gyung's picture
Upload Gemma 4 native SFT 2Epoch checkpoint
a8161b8 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 2042,
"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
},
{
"epoch": 1.000980272025487,
"grad_norm": 0.2949475049972534,
"learning_rate": 5.245833605352246e-06,
"loss": 0.3489718735218048,
"step": 1022
},
{
"epoch": 1.0019605440509742,
"grad_norm": 0.2969910800457001,
"learning_rate": 5.237909579118713e-06,
"loss": 0.3505769968032837,
"step": 1023
},
{
"epoch": 1.0029408160764612,
"grad_norm": 0.2821654975414276,
"learning_rate": 5.2299849539480965e-06,
"loss": 0.36999836564064026,
"step": 1024
},
{
"epoch": 1.0039210881019482,
"grad_norm": 0.2582765221595764,
"learning_rate": 5.2220597497906315e-06,
"loss": 0.3053387701511383,
"step": 1025
},
{
"epoch": 1.0049013601274355,
"grad_norm": 0.28064215183258057,
"learning_rate": 5.214133986598014e-06,
"loss": 0.294108122587204,
"step": 1026
},
{
"epoch": 1.0058816321529225,
"grad_norm": 0.4019564986228943,
"learning_rate": 5.206207684323337e-06,
"loss": 0.28479501605033875,
"step": 1027
},
{
"epoch": 1.0068619041784095,
"grad_norm": 0.28043949604034424,
"learning_rate": 5.198280862921062e-06,
"loss": 0.3343275487422943,
"step": 1028
},
{
"epoch": 1.0078421762038965,
"grad_norm": 0.41693583130836487,
"learning_rate": 5.190353542346951e-06,
"loss": 0.28028419613838196,
"step": 1029
},
{
"epoch": 1.0088224482293837,
"grad_norm": 0.32368141412734985,
"learning_rate": 5.182425742558024e-06,
"loss": 0.39237695932388306,
"step": 1030
},
{
"epoch": 1.0098027202548707,
"grad_norm": 0.32892295718193054,
"learning_rate": 5.174497483512506e-06,
"loss": 0.36585330963134766,
"step": 1031
},
{
"epoch": 1.0107829922803577,
"grad_norm": 0.3015086352825165,
"learning_rate": 5.166568785169781e-06,
"loss": 0.3465660512447357,
"step": 1032
},
{
"epoch": 1.011763264305845,
"grad_norm": 0.24971866607666016,
"learning_rate": 5.15863966749034e-06,
"loss": 0.27902379631996155,
"step": 1033
},
{
"epoch": 1.012743536331332,
"grad_norm": 0.32669827342033386,
"learning_rate": 5.1507101504357225e-06,
"loss": 0.3548927307128906,
"step": 1034
},
{
"epoch": 1.013723808356819,
"grad_norm": 0.3506229519844055,
"learning_rate": 5.142780253968481e-06,
"loss": 0.2632600665092468,
"step": 1035
},
{
"epoch": 1.0147040803823062,
"grad_norm": 0.3122851252555847,
"learning_rate": 5.134849998052123e-06,
"loss": 0.36715859174728394,
"step": 1036
},
{
"epoch": 1.0156843524077932,
"grad_norm": 0.4787794351577759,
"learning_rate": 5.126919402651053e-06,
"loss": 0.3501265048980713,
"step": 1037
},
{
"epoch": 1.0166646244332802,
"grad_norm": 0.3395592272281647,
"learning_rate": 5.118988487730537e-06,
"loss": 0.3890085816383362,
"step": 1038
},
{
"epoch": 1.0176448964587672,
"grad_norm": 0.35633134841918945,
"learning_rate": 5.111057273256648e-06,
"loss": 0.3723975718021393,
"step": 1039
},
{
"epoch": 1.0186251684842544,
"grad_norm": 0.2719672620296478,
"learning_rate": 5.103125779196204e-06,
"loss": 0.34382039308547974,
"step": 1040
},
{
"epoch": 1.0196054405097414,
"grad_norm": 0.46195414662361145,
"learning_rate": 5.095194025516733e-06,
"loss": 0.33529412746429443,
"step": 1041
},
{
"epoch": 1.0205857125352285,
"grad_norm": 0.41559839248657227,
"learning_rate": 5.087262032186418e-06,
"loss": 0.37970829010009766,
"step": 1042
},
{
"epoch": 1.0215659845607157,
"grad_norm": 0.28565263748168945,
"learning_rate": 5.07932981917404e-06,
"loss": 0.3248412013053894,
"step": 1043
},
{
"epoch": 1.0225462565862027,
"grad_norm": 0.2896265387535095,
"learning_rate": 5.071397406448937e-06,
"loss": 0.34457769989967346,
"step": 1044
},
{
"epoch": 1.0235265286116897,
"grad_norm": 0.2948673367500305,
"learning_rate": 5.063464813980948e-06,
"loss": 0.3128063678741455,
"step": 1045
},
{
"epoch": 1.024506800637177,
"grad_norm": 0.2819617986679077,
"learning_rate": 5.055532061740367e-06,
"loss": 0.29185736179351807,
"step": 1046
},
{
"epoch": 1.025487072662664,
"grad_norm": 0.5507665276527405,
"learning_rate": 5.0475991696978844e-06,
"loss": 0.33982205390930176,
"step": 1047
},
{
"epoch": 1.026467344688151,
"grad_norm": 0.3292943835258484,
"learning_rate": 5.039666157824549e-06,
"loss": 0.33232107758522034,
"step": 1048
},
{
"epoch": 1.027447616713638,
"grad_norm": 0.427908331155777,
"learning_rate": 5.03173304609171e-06,
"loss": 0.33932632207870483,
"step": 1049
},
{
"epoch": 1.0284278887391252,
"grad_norm": 0.482431560754776,
"learning_rate": 5.023799854470963e-06,
"loss": 0.3526557683944702,
"step": 1050
},
{
"epoch": 1.0294081607646122,
"grad_norm": 0.2904674708843231,
"learning_rate": 5.015866602934112e-06,
"loss": 0.2847120463848114,
"step": 1051
},
{
"epoch": 1.0303884327900992,
"grad_norm": 0.35751640796661377,
"learning_rate": 5.007933311453108e-06,
"loss": 0.4169718325138092,
"step": 1052
},
{
"epoch": 1.0313687048155864,
"grad_norm": 0.3474958539009094,
"learning_rate": 5e-06,
"loss": 0.33768224716186523,
"step": 1053
},
{
"epoch": 1.0323489768410734,
"grad_norm": 0.34228670597076416,
"learning_rate": 4.992066688546895e-06,
"loss": 0.35064154863357544,
"step": 1054
},
{
"epoch": 1.0333292488665604,
"grad_norm": 0.36021527647972107,
"learning_rate": 4.984133397065889e-06,
"loss": 0.30471956729888916,
"step": 1055
},
{
"epoch": 1.0343095208920476,
"grad_norm": 0.30418914556503296,
"learning_rate": 4.976200145529039e-06,
"loss": 0.30376917123794556,
"step": 1056
},
{
"epoch": 1.0352897929175346,
"grad_norm": 0.31690332293510437,
"learning_rate": 4.9682669539082914e-06,
"loss": 0.3927238881587982,
"step": 1057
},
{
"epoch": 1.0362700649430217,
"grad_norm": 0.29696550965309143,
"learning_rate": 4.960333842175453e-06,
"loss": 0.29678207635879517,
"step": 1058
},
{
"epoch": 1.0372503369685087,
"grad_norm": 0.3309403359889984,
"learning_rate": 4.952400830302117e-06,
"loss": 0.35090988874435425,
"step": 1059
},
{
"epoch": 1.0382306089939959,
"grad_norm": 0.3197896182537079,
"learning_rate": 4.944467938259636e-06,
"loss": 0.24874848127365112,
"step": 1060
},
{
"epoch": 1.039210881019483,
"grad_norm": 0.6880909204483032,
"learning_rate": 4.936535186019053e-06,
"loss": 0.29992103576660156,
"step": 1061
},
{
"epoch": 1.04019115304497,
"grad_norm": 0.6577183604240417,
"learning_rate": 4.928602593551065e-06,
"loss": 0.37265241146087646,
"step": 1062
},
{
"epoch": 1.0411714250704571,
"grad_norm": 0.8678672313690186,
"learning_rate": 4.9206701808259605e-06,
"loss": 0.3217273950576782,
"step": 1063
},
{
"epoch": 1.0421516970959441,
"grad_norm": 0.34309330582618713,
"learning_rate": 4.9127379678135825e-06,
"loss": 0.35799241065979004,
"step": 1064
},
{
"epoch": 1.0431319691214311,
"grad_norm": 0.3836115598678589,
"learning_rate": 4.904805974483267e-06,
"loss": 0.36735689640045166,
"step": 1065
},
{
"epoch": 1.0441122411469184,
"grad_norm": 0.2663360834121704,
"learning_rate": 4.8968742208037976e-06,
"loss": 0.26856791973114014,
"step": 1066
},
{
"epoch": 1.0450925131724054,
"grad_norm": 0.34731292724609375,
"learning_rate": 4.888942726743353e-06,
"loss": 0.2638586163520813,
"step": 1067
},
{
"epoch": 1.0460727851978924,
"grad_norm": 0.5719500184059143,
"learning_rate": 4.881011512269464e-06,
"loss": 0.4483392834663391,
"step": 1068
},
{
"epoch": 1.0470530572233794,
"grad_norm": 0.31212615966796875,
"learning_rate": 4.873080597348948e-06,
"loss": 0.2929500937461853,
"step": 1069
},
{
"epoch": 1.0480333292488666,
"grad_norm": 0.43214935064315796,
"learning_rate": 4.86515000194788e-06,
"loss": 0.335938423871994,
"step": 1070
},
{
"epoch": 1.0490136012743536,
"grad_norm": 0.4487035572528839,
"learning_rate": 4.85721974603152e-06,
"loss": 0.40125352144241333,
"step": 1071
},
{
"epoch": 1.0499938732998406,
"grad_norm": 0.4256312847137451,
"learning_rate": 4.849289849564278e-06,
"loss": 0.3171235918998718,
"step": 1072
},
{
"epoch": 1.0509741453253278,
"grad_norm": 0.25656116008758545,
"learning_rate": 4.841360332509663e-06,
"loss": 0.26315537095069885,
"step": 1073
},
{
"epoch": 1.0519544173508149,
"grad_norm": 0.32400140166282654,
"learning_rate": 4.83343121483022e-06,
"loss": 0.35481417179107666,
"step": 1074
},
{
"epoch": 1.0529346893763019,
"grad_norm": 0.393541157245636,
"learning_rate": 4.825502516487497e-06,
"loss": 0.35062652826309204,
"step": 1075
},
{
"epoch": 1.053914961401789,
"grad_norm": 0.7216739058494568,
"learning_rate": 4.817574257441978e-06,
"loss": 0.3172438144683838,
"step": 1076
},
{
"epoch": 1.054895233427276,
"grad_norm": 0.33664554357528687,
"learning_rate": 4.809646457653051e-06,
"loss": 0.30129069089889526,
"step": 1077
},
{
"epoch": 1.055875505452763,
"grad_norm": 0.30300626158714294,
"learning_rate": 4.801719137078939e-06,
"loss": 0.2866551876068115,
"step": 1078
},
{
"epoch": 1.0568557774782503,
"grad_norm": 0.3296108841896057,
"learning_rate": 4.793792315676665e-06,
"loss": 0.36560168862342834,
"step": 1079
},
{
"epoch": 1.0578360495037373,
"grad_norm": 0.37670421600341797,
"learning_rate": 4.785866013401989e-06,
"loss": 0.3270641565322876,
"step": 1080
},
{
"epoch": 1.0588163215292243,
"grad_norm": 0.29558229446411133,
"learning_rate": 4.777940250209369e-06,
"loss": 0.30972665548324585,
"step": 1081
},
{
"epoch": 1.0597965935547113,
"grad_norm": 0.33294355869293213,
"learning_rate": 4.770015046051904e-06,
"loss": 0.3475317656993866,
"step": 1082
},
{
"epoch": 1.0607768655801986,
"grad_norm": 0.2815360426902771,
"learning_rate": 4.762090420881289e-06,
"loss": 0.31326645612716675,
"step": 1083
},
{
"epoch": 1.0617571376056856,
"grad_norm": 1.8726563453674316,
"learning_rate": 4.754166394647755e-06,
"loss": 0.30086567997932434,
"step": 1084
},
{
"epoch": 1.0627374096311726,
"grad_norm": 0.2989646792411804,
"learning_rate": 4.74624298730003e-06,
"loss": 0.3362056016921997,
"step": 1085
},
{
"epoch": 1.0637176816566598,
"grad_norm": 0.6975619196891785,
"learning_rate": 4.738320218785281e-06,
"loss": 0.36155256628990173,
"step": 1086
},
{
"epoch": 1.0646979536821468,
"grad_norm": 0.239951029419899,
"learning_rate": 4.7303981090490715e-06,
"loss": 0.32886865735054016,
"step": 1087
},
{
"epoch": 1.0656782257076338,
"grad_norm": 0.2613634765148163,
"learning_rate": 4.7224766780353005e-06,
"loss": 0.2992268204689026,
"step": 1088
},
{
"epoch": 1.0666584977331208,
"grad_norm": 0.2885245382785797,
"learning_rate": 4.71455594568616e-06,
"loss": 0.33844658732414246,
"step": 1089
},
{
"epoch": 1.067638769758608,
"grad_norm": 0.2931375801563263,
"learning_rate": 4.706635931942089e-06,
"loss": 0.2772749960422516,
"step": 1090
},
{
"epoch": 1.068619041784095,
"grad_norm": 0.29005083441734314,
"learning_rate": 4.6987166567417085e-06,
"loss": 0.32309579849243164,
"step": 1091
},
{
"epoch": 1.069599313809582,
"grad_norm": 0.2952600121498108,
"learning_rate": 4.690798140021787e-06,
"loss": 0.273627907037735,
"step": 1092
},
{
"epoch": 1.0705795858350693,
"grad_norm": 0.31154772639274597,
"learning_rate": 4.682880401717178e-06,
"loss": 0.34720271825790405,
"step": 1093
},
{
"epoch": 1.0715598578605563,
"grad_norm": 0.38783109188079834,
"learning_rate": 4.674963461760782e-06,
"loss": 0.3123989999294281,
"step": 1094
},
{
"epoch": 1.0725401298860433,
"grad_norm": 0.3247523307800293,
"learning_rate": 4.667047340083481e-06,
"loss": 0.3080710768699646,
"step": 1095
},
{
"epoch": 1.0735204019115305,
"grad_norm": 0.5765203237533569,
"learning_rate": 4.659132056614108e-06,
"loss": 0.3679295480251312,
"step": 1096
},
{
"epoch": 1.0745006739370175,
"grad_norm": 0.5751588940620422,
"learning_rate": 4.651217631279374e-06,
"loss": 0.3422990143299103,
"step": 1097
},
{
"epoch": 1.0754809459625045,
"grad_norm": 0.29886236786842346,
"learning_rate": 4.643304084003839e-06,
"loss": 0.24324765801429749,
"step": 1098
},
{
"epoch": 1.0764612179879918,
"grad_norm": 0.3330618143081665,
"learning_rate": 4.635391434709847e-06,
"loss": 0.3275555968284607,
"step": 1099
},
{
"epoch": 1.0774414900134788,
"grad_norm": 0.8442397713661194,
"learning_rate": 4.627479703317485e-06,
"loss": 0.33885207772254944,
"step": 1100
},
{
"epoch": 1.0784217620389658,
"grad_norm": 0.3028629422187805,
"learning_rate": 4.619568909744524e-06,
"loss": 0.29681622982025146,
"step": 1101
},
{
"epoch": 1.0794020340644528,
"grad_norm": 0.3467601239681244,
"learning_rate": 4.611659073906382e-06,
"loss": 0.33799439668655396,
"step": 1102
},
{
"epoch": 1.08038230608994,
"grad_norm": 0.3561437129974365,
"learning_rate": 4.603750215716057e-06,
"loss": 0.26955825090408325,
"step": 1103
},
{
"epoch": 1.081362578115427,
"grad_norm": 0.35010141134262085,
"learning_rate": 4.595842355084094e-06,
"loss": 0.299543559551239,
"step": 1104
},
{
"epoch": 1.082342850140914,
"grad_norm": 0.32590317726135254,
"learning_rate": 4.587935511918521e-06,
"loss": 0.3415825068950653,
"step": 1105
},
{
"epoch": 1.0833231221664013,
"grad_norm": 0.2841714322566986,
"learning_rate": 4.580029706124804e-06,
"loss": 0.30952975153923035,
"step": 1106
},
{
"epoch": 1.0843033941918883,
"grad_norm": 0.28472036123275757,
"learning_rate": 4.572124957605803e-06,
"loss": 0.3088083565235138,
"step": 1107
},
{
"epoch": 1.0852836662173753,
"grad_norm": 0.3014181852340698,
"learning_rate": 4.564221286261709e-06,
"loss": 0.29532432556152344,
"step": 1108
},
{
"epoch": 1.0862639382428625,
"grad_norm": 0.3145205080509186,
"learning_rate": 4.55631871199001e-06,
"loss": 0.313355028629303,
"step": 1109
},
{
"epoch": 1.0872442102683495,
"grad_norm": 0.3391856849193573,
"learning_rate": 4.548417254685425e-06,
"loss": 0.4632870554924011,
"step": 1110
},
{
"epoch": 1.0882244822938365,
"grad_norm": 0.2976081967353821,
"learning_rate": 4.5405169342398634e-06,
"loss": 0.2705850899219513,
"step": 1111
},
{
"epoch": 1.0892047543193235,
"grad_norm": 0.28754448890686035,
"learning_rate": 4.5326177705423725e-06,
"loss": 0.28869423270225525,
"step": 1112
},
{
"epoch": 1.0901850263448107,
"grad_norm": 0.35979798436164856,
"learning_rate": 4.524719783479088e-06,
"loss": 0.3419206440448761,
"step": 1113
},
{
"epoch": 1.0911652983702977,
"grad_norm": 0.2932913303375244,
"learning_rate": 4.5168229929331805e-06,
"loss": 0.3467756509780884,
"step": 1114
},
{
"epoch": 1.0921455703957847,
"grad_norm": 0.2771269977092743,
"learning_rate": 4.5089274187848144e-06,
"loss": 0.31667637825012207,
"step": 1115
},
{
"epoch": 1.093125842421272,
"grad_norm": 0.29554691910743713,
"learning_rate": 4.501033080911087e-06,
"loss": 0.38721439242362976,
"step": 1116
},
{
"epoch": 1.094106114446759,
"grad_norm": 0.2315177172422409,
"learning_rate": 4.493139999185984e-06,
"loss": 0.3122479319572449,
"step": 1117
},
{
"epoch": 1.095086386472246,
"grad_norm": 0.3093447983264923,
"learning_rate": 4.485248193480328e-06,
"loss": 0.3701137602329254,
"step": 1118
},
{
"epoch": 1.0960666584977332,
"grad_norm": 0.3324609100818634,
"learning_rate": 4.477357683661734e-06,
"loss": 0.3915468752384186,
"step": 1119
},
{
"epoch": 1.0970469305232202,
"grad_norm": 0.4165363907814026,
"learning_rate": 4.469468489594546e-06,
"loss": 0.3073456585407257,
"step": 1120
},
{
"epoch": 1.0980272025487072,
"grad_norm": 0.28519773483276367,
"learning_rate": 4.461580631139806e-06,
"loss": 0.3286278247833252,
"step": 1121
},
{
"epoch": 1.0990074745741945,
"grad_norm": 0.6971908807754517,
"learning_rate": 4.453694128155187e-06,
"loss": 0.2951686978340149,
"step": 1122
},
{
"epoch": 1.0999877465996815,
"grad_norm": 0.33109739422798157,
"learning_rate": 4.445809000494945e-06,
"loss": 0.3409639894962311,
"step": 1123
},
{
"epoch": 1.1009680186251685,
"grad_norm": 1.5810896158218384,
"learning_rate": 4.4379252680098875e-06,
"loss": 0.3432649075984955,
"step": 1124
},
{
"epoch": 1.1019482906506555,
"grad_norm": 0.296281635761261,
"learning_rate": 4.430042950547298e-06,
"loss": 0.32961803674697876,
"step": 1125
},
{
"epoch": 1.1029285626761427,
"grad_norm": 0.3262688219547272,
"learning_rate": 4.4221620679509016e-06,
"loss": 0.2999087870121002,
"step": 1126
},
{
"epoch": 1.1039088347016297,
"grad_norm": 0.3172036111354828,
"learning_rate": 4.414282640060809e-06,
"loss": 0.28766530752182007,
"step": 1127
},
{
"epoch": 1.1048891067271167,
"grad_norm": 0.295867919921875,
"learning_rate": 4.4064046867134755e-06,
"loss": 0.24284593760967255,
"step": 1128
},
{
"epoch": 1.105869378752604,
"grad_norm": 0.3114313781261444,
"learning_rate": 4.398528227741634e-06,
"loss": 0.2893628478050232,
"step": 1129
},
{
"epoch": 1.106849650778091,
"grad_norm": 0.4056475758552551,
"learning_rate": 4.390653282974264e-06,
"loss": 0.3033437132835388,
"step": 1130
},
{
"epoch": 1.107829922803578,
"grad_norm": 0.28321322798728943,
"learning_rate": 4.382779872236527e-06,
"loss": 0.3426106572151184,
"step": 1131
},
{
"epoch": 1.108810194829065,
"grad_norm": 0.30477651953697205,
"learning_rate": 4.374908015349728e-06,
"loss": 0.3231625258922577,
"step": 1132
},
{
"epoch": 1.1097904668545522,
"grad_norm": 0.2864629328250885,
"learning_rate": 4.367037732131254e-06,
"loss": 0.28227946162223816,
"step": 1133
},
{
"epoch": 1.1107707388800392,
"grad_norm": 0.29221442341804504,
"learning_rate": 4.359169042394537e-06,
"loss": 0.3090466856956482,
"step": 1134
},
{
"epoch": 1.1117510109055262,
"grad_norm": 0.32867932319641113,
"learning_rate": 4.3513019659489916e-06,
"loss": 0.3440025746822357,
"step": 1135
},
{
"epoch": 1.1127312829310134,
"grad_norm": 0.309756875038147,
"learning_rate": 4.343436522599975e-06,
"loss": 0.31796810030937195,
"step": 1136
},
{
"epoch": 1.1137115549565004,
"grad_norm": 0.33752650022506714,
"learning_rate": 4.33557273214873e-06,
"loss": 0.3361341655254364,
"step": 1137
},
{
"epoch": 1.1146918269819874,
"grad_norm": 0.9087817072868347,
"learning_rate": 4.327710614392341e-06,
"loss": 0.2776590883731842,
"step": 1138
},
{
"epoch": 1.1156720990074747,
"grad_norm": 0.3308079242706299,
"learning_rate": 4.319850189123681e-06,
"loss": 0.2762799859046936,
"step": 1139
},
{
"epoch": 1.1166523710329617,
"grad_norm": 0.2800014913082123,
"learning_rate": 4.3119914761313566e-06,
"loss": 0.312404602766037,
"step": 1140
},
{
"epoch": 1.1176326430584487,
"grad_norm": 0.29110458493232727,
"learning_rate": 4.304134495199675e-06,
"loss": 0.2578994631767273,
"step": 1141
},
{
"epoch": 1.118612915083936,
"grad_norm": 0.9625495672225952,
"learning_rate": 4.296279266108569e-06,
"loss": 0.3422161340713501,
"step": 1142
},
{
"epoch": 1.119593187109423,
"grad_norm": 0.3016868829727173,
"learning_rate": 4.2884258086335755e-06,
"loss": 0.31202149391174316,
"step": 1143
},
{
"epoch": 1.12057345913491,
"grad_norm": 0.37162452936172485,
"learning_rate": 4.28057414254576e-06,
"loss": 0.3586808443069458,
"step": 1144
},
{
"epoch": 1.121553731160397,
"grad_norm": 0.47081947326660156,
"learning_rate": 4.272724287611684e-06,
"loss": 0.36297255754470825,
"step": 1145
},
{
"epoch": 1.1225340031858841,
"grad_norm": 0.2870137095451355,
"learning_rate": 4.264876263593347e-06,
"loss": 0.3227170705795288,
"step": 1146
},
{
"epoch": 1.1235142752113711,
"grad_norm": 0.31866779923439026,
"learning_rate": 4.257030090248142e-06,
"loss": 0.3754955530166626,
"step": 1147
},
{
"epoch": 1.1244945472368582,
"grad_norm": 0.37888720631599426,
"learning_rate": 4.249185787328798e-06,
"loss": 0.27347415685653687,
"step": 1148
},
{
