noteworthy / trainer_state.json
jon-fernandes's picture
Upload trainer_state.json with huggingface_hub
385eaea verified
Raw
History Blame Contribute Delete
465 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 100,
"global_step": 2408,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0016637204949568473,
"grad_norm": 2.482388973236084,
"learning_rate": 0.0,
"loss": 4.477285861968994,
"step": 1
},
{
"epoch": 0.0033274409899136945,
"grad_norm": 2.6023879051208496,
"learning_rate": 4.132231404958678e-08,
"loss": 4.723363876342773,
"step": 2
},
{
"epoch": 0.004991161484870541,
"grad_norm": 2.5578877925872803,
"learning_rate": 8.264462809917357e-08,
"loss": 4.560671806335449,
"step": 3
},
{
"epoch": 0.006654881979827389,
"grad_norm": 2.4592692852020264,
"learning_rate": 1.2396694214876034e-07,
"loss": 4.479525089263916,
"step": 4
},
{
"epoch": 0.008318602474784236,
"grad_norm": 2.3753788471221924,
"learning_rate": 1.6528925619834713e-07,
"loss": 4.508234024047852,
"step": 5
},
{
"epoch": 0.009982322969741083,
"grad_norm": 2.473618984222412,
"learning_rate": 2.066115702479339e-07,
"loss": 4.489030361175537,
"step": 6
},
{
"epoch": 0.011646043464697931,
"grad_norm": 2.4029815196990967,
"learning_rate": 2.4793388429752067e-07,
"loss": 4.410640716552734,
"step": 7
},
{
"epoch": 0.013309763959654778,
"grad_norm": 2.378934621810913,
"learning_rate": 2.892561983471075e-07,
"loss": 4.521676063537598,
"step": 8
},
{
"epoch": 0.014973484454611625,
"grad_norm": 2.234776496887207,
"learning_rate": 3.3057851239669426e-07,
"loss": 4.27446174621582,
"step": 9
},
{
"epoch": 0.01663720494956847,
"grad_norm": 2.0892786979675293,
"learning_rate": 3.7190082644628103e-07,
"loss": 4.092978000640869,
"step": 10
},
{
"epoch": 0.01830092544452532,
"grad_norm": 2.2632219791412354,
"learning_rate": 4.132231404958678e-07,
"loss": 4.533838272094727,
"step": 11
},
{
"epoch": 0.019964645939482165,
"grad_norm": 1.433074712753296,
"learning_rate": 4.5454545454545457e-07,
"loss": 3.4296646118164062,
"step": 12
},
{
"epoch": 0.021628366434439016,
"grad_norm": 1.3564655780792236,
"learning_rate": 4.958677685950413e-07,
"loss": 3.300870895385742,
"step": 13
},
{
"epoch": 0.023292086929395862,
"grad_norm": 1.2193964719772339,
"learning_rate": 5.371900826446281e-07,
"loss": 3.117865800857544,
"step": 14
},
{
"epoch": 0.02495580742435271,
"grad_norm": 1.2911186218261719,
"learning_rate": 5.78512396694215e-07,
"loss": 3.2935025691986084,
"step": 15
},
{
"epoch": 0.026619527919309556,
"grad_norm": 0.36625346541404724,
"learning_rate": 6.198347107438018e-07,
"loss": 2.1773428916931152,
"step": 16
},
{
"epoch": 0.028283248414266403,
"grad_norm": 0.3553527593612671,
"learning_rate": 6.611570247933885e-07,
"loss": 2.2286384105682373,
"step": 17
},
{
"epoch": 0.02994696890922325,
"grad_norm": 0.3257206380367279,
"learning_rate": 7.024793388429753e-07,
"loss": 2.212399959564209,
"step": 18
},
{
"epoch": 0.031610689404180096,
"grad_norm": 0.28514525294303894,
"learning_rate": 7.438016528925621e-07,
"loss": 2.16679048538208,
"step": 19
},
{
"epoch": 0.03327440989913694,
"grad_norm": 0.28953635692596436,
"learning_rate": 7.851239669421488e-07,
"loss": 2.240231990814209,
"step": 20
},
{
"epoch": 0.03493813039409379,
"grad_norm": 0.2675462067127228,
"learning_rate": 8.264462809917356e-07,
"loss": 2.0887300968170166,
"step": 21
},
{
"epoch": 0.03660185088905064,
"grad_norm": 0.2351558357477188,
"learning_rate": 8.677685950413224e-07,
"loss": 1.871799111366272,
"step": 22
},
{
"epoch": 0.038265571384007484,
"grad_norm": 0.2161591798067093,
"learning_rate": 9.090909090909091e-07,
"loss": 1.8348283767700195,
"step": 23
},
{
"epoch": 0.03992929187896433,
"grad_norm": 0.2249748706817627,
"learning_rate": 9.50413223140496e-07,
"loss": 1.9596117734909058,
"step": 24
},
{
"epoch": 0.041593012373921184,
"grad_norm": 0.21401794254779816,
"learning_rate": 9.917355371900827e-07,
"loss": 1.9094750881195068,
"step": 25
},
{
"epoch": 0.04325673286887803,
"grad_norm": 0.2133910208940506,
"learning_rate": 1.0330578512396695e-06,
"loss": 1.9063758850097656,
"step": 26
},
{
"epoch": 0.04492045336383488,
"grad_norm": 0.20730602741241455,
"learning_rate": 1.0743801652892562e-06,
"loss": 1.8436386585235596,
"step": 27
},
{
"epoch": 0.046584173858791725,
"grad_norm": 0.20181401073932648,
"learning_rate": 1.115702479338843e-06,
"loss": 1.8216067552566528,
"step": 28
},
{
"epoch": 0.04824789435374857,
"grad_norm": 0.2049887478351593,
"learning_rate": 1.15702479338843e-06,
"loss": 1.8613402843475342,
"step": 29
},
{
"epoch": 0.04991161484870542,
"grad_norm": 0.18033020198345184,
"learning_rate": 1.1983471074380167e-06,
"loss": 1.6658318042755127,
"step": 30
},
{
"epoch": 0.051575335343662265,
"grad_norm": 0.15923167765140533,
"learning_rate": 1.2396694214876035e-06,
"loss": 1.532716989517212,
"step": 31
},
{
"epoch": 0.05323905583861911,
"grad_norm": 0.1302931010723114,
"learning_rate": 1.28099173553719e-06,
"loss": 1.4400815963745117,
"step": 32
},
{
"epoch": 0.05490277633357596,
"grad_norm": 0.11385945230722427,
"learning_rate": 1.322314049586777e-06,
"loss": 1.3887157440185547,
"step": 33
},
{
"epoch": 0.056566496828532806,
"grad_norm": 0.11334631592035294,
"learning_rate": 1.3636363636363636e-06,
"loss": 1.3524459600448608,
"step": 34
},
{
"epoch": 0.05823021732348965,
"grad_norm": 0.10479549318552017,
"learning_rate": 1.4049586776859506e-06,
"loss": 1.292414903640747,
"step": 35
},
{
"epoch": 0.0598939378184465,
"grad_norm": 0.09981797635555267,
"learning_rate": 1.4462809917355372e-06,
"loss": 1.2074761390686035,
"step": 36
},
{
"epoch": 0.061557658313403346,
"grad_norm": 0.10237134248018265,
"learning_rate": 1.4876033057851241e-06,
"loss": 1.2024866342544556,
"step": 37
},
{
"epoch": 0.06322137880836019,
"grad_norm": 0.09421642869710922,
"learning_rate": 1.5289256198347107e-06,
"loss": 1.1445802450180054,
"step": 38
},
{
"epoch": 0.06488509930331704,
"grad_norm": 0.08208701759576797,
"learning_rate": 1.5702479338842977e-06,
"loss": 1.114914894104004,
"step": 39
},
{
"epoch": 0.06654881979827389,
"grad_norm": 0.0781988576054573,
"learning_rate": 1.6115702479338842e-06,
"loss": 1.1362968683242798,
"step": 40
},
{
"epoch": 0.06821254029323073,
"grad_norm": 0.06956593692302704,
"learning_rate": 1.6528925619834712e-06,
"loss": 1.050331950187683,
"step": 41
},
{
"epoch": 0.06987626078818758,
"grad_norm": 0.07109206914901733,
"learning_rate": 1.694214876033058e-06,
"loss": 1.0830775499343872,
"step": 42
},
{
"epoch": 0.07153998128314443,
"grad_norm": 0.06287192553281784,
"learning_rate": 1.7355371900826448e-06,
"loss": 1.0035347938537598,
"step": 43
},
{
"epoch": 0.07320370177810127,
"grad_norm": 0.060101497918367386,
"learning_rate": 1.7768595041322315e-06,
"loss": 0.9792177677154541,
"step": 44
},
{
"epoch": 0.07486742227305812,
"grad_norm": 0.05643174424767494,
"learning_rate": 1.8181818181818183e-06,
"loss": 0.9080207347869873,
"step": 45
},
{
"epoch": 0.07653114276801497,
"grad_norm": 0.05656706541776657,
"learning_rate": 1.859504132231405e-06,
"loss": 0.9312663078308105,
"step": 46
},
{
"epoch": 0.07819486326297181,
"grad_norm": 0.04832266643643379,
"learning_rate": 1.900826446280992e-06,
"loss": 0.8578022718429565,
"step": 47
},
{
"epoch": 0.07985858375792866,
"grad_norm": 0.051747020334005356,
"learning_rate": 1.9421487603305786e-06,
"loss": 0.9018316268920898,
"step": 48
},
{
"epoch": 0.08152230425288552,
"grad_norm": 0.04696454107761383,
"learning_rate": 1.9834710743801654e-06,
"loss": 0.8227735757827759,
"step": 49
},
{
"epoch": 0.08318602474784237,
"grad_norm": 0.04074595496058464,
"learning_rate": 2.024793388429752e-06,
"loss": 0.7895236015319824,
"step": 50
},
{
"epoch": 0.08484974524279922,
"grad_norm": 0.041805122047662735,
"learning_rate": 2.066115702479339e-06,
"loss": 0.7831016778945923,
"step": 51
},
{
"epoch": 0.08651346573775606,
"grad_norm": 0.03949985280632973,
"learning_rate": 2.1074380165289257e-06,
"loss": 0.7389830350875854,
"step": 52
},
{
"epoch": 0.08817718623271291,
"grad_norm": 0.041112784296274185,
"learning_rate": 2.1487603305785124e-06,
"loss": 0.7490829229354858,
"step": 53
},
{
"epoch": 0.08984090672766976,
"grad_norm": 0.044654764235019684,
"learning_rate": 2.1900826446280992e-06,
"loss": 0.7309138774871826,
"step": 54
},
{
"epoch": 0.0915046272226266,
"grad_norm": 0.05389357730746269,
"learning_rate": 2.231404958677686e-06,
"loss": 0.7365468740463257,
"step": 55
},
{
"epoch": 0.09316834771758345,
"grad_norm": 0.044067952781915665,
"learning_rate": 2.2727272727272728e-06,
"loss": 0.7065483927726746,
"step": 56
},
{
"epoch": 0.0948320682125403,
"grad_norm": 0.03494361788034439,
"learning_rate": 2.31404958677686e-06,
"loss": 0.7050533294677734,
"step": 57
},
{
"epoch": 0.09649578870749714,
"grad_norm": 0.027442242950201035,
"learning_rate": 2.3553719008264463e-06,
"loss": 0.6887029409408569,
"step": 58
},
{
"epoch": 0.09815950920245399,
"grad_norm": 0.031428925693035126,
"learning_rate": 2.3966942148760335e-06,
"loss": 0.6625376343727112,
"step": 59
},
{
"epoch": 0.09982322969741084,
"grad_norm": 0.032608762383461,
"learning_rate": 2.43801652892562e-06,
"loss": 0.6581825017929077,
"step": 60
},
{
"epoch": 0.10148695019236768,
"grad_norm": 0.02459588088095188,
"learning_rate": 2.479338842975207e-06,
"loss": 0.6294938325881958,
"step": 61
},
{
"epoch": 0.10315067068732453,
"grad_norm": 0.01995225064456463,
"learning_rate": 2.5206611570247934e-06,
"loss": 0.6105325222015381,
"step": 62
},
{
"epoch": 0.10481439118228138,
"grad_norm": 0.028617018833756447,
"learning_rate": 2.56198347107438e-06,
"loss": 0.6707576513290405,
"step": 63
},
{
"epoch": 0.10647811167723822,
"grad_norm": 0.02642955631017685,
"learning_rate": 2.6033057851239673e-06,
"loss": 0.624178409576416,
"step": 64
},
{
"epoch": 0.10814183217219507,
"grad_norm": 0.025090189650654793,
"learning_rate": 2.644628099173554e-06,
"loss": 0.6277080774307251,
"step": 65
},
{
"epoch": 0.10980555266715192,
"grad_norm": 0.03007493168115616,
"learning_rate": 2.6859504132231405e-06,
"loss": 0.6461099982261658,
"step": 66
},
{
"epoch": 0.11146927316210876,
"grad_norm": 0.026639141142368317,
"learning_rate": 2.7272727272727272e-06,
"loss": 0.6154550909996033,
"step": 67
},
{
"epoch": 0.11313299365706561,
"grad_norm": 0.029324783012270927,
"learning_rate": 2.7685950413223144e-06,
"loss": 0.6163111329078674,
"step": 68
},
{
"epoch": 0.11479671415202246,
"grad_norm": 0.02367994748055935,
"learning_rate": 2.809917355371901e-06,
"loss": 0.6023034453392029,
"step": 69
},
{
"epoch": 0.1164604346469793,
"grad_norm": 0.02719041146337986,
"learning_rate": 2.851239669421488e-06,
"loss": 0.6273308992385864,
"step": 70
},
{
"epoch": 0.11812415514193615,
"grad_norm": 0.02549588494002819,
"learning_rate": 2.8925619834710743e-06,
"loss": 0.5701053142547607,
"step": 71
},
{
"epoch": 0.119787875636893,
"grad_norm": 0.022750338539481163,
"learning_rate": 2.9338842975206615e-06,
"loss": 0.5723459720611572,
"step": 72
},
{
"epoch": 0.12145159613184985,
"grad_norm": 0.025367699563503265,
"learning_rate": 2.9752066115702483e-06,
"loss": 0.5892397165298462,
"step": 73
},
{
"epoch": 0.12311531662680669,
"grad_norm": 0.0319160595536232,
"learning_rate": 3.016528925619835e-06,
"loss": 0.5924912691116333,
"step": 74
},
{
"epoch": 0.12477903712176354,
"grad_norm": 0.021769287064671516,
"learning_rate": 3.0578512396694214e-06,
"loss": 0.5665951371192932,
"step": 75
},
{
"epoch": 0.12644275761672039,
"grad_norm": 0.027082569897174835,
"learning_rate": 3.0991735537190086e-06,
"loss": 0.563019871711731,
"step": 76
},
{
"epoch": 0.12810647811167725,
"grad_norm": 0.022641802206635475,
"learning_rate": 3.1404958677685953e-06,
"loss": 0.5703021883964539,
"step": 77
},
{
"epoch": 0.12977019860663408,
"grad_norm": 0.021977661177515984,
"learning_rate": 3.181818181818182e-06,
"loss": 0.5566818118095398,
"step": 78
},
{
"epoch": 0.13143391910159094,
"grad_norm": 0.02996179461479187,
"learning_rate": 3.2231404958677685e-06,
"loss": 0.566483199596405,
"step": 79
},
{
"epoch": 0.13309763959654777,
"grad_norm": 0.02014978975057602,
"learning_rate": 3.264462809917356e-06,
"loss": 0.5580374002456665,
"step": 80
},
{
"epoch": 0.13476136009150463,
"grad_norm": 0.03401867672801018,
"learning_rate": 3.3057851239669424e-06,
"loss": 0.5588016510009766,
"step": 81
},
{
"epoch": 0.13642508058646147,
"grad_norm": 0.022355899214744568,
"learning_rate": 3.347107438016529e-06,
"loss": 0.5484458208084106,
"step": 82
},
{
"epoch": 0.13808880108141833,
"grad_norm": 0.02415446750819683,
"learning_rate": 3.388429752066116e-06,
"loss": 0.5682621002197266,
"step": 83
},
{
"epoch": 0.13975252157637516,
"grad_norm": 0.023742761462926865,
"learning_rate": 3.429752066115703e-06,
"loss": 0.5533846616744995,
"step": 84
},
{
"epoch": 0.14141624207133202,
"grad_norm": 0.022532276809215546,
"learning_rate": 3.4710743801652895e-06,
"loss": 0.5288005471229553,
"step": 85
},
{
"epoch": 0.14307996256628885,
"grad_norm": 0.019874311983585358,
"learning_rate": 3.5123966942148763e-06,
"loss": 0.5370491743087769,
"step": 86
},
{
"epoch": 0.14474368306124571,
"grad_norm": 0.022755231708288193,
"learning_rate": 3.553719008264463e-06,
"loss": 0.511932373046875,
"step": 87
},
{
"epoch": 0.14640740355620255,
"grad_norm": 0.02134534902870655,
"learning_rate": 3.5950413223140502e-06,
"loss": 0.49379733204841614,
"step": 88
},
{
"epoch": 0.1480711240511594,
"grad_norm": 0.021491173654794693,
"learning_rate": 3.6363636363636366e-06,
"loss": 0.4925214648246765,
"step": 89
},
{
"epoch": 0.14973484454611624,
"grad_norm": 0.02258123643696308,
"learning_rate": 3.6776859504132234e-06,
"loss": 0.5131945610046387,
"step": 90
},
{
"epoch": 0.1513985650410731,
"grad_norm": 0.025489268824458122,
"learning_rate": 3.71900826446281e-06,
"loss": 0.4968642592430115,
"step": 91
},
{
"epoch": 0.15306228553602994,
"grad_norm": 0.02109348215162754,
"learning_rate": 3.7603305785123973e-06,
"loss": 0.49502578377723694,
"step": 92
},
{
"epoch": 0.1547260060309868,
"grad_norm": 0.025884000584483147,
"learning_rate": 3.801652892561984e-06,
"loss": 0.4799806475639343,
"step": 93
},
{
"epoch": 0.15638972652594363,
"grad_norm": 0.023585375398397446,
"learning_rate": 3.842975206611571e-06,
"loss": 0.49822694063186646,
"step": 94
},
{
"epoch": 0.1580534470209005,
"grad_norm": 0.031796254217624664,
"learning_rate": 3.884297520661157e-06,
"loss": 0.49923157691955566,
"step": 95
},
{
"epoch": 0.15971716751585732,
"grad_norm": 0.03561358526349068,
"learning_rate": 3.925619834710744e-06,
"loss": 0.4884190261363983,
"step": 96
},
{
"epoch": 0.16138088801081418,
"grad_norm": 0.021946346387267113,
"learning_rate": 3.966942148760331e-06,
"loss": 0.4643127918243408,
"step": 97
},
{
"epoch": 0.16304460850577104,
"grad_norm": 0.028085844591259956,
"learning_rate": 4.008264462809918e-06,
"loss": 0.4594374895095825,
"step": 98
},
{
"epoch": 0.16470832900072788,
"grad_norm": 0.027761919423937798,
"learning_rate": 4.049586776859504e-06,
"loss": 0.4878089725971222,
"step": 99
},
{
"epoch": 0.16637204949568474,
"grad_norm": 0.02483516000211239,
"learning_rate": 4.0909090909090915e-06,
"loss": 0.4365704357624054,
"step": 100
},
{
"epoch": 0.16637204949568474,
"eval_loss": 0.4434735178947449,
"eval_runtime": 398.423,
"eval_samples_per_second": 12.068,
"eval_steps_per_second": 3.017,
"step": 100
},
{
"epoch": 0.16803576999064157,
"grad_norm": 0.026426592841744423,
"learning_rate": 4.132231404958678e-06,
"loss": 0.42997342348098755,
"step": 101
},
{
"epoch": 0.16969949048559843,
"grad_norm": 0.02919408679008484,
"learning_rate": 4.173553719008265e-06,
"loss": 0.4375656843185425,
"step": 102
},
{
"epoch": 0.17136321098055526,
"grad_norm": 0.03021477535367012,
"learning_rate": 4.214876033057851e-06,
"loss": 0.4166721999645233,
"step": 103
},
{
"epoch": 0.17302693147551212,
"grad_norm": 0.0274681244045496,
"learning_rate": 4.2561983471074386e-06,
"loss": 0.41593968868255615,
"step": 104
},
{
"epoch": 0.17469065197046896,
"grad_norm": 0.026254650205373764,
"learning_rate": 4.297520661157025e-06,
"loss": 0.42187049984931946,
"step": 105
},
{
"epoch": 0.17635437246542582,
"grad_norm": 0.04353519156575203,
"learning_rate": 4.338842975206612e-06,
"loss": 0.41187649965286255,
"step": 106
},
{
"epoch": 0.17801809296038265,
"grad_norm": 0.02748814970254898,
"learning_rate": 4.3801652892561984e-06,
"loss": 0.4013650715351105,
"step": 107
},
{
"epoch": 0.1796818134553395,
"grad_norm": 0.053706467151641846,
"learning_rate": 4.421487603305786e-06,
"loss": 0.4117678701877594,
"step": 108
},
{
"epoch": 0.18134553395029634,
"grad_norm": 0.04221956431865692,
"learning_rate": 4.462809917355372e-06,
"loss": 0.3744428753852844,
"step": 109
},
{
"epoch": 0.1830092544452532,
"grad_norm": 0.04144236072897911,
"learning_rate": 4.504132231404959e-06,
"loss": 0.37903785705566406,
"step": 110
},
{
"epoch": 0.18467297494021004,
"grad_norm": 0.038312312215566635,
"learning_rate": 4.5454545454545455e-06,
"loss": 0.3629874885082245,
"step": 111
},
{
"epoch": 0.1863366954351669,
"grad_norm": 0.0359899066388607,
"learning_rate": 4.586776859504133e-06,
"loss": 0.37735822796821594,
"step": 112
},
{
"epoch": 0.18800041593012373,
"grad_norm": 0.030285200104117393,
"learning_rate": 4.62809917355372e-06,
"loss": 0.3504555821418762,
"step": 113
},
{
"epoch": 0.1896641364250806,
"grad_norm": 0.04384183883666992,
"learning_rate": 4.669421487603306e-06,
"loss": 0.36508214473724365,
"step": 114
},
{
"epoch": 0.19132785692003743,
"grad_norm": 0.039419692009687424,
"learning_rate": 4.710743801652893e-06,
"loss": 0.3422982096672058,
"step": 115
},
{
"epoch": 0.1929915774149943,
"grad_norm": 0.0356181338429451,
"learning_rate": 4.75206611570248e-06,
"loss": 0.3440757989883423,
"step": 116
},
{
"epoch": 0.19465529790995112,
"grad_norm": 0.041206084191799164,
"learning_rate": 4.793388429752067e-06,
"loss": 0.32975438237190247,
"step": 117
},
{
"epoch": 0.19631901840490798,
"grad_norm": 0.025390787050127983,
"learning_rate": 4.834710743801653e-06,
"loss": 0.33478546142578125,
"step": 118
},
{
"epoch": 0.1979827388998648,
"grad_norm": 0.03742411360144615,
"learning_rate": 4.87603305785124e-06,
"loss": 0.30944791436195374,
"step": 119
},
{
"epoch": 0.19964645939482167,
"grad_norm": 0.02903994731605053,
"learning_rate": 4.917355371900827e-06,
"loss": 0.2991634011268616,
"step": 120
},
{
"epoch": 0.2013101798897785,
"grad_norm": 0.03681575134396553,
"learning_rate": 4.958677685950414e-06,
"loss": 0.28756508231163025,
"step": 121
},
{
"epoch": 0.20297390038473537,
"grad_norm": 0.031143750995397568,
"learning_rate": 5e-06,
"loss": 0.286681592464447,
"step": 122
},
{
"epoch": 0.2046376208796922,
"grad_norm": 0.030605066567659378,
"learning_rate": 4.999997641274725e-06,
"loss": 0.31048089265823364,
"step": 123
},
{
"epoch": 0.20630134137464906,
"grad_norm": 0.03684311732649803,
"learning_rate": 4.999990565103349e-06,
"loss": 0.2921098470687866,
"step": 124
},
{
"epoch": 0.2079650618696059,
"grad_norm": 0.05223636329174042,
"learning_rate": 4.999978771499224e-06,
"loss": 0.2894522249698639,
"step": 125
},
{
"epoch": 0.20962878236456275,
"grad_norm": 0.02680801972746849,
"learning_rate": 4.999962260484607e-06,
"loss": 0.2657696604728699,
"step": 126
},
{
"epoch": 0.2112925028595196,
"grad_norm": 0.028682326897978783,
"learning_rate": 4.999941032090652e-06,
"loss": 0.27642691135406494,
"step": 127
},
{
"epoch": 0.21295622335447645,
"grad_norm": 0.059440821409225464,
"learning_rate": 4.999915086357417e-06,
"loss": 0.27616798877716064,
"step": 128
},
{
"epoch": 0.2146199438494333,
"grad_norm": 0.038857702165842056,
"learning_rate": 4.99988442333386e-06,
"loss": 0.26294657588005066,
"step": 129
},
{
"epoch": 0.21628366434439014,
"grad_norm": 0.03957323357462883,
"learning_rate": 4.999849043077843e-06,
"loss": 0.2662772536277771,
"step": 130
},
{
"epoch": 0.217947384839347,
"grad_norm": 0.03721378371119499,
"learning_rate": 4.999808945656128e-06,
"loss": 0.25390905141830444,
"step": 131
},
{
"epoch": 0.21961110533430384,
"grad_norm": 0.037432651966810226,
"learning_rate": 4.999764131144377e-06,
"loss": 0.24143514037132263,
"step": 132
},
{
"epoch": 0.2212748258292607,
"grad_norm": 0.029236387461423874,
"learning_rate": 4.999714599627155e-06,
"loss": 0.2406114637851715,
"step": 133
},
{
"epoch": 0.22293854632421753,
"grad_norm": 0.04499300941824913,
"learning_rate": 4.999660351197926e-06,
"loss": 0.24877291917800903,
"step": 134
},
{
"epoch": 0.2246022668191744,
"grad_norm": 0.026736732572317123,
"learning_rate": 4.999601385959056e-06,
"loss": 0.22672352194786072,
"step": 135
},
{
"epoch": 0.22626598731413122,
"grad_norm": 0.046254824846982956,
"learning_rate": 4.999537704021812e-06,
"loss": 0.2213183343410492,
"step": 136
},
{
"epoch": 0.22792970780908808,
"grad_norm": 0.02913353592157364,
"learning_rate": 4.99946930550636e-06,
"loss": 0.21404042840003967,
"step": 137
},
{
"epoch": 0.22959342830404492,
"grad_norm": 0.05235866829752922,
"learning_rate": 4.999396190541766e-06,
"loss": 0.23582462966442108,
"step": 138
},
{
"epoch": 0.23125714879900178,
"grad_norm": 0.030388308688998222,
"learning_rate": 4.999318359265998e-06,
"loss": 0.22350290417671204,
"step": 139
},
{
"epoch": 0.2329208692939586,
"grad_norm": 0.04241563007235527,
"learning_rate": 4.999235811825921e-06,
"loss": 0.2113513946533203,
"step": 140
},
{
"epoch": 0.23458458978891547,
"grad_norm": 0.0259566493332386,
"learning_rate": 4.9991485483773e-06,
"loss": 0.2061123251914978,
"step": 141
},
{
"epoch": 0.2362483102838723,
"grad_norm": 0.04808220639824867,
"learning_rate": 4.999056569084801e-06,
"loss": 0.24066820740699768,
"step": 142
},
{
"epoch": 0.23791203077882916,
"grad_norm": 0.027925971895456314,
"learning_rate": 4.998959874121986e-06,
"loss": 0.2096669226884842,
"step": 143
},
{
"epoch": 0.239575751273786,
"grad_norm": 0.04741431772708893,
"learning_rate": 4.998858463671316e-06,
"loss": 0.2042136937379837,
"step": 144
},
{
"epoch": 0.24123947176874286,
"grad_norm": 0.030547140166163445,
"learning_rate": 4.998752337924152e-06,
"loss": 0.16891774535179138,
"step": 145
},
{
"epoch": 0.2429031922636997,
"grad_norm": 0.041504278779029846,
"learning_rate": 4.998641497080749e-06,
"loss": 0.1880994737148285,
"step": 146
},
{
"epoch": 0.24456691275865655,
"grad_norm": 0.023324500769376755,
"learning_rate": 4.998525941350264e-06,
"loss": 0.18950693309307098,
"step": 147
},
{
"epoch": 0.24623063325361338,
"grad_norm": 0.037083689123392105,
"learning_rate": 4.998405670950747e-06,
"loss": 0.18887647986412048,
"step": 148
},
{
"epoch": 0.24789435374857025,
"grad_norm": 0.023503053933382034,
"learning_rate": 4.998280686109146e-06,
"loss": 0.16628959774971008,
"step": 149
},
{
"epoch": 0.24955807424352708,
"grad_norm": 0.030192900449037552,
"learning_rate": 4.998150987061304e-06,
"loss": 0.1572273075580597,
"step": 150
},
{
"epoch": 0.25122179473848394,
"grad_norm": 0.03396950662136078,
"learning_rate": 4.9980165740519625e-06,
"loss": 0.16307464241981506,
"step": 151
},
{
"epoch": 0.25288551523344077,
"grad_norm": 0.0340605154633522,
"learning_rate": 4.997877447334754e-06,
"loss": 0.16034038364887238,
"step": 152
},
{
"epoch": 0.2545492357283976,
"grad_norm": 0.03912411257624626,
"learning_rate": 4.99773360717221e-06,
"loss": 0.14050956070423126,
"step": 153
},
{
"epoch": 0.2562129562233545,
"grad_norm": 0.03576822578907013,
"learning_rate": 4.997585053835754e-06,
"loss": 0.13379324972629547,
"step": 154
},
{
"epoch": 0.2578766767183113,
"grad_norm": 0.03452559560537338,
"learning_rate": 4.997431787605701e-06,
"loss": 0.1446094661951065,
"step": 155
},
{
"epoch": 0.25954039721326816,
"grad_norm": 0.037528105080127716,
"learning_rate": 4.997273808771263e-06,
"loss": 0.15133655071258545,
"step": 156
},
{
"epoch": 0.261204117708225,
"grad_norm": 0.04468218609690666,
"learning_rate": 4.997111117630543e-06,
"loss": 0.1459997594356537,
"step": 157
},
{
"epoch": 0.2628678382031819,
"grad_norm": 0.034537848085165024,
"learning_rate": 4.996943714490535e-06,
"loss": 0.15663175284862518,
"step": 158
},
{
"epoch": 0.2645315586981387,
"grad_norm": 0.04926777631044388,
"learning_rate": 4.996771599667126e-06,
"loss": 0.1453685462474823,
"step": 159
},
{
"epoch": 0.26619527919309555,
"grad_norm": 0.02621885947883129,
"learning_rate": 4.996594773485093e-06,
"loss": 0.12285202741622925,
"step": 160
},
{
"epoch": 0.26785899968805243,
"grad_norm": 0.046555280685424805,
"learning_rate": 4.996413236278104e-06,
"loss": 0.1389390230178833,
"step": 161
},
{
"epoch": 0.26952272018300927,
"grad_norm": 0.023526031523942947,
"learning_rate": 4.996226988388716e-06,
"loss": 0.13394689559936523,
"step": 162
},
{
"epoch": 0.2711864406779661,
"grad_norm": 0.046248551458120346,
"learning_rate": 4.9960360301683755e-06,
"loss": 0.14260414242744446,
"step": 163
},
{
"epoch": 0.27285016117292293,
"grad_norm": 0.02372513897716999,
"learning_rate": 4.995840361977416e-06,
"loss": 0.11465226113796234,
"step": 164
},
{
"epoch": 0.2745138816678798,
"grad_norm": 0.03185545653104782,
"learning_rate": 4.995639984185059e-06,
"loss": 0.12264078855514526,
"step": 165
},
{
"epoch": 0.27617760216283666,
"grad_norm": 0.03572575002908707,
"learning_rate": 4.9954348971694146e-06,
"loss": 0.11168736964464188,
"step": 166
},
{
"epoch": 0.2778413226577935,
"grad_norm": 0.02216448448598385,
"learning_rate": 4.995225101317478e-06,
"loss": 0.11359298229217529,
"step": 167
},
{
"epoch": 0.2795050431527503,
"grad_norm": 0.035710953176021576,
"learning_rate": 4.99501059702513e-06,
"loss": 0.12145813554525375,
"step": 168
},
{
"epoch": 0.2811687636477072,
"grad_norm": 0.0302723441272974,
"learning_rate": 4.9947913846971345e-06,
"loss": 0.09259926527738571,
"step": 169
},
{
"epoch": 0.28283248414266404,
"grad_norm": 0.018867801874876022,
"learning_rate": 4.994567464747141e-06,
"loss": 0.1060592383146286,
"step": 170
},
{
"epoch": 0.2844962046376209,
"grad_norm": 0.022908039391040802,
"learning_rate": 4.994338837597683e-06,
"loss": 0.09731046855449677,
"step": 171
},
{
"epoch": 0.2861599251325777,
"grad_norm": 0.024193648248910904,
"learning_rate": 4.994105503680176e-06,
"loss": 0.09389811009168625,
"step": 172
},
{
"epoch": 0.2878236456275346,
"grad_norm": 0.01756378822028637,
"learning_rate": 4.993867463434916e-06,
"loss": 0.09718440473079681,
"step": 173
},
{
"epoch": 0.28948736612249143,
"grad_norm": 0.0198496226221323,
"learning_rate": 4.9936247173110785e-06,
"loss": 0.0892348438501358,
"step": 174
},
{
"epoch": 0.29115108661744826,
"grad_norm": 0.023847704753279686,
"learning_rate": 4.993377265766723e-06,
"loss": 0.07874070107936859,
"step": 175
},
{
"epoch": 0.2928148071124051,
"grad_norm": 0.02083473838865757,
"learning_rate": 4.993125109268784e-06,
"loss": 0.09188570082187653,
"step": 176
},
{
"epoch": 0.294478527607362,
"grad_norm": 0.023864734917879105,
"learning_rate": 4.992868248293077e-06,
"loss": 0.08353804051876068,
"step": 177
},
{
"epoch": 0.2961422481023188,
"grad_norm": 0.01996394246816635,
"learning_rate": 4.9926066833242926e-06,
"loss": 0.07662493735551834,
"step": 178
},
{
"epoch": 0.29780596859727565,
"grad_norm": 0.01867249421775341,
"learning_rate": 4.9923404148559995e-06,
"loss": 0.06500700116157532,
"step": 179
},
{
"epoch": 0.2994696890922325,
"grad_norm": 0.020658332854509354,
"learning_rate": 4.992069443390641e-06,
"loss": 0.08731688559055328,
"step": 180
},
{
"epoch": 0.30113340958718937,
"grad_norm": 0.019181935116648674,
"learning_rate": 4.991793769439534e-06,
"loss": 0.06426648795604706,
"step": 181
},
{
"epoch": 0.3027971300821462,
"grad_norm": 0.02977469190955162,
"learning_rate": 4.991513393522871e-06,
"loss": 0.07214197516441345,
"step": 182
},
{
"epoch": 0.30446085057710304,
"grad_norm": 0.01673358865082264,
"learning_rate": 4.991228316169715e-06,
"loss": 0.07281684875488281,
"step": 183
},
{
"epoch": 0.30612457107205987,
"grad_norm": 0.01972797140479088,
"learning_rate": 4.990938537918001e-06,
"loss": 0.06969514489173889,
"step": 184
},
{
"epoch": 0.30778829156701676,
"grad_norm": 0.026684733107686043,
"learning_rate": 4.990644059314536e-06,
"loss": 0.07704263925552368,
"step": 185
},
{
"epoch": 0.3094520120619736,
"grad_norm": 0.01813516579568386,
"learning_rate": 4.990344880914994e-06,
"loss": 0.06715594232082367,
"step": 186
},
{
"epoch": 0.3111157325569304,
"grad_norm": 0.021501829847693443,
"learning_rate": 4.990041003283921e-06,
"loss": 0.06312450766563416,
"step": 187
},
{
"epoch": 0.31277945305188726,
"grad_norm": 0.023688537999987602,
"learning_rate": 4.989732426994725e-06,
"loss": 0.07297050952911377,
"step": 188
},
{
"epoch": 0.31444317354684415,
"grad_norm": 0.020870845764875412,
"learning_rate": 4.989419152629685e-06,
"loss": 0.05973198264837265,
"step": 189
},
{
"epoch": 0.316106894041801,
"grad_norm": 0.031512096524238586,
"learning_rate": 4.9891011807799435e-06,
"loss": 0.06264424324035645,
"step": 190
},
{
"epoch": 0.3177706145367578,
"grad_norm": 0.013584833592176437,
"learning_rate": 4.988778512045507e-06,
"loss": 0.05872301757335663,
"step": 191
},
{
"epoch": 0.31943433503171464,
"grad_norm": 0.022901279851794243,
"learning_rate": 4.9884511470352456e-06,
"loss": 0.058490075170993805,
"step": 192
},
{
"epoch": 0.32109805552667153,
"grad_norm": 0.023982392624020576,
"learning_rate": 4.9881190863668895e-06,
"loss": 0.05945117026567459,
"step": 193
},
{
"epoch": 0.32276177602162837,
"grad_norm": 0.014951342716813087,
"learning_rate": 4.98778233066703e-06,
"loss": 0.04956051707267761,
"step": 194
},
{
"epoch": 0.3244254965165852,
"grad_norm": 0.034372977912425995,
"learning_rate": 4.987440880571121e-06,
"loss": 0.06177012622356415,
"step": 195
},
{
"epoch": 0.3260892170115421,
"grad_norm": 0.01750808022916317,
"learning_rate": 4.98709473672347e-06,
"loss": 0.04706805199384689,
"step": 196
},
{
"epoch": 0.3277529375064989,
"grad_norm": 0.017525361850857735,
"learning_rate": 4.986743899777244e-06,
"loss": 0.05709604173898697,
"step": 197
},
{
"epoch": 0.32941665800145575,
"grad_norm": 0.026951247826218605,
"learning_rate": 4.986388370394466e-06,
"loss": 0.05336014926433563,
"step": 198
},
{
"epoch": 0.3310803784964126,
"grad_norm": 0.019481414929032326,
"learning_rate": 4.986028149246013e-06,
"loss": 0.04981919378042221,
"step": 199
},
{
"epoch": 0.3327440989913695,
"grad_norm": 0.017069054767489433,
"learning_rate": 4.985663237011614e-06,
"loss": 0.04121176898479462,
"step": 200
},
{
"epoch": 0.3327440989913695,
"eval_loss": 0.04646755009889603,
"eval_runtime": 402.3664,
"eval_samples_per_second": 11.949,
"eval_steps_per_second": 2.987,
"step": 200
},
{
"epoch": 0.3344078194863263,
"grad_norm": 0.01519837137311697,
"learning_rate": 4.985293634379852e-06,
"loss": 0.05122241750359535,
"step": 201
},
{
"epoch": 0.33607153998128314,
"grad_norm": 0.018817614763975143,
"learning_rate": 4.984919342048159e-06,
"loss": 0.052712492644786835,
"step": 202
},
{
"epoch": 0.33773526047624,
"grad_norm": 0.015790430828928947,
"learning_rate": 4.984540360722819e-06,
"loss": 0.041432932019233704,
"step": 203
},
{
"epoch": 0.33939898097119686,
"grad_norm": 0.013813985511660576,
"learning_rate": 4.98415669111896e-06,
"loss": 0.044672828167676926,
"step": 204
},
{
"epoch": 0.3410627014661537,
"grad_norm": 0.018357550725340843,
"learning_rate": 4.9837683339605615e-06,
"loss": 0.05308464914560318,
"step": 205
},
{
"epoch": 0.34272642196111053,
"grad_norm": 0.013275988399982452,
"learning_rate": 4.983375289980443e-06,
"loss": 0.05014698952436447,
"step": 206
},
{
"epoch": 0.34439014245606736,
"grad_norm": 0.014597652480006218,
"learning_rate": 4.982977559920273e-06,
"loss": 0.04254170134663582,
"step": 207
},
{
"epoch": 0.34605386295102425,
"grad_norm": 0.014993159100413322,
"learning_rate": 4.982575144530559e-06,
"loss": 0.040170349180698395,
"step": 208
},
{
"epoch": 0.3477175834459811,
"grad_norm": 0.011311318725347519,
"learning_rate": 4.982168044570652e-06,
"loss": 0.04351898282766342,
"step": 209
},
{
"epoch": 0.3493813039409379,
"grad_norm": 0.01357428077608347,
"learning_rate": 4.981756260808741e-06,
"loss": 0.045017972588539124,
"step": 210
},
{
"epoch": 0.35104502443589475,
"grad_norm": 0.011063991114497185,
"learning_rate": 4.981339794021853e-06,
"loss": 0.036168403923511505,
"step": 211
},
{
"epoch": 0.35270874493085164,
"grad_norm": 0.015618620440363884,
"learning_rate": 4.9809186449958536e-06,
"loss": 0.03904163837432861,
"step": 212
},
{
"epoch": 0.35437246542580847,
"grad_norm": 0.010496679693460464,
"learning_rate": 4.980492814525442e-06,
"loss": 0.026133833453059196,
"step": 213
},
{
"epoch": 0.3560361859207653,
"grad_norm": 0.013019095174968243,
"learning_rate": 4.980062303414152e-06,
"loss": 0.03268534690141678,
"step": 214
},
{
"epoch": 0.35769990641572214,
"grad_norm": 0.013511061668395996,
"learning_rate": 4.97962711247435e-06,
"loss": 0.03932594135403633,
"step": 215
},
{
"epoch": 0.359363626910679,
"grad_norm": 0.013383673503994942,
"learning_rate": 4.979187242527233e-06,
"loss": 0.0299623254686594,
"step": 216
},
{
"epoch": 0.36102734740563586,
"grad_norm": 0.030445793643593788,
"learning_rate": 4.978742694402825e-06,
"loss": 0.03553349897265434,
"step": 217
},
{
"epoch": 0.3626910679005927,
"grad_norm": 0.013040585443377495,
"learning_rate": 4.978293468939982e-06,
"loss": 0.035122938454151154,
"step": 218
},
{
"epoch": 0.3643547883955495,
"grad_norm": 0.015449252910912037,
"learning_rate": 4.977839566986382e-06,
"loss": 0.0295286662876606,
"step": 219
},
{
"epoch": 0.3660185088905064,
"grad_norm": 0.026898793876171112,
"learning_rate": 4.977380989398529e-06,
"loss": 0.047140851616859436,
"step": 220
},
{
"epoch": 0.36768222938546324,
"grad_norm": 0.018636824563145638,
"learning_rate": 4.976917737041751e-06,
"loss": 0.03673876076936722,
"step": 221
},
{
"epoch": 0.3693459498804201,
"grad_norm": 0.0145517997443676,
"learning_rate": 4.976449810790196e-06,
"loss": 0.037538688629865646,
"step": 222
},
{
"epoch": 0.3710096703753769,
"grad_norm": 0.012170245870947838,
"learning_rate": 4.97597721152683e-06,
"loss": 0.02781190723180771,
"step": 223
},
{
"epoch": 0.3726733908703338,
"grad_norm": 0.01407319400459528,
"learning_rate": 4.975499940143439e-06,
"loss": 0.03755120187997818,
"step": 224
},
{
"epoch": 0.37433711136529063,
"grad_norm": 0.012709653936326504,
"learning_rate": 4.975017997540625e-06,
"loss": 0.04252926632761955,
"step": 225
},
{
"epoch": 0.37600083186024746,
"grad_norm": 0.013191037811338902,
"learning_rate": 4.974531384627805e-06,
"loss": 0.030559688806533813,
"step": 226
},
{
"epoch": 0.37766455235520435,
"grad_norm": 0.018450967967510223,
"learning_rate": 4.974040102323207e-06,
"loss": 0.030250580981373787,
"step": 227
},
{
"epoch": 0.3793282728501612,
"grad_norm": 0.009675616398453712,
"learning_rate": 4.973544151553869e-06,
"loss": 0.022407352924346924,
"step": 228
},
{
"epoch": 0.380991993345118,
"grad_norm": 0.01245384756475687,
"learning_rate": 4.973043533255645e-06,
"loss": 0.027813564985990524,
"step": 229
},
{
"epoch": 0.38265571384007485,
"grad_norm": 0.013367368839681149,
"learning_rate": 4.972538248373188e-06,
"loss": 0.04025476798415184,
"step": 230
},
{
"epoch": 0.38431943433503174,
"grad_norm": 0.011741326190531254,
"learning_rate": 4.9720282978599625e-06,
"loss": 0.030507639050483704,
"step": 231
},
{
"epoch": 0.3859831548299886,
"grad_norm": 0.009616893716156483,
"learning_rate": 4.971513682678234e-06,
"loss": 0.030176818370819092,
"step": 232
},
{
"epoch": 0.3876468753249454,
"grad_norm": 0.009078212082386017,
"learning_rate": 4.970994403799072e-06,
"loss": 0.021030820906162262,
"step": 233
},
{
"epoch": 0.38931059581990224,
"grad_norm": 0.008414865471422672,
"learning_rate": 4.970470462202343e-06,
"loss": 0.022331478074193,
"step": 234
},
{
"epoch": 0.3909743163148591,
"grad_norm": 0.010737700387835503,
"learning_rate": 4.969941858876719e-06,
"loss": 0.026705848053097725,
"step": 235
},
{
"epoch": 0.39263803680981596,
"grad_norm": 0.009779262356460094,
"learning_rate": 4.96940859481966e-06,
"loss": 0.02494858205318451,
"step": 236
},
{
"epoch": 0.3943017573047728,
"grad_norm": 0.012061214074492455,
"learning_rate": 4.968870671037427e-06,
"loss": 0.02954268828034401,
"step": 237
},
{
"epoch": 0.3959654777997296,
"grad_norm": 0.011517210863530636,
"learning_rate": 4.96832808854507e-06,
"loss": 0.025342591106891632,
"step": 238
},
{
"epoch": 0.3976291982946865,
"grad_norm": 0.009201574139297009,
"learning_rate": 4.967780848366432e-06,
"loss": 0.026798926293849945,
"step": 239
},
{
"epoch": 0.39929291878964335,
"grad_norm": 0.015082907862961292,
"learning_rate": 4.967228951534144e-06,
"loss": 0.02584957890212536,
"step": 240
},
{
"epoch": 0.4009566392846002,
"grad_norm": 0.017343197017908096,
"learning_rate": 4.966672399089626e-06,
"loss": 0.030483413487672806,
"step": 241
},
{
"epoch": 0.402620359779557,
"grad_norm": 0.00961192324757576,
"learning_rate": 4.966111192083081e-06,
"loss": 0.029386430978775024,
"step": 242
},
{
"epoch": 0.4042840802745139,
"grad_norm": 0.011751300655305386,
"learning_rate": 4.965545331573493e-06,
"loss": 0.02346094697713852,
"step": 243
},
{
"epoch": 0.40594780076947073,
"grad_norm": 0.013170548714697361,
"learning_rate": 4.964974818628633e-06,
"loss": 0.02498645707964897,
"step": 244
},
{
"epoch": 0.40761152126442757,
"grad_norm": 0.008634310215711594,
"learning_rate": 4.964399654325045e-06,
"loss": 0.02685185708105564,
"step": 245
},
{
"epoch": 0.4092752417593844,
"grad_norm": 0.010694021359086037,
"learning_rate": 4.963819839748055e-06,
"loss": 0.028527498245239258,
"step": 246
},
{
"epoch": 0.4109389622543413,
"grad_norm": 0.01221507042646408,
"learning_rate": 4.96323537599176e-06,
"loss": 0.0299068596214056,
"step": 247
},
{
"epoch": 0.4126026827492981,
"grad_norm": 0.009258517995476723,
"learning_rate": 4.962646264159031e-06,
"loss": 0.026110809296369553,
"step": 248
},
{
"epoch": 0.41426640324425495,
"grad_norm": 0.009516320191323757,
"learning_rate": 4.962052505361512e-06,
"loss": 0.03480090573430061,
"step": 249
},
{
"epoch": 0.4159301237392118,
"grad_norm": 0.009755507111549377,
"learning_rate": 4.9614541007196136e-06,
"loss": 0.02209116891026497,
"step": 250
},
{
"epoch": 0.4175938442341687,
"grad_norm": 0.009852356277406216,
"learning_rate": 4.960851051362514e-06,
"loss": 0.01930748112499714,
"step": 251
},
{
"epoch": 0.4192575647291255,
"grad_norm": 0.011041955091059208,
"learning_rate": 4.960243358428154e-06,
"loss": 0.029225628823041916,
"step": 252
},
{
"epoch": 0.42092128522408234,
"grad_norm": 0.00869595818221569,
"learning_rate": 4.959631023063238e-06,
"loss": 0.016868792474269867,
"step": 253
},
{
"epoch": 0.4225850057190392,
"grad_norm": 0.016593001782894135,
"learning_rate": 4.959014046423233e-06,
"loss": 0.02676609717309475,
"step": 254
},
{
"epoch": 0.42424872621399606,
"grad_norm": 0.00895164255052805,
"learning_rate": 4.9583924296723606e-06,
"loss": 0.016503386199474335,
"step": 255
},
{
"epoch": 0.4259124467089529,
"grad_norm": 0.009723796509206295,
"learning_rate": 4.957766173983598e-06,
"loss": 0.025092676281929016,
"step": 256
},
{
"epoch": 0.42757616720390973,
"grad_norm": 0.011724260635674,
"learning_rate": 4.9571352805386795e-06,
"loss": 0.018208272755146027,
"step": 257
},
{
"epoch": 0.4292398876988666,
"grad_norm": 0.016934333369135857,
"learning_rate": 4.956499750528086e-06,
"loss": 0.04804690182209015,
"step": 258
},
{
"epoch": 0.43090360819382345,
"grad_norm": 0.009145854972302914,
"learning_rate": 4.955859585151054e-06,
"loss": 0.017941124737262726,
"step": 259
},
{
"epoch": 0.4325673286887803,
"grad_norm": 0.010602661408483982,
"learning_rate": 4.955214785615558e-06,
"loss": 0.027120577171444893,
"step": 260
},
{
"epoch": 0.4342310491837371,
"grad_norm": 0.007691461127251387,
"learning_rate": 4.9545653531383255e-06,
"loss": 0.017808586359024048,
"step": 261
},
{
"epoch": 0.435894769678694,
"grad_norm": 0.007216070778667927,
"learning_rate": 4.953911288944821e-06,
"loss": 0.02255088835954666,
"step": 262
},
{
"epoch": 0.43755849017365084,
"grad_norm": 0.010287751443684101,
"learning_rate": 4.953252594269252e-06,
"loss": 0.02571818232536316,
"step": 263
},
{
"epoch": 0.43922221066860767,
"grad_norm": 0.012414435856044292,
"learning_rate": 4.9525892703545604e-06,
"loss": 0.03532583266496658,
"step": 264
},
{
"epoch": 0.4408859311635645,
"grad_norm": 0.007705547381192446,
"learning_rate": 4.951921318452428e-06,
"loss": 0.0289824940264225,
"step": 265
},
{
"epoch": 0.4425496516585214,
"grad_norm": 0.009268342517316341,
"learning_rate": 4.951248739823264e-06,
"loss": 0.020963333547115326,
"step": 266
},
{
"epoch": 0.4442133721534782,
"grad_norm": 0.009462636895477772,
"learning_rate": 4.950571535736214e-06,
"loss": 0.02270900085568428,
"step": 267
},
{
"epoch": 0.44587709264843506,
"grad_norm": 0.009764865040779114,
"learning_rate": 4.949889707469145e-06,
"loss": 0.020391490310430527,
"step": 268
},
{
"epoch": 0.4475408131433919,
"grad_norm": 0.013523961417376995,
"learning_rate": 4.949203256308658e-06,
"loss": 0.021415244787931442,
"step": 269
},
{
"epoch": 0.4492045336383488,
"grad_norm": 0.014478780329227448,
"learning_rate": 4.948512183550068e-06,
"loss": 0.027904432266950607,
"step": 270
},
{
"epoch": 0.4508682541333056,
"grad_norm": 0.0173338670283556,
"learning_rate": 4.947816490497419e-06,
"loss": 0.01066598016768694,
"step": 271
},
{
"epoch": 0.45253197462826245,
"grad_norm": 0.01282422710210085,
"learning_rate": 4.947116178463469e-06,
"loss": 0.027157828211784363,
"step": 272
},
{
"epoch": 0.4541956951232193,
"grad_norm": 0.012170843780040741,
"learning_rate": 4.946411248769693e-06,
"loss": 0.031752899289131165,
"step": 273
},
{
"epoch": 0.45585941561817617,
"grad_norm": 0.011489557102322578,
"learning_rate": 4.945701702746279e-06,
"loss": 0.0212496779859066,
"step": 274
},
{
"epoch": 0.457523136113133,
"grad_norm": 0.017554108053445816,
"learning_rate": 4.944987541732126e-06,
"loss": 0.024693092331290245,
"step": 275
},
{
"epoch": 0.45918685660808983,
"grad_norm": 0.015555243007838726,
"learning_rate": 4.944268767074842e-06,
"loss": 0.020125292241573334,
"step": 276
},
{
"epoch": 0.46085057710304667,
"grad_norm": 0.011039176024496555,
"learning_rate": 4.943545380130742e-06,
"loss": 0.024579813703894615,
"step": 277
},
{
"epoch": 0.46251429759800355,
"grad_norm": 0.011976576410233974,
"learning_rate": 4.942817382264842e-06,
"loss": 0.014248369261622429,
"step": 278
},
{
"epoch": 0.4641780180929604,
"grad_norm": 0.015246872790157795,
"learning_rate": 4.942084774850858e-06,
"loss": 0.027225427329540253,
"step": 279
},
{
"epoch": 0.4658417385879172,
"grad_norm": 0.013577229343354702,
"learning_rate": 4.941347559271208e-06,
"loss": 0.024211157113313675,
"step": 280
},
{
"epoch": 0.46750545908287405,
"grad_norm": 0.013304123654961586,
"learning_rate": 4.9406057369170015e-06,
"loss": 0.020554721355438232,
"step": 281
},
{
"epoch": 0.46916917957783094,
"grad_norm": 0.019589155912399292,
"learning_rate": 4.939859309188044e-06,
"loss": 0.027771135792136192,
"step": 282
},
{
"epoch": 0.4708329000727878,
"grad_norm": 0.01041374821215868,
"learning_rate": 4.939108277492829e-06,
"loss": 0.020846642553806305,
"step": 283
},
{
"epoch": 0.4724966205677446,
"grad_norm": 0.013426681980490685,
"learning_rate": 4.9383526432485375e-06,
"loss": 0.02354995533823967,
"step": 284
},
{
"epoch": 0.47416034106270144,
"grad_norm": 0.012636139057576656,
"learning_rate": 4.937592407881039e-06,
"loss": 0.01683727838099003,
"step": 285
},
{
"epoch": 0.47582406155765833,
"grad_norm": 0.0078094336204230785,
"learning_rate": 4.93682757282488e-06,
"loss": 0.02656428888440132,
"step": 286
},
{
"epoch": 0.47748778205261516,
"grad_norm": 0.014173673465847969,
"learning_rate": 4.936058139523291e-06,
"loss": 0.02244233526289463,
"step": 287
},
{
"epoch": 0.479151502547572,
"grad_norm": 0.013154593296349049,
"learning_rate": 4.935284109428177e-06,
"loss": 0.014531275257468224,
"step": 288
},
{
"epoch": 0.4808152230425288,
"grad_norm": 0.0084519749507308,
"learning_rate": 4.934505484000116e-06,
"loss": 0.013687445782124996,
"step": 289
},
{
"epoch": 0.4824789435374857,
"grad_norm": 0.008050220087170601,
"learning_rate": 4.93372226470836e-06,
"loss": 0.022462567314505577,
"step": 290
},
{
"epoch": 0.48414266403244255,
"grad_norm": 0.008466185070574284,
"learning_rate": 4.932934453030829e-06,
"loss": 0.01815696805715561,
"step": 291
},
{
"epoch": 0.4858063845273994,
"grad_norm": 0.01022345945239067,
"learning_rate": 4.932142050454107e-06,
"loss": 0.0166168250143528,
"step": 292
},
{
"epoch": 0.48747010502235627,
"grad_norm": 0.010604492388665676,
"learning_rate": 4.931345058473443e-06,
"loss": 0.01394139975309372,
"step": 293
},
{
"epoch": 0.4891338255173131,
"grad_norm": 0.007332377601414919,
"learning_rate": 4.930543478592743e-06,
"loss": 0.017569545656442642,
"step": 294
},
{
"epoch": 0.49079754601226994,
"grad_norm": 0.010255846194922924,
"learning_rate": 4.929737312324574e-06,
"loss": 0.019131770357489586,
"step": 295
},
{
"epoch": 0.49246126650722677,
"grad_norm": 0.007818172685801983,
"learning_rate": 4.928926561190155e-06,
"loss": 0.01540689542889595,
"step": 296
},
{
"epoch": 0.49412498700218366,
"grad_norm": 0.01050427183508873,
"learning_rate": 4.928111226719359e-06,
"loss": 0.018683919683098793,
"step": 297
},
{
"epoch": 0.4957887074971405,
"grad_norm": 0.007114852778613567,
"learning_rate": 4.927291310450705e-06,
"loss": 0.01786038652062416,
"step": 298
},
{
"epoch": 0.4974524279920973,
"grad_norm": 0.008571256883442402,
"learning_rate": 4.926466813931358e-06,
"loss": 0.018031764775514603,
"step": 299
},
{
"epoch": 0.49911614848705416,
"grad_norm": 0.009803226217627525,
"learning_rate": 4.925637738717127e-06,
"loss": 0.020492907613515854,
"step": 300
},
{
"epoch": 0.49911614848705416,
"eval_loss": 0.016338517889380455,
"eval_runtime": 403.1488,
"eval_samples_per_second": 11.926,
"eval_steps_per_second": 2.982,
"step": 300
},
{
"epoch": 0.500779868982011,
"grad_norm": 0.005585647653788328,
"learning_rate": 4.924804086372462e-06,
"loss": 0.013384521938860416,
"step": 301
},
{
"epoch": 0.5024435894769679,
"grad_norm": 0.006244071293622255,
"learning_rate": 4.9239658584704466e-06,
"loss": 0.015450171194970608,
"step": 302
},
{
"epoch": 0.5041073099719248,
"grad_norm": 0.006051548756659031,
"learning_rate": 4.923123056592801e-06,
"loss": 0.00981193408370018,
"step": 303
},
{
"epoch": 0.5057710304668815,
"grad_norm": 0.0061712805181741714,
"learning_rate": 4.922275682329876e-06,
"loss": 0.015286393463611603,
"step": 304
},
{
"epoch": 0.5074347509618384,
"grad_norm": 0.009269554167985916,
"learning_rate": 4.921423737280649e-06,
"loss": 0.014119069091975689,
"step": 305
},
{
"epoch": 0.5090984714567952,
"grad_norm": 0.00824077520519495,
"learning_rate": 4.9205672230527254e-06,
"loss": 0.011779182590544224,
"step": 306
