Experimental100percent_RAG / trainer_state.json
crodri's picture
Version3
ebdbe4e verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9984871406959153,
"eval_steps": 83,
"global_step": 330,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0030257186081694403,
"grad_norm": 1652051746605.6438,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.1213,
"step": 1
},
{
"epoch": 0.006051437216338881,
"grad_norm": 1778479170004.0757,
"learning_rate": 4.000000000000001e-06,
"loss": 1.1062,
"step": 2
},
{
"epoch": 0.009077155824508321,
"grad_norm": 274899.1962126193,
"learning_rate": 6e-06,
"loss": 1.2468,
"step": 3
},
{
"epoch": 0.012102874432677761,
"grad_norm": 247.59631897939568,
"learning_rate": 8.000000000000001e-06,
"loss": 1.159,
"step": 4
},
{
"epoch": 0.015128593040847202,
"grad_norm": 1881.4007972140926,
"learning_rate": 1e-05,
"loss": 0.9184,
"step": 5
},
{
"epoch": 0.018154311649016642,
"grad_norm": 233.1531795666884,
"learning_rate": 1.2e-05,
"loss": 0.8803,
"step": 6
},
{
"epoch": 0.02118003025718608,
"grad_norm": 124.50819835758877,
"learning_rate": 1.4e-05,
"loss": 0.7788,
"step": 7
},
{
"epoch": 0.024205748865355523,
"grad_norm": 171.9635211709154,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.7173,
"step": 8
},
{
"epoch": 0.02723146747352496,
"grad_norm": 88.07379152036668,
"learning_rate": 1.8e-05,
"loss": 0.7042,
"step": 9
},
{
"epoch": 0.030257186081694403,
"grad_norm": 45.22136677949116,
"learning_rate": 2e-05,
"loss": 0.7075,
"step": 10
},
{
"epoch": 0.03328290468986384,
"grad_norm": 51.8972891247439,
"learning_rate": 1.999951808959328e-05,
"loss": 0.7105,
"step": 11
},
{
"epoch": 0.036308623298033284,
"grad_norm": 2.9027805574302725,
"learning_rate": 1.9998072404820648e-05,
"loss": 0.6944,
"step": 12
},
{
"epoch": 0.039334341906202726,
"grad_norm": 3.1018685838791966,
"learning_rate": 1.9995663085020215e-05,
"loss": 0.6577,
"step": 13
},
{
"epoch": 0.04236006051437216,
"grad_norm": 2.1407017926215075,
"learning_rate": 1.9992290362407232e-05,
"loss": 0.6657,
"step": 14
},
{
"epoch": 0.0453857791225416,
"grad_norm": 2.7816290985005927,
"learning_rate": 1.9987954562051724e-05,
"loss": 0.7061,
"step": 15
},
{
"epoch": 0.048411497730711045,
"grad_norm": 3.5947181367716006,
"learning_rate": 1.998265610184716e-05,
"loss": 0.7033,
"step": 16
},
{
"epoch": 0.05143721633888049,
"grad_norm": 2.101093929503267,
"learning_rate": 1.997639549247016e-05,
"loss": 0.6033,
"step": 17
},
{
"epoch": 0.05446293494704992,
"grad_norm": 2.2706543555260597,
"learning_rate": 1.9969173337331283e-05,
"loss": 0.6778,
"step": 18
},
{
"epoch": 0.057488653555219364,
"grad_norm": 1.9878621193889954,
"learning_rate": 1.9960990332516875e-05,
"loss": 0.6914,
"step": 19
},
{
"epoch": 0.060514372163388806,
"grad_norm": 1.9054217558209654,
"learning_rate": 1.995184726672197e-05,
"loss": 0.686,
"step": 20
},
{
"epoch": 0.06354009077155824,
"grad_norm": 2.0927903071386775,
"learning_rate": 1.9941745021174284e-05,
"loss": 0.6513,
"step": 21
},
{
"epoch": 0.06656580937972768,
"grad_norm": 1.9382772994863178,
"learning_rate": 1.9930684569549265e-05,
"loss": 0.6411,
"step": 22
},
{
"epoch": 0.06959152798789713,
"grad_norm": 2.0967174020483164,
"learning_rate": 1.991866697787626e-05,
"loss": 0.7124,
"step": 23
},
{
"epoch": 0.07261724659606657,
"grad_norm": 1.9046195794908993,
"learning_rate": 1.990569340443577e-05,
"loss": 0.6807,
"step": 24
},
{
"epoch": 0.07564296520423601,
"grad_norm": 1.8130874126282714,
"learning_rate": 1.989176509964781e-05,
"loss": 0.6552,
"step": 25
},
{
"epoch": 0.07866868381240545,
"grad_norm": 1.6629030784378982,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.6409,
"step": 26
},
{
"epoch": 0.08169440242057488,
"grad_norm": 1.8058763702424092,
"learning_rate": 1.9861049757675087e-05,
"loss": 0.6484,
"step": 27
},
{
"epoch": 0.08472012102874432,
"grad_norm": 1.7437581075031185,
"learning_rate": 1.9844265680898917e-05,
"loss": 0.6591,
"step": 28
},
{
"epoch": 0.08774583963691376,
"grad_norm": 1.8926901055586824,
"learning_rate": 1.982653279330712e-05,
"loss": 0.6593,
"step": 29
},
{
"epoch": 0.0907715582450832,
"grad_norm": 1.7627809439350672,
"learning_rate": 1.9807852804032306e-05,
"loss": 0.6706,
"step": 30
},
{
"epoch": 0.09379727685325265,
"grad_norm": 1.7294426775990228,
"learning_rate": 1.9788227513490724e-05,
"loss": 0.653,
"step": 31
},
{
"epoch": 0.09682299546142209,
"grad_norm": 1.651276766521114,
"learning_rate": 1.9767658813208725e-05,
"loss": 0.6462,
"step": 32
},
{
"epoch": 0.09984871406959153,
"grad_norm": 1.6707091763587558,
"learning_rate": 1.974614868564045e-05,
"loss": 0.6634,
"step": 33
},
{
"epoch": 0.10287443267776097,
"grad_norm": 1.723552336963672,
"learning_rate": 1.9723699203976768e-05,
"loss": 0.6705,
"step": 34
},
{
"epoch": 0.1059001512859304,
"grad_norm": 1.6975125647885916,
"learning_rate": 1.9700312531945444e-05,
"loss": 0.6382,
"step": 35
},
{
"epoch": 0.10892586989409984,
"grad_norm": 2.010440875780811,
"learning_rate": 1.96759909236026e-05,
"loss": 0.6778,
"step": 36
},
{
"epoch": 0.11195158850226929,
"grad_norm": 1.7330240113206155,
"learning_rate": 1.9650736723115476e-05,
"loss": 0.6554,
"step": 37
},
{
"epoch": 0.11497730711043873,
"grad_norm": 1.7415997994889654,
"learning_rate": 1.9624552364536472e-05,
"loss": 0.6749,
"step": 38
},
{
"epoch": 0.11800302571860817,
"grad_norm": 1.7300907573327364,
"learning_rate": 1.9597440371568576e-05,
"loss": 0.6354,
"step": 39
},
{
"epoch": 0.12102874432677761,
"grad_norm": 1.6556998625513375,
"learning_rate": 1.956940335732209e-05,
"loss": 0.6446,
"step": 40
},
{
"epoch": 0.12405446293494705,
"grad_norm": 1.6449161950791809,
"learning_rate": 1.9540444024062807e-05,
"loss": 0.6545,
"step": 41
},
{