"epoch": 1.1254748192623454,
"grad_norm": 0.3187525272369385,
"learning_rate": 4.241343374583343e-06,
"loss": 0.32907360792160034,
"step": 1149
},
{
"epoch": 1.1264550912878324,
"grad_norm": 0.2819685637950897,
"learning_rate": 4.233502871755036e-06,
"loss": 0.36464327573776245,
"step": 1150
},
{
"epoch": 1.1274353633133194,
"grad_norm": 0.2814306318759918,
"learning_rate": 4.225664298582339e-06,
"loss": 0.2959817051887512,
"step": 1151
},
{
"epoch": 1.1284156353388064,
"grad_norm": 0.3227444291114807,
"learning_rate": 4.217827674798845e-06,
"loss": 0.34794795513153076,
"step": 1152
},
{
"epoch": 1.1293959073642936,
"grad_norm": 0.2617141008377075,
"learning_rate": 4.209993020133251e-06,
"loss": 0.32848256826400757,
"step": 1153
},
{
"epoch": 1.1303761793897806,
"grad_norm": 0.32876184582710266,
"learning_rate": 4.202160354309286e-06,
"loss": 0.3521800637245178,
"step": 1154
},
{
"epoch": 1.1313564514152676,
"grad_norm": 0.38271641731262207,
"learning_rate": 4.194329697045681e-06,
"loss": 0.30995410680770874,
"step": 1155
},
{
"epoch": 1.1323367234407549,
"grad_norm": 0.2933138906955719,
"learning_rate": 4.186501068056102e-06,
"loss": 0.3704349398612976,
"step": 1156
},
{
"epoch": 1.1333169954662419,
"grad_norm": 0.32003486156463623,
"learning_rate": 4.178674487049116e-06,
"loss": 0.3253172039985657,
"step": 1157
},
{
"epoch": 1.1342972674917289,
"grad_norm": 0.43215644359588623,
"learning_rate": 4.1708499737281305e-06,
"loss": 0.2938027083873749,
"step": 1158
},
{
"epoch": 1.135277539517216,
"grad_norm": 0.27677276730537415,
"learning_rate": 4.163027547791347e-06,
"loss": 0.30322253704071045,
"step": 1159
},
{
"epoch": 1.1362578115427031,
"grad_norm": 0.3755778670310974,
"learning_rate": 4.1552072289317185e-06,
"loss": 0.26590603590011597,
"step": 1160
},
{
"epoch": 1.1372380835681901,
"grad_norm": 0.2454625517129898,
"learning_rate": 4.147389036836881e-06,
"loss": 0.27834102511405945,
"step": 1161
},
{
"epoch": 1.1382183555936773,
"grad_norm": 0.3381948471069336,
"learning_rate": 4.139572991189131e-06,
"loss": 0.36362361907958984,
"step": 1162
},
{
"epoch": 1.1391986276191644,
"grad_norm": 0.2932506799697876,
"learning_rate": 4.131759111665349e-06,
"loss": 0.33345091342926025,
"step": 1163
},
{
"epoch": 1.1401788996446514,
"grad_norm": 0.3875240385532379,
"learning_rate": 4.12394741793697e-06,
"loss": 0.3888590335845947,
"step": 1164
},
{
"epoch": 1.1411591716701386,
"grad_norm": 0.32361891865730286,
"learning_rate": 4.116137929669921e-06,
"loss": 0.32534122467041016,
"step": 1165
},
{
"epoch": 1.1421394436956256,
"grad_norm": 0.41653409600257874,
"learning_rate": 4.1083306665245855e-06,
"loss": 0.3440389037132263,
"step": 1166
},
{
"epoch": 1.1431197157211126,
"grad_norm": 0.5101289749145508,
"learning_rate": 4.100525648155731e-06,
"loss": 0.3064062297344208,
"step": 1167
},
{
"epoch": 1.1440999877465996,
"grad_norm": 0.40962323546409607,
"learning_rate": 4.092722894212488e-06,
"loss": 0.39340418577194214,
"step": 1168
},
{
"epoch": 1.1450802597720868,
"grad_norm": 0.33972427248954773,
"learning_rate": 4.084922424338277e-06,
"loss": 0.32673364877700806,
"step": 1169
},
{
"epoch": 1.1460605317975738,
"grad_norm": 0.3114092946052551,
"learning_rate": 4.077124258170774e-06,
"loss": 0.3107997179031372,
"step": 1170
},
{
"epoch": 1.1470408038230608,
"grad_norm": 0.3234725892543793,
"learning_rate": 4.06932841534185e-06,
"loss": 0.28566795587539673,
"step": 1171
},
{
"epoch": 1.148021075848548,
"grad_norm": 0.3056619465351105,
"learning_rate": 4.061534915477533e-06,
"loss": 0.25556519627571106,
"step": 1172
},
{
"epoch": 1.149001347874035,
"grad_norm": 0.37205490469932556,
"learning_rate": 4.053743778197951e-06,
"loss": 0.30245232582092285,
"step": 1173
},
{
"epoch": 1.149981619899522,
"grad_norm": 0.3018168807029724,
"learning_rate": 4.045955023117276e-06,
"loss": 0.3390134572982788,
"step": 1174
},
{
"epoch": 1.150961891925009,
"grad_norm": 0.7401418685913086,
"learning_rate": 4.038168669843698e-06,
"loss": 0.34388044476509094,
"step": 1175
},
{
"epoch": 1.1519421639504963,
"grad_norm": 0.4698144495487213,
"learning_rate": 4.030384737979345e-06,
"loss": 0.3400590121746063,
"step": 1176
},
{
"epoch": 1.1529224359759833,
"grad_norm": 0.2941637337207794,
"learning_rate": 4.02260324712026e-06,
"loss": 0.30593302845954895,
"step": 1177
},
{
"epoch": 1.1539027080014703,
"grad_norm": 0.3559083342552185,
"learning_rate": 4.014824216856336e-06,
"loss": 0.37448573112487793,
"step": 1178
},
{
"epoch": 1.1548829800269576,
"grad_norm": 0.29112058877944946,
"learning_rate": 4.007047666771274e-06,
"loss": 0.3234751224517822,
"step": 1179
},
{
"epoch": 1.1558632520524446,
"grad_norm": 0.36735907196998596,
"learning_rate": 3.999273616442526e-06,
"loss": 0.31429523229599,
"step": 1180
},
{
"epoch": 1.1568435240779316,
"grad_norm": 0.3060409724712372,
"learning_rate": 3.991502085441259e-06,
"loss": 0.31154805421829224,
"step": 1181
},
{
"epoch": 1.1578237961034188,
"grad_norm": 0.3897670805454254,
"learning_rate": 3.9837330933322895e-06,
"loss": 0.3393978774547577,
"step": 1182
},
{
"epoch": 1.1588040681289058,
"grad_norm": 0.3166610300540924,
"learning_rate": 3.975966659674048e-06,
"loss": 0.30177581310272217,
"step": 1183
},
{
"epoch": 1.1597843401543928,
"grad_norm": 0.30921435356140137,
"learning_rate": 3.968202804018519e-06,
"loss": 0.3340331017971039,
"step": 1184
},
{
"epoch": 1.16076461217988,
"grad_norm": 0.3042021095752716,
"learning_rate": 3.960441545911205e-06,
"loss": 0.2912120819091797,
"step": 1185
},
{
"epoch": 1.161744884205367,
"grad_norm": 0.2731899917125702,
"learning_rate": 3.952682904891058e-06,
"loss": 0.33261168003082275,
"step": 1186
},
{
"epoch": 1.162725156230854,
"grad_norm": 0.2941371500492096,
"learning_rate": 3.944926900490452e-06,
"loss": 0.3436034321784973,
"step": 1187
},
{
"epoch": 1.163705428256341,
"grad_norm": 0.3263383209705353,
"learning_rate": 3.937173552235117e-06,
"loss": 0.35178181529045105,
"step": 1188
},
{
"epoch": 1.1646857002818283,
"grad_norm": 0.42471933364868164,
"learning_rate": 3.929422879644099e-06,
"loss": 0.3174160420894623,
"step": 1189
},
{
"epoch": 1.1656659723073153,
"grad_norm": 0.27329784631729126,
"learning_rate": 3.921674902229706e-06,
"loss": 0.27904966473579407,
"step": 1190
},
{
"epoch": 1.1666462443328023,
"grad_norm": 0.29239943623542786,
"learning_rate": 3.913929639497462e-06,
"loss": 0.28838056325912476,
"step": 1191
},
{
"epoch": 1.1676265163582895,
"grad_norm": 0.3087271451950073,
"learning_rate": 3.906187110946058e-06,
"loss": 0.32163307070732117,
"step": 1192
},
{
"epoch": 1.1686067883837765,
"grad_norm": 0.33339235186576843,
"learning_rate": 3.898447336067297e-06,
"loss": 0.2780363857746124,
"step": 1193
},
{
"epoch": 1.1695870604092635,
"grad_norm": 0.3192176818847656,
"learning_rate": 3.890710334346058e-06,
"loss": 0.3871726989746094,
"step": 1194
},
{
"epoch": 1.1705673324347505,
"grad_norm": 0.5433230996131897,
"learning_rate": 3.882976125260229e-06,
"loss": 0.32924360036849976,
"step": 1195
},
{
"epoch": 1.1715476044602378,
"grad_norm": 0.3525082468986511,
"learning_rate": 3.875244728280676e-06,
"loss": 0.31390810012817383,
"step": 1196
},
{
"epoch": 1.1725278764857248,
"grad_norm": 0.2873048186302185,
"learning_rate": 3.867516162871177e-06,
"loss": 0.3036879301071167,
"step": 1197
},
{
"epoch": 1.1735081485112118,
"grad_norm": 0.4844062328338623,
"learning_rate": 3.859790448488394e-06,
"loss": 0.38189056515693665,
"step": 1198
},
{
"epoch": 1.174488420536699,
"grad_norm": 0.33477723598480225,
"learning_rate": 3.8520676045817945e-06,
"loss": 0.329548180103302,
"step": 1199
},
{
"epoch": 1.175468692562186,
"grad_norm": 0.33469077944755554,
"learning_rate": 3.844347650593635e-06,
"loss": 0.3531683087348938,
"step": 1200
},
{
"epoch": 1.176448964587673,
"grad_norm": 0.4027371406555176,
"learning_rate": 3.8366306059588885e-06,
"loss": 0.31449511647224426,
"step": 1201
},
{
"epoch": 1.1774292366131602,
"grad_norm": 1.9771807193756104,
"learning_rate": 3.828916490105206e-06,
"loss": 0.3776460289955139,
"step": 1202
},
{
"epoch": 1.1784095086386472,
"grad_norm": 0.3812496066093445,
"learning_rate": 3.821205322452863e-06,
"loss": 0.37760359048843384,
"step": 1203
},
{
"epoch": 1.1793897806641342,
"grad_norm": 0.3075524568557739,
"learning_rate": 3.8134971224147204e-06,
"loss": 0.3223971128463745,
"step": 1204
},
{
"epoch": 1.1803700526896215,
"grad_norm": 0.33409735560417175,
"learning_rate": 3.8057919093961554e-06,
"loss": 0.3504806458950043,
"step": 1205
},
{
"epoch": 1.1813503247151085,
"grad_norm": 0.8444024920463562,
"learning_rate": 3.798089702795038e-06,
"loss": 0.3162532448768616,
"step": 1206
},
{
"epoch": 1.1823305967405955,
"grad_norm": 0.3940410315990448,
"learning_rate": 3.790390522001662e-06,
"loss": 0.38942915201187134,
"step": 1207
},
{
"epoch": 1.1833108687660825,
"grad_norm": 0.42641228437423706,
"learning_rate": 3.782694386398706e-06,
"loss": 0.3000125586986542,
"step": 1208
},
{
"epoch": 1.1842911407915697,
"grad_norm": 0.3485449552536011,
"learning_rate": 3.775001315361183e-06,
"loss": 0.3856309652328491,
"step": 1209
},
{
"epoch": 1.1852714128170567,
"grad_norm": 0.4671294689178467,
"learning_rate": 3.7673113282563883e-06,
"loss": 0.3539689779281616,
"step": 1210
},
{
"epoch": 1.1862516848425437,
"grad_norm": 0.3236221671104431,
"learning_rate": 3.7596244444438577e-06,
"loss": 0.3083874583244324,
"step": 1211
},
{
"epoch": 1.187231956868031,
"grad_norm": 0.3453122079372406,
"learning_rate": 3.7519406832753085e-06,
"loss": 0.3381679058074951,
"step": 1212
},
{
"epoch": 1.188212228893518,
"grad_norm": 0.2715235650539398,
"learning_rate": 3.7442600640946045e-06,
"loss": 0.26225951313972473,
"step": 1213
},
{
"epoch": 1.189192500919005,
"grad_norm": 0.3354005813598633,
"learning_rate": 3.7365826062376933e-06,
"loss": 0.2832813858985901,
"step": 1214
},
{
"epoch": 1.190172772944492,
"grad_norm": 0.3599259555339813,
"learning_rate": 3.7289083290325668e-06,
"loss": 0.3262927830219269,
"step": 1215
},
{
"epoch": 1.1911530449699792,
"grad_norm": 0.283708393573761,
"learning_rate": 3.721237251799206e-06,
"loss": 0.34195423126220703,
"step": 1216
},
{
"epoch": 1.1921333169954662,
"grad_norm": 0.32946813106536865,
"learning_rate": 3.7135693938495433e-06,
"loss": 0.32708072662353516,
"step": 1217
},
{
"epoch": 1.1931135890209532,
"grad_norm": 0.31973326206207275,
"learning_rate": 3.705904774487396e-06,
"loss": 0.27196550369262695,
"step": 1218
},
{
"epoch": 1.1940938610464404,
"grad_norm": 0.597393810749054,
"learning_rate": 3.69824341300844e-06,
"loss": 0.3495296239852905,
"step": 1219
},
{
"epoch": 1.1950741330719274,
"grad_norm": 0.31360572576522827,
"learning_rate": 3.6905853287001405e-06,
"loss": 0.3529015779495239,
"step": 1220
},
{
"epoch": 1.1960544050974145,
"grad_norm": 0.28495967388153076,
"learning_rate": 3.682930540841717e-06,
"loss": 0.2948427200317383,
"step": 1221
},
{
"epoch": 1.1970346771229017,
"grad_norm": 0.31830504536628723,
"learning_rate": 3.675279068704087e-06,
"loss": 0.3309716582298279,
"step": 1222
},
{
"epoch": 1.1980149491483887,
"grad_norm": 0.5434629917144775,
"learning_rate": 3.667630931549826e-06,
"loss": 0.3288058638572693,
"step": 1223
},
{
"epoch": 1.1989952211738757,
"grad_norm": 0.3231112062931061,
"learning_rate": 3.6599861486331074e-06,
"loss": 0.3353787064552307,
"step": 1224
},
{
"epoch": 1.199975493199363,
"grad_norm": 0.30528584122657776,
"learning_rate": 3.6523447391996613e-06,
"loss": 0.3579012155532837,
"step": 1225
},
{
"epoch": 1.20095576522485,
"grad_norm": 0.27185603976249695,
"learning_rate": 3.6447067224867306e-06,
"loss": 0.28988292813301086,
"step": 1226
},
{
"epoch": 1.201936037250337,
"grad_norm": 0.38071879744529724,
"learning_rate": 3.637072117723012e-06,
"loss": 0.3545647859573364,
"step": 1227
},
{
"epoch": 1.2029163092758242,
"grad_norm": 0.2795783579349518,
"learning_rate": 3.6294409441286133e-06,
"loss": 0.36987143754959106,
"step": 1228
},
{
"epoch": 1.2038965813013112,
"grad_norm": 0.27458617091178894,
"learning_rate": 3.6218132209150047e-06,
"loss": 0.2656174898147583,
"step": 1229
},
{
"epoch": 1.2048768533267982,
"grad_norm": 2.1717491149902344,
"learning_rate": 3.6141889672849728e-06,
"loss": 0.3419286608695984,
"step": 1230
},
{
"epoch": 1.2058571253522852,
"grad_norm": 0.2507125437259674,
"learning_rate": 3.606568202432562e-06,
"loss": 0.28207966685295105,
"step": 1231
},
{
"epoch": 1.2068373973777724,
"grad_norm": 0.30578479170799255,
"learning_rate": 3.598950945543045e-06,
"loss": 0.3086465001106262,
"step": 1232
},
{
"epoch": 1.2078176694032594,
"grad_norm": 0.2804287075996399,
"learning_rate": 3.5913372157928515e-06,
"loss": 0.30075401067733765,
"step": 1233
},
{
"epoch": 1.2087979414287464,
"grad_norm": 0.29239925742149353,
"learning_rate": 3.5837270323495416e-06,
"loss": 0.3367430865764618,
"step": 1234
},
{
"epoch": 1.2097782134542336,
"grad_norm": 0.35808679461479187,
"learning_rate": 3.5761204143717387e-06,
"loss": 0.33775079250335693,
"step": 1235
},
{
"epoch": 1.2107584854797206,
"grad_norm": 0.3731803894042969,
"learning_rate": 3.568517381009099e-06,
"loss": 0.3316825032234192,
"step": 1236
},
{
"epoch": 1.2117387575052077,
"grad_norm": 0.5093199014663696,
"learning_rate": 3.560917951402245e-06,
"loss": 0.2694641053676605,
"step": 1237
},
{
"epoch": 1.2127190295306947,
"grad_norm": 2.2216570377349854,
"learning_rate": 3.553322144682737e-06,
"loss": 0.322288453578949,
"step": 1238
},
{
"epoch": 1.2136993015561819,
"grad_norm": 0.28069257736206055,
"learning_rate": 3.5457299799730047e-06,
"loss": 0.3057772219181061,
"step": 1239
},
{
"epoch": 1.214679573581669,
"grad_norm": 0.25850751996040344,
"learning_rate": 3.538141476386317e-06,
"loss": 0.3064204156398773,
"step": 1240
},
{
"epoch": 1.215659845607156,
"grad_norm": 0.30686894059181213,
"learning_rate": 3.5305566530267217e-06,
"loss": 0.3445843756198883,
"step": 1241
},
{
"epoch": 1.2166401176326431,
"grad_norm": 0.42608144879341125,
"learning_rate": 3.5229755289890005e-06,
"loss": 0.3761751651763916,
"step": 1242
},
{
"epoch": 1.2176203896581301,
"grad_norm": 0.39542150497436523,
"learning_rate": 3.5153981233586277e-06,
"loss": 0.31900158524513245,
"step": 1243
},
{
"epoch": 1.2186006616836171,
"grad_norm": 0.25779104232788086,
"learning_rate": 3.507824455211708e-06,
"loss": 0.3107672333717346,
"step": 1244
},
{
"epoch": 1.2195809337091044,
"grad_norm": 0.3186492621898651,
"learning_rate": 3.5002545436149478e-06,
"loss": 0.24999463558197021,
"step": 1245
},
{
"epoch": 1.2205612057345914,
"grad_norm": 0.34163543581962585,
"learning_rate": 3.4926884076255864e-06,
"loss": 0.3079703748226166,
"step": 1246
},
{
"epoch": 1.2215414777600784,
"grad_norm": 0.321964830160141,
"learning_rate": 3.4851260662913643e-06,
"loss": 0.32029277086257935,
"step": 1247
},
{
"epoch": 1.2225217497855656,
"grad_norm": 0.33328044414520264,
"learning_rate": 3.477567538650466e-06,
"loss": 0.3529759645462036,
"step": 1248
},
{
"epoch": 1.2235020218110526,
"grad_norm": 0.3267224431037903,
"learning_rate": 3.470012843731476e-06,
"loss": 0.38808560371398926,
"step": 1249
},
{
"epoch": 1.2244822938365396,
"grad_norm": 0.26567623019218445,
"learning_rate": 3.4624620005533273e-06,
"loss": 0.23895587027072906,
"step": 1250
},
{
"epoch": 1.2254625658620266,
"grad_norm": 0.2990274727344513,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.28610265254974365,
"step": 1251
},
{
"epoch": 1.2264428378875138,
"grad_norm": 0.2755870521068573,
"learning_rate": 3.4473719454467747e-06,
"loss": 0.2870532274246216,
"step": 1252
},
{
"epoch": 1.2274231099130009,
"grad_norm": 0.2697540521621704,
"learning_rate": 3.439832771507565e-06,
"loss": 0.28703543543815613,
"step": 1253
},
{
"epoch": 1.2284033819384879,
"grad_norm": 0.36372098326683044,
"learning_rate": 3.432297525287495e-06,
"loss": 0.3571392595767975,
"step": 1254
},
{
"epoch": 1.229383653963975,
"grad_norm": 0.680644154548645,
"learning_rate": 3.4247662257565372e-06,
"loss": 0.3493378460407257,
"step": 1255
},
{
"epoch": 1.230363925989462,
"grad_norm": 0.3246632218360901,
"learning_rate": 3.4172388918747278e-06,
"loss": 0.3083442747592926,
"step": 1256
},
{
"epoch": 1.231344198014949,
"grad_norm": 0.2493603378534317,
"learning_rate": 3.4097155425921256e-06,
"loss": 0.30239057540893555,
"step": 1257
},
{
"epoch": 1.232324470040436,
"grad_norm": 0.5045711398124695,
"learning_rate": 3.402196196848751e-06,
"loss": 0.3329586684703827,
"step": 1258
},
{
"epoch": 1.2333047420659233,
"grad_norm": 0.4852418899536133,
"learning_rate": 3.394680873574546e-06,
"loss": 0.3268454372882843,
"step": 1259
},
{
"epoch": 1.2342850140914103,
"grad_norm": 0.2629272937774658,
"learning_rate": 3.3871695916893314e-06,
"loss": 0.310479074716568,
"step": 1260
},
{
"epoch": 1.2352652861168973,
"grad_norm": 0.35888341069221497,
"learning_rate": 3.3796623701027477e-06,
"loss": 0.30436229705810547,
"step": 1261
},
{
"epoch": 1.2362455581423846,
"grad_norm": 0.33824774622917175,
"learning_rate": 3.372159227714218e-06,
"loss": 0.32333600521087646,
"step": 1262
},
{
"epoch": 1.2372258301678716,
"grad_norm": 0.3361474871635437,
"learning_rate": 3.3646601834128924e-06,
"loss": 0.3319692015647888,
"step": 1263
},
{
"epoch": 1.2382061021933586,
"grad_norm": 0.36827853322029114,
"learning_rate": 3.357165256077608e-06,
"loss": 0.26803964376449585,
"step": 1264
},
{
"epoch": 1.2391863742188458,
"grad_norm": 0.43026289343833923,
"learning_rate": 3.349674464576834e-06,
"loss": 0.3155477046966553,
"step": 1265
},
{
"epoch": 1.2401666462443328,
"grad_norm": 0.29646962881088257,
"learning_rate": 3.3421878277686315e-06,
"loss": 0.31426000595092773,
"step": 1266
},
{
"epoch": 1.2411469182698198,
"grad_norm": 0.5364441275596619,
"learning_rate": 3.3347053645005965e-06,
"loss": 0.37239736318588257,
"step": 1267
},
{
"epoch": 1.242127190295307,
"grad_norm": 0.26814860105514526,
"learning_rate": 3.3272270936098243e-06,
"loss": 0.2982516288757324,
"step": 1268
},
{
"epoch": 1.243107462320794,
"grad_norm": 0.2801300883293152,
"learning_rate": 3.319753033922849e-06,
"loss": 0.31472277641296387,
"step": 1269
},
{
"epoch": 1.244087734346281,
"grad_norm": 0.4047705829143524,
"learning_rate": 3.3122832042556118e-06,
"loss": 0.31080782413482666,
"step": 1270
},
{
"epoch": 1.2450680063717683,
"grad_norm": 0.39710474014282227,
"learning_rate": 3.3048176234133967e-06,
"loss": 0.30704551935195923,
"step": 1271
},
{
"epoch": 1.2460482783972553,
"grad_norm": 0.3281836211681366,
"learning_rate": 3.297356310190797e-06,
"loss": 0.31868213415145874,
"step": 1272
},
{
"epoch": 1.2470285504227423,
"grad_norm": 0.3261207044124603,
"learning_rate": 3.289899283371657e-06,
"loss": 0.3119394779205322,
"step": 1273
},
{
"epoch": 1.2480088224482293,
"grad_norm": 0.29138749837875366,
"learning_rate": 3.282446561729035e-06,
"loss": 0.33546704053878784,
"step": 1274
},
{
"epoch": 1.2489890944737165,
"grad_norm": 0.28828164935112,
"learning_rate": 3.274998164025148e-06,
"loss": 0.28582191467285156,
"step": 1275
},
{
"epoch": 1.2499693664992035,
"grad_norm": 0.36214062571525574,
"learning_rate": 3.267554109011327e-06,
"loss": 0.3242008686065674,
