": 1.3479,
"step": 296
},
{
"epoch": 0.001745682814722513,
"grad_norm": 7.904815223532098,
"learning_rate": 5.798237022526935e-07,
"loss": 1.3332,
"step": 297
},
{
"epoch": 0.0017515605346374035,
"grad_norm": 8.907307462389863,
"learning_rate": 5.817825661116553e-07,
"loss": 1.081,
"step": 298
},
{
"epoch": 0.001757438254552294,
"grad_norm": 11.75059259667856,
"learning_rate": 5.83741429970617e-07,
"loss": 1.7135,
"step": 299
},
{
"epoch": 0.0017633159744671848,
"grad_norm": 9.072886413264408,
"learning_rate": 5.857002938295788e-07,
"loss": 1.6713,
"step": 300
},
{
"epoch": 0.0017691936943820753,
"grad_norm": 9.016576648856399,
"learning_rate": 5.876591576885407e-07,
"loss": 1.7791,
"step": 301
},
{
"epoch": 0.001775071414296966,
"grad_norm": 9.023767161545154,
"learning_rate": 5.896180215475025e-07,
"loss": 1.3835,
"step": 302
},
{
"epoch": 0.0017809491342118565,
"grad_norm": 9.962756123160855,
"learning_rate": 5.915768854064643e-07,
"loss": 1.6033,
"step": 303
},
{
"epoch": 0.0017868268541267472,
"grad_norm": 9.133093847326018,
"learning_rate": 5.935357492654261e-07,
"loss": 1.4126,
"step": 304
},
{
"epoch": 0.0017927045740416378,
"grad_norm": 11.5542021679122,
"learning_rate": 5.95494613124388e-07,
"loss": 1.2368,
"step": 305
},
{
"epoch": 0.0017985822939565283,
"grad_norm": 11.907934876986705,
"learning_rate": 5.974534769833497e-07,
"loss": 1.3877,
"step": 306
},
{
"epoch": 0.001804460013871419,
"grad_norm": 7.981708675663852,
"learning_rate": 5.994123408423115e-07,
"loss": 1.4267,
"step": 307
},
{
"epoch": 0.0018103377337863097,
"grad_norm": 8.352686578158465,
"learning_rate": 6.013712047012733e-07,
"loss": 1.1058,
"step": 308
},
{
"epoch": 0.0018162154537012002,
"grad_norm": 9.957225135648658,
"learning_rate": 6.033300685602351e-07,
"loss": 1.6818,
"step": 309
},
{
"epoch": 0.0018220931736160908,
"grad_norm": 9.90328827616875,
"learning_rate": 6.05288932419197e-07,
"loss": 1.5766,
"step": 310
},
{
"epoch": 0.0018279708935309816,
"grad_norm": 11.369184065206666,
"learning_rate": 6.072477962781588e-07,
"loss": 1.096,
"step": 311
},
{
"epoch": 0.0018338486134458721,
"grad_norm": 10.82625712962213,
"learning_rate": 6.092066601371205e-07,
"loss": 1.0124,
"step": 312
},
{
"epoch": 0.0018397263333607627,
"grad_norm": 11.626285416738192,
"learning_rate": 6.111655239960823e-07,
"loss": 1.3971,
"step": 313
},
{
"epoch": 0.0018456040532756532,
"grad_norm": 9.15673251705592,
"learning_rate": 6.131243878550441e-07,
"loss": 1.259,
"step": 314
},
{
"epoch": 0.001851481773190544,
"grad_norm": 11.418733881821664,
"learning_rate": 6.15083251714006e-07,
"loss": 1.4944,
"step": 315
},
{
"epoch": 0.0018573594931054346,
"grad_norm": 8.174275985243646,
"learning_rate": 6.170421155729677e-07,
"loss": 1.1997,
"step": 316
},
{
"epoch": 0.0018632372130203251,
"grad_norm": 9.180576750869545,
"learning_rate": 6.190009794319295e-07,
"loss": 1.3872,
"step": 317
},
{
"epoch": 0.0018691149329352157,
"grad_norm": 12.375955484763304,
"learning_rate": 6.209598432908913e-07,
"loss": 1.324,
"step": 318
},
{
"epoch": 0.0018749926528501065,
"grad_norm": 9.560556934953807,
"learning_rate": 6.229187071498531e-07,
"loss": 1.4166,
"step": 319
},
{
"epoch": 0.001880870372764997,
"grad_norm": 9.492269725987041,
"learning_rate": 6.24877571008815e-07,
"loss": 1.0062,
"step": 320
},
{
"epoch": 0.0018867480926798876,
"grad_norm": 10.202432437740011,
"learning_rate": 6.268364348677768e-07,
"loss": 1.4843,
"step": 321
},
{
"epoch": 0.0018926258125947781,
"grad_norm": 10.254012500981297,