},
{
"epoch": 0.5107621919517521,
"grad_norm": 0.007063789293169975,
"learning_rate": 4.919706141262329e-06,
"loss": 0.011469569057226181,
"step": 307
},
{
"epoch": 0.512425912446709,
"grad_norm": 0.009946978650987148,
"learning_rate": 4.918840493534305e-06,
"loss": 0.020134299993515015,
"step": 308
},
{
"epoch": 0.5140896329416658,
"grad_norm": 0.007792671676725149,
"learning_rate": 4.917970281502112e-06,
"loss": 0.014403749257326126,
"step": 309
},
{
"epoch": 0.5157533534366227,
"grad_norm": 0.010617299936711788,
"learning_rate": 4.917095506807824e-06,
"loss": 0.015932071954011917,
"step": 310
},
{
"epoch": 0.5174170739315795,
"grad_norm": 0.008703233674168587,
"learning_rate": 4.916216171102124e-06,
"loss": 0.022535301744937897,
"step": 311
},
{
"epoch": 0.5190807944265363,
"grad_norm": 0.006629075389355421,
"learning_rate": 4.9153322760443015e-06,
"loss": 0.013775508850812912,
"step": 312
},
{
"epoch": 0.5207445149214932,
"grad_norm": 0.007406014949083328,
"learning_rate": 4.914443823302246e-06,
"loss": 0.01799451746046543,
"step": 313
},
{
"epoch": 0.52240823541645,
"grad_norm": 0.00845777615904808,
"learning_rate": 4.913550814552454e-06,
"loss": 0.018990565091371536,
"step": 314
},
{
"epoch": 0.5240719559114069,
"grad_norm": 0.007644719909876585,
"learning_rate": 4.912653251480013e-06,
"loss": 0.016380393877625465,
"step": 315
},
{
"epoch": 0.5257356764063638,
"grad_norm": 0.005281711462885141,
"learning_rate": 4.9117511357786075e-06,
"loss": 0.011785021051764488,
"step": 316
},
{
"epoch": 0.5273993969013205,
"grad_norm": 0.004629249218851328,
"learning_rate": 4.910844469150512e-06,
"loss": 0.014201750978827477,
"step": 317
},
{
"epoch": 0.5290631173962774,
"grad_norm": 0.006540986709296703,
"learning_rate": 4.909933253306588e-06,
"loss": 0.016089923679828644,
"step": 318
},
{
"epoch": 0.5307268378912343,
"grad_norm": 0.007417282089591026,
"learning_rate": 4.909017489966283e-06,
"loss": 0.015274600125849247,
"step": 319
},
{
"epoch": 0.5323905583861911,
"grad_norm": 0.006533796899020672,
"learning_rate": 4.9080971808576226e-06,
"loss": 0.014022988267242908,
"step": 320
},
{
"epoch": 0.534054278881148,
"grad_norm": 0.005975611042231321,
"learning_rate": 4.907172327717214e-06,
"loss": 0.009007789194583893,
"step": 321
},
{
"epoch": 0.5357179993761049,
"grad_norm": 0.0060996185056865215,
"learning_rate": 4.906242932290234e-06,
"loss": 0.013682247139513493,
"step": 322
},
{
"epoch": 0.5373817198710616,
"grad_norm": 0.004868973977863789,
"learning_rate": 4.905308996330437e-06,
"loss": 0.013650172390043736,
"step": 323
},
{
"epoch": 0.5390454403660185,
"grad_norm": 0.005156499799340963,
"learning_rate": 4.904370521600138e-06,
"loss": 0.008532417006790638,
"step": 324
},
{
"epoch": 0.5407091608609753,
"grad_norm": 0.007405860815197229,
"learning_rate": 4.903427509870222e-06,
"loss": 0.01490192674100399,
"step": 325
},
{
"epoch": 0.5423728813559322,
"grad_norm": 0.009611138142645359,
"learning_rate": 4.902479962920134e-06,
"loss": 0.012371081858873367,
"step": 326
},
{
"epoch": 0.5440366018508891,
"grad_norm": 0.011518461629748344,
"learning_rate": 4.901527882537876e-06,
"loss": 0.014121288433670998,
"step": 327
},
{
"epoch": 0.5457003223458459,
"grad_norm": 0.006337135564535856,
"learning_rate": 4.900571270520004e-06,
"loss": 0.01352706365287304,
"step": 328
},
{
"epoch": 0.5473640428408028,
"grad_norm": 0.006071313749998808,
"learning_rate": 4.899610128671626e-06,
"loss": 0.012974245473742485,
"step": 329
},
{
"epoch": 0.5490277633357596,
"grad_norm": 0.007717214059084654,
"learning_rate": 4.898644458806398e-06,
"loss": 0.010514363646507263,
"step": 330
},
{
"epoch": 0.5506914838307164,
"grad_norm": 0.011666782200336456,
"learning_rate": 4.897674262746522e-06,
"loss": 0.01974363438785076,
"step": 331
},
{
"epoch": 0.5523552043256733,
"grad_norm": 0.006378008518368006,
"learning_rate": 4.896699542322736e-06,
"loss": 0.01286529004573822,
"step": 332
},
{
"epoch": 0.5540189248206301,
"grad_norm": 0.018436580896377563,
"learning_rate": 4.895720299374319e-06,
"loss": 0.011777098290622234,
"step": 333
},
{
"epoch": 0.555682645315587,
"grad_norm": 0.008302860893309116,
"learning_rate": 4.894736535749083e-06,
"loss": 0.014747174456715584,
"step": 334
},
{
"epoch": 0.5573463658105439,
"grad_norm": 0.02435437962412834,
"learning_rate": 4.89374825330337e-06,
"loss": 0.02497844025492668,
"step": 335
},
{
"epoch": 0.5590100863055006,
"grad_norm": 0.007081752642989159,
"learning_rate": 4.892755453902051e-06,
"loss": 0.016742464154958725,
"step": 336
},
{
"epoch": 0.5606738068004575,
"grad_norm": 0.007361928932368755,
"learning_rate": 4.8917581394185175e-06,
"loss": 0.013661217875778675,
"step": 337
},
{
"epoch": 0.5623375272954144,
"grad_norm": 0.008742809295654297,
"learning_rate": 4.890756311734683e-06,
"loss": 0.014807759784162045,
"step": 338
},
{
"epoch": 0.5640012477903712,
"grad_norm": 0.0050587523728609085,
"learning_rate": 4.8897499727409755e-06,
"loss": 0.013927068561315536,
"step": 339
},
{
"epoch": 0.5656649682853281,
"grad_norm": 0.008532224223017693,
"learning_rate": 4.888739124336338e-06,
"loss": 0.012824540957808495,
"step": 340
},
{
"epoch": 0.5673286887802849,
"grad_norm": 0.014145714230835438,
"learning_rate": 4.8877237684282205e-06,
"loss": 0.019548412412405014,
"step": 341
},
{
"epoch": 0.5689924092752418,
"grad_norm": 0.00613539619371295,
"learning_rate": 4.8867039069325804e-06,
"loss": 0.011379221454262733,
"step": 342
},
{
"epoch": 0.5706561297701986,
"grad_norm": 0.006859153043478727,
"learning_rate": 4.8856795417738754e-06,
"loss": 0.013419722206890583,
"step": 343
},
{
"epoch": 0.5723198502651554,
"grad_norm": 0.006972996052354574,
"learning_rate": 4.884650674885062e-06,
"loss": 0.017972735688090324,
"step": 344
},
{
"epoch": 0.5739835707601123,
"grad_norm": 0.007068189792335033,
"learning_rate": 4.883617308207592e-06,
"loss": 0.013157499022781849,
"step": 345
},
{
"epoch": 0.5756472912550692,
"grad_norm": 0.00740961916744709,
"learning_rate": 4.88257944369141e-06,
"loss": 0.011037548072636127,
"step": 346
},
{
"epoch": 0.577311011750026,
"grad_norm": 0.008075796999037266,
"learning_rate": 4.8815370832949425e-06,
"loss": 0.01143650058656931,
"step": 347
},
{
"epoch": 0.5789747322449829,
"grad_norm": 0.005925778299570084,
"learning_rate": 4.880490228985104e-06,
"loss": 0.014801234006881714,
"step": 348
},
{
"epoch": 0.5806384527399396,
"grad_norm": 0.005958295427262783,
"learning_rate": 4.8794388827372884e-06,
"loss": 0.018268218263983727,
"step": 349
},
{
"epoch": 0.5823021732348965,
"grad_norm": 0.005769642069935799,
"learning_rate": 4.878383046535366e-06,
"loss": 0.015677250921726227,
"step": 350
},
{
"epoch": 0.5839658937298534,
"grad_norm": 0.003173237666487694,
"learning_rate": 4.877322722371677e-06,
"loss": 0.007178253494203091,
"step": 351
},
{
"epoch": 0.5856296142248102,
"grad_norm": 0.005019443575292826,
"learning_rate": 4.876257912247033e-06,
"loss": 0.009780826047062874,
"step": 352
},
{
"epoch": 0.5872933347197671,
"grad_norm": 0.0037745118606835604,
"learning_rate": 4.8751886181707105e-06,
"loss": 0.010863647796213627,
"step": 353
},
{
"epoch": 0.588957055214724,
"grad_norm": 0.0066299536265432835,
"learning_rate": 4.874114842160445e-06,
"loss": 0.014221194200217724,
"step": 354
},
{
"epoch": 0.5906207757096807,
"grad_norm": 0.005365578457713127,
"learning_rate": 4.873036586242431e-06,
"loss": 0.018708698451519012,
"step": 355
},
{
"epoch": 0.5922844962046376,
"grad_norm": 0.00685572624206543,
"learning_rate": 4.871953852451316e-06,
"loss": 0.014886384829878807,
"step": 356
},
{
"epoch": 0.5939482166995945,
"grad_norm": 0.006036871578544378,
"learning_rate": 4.8708666428301975e-06,
"loss": 0.015632428228855133,
"step": 357
},
{
"epoch": 0.5956119371945513,
"grad_norm": 0.005431063938885927,
"learning_rate": 4.869774959430619e-06,
"loss": 0.010408837348222733,
"step": 358
},
{
"epoch": 0.5972756576895082,
"grad_norm": 0.0059727029874920845,
"learning_rate": 4.868678804312565e-06,
"loss": 0.018925290554761887,
"step": 359
},
{
"epoch": 0.598939378184465,
"grad_norm": 0.004846646916121244,
"learning_rate": 4.867578179544457e-06,
"loss": 0.010519372299313545,
"step": 360
},
{
"epoch": 0.6006030986794219,
"grad_norm": 0.005863612052053213,
"learning_rate": 4.866473087203154e-06,
"loss": 0.018922332674264908,
"step": 361
},
{
"epoch": 0.6022668191743787,
"grad_norm": 0.004887004382908344,
"learning_rate": 4.865363529373944e-06,
"loss": 0.008048690855503082,
"step": 362
},
{
"epoch": 0.6039305396693355,
"grad_norm": 0.003912640269845724,
"learning_rate": 4.864249508150539e-06,
"loss": 0.008222410455346107,
"step": 363
},
{
"epoch": 0.6055942601642924,
"grad_norm": 0.005772759206593037,
"learning_rate": 4.863131025635076e-06,
"loss": 0.01382957212626934,
"step": 364
},
{
"epoch": 0.6072579806592493,
"grad_norm": 0.006792543921619654,
"learning_rate": 4.862008083938109e-06,
"loss": 0.01627694070339203,
"step": 365
},
{
"epoch": 0.6089217011542061,
"grad_norm": 0.004902017302811146,
"learning_rate": 4.8608806851786075e-06,
"loss": 0.009956824593245983,
"step": 366
},
{
"epoch": 0.610585421649163,
"grad_norm": 0.006868287455290556,
"learning_rate": 4.859748831483949e-06,
"loss": 0.008180022239685059,
"step": 367
},
{
"epoch": 0.6122491421441197,
"grad_norm": 0.008069746196269989,
"learning_rate": 4.858612524989921e-06,
"loss": 0.01568032242357731,
"step": 368
},
{
"epoch": 0.6139128626390766,
"grad_norm": 0.005045538302510977,
"learning_rate": 4.857471767840709e-06,
"loss": 0.008955922909080982,
"step": 369
},
{
"epoch": 0.6155765831340335,
"grad_norm": 0.005875969305634499,
"learning_rate": 4.856326562188902e-06,
"loss": 0.01971004158258438,
"step": 370
},
{
"epoch": 0.6172403036289903,
"grad_norm": 0.005288930144160986,
"learning_rate": 4.855176910195479e-06,
"loss": 0.010098019614815712,
"step": 371
},
{
"epoch": 0.6189040241239472,
"grad_norm": 0.00646026199683547,
"learning_rate": 4.854022814029809e-06,
"loss": 0.014291519299149513,
"step": 372
},
{
"epoch": 0.6205677446189041,
"grad_norm": 0.004560984671115875,
"learning_rate": 4.852864275869652e-06,
"loss": 0.012375302612781525,
"step": 373
},
{
"epoch": 0.6222314651138608,
"grad_norm": 0.007243362721055746,
"learning_rate": 4.851701297901144e-06,
"loss": 0.012917522341012955,
"step": 374
},
{
"epoch": 0.6238951856088177,
"grad_norm": 0.006706541404128075,
"learning_rate": 4.850533882318803e-06,
"loss": 0.011306913569569588,
"step": 375
},
{
"epoch": 0.6255589061037745,
"grad_norm": 0.005922200623899698,
"learning_rate": 4.849362031325518e-06,
"loss": 0.009204890578985214,
"step": 376
},
{
"epoch": 0.6272226265987314,
"grad_norm": 0.005413861479610205,
"learning_rate": 4.8481857471325485e-06,
"loss": 0.010358748957514763,
"step": 377
},
{
"epoch": 0.6288863470936883,
"grad_norm": 0.005771263036876917,
"learning_rate": 4.847005031959521e-06,
"loss": 0.007761461194604635,
"step": 378
},
{
"epoch": 0.6305500675886451,
"grad_norm": 0.007504676003009081,
"learning_rate": 4.84581988803442e-06,
"loss": 0.0132828364148736,
"step": 379
},
{
"epoch": 0.632213788083602,
"grad_norm": 0.007642901036888361,
"learning_rate": 4.84463031759359e-06,
"loss": 0.008316827937960625,
"step": 380
},
{
"epoch": 0.6338775085785588,
"grad_norm": 0.004544087685644627,
"learning_rate": 4.843436322881725e-06,
"loss": 0.00889534130692482,
"step": 381
},
{
"epoch": 0.6355412290735156,
"grad_norm": 0.007763517554849386,
"learning_rate": 4.8422379061518705e-06,
"loss": 0.010410662740468979,
"step": 382
},
{
"epoch": 0.6372049495684725,
"grad_norm": 0.004702461417764425,
"learning_rate": 4.841035069665416e-06,
"loss": 0.012421442195773125,
"step": 383
},
{
"epoch": 0.6388686700634293,
"grad_norm": 0.004728985484689474,
"learning_rate": 4.83982781569209e-06,
"loss": 0.009557276964187622,
"step": 384
},
{
"epoch": 0.6405323905583862,
"grad_norm": 0.006002399139106274,
"learning_rate": 4.838616146509956e-06,
"loss": 0.008976309560239315,
"step": 385
},
{
"epoch": 0.6421961110533431,
"grad_norm": 0.0073576439172029495,
"learning_rate": 4.83740006440541e-06,
"loss": 0.01807243376970291,
"step": 386
},
{
"epoch": 0.6438598315482998,
"grad_norm": 0.004097748547792435,
"learning_rate": 4.8361795716731744e-06,
"loss": 0.009704215452075005,
"step": 387
},
{
"epoch": 0.6455235520432567,
"grad_norm": 0.0038165440782904625,
"learning_rate": 4.8349546706162965e-06,
"loss": 0.0072257500141859055,
"step": 388
},
{
"epoch": 0.6471872725382136,
"grad_norm": 0.007018840406090021,
"learning_rate": 4.833725363546139e-06,
"loss": 0.008374448865652084,
"step": 389
},
{
"epoch": 0.6488509930331704,
"grad_norm": 0.006301086861640215,
"learning_rate": 4.8324916527823795e-06,
"loss": 0.022184694185853004,
"step": 390
},
{
"epoch": 0.6505147135281273,
"grad_norm": 0.004370077047497034,
"learning_rate": 4.831253540653007e-06,
"loss": 0.01324211061000824,
"step": 391
},
{
"epoch": 0.6521784340230842,
"grad_norm": 0.0047884974628686905,
"learning_rate": 4.8300110294943145e-06,
"loss": 0.011861605569720268,
"step": 392
},
{
"epoch": 0.653842154518041,
"grad_norm": 0.0038174018263816833,
"learning_rate": 4.828764121650896e-06,
"loss": 0.006326655857264996,
"step": 393
},
{
"epoch": 0.6555058750129978,
"grad_norm": 0.004340061917901039,
"learning_rate": 4.827512819475641e-06,
"loss": 0.012556311674416065,
"step": 394
},
{
"epoch": 0.6571695955079546,
"grad_norm": 0.0038324350025504827,
"learning_rate": 4.826257125329733e-06,
"loss": 0.007697253488004208,
"step": 395
},
{
"epoch": 0.6588333160029115,
"grad_norm": 0.006832764483988285,
"learning_rate": 4.824997041582641e-06,
"loss": 0.01178714819252491,
"step": 396
},
{
"epoch": 0.6604970364978684,
"grad_norm": 0.004772037733346224,
"learning_rate": 4.82373257061212e-06,
"loss": 0.012505128979682922,
"step": 397
},
{
"epoch": 0.6621607569928252,
"grad_norm": 0.0030513901729136705,
"learning_rate": 4.8224637148042e-06,
"loss": 0.007846292108297348,
"step": 398
},
{
"epoch": 0.6638244774877821,
"grad_norm": 0.0031088057439774275,
"learning_rate": 4.821190476553186e-06,
"loss": 0.004913134500384331,
"step": 399
},
{
"epoch": 0.665488197982739,
"grad_norm": 0.0029906572308391333,
"learning_rate": 4.819912858261656e-06,
"loss": 0.008030945435166359,
"step": 400
},
{
"epoch": 0.665488197982739,
"eval_loss": 0.008412709459662437,
"eval_runtime": 407.1614,
"eval_samples_per_second": 11.809,
"eval_steps_per_second": 2.952,
"step": 400
},
{
"epoch": 0.6671519184776957,
"grad_norm": 0.003743760986253619,
"learning_rate": 4.818630862340449e-06,
"loss": 0.009701243601739407,
"step": 401
},
{
"epoch": 0.6688156389726526,
"grad_norm": 0.003921910189092159,
"learning_rate": 4.817344491208665e-06,
"loss": 0.006662596948444843,
"step": 402
},
{
"epoch": 0.6704793594676094,
"grad_norm": 0.00469661969691515,
"learning_rate": 4.816053747293663e-06,
"loss": 0.008967725560069084,
"step": 403
},
{
"epoch": 0.6721430799625663,
"grad_norm": 0.00589489983394742,
"learning_rate": 4.814758633031049e-06,
"loss": 0.013237478211522102,
"step": 404
},
{
"epoch": 0.6738068004575232,
"grad_norm": 0.00632744375616312,
"learning_rate": 4.813459150864681e-06,
"loss": 0.0064522032625973225,
"step": 405
},
{
"epoch": 0.67547052095248,
"grad_norm": 0.004063498694449663,
"learning_rate": 4.812155303246653e-06,
"loss": 0.0071722520515322685,
"step": 406
},
{
"epoch": 0.6771342414474368,
"grad_norm": 0.005490950774401426,
"learning_rate": 4.810847092637301e-06,
"loss": 0.00855968426913023,
"step": 407
},
{
"epoch": 0.6787979619423937,
"grad_norm": 0.004029491916298866,
"learning_rate": 4.809534521505192e-06,
"loss": 0.0076716807670891285,
"step": 408
},
{
"epoch": 0.6804616824373505,
"grad_norm": 0.004423109348863363,
"learning_rate": 4.8082175923271235e-06,
"loss": 0.009134914726018906,
"step": 409
},
{
"epoch": 0.6821254029323074,
"grad_norm": 0.003607613267377019,
"learning_rate": 4.806896307588113e-06,
"loss": 0.007768969517201185,
"step": 410
},
{
"epoch": 0.6837891234272642,
"grad_norm": 0.0037080564070492983,
"learning_rate": 4.805570669781399e-06,
"loss": 0.008052971214056015,
"step": 411
},
{
"epoch": 0.6854528439222211,
"grad_norm": 0.006682890933007002,
"learning_rate": 4.804240681408434e-06,
"loss": 0.011350871995091438,
"step": 412
},
{
"epoch": 0.6871165644171779,
"grad_norm": 0.00619282154366374,
"learning_rate": 4.802906344978881e-06,
"loss": 0.008059765212237835,
"step": 413
},
{
"epoch": 0.6887802849121347,
"grad_norm": 0.004879550542682409,
"learning_rate": 4.801567663010605e-06,
"loss": 0.011704351752996445,
"step": 414
},
{
"epoch": 0.6904440054070916,
"grad_norm": 0.0034507052041590214,
"learning_rate": 4.800224638029672e-06,
"loss": 0.006550830323249102,
"step": 415
},
{
"epoch": 0.6921077259020485,
"grad_norm": 0.0060535939410328865,
"learning_rate": 4.798877272570343e-06,
"loss": 0.012915357947349548,
"step": 416
},
{
"epoch": 0.6937714463970053,
"grad_norm": 0.0054816617630422115,
"learning_rate": 4.797525569175073e-06,
"loss": 0.008053340017795563,
"step": 417
},
{
"epoch": 0.6954351668919622,
"grad_norm": 0.0040940698236227036,
"learning_rate": 4.796169530394498e-06,
"loss": 0.008299745619297028,
"step": 418
},
{
"epoch": 0.697098887386919,
"grad_norm": 0.003937297500669956,
"learning_rate": 4.7948091587874355e-06,
"loss": 0.006818284280598164,
"step": 419
},
{
"epoch": 0.6987626078818758,
"grad_norm": 0.0057077184319496155,
"learning_rate": 4.793444456920881e-06,
"loss": 0.008323611691594124,
"step": 420
},
{
"epoch": 0.7004263283768327,
"grad_norm": 0.006937435362488031,
"learning_rate": 4.7920754273699985e-06,
"loss": 0.011208346113562584,
"step": 421
},
{
"epoch": 0.7020900488717895,
"grad_norm": 0.00426529860123992,
"learning_rate": 4.790702072718121e-06,
"loss": 0.006764519028365612,
"step": 422
},
{
"epoch": 0.7037537693667464,
"grad_norm": 0.005305912345647812,
"learning_rate": 4.789324395556741e-06,
"loss": 0.014592036604881287,
"step": 423
},
{
"epoch": 0.7054174898617033,
"grad_norm": 0.005696433130651712,
"learning_rate": 4.7879423984855085e-06,
"loss": 0.007089572958648205,
"step": 424
},
{
"epoch": 0.70708121035666,
"grad_norm": 0.005689716432243586,
"learning_rate": 4.786556084112224e-06,
"loss": 0.008991958573460579,
"step": 425
},
{
"epoch": 0.7087449308516169,
"grad_norm": 0.005220255348831415,
"learning_rate": 4.785165455052836e-06,
"loss": 0.007952387444674969,
"step": 426
},
{
"epoch": 0.7104086513465738,
"grad_norm": 0.0035842740908265114,
"learning_rate": 4.783770513931433e-06,
"loss": 0.005702752619981766,
"step": 427
},
{
"epoch": 0.7120723718415306,
"grad_norm": 0.003953936044126749,
"learning_rate": 4.782371263380242e-06,
"loss": 0.008399353362619877,
"step": 428
},
{
"epoch": 0.7137360923364875,
"grad_norm": 0.004692696966230869,
"learning_rate": 4.780967706039622e-06,
"loss": 0.0066901338286697865,
"step": 429
},
{
"epoch": 0.7153998128314443,
"grad_norm": 0.0040604835376143456,
"learning_rate": 4.779559844558056e-06,
"loss": 0.005349840968847275,
"step": 430
},
{
"epoch": 0.7170635333264012,
"grad_norm": 0.005523476283997297,
"learning_rate": 4.778147681592152e-06,
"loss": 0.008249819278717041,
"step": 431
},
{
"epoch": 0.718727253821358,
"grad_norm": 0.004006467759609222,
"learning_rate": 4.776731219806634e-06,
"loss": 0.006997613701969385,
"step": 432
},
{
"epoch": 0.7203909743163148,
"grad_norm": 0.0044905198737978935,
"learning_rate": 4.775310461874337e-06,
"loss": 0.004558880813419819,
"step": 433
},
{
"epoch": 0.7220546948112717,
"grad_norm": 0.004894842393696308,
"learning_rate": 4.773885410476202e-06,
"loss": 0.0108189033344388,
"step": 434
},
{
"epoch": 0.7237184153062286,
"grad_norm": 0.01109437644481659,
"learning_rate": 4.7724560683012735e-06,
"loss": 0.010264476761221886,
"step": 435
},
{
"epoch": 0.7253821358011854,
"grad_norm": 0.005047748796641827,
"learning_rate": 4.771022438046693e-06,
"loss": 0.010578740388154984,
"step": 436
},
{
"epoch": 0.7270458562961423,
"grad_norm": 0.0065911151468753815,
"learning_rate": 4.769584522417691e-06,
"loss": 0.011462481692433357,
"step": 437
},
{
"epoch": 0.728709576791099,
"grad_norm": 0.008467431180179119,
"learning_rate": 4.768142324127586e-06,
"loss": 0.012880068272352219,
"step": 438
},
{
"epoch": 0.7303732972860559,
"grad_norm": 0.004129709675908089,
"learning_rate": 4.766695845897778e-06,
"loss": 0.004490282386541367,
"step": 439
},
{
"epoch": 0.7320370177810128,
"grad_norm": 0.00397163350135088,
"learning_rate": 4.765245090457744e-06,
"loss": 0.006365049630403519,
"step": 440
},
{
"epoch": 0.7337007382759696,
"grad_norm": 0.009726262651383877,
"learning_rate": 4.763790060545028e-06,
"loss": 0.005823507439345121,
"step": 441
},
{
"epoch": 0.7353644587709265,
"grad_norm": 0.00505091343075037,
"learning_rate": 4.762330758905246e-06,
"loss": 0.01049719750881195,
"step": 442
},
{
"epoch": 0.7370281792658834,
"grad_norm": 0.005302008707076311,
"learning_rate": 4.760867188292068e-06,
"loss": 0.012403653934597969,
"step": 443
},
{
"epoch": 0.7386918997608402,
"grad_norm": 0.002863701432943344,
"learning_rate": 4.7593993514672255e-06,
"loss": 0.004971321672201157,
"step": 444
},
{
"epoch": 0.740355620255797,
"grad_norm": 0.0034518903121352196,
"learning_rate": 4.757927251200497e-06,
"loss": 0.00744717987254262,
"step": 445
},
{
"epoch": 0.7420193407507538,
"grad_norm": 0.009052886627614498,
"learning_rate": 4.756450890269705e-06,
"loss": 0.010446527972817421,
"step": 446
},
{
"epoch": 0.7436830612457107,
"grad_norm": 0.005404800642281771,
"learning_rate": 4.754970271460714e-06,
"loss": 0.007915949448943138,
"step": 447
},
{
"epoch": 0.7453467817406676,
"grad_norm": 0.005113918334245682,
"learning_rate": 4.753485397567424e-06,
"loss": 0.011439811438322067,
"step": 448
},
{
"epoch": 0.7470105022356244,
"grad_norm": 0.013223345391452312,
"learning_rate": 4.751996271391761e-06,
"loss": 0.008794520981609821,
"step": 449
},
{
"epoch": 0.7486742227305813,
"grad_norm": 0.0056835380382835865,
"learning_rate": 4.750502895743677e-06,
"loss": 0.010445352643728256,
"step": 450
},
{
"epoch": 0.7503379432255382,
"grad_norm": 0.0067290510050952435,
"learning_rate": 4.749005273441143e-06,
"loss": 0.015891049057245255,
"step": 451
},
{
"epoch": 0.7520016637204949,
"grad_norm": 0.00862638559192419,
"learning_rate": 4.747503407310142e-06,
"loss": 0.014896569773554802,
"step": 452
},
{
"epoch": 0.7536653842154518,
"grad_norm": 0.004024629946798086,
"learning_rate": 4.745997300184666e-06,
"loss": 0.0071207070723176,
"step": 453
},
{
"epoch": 0.7553291047104087,
"grad_norm": 0.00700328778475523,
"learning_rate": 4.744486954906709e-06,
"loss": 0.014192566275596619,
"step": 454
},
{
"epoch": 0.7569928252053655,
"grad_norm": 0.006798639427870512,
"learning_rate": 4.742972374326262e-06,
"loss": 0.011122412979602814,
"step": 455
},
{
"epoch": 0.7586565457003224,
"grad_norm": 0.0045930189080536366,
"learning_rate": 4.74145356130131e-06,
"loss": 0.009697200730443,
"step": 456
},
{
"epoch": 0.7603202661952791,
"grad_norm": 0.0071418690495193005,
"learning_rate": 4.739930518697823e-06,
"loss": 0.012642932124435902,
"step": 457
},
{
"epoch": 0.761983986690236,
"grad_norm": 0.00647162739187479,
"learning_rate": 4.738403249389752e-06,
"loss": 0.013247480615973473,
"step": 458
},
{
"epoch": 0.7636477071851929,
"grad_norm": 0.004375535994768143,
"learning_rate": 4.736871756259023e-06,
"loss": 0.0069526853039860725,
"step": 459
},
{
"epoch": 0.7653114276801497,
"grad_norm": 0.005763936322182417,
"learning_rate": 4.7353360421955345e-06,
"loss": 0.01148962415754795,
"step": 460
},
{
"epoch": 0.7669751481751066,
"grad_norm": 0.004523094277828932,
"learning_rate": 4.733796110097148e-06,
"loss": 0.011522913351655006,
"step": 461
},
{
"epoch": 0.7686388686700635,
"grad_norm": 0.004391341470181942,
"learning_rate": 4.732251962869685e-06,
"loss": 0.00787794217467308,
"step": 462
},
{
"epoch": 0.7703025891650203,
"grad_norm": 0.00304810656234622,
"learning_rate": 4.730703603426921e-06,
"loss": 0.005599565804004669,
"step": 463
},
{
"epoch": 0.7719663096599771,
"grad_norm": 0.00340471975505352,
"learning_rate": 4.729151034690579e-06,
"loss": 0.004527213517576456,
"step": 464
},
{
"epoch": 0.7736300301549339,
"grad_norm": 0.004050911404192448,
"learning_rate": 4.727594259590326e-06,
"loss": 0.0071553378365933895,
"step": 465
},
{
"epoch": 0.7752937506498908,
"grad_norm": 0.0033549219369888306,
"learning_rate": 4.726033281063766e-06,
"loss": 0.0059393346309661865,
"step": 466
},
{
"epoch": 0.7769574711448477,
"grad_norm": 0.004678431898355484,
"learning_rate": 4.724468102056434e-06,
"loss": 0.007732603698968887,
"step": 467
},
{
"epoch": 0.7786211916398045,
"grad_norm": 0.002846264746040106,
"learning_rate": 4.722898725521793e-06,
"loss": 0.00735393725335598,
"step": 468
},
{
"epoch": 0.7802849121347614,
"grad_norm": 0.005869272630661726,
"learning_rate": 4.721325154421224e-06,
"loss": 0.011904164217412472,
"step": 469
},
{
"epoch": 0.7819486326297183,
"grad_norm": 0.004066616762429476,
"learning_rate": 4.7197473917240255e-06,
"loss": 0.005801680032163858,
"step": 470
},
{
"epoch": 0.783612353124675,
"grad_norm": 0.0035934511106461287,
"learning_rate": 4.718165440407404e-06,
"loss": 0.005442335270345211,
"step": 471
},
{
"epoch": 0.7852760736196319,
"grad_norm": 0.0047324420884251595,
"learning_rate": 4.716579303456471e-06,
"loss": 0.00940224714577198,
"step": 472
},
{
"epoch": 0.7869397941145887,
"grad_norm": 0.003724937792867422,
"learning_rate": 4.714988983864235e-06,
"loss": 0.009013586677610874,
"step": 473
},
{
"epoch": 0.7886035146095456,
"grad_norm": 0.004245929419994354,
"learning_rate": 4.713394484631598e-06,
"loss": 0.006094337906688452,
"step": 474
},
{
"epoch": 0.7902672351045025,
"grad_norm": 0.0049941991455852985,
"learning_rate": 4.711795808767348e-06,
"loss": 0.011073598638176918,
"step": 475
},
{
"epoch": 0.7919309555994593,
"grad_norm": 0.0026603529695421457,
"learning_rate": 4.7101929592881545e-06,
"loss": 0.0036363271065056324,
"step": 476
},
{
"epoch": 0.7935946760944161,
"grad_norm": 0.004774894565343857,
"learning_rate": 4.708585939218564e-06,
"loss": 0.01305093988776207,
"step": 477
},
{
"epoch": 0.795258396589373,
"grad_norm": 0.00325093069113791,
"learning_rate": 4.7069747515909905e-06,
"loss": 0.009489022195339203,
"step": 478
},
{
"epoch": 0.7969221170843298,
"grad_norm": 0.006839904002845287,
"learning_rate": 4.7053593994457135e-06,
"loss": 0.012426530010998249,
"step": 479
},
{
"epoch": 0.7985858375792867,
"grad_norm": 0.002349321963265538,
"learning_rate": 4.70373988583087e-06,
"loss": 0.004337068181484938,
"step": 480
},
{
"epoch": 0.8002495580742435,
"grad_norm": 0.0030962666496634483,
"learning_rate": 4.7021162138024524e-06,
"loss": 0.008469605818390846,
"step": 481
},
{
"epoch": 0.8019132785692004,
"grad_norm": 0.003444101894274354,
"learning_rate": 4.700488386424294e-06,
"loss": 0.0059173512272536755,
"step": 482
},
{
"epoch": 0.8035769990641572,
"grad_norm": 0.006231680046766996,
"learning_rate": 4.698856406768076e-06,
"loss": 0.009944312274456024,
"step": 483
},
{
"epoch": 0.805240719559114,
"grad_norm": 0.005668159108608961,
"learning_rate": 4.697220277913311e-06,
"loss": 0.009258732199668884,
"step": 484
},
{
"epoch": 0.8069044400540709,
"grad_norm": 0.005199704319238663,
"learning_rate": 4.695580002947341e-06,
"loss": 0.00513799674808979,
"step": 485
},
{
"epoch": 0.8085681605490278,
"grad_norm": 0.0023570070043206215,
"learning_rate": 4.6939355849653325e-06,
"loss": 0.005083505064249039,
"step": 486
},
{
"epoch": 0.8102318810439846,
"grad_norm": 0.0046600759960711,
"learning_rate": 4.69228702707027e-06,
"loss": 0.005510847084224224,
"step": 487
},
{
"epoch": 0.8118956015389415,
"grad_norm": 0.0029210986103862524,
"learning_rate": 4.69063433237295e-06,
"loss": 0.008221244439482689,
"step": 488
},
{
"epoch": 0.8135593220338984,
"grad_norm": 0.0037554206792265177,
"learning_rate": 4.688977503991975e-06,
"loss": 0.008256680332124233,
"step": 489
},
{
"epoch": 0.8152230425288551,
"grad_norm": 0.003074623178690672,
"learning_rate": 4.687316545053746e-06,
"loss": 0.006913225632160902,
"step": 490
},
{
"epoch": 0.816886763023812,
"grad_norm": 0.002423194469884038,
"learning_rate": 4.68565145869246e-06,
"loss": 0.0056283967569470406,
"step": 491
},
{
"epoch": 0.8185504835187688,
"grad_norm": 0.0032063298858702183,
"learning_rate": 4.683982248050103e-06,
"loss": 0.007283125072717667,
"step": 492
},
{
"epoch": 0.8202142040137257,
"grad_norm": 0.0034963127691298723,
"learning_rate": 4.6823089162764425e-06,
"loss": 0.009975203312933445,
"step": 493
},
{
"epoch": 0.8218779245086826,
"grad_norm": 0.0035684818867594004,
"learning_rate": 4.6806314665290205e-06,
"loss": 0.006531362421810627,
"step": 494
},
{
"epoch": 0.8235416450036394,
"grad_norm": 0.002319071441888809,
"learning_rate": 4.678949901973154e-06,
"loss": 0.006020743865519762,
"step": 495
},
{
"epoch": 0.8252053654985962,
"grad_norm": 0.0032957608345896006,
"learning_rate": 4.677264225781921e-06,
"loss": 0.006968480534851551,
"step": 496
},
{
"epoch": 0.8268690859935531,
"grad_norm": 0.006399734411388636,
"learning_rate": 4.6755744411361585e-06,
"loss": 0.005256010685116053,
"step": 497
},
{
"epoch": 0.8285328064885099,
"grad_norm": 0.002357139950618148,
"learning_rate": 4.6738805512244575e-06,
"loss": 0.004715315066277981,
"step": 498
},
{
"epoch": 0.8301965269834668,
"grad_norm": 0.002225536620244384,
"learning_rate": 4.672182559243155e-06,
"loss": 0.003946832846850157,
"step": 499
},
{
"epoch": 0.8318602474784236,
"grad_norm": 0.007506450638175011,
"learning_rate": 4.670480468396327e-06,
"loss": 0.007973461411893368,
"step": 500
},
{
"epoch": 0.8318602474784236,
"eval_loss": 0.006512382533401251,
"eval_runtime": 404.3517,
"eval_samples_per_second": 11.891,
"eval_steps_per_second": 2.973,
"step": 500
},
{
"epoch": 0.8335239679733805,
"grad_norm": 0.004953257739543915,
"learning_rate": 4.668774281895786e-06,
"loss": 0.007463683374226093,
"step": 501
},
{
"epoch": 0.8351876884683374,
"grad_norm": 0.002295607700943947,
"learning_rate": 4.667064002961073e-06,
"loss": 0.0029762398917227983,
"step": 502
},
{
"epoch": 0.8368514089632941,
"grad_norm": 0.005000600591301918,
"learning_rate": 4.66534963481945e-06,
"loss": 0.00649390509352088,
"step": 503
},
{
"epoch": 0.838515129458251,
"grad_norm": 0.003147828159853816,
"learning_rate": 4.663631180705894e-06,
"loss": 0.007842538878321648,
"step": 504
},
{
"epoch": 0.8401788499532079,
"grad_norm": 0.003913369495421648,
"learning_rate": 4.661908643863096e-06,
"loss": 0.005828930996358395,
"step": 505
},
{
"epoch": 0.8418425704481647,
"grad_norm": 0.004644430708140135,
"learning_rate": 4.66018202754145e-06,
"loss": 0.008422689512372017,
"step": 506
},
{
"epoch": 0.8435062909431216,
"grad_norm": 0.0037597171030938625,
"learning_rate": 4.658451334999043e-06,
"loss": 0.0072099887765944,
"step": 507
},
{
"epoch": 0.8451700114380784,
"grad_norm": 0.007217934355139732,
"learning_rate": 4.656716569501661e-06,
"loss": 0.005425385665148497,
"step": 508
},
{
"epoch": 0.8468337319330352,
"grad_norm": 0.004139721859246492,
"learning_rate": 4.654977734322772e-06,
"loss": 0.005886915139853954,
"step": 509
},
{
"epoch": 0.8484974524279921,
"grad_norm": 0.003015086753293872,
"learning_rate": 4.653234832743521e-06,
"loss": 0.008951975964009762,
"step": 510
},
{
"epoch": 0.8501611729229489,
"grad_norm": 0.0028621021192520857,
"learning_rate": 4.651487868052731e-06,
"loss": 0.0064186276867985725,
"step": 511
},
{
"epoch": 0.8518248934179058,
"grad_norm": 0.0029435402248054743,
"learning_rate": 4.64973684354689e-06,
"loss": 0.005053696688264608,
"step": 512
},
{
"epoch": 0.8534886139128627,
"grad_norm": 0.00349103263579309,
"learning_rate": 4.647981762530145e-06,
"loss": 0.005872088484466076,
"step": 513
},
{
"epoch": 0.8551523344078195,
"grad_norm": 0.005280966870486736,
"learning_rate": 4.6462226283143e-06,
"loss": 0.005002675112336874,
"step": 514
},
{
"epoch": 0.8568160549027763,
"grad_norm": 0.006165878847241402,
"learning_rate": 4.644459444218807e-06,
"loss": 0.00517700519412756,
"step": 515
},
{
"epoch": 0.8584797753977332,
"grad_norm": 0.003584573045372963,
"learning_rate": 4.642692213570759e-06,
"loss": 0.011866070330142975,
"step": 516
},
{
"epoch": 0.86014349589269,
"grad_norm": 0.003298037452623248,
"learning_rate": 4.640920939704885e-06,
"loss": 0.005188156384974718,
"step": 517
},
{
"epoch": 0.8618072163876469,
"grad_norm": 0.00396351283416152,
"learning_rate": 4.639145625963544e-06,
"loss": 0.008152484893798828,
"step": 518
},
{
"epoch": 0.8634709368826037,
"grad_norm": 0.0071251713670790195,
"learning_rate": 4.637366275696718e-06,
"loss": 0.006925216875970364,
"step": 519
},
{
"epoch": 0.8651346573775606,
"grad_norm": 0.006336830090731382,
"learning_rate": 4.635582892262006e-06,
"loss": 0.0077257608063519,
"step": 520
},
{
"epoch": 0.8667983778725175,
"grad_norm": 0.004709393717348576,
"learning_rate": 4.633795479024616e-06,
"loss": 0.007717994041740894,
"step": 521
},
{
"epoch": 0.8684620983674742,
"grad_norm": 0.003954755142331123,
"learning_rate": 4.632004039357364e-06,
"loss": 0.0050251539796590805,
"step": 522
},
{
"epoch": 0.8701258188624311,
"grad_norm": 0.00402754545211792,
"learning_rate": 4.630208576640659e-06,
"loss": 0.00838618166744709,
"step": 523
},
{
"epoch": 0.871789539357388,
"grad_norm": 0.003631069790571928,
"learning_rate": 4.628409094262504e-06,
"loss": 0.006958806421607733,
"step": 524
},
{
"epoch": 0.8734532598523448,
"grad_norm": 0.0026326472871005535,
"learning_rate": 4.6266055956184865e-06,
"loss": 0.005749491509050131,
"step": 525
},
{
"epoch": 0.8751169803473017,
"grad_norm": 0.0023095738142728806,
"learning_rate": 4.624798084111773e-06,
"loss": 0.004408619366586208,
"step": 526
},
{
"epoch": 0.8767807008422585,
"grad_norm": 0.003138443687930703,
"learning_rate": 4.622986563153104e-06,
"loss": 0.0046043386682868,
"step": 527
},
{
"epoch": 0.8784444213372153,
"grad_norm": 0.0023057162761688232,
"learning_rate": 4.621171036160781e-06,
"loss": 0.00428372249007225,
"step": 528
},
{
"epoch": 0.8801081418321722,
"grad_norm": 0.004754107911139727,
"learning_rate": 4.6193515065606675e-06,
"loss": 0.01612052507698536,
"step": 529
},
{
"epoch": 0.881771862327129,
"grad_norm": 0.004191739484667778,
"learning_rate": 4.617527977786182e-06,
"loss": 0.007633868604898453,
"step": 530
},
{
"epoch": 0.8834355828220859,
"grad_norm": 0.003158772364258766,
"learning_rate": 4.615700453278285e-06,
"loss": 0.004591359756886959,
"step": 531
},
{
"epoch": 0.8850993033170428,
"grad_norm": 0.0042033726349473,
"learning_rate": 4.61386893648548e-06,
"loss": 0.005673332139849663,
"step": 532
},
{
"epoch": 0.8867630238119996,
"grad_norm": 0.003734700381755829,
"learning_rate": 4.612033430863804e-06,
"loss": 0.005909692961722612,
"step": 533
},
{
"epoch": 0.8884267443069565,
"grad_norm": 0.0027026941534131765,
"learning_rate": 4.610193939876818e-06,
"loss": 0.006266783457249403,
"step": 534
},
{
"epoch": 0.8900904648019132,
"grad_norm": 0.00583456689491868,
"learning_rate": 4.608350466995606e-06,
"loss": 0.010947637259960175,
"step": 535
},
{
"epoch": 0.8917541852968701,
"grad_norm": 0.0036986898630857468,
"learning_rate": 4.606503015698765e-06,
"loss": 0.00387162109836936,
"step": 536
},
{
"epoch": 0.893417905791827,
"grad_norm": 0.004644804168492556,
"learning_rate": 4.6046515894723985e-06,
"loss": 0.004419853910803795,
"step": 537
},
{
"epoch": 0.8950816262867838,
"grad_norm": 0.005302789155393839,
"learning_rate": 4.602796191810113e-06,
"loss": 0.00521225668489933,
"step": 538
},
{
"epoch": 0.8967453467817407,
"grad_norm": 0.004432398360222578,
"learning_rate": 4.600936826213004e-06,
"loss": 0.0068048653192818165,
"step": 539
},
{
"epoch": 0.8984090672766976,
"grad_norm": 0.0038318608421832323,
"learning_rate": 4.59907349618966e-06,
"loss": 0.006457334849983454,
"step": 540
},
{
"epoch": 0.9000727877716543,
"grad_norm": 0.004747864790260792,
"learning_rate": 4.597206205256147e-06,
"loss": 0.00862425658851862,
"step": 541
},
{
"epoch": 0.9017365082666112,
"grad_norm": 0.002571320626884699,
"learning_rate": 4.595334956936007e-06,
"loss": 0.008329037576913834,
"step": 542
},
{
"epoch": 0.903400228761568,
"grad_norm": 0.004095368552953005,
"learning_rate": 4.593459754760248e-06,
"loss": 0.006220923736691475,
"step": 543
},
{
"epoch": 0.9050639492565249,
"grad_norm": 0.004221968352794647,
"learning_rate": 4.591580602267338e-06,
"loss": 0.006353544071316719,
"step": 544
},
{
"epoch": 0.9067276697514818,
"grad_norm": 0.006588945630937815,
"learning_rate": 4.589697503003203e-06,
"loss": 0.005587139166891575,
"step": 545
},
{
"epoch": 0.9083913902464386,
"grad_norm": 0.002770567312836647,
"learning_rate": 4.587810460521213e-06,
"loss": 0.006217245943844318,
"step": 546
},
{
"epoch": 0.9100551107413954,
"grad_norm": 0.0035460866056382656,
"learning_rate": 4.585919478382178e-06,
"loss": 0.005783476866781712,
"step": 547
},
{
"epoch": 0.9117188312363523,
"grad_norm": 0.0028954341541975737,
"learning_rate": 4.584024560154348e-06,
"loss": 0.005333957262337208,
"step": 548
},
{
"epoch": 0.9133825517313091,
"grad_norm": 0.0033749546855688095,
"learning_rate": 4.582125709413392e-06,
"loss": 0.004037967883050442,
"step": 549
},
{
"epoch": 0.915046272226266,
"grad_norm": 0.004284190014004707,
"learning_rate": 4.580222929742407e-06,
"loss": 0.0068781087175011635,
"step": 550
},
{
"epoch": 0.9167099927212229,
"grad_norm": 0.004720574710518122,
"learning_rate": 4.5783162247318986e-06,
"loss": 0.005199882667511702,
"step": 551
},
{
"epoch": 0.9183737132161797,
"grad_norm": 0.002596945036202669,
"learning_rate": 4.576405597979782e-06,
"loss": 0.005764994770288467,
"step": 552
},
{
"epoch": 0.9200374337111366,
"grad_norm": 0.0030985120683908463,
"learning_rate": 4.5744910530913725e-06,
"loss": 0.004004698246717453,
"step": 553
},
{
"epoch": 0.9217011542060933,
"grad_norm": 0.0047891330905258656,
"learning_rate": 4.572572593679379e-06,
"loss": 0.009581279940903187,
"step": 554
},
{
"epoch": 0.9233648747010502,
"grad_norm": 0.002359537873417139,
"learning_rate": 4.5706502233638935e-06,
"loss": 0.0038718758150935173,
"step": 555
},
{
"epoch": 0.9250285951960071,
"grad_norm": 0.0029691008385270834,
"learning_rate": 4.568723945772394e-06,
"loss": 0.0046098604798316956,
"step": 556
},
{
"epoch": 0.9266923156909639,
"grad_norm": 0.003485879860818386,
"learning_rate": 4.5667937645397276e-06,
"loss": 0.004569823853671551,
"step": 557
},
{
"epoch": 0.9283560361859208,
"grad_norm": 0.010340786539018154,
"learning_rate": 4.564859683308107e-06,
"loss": 0.007273489609360695,
"step": 558
},
{
"epoch": 0.9300197566808777,
"grad_norm": 0.003216391894966364,
"learning_rate": 4.562921705727106e-06,
"loss": 0.005173889454454184,
"step": 559
},
{
"epoch": 0.9316834771758344,
"grad_norm": 0.0023886016570031643,
"learning_rate": 4.5609798354536495e-06,
"loss": 0.0035293418914079666,
"step": 560
},
{
"epoch": 0.9333471976707913,
"grad_norm": 0.0051674325950443745,
"learning_rate": 4.559034076152009e-06,
"loss": 0.007310433778911829,
"step": 561
},
{
"epoch": 0.9350109181657481,
"grad_norm": 0.0061208331026136875,
"learning_rate": 4.557084431493793e-06,
"loss": 0.01674576662480831,
"step": 562
},
{
"epoch": 0.936674638660705,
"grad_norm": 0.0033059692941606045,
"learning_rate": 4.555130905157943e-06,
"loss": 0.00993238016963005,
"step": 563
},
{
"epoch": 0.9383383591556619,
"grad_norm": 0.002794879488646984,
"learning_rate": 4.553173500830724e-06,
"loss": 0.004958215169608593,
"step": 564
},
{
"epoch": 0.9400020796506187,
"grad_norm": 0.006112914066761732,
"learning_rate": 4.55121222220572e-06,
"loss": 0.009362846612930298,
"step": 565
},
{
"epoch": 0.9416658001455755,
"grad_norm": 0.0030292437877506018,
"learning_rate": 4.549247072983825e-06,
"loss": 0.00661174301058054,
"step": 566
},
{
"epoch": 0.9433295206405324,
"grad_norm": 0.002375996205955744,
"learning_rate": 4.5472780568732356e-06,
"loss": 0.0036187791265547276,
"step": 567
},
{
"epoch": 0.9449932411354892,
"grad_norm": 0.0034390664659440517,
"learning_rate": 4.545305177589448e-06,
"loss": 0.0048618377186357975,
"step": 568
},
{
"epoch": 0.9466569616304461,
"grad_norm": 0.003132295561954379,
"learning_rate": 4.5433284388552435e-06,
"loss": 0.004123467952013016,
"step": 569
},
{
"epoch": 0.9483206821254029,
"grad_norm": 0.002180474577471614,
"learning_rate": 4.541347844400692e-06,
"loss": 0.004508224315941334,
"step": 570
},
{
"epoch": 0.9499844026203598,
"grad_norm": 0.0022491244599223137,
"learning_rate": 4.539363397963134e-06,
"loss": 0.004375042859464884,
"step": 571
},
{
"epoch": 0.9516481231153167,
"grad_norm": 0.001735589117743075,
"learning_rate": 4.537375103287183e-06,
"loss": 0.002892045769840479,
"step": 572
},
{
"epoch": 0.9533118436102734,
"grad_norm": 0.0044294800609350204,
"learning_rate": 4.53538296412471e-06,
"loss": 0.006532138213515282,
"step": 573
},
{
"epoch": 0.9549755641052303,
"grad_norm": 0.007447395473718643,
"learning_rate": 4.533386984234841e-06,
"loss": 0.012636320665478706,
"step": 574
},
{
"epoch": 0.9566392846001872,
"grad_norm": 0.0024636040907353163,
"learning_rate": 4.5313871673839525e-06,
"loss": 0.004774153232574463,
"step": 575
},
{
"epoch": 0.958303005095144,
"grad_norm": 0.004148229956626892,
"learning_rate": 4.52938351734566e-06,
"loss": 0.007745377719402313,
"step": 576
},
{
"epoch": 0.9599667255901009,
"grad_norm": 0.0033451179042458534,
"learning_rate": 4.52737603790081e-06,
"loss": 0.00580889917910099,
"step": 577
},
{
"epoch": 0.9616304460850577,
"grad_norm": 0.0032389156986027956,
"learning_rate": 4.525364732837476e-06,
"loss": 0.008575042709708214,
"step": 578
},
{
"epoch": 0.9632941665800145,
"grad_norm": 0.0024506989866495132,
"learning_rate": 4.523349605950953e-06,
"loss": 0.00456388620659709,
"step": 579
},
{
"epoch": 0.9649578870749714,
"grad_norm": 0.002531093545258045,
"learning_rate": 4.521330661043744e-06,
"loss": 0.004104273393750191,
"step": 580
},
{
"epoch": 0.9666216075699282,
"grad_norm": 0.00328632234595716,
"learning_rate": 4.519307901925558e-06,
"loss": 0.007781069725751877,
"step": 581
},
{
"epoch": 0.9682853280648851,
"grad_norm": 0.0028874180279672146,
"learning_rate": 4.517281332413302e-06,
"loss": 0.004761362448334694,
"step": 582
},
{
"epoch": 0.969949048559842,
"grad_norm": 0.0018867182079702616,
"learning_rate": 4.515250956331072e-06,
"loss": 0.00552754569798708,
"step": 583
},
{
"epoch": 0.9716127690547988,
"grad_norm": 0.002540376503020525,
"learning_rate": 4.513216777510149e-06,
"loss": 0.004502263385802507,
"step": 584
},
{
"epoch": 0.9732764895497557,
"grad_norm": 0.0025226708967238665,
"learning_rate": 4.511178799788987e-06,
"loss": 0.004058694466948509,
"step": 585
},
{
"epoch": 0.9749402100447125,
"grad_norm": 0.0028099576011300087,
"learning_rate": 4.50913702701321e-06,
"loss": 0.004276437684893608,
"step": 586
},
{
"epoch": 0.9766039305396693,
"grad_norm": 0.0035055428743362427,
"learning_rate": 4.507091463035601e-06,
"loss": 0.006080637685954571,
"step": 587
},
{
"epoch": 0.9782676510346262,
"grad_norm": 0.0025878252927213907,
"learning_rate": 4.505042111716103e-06,
"loss": 0.0048270029947161674,
"step": 588
},
{
"epoch": 0.979931371529583,
"grad_norm": 0.0019265188602730632,
"learning_rate": 4.502988976921797e-06,
"loss": 0.0017514314968138933,
"step": 589
},
{
"epoch": 0.9815950920245399,
"grad_norm": 0.004005370195955038,
"learning_rate": 4.50093206252691e-06,
"loss": 0.0036932427901774645,
"step": 590
},
{
"epoch": 0.9832588125194968,
"grad_norm": 0.004234121646732092,
"learning_rate": 4.498871372412798e-06,
"loss": 0.005153659265488386,
"step": 591
},
{
"epoch": 0.9849225330144535,
"grad_norm": 0.0056281927973032,
"learning_rate": 4.496806910467944e-06,
"loss": 0.006571756675839424,
"step": 592
},
{
"epoch": 0.9865862535094104,
"grad_norm": 0.0052912007085978985,
"learning_rate": 4.494738680587946e-06,
"loss": 0.012203991413116455,
"step": 593
},
{
"epoch": 0.9882499740043673,
"grad_norm": 0.008162816055119038,
"learning_rate": 4.492666686675511e-06,
"loss": 0.0037072275299578905,
"step": 594
},
{
"epoch": 0.9899136944993241,
"grad_norm": 0.002978774020448327,
"learning_rate": 4.490590932640453e-06,
"loss": 0.004548577591776848,
"step": 595
},
{
"epoch": 0.991577414994281,
"grad_norm": 0.005547270644456148,
"learning_rate": 4.488511422399677e-06,
"loss": 0.0054444968700408936,
"step": 596
},
{
"epoch": 0.9932411354892378,
"grad_norm": 0.006614347919821739,
"learning_rate": 4.48642815987718e-06,
"loss": 0.011927146464586258,
"step": 597
},
{
"epoch": 0.9949048559841946,
"grad_norm": 0.0036901719868183136,
"learning_rate": 4.484341149004035e-06,
"loss": 0.00477896211668849,
"step": 598
},
{
"epoch": 0.9965685764791515,
"grad_norm": 0.004520885646343231,
"learning_rate": 4.482250393718392e-06,
"loss": 0.007755149155855179,
"step": 599
},
{
"epoch": 0.9982322969741083,
"grad_norm": 0.0036596583668142557,
"learning_rate": 4.480155897965463e-06,
"loss": 0.007180670741945505,
"step": 600
},
{
"epoch": 0.9982322969741083,
"eval_loss": 0.007219767663627863,
"eval_runtime": 401.9917,
"eval_samples_per_second": 11.96,
"eval_steps_per_second": 2.99,
"step": 600
},
{
"epoch": 0.9998960174690652,
"grad_norm": 0.0034411894157528877,
"learning_rate": 4.47805766569752e-06,
"loss": 0.007506140973418951,
"step": 601
},
{
"epoch": 1.0,
"grad_norm": 0.04566219821572304,
"learning_rate": 4.475955700873888e-06,
"loss": 0.020605536177754402,
"step": 602
},
{
"epoch": 1.0016637204949568,
"grad_norm": 0.0035529271699488163,
"learning_rate": 4.473850007460932e-06,
"loss": 0.00510389544069767,
"step": 603
},
{
"epoch": 1.0033274409899138,
"grad_norm": 0.008444433100521564,
"learning_rate": 4.471740589432053e-06,
"loss": 0.013971086591482162,
"step": 604
},
{
"epoch": 1.0049911614848706,
"grad_norm": 0.016102664172649384,
"learning_rate": 4.469627450767682e-06,
"loss": 0.04315765202045441,
"step": 605
},
{
"epoch": 1.0066548819798273,
"grad_norm": 0.009336776100099087,
"learning_rate": 4.46751059545527e-06,
"loss": 0.018110543489456177,
"step": 606
},
{
"epoch": 1.0083186024747843,
"grad_norm": 0.003961662296205759,
"learning_rate": 4.465390027489279e-06,
"loss": 0.009482689201831818,
"step": 607
},
{
"epoch": 1.009982322969741,
"grad_norm": 0.0029724633786827326,
"learning_rate": 4.463265750871182e-06,
"loss": 0.004566639196127653,
"step": 608
},
{
"epoch": 1.0116460434646979,
"grad_norm": 0.010566702112555504,
"learning_rate": 4.461137769609445e-06,
"loss": 0.01625455915927887,
"step": 609
},
{
"epoch": 1.0133097639596549,
"grad_norm": 0.0051901922561228275,
"learning_rate": 4.459006087719527e-06,
"loss": 0.008888859301805496,
"step": 610
},
{
"epoch": 1.0149734844546117,
"grad_norm": 0.004209975246340036,
"learning_rate": 4.45687070922387e-06,