"step": 1276
},
{
"epoch": 1.2509496385246905,
"grad_norm": 0.26558277010917664,
"learning_rate": 3.260114415427975e-06,
"loss": 0.2799558639526367,
"step": 1277
},
{
"epoch": 1.2519299105501775,
"grad_norm": 0.3024544417858124,
"learning_rate": 3.252679102004509e-06,
"loss": 0.31685107946395874,
"step": 1278
},
{
"epoch": 1.2529101825756648,
"grad_norm": 0.29641222953796387,
"learning_rate": 3.2452481874593234e-06,
"loss": 0.339430034160614,
"step": 1279
},
{
"epoch": 1.2538904546011518,
"grad_norm": 1.9632152318954468,
"learning_rate": 3.2378216904997345e-06,
"loss": 0.3519091010093689,
"step": 1280
},
{
"epoch": 1.2548707266266388,
"grad_norm": 0.2878991365432739,
"learning_rate": 3.230399629821942e-06,
"loss": 0.32278361916542053,
"step": 1281
},
{
"epoch": 1.255850998652126,
"grad_norm": 0.31732842326164246,
"learning_rate": 3.2229820241109722e-06,
"loss": 0.3337588906288147,
"step": 1282
},
{
"epoch": 1.256831270677613,
"grad_norm": 0.2868574559688568,
"learning_rate": 3.2155688920406415e-06,
"loss": 0.29180842638015747,
"step": 1283
},
{
"epoch": 1.2578115427031,
"grad_norm": 0.5311480164527893,
"learning_rate": 3.2081602522734987e-06,
"loss": 0.36810076236724854,
"step": 1284
},
{
"epoch": 1.2587918147285873,
"grad_norm": 0.2833496332168579,
"learning_rate": 3.200756123460788e-06,
"loss": 0.32042837142944336,
"step": 1285
},
{
"epoch": 1.2597720867540743,
"grad_norm": 0.30482539534568787,
"learning_rate": 3.1933565242423925e-06,
"loss": 0.27995723485946655,
"step": 1286
},
{
"epoch": 1.2607523587795613,
"grad_norm": 0.3214282989501953,
"learning_rate": 3.1859614732467957e-06,
"loss": 0.39340633153915405,
"step": 1287
},
{
"epoch": 1.2617326308050485,
"grad_norm": 0.30645492672920227,
"learning_rate": 3.178570989091028e-06,
"loss": 0.3645610213279724,
"step": 1288
},
{
"epoch": 1.2627129028305355,
"grad_norm": 0.26311758160591125,
"learning_rate": 3.171185090380628e-06,
"loss": 0.3065904378890991,
"step": 1289
},
{
"epoch": 1.2636931748560225,
"grad_norm": 0.26868557929992676,
"learning_rate": 3.1638037957095836e-06,
"loss": 0.322699636220932,
"step": 1290
},
{
"epoch": 1.2646734468815097,
"grad_norm": 0.3596718907356262,
"learning_rate": 3.156427123660297e-06,
"loss": 0.33911019563674927,
"step": 1291
},
{
"epoch": 1.2656537189069967,
"grad_norm": 0.3089102804660797,
"learning_rate": 3.1490550928035312e-06,
"loss": 0.3215683400630951,
"step": 1292
},
{
"epoch": 1.2666339909324837,
"grad_norm": 0.29380035400390625,
"learning_rate": 3.141687721698363e-06,
"loss": 0.3198908865451813,
"step": 1293
},
{
"epoch": 1.267614262957971,
"grad_norm": 0.3018144965171814,
"learning_rate": 3.1343250288921438e-06,
"loss": 0.26924243569374084,
"step": 1294
},
{
"epoch": 1.268594534983458,
"grad_norm": 0.3356236219406128,
"learning_rate": 3.12696703292044e-06,
"loss": 0.3623160719871521,
"step": 1295
},
{
"epoch": 1.269574807008945,
"grad_norm": 0.3001836836338043,
"learning_rate": 3.119613752307002e-06,
"loss": 0.32793790102005005,
"step": 1296
},
{
"epoch": 1.270555079034432,
"grad_norm": 0.2888118028640747,
"learning_rate": 3.1122652055637014e-06,
"loss": 0.3076667785644531,
"step": 1297
},
{
"epoch": 1.271535351059919,
"grad_norm": 0.3300588130950928,
"learning_rate": 3.104921411190499e-06,
"loss": 0.24012883007526398,
"step": 1298
},
{
"epoch": 1.2725156230854062,
"grad_norm": 1.737196922302246,
"learning_rate": 3.097582387675385e-06,
"loss": 0.3130298852920532,
"step": 1299
},
{
"epoch": 1.2734958951108932,
"grad_norm": 0.3157992959022522,
"learning_rate": 3.0902481534943464e-06,
"loss": 0.3188842535018921,
"step": 1300
},
{
"epoch": 1.2744761671363802,
"grad_norm": 0.3161185085773468,
"learning_rate": 3.0829187271113035e-06,
"loss": 0.35474109649658203,
"step": 1301
},
{
"epoch": 1.2754564391618675,
"grad_norm": 0.3295031487941742,
"learning_rate": 3.075594126978084e-06,
"loss": 0.4010552763938904,
"step": 1302
},
{
"epoch": 1.2764367111873545,
"grad_norm": 0.33576318621635437,
"learning_rate": 3.0682743715343565e-06,
"loss": 0.2657952606678009,
"step": 1303
},
{
"epoch": 1.2774169832128415,
"grad_norm": 0.31075185537338257,
"learning_rate": 3.0609594792076e-06,
"loss": 0.27664023637771606,
"step": 1304
},
{
"epoch": 1.2783972552383287,
"grad_norm": 0.29165032505989075,
"learning_rate": 3.053649468413043e-06,
"loss": 0.308698832988739,
"step": 1305
},
{
"epoch": 1.2793775272638157,
"grad_norm": 0.3423086106777191,
"learning_rate": 3.0463443575536324e-06,
"loss": 0.35392165184020996,
"step": 1306
},
{
"epoch": 1.2803577992893027,
"grad_norm": 0.2879369556903839,
"learning_rate": 3.0390441650199727e-06,
"loss": 0.2656344771385193,
"step": 1307
},
{
"epoch": 1.28133807131479,
"grad_norm": 0.2722434401512146,
"learning_rate": 3.0317489091902936e-06,
"loss": 0.3026173710823059,
"step": 1308
},
{
"epoch": 1.282318343340277,
"grad_norm": 0.3254748582839966,
"learning_rate": 3.0244586084303908e-06,
"loss": 0.31657737493515015,
"step": 1309
},
{
"epoch": 1.283298615365764,
"grad_norm": 0.319442480802536,
"learning_rate": 3.0171732810935887e-06,
"loss": 0.3103172779083252,
"step": 1310
},
{
"epoch": 1.2842788873912512,
"grad_norm": 0.8107593059539795,
"learning_rate": 3.0098929455206905e-06,
"loss": 0.2703472375869751,
"step": 1311
},
{
"epoch": 1.2852591594167382,
"grad_norm": 0.2706098258495331,
"learning_rate": 3.00261762003993e-06,
"loss": 0.29308193922042847,
"step": 1312
},
{
"epoch": 1.2862394314422252,
"grad_norm": 0.3512340486049652,
"learning_rate": 2.995347322966933e-06,
"loss": 0.3645230531692505,
"step": 1313
},
{
"epoch": 1.2872197034677124,
"grad_norm": 0.3050675690174103,
"learning_rate": 2.9880820726046613e-06,
"loss": 0.3009169101715088,
"step": 1314
},
{
"epoch": 1.2881999754931994,
"grad_norm": 0.26588180661201477,
"learning_rate": 2.980821887243377e-06,
"loss": 0.29008394479751587,
"step": 1315
},
{
"epoch": 1.2891802475186864,
"grad_norm": 0.3032967746257782,
"learning_rate": 2.9735667851605855e-06,
"loss": 0.32896995544433594,
"step": 1316
},
{
"epoch": 1.2901605195441734,
"grad_norm": 0.3087082803249359,
"learning_rate": 2.966316784621e-06,
"loss": 0.2642725110054016,
"step": 1317
},
{
"epoch": 1.2911407915696607,
"grad_norm": 0.5888882875442505,
"learning_rate": 2.959071903876486e-06,
"loss": 0.30377838015556335,
"step": 1318
},
{
"epoch": 1.2921210635951477,
"grad_norm": 0.30952659249305725,
"learning_rate": 2.951832161166024e-06,
"loss": 0.26346418261528015,
"step": 1319
},
{
"epoch": 1.2931013356206347,
"grad_norm": 0.36168810725212097,
"learning_rate": 2.9445975747156548e-06,
"loss": 0.36242181062698364,
"step": 1320
},
{
"epoch": 1.2940816076461217,
"grad_norm": 0.6191869378089905,
"learning_rate": 2.937368162738445e-06,
"loss": 0.29357028007507324,
"step": 1321
},
{
"epoch": 1.295061879671609,
"grad_norm": 0.40403035283088684,
"learning_rate": 2.9301439434344275e-06,
"loss": 0.3663826882839203,
"step": 1322
},
{
"epoch": 1.296042151697096,
"grad_norm": 0.2266947329044342,
"learning_rate": 2.9229249349905686e-06,
"loss": 0.2353021502494812,
"step": 1323
},
{
"epoch": 1.297022423722583,
"grad_norm": 0.27807435393333435,
"learning_rate": 2.9157111555807103e-06,
"loss": 0.26985543966293335,
"step": 1324
},
{
"epoch": 1.2980026957480701,
"grad_norm": 0.28705501556396484,
"learning_rate": 2.9085026233655367e-06,
"loss": 0.2721678614616394,
"step": 1325
},
{
"epoch": 1.2989829677735572,
"grad_norm": 0.3892923891544342,
"learning_rate": 2.901299356492516e-06,
"loss": 0.30747634172439575,
"step": 1326
},
{
"epoch": 1.2999632397990442,
"grad_norm": 0.25401294231414795,
"learning_rate": 2.8941013730958674e-06,
"loss": 0.3035070300102234,
"step": 1327
},
{
"epoch": 1.3009435118245314,
"grad_norm": 0.5337584018707275,
"learning_rate": 2.886908691296504e-06,
"loss": 0.28601205348968506,
"step": 1328
},
{
"epoch": 1.3019237838500184,
"grad_norm": 0.28078943490982056,
"learning_rate": 2.8797213292019927e-06,
"loss": 0.3016660213470459,
"step": 1329
},
{
"epoch": 1.3029040558755054,
"grad_norm": 0.3134513199329376,
"learning_rate": 2.8725393049065098e-06,
"loss": 0.32474619150161743,
"step": 1330
},
{
"epoch": 1.3038843279009926,
"grad_norm": 0.28749099373817444,
"learning_rate": 2.8653626364907918e-06,
"loss": 0.3196799159049988,
"step": 1331
},
{
"epoch": 1.3048645999264796,
"grad_norm": 0.3583281636238098,
"learning_rate": 2.8581913420220954e-06,
"loss": 0.2746082842350006,
"step": 1332
},
{
"epoch": 1.3058448719519666,
"grad_norm": 0.39170408248901367,
"learning_rate": 2.851025439554142e-06,
"loss": 0.3381309509277344,
"step": 1333
},
{
"epoch": 1.3068251439774539,
"grad_norm": 0.2849377691745758,
"learning_rate": 2.8438649471270876e-06,
"loss": 0.2811689078807831,
"step": 1334
},
{
"epoch": 1.3078054160029409,
"grad_norm": 0.24796049296855927,
"learning_rate": 2.8367098827674575e-06,
"loss": 0.28159859776496887,
"step": 1335
},
{
"epoch": 1.3087856880284279,
"grad_norm": 0.30984076857566833,
"learning_rate": 2.829560264488126e-06,
"loss": 0.2774086892604828,
"step": 1336
},
{
"epoch": 1.3097659600539149,
"grad_norm": 0.3411960303783417,
"learning_rate": 2.82241611028824e-06,
"loss": 0.3824862241744995,
"step": 1337
},
{
"epoch": 1.310746232079402,
"grad_norm": 0.32069236040115356,
"learning_rate": 2.8152774381532033e-06,
"loss": 0.3920864462852478,
"step": 1338
},
{
"epoch": 1.3117265041048891,
"grad_norm": 0.3489685654640198,
"learning_rate": 2.8081442660546126e-06,
"loss": 0.2715360224246979,
"step": 1339
},
{
"epoch": 1.3127067761303761,
"grad_norm": 0.2713499367237091,
"learning_rate": 2.8010166119502238e-06,
"loss": 0.2965244650840759,
"step": 1340
},
{
"epoch": 1.3136870481558631,
"grad_norm": 0.6711511015892029,
"learning_rate": 2.7938944937838924e-06,
"loss": 0.2506905794143677,
"step": 1341
},
{
"epoch": 1.3146673201813504,
"grad_norm": 0.26067718863487244,
"learning_rate": 2.7867779294855502e-06,
"loss": 0.29038554430007935,
"step": 1342
},
{
"epoch": 1.3156475922068374,
"grad_norm": 0.3031093180179596,
"learning_rate": 2.7796669369711294e-06,
"loss": 0.3261447250843048,
"step": 1343
},
{
"epoch": 1.3166278642323244,
"grad_norm": 0.29175373911857605,
"learning_rate": 2.7725615341425526e-06,
"loss": 0.33107811212539673,
"step": 1344
},
{
"epoch": 1.3176081362578116,
"grad_norm": 0.7521331310272217,
"learning_rate": 2.7654617388876612e-06,
"loss": 0.33864927291870117,
"step": 1345
},
{
"epoch": 1.3185884082832986,
"grad_norm": 0.29048389196395874,
"learning_rate": 2.758367569080178e-06,
"loss": 0.3234966993331909,
"step": 1346
},
{
"epoch": 1.3195686803087856,
"grad_norm": 0.35268884897232056,
"learning_rate": 2.751279042579672e-06,
"loss": 0.33619391918182373,
"step": 1347
},
{
"epoch": 1.3205489523342728,
"grad_norm": 0.3249227702617645,
"learning_rate": 2.744196177231498e-06,
"loss": 0.2532762885093689,
"step": 1348
},
{
"epoch": 1.3215292243597598,
"grad_norm": 0.4151158034801483,
"learning_rate": 2.7371189908667604e-06,
"loss": 0.31017154455184937,
"step": 1349
},
{
"epoch": 1.3225094963852468,
"grad_norm": 0.3690338730812073,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.2900843024253845,
"step": 1350
},
{
"epoch": 1.323489768410734,
"grad_norm": 0.2888542413711548,
"learning_rate": 2.722981726340487e-06,
"loss": 0.3149188160896301,
"step": 1351
},
{
"epoch": 1.324470040436221,
"grad_norm": 0.3167160153388977,
"learning_rate": 2.715921683769496e-06,
"loss": 0.2912067770957947,
"step": 1352
},
{
"epoch": 1.325450312461708,
"grad_norm": 0.27939552068710327,
"learning_rate": 2.708867391362948e-06,
"loss": 0.2959906756877899,
"step": 1353
},
{
"epoch": 1.3264305844871953,
"grad_norm": 0.4946768283843994,
"learning_rate": 2.7018188668800123e-06,
"loss": 0.32098856568336487,
"step": 1354
},
{
"epoch": 1.3274108565126823,
"grad_norm": 0.29626569151878357,
"learning_rate": 2.694776128065345e-06,
"loss": 0.2851714789867401,
"step": 1355
},
{
"epoch": 1.3283911285381693,
"grad_norm": 0.31539374589920044,
"learning_rate": 2.687739192649026e-06,
"loss": 0.28555694222450256,
"step": 1356
},
{
"epoch": 1.3293714005636565,
"grad_norm": 0.2927304804325104,
"learning_rate": 2.6807080783465376e-06,
"loss": 0.27567166090011597,
"step": 1357
},
{
"epoch": 1.3303516725891436,
"grad_norm": 0.328565776348114,
"learning_rate": 2.673682802858697e-06,
"loss": 0.27756747603416443,
"step": 1358
},
{
"epoch": 1.3313319446146306,
"grad_norm": 0.28639981150627136,
"learning_rate": 2.6666633838716317e-06,
"loss": 0.3055470883846283,
"step": 1359
},
{
"epoch": 1.3323122166401176,
"grad_norm": 0.3049262464046478,
"learning_rate": 2.659649839056716e-06,
"loss": 0.3017973303794861,
"step": 1360
},
{
"epoch": 1.3332924886656046,
"grad_norm": 0.3556685149669647,
"learning_rate": 2.6526421860705474e-06,
"loss": 0.3087649941444397,
"step": 1361
},
{
"epoch": 1.3342727606910918,
"grad_norm": 0.8926109075546265,
"learning_rate": 2.645640442554878e-06,
"loss": 0.26090800762176514,
"step": 1362
},
{
"epoch": 1.3352530327165788,
"grad_norm": 0.2733709514141083,
"learning_rate": 2.6386446261365874e-06,
"loss": 0.2870309054851532,
"step": 1363
},
{
"epoch": 1.3362333047420658,
"grad_norm": 0.25905874371528625,
"learning_rate": 2.63165475442764e-06,
"loss": 0.2933259606361389,
"step": 1364
},
{
"epoch": 1.337213576767553,
"grad_norm": 0.389791876077652,
"learning_rate": 2.6246708450250256e-06,
"loss": 0.30408746004104614,
"step": 1365
},
{
"epoch": 1.33819384879304,
"grad_norm": 0.4030923545360565,
"learning_rate": 2.617692915510732e-06,
"loss": 0.2746540904045105,
"step": 1366
},
{
"epoch": 1.339174120818527,
"grad_norm": 0.273571640253067,
"learning_rate": 2.6107209834516857e-06,
"loss": 0.3073597252368927,
"step": 1367
},
{
"epoch": 1.3401543928440143,
"grad_norm": 0.2633684575557709,
"learning_rate": 2.603755066399718e-06,
"loss": 0.3117210865020752,
"step": 1368
},
{
"epoch": 1.3411346648695013,
"grad_norm": 0.2624060809612274,
"learning_rate": 2.5967951818915137e-06,
"loss": 0.3079099953174591,
"step": 1369
},
{
"epoch": 1.3421149368949883,
"grad_norm": 0.3099723160266876,
"learning_rate": 2.5898413474485797e-06,
"loss": 0.29641491174697876,
"step": 1370
},
{
"epoch": 1.3430952089204755,
"grad_norm": 0.31890374422073364,
"learning_rate": 2.5828935805771804e-06,
"loss": 0.353140652179718,
"step": 1371
},
{
"epoch": 1.3440754809459625,
"grad_norm": 0.31002503633499146,
"learning_rate": 2.5759518987683154e-06,
"loss": 0.27937400341033936,
"step": 1372
},
{
"epoch": 1.3450557529714495,
"grad_norm": 0.2602542042732239,
"learning_rate": 2.5690163194976576e-06,
"loss": 0.2914687991142273,
"step": 1373
},
{
"epoch": 1.3460360249969368,
"grad_norm": 0.41703489422798157,
"learning_rate": 2.56208686022552e-06,
"loss": 0.3168262541294098,
"step": 1374
},
{
"epoch": 1.3470162970224238,
"grad_norm": 0.4128677248954773,
"learning_rate": 2.5551635383968063e-06,
"loss": 0.27541229128837585,
"step": 1375
},
{
"epoch": 1.3479965690479108,
"grad_norm": 0.41166117787361145,
"learning_rate": 2.5482463714409744e-06,
"loss": 0.27447646856307983,
"step": 1376
},
{
"epoch": 1.348976841073398,
"grad_norm": 0.7910805344581604,
"learning_rate": 2.5413353767719805e-06,
"loss": 0.3449074923992157,
"step": 1377
},
{
"epoch": 1.349957113098885,
"grad_norm": 0.2744949758052826,
"learning_rate": 2.5344305717882487e-06,
"loss": 0.25335192680358887,
"step": 1378
},
{
"epoch": 1.350937385124372,
"grad_norm": 0.34028032422065735,
"learning_rate": 2.527531973872617e-06,
"loss": 0.27056068181991577,
"step": 1379
},
{
"epoch": 1.351917657149859,
"grad_norm": 0.4440552592277527,
"learning_rate": 2.520639600392295e-06,
"loss": 0.3466051518917084,
"step": 1380
},
{
"epoch": 1.3528979291753462,
"grad_norm": 1.3573254346847534,
"learning_rate": 2.5137534686988265e-06,
"loss": 0.3443021774291992,
"step": 1381
},
{
"epoch": 1.3538782012008332,
"grad_norm": 0.26375842094421387,
"learning_rate": 2.5068735961280367e-06,
"loss": 0.24238917231559753,
"step": 1382
},
{
"epoch": 1.3548584732263202,
"grad_norm": 0.30860915780067444,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.3332935571670532,
"step": 1383
},
{
"epoch": 1.3558387452518073,
"grad_norm": 0.2814854681491852,
"learning_rate": 2.493132697618986e-06,
"loss": 0.26244714856147766,
"step": 1384
},
{
"epoch": 1.3568190172772945,
"grad_norm": 0.2804390490055084,
"learning_rate": 2.486271706273421e-06,
"loss": 0.24664735794067383,
"step": 1385
},
{
"epoch": 1.3577992893027815,
"grad_norm": 0.5321319103240967,
"learning_rate": 2.4794170432358415e-06,
"loss": 0.2858157753944397,
"step": 1386
},
{
"epoch": 1.3587795613282685,
"grad_norm": 0.2946712374687195,
"learning_rate": 2.4725687257628533e-06,
"loss": 0.3303595781326294,
"step": 1387
},
{
"epoch": 1.3597598333537557,
"grad_norm": 0.34516823291778564,
"learning_rate": 2.465726771095086e-06,
"loss": 0.31634557247161865,
"step": 1388
},
{
"epoch": 1.3607401053792427,
"grad_norm": 0.24505671858787537,
"learning_rate": 2.4588911964571557e-06,
"loss": 0.2866356372833252,
"step": 1389
},
{
"epoch": 1.3617203774047297,
"grad_norm": 0.35814908146858215,
"learning_rate": 2.4520620190576094e-06,
"loss": 0.33161798119544983,
"step": 1390
},
{
"epoch": 1.362700649430217,
"grad_norm": 0.3869876265525818,
"learning_rate": 2.445239256088898e-06,
"loss": 0.3012307286262512,
"step": 1391
},
{
"epoch": 1.363680921455704,
"grad_norm": 0.2874414622783661,
"learning_rate": 2.4384229247273154e-06,
"loss": 0.2863731384277344,
"step": 1392
},
{
"epoch": 1.364661193481191,
"grad_norm": 0.311261922121048,
"learning_rate": 2.4316130421329696e-06,
"loss": 0.32147401571273804,
"step": 1393
},
{
"epoch": 1.3656414655066782,
"grad_norm": 0.27934855222702026,
"learning_rate": 2.424809625449729e-06,
"loss": 0.29047897458076477,
"step": 1394
},
{
"epoch": 1.3666217375321652,
"grad_norm": 0.6079840064048767,
"learning_rate": 2.418012691805191e-06,
"loss": 0.34937167167663574,
"step": 1395
},
{
"epoch": 1.3676020095576522,
"grad_norm": 0.31358763575553894,
"learning_rate": 2.411222258310626e-06,
"loss": 0.2945541739463806,
"step": 1396
},
{
"epoch": 1.3685822815831394,
"grad_norm": 0.35368451476097107,
"learning_rate": 2.404438342060941e-06,
"loss": 0.3971101641654968,
"step": 1397
},
{
"epoch": 1.3695625536086264,
"grad_norm": 0.3225327134132385,
"learning_rate": 2.3976609601346395e-06,
"loss": 0.31740427017211914,
"step": 1398
},
{
"epoch": 1.3705428256341134,
"grad_norm": 0.2605196535587311,
"learning_rate": 2.3908901295937713e-06,
"loss": 0.3151897192001343,
"step": 1399
},
{
"epoch": 1.3715230976596007,
"grad_norm": 0.3012634813785553,
"learning_rate": 2.3841258674838948e-06,
"loss": 0.2970534563064575,
"step": 1400
},
{
"epoch": 1.3725033696850877,
"grad_norm": 0.32142847776412964,
"learning_rate": 2.3773681908340284e-06,
"loss": 0.4133338928222656,
"step": 1401
},
{
"epoch": 1.3734836417105747,
"grad_norm": 0.3075603246688843,
"learning_rate": 2.3706171166566204e-06,
"loss": 0.2857438921928406,
"step": 1402
},
{
"epoch": 1.3744639137360617,
"grad_norm": 0.3162294030189514,
"learning_rate": 2.363872661947488e-06,
"loss": 0.3295277953147888,
"step": 1403
},
{
"epoch": 1.3754441857615487,
"grad_norm": 0.24163271486759186,
"learning_rate": 2.3571348436857906e-06,