"epoch": 0.12708018154311648,
"grad_norm": 1.8571524167724058,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.6923,
"step": 42
},
{
"epoch": 0.13010590015128592,
"grad_norm": 1.5788885742665648,
"learning_rate": 1.9479769653775107e-05,
"loss": 0.658,
"step": 43
},
{
"epoch": 0.13313161875945537,
"grad_norm": 1.5994342314996153,
"learning_rate": 1.944806046466878e-05,
"loss": 0.6508,
"step": 44
},
{
"epoch": 0.1361573373676248,
"grad_norm": 1.5984139005953688,
"learning_rate": 1.941544065183021e-05,
"loss": 0.6567,
"step": 45
},
{
"epoch": 0.13918305597579425,
"grad_norm": 1.7944744924580214,
"learning_rate": 1.9381913359224844e-05,
"loss": 0.6441,
"step": 46
},
{
"epoch": 0.1422087745839637,
"grad_norm": 1.6884803092360563,
"learning_rate": 1.9347481818282927e-05,
"loss": 0.6576,
"step": 47
},
{
"epoch": 0.14523449319213314,
"grad_norm": 1.720111829128096,
"learning_rate": 1.9312149347588035e-05,
"loss": 0.6597,
"step": 48
},
{
"epoch": 0.14826021180030258,
"grad_norm": 2.063038942605161,
"learning_rate": 1.9275919352557242e-05,
"loss": 0.6741,
"step": 49
},
{
"epoch": 0.15128593040847202,
"grad_norm": 1.7530517631586846,
"learning_rate": 1.9238795325112867e-05,
"loss": 0.6475,
"step": 50
},
{
"epoch": 0.15431164901664146,
"grad_norm": 1.5368739066680468,
"learning_rate": 1.920078084334595e-05,
"loss": 0.631,
"step": 51
},
{
"epoch": 0.1573373676248109,
"grad_norm": 1.5927253485027195,
"learning_rate": 1.916187957117136e-05,
"loss": 0.6558,
"step": 52
},
{
"epoch": 0.16036308623298035,
"grad_norm": 1.4646153212795487,
"learning_rate": 1.9122095257974676e-05,
"loss": 0.6107,
"step": 53
},
{
"epoch": 0.16338880484114976,
"grad_norm": 1.460503618351299,
"learning_rate": 1.9081431738250815e-05,
"loss": 0.6179,
"step": 54
},
{
"epoch": 0.1664145234493192,
"grad_norm": 1.714743332479254,
"learning_rate": 1.9039892931234434e-05,
"loss": 0.6479,
"step": 55
},
{
"epoch": 0.16944024205748864,
"grad_norm": 1.62677711384979,
"learning_rate": 1.8997482840522218e-05,
"loss": 0.6051,
"step": 56
},
{
"epoch": 0.17246596066565809,
"grad_norm": 1.5435569978486443,
"learning_rate": 1.895420555368697e-05,
"loss": 0.658,
"step": 57
},
{
"epoch": 0.17549167927382753,
"grad_norm": 1.6587428756301992,
"learning_rate": 1.891006524188368e-05,
"loss": 0.6637,
"step": 58
},
{
"epoch": 0.17851739788199697,
"grad_norm": 1.5652628120146788,
"learning_rate": 1.8865066159447468e-05,
"loss": 0.6333,
"step": 59
},
{
"epoch": 0.1815431164901664,
"grad_norm": 1.5444986686475162,
"learning_rate": 1.881921264348355e-05,
"loss": 0.6572,
"step": 60
},
{
"epoch": 0.18456883509833585,
"grad_norm": 1.5450148332945652,
"learning_rate": 1.8772509113449243e-05,
"loss": 0.6607,
"step": 61
},
{
"epoch": 0.1875945537065053,
"grad_norm": 1.5710515451809142,
"learning_rate": 1.8724960070727974e-05,
"loss": 0.6836,
"step": 62
},
{
"epoch": 0.19062027231467474,
"grad_norm": 1.5953789189285246,
"learning_rate": 1.8676570098195443e-05,
"loss": 0.6546,
"step": 63
},
{
"epoch": 0.19364599092284418,
"grad_norm": 1.6129769850906486,
"learning_rate": 1.862734385977792e-05,
"loss": 0.6635,
"step": 64
},
{
"epoch": 0.19667170953101362,
"grad_norm": 1.7310590253199207,
"learning_rate": 1.8577286100002723e-05,
"loss": 0.6586,
"step": 65
},
{
"epoch": 0.19969742813918306,
"grad_norm": 1.718165656861673,
"learning_rate": 1.8526401643540924e-05,
"loss": 0.6546,
"step": 66
},
{
"epoch": 0.2027231467473525,
"grad_norm": 1.5564799024390932,
"learning_rate": 1.8474695394742345e-05,
"loss": 0.6664,
"step": 67
},
{
"epoch": 0.20574886535552195,
"grad_norm": 1.577385282543463,
"learning_rate": 1.8422172337162865e-05,
"loss": 0.6255,
"step": 68
},
{
"epoch": 0.2087745839636914,
"grad_norm": 1.4966570870141478,
"learning_rate": 1.8368837533084092e-05,
"loss": 0.6337,
"step": 69
},
{
"epoch": 0.2118003025718608,
"grad_norm": 1.6511795969052077,
"learning_rate": 1.8314696123025456e-05,
"loss": 0.6587,
"step": 70
},
{
"epoch": 0.21482602118003025,
"grad_norm": 1.736697285941769,
"learning_rate": 1.825975332524873e-05,
"loss": 0.6766,
"step": 71
},
{
"epoch": 0.2178517397881997,
"grad_norm": 1.5784517572441206,
"learning_rate": 1.8204014435255136e-05,
"loss": 0.6749,
"step": 72
},
{
"epoch": 0.22087745839636913,
"grad_norm": 1.6000531256473662,
"learning_rate": 1.8147484825274895e-05,
"loss": 0.6276,
"step": 73
},
{
"epoch": 0.22390317700453857,
"grad_norm": 1.6259834986336132,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.6688,
"step": 74
},
{
"epoch": 0.22692889561270801,
"grad_norm": 1.7210972702623966,
"learning_rate": 1.803207531480645e-05,
"loss": 0.6277,
"step": 75
},
{
"epoch": 0.22995461422087746,
"grad_norm": 1.576540948126256,
"learning_rate": 1.797320653772707e-05,
"loss": 0.6134,
"step": 76
},
{
"epoch": 0.2329803328290469,
"grad_norm": 1.4954947441393585,
"learning_rate": 1.7913569286406606e-05,
"loss": 0.5974,
"step": 77
},
{
"epoch": 0.23600605143721634,
"grad_norm": 1.560683044038306,
"learning_rate": 1.785316930880745e-05,
"loss": 0.6411,
"step": 78
},
{
"epoch": 0.23903177004538578,
"grad_norm": 1.6008413583717305,
"learning_rate": 1.779201242640517e-05,
"loss": 0.6461,
"step": 79
},
{
"epoch": 0.24205748865355523,
"grad_norm": 1.4168796682558624,
"learning_rate": 1.773010453362737e-05,
"loss": 0.6266,
"step": 80
},
{
"epoch": 0.24508320726172467,
"grad_norm": 1.537422831490025,
"learning_rate": 1.7667451597285617e-05,
"loss": 0.6436,
"step": 81
},
{
"epoch": 0.2481089258698941,
"grad_norm": 1.5410911101514713,
"learning_rate": 1.7604059656000313e-05,
"loss": 0.6333,
"step": 82
},
{
"epoch": 0.25113464447806355,
"grad_norm": 1.6041340691907475,
"learning_rate": 1.7539934819618696e-05,
"loss": 0.6832,
"step": 83
},
{
"epoch": 0.25113464447806355,