"loss": 0.005752148572355509,
"step": 611
},
{
"epoch": 1.0166372049495684,
"grad_norm": 0.006711136549711227,
"learning_rate": 4.4547316381518905e-06,
"loss": 0.004719719290733337,
"step": 612
},
{
"epoch": 1.0183009254445252,
"grad_norm": 0.0029179907869547606,
"learning_rate": 4.4525888785399725e-06,
"loss": 0.006970624905079603,
"step": 613
},
{
"epoch": 1.0199646459394822,
"grad_norm": 0.002345362678170204,
"learning_rate": 4.450442434431463e-06,
"loss": 0.0034522705245763063,
"step": 614
},
{
"epoch": 1.021628366434439,
"grad_norm": 0.004603294190019369,
"learning_rate": 4.448292309876657e-06,
"loss": 0.004663745407015085,
"step": 615
},
{
"epoch": 1.0232920869293958,
"grad_norm": 0.0023862654343247414,
"learning_rate": 4.4461385089328e-06,
"loss": 0.003139633685350418,
"step": 616
},
{
"epoch": 1.0249558074243528,
"grad_norm": 0.0021421839483082294,
"learning_rate": 4.44398103566407e-06,
"loss": 0.004743538796901703,
"step": 617
},
{
"epoch": 1.0266195279193095,
"grad_norm": 0.0013927406398579478,
"learning_rate": 4.4418198941415756e-06,
"loss": 0.0038331456016749144,
"step": 618
},
{
"epoch": 1.0282832484142663,
"grad_norm": 0.0032657769042998552,
"learning_rate": 4.4396550884433495e-06,
"loss": 0.0025713355280458927,
"step": 619
},
{
"epoch": 1.0299469689092233,
"grad_norm": 0.0034895415883511305,
"learning_rate": 4.437486622654337e-06,
"loss": 0.005330443382263184,
"step": 620
},
{
"epoch": 1.03161068940418,
"grad_norm": 0.014472505077719688,
"learning_rate": 4.43531450086639e-06,
"loss": 0.007480642758309841,
"step": 621
},
{
"epoch": 1.0332744098991369,
"grad_norm": 0.0033677637111395597,
"learning_rate": 4.433138727178259e-06,
"loss": 0.004082255996763706,
"step": 622
},
{
"epoch": 1.0349381303940939,
"grad_norm": 0.0026485221460461617,
"learning_rate": 4.4309593056955865e-06,
"loss": 0.0040733059868216515,
"step": 623
},
{
"epoch": 1.0366018508890507,
"grad_norm": 0.0036252597346901894,
"learning_rate": 4.4287762405308974e-06,
"loss": 0.0023388115223497152,
"step": 624
},
{
"epoch": 1.0382655713840074,
"grad_norm": 0.003635956672951579,
"learning_rate": 4.426589535803593e-06,
"loss": 0.00605718232691288,
"step": 625
},
{
"epoch": 1.0399292918789644,
"grad_norm": 0.007902245968580246,
"learning_rate": 4.424399195639941e-06,
"loss": 0.009553415700793266,
"step": 626
},
{
"epoch": 1.0415930123739212,
"grad_norm": 0.003865851555019617,
"learning_rate": 4.422205224173071e-06,
"loss": 0.008454103022813797,
"step": 627
},
{
"epoch": 1.043256732868878,
"grad_norm": 0.003319124225527048,
"learning_rate": 4.420007625542963e-06,
"loss": 0.0036944844760000706,
"step": 628
},
{
"epoch": 1.0449204533638348,
"grad_norm": 0.0050729489885270596,
"learning_rate": 4.417806403896442e-06,
"loss": 0.005002877209335566,
"step": 629
},
{
"epoch": 1.0465841738587918,
"grad_norm": 0.0038563492707908154,
"learning_rate": 4.41560156338717e-06,
"loss": 0.007914157584309578,
"step": 630
},
{
"epoch": 1.0482478943537485,
"grad_norm": 0.004178495611995459,
"learning_rate": 4.413393108175637e-06,
"loss": 0.004781308118253946,
"step": 631
},
{
"epoch": 1.0499116148487053,
"grad_norm": 0.004806080367416143,
"learning_rate": 4.411181042429156e-06,
"loss": 0.004015027079731226,
"step": 632
},
{
"epoch": 1.0515753353436623,
"grad_norm": 0.0037014191038906574,
"learning_rate": 4.40896537032185e-06,
"loss": 0.004629138857126236,
"step": 633
},
{
"epoch": 1.053239055838619,
"grad_norm": 0.003243440529331565,
"learning_rate": 4.406746096034647e-06,
"loss": 0.005871012806892395,
"step": 634
},
{
"epoch": 1.0549027763335759,
"grad_norm": 0.0017317542806267738,
"learning_rate": 4.4045232237552756e-06,
"loss": 0.003214429132640362,
"step": 635
},
{
"epoch": 1.0565664968285329,
"grad_norm": 0.00374778569675982,
"learning_rate": 4.4022967576782525e-06,
"loss": 0.004213203210383654,
"step": 636
},
{
"epoch": 1.0582302173234897,
"grad_norm": 0.0024602594785392284,
"learning_rate": 4.400066702004874e-06,
"loss": 0.007073926739394665,
"step": 637
},
{
"epoch": 1.0598939378184464,
"grad_norm": 0.002045772271230817,
"learning_rate": 4.39783306094321e-06,
"loss": 0.0070329755544662476,
"step": 638
},
{
"epoch": 1.0615576583134034,
"grad_norm": 0.0015325051499530673,
"learning_rate": 4.395595838708099e-06,
"loss": 0.0028078206814825535,
"step": 639
},
{
"epoch": 1.0632213788083602,
"grad_norm": 0.002612901385873556,
"learning_rate": 4.393355039521134e-06,
"loss": 0.004282036330550909,
"step": 640
},
{
"epoch": 1.064885099303317,
"grad_norm": 0.0035253753885626793,
"learning_rate": 4.391110667610658e-06,
"loss": 0.004646215122193098,
"step": 641
},
{
"epoch": 1.066548819798274,
"grad_norm": 0.0026419060304760933,
"learning_rate": 4.388862727211759e-06,
"loss": 0.006227807141840458,
"step": 642
},
{
"epoch": 1.0682125402932308,
"grad_norm": 0.0018523207399994135,
"learning_rate": 4.386611222566254e-06,
"loss": 0.00431039510294795,
"step": 643
},
{
"epoch": 1.0698762607881875,
"grad_norm": 0.002442911732941866,
"learning_rate": 4.384356157922688e-06,
"loss": 0.003634359687566757,
"step": 644
},
{
"epoch": 1.0715399812831445,
"grad_norm": 0.0021309482399374247,
"learning_rate": 4.382097537536322e-06,
"loss": 0.005421295762062073,
"step": 645
},
{
"epoch": 1.0732037017781013,
"grad_norm": 0.0033649380784481764,
"learning_rate": 4.379835365669132e-06,
"loss": 0.009508009068667889,
"step": 646
},
{
"epoch": 1.074867422273058,
"grad_norm": 0.0021837574895471334,
"learning_rate": 4.377569646589789e-06,
"loss": 0.004345447290688753,
"step": 647
},
{
"epoch": 1.0765311427680149,
"grad_norm": 0.0032484072726219893,
"learning_rate": 4.375300384573659e-06,
"loss": 0.0040743970312178135,
"step": 648
},
{
"epoch": 1.0781948632629719,
"grad_norm": 0.0024877728428691626,
"learning_rate": 4.373027583902796e-06,
"loss": 0.006332829128950834,
"step": 649
},
{
"epoch": 1.0798585837579286,
"grad_norm": 0.0017824380192905664,
"learning_rate": 4.370751248865929e-06,
"loss": 0.0035579982213675976,
"step": 650
},
{
"epoch": 1.0815223042528854,
"grad_norm": 0.0031106146052479744,
"learning_rate": 4.368471383758459e-06,
"loss": 0.007911598309874535,
"step": 651
},
{
"epoch": 1.0831860247478424,
"grad_norm": 0.0014041169779375196,
"learning_rate": 4.366187992882444e-06,
"loss": 0.003576268907636404,
"step": 652
},
{
"epoch": 1.0848497452427992,
"grad_norm": 0.0017640722217038274,
"learning_rate": 4.3639010805466e-06,
"loss": 0.004032103344798088,
"step": 653
},
{
"epoch": 1.086513465737756,
"grad_norm": 0.0037411150988191366,
"learning_rate": 4.361610651066283e-06,
"loss": 0.004132252652198076,
"step": 654
},
{
"epoch": 1.088177186232713,
"grad_norm": 0.0016895364969968796,
"learning_rate": 4.35931670876349e-06,
"loss": 0.0023081060498952866,
"step": 655
},
{
"epoch": 1.0898409067276698,
"grad_norm": 0.00220140116289258,
"learning_rate": 4.357019257966844e-06,
"loss": 0.005424368195235729,
"step": 656
},
{
"epoch": 1.0915046272226265,
"grad_norm": 0.0015891756629571319,
"learning_rate": 4.354718303011588e-06,
"loss": 0.004366081207990646,
"step": 657
},
{
"epoch": 1.0931683477175835,
"grad_norm": 0.0026955781504511833,
"learning_rate": 4.352413848239579e-06,
"loss": 0.0043494547717273235,
"step": 658
},
{
"epoch": 1.0948320682125403,
"grad_norm": 0.0034271839540451765,
"learning_rate": 4.35010589799928e-06,
"loss": 0.0057737985625863075,
"step": 659
},
{
"epoch": 1.096495788707497,
"grad_norm": 0.002068024128675461,
"learning_rate": 4.347794456645744e-06,
"loss": 0.005561623256653547,
"step": 660
},
{
"epoch": 1.098159509202454,
"grad_norm": 0.0031483988277614117,
"learning_rate": 4.345479528540618e-06,
"loss": 0.003334908513352275,
"step": 661
},
{
"epoch": 1.0998232296974109,
"grad_norm": 0.001954516861587763,
"learning_rate": 4.343161118052123e-06,
"loss": 0.003126378171145916,
"step": 662
},
{
"epoch": 1.1014869501923676,
"grad_norm": 0.002192142652347684,
"learning_rate": 4.340839229555056e-06,
"loss": 0.0036504999734461308,
"step": 663
},
{
"epoch": 1.1031506706873246,
"grad_norm": 0.0024770835880190134,
"learning_rate": 4.338513867430773e-06,
"loss": 0.005833543371409178,
"step": 664
},
{
"epoch": 1.1048143911822814,
"grad_norm": 0.0008247373043559492,
"learning_rate": 4.336185036067187e-06,
"loss": 0.0029306081123650074,
"step": 665
},
{
"epoch": 1.1064781116772382,
"grad_norm": 0.002471225569024682,
"learning_rate": 4.3338527398587575e-06,
"loss": 0.0034327320754528046,
"step": 666
},
{
"epoch": 1.108141832172195,
"grad_norm": 0.0027939577121287584,
"learning_rate": 4.33151698320648e-06,
"loss": 0.0042039137333631516,
"step": 667
},
{
"epoch": 1.109805552667152,
"grad_norm": 0.003027573460713029,
"learning_rate": 4.329177770517881e-06,
"loss": 0.005466432776302099,
"step": 668
},
{
"epoch": 1.1114692731621088,
"grad_norm": 0.0014210582012310624,
"learning_rate": 4.32683510620701e-06,
"loss": 0.0019816220737993717,
"step": 669
},
{
"epoch": 1.1131329936570655,
"grad_norm": 0.0018567409133538604,
"learning_rate": 4.324488994694427e-06,
"loss": 0.0032230927608907223,
"step": 670
},
{
"epoch": 1.1147967141520225,
"grad_norm": 0.0019692997448146343,
"learning_rate": 4.322139440407198e-06,
"loss": 0.0035078530199825764,
"step": 671
},
{
"epoch": 1.1164604346469793,
"grad_norm": 0.002447731327265501,
"learning_rate": 4.319786447778887e-06,
"loss": 0.00550186587497592,
"step": 672
},
{
"epoch": 1.118124155141936,
"grad_norm": 0.004330574534833431,
"learning_rate": 4.317430021249543e-06,
"loss": 0.004071483854204416,
"step": 673
},
{
"epoch": 1.119787875636893,
"grad_norm": 0.0016852137632668018,
"learning_rate": 4.315070165265695e-06,
"loss": 0.0034770139027386904,
"step": 674
},
{
"epoch": 1.1214515961318499,
"grad_norm": 0.003290498396381736,
"learning_rate": 4.312706884280349e-06,
"loss": 0.004757840186357498,
"step": 675
},
{
"epoch": 1.1231153166268066,
"grad_norm": 0.0012338663218542933,
"learning_rate": 4.310340182752965e-06,
"loss": 0.0025315401144325733,
"step": 676
},
{
"epoch": 1.1247790371217636,
"grad_norm": 0.0045959739945828915,
"learning_rate": 4.307970065149464e-06,
"loss": 0.006023851688951254,
"step": 677
},
{
"epoch": 1.1264427576167204,
"grad_norm": 0.004816236440092325,
"learning_rate": 4.305596535942211e-06,
"loss": 0.004382929764688015,
"step": 678
},
{
"epoch": 1.1281064781116772,
"grad_norm": 0.002231198363006115,
"learning_rate": 4.303219599610009e-06,
"loss": 0.004849039483815432,
"step": 679
},
{
"epoch": 1.129770198606634,
"grad_norm": 0.004572253208607435,
"learning_rate": 4.300839260638089e-06,
"loss": 0.004676291719079018,
"step": 680
},
{
"epoch": 1.131433919101591,
"grad_norm": 0.0013777704443782568,
"learning_rate": 4.298455523518102e-06,
"loss": 0.0025581195950508118,
"step": 681
},
{
"epoch": 1.1330976395965477,
"grad_norm": 0.0034841045271605253,
"learning_rate": 4.296068392748116e-06,
"loss": 0.006778066046535969,
"step": 682
},
{
"epoch": 1.1347613600915047,
"grad_norm": 0.0038167042657732964,
"learning_rate": 4.293677872832599e-06,
"loss": 0.005039317067712545,
"step": 683
},
{
"epoch": 1.1364250805864615,
"grad_norm": 0.002462555654346943,
"learning_rate": 4.291283968282413e-06,
"loss": 0.004252985585480928,
"step": 684
},
{
"epoch": 1.1380888010814183,
"grad_norm": 0.002770144259557128,
"learning_rate": 4.288886683614809e-06,
"loss": 0.005728703923523426,
"step": 685
},
{
"epoch": 1.139752521576375,
"grad_norm": 0.00589687330648303,
"learning_rate": 4.286486023353417e-06,
"loss": 0.004119477234780788,
"step": 686
},
{
"epoch": 1.141416242071332,
"grad_norm": 0.0018810235196724534,
"learning_rate": 4.284081992028235e-06,
"loss": 0.0024635386653244495,
"step": 687
},
{
"epoch": 1.1430799625662889,
"grad_norm": 0.0035444023087620735,
"learning_rate": 4.281674594175621e-06,
"loss": 0.005115187726914883,
"step": 688
},
{
"epoch": 1.1447436830612456,
"grad_norm": 0.003653658786788583,
"learning_rate": 4.2792638343382894e-06,
"loss": 0.004692722111940384,
"step": 689
},
{
"epoch": 1.1464074035562026,
"grad_norm": 0.0026509996969252825,
"learning_rate": 4.276849717065295e-06,
"loss": 0.004140852950513363,
"step": 690
},
{
"epoch": 1.1480711240511594,
"grad_norm": 0.002603845437988639,
"learning_rate": 4.2744322469120296e-06,
"loss": 0.004025573376566172,
"step": 691
},
{
"epoch": 1.1497348445461162,
"grad_norm": 0.0027512742672115564,
"learning_rate": 4.272011428440212e-06,
"loss": 0.004732303321361542,
"step": 692
},
{
"epoch": 1.1513985650410732,
"grad_norm": 0.002622104249894619,
"learning_rate": 4.269587266217878e-06,
"loss": 0.007063106633722782,
"step": 693
},
{
"epoch": 1.15306228553603,
"grad_norm": 0.0021548105869442225,
"learning_rate": 4.2671597648193745e-06,
"loss": 0.0035763587802648544,
"step": 694
},
{
"epoch": 1.1547260060309867,
"grad_norm": 0.002422524616122246,
"learning_rate": 4.264728928825347e-06,
"loss": 0.007466421462595463,
"step": 695
},
{
"epoch": 1.1563897265259437,
"grad_norm": 0.0017905524000525475,
"learning_rate": 4.262294762822738e-06,
"loss": 0.0030708981212228537,
"step": 696
},
{
"epoch": 1.1580534470209005,
"grad_norm": 0.0022994892206043005,
"learning_rate": 4.259857271404767e-06,
"loss": 0.0022998168133199215,
"step": 697
},
{
"epoch": 1.1597171675158573,
"grad_norm": 0.0017049030866473913,
"learning_rate": 4.257416459170935e-06,
"loss": 0.0031123387161642313,
"step": 698
},
{
"epoch": 1.161380888010814,
"grad_norm": 0.002179944422096014,
"learning_rate": 4.254972330727004e-06,
"loss": 0.003777222940698266,
"step": 699
},
{
"epoch": 1.163044608505771,
"grad_norm": 0.004692301619797945,
"learning_rate": 4.252524890685e-06,
"loss": 0.009404093027114868,
"step": 700
},
{
"epoch": 1.163044608505771,
"eval_loss": 0.004640148486942053,
"eval_runtime": 424.8823,
"eval_samples_per_second": 11.316,
"eval_steps_per_second": 2.829,
"step": 700
},
{
"epoch": 1.1647083290007278,
"grad_norm": 0.002540596527978778,
"learning_rate": 4.250074143663189e-06,
"loss": 0.006953324191272259,
"step": 701
},
{
"epoch": 1.1663720494956848,
"grad_norm": 0.006684183143079281,
"learning_rate": 4.247620094286085e-06,
"loss": 0.004303854890167713,
"step": 702
},
{
"epoch": 1.1680357699906416,
"grad_norm": 0.002960954559966922,
"learning_rate": 4.2451627471844305e-06,
"loss": 0.0035925274714827538,
"step": 703
},
{
"epoch": 1.1696994904855984,
"grad_norm": 0.003732532262802124,
"learning_rate": 4.24270210699519e-06,
"loss": 0.008118562400341034,
"step": 704
},
{
"epoch": 1.1713632109805552,
"grad_norm": 0.004170422907918692,
"learning_rate": 4.240238178361543e-06,
"loss": 0.003138990141451359,
"step": 705
},
{
"epoch": 1.1730269314755122,
"grad_norm": 0.002722041215747595,
"learning_rate": 4.237770965932875e-06,
"loss": 0.0035777364391833544,
"step": 706
},
{
"epoch": 1.174690651970469,
"grad_norm": 0.00289748003706336,
"learning_rate": 4.235300474364766e-06,
"loss": 0.005866762716323137,
"step": 707
},
{
"epoch": 1.1763543724654257,
"grad_norm": 0.0030904258601367474,
"learning_rate": 4.232826708318985e-06,
"loss": 0.0064615714363753796,
"step": 708
},
{
"epoch": 1.1780180929603827,
"grad_norm": 0.004810953047126532,
"learning_rate": 4.230349672463481e-06,
"loss": 0.004663039464503527,
"step": 709
},
{
"epoch": 1.1796818134553395,
"grad_norm": 0.0038745985366404057,
"learning_rate": 4.22786937147237e-06,
"loss": 0.007224446628242731,
"step": 710
},
{
"epoch": 1.1813455339502963,
"grad_norm": 0.0015594467986375093,
"learning_rate": 4.2253858100259304e-06,
"loss": 0.0034856877755373716,
"step": 711
},
{
"epoch": 1.1830092544452533,
"grad_norm": 0.0020052522886544466,
"learning_rate": 4.222898992810596e-06,
"loss": 0.004635176621377468,
"step": 712
},
{
"epoch": 1.18467297494021,
"grad_norm": 0.001965244999155402,
"learning_rate": 4.220408924518939e-06,
"loss": 0.0020197900012135506,
"step": 713
},
{
"epoch": 1.1863366954351668,
"grad_norm": 0.002565849805250764,
"learning_rate": 4.217915609849671e-06,
"loss": 0.005158688873052597,
"step": 714
},
{
"epoch": 1.1880004159301238,
"grad_norm": 0.0020221301820129156,
"learning_rate": 4.215419053507626e-06,
"loss": 0.004507225006818771,
"step": 715
},
{
"epoch": 1.1896641364250806,
"grad_norm": 0.004605746828019619,
"learning_rate": 4.212919260203757e-06,
"loss": 0.006004433147609234,
"step": 716
},
{
"epoch": 1.1913278569200374,
"grad_norm": 0.0044348943047225475,
"learning_rate": 4.210416234655125e-06,
"loss": 0.007470245473086834,
"step": 717
},
{
"epoch": 1.1929915774149942,
"grad_norm": 0.0030243340879678726,
"learning_rate": 4.207909981584889e-06,
"loss": 0.0035587972961366177,
"step": 718
},
{
"epoch": 1.1946552979099512,
"grad_norm": 0.0022278032265603542,
"learning_rate": 4.2054005057223e-06,
"loss": 0.006495712324976921,
"step": 719
},
{
"epoch": 1.196319018404908,
"grad_norm": 0.002851974219083786,
"learning_rate": 4.202887811802687e-06,
"loss": 0.004013349302113056,
"step": 720
},
{
"epoch": 1.1979827388998647,
"grad_norm": 0.0021993438713252544,
"learning_rate": 4.200371904567457e-06,
"loss": 0.0037416350096464157,
"step": 721
},
{
"epoch": 1.1996464593948217,
"grad_norm": 0.004688817076385021,
"learning_rate": 4.197852788764075e-06,
"loss": 0.00362020474858582,
"step": 722
},
{
"epoch": 1.2013101798897785,
"grad_norm": 0.004926290363073349,
"learning_rate": 4.195330469146063e-06,
"loss": 0.007041393779218197,
"step": 723
},
{
"epoch": 1.2029739003847353,
"grad_norm": 0.0027303057722747326,
"learning_rate": 4.1928049504729886e-06,
"loss": 0.004923125728964806,
"step": 724
},
{
"epoch": 1.2046376208796923,
"grad_norm": 0.0019440047908574343,
"learning_rate": 4.1902762375104555e-06,
"loss": 0.0022577960044145584,
"step": 725
},
{
"epoch": 1.206301341374649,
"grad_norm": 0.0027244926895946264,
"learning_rate": 4.187744335030095e-06,
"loss": 0.0067812055349349976,
"step": 726
},
{
"epoch": 1.2079650618696058,
"grad_norm": 0.002273574937134981,
"learning_rate": 4.185209247809557e-06,
"loss": 0.003740078303962946,
"step": 727
},
{
"epoch": 1.2096287823645628,
"grad_norm": 0.0034744152799248695,
"learning_rate": 4.182670980632501e-06,
"loss": 0.013247332535684109,
"step": 728
},
{
"epoch": 1.2112925028595196,
"grad_norm": 0.004806975368410349,
"learning_rate": 4.180129538288587e-06,
"loss": 0.006336216814815998,
"step": 729
},
{
"epoch": 1.2129562233544764,
"grad_norm": 0.0007596082286909223,
"learning_rate": 4.177584925573466e-06,
"loss": 0.0027835089713335037,
"step": 730
},
{
"epoch": 1.2146199438494334,
"grad_norm": 0.002162647433578968,
"learning_rate": 4.175037147288772e-06,
"loss": 0.005248047411441803,
"step": 731
},
{
"epoch": 1.2162836643443902,
"grad_norm": 0.0020637372508645058,
"learning_rate": 4.172486208242113e-06,
"loss": 0.0023893651086837053,
"step": 732
},
{
"epoch": 1.217947384839347,
"grad_norm": 0.004547535441815853,
"learning_rate": 4.169932113247059e-06,
"loss": 0.004172275308519602,
"step": 733
},
{
"epoch": 1.219611105334304,
"grad_norm": 0.0015233661979436874,
"learning_rate": 4.167374867123138e-06,
"loss": 0.0022961003705859184,
"step": 734
},
{
"epoch": 1.2212748258292607,
"grad_norm": 0.001560075324960053,
"learning_rate": 4.164814474695823e-06,
"loss": 0.0037685995921492577,
"step": 735
},
{
"epoch": 1.2229385463242175,
"grad_norm": 0.002407171530649066,
"learning_rate": 4.162250940796523e-06,
"loss": 0.003144836286082864,
"step": 736
},
{
"epoch": 1.2246022668191743,
"grad_norm": 0.0025342016015201807,
"learning_rate": 4.159684270262576e-06,
"loss": 0.005925535224378109,
"step": 737
},
{
"epoch": 1.2262659873141313,
"grad_norm": 0.0020147510804235935,
"learning_rate": 4.157114467937239e-06,
"loss": 0.002619993407279253,
"step": 738
},
{
"epoch": 1.227929707809088,
"grad_norm": 0.002741551026701927,
"learning_rate": 4.154541538669677e-06,
"loss": 0.0038237329572439194,
"step": 739
},
{
"epoch": 1.2295934283040448,
"grad_norm": 0.0026750387623906136,
"learning_rate": 4.151965487314959e-06,
"loss": 0.00299008353613317,
"step": 740
},
{
"epoch": 1.2312571487990018,
"grad_norm": 0.0017631236696615815,
"learning_rate": 4.1493863187340415e-06,
"loss": 0.00323181739076972,
"step": 741
},
{
"epoch": 1.2329208692939586,
"grad_norm": 0.00393708935007453,
"learning_rate": 4.146804037793763e-06,
"loss": 0.0074392640963196754,
"step": 742
},
{
"epoch": 1.2345845897889154,
"grad_norm": 0.002047119429334998,
"learning_rate": 4.144218649366839e-06,
"loss": 0.003371204948052764,
"step": 743
},
{
"epoch": 1.2362483102838724,
"grad_norm": 0.002852058270946145,
"learning_rate": 4.141630158331845e-06,
"loss": 0.005799049511551857,
"step": 744
},
{
"epoch": 1.2379120307788292,
"grad_norm": 0.002589620416983962,
"learning_rate": 4.139038569573213e-06,
"loss": 0.0037557545583695173,
"step": 745
},
{
"epoch": 1.239575751273786,
"grad_norm": 0.0013611821923404932,
"learning_rate": 4.1364438879812194e-06,
"loss": 0.0026148685719817877,
"step": 746
},
{
"epoch": 1.241239471768743,
"grad_norm": 0.0026877254713326693,
"learning_rate": 4.1338461184519776e-06,
"loss": 0.00639910064637661,
"step": 747
},
{
"epoch": 1.2429031922636997,
"grad_norm": 0.004890333395451307,
"learning_rate": 4.131245265887426e-06,
"loss": 0.014384834095835686,
"step": 748
},
{
"epoch": 1.2445669127586565,
"grad_norm": 0.0029951855540275574,
"learning_rate": 4.1286413351953235e-06,
"loss": 0.00565043231472373,
"step": 749
},
{
"epoch": 1.2462306332536133,
"grad_norm": 0.0024212165735661983,
"learning_rate": 4.126034331289235e-06,
"loss": 0.004203692078590393,
"step": 750
},
{
"epoch": 1.2478943537485703,
"grad_norm": 0.0020162512082606554,
"learning_rate": 4.123424259088525e-06,
"loss": 0.0035710574593394995,
"step": 751
},
{
"epoch": 1.249558074243527,
"grad_norm": 0.0020386476535350084,
"learning_rate": 4.120811123518349e-06,
"loss": 0.004487013444304466,
"step": 752
},
{
"epoch": 1.251221794738484,
"grad_norm": 0.002236759290099144,
"learning_rate": 4.1181949295096415e-06,
"loss": 0.004523813258856535,
"step": 753
},
{
"epoch": 1.2528855152334408,
"grad_norm": 0.0019163981778547168,
"learning_rate": 4.11557568199911e-06,
"loss": 0.003014134708791971,
"step": 754
},
{
"epoch": 1.2545492357283976,
"grad_norm": 0.001937209046445787,
"learning_rate": 4.112953385929221e-06,
"loss": 0.003906419035047293,
"step": 755
},
{
"epoch": 1.2562129562233544,
"grad_norm": 0.0027945186011493206,
"learning_rate": 4.110328046248196e-06,
"loss": 0.005981692112982273,
"step": 756
},
{
"epoch": 1.2578766767183114,
"grad_norm": 0.0011745948577299714,
"learning_rate": 4.107699667909999e-06,
"loss": 0.0016988021088764071,
"step": 757
},
{
"epoch": 1.2595403972132682,
"grad_norm": 0.001418607891537249,
"learning_rate": 4.105068255874328e-06,
"loss": 0.003487003967165947,
"step": 758
},
{
"epoch": 1.261204117708225,
"grad_norm": 0.0013838058803230524,
"learning_rate": 4.102433815106606e-06,
"loss": 0.0015933482209220529,
"step": 759
},
{
"epoch": 1.262867838203182,
"grad_norm": 0.002978533273562789,
"learning_rate": 4.09979635057797e-06,
"loss": 0.00687625166028738,
"step": 760
},
{
"epoch": 1.2645315586981387,
"grad_norm": 0.001718941261060536,
"learning_rate": 4.097155867265264e-06,
"loss": 0.003526041517034173,
"step": 761
},
{
"epoch": 1.2661952791930955,
"grad_norm": 0.003133240854367614,
"learning_rate": 4.094512370151027e-06,
"loss": 0.003928614780306816,
"step": 762
},
{
"epoch": 1.2678589996880525,
"grad_norm": 0.0022098422050476074,
"learning_rate": 4.091865864223487e-06,
"loss": 0.0026667716447263956,
"step": 763
},
{
"epoch": 1.2695227201830093,
"grad_norm": 0.0017531696939840913,
"learning_rate": 4.089216354476545e-06,
"loss": 0.001974761486053467,
"step": 764
},
{
"epoch": 1.271186440677966,
"grad_norm": 0.0009637777111493051,
"learning_rate": 4.086563845909779e-06,
"loss": 0.0017917206278070807,
"step": 765
},
{
"epoch": 1.272850161172923,
"grad_norm": 0.0018993336707353592,
"learning_rate": 4.083908343528415e-06,
"loss": 0.00392130296677351,
"step": 766
},
{
"epoch": 1.2745138816678798,
"grad_norm": 0.0016326892655342817,
"learning_rate": 4.081249852343336e-06,
"loss": 0.003777984995394945,
"step": 767
},
{
"epoch": 1.2761776021628366,
"grad_norm": 0.0037745193112641573,
"learning_rate": 4.078588377371062e-06,
"loss": 0.003253435716032982,
"step": 768
},
{
"epoch": 1.2778413226577934,
"grad_norm": 0.005981610622256994,
"learning_rate": 4.075923923633745e-06,
"loss": 0.006038327235728502,
"step": 769
},
{
"epoch": 1.2795050431527504,
"grad_norm": 0.004553422797471285,
"learning_rate": 4.073256496159153e-06,
"loss": 0.0040462627075612545,
"step": 770
},
{
"epoch": 1.2811687636477072,
"grad_norm": 0.0017073902999982238,
"learning_rate": 4.070586099980672e-06,
"loss": 0.0034780718851834536,
"step": 771
},
{
"epoch": 1.2828324841426642,
"grad_norm": 0.0033805659040808678,
"learning_rate": 4.067912740137285e-06,
"loss": 0.00390542671084404,
"step": 772
},
{
"epoch": 1.284496204637621,
"grad_norm": 0.0035299034789204597,
"learning_rate": 4.06523642167357e-06,
"loss": 0.003843412036076188,
"step": 773
},
{
"epoch": 1.2861599251325777,
"grad_norm": 0.004187435377389193,
"learning_rate": 4.062557149639688e-06,
"loss": 0.004727951250970364,
"step": 774
},
{
"epoch": 1.2878236456275345,
"grad_norm": 0.0027708059642463923,
"learning_rate": 4.059874929091369e-06,
"loss": 0.0035429196432232857,
"step": 775
},
{
"epoch": 1.2894873661224915,
"grad_norm": 0.0021549989469349384,
"learning_rate": 4.057189765089914e-06,
"loss": 0.0030433572828769684,
"step": 776
},
{
"epoch": 1.2911510866174483,
"grad_norm": 0.0018761793617159128,
"learning_rate": 4.054501662702172e-06,
"loss": 0.0032656132243573666,
"step": 777
},
{
"epoch": 1.292814807112405,
"grad_norm": 0.0019973707385361195,
"learning_rate": 4.05181062700054e-06,
"loss": 0.0048477197997272015,
"step": 778
},
{
"epoch": 1.294478527607362,
"grad_norm": 0.002293492667376995,
"learning_rate": 4.049116663062949e-06,
"loss": 0.002738622482866049,
"step": 779
},
{
"epoch": 1.2961422481023188,
"grad_norm": 0.0017012613825500011,
"learning_rate": 4.046419775972855e-06,
"loss": 0.002342861844226718,
"step": 780
},
{
"epoch": 1.2978059685972756,
"grad_norm": 0.0008426354033872485,
"learning_rate": 4.043719970819231e-06,
"loss": 0.0015461614821106195,
"step": 781
},
{
"epoch": 1.2994696890922324,
"grad_norm": 0.0017862239619717002,
"learning_rate": 4.041017252696556e-06,
"loss": 0.003129967488348484,
"step": 782
},
{
"epoch": 1.3011334095871894,
"grad_norm": 0.0016204863786697388,
"learning_rate": 4.038311626704806e-06,
"loss": 0.0013765785843133926,
"step": 783
},
{
"epoch": 1.3027971300821461,
"grad_norm": 0.0023099181707948446,
"learning_rate": 4.035603097949444e-06,
"loss": 0.0042774975299835205,
"step": 784
},
{
"epoch": 1.3044608505771031,
"grad_norm": 0.0017065017018467188,
"learning_rate": 4.032891671541409e-06,
"loss": 0.00256139412522316,
"step": 785
},
{
"epoch": 1.30612457107206,
"grad_norm": 0.0013910774141550064,
"learning_rate": 4.030177352597109e-06,
"loss": 0.003075619228184223,
"step": 786
},
{
"epoch": 1.3077882915670167,
"grad_norm": 0.003909314516931772,
"learning_rate": 4.027460146238411e-06,
"loss": 0.006473972462117672,
"step": 787
},
{
"epoch": 1.3094520120619735,
"grad_norm": 0.015759041532874107,
"learning_rate": 4.02474005759263e-06,
"loss": 0.007488309871405363,
"step": 788
},
{
"epoch": 1.3111157325569305,
"grad_norm": 0.002802561968564987,
"learning_rate": 4.022017091792518e-06,
"loss": 0.0052157179452478886,
"step": 789
},
{
"epoch": 1.3127794530518873,
"grad_norm": 0.0033724752720445395,
"learning_rate": 4.01929125397626e-06,
"loss": 0.004162366967648268,
"step": 790
},
{
"epoch": 1.3144431735468443,
"grad_norm": 0.0031333763618022203,
"learning_rate": 4.016562549287455e-06,
"loss": 0.004047158639878035,
"step": 791
},
{
"epoch": 1.316106894041801,
"grad_norm": 0.0016262867720797658,
"learning_rate": 4.013830982875117e-06,
"loss": 0.0020962050184607506,
"step": 792
},
{
"epoch": 1.3177706145367578,
"grad_norm": 0.002156744012609124,
"learning_rate": 4.0110965598936565e-06,
"loss": 0.003945136908441782,
"step": 793
},
{
"epoch": 1.3194343350317146,
"grad_norm": 0.003021260490640998,
"learning_rate": 4.008359285502877e-06,
"loss": 0.00483740633353591,
"step": 794
},
{
"epoch": 1.3210980555266716,
"grad_norm": 0.0018034290987998247,
"learning_rate": 4.005619164867959e-06,
"loss": 0.002895489800721407,
"step": 795
},
{
"epoch": 1.3227617760216284,
"grad_norm": 0.0027613251004368067,
"learning_rate": 4.002876203159458e-06,
"loss": 0.004742368124425411,
"step": 796
},
{
"epoch": 1.3244254965165851,
"grad_norm": 0.004044833593070507,
"learning_rate": 4.000130405553287e-06,
"loss": 0.0023626082111150026,
"step": 797
},
{
"epoch": 1.3260892170115421,
"grad_norm": 0.0029203095473349094,
"learning_rate": 3.997381777230714e-06,
"loss": 0.0063598062843084335,
"step": 798
},
{
"epoch": 1.327752937506499,
"grad_norm": 0.0028201239183545113,
"learning_rate": 3.994630323378344e-06,
"loss": 0.004977297969162464,
"step": 799
},
{
"epoch": 1.3294166580014557,
"grad_norm": 0.0036451423075050116,
"learning_rate": 3.991876049188116e-06,
"loss": 0.0038258880376815796,
"step": 800
},
{
"epoch": 1.3294166580014557,
"eval_loss": 0.004822524730116129,
"eval_runtime": 402.1932,
"eval_samples_per_second": 11.954,
"eval_steps_per_second": 2.989,
"step": 800
},
{
"epoch": 1.3310803784964125,
"grad_norm": 0.0028423636686056852,
"learning_rate": 3.989118959857293e-06,
"loss": 0.0061726756393909454,
"step": 801
},
{
"epoch": 1.3327440989913695,
"grad_norm": 0.002013375284150243,
"learning_rate": 3.986359060588446e-06,
"loss": 0.00220022676512599,
"step": 802
},
{
"epoch": 1.3344078194863263,
"grad_norm": 0.00471561448648572,
"learning_rate": 3.983596356589452e-06,
"loss": 0.0031772053334861994,
"step": 803
},
{
"epoch": 1.3360715399812833,
"grad_norm": 0.002453225664794445,
"learning_rate": 3.980830853073476e-06,
"loss": 0.004086391534656286,
"step": 804
},
{
"epoch": 1.33773526047624,
"grad_norm": 0.002689510118216276,
"learning_rate": 3.978062555258972e-06,
"loss": 0.0053067514672875404,
"step": 805
},
{
"epoch": 1.3393989809711968,
"grad_norm": 0.002451070351526141,
"learning_rate": 3.975291468369661e-06,
"loss": 0.0021515078842639923,
"step": 806
},
{
"epoch": 1.3410627014661536,
"grad_norm": 0.0024362937547266483,
"learning_rate": 3.97251759763453e-06,
"loss": 0.004400534089654684,
"step": 807
},
{
"epoch": 1.3427264219611106,
"grad_norm": 0.0022519261110574007,
"learning_rate": 3.969740948287817e-06,
"loss": 0.0035801767371594906,
"step": 808
},
{
"epoch": 1.3443901424560674,
"grad_norm": 0.004936764016747475,
"learning_rate": 3.966961525569005e-06,
"loss": 0.007511000614613295,
"step": 809
},
{
"epoch": 1.3460538629510244,
"grad_norm": 0.004098286386579275,
"learning_rate": 3.964179334722811e-06,
"loss": 0.0044007147662341595,
"step": 810
},
{
"epoch": 1.3477175834459811,
"grad_norm": 0.004742534831166267,
"learning_rate": 3.961394380999173e-06,
"loss": 0.0067534856498241425,
"step": 811
},
{
"epoch": 1.349381303940938,
"grad_norm": 0.001244925195351243,
"learning_rate": 3.958606669653243e-06,
"loss": 0.0035037680063396692,
"step": 812
},
{
"epoch": 1.3510450244358947,
"grad_norm": 0.0022341671865433455,
"learning_rate": 3.955816205945378e-06,
"loss": 0.005220518913120031,
"step": 813
},
{
"epoch": 1.3527087449308517,
"grad_norm": 0.002156665548682213,
"learning_rate": 3.953022995141128e-06,
"loss": 0.005255310796201229,
"step": 814
},
{
"epoch": 1.3543724654258085,
"grad_norm": 0.0027065190952271223,
"learning_rate": 3.950227042511226e-06,
"loss": 0.004826219752430916,
"step": 815
},
{
"epoch": 1.3560361859207652,
"grad_norm": 0.001812028931453824,
"learning_rate": 3.947428353331579e-06,
"loss": 0.005146780516952276,
"step": 816
},
{
"epoch": 1.3576999064157222,
"grad_norm": 0.0019188732840120792,
"learning_rate": 3.94462693288326e-06,
"loss": 0.002942196559160948,
"step": 817
},
{
"epoch": 1.359363626910679,
"grad_norm": 0.0029397630132734776,
"learning_rate": 3.941822786452491e-06,
"loss": 0.0025993953458964825,
"step": 818
},
{
"epoch": 1.3610273474056358,
"grad_norm": 0.002346345689147711,
"learning_rate": 3.939015919330643e-06,
"loss": 0.003963056020438671,
"step": 819
},
{
"epoch": 1.3626910679005926,
"grad_norm": 0.003207869129255414,
"learning_rate": 3.936206336814219e-06,
"loss": 0.005814023315906525,
"step": 820
},
{
"epoch": 1.3643547883955496,
"grad_norm": 0.0013737737899646163,
"learning_rate": 3.933394044204843e-06,
"loss": 0.0030613031703978777,
"step": 821
},
{
"epoch": 1.3660185088905064,
"grad_norm": 0.0035271679516881704,
"learning_rate": 3.930579046809259e-06,
"loss": 0.005814910866320133,
"step": 822
},
{
"epoch": 1.3676822293854634,
"grad_norm": 0.0025771872606128454,
"learning_rate": 3.92776134993931e-06,
"loss": 0.005833122879266739,
"step": 823
},
{
"epoch": 1.3693459498804201,
"grad_norm": 0.002427369588986039,
"learning_rate": 3.924940958911933e-06,
"loss": 0.004450193606317043,
"step": 824
},
{
"epoch": 1.371009670375377,
"grad_norm": 0.0019530977588146925,
"learning_rate": 3.922117879049152e-06,
"loss": 0.003191176801919937,
"step": 825
},
{
"epoch": 1.3726733908703337,
"grad_norm": 0.003067216370254755,
"learning_rate": 3.91929211567806e-06,
"loss": 0.0036342469975352287,
"step": 826
},
{
"epoch": 1.3743371113652907,
"grad_norm": 0.0018951890524476767,
"learning_rate": 3.916463674130821e-06,
"loss": 0.0027206242084503174,
"step": 827
},
{
"epoch": 1.3760008318602475,
"grad_norm": 0.004039694555103779,
"learning_rate": 3.913632559744645e-06,
"loss": 0.006162284407764673,
"step": 828
},
{
"epoch": 1.3776645523552045,
"grad_norm": 0.003776967292651534,
"learning_rate": 3.910798777861788e-06,
"loss": 0.01263501588255167,
"step": 829
},
{
"epoch": 1.3793282728501612,
"grad_norm": 0.00310622900724411,
"learning_rate": 3.9079623338295436e-06,
"loss": 0.004743027966469526,
"step": 830
},
{
"epoch": 1.380991993345118,
"grad_norm": 0.003913409076631069,
"learning_rate": 3.9051232330002245e-06,
"loss": 0.0044493465684354305,
"step": 831
},
{
"epoch": 1.3826557138400748,
"grad_norm": 0.0017946249572560191,
"learning_rate": 3.902281480731156e-06,
"loss": 0.0026314181741327047,
"step": 832
},
{
"epoch": 1.3843194343350318,
"grad_norm": 0.0020571874920278788,
"learning_rate": 3.899437082384671e-06,
"loss": 0.004028120543807745,
"step": 833
},
{
"epoch": 1.3859831548299886,
"grad_norm": 0.0026965150609612465,
"learning_rate": 3.89659004332809e-06,
"loss": 0.004003521986305714,
"step": 834
},
{
"epoch": 1.3876468753249454,
"grad_norm": 0.003063056617975235,
"learning_rate": 3.893740368933722e-06,
"loss": 0.0029916935600340366,
"step": 835
},
{
"epoch": 1.3893105958199023,
"grad_norm": 0.0036277994513511658,
"learning_rate": 3.8908880645788464e-06,
"loss": 0.007052903063595295,
"step": 836
},
{
"epoch": 1.3909743163148591,
"grad_norm": 0.0017010894371196628,
"learning_rate": 3.888033135645702e-06,
"loss": 0.0030756453052163124,
"step": 837
},
{
"epoch": 1.392638036809816,
"grad_norm": 0.0024809178430587053,
"learning_rate": 3.885175587521486e-06,
"loss": 0.007132468745112419,
"step": 838
},
{
"epoch": 1.3943017573047727,
"grad_norm": 0.0016055714804679155,
"learning_rate": 3.882315425598334e-06,
"loss": 0.0038319770246744156,
"step": 839
},
{
"epoch": 1.3959654777997297,
"grad_norm": 0.0018153567798435688,
"learning_rate": 3.879452655273316e-06,
"loss": 0.001998928841203451,
"step": 840
},
{
"epoch": 1.3976291982946865,
"grad_norm": 0.0028074143920093775,
"learning_rate": 3.876587281948422e-06,
"loss": 0.005024431273341179,
"step": 841
},
{
"epoch": 1.3992929187896435,
"grad_norm": 0.0022321913857012987,
"learning_rate": 3.873719311030556e-06,
"loss": 0.002615418517962098,
"step": 842
},
{
"epoch": 1.4009566392846002,
"grad_norm": 0.0021348358131945133,
"learning_rate": 3.8708487479315204e-06,
"loss": 0.0033859461545944214,
"step": 843
},
{
"epoch": 1.402620359779557,
"grad_norm": 0.005225852597504854,
"learning_rate": 3.867975598068012e-06,
"loss": 0.002719119656831026,
"step": 844
},
{
"epoch": 1.4042840802745138,
"grad_norm": 0.0009146020747721195,
"learning_rate": 3.8650998668616085e-06,
"loss": 0.0021196685265749693,
"step": 845
},
{
"epoch": 1.4059478007694708,
"grad_norm": 0.0017917942022904754,
"learning_rate": 3.862221559738757e-06,
"loss": 0.0072954390197992325,
"step": 846
},
{
"epoch": 1.4076115212644276,
"grad_norm": 0.002280753804370761,
"learning_rate": 3.859340682130766e-06,
"loss": 0.004195714369416237,
"step": 847
},
{
"epoch": 1.4092752417593843,
"grad_norm": 0.0031667782459408045,
"learning_rate": 3.856457239473795e-06,
"loss": 0.002995600923895836,
"step": 848
},
{
"epoch": 1.4109389622543413,
"grad_norm": 0.00300840032286942,
"learning_rate": 3.853571237208843e-06,
"loss": 0.006693383678793907,
"step": 849
},
{
"epoch": 1.4126026827492981,
"grad_norm": 0.0032315480057150126,
"learning_rate": 3.8506826807817395e-06,
"loss": 0.008001086302101612,
"step": 850
},
{
"epoch": 1.414266403244255,
"grad_norm": 0.0031034990679472685,
"learning_rate": 3.847791575643134e-06,
"loss": 0.005453841295093298,
"step": 851
},
{
"epoch": 1.4159301237392117,
"grad_norm": 0.0038969414308667183,
"learning_rate": 3.844897927248483e-06,
"loss": 0.0055399080738425255,
"step": 852
},
{
"epoch": 1.4175938442341687,
"grad_norm": 0.002615910256281495,
"learning_rate": 3.842001741058045e-06,
"loss": 0.0037770913913846016,
"step": 853
},
{
"epoch": 1.4192575647291255,
"grad_norm": 0.001906339661218226,
"learning_rate": 3.839103022536865e-06,
"loss": 0.005109758581966162,
"step": 854
},
{
"epoch": 1.4209212852240825,
"grad_norm": 0.0017247337382286787,
"learning_rate": 3.836201777154769e-06,
"loss": 0.004554691258817911,
"step": 855
},
{
"epoch": 1.4225850057190392,
"grad_norm": 0.0023999211844056845,
"learning_rate": 3.833298010386347e-06,
"loss": 0.006145539693534374,
"step": 856
},
{
"epoch": 1.424248726213996,
"grad_norm": 0.0028406865894794464,
"learning_rate": 3.830391727710954e-06,
"loss": 0.006535790394991636,
"step": 857
},
{
"epoch": 1.4259124467089528,
"grad_norm": 0.002053679432719946,
"learning_rate": 3.827482934612684e-06,
"loss": 0.006250452250242233,
"step": 858
},
{
"epoch": 1.4275761672039098,
"grad_norm": 0.001439629471860826,
"learning_rate": 3.824571636580372e-06,
"loss": 0.0042776912450790405,
"step": 859
},
{
"epoch": 1.4292398876988666,
"grad_norm": 0.0013407040387392044,
"learning_rate": 3.821657839107583e-06,
"loss": 0.002986511215567589,
"step": 860
},
{
"epoch": 1.4309036081938236,
"grad_norm": 0.0020117536187171936,
"learning_rate": 3.818741547692593e-06,
"loss": 0.003501957282423973,
"step": 861
},
{
"epoch": 1.4325673286887803,
"grad_norm": 0.0011241771280765533,
"learning_rate": 3.815822767838386e-06,
"loss": 0.003273077541962266,
"step": 862
},
{
"epoch": 1.4342310491837371,
"grad_norm": 0.0014655692502856255,
"learning_rate": 3.812901505052642e-06,
"loss": 0.002575968625023961,
"step": 863
},
{
"epoch": 1.435894769678694,
"grad_norm": 0.0026631092187017202,
"learning_rate": 3.8099777648477264e-06,
"loss": 0.004115140065550804,
"step": 864
},
{
"epoch": 1.437558490173651,
"grad_norm": 0.0026415600441396236,
"learning_rate": 3.8070515527406803e-06,
"loss": 0.006131789647042751,
"step": 865
},
{
"epoch": 1.4392222106686077,
"grad_norm": 0.0013655349612236023,
"learning_rate": 3.8041228742532064e-06,
"loss": 0.004735004622489214,
"step": 866
},
{
"epoch": 1.4408859311635644,
"grad_norm": 0.0008041858091019094,
"learning_rate": 3.8011917349116633e-06,
"loss": 0.0017425359692424536,
"step": 867
},
{
"epoch": 1.4425496516585214,
"grad_norm": 0.0026084021665155888,
"learning_rate": 3.7982581402470536e-06,
"loss": 0.003985221032053232,
"step": 868
},
{
"epoch": 1.4442133721534782,
"grad_norm": 0.0017706190701574087,
"learning_rate": 3.795322095795012e-06,
"loss": 0.005514065269380808,
"step": 869
},
{
"epoch": 1.445877092648435,
"grad_norm": 0.0013553850585594773,
"learning_rate": 3.7923836070957963e-06,
"loss": 0.0028120880015194416,
"step": 870
},
{
"epoch": 1.4475408131433918,
"grad_norm": 0.0017212049569934607,
"learning_rate": 3.7894426796942773e-06,
"loss": 0.003140087705105543,
"step": 871
},
{
"epoch": 1.4492045336383488,
"grad_norm": 0.002766014775261283,
"learning_rate": 3.786499319139926e-06,
"loss": 0.0036492799408733845,
"step": 872
},
{
"epoch": 1.4508682541333056,
"grad_norm": 0.0016567467246204615,
"learning_rate": 3.7835535309868055e-06,
"loss": 0.0021711941808462143,
"step": 873
},
{
"epoch": 1.4525319746282626,
"grad_norm": 0.001408170210197568,
"learning_rate": 3.78060532079356e-06,
"loss": 0.002645156579092145,
"step": 874
},
{
"epoch": 1.4541956951232193,
"grad_norm": 0.0050718155689537525,
"learning_rate": 3.777654694123404e-06,
"loss": 0.0038958254735916853,
"step": 875
},
{
"epoch": 1.4558594156181761,
"grad_norm": 0.0029584639705717564,
"learning_rate": 3.7747016565441112e-06,
"loss": 0.004871165845543146,
"step": 876
},
{
"epoch": 1.4575231361131329,
"grad_norm": 0.001699764747172594,
"learning_rate": 3.771746213628006e-06,
"loss": 0.0027038324624300003,
"step": 877
},
{
"epoch": 1.4591868566080899,
"grad_norm": 0.0031087801326066256,
"learning_rate": 3.7687883709519496e-06,
"loss": 0.004658531863242388,
"step": 878
},
{
"epoch": 1.4608505771030467,
"grad_norm": 0.0026078419759869576,
"learning_rate": 3.7658281340973336e-06,
"loss": 0.0025535072200000286,
"step": 879
},
{
"epoch": 1.4625142975980037,
"grad_norm": 0.002413343172520399,
"learning_rate": 3.7628655086500654e-06,
"loss": 0.0027896517422050238,
"step": 880
},
{
"epoch": 1.4641780180929604,
"grad_norm": 0.001798558863811195,
"learning_rate": 3.7599005002005616e-06,
"loss": 0.0048567005433142185,
"step": 881
},
{
"epoch": 1.4658417385879172,
"grad_norm": 0.002467783633619547,
"learning_rate": 3.7569331143437336e-06,
"loss": 0.009476818144321442,
"step": 882
},
{
"epoch": 1.467505459082874,
"grad_norm": 0.002080975566059351,
"learning_rate": 3.7539633566789812e-06,
"loss": 0.0033652521669864655,
"step": 883
},
{
"epoch": 1.469169179577831,
"grad_norm": 0.003472168231382966,
"learning_rate": 3.750991232810177e-06,
"loss": 0.00317693455144763,
"step": 884
},
{
"epoch": 1.4708329000727878,
"grad_norm": 0.0024736488703638315,
"learning_rate": 3.7480167483456603e-06,
"loss": 0.0052384547889232635,
"step": 885
},
{
"epoch": 1.4724966205677446,
"grad_norm": 0.0017958751413971186,
"learning_rate": 3.7450399088982247e-06,
"loss": 0.001992971170693636,
"step": 886
},
{
"epoch": 1.4741603410627016,
"grad_norm": 0.003060857066884637,
"learning_rate": 3.742060720085107e-06,
"loss": 0.005137836094945669,
"step": 887
},
{
"epoch": 1.4758240615576583,
"grad_norm": 0.0017445053672417998,
"learning_rate": 3.739079187527978e-06,
"loss": 0.002876036800444126,
"step": 888
},
{
"epoch": 1.477487782052615,
"grad_norm": 0.002199475420638919,
"learning_rate": 3.73609531685293e-06,
"loss": 0.0028552846051752567,
"step": 889
},
{
"epoch": 1.4791515025475719,
"grad_norm": 0.0037617129273712635,
"learning_rate": 3.733109113690469e-06,
"loss": 0.009398862719535828,
"step": 890
},
{
"epoch": 1.4808152230425289,
"grad_norm": 0.0013438520254567266,
"learning_rate": 3.7301205836755006e-06,
"loss": 0.0023825191892683506,
"step": 891
},
{
"epoch": 1.4824789435374857,
"grad_norm": 0.00315672904253006,
"learning_rate": 3.727129732447322e-06,
"loss": 0.004341416992247105,
"step": 892
},
{
"epoch": 1.4841426640324427,
"grad_norm": 0.0012259150389581919,
"learning_rate": 3.7241365656496103e-06,
"loss": 0.0016382039757445455,
"step": 893
},
{
"epoch": 1.4858063845273994,
"grad_norm": 0.0030357290524989367,
"learning_rate": 3.7211410889304117e-06,
"loss": 0.003687058575451374,
"step": 894
},
{
"epoch": 1.4874701050223562,
"grad_norm": 0.002371704438701272,
"learning_rate": 3.7181433079421316e-06,
"loss": 0.0053842682391405106,
"step": 895
},
{
"epoch": 1.489133825517313,
"grad_norm": 0.0021484766621142626,
"learning_rate": 3.7151432283415244e-06,
"loss": 0.002064719330519438,
"step": 896
},
{
"epoch": 1.49079754601227,
"grad_norm": 0.0012575231958180666,
"learning_rate": 3.712140855789679e-06,
"loss": 0.002172749023884535,
"step": 897
},
{
"epoch": 1.4924612665072268,
"grad_norm": 0.0008468807209283113,
"learning_rate": 3.709136195952015e-06,
"loss": 0.002431442029774189,
"step": 898
},
{
"epoch": 1.4941249870021838,
"grad_norm": 0.0007434627623297274,
"learning_rate": 3.706129254498266e-06,
"loss": 0.0017739442409947515,
"step": 899
},
{
"epoch": 1.4957887074971405,
"grad_norm": 0.001650674152188003,
"learning_rate": 3.703120037102469e-06,
"loss": 0.0070024230517446995,
"step": 900
},
{
"epoch": 1.4957887074971405,
"eval_loss": 0.004188983701169491,
"eval_runtime": 395.6086,
"eval_samples_per_second": 12.153,
"eval_steps_per_second": 3.038,
"step": 900
},
{
"epoch": 1.4974524279920973,
"grad_norm": 0.0013096841285005212,
"learning_rate": 3.7001085494429596e-06,
"loss": 0.0030500690918415785,
"step": 901
},
{
"epoch": 1.499116148487054,
"grad_norm": 0.0012824746081605554,
"learning_rate": 3.697094797202355e-06,
"loss": 0.001866386504843831,
"step": 902
},
{
"epoch": 1.5007798689820109,
"grad_norm": 0.0028050485998392105,
"learning_rate": 3.694078786067546e-06,
"loss": 0.005148045253008604,
"step": 903
},
{
"epoch": 1.5024435894769679,
"grad_norm": 0.0022381776943802834,
"learning_rate": 3.691060521729686e-06,
"loss": 0.004079541191458702,
"step": 904
},
{
"epoch": 1.5041073099719249,
"grad_norm": 0.0034186223056167364,
"learning_rate": 3.6880400098841794e-06,
"loss": 0.00605606148019433,
"step": 905
},
{
"epoch": 1.5057710304668817,
"grad_norm": 0.0015580811304971576,
"learning_rate": 3.6850172562306735e-06,
"loss": 0.0045418585650622845,
"step": 906
},
{
"epoch": 1.5074347509618384,
"grad_norm": 0.002241261536255479,
"learning_rate": 3.681992266473044e-06,
"loss": 0.00643511675298214,
"step": 907
},
{
"epoch": 1.5090984714567952,
"grad_norm": 0.002026507630944252,
"learning_rate": 3.6789650463193864e-06,
"loss": 0.003510694019496441,
"step": 908
},
{
"epoch": 1.510762191951752,
"grad_norm": 0.001170490519143641,
"learning_rate": 3.675935601482006e-06,
"loss": 0.0016428721137344837,
"step": 909
},
{
"epoch": 1.512425912446709,
"grad_norm": 0.0025286872405558825,
"learning_rate": 3.6729039376774055e-06,
"loss": 0.003139745444059372,
"step": 910
},
{
"epoch": 1.5140896329416658,
"grad_norm": 0.002121243393048644,
"learning_rate": 3.6698700606262733e-06,
"loss": 0.0036724861711263657,
"step": 911
},
{
"epoch": 1.5157533534366228,
"grad_norm": 0.0019067934481427073,
"learning_rate": 3.6668339760534768e-06,
"loss": 0.0035516428761184216,
"step": 912
},
{
"epoch": 1.5174170739315795,
"grad_norm": 0.002206754172220826,
"learning_rate": 3.6637956896880465e-06,
"loss": 0.0029350381810218096,
"step": 913
},
{
"epoch": 1.5190807944265363,
"grad_norm": 0.0027417195960879326,
"learning_rate": 3.6607552072631685e-06,
"loss": 0.0031718441750854254,