"learning_rate": 6.287952987267386e-07,
"loss": 1.5466,
"step": 322
},
{
"epoch": 0.001898503532509669,
"grad_norm": 11.225147665082007,
"learning_rate": 6.307541625857004e-07,
"loss": 1.7476,
"step": 323
},
{
"epoch": 0.0019043812524245595,
"grad_norm": 8.162781954607768,
"learning_rate": 6.327130264446621e-07,
"loss": 1.4437,
"step": 324
},
{
"epoch": 0.00191025897233945,
"grad_norm": 9.079994444957677,
"learning_rate": 6.34671890303624e-07,
"loss": 1.1562,
"step": 325
},
{
"epoch": 0.0019161366922543408,
"grad_norm": 9.634816296804672,
"learning_rate": 6.366307541625858e-07,
"loss": 1.6029,
"step": 326
},
{
"epoch": 0.0019220144121692314,
"grad_norm": 11.76794240323632,
"learning_rate": 6.385896180215476e-07,
"loss": 1.2548,
"step": 327
},
{
"epoch": 0.001927892132084122,
"grad_norm": 9.010097944627907,
"learning_rate": 6.405484818805094e-07,
"loss": 1.2053,
"step": 328
},
{
"epoch": 0.0019337698519990125,
"grad_norm": 10.400030554541505,
"learning_rate": 6.425073457394713e-07,
"loss": 1.2347,
"step": 329
},
{
"epoch": 0.0019396475719139032,
"grad_norm": 9.580005301294532,
"learning_rate": 6.444662095984329e-07,
"loss": 1.441,
"step": 330
},
{
"epoch": 0.0019455252918287938,
"grad_norm": 95.9734686900897,
"learning_rate": 6.464250734573947e-07,
"loss": 1.6166,
"step": 331
},
{
"epoch": 0.0019514030117436844,
"grad_norm": 10.36525345635369,
"learning_rate": 6.483839373163565e-07,
"loss": 1.2509,
"step": 332
},
{
"epoch": 0.001957280731658575,
"grad_norm": 10.002390744057713,
"learning_rate": 6.503428011753183e-07,
"loss": 1.4298,
"step": 333
},
{
"epoch": 0.0019631584515734657,
"grad_norm": 8.217876579856577,
"learning_rate": 6.523016650342802e-07,
"loss": 1.2985,
"step": 334
},
{
"epoch": 0.001969036171488356,
"grad_norm": 10.009667776909204,
"learning_rate": 6.54260528893242e-07,
"loss": 1.7514,
"step": 335
},
{
"epoch": 0.001974913891403247,
"grad_norm": 8.951831687444894,
"learning_rate": 6.562193927522038e-07,
"loss": 1.5412,
"step": 336
},
{
"epoch": 0.0019807916113181376,
"grad_norm": 9.069755801521628,
"learning_rate": 6.581782566111655e-07,
"loss": 1.1909,
"step": 337
},
{
"epoch": 0.001986669331233028,
"grad_norm": 12.005699769054418,
"learning_rate": 6.601371204701273e-07,
"loss": 1.3262,
"step": 338
},
{
"epoch": 0.0019925470511479187,
"grad_norm": 9.69892721507397,
"learning_rate": 6.620959843290892e-07,
"loss": 1.3983,
"step": 339
},
{
"epoch": 0.0019984247710628095,
"grad_norm": 10.120949574403364,
"learning_rate": 6.64054848188051e-07,
"loss": 1.8783,
"step": 340
},
{
"epoch": 0.0020043024909777,
"grad_norm": 8.432805148499385,
"learning_rate": 6.660137120470128e-07,
"loss": 1.4789,
"step": 341
},
{
"epoch": 0.0020101802108925906,
"grad_norm": 9.690110308833203,
"learning_rate": 6.679725759059746e-07,
"loss": 1.2415,
"step": 342
},
{
"epoch": 0.0020160579308074814,
"grad_norm": 9.80807295529178,
"learning_rate": 6.699314397649363e-07,
"loss": 1.2757,
"step": 343
},
{
"epoch": 0.0020219356507223717,
"grad_norm": 9.956040308785957,
"learning_rate": 6.718903036238982e-07,
"loss": 1.2862,
"step": 344
},
{
"epoch": 0.0020278133706372625,
"grad_norm": 7.842215017092853,
"learning_rate": 6.7384916748286e-07,
"loss": 1.2398,
"step": 345
},
{
"epoch": 0.002033691090552153,
"grad_norm": 7.648481343331776,
"learning_rate": 6.758080313418218e-07,
"loss": 1.1184,
"step": 346
},
{
"epoch": 0.0020395688104670436,
"grad_norm": 12.194721433776252,
"learning_rate": 6.777668952007836e-07,
"loss": 1.7089,
"step": 347
},
{
