gyung's picture
Upload Gemma 4 native SFT 1Epoch checkpoint
70cc982 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1021,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009802720254870727,
"grad_norm": 40.057281494140625,
"learning_rate": 0.0,
"loss": 11.943012237548828,
"step": 1
},
{
"epoch": 0.0019605440509741453,
"grad_norm": 39.88346481323242,
"learning_rate": 1.6129032258064518e-07,
"loss": 11.98295783996582,
"step": 2
},
{
"epoch": 0.002940816076461218,
"grad_norm": 39.42152404785156,
"learning_rate": 3.2258064516129035e-07,
"loss": 11.469547271728516,
"step": 3
},
{
"epoch": 0.003921088101948291,
"grad_norm": 39.722267150878906,
"learning_rate": 4.838709677419355e-07,
"loss": 11.4144287109375,
"step": 4
},
{
"epoch": 0.004901360127435363,
"grad_norm": 39.66934585571289,
"learning_rate": 6.451612903225807e-07,
"loss": 11.902497291564941,
"step": 5
},
{
"epoch": 0.005881632152922436,
"grad_norm": 39.698204040527344,
"learning_rate": 8.064516129032258e-07,
"loss": 10.937923431396484,
"step": 6
},
{
"epoch": 0.006861904178409509,
"grad_norm": 37.903663635253906,
"learning_rate": 9.67741935483871e-07,
"loss": 10.855112075805664,
"step": 7
},
{
"epoch": 0.007842176203896581,
"grad_norm": 37.05473327636719,
"learning_rate": 1.1290322580645162e-06,
"loss": 11.450807571411133,
"step": 8
},
{
"epoch": 0.008822448229383655,
"grad_norm": 28.515823364257812,
"learning_rate": 1.2903225806451614e-06,
"loss": 9.81201171875,
"step": 9
},
{
"epoch": 0.009802720254870726,
"grad_norm": 27.91981315612793,
"learning_rate": 1.4516129032258066e-06,
"loss": 9.86862564086914,
"step": 10
},
{
"epoch": 0.0107829922803578,
"grad_norm": 25.23931312561035,
"learning_rate": 1.6129032258064516e-06,
"loss": 8.997453689575195,
"step": 11
},
{
"epoch": 0.011763264305844873,
"grad_norm": 11.561809539794922,
"learning_rate": 1.774193548387097e-06,
"loss": 6.921294212341309,
"step": 12
},
{
"epoch": 0.012743536331331944,
"grad_norm": 10.728087425231934,
"learning_rate": 1.935483870967742e-06,
"loss": 6.6136064529418945,
"step": 13
},
{
"epoch": 0.013723808356819018,
"grad_norm": 11.228537559509277,
"learning_rate": 2.096774193548387e-06,
"loss": 7.418888092041016,
"step": 14
},
{
"epoch": 0.01470408038230609,
"grad_norm": 8.592437744140625,
"learning_rate": 2.2580645161290324e-06,
"loss": 6.524759292602539,
"step": 15
},
{
"epoch": 0.015684352407793162,
"grad_norm": 5.832430362701416,
"learning_rate": 2.4193548387096776e-06,
"loss": 5.145276069641113,
"step": 16
},
{
"epoch": 0.016664624433280234,
"grad_norm": 5.456412315368652,
"learning_rate": 2.580645161290323e-06,
"loss": 4.973178386688232,
"step": 17
},
{
"epoch": 0.01764489645876731,
"grad_norm": 5.159201622009277,
"learning_rate": 2.7419354838709676e-06,
"loss": 4.893097877502441,
"step": 18
},
{
"epoch": 0.01862516848425438,
"grad_norm": 4.742326259613037,
"learning_rate": 2.903225806451613e-06,
"loss": 4.998117923736572,
"step": 19
},
{
"epoch": 0.019605440509741452,
"grad_norm": 3.7043521404266357,
"learning_rate": 3.0645161290322584e-06,
"loss": 4.396552085876465,
"step": 20
},
{
"epoch": 0.020585712535228527,
"grad_norm": 2.6263623237609863,
"learning_rate": 3.225806451612903e-06,
"loss": 3.8755574226379395,
"step": 21
},
{
"epoch": 0.0215659845607156,
"grad_norm": 2.369710683822632,
"learning_rate": 3.3870967741935484e-06,
"loss": 3.748749256134033,
"step": 22
},
{
"epoch": 0.02254625658620267,
"grad_norm": 2.5380759239196777,
"learning_rate": 3.548387096774194e-06,
"loss": 3.667205810546875,
"step": 23
},
{
"epoch": 0.023526528611689745,
"grad_norm": 2.050586223602295,
"learning_rate": 3.7096774193548392e-06,
"loss": 3.457428216934204,
"step": 24
},
{
"epoch": 0.024506800637176817,
"grad_norm": 1.508134126663208,
"learning_rate": 3.870967741935484e-06,
"loss": 3.092487096786499,
"step": 25
},
{
"epoch": 0.02548707266266389,
"grad_norm": 1.3242005109786987,
"learning_rate": 4.032258064516129e-06,
"loss": 3.138925075531006,
"step": 26
},
{
"epoch": 0.026467344688150964,
"grad_norm": 1.3606603145599365,
"learning_rate": 4.193548387096774e-06,
"loss": 2.8406543731689453,
"step": 27
},
{
"epoch": 0.027447616713638035,
"grad_norm": 1.321070671081543,
"learning_rate": 4.35483870967742e-06,
"loss": 2.524233102798462,
"step": 28
},
{
"epoch": 0.028427888739125107,
"grad_norm": 1.2182248830795288,
"learning_rate": 4.516129032258065e-06,
"loss": 2.5626869201660156,
"step": 29
},
{
"epoch": 0.02940816076461218,
"grad_norm": 1.1130256652832031,
"learning_rate": 4.67741935483871e-06,
"loss": 2.4619808197021484,
"step": 30
},
{
"epoch": 0.030388432790099253,
"grad_norm": 2.651745557785034,
"learning_rate": 4.838709677419355e-06,
"loss": 2.2935352325439453,
"step": 31
},
{
"epoch": 0.031368704815586325,
"grad_norm": 1.0515562295913696,
"learning_rate": 5e-06,
"loss": 2.209563970565796,
"step": 32
},
{
"epoch": 0.0323489768410734,
"grad_norm": 1.356811285018921,
"learning_rate": 5.161290322580646e-06,
"loss": 2.456648349761963,
"step": 33
},
{
"epoch": 0.03332924886656047,
"grad_norm": 0.9561535716056824,
"learning_rate": 5.322580645161291e-06,
"loss": 2.1772305965423584,
"step": 34
},
{
"epoch": 0.03430952089204754,
"grad_norm": 0.909472644329071,
"learning_rate": 5.483870967741935e-06,
"loss": 2.137545108795166,
"step": 35
},
{
"epoch": 0.03528979291753462,
"grad_norm": 0.9167818427085876,
"learning_rate": 5.645161290322582e-06,
"loss": 1.8368252515792847,
"step": 36
},
{
"epoch": 0.036270064943021686,
"grad_norm": 0.820106029510498,
"learning_rate": 5.806451612903226e-06,
"loss": 1.6977366209030151,
"step": 37
},
{
"epoch": 0.03725033696850876,
"grad_norm": 0.9031343460083008,
"learning_rate": 5.967741935483872e-06,
"loss": 1.969242811203003,
"step": 38
},
{
"epoch": 0.038230608993995836,
"grad_norm": 0.8737913966178894,
"learning_rate": 6.129032258064517e-06,
"loss": 1.725028395652771,
"step": 39
},
{
"epoch": 0.039210881019482904,
"grad_norm": 0.7886973023414612,
"learning_rate": 6.290322580645162e-06,
"loss": 1.4492013454437256,
"step": 40
},
{
"epoch": 0.04019115304496998,
"grad_norm": 0.8908096551895142,
"learning_rate": 6.451612903225806e-06,
"loss": 1.5003480911254883,
"step": 41
},
{
"epoch": 0.041171425070457054,
"grad_norm": 0.6863641738891602,
"learning_rate": 6.612903225806452e-06,
"loss": 1.3504502773284912,
"step": 42
},
{
"epoch": 0.04215169709594412,
"grad_norm": 1.157631754875183,
"learning_rate": 6.774193548387097e-06,
"loss": 1.5246672630310059,
"step": 43
},
{
"epoch": 0.0431319691214312,
"grad_norm": 0.7163594365119934,
"learning_rate": 6.935483870967743e-06,
"loss": 1.4303929805755615,
"step": 44
},
{
"epoch": 0.04411224114691827,
"grad_norm": 0.6780818104743958,
"learning_rate": 7.096774193548388e-06,
"loss": 1.1126534938812256,
"step": 45
},
{
"epoch": 0.04509251317240534,
"grad_norm": 0.7238319516181946,
"learning_rate": 7.258064516129033e-06,
"loss": 1.2857847213745117,
"step": 46
},
{
"epoch": 0.046072785197892416,
"grad_norm": 0.6476781964302063,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.188124656677246,
"step": 47
},
{
"epoch": 0.04705305722337949,
"grad_norm": 0.639792263507843,
"learning_rate": 7.580645161290323e-06,
"loss": 1.1587282419204712,
"step": 48
},
{
"epoch": 0.04803332924886656,
"grad_norm": 0.7395796775817871,
"learning_rate": 7.741935483870968e-06,
"loss": 1.3939350843429565,
"step": 49
},
{
"epoch": 0.049013601274353634,
"grad_norm": 0.6860636472702026,
"learning_rate": 7.903225806451613e-06,
"loss": 1.2922598123550415,
"step": 50
},
{
"epoch": 0.04999387329984071,
"grad_norm": 1.0649868249893188,
"learning_rate": 8.064516129032258e-06,
"loss": 1.4464714527130127,
"step": 51
},
{
"epoch": 0.05097414532532778,
"grad_norm": 0.5850899815559387,
"learning_rate": 8.225806451612904e-06,
"loss": 1.0149272680282593,
"step": 52
},
{
"epoch": 0.05195441735081485,
"grad_norm": 0.9045224785804749,
"learning_rate": 8.387096774193549e-06,
"loss": 1.2689924240112305,
"step": 53
},
{
"epoch": 0.05293468937630193,
"grad_norm": 0.6685839891433716,
"learning_rate": 8.548387096774194e-06,
"loss": 1.0457082986831665,
"step": 54
},
{
"epoch": 0.053914961401788995,
"grad_norm": 0.5899534225463867,
"learning_rate": 8.70967741935484e-06,
"loss": 1.0730246305465698,
"step": 55
},
{
"epoch": 0.05489523342727607,
"grad_norm": 0.5628290772438049,
"learning_rate": 8.870967741935484e-06,
"loss": 0.8884706497192383,
"step": 56
},
{
"epoch": 0.05587550545276314,
"grad_norm": 1.536063551902771,
"learning_rate": 9.03225806451613e-06,
"loss": 1.0616618394851685,
"step": 57
},
{
"epoch": 0.05685577747825021,
"grad_norm": 0.8790244460105896,
"learning_rate": 9.193548387096775e-06,
"loss": 1.168012022972107,
"step": 58
},
{
"epoch": 0.05783604950373729,
"grad_norm": 0.6255394816398621,
"learning_rate": 9.35483870967742e-06,
"loss": 0.9364565014839172,
"step": 59
},
{
"epoch": 0.05881632152922436,
"grad_norm": 0.5982190370559692,
"learning_rate": 9.516129032258065e-06,
"loss": 0.8512774705886841,
"step": 60
},
{
"epoch": 0.05979659355471143,
"grad_norm": 0.6719162464141846,
"learning_rate": 9.67741935483871e-06,
"loss": 1.0239495038986206,
"step": 61
},
{
"epoch": 0.06077686558019851,
"grad_norm": 0.6623294353485107,
"learning_rate": 9.838709677419356e-06,
"loss": 0.9806511998176575,
"step": 62
},
{
"epoch": 0.061757137605685575,
"grad_norm": 1.3735675811767578,
"learning_rate": 1e-05,
"loss": 0.9417006969451904,
"step": 63
},
{
"epoch": 0.06273740963117265,
"grad_norm": 0.6799046993255615,
"learning_rate": 9.999993706252978e-06,
"loss": 0.9604445695877075,
"step": 64
},
{
"epoch": 0.06371768165665972,
"grad_norm": 0.592521607875824,
"learning_rate": 9.999974825027756e-06,
"loss": 0.920391857624054,
"step": 65
},
{
"epoch": 0.0646979536821468,
"grad_norm": 0.6740677952766418,
"learning_rate": 9.999943356371867e-06,
"loss": 0.8046770095825195,
"step": 66
},
{
"epoch": 0.06567822570763387,
"grad_norm": 0.5230652689933777,
"learning_rate": 9.999899300364534e-06,
"loss": 0.6587795615196228,
"step": 67
},
{
"epoch": 0.06665849773312094,
"grad_norm": 0.5710243582725525,
"learning_rate": 9.999842657116667e-06,
"loss": 0.8114774227142334,
"step": 68
},
{
"epoch": 0.06763876975860801,
"grad_norm": 0.5118749141693115,
"learning_rate": 9.999773426770864e-06,
"loss": 0.8066937327384949,
"step": 69
},
{
"epoch": 0.06861904178409509,
"grad_norm": 0.7899643182754517,
"learning_rate": 9.999691609501418e-06,
"loss": 0.8518644571304321,
"step": 70
},
{
"epoch": 0.06959931380958216,
"grad_norm": 0.5238215327262878,
"learning_rate": 9.999597205514298e-06,
"loss": 0.7802553176879883,
"step": 71
},
{
"epoch": 0.07057958583506924,
"grad_norm": 0.758558452129364,
"learning_rate": 9.999490215047167e-06,
"loss": 0.7075889706611633,
"step": 72
},
{
"epoch": 0.07155985786055631,
"grad_norm": 0.5630487203598022,
"learning_rate": 9.999370638369377e-06,
"loss": 0.8153989315032959,
"step": 73
},
{
"epoch": 0.07254012988604337,
"grad_norm": 0.5582584142684937,
"learning_rate": 9.999238475781957e-06,
"loss": 0.8560312986373901,
"step": 74
},
{
"epoch": 0.07352040191153045,
"grad_norm": 0.5598907470703125,
"learning_rate": 9.99909372761763e-06,
"loss": 0.8015627861022949,
"step": 75
},
{
"epoch": 0.07450067393701752,
"grad_norm": 0.8003693222999573,
"learning_rate": 9.998936394240798e-06,
"loss": 0.9214512705802917,
"step": 76
},
{
"epoch": 0.0754809459625046,
"grad_norm": 0.46098390221595764,
"learning_rate": 9.998766476047546e-06,
"loss": 0.7340844869613647,
"step": 77
},
{
"epoch": 0.07646121798799167,
"grad_norm": 0.922773540019989,
"learning_rate": 9.998583973465647e-06,
"loss": 0.7753450870513916,
"step": 78
},
{
"epoch": 0.07744149001347873,
"grad_norm": 0.9239602088928223,
"learning_rate": 9.998388886954546e-06,
"loss": 0.7383452653884888,
"step": 79
},
{
"epoch": 0.07842176203896581,
"grad_norm": 0.563861072063446,
"learning_rate": 9.998181217005377e-06,
"loss": 0.7080527544021606,
"step": 80
},
{
"epoch": 0.07940203406445288,
"grad_norm": 0.6129465103149414,
"learning_rate": 9.997960964140946e-06,
"loss": 0.8023513555526733,
"step": 81
},
{
"epoch": 0.08038230608993996,
"grad_norm": 0.9423953890800476,
"learning_rate": 9.997728128915742e-06,
"loss": 0.7608747482299805,
"step": 82
},
{
"epoch": 0.08136257811542703,
"grad_norm": 0.6925708651542664,
"learning_rate": 9.997482711915926e-06,
"loss": 0.8242323994636536,
"step": 83
},
{
"epoch": 0.08234285014091411,
"grad_norm": 0.5418462157249451,
"learning_rate": 9.997224713759335e-06,
"loss": 0.828717827796936,
"step": 84
},
{
"epoch": 0.08332312216640117,
"grad_norm": 0.7097539901733398,
"learning_rate": 9.99695413509548e-06,
"loss": 0.863600492477417,
"step": 85
},
{
"epoch": 0.08430339419188825,
"grad_norm": 0.6111697554588318,
"learning_rate": 9.99667097660554e-06,
"loss": 0.7708598971366882,
"step": 86
},
{
"epoch": 0.08528366621737532,
"grad_norm": 0.5464909672737122,
"learning_rate": 9.99637523900237e-06,
"loss": 0.7777910232543945,
"step": 87
},
{
"epoch": 0.0862639382428624,
"grad_norm": 0.613527774810791,
"learning_rate": 9.996066923030484e-06,
"loss": 0.6923108696937561,
"step": 88
},
{
"epoch": 0.08724421026834947,
"grad_norm": 0.553558886051178,
"learning_rate": 9.995746029466071e-06,
"loss": 0.6494728326797485,
"step": 89
},
{
"epoch": 0.08822448229383655,
"grad_norm": 0.7747089266777039,
"learning_rate": 9.995412559116979e-06,
"loss": 0.6648241281509399,
"step": 90
},
{
"epoch": 0.0892047543193236,
"grad_norm": 0.5550833344459534,
"learning_rate": 9.99506651282272e-06,
"loss": 0.777977705001831,
"step": 91
},
{
"epoch": 0.09018502634481068,
"grad_norm": 0.4516574442386627,
"learning_rate": 9.994707891454466e-06,
"loss": 0.6407644748687744,
"step": 92
},
{
"epoch": 0.09116529837029776,
"grad_norm": 0.651940107345581,
"learning_rate": 9.994336695915041e-06,
"loss": 0.7764822840690613,
"step": 93
},
{
"epoch": 0.09214557039578483,
"grad_norm": 0.5299152135848999,
"learning_rate": 9.993952927138933e-06,
"loss": 0.7924851179122925,
"step": 94
},
{
"epoch": 0.0931258424212719,
"grad_norm": 0.44643956422805786,
"learning_rate": 9.993556586092281e-06,
"loss": 0.6223613619804382,
"step": 95
},
{
"epoch": 0.09410611444675898,
"grad_norm": 0.5021830797195435,
"learning_rate": 9.993147673772869e-06,
"loss": 0.8323131799697876,
"step": 96
},
{
"epoch": 0.09508638647224604,
"grad_norm": 0.4259116053581238,
"learning_rate": 9.992726191210139e-06,
"loss": 0.6581062078475952,
"step": 97
},
{
"epoch": 0.09606665849773312,
"grad_norm": 0.6330034732818604,
"learning_rate": 9.992292139465166e-06,
"loss": 0.630506694316864,
"step": 98
},
{
"epoch": 0.09704693052322019,
"grad_norm": 0.4887732267379761,
"learning_rate": 9.991845519630679e-06,
"loss": 0.8050805330276489,
"step": 99
},
{
"epoch": 0.09802720254870727,
"grad_norm": 0.444149911403656,
"learning_rate": 9.991386332831042e-06,
"loss": 0.6268649101257324,
"step": 100
},
{
"epoch": 0.09900747457419434,
"grad_norm": 2.9208261966705322,
"learning_rate": 9.990914580222258e-06,
"loss": 0.6339569091796875,
"step": 101
},
{
"epoch": 0.09998774659968142,
"grad_norm": 0.5938975811004639,
"learning_rate": 9.990430262991962e-06,
"loss": 0.8084681034088135,
"step": 102
},
{
"epoch": 0.10096801862516848,
"grad_norm": 0.4514586925506592,
"learning_rate": 9.989933382359423e-06,
"loss": 0.6037747859954834,
"step": 103
},
{
"epoch": 0.10194829065065555,
"grad_norm": 0.7299878001213074,
"learning_rate": 9.989423939575538e-06,
"loss": 0.7113616466522217,
"step": 104
},
{
"epoch": 0.10292856267614263,
"grad_norm": 1.390055775642395,
"learning_rate": 9.988901935922826e-06,
"loss": 0.5878804922103882,
"step": 105
},
{
"epoch": 0.1039088347016297,
"grad_norm": 0.5749118328094482,
"learning_rate": 9.988367372715434e-06,
"loss": 0.7539823651313782,
"step": 106
},
{
"epoch": 0.10488910672711678,
"grad_norm": 0.3935142159461975,
"learning_rate": 9.987820251299121e-06,
"loss": 0.6046736240386963,
"step": 107
},
{
"epoch": 0.10586937875260385,
"grad_norm": 0.490558922290802,
"learning_rate": 9.987260573051268e-06,
"loss": 0.6678036451339722,
"step": 108
},
{
"epoch": 0.10684965077809092,
"grad_norm": 0.447732150554657,
"learning_rate": 9.986688339380863e-06,
"loss": 0.6055562496185303,
"step": 109
},
{
"epoch": 0.10782992280357799,
"grad_norm": 0.7977247834205627,
"learning_rate": 9.9861035517285e-06,
"loss": 0.6457165479660034,
"step": 110
},
{
"epoch": 0.10881019482906507,
"grad_norm": 0.553860068321228,
"learning_rate": 9.985506211566388e-06,
"loss": 0.656542181968689,
"step": 111
},
{
"epoch": 0.10979046685455214,
"grad_norm": 0.5382978320121765,
"learning_rate": 9.984896320398326e-06,
"loss": 0.7394854426383972,
"step": 112
},
{
"epoch": 0.11077073888003922,
"grad_norm": 0.4068613350391388,
"learning_rate": 9.984273879759713e-06,
"loss": 0.5744141340255737,
"step": 113
},
{
"epoch": 0.11175101090552628,
"grad_norm": 0.6466748118400574,
"learning_rate": 9.983638891217544e-06,
"loss": 0.8340737223625183,
"step": 114
},
{
"epoch": 0.11273128293101335,
"grad_norm": 0.4718203842639923,
"learning_rate": 9.982991356370404e-06,
"loss": 0.7122345566749573,
"step": 115
},
{
"epoch": 0.11371155495650043,
"grad_norm": 0.5451315641403198,
"learning_rate": 9.982331276848457e-06,
"loss": 0.803488552570343,
"step": 116
},
{
"epoch": 0.1146918269819875,
"grad_norm": 0.568440318107605,
"learning_rate": 9.981658654313458e-06,
"loss": 0.5009474754333496,
"step": 117
},
{
"epoch": 0.11567209900747458,
"grad_norm": 0.5443837642669678,
"learning_rate": 9.980973490458728e-06,
"loss": 0.6596451997756958,
"step": 118
},
{
"epoch": 0.11665237103296165,
"grad_norm": 0.6172768473625183,
"learning_rate": 9.98027578700917e-06,
"loss": 0.7329565286636353,
"step": 119
},
{
"epoch": 0.11763264305844871,
"grad_norm": 0.4362824261188507,
"learning_rate": 9.979565545721249e-06,
"loss": 0.6226569414138794,
"step": 120
},
{
"epoch": 0.11861291508393579,
"grad_norm": 0.42918699979782104,
"learning_rate": 9.978842768382999e-06,
"loss": 0.646345853805542,
"step": 121
},
{
"epoch": 0.11959318710942286,
"grad_norm": 0.7146174907684326,
"learning_rate": 9.978107456814007e-06,
"loss": 0.643134593963623,
"step": 122
},
{
"epoch": 0.12057345913490994,
"grad_norm": 0.6748533844947815,
"learning_rate": 9.977359612865424e-06,
"loss": 0.6500820517539978,
"step": 123
},
{
"epoch": 0.12155373116039701,
"grad_norm": 0.5188791751861572,
"learning_rate": 9.976599238419943e-06,
"loss": 0.6458051204681396,
"step": 124
},
{
"epoch": 0.12253400318588409,
"grad_norm": 0.8062193393707275,
"learning_rate": 9.975826335391808e-06,
"loss": 0.9076640605926514,
"step": 125
},
{
"epoch": 0.12351427521137115,
"grad_norm": 0.6334277987480164,
"learning_rate": 9.975040905726799e-06,
"loss": 0.6395375728607178,
"step": 126
},
{
"epoch": 0.12449454723685822,
"grad_norm": 0.4327232539653778,
"learning_rate": 9.974242951402236e-06,
"loss": 0.605678379535675,
"step": 127
},
{
"epoch": 0.1254748192623453,
"grad_norm": 0.4278698265552521,
"learning_rate": 9.973432474426968e-06,