"eval_loss": 0.6378725171089172,
"eval_runtime": 509.6475,
"eval_samples_per_second": 8.3,
"eval_steps_per_second": 0.261,
"step": 83
},
{
"epoch": 0.25416036308623297,
"grad_norm": 1.462427594152315,
"learning_rate": 1.747508326862597e-05,
"loss": 0.6439,
"step": 84
},
{
"epoch": 0.25718608169440244,
"grad_norm": 1.5821830710705262,
"learning_rate": 1.7409511253549592e-05,
"loss": 0.6352,
"step": 85
},
{
"epoch": 0.26021180030257185,
"grad_norm": 1.5769534092260407,
"learning_rate": 1.7343225094356857e-05,
"loss": 0.6145,
"step": 86
},
{
"epoch": 0.2632375189107413,
"grad_norm": 1.6608776287535334,
"learning_rate": 1.727623117984575e-05,
"loss": 0.6554,
"step": 87
},
{
"epoch": 0.26626323751891073,
"grad_norm": 1.550801637485576,
"learning_rate": 1.720853596702919e-05,
"loss": 0.6284,
"step": 88
},
{
"epoch": 0.2692889561270802,
"grad_norm": 1.6091051095462698,
"learning_rate": 1.7140145980512684e-05,
"loss": 0.6361,
"step": 89
},
{
"epoch": 0.2723146747352496,
"grad_norm": 1.6122796485509625,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.6238,
"step": 90
},
{
"epoch": 0.2753403933434191,
"grad_norm": 1.5705664236414967,
"learning_rate": 1.7001308118985237e-05,
"loss": 0.6622,
"step": 91
},
{
"epoch": 0.2783661119515885,
"grad_norm": 1.6798802272395463,
"learning_rate": 1.6930873625456362e-05,
"loss": 0.6496,
"step": 92
},
{
"epoch": 0.2813918305597579,
"grad_norm": 1.729456660818085,
"learning_rate": 1.685977111990193e-05,
"loss": 0.6469,
"step": 93
},
{
"epoch": 0.2844175491679274,
"grad_norm": 1.4633667149376168,
"learning_rate": 1.678800745532942e-05,
"loss": 0.6133,
"step": 94
},
{
"epoch": 0.2874432677760968,
"grad_norm": 1.578583871152869,
"learning_rate": 1.6715589548470187e-05,
"loss": 0.6344,
"step": 95
},
{
"epoch": 0.29046898638426627,
"grad_norm": 1.5117823930384213,
"learning_rate": 1.664252437911282e-05,
"loss": 0.6268,
"step": 96
},
{
"epoch": 0.2934947049924357,
"grad_norm": 1.55519423900645,
"learning_rate": 1.6568818989430416e-05,
"loss": 0.6618,
"step": 97
},
{
"epoch": 0.29652042360060515,
"grad_norm": 1.568965405473659,
"learning_rate": 1.6494480483301836e-05,
"loss": 0.6282,
"step": 98
},
{
"epoch": 0.29954614220877457,
"grad_norm": 1.5005659642722655,
"learning_rate": 1.641951602562703e-05,
"loss": 0.6242,
"step": 99
},
{
"epoch": 0.30257186081694404,
"grad_norm": 1.4905844492827542,
"learning_rate": 1.6343932841636455e-05,
"loss": 0.6311,
"step": 100
},
{
"epoch": 0.30559757942511345,
"grad_norm": 1.465089233145022,
"learning_rate": 1.6267738216194698e-05,
"loss": 0.6175,
"step": 101
},
{
"epoch": 0.3086232980332829,
"grad_norm": 1.5926794979806516,
"learning_rate": 1.6190939493098344e-05,
"loss": 0.6195,
"step": 102
},
{
"epoch": 0.31164901664145234,
"grad_norm": 1.6211595429202696,
"learning_rate": 1.6113544074368166e-05,
"loss": 0.6428,
"step": 103
},
{
"epoch": 0.3146747352496218,
"grad_norm": 1.4842565634778135,
"learning_rate": 1.6035559419535714e-05,
"loss": 0.6311,
"step": 104
},
{
"epoch": 0.3177004538577912,
"grad_norm": 1.588750243344381,
"learning_rate": 1.5956993044924334e-05,
"loss": 0.6459,
"step": 105
},
{
"epoch": 0.3207261724659607,
"grad_norm": 1.5908076922384171,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.5959,
"step": 106
},
{
"epoch": 0.3237518910741301,
"grad_norm": 1.528918109525865,
"learning_rate": 1.579814548126514e-05,
"loss": 0.6082,
"step": 107
},
{
"epoch": 0.3267776096822995,
"grad_norm": 1.507689796200556,
"learning_rate": 1.5717879602276123e-05,
"loss": 0.6195,
"step": 108
},
{
"epoch": 0.329803328290469,
"grad_norm": 1.5564079492146874,
"learning_rate": 1.5637062622150168e-05,
"loss": 0.6437,
"step": 109
},
{
"epoch": 0.3328290468986384,
"grad_norm": 1.67840601823286,
"learning_rate": 1.5555702330196024e-05,
"loss": 0.6309,
"step": 110
},
{
"epoch": 0.3358547655068079,
"grad_norm": 1.704633390331848,
"learning_rate": 1.547380656808797e-05,
"loss": 0.6229,
"step": 111
},
{
"epoch": 0.3388804841149773,
"grad_norm": 1.4784787495625082,
"learning_rate": 1.5391383229110005e-05,
"loss": 0.6253,
"step": 112
},
{
"epoch": 0.34190620272314676,
"grad_norm": 1.5130041477207583,
"learning_rate": 1.5308440257395095e-05,
"loss": 0.6107,
"step": 113
},
{
"epoch": 0.34493192133131617,
"grad_norm": 1.5130610200511525,
"learning_rate": 1.5224985647159489e-05,
"loss": 0.6287,
"step": 114
},
{
"epoch": 0.34795763993948564,
"grad_norm": 1.743450524980943,
"learning_rate": 1.5141027441932217e-05,
"loss": 0.6479,
"step": 115
},
{
"epoch": 0.35098335854765506,
"grad_norm": 1.5352833649417843,
"learning_rate": 1.5056573733779848e-05,
"loss": 0.6014,
"step": 116
},
{
"epoch": 0.3540090771558245,
"grad_norm": 1.6654822043986846,
"learning_rate": 1.4971632662526545e-05,
"loss": 0.6264,
"step": 117
},
{
"epoch": 0.35703479576399394,
"grad_norm": 1.5869536351236282,
"learning_rate": 1.4886212414969551e-05,
"loss": 0.6407,
"step": 118
},
{
"epoch": 0.3600605143721634,
"grad_norm": 1.5820866980420585,
"learning_rate": 1.4800321224090114e-05,
"loss": 0.6277,
"step": 119
},
{
"epoch": 0.3630862329803328,
"grad_norm": 1.4831553795674497,
"learning_rate": 1.4713967368259981e-05,
"loss": 0.5997,
"step": 120
},
{
"epoch": 0.3661119515885023,
"grad_norm": 1.5852987595773695,
"learning_rate": 1.4627159170443504e-05,
"loss": 0.6348,
"step": 121
},
{
"epoch": 0.3691376701966717,
"grad_norm": 1.5131951272437487,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.6316,
"step": 122
},
{
"epoch": 0.3721633888048411,
"grad_norm": 1.446787025160117,
"learning_rate": 1.4452213258854684e-05,
"loss": 0.6289,
"step": 123
},
{
"epoch": 0.3751891074130106,
"grad_norm": 1.4839285346310753,
"learning_rate": 1.436409240673342e-05,
"loss": 0.6095,
"step": 124
},
{