"loss": 0.27518129348754883,
"step": 1404
},
{
"epoch": 1.376424457787036,
"grad_norm": 0.38715091347694397,
"learning_rate": 2.3504036788339763e-06,
"loss": 0.3373250365257263,
"step": 1405
},
{
"epoch": 1.377404729812523,
"grad_norm": 0.2635200619697571,
"learning_rate": 2.343679184337745e-06,
"loss": 0.30227527022361755,
"step": 1406
},
{
"epoch": 1.37838500183801,
"grad_norm": 0.2817568778991699,
"learning_rate": 2.3369613771260006e-06,
"loss": 0.3550269305706024,
"step": 1407
},
{
"epoch": 1.3793652738634972,
"grad_norm": 0.4502047300338745,
"learning_rate": 2.33025027411082e-06,
"loss": 0.2930881381034851,
"step": 1408
},
{
"epoch": 1.3803455458889842,
"grad_norm": 0.25602367520332336,
"learning_rate": 2.323545892187393e-06,
"loss": 0.26259949803352356,
"step": 1409
},
{
"epoch": 1.3813258179144712,
"grad_norm": 0.3045177161693573,
"learning_rate": 2.3168482482339955e-06,
"loss": 0.35001400113105774,
"step": 1410
},
{
"epoch": 1.3823060899399584,
"grad_norm": 0.4008457064628601,
"learning_rate": 2.310157359111938e-06,
"loss": 0.30838537216186523,
"step": 1411
},
{
"epoch": 1.3832863619654454,
"grad_norm": 0.31439995765686035,
"learning_rate": 2.3034732416655247e-06,
"loss": 0.33484673500061035,
"step": 1412
},
{
"epoch": 1.3842666339909324,
"grad_norm": 0.3156680166721344,
"learning_rate": 2.296795912722014e-06,
"loss": 0.3417940139770508,
"step": 1413
},
{
"epoch": 1.3852469060164196,
"grad_norm": 0.28939467668533325,
"learning_rate": 2.2901253890915714e-06,
"loss": 0.3140500783920288,
"step": 1414
},
{
"epoch": 1.3862271780419066,
"grad_norm": 0.2982892096042633,
"learning_rate": 2.2834616875672362e-06,
"loss": 0.33072012662887573,
"step": 1415
},
{
"epoch": 1.3872074500673937,
"grad_norm": 0.29197433590888977,
"learning_rate": 2.2768048249248648e-06,
"loss": 0.2815811038017273,
"step": 1416
},
{
"epoch": 1.3881877220928809,
"grad_norm": 0.27484625577926636,
"learning_rate": 2.2701548179231048e-06,
"loss": 0.28860288858413696,
"step": 1417
},
{
"epoch": 1.3891679941183679,
"grad_norm": 0.3656412363052368,
"learning_rate": 2.263511683303339e-06,
"loss": 0.42425838112831116,
"step": 1418
},
{
"epoch": 1.390148266143855,
"grad_norm": 0.26825398206710815,
"learning_rate": 2.2568754377896516e-06,
"loss": 0.276248961687088,
"step": 1419
},
{
"epoch": 1.3911285381693421,
"grad_norm": 0.2911313474178314,
"learning_rate": 2.2502460980887807e-06,
"loss": 0.3648977279663086,
"step": 1420
},
{
"epoch": 1.3921088101948291,
"grad_norm": 0.477396160364151,
"learning_rate": 2.2436236808900846e-06,
"loss": 0.3318394422531128,
"step": 1421
},
{
"epoch": 1.3930890822203161,
"grad_norm": 0.369213342666626,
"learning_rate": 2.2370082028654867e-06,
"loss": 0.3071008324623108,
"step": 1422
},
{
"epoch": 1.3940693542458031,
"grad_norm": 0.6697940230369568,
"learning_rate": 2.230399680669449e-06,
"loss": 0.3358566164970398,
"step": 1423
},
{
"epoch": 1.3950496262712904,
"grad_norm": 0.30832651257514954,
"learning_rate": 2.223798130938916e-06,
"loss": 0.3450663685798645,
"step": 1424
},
{
"epoch": 1.3960298982967774,
"grad_norm": 0.34308820962905884,
"learning_rate": 2.2172035702932828e-06,
"loss": 0.37836942076683044,
"step": 1425
},
{
"epoch": 1.3970101703222644,
"grad_norm": 0.26579222083091736,
"learning_rate": 2.210616015334344e-06,
"loss": 0.28905928134918213,
"step": 1426
},
{
"epoch": 1.3979904423477514,
"grad_norm": 0.3129562735557556,
"learning_rate": 2.204035482646267e-06,
"loss": 0.3114771544933319,
"step": 1427
},
{
"epoch": 1.3989707143732386,
"grad_norm": 0.23969358205795288,
"learning_rate": 2.1974619887955294e-06,
"loss": 0.2918145954608917,
"step": 1428
},
{
"epoch": 1.3999509863987256,
"grad_norm": 0.29445183277130127,
"learning_rate": 2.190895550330899e-06,
"loss": 0.32057255506515503,
"step": 1429
},
{
"epoch": 1.4009312584242126,
"grad_norm": 0.31180688738822937,
"learning_rate": 2.1843361837833755e-06,
"loss": 0.3145178556442261,
"step": 1430
},
{
"epoch": 1.4019115304496998,
"grad_norm": 0.29496511816978455,
"learning_rate": 2.1777839056661555e-06,
"loss": 0.31752365827560425,
"step": 1431
},
{
"epoch": 1.4028918024751869,
"grad_norm": 0.34286579489707947,
"learning_rate": 2.171238732474591e-06,
"loss": 0.40405502915382385,
"step": 1432
},
{
"epoch": 1.4038720745006739,
"grad_norm": 0.31270524859428406,
"learning_rate": 2.1647006806861472e-06,
"loss": 0.3145411014556885,
"step": 1433
},
{
"epoch": 1.404852346526161,
"grad_norm": 0.7426226735115051,
"learning_rate": 2.1581697667603634e-06,
"loss": 0.33794814348220825,
"step": 1434
},
{
"epoch": 1.405832618551648,
"grad_norm": 0.35654860734939575,
"learning_rate": 2.1516460071388062e-06,
"loss": 0.30767810344696045,
"step": 1435
},
{
"epoch": 1.406812890577135,
"grad_norm": 0.3136668801307678,
"learning_rate": 2.145129418245035e-06,
"loss": 0.3266030550003052,
"step": 1436
},
{
"epoch": 1.4077931626026223,
"grad_norm": 0.32115745544433594,
"learning_rate": 2.1386200164845527e-06,
"loss": 0.3516904413700104,
"step": 1437
},
{
"epoch": 1.4087734346281093,
"grad_norm": 0.2977611720561981,
"learning_rate": 2.132117818244771e-06,
"loss": 0.337024450302124,
"step": 1438
},
{
"epoch": 1.4097537066535963,
"grad_norm": 0.30378955602645874,
"learning_rate": 2.125622839894964e-06,
"loss": 0.3231973946094513,
"step": 1439
},
{
"epoch": 1.4107339786790836,
"grad_norm": 0.28581324219703674,
"learning_rate": 2.1191350977862362e-06,
"loss": 0.2765822410583496,
"step": 1440
},
{
"epoch": 1.4117142507045706,
"grad_norm": 0.2685025930404663,
"learning_rate": 2.1126546082514665e-06,
"loss": 0.2733510732650757,
"step": 1441
},
{
"epoch": 1.4126945227300576,
"grad_norm": 0.3476353883743286,
"learning_rate": 2.106181387605283e-06,
"loss": 0.37086084485054016,
"step": 1442
},
{
"epoch": 1.4136747947555446,
"grad_norm": 0.6860825419425964,
"learning_rate": 2.09971545214401e-06,
"loss": 0.3570191264152527,
"step": 1443
},
{
"epoch": 1.4146550667810318,
"grad_norm": 0.23985540866851807,
"learning_rate": 2.0932568181456314e-06,
"loss": 0.2715319097042084,
"step": 1444
},
{
"epoch": 1.4156353388065188,
"grad_norm": 0.26984402537345886,
"learning_rate": 2.086805501869749e-06,
"loss": 0.26083263754844666,
"step": 1445
},
{
"epoch": 1.4166156108320058,
"grad_norm": 0.3246079981327057,
"learning_rate": 2.080361519557548e-06,
"loss": 0.26272258162498474,
"step": 1446
},
{
"epoch": 1.4175958828574928,
"grad_norm": 0.3124680519104004,
"learning_rate": 2.073924887431744e-06,
"loss": 0.27763423323631287,
"step": 1447
},
{
"epoch": 1.41857615488298,
"grad_norm": 0.3879992365837097,
"learning_rate": 2.0674956216965484e-06,
"loss": 0.2246568500995636,
"step": 1448
},
{
"epoch": 1.419556426908467,
"grad_norm": 0.2952944338321686,
"learning_rate": 2.061073738537635e-06,
"loss": 0.2683466076850891,
"step": 1449
},
{
"epoch": 1.420536698933954,
"grad_norm": 0.250863254070282,
"learning_rate": 2.054659254122084e-06,
"loss": 0.3093163073062897,
"step": 1450
},
{
"epoch": 1.4215169709594413,
"grad_norm": 0.3000797629356384,
"learning_rate": 2.0482521845983522e-06,
"loss": 0.31663447618484497,
"step": 1451
},
{
"epoch": 1.4224972429849283,
"grad_norm": 0.2764907479286194,
"learning_rate": 2.0418525460962283e-06,
"loss": 0.23023763298988342,
"step": 1452
},
{
"epoch": 1.4234775150104153,
"grad_norm": 0.5278337001800537,
"learning_rate": 2.0354603547267985e-06,
"loss": 0.32687878608703613,
"step": 1453
},
{
"epoch": 1.4244577870359025,
"grad_norm": 0.3522464632987976,
"learning_rate": 2.029075626582393e-06,
"loss": 0.29966071248054504,
"step": 1454
},
{
"epoch": 1.4254380590613895,
"grad_norm": 0.33349722623825073,
"learning_rate": 2.0226983777365604e-06,
"loss": 0.3201775550842285,
"step": 1455
},
{
"epoch": 1.4264183310868765,
"grad_norm": 0.31306883692741394,
"learning_rate": 2.0163286242440155e-06,
"loss": 0.3786449432373047,
"step": 1456
},
{
"epoch": 1.4273986031123638,
"grad_norm": 0.2362661510705948,
"learning_rate": 2.009966382140606e-06,
"loss": 0.2506261467933655,
"step": 1457
},
{
"epoch": 1.4283788751378508,
"grad_norm": 0.280739426612854,
"learning_rate": 2.0036116674432653e-06,
"loss": 0.34754103422164917,
"step": 1458
},
{
"epoch": 1.4293591471633378,
"grad_norm": 0.3180376887321472,
"learning_rate": 1.9972644961499853e-06,
"loss": 0.28927677869796753,
"step": 1459
},
{
"epoch": 1.430339419188825,
"grad_norm": 0.3720912039279938,
"learning_rate": 1.990924884239758e-06,
"loss": 0.3737757205963135,
"step": 1460
},
{
"epoch": 1.431319691214312,
"grad_norm": 0.2702241539955139,
"learning_rate": 1.9845928476725522e-06,
"loss": 0.2926751375198364,
"step": 1461
},
{
"epoch": 1.432299963239799,
"grad_norm": 0.2325724959373474,
"learning_rate": 1.9782684023892613e-06,
"loss": 0.3101198673248291,
"step": 1462
},
{
"epoch": 1.4332802352652863,
"grad_norm": 0.2894973158836365,
"learning_rate": 1.971951564311668e-06,
"loss": 0.28658413887023926,
"step": 1463
},
{
"epoch": 1.4342605072907733,
"grad_norm": 0.25481942296028137,
"learning_rate": 1.9656423493424047e-06,
"loss": 0.2610940635204315,
"step": 1464
},
{
"epoch": 1.4352407793162603,
"grad_norm": 0.32062283158302307,
"learning_rate": 1.959340773364911e-06,
"loss": 0.3723536431789398,
"step": 1465
},
{
"epoch": 1.4362210513417473,
"grad_norm": 0.2787817716598511,
"learning_rate": 1.953046852243401e-06,
"loss": 0.256675124168396,
"step": 1466
},
{
"epoch": 1.4372013233672343,
"grad_norm": 0.3197075426578522,
"learning_rate": 1.946760601822809e-06,
"loss": 0.3052368760108948,
"step": 1467
},
{
"epoch": 1.4381815953927215,
"grad_norm": 0.8772352933883667,
"learning_rate": 1.9404820379287677e-06,
"loss": 0.2730110287666321,
"step": 1468
},
{
"epoch": 1.4391618674182085,
"grad_norm": 0.41473424434661865,
"learning_rate": 1.9342111763675512e-06,
"loss": 0.2705431580543518,
"step": 1469
},
{
"epoch": 1.4401421394436955,
"grad_norm": 0.2209106683731079,
"learning_rate": 1.927948032926047e-06,
"loss": 0.2822519540786743,
"step": 1470
},
{
"epoch": 1.4411224114691827,
"grad_norm": 0.3123934268951416,
"learning_rate": 1.9216926233717087e-06,
"loss": 0.33172059059143066,
"step": 1471
},
{
"epoch": 1.4421026834946697,
"grad_norm": 0.2977900505065918,
"learning_rate": 1.915444963452527e-06,
"loss": 0.29567188024520874,
"step": 1472
},
{
"epoch": 1.4430829555201568,
"grad_norm": 0.42021864652633667,
"learning_rate": 1.9092050688969736e-06,
"loss": 0.2762851119041443,
"step": 1473
},
{
"epoch": 1.444063227545644,
"grad_norm": 0.28360921144485474,
"learning_rate": 1.9029729554139798e-06,
"loss": 0.2697615325450897,
"step": 1474
},
{
"epoch": 1.445043499571131,
"grad_norm": 0.3754095435142517,
"learning_rate": 1.8967486386928819e-06,
"loss": 0.3771976828575134,
"step": 1475
},
{
"epoch": 1.446023771596618,
"grad_norm": 0.33185940980911255,
"learning_rate": 1.89053213440339e-06,
"loss": 0.3262893557548523,
"step": 1476
},
{
"epoch": 1.4470040436221052,
"grad_norm": 0.27861788868904114,
"learning_rate": 1.8843234581955444e-06,
"loss": 0.28204113245010376,
"step": 1477
},
{
"epoch": 1.4479843156475922,
"grad_norm": 0.42855748534202576,
"learning_rate": 1.8781226256996838e-06,
"loss": 0.29269254207611084,
"step": 1478
},
{
"epoch": 1.4489645876730792,
"grad_norm": 0.318498432636261,
"learning_rate": 1.8719296525263925e-06,
"loss": 0.29063719511032104,
"step": 1479
},
{
"epoch": 1.4499448596985665,
"grad_norm": 0.3814685046672821,
"learning_rate": 1.865744554266477e-06,
"loss": 0.307870090007782,
"step": 1480
},
{
"epoch": 1.4509251317240535,
"grad_norm": 0.337492972612381,
"learning_rate": 1.859567346490913e-06,
"loss": 0.34437859058380127,
"step": 1481
},
{
"epoch": 1.4519054037495405,
"grad_norm": 0.40815070271492004,
"learning_rate": 1.8533980447508138e-06,
"loss": 0.285390168428421,
"step": 1482
},
{
"epoch": 1.4528856757750277,
"grad_norm": 0.35439980030059814,
"learning_rate": 1.8472366645773892e-06,
"loss": 0.2774742841720581,
"step": 1483
},
{
"epoch": 1.4538659478005147,
"grad_norm": 0.29884693026542664,
"learning_rate": 1.8410832214819046e-06,
"loss": 0.2602722942829132,
"step": 1484
},
{
"epoch": 1.4548462198260017,
"grad_norm": 0.3463387191295624,
"learning_rate": 1.8349377309556487e-06,
"loss": 0.27696025371551514,
"step": 1485
},
{
"epoch": 1.4558264918514887,
"grad_norm": 0.29056376218795776,
"learning_rate": 1.8288002084698842e-06,
"loss": 0.278798907995224,
"step": 1486
},
{
"epoch": 1.456806763876976,
"grad_norm": 0.30600422620773315,
"learning_rate": 1.8226706694758194e-06,
"loss": 0.29414552450180054,
"step": 1487
},
{
"epoch": 1.457787035902463,
"grad_norm": 0.261929988861084,
"learning_rate": 1.8165491294045596e-06,
"loss": 0.3262881636619568,
"step": 1488
},
{
"epoch": 1.45876730792795,
"grad_norm": 0.8319677114486694,
"learning_rate": 1.810435603667075e-06,
"loss": 0.35338208079338074,
"step": 1489
},
{
"epoch": 1.459747579953437,
"grad_norm": 0.32295849919319153,
"learning_rate": 1.8043301076541575e-06,
"loss": 0.3578181266784668,
"step": 1490
},
{
"epoch": 1.4607278519789242,
"grad_norm": 0.3321084678173065,
"learning_rate": 1.798232656736389e-06,
"loss": 0.3587234914302826,
"step": 1491
},
{
"epoch": 1.4617081240044112,
"grad_norm": 0.26217910647392273,
"learning_rate": 1.792143266264092e-06,
"loss": 0.2982560694217682,
"step": 1492
},
{
"epoch": 1.4626883960298982,
"grad_norm": 0.6092540621757507,
"learning_rate": 1.7860619515673034e-06,
"loss": 0.30992361903190613,
"step": 1493
},
{
"epoch": 1.4636686680553854,
"grad_norm": 0.25786659121513367,
"learning_rate": 1.7799887279557238e-06,
"loss": 0.2732320725917816,
"step": 1494
},
{
"epoch": 1.4646489400808724,
"grad_norm": 0.2791391611099243,
"learning_rate": 1.7739236107186858e-06,
"loss": 0.3341851532459259,
"step": 1495
},
{
"epoch": 1.4656292121063594,
"grad_norm": 0.26698020100593567,
"learning_rate": 1.767866615125114e-06,
"loss": 0.24291715025901794,
"step": 1496
},
{
"epoch": 1.4666094841318467,
"grad_norm": 0.4083669185638428,
"learning_rate": 1.7618177564234907e-06,
"loss": 0.3401205539703369,
"step": 1497
},
{
"epoch": 1.4675897561573337,
"grad_norm": 0.2850293219089508,
"learning_rate": 1.7557770498418087e-06,
"loss": 0.3449852168560028,
"step": 1498
},
{
"epoch": 1.4685700281828207,
"grad_norm": 1.4493825435638428,
"learning_rate": 1.7497445105875377e-06,
"loss": 0.26377227902412415,
"step": 1499
},
{
"epoch": 1.469550300208308,
"grad_norm": 0.23657310009002686,
"learning_rate": 1.7437201538475916e-06,
"loss": 0.2714686989784241,
"step": 1500
},
{
"epoch": 1.470530572233795,
"grad_norm": 0.44274842739105225,
"learning_rate": 1.7377039947882802e-06,
"loss": 0.30843454599380493,
"step": 1501
},
{
"epoch": 1.471510844259282,
"grad_norm": 0.3299300968647003,
"learning_rate": 1.7316960485552758e-06,
"loss": 0.27967724204063416,
"step": 1502
},
{
"epoch": 1.4724911162847691,
"grad_norm": 0.35349297523498535,
"learning_rate": 1.7256963302735752e-06,
"loss": 0.29007840156555176,
"step": 1503
},
{
"epoch": 1.4734713883102561,
"grad_norm": 0.2221602499485016,
"learning_rate": 1.7197048550474643e-06,
"loss": 0.24931935966014862,
"step": 1504
},
{
"epoch": 1.4744516603357432,
"grad_norm": 0.3054089844226837,
"learning_rate": 1.7137216379604727e-06,
"loss": 0.2815810441970825,
"step": 1505
},
{
"epoch": 1.4754319323612302,
"grad_norm": 0.6968994736671448,
"learning_rate": 1.7077466940753446e-06,
"loss": 0.2873488962650299,
"step": 1506
},
{
"epoch": 1.4764122043867174,
"grad_norm": 0.2873939573764801,
"learning_rate": 1.7017800384339928e-06,
"loss": 0.3002283573150635,
"step": 1507
},
{
"epoch": 1.4773924764122044,
"grad_norm": 0.2543400824069977,
"learning_rate": 1.6958216860574661e-06,
"loss": 0.2518739402294159,
"step": 1508
},
{
"epoch": 1.4783727484376914,
"grad_norm": 0.3308892250061035,
"learning_rate": 1.6898716519459074e-06,
"loss": 0.3134765625,
"step": 1509
},
{
"epoch": 1.4793530204631784,
"grad_norm": 0.8625231981277466,
"learning_rate": 1.683929951078524e-06,
"loss": 0.312164306640625,
"step": 1510
},
{
"epoch": 1.4803332924886656,
"grad_norm": 0.29247626662254333,
"learning_rate": 1.6779965984135376e-06,
"loss": 0.35004833340644836,
"step": 1511
},
{
"epoch": 1.4813135645141526,
"grad_norm": 0.294977068901062,
"learning_rate": 1.6720716088881595e-06,
"loss": 0.32739442586898804,
"step": 1512
},
{
"epoch": 1.4822938365396396,
"grad_norm": 0.26107192039489746,
"learning_rate": 1.6661549974185426e-06,
"loss": 0.2478279322385788,
"step": 1513
},
{
"epoch": 1.4832741085651269,
"grad_norm": 0.3958752453327179,
"learning_rate": 1.6602467788997483e-06,
"loss": 0.2893311679363251,
"step": 1514
},
{
"epoch": 1.4842543805906139,
"grad_norm": 0.30146610736846924,
"learning_rate": 1.6543469682057105e-06,
"loss": 0.29788756370544434,
"step": 1515
},
{
"epoch": 1.4852346526161009,
"grad_norm": 0.31995415687561035,
"learning_rate": 1.648455580189194e-06,
"loss": 0.3546915352344513,
"step": 1516
},
{
"epoch": 1.486214924641588,
"grad_norm": 0.3058770000934601,
"learning_rate": 1.6425726296817634e-06,
"loss": 0.27943867444992065,
"step": 1517
},
{
"epoch": 1.4871951966670751,
"grad_norm": 0.24469350278377533,
"learning_rate": 1.6366981314937374e-06,
"loss": 0.2700238525867462,
"step": 1518
},
{
"epoch": 1.4881754686925621,
"grad_norm": 1.5886441469192505,
"learning_rate": 1.6308321004141609e-06,
"loss": 0.30660974979400635,
"step": 1519
},
{
"epoch": 1.4891557407180493,
"grad_norm": 0.2510388493537903,
"learning_rate": 1.624974551210759e-06,
"loss": 0.23549775779247284,
"step": 1520
},
{
"epoch": 1.4901360127435364,
"grad_norm": 0.24097967147827148,
"learning_rate": 1.6191254986299044e-06,
"loss": 0.2824183702468872,
"step": 1521
},
{
"epoch": 1.4911162847690234,
"grad_norm": 0.234966441988945,
"learning_rate": 1.6132849573965781e-06,
"loss": 0.2545176148414612,
"step": 1522
},
{
"epoch": 1.4920965567945106,
"grad_norm": 0.44066134095191956,
"learning_rate": 1.6074529422143398e-06,
"loss": 0.3417361080646515,
"step": 1523
},
{
"epoch": 1.4930768288199976,
"grad_norm": 0.2960100471973419,
"learning_rate": 1.6016294677652772e-06,
"loss": 0.361997127532959,
"step": 1524
},
{
"epoch": 1.4940571008454846,
"grad_norm": 0.3417584300041199,
"learning_rate": 1.5958145487099829e-06,
"loss": 0.3000933527946472,
"step": 1525
},
{
"epoch": 1.4950373728709718,
"grad_norm": 0.2602430582046509,
"learning_rate": 1.5900081996875083e-06,
"loss": 0.2688164710998535,
"step": 1526
},
{
"epoch": 1.4960176448964588,
"grad_norm": 0.39910516142845154,
"learning_rate": 1.5842104353153286e-06,
"loss": 0.3596796989440918,
"step": 1527
},
{
"epoch": 1.4969979169219458,
"grad_norm": 0.6877877712249756,
"learning_rate": 1.5784212701893088e-06,
"loss": 0.27708178758621216,
"step": 1528
},
{
"epoch": 1.4979781889474328,
"grad_norm": 0.44660648703575134,
"learning_rate": 1.5726407188836672e-06,
"loss": 0.28080856800079346,
"step": 1529
},
{
"epoch": 1.4989584609729198,
"grad_norm": 0.24623963236808777,
"learning_rate": 1.566868795950932e-06,
"loss": 0.26157110929489136,
"step": 1530
},
{
"epoch": 1.499938732998407,
"grad_norm": 0.2747780978679657,
"learning_rate": 1.561105515921915e-06,
"loss": 0.33199432492256165,
"step": 1531
},
{
"epoch": 1.500919005023894,
"grad_norm": 0.37532904744148254,