"epoch": 0.0020454465303819344,
"grad_norm": 7.690015793298225,
"learning_rate": 6.797257590597454e-07,
"loss": 1.4625,
"step": 348
},
{
"epoch": 0.0020513242502968247,
"grad_norm": 11.691899229743722,
"learning_rate": 6.816846229187073e-07,
"loss": 1.2942,
"step": 349
},
{
"epoch": 0.0020572019702117155,
"grad_norm": 9.815989117817255,
"learning_rate": 6.83643486777669e-07,
"loss": 1.4441,
"step": 350
},
{
"epoch": 0.0020630796901266063,
"grad_norm": 7.700996093887788,
"learning_rate": 6.856023506366308e-07,
"loss": 1.2803,
"step": 351
},
{
"epoch": 0.0020689574100414966,
"grad_norm": 9.5858324599782,
"learning_rate": 6.875612144955926e-07,
"loss": 1.5798,
"step": 352
},
{
"epoch": 0.0020748351299563874,
"grad_norm": 10.225643669078298,
"learning_rate": 6.895200783545545e-07,
"loss": 1.4671,
"step": 353
},
{
"epoch": 0.002080712849871278,
"grad_norm": 12.74888973456531,
"learning_rate": 6.914789422135163e-07,
"loss": 1.2033,
"step": 354
},
{
"epoch": 0.0020865905697861685,
"grad_norm": 8.747052996779347,
"learning_rate": 6.934378060724781e-07,
"loss": 1.4468,
"step": 355
},
{
"epoch": 0.0020924682897010593,
"grad_norm": 10.224825910436632,
"learning_rate": 6.953966699314397e-07,
"loss": 1.3275,
"step": 356
},
{
"epoch": 0.0020983460096159496,
"grad_norm": 9.76538128069528,
"learning_rate": 6.973555337904015e-07,
"loss": 1.2248,
"step": 357
},
{
"epoch": 0.0021042237295308404,
"grad_norm": 9.343421907990354,
"learning_rate": 6.993143976493634e-07,
"loss": 1.4222,
"step": 358
},
{
"epoch": 0.002110101449445731,
"grad_norm": 11.07607790792727,
"learning_rate": 7.012732615083252e-07,
"loss": 1.4655,
"step": 359
},
{
"epoch": 0.0021159791693606215,
"grad_norm": 9.43492641142549,
"learning_rate": 7.032321253672871e-07,
"loss": 1.3952,
"step": 360
},
{
"epoch": 0.0021218568892755123,
"grad_norm": 10.725935879524329,
"learning_rate": 7.051909892262489e-07,
"loss": 1.3749,
"step": 361
},
{
"epoch": 0.002127734609190403,
"grad_norm": 10.459955763777726,
"learning_rate": 7.071498530852107e-07,
"loss": 1.3638,
"step": 362
},
{
"epoch": 0.0021336123291052934,
"grad_norm": 10.084883781318648,
"learning_rate": 7.091087169441724e-07,
"loss": 1.5208,
"step": 363
},
{
"epoch": 0.002139490049020184,
"grad_norm": 10.037233174623386,
"learning_rate": 7.110675808031342e-07,
"loss": 1.2152,
"step": 364
},
{
"epoch": 0.002145367768935075,
"grad_norm": 8.813983991816277,
"learning_rate": 7.13026444662096e-07,
"loss": 1.6647,
"step": 365
},
{
"epoch": 0.0021512454888499653,
"grad_norm": 12.20631481651128,
"learning_rate": 7.149853085210578e-07,
"loss": 1.6276,
"step": 366
},
{
"epoch": 0.002157123208764856,
"grad_norm": 12.98213610184426,
"learning_rate": 7.169441723800196e-07,
"loss": 1.5915,
"step": 367
},
{
"epoch": 0.0021630009286797464,
"grad_norm": 9.21805497063901,
"learning_rate": 7.189030362389815e-07,
"loss": 1.336,
"step": 368
},
{
"epoch": 0.002168878648594637,
"grad_norm": 9.109226384013539,
"learning_rate": 7.208619000979432e-07,
"loss": 1.6266,
"step": 369
},
{
"epoch": 0.002174756368509528,
"grad_norm": 10.455222812127753,
"learning_rate": 7.22820763956905e-07,
"loss": 1.262,
"step": 370
},
{
"epoch": 0.0021806340884244183,
"grad_norm": 8.391319723021606,
"learning_rate": 7.247796278158668e-07,
"loss": 1.1648,
"step": 371
},
{
"epoch": 0.002186511808339309,
"grad_norm": 120.89278453800495,
"learning_rate": 7.267384916748286e-07,
"loss": 1.4725,
"step": 372
},
{
"epoch": 0.0021923895282542,
"grad_norm": 10.561377041726836,
"learning_rate": 7.286973555337905e-07,