"epoch": 0.37821482602118,
"grad_norm": 1.720340785200561,
"learning_rate": 1.4275550934302822e-05,
"loss": 0.6141,
"step": 125
},
{
"epoch": 0.3812405446293495,
"grad_norm": 1.5007595992236809,
"learning_rate": 1.4186597375374283e-05,
"loss": 0.5946,
"step": 126
},
{
"epoch": 0.3842662632375189,
"grad_norm": 1.601900166031101,
"learning_rate": 1.4097240303476955e-05,
"loss": 0.6099,
"step": 127
},
{
"epoch": 0.38729198184568836,
"grad_norm": 1.4297276715184504,
"learning_rate": 1.4007488331031409e-05,
"loss": 0.601,
"step": 128
},
{
"epoch": 0.3903177004538578,
"grad_norm": 1.5321155661542338,
"learning_rate": 1.391735010851956e-05,
"loss": 0.6529,
"step": 129
},
{
"epoch": 0.39334341906202724,
"grad_norm": 1.5247035157241295,
"learning_rate": 1.3826834323650899e-05,
"loss": 0.6292,
"step": 130
},
{
"epoch": 0.39636913767019666,
"grad_norm": 1.4582854009232522,
"learning_rate": 1.3735949700525164e-05,
"loss": 0.6093,
"step": 131
},
{
"epoch": 0.39939485627836613,
"grad_norm": 1.4893132946281402,
"learning_rate": 1.3644704998791501e-05,
"loss": 0.6642,
"step": 132
},
{
"epoch": 0.40242057488653554,
"grad_norm": 1.4483284402325534,
"learning_rate": 1.3553109012804162e-05,
"loss": 0.633,
"step": 133
},
{
"epoch": 0.405446293494705,
"grad_norm": 1.4598771596801616,
"learning_rate": 1.346117057077493e-05,
"loss": 0.6283,
"step": 134
},
{
"epoch": 0.4084720121028744,
"grad_norm": 1.553386372239155,
"learning_rate": 1.3368898533922202e-05,
"loss": 0.6278,
"step": 135
},
{
"epoch": 0.4114977307110439,
"grad_norm": 1.4953611099827313,
"learning_rate": 1.3276301795616937e-05,
"loss": 0.6314,
"step": 136
},
{
"epoch": 0.4145234493192133,
"grad_norm": 1.4516817565773994,
"learning_rate": 1.3183389280525497e-05,
"loss": 0.6123,
"step": 137
},
{
"epoch": 0.4175491679273828,
"grad_norm": 1.463835831209024,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.6243,
"step": 138
},
{
"epoch": 0.4205748865355522,
"grad_norm": 1.4952603681448646,
"learning_rate": 1.2996652769962567e-05,
"loss": 0.6229,
"step": 139
},
{
"epoch": 0.4236006051437216,
"grad_norm": 1.4911098485246495,
"learning_rate": 1.2902846772544625e-05,
"loss": 0.6237,
"step": 140
},
{
"epoch": 0.4266263237518911,
"grad_norm": 1.4588416492934029,
"learning_rate": 1.2808760992712923e-05,
"loss": 0.6128,
"step": 141
},
{
"epoch": 0.4296520423600605,
"grad_norm": 1.6119208662638396,
"learning_rate": 1.2714404498650743e-05,
"loss": 0.6352,
"step": 142
},
{
"epoch": 0.43267776096822996,
"grad_norm": 1.4138005025925455,
"learning_rate": 1.2619786384633374e-05,
"loss": 0.5785,
"step": 143
},
{
"epoch": 0.4357034795763994,
"grad_norm": 1.5307856089545673,
"learning_rate": 1.252491577015158e-05,
"loss": 0.6228,
"step": 144
},
{
"epoch": 0.43872919818456885,
"grad_norm": 1.5277296504806848,
"learning_rate": 1.242980179903264e-05,
"loss": 0.6306,
"step": 145
},
{
"epoch": 0.44175491679273826,
"grad_norm": 1.3981864936760282,
"learning_rate": 1.2334453638559057e-05,
"loss": 0.5925,
"step": 146
},
{
"epoch": 0.44478063540090773,
"grad_norm": 1.432012218131282,
"learning_rate": 1.2238880478584987e-05,
"loss": 0.6253,
"step": 147
},
{
"epoch": 0.44780635400907715,
"grad_norm": 1.50725822196971,
"learning_rate": 1.2143091530650508e-05,
"loss": 0.6263,
"step": 148
},
{
"epoch": 0.4508320726172466,
"grad_norm": 1.4584993288987338,
"learning_rate": 1.2047096027093798e-05,
"loss": 0.6176,
"step": 149
},
{
"epoch": 0.45385779122541603,
"grad_norm": 1.431477317558029,
"learning_rate": 1.1950903220161286e-05,
"loss": 0.6439,
"step": 150
},
{
"epoch": 0.4568835098335855,
"grad_norm": 1.4640759182723595,
"learning_rate": 1.185452238111591e-05,
"loss": 0.6006,
"step": 151
},
{
"epoch": 0.4599092284417549,
"grad_norm": 1.4339691019212786,
"learning_rate": 1.1757962799343548e-05,
"loss": 0.6,
"step": 152
},
{
"epoch": 0.4629349470499244,
"grad_norm": 1.430696114486699,
"learning_rate": 1.1661233781457655e-05,
"loss": 0.5919,
"step": 153
},
{
"epoch": 0.4659606656580938,
"grad_norm": 1.4354938705216158,
"learning_rate": 1.156434465040231e-05,
"loss": 0.6033,
"step": 154
},
{
"epoch": 0.4689863842662632,
"grad_norm": 1.4537531387316016,
"learning_rate": 1.1467304744553618e-05,
"loss": 0.6111,
"step": 155
},
{
"epoch": 0.4720121028744327,
"grad_norm": 1.4773017265373494,
"learning_rate": 1.1370123416819683e-05,
"loss": 0.5986,
"step": 156
},
{
"epoch": 0.4750378214826021,
"grad_norm": 1.467637701147254,
"learning_rate": 1.1272810033739134e-05,
"loss": 0.6156,
"step": 157
},
{
"epoch": 0.47806354009077157,
"grad_norm": 1.6000029315681117,
"learning_rate": 1.1175373974578378e-05,
"loss": 0.5987,
"step": 158
},
{
"epoch": 0.481089258698941,
"grad_norm": 1.3929295237973425,
"learning_rate": 1.1077824630427593e-05,
"loss": 0.5857,
"step": 159
},
{
"epoch": 0.48411497730711045,
"grad_norm": 1.38453633556067,
"learning_rate": 1.098017140329561e-05,
"loss": 0.6186,
"step": 160
},
{
"epoch": 0.48714069591527986,
"grad_norm": 1.4527687815126924,
"learning_rate": 1.0882423705203698e-05,
"loss": 0.6089,
"step": 161
},
{
"epoch": 0.49016641452344933,
"grad_norm": 1.3595389880404907,
"learning_rate": 1.0784590957278452e-05,
"loss": 0.5803,
"step": 162
},
{
"epoch": 0.49319213313161875,
"grad_norm": 1.4984875304853849,
"learning_rate": 1.0686682588843737e-05,
"loss": 0.6017,
"step": 163
},
{
"epoch": 0.4962178517397882,
"grad_norm": 1.4410230293678532,
"learning_rate": 1.058870803651189e-05,
"loss": 0.6134,
"step": 164
},
{
"epoch": 0.49924357034795763,
"grad_norm": 1.54768953054996,
"learning_rate": 1.0490676743274181e-05,
"loss": 0.6228,
"step": 165
},
{
"epoch": 0.5022692889561271,
"grad_norm": 1.419881622408715,
"learning_rate": 1.0392598157590687e-05,
"loss": 0.5763,
"step": 166
},