"learning_rate": 1.5553508933056665e-06,
"loss": 0.2839410901069641,
"step": 1532
},
{
"epoch": 1.501899277049381,
"grad_norm": 0.28604331612586975,
"learning_rate": 1.549604942589441e-06,
"loss": 0.30254825949668884,
"step": 1533
},
{
"epoch": 1.5028795490748683,
"grad_norm": 0.297220379114151,
"learning_rate": 1.543867678238663e-06,
"loss": 0.36205238103866577,
"step": 1534
},
{
"epoch": 1.5038598211003553,
"grad_norm": 0.347898930311203,
"learning_rate": 1.5381391146968866e-06,
"loss": 0.34542304277420044,
"step": 1535
},
{
"epoch": 1.5048400931258423,
"grad_norm": 0.28700384497642517,
"learning_rate": 1.5324192663857673e-06,
"loss": 0.30575066804885864,
"step": 1536
},
{
"epoch": 1.5058203651513296,
"grad_norm": 0.28909844160079956,
"learning_rate": 1.5267081477050132e-06,
"loss": 0.24873888492584229,
"step": 1537
},
{
"epoch": 1.5068006371768166,
"grad_norm": 0.33241599798202515,
"learning_rate": 1.521005773032362e-06,
"loss": 0.3850703239440918,
"step": 1538
},
{
"epoch": 1.5077809092023036,
"grad_norm": 0.3255611062049866,
"learning_rate": 1.5153121567235334e-06,
"loss": 0.3298748731613159,
"step": 1539
},
{
"epoch": 1.5087611812277908,
"grad_norm": 0.3054555058479309,
"learning_rate": 1.5096273131121997e-06,
"loss": 0.28001922369003296,
"step": 1540
},
{
"epoch": 1.5097414532532778,
"grad_norm": 0.35477396845817566,
"learning_rate": 1.5039512565099468e-06,
"loss": 0.27672362327575684,
"step": 1541
},
{
"epoch": 1.5107217252787648,
"grad_norm": 0.2514999806880951,
"learning_rate": 1.4982840012062428e-06,
"loss": 0.2729589343070984,
"step": 1542
},
{
"epoch": 1.511701997304252,
"grad_norm": 0.39038416743278503,
"learning_rate": 1.4926255614683931e-06,
"loss": 0.31467628479003906,
"step": 1543
},
{
"epoch": 1.512682269329739,
"grad_norm": 1.692759394645691,
"learning_rate": 1.4869759515415167e-06,
"loss": 0.33033663034439087,
"step": 1544
},
{
"epoch": 1.513662541355226,
"grad_norm": 0.2760690748691559,
"learning_rate": 1.4813351856484981e-06,
"loss": 0.27024146914482117,
"step": 1545
},
{
"epoch": 1.5146428133807133,
"grad_norm": 0.3051207363605499,
"learning_rate": 1.4757032779899588e-06,
"loss": 0.3421379625797272,
"step": 1546
},
{
"epoch": 1.5156230854062,
"grad_norm": 0.5136882066726685,
"learning_rate": 1.470080242744218e-06,
"loss": 0.27240028977394104,
"step": 1547
},
{
"epoch": 1.5166033574316873,
"grad_norm": 0.2925591468811035,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.26870036125183105,
"step": 1548
},
{
"epoch": 1.5175836294571745,
"grad_norm": 0.3444752097129822,
"learning_rate": 1.458860846092705e-06,
"loss": 0.30401307344436646,
"step": 1549
},
{
"epoch": 1.5185639014826613,
"grad_norm": 0.33959272503852844,
"learning_rate": 1.453264512931748e-06,
"loss": 0.307686448097229,
"step": 1550
},
{
"epoch": 1.5195441735081485,
"grad_norm": 0.30975255370140076,
"learning_rate": 1.4476771086731567e-06,
"loss": 0.3536240756511688,
"step": 1551
},
{
"epoch": 1.5205244455336357,
"grad_norm": 0.2965373396873474,
"learning_rate": 1.4420986473832127e-06,
"loss": 0.2855833172798157,
"step": 1552
},
{
"epoch": 1.5215047175591225,
"grad_norm": 0.31123054027557373,
"learning_rate": 1.4365291431056871e-06,
"loss": 0.28975018858909607,
"step": 1553
},
{
"epoch": 1.5224849895846098,
"grad_norm": 0.43425288796424866,
"learning_rate": 1.4309686098617975e-06,
"loss": 0.3898746371269226,
"step": 1554
},
{
"epoch": 1.5234652616100968,
"grad_norm": 0.28322139382362366,
"learning_rate": 1.4254170616501828e-06,
"loss": 0.32261791825294495,
"step": 1555
},
{
"epoch": 1.5244455336355838,
"grad_norm": 0.35701265931129456,
"learning_rate": 1.4198745124468565e-06,
"loss": 0.2516995370388031,
"step": 1556
},
{
"epoch": 1.525425805661071,
"grad_norm": 0.3035414218902588,
"learning_rate": 1.4143409762051829e-06,
"loss": 0.2742636799812317,
"step": 1557
},
{
"epoch": 1.526406077686558,
"grad_norm": 0.32761409878730774,
"learning_rate": 1.4088164668558302e-06,
"loss": 0.3257451057434082,
"step": 1558
},
{
"epoch": 1.527386349712045,
"grad_norm": 0.24971483647823334,
"learning_rate": 1.4033009983067454e-06,
"loss": 0.2686963379383087,
"step": 1559
},
{
"epoch": 1.5283666217375322,
"grad_norm": 0.28296390175819397,
"learning_rate": 1.3977945844431118e-06,
"loss": 0.24747627973556519,
"step": 1560
},
{
"epoch": 1.5293468937630192,
"grad_norm": 0.290717750787735,
"learning_rate": 1.3922972391273226e-06,
"loss": 0.2999711334705353,
"step": 1561
},
{
"epoch": 1.5303271657885062,
"grad_norm": 0.2598382532596588,
"learning_rate": 1.3868089761989356e-06,
"loss": 0.25741833448410034,
"step": 1562
},
{
"epoch": 1.5313074378139935,
"grad_norm": 0.23878192901611328,
"learning_rate": 1.3813298094746491e-06,
"loss": 0.2629435062408447,
"step": 1563
},
{
"epoch": 1.5322877098394805,
"grad_norm": 0.40887805819511414,
"learning_rate": 1.375859752748257e-06,
"loss": 0.4089251756668091,
"step": 1564
},
{
"epoch": 1.5332679818649675,
"grad_norm": 0.3845243453979492,
"learning_rate": 1.3703988197906209e-06,
"loss": 0.3369177579879761,
"step": 1565
},
{
"epoch": 1.5342482538904547,
"grad_norm": 0.2645271420478821,
"learning_rate": 1.3649470243496327e-06,
"loss": 0.3359029293060303,
"step": 1566
},
{
"epoch": 1.5352285259159417,
"grad_norm": 0.44226011633872986,
"learning_rate": 1.3595043801501794e-06,
"loss": 0.270320862531662,
"step": 1567
},
{
"epoch": 1.5362087979414287,
"grad_norm": 0.2985086739063263,
"learning_rate": 1.354070900894115e-06,
"loss": 0.2775104343891144,
"step": 1568
},
{
"epoch": 1.537189069966916,
"grad_norm": 0.26536524295806885,
"learning_rate": 1.3486466002602133e-06,
"loss": 0.2881200313568115,
"step": 1569
},
{
"epoch": 1.5381693419924027,
"grad_norm": 0.5512717366218567,
"learning_rate": 1.3432314919041478e-06,
"loss": 0.2949836254119873,
"step": 1570
},
{
"epoch": 1.53914961401789,
"grad_norm": 0.2983371317386627,
"learning_rate": 1.3378255894584463e-06,
"loss": 0.4297691285610199,
"step": 1571
},
{
"epoch": 1.5401298860433772,
"grad_norm": 0.26843464374542236,
"learning_rate": 1.3324289065324608e-06,
"loss": 0.24741193652153015,
"step": 1572
},
{
"epoch": 1.541110158068864,
"grad_norm": 0.8197523951530457,
"learning_rate": 1.3270414567123342e-06,
"loss": 0.35932695865631104,
"step": 1573
},
{
"epoch": 1.5420904300943512,
"grad_norm": 0.2774127721786499,
"learning_rate": 1.3216632535609664e-06,
"loss": 0.24276933073997498,
"step": 1574
},
{
"epoch": 1.5430707021198382,
"grad_norm": 0.4254603981971741,
"learning_rate": 1.3162943106179748e-06,
"loss": 0.3576342761516571,
"step": 1575
},
{
"epoch": 1.5440509741453252,
"grad_norm": 0.33967462182044983,
"learning_rate": 1.310934641399671e-06,
"loss": 0.22038841247558594,
"step": 1576
},
{
"epoch": 1.5450312461708124,
"grad_norm": 0.29422828555107117,
"learning_rate": 1.305584259399013e-06,
"loss": 0.2881976366043091,
"step": 1577
},
{
"epoch": 1.5460115181962995,
"grad_norm": 0.3111098110675812,
"learning_rate": 1.3002431780855817e-06,
"loss": 0.3391568660736084,
"step": 1578
},
{
"epoch": 1.5469917902217865,
"grad_norm": 0.38273900747299194,
"learning_rate": 1.2949114109055417e-06,
"loss": 0.32040834426879883,
"step": 1579
},
{
"epoch": 1.5479720622472737,
"grad_norm": 0.3417913615703583,
"learning_rate": 1.2895889712816129e-06,
"loss": 0.34064000844955444,
"step": 1580
},
{
"epoch": 1.5489523342727607,
"grad_norm": 0.2722271978855133,
"learning_rate": 1.2842758726130283e-06,
"loss": 0.2396983653306961,
"step": 1581
},
{
"epoch": 1.5499326062982477,
"grad_norm": 0.3091174364089966,
"learning_rate": 1.2789721282755102e-06,
"loss": 0.3650130331516266,
"step": 1582
},
{
"epoch": 1.550912878323735,
"grad_norm": 0.2543531060218811,
"learning_rate": 1.2736777516212267e-06,
"loss": 0.2810743451118469,
"step": 1583
},
{
"epoch": 1.551893150349222,
"grad_norm": 0.2622183859348297,
"learning_rate": 1.2683927559787657e-06,
"loss": 0.2828664779663086,
"step": 1584
},
{
"epoch": 1.552873422374709,
"grad_norm": 0.2914018929004669,
"learning_rate": 1.263117154653097e-06,
"loss": 0.2787971496582031,
"step": 1585
},
{
"epoch": 1.5538536944001962,
"grad_norm": 0.5201424956321716,
"learning_rate": 1.2578509609255385e-06,
"loss": 0.30938851833343506,
"step": 1586
},
{
"epoch": 1.5548339664256832,
"grad_norm": 0.2765558660030365,
"learning_rate": 1.2525941880537307e-06,
"loss": 0.2858021855354309,
"step": 1587
},
{
"epoch": 1.5558142384511702,
"grad_norm": 1.301770806312561,
"learning_rate": 1.2473468492715896e-06,
"loss": 0.3091282844543457,
"step": 1588
},
{
"epoch": 1.5567945104766574,
"grad_norm": 0.24136307835578918,
"learning_rate": 1.242108957789287e-06,
"loss": 0.27732300758361816,
"step": 1589
},
{
"epoch": 1.5577747825021442,
"grad_norm": 0.2888999879360199,
"learning_rate": 1.2368805267932072e-06,
"loss": 0.3073512315750122,
"step": 1590
},
{
"epoch": 1.5587550545276314,
"grad_norm": 0.2548321485519409,
"learning_rate": 1.2316615694459188e-06,
"loss": 0.28495168685913086,
"step": 1591
},
{
"epoch": 1.5597353265531186,
"grad_norm": 0.3482649624347687,
"learning_rate": 1.22645209888614e-06,
"loss": 0.3495527505874634,
"step": 1592
},
{
"epoch": 1.5607155985786054,
"grad_norm": 0.2891555428504944,
"learning_rate": 1.2212521282287093e-06,
"loss": 0.27066075801849365,
"step": 1593
},
{
"epoch": 1.5616958706040927,
"grad_norm": 0.41783496737480164,
"learning_rate": 1.2160616705645434e-06,
"loss": 0.30170291662216187,
"step": 1594
},
{
"epoch": 1.5626761426295797,
"grad_norm": 0.31107097864151,
"learning_rate": 1.210880738960616e-06,
"loss": 0.32693082094192505,
"step": 1595
},
{
"epoch": 1.5636564146550667,
"grad_norm": 0.3129148781299591,
"learning_rate": 1.2057093464599156e-06,
"loss": 0.2735086679458618,
"step": 1596
},
{
"epoch": 1.564636686680554,
"grad_norm": 0.3216259181499481,
"learning_rate": 1.200547506081416e-06,
"loss": 0.3297010064125061,
"step": 1597
},
{
"epoch": 1.565616958706041,
"grad_norm": 0.4404817223548889,
"learning_rate": 1.1953952308200428e-06,
"loss": 0.31461334228515625,
"step": 1598
},
{
"epoch": 1.566597230731528,
"grad_norm": 0.323138028383255,
"learning_rate": 1.1902525336466465e-06,
"loss": 0.27893611788749695,
"step": 1599
},
{
"epoch": 1.5675775027570151,
"grad_norm": 0.4638177752494812,
"learning_rate": 1.1851194275079586e-06,
"loss": 0.31542178988456726,
"step": 1600
},
{
"epoch": 1.5685577747825021,
"grad_norm": 0.31430378556251526,
"learning_rate": 1.1799959253265668e-06,
"loss": 0.30155977606773376,
"step": 1601
},
{
"epoch": 1.5695380468079891,
"grad_norm": 0.33770695328712463,
"learning_rate": 1.1748820400008843e-06,
"loss": 0.38152003288269043,
"step": 1602
},
{
"epoch": 1.5705183188334764,
"grad_norm": 0.34815165400505066,
"learning_rate": 1.1697777844051105e-06,
"loss": 0.29367896914482117,
"step": 1603
},
{
"epoch": 1.5714985908589634,
"grad_norm": 0.3172726035118103,
"learning_rate": 1.164683171389202e-06,
"loss": 0.3085186183452606,
"step": 1604
},
{
"epoch": 1.5724788628844504,
"grad_norm": 0.700988233089447,
"learning_rate": 1.1595982137788403e-06,
"loss": 0.35079318284988403,
"step": 1605
},
{
"epoch": 1.5734591349099376,
"grad_norm": 0.37854504585266113,
"learning_rate": 1.154522924375403e-06,
"loss": 0.3135107159614563,
"step": 1606
},
{
"epoch": 1.5744394069354246,
"grad_norm": 0.48977407813072205,
"learning_rate": 1.1494573159559214e-06,
"loss": 0.2882559299468994,
"step": 1607
},
{
"epoch": 1.5754196789609116,
"grad_norm": 0.2940884828567505,
"learning_rate": 1.144401401273062e-06,
"loss": 0.28473877906799316,
"step": 1608
},
{
"epoch": 1.5763999509863988,
"grad_norm": 0.2826167345046997,
"learning_rate": 1.1393551930550828e-06,
"loss": 0.31310757994651794,
"step": 1609
},
{
"epoch": 1.5773802230118859,
"grad_norm": 0.30077439546585083,
"learning_rate": 1.1343187040058068e-06,
"loss": 0.29365789890289307,
"step": 1610
},
{
"epoch": 1.5783604950373729,
"grad_norm": 0.29335930943489075,
"learning_rate": 1.1292919468045876e-06,
"loss": 0.28878748416900635,
"step": 1611
},
{
"epoch": 1.57934076706286,
"grad_norm": 0.37228575348854065,
"learning_rate": 1.1242749341062837e-06,
"loss": 0.33480143547058105,
"step": 1612
},
{
"epoch": 1.5803210390883469,
"grad_norm": 0.4832857847213745,
"learning_rate": 1.1192676785412154e-06,
"loss": 0.37072694301605225,
"step": 1613
},
{
"epoch": 1.581301311113834,
"grad_norm": 0.3139086067676544,
"learning_rate": 1.1142701927151456e-06,
"loss": 0.2953612804412842,
"step": 1614
},
{
"epoch": 1.5822815831393213,
"grad_norm": 0.4169934093952179,
"learning_rate": 1.1092824892092375e-06,
"loss": 0.36215150356292725,
"step": 1615
},
{
"epoch": 1.583261855164808,
"grad_norm": 0.26831310987472534,
"learning_rate": 1.1043045805800285e-06,
"loss": 0.23844678699970245,
"step": 1616
},
{
"epoch": 1.5842421271902953,
"grad_norm": 0.318039208650589,
"learning_rate": 1.099336479359398e-06,
"loss": 0.29555076360702515,
"step": 1617
},
{
"epoch": 1.5852223992157823,
"grad_norm": 0.2937129735946655,
"learning_rate": 1.0943781980545332e-06,
"loss": 0.28672829270362854,
"step": 1618
},
{
"epoch": 1.5862026712412693,
"grad_norm": 0.27550140023231506,
"learning_rate": 1.0894297491479044e-06,
"loss": 0.2887890040874481,
"step": 1619
},
{
"epoch": 1.5871829432667566,
"grad_norm": 0.27732500433921814,
"learning_rate": 1.0844911450972228e-06,
"loss": 0.31291019916534424,
"step": 1620
},
{
"epoch": 1.5881632152922436,
"grad_norm": 0.3701080083847046,
"learning_rate": 1.0795623983354214e-06,
"loss": 0.3425907790660858,
"step": 1621
},
{
"epoch": 1.5891434873177306,
"grad_norm": 0.337433785200119,
"learning_rate": 1.0746435212706125e-06,
"loss": 0.3323116898536682,
"step": 1622
},
{
"epoch": 1.5901237593432178,
"grad_norm": 0.2847992479801178,
"learning_rate": 1.0697345262860638e-06,
"loss": 0.3357844054698944,
"step": 1623
},
{
"epoch": 1.5911040313687048,
"grad_norm": 0.3248917758464813,
"learning_rate": 1.0648354257401622e-06,
"loss": 0.36096757650375366,
"step": 1624
},
{
"epoch": 1.5920843033941918,
"grad_norm": 0.2795041501522064,
"learning_rate": 1.0599462319663906e-06,
"loss": 0.31344178318977356,
"step": 1625
},
{
"epoch": 1.593064575419679,
"grad_norm": 0.29801473021507263,
"learning_rate": 1.0550669572732862e-06,
"loss": 0.311551570892334,
"step": 1626
},
{
"epoch": 1.594044847445166,
"grad_norm": 0.2638187110424042,
"learning_rate": 1.0501976139444191e-06,
"loss": 0.2750704288482666,
"step": 1627
},
{
"epoch": 1.595025119470653,
"grad_norm": 0.26177075505256653,
"learning_rate": 1.0453382142383545e-06,
"loss": 0.2428896427154541,
"step": 1628
},
{
"epoch": 1.5960053914961403,
"grad_norm": 0.6099364757537842,
"learning_rate": 1.0404887703886252e-06,
"loss": 0.29959535598754883,
"step": 1629
},
{
"epoch": 1.5969856635216273,
"grad_norm": 0.26627424359321594,
"learning_rate": 1.0356492946036994e-06,
"loss": 0.2963497042655945,
"step": 1630
},
{
"epoch": 1.5979659355471143,
"grad_norm": 0.3082166612148285,
"learning_rate": 1.0308197990669538e-06,
"loss": 0.28662848472595215,
"step": 1631
},
{
"epoch": 1.5989462075726015,
"grad_norm": 0.3146071434020996,
"learning_rate": 1.026000295936635e-06,
"loss": 0.3462318181991577,
"step": 1632
},
{
"epoch": 1.5999264795980883,
"grad_norm": 0.260776549577713,
"learning_rate": 1.0211907973458391e-06,
"loss": 0.346241295337677,
"step": 1633
},
{
"epoch": 1.6009067516235755,
"grad_norm": 0.3167802691459656,
"learning_rate": 1.0163913154024724e-06,
"loss": 0.284930557012558,
"step": 1634
},
{
"epoch": 1.6018870236490628,
"grad_norm": 0.2441052347421646,
"learning_rate": 1.0116018621892237e-06,
"loss": 0.28023549914360046,
"step": 1635
},
{
"epoch": 1.6028672956745496,
"grad_norm": 0.3004295229911804,
"learning_rate": 1.006822449763537e-06,
"loss": 0.25645917654037476,
"step": 1636
},
{
"epoch": 1.6038475677000368,
"grad_norm": 0.27091652154922485,
"learning_rate": 1.0020530901575754e-06,
"loss": 0.2668631970882416,
"step": 1637
},
{
"epoch": 1.6048278397255238,
"grad_norm": 0.3432711362838745,
"learning_rate": 9.972937953781985e-07,
"loss": 0.33843135833740234,
"step": 1638
},
{
"epoch": 1.6058081117510108,
"grad_norm": 0.31863635778427124,
"learning_rate": 9.925445774069232e-07,
"loss": 0.3187398910522461,
"step": 1639
},
{
"epoch": 1.606788383776498,
"grad_norm": 0.35255932807922363,
"learning_rate": 9.878054481999027e-07,
"loss": 0.299978107213974,
"step": 1640
},
{
"epoch": 1.607768655801985,
"grad_norm": 0.7566986083984375,
"learning_rate": 9.830764196878872e-07,
"loss": 0.3120487928390503,
"step": 1641
},
{
"epoch": 1.608748927827472,
"grad_norm": 0.26897549629211426,
"learning_rate": 9.783575037762006e-07,
"loss": 0.25863292813301086,
"step": 1642
},
{
"epoch": 1.6097291998529593,
"grad_norm": 0.26998525857925415,
"learning_rate": 9.73648712344707e-07,
"loss": 0.265112042427063,
"step": 1643
},
{
"epoch": 1.6107094718784463,
"grad_norm": 0.32196807861328125,
"learning_rate": 9.689500572477857e-07,
"loss": 0.2917467951774597,
"step": 1644
},
{
"epoch": 1.6116897439039333,
"grad_norm": 0.27740079164505005,
"learning_rate": 9.642615503142927e-07,
"loss": 0.3283982276916504,
"step": 1645
},
{
"epoch": 1.6126700159294205,
"grad_norm": 0.29914382100105286,
"learning_rate": 9.595832033475417e-07,
"loss": 0.2946596145629883,
"step": 1646
},
{
"epoch": 1.6136502879549075,
"grad_norm": 0.36208799481391907,
"learning_rate": 9.549150281252633e-07,
"loss": 0.32836201786994934,
"step": 1647
},
{
"epoch": 1.6146305599803945,
"grad_norm": 0.32013386487960815,
"learning_rate": 9.502570363995839e-07,
"loss": 0.31348133087158203,
"step": 1648
},
{
"epoch": 1.6156108320058817,
"grad_norm": 0.2855890393257141,
"learning_rate": 9.456092398969902e-07,
"loss": 0.29105398058891296,
"step": 1649
},
{
"epoch": 1.6165911040313687,
"grad_norm": 0.2615555226802826,
"learning_rate": 9.409716503183075e-07,
"loss": 0.31310999393463135,
"step": 1650
},
{
"epoch": 1.6175713760568557,
"grad_norm": 0.37736764550209045,
"learning_rate": 9.363442793386606e-07,
"loss": 0.36657190322875977,
"step": 1651
},
{
"epoch": 1.618551648082343,
"grad_norm": 0.3742450177669525,
"learning_rate": 9.317271386074489e-07,
"loss": 0.3179669678211212,
"step": 1652
},
{
"epoch": 1.6195319201078298,
"grad_norm": 0.2608967125415802,
"learning_rate": 9.271202397483214e-07,
"loss": 0.2794756293296814,
"step": 1653
},
{
"epoch": 1.620512192133317,
"grad_norm": 0.26353180408477783,
"learning_rate": 9.225235943591398e-07,
"loss": 0.29618656635284424,
"step": 1654
},
{
"epoch": 1.6214924641588042,
"grad_norm": 0.33907854557037354,
"learning_rate": 9.179372140119524e-07,
"loss": 0.29158341884613037,
"step": 1655
},
{
"epoch": 1.622472736184291,
"grad_norm": 0.4554346799850464,
"learning_rate": 9.133611102529655e-07,
"loss": 0.28366580605506897,
"step": 1656
},
{
"epoch": 1.6234530082097782,
"grad_norm": 0.33707955479621887,
"learning_rate": 9.087952946025175e-07,
"loss": 0.25429874658584595,
"step": 1657
},
{
"epoch": 1.6244332802352652,
"grad_norm": 0.29552385210990906,
"learning_rate": 9.042397785550405e-07,
"loss": 0.29397982358932495,
"step": 1658
},
{
"epoch": 1.6254135522607522,
"grad_norm": 0.3059321641921997,
"learning_rate": 8.996945735790447e-07,
"loss": 0.34833550453186035,
"step": 1659
},
{