"loss": 0.6024398803710938,
"step": 128
},
{
"epoch": 0.12645509128783236,
"grad_norm": 0.5286264419555664,
"learning_rate": 9.972609476841368e-06,
"loss": 0.6663965582847595,
"step": 129
},
{
"epoch": 0.12743536331331945,
"grad_norm": 0.5116552710533142,
"learning_rate": 9.971773960717333e-06,
"loss": 0.5636756420135498,
"step": 130
},
{
"epoch": 0.1284156353388065,
"grad_norm": 0.4298281967639923,
"learning_rate": 9.970925928158275e-06,
"loss": 0.6357601881027222,
"step": 131
},
{
"epoch": 0.1293959073642936,
"grad_norm": 0.6595165133476257,
"learning_rate": 9.970065381299112e-06,
"loss": 0.569882869720459,
"step": 132
},
{
"epoch": 0.13037617938978066,
"grad_norm": 0.4833980202674866,
"learning_rate": 9.969192322306271e-06,
"loss": 0.6284748911857605,
"step": 133
},
{
"epoch": 0.13135645141526775,
"grad_norm": 0.42787042260169983,
"learning_rate": 9.968306753377679e-06,
"loss": 0.6308082342147827,
"step": 134
},
{
"epoch": 0.1323367234407548,
"grad_norm": 0.6076167225837708,
"learning_rate": 9.96740867674275e-06,
"loss": 0.8149276971817017,
"step": 135
},
{
"epoch": 0.13331699546624187,
"grad_norm": 0.4894298017024994,
"learning_rate": 9.966498094662398e-06,
"loss": 0.6580473184585571,
"step": 136
},
{
"epoch": 0.13429726749172896,
"grad_norm": 0.7970098853111267,
"learning_rate": 9.965575009429006e-06,
"loss": 0.5278803110122681,
"step": 137
},
{
"epoch": 0.13527753951721602,
"grad_norm": 1.672525405883789,
"learning_rate": 9.964639423366442e-06,
"loss": 0.5920721292495728,
"step": 138
},
{
"epoch": 0.1362578115427031,
"grad_norm": 2.432866334915161,
"learning_rate": 9.963691338830045e-06,
"loss": 0.6551089882850647,
"step": 139
},
{
"epoch": 0.13723808356819017,
"grad_norm": 0.5145550966262817,
"learning_rate": 9.962730758206612e-06,
"loss": 0.6491652727127075,
"step": 140
},
{
"epoch": 0.13821835559367723,
"grad_norm": 0.5592741370201111,
"learning_rate": 9.961757683914406e-06,
"loss": 0.6964321136474609,
"step": 141
},
{
"epoch": 0.13919862761916432,
"grad_norm": 0.44326746463775635,
"learning_rate": 9.960772118403142e-06,
"loss": 0.6217765808105469,
"step": 142
},
{
"epoch": 0.14017889964465138,
"grad_norm": 0.4019976556301117,
"learning_rate": 9.959774064153977e-06,
"loss": 0.5365424156188965,
"step": 143
},
{
"epoch": 0.14115917167013847,
"grad_norm": 0.3979960083961487,
"learning_rate": 9.958763523679515e-06,
"loss": 0.6748864650726318,
"step": 144
},
{
"epoch": 0.14213944369562553,
"grad_norm": 0.4052772521972656,
"learning_rate": 9.957740499523787e-06,
"loss": 0.6118680238723755,
"step": 145
},
{
"epoch": 0.14311971572111262,
"grad_norm": 0.43475887179374695,
"learning_rate": 9.956704994262256e-06,
"loss": 0.6804296970367432,
"step": 146
},
{
"epoch": 0.14409998774659968,
"grad_norm": 0.473827064037323,
"learning_rate": 9.955657010501807e-06,
"loss": 0.621467113494873,
"step": 147
},
{
"epoch": 0.14508025977208674,
"grad_norm": 0.42858877778053284,
"learning_rate": 9.954596550880735e-06,
"loss": 0.5835327506065369,
"step": 148
},
{
"epoch": 0.14606053179757383,
"grad_norm": 0.49841952323913574,
"learning_rate": 9.95352361806875e-06,
"loss": 0.48438161611557007,
"step": 149
},
{
"epoch": 0.1470408038230609,
"grad_norm": 0.4920412302017212,
"learning_rate": 9.952438214766956e-06,
"loss": 0.6087911128997803,
"step": 150
},
{
"epoch": 0.14802107584854798,
"grad_norm": 0.5047152638435364,
"learning_rate": 9.951340343707852e-06,
"loss": 0.7258588075637817,
"step": 151
},
{
"epoch": 0.14900134787403504,
"grad_norm": 0.5514736771583557,
"learning_rate": 9.950230007655333e-06,
"loss": 0.4761975407600403,
"step": 152
},
{
"epoch": 0.1499816198995221,
"grad_norm": 0.44626384973526,
"learning_rate": 9.949107209404664e-06,
"loss": 0.711269736289978,
"step": 153
},
{
"epoch": 0.1509618919250092,
"grad_norm": 0.4138593077659607,
"learning_rate": 9.94797195178249e-06,
"loss": 0.7103148102760315,
"step": 154
},
{
"epoch": 0.15194216395049626,
"grad_norm": 0.8979032635688782,
"learning_rate": 9.946824237646823e-06,
"loss": 0.7115532159805298,
"step": 155
},
{
"epoch": 0.15292243597598335,
"grad_norm": 0.46924495697021484,
"learning_rate": 9.945664069887028e-06,
"loss": 0.740920901298523,
"step": 156
},
{
"epoch": 0.1539027080014704,
"grad_norm": 0.3308998942375183,
"learning_rate": 9.944491451423829e-06,
"loss": 0.5225235223770142,
"step": 157
},
{
"epoch": 0.15488298002695747,
"grad_norm": 0.5165231823921204,
"learning_rate": 9.94330638520929e-06,
"loss": 0.6087205410003662,
"step": 158
},
{
"epoch": 0.15586325205244456,
"grad_norm": 0.4604957103729248,
"learning_rate": 9.942108874226812e-06,
"loss": 0.6101517081260681,
"step": 159
},
{
"epoch": 0.15684352407793162,
"grad_norm": 0.4937458038330078,
"learning_rate": 9.940898921491132e-06,
"loss": 0.5559808015823364,
"step": 160
},
{
"epoch": 0.1578237961034187,
"grad_norm": 0.5843384861946106,
"learning_rate": 9.9396765300483e-06,
"loss": 0.5209529399871826,
"step": 161
},
{
"epoch": 0.15880406812890577,
"grad_norm": 0.5297591686248779,
"learning_rate": 9.938441702975689e-06,
"loss": 0.5647348165512085,
"step": 162
},
{
"epoch": 0.15978434015439286,
"grad_norm": 0.9695852398872375,
"learning_rate": 9.937194443381972e-06,
"loss": 0.7081426978111267,
"step": 163
},
{
"epoch": 0.16076461217987992,
"grad_norm": 0.41647499799728394,
"learning_rate": 9.935934754407125e-06,
"loss": 0.5546735525131226,
"step": 164
},
{
"epoch": 0.16174488420536698,
"grad_norm": 0.41902264952659607,
"learning_rate": 9.934662639222412e-06,
"loss": 0.5788516402244568,
"step": 165
},
{
"epoch": 0.16272515623085407,
"grad_norm": 0.40173494815826416,
"learning_rate": 9.933378101030382e-06,
"loss": 0.5217040181159973,
"step": 166
},
{
"epoch": 0.16370542825634113,
"grad_norm": 0.43726590275764465,
"learning_rate": 9.93208114306486e-06,
"loss": 0.576999306678772,
"step": 167
},
{
"epoch": 0.16468570028182822,
"grad_norm": 0.4090188443660736,
"learning_rate": 9.930771768590934e-06,
"loss": 0.5982215404510498,
"step": 168
},
{
"epoch": 0.16566597230731528,
"grad_norm": 0.4458474814891815,
"learning_rate": 9.929449980904952e-06,
"loss": 0.5174573659896851,
"step": 169
},
{
"epoch": 0.16664624433280234,
"grad_norm": 0.47209373116493225,
"learning_rate": 9.928115783334518e-06,
"loss": 0.6529132127761841,
"step": 170
},
{
"epoch": 0.16762651635828943,
"grad_norm": 0.3903740644454956,
"learning_rate": 9.926769179238467e-06,
"loss": 0.5244958400726318,
"step": 171
},
{
"epoch": 0.1686067883837765,
"grad_norm": 1.6852936744689941,
"learning_rate": 9.925410172006874e-06,
"loss": 0.6307874917984009,
"step": 172
},
{
"epoch": 0.16958706040926358,
"grad_norm": 0.4870506227016449,
"learning_rate": 9.924038765061042e-06,
"loss": 0.6556846499443054,
"step": 173
},
{
"epoch": 0.17056733243475064,
"grad_norm": 0.5423577427864075,
"learning_rate": 9.922654961853483e-06,
"loss": 0.6876533031463623,
"step": 174
},
{
"epoch": 0.17154760446023773,
"grad_norm": 0.5223014950752258,
"learning_rate": 9.921258765867919e-06,
"loss": 0.6682021617889404,
"step": 175
},
{
"epoch": 0.1725278764857248,
"grad_norm": 0.5117551684379578,
"learning_rate": 9.919850180619274e-06,
"loss": 0.6182476282119751,
"step": 176
},
{
"epoch": 0.17350814851121185,
"grad_norm": 0.4597513675689697,
"learning_rate": 9.918429209653662e-06,
"loss": 0.577888548374176,
"step": 177
},
{
"epoch": 0.17448842053669894,
"grad_norm": 1.0743086338043213,
"learning_rate": 9.916995856548371e-06,
"loss": 0.6551996469497681,
"step": 178
},
{
"epoch": 0.175468692562186,
"grad_norm": 0.3808406889438629,
"learning_rate": 9.915550124911866e-06,
"loss": 0.6324294209480286,
"step": 179
},
{
"epoch": 0.1764489645876731,
"grad_norm": 0.3868914246559143,
"learning_rate": 9.914092018383778e-06,
"loss": 0.571523904800415,
"step": 180
},
{
"epoch": 0.17742923661316015,
"grad_norm": 0.4265619218349457,
"learning_rate": 9.912621540634889e-06,
"loss": 0.5399661064147949,
"step": 181
},
{
"epoch": 0.1784095086386472,
"grad_norm": 0.47953376173973083,
"learning_rate": 9.91113869536712e-06,
"loss": 0.4773590564727783,
"step": 182
},
{
"epoch": 0.1793897806641343,
"grad_norm": 0.4811166524887085,
"learning_rate": 9.909643486313533e-06,
"loss": 0.5630605220794678,
"step": 183
},
{
"epoch": 0.18037005268962136,
"grad_norm": 0.5990325808525085,
"learning_rate": 9.908135917238321e-06,
"loss": 0.6958386301994324,
"step": 184
},
{
"epoch": 0.18135032471510845,
"grad_norm": 0.7852551341056824,
"learning_rate": 9.906615991936781e-06,
"loss": 0.5374301671981812,
"step": 185
},
{
"epoch": 0.1823305967405955,
"grad_norm": 0.41726943850517273,
"learning_rate": 9.905083714235326e-06,
"loss": 0.5737833380699158,
"step": 186
},
{
"epoch": 0.18331086876608257,
"grad_norm": 0.4119226932525635,
"learning_rate": 9.903539087991462e-06,
"loss": 0.6594576835632324,
"step": 187
},
{
"epoch": 0.18429114079156966,
"grad_norm": 0.49085819721221924,
"learning_rate": 9.901982117093786e-06,
"loss": 0.6461527347564697,
"step": 188
},
{
"epoch": 0.18527141281705672,
"grad_norm": 0.40030795335769653,
"learning_rate": 9.900412805461968e-06,
"loss": 0.44109809398651123,
"step": 189
},
{
"epoch": 0.1862516848425438,
"grad_norm": 0.4172807037830353,
"learning_rate": 9.898831157046748e-06,
"loss": 0.4879729449748993,
"step": 190
},
{
"epoch": 0.18723195686803087,
"grad_norm": 0.4290069043636322,
"learning_rate": 9.897237175829927e-06,
"loss": 0.5784270167350769,
"step": 191
},
{
"epoch": 0.18821222889351796,
"grad_norm": 0.4286608397960663,
"learning_rate": 9.895630865824347e-06,
"loss": 0.5383234024047852,
"step": 192
},
{
"epoch": 0.18919250091900502,
"grad_norm": 0.4823291003704071,
"learning_rate": 9.894012231073895e-06,
"loss": 0.5863637924194336,
"step": 193
},
{
"epoch": 0.19017277294449209,
"grad_norm": 0.3424108624458313,
"learning_rate": 9.89238127565348e-06,
"loss": 0.5036578178405762,
"step": 194
},
{
"epoch": 0.19115304496997917,
"grad_norm": 1.4271470308303833,
"learning_rate": 9.890738003669029e-06,
"loss": 0.6060590744018555,
"step": 195
},
{
"epoch": 0.19213331699546624,
"grad_norm": 0.5405083298683167,
"learning_rate": 9.88908241925748e-06,
"loss": 0.5896493196487427,
"step": 196
},
{
"epoch": 0.19311358902095332,
"grad_norm": 0.45828577876091003,
"learning_rate": 9.887414526586764e-06,
"loss": 0.6141184568405151,
"step": 197
},
{
"epoch": 0.19409386104644039,
"grad_norm": 0.4599784016609192,
"learning_rate": 9.885734329855798e-06,
"loss": 0.5515347719192505,
"step": 198
},
{
"epoch": 0.19507413307192745,
"grad_norm": 0.4185781478881836,
"learning_rate": 9.884041833294477e-06,
"loss": 0.5514547228813171,
"step": 199
},
{
"epoch": 0.19605440509741454,
"grad_norm": 0.647860586643219,
"learning_rate": 9.882337041163656e-06,
"loss": 0.4992048144340515,
"step": 200
},
{
"epoch": 0.1970346771229016,
"grad_norm": 0.4117739200592041,
"learning_rate": 9.880619957755151e-06,
"loss": 0.6063427925109863,
"step": 201
},
{
"epoch": 0.19801494914838869,
"grad_norm": 0.6844010353088379,
"learning_rate": 9.878890587391715e-06,
"loss": 0.586762547492981,
"step": 202
},
{
"epoch": 0.19899522117387575,
"grad_norm": 0.6205182671546936,
"learning_rate": 9.877148934427037e-06,
"loss": 0.4610150158405304,
"step": 203
},
{
"epoch": 0.19997549319936284,
"grad_norm": 0.8094073534011841,
"learning_rate": 9.875395003245724e-06,
"loss": 0.49249520897865295,
"step": 204
},
{
"epoch": 0.2009557652248499,
"grad_norm": 0.739032506942749,
"learning_rate": 9.873628798263297e-06,
"loss": 0.6425831913948059,
"step": 205
},
{
"epoch": 0.20193603725033696,
"grad_norm": 0.4901730418205261,
"learning_rate": 9.871850323926178e-06,
"loss": 0.5576741695404053,
"step": 206
},
{
"epoch": 0.20291630927582405,
"grad_norm": 0.4465477168560028,
"learning_rate": 9.870059584711668e-06,
"loss": 0.526231050491333,
"step": 207
},
{
"epoch": 0.2038965813013111,
"grad_norm": 0.5906025171279907,
"learning_rate": 9.868256585127956e-06,
"loss": 0.5312085151672363,
"step": 208
},
{
"epoch": 0.2048768533267982,
"grad_norm": 0.7137251496315002,
"learning_rate": 9.86644132971409e-06,
"loss": 0.5133568048477173,
"step": 209
},
{
"epoch": 0.20585712535228526,
"grad_norm": 0.4403294324874878,
"learning_rate": 9.864613823039971e-06,
"loss": 0.5519824028015137,
"step": 210
},
{
"epoch": 0.20683739737777232,
"grad_norm": 0.5797751545906067,
"learning_rate": 9.862774069706346e-06,
"loss": 0.6732640266418457,
"step": 211
},
{
"epoch": 0.2078176694032594,
"grad_norm": 0.5637887120246887,
"learning_rate": 9.860922074344794e-06,
"loss": 0.5482189655303955,
"step": 212
},
{
"epoch": 0.20879794142874647,
"grad_norm": 0.41663646697998047,
"learning_rate": 9.859057841617709e-06,
"loss": 0.4765334725379944,
"step": 213
},
{
"epoch": 0.20977821345423356,
"grad_norm": 0.47632116079330444,
"learning_rate": 9.857181376218295e-06,
"loss": 0.5404700040817261,
"step": 214
},
{
"epoch": 0.21075848547972062,
"grad_norm": 0.3458189070224762,
"learning_rate": 9.855292682870552e-06,
"loss": 0.5178496837615967,
"step": 215
},
{
"epoch": 0.2117387575052077,
"grad_norm": 0.5893949270248413,
"learning_rate": 9.853391766329264e-06,
"loss": 0.5478975772857666,
"step": 216
},
{
"epoch": 0.21271902953069477,
"grad_norm": 0.4302483797073364,
"learning_rate": 9.851478631379982e-06,
"loss": 0.5323036909103394,
"step": 217
},
{
"epoch": 0.21369930155618183,
"grad_norm": 0.3951965570449829,
"learning_rate": 9.849553282839025e-06,
"loss": 0.5403459072113037,
"step": 218
},
{
"epoch": 0.21467957358166892,
"grad_norm": 0.5673352479934692,
"learning_rate": 9.847615725553457e-06,
"loss": 0.526860773563385,
"step": 219
},
{
"epoch": 0.21565984560715598,
"grad_norm": 0.39107856154441833,
"learning_rate": 9.845665964401072e-06,
"loss": 0.5004710555076599,
"step": 220
},
{
"epoch": 0.21664011763264307,
"grad_norm": 0.4703252613544464,
"learning_rate": 9.843704004290393e-06,
"loss": 0.5515217185020447,
"step": 221
},
{
"epoch": 0.21762038965813013,
"grad_norm": 0.3669305145740509,
"learning_rate": 9.841729850160652e-06,
"loss": 0.5492053031921387,
"step": 222
},
{
"epoch": 0.2186006616836172,
"grad_norm": 0.719444990158081,
"learning_rate": 9.839743506981783e-06,
"loss": 0.5406309366226196,
"step": 223
},
{
"epoch": 0.21958093370910428,
"grad_norm": 0.411018043756485,
"learning_rate": 9.837744979754397e-06,
"loss": 0.4943702816963196,
"step": 224
},
{
"epoch": 0.22056120573459134,
"grad_norm": 0.4032668471336365,
"learning_rate": 9.835734273509787e-06,
"loss": 0.5331560969352722,
"step": 225
},
{
"epoch": 0.22154147776007843,
"grad_norm": 0.4198213219642639,
"learning_rate": 9.833711393309902e-06,
"loss": 0.6462162733078003,
"step": 226
},
{
"epoch": 0.2225217497855655,
"grad_norm": 0.5204048156738281,
"learning_rate": 9.831676344247343e-06,
"loss": 0.5799481272697449,
"step": 227
},
{
"epoch": 0.22350202181105255,
"grad_norm": 0.38446101546287537,
"learning_rate": 9.829629131445342e-06,
"loss": 0.49075692892074585,
"step": 228
},
{
"epoch": 0.22448229383653964,
"grad_norm": 0.43012040853500366,
"learning_rate": 9.827569760057755e-06,
"loss": 0.41364920139312744,
"step": 229
},
{
"epoch": 0.2254625658620267,
"grad_norm": 0.5348578691482544,
"learning_rate": 9.825498235269046e-06,
"loss": 0.5324258208274841,
"step": 230
},
{
"epoch": 0.2264428378875138,
"grad_norm": 0.5088627338409424,
"learning_rate": 9.82341456229428e-06,
"loss": 0.5315623879432678,
"step": 231
},
{
"epoch": 0.22742310991300085,
"grad_norm": 0.6464555859565735,
"learning_rate": 9.821318746379097e-06,
"loss": 0.6046267747879028,
"step": 232
},
{
"epoch": 0.22840338193848794,
"grad_norm": 0.6463635563850403,
"learning_rate": 9.819210792799711e-06,
"loss": 0.5082406997680664,
"step": 233
},
{
"epoch": 0.229383653963975,
"grad_norm": 0.36080867052078247,
"learning_rate": 9.817090706862895e-06,
"loss": 0.47844094038009644,
"step": 234
},
{
"epoch": 0.23036392598946207,
"grad_norm": 1.120613932609558,
"learning_rate": 9.814958493905962e-06,
"loss": 0.6177240014076233,
"step": 235
},
{
"epoch": 0.23134419801494915,
"grad_norm": 0.5548020005226135,
"learning_rate": 9.812814159296757e-06,
"loss": 0.5502880811691284,
"step": 236
},
{
"epoch": 0.23232447004043622,
"grad_norm": 4.807314872741699,
"learning_rate": 9.810657708433637e-06,
"loss": 0.62823885679245,
"step": 237
},
{
"epoch": 0.2333047420659233,
"grad_norm": 0.4371277093887329,
"learning_rate": 9.808489146745466e-06,
"loss": 0.5365610718727112,
"step": 238
},
{
"epoch": 0.23428501409141037,
"grad_norm": 0.36411231756210327,
"learning_rate": 9.806308479691595e-06,
"loss": 0.5304818153381348,
"step": 239
},
{
"epoch": 0.23526528611689743,
"grad_norm": 0.5078960061073303,
"learning_rate": 9.804115712761851e-06,
"loss": 0.7285908460617065,
"step": 240
},
{
"epoch": 0.23624555814238452,
"grad_norm": 0.546269416809082,
"learning_rate": 9.801910851476524e-06,
"loss": 0.6330514550209045,
"step": 241
},
{
"epoch": 0.23722583016787158,
"grad_norm": 0.4550075829029083,
"learning_rate": 9.799693901386346e-06,
"loss": 0.6546021699905396,
"step": 242
},
{
"epoch": 0.23820610219335867,
"grad_norm": 0.3764090836048126,
"learning_rate": 9.797464868072489e-06,
"loss": 0.48709845542907715,
"step": 243
},
{
"epoch": 0.23918637421884573,
"grad_norm": 0.6424336433410645,
"learning_rate": 9.79522375714654e-06,
"loss": 0.5066192150115967,
"step": 244
},
{
"epoch": 0.24016664624433282,
"grad_norm": 0.5178307890892029,
"learning_rate": 9.792970574250493e-06,
"loss": 0.5911586880683899,
"step": 245
},
{
"epoch": 0.24114691826981988,
"grad_norm": 0.378834068775177,
"learning_rate": 9.790705325056735e-06,
"loss": 0.5345185399055481,
"step": 246
},
{
"epoch": 0.24212719029530694,
"grad_norm": 0.36339879035949707,
"learning_rate": 9.788428015268027e-06,
"loss": 0.5043888092041016,
"step": 247
},
{
"epoch": 0.24310746232079403,
"grad_norm": 0.6303738355636597,
"learning_rate": 9.786138650617494e-06,
"loss": 0.6438804864883423,
"step": 248
},
{
"epoch": 0.2440877343462811,
"grad_norm": 0.31280583143234253,
"learning_rate": 9.78383723686861e-06,
"loss": 0.4477033019065857,
"step": 249
},
{
"epoch": 0.24506800637176818,
"grad_norm": 0.458517849445343,
"learning_rate": 9.781523779815178e-06,
"loss": 0.47877079248428345,
"step": 250
},
{
"epoch": 0.24604827839725524,
"grad_norm": 0.34510016441345215,
"learning_rate": 9.779198285281326e-06,
"loss": 0.5180049538612366,
"step": 251
},
{
"epoch": 0.2470285504227423,
"grad_norm": 0.4814448356628418,
"learning_rate": 9.776860759121485e-06,
"loss": 0.590415358543396,
"step": 252
},
{
"epoch": 0.2480088224482294,
"grad_norm": 0.4635251462459564,
"learning_rate": 9.774511207220369e-06,
"loss": 0.5447108745574951,
"step": 253
},
{
"epoch": 0.24898909447371645,
"grad_norm": 0.6083935499191284,
"learning_rate": 9.772149635492979e-06,
"loss": 0.5176642537117004,
"step": 254
},
{
"epoch": 0.24996936649920354,
"grad_norm": 0.45982351899147034,
"learning_rate": 9.769776049884564e-06,
"loss": 0.514198899269104,
"step": 255
},
{
"epoch": 0.2509496385246906,
"grad_norm": 2.298872709274292,