{
"epoch": 0.5022692889561271,
"eval_loss": 0.6071931719779968,
"eval_runtime": 498.7669,
"eval_samples_per_second": 8.481,
"eval_steps_per_second": 0.267,
"step": 166
},
{
"epoch": 0.5052950075642966,
"grad_norm": 1.4475547194909217,
"learning_rate": 1.0294481732479635e-05,
"loss": 0.6132,
"step": 167
},
{
"epoch": 0.5083207261724659,
"grad_norm": 1.3657202818657257,
"learning_rate": 1.0196336924606282e-05,
"loss": 0.6088,
"step": 168
},
{
"epoch": 0.5113464447806354,
"grad_norm": 1.439530367007596,
"learning_rate": 1.0098173193371498e-05,
"loss": 0.5775,
"step": 169
},
{
"epoch": 0.5143721633888049,
"grad_norm": 1.4823283030412917,
"learning_rate": 1e-05,
"loss": 0.6035,
"step": 170
},
{
"epoch": 0.5173978819969742,
"grad_norm": 1.424254228900751,
"learning_rate": 9.901826806628505e-06,
"loss": 0.6236,
"step": 171
},
{
"epoch": 0.5204236006051437,
"grad_norm": 1.527483576765537,
"learning_rate": 9.80366307539372e-06,
"loss": 0.6138,
"step": 172
},
{
"epoch": 0.5234493192133132,
"grad_norm": 1.5214269697406193,
"learning_rate": 9.705518267520369e-06,
"loss": 0.6227,
"step": 173
},
{
"epoch": 0.5264750378214826,
"grad_norm": 1.3595619018433223,
"learning_rate": 9.607401842409318e-06,
"loss": 0.5683,
"step": 174
},
{
"epoch": 0.529500756429652,
"grad_norm": 1.4534894324363723,
"learning_rate": 9.50932325672582e-06,
"loss": 0.6108,
"step": 175
},
{
"epoch": 0.5325264750378215,
"grad_norm": 1.386880214106318,
"learning_rate": 9.41129196348811e-06,
"loss": 0.5718,
"step": 176
},
{
"epoch": 0.5355521936459909,
"grad_norm": 1.4675681499043807,
"learning_rate": 9.313317411156265e-06,
"loss": 0.5995,
"step": 177
},
{
"epoch": 0.5385779122541604,
"grad_norm": 1.448341612122338,
"learning_rate": 9.215409042721553e-06,
"loss": 0.5864,
"step": 178
},
{
"epoch": 0.5416036308623298,
"grad_norm": 1.461953036096099,
"learning_rate": 9.117576294796307e-06,
"loss": 0.6023,
"step": 179
},
{
"epoch": 0.5446293494704992,
"grad_norm": 1.425828231836892,
"learning_rate": 9.019828596704394e-06,
"loss": 0.5714,
"step": 180
},
{
"epoch": 0.5476550680786687,
"grad_norm": 1.431412797305995,
"learning_rate": 8.922175369572407e-06,
"loss": 0.6225,
"step": 181
},
{
"epoch": 0.5506807866868382,
"grad_norm": 1.3578648212594675,
"learning_rate": 8.824626025421625e-06,
"loss": 0.5888,
"step": 182
},
{
"epoch": 0.5537065052950075,
"grad_norm": 1.5850872303144425,
"learning_rate": 8.72718996626087e-06,
"loss": 0.6124,
"step": 183
},
{
"epoch": 0.556732223903177,
"grad_norm": 1.4522197835564097,
"learning_rate": 8.629876583180322e-06,
"loss": 0.6128,
"step": 184
},
{
"epoch": 0.5597579425113465,
"grad_norm": 1.355419508837911,
"learning_rate": 8.532695255446384e-06,
"loss": 0.5735,
"step": 185
},
{
"epoch": 0.5627836611195158,
"grad_norm": 1.4251089846754423,
"learning_rate": 8.43565534959769e-06,
"loss": 0.5826,
"step": 186
},
{
"epoch": 0.5658093797276853,
"grad_norm": 1.4688696578023765,
"learning_rate": 8.338766218542348e-06,
"loss": 0.5874,
"step": 187
},
{
"epoch": 0.5688350983358548,
"grad_norm": 1.363483373884317,
"learning_rate": 8.242037200656455e-06,
"loss": 0.5773,
"step": 188
},
{
"epoch": 0.5718608169440242,
"grad_norm": 1.4247250643659577,
"learning_rate": 8.145477618884092e-06,
"loss": 0.5816,
"step": 189
},
{
"epoch": 0.5748865355521936,
"grad_norm": 1.3608178764328125,
"learning_rate": 8.04909677983872e-06,
"loss": 0.5969,
"step": 190
},
{
"epoch": 0.5779122541603631,
"grad_norm": 1.4996759497823389,
"learning_rate": 7.952903972906205e-06,
"loss": 0.6166,
"step": 191
},
{
"epoch": 0.5809379727685325,
"grad_norm": 1.4176668196688376,
"learning_rate": 7.856908469349495e-06,
"loss": 0.5848,
"step": 192
},
{
"epoch": 0.583963691376702,
"grad_norm": 1.408792838362906,
"learning_rate": 7.761119521415017e-06,
"loss": 0.6142,
"step": 193
},
{
"epoch": 0.5869894099848714,
"grad_norm": 1.414367714546386,
"learning_rate": 7.66554636144095e-06,
"loss": 0.6111,
"step": 194
},
{
"epoch": 0.5900151285930408,
"grad_norm": 1.3633466892491897,
"learning_rate": 7.570198200967363e-06,
"loss": 0.6148,
"step": 195
},
{
"epoch": 0.5930408472012103,
"grad_norm": 1.4674036986770749,
"learning_rate": 7.4750842298484205e-06,
"loss": 0.594,
"step": 196
},
{
"epoch": 0.5960665658093798,
"grad_norm": 1.3963931215398027,
"learning_rate": 7.380213615366627e-06,
"loss": 0.6114,
"step": 197
},
{
"epoch": 0.5990922844175491,
"grad_norm": 1.4213688920051264,
"learning_rate": 7.285595501349259e-06,
"loss": 0.6242,
"step": 198
},
{
"epoch": 0.6021180030257186,
"grad_norm": 1.438211556419051,
"learning_rate": 7.191239007287082e-06,
"loss": 0.6156,
"step": 199
},
{
"epoch": 0.6051437216338881,
"grad_norm": 1.3760296910712133,
"learning_rate": 7.097153227455379e-06,
"loss": 0.587,
"step": 200
},
{
"epoch": 0.6081694402420574,
"grad_norm": 1.4196182496053105,
"learning_rate": 7.003347230037434e-06,
"loss": 0.5761,
"step": 201
},
{
"epoch": 0.6111951588502269,
"grad_norm": 1.392660432383428,
"learning_rate": 6.909830056250527e-06,
"loss": 0.5927,
"step": 202
},
{
"epoch": 0.6142208774583964,
"grad_norm": 1.3498375378930192,
"learning_rate": 6.816610719474503e-06,
"loss": 0.5765,
"step": 203
},
{
"epoch": 0.6172465960665658,
"grad_norm": 1.3735452249284985,
"learning_rate": 6.723698204383067e-06,
"loss": 0.583,
"step": 204
},
{
"epoch": 0.6202723146747352,
"grad_norm": 1.348068908404557,
"learning_rate": 6.631101466077801e-06,
"loss": 0.5688,
"step": 205
},
{
"epoch": 0.6232980332829047,
"grad_norm": 1.3808202298057348,
"learning_rate": 6.538829429225068e-06,
"loss": 0.5688,
"step": 206
},
{
"epoch": 0.6263237518910741,
"grad_norm": 1.3923884599746261,
"learning_rate": 6.446890987195842e-06,
"loss": 0.5885,
"step": 207
},
{
"epoch": 0.6293494704992436,