"loss": 1.2098,
"step": 373
},
{
"epoch": 0.00219826724816909,
"grad_norm": 10.679479268602012,
"learning_rate": 7.306562193927523e-07,
"loss": 1.3308,
"step": 374
},
{
"epoch": 0.002204144968083981,
"grad_norm": 8.485373095871307,
"learning_rate": 7.326150832517141e-07,
"loss": 1.5223,
"step": 375
},
{
"epoch": 0.0022100226879988713,
"grad_norm": 8.997398533127315,
"learning_rate": 7.345739471106758e-07,
"loss": 1.4203,
"step": 376
},
{
"epoch": 0.002215900407913762,
"grad_norm": 9.068897975931213,
"learning_rate": 7.365328109696377e-07,
"loss": 1.257,
"step": 377
},
{
"epoch": 0.002221778127828653,
"grad_norm": 9.307758093247212,
"learning_rate": 7.384916748285995e-07,
"loss": 1.0995,
"step": 378
},
{
"epoch": 0.002227655847743543,
"grad_norm": 10.070537945717394,
"learning_rate": 7.404505386875613e-07,
"loss": 1.7374,
"step": 379
},
{
"epoch": 0.002233533567658434,
"grad_norm": 10.366005280163792,
"learning_rate": 7.424094025465231e-07,
"loss": 1.2628,
"step": 380
},
{
"epoch": 0.0022394112875733247,
"grad_norm": 9.08901956278531,
"learning_rate": 7.443682664054849e-07,
"loss": 1.1628,
"step": 381
},
{
"epoch": 0.002245289007488215,
"grad_norm": 9.146350725847851,
"learning_rate": 7.463271302644467e-07,
"loss": 1.5435,
"step": 382
},
{
"epoch": 0.002251166727403106,
"grad_norm": 11.350273071816241,
"learning_rate": 7.482859941234085e-07,
"loss": 1.4359,
"step": 383
},
{
"epoch": 0.0022570444473179966,
"grad_norm": 10.745208607274986,
"learning_rate": 7.502448579823703e-07,
"loss": 1.5371,
"step": 384
},
{
"epoch": 0.002262922167232887,
"grad_norm": 9.133230654048408,
"learning_rate": 7.522037218413321e-07,
"loss": 1.2589,
"step": 385
},
{
"epoch": 0.0022687998871477777,
"grad_norm": 9.919975618980674,
"learning_rate": 7.541625857002939e-07,
"loss": 1.5764,
"step": 386
},
{
"epoch": 0.002274677607062668,
"grad_norm": 10.015249785262874,
"learning_rate": 7.561214495592558e-07,
"loss": 1.2103,
"step": 387
},
{
"epoch": 0.002280555326977559,
"grad_norm": 9.298714026316576,
"learning_rate": 7.580803134182176e-07,
"loss": 1.2527,
"step": 388
},
{
"epoch": 0.0022864330468924496,
"grad_norm": 7.855421719465326,
"learning_rate": 7.600391772771792e-07,
"loss": 0.9689,
"step": 389
},
{
"epoch": 0.00229231076680734,
"grad_norm": 8.19603172426781,
"learning_rate": 7.61998041136141e-07,
"loss": 1.5125,
"step": 390
},
{
"epoch": 0.0022981884867222307,
"grad_norm": 9.355341043777715,
"learning_rate": 7.639569049951028e-07,
"loss": 1.0824,
"step": 391
},
{
"epoch": 0.0023040662066371215,
"grad_norm": 7.6510926135371875,
"learning_rate": 7.659157688540647e-07,
"loss": 1.3634,
"step": 392
},
{
"epoch": 0.002309943926552012,
"grad_norm": 10.424234163569968,
"learning_rate": 7.678746327130265e-07,
"loss": 1.6378,
"step": 393
},
{
"epoch": 0.0023158216464669026,
"grad_norm": 9.89076424202504,
"learning_rate": 7.698334965719883e-07,
"loss": 1.3423,
"step": 394
},
{
"epoch": 0.0023216993663817934,
"grad_norm": 12.940539054812946,
"learning_rate": 7.7179236043095e-07,
"loss": 1.2945,
"step": 395
},
{
"epoch": 0.0023275770862966837,
"grad_norm": 7.680461776345431,
"learning_rate": 7.737512242899118e-07,
"loss": 1.0659,
"step": 396
},
{
"epoch": 0.0023334548062115745,
"grad_norm": 8.63479952985547,
"learning_rate": 7.757100881488737e-07,
"loss": 1.4686,
"step": 397
},
{
"epoch": 0.002339332526126465,
"grad_norm": 8.979262603318334,
"learning_rate": 7.776689520078355e-07,
"loss": 1.0556,
"step": 398
},
{
"epoch": 0.0023452102460413556,
"grad_norm": 10.076704637113531,