"step": 914
},
{
"epoch": 1.520744514921493,
"grad_norm": 0.002210053149610758,
"learning_rate": 3.6577125345161748e-06,
"loss": 0.003069754224270582,
"step": 915
},
{
"epoch": 1.5224082354164499,
"grad_norm": 0.0014066651929169893,
"learning_rate": 3.6546676771885257e-06,
"loss": 0.0033890805207192898,
"step": 916
},
{
"epoch": 1.5240719559114069,
"grad_norm": 0.0011378617491573095,
"learning_rate": 3.6516206410258092e-06,
"loss": 0.002046656096354127,
"step": 917
},
{
"epoch": 1.5257356764063639,
"grad_norm": 0.0013073384761810303,
"learning_rate": 3.6485714317777223e-06,
"loss": 0.002027935115620494,
"step": 918
},
{
"epoch": 1.5273993969013206,
"grad_norm": 0.001576530165039003,
"learning_rate": 3.6455200551980605e-06,
"loss": 0.0013289693742990494,
"step": 919
},
{
"epoch": 1.5290631173962774,
"grad_norm": 0.0018064347095787525,
"learning_rate": 3.642466517044713e-06,
"loss": 0.002037541940808296,
"step": 920
},
{
"epoch": 1.5307268378912342,
"grad_norm": 0.002317433012649417,
"learning_rate": 3.6394108230796455e-06,
"loss": 0.005125693045556545,
"step": 921
},
{
"epoch": 1.532390558386191,
"grad_norm": 0.000983527977950871,
"learning_rate": 3.636352979068891e-06,
"loss": 0.0022681993432343006,
"step": 922
},
{
"epoch": 1.534054278881148,
"grad_norm": 0.0016607396537438035,
"learning_rate": 3.6332929907825426e-06,
"loss": 0.0016774425748735666,
"step": 923
},
{
"epoch": 1.535717999376105,
"grad_norm": 0.002821697387844324,
"learning_rate": 3.630230863994736e-06,
"loss": 0.003204245353117585,
"step": 924
},
{
"epoch": 1.5373817198710618,
"grad_norm": 0.0008679834427312016,
"learning_rate": 3.6271666044836433e-06,
"loss": 0.001236174488440156,
"step": 925
},
{
"epoch": 1.5390454403660185,
"grad_norm": 0.0009674489265307784,
"learning_rate": 3.624100218031464e-06,
"loss": 0.001415852690115571,
"step": 926
},
{
"epoch": 1.5407091608609753,
"grad_norm": 0.0015830635093152523,
"learning_rate": 3.621031710424407e-06,
"loss": 0.0019250408513471484,
"step": 927
},
{
"epoch": 1.542372881355932,
"grad_norm": 0.002309768460690975,
"learning_rate": 3.6179610874526856e-06,
"loss": 0.004901132546365261,
"step": 928
},
{
"epoch": 1.544036601850889,
"grad_norm": 0.0043786936439573765,
"learning_rate": 3.614888354910505e-06,
"loss": 0.0032006818801164627,
"step": 929
},
{
"epoch": 1.5457003223458459,
"grad_norm": 0.0012186798267066479,
"learning_rate": 3.6118135185960507e-06,
"loss": 0.002999824471771717,
"step": 930
},
{
"epoch": 1.5473640428408029,
"grad_norm": 0.0018106299685314298,
"learning_rate": 3.6087365843114773e-06,
"loss": 0.0023737032897770405,
"step": 931
},
{
"epoch": 1.5490277633357596,
"grad_norm": 0.0017335520824417472,
"learning_rate": 3.6056575578629006e-06,
"loss": 0.0030579883605241776,
"step": 932
},
{
"epoch": 1.5506914838307164,
"grad_norm": 0.004123002290725708,
"learning_rate": 3.6025764450603808e-06,
"loss": 0.003146671922877431,
"step": 933
},
{
"epoch": 1.5523552043256732,
"grad_norm": 0.004050532355904579,
"learning_rate": 3.5994932517179182e-06,
"loss": 0.005657460540533066,
"step": 934
},
{
"epoch": 1.55401892482063,
"grad_norm": 0.0015287046553567052,
"learning_rate": 3.596407983653436e-06,
"loss": 0.0027307921554893255,
"step": 935
},
{
"epoch": 1.555682645315587,
"grad_norm": 0.0017233432736247778,
"learning_rate": 3.5933206466887755e-06,
"loss": 0.002712239045649767,
"step": 936
},
{
"epoch": 1.557346365810544,
"grad_norm": 0.001591035514138639,
"learning_rate": 3.59023124664968e-06,
"loss": 0.004103376530110836,
"step": 937
},
{
"epoch": 1.5590100863055008,
"grad_norm": 0.0024902494624257088,
"learning_rate": 3.5871397893657867e-06,
"loss": 0.0025554909370839596,
"step": 938
},
{
"epoch": 1.5606738068004575,
"grad_norm": 0.0032940879464149475,
"learning_rate": 3.5840462806706126e-06,
"loss": 0.004046164453029633,
"step": 939
},
{
"epoch": 1.5623375272954143,
"grad_norm": 0.002541156252846122,
"learning_rate": 3.5809507264015502e-06,
"loss": 0.0029931843746453524,
"step": 940
},
{
"epoch": 1.564001247790371,
"grad_norm": 0.0018427117029204965,
"learning_rate": 3.5778531323998465e-06,
"loss": 0.003148130141198635,
"step": 941
},
{
"epoch": 1.565664968285328,
"grad_norm": 0.0030140052549540997,
"learning_rate": 3.574753504510602e-06,
"loss": 0.0034573671873658895,
"step": 942
},
{
"epoch": 1.5673286887802849,
"grad_norm": 0.001244527637027204,
"learning_rate": 3.571651848582753e-06,
"loss": 0.003539226483553648,
"step": 943
},
{
"epoch": 1.5689924092752419,
"grad_norm": 0.002971372799947858,
"learning_rate": 3.5685481704690617e-06,
"loss": 0.003306737169623375,
"step": 944
},
{
"epoch": 1.5706561297701986,
"grad_norm": 0.0023480779491364956,
"learning_rate": 3.5654424760261082e-06,
"loss": 0.0052779726684093475,
"step": 945
},
{
"epoch": 1.5723198502651554,
"grad_norm": 0.0023597951512783766,
"learning_rate": 3.5623347711142764e-06,
"loss": 0.0025963601656258106,
"step": 946
},
{
"epoch": 1.5739835707601122,
"grad_norm": 0.001831749570555985,
"learning_rate": 3.5592250615977434e-06,
"loss": 0.00302727147936821,
"step": 947
},
{
"epoch": 1.5756472912550692,
"grad_norm": 0.0016239514807239175,
"learning_rate": 3.5561133533444703e-06,
"loss": 0.0025464145001024008,
"step": 948
},
{
"epoch": 1.577311011750026,
"grad_norm": 0.00151860062032938,
"learning_rate": 3.552999652226189e-06,
"loss": 0.0034899080637842417,
"step": 949
},
{
"epoch": 1.578974732244983,
"grad_norm": 0.0012699831277132034,
"learning_rate": 3.549883964118392e-06,
"loss": 0.001983851194381714,
"step": 950
},
{
"epoch": 1.5806384527399397,
"grad_norm": 0.0034781943541020155,
"learning_rate": 3.54676629490032e-06,
"loss": 0.003068584017455578,
"step": 951
},
{
"epoch": 1.5823021732348965,
"grad_norm": 0.002003491623327136,
"learning_rate": 3.543646650454955e-06,
"loss": 0.0037051017861813307,
"step": 952
},
{
"epoch": 1.5839658937298533,
"grad_norm": 0.0011521793203428388,
"learning_rate": 3.5405250366690023e-06,
"loss": 0.00235694064758718,
"step": 953
},
{
"epoch": 1.58562961422481,
"grad_norm": 0.0016058378387242556,
"learning_rate": 3.5374014594328877e-06,
"loss": 0.0022253813222050667,
"step": 954
},
{
"epoch": 1.587293334719767,
"grad_norm": 0.0031109461560845375,
"learning_rate": 3.5342759246407378e-06,
"loss": 0.008151844143867493,
"step": 955
},
{
"epoch": 1.588957055214724,
"grad_norm": 0.0023460695520043373,
"learning_rate": 3.5311484381903754e-06,
"loss": 0.006699217949062586,
"step": 956
},
{
"epoch": 1.5906207757096809,
"grad_norm": 0.0008898034575395286,
"learning_rate": 3.528019005983306e-06,
"loss": 0.0008910002070479095,
"step": 957
},
{
"epoch": 1.5922844962046376,
"grad_norm": 0.0015714402543380857,
"learning_rate": 3.5248876339247053e-06,
"loss": 0.002021456602960825,
"step": 958
},
{
"epoch": 1.5939482166995944,
"grad_norm": 0.0008527041645720601,
"learning_rate": 3.521754327923412e-06,
"loss": 0.0010136617347598076,
"step": 959
},
{
"epoch": 1.5956119371945512,
"grad_norm": 0.0016987391281872988,
"learning_rate": 3.5186190938919106e-06,
"loss": 0.002373310038819909,
"step": 960
},
{
"epoch": 1.5972756576895082,
"grad_norm": 0.0019705703016370535,
"learning_rate": 3.515481937746327e-06,
"loss": 0.0033961357548832893,
"step": 961
},
{
"epoch": 1.598939378184465,
"grad_norm": 0.0015687725972384214,
"learning_rate": 3.5123428654064134e-06,
"loss": 0.0030006766319274902,
"step": 962
},
{
"epoch": 1.600603098679422,
"grad_norm": 0.0016754200914874673,
"learning_rate": 3.509201882795536e-06,
"loss": 0.002887715119868517,
"step": 963
},
{
"epoch": 1.6022668191743787,
"grad_norm": 0.0008902286062948406,
"learning_rate": 3.5060589958406677e-06,
"loss": 0.0015371054178103805,
"step": 964
},
{
"epoch": 1.6039305396693355,
"grad_norm": 0.0012788430321961641,
"learning_rate": 3.5029142104723725e-06,
"loss": 0.0015881682047620416,
"step": 965
},
{
"epoch": 1.6055942601642923,
"grad_norm": 0.001669449033215642,
"learning_rate": 3.4997675326247993e-06,
"loss": 0.0019540064968168736,
"step": 966
},
{
"epoch": 1.6072579806592493,
"grad_norm": 0.0008664173074066639,
"learning_rate": 3.4966189682356677e-06,
"loss": 0.0009041058365255594,
"step": 967
},
{
"epoch": 1.608921701154206,
"grad_norm": 0.0017394538735970855,
"learning_rate": 3.493468523246255e-06,
"loss": 0.0033552986569702625,
"step": 968
},
{
"epoch": 1.610585421649163,
"grad_norm": 0.0015175569569692016,
"learning_rate": 3.4903162036013894e-06,
"loss": 0.004245993215590715,
"step": 969
},
{
"epoch": 1.6122491421441199,
"grad_norm": 0.0025712379720062017,
"learning_rate": 3.487162015249436e-06,
"loss": 0.006007185205817223,
"step": 970
},
{
"epoch": 1.6139128626390766,
"grad_norm": 0.0024991503451019526,
"learning_rate": 3.484005964142285e-06,
"loss": 0.0040008206851780415,
"step": 971
},
{
"epoch": 1.6155765831340334,
"grad_norm": 0.0014480374520644546,
"learning_rate": 3.4808480562353426e-06,
"loss": 0.0017751908162608743,
"step": 972
},
{
"epoch": 1.6172403036289902,
"grad_norm": 0.0022420494351536036,
"learning_rate": 3.477688297487519e-06,
"loss": 0.002922196639701724,
"step": 973
},
{
"epoch": 1.6189040241239472,
"grad_norm": 0.0012288083089515567,
"learning_rate": 3.474526693861216e-06,
"loss": 0.0026381013449281454,
"step": 974
},
{
"epoch": 1.6205677446189042,
"grad_norm": 0.00684167118743062,
"learning_rate": 3.4713632513223178e-06,
"loss": 0.009229960851371288,
"step": 975
},
{
"epoch": 1.622231465113861,
"grad_norm": 0.0032470698934048414,
"learning_rate": 3.4681979758401767e-06,
"loss": 0.00305845495313406,
"step": 976
},
{
"epoch": 1.6238951856088177,
"grad_norm": 0.001162294065579772,
"learning_rate": 3.465030873387606e-06,
"loss": 0.0036955790128558874,
"step": 977
},
{
"epoch": 1.6255589061037745,
"grad_norm": 0.0008722259080968797,
"learning_rate": 3.461861949940865e-06,
"loss": 0.0015360023826360703,
"step": 978
},
{
"epoch": 1.6272226265987313,
"grad_norm": 0.0019653120543807745,
"learning_rate": 3.458691211479649e-06,
"loss": 0.0034643567632883787,
"step": 979
},
{
"epoch": 1.6288863470936883,
"grad_norm": 0.0019016037695109844,
"learning_rate": 3.4555186639870795e-06,
"loss": 0.00340990349650383,
"step": 980
},
{
"epoch": 1.630550067588645,
"grad_norm": 0.002036824356764555,
"learning_rate": 3.4523443134496916e-06,
"loss": 0.0035561122931540012,
"step": 981
},
{
"epoch": 1.632213788083602,
"grad_norm": 0.0014058734523132443,
"learning_rate": 3.4491681658574205e-06,
"loss": 0.0019305794266983867,
"step": 982
},
{
"epoch": 1.6338775085785588,
"grad_norm": 0.0015478340210393071,
"learning_rate": 3.445990227203594e-06,
"loss": 0.0024740060325711966,
"step": 983
},
{
"epoch": 1.6355412290735156,
"grad_norm": 0.0011504017747938633,
"learning_rate": 3.442810503484921e-06,
"loss": 0.0033877063542604446,
"step": 984
},
{
"epoch": 1.6372049495684724,
"grad_norm": 0.002352361800149083,
"learning_rate": 3.4396290007014752e-06,
"loss": 0.0032197399996221066,
"step": 985
},
{
"epoch": 1.6388686700634292,
"grad_norm": 0.0008741469937376678,
"learning_rate": 3.4364457248566913e-06,
"loss": 0.0018215301679447293,
"step": 986
},
{
"epoch": 1.6405323905583862,
"grad_norm": 0.0013750221114605665,
"learning_rate": 3.433260681957346e-06,
"loss": 0.00448352936655283,
"step": 987
},
{
"epoch": 1.6421961110533432,
"grad_norm": 0.0018812837079167366,
"learning_rate": 3.430073878013554e-06,
"loss": 0.002898102393373847,
"step": 988
},
{
"epoch": 1.6438598315483,
"grad_norm": 0.0017209555953741074,
"learning_rate": 3.4268853190387496e-06,
"loss": 0.003182763233780861,
"step": 989
},
{
"epoch": 1.6455235520432567,
"grad_norm": 0.00356616685166955,
"learning_rate": 3.423695011049683e-06,
"loss": 0.006524332799017429,
"step": 990
},
{
"epoch": 1.6471872725382135,
"grad_norm": 0.0020170877687633038,
"learning_rate": 3.4205029600663996e-06,
"loss": 0.0026602193247526884,
"step": 991
},
{
"epoch": 1.6488509930331703,
"grad_norm": 0.0013372714165598154,
"learning_rate": 3.4173091721122375e-06,
"loss": 0.0021778345108032227,
"step": 992
},
{
"epoch": 1.6505147135281273,
"grad_norm": 0.0016821332974359393,
"learning_rate": 3.414113653213812e-06,
"loss": 0.0026598330587148666,
"step": 993
},
{
"epoch": 1.6521784340230843,
"grad_norm": 0.0012136506848037243,
"learning_rate": 3.410916409401004e-06,
"loss": 0.0019583716057240963,
"step": 994
},
{
"epoch": 1.653842154518041,
"grad_norm": 0.003831377485767007,
"learning_rate": 3.407717446706948e-06,
"loss": 0.00612489553168416,
"step": 995
},
{
"epoch": 1.6555058750129978,
"grad_norm": 0.002008938929066062,
"learning_rate": 3.4045167711680244e-06,
"loss": 0.0036759148351848125,
"step": 996
},
{
"epoch": 1.6571695955079546,
"grad_norm": 0.0018878724658861756,
"learning_rate": 3.4013143888238455e-06,
"loss": 0.0036940951831638813,
"step": 997
},
{
"epoch": 1.6588333160029114,
"grad_norm": 0.0028713566716760397,
"learning_rate": 3.398110305717241e-06,
"loss": 0.0030482043512165546,
"step": 998
},
{
"epoch": 1.6604970364978684,
"grad_norm": 0.0028156277257949114,
"learning_rate": 3.3949045278942545e-06,
"loss": 0.0030587539076805115,
"step": 999
},
{
"epoch": 1.6621607569928252,
"grad_norm": 0.0009347902378067374,
"learning_rate": 3.3916970614041244e-06,
"loss": 0.0013425936922430992,
"step": 1000
},
{
"epoch": 1.6621607569928252,
"eval_loss": 0.0032257065176963806,
"eval_runtime": 490.8151,
"eval_samples_per_second": 9.796,
"eval_steps_per_second": 2.449,
"step": 1000
},
{
"epoch": 1.6638244774877822,
"grad_norm": 0.001645615790039301,
"learning_rate": 3.3884879122992762e-06,
"loss": 0.004788869060575962,
"step": 1001
},
{
"epoch": 1.665488197982739,
"grad_norm": 0.0014216083800420165,
"learning_rate": 3.3852770866353125e-06,
"loss": 0.00593174621462822,
"step": 1002
},
{
"epoch": 1.6671519184776957,
"grad_norm": 0.002447444712743163,
"learning_rate": 3.382064590470996e-06,
"loss": 0.0037248479202389717,
"step": 1003
},
{
"epoch": 1.6688156389726525,
"grad_norm": 0.002557245083153248,
"learning_rate": 3.378850429868244e-06,
"loss": 0.0015867094043642282,
"step": 1004
},
{
"epoch": 1.6704793594676093,
"grad_norm": 0.0016936559695750475,
"learning_rate": 3.3756346108921145e-06,
"loss": 0.00401366176083684,
"step": 1005
},
{
"epoch": 1.6721430799625663,
"grad_norm": 0.0017694927519187331,
"learning_rate": 3.372417139610793e-06,
"loss": 0.004327396862208843,
"step": 1006
},
{
"epoch": 1.6738068004575233,
"grad_norm": 0.002197755267843604,
"learning_rate": 3.369198022095585e-06,
"loss": 0.002049348782747984,
"step": 1007
},
{
"epoch": 1.67547052095248,
"grad_norm": 0.0014030742458999157,
"learning_rate": 3.3659772644209023e-06,
"loss": 0.002163403434678912,
"step": 1008
},
{
"epoch": 1.6771342414474368,
"grad_norm": 0.001984333386644721,
"learning_rate": 3.36275487266425e-06,
"loss": 0.00507196132093668,
"step": 1009
},
{
"epoch": 1.6787979619423936,
"grad_norm": 0.0017958490643650293,
"learning_rate": 3.3595308529062176e-06,
"loss": 0.003381962887942791,
"step": 1010
},
{
"epoch": 1.6804616824373504,
"grad_norm": 0.0017129637999460101,
"learning_rate": 3.3563052112304674e-06,
"loss": 0.007319668307900429,
"step": 1011
},
{
"epoch": 1.6821254029323074,
"grad_norm": 0.0027047405019402504,
"learning_rate": 3.3530779537237194e-06,
"loss": 0.004077594727277756,
"step": 1012
},
{
"epoch": 1.6837891234272642,
"grad_norm": 0.0012556860456243157,
"learning_rate": 3.349849086475747e-06,
"loss": 0.0024214675650000572,
"step": 1013
},
{
"epoch": 1.6854528439222212,
"grad_norm": 0.002287906827405095,
"learning_rate": 3.346618615579359e-06,
"loss": 0.0034168255515396595,
"step": 1014
},
{
"epoch": 1.687116564417178,
"grad_norm": 0.0010098470374941826,
"learning_rate": 3.3433865471303876e-06,
"loss": 0.0009828556794673204,
"step": 1015
},
{
"epoch": 1.6887802849121347,
"grad_norm": 0.0012711078161373734,
"learning_rate": 3.3401528872276847e-06,
"loss": 0.0018059266731142998,
"step": 1016
},
{
"epoch": 1.6904440054070915,
"grad_norm": 0.0008985710446722806,
"learning_rate": 3.3369176419731004e-06,
"loss": 0.0007820754544809461,
"step": 1017
},
{
"epoch": 1.6921077259020485,
"grad_norm": 0.0005346555262804031,
"learning_rate": 3.33368081747148e-06,
"loss": 0.0012233923189342022,
"step": 1018
},
{
"epoch": 1.6937714463970053,
"grad_norm": 0.002259831642732024,
"learning_rate": 3.3304424198306464e-06,
"loss": 0.006927793845534325,
"step": 1019
},
{
"epoch": 1.6954351668919623,
"grad_norm": 0.0024394611828029156,
"learning_rate": 3.3272024551613926e-06,
"loss": 0.0051927645690739155,
"step": 1020
},
{
"epoch": 1.697098887386919,
"grad_norm": 0.0016627139411866665,
"learning_rate": 3.3239609295774667e-06,
"loss": 0.0044672624208033085,
"step": 1021
},
{
"epoch": 1.6987626078818758,
"grad_norm": 0.002487657591700554,
"learning_rate": 3.3207178491955656e-06,
"loss": 0.005251655355095863,
"step": 1022
},
{
"epoch": 1.7004263283768326,
"grad_norm": 0.0012484157923609018,
"learning_rate": 3.3174732201353155e-06,
"loss": 0.0025723387952893972,
"step": 1023
},
{
"epoch": 1.7020900488717894,
"grad_norm": 0.0019987961277365685,
"learning_rate": 3.3142270485192683e-06,
"loss": 0.0029333368875086308,
"step": 1024
},
{
"epoch": 1.7037537693667464,
"grad_norm": 0.0024946676567196846,
"learning_rate": 3.3109793404728855e-06,
"loss": 0.002550342585891485,
"step": 1025
},
{
"epoch": 1.7054174898617034,
"grad_norm": 0.0016121126245707273,
"learning_rate": 3.3077301021245285e-06,
"loss": 0.0028350313659757376,
"step": 1026
},
{
"epoch": 1.7070812103566602,
"grad_norm": 0.0029108431190252304,
"learning_rate": 3.3044793396054447e-06,
"loss": 0.004767598584294319,
"step": 1027
},
{
"epoch": 1.708744930851617,
"grad_norm": 0.001531237387098372,
"learning_rate": 3.3012270590497596e-06,
"loss": 0.00230689300224185,
"step": 1028
},
{
"epoch": 1.7104086513465737,
"grad_norm": 0.0027158576995134354,
"learning_rate": 3.2979732665944615e-06,
"loss": 0.007252411916851997,
"step": 1029
},
{
"epoch": 1.7120723718415305,
"grad_norm": 0.001231711357831955,
"learning_rate": 3.2947179683793928e-06,
"loss": 0.001815581344999373,
"step": 1030
},
{
"epoch": 1.7137360923364875,
"grad_norm": 0.002293652854859829,
"learning_rate": 3.291461170547237e-06,
"loss": 0.004373159259557724,
"step": 1031
},
{
"epoch": 1.7153998128314443,
"grad_norm": 0.0017853020690381527,
"learning_rate": 3.2882028792435072e-06,
"loss": 0.004501050338149071,
"step": 1032
},
{
"epoch": 1.7170635333264013,
"grad_norm": 0.0012384027941152453,
"learning_rate": 3.2849431006165343e-06,
"loss": 0.0017943148268386722,
"step": 1033
},
{
"epoch": 1.718727253821358,
"grad_norm": 0.0015722275711596012,
"learning_rate": 3.2816818408174567e-06,
"loss": 0.002293921774253249,
"step": 1034
},
{
"epoch": 1.7203909743163148,
"grad_norm": 0.0022003373596817255,
"learning_rate": 3.278419106000206e-06,
"loss": 0.0037125535309314728,
"step": 1035
},
{
"epoch": 1.7220546948112716,
"grad_norm": 0.0007672826759517193,
"learning_rate": 3.2751549023214995e-06,
"loss": 0.0020989507902413607,
"step": 1036
},
{
"epoch": 1.7237184153062286,
"grad_norm": 0.0009453165694139898,
"learning_rate": 3.2718892359408245e-06,
"loss": 0.003136922139674425,
"step": 1037
},
{
"epoch": 1.7253821358011854,
"grad_norm": 0.0012431323993951082,
"learning_rate": 3.2686221130204287e-06,
"loss": 0.0036514741368591785,
"step": 1038
},
{
"epoch": 1.7270458562961424,
"grad_norm": 0.0009162081405520439,
"learning_rate": 3.265353539725309e-06,
"loss": 0.0027881120331585407,
"step": 1039
},
{
"epoch": 1.7287095767910992,
"grad_norm": 0.0009822389110922813,
"learning_rate": 3.2620835222231972e-06,
"loss": 0.002034775447100401,
"step": 1040
},
{
"epoch": 1.730373297286056,
"grad_norm": 0.001126887509599328,
"learning_rate": 3.2588120666845534e-06,
"loss": 0.003957305569201708,
"step": 1041
},
{
"epoch": 1.7320370177810127,
"grad_norm": 0.001378571498207748,
"learning_rate": 3.255539179282548e-06,
"loss": 0.0029517512302845716,
"step": 1042
},
{
"epoch": 1.7337007382759695,
"grad_norm": 0.0016041906783357263,
"learning_rate": 3.2522648661930558e-06,
"loss": 0.0043833935633301735,
"step": 1043
},
{
"epoch": 1.7353644587709265,
"grad_norm": 0.0011807052651420236,
"learning_rate": 3.2489891335946413e-06,
"loss": 0.0019419104792177677,
"step": 1044
},
{
"epoch": 1.7370281792658835,
"grad_norm": 0.0009768899763002992,
"learning_rate": 3.245711987668545e-06,
"loss": 0.002313793171197176,
"step": 1045
},
{
"epoch": 1.7386918997608403,
"grad_norm": 0.0015365976141765714,
"learning_rate": 3.2424334345986787e-06,
"loss": 0.002823270158842206,
"step": 1046
},
{
"epoch": 1.740355620255797,
"grad_norm": 0.0016022294294089079,
"learning_rate": 3.239153480571605e-06,
"loss": 0.0011497450759634376,
"step": 1047
},
{
"epoch": 1.7420193407507538,
"grad_norm": 0.002018695930019021,
"learning_rate": 3.2358721317765344e-06,
"loss": 0.0035544075071811676,
"step": 1048
},
{
"epoch": 1.7436830612457106,
"grad_norm": 0.002275625476613641,
"learning_rate": 3.2325893944053066e-06,
"loss": 0.0036176592111587524,
"step": 1049
},
{
"epoch": 1.7453467817406676,
"grad_norm": 0.0009655337198637426,
"learning_rate": 3.2293052746523814e-06,
"loss": 0.0015187448589131236,
"step": 1050
},
{
"epoch": 1.7470105022356244,
"grad_norm": 0.0012670238502323627,
"learning_rate": 3.2260197787148277e-06,
"loss": 0.0028458700980991125,
"step": 1051
},
{
"epoch": 1.7486742227305814,
"grad_norm": 0.002225345466285944,
"learning_rate": 3.222732912792313e-06,
"loss": 0.0032379874028265476,
"step": 1052
},
{
"epoch": 1.7503379432255382,
"grad_norm": 0.00402183597907424,
"learning_rate": 3.2194446830870865e-06,
"loss": 0.0032164682634174824,
"step": 1053
},
{
"epoch": 1.752001663720495,
"grad_norm": 0.0012006743345409632,
"learning_rate": 3.2161550958039732e-06,
"loss": 0.004632690455764532,
"step": 1054
},
{
"epoch": 1.7536653842154517,
"grad_norm": 0.005980351008474827,
"learning_rate": 3.2128641571503594e-06,
"loss": 0.003166868118569255,
"step": 1055
},
{
"epoch": 1.7553291047104087,
"grad_norm": 0.002437678398564458,
"learning_rate": 3.2095718733361803e-06,
"loss": 0.008781610988080502,
"step": 1056
},
{
"epoch": 1.7569928252053655,
"grad_norm": 0.0013743985909968615,
"learning_rate": 3.2062782505739125e-06,
"loss": 0.0027862025890499353,
"step": 1057
},
{
"epoch": 1.7586565457003225,
"grad_norm": 0.0014083994319662452,
"learning_rate": 3.202983295078555e-06,
"loss": 0.004423089791089296,
"step": 1058
},
{
"epoch": 1.7603202661952793,
"grad_norm": 0.0012103941990062594,
"learning_rate": 3.199687013067624e-06,
"loss": 0.0016060488997027278,
"step": 1059
},
{
"epoch": 1.761983986690236,
"grad_norm": 0.001576002687215805,
"learning_rate": 3.1963894107611395e-06,
"loss": 0.002927867928519845,
"step": 1060
},
{
"epoch": 1.7636477071851928,
"grad_norm": 0.0017253441037610173,
"learning_rate": 3.1930904943816104e-06,
"loss": 0.0021498226560652256,
"step": 1061
},
{
"epoch": 1.7653114276801496,
"grad_norm": 0.0037647176068276167,
"learning_rate": 3.189790270154028e-06,
"loss": 0.004621579311788082,
"step": 1062
},
{
"epoch": 1.7669751481751066,
"grad_norm": 0.0019720825366675854,
"learning_rate": 3.186488744305849e-06,
"loss": 0.0062108938582241535,
"step": 1063
},
{
"epoch": 1.7686388686700636,
"grad_norm": 0.0019118713680654764,
"learning_rate": 3.183185923066988e-06,
"loss": 0.003158648731186986,
"step": 1064
},
{
"epoch": 1.7703025891650204,
"grad_norm": 0.0035292403772473335,
"learning_rate": 3.179881812669804e-06,
"loss": 0.003806828986853361,
"step": 1065
},
{
"epoch": 1.7719663096599771,
"grad_norm": 0.0009236359037458897,
"learning_rate": 3.1765764193490863e-06,
"loss": 0.0024340436793863773,
"step": 1066
},
{
"epoch": 1.773630030154934,
"grad_norm": 0.0012912286911159754,
"learning_rate": 3.173269749342047e-06,
"loss": 0.0013527884148061275,
"step": 1067
},
{
"epoch": 1.7752937506498907,
"grad_norm": 0.0019723335281014442,
"learning_rate": 3.1699618088883094e-06,
"loss": 0.0035508442670106888,
"step": 1068
},
{
"epoch": 1.7769574711448477,
"grad_norm": 0.0014623295282945037,
"learning_rate": 3.1666526042298883e-06,
"loss": 0.00359426811337471,
"step": 1069
},
{
"epoch": 1.7786211916398045,
"grad_norm": 0.0019803629256784916,
"learning_rate": 3.16334214161119e-06,
"loss": 0.0036465353332459927,
"step": 1070
},
{
"epoch": 1.7802849121347615,
"grad_norm": 0.0022365187760442495,
"learning_rate": 3.1600304272789904e-06,
"loss": 0.002207161160185933,
"step": 1071
},
{
"epoch": 1.7819486326297183,
"grad_norm": 0.0012209509732201695,
"learning_rate": 3.1567174674824303e-06,
"loss": 0.0025660775136202574,
"step": 1072
},
{
"epoch": 1.783612353124675,
"grad_norm": 0.0028881127946078777,
"learning_rate": 3.1534032684729978e-06,
"loss": 0.004914230667054653,
"step": 1073
},
{
"epoch": 1.7852760736196318,
"grad_norm": 0.001741714426316321,
"learning_rate": 3.1500878365045217e-06,
"loss": 0.0034641423262655735,
"step": 1074
},
{
"epoch": 1.7869397941145886,
"grad_norm": 0.00211711716838181,
"learning_rate": 3.1467711778331573e-06,
"loss": 0.004626632202416658,
"step": 1075
},
{
"epoch": 1.7886035146095456,
"grad_norm": 0.002563687739893794,
"learning_rate": 3.143453298717373e-06,
"loss": 0.00369662931188941,
"step": 1076
},
{
"epoch": 1.7902672351045026,
"grad_norm": 0.001437278580851853,
"learning_rate": 3.14013420541794e-06,
"loss": 0.0028063696809113026,
"step": 1077
},
{
"epoch": 1.7919309555994594,
"grad_norm": 0.003704370930790901,
"learning_rate": 3.1368139041979235e-06,
"loss": 0.0021443888545036316,
"step": 1078
},
{
"epoch": 1.7935946760944161,
"grad_norm": 0.0024365736171603203,
"learning_rate": 3.133492401322666e-06,
"loss": 0.0016761664301156998,
"step": 1079
},
{
"epoch": 1.795258396589373,
"grad_norm": 0.001835328177548945,
"learning_rate": 3.1301697030597772e-06,
"loss": 0.003327090060338378,
"step": 1080
},
{
"epoch": 1.7969221170843297,
"grad_norm": 0.0011926194420084357,
"learning_rate": 3.126845815679123e-06,
"loss": 0.002464383840560913,
"step": 1081
},
{
"epoch": 1.7985858375792867,
"grad_norm": 0.0010406651999801397,
"learning_rate": 3.1235207454528137e-06,
"loss": 0.0008981494465842843,
"step": 1082
},
{
"epoch": 1.8002495580742435,
"grad_norm": 0.0010186567669734359,
"learning_rate": 3.12019449865519e-06,
"loss": 0.003182903630658984,
"step": 1083
},
{
"epoch": 1.8019132785692005,
"grad_norm": 0.0012448715278878808,
"learning_rate": 3.116867081562815e-06,
"loss": 0.0023764080833643675,
"step": 1084
},
{
"epoch": 1.8035769990641572,
"grad_norm": 0.007940222509205341,
"learning_rate": 3.1135385004544584e-06,
"loss": 0.0053514945320785046,
"step": 1085
},
{
"epoch": 1.805240719559114,
"grad_norm": 0.0014136300887912512,
"learning_rate": 3.1102087616110866e-06,
"loss": 0.00209840782918036,
"step": 1086
},
{
"epoch": 1.8069044400540708,
"grad_norm": 0.002945336978882551,
"learning_rate": 3.1068778713158515e-06,
"loss": 0.002496186876669526,
"step": 1087
},
{
"epoch": 1.8085681605490278,
"grad_norm": 0.001605421188287437,
"learning_rate": 3.1035458358540764e-06,
"loss": 0.002253869315609336,
"step": 1088
},
{
"epoch": 1.8102318810439846,
"grad_norm": 0.0010100161889567971,
"learning_rate": 3.100212661513247e-06,
"loss": 0.001364353229291737,
"step": 1089
},
{
"epoch": 1.8118956015389416,
"grad_norm": 0.001074383850209415,
"learning_rate": 3.096878354582998e-06,
"loss": 0.0016808705404400826,
"step": 1090
},
{
"epoch": 1.8135593220338984,
"grad_norm": 0.0030826677102595568,
"learning_rate": 3.093542921355099e-06,
"loss": 0.0047744084149599075,
"step": 1091
},
{
"epoch": 1.8152230425288551,
"grad_norm": 0.0025080074556171894,
"learning_rate": 3.0902063681234473e-06,
"loss": 0.0018723970279097557,
"step": 1092
},
{
"epoch": 1.816886763023812,
"grad_norm": 0.0015788818709552288,
"learning_rate": 3.086868701184054e-06,
"loss": 0.001651267521083355,
"step": 1093
},
{
"epoch": 1.8185504835187687,
"grad_norm": 0.0016100323991850019,
"learning_rate": 3.083529926835028e-06,
"loss": 0.003768754657357931,
"step": 1094
},
{
"epoch": 1.8202142040137257,
"grad_norm": 0.003728320589289069,
"learning_rate": 3.0801900513765732e-06,
"loss": 0.0027151263784617186,
"step": 1095
},
{
"epoch": 1.8218779245086827,
"grad_norm": 0.0030173619743436575,
"learning_rate": 3.076849081110967e-06,
"loss": 0.003926532343029976,
"step": 1096
},
{
"epoch": 1.8235416450036395,
"grad_norm": 0.002205587923526764,
"learning_rate": 3.073507022342554e-06,
"loss": 0.002268116921186447,
"step": 1097
},
{
"epoch": 1.8252053654985962,
"grad_norm": 0.002099486766383052,
"learning_rate": 3.070163881377734e-06,
"loss": 0.0035694725811481476,
"step": 1098
},
{
"epoch": 1.826869085993553,
"grad_norm": 0.0023339507170021534,
"learning_rate": 3.066819664524947e-06,
"loss": 0.0021657149773091078,
"step": 1099
},
{
"epoch": 1.8285328064885098,
"grad_norm": 0.0017025353154167533,
"learning_rate": 3.063474378094665e-06,
"loss": 0.001973213627934456,
"step": 1100
},
{
"epoch": 1.8285328064885098,
"eval_loss": 0.0031945928931236267,
"eval_runtime": 492.7979,
"eval_samples_per_second": 9.757,
"eval_steps_per_second": 2.439,
"step": 1100
},
{
"epoch": 1.8301965269834668,
"grad_norm": 0.001167317503131926,
"learning_rate": 3.060128028399376e-06,
"loss": 0.002630096860229969,
"step": 1101
},
{
"epoch": 1.8318602474784236,
"grad_norm": 0.001953288447111845,
"learning_rate": 3.056780621753577e-06,
"loss": 0.0033156988210976124,
"step": 1102
},
{
"epoch": 1.8335239679733806,
"grad_norm": 0.0020119582768529654,
"learning_rate": 3.0534321644737574e-06,
"loss": 0.001855833805166185,
"step": 1103
},
{
"epoch": 1.8351876884683374,
"grad_norm": 0.002142886398360133,
"learning_rate": 3.0500826628783903e-06,
"loss": 0.002585946349427104,
"step": 1104
},
{
"epoch": 1.8368514089632941,
"grad_norm": 0.0028928187675774097,
"learning_rate": 3.046732123287918e-06,
"loss": 0.005204905290156603,
"step": 1105
},
{
"epoch": 1.838515129458251,
"grad_norm": 0.0011265117209404707,
"learning_rate": 3.043380552024744e-06,
"loss": 0.0032711224630475044,
"step": 1106
},
{
"epoch": 1.840178849953208,
"grad_norm": 0.0027929674834012985,
"learning_rate": 3.0400279554132157e-06,
"loss": 0.0024414071813225746,
"step": 1107
},
{
"epoch": 1.8418425704481647,
"grad_norm": 0.0014338293112814426,
"learning_rate": 3.0366743397796166e-06,
"loss": 0.002678456250578165,
"step": 1108
},
{
"epoch": 1.8435062909431217,
"grad_norm": 0.0025491593405604362,
"learning_rate": 3.033319711452154e-06,
"loss": 0.0031573877204209566,
"step": 1109
},
{
"epoch": 1.8451700114380785,
"grad_norm": 0.0013114232569932938,
"learning_rate": 3.0299640767609447e-06,
"loss": 0.002688421867787838,
"step": 1110
},
{
"epoch": 1.8468337319330352,
"grad_norm": 0.002950844122096896,
"learning_rate": 3.0266074420380043e-06,
"loss": 0.0028441757895052433,
"step": 1111
},
{
"epoch": 1.848497452427992,
"grad_norm": 0.002961844904348254,
"learning_rate": 3.023249813617238e-06,
"loss": 0.0035082674585282803,
"step": 1112
},
{
"epoch": 1.8501611729229488,
"grad_norm": 0.001897338661365211,
"learning_rate": 3.0198911978344213e-06,
"loss": 0.0033026670571416616,
"step": 1113
},
{
"epoch": 1.8518248934179058,
"grad_norm": 0.0018700726795941591,
"learning_rate": 3.0165316010271982e-06,
"loss": 0.0023832167498767376,
"step": 1114
},
{
"epoch": 1.8534886139128628,
"grad_norm": 0.0016329766949638724,
"learning_rate": 3.0131710295350615e-06,
"loss": 0.0044321902096271515,
"step": 1115
},
{
"epoch": 1.8551523344078196,
"grad_norm": 0.00122320675291121,
"learning_rate": 3.0098094896993413e-06,
"loss": 0.0025666982401162386,
"step": 1116
},
{
"epoch": 1.8568160549027763,
"grad_norm": 0.0013701791176572442,
"learning_rate": 3.0064469878631986e-06,
"loss": 0.0031274936627596617,
"step": 1117
},
{
"epoch": 1.8584797753977331,
"grad_norm": 0.0023116201627999544,
"learning_rate": 3.003083530371606e-06,
"loss": 0.0033771577291190624,
"step": 1118
},
{
"epoch": 1.86014349589269,
"grad_norm": 0.0016470644623041153,
"learning_rate": 2.9997191235713435e-06,
"loss": 0.004839047789573669,
"step": 1119
},
{
"epoch": 1.861807216387647,
"grad_norm": 0.002336474135518074,
"learning_rate": 2.9963537738109783e-06,
"loss": 0.008031599223613739,
"step": 1120
},
{
"epoch": 1.8634709368826037,
"grad_norm": 0.0022251568734645844,
"learning_rate": 2.9929874874408595e-06,
"loss": 0.0022900504991412163,
"step": 1121
},
{
"epoch": 1.8651346573775607,
"grad_norm": 0.00211269105784595,
"learning_rate": 2.9896202708131027e-06,
"loss": 0.00309790950268507,
"step": 1122
},
{
"epoch": 1.8667983778725175,
"grad_norm": 0.002572975354269147,
"learning_rate": 2.98625213028158e-06,
"loss": 0.0036967257037758827,
"step": 1123
},
{
"epoch": 1.8684620983674742,
"grad_norm": 0.0022551484871655703,
"learning_rate": 2.9828830722019046e-06,
"loss": 0.00239252345636487,
"step": 1124
},
{
"epoch": 1.870125818862431,
"grad_norm": 0.0015825015725567937,
"learning_rate": 2.979513102931424e-06,
"loss": 0.002202708274126053,
"step": 1125
},
{
"epoch": 1.871789539357388,
"grad_norm": 0.0018934940453618765,
"learning_rate": 2.9761422288292017e-06,
"loss": 0.002752728760242462,
"step": 1126
},
{
"epoch": 1.8734532598523448,
"grad_norm": 0.002726193517446518,
"learning_rate": 2.9727704562560124e-06,
"loss": 0.003436418017372489,
"step": 1127
},
{
"epoch": 1.8751169803473018,
"grad_norm": 0.0008292084676213562,
"learning_rate": 2.9693977915743227e-06,
"loss": 0.0028455648571252823,
"step": 1128
},
{
"epoch": 1.8767807008422586,
"grad_norm": 0.00249700085259974,
"learning_rate": 2.9660242411482848e-06,
"loss": 0.0030478155240416527,
"step": 1129
},
{
"epoch": 1.8784444213372153,
"grad_norm": 0.0016656204825267196,
"learning_rate": 2.9626498113437215e-06,
"loss": 0.0029142191633582115,
"step": 1130
},
{
"epoch": 1.8801081418321721,
"grad_norm": 0.0011652596294879913,
"learning_rate": 2.9592745085281154e-06,
"loss": 0.0027298261411488056,
"step": 1131
},
{
"epoch": 1.881771862327129,
"grad_norm": 0.00196881964802742,
"learning_rate": 2.955898339070596e-06,
"loss": 0.0025863386690616608,
"step": 1132
},
{
"epoch": 1.883435582822086,
"grad_norm": 0.0018504951149225235,
"learning_rate": 2.9525213093419275e-06,
"loss": 0.002715812064707279,
"step": 1133
},
{
"epoch": 1.885099303317043,
"grad_norm": 0.0023180446587502956,
"learning_rate": 2.9491434257144995e-06,
"loss": 0.0040229130536317825,
"step": 1134
},
{
"epoch": 1.8867630238119997,
"grad_norm": 0.002672442002221942,
"learning_rate": 2.9457646945623107e-06,
"loss": 0.0018784047570079565,
"step": 1135
},
{
"epoch": 1.8884267443069565,
"grad_norm": 0.0018968256190419197,
"learning_rate": 2.9423851222609607e-06,
"loss": 0.002349269110709429,
"step": 1136
},
{
"epoch": 1.8900904648019132,
"grad_norm": 0.0006651251460425556,
"learning_rate": 2.939004715187635e-06,
"loss": 0.0016391770914196968,
"step": 1137
},
{
"epoch": 1.89175418529687,
"grad_norm": 0.0026850176509469748,
"learning_rate": 2.935623479721095e-06,
"loss": 0.004446036648005247,
"step": 1138
},
{
"epoch": 1.893417905791827,
"grad_norm": 0.001719061634503305,
"learning_rate": 2.932241422241665e-06,
"loss": 0.00394339207559824,
"step": 1139
},
{
"epoch": 1.8950816262867838,
"grad_norm": 0.0012696747435256839,
"learning_rate": 2.9288585491312206e-06,
"loss": 0.0010805041529238224,
"step": 1140
},
{
"epoch": 1.8967453467817408,
"grad_norm": 0.0013430069666355848,
"learning_rate": 2.925474866773176e-06,
"loss": 0.0016727771144360304,
"step": 1141
},
{
"epoch": 1.8984090672766976,
"grad_norm": 0.0015959268203005195,
"learning_rate": 2.922090381552475e-06,
"loss": 0.0031905556097626686,
"step": 1142
},
{
"epoch": 1.9000727877716543,
"grad_norm": 0.0014629323268309236,
"learning_rate": 2.9187050998555715e-06,
"loss": 0.002370205009356141,
"step": 1143
},
{
"epoch": 1.9017365082666111,
"grad_norm": 0.0007864710642024875,
"learning_rate": 2.915319028070427e-06,
"loss": 0.0023790618870407343,
"step": 1144
},
{
"epoch": 1.903400228761568,
"grad_norm": 0.0021873966325074434,
"learning_rate": 2.9119321725864914e-06,
"loss": 0.00319493655115366,
"step": 1145
},
{
"epoch": 1.905063949256525,
"grad_norm": 0.0016217492520809174,
"learning_rate": 2.908544539794693e-06,
"loss": 0.0021452787332236767,
"step": 1146
},
{
"epoch": 1.906727669751482,
"grad_norm": 0.0013570383889600635,
"learning_rate": 2.9051561360874297e-06,
"loss": 0.004207832273095846,
"step": 1147
},
{
"epoch": 1.9083913902464387,
"grad_norm": 0.0018747537396848202,
"learning_rate": 2.901766967858551e-06,
"loss": 0.003897372167557478,
"step": 1148
},
{
"epoch": 1.9100551107413954,
"grad_norm": 0.001689862459897995,
"learning_rate": 2.8983770415033507e-06,
"loss": 0.002389749512076378,
"step": 1149
},
{
"epoch": 1.9117188312363522,
"grad_norm": 0.0020829078275710344,
"learning_rate": 2.8949863634185533e-06,
"loss": 0.004125273786485195,
"step": 1150
},
{
"epoch": 1.913382551731309,
"grad_norm": 0.0015352623304352164,
"learning_rate": 2.8915949400022995e-06,
"loss": 0.002619950333610177,
"step": 1151
},
{
"epoch": 1.915046272226266,
"grad_norm": 0.0009680234943516552,
"learning_rate": 2.8882027776541406e-06,
"loss": 0.0014335473533719778,
"step": 1152
},
{
"epoch": 1.916709992721223,
"grad_norm": 0.0008594352402724326,
"learning_rate": 2.8848098827750186e-06,
"loss": 0.0028461103793233633,
"step": 1153
},
{
"epoch": 1.9183737132161798,
"grad_norm": 0.0011779307387769222,
"learning_rate": 2.8814162617672586e-06,
"loss": 0.0027607521042227745,
"step": 1154
},
{
"epoch": 1.9200374337111366,
"grad_norm": 0.0011340143391862512,
"learning_rate": 2.8780219210345573e-06,
"loss": 0.0025814450345933437,
"step": 1155
},
{
"epoch": 1.9217011542060933,
"grad_norm": 0.0008464404963888228,
"learning_rate": 2.8746268669819676e-06,
"loss": 0.0018523922190070152,
"step": 1156
},
{
"epoch": 1.92336487470105,
"grad_norm": 0.001362015143968165,
"learning_rate": 2.8712311060158904e-06,
"loss": 0.0011739463079720736,
"step": 1157
},
{
"epoch": 1.925028595196007,
"grad_norm": 0.0014083880232647061,
"learning_rate": 2.8678346445440588e-06,
"loss": 0.003657055553048849,
"step": 1158
},
{
"epoch": 1.9266923156909639,
"grad_norm": 0.0016203102422878146,
"learning_rate": 2.8644374889755284e-06,
"loss": 0.0018270399887114763,
"step": 1159
},
{
"epoch": 1.9283560361859209,
"grad_norm": 0.0005126083851791918,
"learning_rate": 2.861039645720664e-06,
"loss": 0.0013917206088081002,
"step": 1160
},
{
"epoch": 1.9300197566808777,
"grad_norm": 0.0013629321474581957,
"learning_rate": 2.85764112119113e-06,
"loss": 0.00415899557992816,
"step": 1161
},
{
"epoch": 1.9316834771758344,
"grad_norm": 0.0008958380785770714,
"learning_rate": 2.854241921799874e-06,
"loss": 0.0022279848344624043,
"step": 1162
},
{
"epoch": 1.9333471976707912,
"grad_norm": 0.000760416267439723,
"learning_rate": 2.850842053961119e-06,
"loss": 0.0014205033658072352,
"step": 1163
},
{
"epoch": 1.935010918165748,
"grad_norm": 0.0006858188426122069,
"learning_rate": 2.847441524090347e-06,
"loss": 0.0008534875814802945,
"step": 1164
},
{
"epoch": 1.936674638660705,
"grad_norm": 0.0022949844133108854,
"learning_rate": 2.844040338604291e-06,
"loss": 0.003076274413615465,
"step": 1165
},
{
"epoch": 1.938338359155662,
"grad_norm": 0.0010439804755151272,
"learning_rate": 2.8406385039209217e-06,
"loss": 0.004112100228667259,
"step": 1166
},
{
"epoch": 1.9400020796506188,
"grad_norm": 0.002190642524510622,
"learning_rate": 2.837236026459432e-06,
"loss": 0.007742525544017553,
"step": 1167
},
{
"epoch": 1.9416658001455755,
"grad_norm": 0.0006106036016717553,
"learning_rate": 2.833832912640232e-06,
"loss": 0.0012337150983512402,
"step": 1168
},
{
"epoch": 1.9433295206405323,
"grad_norm": 0.0016449900576844811,
"learning_rate": 2.8304291688849283e-06,
"loss": 0.002717619063332677,
"step": 1169
},
{
"epoch": 1.944993241135489,
"grad_norm": 0.0008701571496203542,
"learning_rate": 2.827024801616319e-06,
"loss": 0.002019080799072981,
"step": 1170
},
{
"epoch": 1.946656961630446,
"grad_norm": 0.0016879767645150423,
"learning_rate": 2.8236198172583765e-06,
"loss": 0.004312856588512659,
"step": 1171
},
{
"epoch": 1.9483206821254029,
"grad_norm": 0.0020904159173369408,
"learning_rate": 2.820214222236241e-06,
"loss": 0.0037140892818570137,
"step": 1172
},
{
"epoch": 1.9499844026203599,
"grad_norm": 0.002312425058335066,
"learning_rate": 2.816808022976201e-06,
"loss": 0.003156473394483328,
"step": 1173
},
{
"epoch": 1.9516481231153167,
"grad_norm": 0.00271781743504107,
"learning_rate": 2.813401225905688e-06,
"loss": 0.005529695190489292,
"step": 1174
},
{
"epoch": 1.9533118436102734,
"grad_norm": 0.001697467640042305,
"learning_rate": 2.8099938374532615e-06,
"loss": 0.004889095202088356,
"step": 1175
},
{
"epoch": 1.9549755641052302,
"grad_norm": 0.0007833896670490503,
"learning_rate": 2.806585864048594e-06,
"loss": 0.0024441727437078953,
"step": 1176
},
{
"epoch": 1.9566392846001872,
"grad_norm": 0.0011363000376150012,
"learning_rate": 2.8031773121224665e-06,
"loss": 0.0024421929847449064,
"step": 1177
},
{
"epoch": 1.958303005095144,
"grad_norm": 0.001958192326128483,
"learning_rate": 2.799768188106747e-06,
"loss": 0.0029815295711159706,
"step": 1178
},
{
"epoch": 1.959966725590101,
"grad_norm": 0.0009831757051870227,
"learning_rate": 2.7963584984343856e-06,
"loss": 0.0018389209872111678,
"step": 1179
},
{
"epoch": 1.9616304460850578,
"grad_norm": 0.0021405015140771866,
"learning_rate": 2.7929482495393995e-06,
"loss": 0.002608387963846326,
"step": 1180
},
{
"epoch": 1.9632941665800145,
"grad_norm": 0.0015074997209012508,
"learning_rate": 2.7895374478568608e-06,
"loss": 0.0015620214398950338,
"step": 1181
},
{
"epoch": 1.9649578870749713,
"grad_norm": 0.0005247404333204031,
"learning_rate": 2.786126099822885e-06,
"loss": 0.0012147929519414902,
"step": 1182
},
{
"epoch": 1.966621607569928,
"grad_norm": 0.0006081117899157107,
"learning_rate": 2.7827142118746187e-06,
"loss": 0.00113436218816787,
"step": 1183
},
{
"epoch": 1.968285328064885,
"grad_norm": 0.0008387586567550898,
"learning_rate": 2.779301790450226e-06,
"loss": 0.0012308708392083645,
"step": 1184
},
{
"epoch": 1.969949048559842,
"grad_norm": 0.003745946567505598,
"learning_rate": 2.7758888419888797e-06,
"loss": 0.012694332748651505,
"step": 1185
},
{
"epoch": 1.9716127690547989,
"grad_norm": 0.0008437780779786408,
"learning_rate": 2.7724753729307454e-06,
"loss": 0.0010596851352602243,
"step": 1186
},
{
"epoch": 1.9732764895497557,
"grad_norm": 0.0026630330830812454,
"learning_rate": 2.769061389716971e-06,
"loss": 0.0029301990289241076,
"step": 1187
},
{
"epoch": 1.9749402100447124,
"grad_norm": 0.0033500911667943,
"learning_rate": 2.765646898789677e-06,
"loss": 0.006291859783232212,
"step": 1188
},
{
"epoch": 1.9766039305396692,
"grad_norm": 0.0017212156672030687,
"learning_rate": 2.762231906591939e-06,
"loss": 0.004028333351016045,
"step": 1189
},
{
"epoch": 1.9782676510346262,
"grad_norm": 0.0007014305447228253,
"learning_rate": 2.75881641956778e-06,
"loss": 0.0016968459822237492,
"step": 1190
},
{
"epoch": 1.979931371529583,
"grad_norm": 0.0030681570060551167,
"learning_rate": 2.7554004441621562e-06,
"loss": 0.0026011753361672163,
"step": 1191
},
{
"epoch": 1.98159509202454,
"grad_norm": 0.002008343581110239,
"learning_rate": 2.7519839868209462e-06,
"loss": 0.005046532489359379,
"step": 1192
},
{
"epoch": 1.9832588125194968,
"grad_norm": 0.0019504876108840108,
"learning_rate": 2.748567053990937e-06,
"loss": 0.003577637020498514,
"step": 1193
},
{
"epoch": 1.9849225330144535,
"grad_norm": 0.002227432793006301,
"learning_rate": 2.7451496521198144e-06,
"loss": 0.005140396300703287,
"step": 1194
},
{
"epoch": 1.9865862535094103,
"grad_norm": 0.0017220035661011934,
"learning_rate": 2.741731787656146e-06,
"loss": 0.0027806637808680534,
"step": 1195
},
{
"epoch": 1.9882499740043673,
"grad_norm": 0.0010862553026527166,
"learning_rate": 2.7383134670493765e-06,
"loss": 0.0017553982324898243,
"step": 1196
},
{
"epoch": 1.989913694499324,
"grad_norm": 0.0007901210919953883,
"learning_rate": 2.734894696749808e-06,
"loss": 0.0015418350230902433,
"step": 1197
},
{
"epoch": 1.991577414994281,
"grad_norm": 0.0011503990972414613,
"learning_rate": 2.7314754832085926e-06,
"loss": 0.0023582661524415016,
"step": 1198
},
{
"epoch": 1.9932411354892379,
"grad_norm": 0.002424992388114333,
"learning_rate": 2.728055832877719e-06,
"loss": 0.0024170917458832264,
"step": 1199
},
{
"epoch": 1.9949048559841946,
"grad_norm": 0.0009666327969171107,
"learning_rate": 2.7246357522099996e-06,
"loss": 0.0011229885276407003,
"step": 1200
},
{
"epoch": 1.9949048559841946,
"eval_loss": 0.0028202177491039038,
"eval_runtime": 492.8314,
"eval_samples_per_second": 9.756,
"eval_steps_per_second": 2.439,
"step": 1200
},
{
"epoch": 1.9965685764791514,
"grad_norm": 0.0008989392081275582,
"learning_rate": 2.721215247659059e-06,
"loss": 0.0006293777259998024,
"step": 1201
},
{
"epoch": 1.9982322969741082,
"grad_norm": 0.0019021104089915752,
"learning_rate": 2.7177943256793214e-06,
"loss": 0.0019024854991585016,
"step": 1202
},
{
"epoch": 1.9998960174690652,
"grad_norm": 0.001411781762726605,
"learning_rate": 2.7143729927259992e-06,
"loss": 0.003327167360112071,
"step": 1203
},
{
"epoch": 2.0,
"grad_norm": 0.000364909996278584,
"learning_rate": 2.7109512552550804e-06,
"loss": 0.00020373324514366686,
"step": 1204
},
{
"epoch": 2.0016637204949568,
"grad_norm": 0.0010985415428876877,
"learning_rate": 2.707529119723315e-06,
"loss": 0.0023868954740464687,
"step": 1205
},
{
"epoch": 2.0033274409899136,
"grad_norm": 0.0006375758093781769,
"learning_rate": 2.7041065925882054e-06,
"loss": 0.0005520120030269027,
"step": 1206
},
{
"epoch": 2.0049911614848703,
"grad_norm": 0.0005418790387921035,
"learning_rate": 2.7006836803079934e-06,
"loss": 0.0009355536894872785,
"step": 1207
},
{
"epoch": 2.0066548819798276,
"grad_norm": 0.0011295054573565722,
"learning_rate": 2.697260389341645e-06,
"loss": 0.0026254295371472836,
"step": 1208
},
{
"epoch": 2.0083186024747843,
"grad_norm": 0.00038410667912103236,
"learning_rate": 2.693836726148844e-06,
"loss": 0.0011726864613592625,
"step": 1209
},
{
"epoch": 2.009982322969741,
"grad_norm": 0.00015486196207348257,
"learning_rate": 2.6904126971899754e-06,
"loss": 0.00015572525444440544,
"step": 1210
},
{
"epoch": 2.011646043464698,
"grad_norm": 0.001346680917777121,
"learning_rate": 2.686988308926112e-06,
"loss": 0.0015534437261521816,
"step": 1211
},
{
"epoch": 2.0133097639596547,
"grad_norm": 0.0009657160262577236,
"learning_rate": 2.68356356781901e-06,
"loss": 0.003846096573397517,
"step": 1212
},
{
"epoch": 2.0149734844546114,
"grad_norm": 0.0003977494779974222,
"learning_rate": 2.6801384803310855e-06,
"loss": 0.0006163832149468362,
"step": 1213
},
{
"epoch": 2.0166372049495687,
"grad_norm": 0.000773407518863678,
"learning_rate": 2.676713052925411e-06,
"loss": 0.0017007028218358755,
"step": 1214
},
{
"epoch": 2.0183009254445254,
