gyung's picture
Upload Gemma 4 native SFT 1Epoch checkpoint
d558150 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1021,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009802720254870727,
"grad_norm": 317.54534912109375,
"learning_rate": 0.0,
"loss": 13.04903793334961,
"step": 1
},
{
"epoch": 0.0019605440509741453,
"grad_norm": 800.3175048828125,
"learning_rate": 1.6129032258064518e-07,
"loss": 12.59231185913086,
"step": 2
},
{
"epoch": 0.002940816076461218,
"grad_norm": 107.61616516113281,
"learning_rate": 3.2258064516129035e-07,
"loss": 12.499323844909668,
"step": 3
},
{
"epoch": 0.003921088101948291,
"grad_norm": 159.5131378173828,
"learning_rate": 4.838709677419355e-07,
"loss": 12.74274730682373,
"step": 4
},
{
"epoch": 0.004901360127435363,
"grad_norm": 130.99020385742188,
"learning_rate": 6.451612903225807e-07,
"loss": 13.881667137145996,
"step": 5
},
{
"epoch": 0.005881632152922436,
"grad_norm": 259.9962158203125,
"learning_rate": 8.064516129032258e-07,
"loss": 12.893087387084961,
"step": 6
},
{
"epoch": 0.006861904178409509,
"grad_norm": 106.34725952148438,
"learning_rate": 9.67741935483871e-07,
"loss": 12.537609100341797,
"step": 7
},
{
"epoch": 0.007842176203896581,
"grad_norm": 109.43804931640625,
"learning_rate": 1.1290322580645162e-06,
"loss": 12.597084045410156,
"step": 8
},
{
"epoch": 0.008822448229383655,
"grad_norm": 82.59698486328125,
"learning_rate": 1.2903225806451614e-06,
"loss": 11.605478286743164,
"step": 9
},
{
"epoch": 0.009802720254870726,
"grad_norm": 77.49324035644531,
"learning_rate": 1.4516129032258066e-06,
"loss": 11.378276824951172,
"step": 10
},
{
"epoch": 0.0107829922803578,
"grad_norm": 62.177425384521484,
"learning_rate": 1.6129032258064516e-06,
"loss": 10.7266845703125,
"step": 11
},
{
"epoch": 0.011763264305844873,
"grad_norm": 26.330503463745117,
"learning_rate": 1.774193548387097e-06,
"loss": 9.117469787597656,
"step": 12
},
{
"epoch": 0.012743536331331944,
"grad_norm": 30.31421661376953,
"learning_rate": 1.935483870967742e-06,
"loss": 8.640227317810059,
"step": 13
},
{
"epoch": 0.013723808356819018,
"grad_norm": 20.590795516967773,
"learning_rate": 2.096774193548387e-06,
"loss": 8.7728271484375,
"step": 14
},
{
"epoch": 0.01470408038230609,
"grad_norm": 64.54769134521484,
"learning_rate": 2.2580645161290324e-06,
"loss": 8.358410835266113,
"step": 15
},
{
"epoch": 0.015684352407793162,
"grad_norm": 26.623218536376953,
"learning_rate": 2.4193548387096776e-06,
"loss": 7.563622951507568,
"step": 16
},
{
"epoch": 0.016664624433280234,
"grad_norm": 26.686445236206055,
"learning_rate": 2.580645161290323e-06,
"loss": 7.325545310974121,
"step": 17
},
{
"epoch": 0.01764489645876731,
"grad_norm": 22.71994400024414,
"learning_rate": 2.7419354838709676e-06,
"loss": 7.185205459594727,
"step": 18
},
{
"epoch": 0.01862516848425438,
"grad_norm": 20.37201690673828,
"learning_rate": 2.903225806451613e-06,
"loss": 7.332908630371094,
"step": 19
},
{
"epoch": 0.019605440509741452,
"grad_norm": 17.708080291748047,
"learning_rate": 3.0645161290322584e-06,
"loss": 6.68546199798584,
"step": 20
},
{
"epoch": 0.020585712535228527,
"grad_norm": 13.734796524047852,
"learning_rate": 3.225806451612903e-06,
"loss": 6.269188404083252,
"step": 21
},
{
"epoch": 0.0215659845607156,