"learning_rate": 7.796278158667973e-07,
"loss": 1.4495,
"step": 399
},
{
"epoch": 0.0023510879659562464,
"grad_norm": 8.951312030590396,
"learning_rate": 7.815866797257591e-07,
"loss": 1.6132,
"step": 400
},
{
"epoch": 0.0023569656858711367,
"grad_norm": 8.979873542188797,
"learning_rate": 7.83545543584721e-07,
"loss": 1.3542,
"step": 401
},
{
"epoch": 0.0023628434057860275,
"grad_norm": 9.51825670965823,
"learning_rate": 7.855044074436827e-07,
"loss": 1.402,
"step": 402
},
{
"epoch": 0.0023687211257009183,
"grad_norm": 8.896133514148623,
"learning_rate": 7.874632713026445e-07,
"loss": 1.2905,
"step": 403
},
{
"epoch": 0.0023745988456158086,
"grad_norm": 9.451231888455286,
"learning_rate": 7.894221351616063e-07,
"loss": 1.4826,
"step": 404
},
{
"epoch": 0.0023804765655306994,
"grad_norm": 10.471983965291152,
"learning_rate": 7.913809990205681e-07,
"loss": 1.3434,
"step": 405
},
{
"epoch": 0.00238635428544559,
"grad_norm": 8.795466845801405,
"learning_rate": 7.9333986287953e-07,
"loss": 1.4804,
"step": 406
},
{
"epoch": 0.0023922320053604805,
"grad_norm": 8.395943785928992,
"learning_rate": 7.952987267384918e-07,
"loss": 1.3481,
"step": 407
},
{
"epoch": 0.0023981097252753713,
"grad_norm": 8.57882860152985,
"learning_rate": 7.972575905974535e-07,
"loss": 1.4464,
"step": 408
},
{
"epoch": 0.0024039874451902616,
"grad_norm": 9.311372146946045,
"learning_rate": 7.992164544564153e-07,
"loss": 1.4611,
"step": 409
},
{
"epoch": 0.0024098651651051524,
"grad_norm": 8.618542789091416,
"learning_rate": 8.011753183153771e-07,
"loss": 1.323,
"step": 410
},
{
"epoch": 0.002415742885020043,
"grad_norm": 9.51907141373062,
"learning_rate": 8.03134182174339e-07,
"loss": 1.1481,
"step": 411
},
{
"epoch": 0.0024216206049349335,
"grad_norm": 10.019154850223428,
"learning_rate": 8.050930460333008e-07,
"loss": 1.243,
"step": 412
},
{
"epoch": 0.0024274983248498243,
"grad_norm": 9.326959371627055,
"learning_rate": 8.070519098922626e-07,
"loss": 1.3822,
"step": 413
},
{
"epoch": 0.002433376044764715,
"grad_norm": 9.228495426900992,
"learning_rate": 8.090107737512244e-07,
"loss": 1.1939,
"step": 414
},
{
"epoch": 0.0024392537646796054,
"grad_norm": 9.307552867358865,
"learning_rate": 8.109696376101861e-07,
"loss": 1.3198,
"step": 415
},
{
"epoch": 0.002445131484594496,
"grad_norm": 8.463031761866146,
"learning_rate": 8.12928501469148e-07,
"loss": 1.4693,
"step": 416
},
{
"epoch": 0.0024510092045093865,
"grad_norm": 9.192023455695411,
"learning_rate": 8.148873653281098e-07,
"loss": 1.2851,
"step": 417
},
{
"epoch": 0.0024568869244242773,
"grad_norm": 8.852688157505836,
"learning_rate": 8.168462291870716e-07,
"loss": 1.3559,
"step": 418
},
{
"epoch": 0.002462764644339168,
"grad_norm": 10.27985063998931,
"learning_rate": 8.188050930460334e-07,
"loss": 1.5606,
"step": 419
},
{
"epoch": 0.0024686423642540584,
"grad_norm": 8.299719168895864,
"learning_rate": 8.207639569049953e-07,
"loss": 1.1143,
"step": 420
},
{
"epoch": 0.002474520084168949,
"grad_norm": 9.944928729713705,
"learning_rate": 8.227228207639569e-07,
"loss": 1.3194,
"step": 421
},
{
"epoch": 0.00248039780408384,
"grad_norm": 12.03407092301725,
"learning_rate": 8.246816846229187e-07,
"loss": 1.4087,
"step": 422
},
{
"epoch": 0.0024862755239987303,
"grad_norm": 14.891508805712057,
"learning_rate": 8.266405484818805e-07,
"loss": 1.4645,
"step": 423
},
{
"epoch": 0.002492153243913621,
"grad_norm": 9.1187340513171,
"learning_rate": 8.285994123408423e-07,
"loss": 1.1926,
"step": 424
},
{
"epoch": 0.002498030963828512,