"grad_norm": 1.374737599408344,
"learning_rate": 6.355295001208504e-06,
"loss": 0.5898,
"step": 208
},
{
"epoch": 0.632375189107413,
"grad_norm": 1.3989565705834308,
"learning_rate": 6.2640502994748375e-06,
"loss": 0.5576,
"step": 209
},
{
"epoch": 0.6354009077155824,
"grad_norm": 1.4448280944807503,
"learning_rate": 6.173165676349103e-06,
"loss": 0.6071,
"step": 210
},
{
"epoch": 0.6384266263237519,
"grad_norm": 1.3927626833954307,
"learning_rate": 6.082649891480441e-06,
"loss": 0.595,
"step": 211
},
{
"epoch": 0.6414523449319214,
"grad_norm": 1.4916401062280373,
"learning_rate": 5.9925116689685925e-06,
"loss": 0.6004,
"step": 212
},
{
"epoch": 0.6444780635400907,
"grad_norm": 1.3743489632421992,
"learning_rate": 5.902759696523046e-06,
"loss": 0.586,
"step": 213
},
{
"epoch": 0.6475037821482602,
"grad_norm": 1.4595349210684398,
"learning_rate": 5.813402624625722e-06,
"loss": 0.5807,
"step": 214
},
{
"epoch": 0.6505295007564297,
"grad_norm": 1.4546859536718255,
"learning_rate": 5.724449065697182e-06,
"loss": 0.6084,
"step": 215
},
{
"epoch": 0.653555219364599,
"grad_norm": 1.5016556623739812,
"learning_rate": 5.635907593266578e-06,
"loss": 0.6237,
"step": 216
},
{
"epoch": 0.6565809379727685,
"grad_norm": 1.4154460915318592,
"learning_rate": 5.54778674114532e-06,
"loss": 0.5839,
"step": 217
},
{
"epoch": 0.659606656580938,
"grad_norm": 1.4088041488582226,
"learning_rate": 5.460095002604533e-06,
"loss": 0.5825,
"step": 218
},
{
"epoch": 0.6626323751891074,
"grad_norm": 1.3599062363064875,
"learning_rate": 5.3728408295565e-06,
"loss": 0.5714,
"step": 219
},
{
"epoch": 0.6656580937972768,
"grad_norm": 1.3293624699751176,
"learning_rate": 5.286032631740023e-06,
"loss": 0.5861,
"step": 220
},
{
"epoch": 0.6686838124054463,
"grad_norm": 1.4021379502923315,
"learning_rate": 5.199678775909889e-06,
"loss": 0.6036,
"step": 221
},
{
"epoch": 0.6717095310136157,
"grad_norm": 1.3668189281210377,
"learning_rate": 5.1137875850304545e-06,
"loss": 0.5787,
"step": 222
},
{
"epoch": 0.6747352496217852,
"grad_norm": 1.3566258300920009,
"learning_rate": 5.0283673374734546e-06,
"loss": 0.5849,
"step": 223
},
{
"epoch": 0.6777609682299546,
"grad_norm": 1.3884171461497674,
"learning_rate": 4.943426266220156e-06,
"loss": 0.5746,
"step": 224
},
{
"epoch": 0.680786686838124,
"grad_norm": 1.4208065860811094,
"learning_rate": 4.858972558067784e-06,
"loss": 0.5969,
"step": 225
},
{
"epoch": 0.6838124054462935,
"grad_norm": 1.479579795754236,
"learning_rate": 4.775014352840512e-06,
"loss": 0.5928,
"step": 226
},
{
"epoch": 0.686838124054463,
"grad_norm": 1.4412917161056067,
"learning_rate": 4.691559742604906e-06,
"loss": 0.6048,
"step": 227
},
{
"epoch": 0.6898638426626323,
"grad_norm": 1.3640710120709136,
"learning_rate": 4.608616770889998e-06,
"loss": 0.5809,
"step": 228
},
{
"epoch": 0.6928895612708018,
"grad_norm": 1.382969377214025,
"learning_rate": 4.526193431912038e-06,
"loss": 0.5775,
"step": 229
},
{
"epoch": 0.6959152798789713,
"grad_norm": 1.4572381884526804,
"learning_rate": 4.444297669803981e-06,
"loss": 0.5955,
"step": 230
},
{
"epoch": 0.6989409984871406,
"grad_norm": 1.3965942128271758,
"learning_rate": 4.362937377849832e-06,
"loss": 0.5717,
"step": 231
},
{
"epoch": 0.7019667170953101,
"grad_norm": 1.363037831585429,
"learning_rate": 4.282120397723879e-06,
"loss": 0.5905,
"step": 232
},
{
"epoch": 0.7049924357034796,
"grad_norm": 1.4211039008481519,
"learning_rate": 4.2018545187348645e-06,
"loss": 0.6126,
"step": 233
},
{
"epoch": 0.708018154311649,
"grad_norm": 1.4516183713570143,
"learning_rate": 4.12214747707527e-06,
"loss": 0.601,
"step": 234
},
{
"epoch": 0.7110438729198184,
"grad_norm": 1.4070386843255802,
"learning_rate": 4.043006955075667e-06,
"loss": 0.5817,
"step": 235
},
{
"epoch": 0.7140695915279879,
"grad_norm": 1.4058491023144444,
"learning_rate": 3.964440580464286e-06,
"loss": 0.5657,
"step": 236
},
{
"epoch": 0.7170953101361573,
"grad_norm": 1.3767493634042243,
"learning_rate": 3.8864559256318375e-06,
"loss": 0.5844,
"step": 237
},
{
"epoch": 0.7201210287443268,
"grad_norm": 1.4340353098948653,
"learning_rate": 3.8090605069016596e-06,
"loss": 0.6049,
"step": 238
},
{
"epoch": 0.7231467473524962,
"grad_norm": 1.3897630441399809,
"learning_rate": 3.7322617838053066e-06,
"loss": 0.5721,
"step": 239
},
{
"epoch": 0.7261724659606656,
"grad_norm": 1.458875523946377,
"learning_rate": 3.6560671583635467e-06,
"loss": 0.5804,
"step": 240
},
{
"epoch": 0.7291981845688351,
"grad_norm": 1.416661390657441,
"learning_rate": 3.58048397437297e-06,
"loss": 0.5759,
"step": 241
},
{
"epoch": 0.7322239031770046,
"grad_norm": 1.4765471008195479,
"learning_rate": 3.505519516698165e-06,
"loss": 0.5913,
"step": 242
},
{
"epoch": 0.735249621785174,
"grad_norm": 1.3479820090813943,
"learning_rate": 3.4311810105695875e-06,
"loss": 0.5726,
"step": 243
},
{
"epoch": 0.7382753403933434,
"grad_norm": 1.3950132266291524,
"learning_rate": 3.3574756208871862e-06,
"loss": 0.5817,
"step": 244
},
{
"epoch": 0.7413010590015129,
"grad_norm": 1.5164531444092535,
"learning_rate": 3.284410451529816e-06,
"loss": 0.6176,
"step": 245
},
{
"epoch": 0.7443267776096822,
"grad_norm": 1.3961701000222428,
"learning_rate": 3.2119925446705824e-06,
"loss": 0.5523,
"step": 246
},
{
"epoch": 0.7473524962178517,
"grad_norm": 1.4261029837453818,
"learning_rate": 3.140228880098074e-06,
"loss": 0.5904,
"step": 247
},
{
"epoch": 0.7503782148260212,
"grad_norm": 1.3718208369891849,
"learning_rate": 3.069126374543643e-06,
"loss": 0.5746,
"step": 248
},
{
"epoch": 0.7534039334341907,
"grad_norm": 1.3718247539733979,
"learning_rate": 2.998691881014765e-06,
"loss": 0.5898,
"step": 249
},
{
"epoch": 0.7534039334341907,
"eval_loss": 0.5826606154441833,