"learning_rate": 9.767390456370625e-06,
"loss": 0.8669224381446838,
"step": 256
},
{
"epoch": 0.25192991055017766,
"grad_norm": 0.6136132478713989,
"learning_rate": 9.76499286095689e-06,
"loss": 0.552510142326355,
"step": 257
},
{
"epoch": 0.2529101825756647,
"grad_norm": 0.48572230339050293,
"learning_rate": 9.762583269679304e-06,
"loss": 0.47806376218795776,
"step": 258
},
{
"epoch": 0.25389045460115184,
"grad_norm": 0.4945577383041382,
"learning_rate": 9.760161688604008e-06,
"loss": 0.5923348665237427,
"step": 259
},
{
"epoch": 0.2548707266266389,
"grad_norm": 0.44879934191703796,
"learning_rate": 9.75772812382733e-06,
"loss": 0.475081205368042,
"step": 260
},
{
"epoch": 0.25585099865212596,
"grad_norm": 0.5548574328422546,
"learning_rate": 9.755282581475769e-06,
"loss": 0.5204034447669983,
"step": 261
},
{
"epoch": 0.256831270677613,
"grad_norm": 0.38964948058128357,
"learning_rate": 9.75282506770597e-06,
"loss": 0.4777343273162842,
"step": 262
},
{
"epoch": 0.2578115427031001,
"grad_norm": 0.36434146761894226,
"learning_rate": 9.750355588704728e-06,
"loss": 0.4324755370616913,
"step": 263
},
{
"epoch": 0.2587918147285872,
"grad_norm": 0.5199502110481262,
"learning_rate": 9.747874150688948e-06,
"loss": 0.5141733884811401,
"step": 264
},
{
"epoch": 0.25977208675407426,
"grad_norm": 1.9240013360977173,
"learning_rate": 9.745380759905648e-06,
"loss": 0.5489879250526428,
"step": 265
},
{
"epoch": 0.2607523587795613,
"grad_norm": 0.43477994203567505,
"learning_rate": 9.742875422631937e-06,
"loss": 0.4643503427505493,
"step": 266
},
{
"epoch": 0.2617326308050484,
"grad_norm": 0.4621925950050354,
"learning_rate": 9.740358145174999e-06,
"loss": 0.5075932145118713,
"step": 267
},
{
"epoch": 0.2627129028305355,
"grad_norm": 0.3986125588417053,
"learning_rate": 9.737828933872076e-06,
"loss": 0.5133166313171387,
"step": 268
},
{
"epoch": 0.26369317485602256,
"grad_norm": 0.3705284893512726,
"learning_rate": 9.735287795090455e-06,
"loss": 0.5779547691345215,
"step": 269
},
{
"epoch": 0.2646734468815096,
"grad_norm": 0.5568776726722717,
"learning_rate": 9.73273473522745e-06,
"loss": 0.47999048233032227,
"step": 270
},
{
"epoch": 0.2656537189069967,
"grad_norm": 0.7552224397659302,
"learning_rate": 9.730169760710385e-06,
"loss": 0.5507869124412537,
"step": 271
},
{
"epoch": 0.26663399093248374,
"grad_norm": 0.37253665924072266,
"learning_rate": 9.727592877996585e-06,
"loss": 0.5092483758926392,
"step": 272
},
{
"epoch": 0.26761426295797086,
"grad_norm": 0.40661561489105225,
"learning_rate": 9.725004093573343e-06,
"loss": 0.5572633743286133,
"step": 273
},
{
"epoch": 0.2685945349834579,
"grad_norm": 0.49091172218322754,
"learning_rate": 9.722403413957924e-06,
"loss": 0.48622071743011475,
"step": 274
},
{
"epoch": 0.269574807008945,
"grad_norm": 0.5230619311332703,
"learning_rate": 9.719790845697534e-06,
"loss": 0.5598489046096802,
"step": 275
},
{
"epoch": 0.27055507903443204,
"grad_norm": 0.39431917667388916,
"learning_rate": 9.717166395369312e-06,
"loss": 0.5124005675315857,
"step": 276
},
{
"epoch": 0.2715353510599191,
"grad_norm": 0.41142791509628296,
"learning_rate": 9.71453006958031e-06,
"loss": 0.48253190517425537,
"step": 277
},
{
"epoch": 0.2725156230854062,
"grad_norm": 0.3350473642349243,
"learning_rate": 9.711881874967471e-06,
"loss": 0.3939792513847351,
"step": 278
},
{
"epoch": 0.2734958951108933,
"grad_norm": 0.38885241746902466,
"learning_rate": 9.709221818197626e-06,
"loss": 0.539023220539093,
"step": 279
},
{
"epoch": 0.27447616713638034,
"grad_norm": 0.6647607684135437,
"learning_rate": 9.70654990596746e-06,
"loss": 0.6396561861038208,
"step": 280
},
{
"epoch": 0.2754564391618674,
"grad_norm": 0.43505945801734924,
"learning_rate": 9.703866145003512e-06,
"loss": 0.5345339179039001,
"step": 281
},
{
"epoch": 0.27643671118735447,
"grad_norm": 0.6203579306602478,
"learning_rate": 9.701170542062149e-06,
"loss": 0.47985559701919556,
"step": 282
},
{
"epoch": 0.2774169832128416,
"grad_norm": 0.484285831451416,
"learning_rate": 9.698463103929542e-06,
"loss": 0.5870672464370728,
"step": 283
},
{
"epoch": 0.27839725523832864,
"grad_norm": 0.3385967016220093,
"learning_rate": 9.69574383742167e-06,
"loss": 0.4503611624240875,
"step": 284
},
{
"epoch": 0.2793775272638157,
"grad_norm": 0.41106754541397095,
"learning_rate": 9.69301274938428e-06,
"loss": 0.5788821578025818,
"step": 285
},
{
"epoch": 0.28035779928930277,
"grad_norm": 0.3710075616836548,
"learning_rate": 9.690269846692882e-06,
"loss": 0.5564764738082886,
"step": 286
},
{
"epoch": 0.28133807131478983,
"grad_norm": 0.4804702699184418,
"learning_rate": 9.687515136252732e-06,
"loss": 0.5230247974395752,
"step": 287
},
{
"epoch": 0.28231834334027694,
"grad_norm": 0.36611407995224,
"learning_rate": 9.68474862499881e-06,
"loss": 0.5235856175422668,
"step": 288
},
{
"epoch": 0.283298615365764,
"grad_norm": 0.42279067635536194,
"learning_rate": 9.681970319895804e-06,
"loss": 0.5690290927886963,
"step": 289
},
{
"epoch": 0.28427888739125107,
"grad_norm": 0.4615001678466797,
"learning_rate": 9.679180227938095e-06,
"loss": 0.5438805818557739,
"step": 290
},
{
"epoch": 0.28525915941673813,
"grad_norm": 0.4851062595844269,
"learning_rate": 9.676378356149733e-06,
"loss": 0.5331500768661499,
"step": 291
},
{
"epoch": 0.28623943144222525,
"grad_norm": 0.5172120928764343,
"learning_rate": 9.67356471158443e-06,
"loss": 0.4695644974708557,
"step": 292
},
{
"epoch": 0.2872197034677123,
"grad_norm": 0.49664053320884705,
"learning_rate": 9.670739301325534e-06,
"loss": 0.5288171768188477,
"step": 293
},
{
"epoch": 0.28819997549319937,
"grad_norm": 0.47372597455978394,
"learning_rate": 9.667902132486009e-06,
"loss": 0.6496613025665283,
"step": 294
},
{
"epoch": 0.28918024751868643,
"grad_norm": 0.35643258690834045,
"learning_rate": 9.665053212208426e-06,
"loss": 0.4534105658531189,
"step": 295
},
{
"epoch": 0.2901605195441735,
"grad_norm": 0.3932669758796692,
"learning_rate": 9.662192547664938e-06,
"loss": 0.5261329412460327,
"step": 296
},
{
"epoch": 0.2911407915696606,
"grad_norm": 1.0437190532684326,
"learning_rate": 9.659320146057263e-06,
"loss": 0.5095226764678955,
"step": 297
},
{
"epoch": 0.29212106359514767,
"grad_norm": 0.4327160716056824,
"learning_rate": 9.65643601461667e-06,
"loss": 0.42723172903060913,
"step": 298
},
{
"epoch": 0.29310133562063473,
"grad_norm": 0.7766056060791016,
"learning_rate": 9.653540160603956e-06,
"loss": 0.5549726486206055,
"step": 299
},
{
"epoch": 0.2940816076461218,
"grad_norm": 0.44525232911109924,
"learning_rate": 9.650632591309431e-06,
"loss": 0.5621340274810791,
"step": 300
},
{
"epoch": 0.29506187967160885,
"grad_norm": 0.7762605547904968,
"learning_rate": 9.647713314052896e-06,
"loss": 0.48529478907585144,
"step": 301
},
{
"epoch": 0.29604215169709597,
"grad_norm": 0.6724576950073242,
"learning_rate": 9.64478233618363e-06,
"loss": 0.5356699228286743,
"step": 302
},
{
"epoch": 0.29702242372258303,
"grad_norm": 0.41225090622901917,
"learning_rate": 9.641839665080363e-06,
"loss": 0.5188729763031006,
"step": 303
},
{
"epoch": 0.2980026957480701,
"grad_norm": 1.4071273803710938,
"learning_rate": 9.63888530815127e-06,
"loss": 0.42046862840652466,
"step": 304
},
{
"epoch": 0.29898296777355715,
"grad_norm": 0.39619553089141846,
"learning_rate": 9.635919272833938e-06,
"loss": 0.4914689064025879,
"step": 305
},
{
"epoch": 0.2999632397990442,
"grad_norm": 0.4301491975784302,
"learning_rate": 9.632941566595357e-06,
"loss": 0.5850746631622314,
"step": 306
},
{
"epoch": 0.30094351182453133,
"grad_norm": 0.35922926664352417,
"learning_rate": 9.629952196931902e-06,
"loss": 0.46321457624435425,
"step": 307
},
{
"epoch": 0.3019237838500184,
"grad_norm": 0.687816321849823,
"learning_rate": 9.626951171369306e-06,
"loss": 0.5080117583274841,
"step": 308
},
{
"epoch": 0.30290405587550545,
"grad_norm": 0.7874114513397217,
"learning_rate": 9.623938497462647e-06,
"loss": 0.7459219694137573,
"step": 309
},
{
"epoch": 0.3038843279009925,
"grad_norm": 0.40115970373153687,
"learning_rate": 9.620914182796327e-06,
"loss": 0.40731143951416016,
"step": 310
},
{
"epoch": 0.3048645999264796,
"grad_norm": 0.5670820474624634,
"learning_rate": 9.617878234984056e-06,
"loss": 0.4911518394947052,
"step": 311
},
{
"epoch": 0.3058448719519667,
"grad_norm": 0.406183660030365,
"learning_rate": 9.61483066166883e-06,
"loss": 0.485312283039093,
"step": 312
},
{
"epoch": 0.30682514397745375,
"grad_norm": 0.5942997932434082,
"learning_rate": 9.611771470522908e-06,
"loss": 0.5256763696670532,
"step": 313
},
{
"epoch": 0.3078054160029408,
"grad_norm": 0.4682798683643341,
"learning_rate": 9.608700669247803e-06,
"loss": 0.4943842589855194,
"step": 314
},
{
"epoch": 0.3087856880284279,
"grad_norm": 0.4411674439907074,
"learning_rate": 9.60561826557425e-06,
"loss": 0.4751124978065491,
"step": 315
},
{
"epoch": 0.30976596005391493,
"grad_norm": 0.42140281200408936,
"learning_rate": 9.602524267262202e-06,
"loss": 0.48478978872299194,
"step": 316
},
{
"epoch": 0.31074623207940205,
"grad_norm": 0.9000359773635864,
"learning_rate": 9.599418682100793e-06,
"loss": 0.49367135763168335,
"step": 317
},
{
"epoch": 0.3117265041048891,
"grad_norm": 0.44587165117263794,
"learning_rate": 9.596301517908329e-06,
"loss": 0.4780866503715515,
"step": 318
},
{
"epoch": 0.3127067761303762,
"grad_norm": 0.9870256781578064,
"learning_rate": 9.59317278253227e-06,
"loss": 0.5383126735687256,
"step": 319
},
{
"epoch": 0.31368704815586324,
"grad_norm": 0.3557277321815491,
"learning_rate": 9.5900324838492e-06,
"loss": 0.5207388401031494,
"step": 320
},
{
"epoch": 0.31466732018135035,
"grad_norm": 0.4181939959526062,
"learning_rate": 9.586880629764817e-06,
"loss": 0.4727376699447632,
"step": 321
},
{
"epoch": 0.3156475922068374,
"grad_norm": 0.42442455887794495,
"learning_rate": 9.583717228213914e-06,
"loss": 0.4959885776042938,
"step": 322
},
{
"epoch": 0.3166278642323245,
"grad_norm": 0.5203853845596313,
"learning_rate": 9.580542287160348e-06,
"loss": 0.5454490184783936,
"step": 323
},
{
"epoch": 0.31760813625781154,
"grad_norm": 0.46688640117645264,
"learning_rate": 9.577355814597031e-06,
"loss": 0.5831551551818848,
"step": 324
},
{
"epoch": 0.3185884082832986,
"grad_norm": 0.4086481034755707,
"learning_rate": 9.574157818545902e-06,
"loss": 0.48657554388046265,
"step": 325
},
{
"epoch": 0.3195686803087857,
"grad_norm": 0.443040668964386,
"learning_rate": 9.570948307057912e-06,
"loss": 0.5815020799636841,
"step": 326
},
{
"epoch": 0.3205489523342728,
"grad_norm": 0.4315493702888489,
"learning_rate": 9.567727288213005e-06,
"loss": 0.4193767309188843,
"step": 327
},
{
"epoch": 0.32152922435975984,
"grad_norm": 0.4256093204021454,
"learning_rate": 9.564494770120089e-06,
"loss": 0.4679313004016876,
"step": 328
},
{
"epoch": 0.3225094963852469,
"grad_norm": 0.510225236415863,
"learning_rate": 9.561250760917026e-06,
"loss": 0.5247933268547058,
"step": 329
},
{
"epoch": 0.32348976841073396,
"grad_norm": 0.72126305103302,
"learning_rate": 9.557995268770609e-06,
"loss": 0.5531257390975952,
"step": 330
},
{
"epoch": 0.3244700404362211,
"grad_norm": 0.3419378399848938,
"learning_rate": 9.554728301876525e-06,
"loss": 0.4606001377105713,
"step": 331
},
{
"epoch": 0.32545031246170814,
"grad_norm": 0.38200464844703674,
"learning_rate": 9.551449868459371e-06,
"loss": 0.5388354659080505,
"step": 332
},
{
"epoch": 0.3264305844871952,
"grad_norm": 0.5292008519172668,
"learning_rate": 9.548159976772593e-06,
"loss": 0.5080040097236633,
"step": 333
},
{
"epoch": 0.32741085651268226,
"grad_norm": 0.4875475764274597,
"learning_rate": 9.544858635098491e-06,
"loss": 0.45527976751327515,
"step": 334
},
{
"epoch": 0.3283911285381693,
"grad_norm": 0.4285490810871124,
"learning_rate": 9.541545851748186e-06,
"loss": 0.5484116673469543,
"step": 335
},
{
"epoch": 0.32937140056365644,
"grad_norm": 0.44281134009361267,
"learning_rate": 9.538221635061611e-06,
"loss": 0.4413299560546875,
"step": 336
},
{
"epoch": 0.3303516725891435,
"grad_norm": 0.40900593996047974,
"learning_rate": 9.534885993407474e-06,
"loss": 0.5252339839935303,
"step": 337
},
{
"epoch": 0.33133194461463056,
"grad_norm": 0.42202669382095337,
"learning_rate": 9.531538935183252e-06,
"loss": 0.524871826171875,
"step": 338
},
{
"epoch": 0.3323122166401176,
"grad_norm": 0.4463450610637665,
"learning_rate": 9.528180468815155e-06,
"loss": 0.5509671568870544,
"step": 339
},
{
"epoch": 0.3332924886656047,
"grad_norm": 0.38109633326530457,
"learning_rate": 9.524810602758122e-06,
"loss": 0.43840712308883667,
"step": 340
},
{
"epoch": 0.3342727606910918,
"grad_norm": 0.37860217690467834,
"learning_rate": 9.521429345495787e-06,
"loss": 0.4627377390861511,
"step": 341
},
{
"epoch": 0.33525303271657886,
"grad_norm": 1.67256760597229,
"learning_rate": 9.518036705540459e-06,
"loss": 0.6440774202346802,
"step": 342
},
{
"epoch": 0.3362333047420659,
"grad_norm": 0.42276817560195923,
"learning_rate": 9.514632691433108e-06,
"loss": 0.5407246351242065,
"step": 343
},
{
"epoch": 0.337213576767553,
"grad_norm": 0.3594137728214264,
"learning_rate": 9.51121731174333e-06,
"loss": 0.502845287322998,
"step": 344
},
{
"epoch": 0.33819384879304004,
"grad_norm": 0.6008448600769043,
"learning_rate": 9.507790575069347e-06,
"loss": 0.5553073883056641,
"step": 345
},
{
"epoch": 0.33917412081852716,
"grad_norm": 0.3711321949958801,
"learning_rate": 9.504352490037957e-06,
"loss": 0.44527333974838257,
"step": 346
},
{
"epoch": 0.3401543928440142,
"grad_norm": 0.46572935581207275,
"learning_rate": 9.50090306530454e-06,
"loss": 0.5924540758132935,
"step": 347
},
{
"epoch": 0.3411346648695013,
"grad_norm": 0.7329858541488647,
"learning_rate": 9.497442309553017e-06,
"loss": 0.46114006638526917,
"step": 348
},
{
"epoch": 0.34211493689498834,
"grad_norm": 0.44977322220802307,
"learning_rate": 9.493970231495836e-06,
"loss": 0.6232258677482605,
"step": 349
},
{
"epoch": 0.34309520892047546,
"grad_norm": 0.4757539629936218,
"learning_rate": 9.490486839873949e-06,
"loss": 0.57889324426651,
"step": 350
},
{
"epoch": 0.3440754809459625,
"grad_norm": 0.31424209475517273,
"learning_rate": 9.486992143456792e-06,
"loss": 0.4667109549045563,
"step": 351
},
{
"epoch": 0.3450557529714496,
"grad_norm": 0.3532857894897461,
"learning_rate": 9.483486151042258e-06,
"loss": 0.46391308307647705,
"step": 352
},
{
"epoch": 0.34603602499693664,
"grad_norm": 0.34860628843307495,
"learning_rate": 9.47996887145668e-06,
"loss": 0.4222821295261383,
"step": 353
},
{
"epoch": 0.3470162970224237,
"grad_norm": 0.39144259691238403,
"learning_rate": 9.476440313554803e-06,
"loss": 0.5101709365844727,
"step": 354
},
{
"epoch": 0.3479965690479108,
"grad_norm": 0.5720930099487305,
"learning_rate": 9.47290048621977e-06,
"loss": 0.4800007939338684,
"step": 355
},
{
"epoch": 0.3489768410733979,
"grad_norm": 1.3871150016784668,
"learning_rate": 9.469349398363088e-06,
"loss": 0.4612959027290344,
"step": 356
},
{
"epoch": 0.34995711309888494,
"grad_norm": 0.623758852481842,
"learning_rate": 9.46578705892462e-06,
"loss": 0.5273652672767639,
"step": 357
},
{
"epoch": 0.350937385124372,
"grad_norm": 0.46722686290740967,
"learning_rate": 9.46221347687255e-06,
"loss": 0.5567481517791748,
"step": 358
},
{
"epoch": 0.35191765714985906,
"grad_norm": 0.37911084294319153,
"learning_rate": 9.458628661203368e-06,
"loss": 0.46729549765586853,
"step": 359
},
{
"epoch": 0.3528979291753462,
"grad_norm": 0.44079819321632385,
"learning_rate": 9.45503262094184e-06,
"loss": 0.5010742545127869,
"step": 360
},
{
"epoch": 0.35387820120083324,
"grad_norm": 0.3892115652561188,
"learning_rate": 9.451425365140997e-06,
"loss": 0.5010262727737427,
"step": 361
},
{
"epoch": 0.3548584732263203,
"grad_norm": 0.2773195505142212,
"learning_rate": 9.447806902882098e-06,
"loss": 0.43856537342071533,
"step": 362
},
{
"epoch": 0.35583874525180736,
"grad_norm": 0.5443810224533081,
"learning_rate": 9.444177243274619e-06,
"loss": 0.4659033417701721,
"step": 363
},
{
"epoch": 0.3568190172772944,
"grad_norm": 0.4063398540019989,
"learning_rate": 9.440536395456223e-06,
"loss": 0.5410536527633667,
"step": 364
},
{
"epoch": 0.35779928930278154,
"grad_norm": 0.38613906502723694,
"learning_rate": 9.43688436859274e-06,
"loss": 0.5363536477088928,
"step": 365
},
{
"epoch": 0.3587795613282686,
"grad_norm": 0.36793196201324463,
"learning_rate": 9.433221171878144e-06,
"loss": 0.433261513710022,
"step": 366
},
{
"epoch": 0.35975983335375566,
"grad_norm": 0.423566997051239,
"learning_rate": 9.429546814534528e-06,
"loss": 0.5269312262535095,
"step": 367
},
{
"epoch": 0.3607401053792427,
"grad_norm": 1.0134918689727783,
"learning_rate": 9.425861305812083e-06,
"loss": 0.6337320804595947,
"step": 368
},
{
"epoch": 0.3617203774047298,
"grad_norm": 0.35793018341064453,
"learning_rate": 9.422164654989073e-06,
"loss": 0.49202513694763184,
"step": 369
},
{
"epoch": 0.3627006494302169,
"grad_norm": 0.35067111253738403,
"learning_rate": 9.41845687137181e-06,
"loss": 0.5590043067932129,
"step": 370
},
{
"epoch": 0.36368092145570396,
"grad_norm": 0.6976906657218933,
"learning_rate": 9.414737964294636e-06,
"loss": 0.44993770122528076,
"step": 371
},
{
"epoch": 0.364661193481191,
"grad_norm": 0.5873794555664062,
"learning_rate": 9.411007943119895e-06,
"loss": 0.7101871371269226,
"step": 372
},
{
"epoch": 0.3656414655066781,
"grad_norm": 0.3841715157032013,
"learning_rate": 9.40726681723791e-06,
"loss": 0.5202108025550842,
"step": 373
},
{
"epoch": 0.36662173753216515,
"grad_norm": 0.35475191473960876,
"learning_rate": 9.403514596066964e-06,
"loss": 0.482761949300766,
"step": 374
},
{
"epoch": 0.36760200955765227,
"grad_norm": 0.3288975954055786,
"learning_rate": 9.399751289053267e-06,
"loss": 0.48916640877723694,
"step": 375
},
{
"epoch": 0.3685822815831393,
"grad_norm": 0.30811378359794617,
"learning_rate": 9.39597690567094e-06,
"loss": 0.40215981006622314,
"step": 376
},
{
"epoch": 0.3695625536086264,
"grad_norm": 0.40600016713142395,
"learning_rate": 9.392191455421989e-06,
"loss": 0.4360930025577545,
"step": 377
},
{
"epoch": 0.37054282563411345,
"grad_norm": 0.37359315156936646,
"learning_rate": 9.388394947836278e-06,
"loss": 0.4261525869369507,
"step": 378
},
{
"epoch": 0.37152309765960057,
"grad_norm": 0.375644326210022,
"learning_rate": 9.384587392471516e-06,
"loss": 0.46239417791366577,
"step": 379
},
{
"epoch": 0.3725033696850876,
"grad_norm": 0.33713436126708984,
"learning_rate": 9.380768798913213e-06,
"loss": 0.4638916552066803,
"step": 380
},
{
"epoch": 0.3734836417105747,
"grad_norm": 0.4940984547138214,
"learning_rate": 9.376939176774678e-06,
"loss": 0.46083444356918335,
"step": 381
},
{
"epoch": 0.37446391373606175,
"grad_norm": 0.3293454945087433,
"learning_rate": 9.37309853569698e-06,
"loss": 0.4653550386428833,
"step": 382
},
{
"epoch": 0.3754441857615488,
"grad_norm": 0.463020384311676,
"learning_rate": 9.369246885348926e-06,
"loss": 0.5448740124702454,
"step": 383
},
{
"epoch": 0.3764244577870359,