"epoch": 1.6263938242862395,
"grad_norm": 0.27479004859924316,
"learning_rate": 8.951596911170756e-07,
"loss": 0.32173416018486023,
"step": 1660
},
{
"epoch": 1.6273740963117265,
"grad_norm": 0.26350149512290955,
"learning_rate": 8.906351425856952e-07,
"loss": 0.2980935573577881,
"step": 1661
},
{
"epoch": 1.6283543683372135,
"grad_norm": 0.3240721821784973,
"learning_rate": 8.861209393754478e-07,
"loss": 0.30279961228370667,
"step": 1662
},
{
"epoch": 1.6293346403627007,
"grad_norm": 0.367153137922287,
"learning_rate": 8.816170928508367e-07,
"loss": 0.32025113701820374,
"step": 1663
},
{
"epoch": 1.6303149123881877,
"grad_norm": 0.4977502226829529,
"learning_rate": 8.771236143502876e-07,
"loss": 0.3054382801055908,
"step": 1664
},
{
"epoch": 1.6312951844136747,
"grad_norm": 0.3074219524860382,
"learning_rate": 8.7264051518613e-07,
"loss": 0.3213973641395569,
"step": 1665
},
{
"epoch": 1.632275456439162,
"grad_norm": 0.3323941230773926,
"learning_rate": 8.681678066445598e-07,
"loss": 0.33570054173469543,
"step": 1666
},
{
"epoch": 1.633255728464649,
"grad_norm": 0.36144256591796875,
"learning_rate": 8.637054999856148e-07,
"loss": 0.28618285059928894,
"step": 1667
},
{
"epoch": 1.634236000490136,
"grad_norm": 0.34499239921569824,
"learning_rate": 8.592536064431467e-07,
"loss": 0.3219583034515381,
"step": 1668
},
{
"epoch": 1.6352162725156232,
"grad_norm": 5.040250301361084,
"learning_rate": 8.54812137224792e-07,
"loss": 0.28886258602142334,
"step": 1669
},
{
"epoch": 1.6361965445411102,
"grad_norm": 0.2714666426181793,
"learning_rate": 8.503811035119452e-07,
"loss": 0.2672621011734009,
"step": 1670
},
{
"epoch": 1.6371768165665972,
"grad_norm": 0.3235977590084076,
"learning_rate": 8.459605164597268e-07,
"loss": 0.34967297315597534,
"step": 1671
},
{
"epoch": 1.6381570885920844,
"grad_norm": 0.32485300302505493,
"learning_rate": 8.415503871969616e-07,
"loss": 0.24465565383434296,
"step": 1672
},
{
"epoch": 1.6391373606175714,
"grad_norm": 0.2748887836933136,
"learning_rate": 8.371507268261436e-07,
"loss": 0.2840154767036438,
"step": 1673
},
{
"epoch": 1.6401176326430584,
"grad_norm": 0.32614773511886597,
"learning_rate": 8.327615464234129e-07,
"loss": 0.3462795913219452,
"step": 1674
},
{
"epoch": 1.6410979046685457,
"grad_norm": 0.257012277841568,
"learning_rate": 8.283828570385239e-07,
"loss": 0.29747673869132996,
"step": 1675
},
{
"epoch": 1.6420781766940324,
"grad_norm": 0.3337796926498413,
"learning_rate": 8.240146696948254e-07,
"loss": 0.35426998138427734,
"step": 1676
},
{
"epoch": 1.6430584487195197,
"grad_norm": 0.3709498643875122,
"learning_rate": 8.196569953892202e-07,
"loss": 0.32646411657333374,
"step": 1677
},
{
"epoch": 1.644038720745007,
"grad_norm": 0.2858070135116577,
"learning_rate": 8.153098450921514e-07,
"loss": 0.368857741355896,
"step": 1678
},
{
"epoch": 1.6450189927704937,
"grad_norm": 0.33654725551605225,
"learning_rate": 8.109732297475637e-07,
"loss": 0.29511749744415283,
"step": 1679
},
{
"epoch": 1.645999264795981,
"grad_norm": 0.3183237612247467,
"learning_rate": 8.066471602728804e-07,
"loss": 0.2874103784561157,
"step": 1680
},
{
"epoch": 1.646979536821468,
"grad_norm": 0.2795041501522064,
"learning_rate": 8.023316475589754e-07,
"loss": 0.26365625858306885,
"step": 1681
},
{
"epoch": 1.647959808846955,
"grad_norm": 0.3034282624721527,
"learning_rate": 7.980267024701494e-07,
"loss": 0.34164029359817505,
"step": 1682
},
{
"epoch": 1.6489400808724421,
"grad_norm": 0.40935438871383667,
"learning_rate": 7.937323358440935e-07,
"loss": 0.3241567015647888,
"step": 1683
},
{
"epoch": 1.6499203528979292,
"grad_norm": 0.3500325679779053,
"learning_rate": 7.894485584918737e-07,
"loss": 0.3543739318847656,
"step": 1684
},
{
"epoch": 1.6509006249234162,
"grad_norm": 0.2601185739040375,
"learning_rate": 7.851753811978924e-07,
"loss": 0.2537991404533386,
"step": 1685
},
{
"epoch": 1.6518808969489034,
"grad_norm": 0.2820446789264679,
"learning_rate": 7.809128147198692e-07,
"loss": 0.24991591274738312,
"step": 1686
},
{
"epoch": 1.6528611689743904,
"grad_norm": 0.3387594223022461,
"learning_rate": 7.766608697888095e-07,
"loss": 0.3335416913032532,
"step": 1687
},
{
"epoch": 1.6538414409998774,
"grad_norm": 0.3065255284309387,
"learning_rate": 7.724195571089787e-07,
"loss": 0.36776840686798096,
"step": 1688
},
{
"epoch": 1.6548217130253646,
"grad_norm": 0.31597450375556946,
"learning_rate": 7.681888873578786e-07,
"loss": 0.35172754526138306,
"step": 1689
},
{
"epoch": 1.6558019850508516,
"grad_norm": 0.33000609278678894,
"learning_rate": 7.639688711862131e-07,
"loss": 0.31090372800827026,
"step": 1690
},
{
"epoch": 1.6567822570763386,
"grad_norm": 0.4180420935153961,
"learning_rate": 7.597595192178702e-07,
"loss": 0.3139837682247162,
"step": 1691
},
{
"epoch": 1.6577625291018259,
"grad_norm": 0.2650761008262634,
"learning_rate": 7.555608420498872e-07,
"loss": 0.2890176773071289,
"step": 1692
},
{
"epoch": 1.6587428011273129,
"grad_norm": 0.3677182197570801,
"learning_rate": 7.513728502524286e-07,
"loss": 0.2927079200744629,
"step": 1693
},
{
"epoch": 1.6597230731527999,
"grad_norm": 0.2937071919441223,
"learning_rate": 7.471955543687586e-07,
"loss": 0.29945477843284607,
"step": 1694
},
{
"epoch": 1.660703345178287,
"grad_norm": 0.30642956495285034,
"learning_rate": 7.430289649152156e-07,
"loss": 0.2288358509540558,
"step": 1695
},
{
"epoch": 1.6616836172037739,
"grad_norm": 0.2731684148311615,
"learning_rate": 7.388730923811827e-07,
"loss": 0.29736006259918213,
"step": 1696
},
{
"epoch": 1.6626638892292611,
"grad_norm": 0.2501773536205292,
"learning_rate": 7.347279472290647e-07,
"loss": 0.2976704239845276,
"step": 1697
},
{
"epoch": 1.6636441612547483,
"grad_norm": 0.27353107929229736,
"learning_rate": 7.305935398942598e-07,
"loss": 0.3099774122238159,
"step": 1698
},
{
"epoch": 1.6646244332802351,
"grad_norm": 0.42144420742988586,
"learning_rate": 7.264698807851328e-07,
"loss": 0.35031384229660034,
"step": 1699
},
{
"epoch": 1.6656047053057224,
"grad_norm": 0.5568183660507202,
"learning_rate": 7.2235698028299e-07,
"loss": 0.28940969705581665,
"step": 1700
},
{
"epoch": 1.6665849773312094,
"grad_norm": 0.30815014243125916,
"learning_rate": 7.182548487420555e-07,
"loss": 0.27524593472480774,
"step": 1701
},
{
"epoch": 1.6675652493566964,
"grad_norm": 0.33465591073036194,
"learning_rate": 7.141634964894389e-07,
"loss": 0.24376648664474487,
"step": 1702
},
{
"epoch": 1.6685455213821836,
"grad_norm": 0.2837412655353546,
"learning_rate": 7.100829338251147e-07,
"loss": 0.3499067425727844,
"step": 1703
},
{
"epoch": 1.6695257934076706,
"grad_norm": 0.31930503249168396,
"learning_rate": 7.060131710218959e-07,
"loss": 0.35859760642051697,
"step": 1704
},
{
"epoch": 1.6705060654331576,
"grad_norm": 0.27367374300956726,
"learning_rate": 7.019542183254047e-07,
"loss": 0.25798752903938293,
"step": 1705
},
{
"epoch": 1.6714863374586448,
"grad_norm": 0.24604664742946625,
"learning_rate": 6.979060859540493e-07,
"loss": 0.2215338498353958,
"step": 1706
},
{
"epoch": 1.6724666094841318,
"grad_norm": 0.43441060185432434,
"learning_rate": 6.938687840989972e-07,
"loss": 0.2625845670700073,
"step": 1707
},
{
"epoch": 1.6734468815096188,
"grad_norm": 0.2783740162849426,
"learning_rate": 6.898423229241535e-07,
"loss": 0.3101246953010559,
"step": 1708
},
{
"epoch": 1.674427153535106,
"grad_norm": 0.399839848279953,
"learning_rate": 6.858267125661272e-07,
"loss": 0.32933682203292847,
"step": 1709
},
{
"epoch": 1.675407425560593,
"grad_norm": 0.279772013425827,
"learning_rate": 6.818219631342149e-07,
"loss": 0.25054386258125305,
"step": 1710
},
{
"epoch": 1.67638769758608,
"grad_norm": 0.2902394235134125,
"learning_rate": 6.778280847103668e-07,
"loss": 0.2735682427883148,
"step": 1711
},
{
"epoch": 1.6773679696115673,
"grad_norm": 0.2759038209915161,
"learning_rate": 6.738450873491675e-07,
"loss": 0.2931666076183319,
"step": 1712
},
{
"epoch": 1.6783482416370543,
"grad_norm": 0.2626618444919586,
"learning_rate": 6.698729810778065e-07,
"loss": 0.21225613355636597,
"step": 1713
},
{
"epoch": 1.6793285136625413,
"grad_norm": 0.731498658657074,
"learning_rate": 6.659117758960598e-07,
"loss": 0.3383631706237793,
"step": 1714
},
{
"epoch": 1.6803087856880285,
"grad_norm": 0.3369990289211273,
"learning_rate": 6.619614817762537e-07,
"loss": 0.27933913469314575,
"step": 1715
},
{
"epoch": 1.6812890577135156,
"grad_norm": 0.3137538433074951,
"learning_rate": 6.580221086632516e-07,
"loss": 0.2801979184150696,
"step": 1716
},
{
"epoch": 1.6822693297390026,
"grad_norm": 0.3229953348636627,
"learning_rate": 6.540936664744197e-07,
"loss": 0.376849889755249,
"step": 1717
},
{
"epoch": 1.6832496017644898,
"grad_norm": 0.2665906250476837,
"learning_rate": 6.501761650996052e-07,
"loss": 0.3020632565021515,
"step": 1718
},
{
"epoch": 1.6842298737899766,
"grad_norm": 0.4033171832561493,
"learning_rate": 6.462696144011149e-07,
"loss": 0.2799711227416992,
"step": 1719
},
{
"epoch": 1.6852101458154638,
"grad_norm": 0.2735978066921234,
"learning_rate": 6.423740242136833e-07,
"loss": 0.2823280394077301,
"step": 1720
},
{
"epoch": 1.686190417840951,
"grad_norm": 0.23379062116146088,
"learning_rate": 6.384894043444568e-07,
"loss": 0.23845013976097107,
"step": 1721
},
{
"epoch": 1.6871706898664378,
"grad_norm": 0.3043540418148041,
"learning_rate": 6.34615764572959e-07,
"loss": 0.3117551803588867,
"step": 1722
},
{
"epoch": 1.688150961891925,
"grad_norm": 0.28605177998542786,
"learning_rate": 6.307531146510754e-07,
"loss": 0.3261862099170685,
"step": 1723
},
{
"epoch": 1.689131233917412,
"grad_norm": 0.3018704354763031,
"learning_rate": 6.269014643030214e-07,
"loss": 0.30671772360801697,
"step": 1724
},
{
"epoch": 1.690111505942899,
"grad_norm": 0.2939828336238861,
"learning_rate": 6.230608232253227e-07,
"loss": 0.3112119436264038,
"step": 1725
},
{
"epoch": 1.6910917779683863,
"grad_norm": 0.33668121695518494,
"learning_rate": 6.192312010867868e-07,
"loss": 0.2780793309211731,
"step": 1726
},
{
"epoch": 1.6920720499938733,
"grad_norm": 0.31729912757873535,
"learning_rate": 6.154126075284855e-07,
"loss": 0.3087517023086548,
"step": 1727
},
{
"epoch": 1.6930523220193603,
"grad_norm": 0.44336259365081787,
"learning_rate": 6.116050521637218e-07,
"loss": 0.36683234572410583,
"step": 1728
},
{
"epoch": 1.6940325940448475,
"grad_norm": 0.43094173073768616,
"learning_rate": 6.07808544578013e-07,
"loss": 0.4434608817100525,
"step": 1729
},
{
"epoch": 1.6950128660703345,
"grad_norm": 0.2559085190296173,
"learning_rate": 6.04023094329062e-07,
"loss": 0.27192267775535583,
"step": 1730
},
{
"epoch": 1.6959931380958215,
"grad_norm": 0.2534147799015045,
"learning_rate": 6.002487109467347e-07,
"loss": 0.2602151036262512,
"step": 1731
},
{
"epoch": 1.6969734101213088,
"grad_norm": 0.2544151544570923,
"learning_rate": 5.964854039330365e-07,
"loss": 0.24186502397060394,
"step": 1732
},
{
"epoch": 1.6979536821467958,
"grad_norm": 0.5423831343650818,
"learning_rate": 5.927331827620902e-07,
"loss": 0.3565855026245117,
"step": 1733
},
{
"epoch": 1.6989339541722828,
"grad_norm": 0.36293214559555054,
"learning_rate": 5.889920568801055e-07,
"loss": 0.34965237975120544,
"step": 1734
},
{
"epoch": 1.69991422619777,
"grad_norm": 0.422600656747818,
"learning_rate": 5.852620357053651e-07,
"loss": 0.3058886229991913,
"step": 1735
},
{
"epoch": 1.700894498223257,
"grad_norm": 0.2973859906196594,
"learning_rate": 5.815431286281925e-07,
"loss": 0.342462956905365,
"step": 1736
},
{
"epoch": 1.701874770248744,
"grad_norm": 0.24987950921058655,
"learning_rate": 5.778353450109286e-07,
"loss": 0.2796635925769806,
"step": 1737
},
{
"epoch": 1.7028550422742312,
"grad_norm": 0.29093554615974426,
"learning_rate": 5.741386941879179e-07,
"loss": 0.29703834652900696,
"step": 1738
},
{
"epoch": 1.703835314299718,
"grad_norm": 0.284636914730072,
"learning_rate": 5.704531854654721e-07,
"loss": 0.2420828938484192,
"step": 1739
},
{
"epoch": 1.7048155863252052,
"grad_norm": 0.27879565954208374,
"learning_rate": 5.667788281218567e-07,
"loss": 0.3005129396915436,
"step": 1740
},
{
"epoch": 1.7057958583506925,
"grad_norm": 0.33010444045066833,
"learning_rate": 5.631156314072605e-07,
"loss": 0.33327576518058777,
"step": 1741
},
{
"epoch": 1.7067761303761793,
"grad_norm": 0.4578080475330353,
"learning_rate": 5.594636045437796e-07,
"loss": 0.26480650901794434,
"step": 1742
},
{
"epoch": 1.7077564024016665,
"grad_norm": 0.3928011357784271,
"learning_rate": 5.558227567253832e-07,
"loss": 0.28785082697868347,
"step": 1743
},
{
"epoch": 1.7087366744271535,
"grad_norm": 0.31983065605163574,
"learning_rate": 5.521930971179041e-07,
"loss": 0.33132898807525635,
"step": 1744
},
{
"epoch": 1.7097169464526405,
"grad_norm": 0.46544909477233887,
"learning_rate": 5.485746348590048e-07,
"loss": 0.3230981230735779,
"step": 1745
},
{
"epoch": 1.7106972184781277,
"grad_norm": 0.7985524535179138,
"learning_rate": 5.449673790581611e-07,
"loss": 0.2958715558052063,
"step": 1746
},
{
"epoch": 1.7116774905036147,
"grad_norm": 0.4226300120353699,
"learning_rate": 5.413713387966329e-07,
"loss": 0.3325260281562805,
"step": 1747
},
{
"epoch": 1.7126577625291017,
"grad_norm": 0.28626784682273865,
"learning_rate": 5.377865231274504e-07,
"loss": 0.2998458445072174,
"step": 1748
},
{
"epoch": 1.713638034554589,
"grad_norm": 0.3840333819389343,
"learning_rate": 5.34212941075381e-07,
"loss": 0.3688451051712036,
"step": 1749
},
{
"epoch": 1.714618306580076,
"grad_norm": 0.2735205888748169,
"learning_rate": 5.306506016369134e-07,
"loss": 0.3664025068283081,
"step": 1750
},
{
"epoch": 1.715598578605563,
"grad_norm": 0.3691958785057068,
"learning_rate": 5.270995137802315e-07,
"loss": 0.3300861120223999,
"step": 1751
},
{
"epoch": 1.7165788506310502,
"grad_norm": 0.24028199911117554,
"learning_rate": 5.235596864451975e-07,
"loss": 0.22591480612754822,
"step": 1752
},
{
"epoch": 1.7175591226565372,
"grad_norm": 0.4072524905204773,
"learning_rate": 5.200311285433213e-07,
"loss": 0.34179577231407166,
"step": 1753
},
{
"epoch": 1.7185393946820242,
"grad_norm": 0.3562641143798828,
"learning_rate": 5.165138489577415e-07,
"loss": 0.3574969470500946,
"step": 1754
},
{
"epoch": 1.7195196667075114,
"grad_norm": 0.2758108675479889,
"learning_rate": 5.130078565432089e-07,
"loss": 0.2715908885002136,
"step": 1755
},
{
"epoch": 1.7204999387329984,
"grad_norm": 0.41939201951026917,
"learning_rate": 5.095131601260517e-07,
"loss": 0.3063475489616394,
"step": 1756
},
{
"epoch": 1.7214802107584855,
"grad_norm": 0.3254632353782654,
"learning_rate": 5.06029768504166e-07,
"loss": 0.2959940731525421,
"step": 1757
},
{
"epoch": 1.7224604827839727,
"grad_norm": 0.2487296760082245,
"learning_rate": 5.025576904469842e-07,
"loss": 0.26920783519744873,
"step": 1758
},
{
"epoch": 1.7234407548094595,
"grad_norm": 0.3512011468410492,
"learning_rate": 4.990969346954611e-07,
"loss": 0.2627406120300293,
"step": 1759
},
{
"epoch": 1.7244210268349467,
"grad_norm": 0.52886563539505,
"learning_rate": 4.956475099620433e-07,
"loss": 0.2507293224334717,
"step": 1760
},
{
"epoch": 1.725401298860434,
"grad_norm": 1.0871943235397339,
"learning_rate": 4.922094249306559e-07,
"loss": 0.29652100801467896,
"step": 1761
},
{
"epoch": 1.7263815708859207,
"grad_norm": 0.26580196619033813,
"learning_rate": 4.887826882566699e-07,
"loss": 0.2896541655063629,
"step": 1762
},
{
"epoch": 1.727361842911408,
"grad_norm": 0.4702354073524475,
"learning_rate": 4.853673085668947e-07,
"loss": 0.2931060194969177,
"step": 1763
},
{
"epoch": 1.728342114936895,
"grad_norm": 0.3238118588924408,
"learning_rate": 4.819632944595415e-07,
"loss": 0.33639052510261536,
"step": 1764
},
{
"epoch": 1.729322386962382,
"grad_norm": 0.40850332379341125,
"learning_rate": 4.785706545042141e-07,
"loss": 0.2757587432861328,
"step": 1765
},
{
"epoch": 1.7303026589878692,
"grad_norm": 0.5026625990867615,
"learning_rate": 4.751893972418781e-07,
"loss": 0.30373257398605347,
"step": 1766
},
{
"epoch": 1.7312829310133562,
"grad_norm": 0.36209017038345337,
"learning_rate": 4.7181953118484556e-07,
"loss": 0.3439788222312927,
"step": 1767
},
{
"epoch": 1.7322632030388432,
"grad_norm": 0.257283478975296,
"learning_rate": 4.6846106481675035e-07,
"loss": 0.257625937461853,
"step": 1768
},
{
"epoch": 1.7332434750643304,
"grad_norm": 0.29908475279808044,
"learning_rate": 4.651140065925269e-07,
"loss": 0.2816242575645447,
"step": 1769
},
{
"epoch": 1.7342237470898174,
"grad_norm": 0.3468051850795746,
"learning_rate": 4.617783649383906e-07,
"loss": 0.2884095311164856,
"step": 1770
},
{
"epoch": 1.7352040191153044,
"grad_norm": 0.26915886998176575,
"learning_rate": 4.58454148251814e-07,
"loss": 0.26531949639320374,
"step": 1771
},
{
"epoch": 1.7361842911407916,
"grad_norm": 0.3561387360095978,
"learning_rate": 4.551413649015113e-07,
"loss": 0.2574940323829651,
"step": 1772
},
{
"epoch": 1.7371645631662787,
"grad_norm": 0.6052283048629761,
"learning_rate": 4.5184002322740784e-07,
"loss": 0.30162250995635986,
"step": 1773
},
{
"epoch": 1.7381448351917657,
"grad_norm": 0.309078574180603,
"learning_rate": 4.48550131540631e-07,
"loss": 0.3131977319717407,
"step": 1774
},
{
"epoch": 1.7391251072172529,
"grad_norm": 0.31269189715385437,
"learning_rate": 4.4527169812347446e-07,
"loss": 0.30090945959091187,
"step": 1775
},
{
"epoch": 1.74010537924274,
"grad_norm": 0.280923992395401,
"learning_rate": 4.420047312293946e-07,
"loss": 0.3295038938522339,
"step": 1776
},
{
"epoch": 1.741085651268227,
"grad_norm": 0.26941558718681335,
"learning_rate": 4.387492390829734e-07,
"loss": 0.30732887983322144,
"step": 1777
},
{
"epoch": 1.7420659232937141,
"grad_norm": 0.29161518812179565,
"learning_rate": 4.355052298799112e-07,
"loss": 0.29111772775650024,
"step": 1778
},
{
"epoch": 1.7430461953192011,
"grad_norm": 0.9925495982170105,
"learning_rate": 4.322727117869951e-07,
"loss": 0.2879437208175659,
"step": 1779
},
{
"epoch": 1.7440264673446881,
"grad_norm": 0.5777062177658081,
"learning_rate": 4.2905169294208846e-07,
"loss": 0.28510749340057373,
"step": 1780
},
{
"epoch": 1.7450067393701754,
"grad_norm": 0.2889922559261322,
"learning_rate": 4.2584218145409916e-07,
"loss": 0.2741391360759735,
"step": 1781
},
{
"epoch": 1.7459870113956621,
"grad_norm": 0.31145453453063965,
"learning_rate": 4.2264418540297e-07,
"loss": 0.3046385645866394,
"step": 1782
},
{
"epoch": 1.7469672834211494,
"grad_norm": 0.28848639130592346,
"learning_rate": 4.194577128396521e-07,
"loss": 0.26749563217163086,
"step": 1783
},
{
"epoch": 1.7479475554466366,
"grad_norm": 0.3972063958644867,
"learning_rate": 4.1628277178608677e-07,
"loss": 0.4035048484802246,
"step": 1784
},
{
"epoch": 1.7489278274721234,
"grad_norm": 0.3292684555053711,
"learning_rate": 4.131193702351827e-07,
"loss": 0.3437063992023468,
"step": 1785
},
{
"epoch": 1.7499080994976106,
"grad_norm": 0.38713040947914124,
"learning_rate": 4.099675161508021e-07,
"loss": 0.2893226146697998,
"step": 1786
},
{
"epoch": 1.7508883715230976,
"grad_norm": 0.5345596075057983,
"learning_rate": 4.0682721746773346e-07,
"loss": 0.3005547523498535,
"step": 1787
},
{
"epoch": 1.7518686435485846,
"grad_norm": 0.33142930269241333,