"grad_norm": 9.319764152313796,
"learning_rate": 8.305582761998042e-07,
"loss": 1.2343,
"step": 425
},
{
"epoch": 0.002503908683743402,
"grad_norm": 10.484381391020465,
"learning_rate": 8.32517140058766e-07,
"loss": 1.2541,
"step": 426
},
{
"epoch": 0.002509786403658293,
"grad_norm": 8.69167001436625,
"learning_rate": 8.344760039177278e-07,
"loss": 1.2757,
"step": 427
},
{
"epoch": 0.0025156641235731833,
"grad_norm": 10.430608969858419,
"learning_rate": 8.364348677766895e-07,
"loss": 1.3012,
"step": 428
},
{
"epoch": 0.002521541843488074,
"grad_norm": 9.637826577295606,
"learning_rate": 8.383937316356513e-07,
"loss": 1.4526,
"step": 429
},
{
"epoch": 0.002527419563402965,
"grad_norm": 9.875465863294172,
"learning_rate": 8.403525954946132e-07,
"loss": 1.3497,
"step": 430
},
{
"epoch": 0.002533297283317855,
"grad_norm": 10.327297011939747,
"learning_rate": 8.42311459353575e-07,
"loss": 1.6876,
"step": 431
},
{
"epoch": 0.002539175003232746,
"grad_norm": 8.944459341621092,
"learning_rate": 8.442703232125368e-07,
"loss": 1.557,
"step": 432
},
{
"epoch": 0.0025450527231476367,
"grad_norm": 9.38025206199169,
"learning_rate": 8.462291870714986e-07,
"loss": 1.4929,
"step": 433
},
{
"epoch": 0.002550930443062527,
"grad_norm": 6.7474803815775894,
"learning_rate": 8.481880509304604e-07,
"loss": 1.1632,
"step": 434
},
{
"epoch": 0.002556808162977418,
"grad_norm": 9.316752827855126,
"learning_rate": 8.501469147894222e-07,
"loss": 1.555,
"step": 435
},
{
"epoch": 0.0025626858828923086,
"grad_norm": 8.950472923863874,
"learning_rate": 8.52105778648384e-07,
"loss": 1.5023,
"step": 436
},
{
"epoch": 0.002568563602807199,
"grad_norm": 7.877832954575786,
"learning_rate": 8.540646425073458e-07,
"loss": 1.2223,
"step": 437
},
{
"epoch": 0.0025744413227220897,
"grad_norm": 11.13541528348567,
"learning_rate": 8.560235063663076e-07,
"loss": 1.4983,
"step": 438
},
{
"epoch": 0.00258031904263698,
"grad_norm": 10.284682190541375,
"learning_rate": 8.579823702252694e-07,
"loss": 1.4565,
"step": 439
},
{
"epoch": 0.002586196762551871,
"grad_norm": 7.154879985048399,
"learning_rate": 8.599412340842313e-07,
"loss": 0.9219,
"step": 440
},
{
"epoch": 0.0025920744824667616,
"grad_norm": 12.374858918870638,
"learning_rate": 8.61900097943193e-07,
"loss": 1.3466,
"step": 441
},
{
"epoch": 0.002597952202381652,
"grad_norm": 9.530586145055318,
"learning_rate": 8.638589618021548e-07,
"loss": 1.2966,
"step": 442
},
{
"epoch": 0.0026038299222965427,
"grad_norm": 10.52530275859048,
"learning_rate": 8.658178256611166e-07,
"loss": 1.535,
"step": 443
},
{
"epoch": 0.0026097076422114335,
"grad_norm": 11.113832609940571,
"learning_rate": 8.677766895200785e-07,
"loss": 1.2633,
"step": 444
},
{
"epoch": 0.002615585362126324,
"grad_norm": 9.893958350977293,
"learning_rate": 8.697355533790403e-07,
"loss": 1.1753,
"step": 445
},
{
"epoch": 0.0026214630820412146,
"grad_norm": 9.699442687702016,
"learning_rate": 8.716944172380021e-07,
"loss": 1.2883,
"step": 446
},
{
"epoch": 0.0026273408019561054,
"grad_norm": 8.627446824111422,
"learning_rate": 8.736532810969639e-07,
"loss": 1.2994,
"step": 447
},
{
"epoch": 0.0026332185218709957,
"grad_norm": 9.18927129378313,
"learning_rate": 8.756121449559256e-07,
"loss": 1.0687,
"step": 448
},
{
"epoch": 0.0026390962417858865,
"grad_norm": 11.084942651516588,
"learning_rate": 8.775710088148875e-07,
"loss": 1.2795,
"step": 449
},
{
"epoch": 0.002644973961700777,
"grad_norm": 9.129354706792896,
"learning_rate": 8.795298726738493e-07,
"loss": 1.3835,