"grad_norm": 0.3816799819469452,
"learning_rate": 9.365384235427042e-06,
"loss": 0.4663521647453308,
"step": 384
},
{
"epoch": 0.377404729812523,
"grad_norm": 0.5458483099937439,
"learning_rate": 9.361510595655545e-06,
"loss": 0.5383090972900391,
"step": 385
},
{
"epoch": 0.37838500183801005,
"grad_norm": 0.4569418132305145,
"learning_rate": 9.357625975786318e-06,
"loss": 0.509972095489502,
"step": 386
},
{
"epoch": 0.3793652738634971,
"grad_norm": 0.34564098715782166,
"learning_rate": 9.353730385598887e-06,
"loss": 0.5029498338699341,
"step": 387
},
{
"epoch": 0.38034554588898417,
"grad_norm": 0.3261181712150574,
"learning_rate": 9.349823834900396e-06,
"loss": 0.46546491980552673,
"step": 388
},
{
"epoch": 0.3813258179144713,
"grad_norm": 0.3854961693286896,
"learning_rate": 9.345906333525582e-06,
"loss": 0.4260493218898773,
"step": 389
},
{
"epoch": 0.38230608993995835,
"grad_norm": 0.40638965368270874,
"learning_rate": 9.341977891336748e-06,
"loss": 0.5061936974525452,
"step": 390
},
{
"epoch": 0.3832863619654454,
"grad_norm": 0.3685773015022278,
"learning_rate": 9.338038518223746e-06,
"loss": 0.4264472723007202,
"step": 391
},
{
"epoch": 0.38426663399093247,
"grad_norm": 0.38765469193458557,
"learning_rate": 9.334088224103942e-06,
"loss": 0.4498711824417114,
"step": 392
},
{
"epoch": 0.38524690601641953,
"grad_norm": 0.5639768242835999,
"learning_rate": 9.330127018922195e-06,
"loss": 0.5633841753005981,
"step": 393
},
{
"epoch": 0.38622717804190665,
"grad_norm": 0.38081955909729004,
"learning_rate": 9.326154912650834e-06,
"loss": 0.45630359649658203,
"step": 394
},
{
"epoch": 0.3872074500673937,
"grad_norm": 0.370786190032959,
"learning_rate": 9.322171915289635e-06,
"loss": 0.531493604183197,
"step": 395
},
{
"epoch": 0.38818772209288077,
"grad_norm": 0.6124679446220398,
"learning_rate": 9.318178036865786e-06,
"loss": 0.45353901386260986,
"step": 396
},
{
"epoch": 0.38916799411836783,
"grad_norm": 0.37071406841278076,
"learning_rate": 9.314173287433874e-06,
"loss": 0.5473002195358276,
"step": 397
},
{
"epoch": 0.3901482661438549,
"grad_norm": 0.42948129773139954,
"learning_rate": 9.310157677075847e-06,
"loss": 0.47444501519203186,
"step": 398
},
{
"epoch": 0.391128538169342,
"grad_norm": 0.5061194896697998,
"learning_rate": 9.306131215901004e-06,
"loss": 0.5286732316017151,
"step": 399
},
{
"epoch": 0.39210881019482907,
"grad_norm": 0.2663268744945526,
"learning_rate": 9.302093914045953e-06,
"loss": 0.3606998324394226,
"step": 400
},
{
"epoch": 0.39308908222031613,
"grad_norm": 0.4448438584804535,
"learning_rate": 9.298045781674595e-06,
"loss": 0.48375701904296875,
"step": 401
},
{
"epoch": 0.3940693542458032,
"grad_norm": 0.4498058259487152,
"learning_rate": 9.293986828978106e-06,
"loss": 0.5115647912025452,
"step": 402
},
{
"epoch": 0.3950496262712903,
"grad_norm": 0.3477303087711334,
"learning_rate": 9.289917066174887e-06,
"loss": 0.497074693441391,
"step": 403
},
{
"epoch": 0.39602989829677737,
"grad_norm": 0.34829002618789673,
"learning_rate": 9.285836503510562e-06,
"loss": 0.4563921093940735,
"step": 404
},
{
"epoch": 0.39701017032226443,
"grad_norm": 0.3950130045413971,
"learning_rate": 9.281745151257946e-06,
"loss": 0.4813639521598816,
"step": 405
},
{
"epoch": 0.3979904423477515,
"grad_norm": 0.3934921622276306,
"learning_rate": 9.277643019717011e-06,
"loss": 0.5899122953414917,
"step": 406
},
{
"epoch": 0.39897071437323856,
"grad_norm": 0.39351391792297363,
"learning_rate": 9.273530119214868e-06,
"loss": 0.46426674723625183,
"step": 407
},
{
"epoch": 0.39995098639872567,
"grad_norm": 0.3440612256526947,
"learning_rate": 9.269406460105742e-06,
"loss": 0.42837169766426086,
"step": 408
},
{
"epoch": 0.40093125842421273,
"grad_norm": 0.3579244315624237,
"learning_rate": 9.265272052770936e-06,
"loss": 0.4589236080646515,
"step": 409
},
{
"epoch": 0.4019115304496998,
"grad_norm": 0.297066330909729,
"learning_rate": 9.261126907618818e-06,
"loss": 0.4442852735519409,
"step": 410
},
{
"epoch": 0.40289180247518686,
"grad_norm": 0.42027294635772705,
"learning_rate": 9.256971035084786e-06,
"loss": 0.5384013056755066,
"step": 411
},
{
"epoch": 0.4038720745006739,
"grad_norm": 0.39028677344322205,
"learning_rate": 9.252804445631243e-06,
"loss": 0.42798590660095215,
"step": 412
},
{
"epoch": 0.40485234652616103,
"grad_norm": 0.43818506598472595,
"learning_rate": 9.248627149747573e-06,
"loss": 0.617618203163147,
"step": 413
},
{
"epoch": 0.4058326185516481,
"grad_norm": 0.36370301246643066,
"learning_rate": 9.244439157950114e-06,
"loss": 0.6180145740509033,
"step": 414
},
{
"epoch": 0.40681289057713516,
"grad_norm": 0.35960403084754944,
"learning_rate": 9.24024048078213e-06,
"loss": 0.45949965715408325,
"step": 415
},
{
"epoch": 0.4077931626026222,
"grad_norm": 0.44299089908599854,
"learning_rate": 9.236031128813786e-06,
"loss": 0.5033418536186218,
"step": 416
},
{
"epoch": 0.4087734346281093,
"grad_norm": 0.35620710253715515,
"learning_rate": 9.231811112642121e-06,
"loss": 0.4707130193710327,
"step": 417
},
{
"epoch": 0.4097537066535964,
"grad_norm": 0.3033544421195984,
"learning_rate": 9.227580442891022e-06,
"loss": 0.4644666314125061,
"step": 418
},
{
"epoch": 0.41073397867908346,
"grad_norm": 0.3623369634151459,
"learning_rate": 9.223339130211194e-06,
"loss": 0.44792062044143677,
"step": 419
},
{
"epoch": 0.4117142507045705,
"grad_norm": 0.37476927042007446,
"learning_rate": 9.219087185280132e-06,
"loss": 0.5332671403884888,
"step": 420
},
{
"epoch": 0.4126945227300576,
"grad_norm": 0.5440754294395447,
"learning_rate": 9.214824618802108e-06,
"loss": 0.5702677369117737,
"step": 421
},
{
"epoch": 0.41367479475554464,
"grad_norm": 0.3983120918273926,
"learning_rate": 9.210551441508126e-06,
"loss": 0.4130643606185913,
"step": 422
},
{
"epoch": 0.41465506678103176,
"grad_norm": 0.39142894744873047,
"learning_rate": 9.206267664155906e-06,
"loss": 0.39005011320114136,
"step": 423
},
{
"epoch": 0.4156353388065188,
"grad_norm": 0.3855965733528137,
"learning_rate": 9.201973297529852e-06,
"loss": 0.4736930727958679,
"step": 424
},
{
"epoch": 0.4166156108320059,
"grad_norm": 0.3965217173099518,
"learning_rate": 9.197668352441025e-06,
"loss": 0.44794413447380066,
"step": 425
},
{
"epoch": 0.41759588285749294,
"grad_norm": 0.4432395398616791,
"learning_rate": 9.193352839727122e-06,
"loss": 0.4554362893104553,
"step": 426
},
{
"epoch": 0.41857615488298,
"grad_norm": 0.35262706875801086,
"learning_rate": 9.189026770252437e-06,
"loss": 0.4812745749950409,
"step": 427
},
{
"epoch": 0.4195564269084671,
"grad_norm": 0.3891749680042267,
"learning_rate": 9.18469015490785e-06,
"loss": 0.5299009680747986,
"step": 428
},
{
"epoch": 0.4205366989339542,
"grad_norm": 0.5775208473205566,
"learning_rate": 9.18034300461078e-06,
"loss": 0.5198730826377869,
"step": 429
},
{
"epoch": 0.42151697095944124,
"grad_norm": 0.43263643980026245,
"learning_rate": 9.175985330305176e-06,
"loss": 0.5285135507583618,
"step": 430
},
{
"epoch": 0.4224972429849283,
"grad_norm": 0.4449794888496399,
"learning_rate": 9.171617142961477e-06,
"loss": 0.47849366068840027,
"step": 431
},
{
"epoch": 0.4234775150104154,
"grad_norm": 0.3742443919181824,
"learning_rate": 9.167238453576589e-06,
"loss": 0.4573095440864563,
"step": 432
},
{
"epoch": 0.4244577870359025,
"grad_norm": 0.37179848551750183,
"learning_rate": 9.162849273173857e-06,
"loss": 0.48387712240219116,
"step": 433
},
{
"epoch": 0.42543805906138954,
"grad_norm": 0.3982168734073639,
"learning_rate": 9.158449612803039e-06,
"loss": 0.5581054091453552,
"step": 434
},
{
"epoch": 0.4264183310868766,
"grad_norm": 0.36328697204589844,
"learning_rate": 9.154039483540273e-06,
"loss": 0.494734525680542,
"step": 435
},
{
"epoch": 0.42739860311236366,
"grad_norm": 0.4150455892086029,
"learning_rate": 9.149618896488056e-06,
"loss": 0.48160862922668457,
"step": 436
},
{
"epoch": 0.4283788751378508,
"grad_norm": 0.3177533745765686,
"learning_rate": 9.145187862775208e-06,
"loss": 0.3643690347671509,
"step": 437
},
{
"epoch": 0.42935914716333784,
"grad_norm": 0.453244149684906,
"learning_rate": 9.140746393556853e-06,
"loss": 0.5189401507377625,
"step": 438
},
{
"epoch": 0.4303394191888249,
"grad_norm": 0.3507213592529297,
"learning_rate": 9.136294500014387e-06,
"loss": 0.5552747845649719,
"step": 439
},
{
"epoch": 0.43131969121431196,
"grad_norm": 0.4679151475429535,
"learning_rate": 9.13183219335544e-06,
"loss": 0.47891873121261597,
"step": 440
},
{
"epoch": 0.432299963239799,
"grad_norm": 0.3539535105228424,
"learning_rate": 9.12735948481387e-06,
"loss": 0.493435263633728,
"step": 441
},
{
"epoch": 0.43328023526528614,
"grad_norm": 0.3762354254722595,
"learning_rate": 9.122876385649713e-06,
"loss": 0.42575687170028687,
"step": 442
},
{
"epoch": 0.4342605072907732,
"grad_norm": 0.34173211455345154,
"learning_rate": 9.118382907149164e-06,
"loss": 0.5345132350921631,
"step": 443
},
{
"epoch": 0.43524077931626026,
"grad_norm": 0.45629680156707764,
"learning_rate": 9.113879060624554e-06,
"loss": 0.447407603263855,
"step": 444
},
{
"epoch": 0.4362210513417473,
"grad_norm": 0.34496432542800903,
"learning_rate": 9.109364857414306e-06,
"loss": 0.4799361228942871,
"step": 445
},
{
"epoch": 0.4372013233672344,
"grad_norm": 0.3275192975997925,
"learning_rate": 9.104840308882925e-06,
"loss": 0.46993952989578247,
"step": 446
},
{
"epoch": 0.4381815953927215,
"grad_norm": 0.47728461027145386,
"learning_rate": 9.100305426420957e-06,
"loss": 0.5420528650283813,
"step": 447
},
{
"epoch": 0.43916186741820856,
"grad_norm": 0.37131267786026,
"learning_rate": 9.09576022144496e-06,
"loss": 0.4462127089500427,
"step": 448
},
{
"epoch": 0.4401421394436956,
"grad_norm": 0.3171160817146301,
"learning_rate": 9.091204705397485e-06,
"loss": 0.4324248135089874,
"step": 449
},
{
"epoch": 0.4411224114691827,
"grad_norm": 0.3588997721672058,
"learning_rate": 9.086638889747034e-06,
"loss": 0.48427823185920715,
"step": 450
},
{
"epoch": 0.44210268349466975,
"grad_norm": 0.2989208400249481,
"learning_rate": 9.08206278598805e-06,
"loss": 0.33199605345726013,
"step": 451
},
{
"epoch": 0.44308295552015686,
"grad_norm": 0.44509533047676086,
"learning_rate": 9.077476405640862e-06,
"loss": 0.553246021270752,
"step": 452
},
{
"epoch": 0.4440632275456439,
"grad_norm": 0.3120643198490143,
"learning_rate": 9.07287976025168e-06,
"loss": 0.5017093420028687,
"step": 453
},
{
"epoch": 0.445043499571131,
"grad_norm": 0.46251463890075684,
"learning_rate": 9.06827286139255e-06,
"loss": 0.45640844106674194,
"step": 454
},
{
"epoch": 0.44602377159661805,
"grad_norm": 0.506580650806427,
"learning_rate": 9.06365572066134e-06,
"loss": 0.5286303758621216,
"step": 455
},
{
"epoch": 0.4470040436221051,
"grad_norm": 0.3329618573188782,
"learning_rate": 9.059028349681693e-06,
"loss": 0.48349183797836304,
"step": 456
},
{
"epoch": 0.4479843156475922,
"grad_norm": 0.338229238986969,
"learning_rate": 9.05439076010301e-06,
"loss": 0.4085380733013153,
"step": 457
},
{
"epoch": 0.4489645876730793,
"grad_norm": 0.5997896790504456,
"learning_rate": 9.04974296360042e-06,
"loss": 0.4864353537559509,
"step": 458
},
{
"epoch": 0.44994485969856635,
"grad_norm": 0.31409236788749695,
"learning_rate": 9.045084971874738e-06,
"loss": 0.4898025393486023,
"step": 459
},
{
"epoch": 0.4509251317240534,
"grad_norm": 0.497396320104599,
"learning_rate": 9.040416796652459e-06,
"loss": 0.5317016243934631,
"step": 460
},
{
"epoch": 0.4519054037495405,
"grad_norm": 0.4312289357185364,
"learning_rate": 9.035738449685707e-06,
"loss": 0.44609835743904114,
"step": 461
},
{
"epoch": 0.4528856757750276,
"grad_norm": 0.37432053685188293,
"learning_rate": 9.031049942752215e-06,
"loss": 0.5482521057128906,
"step": 462
},
{
"epoch": 0.45386594780051465,
"grad_norm": 0.3643771708011627,
"learning_rate": 9.026351287655294e-06,
"loss": 0.444631963968277,
"step": 463
},
{
"epoch": 0.4548462198260017,
"grad_norm": 0.8187288641929626,
"learning_rate": 9.021642496223802e-06,
"loss": 0.4739605784416199,
"step": 464
},
{
"epoch": 0.45582649185148877,
"grad_norm": 0.4810348153114319,
"learning_rate": 9.016923580312114e-06,
"loss": 0.4590006172657013,
"step": 465
},
{
"epoch": 0.4568067638769759,
"grad_norm": 0.33898085355758667,
"learning_rate": 9.012194551800098e-06,
"loss": 0.44440728425979614,
"step": 466
},
{
"epoch": 0.45778703590246295,
"grad_norm": 0.48001205921173096,
"learning_rate": 9.007455422593077e-06,
"loss": 0.4935011863708496,
"step": 467
},
{
"epoch": 0.45876730792795,
"grad_norm": 0.2779478430747986,
"learning_rate": 9.002706204621802e-06,
"loss": 0.44410941004753113,
"step": 468
},
{
"epoch": 0.45974757995343707,
"grad_norm": 0.34186968207359314,
"learning_rate": 8.997946909842426e-06,
"loss": 0.43802377581596375,
"step": 469
},
{
"epoch": 0.46072785197892413,
"grad_norm": 0.3661576807498932,
"learning_rate": 8.993177550236464e-06,
"loss": 0.45635926723480225,
"step": 470
},
{
"epoch": 0.46170812400441125,
"grad_norm": 1.0663844347000122,
"learning_rate": 8.988398137810778e-06,
"loss": 0.43027034401893616,
"step": 471
},
{
"epoch": 0.4626883960298983,
"grad_norm": 0.5256292819976807,
"learning_rate": 8.983608684597529e-06,
"loss": 0.5046781301498413,
"step": 472
},
{
"epoch": 0.46366866805538537,
"grad_norm": 0.5332661271095276,
"learning_rate": 8.978809202654161e-06,
"loss": 0.5197567939758301,
"step": 473
},
{
"epoch": 0.46464894008087243,
"grad_norm": 0.657674252986908,
"learning_rate": 8.973999704063365e-06,
"loss": 0.4685856103897095,
"step": 474
},
{
"epoch": 0.4656292121063595,
"grad_norm": 0.42174631357192993,
"learning_rate": 8.969180200933048e-06,
"loss": 0.5563768148422241,
"step": 475
},
{
"epoch": 0.4666094841318466,
"grad_norm": 0.4088301956653595,
"learning_rate": 8.9643507053963e-06,
"loss": 0.4976760745048523,
"step": 476
},
{
"epoch": 0.46758975615733367,
"grad_norm": 0.37379127740859985,
"learning_rate": 8.959511229611377e-06,
"loss": 0.587893545627594,
"step": 477
},
{
"epoch": 0.46857002818282073,
"grad_norm": 0.2950038015842438,
"learning_rate": 8.954661785761648e-06,
"loss": 0.5042154788970947,
"step": 478
},
{
"epoch": 0.4695503002083078,
"grad_norm": 0.5372236371040344,
"learning_rate": 8.949802386055582e-06,
"loss": 0.5769034028053284,
"step": 479
},
{
"epoch": 0.47053057223379485,
"grad_norm": 0.3698883056640625,
"learning_rate": 8.944933042726715e-06,
"loss": 0.4811990261077881,
"step": 480
},
{
"epoch": 0.47151084425928197,
"grad_norm": 0.3767825663089752,
"learning_rate": 8.94005376803361e-06,
"loss": 0.5653142929077148,
"step": 481
},
{
"epoch": 0.47249111628476903,
"grad_norm": 0.5690767168998718,
"learning_rate": 8.935164574259838e-06,
"loss": 0.5382915735244751,
"step": 482
},
{
"epoch": 0.4734713883102561,
"grad_norm": 0.4781832695007324,
"learning_rate": 8.930265473713939e-06,
"loss": 0.504697322845459,
"step": 483
},
{
"epoch": 0.47445166033574315,
"grad_norm": 0.45350706577301025,
"learning_rate": 8.925356478729389e-06,
"loss": 0.4871259927749634,
"step": 484
},
{
"epoch": 0.4754319323612302,
"grad_norm": 0.3693767786026001,
"learning_rate": 8.92043760166458e-06,
"loss": 0.5031224489212036,
"step": 485
},
{
"epoch": 0.47641220438671733,
"grad_norm": 0.3703589141368866,
"learning_rate": 8.915508854902778e-06,
"loss": 0.4236990809440613,
"step": 486
},
{
"epoch": 0.4773924764122044,
"grad_norm": 0.35352823138237,
"learning_rate": 8.910570250852098e-06,
"loss": 0.3845526874065399,
"step": 487
},
{
"epoch": 0.47837274843769145,
"grad_norm": 0.29366472363471985,
"learning_rate": 8.905621801945467e-06,
"loss": 0.44821709394454956,
"step": 488
},
{
"epoch": 0.4793530204631785,
"grad_norm": 0.4202713370323181,
"learning_rate": 8.900663520640605e-06,
"loss": 0.5091155767440796,
"step": 489
},
{
"epoch": 0.48033329248866563,
"grad_norm": 0.37860098481178284,
"learning_rate": 8.895695419419973e-06,
"loss": 0.6661853194236755,
"step": 490
},
{
"epoch": 0.4813135645141527,
"grad_norm": 0.3347987234592438,
"learning_rate": 8.890717510790763e-06,
"loss": 0.46972882747650146,
"step": 491
},
{
"epoch": 0.48229383653963975,
"grad_norm": 0.43081235885620117,
"learning_rate": 8.885729807284855e-06,
"loss": 0.5169898271560669,
"step": 492
},
{
"epoch": 0.4832741085651268,
"grad_norm": 0.39748743176460266,
"learning_rate": 8.880732321458785e-06,
"loss": 0.46452099084854126,
"step": 493
},
{
"epoch": 0.4842543805906139,
"grad_norm": 0.38093870878219604,
"learning_rate": 8.875725065893717e-06,
"loss": 0.4864256680011749,
"step": 494
},
{
"epoch": 0.485234652616101,
"grad_norm": 0.3567042052745819,
"learning_rate": 8.870708053195414e-06,
"loss": 0.5172852277755737,
"step": 495
},
{
"epoch": 0.48621492464158805,
"grad_norm": 0.6800740957260132,
"learning_rate": 8.865681295994195e-06,
"loss": 0.5141050815582275,
"step": 496
},
{
"epoch": 0.4871951966670751,
"grad_norm": 0.3330962061882019,
"learning_rate": 8.860644806944917e-06,
"loss": 0.4011129140853882,
"step": 497
},
{
"epoch": 0.4881754686925622,
"grad_norm": 0.384223073720932,
"learning_rate": 8.85559859872694e-06,
"loss": 0.4802374243736267,
"step": 498
},
{
"epoch": 0.48915574071804924,
"grad_norm": 0.37003326416015625,
"learning_rate": 8.850542684044078e-06,
"loss": 0.42417556047439575,
"step": 499
},
{
"epoch": 0.49013601274353635,
"grad_norm": 0.39030569791793823,
"learning_rate": 8.845477075624598e-06,
"loss": 0.49241891503334045,
"step": 500
},
{
"epoch": 0.4911162847690234,
"grad_norm": 0.35072997212409973,
"learning_rate": 8.84040178622116e-06,
"loss": 0.5204368829727173,
"step": 501
},
{
"epoch": 0.4920965567945105,
"grad_norm": 0.3954608142375946,
"learning_rate": 8.835316828610801e-06,
"loss": 0.5243607759475708,
"step": 502
},
{
"epoch": 0.49307682881999754,
"grad_norm": 0.369228720664978,
"learning_rate": 8.83022221559489e-06,
"loss": 0.5085795521736145,
"step": 503
},
{
"epoch": 0.4940571008454846,
"grad_norm": 0.431448757648468,
"learning_rate": 8.825117959999117e-06,
"loss": 0.4066562056541443,
"step": 504
},
{
"epoch": 0.4950373728709717,
"grad_norm": 0.5024880170822144,
"learning_rate": 8.820004074673433e-06,
"loss": 0.4971996545791626,
"step": 505
},
{
"epoch": 0.4960176448964588,
"grad_norm": 0.9446320533752441,
"learning_rate": 8.814880572492043e-06,
"loss": 0.45391514897346497,
"step": 506
},
{
"epoch": 0.49699791692194584,
"grad_norm": 0.49279657006263733,
"learning_rate": 8.809747466353356e-06,
"loss": 0.5771381855010986,
"step": 507
},
{
"epoch": 0.4979781889474329,
"grad_norm": 0.33148542046546936,
"learning_rate": 8.804604769179958e-06,
"loss": 0.3664973974227905,
"step": 508
},
{
"epoch": 0.49895846097291996,
"grad_norm": 0.46150439977645874,
"learning_rate": 8.799452493918586e-06,
"loss": 0.5991448163986206,
"step": 509
},
{
"epoch": 0.4999387329984071,
"grad_norm": 0.41725149750709534,
"learning_rate": 8.794290653540086e-06,
"loss": 0.5441845655441284,
"step": 510
},
{
"epoch": 0.5009190050238941,
"grad_norm": 0.6852184534072876,
"learning_rate": 8.789119261039385e-06,
"loss": 0.39037442207336426,