"learning_rate": 4.036984820916723e-07,
"loss": 0.32711589336395264,
"step": 1788
},
{
"epoch": 1.7528489155740719,
"grad_norm": 0.26989975571632385,
"learning_rate": 4.005813178992091e-07,
"loss": 0.2880594730377197,
"step": 1789
},
{
"epoch": 1.7538291875995589,
"grad_norm": 0.31765440106391907,
"learning_rate": 3.9747573273779816e-07,
"loss": 0.3588324785232544,
"step": 1790
},
{
"epoch": 1.7548094596250459,
"grad_norm": 0.31180277466773987,
"learning_rate": 3.9438173442575e-07,
"loss": 0.3017778992652893,
"step": 1791
},
{
"epoch": 1.755789731650533,
"grad_norm": 0.26902875304222107,
"learning_rate": 3.9129933075219837e-07,
"loss": 0.26640191674232483,
"step": 1792
},
{
"epoch": 1.75677000367602,
"grad_norm": 0.4574669301509857,
"learning_rate": 3.882285294770938e-07,
"loss": 0.32540130615234375,
"step": 1793
},
{
"epoch": 1.757750275701507,
"grad_norm": 0.2612755298614502,
"learning_rate": 3.851693383311722e-07,
"loss": 0.27157357335090637,
"step": 1794
},
{
"epoch": 1.7587305477269943,
"grad_norm": 0.3225047290325165,
"learning_rate": 3.821217650159453e-07,
"loss": 0.37628838419914246,
"step": 1795
},
{
"epoch": 1.7597108197524813,
"grad_norm": 0.41944748163223267,
"learning_rate": 3.790858172036738e-07,
"loss": 0.31284916400909424,
"step": 1796
},
{
"epoch": 1.7606910917779683,
"grad_norm": 0.3646566569805145,
"learning_rate": 3.760615025373543e-07,
"loss": 0.30310750007629395,
"step": 1797
},
{
"epoch": 1.7616713638034556,
"grad_norm": 0.2894865870475769,
"learning_rate": 3.730488286306944e-07,
"loss": 0.3324729800224304,
"step": 1798
},
{
"epoch": 1.7626516358289426,
"grad_norm": 0.28256338834762573,
"learning_rate": 3.7004780306809873e-07,
"loss": 0.27221590280532837,
"step": 1799
},
{
"epoch": 1.7636319078544296,
"grad_norm": 0.3326883614063263,
"learning_rate": 3.6705843340464287e-07,
"loss": 0.3260238766670227,
"step": 1800
},
{
"epoch": 1.7646121798799168,
"grad_norm": 0.2526184022426605,
"learning_rate": 3.6408072716606346e-07,
"loss": 0.2859748899936676,
"step": 1801
},
{
"epoch": 1.7655924519054036,
"grad_norm": 0.319115549325943,
"learning_rate": 3.611146918487307e-07,
"loss": 0.29765060544013977,
"step": 1802
},
{
"epoch": 1.7665727239308908,
"grad_norm": 0.3791467547416687,
"learning_rate": 3.581603349196372e-07,
"loss": 0.2909999489784241,
"step": 1803
},
{
"epoch": 1.767552995956378,
"grad_norm": 0.22214040160179138,
"learning_rate": 3.552176638163707e-07,
"loss": 0.2361432909965515,
"step": 1804
},
{
"epoch": 1.7685332679818648,
"grad_norm": 0.34701231122016907,
"learning_rate": 3.522866859471047e-07,
"loss": 0.3561060428619385,
"step": 1805
},
{
"epoch": 1.769513540007352,
"grad_norm": 0.259863018989563,
"learning_rate": 3.4936740869057075e-07,
"loss": 0.3169648051261902,
"step": 1806
},
{
"epoch": 1.770493812032839,
"grad_norm": 0.26312339305877686,
"learning_rate": 3.46459839396045e-07,
"loss": 0.25871801376342773,
"step": 1807
},
{
"epoch": 1.771474084058326,
"grad_norm": 0.30160316824913025,
"learning_rate": 3.435639853833317e-07,
"loss": 0.2903321385383606,
"step": 1808
},
{
"epoch": 1.7724543560838133,
"grad_norm": 0.2984578609466553,
"learning_rate": 3.406798539427386e-07,
"loss": 0.3002643585205078,
"step": 1809
},
{
"epoch": 1.7734346281093003,
"grad_norm": 0.26575908064842224,
"learning_rate": 3.3780745233506384e-07,
"loss": 0.29481959342956543,
"step": 1810
},
{
"epoch": 1.7744149001347873,
"grad_norm": 0.33584773540496826,
"learning_rate": 3.3494678779157464e-07,
"loss": 0.3179909586906433,
"step": 1811
},
{
"epoch": 1.7753951721602745,
"grad_norm": 0.35129356384277344,
"learning_rate": 3.320978675139919e-07,
"loss": 0.3006082773208618,
"step": 1812
},
{
"epoch": 1.7763754441857615,
"grad_norm": 0.2673183083534241,
"learning_rate": 3.2926069867446673e-07,
"loss": 0.31049245595932007,
"step": 1813
},
{
"epoch": 1.7773557162112485,
"grad_norm": 0.3915678560733795,
"learning_rate": 3.2643528841556973e-07,
"loss": 0.31491610407829285,
"step": 1814
},
{
"epoch": 1.7783359882367358,
"grad_norm": 0.3239181637763977,
"learning_rate": 3.2362164385026704e-07,
"loss": 0.34335559606552124,
"step": 1815
},
{
"epoch": 1.7793162602622228,
"grad_norm": 0.3199029266834259,
"learning_rate": 3.2081977206190715e-07,
"loss": 0.4266231656074524,
"step": 1816
},
{
"epoch": 1.7802965322877098,
"grad_norm": 0.30837273597717285,
"learning_rate": 3.180296801041971e-07,
"loss": 0.2728562653064728,
"step": 1817
},
{
"epoch": 1.781276804313197,
"grad_norm": 0.3265760540962219,
"learning_rate": 3.1525137500119207e-07,
"loss": 0.357178270816803,
"step": 1818
},
{
"epoch": 1.782257076338684,
"grad_norm": 0.2554183304309845,
"learning_rate": 3.1248486374726884e-07,
"loss": 0.22811266779899597,
"step": 1819
},
{
"epoch": 1.783237348364171,
"grad_norm": 0.31561270356178284,
"learning_rate": 3.09730153307119e-07,
"loss": 0.3208150863647461,
"step": 1820
},
{
"epoch": 1.7842176203896583,
"grad_norm": 0.45564520359039307,
"learning_rate": 3.069872506157212e-07,
"loss": 0.24927136301994324,
"step": 1821
},
{
"epoch": 1.785197892415145,
"grad_norm": 0.23808327317237854,
"learning_rate": 3.042561625783308e-07,
"loss": 0.22433635592460632,
"step": 1822
},
{
"epoch": 1.7861781644406323,
"grad_norm": 0.38921868801116943,
"learning_rate": 3.015368960704584e-07,
"loss": 0.32440510392189026,
"step": 1823
},
{
"epoch": 1.7871584364661195,
"grad_norm": 0.3089323341846466,
"learning_rate": 2.988294579378537e-07,
"loss": 0.26491740345954895,
"step": 1824
},
{
"epoch": 1.7881387084916063,
"grad_norm": 0.37726542353630066,
"learning_rate": 2.9613385499648926e-07,
"loss": 0.3237096667289734,
"step": 1825
},
{
"epoch": 1.7891189805170935,
"grad_norm": 0.28678110241889954,
"learning_rate": 2.934500940325413e-07,
"loss": 0.3018602132797241,
"step": 1826
},
{
"epoch": 1.7900992525425807,
"grad_norm": 0.3234625458717346,
"learning_rate": 2.9077818180237693e-07,
"loss": 0.30036652088165283,
"step": 1827
},
{
"epoch": 1.7910795245680675,
"grad_norm": 0.28121933341026306,
"learning_rate": 2.881181250325299e-07,
"loss": 0.2576819956302643,
"step": 1828
},
{
"epoch": 1.7920597965935547,
"grad_norm": 0.2533399760723114,
"learning_rate": 2.8546993041969173e-07,
"loss": 0.2615036368370056,
"step": 1829
},
{
"epoch": 1.7930400686190417,
"grad_norm": 0.287164568901062,
"learning_rate": 2.8283360463068786e-07,
"loss": 0.26755183935165405,
"step": 1830
},
{
"epoch": 1.7940203406445288,
"grad_norm": 0.24917367100715637,
"learning_rate": 2.802091543024671e-07,
"loss": 0.28778645396232605,
"step": 1831
},
{
"epoch": 1.795000612670016,
"grad_norm": 0.3462033271789551,
"learning_rate": 2.7759658604207705e-07,
"loss": 0.40104758739471436,
"step": 1832
},
{
"epoch": 1.795980884695503,
"grad_norm": 0.7626352310180664,
"learning_rate": 2.7499590642665773e-07,
"loss": 0.3287285566329956,
"step": 1833
},
{
"epoch": 1.79696115672099,
"grad_norm": 0.3454156219959259,
"learning_rate": 2.724071220034158e-07,
"loss": 0.3017869293689728,
"step": 1834
},
{
"epoch": 1.7979414287464772,
"grad_norm": 0.2537349462509155,
"learning_rate": 2.6983023928961406e-07,
"loss": 0.24381256103515625,
"step": 1835
},
{
"epoch": 1.7989217007719642,
"grad_norm": 0.32373178005218506,
"learning_rate": 2.6726526477254986e-07,
"loss": 0.3076136112213135,
"step": 1836
},
{
"epoch": 1.7999019727974512,
"grad_norm": 1.0752116441726685,
"learning_rate": 2.647122049095463e-07,
"loss": 0.29388776421546936,
"step": 1837
},
{
"epoch": 1.8008822448229385,
"grad_norm": 0.3435729742050171,
"learning_rate": 2.621710661279253e-07,
"loss": 0.3035465478897095,
"step": 1838
},
{
"epoch": 1.8018625168484255,
"grad_norm": 0.2958340048789978,
"learning_rate": 2.596418548250029e-07,
"loss": 0.30120569467544556,
"step": 1839
},
{
"epoch": 1.8028427888739125,
"grad_norm": 0.25627389550209045,
"learning_rate": 2.571245773680647e-07,
"loss": 0.25132662057876587,
"step": 1840
},
{
"epoch": 1.8038230608993997,
"grad_norm": 0.2710166275501251,
"learning_rate": 2.546192400943537e-07,
"loss": 0.3297197222709656,
"step": 1841
},
{
"epoch": 1.8048033329248867,
"grad_norm": 0.4414254128932953,
"learning_rate": 2.52125849311054e-07,
"loss": 0.33345165848731995,
"step": 1842
},
{
"epoch": 1.8057836049503737,
"grad_norm": 0.37883245944976807,
"learning_rate": 2.4964441129527337e-07,
"loss": 0.4159780740737915,
"step": 1843
},
{
"epoch": 1.806763876975861,
"grad_norm": 0.31123027205467224,
"learning_rate": 2.4717493229402976e-07,
"loss": 0.22239896655082703,
"step": 1844
},
{
"epoch": 1.8077441490013477,
"grad_norm": 0.28076764941215515,
"learning_rate": 2.447174185242324e-07,
"loss": 0.25164616107940674,
"step": 1845
},
{
"epoch": 1.808724421026835,
"grad_norm": 0.449293315410614,
"learning_rate": 2.4227187617267057e-07,
"loss": 0.32495129108428955,
"step": 1846
},
{
"epoch": 1.8097046930523222,
"grad_norm": 0.2799157202243805,
"learning_rate": 2.398383113959929e-07,
"loss": 0.27509021759033203,
"step": 1847
},
{
"epoch": 1.810684965077809,
"grad_norm": 0.2735052704811096,
"learning_rate": 2.3741673032069757e-07,
"loss": 0.26909664273262024,
"step": 1848
},
{
"epoch": 1.8116652371032962,
"grad_norm": 0.3256213963031769,
"learning_rate": 2.3500713904311023e-07,
"loss": 0.2873234450817108,
"step": 1849
},
{
"epoch": 1.8126455091287832,
"grad_norm": 0.3012024164199829,
"learning_rate": 2.3260954362937626e-07,
"loss": 0.2689579129219055,
"step": 1850
},
{
"epoch": 1.8136257811542702,
"grad_norm": 0.37649938464164734,
"learning_rate": 2.3022395011543687e-07,
"loss": 0.36283984780311584,
"step": 1851
},
{
"epoch": 1.8146060531797574,
"grad_norm": 0.5700175762176514,
"learning_rate": 2.2785036450702237e-07,
"loss": 0.29013875126838684,
"step": 1852
},
{
"epoch": 1.8155863252052444,
"grad_norm": 0.32501351833343506,
"learning_rate": 2.2548879277963065e-07,
"loss": 0.3003235459327698,
"step": 1853
},
{
"epoch": 1.8165665972307314,
"grad_norm": 0.29464301466941833,
"learning_rate": 2.2313924087851657e-07,
"loss": 0.2945302724838257,
"step": 1854
},
{
"epoch": 1.8175468692562187,
"grad_norm": 0.24878771603107452,
"learning_rate": 2.2080171471867362e-07,
"loss": 0.31488776206970215,
"step": 1855
},
{
"epoch": 1.8185271412817057,
"grad_norm": 0.29707762598991394,
"learning_rate": 2.1847622018482283e-07,
"loss": 0.3090899586677551,
"step": 1856
},
{
"epoch": 1.8195074133071927,
"grad_norm": 0.39142900705337524,
"learning_rate": 2.161627631313923e-07,
"loss": 0.2969425916671753,
"step": 1857
},
{
"epoch": 1.82048768533268,
"grad_norm": 1.6219656467437744,
"learning_rate": 2.1386134938250645e-07,
"loss": 0.3006933927536011,
"step": 1858
},
{
"epoch": 1.821467957358167,
"grad_norm": 0.26177123188972473,
"learning_rate": 2.1157198473197417e-07,
"loss": 0.25019460916519165,
"step": 1859
},
{
"epoch": 1.822448229383654,
"grad_norm": 0.26025912165641785,
"learning_rate": 2.0929467494326616e-07,
"loss": 0.23798510432243347,
"step": 1860
},
{
"epoch": 1.8234285014091411,
"grad_norm": 0.27657848596572876,
"learning_rate": 2.0702942574950812e-07,
"loss": 0.3156519830226898,
"step": 1861
},
{
"epoch": 1.8244087734346282,
"grad_norm": 0.3151538372039795,
"learning_rate": 2.047762428534622e-07,
"loss": 0.36792927980422974,
"step": 1862
},
{
"epoch": 1.8253890454601152,
"grad_norm": 0.3295036852359772,
"learning_rate": 2.0253513192751374e-07,
"loss": 0.35760486125946045,
"step": 1863
},
{
"epoch": 1.8263693174856024,
"grad_norm": 0.33724159002304077,
"learning_rate": 2.0030609861365568e-07,
"loss": 0.30442699790000916,
"step": 1864
},
{
"epoch": 1.8273495895110892,
"grad_norm": 0.2889494299888611,
"learning_rate": 1.9808914852347817e-07,
"loss": 0.3131023943424225,
"step": 1865
},
{
"epoch": 1.8283298615365764,
"grad_norm": 0.2924818992614746,
"learning_rate": 1.9588428723814945e-07,
"loss": 0.32149678468704224,
"step": 1866
},
{
"epoch": 1.8293101335620636,
"grad_norm": 0.2839803695678711,
"learning_rate": 1.9369152030840553e-07,
"loss": 0.2898765206336975,
"step": 1867
},
{
"epoch": 1.8302904055875504,
"grad_norm": 0.29278966784477234,
"learning_rate": 1.9151085325453512e-07,
"loss": 0.2627498507499695,
"step": 1868
},
{
"epoch": 1.8312706776130376,
"grad_norm": 0.22664348781108856,
"learning_rate": 1.8934229156636453e-07,
"loss": 0.2446165680885315,
"step": 1869
},
{
"epoch": 1.8322509496385246,
"grad_norm": 0.4188863933086395,
"learning_rate": 1.8718584070324452e-07,
"loss": 0.32859066128730774,
"step": 1870
},
{
"epoch": 1.8332312216640116,
"grad_norm": 0.3696195185184479,
"learning_rate": 1.8504150609403858e-07,
"loss": 0.3229065537452698,
"step": 1871
},
{
"epoch": 1.8342114936894989,
"grad_norm": 1.0871840715408325,
"learning_rate": 1.8290929313710514e-07,
"loss": 0.3121337294578552,
"step": 1872
},
{
"epoch": 1.8351917657149859,
"grad_norm": 0.25832363963127136,
"learning_rate": 1.807892072002898e-07,
"loss": 0.26344549655914307,
"step": 1873
},
{
"epoch": 1.8361720377404729,
"grad_norm": 0.43513187766075134,
"learning_rate": 1.7868125362090482e-07,
"loss": 0.2793174684047699,
"step": 1874
},
{
"epoch": 1.8371523097659601,
"grad_norm": 0.2573612630367279,
"learning_rate": 1.765854377057219e-07,
"loss": 0.24476739764213562,
"step": 1875
},
{
"epoch": 1.8381325817914471,
"grad_norm": 0.4029923975467682,
"learning_rate": 1.7450176473095436e-07,
"loss": 0.2533695697784424,
"step": 1876
},
{
"epoch": 1.8391128538169341,
"grad_norm": 0.6414059996604919,
"learning_rate": 1.724302399422456e-07,
"loss": 0.3746204972267151,
"step": 1877
},
{
"epoch": 1.8400931258424214,
"grad_norm": 0.2549140751361847,
"learning_rate": 1.7037086855465902e-07,
"loss": 0.2832810878753662,
"step": 1878
},
{
"epoch": 1.8410733978679084,
"grad_norm": 0.27323755621910095,
"learning_rate": 1.6832365575265742e-07,
"loss": 0.25520244240760803,
"step": 1879
},
{
"epoch": 1.8420536698933954,
"grad_norm": 0.35951438546180725,
"learning_rate": 1.6628860669009816e-07,
"loss": 0.3030947148799896,
"step": 1880
},
{
"epoch": 1.8430339419188826,
"grad_norm": 0.27295824885368347,
"learning_rate": 1.6426572649021477e-07,
"loss": 0.3106752336025238,
"step": 1881
},
{
"epoch": 1.8440142139443696,
"grad_norm": 0.34768903255462646,
"learning_rate": 1.6225502024560468e-07,
"loss": 0.36633622646331787,
"step": 1882
},
{
"epoch": 1.8449944859698566,
"grad_norm": 0.3569660782814026,
"learning_rate": 1.6025649301821877e-07,
"loss": 0.30374157428741455,
"step": 1883
},
{
"epoch": 1.8459747579953438,
"grad_norm": 0.26604169607162476,
"learning_rate": 1.582701498393474e-07,
"loss": 0.3108394145965576,
"step": 1884
},
{
"epoch": 1.8469550300208308,
"grad_norm": 0.2547454237937927,
"learning_rate": 1.562959957096072e-07,
"loss": 0.2749966084957123,
"step": 1885
},
{
"epoch": 1.8479353020463178,
"grad_norm": 0.45910000801086426,
"learning_rate": 1.543340355989287e-07,
"loss": 0.33513882756233215,
"step": 1886
},
{
"epoch": 1.848915574071805,
"grad_norm": 0.3073311746120453,
"learning_rate": 1.5238427444654368e-07,
"loss": 0.3324298858642578,
"step": 1887
},
{
"epoch": 1.8498958460972919,
"grad_norm": 0.3074125647544861,
"learning_rate": 1.5044671716097414e-07,
"loss": 0.3255133628845215,
"step": 1888
},
{
"epoch": 1.850876118122779,
"grad_norm": 0.290433406829834,
"learning_rate": 1.4852136862001766e-07,
"loss": 0.30217599868774414,
"step": 1889
},
{
"epoch": 1.8518563901482663,
"grad_norm": 0.313102662563324,
"learning_rate": 1.466082336707375e-07,
"loss": 0.3226086497306824,
"step": 1890
},
{
"epoch": 1.852836662173753,
"grad_norm": 0.35070061683654785,
"learning_rate": 1.4470731712944885e-07,
"loss": 0.3071075677871704,
"step": 1891
},
{
"epoch": 1.8538169341992403,
"grad_norm": 0.5057258605957031,
"learning_rate": 1.428186237817053e-07,
"loss": 0.2564800977706909,
"step": 1892
},
{
"epoch": 1.8547972062247273,
"grad_norm": 0.28655409812927246,
"learning_rate": 1.4094215838229176e-07,
"loss": 0.28801411390304565,
"step": 1893
},
{
"epoch": 1.8557774782502143,
"grad_norm": 0.5691519975662231,
"learning_rate": 1.3907792565520716e-07,
"loss": 0.3079022467136383,
"step": 1894
},
{
"epoch": 1.8567577502757016,
"grad_norm": 0.2758004665374756,
"learning_rate": 1.372259302936546e-07,
"loss": 0.28049877285957336,
"step": 1895
},
{
"epoch": 1.8577380223011886,
"grad_norm": 0.30326518416404724,
"learning_rate": 1.3538617696003066e-07,
"loss": 0.25389379262924194,
"step": 1896
},
{
"epoch": 1.8587182943266756,
"grad_norm": 0.28356459736824036,
"learning_rate": 1.3355867028591209e-07,
"loss": 0.2727922201156616,
"step": 1897
},
{
"epoch": 1.8596985663521628,
"grad_norm": 0.27547943592071533,
"learning_rate": 1.3174341487204478e-07,
"loss": 0.27276551723480225,
"step": 1898
},
{
"epoch": 1.8606788383776498,
"grad_norm": 0.29917097091674805,
"learning_rate": 1.2994041528833267e-07,
"loss": 0.28016000986099243,
"step": 1899
},
{
"epoch": 1.8616591104031368,
"grad_norm": 0.2904070019721985,
"learning_rate": 1.2814967607382433e-07,
"loss": 0.28806230425834656,
"step": 1900
},
{
"epoch": 1.862639382428624,
"grad_norm": 2.247464179992676,
"learning_rate": 1.263712017367036e-07,
"loss": 0.2650207281112671,
"step": 1901
},
{
"epoch": 1.863619654454111,
"grad_norm": 1.078545093536377,
"learning_rate": 1.246049967542773e-07,
"loss": 0.3239307403564453,
"step": 1902
},
{
"epoch": 1.864599926479598,
"grad_norm": 0.336080402135849,
"learning_rate": 1.2285106557296479e-07,
"loss": 0.297506719827652,
"step": 1903
},
{
"epoch": 1.8655801985050853,
"grad_norm": 0.29012835025787354,
"learning_rate": 1.211094126082857e-07,
"loss": 0.28052985668182373,
"step": 1904
},
{
"epoch": 1.8665604705305723,
"grad_norm": 0.29115766286849976,
"learning_rate": 1.193800422448499e-07,
"loss": 0.30387037992477417,
"step": 1905
},
{
"epoch": 1.8675407425560593,
"grad_norm": 0.38615137338638306,
"learning_rate": 1.1766295883634427e-07,
"loss": 0.31118592619895935,
"step": 1906
},
{
"epoch": 1.8685210145815465,
"grad_norm": 0.3901364505290985,
"learning_rate": 1.1595816670552429e-07,
"loss": 0.34375613927841187,
"step": 1907
},
{
"epoch": 1.8695012866070333,
"grad_norm": 0.2811959683895111,
"learning_rate": 1.1426567014420297e-07,
"loss": 0.30543720722198486,
"step": 1908
},
{
"epoch": 1.8704815586325205,
"grad_norm": 0.2971404492855072,
"learning_rate": 1.12585473413237e-07,
"loss": 0.3038825988769531,
"step": 1909
},
{
"epoch": 1.8714618306580078,
"grad_norm": 0.30307215452194214,
"learning_rate": 1.1091758074252112e-07,
"loss": 0.30602410435676575,
"step": 1910
},
{
"epoch": 1.8724421026834945,
"grad_norm": 0.33499452471733093,
"learning_rate": 1.0926199633097156e-07,
"loss": 0.30949532985687256,
"step": 1911
},
{
"epoch": 1.8734223747089818,
"grad_norm": 0.25841644406318665,
"learning_rate": 1.076187243465221e-07,
"loss": 0.2441016286611557,
"step": 1912
},
{
"epoch": 1.8744026467344688,
"grad_norm": 0.2458769530057907,
"learning_rate": 1.0598776892610685e-07,
"loss": 0.24789434671401978,
"step": 1913
},
{
"epoch": 1.8753829187599558,
"grad_norm": 0.38307449221611023,
"learning_rate": 1.0436913417565364e-07,
"loss": 0.3137006163597107,
"step": 1914
},
{
"epoch": 1.876363190785443,
"grad_norm": 0.24433670938014984,
"learning_rate": 1.0276282417007399e-07,
"loss": 0.21425378322601318,
"step": 1915
},
{