"grad_norm": 0.0014128134353086352,
"learning_rate": 2.6732872920657018e-06,
"loss": 0.0023884056136012077,
"step": 1215
},
{
"epoch": 2.019964645939482,
"grad_norm": 0.0012431552167981863,
"learning_rate": 2.6698612042162995e-06,
"loss": 0.00187435292173177,
"step": 1216
},
{
"epoch": 2.021628366434439,
"grad_norm": 0.0012625644449144602,
"learning_rate": 2.6664347958421647e-06,
"loss": 0.0011748441029340029,
"step": 1217
},
{
"epoch": 2.0232920869293958,
"grad_norm": 0.0012824207078665495,
"learning_rate": 2.6630080734088625e-06,
"loss": 0.001920399023219943,
"step": 1218
},
{
"epoch": 2.0249558074243525,
"grad_norm": 0.0007549470174126327,
"learning_rate": 2.6595810433825496e-06,
"loss": 0.0016656176885589957,
"step": 1219
},
{
"epoch": 2.0266195279193098,
"grad_norm": 0.001832195557653904,
"learning_rate": 2.6561537122299647e-06,
"loss": 0.002855651779100299,
"step": 1220
},
{
"epoch": 2.0282832484142665,
"grad_norm": 0.0010537878843024373,
"learning_rate": 2.6527260864184135e-06,
"loss": 0.0023729479871690273,
"step": 1221
},
{
"epoch": 2.0299469689092233,
"grad_norm": 0.0011735305888578296,
"learning_rate": 2.6492981724157576e-06,
"loss": 0.0042937640100717545,
"step": 1222
},
{
"epoch": 2.03161068940418,
"grad_norm": 0.0014465588610619307,
"learning_rate": 2.6458699766904033e-06,
"loss": 0.0026445563416928053,
"step": 1223
},
{
"epoch": 2.033274409899137,
"grad_norm": 0.001032729516737163,
"learning_rate": 2.6424415057112883e-06,
"loss": 0.0027413975913077593,
"step": 1224
},
{
"epoch": 2.0349381303940937,
"grad_norm": 0.000898035941645503,
"learning_rate": 2.6390127659478698e-06,
"loss": 0.0032835854217410088,
"step": 1225
},
{
"epoch": 2.0366018508890504,
"grad_norm": 0.0013310214271768928,
"learning_rate": 2.6355837638701115e-06,
"loss": 0.0020950636826455593,
"step": 1226
},
{
"epoch": 2.0382655713840077,
"grad_norm": 0.00045725522795692086,
"learning_rate": 2.632154505948472e-06,
"loss": 0.0014126868918538094,
"step": 1227
},
{
"epoch": 2.0399292918789644,
"grad_norm": 0.0009511184180155396,
"learning_rate": 2.6287249986538944e-06,
"loss": 0.0018513341201469302,
"step": 1228
},
{
"epoch": 2.041593012373921,
"grad_norm": 0.0011876067146658897,
"learning_rate": 2.62529524845779e-06,
"loss": 0.002044444438070059,
"step": 1229
},
{
"epoch": 2.043256732868878,
"grad_norm": 0.0030956417322158813,
"learning_rate": 2.6218652618320306e-06,
"loss": 0.002602633088827133,
"step": 1230
},
{
"epoch": 2.0449204533638348,
"grad_norm": 0.0008830548031255603,
"learning_rate": 2.6184350452489317e-06,
"loss": 0.0022057429887354374,
"step": 1231
},
{
"epoch": 2.0465841738587915,
"grad_norm": 0.0007592173642478883,
"learning_rate": 2.615004605181246e-06,
"loss": 0.001316930167376995,
"step": 1232
},
{
"epoch": 2.0482478943537488,
"grad_norm": 0.0010199942626059055,
"learning_rate": 2.611573948102144e-06,
"loss": 0.0023947772569954395,
"step": 1233
},
{
"epoch": 2.0499116148487055,
"grad_norm": 0.000911121373064816,
"learning_rate": 2.6081430804852093e-06,
"loss": 0.001219225930981338,
"step": 1234
},
{
"epoch": 2.0515753353436623,
"grad_norm": 0.0014884383417665958,
"learning_rate": 2.604712008804421e-06,
"loss": 0.0033853487111628056,
"step": 1235
},
{
"epoch": 2.053239055838619,
"grad_norm": 0.0007127722492441535,
"learning_rate": 2.601280739534143e-06,
"loss": 0.0009231525473296642,
"step": 1236
},
{
"epoch": 2.054902776333576,
"grad_norm": 0.0017329002730548382,
"learning_rate": 2.5978492791491126e-06,
"loss": 0.003366638207808137,
"step": 1237
},
{
"epoch": 2.0565664968285327,
"grad_norm": 0.0020664972253143787,
"learning_rate": 2.594417634124428e-06,
"loss": 0.00315756443887949,
"step": 1238
},
{
"epoch": 2.05823021732349,
"grad_norm": 0.0015665602404624224,
"learning_rate": 2.590985810935535e-06,
"loss": 0.001807854394428432,
"step": 1239
},
{
"epoch": 2.0598939378184467,
"grad_norm": 0.0004127174324821681,
"learning_rate": 2.5875538160582176e-06,
"loss": 0.0014860054943710566,
"step": 1240
},
{
"epoch": 2.0615576583134034,
"grad_norm": 0.0011286535300314426,
"learning_rate": 2.58412165596858e-06,
"loss": 0.0014345969539135695,
"step": 1241
},
{
"epoch": 2.06322137880836,
"grad_norm": 0.0011797641636803746,
"learning_rate": 2.5806893371430413e-06,
"loss": 0.0024553739931434393,
"step": 1242
},
{
"epoch": 2.064885099303317,
"grad_norm": 0.002415080787613988,
"learning_rate": 2.57725686605832e-06,
"loss": 0.003503277897834778,
"step": 1243
},
{
"epoch": 2.0665488197982738,
"grad_norm": 0.001448194496333599,
"learning_rate": 2.5738242491914206e-06,
"loss": 0.0021282355301082134,
"step": 1244
},
{
"epoch": 2.0682125402932305,
"grad_norm": 0.0005141524598002434,
"learning_rate": 2.5703914930196227e-06,
"loss": 0.0011137824039906263,
"step": 1245
},
{
"epoch": 2.0698762607881878,
"grad_norm": 0.0024350290186703205,
"learning_rate": 2.5669586040204697e-06,
"loss": 0.002408391796052456,
"step": 1246
},
{
"epoch": 2.0715399812831445,
"grad_norm": 0.0010828844970092177,
"learning_rate": 2.5635255886717553e-06,
"loss": 0.0017078034579753876,
"step": 1247
},
{
"epoch": 2.0732037017781013,
"grad_norm": 0.0007732762023806572,
"learning_rate": 2.560092453451512e-06,
"loss": 0.0011814554454758763,
"step": 1248
},
{
"epoch": 2.074867422273058,
"grad_norm": 0.004280226770788431,
"learning_rate": 2.5566592048379975e-06,
"loss": 0.004877451341599226,
"step": 1249
},
{
"epoch": 2.076531142768015,
"grad_norm": 0.0010332359233871102,
"learning_rate": 2.553225849309684e-06,
"loss": 0.0014195141848176718,
"step": 1250
},
{
"epoch": 2.0781948632629716,
"grad_norm": 0.0026121155824512243,
"learning_rate": 2.5497923933452464e-06,
"loss": 0.001718388288281858,
"step": 1251
},
{
"epoch": 2.079858583757929,
"grad_norm": 0.0010297087719663978,
"learning_rate": 2.5463588434235463e-06,
"loss": 0.002603790257126093,
"step": 1252
},
{
"epoch": 2.0815223042528856,
"grad_norm": 0.0010890235425904393,
"learning_rate": 2.542925206023626e-06,
"loss": 0.001935492386110127,
"step": 1253
},
{
"epoch": 2.0831860247478424,
"grad_norm": 0.0012174684088677168,
"learning_rate": 2.5394914876246916e-06,
"loss": 0.0012549603125080466,
"step": 1254
},
{
"epoch": 2.084849745242799,
"grad_norm": 0.0012653002049773932,
"learning_rate": 2.5360576947061004e-06,
"loss": 0.0020303502678871155,
"step": 1255
},
{
"epoch": 2.086513465737756,
"grad_norm": 0.001454854034818709,
"learning_rate": 2.5326238337473537e-06,
"loss": 0.001309222192503512,
"step": 1256
},
{
"epoch": 2.0881771862327128,
"grad_norm": 0.001833008136600256,
"learning_rate": 2.5291899112280765e-06,
"loss": 0.002758257556706667,
"step": 1257
},
{
"epoch": 2.0898409067276695,
"grad_norm": 0.0009782977867871523,
"learning_rate": 2.5257559336280145e-06,
"loss": 0.0014265577774494886,
"step": 1258
},
{
"epoch": 2.0915046272226268,
"grad_norm": 0.0008711738046258688,
"learning_rate": 2.522321907427016e-06,
"loss": 0.0021401201374828815,
"step": 1259
},
{
"epoch": 2.0931683477175835,
"grad_norm": 0.0015375674702227116,
"learning_rate": 2.5188878391050187e-06,
"loss": 0.003357475157827139,
"step": 1260
},
{
"epoch": 2.0948320682125403,
"grad_norm": 0.0019423742778599262,
"learning_rate": 2.515453735142043e-06,
"loss": 0.0037136247847229242,
"step": 1261
},
{
"epoch": 2.096495788707497,
"grad_norm": 0.0012821252457797527,
"learning_rate": 2.5120196020181752e-06,
"loss": 0.0030806218273937702,
"step": 1262
},
{
"epoch": 2.098159509202454,
"grad_norm": 0.00048500695265829563,
"learning_rate": 2.5085854462135556e-06,
"loss": 0.001238503959029913,
"step": 1263
},
{
"epoch": 2.0998232296974106,
"grad_norm": 0.0020267192739993334,
"learning_rate": 2.505151274208369e-06,
"loss": 0.002566360402852297,
"step": 1264
},
{
"epoch": 2.101486950192368,
"grad_norm": 0.0007626867154613137,
"learning_rate": 2.50171709248283e-06,
"loss": 0.0007100152433849871,
"step": 1265
},
{
"epoch": 2.1031506706873246,
"grad_norm": 0.001000856515020132,
"learning_rate": 2.4982829075171714e-06,
"loss": 0.001992270816117525,
"step": 1266
},
{
"epoch": 2.1048143911822814,
"grad_norm": 0.0008598253480158746,
"learning_rate": 2.494848725791632e-06,
"loss": 0.0017945771105587482,
"step": 1267
},
{
"epoch": 2.106478111677238,
"grad_norm": 0.00042779420618899167,
"learning_rate": 2.4914145537864453e-06,
"loss": 0.001541482168249786,
"step": 1268
},
{
"epoch": 2.108141832172195,
"grad_norm": 0.0016860082978382707,
"learning_rate": 2.4879803979818256e-06,
"loss": 0.002277143532410264,
"step": 1269
},
{
"epoch": 2.1098055526671518,
"grad_norm": 0.0016471690032631159,
"learning_rate": 2.4845462648579573e-06,
"loss": 0.0015763354022055864,
"step": 1270
},
{
"epoch": 2.111469273162109,
"grad_norm": 0.006739214528352022,
"learning_rate": 2.481112160894982e-06,
"loss": 0.0013695925008505583,
"step": 1271
},
{
"epoch": 2.1131329936570658,
"grad_norm": 0.000667403161060065,
"learning_rate": 2.4776780925729853e-06,
"loss": 0.001070688245818019,
"step": 1272
},
{
"epoch": 2.1147967141520225,
"grad_norm": 0.0034296319354325533,
"learning_rate": 2.474244066371986e-06,
"loss": 0.0019356642151251435,
"step": 1273
},
{
"epoch": 2.1164604346469793,
"grad_norm": 0.0010742850136011839,
"learning_rate": 2.4708100887719243e-06,
"loss": 0.00449890922755003,
"step": 1274
},
{
"epoch": 2.118124155141936,
"grad_norm": 0.0008409210131503642,
"learning_rate": 2.4673761662526475e-06,
"loss": 0.0014720135368406773,
"step": 1275
},
{
"epoch": 2.119787875636893,
"grad_norm": 0.0013643614947795868,
"learning_rate": 2.4639423052938995e-06,
"loss": 0.0017583542503416538,
"step": 1276
},
{
"epoch": 2.1214515961318496,
"grad_norm": 0.001211738446727395,
"learning_rate": 2.4605085123753097e-06,
"loss": 0.0019427321385592222,
"step": 1277
},
{
"epoch": 2.123115316626807,
"grad_norm": 0.0016727972542867064,
"learning_rate": 2.4570747939763745e-06,
"loss": 0.0026699069421738386,
"step": 1278
},
{
"epoch": 2.1247790371217636,
"grad_norm": 0.0009743532282300293,
"learning_rate": 2.453641156576454e-06,
"loss": 0.0024324795231223106,
"step": 1279
},
{
"epoch": 2.1264427576167204,
"grad_norm": 0.0011676120338961482,
"learning_rate": 2.4502076066547545e-06,
"loss": 0.001394494785927236,
"step": 1280
},
{
"epoch": 2.128106478111677,
"grad_norm": 0.001042691059410572,
"learning_rate": 2.4467741506903162e-06,
"loss": 0.0019729523919522762,
"step": 1281
},
{
"epoch": 2.129770198606634,
"grad_norm": 0.0010410549584776163,
"learning_rate": 2.443340795162003e-06,
"loss": 0.0013870508410036564,
"step": 1282
},
{
"epoch": 2.1314339191015907,
"grad_norm": 0.0010129284346476197,
"learning_rate": 2.4399075465484883e-06,
"loss": 0.001577532384544611,
"step": 1283
},
{
"epoch": 2.133097639596548,
"grad_norm": 0.0017204082105308771,
"learning_rate": 2.4364744113282455e-06,
"loss": 0.0015860439743846655,
"step": 1284
},
{
"epoch": 2.1347613600915047,
"grad_norm": 0.002252376638352871,
"learning_rate": 2.433041395979531e-06,
"loss": 0.0027545292396098375,
"step": 1285
},
{
"epoch": 2.1364250805864615,
"grad_norm": 0.0013701152056455612,
"learning_rate": 2.429608506980378e-06,
"loss": 0.0038595260120928288,
"step": 1286
},
{
"epoch": 2.1380888010814183,
"grad_norm": 0.0010166936554014683,
"learning_rate": 2.4261757508085803e-06,
"loss": 0.0021340944804251194,
"step": 1287
},
{
"epoch": 2.139752521576375,
"grad_norm": 0.001493407879024744,
"learning_rate": 2.422743133941681e-06,
"loss": 0.00273123593069613,
"step": 1288
},
{
"epoch": 2.141416242071332,
"grad_norm": 0.0010118827922269702,
"learning_rate": 2.419310662856959e-06,
"loss": 0.0008498326642438769,
"step": 1289
},
{
"epoch": 2.143079962566289,
"grad_norm": 0.002250266494229436,
"learning_rate": 2.415878344031421e-06,
"loss": 0.0029660004656761885,
"step": 1290
},
{
"epoch": 2.144743683061246,
"grad_norm": 0.0014006498968228698,
"learning_rate": 2.4124461839417832e-06,
"loss": 0.0014974691439419985,
"step": 1291
},
{
"epoch": 2.1464074035562026,
"grad_norm": 0.0015807535964995623,
"learning_rate": 2.4090141890644654e-06,
"loss": 0.0029796557500958443,
"step": 1292
},
{
"epoch": 2.1480711240511594,
"grad_norm": 0.0005720490007661283,
"learning_rate": 2.405582365875573e-06,
"loss": 0.0011643368052318692,
"step": 1293
},
{
"epoch": 2.149734844546116,
"grad_norm": 0.0018117536092177033,
"learning_rate": 2.4021507208508882e-06,
"loss": 0.002712543122470379,
"step": 1294
},
{
"epoch": 2.151398565041073,
"grad_norm": 0.0020061598625034094,
"learning_rate": 2.398719260465858e-06,
"loss": 0.0019743884913623333,
"step": 1295
},
{
"epoch": 2.1530622855360297,
"grad_norm": 0.0008923659333959222,
"learning_rate": 2.3952879911955794e-06,
"loss": 0.0012087312061339617,
"step": 1296
},
{
"epoch": 2.154726006030987,
"grad_norm": 0.00046620931243523955,
"learning_rate": 2.391856919514791e-06,
"loss": 0.0005710732657462358,
"step": 1297
},
{
"epoch": 2.1563897265259437,
"grad_norm": 0.000859422842040658,
"learning_rate": 2.3884260518978562e-06,
"loss": 0.0024305065162479877,
"step": 1298
},
{
"epoch": 2.1580534470209005,
"grad_norm": 0.00048544761375524104,
"learning_rate": 2.3849953948187552e-06,
"loss": 0.002606534631922841,
"step": 1299
},
{
"epoch": 2.1597171675158573,
"grad_norm": 0.0009777328232303262,
"learning_rate": 2.3815649547510687e-06,
"loss": 0.0029517700895667076,
"step": 1300
},
{
"epoch": 2.1597171675158573,
"eval_loss": 0.0023095491342246532,
"eval_runtime": 491.4412,
"eval_samples_per_second": 9.783,
"eval_steps_per_second": 2.446,
"step": 1300
},
{
"epoch": 2.161380888010814,
"grad_norm": 0.000683385122101754,
"learning_rate": 2.37813473816797e-06,
"loss": 0.0020315079018473625,
"step": 1301
},
{
"epoch": 2.163044608505771,
"grad_norm": 0.000580743420869112,
"learning_rate": 2.3747047515422102e-06,
"loss": 0.0006980924517847598,
"step": 1302
},
{
"epoch": 2.164708329000728,
"grad_norm": 0.0011160477297380567,
"learning_rate": 2.371275001346106e-06,
"loss": 0.0017467639409005642,
"step": 1303
},
{
"epoch": 2.166372049495685,
"grad_norm": 0.0030428199097514153,
"learning_rate": 2.367845494051529e-06,
"loss": 0.0016722860746085644,
"step": 1304
},
{
"epoch": 2.1680357699906416,
"grad_norm": 0.0008094862569123507,
"learning_rate": 2.3644162361298897e-06,
"loss": 0.0006268250290304422,
"step": 1305
},
{
"epoch": 2.1696994904855984,
"grad_norm": 0.0016655492363497615,
"learning_rate": 2.360987234052131e-06,
"loss": 0.0021564862690865993,
"step": 1306
},
{
"epoch": 2.171363210980555,
"grad_norm": 0.0011375686153769493,
"learning_rate": 2.357558494288712e-06,
"loss": 0.0009511542157270014,
"step": 1307
},
{
"epoch": 2.173026931475512,
"grad_norm": 0.0017173878150060773,
"learning_rate": 2.354130023309597e-06,
"loss": 0.0010811176616698503,
"step": 1308
},
{
"epoch": 2.1746906519704687,
"grad_norm": 0.0007098483620211482,
"learning_rate": 2.350701827584243e-06,
"loss": 0.0018661936046555638,
"step": 1309
},
{
"epoch": 2.176354372465426,
"grad_norm": 0.0013119137147441506,
"learning_rate": 2.3472739135815877e-06,
"loss": 0.0024188230745494366,
"step": 1310
},
{
"epoch": 2.1780180929603827,
"grad_norm": 0.0021531626116484404,
"learning_rate": 2.343846287770036e-06,
"loss": 0.001036312896758318,
"step": 1311
},
{
"epoch": 2.1796818134553395,
"grad_norm": 0.0008549646590836346,
"learning_rate": 2.340418956617451e-06,
"loss": 0.0008003504481166601,
"step": 1312
},
{
"epoch": 2.1813455339502963,
"grad_norm": 0.00044107611756771803,
"learning_rate": 2.336991926591138e-06,
"loss": 0.001624225522391498,
"step": 1313
},
{
"epoch": 2.183009254445253,
"grad_norm": 0.0014437201898545027,
"learning_rate": 2.3335652041578352e-06,
"loss": 0.001267612329684198,
"step": 1314
},
{
"epoch": 2.18467297494021,
"grad_norm": 0.0021631629206240177,
"learning_rate": 2.3301387957837017e-06,
"loss": 0.001268639462068677,
"step": 1315
},
{
"epoch": 2.186336695435167,
"grad_norm": 0.000520045985467732,
"learning_rate": 2.326712707934299e-06,
"loss": 0.0013215752551332116,
"step": 1316
},
{
"epoch": 2.188000415930124,
"grad_norm": 0.0011714874999597669,
"learning_rate": 2.3232869470745893e-06,
"loss": 0.001322682248428464,
"step": 1317
},
{
"epoch": 2.1896641364250806,
"grad_norm": 0.0010334184626117349,
"learning_rate": 2.3198615196689153e-06,
"loss": 0.001695943996310234,
"step": 1318
},
{
"epoch": 2.1913278569200374,
"grad_norm": 0.0021170424297451973,
"learning_rate": 2.3164364321809906e-06,
"loss": 0.0025002877227962017,
"step": 1319
},
{
"epoch": 2.192991577414994,
"grad_norm": 0.0027872321661561728,
"learning_rate": 2.3130116910738874e-06,
"loss": 0.0025419231969863176,
"step": 1320
},
{
"epoch": 2.194655297909951,
"grad_norm": 0.0006817941321060061,
"learning_rate": 2.309587302810026e-06,
"loss": 0.0017084216233342886,
"step": 1321
},
{
"epoch": 2.196319018404908,
"grad_norm": 0.001300262869335711,
"learning_rate": 2.306163273851157e-06,
"loss": 0.0018340634414926171,
"step": 1322
},
{
"epoch": 2.197982738899865,
"grad_norm": 0.000826711009722203,
"learning_rate": 2.302739610658356e-06,
"loss": 0.0016389775555580854,
"step": 1323
},
{
"epoch": 2.1996464593948217,
"grad_norm": 0.005244811996817589,
"learning_rate": 2.2993163196920075e-06,
"loss": 0.005631591659039259,
"step": 1324
},
{
"epoch": 2.2013101798897785,
"grad_norm": 0.0012326426804065704,
"learning_rate": 2.295893407411795e-06,
"loss": 0.0012298778165131807,
"step": 1325
},
{
"epoch": 2.2029739003847353,
"grad_norm": 0.0007586965220980346,
"learning_rate": 2.2924708802766857e-06,
"loss": 0.000672422640491277,
"step": 1326
},
{
"epoch": 2.204637620879692,
"grad_norm": 0.0018974530976265669,
"learning_rate": 2.2890487447449204e-06,
"loss": 0.00189477507956326,
"step": 1327
},
{
"epoch": 2.2063013413746493,
"grad_norm": 0.001914669992402196,
"learning_rate": 2.285627007274001e-06,
"loss": 0.0016375902341678739,
"step": 1328
},
{
"epoch": 2.207965061869606,
"grad_norm": 0.0006555395666509867,
"learning_rate": 2.282205674320679e-06,
"loss": 0.0009445602772757411,
"step": 1329
},
{
"epoch": 2.209628782364563,
"grad_norm": 0.0006078650476410985,
"learning_rate": 2.2787847523409416e-06,
"loss": 0.0010109645081683993,
"step": 1330
},
{
"epoch": 2.2112925028595196,
"grad_norm": 0.0008630360243842006,
"learning_rate": 2.2753642477900012e-06,
"loss": 0.0009613513248041272,
"step": 1331
},
{
"epoch": 2.2129562233544764,
"grad_norm": 0.0013055541785433888,
"learning_rate": 2.2719441671222815e-06,
"loss": 0.0035888648126274347,
"step": 1332
},
{
"epoch": 2.214619943849433,
"grad_norm": 0.0025269987527281046,
"learning_rate": 2.268524516791408e-06,
"loss": 0.003346838988363743,
"step": 1333
},
{
"epoch": 2.21628366434439,
"grad_norm": 0.0008320605847984552,
"learning_rate": 2.2651053032501928e-06,
"loss": 0.002192307496443391,
"step": 1334
},
{
"epoch": 2.217947384839347,
"grad_norm": 0.0015165675431489944,
"learning_rate": 2.261686532950624e-06,
"loss": 0.0017252475954592228,
"step": 1335
},
{
"epoch": 2.219611105334304,
"grad_norm": 0.0014292318373918533,
"learning_rate": 2.2582682123438547e-06,
"loss": 0.002852159086614847,
"step": 1336
},
{
"epoch": 2.2212748258292607,
"grad_norm": 0.002281657885760069,
"learning_rate": 2.254850347880187e-06,
"loss": 0.004102081526070833,
"step": 1337
},
{
"epoch": 2.2229385463242175,
"grad_norm": 0.0006204011733643711,
"learning_rate": 2.2514329460090633e-06,
"loss": 0.000904565560631454,
"step": 1338
},
{
"epoch": 2.2246022668191743,
"grad_norm": 0.0009091427782550454,
"learning_rate": 2.248016013179054e-06,
"loss": 0.0014529551845043898,
"step": 1339
},
{
"epoch": 2.226265987314131,
"grad_norm": 0.0004098558274563402,
"learning_rate": 2.244599555837844e-06,
"loss": 0.0008046920411288738,
"step": 1340
},
{
"epoch": 2.2279297078090883,
"grad_norm": 0.000955933821387589,
"learning_rate": 2.2411835804322206e-06,
"loss": 0.0019180045928806067,
"step": 1341
},
{
"epoch": 2.229593428304045,
"grad_norm": 0.0008697609300725162,
"learning_rate": 2.2377680934080625e-06,
"loss": 0.0011718597961589694,
"step": 1342
},
{
"epoch": 2.231257148799002,
"grad_norm": 0.0013697652611881495,
"learning_rate": 2.2343531012103244e-06,
"loss": 0.0017025491688400507,
"step": 1343
},
{
"epoch": 2.2329208692939586,
"grad_norm": 0.001512395218014717,
"learning_rate": 2.2309386102830295e-06,
"loss": 0.004064799752086401,
"step": 1344
},
{
"epoch": 2.2345845897889154,
"grad_norm": 0.0010302969021722674,
"learning_rate": 2.227524627069256e-06,
"loss": 0.0012793681817129254,
"step": 1345
},
{
"epoch": 2.236248310283872,
"grad_norm": 0.0011541899293661118,
"learning_rate": 2.2241111580111207e-06,
"loss": 0.0014875237829983234,
"step": 1346
},
{
"epoch": 2.237912030778829,
"grad_norm": 0.0012230722932145,
"learning_rate": 2.220698209549774e-06,
"loss": 0.0022882595658302307,
"step": 1347
},
{
"epoch": 2.239575751273786,
"grad_norm": 0.0009680510847829282,
"learning_rate": 2.2172857881253825e-06,
"loss": 0.0015226060058921576,
"step": 1348
},
{
"epoch": 2.241239471768743,
"grad_norm": 0.0006873685633763671,
"learning_rate": 2.2138739001771157e-06,
"loss": 0.0015360141405835748,
"step": 1349
},
{
"epoch": 2.2429031922636997,
"grad_norm": 0.0008744509541429579,
"learning_rate": 2.2104625521431396e-06,
"loss": 0.00249713147059083,
"step": 1350
},
{
"epoch": 2.2445669127586565,
"grad_norm": 0.0005692843697033823,
"learning_rate": 2.207051750460601e-06,
"loss": 0.0018285006517544389,
"step": 1351
},
{
"epoch": 2.2462306332536133,
"grad_norm": 0.0012063119793310761,
"learning_rate": 2.2036415015656148e-06,
"loss": 0.001575134927406907,
"step": 1352
},
{
"epoch": 2.24789435374857,
"grad_norm": 0.0007556597120128572,
"learning_rate": 2.2002318118932543e-06,
"loss": 0.0016872566193342209,
"step": 1353
},
{
"epoch": 2.2495580742435273,
"grad_norm": 0.0036170482635498047,
"learning_rate": 2.1968226878775347e-06,
"loss": 0.004833456128835678,
"step": 1354
},
{
"epoch": 2.251221794738484,
"grad_norm": 0.0002517940883990377,
"learning_rate": 2.1934141359514062e-06,
"loss": 0.0006511223036795855,
"step": 1355
},
{
"epoch": 2.252885515233441,
"grad_norm": 0.0014082285342738032,
"learning_rate": 2.1900061625467393e-06,
"loss": 0.003841613419353962,
"step": 1356
},
{
"epoch": 2.2545492357283976,
"grad_norm": 0.0004671132192015648,
"learning_rate": 2.1865987740943116e-06,
"loss": 0.000753692933358252,
"step": 1357
},
{
"epoch": 2.2562129562233544,
"grad_norm": 0.0010657520033419132,
"learning_rate": 2.183191977023799e-06,
"loss": 0.0016676551895216107,
"step": 1358
},
{
"epoch": 2.257876676718311,
"grad_norm": 0.0010002084309235215,
"learning_rate": 2.17978577776376e-06,
"loss": 0.0006241835653781891,
"step": 1359
},
{
"epoch": 2.259540397213268,
"grad_norm": 0.0013890896225348115,
"learning_rate": 2.176380182741624e-06,
"loss": 0.0022400356829166412,
"step": 1360
},
{
"epoch": 2.261204117708225,
"grad_norm": 0.00048793599125929177,
"learning_rate": 2.172975198383682e-06,
"loss": 0.0003696032799780369,
"step": 1361
},
{
"epoch": 2.262867838203182,
"grad_norm": 0.002617779653519392,
"learning_rate": 2.169570831115072e-06,
"loss": 0.002739534480497241,
"step": 1362
},
{
"epoch": 2.2645315586981387,
"grad_norm": 0.0011962675489485264,
"learning_rate": 2.1661670873597686e-06,
"loss": 0.002317224396392703,
"step": 1363
},
{
"epoch": 2.2661952791930955,
"grad_norm": 0.0007817652076482773,
"learning_rate": 2.1627639735405683e-06,
"loss": 0.0005807623383589089,
"step": 1364
},
{
"epoch": 2.2678589996880523,
"grad_norm": 0.0009937586728483438,
"learning_rate": 2.1593614960790795e-06,
"loss": 0.0020293928682804108,
"step": 1365
},
{
"epoch": 2.2695227201830095,
"grad_norm": 0.0014677675208076835,
"learning_rate": 2.15595966139571e-06,
"loss": 0.0034731775522232056,
"step": 1366
},
{
"epoch": 2.2711864406779663,
"grad_norm": 0.0005615411209873855,
"learning_rate": 2.152558475909654e-06,
"loss": 0.0013579119695350528,
"step": 1367
},
{
"epoch": 2.272850161172923,
"grad_norm": 0.001657299930229783,
"learning_rate": 2.149157946038882e-06,
"loss": 0.0017475045751780272,
"step": 1368
},
{
"epoch": 2.27451388166788,
"grad_norm": 0.0018001466523855925,
"learning_rate": 2.145758078200126e-06,
"loss": 0.0034491694532334805,
"step": 1369
},
{
"epoch": 2.2761776021628366,
"grad_norm": 0.0018540103919804096,
"learning_rate": 2.1423588788088704e-06,
"loss": 0.00254809926263988,
"step": 1370
},
{
"epoch": 2.2778413226577934,
"grad_norm": 0.0010992720490321517,
"learning_rate": 2.1389603542793364e-06,
"loss": 0.0014880215749144554,
"step": 1371
},
{
"epoch": 2.27950504315275,
"grad_norm": 0.0004704767488874495,
"learning_rate": 2.1355625110244725e-06,
"loss": 0.0011282748309895396,
"step": 1372
},
{
"epoch": 2.2811687636477074,
"grad_norm": 0.0004566149436868727,
"learning_rate": 2.1321653554559425e-06,
"loss": 0.0008396217599511147,
"step": 1373
},
{
"epoch": 2.282832484142664,
"grad_norm": 0.0017184752505272627,
"learning_rate": 2.1287688939841104e-06,
"loss": 0.0032348756212741137,
"step": 1374
},
{
"epoch": 2.284496204637621,
"grad_norm": 0.0005679796449840069,
"learning_rate": 2.125373133018033e-06,
"loss": 0.0008454375201836228,
"step": 1375
},
{
"epoch": 2.2861599251325777,
"grad_norm": 0.0016176571371033788,
"learning_rate": 2.1219780789654436e-06,
"loss": 0.0013313901145011187,
"step": 1376
},
{
"epoch": 2.2878236456275345,
"grad_norm": 0.0033084191381931305,
"learning_rate": 2.1185837382327422e-06,
"loss": 0.0016603253316134214,
"step": 1377
},
{
"epoch": 2.2894873661224913,
"grad_norm": 0.0013860368635505438,
"learning_rate": 2.1151901172249823e-06,
"loss": 0.0009522234322503209,
"step": 1378
},
{
"epoch": 2.2911510866174485,
"grad_norm": 0.0012617844622582197,
"learning_rate": 2.1117972223458598e-06,
"loss": 0.00269622216001153,
"step": 1379
},
{
"epoch": 2.2928148071124053,
"grad_norm": 0.0007739314460195601,
"learning_rate": 2.108405059997701e-06,
"loss": 0.0008796498877927661,
"step": 1380
},
{
"epoch": 2.294478527607362,
"grad_norm": 0.0017251380486413836,
"learning_rate": 2.1050136365814484e-06,
"loss": 0.0015982100740075111,
"step": 1381
},
{
"epoch": 2.296142248102319,
"grad_norm": 0.0005771266296505928,
"learning_rate": 2.10162295849665e-06,
"loss": 0.0005673667183145881,
"step": 1382
},
{
"epoch": 2.2978059685972756,
"grad_norm": 0.0017698346637189388,
"learning_rate": 2.0982330321414495e-06,
"loss": 0.0014651466626673937,
"step": 1383
},
{
"epoch": 2.2994696890922324,
"grad_norm": 0.0010502157965674996,
"learning_rate": 2.094843863912571e-06,
"loss": 0.002775599481537938,
"step": 1384
},
{
"epoch": 2.301133409587189,
"grad_norm": 0.00099456706084311,
"learning_rate": 2.0914554602053072e-06,
"loss": 0.0009711871389299631,
"step": 1385
},
{
"epoch": 2.3027971300821464,
"grad_norm": 0.0018341030227020383,
"learning_rate": 2.0880678274135103e-06,
"loss": 0.002917747711762786,
"step": 1386
},
{
"epoch": 2.304460850577103,
"grad_norm": 0.0009427666082046926,
"learning_rate": 2.084680971929574e-06,
"loss": 0.0017029845621436834,
"step": 1387
},
{
"epoch": 2.30612457107206,
"grad_norm": 0.004653818905353546,
"learning_rate": 2.0812949001444293e-06,
"loss": 0.002260258886963129,
"step": 1388
},
{
"epoch": 2.3077882915670167,
"grad_norm": 0.0009278209763579071,
"learning_rate": 2.077909618447526e-06,
"loss": 0.0015303139807656407,
"step": 1389
},
{
"epoch": 2.3094520120619735,
"grad_norm": 0.002043406944721937,
"learning_rate": 2.0745251332268238e-06,
"loss": 0.0017000572988763452,
"step": 1390
},
{
"epoch": 2.3111157325569303,
"grad_norm": 0.0007454622536897659,
"learning_rate": 2.07114145086878e-06,
"loss": 0.001288210623897612,
"step": 1391
},
{
"epoch": 2.3127794530518875,
"grad_norm": 0.0007290096255019307,
"learning_rate": 2.0677585777583366e-06,
"loss": 0.0012519609881564975,
"step": 1392
},
{
"epoch": 2.3144431735468443,
"grad_norm": 0.001200285041704774,
"learning_rate": 2.0643765202789064e-06,
"loss": 0.0011761067435145378,
"step": 1393
},
{
"epoch": 2.316106894041801,
"grad_norm": 0.00094900670228526,
"learning_rate": 2.060995284812366e-06,
"loss": 0.0020300294272601604,
"step": 1394
},
{
"epoch": 2.317770614536758,
"grad_norm": 0.00196209829300642,
"learning_rate": 2.0576148777390397e-06,
"loss": 0.0012100592721253633,
"step": 1395
},
{
"epoch": 2.3194343350317146,
"grad_norm": 0.0006102774059399962,
"learning_rate": 2.0542353054376893e-06,
"loss": 0.0009437316912226379,
"step": 1396
},
{
"epoch": 2.3210980555266714,
"grad_norm": 0.001464645960368216,
"learning_rate": 2.0508565742855017e-06,
"loss": 0.0018985284259542823,
"step": 1397
},
{
"epoch": 2.322761776021628,
"grad_norm": 0.0013890875270590186,
"learning_rate": 2.0474786906580733e-06,
"loss": 0.0029032896272838116,
"step": 1398
},
{
"epoch": 2.3244254965165854,
"grad_norm": 0.0017763370415195823,
"learning_rate": 2.044101660929405e-06,
"loss": 0.0016929400153458118,
"step": 1399
},
{
"epoch": 2.326089217011542,
"grad_norm": 0.000577530125156045,
"learning_rate": 2.040725491471885e-06,
"loss": 0.0017563023138791323,
"step": 1400
},
{
"epoch": 2.326089217011542,
"eval_loss": 0.002368729095906019,
"eval_runtime": 491.5856,
"eval_samples_per_second": 9.781,
"eval_steps_per_second": 2.445,
"step": 1400
},
{
"epoch": 2.327752937506499,
"grad_norm": 0.00043292154441587627,
"learning_rate": 2.037350188656279e-06,
"loss": 0.0007160247769206762,
"step": 1401
},
{
"epoch": 2.3294166580014557,
"grad_norm": 0.0012572737177833915,
"learning_rate": 2.0339757588517165e-06,
"loss": 0.0022942391224205494,
"step": 1402
},
{
"epoch": 2.3310803784964125,
"grad_norm": 0.0006292902980931103,
"learning_rate": 2.0306022084256786e-06,
"loss": 0.0004353314870968461,
"step": 1403
},
{
"epoch": 2.3327440989913697,
"grad_norm": 0.001153657678514719,
"learning_rate": 2.027229543743989e-06,
"loss": 0.0024223560467362404,
"step": 1404
},
{
"epoch": 2.3344078194863265,
"grad_norm": 0.000907072564586997,
"learning_rate": 2.0238577711707987e-06,
"loss": 0.0014202623860910535,
"step": 1405
},
{
"epoch": 2.3360715399812833,
"grad_norm": 0.0012798805255442858,
"learning_rate": 2.0204868970685764e-06,
"loss": 0.0024821199476718903,
"step": 1406
},
{
"epoch": 2.33773526047624,
"grad_norm": 0.0017148812767118216,
"learning_rate": 2.0171169277980954e-06,
"loss": 0.0014755388256162405,
"step": 1407
},
{
"epoch": 2.339398980971197,
"grad_norm": 0.0031123829539865255,
"learning_rate": 2.0137478697184205e-06,
"loss": 0.004206657875329256,
"step": 1408
},
{
"epoch": 2.3410627014661536,
"grad_norm": 0.001084084389731288,
"learning_rate": 2.0103797291868977e-06,
"loss": 0.0021153988782316446,
"step": 1409
},
{
"epoch": 2.3427264219611104,
"grad_norm": 0.002781338058412075,
"learning_rate": 2.0070125125591414e-06,
"loss": 0.001565725775435567,
"step": 1410
},
{
"epoch": 2.344390142456067,
"grad_norm": 0.0021625482477247715,
"learning_rate": 2.0036462261890225e-06,
"loss": 0.003212597919628024,
"step": 1411
},
{
"epoch": 2.3460538629510244,
"grad_norm": 0.0009133850689977407,
"learning_rate": 2.0002808764286573e-06,
"loss": 0.0007037279428914189,
"step": 1412
},
{
"epoch": 2.347717583445981,
"grad_norm": 0.0021134146954864264,
"learning_rate": 1.9969164696283945e-06,
"loss": 0.0028274108190089464,
"step": 1413
},
{
"epoch": 2.349381303940938,
"grad_norm": 0.0014362254878506064,
"learning_rate": 1.9935530121368023e-06,
"loss": 0.002772002946585417,
"step": 1414
},
{
"epoch": 2.3510450244358947,
"grad_norm": 0.001166126225143671,
"learning_rate": 1.990190510300659e-06,
"loss": 0.0007662165444344282,
"step": 1415
},
{
"epoch": 2.3527087449308515,
"grad_norm": 0.007451492827385664,
"learning_rate": 1.986828970464939e-06,
"loss": 0.0043607125990092754,
"step": 1416
},
{
"epoch": 2.3543724654258087,
"grad_norm": 0.0006244085379876196,
"learning_rate": 1.983468398972802e-06,
"loss": 0.001204540254548192,
"step": 1417
},
{
"epoch": 2.3560361859207655,
"grad_norm": 0.0016244313446804881,
"learning_rate": 1.980108802165579e-06,
"loss": 0.003739001462236047,
"step": 1418
},
{
"epoch": 2.3576999064157222,
"grad_norm": 0.0021900099236518145,
"learning_rate": 1.976750186382764e-06,
"loss": 0.0014113308861851692,
"step": 1419
},
{
"epoch": 2.359363626910679,
"grad_norm": 0.0023189790081232786,
"learning_rate": 1.9733925579619965e-06,
"loss": 0.003480505431070924,
"step": 1420
},
{
"epoch": 2.361027347405636,
"grad_norm": 0.002194036729633808,
"learning_rate": 1.970035923239056e-06,
"loss": 0.002109128050506115,
"step": 1421
},
{
"epoch": 2.3626910679005926,
"grad_norm": 0.0006388880428858101,
"learning_rate": 1.9666802885478463e-06,
"loss": 0.00182803301140666,
"step": 1422
},
{
"epoch": 2.3643547883955494,
"grad_norm": 0.0022078445181250572,
"learning_rate": 1.963325660220384e-06,
"loss": 0.0020670914091169834,
"step": 1423
},
{
"epoch": 2.3660185088905066,
"grad_norm": 0.008101390674710274,
"learning_rate": 1.9599720445867856e-06,
"loss": 0.0016147526912391186,
"step": 1424
},
{
"epoch": 2.3676822293854634,
"grad_norm": 0.0006417955737560987,
"learning_rate": 1.956619447975257e-06,
"loss": 0.001665701624006033,
"step": 1425
},
{
"epoch": 2.36934594988042,
"grad_norm": 0.001856435090303421,
"learning_rate": 1.9532678767120827e-06,
"loss": 0.001060395734384656,
"step": 1426
},
{
"epoch": 2.371009670375377,
"grad_norm": 0.002327260095626116,
"learning_rate": 1.9499173371216105e-06,
"loss": 0.005265223328024149,
"step": 1427
},
{
"epoch": 2.3726733908703337,
"grad_norm": 0.0020023256074637175,
"learning_rate": 1.946567835526243e-06,
"loss": 0.0014220771845430136,
"step": 1428
},
{
"epoch": 2.3743371113652905,
"grad_norm": 0.0010369986994192004,
"learning_rate": 1.943219378246423e-06,
"loss": 0.0021268511191010475,
"step": 1429
},
{
"epoch": 2.3760008318602477,
"grad_norm": 0.001503292704001069,
"learning_rate": 1.9398719716006246e-06,
"loss": 0.0016041295602917671,
"step": 1430
},
{
"epoch": 2.3776645523552045,
"grad_norm": 0.0012758244993165135,
"learning_rate": 1.936525621905336e-06,
"loss": 0.002271541627123952,
"step": 1431
},
{
"epoch": 2.3793282728501612,
"grad_norm": 0.0009672525920905173,
"learning_rate": 1.9331803354750537e-06,
"loss": 0.0027922196313738823,
"step": 1432
},
{
"epoch": 2.380991993345118,
"grad_norm": 0.0017020419472828507,
"learning_rate": 1.9298361186222665e-06,
"loss": 0.003419260261580348,
"step": 1433
},
{
"epoch": 2.382655713840075,
"grad_norm": 0.0012927661882713437,
"learning_rate": 1.926492977657446e-06,
"loss": 0.0022287052124738693,
"step": 1434
},
{
"epoch": 2.3843194343350316,
"grad_norm": 0.0008178834686987102,
"learning_rate": 1.9231509188890345e-06,
"loss": 0.001787499408237636,
"step": 1435
},
{
"epoch": 2.3859831548299884,
"grad_norm": 0.0016086464747786522,
"learning_rate": 1.919809948623428e-06,
"loss": 0.001454781275242567,
"step": 1436
},
{
"epoch": 2.3876468753249456,
"grad_norm": 0.0008009792654775083,
"learning_rate": 1.9164700731649723e-06,
"loss": 0.0010338013526052237,
"step": 1437
},
{
"epoch": 2.3893105958199023,
"grad_norm": 0.0006302039255388081,
"learning_rate": 1.913131298815947e-06,
"loss": 0.0007497619953937829,
"step": 1438
},
{
"epoch": 2.390974316314859,
"grad_norm": 0.00047956148046068847,
"learning_rate": 1.9097936318765527e-06,
"loss": 0.0007028962718322873,
"step": 1439
},
{
"epoch": 2.392638036809816,
"grad_norm": 0.0009208683623000979,
"learning_rate": 1.906457078644901e-06,
"loss": 0.0025395380798727274,
"step": 1440
},
{
"epoch": 2.3943017573047727,
"grad_norm": 0.0004860176413785666,
"learning_rate": 1.903121645417003e-06,
"loss": 0.0011189526412636042,
"step": 1441
},
{
"epoch": 2.3959654777997295,
"grad_norm": 0.0008918407256715,
"learning_rate": 1.8997873384867534e-06,
"loss": 0.0019672035705298185,
"step": 1442
},
{
"epoch": 2.3976291982946867,
"grad_norm": 0.0006723460974171758,
"learning_rate": 1.8964541641459242e-06,
"loss": 0.001565326121635735,
"step": 1443
},
{
"epoch": 2.3992929187896435,
"grad_norm": 0.0010983988177031279,
"learning_rate": 1.893122128684149e-06,
"loss": 0.0020677861757576466,
"step": 1444
},
{
"epoch": 2.4009566392846002,
"grad_norm": 0.000979145523160696,
"learning_rate": 1.8897912383889138e-06,
"loss": 0.0024629514664411545,
"step": 1445
},
{
"epoch": 2.402620359779557,
"grad_norm": 0.0004037956823594868,
"learning_rate": 1.886461499545543e-06,
"loss": 0.001216419623233378,
"step": 1446
},
{
"epoch": 2.404284080274514,
"grad_norm": 0.003060317598283291,
"learning_rate": 1.883132918437186e-06,
"loss": 0.003621977288275957,
"step": 1447
},
{
"epoch": 2.4059478007694706,
"grad_norm": 0.0002842900576069951,
"learning_rate": 1.8798055013448105e-06,
"loss": 0.0003931760147679597,
"step": 1448
},
{
"epoch": 2.4076115212644273,
"grad_norm": 0.0008548618061468005,
"learning_rate": 1.8764792545471872e-06,
"loss": 0.0012658205814659595,
"step": 1449
},
{
"epoch": 2.4092752417593846,
"grad_norm": 0.001995617290958762,
"learning_rate": 1.8731541843208772e-06,
"loss": 0.004006143659353256,
"step": 1450
},
{
"epoch": 2.4109389622543413,
"grad_norm": 0.000583532324526459,
"learning_rate": 1.869830296940223e-06,
"loss": 0.0003809987974818796,
"step": 1451
},
{
"epoch": 2.412602682749298,
"grad_norm": 0.0008577098487876356,
"learning_rate": 1.8665075986773346e-06,
"loss": 0.0008792573353275657,
"step": 1452
},
{
"epoch": 2.414266403244255,
"grad_norm": 0.0004475556779652834,
"learning_rate": 1.863186095802077e-06,
"loss": 0.0011853489559143782,
"step": 1453
},
{
"epoch": 2.4159301237392117,
"grad_norm": 0.00038343455526046455,
"learning_rate": 1.8598657945820605e-06,
"loss": 0.00043216568883508444,
"step": 1454
},
{
"epoch": 2.417593844234169,
"grad_norm": 0.0008230686653405428,
"learning_rate": 1.8565467012826282e-06,
"loss": 0.0017967612948268652,
"step": 1455
},
{
"epoch": 2.4192575647291257,
"grad_norm": 0.00038567421142943203,
"learning_rate": 1.853228822166843e-06,
"loss": 0.0010309724602848291,
"step": 1456
},
{
"epoch": 2.4209212852240825,
"grad_norm": 0.0013990384759381413,
"learning_rate": 1.849912163495479e-06,
"loss": 0.0018748044967651367,
"step": 1457
},
{
"epoch": 2.4225850057190392,
"grad_norm": 0.002193356631323695,
"learning_rate": 1.8465967315270029e-06,
"loss": 0.002720393007621169,
"step": 1458
},
{
"epoch": 2.424248726213996,
"grad_norm": 0.000335589109454304,
"learning_rate": 1.8432825325175707e-06,
"loss": 0.00045209412928670645,
"step": 1459
},
{
"epoch": 2.425912446708953,
"grad_norm": 0.001692134770564735,
"learning_rate": 1.8399695727210098e-06,
"loss": 0.005261164158582687,
"step": 1460
},
{
"epoch": 2.4275761672039096,
"grad_norm": 0.0018994332058355212,
"learning_rate": 1.836657858388811e-06,
"loss": 0.0019658301025629044,
"step": 1461
},
{
"epoch": 2.429239887698867,
"grad_norm": 0.0006664514658041298,
"learning_rate": 1.8333473957701126e-06,
"loss": 0.0028958944603800774,
"step": 1462
},
{
"epoch": 2.4309036081938236,
"grad_norm": 0.0017626987537369132,
"learning_rate": 1.830038191111692e-06,
"loss": 0.0033536856062710285,
"step": 1463
},
{
"epoch": 2.4325673286887803,
"grad_norm": 0.0005109024350531399,
"learning_rate": 1.8267302506579532e-06,
"loss": 0.0013197731459513307,
"step": 1464
},
{
"epoch": 2.434231049183737,
"grad_norm": 0.0014629487413913012,
"learning_rate": 1.8234235806509145e-06,
"loss": 0.0009660832583904266,
"step": 1465
},
{
"epoch": 2.435894769678694,
"grad_norm": 0.0006256970809772611,
"learning_rate": 1.8201181873301967e-06,
"loss": 0.0014558584662154317,
"step": 1466
},
{
"epoch": 2.4375584901736507,
"grad_norm": 0.0007523475214838982,
"learning_rate": 1.816814076933012e-06,
"loss": 0.000893545220606029,
"step": 1467
},
{
"epoch": 2.439222210668608,
"grad_norm": 0.0005765758687630296,
"learning_rate": 1.813511255694152e-06,
"loss": 0.0009665571851655841,
"step": 1468
},
{
"epoch": 2.4408859311635647,
"grad_norm": 0.0018482616869732738,
"learning_rate": 1.8102097298459732e-06,
"loss": 0.002785992342978716,
"step": 1469
},
{
"epoch": 2.4425496516585214,
"grad_norm": 0.0007372192339971662,
"learning_rate": 1.80690950561839e-06,
"loss": 0.0010839628521353006,
"step": 1470
},
{
"epoch": 2.4442133721534782,
"grad_norm": 0.0007215483929030597,
"learning_rate": 1.8036105892388611e-06,
"loss": 0.001313893124461174,
"step": 1471
},
{
"epoch": 2.445877092648435,
"grad_norm": 0.001106858253479004,
"learning_rate": 1.800312986932376e-06,
"loss": 0.0017755581066012383,
"step": 1472
},
{
"epoch": 2.447540813143392,
"grad_norm": 0.0011772233992815018,
"learning_rate": 1.7970167049214466e-06,
"loss": 0.002397167030721903,
"step": 1473
},
{
"epoch": 2.4492045336383486,
"grad_norm": 0.0008078587707132101,
"learning_rate": 1.7937217494260888e-06,
"loss": 0.0017798661720007658,
"step": 1474
},
{
"epoch": 2.450868254133306,
"grad_norm": 0.0008612503297626972,
"learning_rate": 1.7904281266638201e-06,
"loss": 0.002131294459104538,
"step": 1475
},
{
"epoch": 2.4525319746282626,
"grad_norm": 0.000806903641205281,
"learning_rate": 1.7871358428496416e-06,
"loss": 0.0010623048292472959,
"step": 1476
},
{
"epoch": 2.4541956951232193,
"grad_norm": 0.0006560089532285929,
"learning_rate": 1.7838449041960276e-06,
"loss": 0.0005079662660136819,
"step": 1477
},
{
"epoch": 2.455859415618176,
"grad_norm": 0.0010120128281414509,
"learning_rate": 1.7805553169129142e-06,
"loss": 0.001596860121935606,
"step": 1478
},
{
"epoch": 2.457523136113133,
"grad_norm": 0.0013024996733292937,
"learning_rate": 1.7772670872076883e-06,
"loss": 0.004441859200596809,
"step": 1479
},
{
"epoch": 2.4591868566080897,
"grad_norm": 0.0021134542766958475,
"learning_rate": 1.773980221285173e-06,
"loss": 0.0017620328580960631,
"step": 1480
},
{
"epoch": 2.460850577103047,
"grad_norm": 0.0005604978650808334,
"learning_rate": 1.7706947253476194e-06,
"loss": 0.0015471463557332754,
"step": 1481
},
{
"epoch": 2.4625142975980037,
"grad_norm": 0.0011455026688054204,
"learning_rate": 1.767410605594694e-06,
"loss": 0.0016314920503646135,
"step": 1482
},
{
"epoch": 2.4641780180929604,
"grad_norm": 0.0012556511210277677,
"learning_rate": 1.7641278682234658e-06,
"loss": 0.0021901875734329224,
"step": 1483
},
{
"epoch": 2.465841738587917,
"grad_norm": 0.0010981862433254719,
"learning_rate": 1.7608465194283958e-06,
"loss": 0.002816542750224471,
"step": 1484
},
{
"epoch": 2.467505459082874,
"grad_norm": 0.001629547798074782,
"learning_rate": 1.757566565401323e-06,
"loss": 0.003777154488489032,
"step": 1485
},
{
"epoch": 2.4691691795778308,
"grad_norm": 0.0017756317974999547,
"learning_rate": 1.7542880123314559e-06,
"loss": 0.0037420601584017277,
"step": 1486
},
{
"epoch": 2.4708329000727876,
"grad_norm": 0.0009217644692398608,
"learning_rate": 1.75101086640536e-06,
"loss": 0.0017659610603004694,
"step": 1487
},
{
"epoch": 2.4724966205677448,
"grad_norm": 0.0008854199550114572,
"learning_rate": 1.7477351338069442e-06,
"loss": 0.0025656609795987606,
"step": 1488
},
{
"epoch": 2.4741603410627016,
"grad_norm": 0.0008711411501280963,
"learning_rate": 1.7444608207174519e-06,
"loss": 0.001252155052497983,
"step": 1489
},
{
"epoch": 2.4758240615576583,
"grad_norm": 0.0007815634598955512,
"learning_rate": 1.741187933315448e-06,
"loss": 0.0015573770506307483,
"step": 1490
},
{
"epoch": 2.477487782052615,
"grad_norm": 0.0009990883991122246,
"learning_rate": 1.7379164777768038e-06,
"loss": 0.0016516413306817412,
"step": 1491
},
{
"epoch": 2.479151502547572,
"grad_norm": 0.0006764251738786697,
"learning_rate": 1.734646460274692e-06,
"loss": 0.0020175199024379253,
"step": 1492
},
{
"epoch": 2.4808152230425287,
"grad_norm": 0.0008993525989353657,
"learning_rate": 1.7313778869795717e-06,
"loss": 0.0015614933799952269,
"step": 1493
},
{
"epoch": 2.482478943537486,
"grad_norm": 0.001278067473322153,
"learning_rate": 1.728110764059176e-06,
"loss": 0.0010163411498069763,
"step": 1494
},
{
"epoch": 2.4841426640324427,
"grad_norm": 0.0007695929380133748,
"learning_rate": 1.7248450976785011e-06,
"loss": 0.001716417958959937,
"step": 1495
},
{
"epoch": 2.4858063845273994,
"grad_norm": 0.00044241335126571357,
"learning_rate": 1.7215808939997945e-06,
"loss": 0.002265610732138157,
"step": 1496
},
{
"epoch": 2.487470105022356,
"grad_norm": 0.0005241918261162937,
"learning_rate": 1.7183181591825437e-06,
"loss": 0.0009319846867583692,
"step": 1497
},
{
"epoch": 2.489133825517313,
"grad_norm": 0.0007762466557323933,
"learning_rate": 1.7150568993834666e-06,
"loss": 0.001524902181699872,
"step": 1498
},
{
"epoch": 2.4907975460122698,
"grad_norm": 0.0008093323558568954,
"learning_rate": 1.7117971207564934e-06,
"loss": 0.0015047211199998856,
"step": 1499
},
{
"epoch": 2.4924612665072265,
"grad_norm": 0.0007091707666404545,
"learning_rate": 1.7085388294527632e-06,
"loss": 0.0006919258739799261,
"step": 1500
},
{
"epoch": 2.4924612665072265,
"eval_loss": 0.0021714933682233095,
"eval_runtime": 490.2004,
"eval_samples_per_second": 9.808,
"eval_steps_per_second": 2.452,
"step": 1500
},
{
"epoch": 2.4941249870021838,
"grad_norm": 0.0005224543274380267,
"learning_rate": 1.705282031620608e-06,
"loss": 0.0010630788747221231,
"step": 1501
},
{
"epoch": 2.4957887074971405,
"grad_norm": 0.0009647620609030128,
"learning_rate": 1.7020267334055393e-06,
"loss": 0.0013143199030309916,
"step": 1502
},
{
"epoch": 2.4974524279920973,
"grad_norm": 0.0020491466857492924,
"learning_rate": 1.6987729409502412e-06,
"loss": 0.0006549229728989303,
"step": 1503
},
{
"epoch": 2.499116148487054,
"grad_norm": 0.0010662437416613102,
"learning_rate": 1.6955206603945557e-06,
"loss": 0.0024829788599163294,
"step": 1504
},
{
"epoch": 2.500779868982011,
"grad_norm": 0.0021262792870402336,
"learning_rate": 1.6922698978754726e-06,
"loss": 0.0022876132279634476,
"step": 1505
},
{
"epoch": 2.502443589476968,
"grad_norm": 0.0037959532346576452,
"learning_rate": 1.6890206595271153e-06,
"loss": 0.0065306369215250015,
"step": 1506
},
{
"epoch": 2.504107309971925,
"grad_norm": 0.0008538314723409712,
"learning_rate": 1.6857729514807325e-06,
"loss": 0.0015486277407035232,
"step": 1507
},
{
"epoch": 2.5057710304668817,
"grad_norm": 0.0008208471699617803,
"learning_rate": 1.6825267798646851e-06,
"loss": 0.0010096883634105325,
"step": 1508
},
{
"epoch": 2.5074347509618384,
"grad_norm": 0.0006657431949861348,
"learning_rate": 1.6792821508044352e-06,
"loss": 0.0013219774700701237,
"step": 1509
},
{
"epoch": 2.509098471456795,
"grad_norm": 0.0006319488165900111,
"learning_rate": 1.6760390704225333e-06,
"loss": 0.0018334081396460533,
"step": 1510
},
{
"epoch": 2.510762191951752,
"grad_norm": 0.0003473910328466445,
"learning_rate": 1.672797544838608e-06,
"loss": 0.0008822108502499759,
"step": 1511
},
{
"epoch": 2.5124259124467088,
"grad_norm": 0.001212777686305344,
"learning_rate": 1.6695575801693549e-06,
"loss": 0.0028206915594637394,
"step": 1512
},
{
"epoch": 2.5140896329416655,
"grad_norm": 0.0005771344294771552,
"learning_rate": 1.6663191825285214e-06,
"loss": 0.0005732635036110878,
"step": 1513
},
{
"epoch": 2.5157533534366228,
"grad_norm": 0.0009036052506417036,
"learning_rate": 1.6630823580269005e-06,