"step": 511
},
{
"epoch": 0.5018992770493812,
"grad_norm": 0.5998321175575256,
"learning_rate": 8.783938329435458e-06,
"loss": 0.41170352697372437,
"step": 512
},
{
"epoch": 0.5028795490748683,
"grad_norm": 0.35790279507637024,
"learning_rate": 8.778747871771293e-06,
"loss": 0.40110185742378235,
"step": 513
},
{
"epoch": 0.5038598211003553,
"grad_norm": 0.45980432629585266,
"learning_rate": 8.773547901113862e-06,
"loss": 0.4221269190311432,
"step": 514
},
{
"epoch": 0.5048400931258424,
"grad_norm": 0.30438631772994995,
"learning_rate": 8.768338430554083e-06,
"loss": 0.4353146553039551,
"step": 515
},
{
"epoch": 0.5058203651513294,
"grad_norm": 0.6454721689224243,
"learning_rate": 8.763119473206795e-06,
"loss": 0.49284058809280396,
"step": 516
},
{
"epoch": 0.5068006371768166,
"grad_norm": 0.3560524582862854,
"learning_rate": 8.757891042210713e-06,
"loss": 0.5957845449447632,
"step": 517
},
{
"epoch": 0.5077809092023037,
"grad_norm": 0.4516677260398865,
"learning_rate": 8.752653150728412e-06,
"loss": 0.42609721422195435,
"step": 518
},
{
"epoch": 0.5087611812277907,
"grad_norm": 0.3916192650794983,
"learning_rate": 8.747405811946272e-06,
"loss": 0.4104868173599243,
"step": 519
},
{
"epoch": 0.5097414532532778,
"grad_norm": 0.2893811762332916,
"learning_rate": 8.742149039074463e-06,
"loss": 0.39205431938171387,
"step": 520
},
{
"epoch": 0.5107217252787648,
"grad_norm": 0.44901058077812195,
"learning_rate": 8.736882845346906e-06,
"loss": 0.5412840843200684,
"step": 521
},
{
"epoch": 0.5117019973042519,
"grad_norm": 0.5004028081893921,
"learning_rate": 8.731607244021236e-06,
"loss": 0.5038070678710938,
"step": 522
},
{
"epoch": 0.512682269329739,
"grad_norm": 0.5647842884063721,
"learning_rate": 8.726322248378775e-06,
"loss": 0.4985129237174988,
"step": 523
},
{
"epoch": 0.513662541355226,
"grad_norm": 0.4644056260585785,
"learning_rate": 8.72102787172449e-06,
"loss": 0.4811592400074005,
"step": 524
},
{
"epoch": 0.5146428133807132,
"grad_norm": 0.3694928288459778,
"learning_rate": 8.715724127386971e-06,
"loss": 0.5067281723022461,
"step": 525
},
{
"epoch": 0.5156230854062002,
"grad_norm": 0.3387005031108856,
"learning_rate": 8.710411028718388e-06,
"loss": 0.4230249524116516,
"step": 526
},
{
"epoch": 0.5166033574316873,
"grad_norm": 0.286080926656723,
"learning_rate": 8.705088589094458e-06,
"loss": 0.3872826099395752,
"step": 527
},
{
"epoch": 0.5175836294571744,
"grad_norm": 0.43313509225845337,
"learning_rate": 8.69975682191442e-06,
"loss": 0.5232932567596436,
"step": 528
},
{
"epoch": 0.5185639014826614,
"grad_norm": 0.29711467027664185,
"learning_rate": 8.69441574060099e-06,
"loss": 0.38450002670288086,
"step": 529
},
{
"epoch": 0.5195441735081485,
"grad_norm": 0.40762242674827576,
"learning_rate": 8.68906535860033e-06,
"loss": 0.45559126138687134,
"step": 530
},
{
"epoch": 0.5205244455336356,
"grad_norm": 0.3982981741428375,
"learning_rate": 8.683705689382025e-06,
"loss": 0.525196373462677,
"step": 531
},
{
"epoch": 0.5215047175591226,
"grad_norm": 0.36136937141418457,
"learning_rate": 8.678336746439036e-06,
"loss": 0.5484825372695923,
"step": 532
},
{
"epoch": 0.5224849895846098,
"grad_norm": 0.3282698094844818,
"learning_rate": 8.672958543287666e-06,
"loss": 0.44835835695266724,
"step": 533
},
{
"epoch": 0.5234652616100968,
"grad_norm": 0.3501221537590027,
"learning_rate": 8.667571093467542e-06,
"loss": 0.436707079410553,
"step": 534
},
{
"epoch": 0.5244455336355839,
"grad_norm": 0.35575124621391296,
"learning_rate": 8.662174410541556e-06,
"loss": 0.48804348707199097,
"step": 535
},
{
"epoch": 0.525425805661071,
"grad_norm": 0.33241739869117737,
"learning_rate": 8.656768508095853e-06,
"loss": 0.49325263500213623,
"step": 536
},
{
"epoch": 0.526406077686558,
"grad_norm": 0.5140376091003418,
"learning_rate": 8.651353399739787e-06,
"loss": 0.4976520538330078,
"step": 537
},
{
"epoch": 0.5273863497120451,
"grad_norm": 0.4093848168849945,
"learning_rate": 8.645929099105886e-06,
"loss": 0.46503710746765137,
"step": 538
},
{
"epoch": 0.5283666217375321,
"grad_norm": 0.67371666431427,
"learning_rate": 8.640495619849821e-06,
"loss": 0.50223708152771,
"step": 539
},
{
"epoch": 0.5293468937630192,
"grad_norm": 0.38159266114234924,
"learning_rate": 8.63505297565037e-06,
"loss": 0.4161151647567749,
"step": 540
},
{
"epoch": 0.5303271657885064,
"grad_norm": 0.3228311240673065,
"learning_rate": 8.629601180209382e-06,
"loss": 0.4917384684085846,
"step": 541
},
{
"epoch": 0.5313074378139934,
"grad_norm": 0.5631666779518127,
"learning_rate": 8.624140247251745e-06,
"loss": 0.4398884177207947,
"step": 542
},
{
"epoch": 0.5322877098394805,
"grad_norm": 0.4087778627872467,
"learning_rate": 8.61867019052535e-06,
"loss": 0.4229411780834198,
"step": 543
},
{
"epoch": 0.5332679818649675,
"grad_norm": 0.382101833820343,
"learning_rate": 8.613191023801065e-06,
"loss": 0.40180808305740356,
"step": 544
},
{
"epoch": 0.5342482538904546,
"grad_norm": 0.34626108407974243,
"learning_rate": 8.607702760872679e-06,
"loss": 0.4650096893310547,
"step": 545
},
{
"epoch": 0.5352285259159417,
"grad_norm": 0.44842782616615295,
"learning_rate": 8.602205415556889e-06,
"loss": 0.43995392322540283,
"step": 546
},
{
"epoch": 0.5362087979414287,
"grad_norm": 0.33441150188446045,
"learning_rate": 8.596699001693257e-06,
"loss": 0.42453593015670776,
"step": 547
},
{
"epoch": 0.5371890699669158,
"grad_norm": 0.34509655833244324,
"learning_rate": 8.591183533144172e-06,
"loss": 0.47213613986968994,
"step": 548
},
{
"epoch": 0.5381693419924028,
"grad_norm": 0.3117007613182068,
"learning_rate": 8.585659023794818e-06,
"loss": 0.4197375476360321,
"step": 549
},
{
"epoch": 0.53914961401789,
"grad_norm": 0.3563596308231354,
"learning_rate": 8.580125487553143e-06,
"loss": 0.4717135429382324,
"step": 550
},
{
"epoch": 0.5401298860433771,
"grad_norm": 0.41597846150398254,
"learning_rate": 8.574582938349818e-06,
"loss": 0.4353729784488678,
"step": 551
},
{
"epoch": 0.5411101580688641,
"grad_norm": 0.3439650535583496,
"learning_rate": 8.569031390138202e-06,
"loss": 0.553699791431427,
"step": 552
},
{
"epoch": 0.5420904300943512,
"grad_norm": 0.32140955328941345,
"learning_rate": 8.563470856894316e-06,
"loss": 0.43937599658966064,
"step": 553
},
{
"epoch": 0.5430707021198382,
"grad_norm": 0.3415094017982483,
"learning_rate": 8.557901352616788e-06,
"loss": 0.5031224489212036,
"step": 554
},
{
"epoch": 0.5440509741453253,
"grad_norm": 0.32968753576278687,
"learning_rate": 8.552322891326846e-06,
"loss": 0.4652668833732605,
"step": 555
},
{
"epoch": 0.5450312461708124,
"grad_norm": 0.36093825101852417,
"learning_rate": 8.546735487068253e-06,
"loss": 0.49096938967704773,
"step": 556
},
{
"epoch": 0.5460115181962995,
"grad_norm": 0.33851516246795654,
"learning_rate": 8.541139153907296e-06,
"loss": 0.42208337783813477,
"step": 557
},
{
"epoch": 0.5469917902217866,
"grad_norm": 0.29839521646499634,
"learning_rate": 8.535533905932739e-06,
"loss": 0.3912207782268524,
"step": 558
},
{
"epoch": 0.5479720622472736,
"grad_norm": 0.5822285413742065,
"learning_rate": 8.529919757255783e-06,
"loss": 0.4395484924316406,
"step": 559
},
{
"epoch": 0.5489523342727607,
"grad_norm": 1.7440687417984009,
"learning_rate": 8.524296722010042e-06,
"loss": 0.5006183981895447,
"step": 560
},
{
"epoch": 0.5499326062982478,
"grad_norm": 0.3179253041744232,
"learning_rate": 8.518664814351502e-06,
"loss": 0.43906790018081665,
"step": 561
},
{
"epoch": 0.5509128783237348,
"grad_norm": 0.818109929561615,
"learning_rate": 8.513024048458484e-06,
"loss": 0.4221283495426178,
"step": 562
},
{
"epoch": 0.5518931503492219,
"grad_norm": 0.3536345660686493,
"learning_rate": 8.507374438531606e-06,
"loss": 0.4571268856525421,
"step": 563
},
{
"epoch": 0.5528734223747089,
"grad_norm": 0.7931411266326904,
"learning_rate": 8.501715998793758e-06,
"loss": 0.3945927619934082,
"step": 564
},
{
"epoch": 0.553853694400196,
"grad_norm": 0.40802615880966187,
"learning_rate": 8.496048743490053e-06,
"loss": 0.46674811840057373,
"step": 565
},
{
"epoch": 0.5548339664256832,
"grad_norm": 0.35188597440719604,
"learning_rate": 8.490372686887802e-06,
"loss": 0.4779481291770935,
"step": 566
},
{
"epoch": 0.5558142384511702,
"grad_norm": 0.4562278389930725,
"learning_rate": 8.48468784327647e-06,
"loss": 0.5522409677505493,
"step": 567
},
{
"epoch": 0.5567945104766573,
"grad_norm": 0.5045914053916931,
"learning_rate": 8.478994226967638e-06,
"loss": 0.4797571003437042,
"step": 568
},
{
"epoch": 0.5577747825021443,
"grad_norm": 0.4669787883758545,
"learning_rate": 8.473291852294986e-06,
"loss": 0.5092311501502991,
"step": 569
},
{
"epoch": 0.5587550545276314,
"grad_norm": 0.37657758593559265,
"learning_rate": 8.467580733614233e-06,
"loss": 0.48225507140159607,
"step": 570
},
{
"epoch": 0.5597353265531185,
"grad_norm": 0.37402990460395813,
"learning_rate": 8.461860885303116e-06,
"loss": 0.4428871273994446,
"step": 571
},
{
"epoch": 0.5607155985786055,
"grad_norm": 0.40197864174842834,
"learning_rate": 8.45613232176134e-06,
"loss": 0.42979440093040466,
"step": 572
},
{
"epoch": 0.5616958706040927,
"grad_norm": 0.4038158357143402,
"learning_rate": 8.450395057410561e-06,
"loss": 0.4812285006046295,
"step": 573
},
{
"epoch": 0.5626761426295797,
"grad_norm": 0.3953388035297394,
"learning_rate": 8.444649106694335e-06,
"loss": 0.4984673857688904,
"step": 574
},
{
"epoch": 0.5636564146550668,
"grad_norm": 0.33300045132637024,
"learning_rate": 8.438894484078086e-06,
"loss": 0.4090971052646637,
"step": 575
},
{
"epoch": 0.5646366866805539,
"grad_norm": 0.40560683608055115,
"learning_rate": 8.433131204049067e-06,
"loss": 0.47987645864486694,
"step": 576
},
{
"epoch": 0.5656169587060409,
"grad_norm": 0.38534432649612427,
"learning_rate": 8.427359281116335e-06,
"loss": 0.41153815388679504,
"step": 577
},
{
"epoch": 0.566597230731528,
"grad_norm": 0.4080407917499542,
"learning_rate": 8.421578729810693e-06,
"loss": 0.4992625117301941,
"step": 578
},
{
"epoch": 0.567577502757015,
"grad_norm": 0.43664252758026123,
"learning_rate": 8.415789564684673e-06,
"loss": 0.5746076703071594,
"step": 579
},
{
"epoch": 0.5685577747825021,
"grad_norm": 0.6770800948143005,
"learning_rate": 8.409991800312493e-06,
"loss": 0.481756329536438,
"step": 580
},
{
"epoch": 0.5695380468079893,
"grad_norm": 0.5344410538673401,
"learning_rate": 8.404185451290017e-06,
"loss": 0.49433791637420654,
"step": 581
},
{
"epoch": 0.5705183188334763,
"grad_norm": 0.45804905891418457,
"learning_rate": 8.398370532234723e-06,
"loss": 0.5353056192398071,
"step": 582
},
{
"epoch": 0.5714985908589634,
"grad_norm": 0.2677948772907257,
"learning_rate": 8.392547057785662e-06,
"loss": 0.3388299345970154,
"step": 583
},
{
"epoch": 0.5724788628844505,
"grad_norm": 0.3658766746520996,
"learning_rate": 8.386715042603424e-06,
"loss": 0.46216684579849243,
"step": 584
},
{
"epoch": 0.5734591349099375,
"grad_norm": 1.5290638208389282,
"learning_rate": 8.380874501370098e-06,
"loss": 0.44735243916511536,
"step": 585
},
{
"epoch": 0.5744394069354246,
"grad_norm": 0.34169498085975647,
"learning_rate": 8.375025448789242e-06,
"loss": 0.40964770317077637,
"step": 586
},
{
"epoch": 0.5754196789609116,
"grad_norm": 0.4692463278770447,
"learning_rate": 8.36916789958584e-06,
"loss": 0.4502611458301544,
"step": 587
},
{
"epoch": 0.5763999509863987,
"grad_norm": 0.3461305797100067,
"learning_rate": 8.363301868506264e-06,
"loss": 0.49499672651290894,
"step": 588
},
{
"epoch": 0.5773802230118859,
"grad_norm": 0.33883342146873474,
"learning_rate": 8.357427370318239e-06,
"loss": 0.44890397787094116,
"step": 589
},
{
"epoch": 0.5783604950373729,
"grad_norm": 0.5561463832855225,
"learning_rate": 8.351544419810808e-06,
"loss": 0.43895572423934937,
"step": 590
},
{
"epoch": 0.57934076706286,
"grad_norm": 0.3551168143749237,
"learning_rate": 8.345653031794292e-06,
"loss": 0.455233633518219,
"step": 591
},
{
"epoch": 0.580321039088347,
"grad_norm": 0.44089698791503906,
"learning_rate": 8.339753221100254e-06,
"loss": 0.4776260256767273,
"step": 592
},
{
"epoch": 0.5813013111138341,
"grad_norm": 0.3547059893608093,
"learning_rate": 8.33384500258146e-06,
"loss": 0.6100218296051025,
"step": 593
},
{
"epoch": 0.5822815831393212,
"grad_norm": 0.3045068383216858,
"learning_rate": 8.327928391111841e-06,
"loss": 0.47258082032203674,
"step": 594
},
{
"epoch": 0.5832618551648082,
"grad_norm": 0.4481658339500427,
"learning_rate": 8.322003401586463e-06,
"loss": 0.5098180770874023,
"step": 595
},
{
"epoch": 0.5842421271902953,
"grad_norm": 0.2963522970676422,
"learning_rate": 8.316070048921477e-06,
"loss": 0.3966555595397949,
"step": 596
},
{
"epoch": 0.5852223992157823,
"grad_norm": 0.29471850395202637,
"learning_rate": 8.310128348054093e-06,
"loss": 0.37038129568099976,
"step": 597
},
{
"epoch": 0.5862026712412695,
"grad_norm": 0.5274387001991272,
"learning_rate": 8.304178313942536e-06,
"loss": 0.444481760263443,
"step": 598
},
{
"epoch": 0.5871829432667566,
"grad_norm": 0.33825987577438354,
"learning_rate": 8.298219961566008e-06,
"loss": 0.4450898766517639,
"step": 599
},
{
"epoch": 0.5881632152922436,
"grad_norm": 0.4119406044483185,
"learning_rate": 8.292253305924656e-06,
"loss": 0.46286869049072266,
"step": 600
},
{
"epoch": 0.5891434873177307,
"grad_norm": 0.410395085811615,
"learning_rate": 8.286278362039527e-06,
"loss": 0.5062334537506104,
"step": 601
},
{
"epoch": 0.5901237593432177,
"grad_norm": 0.3780165910720825,
"learning_rate": 8.280295144952537e-06,
"loss": 0.5119242072105408,
"step": 602
},
{
"epoch": 0.5911040313687048,
"grad_norm": 0.4394252300262451,
"learning_rate": 8.274303669726427e-06,
"loss": 0.4055041968822479,
"step": 603
},
{
"epoch": 0.5920843033941919,
"grad_norm": 0.3775176405906677,
"learning_rate": 8.268303951444727e-06,
"loss": 0.3980475068092346,
"step": 604
},
{
"epoch": 0.5930645754196789,
"grad_norm": 0.34429699182510376,
"learning_rate": 8.262296005211722e-06,
"loss": 0.3743438124656677,
"step": 605
},
{
"epoch": 0.5940448474451661,
"grad_norm": 0.3745001256465912,
"learning_rate": 8.25627984615241e-06,
"loss": 0.5286900401115417,
"step": 606
},
{
"epoch": 0.5950251194706531,
"grad_norm": 0.31092026829719543,
"learning_rate": 8.250255489412464e-06,
"loss": 0.4566567540168762,
"step": 607
},
{
"epoch": 0.5960053914961402,
"grad_norm": 0.33834308385849,
"learning_rate": 8.244222950158194e-06,
"loss": 0.5007248520851135,
"step": 608
},
{
"epoch": 0.5969856635216273,
"grad_norm": 0.3195263147354126,
"learning_rate": 8.238182243576512e-06,
"loss": 0.3759029507637024,
"step": 609
},
{
"epoch": 0.5979659355471143,
"grad_norm": 0.3228251039981842,
"learning_rate": 8.232133384874887e-06,
"loss": 0.4775990843772888,
"step": 610
},
{
"epoch": 0.5989462075726014,
"grad_norm": 0.3790968656539917,
"learning_rate": 8.226076389281316e-06,
"loss": 0.5054103136062622,
"step": 611
},
{
"epoch": 0.5999264795980884,
"grad_norm": 0.36363181471824646,
"learning_rate": 8.220011272044276e-06,
"loss": 0.4811822175979614,
"step": 612
},
{
"epoch": 0.6009067516235755,
"grad_norm": 0.3438539206981659,
"learning_rate": 8.213938048432697e-06,
"loss": 0.46697765588760376,
"step": 613
},
{
"epoch": 0.6018870236490627,
"grad_norm": 0.2779245376586914,
"learning_rate": 8.207856733735908e-06,
"loss": 0.4356715977191925,
"step": 614
},
{
"epoch": 0.6028672956745497,
"grad_norm": 0.39573705196380615,
"learning_rate": 8.201767343263612e-06,
"loss": 0.4438568353652954,
"step": 615
},
{
"epoch": 0.6038475677000368,
"grad_norm": 0.34524670243263245,
"learning_rate": 8.195669892345842e-06,
"loss": 0.434609591960907,
"step": 616
},
{
"epoch": 0.6048278397255238,
"grad_norm": 0.3967142105102539,
"learning_rate": 8.189564396332927e-06,
"loss": 0.4618009924888611,
"step": 617
},
{
"epoch": 0.6058081117510109,
"grad_norm": 0.34043848514556885,
"learning_rate": 8.183450870595443e-06,
"loss": 0.41299062967300415,
"step": 618
},
{
"epoch": 0.606788383776498,
"grad_norm": 0.337545245885849,
"learning_rate": 8.177329330524182e-06,
"loss": 0.44722437858581543,
"step": 619
},
{
"epoch": 0.607768655801985,
"grad_norm": 0.3154465854167938,
"learning_rate": 8.171199791530117e-06,
"loss": 0.47247111797332764,
"step": 620
},
{
"epoch": 0.6087489278274721,
"grad_norm": 0.3351287543773651,
"learning_rate": 8.165062269044353e-06,
"loss": 0.4353683888912201,
"step": 621
},
{
"epoch": 0.6097291998529591,
"grad_norm": 0.3771119713783264,
"learning_rate": 8.158916778518097e-06,
"loss": 0.46275755763053894,
"step": 622
},
{
"epoch": 0.6107094718784463,
"grad_norm": 1.5629202127456665,
"learning_rate": 8.152763335422612e-06,
"loss": 0.4777594804763794,
"step": 623
},
{
"epoch": 0.6116897439039334,
"grad_norm": 0.33430737257003784,
"learning_rate": 8.146601955249187e-06,
"loss": 0.4601077437400818,
"step": 624
},
{
"epoch": 0.6126700159294204,
"grad_norm": 0.3842839300632477,
"learning_rate": 8.140432653509089e-06,
"loss": 0.4376612603664398,
"step": 625
},
{
"epoch": 0.6136502879549075,
"grad_norm": 0.4760330617427826,
"learning_rate": 8.134255445733524e-06,
"loss": 0.5028945207595825,
"step": 626
},
{
"epoch": 0.6146305599803945,
"grad_norm": 0.31411486864089966,
"learning_rate": 8.128070347473609e-06,
"loss": 0.4536152184009552,
"step": 627
},
{
"epoch": 0.6156108320058816,
"grad_norm": 0.30326178669929504,
"learning_rate": 8.121877374300318e-06,
"loss": 0.407229483127594,
"step": 628
},
{
"epoch": 0.6165911040313687,
"grad_norm": 1.2420125007629395,
"learning_rate": 8.115676541804456e-06,
"loss": 0.41989070177078247,
"step": 629
},
{
"epoch": 0.6175713760568557,
"grad_norm": 0.31548216938972473,
"learning_rate": 8.109467865596612e-06,
"loss": 0.4348871111869812,
"step": 630
},
{
"epoch": 0.6185516480823429,
"grad_norm": 0.27843648195266724,
"learning_rate": 8.10325136130712e-06,
"loss": 0.4506821036338806,
"step": 631
},
{
"epoch": 0.6195319201078299,
"grad_norm": 0.7742169499397278,
"learning_rate": 8.097027044586021e-06,
"loss": 0.5676933526992798,
"step": 632
},
{
"epoch": 0.620512192133317,
"grad_norm": 0.37222349643707275,
"learning_rate": 8.090794931103026e-06,
"loss": 0.42997249960899353,
"step": 633
},
{
"epoch": 0.6214924641588041,
"grad_norm": 0.46167317032814026,
"learning_rate": 8.084555036547475e-06,
"loss": 0.4906580448150635,
"step": 634
},
{
"epoch": 0.6224727361842911,
"grad_norm": 0.28733551502227783,
"learning_rate": 8.078307376628292e-06,
"loss": 0.3763286769390106,
"step": 635
},
{
"epoch": 0.6234530082097782,
"grad_norm": 0.3981354236602783,
"learning_rate": 8.072051967073955e-06,
"loss": 0.4578377604484558,
"step": 636
},
{
"epoch": 0.6244332802352652,
"grad_norm": 0.26882219314575195,
"learning_rate": 8.065788823632451e-06,
"loss": 0.3310215473175049,
"step": 637
},
{
"epoch": 0.6254135522607523,
"grad_norm": 0.2641872465610504,
"learning_rate": 8.059517962071234e-06,
"loss": 0.3867686986923218,
"step": 638
},
{
"epoch": 0.6263938242862395,
"grad_norm": 0.3582305312156677,
"learning_rate": 8.053239398177191e-06,