"step": 450
},
{
"epoch": 0.0026508516816156676,
"grad_norm": 12.327908421949875,
"learning_rate": 8.814887365328111e-07,
"loss": 1.5225,
"step": 451
},
{
"epoch": 0.0026567294015305584,
"grad_norm": 8.31873640235426,
"learning_rate": 8.834476003917729e-07,
"loss": 0.8396,
"step": 452
},
{
"epoch": 0.0026626071214454487,
"grad_norm": 9.854541808925802,
"learning_rate": 8.854064642507347e-07,
"loss": 1.5565,
"step": 453
},
{
"epoch": 0.0026684848413603395,
"grad_norm": 9.985607380778386,
"learning_rate": 8.873653281096964e-07,
"loss": 1.342,
"step": 454
},
{
"epoch": 0.0026743625612752303,
"grad_norm": 8.500019891152785,
"learning_rate": 8.893241919686582e-07,
"loss": 1.2752,
"step": 455
},
{
"epoch": 0.0026802402811901206,
"grad_norm": 9.733680853033052,
"learning_rate": 8.9128305582762e-07,
"loss": 1.337,
"step": 456
},
{
"epoch": 0.0026861180011050114,
"grad_norm": 8.956767161220945,
"learning_rate": 8.932419196865818e-07,
"loss": 1.2516,
"step": 457
},
{
"epoch": 0.0026919957210199017,
"grad_norm": 8.920962685189888,
"learning_rate": 8.952007835455436e-07,
"loss": 1.2214,
"step": 458
},
{
"epoch": 0.0026978734409347925,
"grad_norm": 7.890385891017338,
"learning_rate": 8.971596474045055e-07,
"loss": 1.2096,
"step": 459
},
{
"epoch": 0.0027037511608496833,
"grad_norm": 8.718195916912373,
"learning_rate": 8.991185112634673e-07,
"loss": 1.4737,
"step": 460
},
{
"epoch": 0.0027096288807645736,
"grad_norm": 9.678136630934903,
"learning_rate": 9.01077375122429e-07,
"loss": 1.4187,
"step": 461
},
{
"epoch": 0.0027155066006794644,
"grad_norm": 9.23524546999622,
"learning_rate": 9.030362389813908e-07,
"loss": 1.6633,
"step": 462
},
{
"epoch": 0.002721384320594355,
"grad_norm": 9.32560092969864,
"learning_rate": 9.049951028403526e-07,
"loss": 1.1974,
"step": 463
},
{
"epoch": 0.0027272620405092455,
"grad_norm": 9.528071260395997,
"learning_rate": 9.069539666993145e-07,
"loss": 1.1997,
"step": 464
},
{
"epoch": 0.0027331397604241363,
"grad_norm": 10.425385414374023,
"learning_rate": 9.089128305582763e-07,
"loss": 1.3841,
"step": 465
},
{
"epoch": 0.002739017480339027,
"grad_norm": 9.650811170590984,
"learning_rate": 9.108716944172381e-07,
"loss": 1.445,
"step": 466
},
{
"epoch": 0.0027448952002539174,
"grad_norm": 9.252989896370744,
"learning_rate": 9.128305582761998e-07,
"loss": 1.4595,
"step": 467
},
{
"epoch": 0.002750772920168808,
"grad_norm": 8.799240937283905,
"learning_rate": 9.147894221351617e-07,
"loss": 1.3579,
"step": 468
},
{
"epoch": 0.0027566506400836985,
"grad_norm": 9.687119531664132,
"learning_rate": 9.167482859941235e-07,
"loss": 1.3415,
"step": 469
},
{
"epoch": 0.0027625283599985893,
"grad_norm": 8.093757997436107,
"learning_rate": 9.187071498530853e-07,
"loss": 1.0167,
"step": 470
},
{
"epoch": 0.00276840607991348,
"grad_norm": 8.141074348047649,
"learning_rate": 9.206660137120471e-07,
"loss": 1.5446,
"step": 471
},
{
"epoch": 0.0027742837998283704,
"grad_norm": 8.677237791994125,
"learning_rate": 9.226248775710089e-07,
"loss": 1.5067,
"step": 472
},
{
"epoch": 0.002780161519743261,
"grad_norm": 10.61789466008841,
"learning_rate": 9.245837414299708e-07,
"loss": 1.3241,
"step": 473
},
{
"epoch": 0.002786039239658152,
"grad_norm": 8.683931156910312,
"learning_rate": 9.265426052889325e-07,
"loss": 1.8794,
"step": 474
},
{
"epoch": 0.0027919169595730423,
"grad_norm": 12.103438115936399,
"learning_rate": 9.285014691478943e-07,
"loss": 1.5244,
"step": 475
},
{
"epoch": 0.002797794679487933,
"grad_norm": 9.337103354710932,