"eval_runtime": 503.0413,
"eval_samples_per_second": 8.409,
"eval_steps_per_second": 0.264,
"step": 249
},
{
"epoch": 0.75642965204236,
"grad_norm": 1.4185862364935742,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.6095,
"step": 250
},
{
"epoch": 0.7594553706505295,
"grad_norm": 1.4581760073985484,
"learning_rate": 2.859854019487318e-06,
"loss": 0.594,
"step": 251
},
{
"epoch": 0.762481089258699,
"grad_norm": 1.3845699796654831,
"learning_rate": 2.791464032970812e-06,
"loss": 0.6002,
"step": 252
},
{
"epoch": 0.7655068078668684,
"grad_norm": 1.3251469772846132,
"learning_rate": 2.723768820154251e-06,
"loss": 0.5454,
"step": 253
},
{
"epoch": 0.7685325264750378,
"grad_norm": 1.334450098425696,
"learning_rate": 2.656774905643147e-06,
"loss": 0.5552,
"step": 254
},
{
"epoch": 0.7715582450832073,
"grad_norm": 1.3718433239430523,
"learning_rate": 2.5904887464504115e-06,
"loss": 0.5894,
"step": 255
},
{
"epoch": 0.7745839636913767,
"grad_norm": 1.4263164476517571,
"learning_rate": 2.5249167313740307e-06,
"loss": 0.5675,
"step": 256
},
{
"epoch": 0.7776096822995462,
"grad_norm": 1.4139381321578703,
"learning_rate": 2.4600651803813057e-06,
"loss": 0.5869,
"step": 257
},
{
"epoch": 0.7806354009077155,
"grad_norm": 1.40674475404999,
"learning_rate": 2.395940343999691e-06,
"loss": 0.5755,
"step": 258
},
{
"epoch": 0.783661119515885,
"grad_norm": 1.3109070005496368,
"learning_rate": 2.332548402714385e-06,
"loss": 0.5474,
"step": 259
},
{
"epoch": 0.7866868381240545,
"grad_norm": 1.3297333386147439,
"learning_rate": 2.26989546637263e-06,
"loss": 0.544,
"step": 260
},
{
"epoch": 0.789712556732224,
"grad_norm": 1.3851600194590172,
"learning_rate": 2.207987573594833e-06,
"loss": 0.5953,
"step": 261
},
{
"epoch": 0.7927382753403933,
"grad_norm": 1.3499760384279826,
"learning_rate": 2.146830691192553e-06,
"loss": 0.5843,
"step": 262
},
{
"epoch": 0.7957639939485628,
"grad_norm": 1.380248534318227,
"learning_rate": 2.086430713593397e-06,
"loss": 0.5764,
"step": 263
},
{
"epoch": 0.7987897125567323,
"grad_norm": 1.4147089150561032,
"learning_rate": 2.02679346227293e-06,
"loss": 0.6039,
"step": 264
},
{
"epoch": 0.8018154311649016,
"grad_norm": 1.3831525637034685,
"learning_rate": 1.967924685193552e-06,
"loss": 0.5897,
"step": 265
},
{
"epoch": 0.8048411497730711,
"grad_norm": 1.3608060023356416,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.5777,
"step": 266
},
{
"epoch": 0.8078668683812406,
"grad_norm": 1.3703430718414862,
"learning_rate": 1.8525151747251058e-06,
"loss": 0.5719,
"step": 267
},
{
"epoch": 0.81089258698941,
"grad_norm": 1.4214961547733749,
"learning_rate": 1.7959855647448642e-06,
"loss": 0.5656,
"step": 268
},
{
"epoch": 0.8139183055975794,
"grad_norm": 1.3995513064696632,
"learning_rate": 1.7402466747512704e-06,
"loss": 0.5709,
"step": 269
},
{
"epoch": 0.8169440242057489,
"grad_norm": 1.392923658058648,
"learning_rate": 1.6853038769745466e-06,
"loss": 0.5879,
"step": 270
},
{
"epoch": 0.8199697428139183,
"grad_norm": 1.357249339017749,
"learning_rate": 1.6311624669159064e-06,
"loss": 0.5846,
"step": 271
},
{
"epoch": 0.8229954614220878,
"grad_norm": 1.3759453897612757,
"learning_rate": 1.577827662837136e-06,
"loss": 0.5702,
"step": 272
},
{
"epoch": 0.8260211800302572,
"grad_norm": 1.367351688015586,
"learning_rate": 1.5253046052576559e-06,
"loss": 0.5736,
"step": 273
},
{
"epoch": 0.8290468986384266,
"grad_norm": 1.4233508625485896,
"learning_rate": 1.4735983564590784e-06,
"loss": 0.5921,
"step": 274
},
{
"epoch": 0.8320726172465961,
"grad_norm": 1.3905578571681132,
"learning_rate": 1.4227138999972801e-06,
"loss": 0.595,
"step": 275
},
{
"epoch": 0.8350983358547656,
"grad_norm": 1.3922991422194104,
"learning_rate": 1.3726561402220818e-06,
"loss": 0.5404,
"step": 276
},
{
"epoch": 0.8381240544629349,
"grad_norm": 1.3853933022135794,
"learning_rate": 1.3234299018045615e-06,
"loss": 0.5698,
"step": 277
},
{
"epoch": 0.8411497730711044,
"grad_norm": 1.4031955507773741,
"learning_rate": 1.2750399292720284e-06,
"loss": 0.5405,
"step": 278
},
{
"epoch": 0.8441754916792739,
"grad_norm": 1.3652661130652248,
"learning_rate": 1.2274908865507595e-06,
"loss": 0.5497,
"step": 279
},
{
"epoch": 0.8472012102874432,
"grad_norm": 1.315497749826689,
"learning_rate": 1.1807873565164507e-06,
"loss": 0.5274,
"step": 280
},
{
"epoch": 0.8502269288956127,
"grad_norm": 1.448804711972974,
"learning_rate": 1.1349338405525368e-06,
"loss": 0.5944,
"step": 281
},
{
"epoch": 0.8532526475037822,
"grad_norm": 1.4761576990708463,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.5939,
"step": 282
},
{
"epoch": 0.8562783661119516,
"grad_norm": 1.3754542350921957,
"learning_rate": 1.045794446313031e-06,
"loss": 0.5809,
"step": 283
},
{
"epoch": 0.859304084720121,
"grad_norm": 1.344098072725609,
"learning_rate": 1.0025171594777872e-06,
"loss": 0.5609,
"step": 284
},
{
"epoch": 0.8623298033282905,
"grad_norm": 1.4193572796450025,
"learning_rate": 9.601070687655667e-07,
"loss": 0.5933,
"step": 285
},
{
"epoch": 0.8653555219364599,
"grad_norm": 1.3750840657263779,
"learning_rate": 9.185682617491865e-07,
"loss": 0.576,
"step": 286
},
{
"epoch": 0.8683812405446294,
"grad_norm": 1.3433950119608755,
"learning_rate": 8.779047420253239e-07,
"loss": 0.5604,
"step": 287
},
{
"epoch": 0.8714069591527988,
"grad_norm": 1.3115795660045166,
"learning_rate": 8.381204288286415e-07,
"loss": 0.5657,
"step": 288
},
{
"epoch": 0.8744326777609682,
"grad_norm": 1.338233097499156,
"learning_rate": 7.992191566540519e-07,
"loss": 0.5454,
"step": 289
},
{
"epoch": 0.8774583963691377,
"grad_norm": 1.3610318453780819,
"learning_rate": 7.612046748871327e-07,
"loss": 0.564,
"step": 290
},
{
"epoch": 0.8804841149773072,
"grad_norm": 1.3964020958472012,