"loss": 0.43811437487602234,
"step": 639
},
{
"epoch": 0.6273740963117265,
"grad_norm": 0.7451300024986267,
"learning_rate": 8.046953147756601e-06,
"loss": 0.4913032054901123,
"step": 640
},
{
"epoch": 0.6283543683372136,
"grad_norm": 0.2916222810745239,
"learning_rate": 8.04065922663509e-06,
"loss": 0.38071513175964355,
"step": 641
},
{
"epoch": 0.6293346403627007,
"grad_norm": 0.4061773717403412,
"learning_rate": 8.034357650657597e-06,
"loss": 0.6400339603424072,
"step": 642
},
{
"epoch": 0.6303149123881877,
"grad_norm": 0.34440261125564575,
"learning_rate": 8.028048435688333e-06,
"loss": 0.41918694972991943,
"step": 643
},
{
"epoch": 0.6312951844136748,
"grad_norm": 0.382970929145813,
"learning_rate": 8.02173159761074e-06,
"loss": 0.4560946822166443,
"step": 644
},
{
"epoch": 0.6322754564391618,
"grad_norm": 0.41820210218429565,
"learning_rate": 8.015407152327448e-06,
"loss": 0.5259404182434082,
"step": 645
},
{
"epoch": 0.633255728464649,
"grad_norm": 0.3859950006008148,
"learning_rate": 8.009075115760243e-06,
"loss": 0.5085136294364929,
"step": 646
},
{
"epoch": 0.6342360004901361,
"grad_norm": 0.39102399349212646,
"learning_rate": 8.002735503850016e-06,
"loss": 0.41963809728622437,
"step": 647
},
{
"epoch": 0.6352162725156231,
"grad_norm": 0.4935574233531952,
"learning_rate": 7.996388332556735e-06,
"loss": 0.5284677743911743,
"step": 648
},
{
"epoch": 0.6361965445411102,
"grad_norm": 0.33174291253089905,
"learning_rate": 7.990033617859396e-06,
"loss": 0.40616124868392944,
"step": 649
},
{
"epoch": 0.6371768165665972,
"grad_norm": 0.358569473028183,
"learning_rate": 7.983671375755986e-06,
"loss": 0.4649643898010254,
"step": 650
},
{
"epoch": 0.6381570885920843,
"grad_norm": 0.3316868245601654,
"learning_rate": 7.97730162226344e-06,
"loss": 0.5194536447525024,
"step": 651
},
{
"epoch": 0.6391373606175714,
"grad_norm": 0.3334660530090332,
"learning_rate": 7.970924373417608e-06,
"loss": 0.4895654320716858,
"step": 652
},
{
"epoch": 0.6401176326430584,
"grad_norm": 0.34670495986938477,
"learning_rate": 7.964539645273204e-06,
"loss": 0.407070517539978,
"step": 653
},
{
"epoch": 0.6410979046685455,
"grad_norm": 0.2850871980190277,
"learning_rate": 7.958147453903773e-06,
"loss": 0.373299777507782,
"step": 654
},
{
"epoch": 0.6420781766940326,
"grad_norm": 0.29412031173706055,
"learning_rate": 7.951747815401651e-06,
"loss": 0.4365443289279938,
"step": 655
},
{
"epoch": 0.6430584487195197,
"grad_norm": 0.31543809175491333,
"learning_rate": 7.945340745877919e-06,
"loss": 0.45521238446235657,
"step": 656
},
{
"epoch": 0.6440387207450068,
"grad_norm": 0.3354298174381256,
"learning_rate": 7.938926261462366e-06,
"loss": 0.5221863985061646,
"step": 657
},
{
"epoch": 0.6450189927704938,
"grad_norm": 0.3284989893436432,
"learning_rate": 7.932504378303452e-06,
"loss": 0.5369656085968018,
"step": 658
},
{
"epoch": 0.6459992647959809,
"grad_norm": 0.3458614945411682,
"learning_rate": 7.92607511256826e-06,
"loss": 0.41837698221206665,
"step": 659
},
{
"epoch": 0.6469795368214679,
"grad_norm": 0.3532955050468445,
"learning_rate": 7.919638480442453e-06,
"loss": 0.44732317328453064,
"step": 660
},
{
"epoch": 0.647959808846955,
"grad_norm": 0.3655622601509094,
"learning_rate": 7.913194498130252e-06,
"loss": 0.4040653109550476,
"step": 661
},
{
"epoch": 0.6489400808724421,
"grad_norm": 0.8275192975997925,
"learning_rate": 7.90674318185437e-06,
"loss": 0.5069531798362732,
"step": 662
},
{
"epoch": 0.6499203528979292,
"grad_norm": 0.47497305274009705,
"learning_rate": 7.900284547855992e-06,
"loss": 0.4908417761325836,
"step": 663
},
{
"epoch": 0.6509006249234163,
"grad_norm": 0.4310746490955353,
"learning_rate": 7.893818612394717e-06,
"loss": 0.44629257917404175,
"step": 664
},
{
"epoch": 0.6518808969489033,
"grad_norm": 0.9688156843185425,
"learning_rate": 7.887345391748533e-06,
"loss": 0.44289934635162354,
"step": 665
},
{
"epoch": 0.6528611689743904,
"grad_norm": 0.5437262058258057,
"learning_rate": 7.880864902213765e-06,
"loss": 0.5355626940727234,
"step": 666
},
{
"epoch": 0.6538414409998775,
"grad_norm": 3.0646140575408936,
"learning_rate": 7.874377160105037e-06,
"loss": 0.43053728342056274,
"step": 667
},
{
"epoch": 0.6548217130253645,
"grad_norm": 0.9949020743370056,
"learning_rate": 7.86788218175523e-06,
"loss": 0.5058865547180176,
"step": 668
},
{
"epoch": 0.6558019850508516,
"grad_norm": 1.1058770418167114,
"learning_rate": 7.861379983515449e-06,
"loss": 0.5040953159332275,
"step": 669
},
{
"epoch": 0.6567822570763386,
"grad_norm": 0.5278763771057129,
"learning_rate": 7.854870581754967e-06,
"loss": 0.4105257987976074,
"step": 670
},
{
"epoch": 0.6577625291018258,
"grad_norm": 0.32923558354377747,
"learning_rate": 7.848353992861195e-06,
"loss": 0.46031761169433594,
"step": 671
},
{
"epoch": 0.6587428011273129,
"grad_norm": 0.2733277678489685,
"learning_rate": 7.841830233239638e-06,
"loss": 0.4199182987213135,
"step": 672
},
{
"epoch": 0.6597230731527999,
"grad_norm": 0.3369251787662506,
"learning_rate": 7.835299319313854e-06,
"loss": 0.4194727838039398,
"step": 673
},
{
"epoch": 0.660703345178287,
"grad_norm": 0.28543683886528015,
"learning_rate": 7.82876126752541e-06,
"loss": 0.4160378575325012,
"step": 674
},
{
"epoch": 0.661683617203774,
"grad_norm": 0.39933449029922485,
"learning_rate": 7.822216094333847e-06,
"loss": 0.38448667526245117,
"step": 675
},
{
"epoch": 0.6626638892292611,
"grad_norm": 0.37339338660240173,
"learning_rate": 7.815663816216626e-06,
"loss": 0.5100721120834351,
"step": 676
},
{
"epoch": 0.6636441612547482,
"grad_norm": 0.3124563992023468,
"learning_rate": 7.8091044496691e-06,
"loss": 0.4392182528972626,
"step": 677
},
{
"epoch": 0.6646244332802352,
"grad_norm": 0.4176885187625885,
"learning_rate": 7.80253801120447e-06,
"loss": 0.44882988929748535,
"step": 678
},
{
"epoch": 0.6656047053057224,
"grad_norm": 0.4605705738067627,
"learning_rate": 7.795964517353734e-06,
"loss": 0.4024580121040344,
"step": 679
},
{
"epoch": 0.6665849773312094,
"grad_norm": 0.35026490688323975,
"learning_rate": 7.789383984665658e-06,
"loss": 0.403725802898407,
"step": 680
},
{
"epoch": 0.6675652493566965,
"grad_norm": 0.4219362139701843,
"learning_rate": 7.782796429706721e-06,
"loss": 0.4156259596347809,
"step": 681
},
{
"epoch": 0.6685455213821836,
"grad_norm": 0.4736144244670868,
"learning_rate": 7.776201869061086e-06,
"loss": 0.48734694719314575,
"step": 682
},
{
"epoch": 0.6695257934076706,
"grad_norm": 0.30035045742988586,
"learning_rate": 7.769600319330553e-06,
"loss": 0.4142364263534546,
"step": 683
},
{
"epoch": 0.6705060654331577,
"grad_norm": 0.448886901140213,
"learning_rate": 7.762991797134513e-06,
"loss": 0.4432128071784973,
"step": 684
},
{
"epoch": 0.6714863374586447,
"grad_norm": 0.8460773229598999,
"learning_rate": 7.756376319109917e-06,
"loss": 0.431024432182312,
"step": 685
},
{
"epoch": 0.6724666094841318,
"grad_norm": 0.36831435561180115,
"learning_rate": 7.74975390191122e-06,
"loss": 0.4459499418735504,
"step": 686
},
{
"epoch": 0.673446881509619,
"grad_norm": 0.3132564127445221,
"learning_rate": 7.743124562210351e-06,
"loss": 0.5341706275939941,
"step": 687
},
{
"epoch": 0.674427153535106,
"grad_norm": 0.334623247385025,
"learning_rate": 7.736488316696663e-06,
"loss": 0.34817469120025635,
"step": 688
},
{
"epoch": 0.6754074255605931,
"grad_norm": 2.4007210731506348,
"learning_rate": 7.729845182076896e-06,
"loss": 0.5143555402755737,
"step": 689
},
{
"epoch": 0.6763876975860801,
"grad_norm": 0.46384498476982117,
"learning_rate": 7.723195175075136e-06,
"loss": 0.42345941066741943,
"step": 690
},
{
"epoch": 0.6773679696115672,
"grad_norm": 0.35396480560302734,
"learning_rate": 7.716538312432767e-06,
"loss": 0.45579758286476135,
"step": 691
},
{
"epoch": 0.6783482416370543,
"grad_norm": 0.3271118104457855,
"learning_rate": 7.70987461090843e-06,
"loss": 0.47848811745643616,
"step": 692
},
{
"epoch": 0.6793285136625413,
"grad_norm": 0.3389183580875397,
"learning_rate": 7.703204087277989e-06,
"loss": 0.4893483519554138,
"step": 693
},
{
"epoch": 0.6803087856880284,
"grad_norm": 0.3506358861923218,
"learning_rate": 7.696526758334477e-06,
"loss": 0.45978349447250366,
"step": 694
},
{
"epoch": 0.6812890577135156,
"grad_norm": 0.2780469059944153,
"learning_rate": 7.689842640888063e-06,
"loss": 0.3461371660232544,
"step": 695
},
{
"epoch": 0.6822693297390026,
"grad_norm": 0.9676852226257324,
"learning_rate": 7.683151751766005e-06,
"loss": 0.4354704022407532,
"step": 696
},
{
"epoch": 0.6832496017644897,
"grad_norm": 0.34818094968795776,
"learning_rate": 7.676454107812608e-06,
"loss": 0.5859825611114502,
"step": 697
},
{
"epoch": 0.6842298737899767,
"grad_norm": 0.3777989149093628,
"learning_rate": 7.669749725889182e-06,
"loss": 0.46829038858413696,
"step": 698
},
{
"epoch": 0.6852101458154638,
"grad_norm": 0.37987858057022095,
"learning_rate": 7.663038622873999e-06,
"loss": 0.48585087060928345,
"step": 699
},
{
"epoch": 0.6861904178409509,
"grad_norm": 0.5132353901863098,
"learning_rate": 7.656320815662257e-06,
"loss": 0.4944048523902893,
"step": 700
},
{
"epoch": 0.6871706898664379,
"grad_norm": 0.3148539960384369,
"learning_rate": 7.649596321166024e-06,
"loss": 0.42746883630752563,
"step": 701
},
{
"epoch": 0.688150961891925,
"grad_norm": 0.3091593384742737,
"learning_rate": 7.64286515631421e-06,
"loss": 0.3888963460922241,
"step": 702
},
{
"epoch": 0.689131233917412,
"grad_norm": 0.3705068528652191,
"learning_rate": 7.636127338052513e-06,
"loss": 0.48022907972335815,
"step": 703
},
{
"epoch": 0.6901115059428992,
"grad_norm": 0.5678703784942627,
"learning_rate": 7.629382883343381e-06,
"loss": 0.47102195024490356,
"step": 704
},
{
"epoch": 0.6910917779683863,
"grad_norm": 0.3253324627876282,
"learning_rate": 7.622631809165972e-06,
"loss": 0.32840976119041443,
"step": 705
},
{
"epoch": 0.6920720499938733,
"grad_norm": 0.30027759075164795,
"learning_rate": 7.615874132516108e-06,
"loss": 0.319786012172699,
"step": 706
},
{
"epoch": 0.6930523220193604,
"grad_norm": 0.358069509267807,
"learning_rate": 7.60910987040623e-06,
"loss": 0.39983999729156494,
"step": 707
},
{
"epoch": 0.6940325940448474,
"grad_norm": 0.5945205688476562,
"learning_rate": 7.602339039865362e-06,
"loss": 0.42683470249176025,
"step": 708
},
{
"epoch": 0.6950128660703345,
"grad_norm": 0.5269178748130798,
"learning_rate": 7.595561657939061e-06,
"loss": 0.43843698501586914,
"step": 709
},
{
"epoch": 0.6959931380958216,
"grad_norm": 1.1698843240737915,
"learning_rate": 7.588777741689376e-06,
"loss": 0.45243722200393677,
"step": 710
},
{
"epoch": 0.6969734101213086,
"grad_norm": 0.39858272671699524,
"learning_rate": 7.5819873081948105e-06,
"loss": 0.4786381721496582,
"step": 711
},
{
"epoch": 0.6979536821467958,
"grad_norm": 0.6274183988571167,
"learning_rate": 7.575190374550272e-06,
"loss": 0.4176969528198242,
"step": 712
},
{
"epoch": 0.6989339541722828,
"grad_norm": 0.3426187038421631,
"learning_rate": 7.568386957867033e-06,
"loss": 0.46479901671409607,
"step": 713
},
{
"epoch": 0.6999142261977699,
"grad_norm": 0.3482237458229065,
"learning_rate": 7.561577075272686e-06,
"loss": 0.48701000213623047,
"step": 714
},
{
"epoch": 0.700894498223257,
"grad_norm": 0.290752649307251,
"learning_rate": 7.554760743911104e-06,
"loss": 0.38350632786750793,
"step": 715
},
{
"epoch": 0.701874770248744,
"grad_norm": 0.32522499561309814,
"learning_rate": 7.5479379809423905e-06,
"loss": 0.45261746644973755,
"step": 716
},
{
"epoch": 0.7028550422742311,
"grad_norm": 0.4394126534461975,
"learning_rate": 7.541108803542846e-06,
"loss": 0.4414372444152832,
"step": 717
},
{
"epoch": 0.7038353142997181,
"grad_norm": 0.3642404079437256,
"learning_rate": 7.534273228904916e-06,
"loss": 0.3970096707344055,
"step": 718
},
{
"epoch": 0.7048155863252052,
"grad_norm": 0.4446166455745697,
"learning_rate": 7.527431274237149e-06,
"loss": 0.3591592311859131,
"step": 719
},
{
"epoch": 0.7057958583506924,
"grad_norm": 0.34355565905570984,
"learning_rate": 7.52058295676416e-06,
"loss": 0.4559618830680847,
"step": 720
},
{
"epoch": 0.7067761303761794,
"grad_norm": 0.397384375333786,
"learning_rate": 7.5137282937265796e-06,
"loss": 0.4717535972595215,
"step": 721
},
{
"epoch": 0.7077564024016665,
"grad_norm": 0.3389602303504944,
"learning_rate": 7.506867302381016e-06,
"loss": 0.41275453567504883,
"step": 722
},
{
"epoch": 0.7087366744271535,
"grad_norm": 0.4157849848270416,
"learning_rate": 7.500000000000001e-06,
"loss": 0.38644737005233765,
"step": 723
},
{
"epoch": 0.7097169464526406,
"grad_norm": 0.3657962679862976,
"learning_rate": 7.493126403871964e-06,
"loss": 0.3874867558479309,
"step": 724
},
{
"epoch": 0.7106972184781277,
"grad_norm": 0.44123581051826477,
"learning_rate": 7.486246531301178e-06,
"loss": 0.5014352798461914,
"step": 725
},
{
"epoch": 0.7116774905036147,
"grad_norm": 0.3996449112892151,
"learning_rate": 7.479360399607707e-06,
"loss": 0.41623663902282715,
"step": 726
},
{
"epoch": 0.7126577625291018,
"grad_norm": 0.3702757954597473,
"learning_rate": 7.472468026127385e-06,
"loss": 0.4651917815208435,
"step": 727
},
{
"epoch": 0.7136380345545889,
"grad_norm": 0.4727034568786621,
"learning_rate": 7.465569428211752e-06,
"loss": 0.4840661287307739,
"step": 728
},
{
"epoch": 0.714618306580076,
"grad_norm": 0.4097318947315216,
"learning_rate": 7.45866462322802e-06,
"loss": 0.4175463616847992,
"step": 729
},
{
"epoch": 0.7155985786055631,
"grad_norm": 0.4900283217430115,
"learning_rate": 7.451753628559027e-06,
"loss": 0.37979835271835327,
"step": 730
},
{
"epoch": 0.7165788506310501,
"grad_norm": 1.1253916025161743,
"learning_rate": 7.444836461603195e-06,
"loss": 0.5262601375579834,
"step": 731
},
{
"epoch": 0.7175591226565372,
"grad_norm": 0.32169947028160095,
"learning_rate": 7.437913139774482e-06,
"loss": 0.39289093017578125,
"step": 732
},
{
"epoch": 0.7185393946820242,
"grad_norm": 0.2856983244419098,
"learning_rate": 7.430983680502344e-06,
"loss": 0.3717386722564697,
"step": 733
},
{
"epoch": 0.7195196667075113,
"grad_norm": 0.5559130907058716,
"learning_rate": 7.424048101231687e-06,
"loss": 0.4056721031665802,
"step": 734
},
{
"epoch": 0.7204999387329984,
"grad_norm": 0.3144131302833557,
"learning_rate": 7.4171064194228196e-06,
"loss": 0.3691323399543762,
"step": 735
},
{
"epoch": 0.7214802107584855,
"grad_norm": 0.31543290615081787,
"learning_rate": 7.410158652551421e-06,
"loss": 0.4696192145347595,
"step": 736
},
{
"epoch": 0.7224604827839726,
"grad_norm": 0.34641218185424805,
"learning_rate": 7.403204818108487e-06,
"loss": 0.4930548667907715,
"step": 737
},
{
"epoch": 0.7234407548094596,
"grad_norm": 0.33608800172805786,
"learning_rate": 7.396244933600285e-06,
"loss": 0.47634634375572205,
"step": 738
},
{
"epoch": 0.7244210268349467,
"grad_norm": 0.46371525526046753,
"learning_rate": 7.3892790165483164e-06,
"loss": 0.446674108505249,
"step": 739
},
{
"epoch": 0.7254012988604338,
"grad_norm": 0.4306736886501312,
"learning_rate": 7.382307084489269e-06,
"loss": 0.45234882831573486,
"step": 740
},
{
"epoch": 0.7263815708859208,
"grad_norm": 0.4029739797115326,
"learning_rate": 7.3753291549749764e-06,
"loss": 0.4558042287826538,
"step": 741
},
{
"epoch": 0.7273618429114079,
"grad_norm": 0.348371297121048,
"learning_rate": 7.368345245572362e-06,
"loss": 0.35345882177352905,
"step": 742
},
{
"epoch": 0.7283421149368949,
"grad_norm": 0.5065837502479553,
"learning_rate": 7.361355373863415e-06,
"loss": 0.3955862820148468,
"step": 743
},
{
"epoch": 0.729322386962382,
"grad_norm": 0.6376745700836182,
"learning_rate": 7.354359557445125e-06,
"loss": 0.47946324944496155,
"step": 744
},
{
"epoch": 0.7303026589878692,
"grad_norm": 0.31279855966567993,
"learning_rate": 7.347357813929455e-06,
"loss": 0.5491936206817627,
"step": 745
},
{
"epoch": 0.7312829310133562,
"grad_norm": 0.264043927192688,
"learning_rate": 7.3403501609432835e-06,
"loss": 0.44447633624076843,
"step": 746
},
{
"epoch": 0.7322632030388433,
"grad_norm": 0.35330650210380554,
"learning_rate": 7.333336616128369e-06,
"loss": 0.45426321029663086,
"step": 747
},
{
"epoch": 0.7332434750643303,
"grad_norm": 0.5701055526733398,
"learning_rate": 7.326317197141304e-06,
"loss": 0.5213590264320374,
"step": 748
},
{
"epoch": 0.7342237470898174,
"grad_norm": 0.3818702697753906,
"learning_rate": 7.319291921653464e-06,
"loss": 0.4960825443267822,
"step": 749
},
{
"epoch": 0.7352040191153045,
"grad_norm": 0.33272963762283325,
"learning_rate": 7.312260807350975e-06,
"loss": 0.46506649255752563,
"step": 750
},
{
"epoch": 0.7361842911407915,
"grad_norm": 0.373261034488678,
"learning_rate": 7.305223871934657e-06,
"loss": 0.4264051020145416,
"step": 751
},
{
"epoch": 0.7371645631662787,
"grad_norm": 0.36157459020614624,
"learning_rate": 7.298181133119987e-06,
"loss": 0.40642160177230835,
"step": 752
},
{
"epoch": 0.7381448351917658,
"grad_norm": 0.3849131762981415,
"learning_rate": 7.291132608637053e-06,
"loss": 0.4864940047264099,
"step": 753
},
{
"epoch": 0.7391251072172528,
"grad_norm": 0.29483067989349365,
"learning_rate": 7.284078316230504e-06,
"loss": 0.39711129665374756,
"step": 754
},
{
"epoch": 0.7401053792427399,
"grad_norm": 0.3704259395599365,
"learning_rate": 7.2770182736595164e-06,
"loss": 0.3901249170303345,
"step": 755
},
{
"epoch": 0.7410856512682269,
"grad_norm": 0.33157676458358765,
"learning_rate": 7.269952498697734e-06,
"loss": 0.4264463186264038,
"step": 756
},
{
"epoch": 0.742065923293714,
"grad_norm": 0.43318718671798706,
"learning_rate": 7.262881009133242e-06,
"loss": 0.5198872089385986,
"step": 757
},
{
"epoch": 0.7430461953192011,
"grad_norm": 0.6015837788581848,
"learning_rate": 7.255803822768504e-06,
"loss": 0.4003225862979889,
"step": 758
},
{
"epoch": 0.7440264673446881,
"grad_norm": 0.45841240882873535,
"learning_rate": 7.24872095742033e-06,
"loss": 0.4293782413005829,
"step": 759
},
{
"epoch": 0.7450067393701753,
"grad_norm": 0.354641854763031,
"learning_rate": 7.241632430919823e-06,
"loss": 0.4521189033985138,
"step": 760
},
{
"epoch": 0.7459870113956623,
"grad_norm": 0.7112554907798767,
"learning_rate": 7.234538261112342e-06,
"loss": 0.35923296213150024,
"step": 761
},
{
"epoch": 0.7469672834211494,
"grad_norm": 0.34490618109703064,
"learning_rate": 7.227438465857448e-06,
"loss": 0.49387502670288086,
"step": 762
},
{
"epoch": 0.7479475554466365,
"grad_norm": 0.29901543259620667,
"learning_rate": 7.2203330630288714e-06,
"loss": 0.36354517936706543,
"step": 763
},
{
"epoch": 0.7489278274721235,
"grad_norm": 0.364378422498703,
"learning_rate": 7.213222070514453e-06,
"loss": 0.5775377750396729,
"step": 764
},
{
"epoch": 0.7499080994976106,
"grad_norm": 0.3677915036678314,
"learning_rate": 7.206105506216107e-06,
"loss": 0.4223652780056,
"step": 765
},
{
"epoch": 0.7508883715230976,
"grad_norm": 0.4173823297023773,
"learning_rate": 7.198983388049779e-06,
"loss": 0.4367346167564392,
"step": 766
},
{
"epoch": 0.7518686435485847,
"grad_norm": 0.29759371280670166,
"learning_rate": 7.191855733945388e-06,