"epoch": 1.87734346281093,
"grad_norm": 0.2898837625980377,
"learning_rate": 1.0116884295325202e-07,
"loss": 0.2932490110397339,
"step": 1916
},
{
"epoch": 1.878323734836417,
"grad_norm": 0.25592657923698425,
"learning_rate": 9.958719453803278e-08,
"loss": 0.24505218863487244,
"step": 1917
},
{
"epoch": 1.8793040068619042,
"grad_norm": 0.35855525732040405,
"learning_rate": 9.801788290621505e-08,
"loss": 0.3714182376861572,
"step": 1918
},
{
"epoch": 1.8802842788873912,
"grad_norm": 0.32341647148132324,
"learning_rate": 9.646091200853802e-08,
"loss": 0.3497477173805237,
"step": 1919
},
{
"epoch": 1.8812645509128783,
"grad_norm": 0.5551256537437439,
"learning_rate": 9.491628576467516e-08,
"loss": 0.2258259654045105,
"step": 1920
},
{
"epoch": 1.8822448229383655,
"grad_norm": 0.3139249384403229,
"learning_rate": 9.338400806321979e-08,
"loss": 0.3736576735973358,
"step": 1921
},
{
"epoch": 1.8832250949638525,
"grad_norm": 0.2844786047935486,
"learning_rate": 9.186408276168012e-08,
"loss": 0.29438072443008423,
"step": 1922
},
{
"epoch": 1.8842053669893395,
"grad_norm": 0.26426976919174194,
"learning_rate": 9.035651368646647e-08,
"loss": 0.2981555759906769,
"step": 1923
},
{
"epoch": 1.8851856390148267,
"grad_norm": 0.294452965259552,
"learning_rate": 8.886130463288178e-08,
"loss": 0.3127821385860443,
"step": 1924
},
{
"epoch": 1.8861659110403137,
"grad_norm": 0.30885639786720276,
"learning_rate": 8.737845936511335e-08,
"loss": 0.31298965215682983,
"step": 1925
},
{
"epoch": 1.8871461830658007,
"grad_norm": 0.38961634039878845,
"learning_rate": 8.590798161622227e-08,
"loss": 0.3624933660030365,
"step": 1926
},
{
"epoch": 1.888126455091288,
"grad_norm": 0.29062607884407043,
"learning_rate": 8.444987508813451e-08,
"loss": 0.35337382555007935,
"step": 1927
},
{
"epoch": 1.8891067271167747,
"grad_norm": 0.3214566707611084,
"learning_rate": 8.300414345163043e-08,
"loss": 0.3193872570991516,
"step": 1928
},
{
"epoch": 1.890086999142262,
"grad_norm": 0.2936144471168518,
"learning_rate": 8.157079034633974e-08,
"loss": 0.3158854842185974,
"step": 1929
},
{
"epoch": 1.8910672711677492,
"grad_norm": 0.2875578701496124,
"learning_rate": 8.014981938072542e-08,
"loss": 0.2592881917953491,
"step": 1930
},
{
"epoch": 1.892047543193236,
"grad_norm": 0.42762941122055054,
"learning_rate": 7.874123413208145e-08,
"loss": 0.2912369966506958,
"step": 1931
},
{
"epoch": 1.8930278152187232,
"grad_norm": 0.34188103675842285,
"learning_rate": 7.734503814651906e-08,
"loss": 0.3902429938316345,
"step": 1932
},
{
"epoch": 1.8940080872442102,
"grad_norm": 0.29489389061927795,
"learning_rate": 7.59612349389599e-08,
"loss": 0.2799641489982605,
"step": 1933
},
{
"epoch": 1.8949883592696972,
"grad_norm": 0.3210509419441223,
"learning_rate": 7.458982799312676e-08,
"loss": 0.2712163031101227,
"step": 1934
},
{
"epoch": 1.8959686312951844,
"grad_norm": 0.34548133611679077,
"learning_rate": 7.32308207615351e-08,
"loss": 0.3382522463798523,
"step": 1935
},
{
"epoch": 1.8969489033206715,
"grad_norm": 0.35671624541282654,
"learning_rate": 7.188421666548373e-08,
"loss": 0.29560160636901855,
"step": 1936
},
{
"epoch": 1.8979291753461585,
"grad_norm": 0.2601449489593506,
"learning_rate": 7.055001909504755e-08,
"loss": 0.22935447096824646,
"step": 1937
},
{
"epoch": 1.8989094473716457,
"grad_norm": 0.2522633969783783,
"learning_rate": 6.922823140906754e-08,
"loss": 0.27867788076400757,
"step": 1938
},
{
"epoch": 1.8998897193971327,
"grad_norm": 0.3229309618473053,
"learning_rate": 6.791885693514134e-08,
"loss": 0.31017035245895386,
"step": 1939
},
{
"epoch": 1.9008699914226197,
"grad_norm": 0.2899870276451111,
"learning_rate": 6.662189896961824e-08,
"loss": 0.2620551586151123,
"step": 1940
},
{
"epoch": 1.901850263448107,
"grad_norm": 0.26009392738342285,
"learning_rate": 6.533736077758868e-08,
"loss": 0.28719398379325867,
"step": 1941
},
{
"epoch": 1.902830535473594,
"grad_norm": 0.2573067247867584,
"learning_rate": 6.406524559287641e-08,
"loss": 0.28080788254737854,
"step": 1942
},
{
"epoch": 1.903810807499081,
"grad_norm": 0.4719696342945099,
"learning_rate": 6.280555661802857e-08,
"loss": 0.3660171627998352,
"step": 1943
},
{
"epoch": 1.9047910795245682,
"grad_norm": 0.38690435886383057,
"learning_rate": 6.15582970243117e-08,
"loss": 0.33985990285873413,
"step": 1944
},
{
"epoch": 1.9057713515500552,
"grad_norm": 0.3255194425582886,
"learning_rate": 6.032346995169968e-08,
"loss": 0.3487090468406677,
"step": 1945
},
{
"epoch": 1.9067516235755422,
"grad_norm": 0.30564531683921814,
"learning_rate": 5.9101078508869144e-08,
"loss": 0.33180680871009827,
"step": 1946
},
{
"epoch": 1.9077318956010294,
"grad_norm": 0.2677140533924103,
"learning_rate": 5.7891125773187896e-08,
"loss": 0.2766750454902649,
"step": 1947
},
{
"epoch": 1.9087121676265164,
"grad_norm": 0.3282104432582855,
"learning_rate": 5.669361479071156e-08,
"loss": 0.2713039219379425,
"step": 1948
},
{
"epoch": 1.9096924396520034,
"grad_norm": 0.3672855496406555,
"learning_rate": 5.550854857617194e-08,
"loss": 0.3334164619445801,
"step": 1949
},
{
"epoch": 1.9106727116774906,
"grad_norm": 0.24114011228084564,
"learning_rate": 5.4335930112972e-08,
"loss": 0.3081708252429962,
"step": 1950
},
{
"epoch": 1.9116529837029774,
"grad_norm": 1.5846797227859497,
"learning_rate": 5.3175762353177563e-08,
"loss": 0.27858009934425354,
"step": 1951
},
{
"epoch": 1.9126332557284647,
"grad_norm": 0.3412085771560669,
"learning_rate": 5.202804821750951e-08,
"loss": 0.24326318502426147,
"step": 1952
},
{
"epoch": 1.9136135277539519,
"grad_norm": 0.3048664331436157,
"learning_rate": 5.089279059533658e-08,
"loss": 0.27641135454177856,
"step": 1953
},
{
"epoch": 1.9145937997794387,
"grad_norm": 0.29281336069107056,
"learning_rate": 4.9769992344668174e-08,
"loss": 0.25403887033462524,
"step": 1954
},
{
"epoch": 1.915574071804926,
"grad_norm": 0.32677462697029114,
"learning_rate": 4.865965629214819e-08,
"loss": 0.2981630563735962,
"step": 1955
},
{
"epoch": 1.916554343830413,
"grad_norm": 0.504023015499115,
"learning_rate": 4.756178523304622e-08,
"loss": 0.30510103702545166,
"step": 1956
},
{
"epoch": 1.9175346158559,
"grad_norm": 0.3021174669265747,
"learning_rate": 4.6476381931251366e-08,
"loss": 0.32467198371887207,
"step": 1957
},
{
"epoch": 1.9185148878813871,
"grad_norm": 0.26853665709495544,
"learning_rate": 4.5403449119265085e-08,
"loss": 0.3218815326690674,
"step": 1958
},
{
"epoch": 1.9194951599068741,
"grad_norm": 0.35650959610939026,
"learning_rate": 4.434298949819449e-08,
"loss": 0.34601783752441406,
"step": 1959
},
{
"epoch": 1.9204754319323611,
"grad_norm": 0.3097904622554779,
"learning_rate": 4.329500573774459e-08,
"loss": 0.3032281994819641,
"step": 1960
},
{
"epoch": 1.9214557039578484,
"grad_norm": 0.3040037453174591,
"learning_rate": 4.225950047621441e-08,
"loss": 0.2551646828651428,
"step": 1961
},
{
"epoch": 1.9224359759833354,
"grad_norm": 0.24835410714149475,
"learning_rate": 4.123647632048644e-08,
"loss": 0.25921785831451416,
"step": 1962
},
{
"epoch": 1.9234162480088224,
"grad_norm": 0.2835935950279236,
"learning_rate": 4.02259358460233e-08,
"loss": 0.28769272565841675,
"step": 1963
},
{
"epoch": 1.9243965200343096,
"grad_norm": 0.36556360125541687,
"learning_rate": 3.922788159685942e-08,
"loss": 0.3693891167640686,
"step": 1964
},
{
"epoch": 1.9253767920597966,
"grad_norm": 0.34024298191070557,
"learning_rate": 3.8242316085594923e-08,
"loss": 0.3239486813545227,
"step": 1965
},
{
"epoch": 1.9263570640852836,
"grad_norm": 0.2928847372531891,
"learning_rate": 3.726924179339009e-08,
"loss": 0.30624955892562866,
"step": 1966
},
{
"epoch": 1.9273373361107708,
"grad_norm": 0.24757793545722961,
"learning_rate": 3.630866116995757e-08,
"loss": 0.26342934370040894,
"step": 1967
},
{
"epoch": 1.9283176081362579,
"grad_norm": 0.3188454210758209,
"learning_rate": 3.536057663355852e-08,
"loss": 0.34307852387428284,
"step": 1968
},
{
"epoch": 1.9292978801617449,
"grad_norm": 0.2783260941505432,
"learning_rate": 3.44249905709948e-08,
"loss": 0.2881426513195038,
"step": 1969
},
{
"epoch": 1.930278152187232,
"grad_norm": 0.32684046030044556,
"learning_rate": 3.350190533760289e-08,
"loss": 0.3314821422100067,
"step": 1970
},
{
"epoch": 1.9312584242127189,
"grad_norm": 0.32812216877937317,
"learning_rate": 3.25913232572489e-08,
"loss": 0.3241310715675354,
"step": 1971
},
{
"epoch": 1.932238696238206,
"grad_norm": 0.26690545678138733,
"learning_rate": 3.169324662232187e-08,
"loss": 0.2864891290664673,
"step": 1972
},
{
"epoch": 1.9332189682636933,
"grad_norm": 0.2353801280260086,
"learning_rate": 3.080767769372939e-08,
"loss": 0.2477440983057022,
"step": 1973
},
{
"epoch": 1.93419924028918,
"grad_norm": 0.24104543030261993,
"learning_rate": 2.9934618700889205e-08,
"loss": 0.23625187575817108,
"step": 1974
},
{
"epoch": 1.9351795123146673,
"grad_norm": 0.32768771052360535,
"learning_rate": 2.907407184172706e-08,
"loss": 0.3142569661140442,
"step": 1975
},
{
"epoch": 1.9361597843401543,
"grad_norm": 0.36556240916252136,
"learning_rate": 2.8226039282667762e-08,
"loss": 0.302887499332428,
"step": 1976
},
{
"epoch": 1.9371400563656413,
"grad_norm": 0.30446141958236694,
"learning_rate": 2.7390523158633552e-08,
"loss": 0.30839449167251587,
"step": 1977
},
{
"epoch": 1.9381203283911286,
"grad_norm": 0.3645970821380615,
"learning_rate": 2.6567525573034102e-08,
"loss": 0.30195772647857666,
"step": 1978
},
{
"epoch": 1.9391006004166156,
"grad_norm": 0.2804482877254486,
"learning_rate": 2.57570485977654e-08,
"loss": 0.31752097606658936,
"step": 1979
},
{
"epoch": 1.9400808724421026,
"grad_norm": 0.2640190124511719,
"learning_rate": 2.495909427320198e-08,
"loss": 0.26195472478866577,
"step": 1980
},
{
"epoch": 1.9410611444675898,
"grad_norm": 0.3151174485683441,
"learning_rate": 2.4173664608193592e-08,
"loss": 0.3523339033126831,
"step": 1981
},
{
"epoch": 1.9420414164930768,
"grad_norm": 0.3739902377128601,
"learning_rate": 2.3400761580057442e-08,
"loss": 0.3262278139591217,
"step": 1982
},
{
"epoch": 1.9430216885185638,
"grad_norm": 0.3578769862651825,
"learning_rate": 2.264038713457706e-08,
"loss": 0.28649091720581055,
"step": 1983
},
{
"epoch": 1.944001960544051,
"grad_norm": 0.2557157278060913,
"learning_rate": 2.189254318599343e-08,
"loss": 0.2472064197063446,
"step": 1984
},
{
"epoch": 1.944982232569538,
"grad_norm": 0.3018450140953064,
"learning_rate": 2.1157231617002783e-08,
"loss": 0.26410800218582153,
"step": 1985
},
{
"epoch": 1.945962504595025,
"grad_norm": 0.33849969506263733,
"learning_rate": 2.0434454278752126e-08,
"loss": 0.32833996415138245,
"step": 1986
},
{
"epoch": 1.9469427766205123,
"grad_norm": 0.2271992564201355,
"learning_rate": 1.9724212990830938e-08,
"loss": 0.23445779085159302,
"step": 1987
},
{
"epoch": 1.9479230486459993,
"grad_norm": 0.3064795434474945,
"learning_rate": 1.9026509541272276e-08,
"loss": 0.30196964740753174,
"step": 1988
},
{
"epoch": 1.9489033206714863,
"grad_norm": 0.2898283898830414,
"learning_rate": 1.834134568654333e-08,
"loss": 0.32186174392700195,
"step": 1989
},
{
"epoch": 1.9498835926969735,
"grad_norm": 1.0373432636260986,
"learning_rate": 1.7668723151542667e-08,
"loss": 0.2805105745792389,
"step": 1990
},
{
"epoch": 1.9508638647224603,
"grad_norm": 0.318493127822876,
"learning_rate": 1.7008643629596866e-08,
"loss": 0.4004296660423279,
"step": 1991
},
{
"epoch": 1.9518441367479475,
"grad_norm": 0.29151806235313416,
"learning_rate": 1.6361108782456113e-08,
"loss": 0.28746330738067627,
"step": 1992
},
{
"epoch": 1.9528244087734348,
"grad_norm": 0.3635181486606598,
"learning_rate": 1.5726120240288632e-08,
"loss": 0.27345967292785645,
"step": 1993
},
{
"epoch": 1.9538046807989216,
"grad_norm": 0.2018825262784958,
"learning_rate": 1.510367960167569e-08,
"loss": 0.21855270862579346,
"step": 1994
},
{
"epoch": 1.9547849528244088,
"grad_norm": 0.784898579120636,
"learning_rate": 1.449378843361271e-08,
"loss": 0.28332945704460144,
"step": 1995
},
{
"epoch": 1.955765224849896,
"grad_norm": 0.3147640824317932,
"learning_rate": 1.3896448271499275e-08,
"loss": 0.3062807023525238,
"step": 1996
},
{
"epoch": 1.9567454968753828,
"grad_norm": 0.3980037271976471,
"learning_rate": 1.3311660619138578e-08,
"loss": 0.29567664861679077,
"step": 1997
},
{
"epoch": 1.95772576890087,
"grad_norm": 0.29037508368492126,
"learning_rate": 1.2739426948732426e-08,
"loss": 0.24397769570350647,
"step": 1998
},
{
"epoch": 1.958706040926357,
"grad_norm": 0.500783383846283,
"learning_rate": 1.2179748700879013e-08,
"loss": 0.3210737705230713,
"step": 1999
},
{
"epoch": 1.959686312951844,
"grad_norm": 0.23673944175243378,
"learning_rate": 1.1632627284566822e-08,
"loss": 0.2560279369354248,
"step": 2000
},
{
"epoch": 1.9606665849773313,
"grad_norm": 0.2513909339904785,
"learning_rate": 1.109806407717462e-08,
"loss": 0.30129140615463257,
"step": 2001
},
{
"epoch": 1.9616468570028183,
"grad_norm": 0.38034820556640625,
"learning_rate": 1.0576060424463686e-08,
"loss": 0.3537079989910126,
"step": 2002
},
{
"epoch": 1.9626271290283053,
"grad_norm": 0.3011312782764435,
"learning_rate": 1.006661764057837e-08,
"loss": 0.26366227865219116,
"step": 2003
},
{
"epoch": 1.9636074010537925,
"grad_norm": 0.3710026443004608,
"learning_rate": 9.56973700803887e-09,
"loss": 0.29284441471099854,
"step": 2004
},
{
"epoch": 1.9645876730792795,
"grad_norm": 0.35956400632858276,
"learning_rate": 9.085419777743465e-09,
"loss": 0.38616102933883667,
"step": 2005
},
{
"epoch": 1.9655679451047665,
"grad_norm": 1.3838956356048584,
"learning_rate": 8.613667168958506e-09,
"loss": 0.3074329197406769,
"step": 2006
},
{
"epoch": 1.9665482171302537,
"grad_norm": 0.3716168701648712,
"learning_rate": 8.15448036932176e-09,
"loss": 0.3507290780544281,
"step": 2007
},
{
"epoch": 1.9675284891557407,
"grad_norm": 0.8683135509490967,
"learning_rate": 7.707860534834632e-09,
"loss": 0.2847939431667328,
"step": 2008
},
{
"epoch": 1.9685087611812278,
"grad_norm": 0.25470155477523804,
"learning_rate": 7.273808789862724e-09,
"loss": 0.2504078447818756,
"step": 2009
},
{
"epoch": 1.969489033206715,
"grad_norm": 0.304253488779068,
"learning_rate": 6.852326227130835e-09,
"loss": 0.3440891206264496,
"step": 2010
},
{
"epoch": 1.970469305232202,
"grad_norm": 0.2823781967163086,
"learning_rate": 6.4434139077201865e-09,
"loss": 0.3278113007545471,
"step": 2011
},
{
"epoch": 1.971449577257689,
"grad_norm": 0.3159235715866089,
"learning_rate": 6.0470728610673155e-09,
"loss": 0.30382871627807617,
"step": 2012
},
{
"epoch": 1.9724298492831762,
"grad_norm": 0.3064816892147064,
"learning_rate": 5.6633040849601865e-09,
"loss": 0.34606504440307617,
"step": 2013
},
{
"epoch": 1.973410121308663,
"grad_norm": 0.27829864621162415,
"learning_rate": 5.292108545536523e-09,
"loss": 0.28693780303001404,
"step": 2014
},
{
"epoch": 1.9743903933341502,
"grad_norm": 0.34374088048934937,
"learning_rate": 4.933487177280483e-09,
"loss": 0.35004740953445435,
"step": 2015
},
{
"epoch": 1.9753706653596375,
"grad_norm": 0.5050590634346008,
"learning_rate": 4.587440883021543e-09,
"loss": 0.2870016098022461,
"step": 2016
},
{
"epoch": 1.9763509373851242,
"grad_norm": 0.3248060643672943,
"learning_rate": 4.253970533929508e-09,
"loss": 0.2557294964790344,
"step": 2017
},
{
"epoch": 1.9773312094106115,
"grad_norm": 0.3418165445327759,
"learning_rate": 3.9330769695167245e-09,
"loss": 0.2562110424041748,
"step": 2018
},
{
"epoch": 1.9783114814360985,
"grad_norm": 0.2986294627189636,
"learning_rate": 3.6247609976319818e-09,
"loss": 0.3390492796897888,
"step": 2019
},
{
"epoch": 1.9792917534615855,
"grad_norm": 0.4446433484554291,
"learning_rate": 3.3290233944605067e-09,
"loss": 0.2774254083633423,
"step": 2020
},
{
"epoch": 1.9802720254870727,
"grad_norm": 0.5931270122528076,
"learning_rate": 3.0458649045211897e-09,
"loss": 0.2819487452507019,
"step": 2021
},
{
"epoch": 1.9812522975125597,
"grad_norm": 0.4004150331020355,
"learning_rate": 2.7752862406654755e-09,
"loss": 0.311737596988678,
"step": 2022
},
{
"epoch": 1.9822325695380467,
"grad_norm": 0.2874223589897156,
"learning_rate": 2.5172880840745873e-09,
"loss": 0.26529502868652344,
"step": 2023
},
{
"epoch": 1.983212841563534,
"grad_norm": 0.2985301911830902,
"learning_rate": 2.2718710842584145e-09,
"loss": 0.27759504318237305,
"step": 2024
},
{
"epoch": 1.984193113589021,
"grad_norm": 0.2927117943763733,
"learning_rate": 2.0390358590538507e-09,
"loss": 0.28353866934776306,
"step": 2025
},
{
"epoch": 1.985173385614508,
"grad_norm": 0.3729081153869629,
"learning_rate": 1.8187829946236802e-09,
"loss": 0.3370395600795746,
"step": 2026
},
{
"epoch": 1.9861536576399952,
"grad_norm": 0.8061376810073853,
"learning_rate": 1.61111304545436e-09,
"loss": 0.23381732404232025,
"step": 2027
},
{
"epoch": 1.9871339296654822,
"grad_norm": 0.34020355343818665,
"learning_rate": 1.4160265343549084e-09,
"loss": 0.2879725694656372,
"step": 2028
},
{
"epoch": 1.9881142016909692,
"grad_norm": 0.2938496768474579,
"learning_rate": 1.2335239524541298e-09,
"loss": 0.3000931143760681,
"step": 2029
},
{
"epoch": 1.9890944737164564,
"grad_norm": 0.2332998365163803,
"learning_rate": 1.0636057592033899e-09,
"loss": 0.28940385580062866,
"step": 2030
},
{
"epoch": 1.9900747457419434,
"grad_norm": 0.27316349744796753,
"learning_rate": 9.062723823710651e-10,
"loss": 0.24524575471878052,
"step": 2031
},
{
"epoch": 1.9910550177674304,
"grad_norm": 0.41514623165130615,
"learning_rate": 7.615242180436521e-10,
"loss": 0.33912310004234314,
"step": 2032
},
{
"epoch": 1.9920352897929177,
"grad_norm": 0.5395570397377014,
"learning_rate": 6.293616306246586e-10,
"loss": 0.3492683172225952,
"step": 2033
},
{
"epoch": 1.9930155618184044,
"grad_norm": 0.2812623083591461,
"learning_rate": 5.09784952833492e-10,
"loss": 0.27273938059806824,
"step": 2034
},
{
"epoch": 1.9939958338438917,
"grad_norm": 0.31485092639923096,
"learning_rate": 4.027944857032395e-10,
"loss": 0.33989644050598145,
"step": 2035
},
{
"epoch": 1.994976105869379,
"grad_norm": 0.3575870990753174,
"learning_rate": 3.0839049858344406e-10,
"loss": 0.2723734378814697,
"step": 2036
},
{
"epoch": 1.9959563778948657,
"grad_norm": 0.3108024299144745,
"learning_rate": 2.265732291356626e-10,
"loss": 0.32159334421157837,
"step": 2037
},
{
"epoch": 1.996936649920353,
"grad_norm": 0.27574658393859863,
"learning_rate": 1.5734288333457692e-10,
"loss": 0.3451283574104309,
"step": 2038
},
{
"epoch": 1.99791692194584,
"grad_norm": 0.27294427156448364,
"learning_rate": 1.0069963546743833e-10,
"loss": 0.29713353514671326,
"step": 2039
},
{
"epoch": 1.998897193971327,
"grad_norm": 0.38074108958244324,
"learning_rate": 5.6643628134067655e-11,
"loss": 0.29969358444213867,
"step": 2040
},
{
"epoch": 1.9998774659968142,
"grad_norm": 0.2569642663002014,
"learning_rate": 2.5174972244634834e-11,
"loss": 0.26567474007606506,
"step": 2041
},
{
"epoch": 2.0,
"grad_norm": 1.0074602365493774,
"learning_rate": 6.2937470224344596e-12,
"loss": 0.6110332012176514,
"step": 2042
}
],
"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": true
},
"attributes": {}
}
},
"total_flos": 2.0191018389843476e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}