"loss": 0.0025500047486275434,
"step": 1514
},
{
"epoch": 2.5174170739315795,
"grad_norm": 0.000790099729783833,
"learning_rate": 1.6598471127723162e-06,
"loss": 0.00228320574387908,
"step": 1515
},
{
"epoch": 2.5190807944265363,
"grad_norm": 0.0009197267354466021,
"learning_rate": 1.6566134528696126e-06,
"loss": 0.0007472310098819435,
"step": 1516
},
{
"epoch": 2.520744514921493,
"grad_norm": 0.0008512712665833533,
"learning_rate": 1.6533813844206426e-06,
"loss": 0.001149134011939168,
"step": 1517
},
{
"epoch": 2.52240823541645,
"grad_norm": 0.0008269053651019931,
"learning_rate": 1.6501509135242533e-06,
"loss": 0.0011230545351281762,
"step": 1518
},
{
"epoch": 2.524071955911407,
"grad_norm": 0.0018875693203881383,
"learning_rate": 1.6469220462762807e-06,
"loss": 0.005237270146608353,
"step": 1519
},
{
"epoch": 2.525735676406364,
"grad_norm": 0.0010742016602307558,
"learning_rate": 1.6436947887695336e-06,
"loss": 0.0016066599637269974,
"step": 1520
},
{
"epoch": 2.5273993969013206,
"grad_norm": 0.0008471138426102698,
"learning_rate": 1.6404691470937829e-06,
"loss": 0.0014250362291932106,
"step": 1521
},
{
"epoch": 2.5290631173962774,
"grad_norm": 0.00104066904168576,
"learning_rate": 1.6372451273357504e-06,
"loss": 0.0016985085094347596,
"step": 1522
},
{
"epoch": 2.530726837891234,
"grad_norm": 0.0012702214298769832,
"learning_rate": 1.6340227355790988e-06,
"loss": 0.001974192913621664,
"step": 1523
},
{
"epoch": 2.532390558386191,
"grad_norm": 0.00034718395909294486,
"learning_rate": 1.6308019779044154e-06,
"loss": 0.000869637297000736,
"step": 1524
},
{
"epoch": 2.5340542788811478,
"grad_norm": 0.000563792185857892,
"learning_rate": 1.6275828603892078e-06,
"loss": 0.001410376513376832,
"step": 1525
},
{
"epoch": 2.535717999376105,
"grad_norm": 0.0015897756675258279,
"learning_rate": 1.6243653891078864e-06,
"loss": 0.0031209930311888456,
"step": 1526
},
{
"epoch": 2.5373817198710618,
"grad_norm": 0.0014952553901821375,
"learning_rate": 1.6211495701317565e-06,
"loss": 0.002497080946341157,
"step": 1527
},
{
"epoch": 2.5390454403660185,
"grad_norm": 0.0002881517866626382,
"learning_rate": 1.6179354095290051e-06,
"loss": 0.00020394183229655027,
"step": 1528
},
{
"epoch": 2.5407091608609753,
"grad_norm": 0.0009688301361165941,
"learning_rate": 1.6147229133646885e-06,
"loss": 0.001731213997118175,
"step": 1529
},
{
"epoch": 2.542372881355932,
"grad_norm": 0.0005621787277050316,
"learning_rate": 1.611512087700724e-06,
"loss": 0.001863375655375421,
"step": 1530
},
{
"epoch": 2.5440366018508893,
"grad_norm": 0.0005942243151366711,
"learning_rate": 1.6083029385958762e-06,
"loss": 0.001412746263667941,
"step": 1531
},
{
"epoch": 2.545700322345846,
"grad_norm": 0.0009608308901078999,
"learning_rate": 1.6050954721057461e-06,
"loss": 0.002480145078152418,
"step": 1532
},
{
"epoch": 2.547364042840803,
"grad_norm": 0.00087307597277686,
"learning_rate": 1.6018896942827595e-06,
"loss": 0.0018921629525721073,
"step": 1533
},
{
"epoch": 2.5490277633357596,
"grad_norm": 0.0017327595269307494,
"learning_rate": 1.5986856111761562e-06,
"loss": 0.004610971547663212,
"step": 1534
},
{
"epoch": 2.5506914838307164,
"grad_norm": 0.000762734969612211,
"learning_rate": 1.595483228831976e-06,
"loss": 0.002128410153090954,
"step": 1535
},
{
"epoch": 2.552355204325673,
"grad_norm": 0.0002462914271745831,
"learning_rate": 1.5922825532930526e-06,
"loss": 0.0005376354092732072,
"step": 1536
},
{
"epoch": 2.55401892482063,
"grad_norm": 0.001250227796845138,
"learning_rate": 1.5890835905989969e-06,
"loss": 0.0013491283170878887,
"step": 1537
},
{
"epoch": 2.5556826453155868,
"grad_norm": 0.0010119171347469091,
"learning_rate": 1.5858863467861882e-06,
"loss": 0.0018797065131366253,
"step": 1538
},
{
"epoch": 2.557346365810544,
"grad_norm": 0.0011957940878346562,
"learning_rate": 1.582690827887763e-06,
"loss": 0.000866060727275908,
"step": 1539
},
{
"epoch": 2.5590100863055008,
"grad_norm": 0.0012566793011501431,
"learning_rate": 1.5794970399336012e-06,
"loss": 0.0031211625318974257,
"step": 1540
},
{
"epoch": 2.5606738068004575,
"grad_norm": 0.0006489444640465081,
"learning_rate": 1.576304988950318e-06,
"loss": 0.0026007932610809803,
"step": 1541
},
{
"epoch": 2.5623375272954143,
"grad_norm": 0.001092761754989624,
"learning_rate": 1.5731146809612508e-06,
"loss": 0.001698095933534205,
"step": 1542
},
{
"epoch": 2.564001247790371,
"grad_norm": 0.0014394314493983984,
"learning_rate": 1.569926121986447e-06,
"loss": 0.0019551399163901806,
"step": 1543
},
{
"epoch": 2.5656649682853283,
"grad_norm": 0.001035700086504221,
"learning_rate": 1.566739318042655e-06,
"loss": 0.002465068595483899,
"step": 1544
},
{
"epoch": 2.567328688780285,
"grad_norm": 0.0009444122551940382,
"learning_rate": 1.56355427514331e-06,
"loss": 0.001828762935474515,
"step": 1545
},
{
"epoch": 2.568992409275242,
"grad_norm": 0.0007945806719362736,
"learning_rate": 1.5603709992985256e-06,
"loss": 0.0018355362117290497,
"step": 1546
},
{
"epoch": 2.5706561297701986,
"grad_norm": 0.0015775883803144097,
"learning_rate": 1.5571894965150796e-06,
"loss": 0.002111088950186968,
"step": 1547
},
{
"epoch": 2.5723198502651554,
"grad_norm": 0.001175825484097004,
"learning_rate": 1.554009772796406e-06,
"loss": 0.0020981403067708015,
"step": 1548
},
{
"epoch": 2.573983570760112,
"grad_norm": 0.0011401629308238626,
"learning_rate": 1.55083183414258e-06,
"loss": 0.0018020938150584698,
"step": 1549
},
{
"epoch": 2.575647291255069,
"grad_norm": 0.00028235308127477765,
"learning_rate": 1.5476556865503095e-06,
"loss": 0.0011889156885445118,
"step": 1550
},
{
"epoch": 2.5773110117500257,
"grad_norm": 0.0009733039187267423,
"learning_rate": 1.5444813360129207e-06,
"loss": 0.0014338747132569551,
"step": 1551
},
{
"epoch": 2.578974732244983,
"grad_norm": 0.0006195545429363847,
"learning_rate": 1.5413087885203515e-06,
"loss": 0.0003205559041816741,
"step": 1552
},
{
"epoch": 2.5806384527399397,
"grad_norm": 0.000734661181923002,
"learning_rate": 1.538138050059136e-06,
"loss": 0.0014326430391520262,
"step": 1553
},
{
"epoch": 2.5823021732348965,
"grad_norm": 0.0009669451974332333,
"learning_rate": 1.5349691266123946e-06,
"loss": 0.0034975572489202023,
"step": 1554
},
{
"epoch": 2.5839658937298533,
"grad_norm": 0.0009032751549966633,
"learning_rate": 1.5318020241598248e-06,
"loss": 0.0009406222961843014,
"step": 1555
},
{
"epoch": 2.58562961422481,
"grad_norm": 0.0013129219878464937,
"learning_rate": 1.5286367486776835e-06,
"loss": 0.003194645280018449,
"step": 1556
},
{
"epoch": 2.5872933347197673,
"grad_norm": 0.0017833646852523088,
"learning_rate": 1.5254733061387846e-06,
"loss": 0.0013546156696975231,
"step": 1557
},
{
"epoch": 2.588957055214724,
"grad_norm": 0.0005147892516106367,
"learning_rate": 1.5223117025124817e-06,
"loss": 0.0015783794224262238,
"step": 1558
},
{
"epoch": 2.590620775709681,
"grad_norm": 0.0008144800085574389,
"learning_rate": 1.5191519437646576e-06,
"loss": 0.0038211564533412457,
"step": 1559
},
{
"epoch": 2.5922844962046376,
"grad_norm": 0.000820773362647742,
"learning_rate": 1.5159940358577151e-06,
"loss": 0.0019291674252599478,
"step": 1560
},
{
"epoch": 2.5939482166995944,
"grad_norm": 0.001188530120998621,
"learning_rate": 1.512837984750565e-06,
"loss": 0.0013952332083135843,
"step": 1561
},
{
"epoch": 2.595611937194551,
"grad_norm": 0.0013983040116727352,
"learning_rate": 1.5096837963986112e-06,
"loss": 0.0018704570829868317,
"step": 1562
},
{
"epoch": 2.597275657689508,
"grad_norm": 0.0008498321985825896,
"learning_rate": 1.5065314767537453e-06,
"loss": 0.0011971730273216963,
"step": 1563
},
{
"epoch": 2.5989393781844647,
"grad_norm": 0.00019363206229172647,
"learning_rate": 1.5033810317643327e-06,
"loss": 0.00046173634473234415,
"step": 1564
},
{
"epoch": 2.600603098679422,
"grad_norm": 0.0007203574641607702,
"learning_rate": 1.5002324673752006e-06,
"loss": 0.0010821465402841568,
"step": 1565
},
{
"epoch": 2.6022668191743787,
"grad_norm": 0.00038083441904745996,
"learning_rate": 1.4970857895276285e-06,
"loss": 0.0011750578414648771,
"step": 1566
},
{
"epoch": 2.6039305396693355,
"grad_norm": 0.0007648352766409516,
"learning_rate": 1.4939410041593338e-06,
"loss": 0.0026476874481886625,
"step": 1567
},
{
"epoch": 2.6055942601642923,
"grad_norm": 0.0018421369604766369,
"learning_rate": 1.4907981172044647e-06,
"loss": 0.002846265211701393,
"step": 1568
},
{
"epoch": 2.6072579806592495,
"grad_norm": 0.0003029352519661188,
"learning_rate": 1.487657134593587e-06,
"loss": 0.0008935644291341305,
"step": 1569
},
{
"epoch": 2.6089217011542063,
"grad_norm": 0.0004757127317134291,
"learning_rate": 1.4845180622536728e-06,
"loss": 0.0011599808931350708,
"step": 1570
},
{
"epoch": 2.610585421649163,
"grad_norm": 0.0005109715275466442,
"learning_rate": 1.4813809061080893e-06,
"loss": 0.0024393724743276834,
"step": 1571
},
{
"epoch": 2.61224914214412,
"grad_norm": 0.0002291719283675775,
"learning_rate": 1.4782456720765895e-06,
"loss": 0.00032998324604704976,
"step": 1572
},
{
"epoch": 2.6139128626390766,
"grad_norm": 0.0007486707763746381,
"learning_rate": 1.4751123660752955e-06,
"loss": 0.002003737259656191,
"step": 1573
},
{
"epoch": 2.6155765831340334,
"grad_norm": 0.0005735998274758458,
"learning_rate": 1.4719809940166952e-06,
"loss": 0.001227132510393858,
"step": 1574
},
{
"epoch": 2.61724030362899,
"grad_norm": 0.00040597538463771343,
"learning_rate": 1.4688515618096252e-06,
"loss": 0.0011647650972008705,
"step": 1575
},
{
"epoch": 2.618904024123947,
"grad_norm": 0.0011245380155742168,
"learning_rate": 1.4657240753592627e-06,
"loss": 0.000804451119620353,
"step": 1576
},
{
"epoch": 2.620567744618904,
"grad_norm": 0.00036305448156781495,
"learning_rate": 1.462598540567113e-06,
"loss": 0.00045339000644162297,
"step": 1577
},
{
"epoch": 2.622231465113861,
"grad_norm": 0.0013733444502577186,
"learning_rate": 1.4594749633309981e-06,
"loss": 0.00537478644400835,
"step": 1578
},
{
"epoch": 2.6238951856088177,
"grad_norm": 0.0005763823282904923,
"learning_rate": 1.456353349545046e-06,
"loss": 0.0016458909958600998,
"step": 1579
},
{
"epoch": 2.6255589061037745,
"grad_norm": 0.0003702778776641935,
"learning_rate": 1.4532337050996804e-06,
"loss": 0.0006556995795108378,
"step": 1580
},
{
"epoch": 2.6272226265987313,
"grad_norm": 0.00021504143660422415,
"learning_rate": 1.4501160358816085e-06,
"loss": 0.0008495253277942538,
"step": 1581
},
{
"epoch": 2.6288863470936885,
"grad_norm": 0.0007969992002472281,
"learning_rate": 1.4470003477738111e-06,
"loss": 0.0007396379369311035,
"step": 1582
},
{
"epoch": 2.6305500675886453,
"grad_norm": 0.0006778067909181118,
"learning_rate": 1.4438866466555308e-06,
"loss": 0.00211876118555665,
"step": 1583
},
{
"epoch": 2.632213788083602,
"grad_norm": 0.0008974986267276108,
"learning_rate": 1.4407749384022576e-06,
"loss": 0.0029981655534356833,
"step": 1584
},
{
"epoch": 2.633877508578559,
"grad_norm": 0.0011483733542263508,
"learning_rate": 1.4376652288857249e-06,
"loss": 0.000901217048522085,
"step": 1585
},
{
"epoch": 2.6355412290735156,
"grad_norm": 0.0005572153022512794,
"learning_rate": 1.4345575239738928e-06,
"loss": 0.0009018393466249108,
"step": 1586
},
{
"epoch": 2.6372049495684724,
"grad_norm": 0.000704800128005445,
"learning_rate": 1.431451829530939e-06,
"loss": 0.0024085640907287598,
"step": 1587
},
{
"epoch": 2.638868670063429,
"grad_norm": 0.0006277773063629866,
"learning_rate": 1.4283481514172487e-06,
"loss": 0.0006163071375340223,
"step": 1588
},
{
"epoch": 2.640532390558386,
"grad_norm": 0.0006596968742087483,
"learning_rate": 1.425246495489399e-06,
"loss": 0.0014800459612160921,
"step": 1589
},
{
"epoch": 2.642196111053343,
"grad_norm": 0.0012964921770617366,
"learning_rate": 1.4221468676001544e-06,
"loss": 0.0024225322995334864,
"step": 1590
},
{
"epoch": 2.6438598315483,
"grad_norm": 0.0008854639600031078,
"learning_rate": 1.419049273598451e-06,
"loss": 0.0015666240360587835,
"step": 1591
},
{
"epoch": 2.6455235520432567,
"grad_norm": 0.00041664467426016927,
"learning_rate": 1.4159537193293876e-06,
"loss": 0.0004793188127223402,
"step": 1592
},
{
"epoch": 2.6471872725382135,
"grad_norm": 0.00037037362926639616,
"learning_rate": 1.4128602106342154e-06,
"loss": 0.0007164277485571802,
"step": 1593
},
{
"epoch": 2.6488509930331703,
"grad_norm": 0.0005810911534354091,
"learning_rate": 1.4097687533503213e-06,
"loss": 0.0017024344997480512,
"step": 1594
},
{
"epoch": 2.6505147135281275,
"grad_norm": 0.0015127704245969653,
"learning_rate": 1.4066793533112255e-06,
"loss": 0.003262751968577504,
"step": 1595
},
{
"epoch": 2.6521784340230843,
"grad_norm": 0.0008967657340690494,
"learning_rate": 1.4035920163465648e-06,
"loss": 0.0025144058745354414,
"step": 1596
},
{
"epoch": 2.653842154518041,
"grad_norm": 0.0009657472255639732,
"learning_rate": 1.400506748282083e-06,
"loss": 0.002190806670114398,
"step": 1597
},
{
"epoch": 2.655505875012998,
"grad_norm": 0.0011267064837738872,
"learning_rate": 1.3974235549396198e-06,
"loss": 0.0018630200065672398,
"step": 1598
},
{
"epoch": 2.6571695955079546,
"grad_norm": 0.0006663509993813932,
"learning_rate": 1.3943424421370998e-06,
"loss": 0.0017727947561070323,
"step": 1599
},
{
"epoch": 2.6588333160029114,
"grad_norm": 0.0015259669162333012,
"learning_rate": 1.3912634156885235e-06,
"loss": 0.004836976062506437,
"step": 1600
},
{
"epoch": 2.6588333160029114,
"eval_loss": 0.002130915177986026,
"eval_runtime": 491.4387,
"eval_samples_per_second": 9.784,
"eval_steps_per_second": 2.446,
"step": 1600
},
{
"epoch": 2.660497036497868,
"grad_norm": 0.0005084536387585104,
"learning_rate": 1.3881864814039503e-06,
"loss": 0.0009671523002907634,
"step": 1601
},
{
"epoch": 2.662160756992825,
"grad_norm": 0.000593173666857183,
"learning_rate": 1.3851116450894959e-06,
"loss": 0.0017778316978365183,
"step": 1602
},
{
"epoch": 2.663824477487782,
"grad_norm": 0.0002884052519220859,
"learning_rate": 1.382038912547315e-06,
"loss": 0.0005450017051771283,
"step": 1603
},
{
"epoch": 2.665488197982739,
"grad_norm": 0.0009035166003741324,
"learning_rate": 1.3789682895755935e-06,
"loss": 0.0014512368943542242,
"step": 1604
},
{
"epoch": 2.6671519184776957,
"grad_norm": 0.0010885243536904454,
"learning_rate": 1.3758997819685366e-06,
"loss": 0.002957823220640421,
"step": 1605
},
{
"epoch": 2.6688156389726525,
"grad_norm": 0.0012139276368543506,
"learning_rate": 1.3728333955163565e-06,
"loss": 0.0007539030630141497,
"step": 1606
},
{
"epoch": 2.6704793594676093,
"grad_norm": 0.0011482624104246497,
"learning_rate": 1.3697691360052646e-06,
"loss": 0.0011043865233659744,
"step": 1607
},
{
"epoch": 2.6721430799625665,
"grad_norm": 0.00046592974103987217,
"learning_rate": 1.3667070092174587e-06,
"loss": 0.0020422644447535276,
"step": 1608
},
{
"epoch": 2.6738068004575233,
"grad_norm": 0.0017777342582121491,
"learning_rate": 1.3636470209311093e-06,
"loss": 0.0010365403722971678,
"step": 1609
},
{
"epoch": 2.67547052095248,
"grad_norm": 0.001754077966324985,
"learning_rate": 1.360589176920355e-06,
"loss": 0.0025688004679977894,
"step": 1610
},
{
"epoch": 2.677134241447437,
"grad_norm": 0.0008552870131097734,
"learning_rate": 1.357533482955287e-06,
"loss": 0.0021906231995671988,
"step": 1611
},
{
"epoch": 2.6787979619423936,
"grad_norm": 0.0007275700918398798,
"learning_rate": 1.354479944801939e-06,
"loss": 0.0016428211238235235,
"step": 1612
},
{
"epoch": 2.6804616824373504,
"grad_norm": 0.0017000288935378194,
"learning_rate": 1.3514285682222777e-06,
"loss": 0.0026704478077590466,
"step": 1613
},
{
"epoch": 2.682125402932307,
"grad_norm": 0.0007832667324692011,
"learning_rate": 1.3483793589741901e-06,
"loss": 0.001527699758298695,
"step": 1614
},
{
"epoch": 2.683789123427264,
"grad_norm": 0.0010062839137390256,
"learning_rate": 1.3453323228114745e-06,
"loss": 0.0007239821134135127,
"step": 1615
},
{
"epoch": 2.685452843922221,
"grad_norm": 0.0005412401515059173,
"learning_rate": 1.3422874654838263e-06,
"loss": 0.0009945163037627935,
"step": 1616
},
{
"epoch": 2.687116564417178,
"grad_norm": 0.0015882287407293916,
"learning_rate": 1.3392447927368315e-06,
"loss": 0.001623464049771428,
"step": 1617
},
{
"epoch": 2.6887802849121347,
"grad_norm": 0.0002611389209050685,
"learning_rate": 1.3362043103119537e-06,
"loss": 0.0006480780430138111,
"step": 1618
},
{
"epoch": 2.6904440054070915,
"grad_norm": 0.00046936588478274643,
"learning_rate": 1.3331660239465232e-06,
"loss": 0.0008513991488143802,
"step": 1619
},
{
"epoch": 2.6921077259020487,
"grad_norm": 0.0010908697731792927,
"learning_rate": 1.3301299393737262e-06,
"loss": 0.003656618529930711,
"step": 1620
},
{
"epoch": 2.6937714463970055,
"grad_norm": 0.0030978077556937933,
"learning_rate": 1.3270960623225953e-06,
"loss": 0.0021632835268974304,
"step": 1621
},
{
"epoch": 2.6954351668919623,
"grad_norm": 0.0015169283142313361,
"learning_rate": 1.324064398517994e-06,
"loss": 0.0006730949389748275,
"step": 1622
},
{
"epoch": 2.697098887386919,
"grad_norm": 0.0005174805992282927,
"learning_rate": 1.3210349536806138e-06,
"loss": 0.0017688910011202097,
"step": 1623
},
{
"epoch": 2.698762607881876,
"grad_norm": 0.00047087905113585293,
"learning_rate": 1.3180077335269565e-06,
"loss": 0.0015057477867230773,
"step": 1624
},
{
"epoch": 2.7004263283768326,
"grad_norm": 0.0007247462053783238,
"learning_rate": 1.3149827437693267e-06,
"loss": 0.0026692950632423162,
"step": 1625
},
{
"epoch": 2.7020900488717894,
"grad_norm": 0.0009361693519167602,
"learning_rate": 1.3119599901158214e-06,
"loss": 0.001273240428417921,
"step": 1626
},
{
"epoch": 2.703753769366746,
"grad_norm": 0.0006839700508862734,
"learning_rate": 1.3089394782703152e-06,
"loss": 0.002276037121191621,
"step": 1627
},
{
"epoch": 2.7054174898617034,
"grad_norm": 0.00048088282346725464,
"learning_rate": 1.3059212139324548e-06,
"loss": 0.0010588760487735271,
"step": 1628
},
{
"epoch": 2.70708121035666,
"grad_norm": 0.0007159346132539213,
"learning_rate": 1.3029052027976457e-06,
"loss": 0.0015596949961036444,
"step": 1629
},
{
"epoch": 2.708744930851617,
"grad_norm": 0.0007666469318792224,
"learning_rate": 1.299891450557041e-06,
"loss": 0.003373862709850073,
"step": 1630
},
{
"epoch": 2.7104086513465737,
"grad_norm": 0.0009571753907948732,
"learning_rate": 1.2968799628975311e-06,
"loss": 0.0008542872965335846,
"step": 1631
},
{
"epoch": 2.7120723718415305,
"grad_norm": 0.002131329383701086,
"learning_rate": 1.2938707455017358e-06,
"loss": 0.002327835885807872,
"step": 1632
},
{
"epoch": 2.7137360923364877,
"grad_norm": 0.0014738694299012423,
"learning_rate": 1.2908638040479855e-06,
"loss": 0.0012567926896736026,
"step": 1633
},
{
"epoch": 2.7153998128314445,
"grad_norm": 0.0009184948867186904,
"learning_rate": 1.2878591442103215e-06,
"loss": 0.0025412426330149174,
"step": 1634
},
{
"epoch": 2.7170635333264013,
"grad_norm": 0.0003600319323595613,
"learning_rate": 1.2848567716584764e-06,
"loss": 0.0004601909895427525,
"step": 1635
},
{
"epoch": 2.718727253821358,
"grad_norm": 0.0003036624111700803,
"learning_rate": 1.2818566920578684e-06,
"loss": 0.0005614445544779301,
"step": 1636
},
{
"epoch": 2.720390974316315,
"grad_norm": 0.0017151448410004377,
"learning_rate": 1.2788589110695896e-06,
"loss": 0.0019237557426095009,
"step": 1637
},
{
"epoch": 2.7220546948112716,
"grad_norm": 0.00035357868182472885,
"learning_rate": 1.275863434350391e-06,
"loss": 0.0011692875996232033,
"step": 1638
},
{
"epoch": 2.7237184153062284,
"grad_norm": 0.0005569895729422569,
"learning_rate": 1.2728702675526788e-06,
"loss": 0.001136164297349751,
"step": 1639
},
{
"epoch": 2.725382135801185,
"grad_norm": 0.0008662780746817589,
"learning_rate": 1.2698794163244998e-06,
"loss": 0.00196397234685719,
"step": 1640
},
{
"epoch": 2.7270458562961424,
"grad_norm": 0.0009904677281156182,
"learning_rate": 1.2668908863095311e-06,
"loss": 0.00174479850102216,
"step": 1641
},
{
"epoch": 2.728709576791099,
"grad_norm": 0.001673137187026441,
"learning_rate": 1.2639046831470697e-06,
"loss": 0.0008900794200599194,
"step": 1642
},
{
"epoch": 2.730373297286056,
"grad_norm": 0.0008164556929841638,
"learning_rate": 1.2609208124720228e-06,
"loss": 0.0016282034339383245,
"step": 1643
},
{
"epoch": 2.7320370177810127,
"grad_norm": 0.000763962569180876,
"learning_rate": 1.2579392799148938e-06,
"loss": 0.00045784079702571034,
"step": 1644
},
{
"epoch": 2.7337007382759695,
"grad_norm": 0.0025477285962551832,
"learning_rate": 1.2549600911017761e-06,
"loss": 0.002230012556537986,
"step": 1645
},
{
"epoch": 2.7353644587709267,
"grad_norm": 0.0005968649056740105,
"learning_rate": 1.25198325165434e-06,
"loss": 0.0012856629909947515,
"step": 1646
},
{
"epoch": 2.7370281792658835,
"grad_norm": 0.0015948619693517685,
"learning_rate": 1.2490087671898234e-06,
"loss": 0.003677614266052842,
"step": 1647
},
{
"epoch": 2.7386918997608403,
"grad_norm": 0.0007306319894269109,
"learning_rate": 1.24603664332102e-06,
"loss": 0.0014641217421740294,
"step": 1648
},
{
"epoch": 2.740355620255797,
"grad_norm": 0.0003455659607425332,
"learning_rate": 1.243066885656267e-06,
"loss": 0.0004623204004019499,
"step": 1649
},
{
"epoch": 2.742019340750754,
"grad_norm": 0.0006565743242390454,
"learning_rate": 1.240099499799439e-06,
"loss": 0.0017107607563957572,
"step": 1650
},
{
"epoch": 2.7436830612457106,
"grad_norm": 0.0005443418631330132,
"learning_rate": 1.237134491349935e-06,
"loss": 0.001200504251755774,
"step": 1651
},
{
"epoch": 2.7453467817406674,
"grad_norm": 0.000680884812027216,
"learning_rate": 1.234171865902667e-06,
"loss": 0.0012294517364352942,
"step": 1652
},
{
"epoch": 2.747010502235624,
"grad_norm": 0.0007219392573460937,
"learning_rate": 1.2312116290480506e-06,
"loss": 0.001784635242074728,
"step": 1653
},
{
"epoch": 2.7486742227305814,
"grad_norm": 0.001259760232642293,
"learning_rate": 1.228253786371995e-06,
"loss": 0.002943810075521469,
"step": 1654
},
{
"epoch": 2.750337943225538,
"grad_norm": 0.0005466834991239011,
"learning_rate": 1.2252983434558894e-06,
"loss": 0.0013657533563673496,
"step": 1655
},
{
"epoch": 2.752001663720495,
"grad_norm": 0.0007785367197357118,
"learning_rate": 1.2223453058765966e-06,
"loss": 0.0010540152434259653,
"step": 1656
},
{
"epoch": 2.7536653842154517,
"grad_norm": 0.0008473260677419603,
"learning_rate": 1.2193946792064403e-06,
"loss": 0.0014812792651355267,
"step": 1657
},
{
"epoch": 2.755329104710409,
"grad_norm": 0.0009911999804899096,
"learning_rate": 1.2164464690131947e-06,
"loss": 0.0032296220306307077,
"step": 1658
},
{
"epoch": 2.7569928252053657,
"grad_norm": 0.0007574434275738895,
"learning_rate": 1.2135006808600752e-06,
"loss": 0.001332910731434822,
"step": 1659
},
{
"epoch": 2.7586565457003225,
"grad_norm": 0.00160769815556705,
"learning_rate": 1.2105573203057233e-06,
"loss": 0.0020911500323563814,
"step": 1660
},
{
"epoch": 2.7603202661952793,
"grad_norm": 0.000817932712379843,
"learning_rate": 1.207616392904204e-06,
"loss": 0.0013863914646208286,
"step": 1661
},
{
"epoch": 2.761983986690236,
"grad_norm": 0.000839824031572789,
"learning_rate": 1.2046779042049883e-06,
"loss": 0.002464557997882366,
"step": 1662
},
{
"epoch": 2.763647707185193,
"grad_norm": 0.0009035566472448409,
"learning_rate": 1.2017418597529464e-06,
"loss": 0.0007109399302862585,
"step": 1663
},
{
"epoch": 2.7653114276801496,
"grad_norm": 0.0004544042458292097,
"learning_rate": 1.1988082650883376e-06,
"loss": 0.0015377543168142438,
"step": 1664
},
{
"epoch": 2.7669751481751064,
"grad_norm": 0.0002455816720612347,
"learning_rate": 1.1958771257467946e-06,
"loss": 0.0011990536004304886,
"step": 1665
},
{
"epoch": 2.7686388686700636,
"grad_norm": 0.00031565967947244644,
"learning_rate": 1.1929484472593205e-06,
"loss": 0.0004553858598228544,
"step": 1666
},
{
"epoch": 2.7703025891650204,
"grad_norm": 0.0012521336320787668,
"learning_rate": 1.190022235152274e-06,
"loss": 0.0022942747455090284,
"step": 1667
},
{
"epoch": 2.771966309659977,
"grad_norm": 0.0004633648495655507,
"learning_rate": 1.1870984949473586e-06,
"loss": 0.0014716461300849915,
"step": 1668
},
{
"epoch": 2.773630030154934,
"grad_norm": 0.0006363751017488539,
"learning_rate": 1.184177232161615e-06,
"loss": 0.0014855836052447557,
"step": 1669
},
{
"epoch": 2.7752937506498907,
"grad_norm": 0.0014003062387928367,
"learning_rate": 1.1812584523074089e-06,
"loss": 0.002873154589906335,
"step": 1670
},
{
"epoch": 2.776957471144848,
"grad_norm": 0.0004970761365257204,
"learning_rate": 1.1783421608924183e-06,
"loss": 0.0014596062246710062,
"step": 1671
},
{
"epoch": 2.7786211916398047,
"grad_norm": 0.0008416569326072931,
"learning_rate": 1.1754283634196285e-06,
"loss": 0.0005961301503702998,
"step": 1672
},
{
"epoch": 2.7802849121347615,
"grad_norm": 0.000693576701451093,
"learning_rate": 1.1725170653873174e-06,
"loss": 0.0010480983182787895,
"step": 1673
},
{
"epoch": 2.7819486326297183,
"grad_norm": 0.0005453251651488245,
"learning_rate": 1.1696082722890474e-06,
"loss": 0.0005093438667245209,
"step": 1674
},
{
"epoch": 2.783612353124675,
"grad_norm": 0.001792008988559246,
"learning_rate": 1.1667019896136539e-06,
"loss": 0.004601791501045227,
"step": 1675
},
{
"epoch": 2.785276073619632,
"grad_norm": 0.0014838258503004909,
"learning_rate": 1.1637982228452329e-06,
"loss": 0.002853119745850563,
"step": 1676
},
{
"epoch": 2.7869397941145886,
"grad_norm": 0.0009499151492491364,
"learning_rate": 1.1608969774631366e-06,
"loss": 0.002817405853420496,
"step": 1677
},
{
"epoch": 2.7886035146095454,
"grad_norm": 0.0015111935790628195,
"learning_rate": 1.1579982589419568e-06,
"loss": 0.0013510992284864187,
"step": 1678
},
{
"epoch": 2.7902672351045026,
"grad_norm": 0.0015526446513831615,
"learning_rate": 1.155102072751518e-06,
"loss": 0.0025683874264359474,
"step": 1679
},
{
"epoch": 2.7919309555994594,
"grad_norm": 0.0003567964013200253,
"learning_rate": 1.152208424356867e-06,
"loss": 0.0005328885745257139,
"step": 1680
},
{
"epoch": 2.793594676094416,
"grad_norm": 0.0012884798925369978,
"learning_rate": 1.1493173192182613e-06,
"loss": 0.0021837963722646236,
"step": 1681
},
{
"epoch": 2.795258396589373,
"grad_norm": 0.0006896915147081017,
"learning_rate": 1.1464287627911577e-06,
"loss": 0.0025080733466893435,
"step": 1682
},
{
"epoch": 2.7969221170843297,
"grad_norm": 0.001546475337818265,
"learning_rate": 1.1435427605262057e-06,
"loss": 0.0012556775473058224,
"step": 1683
},
{
"epoch": 2.798585837579287,
"grad_norm": 0.0012773580383509398,
"learning_rate": 1.1406593178692346e-06,
"loss": 0.0026327914092689753,
"step": 1684
},
{
"epoch": 2.8002495580742437,
"grad_norm": 0.0011847744463011622,
"learning_rate": 1.1377784402612439e-06,
"loss": 0.002403012476861477,
"step": 1685
},
{
"epoch": 2.8019132785692005,
"grad_norm": 0.0005747679970227182,
"learning_rate": 1.1349001331383921e-06,
"loss": 0.001767489593476057,
"step": 1686
},
{
"epoch": 2.8035769990641572,
"grad_norm": 0.0015463011804968119,
"learning_rate": 1.132024401931988e-06,
"loss": 0.002241876907646656,
"step": 1687
},
{
"epoch": 2.805240719559114,
"grad_norm": 0.0010779668809846044,
"learning_rate": 1.12915125206848e-06,
"loss": 0.0010287202894687653,
"step": 1688
},
{
"epoch": 2.806904440054071,
"grad_norm": 0.0007248444017022848,
"learning_rate": 1.1262806889694455e-06,
"loss": 0.0018740869127213955,
"step": 1689
},
{
"epoch": 2.8085681605490276,
"grad_norm": 0.000944844214245677,
"learning_rate": 1.1234127180515787e-06,
"loss": 0.0012290524318814278,
"step": 1690
},
{
"epoch": 2.8102318810439844,
"grad_norm": 0.0008783553494140506,
"learning_rate": 1.1205473447266843e-06,
"loss": 0.0016621832037344575,
"step": 1691
},
{
"epoch": 2.8118956015389416,
"grad_norm": 0.0011562372092157602,
"learning_rate": 1.117684574401666e-06,
"loss": 0.0015901480801403522,
"step": 1692
},
{
"epoch": 2.8135593220338984,
"grad_norm": 0.0003519030869938433,
"learning_rate": 1.1148244124785143e-06,
"loss": 0.0010819558519870043,
"step": 1693
},
{
"epoch": 2.815223042528855,
"grad_norm": 0.0007611211040057242,
"learning_rate": 1.111966864354298e-06,
"loss": 0.0018490708898752928,
"step": 1694
},
{
"epoch": 2.816886763023812,
"grad_norm": 0.000434171553933993,
"learning_rate": 1.1091119354211544e-06,
"loss": 0.0010143747786059976,
"step": 1695
},
{
"epoch": 2.8185504835187687,
"grad_norm": 0.0005083135329186916,
"learning_rate": 1.1062596310662775e-06,
"loss": 0.0012547963997349143,
"step": 1696
},
{
"epoch": 2.820214204013726,
"grad_norm": 0.000701703829690814,
"learning_rate": 1.1034099566719104e-06,
"loss": 0.001071410602889955,
"step": 1697
},
{
"epoch": 2.8218779245086827,
"grad_norm": 0.00045851178583689034,
"learning_rate": 1.1005629176153302e-06,
"loss": 0.0013043745420873165,
"step": 1698
},
{
"epoch": 2.8235416450036395,
"grad_norm": 0.0009692832245491445,
"learning_rate": 1.097718519268844e-06,
"loss": 0.0008316121529787779,
"step": 1699
},
{
"epoch": 2.8252053654985962,
"grad_norm": 0.0006598404725082219,
"learning_rate": 1.0948767669997762e-06,
"loss": 0.0016597459325566888,
"step": 1700
},
{
"epoch": 2.8252053654985962,
"eval_loss": 0.0020048110745847225,
"eval_runtime": 491.3675,
"eval_samples_per_second": 9.785,
"eval_steps_per_second": 2.446,
"step": 1700
},
{
"epoch": 2.826869085993553,
"grad_norm": 0.001109412987716496,
"learning_rate": 1.092037666170456e-06,
"loss": 0.004444441292434931,
"step": 1701
},
{
"epoch": 2.82853280648851,
"grad_norm": 0.0004603521665558219,
"learning_rate": 1.0892012221382115e-06,
"loss": 0.0005218712612986565,
"step": 1702
},
{
"epoch": 2.8301965269834666,
"grad_norm": 0.0006472832174040377,
"learning_rate": 1.0863674402553564e-06,
"loss": 0.0018780764658004045,
"step": 1703
},
{
"epoch": 2.8318602474784234,
"grad_norm": 0.00071156449848786,
"learning_rate": 1.08353632586918e-06,
"loss": 0.0007375496788881719,
"step": 1704
},
{
"epoch": 2.8335239679733806,
"grad_norm": 0.000391888344893232,
"learning_rate": 1.0807078843219395e-06,
"loss": 0.0014355393359437585,
"step": 1705
},
{
"epoch": 2.8351876884683374,
"grad_norm": 0.0008743875660002232,
"learning_rate": 1.077882120950849e-06,
"loss": 0.001603996381163597,
"step": 1706
},
{
"epoch": 2.836851408963294,
"grad_norm": 0.0015048839850351214,
"learning_rate": 1.0750590410880671e-06,
"loss": 0.001089327735826373,
"step": 1707
},
{
"epoch": 2.838515129458251,
"grad_norm": 0.0005386308184824884,
"learning_rate": 1.072238650060691e-06,
"loss": 0.0015150413382798433,
"step": 1708
},
{
"epoch": 2.840178849953208,
"grad_norm": 0.0011673681437969208,
"learning_rate": 1.0694209531907412e-06,
"loss": 0.0028161397203803062,
"step": 1709
},
{
"epoch": 2.841842570448165,
"grad_norm": 0.0003765317960642278,
"learning_rate": 1.0666059557951566e-06,
"loss": 0.0004952827002853155,
"step": 1710
},
{
"epoch": 2.8435062909431217,
"grad_norm": 0.0005612285458482802,
"learning_rate": 1.0637936631857815e-06,
"loss": 0.001651897095143795,
"step": 1711
},
{
"epoch": 2.8451700114380785,
"grad_norm": 0.0003087786608375609,
"learning_rate": 1.0609840806693567e-06,
"loss": 0.0002589740033727139,
"step": 1712
},
{
"epoch": 2.8468337319330352,
"grad_norm": 0.00024726163246668875,
"learning_rate": 1.0581772135475089e-06,
"loss": 0.0003322141710668802,
"step": 1713
},
{
"epoch": 2.848497452427992,
"grad_norm": 0.0005495923687703907,
"learning_rate": 1.0553730671167412e-06,
"loss": 0.000860578496940434,
"step": 1714
},
{
"epoch": 2.850161172922949,
"grad_norm": 0.0003994822036474943,
"learning_rate": 1.052571646668421e-06,
"loss": 0.00059411337133497,
"step": 1715
},
{
"epoch": 2.8518248934179056,
"grad_norm": 0.0005885502323508263,
"learning_rate": 1.0497729574887744e-06,
"loss": 0.0014929859898984432,
"step": 1716
},
{
"epoch": 2.853488613912863,
"grad_norm": 0.000602265412453562,
"learning_rate": 1.0469770048588723e-06,
"loss": 0.0020361102651804686,
"step": 1717
},
{
"epoch": 2.8551523344078196,
"grad_norm": 0.000675926567055285,
"learning_rate": 1.0441837940546217e-06,
"loss": 0.002011209260672331,
"step": 1718
},
{
"epoch": 2.8568160549027763,
"grad_norm": 0.0006539134192280471,
"learning_rate": 1.0413933303467578e-06,
"loss": 0.001942803617566824,
"step": 1719
},
{
"epoch": 2.858479775397733,
"grad_norm": 0.0003630979626905173,
"learning_rate": 1.038605619000828e-06,
"loss": 0.0002884640416596085,
"step": 1720
},
{
"epoch": 2.86014349589269,
"grad_norm": 0.00048402079846709967,
"learning_rate": 1.0358206652771896e-06,
"loss": 0.0006971699185669422,
"step": 1721
},
{
"epoch": 2.861807216387647,
"grad_norm": 0.000931443297304213,
"learning_rate": 1.033038474430995e-06,
"loss": 0.0030923194717615843,
"step": 1722
},
{
"epoch": 2.863470936882604,
"grad_norm": 0.0006461223820224404,
"learning_rate": 1.0302590517121835e-06,
"loss": 0.0008790866122581065,
"step": 1723
},
{
"epoch": 2.8651346573775607,
"grad_norm": 0.0007613158086314797,
"learning_rate": 1.0274824023654717e-06,
"loss": 0.001279466669075191,
"step": 1724
},
{
"epoch": 2.8667983778725175,
"grad_norm": 0.010180974379181862,
"learning_rate": 1.0247085316303401e-06,
"loss": 0.007522291969507933,
"step": 1725
},
{
"epoch": 2.8684620983674742,
"grad_norm": 0.00044024086673744023,
"learning_rate": 1.0219374447410289e-06,
"loss": 0.0016961991786956787,
"step": 1726
},
{
"epoch": 2.870125818862431,
"grad_norm": 0.00041707613854669034,
"learning_rate": 1.019169146926524e-06,
"loss": 0.0007171843899413943,
"step": 1727
},
{
"epoch": 2.871789539357388,
"grad_norm": 0.0010140544036403298,
"learning_rate": 1.016403643410549e-06,
"loss": 0.002007360104471445,
"step": 1728
},
{
"epoch": 2.8734532598523446,
"grad_norm": 0.0014105895534157753,
"learning_rate": 1.013640939411554e-06,
"loss": 0.0018892575753852725,
"step": 1729
},
{
"epoch": 2.875116980347302,
"grad_norm": 0.0012860961724072695,
"learning_rate": 1.010881040142708e-06,
"loss": 0.0023204863537102938,
"step": 1730
},
{
"epoch": 2.8767807008422586,
"grad_norm": 0.0006491722888313234,
"learning_rate": 1.0081239508118842e-06,
"loss": 0.0016327498015016317,
"step": 1731
},
{
"epoch": 2.8784444213372153,
"grad_norm": 0.0014453501207754016,
"learning_rate": 1.0053696766216566e-06,
"loss": 0.002087503205984831,
"step": 1732
},
{
"epoch": 2.880108141832172,
"grad_norm": 0.00037019955925643444,
"learning_rate": 1.0026182227692865e-06,
"loss": 0.0004213819629512727,
"step": 1733
},
{
"epoch": 2.881771862327129,
"grad_norm": 0.0007157890941016376,
"learning_rate": 9.998695944467127e-07,
"loss": 0.0005501360283233225,
"step": 1734
},
{
"epoch": 2.883435582822086,
"grad_norm": 0.000919701240491122,
"learning_rate": 9.97123796840543e-07,
"loss": 0.0014758924953639507,
"step": 1735
},
{
"epoch": 2.885099303317043,
"grad_norm": 0.000745898752938956,
"learning_rate": 9.943808351320418e-07,
"loss": 0.0012118774466216564,
"step": 1736
},
{
"epoch": 2.8867630238119997,
"grad_norm": 0.0008136593387462199,
"learning_rate": 9.916407144971245e-07,
"loss": 0.0015056411502882838,
"step": 1737
},
{
"epoch": 2.8884267443069565,
"grad_norm": 0.0004307371564209461,
"learning_rate": 9.889034401063443e-07,
"loss": 0.0003619653289206326,
"step": 1738
},
{
"epoch": 2.8900904648019132,
"grad_norm": 0.0006141100311651826,
"learning_rate": 9.861690171248841e-07,
"loss": 0.001559469848871231,
"step": 1739
},
{
"epoch": 2.89175418529687,
"grad_norm": 0.0005701344343833625,
"learning_rate": 9.834374507125458e-07,
"loss": 0.0010183728300035,
"step": 1740
},
{
"epoch": 2.893417905791827,
"grad_norm": 0.0004537877975963056,
"learning_rate": 9.807087460237419e-07,
"loss": 0.000840195338241756,
"step": 1741
},
{
"epoch": 2.8950816262867836,
"grad_norm": 0.00036333585740067065,
"learning_rate": 9.779829082074827e-07,
"loss": 0.0005079791881144047,
"step": 1742
},
{
"epoch": 2.896745346781741,
"grad_norm": 0.0007203708519227803,
"learning_rate": 9.752599424073707e-07,
"loss": 0.0018293685279786587,
"step": 1743
},
{
"epoch": 2.8984090672766976,
"grad_norm": 0.0016104558017104864,
"learning_rate": 9.725398537615894e-07,
"loss": 0.0006681864615529776,
"step": 1744
},
{
"epoch": 2.9000727877716543,
"grad_norm": 0.00028292901697568595,
"learning_rate": 9.698226474028913e-07,
"loss": 0.0003127449890598655,
"step": 1745
},
{
"epoch": 2.901736508266611,
"grad_norm": 0.0007463365909643471,
"learning_rate": 9.671083284585925e-07,
"loss": 0.0006199987255968153,
"step": 1746
},
{
"epoch": 2.903400228761568,
"grad_norm": 0.0015855298843234777,
"learning_rate": 9.643969020505573e-07,
"loss": 0.0009128288365900517,
"step": 1747
},
{
"epoch": 2.905063949256525,
"grad_norm": 0.0008573647937737405,
"learning_rate": 9.616883732951945e-07,
"loss": 0.002327641937881708,
"step": 1748
},
{
"epoch": 2.906727669751482,
"grad_norm": 0.0010641596745699644,
"learning_rate": 9.589827473034443e-07,
"loss": 0.0026898360811173916,
"step": 1749
},
{
"epoch": 2.9083913902464387,
"grad_norm": 0.00045034498907625675,
"learning_rate": 9.562800291807695e-07,
"loss": 0.000989622320048511,
"step": 1750
},
{
"epoch": 2.9100551107413954,
"grad_norm": 0.00027441600104793906,
"learning_rate": 9.535802240271455e-07,
"loss": 0.0004929627757519484,
"step": 1751
},
{
"epoch": 2.9117188312363522,
"grad_norm": 0.000503926130477339,
"learning_rate": 9.508833369370524e-07,
"loss": 0.0007410001708194613,
"step": 1752
},
{
"epoch": 2.913382551731309,
"grad_norm": 0.00037136339233256876,
"learning_rate": 9.481893729994609e-07,
"loss": 0.0009830060880631208,
"step": 1753
},
{
"epoch": 2.9150462722262658,
"grad_norm": 0.0011238381266593933,
"learning_rate": 9.454983372978288e-07,
"loss": 0.0027857953682541847,
"step": 1754
},
{
"epoch": 2.916709992721223,
"grad_norm": 0.0020219890866428614,
"learning_rate": 9.428102349100868e-07,
"loss": 0.002705494174733758,
"step": 1755
},
{
"epoch": 2.9183737132161798,
"grad_norm": 0.0005653185653500259,
"learning_rate": 9.40125070908631e-07,
"loss": 0.001744154840707779,
"step": 1756
},
{
"epoch": 2.9200374337111366,
"grad_norm": 0.0004890793352387846,
"learning_rate": 9.374428503603139e-07,
"loss": 0.0008477336959913373,
"step": 1757
},
{
"epoch": 2.9217011542060933,
"grad_norm": 0.0013079852797091007,
"learning_rate": 9.347635783264309e-07,
"loss": 0.004897422157227993,
"step": 1758
},
{
"epoch": 2.92336487470105,
"grad_norm": 0.000551872537471354,
"learning_rate": 9.32087259862716e-07,
"loss": 0.0006481015589088202,
"step": 1759
},
{
"epoch": 2.9250285951960073,
"grad_norm": 0.0008821063674986362,
"learning_rate": 9.294139000193292e-07,
"loss": 0.0007773377583362162,
"step": 1760
},
{
"epoch": 2.926692315690964,
"grad_norm": 0.0018709349678829312,
"learning_rate": 9.267435038408479e-07,
"loss": 0.002509012818336487,
"step": 1761
},
{
"epoch": 2.928356036185921,
"grad_norm": 0.0023014312610030174,
"learning_rate": 9.240760763662562e-07,
"loss": 0.0031186698470264673,
"step": 1762
},
{
"epoch": 2.9300197566808777,
"grad_norm": 0.0033249291591346264,
"learning_rate": 9.214116226289388e-07,
"loss": 0.004345408175140619,
"step": 1763
},
{
"epoch": 2.9316834771758344,
"grad_norm": 0.0006871186196804047,
"learning_rate": 9.187501476566648e-07,
"loss": 0.0013219594256952405,
"step": 1764
},
{
"epoch": 2.933347197670791,
"grad_norm": 0.0013599384110420942,
"learning_rate": 9.16091656471586e-07,
"loss": 0.0026010123547166586,
"step": 1765
},
{
"epoch": 2.935010918165748,
"grad_norm": 0.001112336409278214,
"learning_rate": 9.134361540902225e-07,
"loss": 0.001636276487261057,
"step": 1766
},
{
"epoch": 2.9366746386607048,
"grad_norm": 0.00034071304253302515,
"learning_rate": 9.10783645523455e-07,
"loss": 0.0011249319650232792,
"step": 1767
},
{
"epoch": 2.938338359155662,
"grad_norm": 0.0002553804370108992,
"learning_rate": 9.081341357765145e-07,
"loss": 0.0011264397762715816,
"step": 1768
},
{
"epoch": 2.9400020796506188,
"grad_norm": 0.0016837348230183125,
"learning_rate": 9.054876298489742e-07,
"loss": 0.0035460232757031918,
"step": 1769
},
{
"epoch": 2.9416658001455755,
"grad_norm": 0.0009260933729819953,
"learning_rate": 9.02844132734737e-07,
"loss": 0.0019019779283553362,
"step": 1770
},
{
"epoch": 2.9433295206405323,
"grad_norm": 0.0003123321512248367,
"learning_rate": 9.002036494220306e-07,
"loss": 0.00030813238117843866,
"step": 1771
},
{
"epoch": 2.944993241135489,
"grad_norm": 0.0014501853147521615,
"learning_rate": 8.975661848933945e-07,
"loss": 0.0015067444182932377,
"step": 1772
},
{
"epoch": 2.9466569616304463,
"grad_norm": 0.0007959870272316039,
"learning_rate": 8.949317441256724e-07,
"loss": 0.0015829142648726702,
"step": 1773
},
{
"epoch": 2.948320682125403,
"grad_norm": 0.0006247299024835229,
"learning_rate": 8.923003320900014e-07,
"loss": 0.0010385081404820085,
"step": 1774
},
{
"epoch": 2.94998440262036,
"grad_norm": 0.0006999396136961877,
"learning_rate": 8.896719537518048e-07,
"loss": 0.0003225843538530171,
"step": 1775
},
{
"epoch": 2.9516481231153167,
"grad_norm": 0.001017800415866077,
"learning_rate": 8.870466140707795e-07,
"loss": 0.0015760234091430902,
"step": 1776
},
{
"epoch": 2.9533118436102734,
"grad_norm": 0.0008578760316595435,
"learning_rate": 8.844243180008913e-07,
"loss": 0.00260445149615407,
"step": 1777
},
{
"epoch": 2.95497556410523,
"grad_norm": 0.0011144928866997361,
"learning_rate": 8.818050704903589e-07,
"loss": 0.0008055091602727771,
"step": 1778
},
{
"epoch": 2.956639284600187,
"grad_norm": 0.0011398998321965337,
"learning_rate": 8.791888764816514e-07,
"loss": 0.0029183195438236,
"step": 1779
},
{
"epoch": 2.9583030050951438,
"grad_norm": 0.0003928740625269711,
"learning_rate": 8.765757409114753e-07,
"loss": 0.0009769850876182318,
"step": 1780
},
{
"epoch": 2.959966725590101,
"grad_norm": 0.0003030920051969588,
"learning_rate": 8.739656687107656e-07,
"loss": 0.00029227673076093197,
"step": 1781
},
{
"epoch": 2.9616304460850578,
"grad_norm": 0.00041401237831451,
"learning_rate": 8.713586648046768e-07,
"loss": 0.001756420824676752,
"step": 1782
},
{
"epoch": 2.9632941665800145,
"grad_norm": 0.0010973252356052399,
"learning_rate": 8.68754734112574e-07,
"loss": 0.004302047658711672,
"step": 1783
},
{
"epoch": 2.9649578870749713,
"grad_norm": 0.0006523855263367295,
"learning_rate": 8.661538815480228e-07,
"loss": 0.0008866323623806238,
"step": 1784
},
{
"epoch": 2.966621607569928,
"grad_norm": 0.0012584852520376444,
"learning_rate": 8.635561120187813e-07,
"loss": 0.001941533526405692,
"step": 1785
},
{
"epoch": 2.9682853280648853,
"grad_norm": 0.0008230031817220151,
"learning_rate": 8.609614304267877e-07,
"loss": 0.002135834889486432,
"step": 1786
},
{
"epoch": 2.969949048559842,
"grad_norm": 0.0012836785754188895,
"learning_rate": 8.583698416681555e-07,
"loss": 0.0027004701551049948,
"step": 1787
},
{
"epoch": 2.971612769054799,
"grad_norm": 0.00196822895668447,
"learning_rate": 8.557813506331616e-07,
"loss": 0.0021796643268316984,
"step": 1788
},
{
"epoch": 2.9732764895497557,
"grad_norm": 0.00046909775119274855,
"learning_rate": 8.531959622062372e-07,
"loss": 0.0008232493419200182,
"step": 1789
},
{
"epoch": 2.9749402100447124,
"grad_norm": 0.0011132569052278996,
"learning_rate": 8.506136812659601e-07,
"loss": 0.001676748739555478,
"step": 1790
},
{
"epoch": 2.976603930539669,
"grad_norm": 0.001434841426089406,
"learning_rate": 8.480345126850414e-07,
"loss": 0.0022043860517442226,
"step": 1791
},
{
"epoch": 2.978267651034626,
"grad_norm": 0.0004792559484485537,
"learning_rate": 8.454584613303227e-07,
"loss": 0.0009898152202367783,
"step": 1792
},
{
"epoch": 2.9799313715295828,
"grad_norm": 0.0005377150373533368,
"learning_rate": 8.428855320627613e-07,
"loss": 0.0008193760877475142,
"step": 1793
},
{
"epoch": 2.98159509202454,
"grad_norm": 0.00031275334185920656,
"learning_rate": 8.403157297374239e-07,
"loss": 0.0009223963716067374,
"step": 1794
},
{
"epoch": 2.9832588125194968,
"grad_norm": 0.0010333925019949675,
"learning_rate": 8.377490592034779e-07,
"loss": 0.0013084551319479942,
"step": 1795
},
{
"epoch": 2.9849225330144535,
"grad_norm": 0.0006100983009673655,
"learning_rate": 8.35185525304178e-07,
"loss": 0.0019747803453356028,
"step": 1796
},
{
"epoch": 2.9865862535094103,
"grad_norm": 0.000807909993454814,
"learning_rate": 8.326251328768626e-07,
"loss": 0.0023377123288810253,
"step": 1797
},
{
"epoch": 2.9882499740043675,
"grad_norm": 0.0010610126191750169,
"learning_rate": 8.300678867529415e-07,
"loss": 0.001045091892592609,
"step": 1798
},
{
"epoch": 2.9899136944993243,
"grad_norm": 0.0007304865284822881,
"learning_rate": 8.275137917578879e-07,
"loss": 0.0006219320348463953,
"step": 1799
},
{
"epoch": 2.991577414994281,
"grad_norm": 0.0004164436541032046,
"learning_rate": 8.249628527112282e-07,
"loss": 0.0018955947598442435,
"step": 1800
},
{
"epoch": 2.991577414994281,
"eval_loss": 0.0018649041885510087,
"eval_runtime": 490.6057,
"eval_samples_per_second": 9.8,
"eval_steps_per_second": 2.45,
"step": 1800
},
{
"epoch": 2.993241135489238,
"grad_norm": 0.000529412820469588,
"learning_rate": 8.224150744265352e-07,
"loss": 0.0011494626523926854,
"step": 1801
},
{
"epoch": 2.9949048559841946,
"grad_norm": 0.0014844761462882161,
"learning_rate": 8.198704617114143e-07,
"loss": 0.001377538195811212,
"step": 1802
},
{
"epoch": 2.9965685764791514,
"grad_norm": 0.0004270290955901146,
"learning_rate": 8.173290193674996e-07,
"loss": 0.0015947625506669283,
"step": 1803
},
{
"epoch": 2.998232296974108,
"grad_norm": 0.00046261935494840145,
"learning_rate": 8.147907521904433e-07,
"loss": 0.0006508635706268251,
"step": 1804
},
{
"epoch": 2.999896017469065,
"grad_norm": 0.0010813989210873842,
"learning_rate": 8.122556649699051e-07,
"loss": 0.0012900270521640778,
"step": 1805
},
{
"epoch": 3.0,
"grad_norm": 0.0009801322594285011,
"learning_rate": 8.097237624895452e-07,
"loss": 0.0008373880991712213,
"step": 1806
},
{
"epoch": 3.0016637204949568,
"grad_norm": 0.0006377885001711547,
"learning_rate": 8.07195049527012e-07,
"loss": 0.0012004249729216099,
"step": 1807
},
{
"epoch": 3.0033274409899136,
"grad_norm": 0.0005789811257272959,
"learning_rate": 8.046695308539376e-07,
"loss": 0.0004920345381833613,
"step": 1808
},
{
"epoch": 3.0049911614848703,
"grad_norm": 0.0002769168349914253,
"learning_rate": 8.021472112359255e-07,
"loss": 0.0006314446218311787,
"step": 1809
},
{
"epoch": 3.0066548819798276,
"grad_norm": 0.00032566802110522985,
"learning_rate": 7.996280954325433e-07,
"loss": 0.0009013297967612743,
"step": 1810
},
{
"epoch": 3.0083186024747843,
"grad_norm": 0.0004201479896437377,
"learning_rate": 7.971121881973126e-07,
"loss": 0.0005803474923595786,
"step": 1811
},
{
"epoch": 3.009982322969741,
"grad_norm": 0.0005800558137707412,
"learning_rate": 7.945994942777016e-07,
"loss": 0.0009451241930946708,
"step": 1812
},
{
"epoch": 3.011646043464698,
"grad_norm": 0.000676999450661242,
"learning_rate": 7.92090018415112e-07,
"loss": 0.0016937021864578128,
"step": 1813
},
{
"epoch": 3.0133097639596547,