"loss": 0.4289306402206421,
"step": 767
},
{
"epoch": 0.7528489155740719,
"grad_norm": 0.3195274770259857,
"learning_rate": 7.1847225618467975e-06,
"loss": 0.3497508764266968,
"step": 768
},
{
"epoch": 0.7538291875995589,
"grad_norm": 0.29872044920921326,
"learning_rate": 7.177583889711763e-06,
"loss": 0.4069169759750366,
"step": 769
},
{
"epoch": 0.754809459625046,
"grad_norm": 0.3418881297111511,
"learning_rate": 7.170439735511876e-06,
"loss": 0.47802841663360596,
"step": 770
},
{
"epoch": 0.755789731650533,
"grad_norm": 0.2910369038581848,
"learning_rate": 7.163290117232542e-06,
"loss": 0.41970184445381165,
"step": 771
},
{
"epoch": 0.7567700036760201,
"grad_norm": 0.4719921350479126,
"learning_rate": 7.1561350528729145e-06,
"loss": 0.4587634801864624,
"step": 772
},
{
"epoch": 0.7577502757015072,
"grad_norm": 0.30202654004096985,
"learning_rate": 7.148974560445859e-06,
"loss": 0.33863359689712524,
"step": 773
},
{
"epoch": 0.7587305477269942,
"grad_norm": 0.29719576239585876,
"learning_rate": 7.1418086579779075e-06,
"loss": 0.40096843242645264,
"step": 774
},
{
"epoch": 0.7597108197524813,
"grad_norm": 0.34799328446388245,
"learning_rate": 7.1346373635092095e-06,
"loss": 0.4639182686805725,
"step": 775
},
{
"epoch": 0.7606910917779683,
"grad_norm": 0.4758872985839844,
"learning_rate": 7.127460695093493e-06,
"loss": 0.4926512539386749,
"step": 776
},
{
"epoch": 0.7616713638034555,
"grad_norm": 0.37746936082839966,
"learning_rate": 7.12027867079801e-06,
"loss": 0.4766829013824463,
"step": 777
},
{
"epoch": 0.7626516358289426,
"grad_norm": 0.5357100963592529,
"learning_rate": 7.113091308703498e-06,
"loss": 0.4351813495159149,
"step": 778
},
{
"epoch": 0.7636319078544296,
"grad_norm": 0.42104777693748474,
"learning_rate": 7.105898626904134e-06,
"loss": 0.44934508204460144,
"step": 779
},
{
"epoch": 0.7646121798799167,
"grad_norm": 0.3015410304069519,
"learning_rate": 7.098700643507485e-06,
"loss": 0.4699820876121521,
"step": 780
},
{
"epoch": 0.7655924519054037,
"grad_norm": 0.31635135412216187,
"learning_rate": 7.0914973766344645e-06,
"loss": 0.3651800751686096,
"step": 781
},
{
"epoch": 0.7665727239308908,
"grad_norm": 0.3795527517795563,
"learning_rate": 7.0842888444192905e-06,
"loss": 0.5565624237060547,
"step": 782
},
{
"epoch": 0.7675529959563779,
"grad_norm": 0.3574817478656769,
"learning_rate": 7.0770750650094335e-06,
"loss": 0.45820948481559753,
"step": 783
},
{
"epoch": 0.7685332679818649,
"grad_norm": 0.3084055781364441,
"learning_rate": 7.069856056565573e-06,
"loss": 0.3995293378829956,
"step": 784
},
{
"epoch": 0.7695135400073521,
"grad_norm": 0.31401678919792175,
"learning_rate": 7.062631837261556e-06,
"loss": 0.42256277799606323,
"step": 785
},
{
"epoch": 0.7704938120328391,
"grad_norm": 0.27648094296455383,
"learning_rate": 7.055402425284346e-06,
"loss": 0.37582093477249146,
"step": 786
},
{
"epoch": 0.7714740840583262,
"grad_norm": 0.39012810587882996,
"learning_rate": 7.048167838833977e-06,
"loss": 0.46207207441329956,
"step": 787
},
{
"epoch": 0.7724543560838133,
"grad_norm": 0.3592337965965271,
"learning_rate": 7.040928096123516e-06,
"loss": 0.42086684703826904,
"step": 788
},
{
"epoch": 0.7734346281093003,
"grad_norm": 0.416060209274292,
"learning_rate": 7.033683215379002e-06,
"loss": 0.4350088834762573,
"step": 789
},
{
"epoch": 0.7744149001347874,
"grad_norm": 0.41734376549720764,
"learning_rate": 7.026433214839416e-06,
"loss": 0.43400758504867554,
"step": 790
},
{
"epoch": 0.7753951721602744,
"grad_norm": 0.32856181263923645,
"learning_rate": 7.019178112756625e-06,
"loss": 0.3643208146095276,
"step": 791
},
{
"epoch": 0.7763754441857615,
"grad_norm": 0.2868272066116333,
"learning_rate": 7.011917927395341e-06,
"loss": 0.4030511975288391,
"step": 792
},
{
"epoch": 0.7773557162112487,
"grad_norm": 0.29374998807907104,
"learning_rate": 7.004652677033069e-06,
"loss": 0.4327206611633301,
"step": 793
},
{
"epoch": 0.7783359882367357,
"grad_norm": 0.4732438027858734,
"learning_rate": 6.997382379960072e-06,
"loss": 0.44608888030052185,
"step": 794
},
{
"epoch": 0.7793162602622228,
"grad_norm": 0.3050212860107422,
"learning_rate": 6.990107054479313e-06,
"loss": 0.41696059703826904,
"step": 795
},
{
"epoch": 0.7802965322877098,
"grad_norm": 0.2738250195980072,
"learning_rate": 6.9828267189064125e-06,
"loss": 0.44417643547058105,
"step": 796
},
{
"epoch": 0.7812768043131969,
"grad_norm": 0.35405758023262024,
"learning_rate": 6.9755413915696105e-06,
"loss": 0.48841744661331177,
"step": 797
},
{
"epoch": 0.782257076338684,
"grad_norm": 0.3926083445549011,
"learning_rate": 6.968251090809708e-06,
"loss": 0.34732434153556824,
"step": 798
},
{
"epoch": 0.783237348364171,
"grad_norm": 0.6493455171585083,
"learning_rate": 6.960955834980028e-06,
"loss": 0.41161495447158813,
"step": 799
},
{
"epoch": 0.7842176203896581,
"grad_norm": 0.38331207633018494,
"learning_rate": 6.953655642446368e-06,
"loss": 0.48153337836265564,
"step": 800
},
{
"epoch": 0.7851978924151451,
"grad_norm": 0.3462029993534088,
"learning_rate": 6.946350531586959e-06,
"loss": 0.39161407947540283,
"step": 801
},
{
"epoch": 0.7861781644406323,
"grad_norm": 0.29707103967666626,
"learning_rate": 6.9390405207924014e-06,
"loss": 0.43831056356430054,
"step": 802
},
{
"epoch": 0.7871584364661194,
"grad_norm": 0.34447264671325684,
"learning_rate": 6.931725628465643e-06,
"loss": 0.3981401324272156,
"step": 803
},
{
"epoch": 0.7881387084916064,
"grad_norm": 0.38846179842948914,
"learning_rate": 6.9244058730219176e-06,
"loss": 0.48101720213890076,
"step": 804
},
{
"epoch": 0.7891189805170935,
"grad_norm": 0.22997339069843292,
"learning_rate": 6.917081272888697e-06,
"loss": 0.30385780334472656,
"step": 805
},
{
"epoch": 0.7900992525425806,
"grad_norm": 0.3154858648777008,
"learning_rate": 6.909751846505657e-06,
"loss": 0.36327075958251953,
"step": 806
},
{
"epoch": 0.7910795245680676,
"grad_norm": 0.42443451285362244,
"learning_rate": 6.902417612324615e-06,
"loss": 0.45198142528533936,
"step": 807
},
{
"epoch": 0.7920597965935547,
"grad_norm": 0.4964180588722229,
"learning_rate": 6.895078588809503e-06,
"loss": 0.42241811752319336,
"step": 808
},
{
"epoch": 0.7930400686190417,
"grad_norm": 0.2849682569503784,
"learning_rate": 6.887734794436301e-06,
"loss": 0.3411692976951599,
"step": 809
},
{
"epoch": 0.7940203406445289,
"grad_norm": 0.322028785943985,
"learning_rate": 6.880386247692999e-06,
"loss": 0.35893434286117554,
"step": 810
},
{
"epoch": 0.795000612670016,
"grad_norm": 0.283246248960495,
"learning_rate": 6.873032967079562e-06,
"loss": 0.39699044823646545,
"step": 811
},
{
"epoch": 0.795980884695503,
"grad_norm": 0.33114364743232727,
"learning_rate": 6.865674971107859e-06,
"loss": 0.4816048741340637,
"step": 812
},
{
"epoch": 0.7969611567209901,
"grad_norm": 0.3033636510372162,
"learning_rate": 6.858312278301638e-06,
"loss": 0.4470120668411255,
"step": 813
},
{
"epoch": 0.7979414287464771,
"grad_norm": 0.5208700895309448,
"learning_rate": 6.850944907196472e-06,
"loss": 0.399263858795166,
"step": 814
},
{
"epoch": 0.7989217007719642,
"grad_norm": 0.31744447350502014,
"learning_rate": 6.8435728763397045e-06,
"loss": 0.3862346410751343,
"step": 815
},
{
"epoch": 0.7999019727974513,
"grad_norm": 0.3611550033092499,
"learning_rate": 6.836196204290417e-06,
"loss": 0.4473588466644287,
"step": 816
},
{
"epoch": 0.8008822448229383,
"grad_norm": 0.38386568427085876,
"learning_rate": 6.828814909619374e-06,
"loss": 0.4471675157546997,
"step": 817
},
{
"epoch": 0.8018625168484255,
"grad_norm": 0.2766851484775543,
"learning_rate": 6.821429010908972e-06,
"loss": 0.39848029613494873,
"step": 818
},
{
"epoch": 0.8028427888739125,
"grad_norm": 0.33408650755882263,
"learning_rate": 6.814038526753205e-06,
"loss": 0.43641579151153564,
"step": 819
},
{
"epoch": 0.8038230608993996,
"grad_norm": 0.5727699995040894,
"learning_rate": 6.806643475757608e-06,
"loss": 0.38241153955459595,
"step": 820
},
{
"epoch": 0.8048033329248867,
"grad_norm": 0.2858707308769226,
"learning_rate": 6.799243876539213e-06,
"loss": 0.4386863708496094,
"step": 821
},
{
"epoch": 0.8057836049503737,
"grad_norm": 0.3390565812587738,
"learning_rate": 6.7918397477265e-06,
"loss": 0.49160879850387573,
"step": 822
},
{
"epoch": 0.8067638769758608,
"grad_norm": 0.3348511755466461,
"learning_rate": 6.78443110795936e-06,
"loss": 0.40226948261260986,
"step": 823
},
{
"epoch": 0.8077441490013478,
"grad_norm": 0.34619420766830444,
"learning_rate": 6.77701797588903e-06,
"loss": 0.40227335691452026,
"step": 824
},
{
"epoch": 0.808724421026835,
"grad_norm": 0.39033299684524536,
"learning_rate": 6.76960037017806e-06,
"loss": 0.3632168471813202,
"step": 825
},
{
"epoch": 0.8097046930523221,
"grad_norm": 0.4732474684715271,
"learning_rate": 6.762178309500267e-06,
"loss": 0.48236578702926636,
"step": 826
},
{
"epoch": 0.8106849650778091,
"grad_norm": 0.32661864161491394,
"learning_rate": 6.75475181254068e-06,
"loss": 0.4657779932022095,
"step": 827
},
{
"epoch": 0.8116652371032962,
"grad_norm": 0.3059816360473633,
"learning_rate": 6.747320897995493e-06,
"loss": 0.4491598606109619,
"step": 828
},
{
"epoch": 0.8126455091287832,
"grad_norm": 0.331331342458725,
"learning_rate": 6.739885584572026e-06,
"loss": 0.4129459857940674,
"step": 829
},
{
"epoch": 0.8136257811542703,
"grad_norm": 0.31637993454933167,
"learning_rate": 6.732445890988675e-06,
"loss": 0.547048807144165,
"step": 830
},
{
"epoch": 0.8146060531797574,
"grad_norm": 0.6699824929237366,
"learning_rate": 6.725001835974854e-06,
"loss": 0.43884965777397156,
"step": 831
},
{
"epoch": 0.8155863252052444,
"grad_norm": 0.4328043460845947,
"learning_rate": 6.717553438270966e-06,
"loss": 0.36510759592056274,
"step": 832
},
{
"epoch": 0.8165665972307316,
"grad_norm": 0.29438063502311707,
"learning_rate": 6.710100716628345e-06,
"loss": 0.4367470145225525,
"step": 833
},
{
"epoch": 0.8175468692562186,
"grad_norm": 0.28324782848358154,
"learning_rate": 6.702643689809205e-06,
"loss": 0.34590137004852295,
"step": 834
},
{
"epoch": 0.8185271412817057,
"grad_norm": 0.4410516917705536,
"learning_rate": 6.695182376586603e-06,
"loss": 0.4005008339881897,
"step": 835
},
{
"epoch": 0.8195074133071928,
"grad_norm": 0.4737011194229126,
"learning_rate": 6.68771679574439e-06,
"loss": 0.49619442224502563,
"step": 836
},
{
"epoch": 0.8204876853326798,
"grad_norm": 0.33376845717430115,
"learning_rate": 6.680246966077151e-06,
"loss": 0.43638962507247925,
"step": 837
},
{
"epoch": 0.8214679573581669,
"grad_norm": 0.31139299273490906,
"learning_rate": 6.672772906390177e-06,
"loss": 0.41185882687568665,
"step": 838
},
{
"epoch": 0.8224482293836539,
"grad_norm": 0.444981187582016,
"learning_rate": 6.665294635499404e-06,
"loss": 0.3704143166542053,
"step": 839
},
{
"epoch": 0.823428501409141,
"grad_norm": 0.37989741563796997,
"learning_rate": 6.65781217223137e-06,
"loss": 0.4551524817943573,
"step": 840
},
{
"epoch": 0.8244087734346282,
"grad_norm": 0.810689389705658,
"learning_rate": 6.650325535423166e-06,
"loss": 0.5108350515365601,
"step": 841
},
{
"epoch": 0.8253890454601152,
"grad_norm": 0.3536842465400696,
"learning_rate": 6.642834743922393e-06,
"loss": 0.5772078037261963,
"step": 842
},
{
"epoch": 0.8263693174856023,
"grad_norm": 0.31637224555015564,
"learning_rate": 6.635339816587109e-06,
"loss": 0.42627257108688354,
"step": 843
},
{
"epoch": 0.8273495895110893,
"grad_norm": 0.34329280257225037,
"learning_rate": 6.627840772285784e-06,
"loss": 0.4413832724094391,
"step": 844
},
{
"epoch": 0.8283298615365764,
"grad_norm": 0.28939110040664673,
"learning_rate": 6.6203376298972535e-06,
"loss": 0.43398407101631165,
"step": 845
},
{
"epoch": 0.8293101335620635,
"grad_norm": 0.5527605414390564,
"learning_rate": 6.612830408310671e-06,
"loss": 0.46303272247314453,
"step": 846
},
{
"epoch": 0.8302904055875505,
"grad_norm": 0.3042815923690796,
"learning_rate": 6.605319126425455e-06,
"loss": 0.4387814700603485,
"step": 847
},
{
"epoch": 0.8312706776130376,
"grad_norm": 0.3447803556919098,
"learning_rate": 6.5978038031512505e-06,
"loss": 0.34531593322753906,
"step": 848
},
{
"epoch": 0.8322509496385246,
"grad_norm": 0.3752855658531189,
"learning_rate": 6.590284457407876e-06,
"loss": 0.5217232704162598,
"step": 849
},
{
"epoch": 0.8332312216640118,
"grad_norm": 0.3138746917247772,
"learning_rate": 6.582761108125272e-06,
"loss": 0.37713712453842163,
"step": 850
},
{
"epoch": 0.8342114936894989,
"grad_norm": 0.313753604888916,
"learning_rate": 6.5752337742434644e-06,
"loss": 0.36168724298477173,
"step": 851
},
{
"epoch": 0.8351917657149859,
"grad_norm": 0.5864412188529968,
"learning_rate": 6.567702474712507e-06,
"loss": 0.45619457960128784,
"step": 852
},
{
"epoch": 0.836172037740473,
"grad_norm": 0.33605337142944336,
"learning_rate": 6.560167228492436e-06,
"loss": 0.40912872552871704,
"step": 853
},
{
"epoch": 0.83715230976596,
"grad_norm": 0.32959526777267456,
"learning_rate": 6.552628054553225e-06,
"loss": 0.3917942941188812,
"step": 854
},
{
"epoch": 0.8381325817914471,
"grad_norm": 0.3304107189178467,
"learning_rate": 6.545084971874738e-06,
"loss": 0.4579445719718933,
"step": 855
},
{
"epoch": 0.8391128538169342,
"grad_norm": 0.31899625062942505,
"learning_rate": 6.537537999446673e-06,
"loss": 0.4087638854980469,
"step": 856
},
{
"epoch": 0.8400931258424212,
"grad_norm": 2.0095064640045166,
"learning_rate": 6.529987156268527e-06,
"loss": 0.4091487228870392,
"step": 857
},
{
"epoch": 0.8410733978679084,
"grad_norm": 0.6623659133911133,
"learning_rate": 6.522432461349536e-06,
"loss": 0.4094298481941223,
"step": 858
},
{
"epoch": 0.8420536698933954,
"grad_norm": 0.4556414783000946,
"learning_rate": 6.514873933708637e-06,
"loss": 0.527363121509552,
"step": 859
},
{
"epoch": 0.8430339419188825,
"grad_norm": 0.3485345244407654,
"learning_rate": 6.507311592374415e-06,
"loss": 0.4396172761917114,
"step": 860
},
{
"epoch": 0.8440142139443696,
"grad_norm": 0.6599644422531128,
"learning_rate": 6.499745456385054e-06,
"loss": 0.4917178153991699,
"step": 861
},
{
"epoch": 0.8449944859698566,
"grad_norm": 0.3528061807155609,
"learning_rate": 6.4921755447882926e-06,
"loss": 0.4211848974227905,
"step": 862
},
{
"epoch": 0.8459747579953437,
"grad_norm": 0.4127756357192993,
"learning_rate": 6.484601876641375e-06,
"loss": 0.4486020803451538,
"step": 863
},
{
"epoch": 0.8469550300208308,
"grad_norm": 0.3647272288799286,
"learning_rate": 6.477024471011001e-06,
"loss": 0.40438419580459595,
"step": 864
},
{
"epoch": 0.8479353020463178,
"grad_norm": 1.1576286554336548,
"learning_rate": 6.469443346973281e-06,
"loss": 0.41841182112693787,
"step": 865
},
{
"epoch": 0.848915574071805,
"grad_norm": 0.43161657452583313,
"learning_rate": 6.461858523613684e-06,
"loss": 0.4205344319343567,
"step": 866
},
{
"epoch": 0.849895846097292,
"grad_norm": 0.4292677044868469,
"learning_rate": 6.454270020026996e-06,
"loss": 0.47648191452026367,
"step": 867
},
{
"epoch": 0.8508761181227791,
"grad_norm": 0.27494698762893677,
"learning_rate": 6.446677855317265e-06,
"loss": 0.3918209373950958,
"step": 868
},
{
"epoch": 0.8518563901482662,
"grad_norm": 0.33009472489356995,
"learning_rate": 6.439082048597755e-06,
"loss": 0.39804190397262573,
"step": 869
},
{
"epoch": 0.8528366621737532,
"grad_norm": 0.3878440260887146,
"learning_rate": 6.431482618990902e-06,
"loss": 0.39374232292175293,
"step": 870
},
{
"epoch": 0.8538169341992403,
"grad_norm": 0.39591407775878906,
"learning_rate": 6.423879585628262e-06,
"loss": 0.49599969387054443,
"step": 871
},
{
"epoch": 0.8547972062247273,
"grad_norm": 0.3894732594490051,
"learning_rate": 6.41627296765046e-06,
"loss": 0.5123536586761475,
"step": 872
},
{
"epoch": 0.8557774782502144,
"grad_norm": 0.32534563541412354,
"learning_rate": 6.408662784207149e-06,
"loss": 0.4203255772590637,
"step": 873
},
{
"epoch": 0.8567577502757016,
"grad_norm": 0.438992977142334,
"learning_rate": 6.401049054456957e-06,
"loss": 0.5640689134597778,
"step": 874
},
{
"epoch": 0.8577380223011886,
"grad_norm": 0.2719486653804779,
"learning_rate": 6.39343179756744e-06,
"loss": 0.3920605182647705,
"step": 875
},
{
"epoch": 0.8587182943266757,
"grad_norm": 0.4246402382850647,
"learning_rate": 6.385811032715031e-06,
"loss": 0.37895235419273376,
"step": 876
},
{
"epoch": 0.8596985663521627,
"grad_norm": 0.45037129521369934,
"learning_rate": 6.378186779084996e-06,
"loss": 0.4424005448818207,
"step": 877
},
{
"epoch": 0.8606788383776498,
"grad_norm": 0.574342668056488,
"learning_rate": 6.370559055871389e-06,
"loss": 0.48311787843704224,
"step": 878
},
{
"epoch": 0.8616591104031369,
"grad_norm": 0.4536021053791046,
"learning_rate": 6.362927882276991e-06,
"loss": 0.5810052156448364,
"step": 879
},
{
"epoch": 0.8626393824286239,
"grad_norm": 0.26062390208244324,
"learning_rate": 6.355293277513271e-06,
"loss": 0.3816462457180023,
"step": 880
},
{
"epoch": 0.863619654454111,
"grad_norm": 0.2729369103908539,
"learning_rate": 6.34765526080034e-06,
"loss": 0.3641537129878998,
"step": 881
},
{
"epoch": 0.864599926479598,
"grad_norm": 0.4085536301136017,
"learning_rate": 6.3400138513668955e-06,
"loss": 0.3604713976383209,
"step": 882
},
{
"epoch": 0.8655801985050852,
"grad_norm": 0.4680600166320801,
"learning_rate": 6.332369068450175e-06,
"loss": 0.4618578553199768,
"step": 883
},
{
"epoch": 0.8665604705305723,
"grad_norm": 0.25370433926582336,
"learning_rate": 6.324720931295913e-06,
"loss": 0.3337012529373169,
"step": 884
},
{
"epoch": 0.8675407425560593,
"grad_norm": 0.4016103744506836,
"learning_rate": 6.317069459158284e-06,
"loss": 0.4735439121723175,
"step": 885
},
{
"epoch": 0.8685210145815464,
"grad_norm": 0.26672765612602234,
"learning_rate": 6.3094146712998595e-06,
"loss": 0.44321590662002563,
"step": 886
},
{
"epoch": 0.8695012866070334,
"grad_norm": 0.29990482330322266,
"learning_rate": 6.301756586991561e-06,
"loss": 0.44410741329193115,
"step": 887
},
{
"epoch": 0.8704815586325205,
"grad_norm": 0.8938698172569275,
"learning_rate": 6.294095225512604e-06,
"loss": 0.4193858802318573,
"step": 888
},
{
"epoch": 0.8714618306580076,
"grad_norm": 0.2698467969894409,
"learning_rate": 6.286430606150458e-06,
"loss": 0.4426466226577759,
"step": 889
},
{
"epoch": 0.8724421026834946,
"grad_norm": 0.3382076323032379,
"learning_rate": 6.278762748200795e-06,
"loss": 0.37501290440559387,
"step": 890
},
{
"epoch": 0.8734223747089818,
"grad_norm": 0.3385145664215088,
"learning_rate": 6.271091670967437e-06,
"loss": 0.43820109963417053,
"step": 891
},
{
"epoch": 0.8744026467344688,
"grad_norm": 0.3152020573616028,
"learning_rate": 6.2634173937623075e-06,
"loss": 0.37993699312210083,
"step": 892
},
{
"epoch": 0.8753829187599559,
"grad_norm": 1.5932375192642212,
"learning_rate": 6.255739935905396e-06,
"loss": 0.4053550362586975,
"step": 893
},
{
"epoch": 0.876363190785443,
"grad_norm": 0.5584542155265808,
"learning_rate": 6.248059316724693e-06,
"loss": 0.35968148708343506,
"step": 894
},
{
"epoch": 0.87734346281093,
"grad_norm": 0.39956364035606384,