"learning_rate": 9.304603330068561e-07,
"loss": 1.1772,
"step": 476
},
{
"epoch": 0.002803672399402824,
"grad_norm": 7.2667453744292265,
"learning_rate": 9.324191968658179e-07,
"loss": 1.336,
"step": 477
},
{
"epoch": 0.002809550119317714,
"grad_norm": 7.7598194126940205,
"learning_rate": 9.343780607247798e-07,
"loss": 1.1973,
"step": 478
},
{
"epoch": 0.002815427839232605,
"grad_norm": 8.149477213248854,
"learning_rate": 9.363369245837416e-07,
"loss": 0.97,
"step": 479
},
{
"epoch": 0.0028213055591474953,
"grad_norm": 9.690960617524013,
"learning_rate": 9.382957884427032e-07,
"loss": 1.5165,
"step": 480
},
{
"epoch": 0.002827183279062386,
"grad_norm": 7.931888369964113,
"learning_rate": 9.40254652301665e-07,
"loss": 1.2378,
"step": 481
},
{
"epoch": 0.002833060998977277,
"grad_norm": 10.600482057686264,
"learning_rate": 9.42213516160627e-07,
"loss": 1.4659,
"step": 482
},
{
"epoch": 0.002838938718892167,
"grad_norm": 7.766775925211392,
"learning_rate": 9.441723800195888e-07,
"loss": 1.3378,
"step": 483
},
{
"epoch": 0.002844816438807058,
"grad_norm": 8.567731069736155,
"learning_rate": 9.461312438785506e-07,
"loss": 1.2967,
"step": 484
},
{
"epoch": 0.0028506941587219487,
"grad_norm": 8.798070275014695,
"learning_rate": 9.480901077375124e-07,
"loss": 1.696,
"step": 485
},
{
"epoch": 0.002856571878636839,
"grad_norm": 9.728901498213226,
"learning_rate": 9.500489715964742e-07,
"loss": 1.412,
"step": 486
},
{
"epoch": 0.00286244959855173,
"grad_norm": 11.037063609664564,
"learning_rate": 9.520078354554358e-07,
"loss": 1.5683,
"step": 487
},
{
"epoch": 0.0028683273184666206,
"grad_norm": 10.07693872625589,
"learning_rate": 9.539666993143978e-07,
"loss": 1.3795,
"step": 488
},
{
"epoch": 0.002874205038381511,
"grad_norm": 9.198521301044554,
"learning_rate": 9.559255631733596e-07,
"loss": 1.3621,
"step": 489
},
{
"epoch": 0.0028800827582964018,
"grad_norm": 9.072594042611673,
"learning_rate": 9.578844270323214e-07,
"loss": 1.2669,
"step": 490
},
{
"epoch": 0.002885960478211292,
"grad_norm": 10.420415483294377,
"learning_rate": 9.598432908912832e-07,
"loss": 1.2451,
"step": 491
},
{
"epoch": 0.002891838198126183,
"grad_norm": 8.650756741993991,
"learning_rate": 9.61802154750245e-07,
"loss": 1.3327,
"step": 492
},
{
"epoch": 0.0028977159180410736,
"grad_norm": 10.699415230564549,
"learning_rate": 9.637610186092067e-07,
"loss": 1.3475,
"step": 493
},
{
"epoch": 0.002903593637955964,
"grad_norm": 10.252178027073803,
"learning_rate": 9.657198824681685e-07,
"loss": 1.6232,
"step": 494
},
{
"epoch": 0.0029094713578708548,
"grad_norm": 8.058950860691148,
"learning_rate": 9.676787463271303e-07,
"loss": 1.3211,
"step": 495
},
{
"epoch": 0.0029153490777857455,
"grad_norm": 10.007337786160015,
"learning_rate": 9.696376101860921e-07,
"loss": 1.5046,
"step": 496
},
{
"epoch": 0.002921226797700636,
"grad_norm": 7.85615985186399,
"learning_rate": 9.71596474045054e-07,
"loss": 1.1484,
"step": 497
},
{
"epoch": 0.0029271045176155266,
"grad_norm": 8.540344613450793,
"learning_rate": 9.735553379040158e-07,
"loss": 1.4431,
"step": 498
},
{
"epoch": 0.002932982237530417,
"grad_norm": 8.896907762371791,
"learning_rate": 9.755142017629776e-07,
"loss": 1.1179,
"step": 499
},
{
"epoch": 0.0029388599574453078,
"grad_norm": 8.91483404955683,
"learning_rate": 9.774730656219392e-07,
"loss": 1.9639,
"step": 500
}
],
"logging_steps": 1.0,
"max_steps": 170134,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7987300275200.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}