"grad_norm": 0.0013515023747459054,
"learning_rate": 7.895837653448759e-07,
"loss": 0.0018922430463135242,
"step": 1814
},
{
"epoch": 3.0149734844546114,
"grad_norm": 0.0002872826880775392,
"learning_rate": 7.870807397962438e-07,
"loss": 0.0008894064230844378,
"step": 1815
},
{
"epoch": 3.0166372049495687,
"grad_norm": 0.0007774510886520147,
"learning_rate": 7.845809464923748e-07,
"loss": 0.0020803227089345455,
"step": 1816
},
{
"epoch": 3.0183009254445254,
"grad_norm": 0.0004965619882568717,
"learning_rate": 7.820843901503308e-07,
"loss": 0.001645309617742896,
"step": 1817
},
{
"epoch": 3.019964645939482,
"grad_norm": 0.00048233833513222635,
"learning_rate": 7.79591075481062e-07,
"loss": 0.002242596121504903,
"step": 1818
},
{
"epoch": 3.021628366434439,
"grad_norm": 0.0013223245041444898,
"learning_rate": 7.771010071894052e-07,
"loss": 0.0007967266719788313,
"step": 1819
},
{
"epoch": 3.0232920869293958,
"grad_norm": 0.00044782142504118383,
"learning_rate": 7.7461418997407e-07,
"loss": 0.0008925136062316597,
"step": 1820
},
{
"epoch": 3.0249558074243525,
"grad_norm": 0.0006108682719059289,
"learning_rate": 7.721306285276309e-07,
"loss": 0.000799420871771872,
"step": 1821
},
{
"epoch": 3.0266195279193098,
"grad_norm": 0.0005464836140163243,
"learning_rate": 7.696503275365194e-07,
"loss": 0.0025266679003834724,
"step": 1822
},
{
"epoch": 3.0282832484142665,
"grad_norm": 0.00018666086543817073,
"learning_rate": 7.671732916810154e-07,
"loss": 0.00021197514433879405,
"step": 1823
},
{
"epoch": 3.0299469689092233,
"grad_norm": 0.00032212541555054486,
"learning_rate": 7.646995256352346e-07,
"loss": 0.001422622473910451,
"step": 1824
},
{
"epoch": 3.03161068940418,
"grad_norm": 0.0002873683115467429,
"learning_rate": 7.622290340671256e-07,
"loss": 0.0008930754847824574,
"step": 1825
},
{
"epoch": 3.033274409899137,
"grad_norm": 0.0004030826676171273,
"learning_rate": 7.597618216384576e-07,
"loss": 0.0012631858699023724,
"step": 1826
},
{
"epoch": 3.0349381303940937,
"grad_norm": 0.0005824784748256207,
"learning_rate": 7.572978930048108e-07,
"loss": 0.000932716007810086,
"step": 1827
},
{
"epoch": 3.0366018508890504,
"grad_norm": 0.0015588699607178569,
"learning_rate": 7.54837252815571e-07,
"loss": 0.0011526403250172734,
"step": 1828
},
{
"epoch": 3.0382655713840077,
"grad_norm": 0.0006636729813180864,
"learning_rate": 7.523799057139158e-07,
"loss": 0.0007361185271292925,
"step": 1829
},
{
"epoch": 3.0399292918789644,
"grad_norm": 0.0008180178119800985,
"learning_rate": 7.49925856336812e-07,
"loss": 0.0009815450757741928,
"step": 1830
},
{
"epoch": 3.041593012373921,
"grad_norm": 0.0004374012351036072,
"learning_rate": 7.474751093150015e-07,
"loss": 0.0007682805880904198,
"step": 1831
},
{
"epoch": 3.043256732868878,
"grad_norm": 0.00045620554010383785,
"learning_rate": 7.450276692729957e-07,
"loss": 0.00044830897240899503,
"step": 1832
},
{
"epoch": 3.0449204533638348,
"grad_norm": 0.00020274176495149732,
"learning_rate": 7.425835408290655e-07,
"loss": 0.0008629217045381665,
"step": 1833
},
{
"epoch": 3.0465841738587915,
"grad_norm": 0.0003887891652993858,
"learning_rate": 7.40142728595234e-07,
"loss": 0.0013906953390687704,
"step": 1834
},
{
"epoch": 3.0482478943537488,
"grad_norm": 0.0005733828875236213,
"learning_rate": 7.377052371772637e-07,
"loss": 0.0016839816235005856,
"step": 1835
},
{
"epoch": 3.0499116148487055,
"grad_norm": 0.0010170930763706565,
"learning_rate": 7.352710711746536e-07,
"loss": 0.0007546496344730258,
"step": 1836
},
{
"epoch": 3.0515753353436623,
"grad_norm": 0.0003123862552456558,
"learning_rate": 7.328402351806269e-07,
"loss": 0.0006829963531345129,
"step": 1837
},
{
"epoch": 3.053239055838619,
"grad_norm": 0.0003148913965560496,
"learning_rate": 7.304127337821229e-07,
"loss": 0.000566673232242465,
"step": 1838
},
{
"epoch": 3.054902776333576,
"grad_norm": 0.00047699068090878427,
"learning_rate": 7.279885715597896e-07,
"loss": 0.0013908733380958438,
"step": 1839
},
{
"epoch": 3.0565664968285327,
"grad_norm": 0.0006750078173354268,
"learning_rate": 7.255677530879713e-07,
"loss": 0.0002647516957949847,
"step": 1840
},
{
"epoch": 3.05823021732349,
"grad_norm": 0.0004247261967975646,
"learning_rate": 7.231502829347056e-07,
"loss": 0.001544360420666635,
"step": 1841
},
{
"epoch": 3.0598939378184467,
"grad_norm": 0.00041044820682145655,
"learning_rate": 7.207361656617112e-07,
"loss": 0.0008406200795434415,
"step": 1842
},
{
"epoch": 3.0615576583134034,
"grad_norm": 0.0009824121370911598,
"learning_rate": 7.183254058243791e-07,
"loss": 0.0012988644884899259,
"step": 1843
},
{
"epoch": 3.06322137880836,
"grad_norm": 0.0019123097881674767,
"learning_rate": 7.159180079717656e-07,
"loss": 0.0016793514369055629,
"step": 1844
},
{
"epoch": 3.064885099303317,
"grad_norm": 0.000979103846475482,
"learning_rate": 7.135139766465838e-07,
"loss": 0.001676307525485754,
"step": 1845
},
{
"epoch": 3.0665488197982738,
"grad_norm": 0.0005059174727648497,
"learning_rate": 7.111133163851916e-07,
"loss": 0.001257180469110608,
"step": 1846
},
{
"epoch": 3.0682125402932305,
"grad_norm": 0.00040546999662183225,
"learning_rate": 7.087160317175881e-07,
"loss": 0.0006301425164565444,
"step": 1847
},
{
"epoch": 3.0698762607881878,
"grad_norm": 0.0005103031289763749,
"learning_rate": 7.06322127167402e-07,
"loss": 0.000969672342762351,
"step": 1848
},
{
"epoch": 3.0715399812831445,
"grad_norm": 0.0009169050026685,
"learning_rate": 7.03931607251884e-07,
"loss": 0.000986608793027699,
"step": 1849
},
{
"epoch": 3.0732037017781013,
"grad_norm": 0.0005506612360477448,
"learning_rate": 7.015444764818988e-07,
"loss": 0.0008717355667613447,
"step": 1850
},
{
"epoch": 3.074867422273058,
"grad_norm": 0.0004976927884854376,
"learning_rate": 6.991607393619129e-07,
"loss": 0.000616822624579072,
"step": 1851
},
{
"epoch": 3.076531142768015,
"grad_norm": 0.00012596679152920842,
"learning_rate": 6.967804003899925e-07,
"loss": 0.00013015244621783495,
"step": 1852
},
{
"epoch": 3.0781948632629716,
"grad_norm": 0.0006843673181720078,
"learning_rate": 6.944034640577896e-07,
"loss": 0.001166867557913065,
"step": 1853
},
{
"epoch": 3.079858583757929,
"grad_norm": 0.0006668114219792187,
"learning_rate": 6.920299348505365e-07,
"loss": 0.0013408659724518657,
"step": 1854
},
{
"epoch": 3.0815223042528856,
"grad_norm": 0.0005010065506212413,
"learning_rate": 6.896598172470356e-07,
"loss": 0.0011945900041610003,
"step": 1855
},
{
"epoch": 3.0831860247478424,
"grad_norm": 0.00048501609126105905,
"learning_rate": 6.872931157196519e-07,
"loss": 0.0008071433403529227,
"step": 1856
},
{
"epoch": 3.084849745242799,
"grad_norm": 0.00019211566541343927,
"learning_rate": 6.849298347343044e-07,
"loss": 0.00031206224230118096,
"step": 1857
},
{
"epoch": 3.086513465737756,
"grad_norm": 0.0005477959057316184,
"learning_rate": 6.825699787504586e-07,
"loss": 0.0014917771331965923,
"step": 1858
},
{
"epoch": 3.0881771862327128,
"grad_norm": 0.0004196336667519063,
"learning_rate": 6.802135522211142e-07,
"loss": 0.0007323448662646115,
"step": 1859
},
{
"epoch": 3.0898409067276695,
"grad_norm": 0.0008467534207738936,
"learning_rate": 6.778605595928025e-07,
"loss": 0.0009000095305964351,
"step": 1860
},
{
"epoch": 3.0915046272226268,
"grad_norm": 0.0008993221563287079,
"learning_rate": 6.755110053055738e-07,
"loss": 0.0020232279784977436,
"step": 1861
},
{
"epoch": 3.0931683477175835,
"grad_norm": 0.0012050456134602427,
"learning_rate": 6.731648937929911e-07,
"loss": 0.002171237487345934,
"step": 1862
},
{
"epoch": 3.0948320682125403,
"grad_norm": 0.000659894838463515,
"learning_rate": 6.708222294821196e-07,
"loss": 0.0014221647288650274,
"step": 1863
},
{
"epoch": 3.096495788707497,
"grad_norm": 0.0002312004507984966,
"learning_rate": 6.684830167935207e-07,
"loss": 0.0005635049310512841,
"step": 1864
},
{
"epoch": 3.098159509202454,
"grad_norm": 0.0006758279632776976,
"learning_rate": 6.66147260141243e-07,
"loss": 0.0010983545798808336,
"step": 1865
},
{
"epoch": 3.0998232296974106,
"grad_norm": 0.0014547808095812798,
"learning_rate": 6.638149639328134e-07,
"loss": 0.0019277246901765466,
"step": 1866
},
{
"epoch": 3.101486950192368,
"grad_norm": 0.0007520727231167257,
"learning_rate": 6.614861325692277e-07,
"loss": 0.0018726043635979295,
"step": 1867
},
{
"epoch": 3.1031506706873246,
"grad_norm": 0.0008481996483169496,
"learning_rate": 6.591607704449446e-07,
"loss": 0.0021393303759396076,
"step": 1868
},
{
"epoch": 3.1048143911822814,
"grad_norm": 0.0004936470650136471,
"learning_rate": 6.568388819478769e-07,
"loss": 0.0007193313795141876,
"step": 1869
},
{
"epoch": 3.106478111677238,
"grad_norm": 0.0017772550927475095,
"learning_rate": 6.545204714593825e-07,
"loss": 0.0028288543689996004,
"step": 1870
},
{
"epoch": 3.108141832172195,
"grad_norm": 0.0007597859366796911,
"learning_rate": 6.522055433542557e-07,
"loss": 0.002685034414753318,
"step": 1871
},
{
"epoch": 3.1098055526671518,
"grad_norm": 0.0007303763413801789,
"learning_rate": 6.49894102000721e-07,
"loss": 0.0016073291189968586,
"step": 1872
},
{
"epoch": 3.111469273162109,
"grad_norm": 0.0007561154197901487,
"learning_rate": 6.47586151760421e-07,
"loss": 0.0009440925787203014,
"step": 1873
},
{
"epoch": 3.1131329936570658,
"grad_norm": 0.0003546578227542341,
"learning_rate": 6.452816969884127e-07,
"loss": 0.0004244074516464025,
"step": 1874
},
{
"epoch": 3.1147967141520225,
"grad_norm": 0.00033709831768646836,
"learning_rate": 6.429807420331568e-07,
"loss": 0.00048436602810397744,
"step": 1875
},
{
"epoch": 3.1164604346469793,
"grad_norm": 0.0008623041212558746,
"learning_rate": 6.406832912365101e-07,
"loss": 0.0015984228812158108,
"step": 1876
},
{
"epoch": 3.118124155141936,
"grad_norm": 0.000693855807185173,
"learning_rate": 6.383893489337172e-07,
"loss": 0.0021479884162545204,
"step": 1877
},
{
"epoch": 3.119787875636893,
"grad_norm": 0.0004894212470389903,
"learning_rate": 6.360989194534004e-07,
"loss": 0.0007753630052320659,
"step": 1878
},
{
"epoch": 3.1214515961318496,
"grad_norm": 0.0005258218152448535,
"learning_rate": 6.338120071175558e-07,
"loss": 0.0011958642862737179,
"step": 1879
},
{
"epoch": 3.123115316626807,
"grad_norm": 0.00035203678999096155,
"learning_rate": 6.315286162415412e-07,
"loss": 0.0011395043693482876,
"step": 1880
},
{
"epoch": 3.1247790371217636,
"grad_norm": 0.0011336497263982892,
"learning_rate": 6.292487511340709e-07,
"loss": 0.0007257706602104008,
"step": 1881
},
{
"epoch": 3.1264427576167204,
"grad_norm": 0.0004972171154804528,
"learning_rate": 6.269724160972043e-07,
"loss": 0.0012949982192367315,
"step": 1882
},
{
"epoch": 3.128106478111677,
"grad_norm": 0.00041867265827022493,
"learning_rate": 6.246996154263421e-07,
"loss": 0.0004262304864823818,
"step": 1883
},
{
"epoch": 3.129770198606634,
"grad_norm": 0.0005103400908410549,
"learning_rate": 6.224303534102125e-07,
"loss": 0.0006815506494604051,
"step": 1884
},
{
"epoch": 3.1314339191015907,
"grad_norm": 0.0010413973359391093,
"learning_rate": 6.201646343308685e-07,
"loss": 0.0015371940098702908,
"step": 1885
},
{
"epoch": 3.133097639596548,
"grad_norm": 0.0015511265955865383,
"learning_rate": 6.179024624636772e-07,
"loss": 0.0027551804669201374,
"step": 1886
},
{
"epoch": 3.1347613600915047,
"grad_norm": 0.000395275535993278,
"learning_rate": 6.156438420773125e-07,
"loss": 0.0005117086693644524,
"step": 1887
},
{
"epoch": 3.1364250805864615,
"grad_norm": 0.0002733297587838024,
"learning_rate": 6.133887774337471e-07,
"loss": 0.0005519022815860808,
"step": 1888
},
{
"epoch": 3.1380888010814183,
"grad_norm": 0.0009390793857164681,
"learning_rate": 6.111372727882417e-07,
"loss": 0.0010190055472776294,
"step": 1889
},
{
"epoch": 3.139752521576375,
"grad_norm": 0.0008794846362434328,
"learning_rate": 6.088893323893419e-07,
"loss": 0.0028912704437971115,
"step": 1890
},
{
"epoch": 3.141416242071332,
"grad_norm": 0.00033567583886906505,
"learning_rate": 6.066449604788666e-07,
"loss": 0.0005888398736715317,
"step": 1891
},
{
"epoch": 3.143079962566289,
"grad_norm": 0.0005849747103638947,
"learning_rate": 6.044041612919016e-07,
"loss": 0.00027158469310961664,
"step": 1892
},
{
"epoch": 3.144743683061246,
"grad_norm": 0.0003590272390283644,
"learning_rate": 6.021669390567902e-07,
"loss": 0.0011505287839099765,
"step": 1893
},
{
"epoch": 3.1464074035562026,
"grad_norm": 0.00037003972101956606,
"learning_rate": 5.999332979951272e-07,
"loss": 0.001266230596229434,
"step": 1894
},
{
"epoch": 3.1480711240511594,
"grad_norm": 0.0006911141099408269,
"learning_rate": 5.977032423217482e-07,
"loss": 0.0023922715336084366,
"step": 1895
},
{
"epoch": 3.149734844546116,
"grad_norm": 0.00020799963385798037,
"learning_rate": 5.954767762447244e-07,
"loss": 0.0005746656679548323,
"step": 1896
},
{
"epoch": 3.151398565041073,
"grad_norm": 0.0002363747189519927,
"learning_rate": 5.932539039653535e-07,
"loss": 0.0006022070301696658,
"step": 1897
},
{
"epoch": 3.1530622855360297,
"grad_norm": 0.000486257195007056,
"learning_rate": 5.910346296781511e-07,
"loss": 0.0011896261712536216,
"step": 1898
},
{
"epoch": 3.154726006030987,
"grad_norm": 0.0006207071710377932,
"learning_rate": 5.888189575708453e-07,
"loss": 0.0014273447450250387,
"step": 1899
},
{
"epoch": 3.1563897265259437,
"grad_norm": 0.0006188419065438211,
"learning_rate": 5.866068918243634e-07,
"loss": 0.0006204223609529436,
"step": 1900
},
{
"epoch": 3.1563897265259437,
"eval_loss": 0.0018834135262295604,
"eval_runtime": 491.4511,
"eval_samples_per_second": 9.783,
"eval_steps_per_second": 2.446,
"step": 1900
},
{
"epoch": 3.1580534470209005,
"grad_norm": 0.0005833645700477064,
"learning_rate": 5.843984366128308e-07,
"loss": 0.0017470639431849122,
"step": 1901
},
{
"epoch": 3.1597171675158573,
"grad_norm": 0.0005380320362746716,
"learning_rate": 5.821935961035589e-07,
"loss": 0.0008071457268670201,
"step": 1902
},
{
"epoch": 3.161380888010814,
"grad_norm": 0.0004519163048826158,
"learning_rate": 5.799923744570376e-07,
"loss": 0.0010613186750561,
"step": 1903
},
{
"epoch": 3.163044608505771,
"grad_norm": 0.0016710092313587666,
"learning_rate": 5.777947758269295e-07,
"loss": 0.0009790394688025117,
"step": 1904
},
{
"epoch": 3.164708329000728,
"grad_norm": 0.0010518544586375356,
"learning_rate": 5.756008043600594e-07,
"loss": 0.0018751542083919048,
"step": 1905
},
{
"epoch": 3.166372049495685,
"grad_norm": 0.00041189463809132576,
"learning_rate": 5.734104641964075e-07,
"loss": 0.0011019426165148616,
"step": 1906
},
{
"epoch": 3.1680357699906416,
"grad_norm": 0.0005437190993689001,
"learning_rate": 5.712237594691028e-07,
"loss": 0.0007574702030979097,
"step": 1907
},
{
"epoch": 3.1696994904855984,
"grad_norm": 0.0003270143934059888,
"learning_rate": 5.690406943044138e-07,
"loss": 0.00037680071545764804,
"step": 1908
},
{
"epoch": 3.171363210980555,
"grad_norm": 0.00015396956587210298,
"learning_rate": 5.668612728217412e-07,
"loss": 0.0007919893832877278,
"step": 1909
},
{
"epoch": 3.173026931475512,
"grad_norm": 0.0005151602090336382,
"learning_rate": 5.646854991336112e-07,
"loss": 0.0012475823750719428,
"step": 1910
},
{
"epoch": 3.1746906519704687,
"grad_norm": 0.000625112559646368,
"learning_rate": 5.625133773456639e-07,
"loss": 0.0008614555117674172,
"step": 1911
},
{
"epoch": 3.176354372465426,
"grad_norm": 0.0008656211430206895,
"learning_rate": 5.603449115566511e-07,
"loss": 0.0023587134201079607,
"step": 1912
},
{
"epoch": 3.1780180929603827,
"grad_norm": 0.0003512735420372337,
"learning_rate": 5.581801058584252e-07,
"loss": 0.00029785794322378933,
"step": 1913
},
{
"epoch": 3.1796818134553395,
"grad_norm": 0.00034385884646326303,
"learning_rate": 5.560189643359312e-07,
"loss": 0.0004459562769625336,
"step": 1914
},
{
"epoch": 3.1813455339502963,
"grad_norm": 0.0002447869919706136,
"learning_rate": 5.538614910672005e-07,
"loss": 0.00018510760855861008,
"step": 1915
},
{
"epoch": 3.183009254445253,
"grad_norm": 0.0010491593275219202,
"learning_rate": 5.517076901233434e-07,
"loss": 0.0005361376097425818,
"step": 1916
},
{
"epoch": 3.18467297494021,
"grad_norm": 0.0007156440988183022,
"learning_rate": 5.495575655685382e-07,
"loss": 0.0016970396973192692,
"step": 1917
},
{
"epoch": 3.186336695435167,
"grad_norm": 0.001344272750429809,
"learning_rate": 5.474111214600278e-07,
"loss": 0.0010842399206012487,
"step": 1918
},
{
"epoch": 3.188000415930124,
"grad_norm": 0.0005655754939652979,
"learning_rate": 5.452683618481103e-07,
"loss": 0.001635242602787912,
"step": 1919
},
{
"epoch": 3.1896641364250806,
"grad_norm": 0.00028636507340706885,
"learning_rate": 5.431292907761305e-07,
"loss": 0.0009821431012824178,
"step": 1920
},
{
"epoch": 3.1913278569200374,
"grad_norm": 0.00041335669811815023,
"learning_rate": 5.409939122804736e-07,
"loss": 0.0004407725646160543,
"step": 1921
},
{
"epoch": 3.192991577414994,
"grad_norm": 0.0008381790830753744,
"learning_rate": 5.388622303905558e-07,
"loss": 0.0011460300302132964,
"step": 1922
},
{
"epoch": 3.194655297909951,
"grad_norm": 0.0005317120812833309,
"learning_rate": 5.367342491288186e-07,
"loss": 0.00032618624391034245,
"step": 1923
},
{
"epoch": 3.196319018404908,
"grad_norm": 0.0011242582695558667,
"learning_rate": 5.346099725107213e-07,
"loss": 0.002115800976753235,
"step": 1924
},
{
"epoch": 3.197982738899865,
"grad_norm": 0.00022066081874072552,
"learning_rate": 5.324894045447312e-07,
"loss": 0.0003675711923278868,
"step": 1925
},
{
"epoch": 3.1996464593948217,
"grad_norm": 0.00032584325526840985,
"learning_rate": 5.303725492323194e-07,
"loss": 0.0008598530548624694,
"step": 1926
},
{
"epoch": 3.2013101798897785,
"grad_norm": 0.0003403827140573412,
"learning_rate": 5.282594105679481e-07,
"loss": 0.0009377440437674522,
"step": 1927
},
{
"epoch": 3.2029739003847353,
"grad_norm": 0.000445442070486024,
"learning_rate": 5.261499925390692e-07,
"loss": 0.0010707492474466562,
"step": 1928
},
{
"epoch": 3.204637620879692,
"grad_norm": 0.00046939656022004783,
"learning_rate": 5.240442991261127e-07,
"loss": 0.0007110872538760304,
"step": 1929
},
{
"epoch": 3.2063013413746493,
"grad_norm": 0.00022708141477778554,
"learning_rate": 5.219423343024804e-07,
"loss": 0.000206417273147963,
"step": 1930
},
{
"epoch": 3.207965061869606,
"grad_norm": 0.0007522849482484162,
"learning_rate": 5.198441020345382e-07,
"loss": 0.00209865253418684,
"step": 1931
},
{
"epoch": 3.209628782364563,
"grad_norm": 0.0005657677538692951,
"learning_rate": 5.177496062816101e-07,
"loss": 0.0007334444671869278,
"step": 1932
},
{
"epoch": 3.2112925028595196,
"grad_norm": 0.00067489332286641,
"learning_rate": 5.156588509959659e-07,
"loss": 0.0007933019660413265,
"step": 1933
},
{
"epoch": 3.2129562233544764,
"grad_norm": 0.0009653370943851769,
"learning_rate": 5.13571840122821e-07,
"loss": 0.0015348546439781785,
"step": 1934
},
{
"epoch": 3.214619943849433,
"grad_norm": 0.0008980388520285487,
"learning_rate": 5.114885776003234e-07,
"loss": 0.002591199241578579,
"step": 1935
},
{
"epoch": 3.21628366434439,
"grad_norm": 0.0007487988914363086,
"learning_rate": 5.094090673595478e-07,
"loss": 0.0008501316187903285,
"step": 1936
},
{
"epoch": 3.217947384839347,
"grad_norm": 0.0011044201673939824,
"learning_rate": 5.073333133244896e-07,
"loss": 0.002894491655752063,
"step": 1937
},
{
"epoch": 3.219611105334304,
"grad_norm": 0.0007839900208637118,
"learning_rate": 5.052613194120554e-07,
"loss": 0.0013367477804422379,
"step": 1938
},
{
"epoch": 3.2212748258292607,
"grad_norm": 0.000503472751006484,
"learning_rate": 5.031930895320569e-07,
"loss": 0.000841394648887217,
"step": 1939
},
{
"epoch": 3.2229385463242175,
"grad_norm": 0.002264294307678938,
"learning_rate": 5.011286275872021e-07,
"loss": 0.0027972303796559572,
"step": 1940
},
{
"epoch": 3.2246022668191743,
"grad_norm": 0.00025088360416702926,
"learning_rate": 4.990679374730905e-07,
"loss": 0.0010514447931200266,
"step": 1941
},
{
"epoch": 3.226265987314131,
"grad_norm": 0.0012186086969450116,
"learning_rate": 4.970110230782035e-07,
"loss": 0.00212580687366426,
"step": 1942
},
{
"epoch": 3.2279297078090883,
"grad_norm": 0.0005759792984463274,
"learning_rate": 4.949578882838982e-07,
"loss": 0.0012910933000966907,
"step": 1943
},
{
"epoch": 3.229593428304045,
"grad_norm": 0.0003267624997533858,
"learning_rate": 4.929085369643988e-07,
"loss": 0.0011039895471185446,
"step": 1944
},
{
"epoch": 3.231257148799002,
"grad_norm": 0.00042506324825808406,
"learning_rate": 4.908629729867908e-07,
"loss": 0.0007888518157415092,
"step": 1945
},
{
"epoch": 3.2329208692939586,
"grad_norm": 0.0010529316496104002,
"learning_rate": 4.88821200211014e-07,
"loss": 0.0011536410311236978,
"step": 1946
},
{
"epoch": 3.2345845897889154,
"grad_norm": 0.0015189244877547026,
"learning_rate": 4.867832224898517e-07,
"loss": 0.0011142066214233637,
"step": 1947
},
{
"epoch": 3.236248310283872,
"grad_norm": 0.00029045421979390085,
"learning_rate": 4.847490436689281e-07,
"loss": 0.000906751025468111,
"step": 1948
},
{
"epoch": 3.237912030778829,
"grad_norm": 0.00041969105950556695,
"learning_rate": 4.827186675866985e-07,
"loss": 0.001455049030482769,
"step": 1949
},
{
"epoch": 3.239575751273786,
"grad_norm": 0.00031135749304667115,
"learning_rate": 4.806920980744426e-07,
"loss": 0.00046373007353395224,
"step": 1950
},
{
"epoch": 3.241239471768743,
"grad_norm": 0.0017488739686086774,
"learning_rate": 4.786693389562566e-07,
"loss": 0.003784562461078167,
"step": 1951
},
{
"epoch": 3.2429031922636997,
"grad_norm": 0.0002576555125415325,
"learning_rate": 4.7665039404904747e-07,
"loss": 0.0005821231752634048,
"step": 1952
},
{
"epoch": 3.2445669127586565,
"grad_norm": 0.0005598700372502208,
"learning_rate": 4.746352671625237e-07,
"loss": 0.0012865583412349224,
"step": 1953
},
{
"epoch": 3.2462306332536133,
"grad_norm": 0.0003530400281306356,
"learning_rate": 4.72623962099191e-07,
"loss": 0.0007247515022754669,
"step": 1954
},
{
"epoch": 3.24789435374857,
"grad_norm": 0.0004015358863398433,
"learning_rate": 4.7061648265434053e-07,
"loss": 0.00036378641379997134,
"step": 1955
},
{
"epoch": 3.2495580742435273,
"grad_norm": 0.0006843737210147083,
"learning_rate": 4.6861283261604745e-07,
"loss": 0.0021625924855470657,
"step": 1956
},
{
"epoch": 3.251221794738484,
"grad_norm": 0.0014490418834611773,
"learning_rate": 4.666130157651594e-07,
"loss": 0.0012110623065382242,
"step": 1957
},
{
"epoch": 3.252885515233441,
"grad_norm": 0.0009911368833854795,
"learning_rate": 4.6461703587529106e-07,
"loss": 0.0018840692937374115,
"step": 1958
},
{
"epoch": 3.2545492357283976,
"grad_norm": 0.00027222695644013584,
"learning_rate": 4.62624896712818e-07,
"loss": 0.0001918284106068313,
"step": 1959
},
{
"epoch": 3.2562129562233544,
"grad_norm": 0.0007084309472702444,
"learning_rate": 4.6063660203686635e-07,
"loss": 0.0016576352063566446,
"step": 1960
},
{
"epoch": 3.257876676718311,
"grad_norm": 0.00021952651150058955,
"learning_rate": 4.586521555993087e-07,
"loss": 0.0004061336221639067,
"step": 1961
},
{
"epoch": 3.259540397213268,
"grad_norm": 0.0003330149920657277,
"learning_rate": 4.5667156114475695e-07,
"loss": 0.0006381084094755352,
"step": 1962
},
{
"epoch": 3.261204117708225,
"grad_norm": 0.0007844572537578642,
"learning_rate": 4.5469482241055324e-07,
"loss": 0.002945750020444393,
"step": 1963
},
{
"epoch": 3.262867838203182,
"grad_norm": 0.0007807101937942207,
"learning_rate": 4.527219431267646e-07,
"loss": 0.0010229629697278142,
"step": 1964
},
{
"epoch": 3.2645315586981387,
"grad_norm": 0.002043907530605793,
"learning_rate": 4.507529270161759e-07,
"loss": 0.004027180373668671,
"step": 1965
},
{
"epoch": 3.2661952791930955,
"grad_norm": 0.0008136004325933754,
"learning_rate": 4.4878777779428034e-07,
"loss": 0.0009637029143050313,
"step": 1966
},
{
"epoch": 3.2678589996880523,
"grad_norm": 0.0036568024661391973,
"learning_rate": 4.4682649916927614e-07,
"loss": 0.0013857954181730747,
"step": 1967
},
{
"epoch": 3.2695227201830095,
"grad_norm": 0.0008362821536138654,
"learning_rate": 4.4486909484205725e-07,
"loss": 0.0020736800506711006,
"step": 1968
},
{
"epoch": 3.2711864406779663,
"grad_norm": 0.0007145930430851877,
"learning_rate": 4.429155685062073e-07,
"loss": 0.00082812225446105,
"step": 1969
},
{
"epoch": 3.272850161172923,
"grad_norm": 0.0006530311075039208,
"learning_rate": 4.409659238479919e-07,
"loss": 0.0017145945457741618,
"step": 1970
},
{
"epoch": 3.27451388166788,
"grad_norm": 0.00036001464468427,
"learning_rate": 4.39020164546351e-07,
"loss": 0.0006193873705342412,
"step": 1971
},
{
"epoch": 3.2761776021628366,
"grad_norm": 0.0005106754833832383,
"learning_rate": 4.370782942728946e-07,
"loss": 0.0015185193624347448,
"step": 1972
},
{
"epoch": 3.2778413226577934,
"grad_norm": 0.0010290957288816571,
"learning_rate": 4.3514031669189325e-07,
"loss": 0.0004486526013351977,
"step": 1973
},
{
"epoch": 3.27950504315275,
"grad_norm": 0.000586140900850296,
"learning_rate": 4.3320623546027283e-07,
"loss": 0.00030024960869923234,
"step": 1974
},
{
"epoch": 3.2811687636477074,
"grad_norm": 0.002518952591344714,
"learning_rate": 4.312760542276059e-07,
"loss": 0.0011890861205756664,
"step": 1975
},
{
"epoch": 3.282832484142664,
"grad_norm": 0.0009471022640354931,
"learning_rate": 4.293497766361068e-07,
"loss": 0.002298705279827118,
"step": 1976
},
{
"epoch": 3.284496204637621,
"grad_norm": 0.00030820450047031045,
"learning_rate": 4.2742740632062243e-07,
"loss": 0.0003215964825358242,
"step": 1977
},
{
"epoch": 3.2861599251325777,
"grad_norm": 0.0006197311449795961,
"learning_rate": 4.255089469086279e-07,
"loss": 0.0013243852881714702,
"step": 1978
},
{
"epoch": 3.2878236456275345,
"grad_norm": 0.0003548903623595834,
"learning_rate": 4.235944020202182e-07,
"loss": 0.0005627534701488912,
"step": 1979
},
{
"epoch": 3.2894873661224913,
"grad_norm": 0.000911902985535562,
"learning_rate": 4.216837752681019e-07,
"loss": 0.0016927756369113922,
"step": 1980
},
{
"epoch": 3.2911510866174485,
"grad_norm": 0.00043464102782309055,
"learning_rate": 4.19777070257594e-07,
"loss": 0.001056357054039836,
"step": 1981
},
{
"epoch": 3.2928148071124053,
"grad_norm": 0.0005667249206453562,
"learning_rate": 4.1787429058660845e-07,
"loss": 0.0005767525872215629,
"step": 1982
},
{
"epoch": 3.294478527607362,
"grad_norm": 0.0024056362453848124,
"learning_rate": 4.159754398456531e-07,
"loss": 0.0006031446973793209,
"step": 1983
},
{
"epoch": 3.296142248102319,
"grad_norm": 0.0004384003404993564,
"learning_rate": 4.14080521617822e-07,
"loss": 0.0008626229246146977,
"step": 1984
},
{
"epoch": 3.2978059685972756,
"grad_norm": 0.0007387603400275111,
"learning_rate": 4.121895394787881e-07,
"loss": 0.0014089690521359444,
"step": 1985
},
{
"epoch": 3.2994696890922324,
"grad_norm": 0.0007006418891251087,
"learning_rate": 4.103024969967981e-07,
"loss": 0.0016673547215759754,
"step": 1986
},
{
"epoch": 3.301133409587189,
"grad_norm": 0.0003609545237850398,
"learning_rate": 4.084193977326625e-07,
"loss": 0.0004953519091941416,
"step": 1987
},
{
"epoch": 3.3027971300821464,
"grad_norm": 0.0007110030273906887,
"learning_rate": 4.0654024523975323e-07,
"loss": 0.0015487212222069502,
"step": 1988
},
{
"epoch": 3.304460850577103,
"grad_norm": 0.0011667604558169842,
"learning_rate": 4.0466504306399366e-07,
"loss": 0.0027107703499495983,
"step": 1989
},
{
"epoch": 3.30612457107206,
"grad_norm": 0.00033490470377728343,
"learning_rate": 4.027937947438532e-07,
"loss": 0.0012670408468693495,
"step": 1990
},
{
"epoch": 3.3077882915670167,
"grad_norm": 0.0005540425190702081,
"learning_rate": 4.009265038103402e-07,
"loss": 0.0018498932477086782,
"step": 1991
},
{
"epoch": 3.3094520120619735,
"grad_norm": 0.0005201280582696199,
"learning_rate": 3.9906317378699684e-07,
"loss": 0.0012013057712465525,
"step": 1992
},
{
"epoch": 3.3111157325569303,
"grad_norm": 0.0007871100679039955,
"learning_rate": 3.972038081898885e-07,
"loss": 0.0019484572112560272,
"step": 1993
},
{
"epoch": 3.3127794530518875,
"grad_norm": 0.0008555342210456729,
"learning_rate": 3.9534841052760174e-07,
"loss": 0.0034242752008140087,
"step": 1994
},
{
"epoch": 3.3144431735468443,
"grad_norm": 0.0008200041484087706,
"learning_rate": 3.9349698430123566e-07,
"loss": 0.0005676429718732834,
"step": 1995
},
{
"epoch": 3.316106894041801,
"grad_norm": 0.0006836490356363356,
"learning_rate": 3.9164953300439456e-07,
"loss": 0.0011167812626808882,
"step": 1996
},
{
"epoch": 3.317770614536758,
"grad_norm": 0.0015542430337518454,
"learning_rate": 3.898060601231832e-07,
"loss": 0.0025796834379434586,
"step": 1997
},
{
"epoch": 3.3194343350317146,
"grad_norm": 0.0003486530913505703,
"learning_rate": 3.879665691361975e-07,
"loss": 0.00042451033368706703,
"step": 1998
},
{
"epoch": 3.3210980555266714,
"grad_norm": 0.000579525250941515,
"learning_rate": 3.861310635145207e-07,
"loss": 0.0010016821324825287,
"step": 1999
},
{
"epoch": 3.322761776021628,
"grad_norm": 0.0004743239260278642,
"learning_rate": 3.8429954672171613e-07,
"loss": 0.00039713463047519326,
"step": 2000
},
{
"epoch": 3.322761776021628,
"eval_loss": 0.00184919114690274,
"eval_runtime": 491.8051,
"eval_samples_per_second": 9.776,
"eval_steps_per_second": 2.444,
"step": 2000
},
{
"epoch": 3.3244254965165854,
"grad_norm": 0.0007103593670763075,
"learning_rate": 3.824720222138192e-07,
"loss": 0.0009300188394263387,
"step": 2001
},
{
"epoch": 3.326089217011542,
"grad_norm": 0.0003298052179161459,
"learning_rate": 3.806484934393331e-07,
"loss": 0.0002536668907850981,
"step": 2002
},
{
"epoch": 3.327752937506499,
"grad_norm": 0.00027529304497875273,
"learning_rate": 3.788289638392206e-07,
"loss": 0.000571264186874032,
"step": 2003
},
{
"epoch": 3.3294166580014557,
"grad_norm": 0.0007259314879775047,
"learning_rate": 3.7701343684689725e-07,
"loss": 0.0005468014860525727,
"step": 2004
},
{
"epoch": 3.3310803784964125,
"grad_norm": 0.0003917735884897411,
"learning_rate": 3.7520191588822695e-07,
"loss": 0.0009275685297325253,
"step": 2005
},
{
"epoch": 3.3327440989913697,
"grad_norm": 0.0007926534744910896,
"learning_rate": 3.7339440438151383e-07,
"loss": 0.0013022706843912601,
"step": 2006
},
{
"epoch": 3.3344078194863265,
"grad_norm": 0.0012599979527294636,
"learning_rate": 3.7159090573749693e-07,
"loss": 0.000500794849358499,
"step": 2007
},
{
"epoch": 3.3360715399812833,
"grad_norm": 0.00033590139355510473,
"learning_rate": 3.6979142335934246e-07,
"loss": 0.0007862839847803116,
"step": 2008
},
{
"epoch": 3.33773526047624,
"grad_norm": 0.0002938685938715935,
"learning_rate": 3.67995960642637e-07,
"loss": 0.00045850701280869544,
"step": 2009
},
{
"epoch": 3.339398980971197,
"grad_norm": 0.00036255698068998754,
"learning_rate": 3.6620452097538424e-07,
"loss": 0.001345583237707615,
"step": 2010
},
{
"epoch": 3.3410627014661536,
"grad_norm": 0.0007394360145553946,
"learning_rate": 3.644171077379949e-07,
"loss": 0.001691554207354784,
"step": 2011
},
{
"epoch": 3.3427264219611104,
"grad_norm": 0.00044334522681310773,
"learning_rate": 3.6263372430328266e-07,
"loss": 0.001117645064368844,
"step": 2012
},
{
"epoch": 3.344390142456067,
"grad_norm": 0.00046220130752772093,
"learning_rate": 3.6085437403645645e-07,
"loss": 0.0003471467352937907,
"step": 2013
},
{
"epoch": 3.3460538629510244,
"grad_norm": 0.00021781664690934122,
"learning_rate": 3.5907906029511606e-07,
"loss": 0.00023465295089408755,
"step": 2014
},
{
"epoch": 3.347717583445981,
"grad_norm": 0.00014978775288909674,
"learning_rate": 3.573077864292421e-07,
"loss": 0.00019691628403961658,
"step": 2015
},
{
"epoch": 3.349381303940938,
"grad_norm": 0.000697939598467201,
"learning_rate": 3.555405557811936e-07,
"loss": 0.0011822873493656516,
"step": 2016
},
{
"epoch": 3.3510450244358947,
"grad_norm": 0.0019957164768129587,
"learning_rate": 3.537773716857004e-07,
"loss": 0.0014215171104297042,
"step": 2017
},
{
"epoch": 3.3527087449308515,
"grad_norm": 0.00011045035353163257,
"learning_rate": 3.5201823746985554e-07,
"loss": 8.753546717343852e-05,
"step": 2018
},
{
"epoch": 3.3543724654258087,
"grad_norm": 0.00021515815751627088,
"learning_rate": 3.5026315645311114e-07,
"loss": 0.0002639327722135931,
"step": 2019
},
{
"epoch": 3.3560361859207655,
"grad_norm": 0.00012338242959231138,
"learning_rate": 3.485121319472695e-07,
"loss": 0.00012567051453515887,
"step": 2020
},
{
"epoch": 3.3576999064157222,
"grad_norm": 0.00023669407528359443,
"learning_rate": 3.4676516725647953e-07,
"loss": 0.0004816576838493347,
"step": 2021
},
{
"epoch": 3.359363626910679,
"grad_norm": 0.0004548949364107102,
"learning_rate": 3.450222656772292e-07,
"loss": 0.0005232831463217735,
"step": 2022
},
{
"epoch": 3.361027347405636,
"grad_norm": 0.0003678028006106615,
"learning_rate": 3.43283430498339e-07,
"loss": 0.0003801046696025878,
"step": 2023
},
{
"epoch": 3.3626910679005926,
"grad_norm": 0.00035056608612649143,
"learning_rate": 3.4154866500095695e-07,
"loss": 0.0007555793272331357,
"step": 2024
},
{
"epoch": 3.3643547883955494,
"grad_norm": 0.0005938123795203865,
"learning_rate": 3.3981797245855096e-07,
"loss": 0.001063125324435532,
"step": 2025
},
{
"epoch": 3.3660185088905066,
"grad_norm": 0.0010471873683854938,
"learning_rate": 3.380913561369037e-07,
"loss": 0.0012568510137498379,
"step": 2026
},
{
"epoch": 3.3676822293854634,
"grad_norm": 0.000528757693246007,
"learning_rate": 3.363688192941067e-07,
"loss": 0.0007248009787872434,
"step": 2027
},
{
"epoch": 3.36934594988042,
"grad_norm": 0.0006465297774411738,
"learning_rate": 3.346503651805513e-07,
"loss": 0.0006219932110980153,
"step": 2028
},
{
"epoch": 3.371009670375377,
"grad_norm": 0.001121693872846663,
"learning_rate": 3.329359970389279e-07,
"loss": 0.0017972507048398256,
"step": 2029
},
{
"epoch": 3.3726733908703337,
"grad_norm": 0.0007588271400891244,
"learning_rate": 3.312257181042142e-07,
"loss": 0.0011564313899725676,
"step": 2030
},
{
"epoch": 3.3743371113652905,
"grad_norm": 0.00013510511780623347,
"learning_rate": 3.2951953160367365e-07,
"loss": 0.00013225802103988826,
"step": 2031
},
{
"epoch": 3.3760008318602477,
"grad_norm": 0.000917671131901443,
"learning_rate": 3.2781744075684576e-07,
"loss": 0.0008554297965019941,
"step": 2032
},
{
"epoch": 3.3776645523552045,
"grad_norm": 0.0003718238731380552,
"learning_rate": 3.261194487755426e-07,
"loss": 0.00021981318423058838,
"step": 2033
},
{
"epoch": 3.3793282728501612,
"grad_norm": 0.00033087420160882175,
"learning_rate": 3.2442555886384145e-07,
"loss": 0.0009502323227934539,
"step": 2034
},
{
"epoch": 3.380991993345118,
"grad_norm": 0.0007806571666151285,
"learning_rate": 3.2273577421807976e-07,
"loss": 0.0017600739374756813,
"step": 2035
},
{
"epoch": 3.382655713840075,
"grad_norm": 0.0003112622071057558,
"learning_rate": 3.2105009802684636e-07,
"loss": 0.0007383651682175696,
"step": 2036
},
{
"epoch": 3.3843194343350316,
"grad_norm": 0.0006200448842719197,
"learning_rate": 3.1936853347097923e-07,
"loss": 0.0013696863315999508,
"step": 2037
},
{
"epoch": 3.3859831548299884,
"grad_norm": 0.0008885011775419116,
"learning_rate": 3.1769108372355804e-07,
"loss": 0.001769607188180089,
"step": 2038
},
{
"epoch": 3.3876468753249456,
"grad_norm": 0.000550445111002773,
"learning_rate": 3.1601775194989693e-07,
"loss": 0.0013523256639018655,
"step": 2039
},
{
"epoch": 3.3893105958199023,
"grad_norm": 0.00022062704374548048,
"learning_rate": 3.143485413075398e-07,
"loss": 0.0003513278206810355,
"step": 2040
},
{
"epoch": 3.390974316314859,
"grad_norm": 0.0005184001056477427,
"learning_rate": 3.1268345494625486e-07,
"loss": 0.0006000545108690858,
"step": 2041
},
{
"epoch": 3.392638036809816,
"grad_norm": 0.000451189698651433,
"learning_rate": 3.1102249600802573e-07,
"loss": 0.0012537793954834342,
"step": 2042
},
{
"epoch": 3.3943017573047727,
"grad_norm": 0.0005222734180279076,
"learning_rate": 3.093656676270501e-07,
"loss": 0.0004908693372271955,
"step": 2043
},
{
"epoch": 3.3959654777997295,
"grad_norm": 0.0006829127087257802,
"learning_rate": 3.0771297292972986e-07,
"loss": 0.0016357006970793009,
"step": 2044
},
{
"epoch": 3.3976291982946867,
"grad_norm": 0.00043566894601099193,
"learning_rate": 3.0606441503466753e-07,
"loss": 0.0005723870126530528,
"step": 2045
},
{
"epoch": 3.3992929187896435,
"grad_norm": 0.00031372217927128077,
"learning_rate": 3.044199970526593e-07,
"loss": 0.0005421612877398729,
"step": 2046
},
{
"epoch": 3.4009566392846002,
"grad_norm": 0.0005682411137968302,
"learning_rate": 3.027797220866896e-07,
"loss": 0.0015814367216080427,
"step": 2047
},
{
"epoch": 3.402620359779557,
"grad_norm": 0.00031971544376574457,
"learning_rate": 3.01143593231924e-07,
"loss": 0.0011364194797351956,
"step": 2048
},
{
"epoch": 3.404284080274514,
"grad_norm": 0.0009130490943789482,
"learning_rate": 2.995116135757059e-07,
"loss": 0.00183593831025064,
"step": 2049
},
{
"epoch": 3.4059478007694706,
"grad_norm": 0.0004094933101441711,
"learning_rate": 2.978837861975484e-07,
"loss": 0.0007758493884466588,
"step": 2050
},
{
"epoch": 3.4076115212644273,
"grad_norm": 0.00035802522324956954,
"learning_rate": 2.962601141691296e-07,
"loss": 0.0007563665276393294,
"step": 2051
},
{
"epoch": 3.4092752417593846,
"grad_norm": 0.00038640270940959454,
"learning_rate": 2.9464060055428703e-07,
"loss": 0.0003179998602718115,
"step": 2052
},
{
"epoch": 3.4109389622543413,
"grad_norm": 0.0005746350507251918,
"learning_rate": 2.930252484090101e-07,
"loss": 0.0010296654654666781,
"step": 2053
},
{
"epoch": 3.412602682749298,
"grad_norm": 0.00018986756913363934,
"learning_rate": 2.9141406078143644e-07,
"loss": 0.0005916217342019081,
"step": 2054
},
{
"epoch": 3.414266403244255,
"grad_norm": 0.00030949199572205544,
"learning_rate": 2.8980704071184557e-07,
"loss": 0.0005941985873505473,
"step": 2055
},
{
"epoch": 3.4159301237392117,
"grad_norm": 0.000416825816500932,
"learning_rate": 2.882041912326525e-07,
"loss": 0.0003663770330604166,
"step": 2056
},
{
"epoch": 3.417593844234169,
"grad_norm": 0.0006616072496399283,
"learning_rate": 2.8660551536840277e-07,
"loss": 0.0013288824120536447,
"step": 2057
},
{
"epoch": 3.4192575647291257,
"grad_norm": 0.0004903053632006049,
"learning_rate": 2.8501101613576526e-07,
"loss": 0.0007218238897621632,
"step": 2058
},
{
"epoch": 3.4209212852240825,
"grad_norm": 0.0002515607629902661,
"learning_rate": 2.834206965435293e-07,
"loss": 0.00033488430199213326,
"step": 2059
},
{
"epoch": 3.4225850057190392,
"grad_norm": 0.0004987838328815997,
"learning_rate": 2.818345595925959e-07,
"loss": 0.0005008272128179669,
"step": 2060
},
{
"epoch": 3.424248726213996,
"grad_norm": 0.0004810976970475167,
"learning_rate": 2.8025260827597463e-07,
"loss": 0.0014125597663223743,
"step": 2061
},
{
"epoch": 3.425912446708953,
"grad_norm": 0.0006654640892520547,
"learning_rate": 2.7867484557877607e-07,
"loss": 0.0020283989142626524,
"step": 2062
},
{
"epoch": 3.4275761672039096,
"grad_norm": 0.0005379806971177459,
"learning_rate": 2.7710127447820783e-07,
"loss": 0.0003175054444000125,
"step": 2063
},
{
"epoch": 3.429239887698867,
"grad_norm": 0.0003664498508442193,
"learning_rate": 2.7553189794356615e-07,
"loss": 0.0009711123420856893,
"step": 2064
},
{
"epoch": 3.4309036081938236,
"grad_norm": 0.0006012935773469508,
"learning_rate": 2.739667189362347e-07,
"loss": 0.0008616911945864558,
"step": 2065
},
{
"epoch": 3.4325673286887803,
"grad_norm": 0.0003376624081283808,
"learning_rate": 2.724057404096744e-07,
"loss": 0.0007702240254729986,
"step": 2066
},
{
"epoch": 3.434231049183737,
"grad_norm": 0.0005580992437899113,
"learning_rate": 2.708489653094218e-07,
"loss": 0.00029471275047399104,
"step": 2067
},
{
"epoch": 3.435894769678694,
"grad_norm": 0.0006376696401275694,
"learning_rate": 2.692963965730805e-07,
"loss": 0.0015568721573799849,
"step": 2068
},
{
"epoch": 3.4375584901736507,
"grad_norm": 0.00039179300074465573,
"learning_rate": 2.677480371303162e-07,
"loss": 0.0004968352150171995,
"step": 2069
},
{
"epoch": 3.439222210668608,
"grad_norm": 0.0004967460408806801,
"learning_rate": 2.662038899028532e-07,
"loss": 0.0016945095267146826,
"step": 2070
},
{
"epoch": 3.4408859311635647,
"grad_norm": 0.0004912624717690051,
"learning_rate": 2.6466395780446657e-07,
"loss": 0.00030252139549702406,
"step": 2071
},
{
"epoch": 3.4425496516585214,
"grad_norm": 0.0005200419691391289,
"learning_rate": 2.6312824374097794e-07,
"loss": 0.0014322304632514715,
"step": 2072
},
{
"epoch": 3.4442133721534782,
"grad_norm": 0.0004998520016670227,
"learning_rate": 2.6159675061024905e-07,
"loss": 0.0020267111249268055,
"step": 2073
},
{
"epoch": 3.445877092648435,
"grad_norm": 0.00022210314637050033,
"learning_rate": 2.6006948130217815e-07,
"loss": 0.00020690905512310565,
"step": 2074
},
{
"epoch": 3.447540813143392,
"grad_norm": 0.0008716708398424089,
"learning_rate": 2.585464386986908e-07,
"loss": 0.001405320130288601,
"step": 2075
},
{
"epoch": 3.4492045336383486,
"grad_norm": 0.0002139442804036662,
"learning_rate": 2.570276256737386e-07,
"loss": 0.0003679142100736499,
"step": 2076
},
{
"epoch": 3.450868254133306,
"grad_norm": 0.0001617825764697045,
"learning_rate": 2.555130450932922e-07,
"loss": 0.00012385935406200588,
"step": 2077
},
{
"epoch": 3.4525319746282626,
"grad_norm": 0.000477365858387202,
"learning_rate": 2.54002699815335e-07,
"loss": 0.0016203195555135608,
"step": 2078
},
{
"epoch": 3.4541956951232193,
"grad_norm": 0.0006356213707476854,
"learning_rate": 2.52496592689859e-07,
"loss": 0.0019338249694555998,
"step": 2079
},
{
"epoch": 3.455859415618176,
"grad_norm": 0.0006331139011308551,
"learning_rate": 2.5099472655885777e-07,
"loss": 0.0004027165996376425,
"step": 2080
},
{
"epoch": 3.457523136113133,
"grad_norm": 0.00026520664687268436,
"learning_rate": 2.4949710425632353e-07,
"loss": 0.0004268680640961975,
"step": 2081
},
{
"epoch": 3.4591868566080897,
"grad_norm": 0.0012745996937155724,
"learning_rate": 2.4800372860823956e-07,
"loss": 0.00493782851845026,
"step": 2082
},
{
"epoch": 3.460850577103047,
"grad_norm": 0.001241783844307065,
"learning_rate": 2.465146024325765e-07,
"loss": 0.001954779028892517,
"step": 2083
},
{
"epoch": 3.4625142975980037,
"grad_norm": 0.0007233137730509043,
"learning_rate": 2.4502972853928606e-07,
"loss": 0.003158936742693186,
"step": 2084
},
{
"epoch": 3.4641780180929604,
"grad_norm": 0.0004373241390567273,
"learning_rate": 2.435491097302961e-07,
"loss": 0.0005493911448866129,
"step": 2085
},
{
"epoch": 3.465841738587917,
"grad_norm": 0.00048215530114248395,
"learning_rate": 2.420727487995045e-07,
"loss": 0.0003778711543418467,
"step": 2086
},
{
"epoch": 3.467505459082874,
"grad_norm": 0.0008155179093591869,
"learning_rate": 2.40600648532775e-07,
"loss": 0.0015070312656462193,
"step": 2087
},
{
"epoch": 3.4691691795778308,
"grad_norm": 0.0006131388363428414,
"learning_rate": 2.3913281170793196e-07,
"loss": 0.0013324524043127894,
"step": 2088
},
{
"epoch": 3.4708329000727876,
"grad_norm": 0.0007873099530115724,
"learning_rate": 2.376692410947548e-07,
"loss": 0.0010801044991239905,
"step": 2089
},
{
"epoch": 3.4724966205677448,
"grad_norm": 0.0003397899563424289,
"learning_rate": 2.3620993945497217e-07,
"loss": 0.0005977149703539908,
"step": 2090
},
{
"epoch": 3.4741603410627016,
"grad_norm": 0.000378866825485602,
"learning_rate": 2.347549095422569e-07,
"loss": 0.0011136323446407914,
"step": 2091
},
{
"epoch": 3.4758240615576583,
"grad_norm": 0.0004786884237546474,
"learning_rate": 2.3330415410222212e-07,
"loss": 0.0007871107663959265,
"step": 2092
},
{
"epoch": 3.477487782052615,
"grad_norm": 0.0007962996023707092,
"learning_rate": 2.3185767587241447e-07,
"loss": 0.001519782468676567,
"step": 2093
},
{
"epoch": 3.479151502547572,
"grad_norm": 0.00041777914157137275,
"learning_rate": 2.3041547758230977e-07,
"loss": 0.001065521501004696,
"step": 2094
},
{
"epoch": 3.4808152230425287,
"grad_norm": 0.0003879894211422652,
"learning_rate": 2.2897756195330773e-07,
"loss": 0.001152183162048459,
"step": 2095
},
{
"epoch": 3.482478943537486,
"grad_norm": 0.0003022203454747796,
"learning_rate": 2.2754393169872685e-07,
"loss": 0.0007047376129776239,
"step": 2096
},
{
"epoch": 3.4841426640324427,
"grad_norm": 0.0007938887574709952,
"learning_rate": 2.2611458952379872e-07,
"loss": 0.001837533782236278,
"step": 2097
},
{
"epoch": 3.4858063845273994,
"grad_norm": 0.0003543386119417846,
"learning_rate": 2.246895381256639e-07,
"loss": 0.0008702923660166562,
"step": 2098
},
{
"epoch": 3.487470105022356,
"grad_norm": 0.0016991429729387164,
"learning_rate": 2.232687801933664e-07,
"loss": 0.002679019933566451,
"step": 2099
},
{
"epoch": 3.489133825517313,
"grad_norm": 0.0003651150909718126,
"learning_rate": 2.2185231840784778e-07,
"loss": 0.0006635597092099488,
"step": 2100
},
{
"epoch": 3.489133825517313,
"eval_loss": 0.001800627913326025,
"eval_runtime": 491.0251,
"eval_samples_per_second": 9.792,
"eval_steps_per_second": 2.448,
"step": 2100
},
{
"epoch": 3.4907975460122698,
"grad_norm": 0.0009423141600564122,
"learning_rate": 2.204401554419444e-07,
"loss": 0.0022930323611944914,
"step": 2101
},
{
"epoch": 3.4924612665072265,
"grad_norm": 0.00046992572606541216,
"learning_rate": 2.1903229396037896e-07,
"loss": 0.0006827316246926785,
"step": 2102
},
{
"epoch": 3.4941249870021838,
"grad_norm": 0.00031514756847172976,
"learning_rate": 2.1762873661975825e-07,
"loss": 0.001143246074207127,
"step": 2103
},
{
"epoch": 3.4957887074971405,
"grad_norm": 0.0005974940140731633,
"learning_rate": 2.1622948606856765e-07,
"loss": 0.0005027040606364608,
"step": 2104
},
{
"epoch": 3.4974524279920973,
"grad_norm": 0.0007181466207839549,
"learning_rate": 2.1483454494716504e-07,
"loss": 0.0008611283265054226,
"step": 2105
},
{
"epoch": 3.499116148487054,
"grad_norm": 0.0006843828014098108,
"learning_rate": 2.1344391588777658e-07,
"loss": 0.0006472602253779769,
"step": 2106
},
{
"epoch": 3.500779868982011,
"grad_norm": 0.0006242678500711918,
"learning_rate": 2.1205760151449206e-07,
"loss": 0.0026610223576426506,
"step": 2107
},
{
"epoch": 3.502443589476968,
"grad_norm": 0.0006409132038243115,
"learning_rate": 2.106756044432598e-07,
"loss": 0.0014607177581638098,
"step": 2108
},
{
"epoch": 3.504107309971925,
"grad_norm": 0.00031891578692011535,
"learning_rate": 2.0929792728187986e-07,
"loss": 0.0004305794136598706,
"step": 2109
},
{
"epoch": 3.5057710304668817,
"grad_norm": 0.00021313800243660808,
"learning_rate": 2.079245726300022e-07,
"loss": 0.0005588196218013763,
"step": 2110
},
{
"epoch": 3.5074347509618384,
"grad_norm": 0.0005837749922648072,
"learning_rate": 2.0655554307911997e-07,
"loss": 0.0011885692365467548,
"step": 2111
},
{
"epoch": 3.509098471456795,
"grad_norm": 0.00048057257663458586,
"learning_rate": 2.05190841212565e-07,
"loss": 0.0007919726194813848,
"step": 2112
},
{
"epoch": 3.510762191951752,
"grad_norm": 0.0004323796893004328,