"learning_rate": 6.240375555556145e-06,
"loss": 0.4879263639450073,
"step": 895
},
{
"epoch": 0.8783237348364171,
"grad_norm": 0.3270205557346344,
"learning_rate": 6.232688671743613e-06,
"loss": 0.3840183615684509,
"step": 896
},
{
"epoch": 0.8793040068619041,
"grad_norm": 0.3887234330177307,
"learning_rate": 6.22499868463882e-06,
"loss": 0.44902151823043823,
"step": 897
},
{
"epoch": 0.8802842788873912,
"grad_norm": 0.3597409725189209,
"learning_rate": 6.217305613601296e-06,
"loss": 0.5107446908950806,
"step": 898
},
{
"epoch": 0.8812645509128784,
"grad_norm": 0.5276545286178589,
"learning_rate": 6.209609477998339e-06,
"loss": 0.4594070017337799,
"step": 899
},
{
"epoch": 0.8822448229383654,
"grad_norm": 1.396403431892395,
"learning_rate": 6.201910297204963e-06,
"loss": 0.3972088396549225,
"step": 900
},
{
"epoch": 0.8832250949638525,
"grad_norm": 0.3442447781562805,
"learning_rate": 6.194208090603845e-06,
"loss": 0.4140443205833435,
"step": 901
},
{
"epoch": 0.8842053669893395,
"grad_norm": 0.3070739507675171,
"learning_rate": 6.186502877585281e-06,
"loss": 0.37739214301109314,
"step": 902
},
{
"epoch": 0.8851856390148266,
"grad_norm": 0.3227234184741974,
"learning_rate": 6.178794677547138e-06,
"loss": 0.4161836504936218,
"step": 903
},
{
"epoch": 0.8861659110403137,
"grad_norm": 0.3102916181087494,
"learning_rate": 6.171083509894795e-06,
"loss": 0.38741618394851685,
"step": 904
},
{
"epoch": 0.8871461830658007,
"grad_norm": 0.3296877145767212,
"learning_rate": 6.163369394041112e-06,
"loss": 0.3972170054912567,
"step": 905
},
{
"epoch": 0.8881264550912878,
"grad_norm": 0.33786922693252563,
"learning_rate": 6.155652349406366e-06,
"loss": 0.4064939320087433,
"step": 906
},
{
"epoch": 0.8891067271167749,
"grad_norm": 0.3524640202522278,
"learning_rate": 6.1479323954182055e-06,
"loss": 0.40416407585144043,
"step": 907
},
{
"epoch": 0.890086999142262,
"grad_norm": 0.4218555986881256,
"learning_rate": 6.140209551511609e-06,
"loss": 0.46490415930747986,
"step": 908
},
{
"epoch": 0.8910672711677491,
"grad_norm": 0.7574635148048401,
"learning_rate": 6.132483837128823e-06,
"loss": 0.3996220827102661,
"step": 909
},
{
"epoch": 0.8920475431932361,
"grad_norm": 0.320833295583725,
"learning_rate": 6.124755271719326e-06,
"loss": 0.5552396774291992,
"step": 910
},
{
"epoch": 0.8930278152187232,
"grad_norm": 0.6593910455703735,
"learning_rate": 6.1170238747397715e-06,
"loss": 0.4720093905925751,
"step": 911
},
{
"epoch": 0.8940080872442102,
"grad_norm": 0.33652421832084656,
"learning_rate": 6.109289665653944e-06,
"loss": 0.4044809341430664,
"step": 912
},
{
"epoch": 0.8949883592696973,
"grad_norm": 0.37871572375297546,
"learning_rate": 6.101552663932704e-06,
"loss": 0.42648154497146606,
"step": 913
},
{
"epoch": 0.8959686312951844,
"grad_norm": 0.33748161792755127,
"learning_rate": 6.093812889053945e-06,
"loss": 0.4487994313240051,
"step": 914
},
{
"epoch": 0.8969489033206715,
"grad_norm": 0.39295104146003723,
"learning_rate": 6.08607036050254e-06,
"loss": 0.3864741325378418,
"step": 915
},
{
"epoch": 0.8979291753461586,
"grad_norm": 1.3106187582015991,
"learning_rate": 6.078325097770296e-06,
"loss": 0.4163864850997925,
"step": 916
},
{
"epoch": 0.8989094473716457,
"grad_norm": 0.5112157464027405,
"learning_rate": 6.070577120355903e-06,
"loss": 0.5529147386550903,
"step": 917
},
{
"epoch": 0.8998897193971327,
"grad_norm": 0.330213725566864,
"learning_rate": 6.062826447764883e-06,
"loss": 0.4116767644882202,
"step": 918
},
{
"epoch": 0.9008699914226198,
"grad_norm": 0.3667148947715759,
"learning_rate": 6.055073099509549e-06,
"loss": 0.4220765233039856,
"step": 919
},
{
"epoch": 0.9018502634481068,
"grad_norm": 0.3062370717525482,
"learning_rate": 6.047317095108943e-06,
"loss": 0.49312081933021545,
"step": 920
},
{
"epoch": 0.9028305354735939,
"grad_norm": 0.29760825634002686,
"learning_rate": 6.039558454088796e-06,
"loss": 0.3371737003326416,
"step": 921
},
{
"epoch": 0.903810807499081,
"grad_norm": 0.2954891324043274,
"learning_rate": 6.031797195981481e-06,
"loss": 0.47716888785362244,
"step": 922
},
{
"epoch": 0.904791079524568,
"grad_norm": 0.28700292110443115,
"learning_rate": 6.024033340325954e-06,
"loss": 0.39171314239501953,
"step": 923
},
{
"epoch": 0.9057713515500552,
"grad_norm": 0.2711394727230072,
"learning_rate": 6.016266906667711e-06,
"loss": 0.41017502546310425,
"step": 924
},
{
"epoch": 0.9067516235755422,
"grad_norm": 0.30363714694976807,
"learning_rate": 6.0084979145587444e-06,
"loss": 0.4056490659713745,
"step": 925
},
{
"epoch": 0.9077318956010293,
"grad_norm": 0.45842301845550537,
"learning_rate": 6.000726383557476e-06,
"loss": 0.3898978531360626,
"step": 926
},
{
"epoch": 0.9087121676265164,
"grad_norm": 0.2909232974052429,
"learning_rate": 5.9929523332287275e-06,
"loss": 0.4011874198913574,
"step": 927
},
{
"epoch": 0.9096924396520034,
"grad_norm": 0.2813899517059326,
"learning_rate": 5.985175783143666e-06,
"loss": 0.41717562079429626,
"step": 928
},
{
"epoch": 0.9106727116774905,
"grad_norm": 0.2974295914173126,
"learning_rate": 5.977396752879742e-06,
"loss": 0.49156323075294495,
"step": 929
},
{
"epoch": 0.9116529837029775,
"grad_norm": 0.23285380005836487,
"learning_rate": 5.9696152620206575e-06,
"loss": 0.359454870223999,
"step": 930
},
{
"epoch": 0.9126332557284647,
"grad_norm": 0.33687731623649597,
"learning_rate": 5.961831330156306e-06,
"loss": 0.523488461971283,
"step": 931
},
{
"epoch": 0.9136135277539518,
"grad_norm": 0.39405179023742676,
"learning_rate": 5.954044976882725e-06,
"loss": 0.4537978172302246,
"step": 932
},
{
"epoch": 0.9145937997794388,
"grad_norm": 0.5250006914138794,
"learning_rate": 5.946256221802052e-06,
"loss": 0.42208331823349,
"step": 933
},
{
"epoch": 0.9155740718049259,
"grad_norm": 0.45084014534950256,
"learning_rate": 5.938465084522468e-06,
"loss": 0.4873068332672119,
"step": 934
},
{
"epoch": 0.9165543438304129,
"grad_norm": 0.35132575035095215,
"learning_rate": 5.930671584658151e-06,
"loss": 0.3839843273162842,
"step": 935
},
{
"epoch": 0.9175346158559,
"grad_norm": 0.3010925054550171,
"learning_rate": 5.922875741829227e-06,
"loss": 0.45875054597854614,
"step": 936
},
{
"epoch": 0.9185148878813871,
"grad_norm": 0.3349578380584717,
"learning_rate": 5.915077575661723e-06,
"loss": 0.5047145485877991,
"step": 937
},
{
"epoch": 0.9194951599068741,
"grad_norm": 0.2819299101829529,
"learning_rate": 5.907277105787513e-06,
"loss": 0.3440694212913513,
"step": 938
},
{
"epoch": 0.9204754319323613,
"grad_norm": 0.31048014760017395,
"learning_rate": 5.89947435184427e-06,
"loss": 0.3258098363876343,
"step": 939
},
{
"epoch": 0.9214557039578483,
"grad_norm": 0.2895377576351166,
"learning_rate": 5.891669333475416e-06,
"loss": 0.52451491355896,
"step": 940
},
{
"epoch": 0.9224359759833354,
"grad_norm": 0.3913104832172394,
"learning_rate": 5.883862070330079e-06,
"loss": 0.5066572427749634,
"step": 941
},
{
"epoch": 0.9234162480088225,
"grad_norm": 0.31948184967041016,
"learning_rate": 5.876052582063031e-06,
"loss": 0.4451715350151062,
"step": 942
},
{
"epoch": 0.9243965200343095,
"grad_norm": 0.35596105456352234,
"learning_rate": 5.8682408883346535e-06,
"loss": 0.3669154644012451,
"step": 943
},
{
"epoch": 0.9253767920597966,
"grad_norm": 0.3608701229095459,
"learning_rate": 5.860427008810872e-06,
"loss": 0.4023933410644531,
"step": 944
},
{
"epoch": 0.9263570640852836,
"grad_norm": 0.562224805355072,
"learning_rate": 5.85261096316312e-06,
"loss": 0.4519593417644501,
"step": 945
},
{
"epoch": 0.9273373361107707,
"grad_norm": 0.386809766292572,
"learning_rate": 5.844792771068284e-06,
"loss": 0.49128207564353943,
"step": 946
},
{
"epoch": 0.9283176081362579,
"grad_norm": 0.8160575032234192,
"learning_rate": 5.8369724522086545e-06,
"loss": 0.38714125752449036,
"step": 947
},
{
"epoch": 0.9292978801617449,
"grad_norm": 0.32722610235214233,
"learning_rate": 5.829150026271871e-06,
"loss": 0.4942607581615448,
"step": 948
},
{
"epoch": 0.930278152187232,
"grad_norm": 0.3709994852542877,
"learning_rate": 5.821325512950886e-06,
"loss": 0.39199280738830566,
"step": 949
},
{
"epoch": 0.931258424212719,
"grad_norm": 0.4097587764263153,
"learning_rate": 5.8134989319439004e-06,
"loss": 0.460847407579422,
"step": 950
},
{
"epoch": 0.9322386962382061,
"grad_norm": 0.2851768434047699,
"learning_rate": 5.805670302954322e-06,
"loss": 0.4190582036972046,
"step": 951
},
{
"epoch": 0.9332189682636932,
"grad_norm": 0.45305484533309937,
"learning_rate": 5.797839645690715e-06,
"loss": 0.4143264591693878,
"step": 952
},
{
"epoch": 0.9341992402891802,
"grad_norm": 0.3221477270126343,
"learning_rate": 5.79000697986675e-06,
"loss": 0.40811049938201904,
"step": 953
},
{
"epoch": 0.9351795123146673,
"grad_norm": 0.31661131978034973,
"learning_rate": 5.782172325201155e-06,
"loss": 0.41623321175575256,
"step": 954
},
{
"epoch": 0.9361597843401543,
"grad_norm": 0.4302314221858978,
"learning_rate": 5.774335701417662e-06,
"loss": 0.37656867504119873,
"step": 955
},
{
"epoch": 0.9371400563656415,
"grad_norm": 0.31532588601112366,
"learning_rate": 5.766497128244963e-06,
"loss": 0.4484158158302307,
"step": 956
},
{
"epoch": 0.9381203283911286,
"grad_norm": 0.32827454805374146,
"learning_rate": 5.758656625416659e-06,
"loss": 0.42544180154800415,
"step": 957
},
{
"epoch": 0.9391006004166156,
"grad_norm": 0.3877761662006378,
"learning_rate": 5.750814212671202e-06,
"loss": 0.4354080855846405,
"step": 958
},
{
"epoch": 0.9400808724421027,
"grad_norm": 0.33081963658332825,
"learning_rate": 5.7429699097518585e-06,
"loss": 0.43439674377441406,
"step": 959
},
{
"epoch": 0.9410611444675897,
"grad_norm": 0.27640819549560547,
"learning_rate": 5.735123736406655e-06,
"loss": 0.38663074374198914,
"step": 960
},
{
"epoch": 0.9420414164930768,
"grad_norm": 0.342392235994339,
"learning_rate": 5.727275712388318e-06,
"loss": 0.41152411699295044,
"step": 961
},
{
"epoch": 0.9430216885185639,
"grad_norm": 0.4459559917449951,
"learning_rate": 5.719425857454242e-06,
"loss": 0.5751574039459229,
"step": 962
},
{
"epoch": 0.9440019605440509,
"grad_norm": 0.2561616897583008,
"learning_rate": 5.711574191366427e-06,
"loss": 0.3863458037376404,
"step": 963
},
{
"epoch": 0.9449822325695381,
"grad_norm": 0.3568366765975952,
"learning_rate": 5.7037207338914326e-06,
"loss": 0.42360058426856995,
"step": 964
},
{
"epoch": 0.9459625045950251,
"grad_norm": 0.35666730999946594,
"learning_rate": 5.695865504800328e-06,
"loss": 0.48775383830070496,
"step": 965
},
{
"epoch": 0.9469427766205122,
"grad_norm": 0.32493409514427185,
"learning_rate": 5.688008523868646e-06,
"loss": 0.43105781078338623,
"step": 966
},
{
"epoch": 0.9479230486459993,
"grad_norm": 0.40387213230133057,
"learning_rate": 5.680149810876322e-06,
"loss": 0.5018634796142578,
"step": 967
},
{
"epoch": 0.9489033206714863,
"grad_norm": 0.3793201148509979,
"learning_rate": 5.6722893856076596e-06,
"loss": 0.3920091390609741,
"step": 968
},
{
"epoch": 0.9498835926969734,
"grad_norm": 0.3202980160713196,
"learning_rate": 5.664427267851271e-06,
"loss": 0.39649468660354614,
"step": 969
},
{
"epoch": 0.9508638647224604,
"grad_norm": 0.3573557436466217,
"learning_rate": 5.656563477400026e-06,
"loss": 0.4525819420814514,
"step": 970
},
{
"epoch": 0.9518441367479475,
"grad_norm": 0.31116071343421936,
"learning_rate": 5.648698034051009e-06,
"loss": 0.4574243426322937,
"step": 971
},
{
"epoch": 0.9528244087734347,
"grad_norm": 0.28950875997543335,
"learning_rate": 5.640830957605466e-06,
"loss": 0.4106305241584778,
"step": 972
},
{
"epoch": 0.9538046807989217,
"grad_norm": 0.28239327669143677,
"learning_rate": 5.632962267868747e-06,
"loss": 0.38989928364753723,
"step": 973
},
{
"epoch": 0.9547849528244088,
"grad_norm": 0.3322334289550781,
"learning_rate": 5.625091984650273e-06,
"loss": 0.4464173913002014,
"step": 974
},
{
"epoch": 0.9557652248498959,
"grad_norm": 0.26644036173820496,
"learning_rate": 5.617220127763474e-06,
"loss": 0.35272306203842163,
"step": 975
},
{
"epoch": 0.9567454968753829,
"grad_norm": 0.305325448513031,
"learning_rate": 5.609346717025738e-06,
"loss": 0.38712137937545776,
"step": 976
},
{
"epoch": 0.95772576890087,
"grad_norm": 0.29020068049430847,
"learning_rate": 5.601471772258368e-06,
"loss": 0.41181060671806335,
"step": 977
},
{
"epoch": 0.958706040926357,
"grad_norm": 0.29625049233436584,
"learning_rate": 5.593595313286526e-06,
"loss": 0.3349764943122864,
"step": 978
},
{
"epoch": 0.9596863129518441,
"grad_norm": 0.3069136440753937,
"learning_rate": 5.585717359939192e-06,
"loss": 0.3725668787956238,
"step": 979
},
{
"epoch": 0.9606665849773313,
"grad_norm": 0.4017685353755951,
"learning_rate": 5.5778379320491e-06,
"loss": 0.3544464111328125,
"step": 980
},
{
"epoch": 0.9616468570028183,
"grad_norm": 0.32980185747146606,
"learning_rate": 5.569957049452703e-06,
"loss": 0.537687361240387,
"step": 981
},
{
"epoch": 0.9626271290283054,
"grad_norm": 0.46896466612815857,
"learning_rate": 5.562074731990115e-06,
"loss": 0.4130178689956665,
"step": 982
},
{
"epoch": 0.9636074010537924,
"grad_norm": 0.26929736137390137,
"learning_rate": 5.5541909995050554e-06,
"loss": 0.3820406198501587,
"step": 983
},
{
"epoch": 0.9645876730792795,
"grad_norm": 0.2751615047454834,
"learning_rate": 5.546305871844816e-06,
"loss": 0.31729909777641296,
"step": 984
},
{
"epoch": 0.9655679451047666,
"grad_norm": 0.3322264850139618,
"learning_rate": 5.538419368860196e-06,
"loss": 0.3638133406639099,
"step": 985
},
{
"epoch": 0.9665482171302536,
"grad_norm": 0.27276843786239624,
"learning_rate": 5.530531510405454e-06,
"loss": 0.36567509174346924,
"step": 986
},
{
"epoch": 0.9675284891557407,
"grad_norm": 0.31489068269729614,
"learning_rate": 5.522642316338268e-06,
"loss": 0.37623077630996704,
"step": 987
},
{
"epoch": 0.9685087611812278,
"grad_norm": 0.3067493736743927,
"learning_rate": 5.514751806519673e-06,
"loss": 0.44304215908050537,
"step": 988
},
{
"epoch": 0.9694890332067149,
"grad_norm": 0.5795502066612244,
"learning_rate": 5.506860000814017e-06,
"loss": 0.3512406051158905,
"step": 989
},
{
"epoch": 0.970469305232202,
"grad_norm": 0.35392746329307556,
"learning_rate": 5.498966919088914e-06,
"loss": 0.41662225127220154,
"step": 990
},
{
"epoch": 0.971449577257689,
"grad_norm": 0.2693781554698944,
"learning_rate": 5.491072581215186e-06,
"loss": 0.42721617221832275,
"step": 991
},
{
"epoch": 0.9724298492831761,
"grad_norm": 0.27147525548934937,
"learning_rate": 5.48317700706682e-06,
"loss": 0.4017362594604492,
"step": 992
},
{
"epoch": 0.9734101213086631,
"grad_norm": 0.29802027344703674,
"learning_rate": 5.475280216520913e-06,
"loss": 0.37348949909210205,
"step": 993
},
{
"epoch": 0.9743903933341502,
"grad_norm": 1.0046125650405884,
"learning_rate": 5.467382229457629e-06,
"loss": 0.4139745533466339,
"step": 994
},
{
"epoch": 0.9753706653596373,
"grad_norm": 0.25824400782585144,
"learning_rate": 5.459483065760138e-06,
"loss": 0.4171868562698364,
"step": 995
},
{
"epoch": 0.9763509373851244,
"grad_norm": 0.33299964666366577,
"learning_rate": 5.451582745314576e-06,
"loss": 0.41665223240852356,
"step": 996
},
{
"epoch": 0.9773312094106115,
"grad_norm": 0.30547717213630676,
"learning_rate": 5.443681288009991e-06,
"loss": 0.4325626492500305,
"step": 997
},
{
"epoch": 0.9783114814360985,
"grad_norm": 0.39693406224250793,
"learning_rate": 5.435778713738292e-06,
"loss": 0.3989555835723877,
"step": 998
},
{
"epoch": 0.9792917534615856,
"grad_norm": 0.9134981036186218,
"learning_rate": 5.4278750423942e-06,
"loss": 0.45765018463134766,
"step": 999
},
{
"epoch": 0.9802720254870727,
"grad_norm": 0.2770118713378906,
"learning_rate": 5.419970293875197e-06,
"loss": 0.34857940673828125,
"step": 1000
},
{
"epoch": 0.9812522975125597,
"grad_norm": 0.34108588099479675,
"learning_rate": 5.412064488081482e-06,
"loss": 0.41997697949409485,
"step": 1001
},
{
"epoch": 0.9822325695380468,
"grad_norm": 0.37812215089797974,
"learning_rate": 5.404157644915907e-06,
"loss": 0.4056851267814636,
"step": 1002
},
{
"epoch": 0.9832128415635338,
"grad_norm": 0.296744704246521,
"learning_rate": 5.396249784283943e-06,
"loss": 0.3869246542453766,
"step": 1003
},
{
"epoch": 0.984193113589021,
"grad_norm": 0.31195154786109924,
"learning_rate": 5.388340926093619e-06,
"loss": 0.39322972297668457,
"step": 1004
},
{
"epoch": 0.9851733856145081,
"grad_norm": 0.5937031507492065,
"learning_rate": 5.380431090255475e-06,
"loss": 0.3852788209915161,
"step": 1005
},
{
"epoch": 0.9861536576399951,
"grad_norm": 0.32535937428474426,
"learning_rate": 5.372520296682516e-06,
"loss": 0.5197076797485352,
"step": 1006
},
{
"epoch": 0.9871339296654822,
"grad_norm": 0.2741507291793823,
"learning_rate": 5.364608565290154e-06,
"loss": 0.3643001317977905,
"step": 1007
},
{
"epoch": 0.9881142016909692,
"grad_norm": 0.3201862871646881,
"learning_rate": 5.356695915996162e-06,
"loss": 0.37721192836761475,
"step": 1008
},
{
"epoch": 0.9890944737164563,
"grad_norm": 0.29314473271369934,
"learning_rate": 5.348782368720627e-06,
"loss": 0.42935293912887573,
"step": 1009
},
{
"epoch": 0.9900747457419434,
"grad_norm": 0.2943647801876068,
"learning_rate": 5.340867943385895e-06,
"loss": 0.39981311559677124,
"step": 1010
},
{
"epoch": 0.9910550177674304,
"grad_norm": 0.3196339011192322,
"learning_rate": 5.33295265991652e-06,
"loss": 0.4453566074371338,
"step": 1011
},
{
"epoch": 0.9920352897929176,
"grad_norm": 0.3288740813732147,
"learning_rate": 5.325036538239221e-06,
"loss": 0.37364131212234497,
"step": 1012
},
{
"epoch": 0.9930155618184046,
"grad_norm": 0.3059570789337158,
"learning_rate": 5.317119598282823e-06,
"loss": 0.3723534643650055,
"step": 1013
},
{
"epoch": 0.9939958338438917,
"grad_norm": 0.3757978081703186,
"learning_rate": 5.3092018599782155e-06,
"loss": 0.4581247568130493,
"step": 1014
},
{
"epoch": 0.9949761058693788,
"grad_norm": 0.3322988450527191,
"learning_rate": 5.301283343258293e-06,
"loss": 0.42948096990585327,
"step": 1015
},
{
"epoch": 0.9959563778948658,
"grad_norm": 0.3131640553474426,
"learning_rate": 5.293364068057912e-06,
"loss": 0.4125756025314331,
"step": 1016
},
{
"epoch": 0.9969366499203529,
"grad_norm": 0.23614847660064697,
"learning_rate": 5.285444054313841e-06,
"loss": 0.2991466522216797,
"step": 1017
},
{
"epoch": 0.9979169219458399,
"grad_norm": 0.4450140595436096,
"learning_rate": 5.277523321964701e-06,
"loss": 0.4575870633125305,
"step": 1018
},
{
"epoch": 0.998897193971327,
"grad_norm": 0.3129064440727234,
"learning_rate": 5.26960189095093e-06,
"loss": 0.3422028422355652,
"step": 1019
},
{
"epoch": 0.9998774659968142,
"grad_norm": 0.26912441849708557,
"learning_rate": 5.2616797812147205e-06,
"loss": 0.3474966585636139,
"step": 1020
},
{
"epoch": 1.0,
"grad_norm": 1.288994550704956,
"learning_rate": 5.253757012699972e-06,
"loss": 0.5389457941055298,
"step": 1021
}
],
"logging_steps": 1,
"max_steps": 2042,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.0095509197270548e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}