"learning_rate": 7.240806474427598e-07,
"loss": 0.5923,
"step": 291
},
{
"epoch": 0.8835098335854765,
"grad_norm": 1.376395373330586,
"learning_rate": 6.878506524119644e-07,
"loss": 0.6,
"step": 292
},
{
"epoch": 0.886535552193646,
"grad_norm": 1.3683565764155636,
"learning_rate": 6.525181817170756e-07,
"loss": 0.5606,
"step": 293
},
{
"epoch": 0.8895612708018155,
"grad_norm": 1.455136360910936,
"learning_rate": 6.180866407751595e-07,
"loss": 0.5692,
"step": 294
},
{
"epoch": 0.8925869894099848,
"grad_norm": 1.4316392685854125,
"learning_rate": 5.845593481697931e-07,
"loss": 0.5918,
"step": 295
},
{
"epoch": 0.8956127080181543,
"grad_norm": 1.4466663140505351,
"learning_rate": 5.519395353312195e-07,
"loss": 0.5658,
"step": 296
},
{
"epoch": 0.8986384266263238,
"grad_norm": 1.2955768617089864,
"learning_rate": 5.20230346224897e-07,
"loss": 0.5335,
"step": 297
},
{
"epoch": 0.9016641452344932,
"grad_norm": 1.3392764774102477,
"learning_rate": 4.894348370484648e-07,
"loss": 0.5512,
"step": 298
},
{
"epoch": 0.9046898638426626,
"grad_norm": 1.3740536674819088,
"learning_rate": 4.5955597593719593e-07,
"loss": 0.5832,
"step": 299
},
{
"epoch": 0.9077155824508321,
"grad_norm": 1.348840581020888,
"learning_rate": 4.305966426779118e-07,
"loss": 0.5618,
"step": 300
},
{
"epoch": 0.9107413010590015,
"grad_norm": 1.374168464539394,
"learning_rate": 4.025596284314259e-07,
"loss": 0.5964,
"step": 301
},
{
"epoch": 0.913767019667171,
"grad_norm": 1.3566460139365393,
"learning_rate": 3.7544763546352834e-07,
"loss": 0.5796,
"step": 302
},
{
"epoch": 0.9167927382753404,
"grad_norm": 1.357469688403083,
"learning_rate": 3.492632768845261e-07,
"loss": 0.5624,
"step": 303
},
{
"epoch": 0.9198184568835098,
"grad_norm": 1.3621590802508963,
"learning_rate": 3.2400907639740243e-07,
"loss": 0.575,
"step": 304
},
{
"epoch": 0.9228441754916793,
"grad_norm": 1.395351187361393,
"learning_rate": 2.996874680545603e-07,
"loss": 0.5803,
"step": 305
},
{
"epoch": 0.9258698940998488,
"grad_norm": 1.366363269690252,
"learning_rate": 2.7630079602323447e-07,
"loss": 0.565,
"step": 306
},
{
"epoch": 0.9288956127080181,
"grad_norm": 1.3982055909281412,
"learning_rate": 2.5385131435955e-07,
"loss": 0.5712,
"step": 307
},
{
"epoch": 0.9319213313161876,
"grad_norm": 1.3842502876081713,
"learning_rate": 2.3234118679127615e-07,
"loss": 0.5941,
"step": 308
},
{
"epoch": 0.9349470499243571,
"grad_norm": 1.3884789864847773,
"learning_rate": 2.117724865092774e-07,
"loss": 0.5943,
"step": 309
},
{
"epoch": 0.9379727685325264,
"grad_norm": 1.3760759747686648,
"learning_rate": 1.921471959676957e-07,
"loss": 0.5796,
"step": 310
},
{
"epoch": 0.9409984871406959,
"grad_norm": 1.3319794266214655,
"learning_rate": 1.734672066928822e-07,
"loss": 0.5549,
"step": 311
},
{
"epoch": 0.9440242057488654,
"grad_norm": 1.4120137218589193,
"learning_rate": 1.5573431910108404e-07,
"loss": 0.5931,
"step": 312
},
{
"epoch": 0.9470499243570348,
"grad_norm": 1.4122176009144072,
"learning_rate": 1.3895024232491338e-07,
"loss": 0.5845,
"step": 313
},
{
"epoch": 0.9500756429652042,
"grad_norm": 1.3674865401057519,
"learning_rate": 1.231165940486234e-07,
"loss": 0.5545,
"step": 314
},
{
"epoch": 0.9531013615733737,
"grad_norm": 1.5231006561759435,
"learning_rate": 1.0823490035218986e-07,
"loss": 0.5614,
"step": 315
},
{
"epoch": 0.9561270801815431,
"grad_norm": 1.3854136210324859,
"learning_rate": 9.43065955642275e-08,
"loss": 0.5816,
"step": 316
},
{
"epoch": 0.9591527987897126,
"grad_norm": 1.3429626155958174,
"learning_rate": 8.133302212373961e-08,
"loss": 0.5562,
"step": 317
},
{
"epoch": 0.962178517397882,
"grad_norm": 1.418941752316243,
"learning_rate": 6.931543045073708e-08,
"loss": 0.5927,
"step": 318
},
{
"epoch": 0.9652042360060514,
"grad_norm": 1.387912020644781,
"learning_rate": 5.8254978825718065e-08,
"loss": 0.5691,
"step": 319
},
{
"epoch": 0.9682299546142209,
"grad_norm": 1.4101899842961938,
"learning_rate": 4.815273327803183e-08,
"loss": 0.5783,
"step": 320
},
{
"epoch": 0.9712556732223904,
"grad_norm": 1.3675769827451856,
"learning_rate": 3.900966748312862e-08,
"loss": 0.5767,
"step": 321
},
{
"epoch": 0.9742813918305597,
"grad_norm": 1.4617606152163445,
"learning_rate": 3.082666266872036e-08,
"loss": 0.5939,
"step": 322
},
{
"epoch": 0.9773071104387292,
"grad_norm": 1.3821813052772716,
"learning_rate": 2.3604507529843e-08,
"loss": 0.5736,
"step": 323
},
{
"epoch": 0.9803328290468987,
"grad_norm": 1.408806118174802,
"learning_rate": 1.7343898152841765e-08,
"loss": 0.6041,
"step": 324
},
{
"epoch": 0.983358547655068,
"grad_norm": 1.3572720348697853,
"learning_rate": 1.2045437948275952e-08,
"loss": 0.5607,
"step": 325
},
{
"epoch": 0.9863842662632375,
"grad_norm": 1.3285080494951755,
"learning_rate": 7.70963759277099e-09,
"loss": 0.572,
"step": 326
},
{
"epoch": 0.989409984871407,
"grad_norm": 1.396106282425076,
"learning_rate": 4.336914979787832e-09,
"loss": 0.5947,
"step": 327
},
{
"epoch": 0.9924357034795764,
"grad_norm": 1.363144404869099,
"learning_rate": 1.9275951793518154e-09,
"loss": 0.5798,
"step": 328
},
{
"epoch": 0.9954614220877458,
"grad_norm": 1.3108458054196304,
"learning_rate": 4.819104067199653e-10,
"loss": 0.5559,
"step": 329
},
{
"epoch": 0.9984871406959153,
"grad_norm": 1.435662614175239,
"learning_rate": 0.0,
"loss": 0.5648,
"step": 330
},
{
"epoch": 0.9984871406959153,
"step": 330,
"total_flos": 276171765841920.0,
"train_loss": 0.6195345849701852,
"train_runtime": 13064.0957,
"train_samples_per_second": 3.238,
"train_steps_per_second": 0.025
}
],
"logging_steps": 1.0,
"max_steps": 330,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 83,
"total_flos": 276171765841920.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}