"learning_rate": 2.038304696055024e-07,
"loss": 0.0009170786943286657,
"step": 2113
},
{
"epoch": 3.5124259124467088,
"grad_norm": 0.00035488922731019557,
"learning_rate": 2.0247443082492686e-07,
"loss": 0.0006726681021973491,
"step": 2114
},
{
"epoch": 3.5140896329416655,
"grad_norm": 0.0003869618522003293,
"learning_rate": 2.0112272742965678e-07,
"loss": 0.0006815298111177981,
"step": 2115
},
{
"epoch": 3.5157533534366228,
"grad_norm": 0.00034062107442878187,
"learning_rate": 1.997753619703291e-07,
"loss": 0.0008145941537804902,
"step": 2116
},
{
"epoch": 3.5174170739315795,
"grad_norm": 0.0014651084784418344,
"learning_rate": 1.9843233698939617e-07,
"loss": 0.000923621584661305,
"step": 2117
},
{
"epoch": 3.5190807944265363,
"grad_norm": 0.0005609922809526324,
"learning_rate": 1.9709365502111944e-07,
"loss": 0.0010176009964197874,
"step": 2118
},
{
"epoch": 3.520744514921493,
"grad_norm": 0.0006454312242567539,
"learning_rate": 1.957593185915657e-07,
"loss": 0.0017813025042414665,
"step": 2119
},
{
"epoch": 3.52240823541645,
"grad_norm": 0.00036230773548595607,
"learning_rate": 1.9442933021860095e-07,
"loss": 0.0005436648498289287,
"step": 2120
},
{
"epoch": 3.524071955911407,
"grad_norm": 0.0003991083358414471,
"learning_rate": 1.9310369241188732e-07,
"loss": 0.0004445482627488673,
"step": 2121
},
{
"epoch": 3.525735676406364,
"grad_norm": 0.000983903999440372,
"learning_rate": 1.9178240767287666e-07,
"loss": 0.0016248344909399748,
"step": 2122
},
{
"epoch": 3.5273993969013206,
"grad_norm": 0.00019582314416766167,
"learning_rate": 1.904654784948079e-07,
"loss": 0.0008514729561284184,
"step": 2123
},
{
"epoch": 3.5290631173962774,
"grad_norm": 0.0004321737796999514,
"learning_rate": 1.8915290736269965e-07,
"loss": 0.0014025006676092744,
"step": 2124
},
{
"epoch": 3.530726837891234,
"grad_norm": 0.00023967149900272489,
"learning_rate": 1.878446967533476e-07,
"loss": 0.0005360299255698919,
"step": 2125
},
{
"epoch": 3.532390558386191,
"grad_norm": 0.00028346243198029697,
"learning_rate": 1.865408491353199e-07,
"loss": 0.0008053178898990154,
"step": 2126
},
{
"epoch": 3.5340542788811478,
"grad_norm": 0.0009381191339343786,
"learning_rate": 1.8524136696895068e-07,
"loss": 0.00040692309266887605,
"step": 2127
},
{
"epoch": 3.535717999376105,
"grad_norm": 0.00023939047241583467,
"learning_rate": 1.8394625270633793e-07,
"loss": 0.0006541266338899732,
"step": 2128
},
{
"epoch": 3.5373817198710618,
"grad_norm": 0.000494911044370383,
"learning_rate": 1.8265550879133538e-07,
"loss": 0.0005820249207317829,
"step": 2129
},
{
"epoch": 3.5390454403660185,
"grad_norm": 0.00031910924008116126,
"learning_rate": 1.8136913765955195e-07,
"loss": 0.00022852102119941264,
"step": 2130
},
{
"epoch": 3.5407091608609753,
"grad_norm": 0.0003672874590847641,
"learning_rate": 1.8008714173834456e-07,
"loss": 0.0016192167531698942,
"step": 2131
},
{
"epoch": 3.542372881355932,
"grad_norm": 0.0003836368559859693,
"learning_rate": 1.7880952344681402e-07,
"loss": 0.0009387480677105486,
"step": 2132
},
{
"epoch": 3.5440366018508893,
"grad_norm": 0.0003523621999192983,
"learning_rate": 1.7753628519580097e-07,
"loss": 0.0005441518733277917,
"step": 2133
},
{
"epoch": 3.545700322345846,
"grad_norm": 0.0005042285774834454,
"learning_rate": 1.7626742938788105e-07,
"loss": 0.000680170429404825,
"step": 2134
},
{
"epoch": 3.547364042840803,
"grad_norm": 0.0006183210061863065,
"learning_rate": 1.7500295841735905e-07,
"loss": 0.0006854024832136929,
"step": 2135
},
{
"epoch": 3.5490277633357596,
"grad_norm": 0.0004747119382955134,
"learning_rate": 1.7374287467026767e-07,
"loss": 0.0008972194045782089,
"step": 2136
},
{
"epoch": 3.5506914838307164,
"grad_norm": 0.00047969879233278334,
"learning_rate": 1.7248718052435942e-07,
"loss": 0.000863411754835397,
"step": 2137
},
{
"epoch": 3.552355204325673,
"grad_norm": 0.00033154431730508804,
"learning_rate": 1.712358783491047e-07,
"loss": 0.00024044688325375319,
"step": 2138
},
{
"epoch": 3.55401892482063,
"grad_norm": 0.00040625245310366154,
"learning_rate": 1.6998897050568618e-07,
"loss": 0.00100037083029747,
"step": 2139
},
{
"epoch": 3.5556826453155868,
"grad_norm": 0.002271288773044944,
"learning_rate": 1.6874645934699342e-07,
"loss": 0.0009121758048422635,
"step": 2140
},
{
"epoch": 3.557346365810544,
"grad_norm": 0.0003006242332048714,
"learning_rate": 1.6750834721762117e-07,
"loss": 0.00045710286940447986,
"step": 2141
},
{
"epoch": 3.5590100863055008,
"grad_norm": 0.00031541401403956115,
"learning_rate": 1.6627463645386199e-07,
"loss": 0.0009236993500962853,
"step": 2142
},
{
"epoch": 3.5606738068004575,
"grad_norm": 0.0012363658752292395,
"learning_rate": 1.6504532938370427e-07,
"loss": 0.0014181968290358782,
"step": 2143
},
{
"epoch": 3.5623375272954143,
"grad_norm": 0.0006429841741919518,
"learning_rate": 1.6382042832682577e-07,
"loss": 0.002108179498463869,
"step": 2144
},
{
"epoch": 3.564001247790371,
"grad_norm": 0.0008274697465822101,
"learning_rate": 1.6259993559459091e-07,
"loss": 0.0012376433005556464,
"step": 2145
},
{
"epoch": 3.5656649682853283,
"grad_norm": 0.00020373162988107651,
"learning_rate": 1.613838534900447e-07,
"loss": 0.0004814681888092309,
"step": 2146
},
{
"epoch": 3.567328688780285,
"grad_norm": 0.0003222871746402234,
"learning_rate": 1.601721843079107e-07,
"loss": 0.00038049567956477404,
"step": 2147
},
{
"epoch": 3.568992409275242,
"grad_norm": 0.00017242447938770056,
"learning_rate": 1.5896493033458416e-07,
"loss": 0.00020702739129774272,
"step": 2148
},
{
"epoch": 3.5706561297701986,
"grad_norm": 0.0004510879225563258,
"learning_rate": 1.5776209384812946e-07,
"loss": 0.0011956410016864538,
"step": 2149
},
{
"epoch": 3.5723198502651554,
"grad_norm": 0.0005464230780489743,
"learning_rate": 1.5656367711827602e-07,
"loss": 0.0018636614549905062,
"step": 2150
},
{
"epoch": 3.573983570760112,
"grad_norm": 0.0003373275394551456,
"learning_rate": 1.553696824064116e-07,
"loss": 0.0006840836722403765,
"step": 2151
},
{
"epoch": 3.575647291255069,
"grad_norm": 0.000650634232442826,
"learning_rate": 1.5418011196558085e-07,
"loss": 0.0015320621896535158,
"step": 2152
},
{
"epoch": 3.5773110117500257,
"grad_norm": 0.0008834419422782958,
"learning_rate": 1.529949680404799e-07,
"loss": 0.0005219897720962763,
"step": 2153
},
{
"epoch": 3.578974732244983,
"grad_norm": 0.00017787229444365948,
"learning_rate": 1.5181425286745155e-07,
"loss": 0.0001635367952985689,
"step": 2154
},
{
"epoch": 3.5806384527399397,
"grad_norm": 0.0002696933224797249,
"learning_rate": 1.5063796867448243e-07,
"loss": 0.0004062880761921406,
"step": 2155
},
{
"epoch": 3.5823021732348965,
"grad_norm": 0.000314901873935014,
"learning_rate": 1.4946611768119763e-07,
"loss": 0.0002677102165762335,
"step": 2156
},
{
"epoch": 3.5839658937298533,
"grad_norm": 0.0017086658626794815,
"learning_rate": 1.4829870209885605e-07,
"loss": 0.003296004142612219,
"step": 2157
},
{
"epoch": 3.58562961422481,
"grad_norm": 0.0006046981434337795,
"learning_rate": 1.471357241303481e-07,
"loss": 0.0017653173999860883,
"step": 2158
},
{
"epoch": 3.5872933347197673,
"grad_norm": 0.0014973668148741126,
"learning_rate": 1.4597718597019055e-07,
"loss": 0.0022984712850302458,
"step": 2159
},
{
"epoch": 3.588957055214724,
"grad_norm": 0.000747077923733741,
"learning_rate": 1.4482308980452164e-07,
"loss": 0.0016120221698656678,
"step": 2160
},
{
"epoch": 3.590620775709681,
"grad_norm": 0.0008344343514181674,
"learning_rate": 1.436734378110985e-07,
"loss": 0.001747865928336978,
"step": 2161
},
{
"epoch": 3.5922844962046376,
"grad_norm": 0.0002902126288972795,
"learning_rate": 1.425282321592908e-07,
"loss": 0.0007139051449485123,
"step": 2162
},
{
"epoch": 3.5939482166995944,
"grad_norm": 0.00047769720549695194,
"learning_rate": 1.4138747501007966e-07,
"loss": 0.00109275639988482,
"step": 2163
},
{
"epoch": 3.595611937194551,
"grad_norm": 0.0005414061597548425,
"learning_rate": 1.4025116851605125e-07,
"loss": 0.001188867725431919,
"step": 2164
},
{
"epoch": 3.597275657689508,
"grad_norm": 0.0002681358018890023,
"learning_rate": 1.3911931482139317e-07,
"loss": 0.0007024380029179156,
"step": 2165
},
{
"epoch": 3.5989393781844647,
"grad_norm": 0.00043560643098317087,
"learning_rate": 1.379919160618909e-07,
"loss": 0.0007490735151804984,
"step": 2166
},
{
"epoch": 3.600603098679422,
"grad_norm": 0.0005028645973652601,
"learning_rate": 1.368689743649243e-07,
"loss": 0.0008132288348861039,
"step": 2167
},
{
"epoch": 3.6022668191743787,
"grad_norm": 0.0006175213493406773,
"learning_rate": 1.3575049184946122e-07,
"loss": 0.0014601877192035317,
"step": 2168
},
{
"epoch": 3.6039305396693355,
"grad_norm": 0.0002806995762512088,
"learning_rate": 1.346364706260564e-07,
"loss": 0.0002837897918652743,
"step": 2169
},
{
"epoch": 3.6055942601642923,
"grad_norm": 0.0003279212396591902,
"learning_rate": 1.3352691279684582e-07,
"loss": 0.0011922994162887335,
"step": 2170
},
{
"epoch": 3.6072579806592495,
"grad_norm": 0.00029519712552428246,
"learning_rate": 1.324218204555433e-07,
"loss": 0.0010329822544008493,
"step": 2171
},
{
"epoch": 3.6089217011542063,
"grad_norm": 0.0012609957484528422,
"learning_rate": 1.3132119568743662e-07,
"loss": 0.0010717506520450115,
"step": 2172
},
{
"epoch": 3.610585421649163,
"grad_norm": 0.000270798773271963,
"learning_rate": 1.3022504056938196e-07,
"loss": 0.0009245692635886371,
"step": 2173
},
{
"epoch": 3.61224914214412,
"grad_norm": 0.0003105451469309628,
"learning_rate": 1.2913335716980307e-07,
"loss": 0.000851097924169153,
"step": 2174
},
{
"epoch": 3.6139128626390766,
"grad_norm": 0.0004231147759128362,
"learning_rate": 1.2804614754868466e-07,
"loss": 0.001083697541616857,
"step": 2175
},
{
"epoch": 3.6155765831340334,
"grad_norm": 0.0006465599872171879,
"learning_rate": 1.2696341375756982e-07,
"loss": 0.0018808020977303386,
"step": 2176
},
{
"epoch": 3.61724030362899,
"grad_norm": 0.00040738159441389143,
"learning_rate": 1.2588515783955564e-07,
"loss": 0.0010184545535594225,
"step": 2177
},
{
"epoch": 3.618904024123947,
"grad_norm": 0.0005552941584028304,
"learning_rate": 1.2481138182929065e-07,
"loss": 0.0013670467305928469,
"step": 2178
},
{
"epoch": 3.620567744618904,
"grad_norm": 0.0006208255654200912,
"learning_rate": 1.2374208775296742e-07,
"loss": 0.0009191983845084906,
"step": 2179
},
{
"epoch": 3.622231465113861,
"grad_norm": 0.0006596720195375383,
"learning_rate": 1.2267727762832388e-07,
"loss": 0.0017082467675209045,
"step": 2180
},
{
"epoch": 3.6238951856088177,
"grad_norm": 0.0013268449110910296,
"learning_rate": 1.2161695346463498e-07,
"loss": 0.002551913959905505,
"step": 2181
},
{
"epoch": 3.6255589061037745,
"grad_norm": 0.0002721058845054358,
"learning_rate": 1.2056111726271192e-07,
"loss": 0.0005901381373405457,
"step": 2182
},
{
"epoch": 3.6272226265987313,
"grad_norm": 0.0006825258606113493,
"learning_rate": 1.195097710148968e-07,
"loss": 0.0011782599613070488,
"step": 2183
},
{
"epoch": 3.6288863470936885,
"grad_norm": 0.00038542994298040867,
"learning_rate": 1.1846291670505855e-07,
"loss": 0.0003281470271758735,
"step": 2184
},
{
"epoch": 3.6305500675886453,
"grad_norm": 0.0006535347201861441,
"learning_rate": 1.1742055630859117e-07,
"loss": 0.0009970370447263122,
"step": 2185
},
{
"epoch": 3.632213788083602,
"grad_norm": 0.0002451266336720437,
"learning_rate": 1.1638269179240796e-07,
"loss": 0.0005679831374436617,
"step": 2186
},
{
"epoch": 3.633877508578559,
"grad_norm": 0.00044009933480992913,
"learning_rate": 1.1534932511493846e-07,
"loss": 0.0011061355471611023,
"step": 2187
},
{
"epoch": 3.6355412290735156,
"grad_norm": 0.00046958556049503386,
"learning_rate": 1.1432045822612564e-07,
"loss": 0.00047064467798918486,
"step": 2188
},
{
"epoch": 3.6372049495684724,
"grad_norm": 0.000384511862648651,
"learning_rate": 1.132960930674204e-07,
"loss": 0.001256321556866169,
"step": 2189
},
{
"epoch": 3.638868670063429,
"grad_norm": 0.0004960609949193895,
"learning_rate": 1.1227623157177986e-07,
"loss": 0.001438457635231316,
"step": 2190
},
{
"epoch": 3.640532390558386,
"grad_norm": 0.0005817452911287546,
"learning_rate": 1.1126087566366266e-07,
"loss": 0.0007497647311538458,
"step": 2191
},
{
"epoch": 3.642196111053343,
"grad_norm": 0.0005362757365219295,
"learning_rate": 1.1025002725902484e-07,
"loss": 0.0009658780181780457,
"step": 2192
},
{
"epoch": 3.6438598315483,
"grad_norm": 0.0004179330717306584,
"learning_rate": 1.0924368826531751e-07,
"loss": 0.00042840588139370084,
"step": 2193
},
{
"epoch": 3.6455235520432567,
"grad_norm": 0.00041192653588950634,
"learning_rate": 1.0824186058148278e-07,
"loss": 0.00033956009428948164,
"step": 2194
},
{
"epoch": 3.6471872725382135,
"grad_norm": 0.0011990912025794387,
"learning_rate": 1.0724454609794931e-07,
"loss": 0.0005998517153784633,
"step": 2195
},
{
"epoch": 3.6488509930331703,
"grad_norm": 0.000372793001588434,
"learning_rate": 1.0625174669663036e-07,
"loss": 0.000853288802318275,
"step": 2196
},
{
"epoch": 3.6505147135281275,
"grad_norm": 0.0004351623938418925,
"learning_rate": 1.0526346425091815e-07,
"loss": 0.0008543959120288491,
"step": 2197
},
{
"epoch": 3.6521784340230843,
"grad_norm": 0.0024484992027282715,
"learning_rate": 1.042797006256821e-07,
"loss": 0.0028258298989385366,
"step": 2198
},
{
"epoch": 3.653842154518041,
"grad_norm": 0.00040658764191903174,
"learning_rate": 1.0330045767726504e-07,
"loss": 0.0016119135543704033,
"step": 2199
},
{
"epoch": 3.655505875012998,
"grad_norm": 0.0003821366699412465,
"learning_rate": 1.023257372534786e-07,
"loss": 0.0004272300866432488,
"step": 2200
},
{
"epoch": 3.655505875012998,
"eval_loss": 0.0017772268038243055,
"eval_runtime": 491.272,
"eval_samples_per_second": 9.787,
"eval_steps_per_second": 2.447,
"step": 2200
},
{
"epoch": 3.6571695955079546,
"grad_norm": 0.0009488234645687044,
"learning_rate": 1.0135554119360153e-07,
"loss": 0.0017828824929893017,
"step": 2201
},
{
"epoch": 3.6588333160029114,
"grad_norm": 0.00019926000095438212,
"learning_rate": 1.0038987132837435e-07,
"loss": 0.000897647812962532,
"step": 2202
},
{
"epoch": 3.660497036497868,
"grad_norm": 0.0006051113014109433,
"learning_rate": 9.942872947999672e-08,
"loss": 0.0019816546700894833,
"step": 2203
},
{
"epoch": 3.662160756992825,
"grad_norm": 0.00035018715425394475,
"learning_rate": 9.847211746212504e-08,
"loss": 0.0008289602119475603,
"step": 2204
},
{
"epoch": 3.663824477487782,
"grad_norm": 0.0005620543961413205,
"learning_rate": 9.752003707986652e-08,
"loss": 0.0011024302802979946,
"step": 2205
},
{
"epoch": 3.665488197982739,
"grad_norm": 0.00018053792882710695,
"learning_rate": 9.657249012977821e-08,
"loss": 0.0006109848036430776,
"step": 2206
},
{
"epoch": 3.6671519184776957,
"grad_norm": 0.000838694570120424,
"learning_rate": 9.562947839986264e-08,
"loss": 0.0028103888034820557,
"step": 2207
},
{
"epoch": 3.6688156389726525,
"grad_norm": 0.0004535977204795927,
"learning_rate": 9.469100366956391e-08,
"loss": 0.000537961081136018,
"step": 2208
},
{
"epoch": 3.6704793594676093,
"grad_norm": 0.0015345853753387928,
"learning_rate": 9.375706770976573e-08,
"loss": 0.0016995491459965706,
"step": 2209
},
{
"epoch": 3.6721430799625665,
"grad_norm": 0.00038879108615219593,
"learning_rate": 9.282767228278672e-08,
"loss": 0.0005763666122220457,
"step": 2210
},
{
"epoch": 3.6738068004575233,
"grad_norm": 0.00027304861578159034,
"learning_rate": 9.190281914237736e-08,
"loss": 0.00019859435269609094,
"step": 2211
},
{
"epoch": 3.67547052095248,
"grad_norm": 0.00028828351059928536,
"learning_rate": 9.09825100337175e-08,
"loss": 0.00034817203413695097,
"step": 2212
},
{
"epoch": 3.677134241447437,
"grad_norm": 0.0007686371682211757,
"learning_rate": 9.006674669341214e-08,
"loss": 0.0016276794485747814,
"step": 2213
},
{
"epoch": 3.6787979619423936,
"grad_norm": 0.000721732503734529,
"learning_rate": 8.915553084948847e-08,
"loss": 0.0017058923840522766,
"step": 2214
},
{
"epoch": 3.6804616824373504,
"grad_norm": 0.0006037505227141082,
"learning_rate": 8.824886422139273e-08,
"loss": 0.001759952399879694,
"step": 2215
},
{
"epoch": 3.682125402932307,
"grad_norm": 0.00021061238658148795,
"learning_rate": 8.734674851998748e-08,
"loss": 0.0008622463792562485,
"step": 2216
},
{
"epoch": 3.683789123427264,
"grad_norm": 0.0003846586332656443,
"learning_rate": 8.64491854475466e-08,
"loss": 0.0010997102363035083,
"step": 2217
},
{
"epoch": 3.685452843922221,
"grad_norm": 0.000777195324189961,
"learning_rate": 8.55561766977539e-08,
"loss": 0.00318273832090199,
"step": 2218
},
{
"epoch": 3.687116564417178,
"grad_norm": 0.0010364966001361609,
"learning_rate": 8.46677239556995e-08,
"loss": 0.001554339425638318,
"step": 2219
},
{
"epoch": 3.6887802849121347,
"grad_norm": 0.0006480670999735594,
"learning_rate": 8.378382889787596e-08,
"loss": 0.0019870519172400236,
"step": 2220
},
{
"epoch": 3.6904440054070915,
"grad_norm": 0.0005411705933511257,
"learning_rate": 8.290449319217603e-08,
"loss": 0.001172770163975656,
"step": 2221
},
{
"epoch": 3.6921077259020487,
"grad_norm": 0.0001723592431517318,
"learning_rate": 8.202971849788854e-08,
"loss": 0.000567898852750659,
"step": 2222
},
{
"epoch": 3.6937714463970055,
"grad_norm": 0.0002687703527044505,
"learning_rate": 8.115950646569587e-08,
"loss": 0.0005061675328761339,
"step": 2223
},
{
"epoch": 3.6954351668919623,
"grad_norm": 0.0002580997534096241,
"learning_rate": 8.029385873767115e-08,
"loss": 0.0005549566121771932,
"step": 2224
},
{
"epoch": 3.697098887386919,
"grad_norm": 0.0005996919353492558,
"learning_rate": 7.943277694727469e-08,
"loss": 0.0011159824207425117,
"step": 2225
},
{
"epoch": 3.698762607881876,
"grad_norm": 0.00021931962692178786,
"learning_rate": 7.857626271935037e-08,
"loss": 0.00018094456754624844,
"step": 2226
},
{
"epoch": 3.7004263283768326,
"grad_norm": 0.0003673741302918643,
"learning_rate": 7.772431767012423e-08,
"loss": 0.0005871973698958755,
"step": 2227
},
{
"epoch": 3.7020900488717894,
"grad_norm": 0.0002457849041093141,
"learning_rate": 7.68769434071992e-08,
"loss": 0.0002511520578991622,
"step": 2228
},
{
"epoch": 3.703753769366746,
"grad_norm": 0.000532071222551167,
"learning_rate": 7.603414152955374e-08,
"loss": 0.001324426382780075,
"step": 2229
},
{
"epoch": 3.7054174898617034,
"grad_norm": 0.00074821210000664,
"learning_rate": 7.519591362753848e-08,
"loss": 0.0009245446999557316,
"step": 2230
},
{
"epoch": 3.70708121035666,
"grad_norm": 0.000523492693901062,
"learning_rate": 7.436226128287288e-08,
"loss": 0.0013367345090955496,
"step": 2231
},
{
"epoch": 3.708744930851617,
"grad_norm": 0.0005014461930841208,
"learning_rate": 7.35331860686428e-08,
"loss": 0.00041717401472851634,
"step": 2232
},
{
"epoch": 3.7104086513465737,
"grad_norm": 0.00034521002089604735,
"learning_rate": 7.270868954929595e-08,
"loss": 0.0006486955680884421,
"step": 2233
},
{
"epoch": 3.7120723718415305,
"grad_norm": 0.0002621794992592186,
"learning_rate": 7.188877328064142e-08,
"loss": 0.0001665848249103874,
"step": 2234
},
{
"epoch": 3.7137360923364877,
"grad_norm": 0.000527933007106185,
"learning_rate": 7.107343880984496e-08,
"loss": 0.0015387427993118763,
"step": 2235
},
{
"epoch": 3.7153998128314445,
"grad_norm": 0.0006754556088708341,
"learning_rate": 7.026268767542671e-08,
"loss": 0.0019412613473832607,
"step": 2236
},
{
"epoch": 3.7170635333264013,
"grad_norm": 0.00033381491084583104,
"learning_rate": 6.94565214072579e-08,
"loss": 0.000849235279019922,
"step": 2237
},
{
"epoch": 3.718727253821358,
"grad_norm": 0.00031778853735886514,
"learning_rate": 6.86549415265586e-08,
"loss": 0.0003436643164604902,
"step": 2238
},
{
"epoch": 3.720390974316315,
"grad_norm": 0.000337104604113847,
"learning_rate": 6.785794954589365e-08,
"loss": 0.00036453024949878454,
"step": 2239
},
{
"epoch": 3.7220546948112716,
"grad_norm": 0.0006949629751034081,
"learning_rate": 6.706554696917139e-08,
"loss": 0.0010551823070272803,
"step": 2240
},
{
"epoch": 3.7237184153062284,
"grad_norm": 0.00045205987407825887,
"learning_rate": 6.627773529163994e-08,
"loss": 0.00048298074398189783,
"step": 2241
},
{
"epoch": 3.725382135801185,
"grad_norm": 0.0003112118865828961,
"learning_rate": 6.549451599988432e-08,
"loss": 0.0006346139125525951,
"step": 2242
},
{
"epoch": 3.7270458562961424,
"grad_norm": 0.0003499843296594918,
"learning_rate": 6.471589057182398e-08,
"loss": 0.001009125029668212,
"step": 2243
},
{
"epoch": 3.728709576791099,
"grad_norm": 0.00034769653575494885,
"learning_rate": 6.394186047670947e-08,
"loss": 0.0003636126348283142,
"step": 2244
},
{
"epoch": 3.730373297286056,
"grad_norm": 0.00042841010144911706,
"learning_rate": 6.317242717511995e-08,
"loss": 0.00031774997478350997,
"step": 2245
},
{
"epoch": 3.7320370177810127,
"grad_norm": 0.000637256249319762,
"learning_rate": 6.240759211896153e-08,
"loss": 0.000800920941401273,
"step": 2246
},
{
"epoch": 3.7337007382759695,
"grad_norm": 0.00034591983421705663,
"learning_rate": 6.16473567514625e-08,
"loss": 0.0005393112078309059,
"step": 2247
},
{
"epoch": 3.7353644587709267,
"grad_norm": 0.001043170690536499,
"learning_rate": 6.089172250717201e-08,
"loss": 0.002363135339692235,
"step": 2248
},
{
"epoch": 3.7370281792658835,
"grad_norm": 0.0002499793190509081,
"learning_rate": 6.014069081195673e-08,
"loss": 0.0002837787615135312,
"step": 2249
},
{
"epoch": 3.7386918997608403,
"grad_norm": 0.0007483753724955022,
"learning_rate": 5.9394263082998836e-08,
"loss": 0.0016868056263774633,
"step": 2250
},
{
"epoch": 3.740355620255797,
"grad_norm": 0.0009630986023694277,
"learning_rate": 5.8652440728792504e-08,
"loss": 0.0023214765824377537,
"step": 2251
},
{
"epoch": 3.742019340750754,
"grad_norm": 0.0007593136979267001,
"learning_rate": 5.791522514914216e-08,
"loss": 0.0014121567364782095,
"step": 2252
},
{
"epoch": 3.7436830612457106,
"grad_norm": 0.00046497341827489436,
"learning_rate": 5.718261773515865e-08,
"loss": 0.001239665667526424,
"step": 2253
},
{
"epoch": 3.7453467817406674,
"grad_norm": 0.0004559413646347821,
"learning_rate": 5.64546198692581e-08,
"loss": 0.0008746258681640029,
"step": 2254
},
{
"epoch": 3.747010502235624,
"grad_norm": 0.00031668288283981383,
"learning_rate": 5.573123292515775e-08,
"loss": 0.00044546768185682595,
"step": 2255
},
{
"epoch": 3.7486742227305814,
"grad_norm": 0.0006626700051128864,
"learning_rate": 5.50124582678746e-08,
"loss": 0.0009660154464654624,
"step": 2256
},
{
"epoch": 3.750337943225538,
"grad_norm": 0.0002030782779911533,
"learning_rate": 5.429829725372204e-08,
"loss": 0.00020031498570460826,
"step": 2257
},
{
"epoch": 3.752001663720495,
"grad_norm": 0.0021811239421367645,
"learning_rate": 5.3588751230307935e-08,
"loss": 0.0007013549329712987,
"step": 2258
},
{
"epoch": 3.7536653842154517,
"grad_norm": 0.0006810147897340357,
"learning_rate": 5.2883821536531545e-08,
"loss": 0.0013047631364315748,
"step": 2259
},
{
"epoch": 3.755329104710409,
"grad_norm": 0.0003946752694901079,
"learning_rate": 5.218350950258133e-08,
"loss": 0.00029323113267309964,
"step": 2260
},
{
"epoch": 3.7569928252053657,
"grad_norm": 0.0013763107126578689,
"learning_rate": 5.1487816449932174e-08,
"loss": 0.0007352272514253855,
"step": 2261
},
{
"epoch": 3.7586565457003225,
"grad_norm": 0.000568808987736702,
"learning_rate": 5.079674369134313e-08,
"loss": 0.0008357037440873682,
"step": 2262
},
{
"epoch": 3.7603202661952793,
"grad_norm": 0.0008823948446661234,
"learning_rate": 5.0110292530854696e-08,
"loss": 0.002022108295932412,
"step": 2263
},
{
"epoch": 3.761983986690236,
"grad_norm": 0.00047194812214002013,
"learning_rate": 4.942846426378683e-08,
"loss": 0.0012379256077110767,
"step": 2264
},
{
"epoch": 3.763647707185193,
"grad_norm": 0.00037526569212786853,
"learning_rate": 4.875126017673593e-08,
"loss": 0.0010932236909866333,
"step": 2265
},
{
"epoch": 3.7653114276801496,
"grad_norm": 0.001681567169725895,
"learning_rate": 4.807868154757284e-08,
"loss": 0.001502260216511786,
"step": 2266
},
{
"epoch": 3.7669751481751064,
"grad_norm": 0.00038177220267243683,
"learning_rate": 4.741072964543958e-08,
"loss": 0.0011915108188986778,
"step": 2267
},
{
"epoch": 3.7686388686700636,
"grad_norm": 0.0006360916304402053,
"learning_rate": 4.6747405730748765e-08,
"loss": 0.0012110350653529167,
"step": 2268
},
{
"epoch": 3.7703025891650204,
"grad_norm": 0.0014272902626544237,
"learning_rate": 4.6088711055179426e-08,
"loss": 0.0026597613468766212,
"step": 2269
},
{
"epoch": 3.771966309659977,
"grad_norm": 0.0002763984084594995,
"learning_rate": 4.543464686167537e-08,
"loss": 0.0004325263435021043,
"step": 2270
},
{
"epoch": 3.773630030154934,
"grad_norm": 0.00036100976285524666,
"learning_rate": 4.478521438444267e-08,
"loss": 0.001015659305267036,
"step": 2271
},
{
"epoch": 3.7752937506498907,
"grad_norm": 0.00018730417650658637,
"learning_rate": 4.414041484894743e-08,
"loss": 0.00018422273569740355,
"step": 2272
},
{
"epoch": 3.776957471144848,
"grad_norm": 0.0004674112715292722,
"learning_rate": 4.3500249471913616e-08,
"loss": 0.001334009226411581,
"step": 2273
},
{
"epoch": 3.7786211916398047,
"grad_norm": 0.0006042693858034909,
"learning_rate": 4.2864719461321036e-08,
"loss": 0.0009468183270655572,
"step": 2274
},
{
"epoch": 3.7802849121347615,
"grad_norm": 0.000509067380335182,
"learning_rate": 4.223382601640208e-08,
"loss": 0.0006977926241233945,
"step": 2275
},
{
"epoch": 3.7819486326297183,
"grad_norm": 0.0004617350932676345,
"learning_rate": 4.160757032764001e-08,
"loss": 0.001105066854506731,
"step": 2276
},
{
"epoch": 3.783612353124675,
"grad_norm": 0.0007248525507748127,
"learning_rate": 4.098595357676732e-08,
"loss": 0.002765103243291378,
"step": 2277
},
{
"epoch": 3.785276073619632,
"grad_norm": 0.0004243562580086291,
"learning_rate": 4.036897693676184e-08,
"loss": 0.0010327072814106941,
"step": 2278
},
{
"epoch": 3.7869397941145886,
"grad_norm": 0.0010116234188899398,
"learning_rate": 3.9756641571847e-08,
"loss": 0.0029219533316791058,
"step": 2279
},
{
"epoch": 3.7886035146095454,
"grad_norm": 0.00012117379083065316,
"learning_rate": 3.914894863748714e-08,
"loss": 0.0004345136694610119,
"step": 2280
},
{
"epoch": 3.7902672351045026,
"grad_norm": 0.0007400370086543262,
"learning_rate": 3.854589928038666e-08,
"loss": 0.0018152652774006128,
"step": 2281
},
{
"epoch": 3.7919309555994594,
"grad_norm": 0.0004178027156740427,
"learning_rate": 3.794749463848835e-08,
"loss": 0.0008192003006115556,
"step": 2282
},
{
"epoch": 3.793594676094416,
"grad_norm": 0.00044253907981328666,
"learning_rate": 3.735373584096924e-08,
"loss": 0.0011070969048887491,
"step": 2283
},
{
"epoch": 3.795258396589373,
"grad_norm": 0.0009567950037308037,
"learning_rate": 3.676462400824088e-08,
"loss": 0.0016953707672655582,
"step": 2284
},
{
"epoch": 3.7969221170843297,
"grad_norm": 0.0004708846681751311,
"learning_rate": 3.618016025194598e-08,
"loss": 0.0005484845023602247,
"step": 2285
},
{
"epoch": 3.798585837579287,
"grad_norm": 0.0005545560852624476,
"learning_rate": 3.560034567495513e-08,
"loss": 0.0006607613759115338,
"step": 2286
},
{
"epoch": 3.8002495580742437,
"grad_norm": 0.0006822629366070032,
"learning_rate": 3.5025181371367844e-08,
"loss": 0.0019367439672350883,
"step": 2287
},
{
"epoch": 3.8019132785692005,
"grad_norm": 0.0006036330014467239,
"learning_rate": 3.4454668426507076e-08,
"loss": 0.000651317008305341,
"step": 2288
},
{
"epoch": 3.8035769990641572,
"grad_norm": 0.00024399223912041634,
"learning_rate": 3.388880791692001e-08,
"loss": 0.0004422190831974149,
"step": 2289
},
{
"epoch": 3.805240719559114,
"grad_norm": 0.0008544662268832326,
"learning_rate": 3.33276009103739e-08,
"loss": 0.0014506196603178978,
"step": 2290
},
{
"epoch": 3.806904440054071,
"grad_norm": 0.0008960445993579924,
"learning_rate": 3.2771048465855546e-08,
"loss": 0.0016973534366115928,
"step": 2291
},
{
"epoch": 3.8085681605490276,
"grad_norm": 0.0002961951831821352,
"learning_rate": 3.221915163356848e-08,
"loss": 0.0009070704109035432,
"step": 2292
},
{
"epoch": 3.8102318810439844,
"grad_norm": 0.0005132612423039973,
"learning_rate": 3.167191145493076e-08,
"loss": 0.001032351516187191,
"step": 2293
},
{
"epoch": 3.8118956015389416,
"grad_norm": 0.00018734767218120396,
"learning_rate": 3.1129328962573865e-08,
"loss": 0.00034245982533320785,
"step": 2294
},
{
"epoch": 3.8135593220338984,
"grad_norm": 0.0005684709176421165,
"learning_rate": 3.05914051803402e-08,
"loss": 0.0010232322383672,
"step": 2295
},
{
"epoch": 3.815223042528855,
"grad_norm": 0.0007894489681348205,
"learning_rate": 3.005814112328143e-08,
"loss": 0.0005177940474823117,
"step": 2296
},
{
"epoch": 3.816886763023812,
"grad_norm": 0.0005052572232671082,
"learning_rate": 2.9529537797656215e-08,
"loss": 0.00123726692982018,
"step": 2297
},
{
"epoch": 3.8185504835187687,
"grad_norm": 0.0003198084014002234,
"learning_rate": 2.900559620092891e-08,
"loss": 0.0007540383376181126,
"step": 2298
},
{
"epoch": 3.820214204013726,
"grad_norm": 0.00037152034929022193,
"learning_rate": 2.8486317321766432e-08,
"loss": 0.0010864018695428967,
"step": 2299
},
{
"epoch": 3.8218779245086827,
"grad_norm": 0.0003372343198861927,
"learning_rate": 2.797170214003775e-08,
"loss": 0.0008511084015481174,
"step": 2300
},
{
"epoch": 3.8218779245086827,
"eval_loss": 0.0017695488641038537,
"eval_runtime": 490.2862,
"eval_samples_per_second": 9.807,
"eval_steps_per_second": 2.452,
"step": 2300
},
{
"epoch": 3.8235416450036395,
"grad_norm": 0.0007120806840248406,
"learning_rate": 2.7461751626811916e-08,
"loss": 0.0017852610908448696,
"step": 2301
},
{
"epoch": 3.8252053654985962,
"grad_norm": 0.0023090168833732605,
"learning_rate": 2.6956466744355315e-08,
"loss": 0.0010553663596510887,
"step": 2302
},
{
"epoch": 3.826869085993553,
"grad_norm": 0.000529763288795948,
"learning_rate": 2.6455848446130526e-08,
"loss": 0.0010059673804789782,
"step": 2303
},
{
"epoch": 3.82853280648851,
"grad_norm": 0.0005984203307889402,
"learning_rate": 2.5959897676794134e-08,
"loss": 0.0006512810941785574,
"step": 2304
},
{
"epoch": 3.8301965269834666,
"grad_norm": 0.0004964310792274773,
"learning_rate": 2.546861537219586e-08,
"loss": 0.0004773701657541096,
"step": 2305
},
{
"epoch": 3.8318602474784234,
"grad_norm": 0.0003719302185345441,
"learning_rate": 2.4982002459375265e-08,
"loss": 0.0005666849319823086,
"step": 2306
},
{
"epoch": 3.8335239679733806,
"grad_norm": 0.0004134979972150177,
"learning_rate": 2.450005985656173e-08,
"loss": 0.0005266037187539041,
"step": 2307
},
{
"epoch": 3.8351876884683374,
"grad_norm": 0.0006398731493391097,
"learning_rate": 2.4022788473170853e-08,
"loss": 0.0011300542391836643,
"step": 2308
},
{
"epoch": 3.836851408963294,
"grad_norm": 0.0009114894201047719,
"learning_rate": 2.355018920980501e-08,
"loss": 0.0019518604967743158,
"step": 2309
},
{
"epoch": 3.838515129458251,
"grad_norm": 0.0005688403616659343,
"learning_rate": 2.308226295824917e-08,
"loss": 0.00036758609348908067,
"step": 2310
},
{
"epoch": 3.840178849953208,
"grad_norm": 0.0004458473122213036,
"learning_rate": 2.2619010601470925e-08,
"loss": 0.0006280520465224981,
"step": 2311
},
{
"epoch": 3.841842570448165,
"grad_norm": 0.00043085412471555173,
"learning_rate": 2.2160433013618533e-08,
"loss": 0.0002646467764861882,
"step": 2312
},
{
"epoch": 3.8435062909431217,
"grad_norm": 0.00022731930948793888,
"learning_rate": 2.170653106001841e-08,
"loss": 0.0002789350401144475,
"step": 2313
},
{
"epoch": 3.8451700114380785,
"grad_norm": 0.0002457225928083062,
"learning_rate": 2.1257305597175428e-08,
"loss": 0.0006970633985474706,
"step": 2314
},
{
"epoch": 3.8468337319330352,
"grad_norm": 0.00033074014936573803,
"learning_rate": 2.0812757472768175e-08,
"loss": 0.0007895432645455003,
"step": 2315
},
{
"epoch": 3.848497452427992,
"grad_norm": 0.0004209664766676724,
"learning_rate": 2.037288752565064e-08,
"loss": 0.0005390758742578328,
"step": 2316
},
{
"epoch": 3.850161172922949,
"grad_norm": 0.0006175689049996436,
"learning_rate": 1.99376965858486e-08,
"loss": 0.00044948633876629174,
"step": 2317
},
{
"epoch": 3.8518248934179056,
"grad_norm": 0.0005152714438736439,
"learning_rate": 1.9507185474558765e-08,
"loss": 0.00044237388647161424,
"step": 2318
},
{
"epoch": 3.853488613912863,
"grad_norm": 0.00018733202887233347,
"learning_rate": 1.908135500414743e-08,
"loss": 0.0006890572840347886,
"step": 2319
},
{
"epoch": 3.8551523344078196,
"grad_norm": 0.0010485963430255651,
"learning_rate": 1.866020597814766e-08,
"loss": 0.0014118121471256018,
"step": 2320
},
{
"epoch": 3.8568160549027763,
"grad_norm": 0.0004583117552101612,
"learning_rate": 1.8243739191259603e-08,
"loss": 0.0010312426602467895,
"step": 2321
},
{
"epoch": 3.858479775397733,
"grad_norm": 0.00041788682574406266,
"learning_rate": 1.7831955429348235e-08,
"loss": 0.0004907081602141261,
"step": 2322
},
{
"epoch": 3.86014349589269,
"grad_norm": 0.0006742252153344452,
"learning_rate": 1.7424855469440617e-08,
"loss": 0.00199217745102942,
"step": 2323
},
{
"epoch": 3.861807216387647,
"grad_norm": 0.0006253820611163974,
"learning_rate": 1.7022440079726976e-08,
"loss": 0.0010176701471209526,
"step": 2324
},
{
"epoch": 3.863470936882604,
"grad_norm": 0.0006420851568691432,
"learning_rate": 1.6624710019556844e-08,
"loss": 0.000928575056605041,
"step": 2325
},
{
"epoch": 3.8651346573775607,
"grad_norm": 0.0007003938080742955,
"learning_rate": 1.623166603943932e-08,
"loss": 0.0018214373849332333,
"step": 2326
},
{
"epoch": 3.8667983778725175,
"grad_norm": 0.0003563383943401277,
"learning_rate": 1.584330888104002e-08,
"loss": 0.00042250618571415544,
"step": 2327
},
{
"epoch": 3.8684620983674742,
"grad_norm": 0.0006026191986165941,
"learning_rate": 1.5459639277181637e-08,
"loss": 0.0012118013110011816,
"step": 2328
},
{
"epoch": 3.870125818862431,
"grad_norm": 0.0019928766414523125,
"learning_rate": 1.508065795184116e-08,
"loss": 0.001427660696208477,
"step": 2329
},
{
"epoch": 3.871789539357388,
"grad_norm": 0.0011689052917063236,
"learning_rate": 1.4706365620149043e-08,
"loss": 0.0008310202974826097,
"step": 2330
},
{
"epoch": 3.8734532598523446,
"grad_norm": 0.00027448718901723623,
"learning_rate": 1.433676298838671e-08,
"loss": 0.0003550715046003461,
"step": 2331
},
{
"epoch": 3.875116980347302,
"grad_norm": 0.00040693397750146687,
"learning_rate": 1.3971850753987936e-08,
"loss": 0.0009889578213915229,
"step": 2332
},
{
"epoch": 3.8767807008422586,
"grad_norm": 0.00041427675751037896,
"learning_rate": 1.3611629605534139e-08,
"loss": 0.000923509243875742,
"step": 2333
},
{
"epoch": 3.8784444213372153,
"grad_norm": 0.0012977409642189741,
"learning_rate": 1.325610022275603e-08,
"loss": 0.0011391546577215195,
"step": 2334
},
{
"epoch": 3.880108141832172,
"grad_norm": 0.0003560408076737076,
"learning_rate": 1.29052632765303e-08,
"loss": 0.00028428417863324285,
"step": 2335
},
{
"epoch": 3.881771862327129,
"grad_norm": 0.0002894706267397851,
"learning_rate": 1.2559119428879607e-08,
"loss": 0.0006873887032270432,
"step": 2336
},
{
"epoch": 3.883435582822086,
"grad_norm": 0.0006624959642067552,
"learning_rate": 1.2217669332970084e-08,
"loss": 0.0011978786205872893,
"step": 2337
},
{
"epoch": 3.885099303317043,
"grad_norm": 0.0004509827704168856,
"learning_rate": 1.1880913633111335e-08,
"loss": 0.0003517848963383585,
"step": 2338
},
{
"epoch": 3.8867630238119997,
"grad_norm": 0.0004664452571887523,
"learning_rate": 1.1548852964755053e-08,
"loss": 0.0006236121407710016,
"step": 2339
},
{
"epoch": 3.8884267443069565,
"grad_norm": 0.0010952012380585074,
"learning_rate": 1.122148795449307e-08,
"loss": 0.001222100923769176,
"step": 2340
},
{
"epoch": 3.8900904648019132,
"grad_norm": 0.0004181310359854251,
"learning_rate": 1.0898819220056811e-08,
"loss": 0.0012167840031906962,
"step": 2341
},
{
"epoch": 3.89175418529687,
"grad_norm": 0.0009198611951433122,
"learning_rate": 1.058084737031534e-08,
"loss": 0.001886456273496151,
"step": 2342
},
{
"epoch": 3.893417905791827,
"grad_norm": 0.0006522826733998954,
"learning_rate": 1.0267573005275645e-08,
"loss": 0.002321903593838215,
"step": 2343
},
{
"epoch": 3.8950816262867836,
"grad_norm": 0.0005509674083441496,
"learning_rate": 9.95899671607986e-09,
"loss": 0.0020275390706956387,
"step": 2344
},
{
"epoch": 3.896745346781741,
"grad_norm": 0.0004306949267629534,
"learning_rate": 9.655119085005827e-09,
"loss": 0.0005395016050897539,
"step": 2345
},
{
"epoch": 3.8984090672766976,
"grad_norm": 0.0005223198095336556,
"learning_rate": 9.355940685464305e-09,
"loss": 0.0008524530567228794,
"step": 2346
},
{
"epoch": 3.9000727877716543,
"grad_norm": 0.0007850914262235165,
"learning_rate": 9.061462081999262e-09,
"loss": 0.001647283905185759,
"step": 2347
},
{
"epoch": 3.901736508266611,
"grad_norm": 0.0004000430053565651,
"learning_rate": 8.771683830285649e-09,
"loss": 0.0007437191670760512,
"step": 2348
},
{
"epoch": 3.903400228761568,
"grad_norm": 0.0003827679029200226,
"learning_rate": 8.486606477129677e-09,
"loss": 0.00033460953272879124,
"step": 2349
},
{
"epoch": 3.905063949256525,
"grad_norm": 0.0010400597238913178,
"learning_rate": 8.206230560466322e-09,
"loss": 0.001289241947233677,
"step": 2350
},
{
"epoch": 3.906727669751482,
"grad_norm": 0.0002617063873913139,
"learning_rate": 7.930556609359596e-09,
"loss": 0.00028740576817654073,
"step": 2351
},
{
"epoch": 3.9083913902464387,
"grad_norm": 0.00038715035771019757,
"learning_rate": 7.659585144000892e-09,
"loss": 0.0006691685412079096,
"step": 2352
},
{
"epoch": 3.9100551107413954,
"grad_norm": 0.00024767228751443326,
"learning_rate": 7.393316675707584e-09,
"loss": 0.0009464296163059771,
"step": 2353
},
{
"epoch": 3.9117188312363522,
"grad_norm": 0.0005090117338113487,
"learning_rate": 7.131751706923595e-09,
"loss": 0.001270594191737473,
"step": 2354
},
{
"epoch": 3.913382551731309,
"grad_norm": 0.0008436394273303449,
"learning_rate": 6.8748907312163325e-09,
"loss": 0.0014107085298746824,
"step": 2355
},
{
"epoch": 3.9150462722262658,
"grad_norm": 0.00042344772373326123,
"learning_rate": 6.622734233277528e-09,
"loss": 0.00029759103199467063,
"step": 2356
},
{
"epoch": 3.916709992721223,
"grad_norm": 0.0011615968542173505,
"learning_rate": 6.375282688921569e-09,
"loss": 0.0027657081373035908,
"step": 2357
},
{
"epoch": 3.9183737132161798,
"grad_norm": 0.0006488868384622037,
"learning_rate": 6.132536565084945e-09,
"loss": 0.0007090591825544834,
"step": 2358
},
{
"epoch": 3.9200374337111366,
"grad_norm": 0.00039018920506350696,
"learning_rate": 5.894496319824306e-09,
"loss": 0.0009499025763943791,
"step": 2359
},
{
"epoch": 3.9217011542060933,
"grad_norm": 0.0005614451947622001,
"learning_rate": 5.661162402316733e-09,
"loss": 0.0015096960123628378,
"step": 2360
},
{
"epoch": 3.92336487470105,
"grad_norm": 0.0004525469848886132,
"learning_rate": 5.432535252859472e-09,
"loss": 0.0007028505788184702,
"step": 2361
},
{
"epoch": 3.9250285951960073,
"grad_norm": 0.000415037851780653,
"learning_rate": 5.208615302866593e-09,
"loss": 0.0008837031782604754,
"step": 2362
},
{
"epoch": 3.926692315690964,
"grad_norm": 0.0017863917164504528,
"learning_rate": 4.989402974871216e-09,
"loss": 0.0014681301545351744,
"step": 2363
},
{
"epoch": 3.928356036185921,
"grad_norm": 0.000572762219235301,
"learning_rate": 4.774898682522455e-09,
"loss": 0.0012490102089941502,
"step": 2364
},
{
"epoch": 3.9300197566808777,
"grad_norm": 0.0006480145384557545,
"learning_rate": 4.565102830585699e-09,
"loss": 0.000560228421818465,
"step": 2365
},
{
"epoch": 3.9316834771758344,
"grad_norm": 0.0008436389616690576,
"learning_rate": 4.360015814941498e-09,
"loss": 0.0010321622248739004,
"step": 2366
},
{
"epoch": 3.933347197670791,
"grad_norm": 0.0009871098445728421,
"learning_rate": 4.159638022585011e-09,
"loss": 0.001806100714020431,
"step": 2367
},
{
"epoch": 3.935010918165748,
"grad_norm": 0.0009362440323457122,
"learning_rate": 3.96396983162517e-09,
"loss": 0.0003677202039398253,
"step": 2368
},
{
"epoch": 3.9366746386607048,
"grad_norm": 0.0002989772765431553,
"learning_rate": 3.773011611284128e-09,
"loss": 0.0011360242497175932,
"step": 2369
},
{
"epoch": 3.938338359155662,
"grad_norm": 0.0009438424604013562,
"learning_rate": 3.586763721896147e-09,
"loss": 0.0016448276583105326,
"step": 2370
},
{
"epoch": 3.9400020796506188,
"grad_norm": 0.0011332824360579252,
"learning_rate": 3.4052265149070453e-09,
"loss": 0.0027621465269476175,
"step": 2371
},
{
"epoch": 3.9416658001455755,
"grad_norm": 0.0005722562782466412,
"learning_rate": 3.2284003328744706e-09,
"loss": 0.0005765999085269868,
"step": 2372
},
{
"epoch": 3.9433295206405323,
"grad_norm": 0.00041134137427434325,
"learning_rate": 3.056285509465684e-09,
"loss": 0.00039809662848711014,
"step": 2373
},
{
"epoch": 3.944993241135489,
"grad_norm": 0.002205124357715249,
"learning_rate": 2.888882369457835e-09,
"loss": 0.0011762338690459728,
"step": 2374
},
{
"epoch": 3.9466569616304463,
"grad_norm": 0.0006662149680778384,
"learning_rate": 2.726191228737407e-09,
"loss": 0.0012037127744406462,
"step": 2375
},
{
"epoch": 3.948320682125403,
"grad_norm": 0.00028259915416128933,
"learning_rate": 2.5682123942993852e-09,
"loss": 0.0002108765474986285,
"step": 2376
},
{
"epoch": 3.94998440262036,
"grad_norm": 0.00035649127676151693,
"learning_rate": 2.414946164246701e-09,
"loss": 0.0012281450908631086,
"step": 2377
},
{
"epoch": 3.9516481231153167,
"grad_norm": 0.0008716229349374771,
"learning_rate": 2.2663928277896763e-09,
"loss": 0.0019710322376340628,
"step": 2378
},
{
"epoch": 3.9533118436102734,
"grad_norm": 0.00026112596970051527,
"learning_rate": 2.122552665245747e-09,
"loss": 0.0003657359629869461,
"step": 2379
},
{
"epoch": 3.95497556410523,
"grad_norm": 0.0003548185050021857,
"learning_rate": 1.9834259480380756e-09,
"loss": 0.00039860629476606846,
"step": 2380
},
{
"epoch": 3.956639284600187,
"grad_norm": 0.0009071942768059671,
"learning_rate": 1.8490129386963818e-09,
"loss": 0.00256812060251832,
"step": 2381
},
{
"epoch": 3.9583030050951438,
"grad_norm": 0.00014368511619977653,
"learning_rate": 1.719313890855001e-09,
"loss": 0.000299682782497257,
"step": 2382
},
{
"epoch": 3.959966725590101,
"grad_norm": 0.0008304541115649045,
"learning_rate": 1.5943290492539953e-09,
"loss": 0.0019825128838419914,
"step": 2383
},
{
"epoch": 3.9616304460850578,
"grad_norm": 0.0004863161302637309,
"learning_rate": 1.4740586497366538e-09,
"loss": 0.0006778284441679716,
"step": 2384
},
{
"epoch": 3.9632941665800145,
"grad_norm": 0.00045845622662454844,
"learning_rate": 1.358502919251159e-09,
"loss": 0.0009040733566507697,
"step": 2385
},
{
"epoch": 3.9649578870749713,
"grad_norm": 0.0006530190003104508,
"learning_rate": 1.247662075848921e-09,
"loss": 0.0013933960581198335,
"step": 2386
},
{
"epoch": 3.966621607569928,
"grad_norm": 0.00040783261647447944,
"learning_rate": 1.1415363286843007e-09,
"loss": 0.0008399755461141467,
"step": 2387
},
{
"epoch": 3.9682853280648853,
"grad_norm": 0.0008198352297767997,
"learning_rate": 1.0401258780146084e-09,
"loss": 0.001252037356607616,
"step": 2388
},
{
"epoch": 3.969949048559842,
"grad_norm": 0.0005774135352112353,
"learning_rate": 9.434309151992727e-10,
"loss": 0.001550138695165515,
"step": 2389
},
{
"epoch": 3.971612769054799,
"grad_norm": 0.001226370339281857,
"learning_rate": 8.514516226998393e-10,
"loss": 0.001360030728392303,
"step": 2390
},
{
"epoch": 3.9732764895497557,
"grad_norm": 0.0003287496219854802,
"learning_rate": 7.641881740794166e-10,
"loss": 0.00078690325608477,
"step": 2391
},
{
"epoch": 3.9749402100447124,
"grad_norm": 0.0006232394371181726,
"learning_rate": 6.816407340023978e-10,
"loss": 0.0008975286036729813,
"step": 2392
},
{
"epoch": 3.976603930539669,
"grad_norm": 0.000568083138205111,
"learning_rate": 6.03809458233906e-10,
"loss": 0.0012704274849966168,
"step": 2393
},
{
"epoch": 3.978267651034626,
"grad_norm": 0.00043166911927983165,
"learning_rate": 5.306944936406266e-10,
"loss": 0.0010767162311822176,
"step": 2394
},
{
"epoch": 3.9799313715295828,
"grad_norm": 0.0003333722706884146,
"learning_rate": 4.622959781883096e-10,
"loss": 0.0009152543498203158,
"step": 2395
},
{
"epoch": 3.98159509202454,
"grad_norm": 0.0008719568722881377,
"learning_rate": 3.9861404094426734e-10,
"loss": 0.0020166838075965643,
"step": 2396
},
{
"epoch": 3.9832588125194968,
"grad_norm": 0.00061134371208027,
"learning_rate": 3.3964880207459916e-10,
"loss": 0.0016927790129557252,
"step": 2397
},
{
"epoch": 3.9849225330144535,
"grad_norm": 0.00016088728443719447,
"learning_rate": 2.8540037284557897e-10,
"loss": 0.0004834794672206044,
"step": 2398
},
{
"epoch": 3.9865862535094103,
"grad_norm": 0.00041522938408888876,
"learning_rate": 2.358688556233779e-10,
"loss": 0.0015406750608235598,
"step": 2399
},
{
"epoch": 3.9882499740043675,
"grad_norm": 0.0003348875616211444,
"learning_rate": 1.9105434387239886e-10,
"loss": 0.0004655818920582533,
"step": 2400
},
{
"epoch": 3.9882499740043675,
"eval_loss": 0.001770942471921444,
"eval_runtime": 490.0651,
"eval_samples_per_second": 9.811,
"eval_steps_per_second": 2.453,
"step": 2400
},
{
"epoch": 3.9899136944993243,
"grad_norm": 0.0005287445383146405,
"learning_rate": 1.509569221569418e-10,
"loss": 0.0008558098925277591,
"step": 2401
},
{
"epoch": 3.991577414994281,
"grad_norm": 0.000635623000562191,
"learning_rate": 1.1557666614037122e-10,
"loss": 0.001685410039499402,
"step": 2402
},
{
"epoch": 3.993241135489238,
"grad_norm": 0.001174519187770784,
"learning_rate": 8.49136425840058e-11,
"loss": 0.0027791480533778667,
"step": 2403
},
{
"epoch": 3.9949048559841946,
"grad_norm": 0.0006644176319241524,
"learning_rate": 5.896790934878383e-11,
"loss": 0.0005844932748004794,
"step": 2404
},
{
"epoch": 3.9965685764791514,
"grad_norm": 0.0004365496861282736,
"learning_rate": 3.7739515393320215e-11,
"loss": 0.0005024238489568233,
"step": 2405
},
{
"epoch": 3.998232296974108,
"grad_norm": 0.00023726493236608803,
"learning_rate": 2.122850077584948e-11,
"loss": 0.00023512027109973133,
"step": 2406
},
{
"epoch": 3.999896017469065,
"grad_norm": 0.0007440268527716398,
"learning_rate": 9.434896651727699e-12,
"loss": 0.0026000516954809427,
"step": 2407
},
{
"epoch": 4.0,
"grad_norm": 0.0002130132488673553,
"learning_rate": 2.358725275652951e-12,
"loss": 9.920789307216182e-05,
"step": 2408
},
{
"epoch": 4.0,
"eval_loss": 0.0017681519966572523,
"eval_runtime": 491.0341,
"eval_samples_per_second": 9.792,
"eval_steps_per_second": 2.448,
"step": 2408
},
{
"epoch": 4.0,
"step": 2408,
"total_flos": 6.475362492187935e+17,
"train_loss": 0.0011171495650682802,
"train_runtime": 68951.8007,
"train_samples_per_second": 2.231,
"train_steps_per_second": 0.035
}
],
"logging_steps": 1,
"max_steps": 2408,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.475362492187935e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}