"grad_norm": 14.581665992736816,
"learning_rate": 3.3870967741935484e-06,
"loss": 5.984618186950684,
"step": 22
},
{
"epoch": 0.02254625658620267,
"grad_norm": 12.097245216369629,
"learning_rate": 3.548387096774194e-06,
"loss": 5.684147357940674,
"step": 23
},
{
"epoch": 0.023526528611689745,
"grad_norm": 13.412657737731934,
"learning_rate": 3.7096774193548392e-06,
"loss": 5.540960788726807,
"step": 24
},
{
"epoch": 0.024506800637176817,
"grad_norm": 8.56202507019043,
"learning_rate": 3.870967741935484e-06,
"loss": 5.190829277038574,
"step": 25
},
{
"epoch": 0.02548707266266389,
"grad_norm": 7.289759635925293,
"learning_rate": 4.032258064516129e-06,
"loss": 5.197663307189941,
"step": 26
},
{
"epoch": 0.026467344688150964,
"grad_norm": 5.432161808013916,
"learning_rate": 4.193548387096774e-06,
"loss": 4.925121784210205,
"step": 27
},
{
"epoch": 0.027447616713638035,
"grad_norm": 3.859867811203003,
"learning_rate": 4.35483870967742e-06,
"loss": 4.4424943923950195,
"step": 28
},
{
"epoch": 0.028427888739125107,
"grad_norm": 3.7175097465515137,
"learning_rate": 4.516129032258065e-06,
"loss": 4.535539150238037,
"step": 29
},
{
"epoch": 0.02940816076461218,
"grad_norm": 3.74261474609375,
"learning_rate": 4.67741935483871e-06,
"loss": 4.304415702819824,
"step": 30
},
{
"epoch": 0.030388432790099253,
"grad_norm": 4.555102825164795,
"learning_rate": 4.838709677419355e-06,
"loss": 3.901155948638916,
"step": 31
},
{
"epoch": 0.031368704815586325,
"grad_norm": 2.8209569454193115,
"learning_rate": 5e-06,
"loss": 3.726323127746582,
"step": 32
},
{
"epoch": 0.0323489768410734,
"grad_norm": 7.080472469329834,
"learning_rate": 5.161290322580646e-06,
"loss": 4.056514739990234,
"step": 33
},
{
"epoch": 0.03332924886656047,
"grad_norm": 5.781238079071045,
"learning_rate": 5.322580645161291e-06,
"loss": 3.8636646270751953,
"step": 34
},
{
"epoch": 0.03430952089204754,
"grad_norm": 2.629904270172119,
"learning_rate": 5.483870967741935e-06,
"loss": 3.6947147846221924,
"step": 35
},
{
"epoch": 0.03528979291753462,
"grad_norm": 3.0892622470855713,
"learning_rate": 5.645161290322582e-06,
"loss": 3.5093894004821777,
"step": 36
},
{
"epoch": 0.036270064943021686,
"grad_norm": 2.350789785385132,
"learning_rate": 5.806451612903226e-06,
"loss": 3.1116795539855957,
"step": 37
},
{
"epoch": 0.03725033696850876,
"grad_norm": 2.9521284103393555,
"learning_rate": 5.967741935483872e-06,
"loss": 3.526960611343384,
"step": 38
},
{
"epoch": 0.038230608993995836,
"grad_norm": 2.6192679405212402,
"learning_rate": 6.129032258064517e-06,
"loss": 3.220268487930298,
"step": 39
},
{
"epoch": 0.039210881019482904,
"grad_norm": 2.757742404937744,
"learning_rate": 6.290322580645162e-06,
"loss": 2.7428488731384277,
"step": 40
},
{
"epoch": 0.04019115304496998,
"grad_norm": 2.104900598526001,
"learning_rate": 6.451612903225806e-06,
"loss": 2.9762954711914062,
"step": 41
},
{
"epoch": 0.041171425070457054,
"grad_norm": 1.664154291152954,
"learning_rate": 6.612903225806452e-06,
"loss": 2.6892447471618652,
"step": 42
},
{
"epoch": 0.04215169709594412,
"grad_norm": 3.2438557147979736,
"learning_rate": 6.774193548387097e-06,
"loss": 2.8475565910339355,
"step": 43
},
{
"epoch": 0.0431319691214312,
"grad_norm": 2.034782648086548,
"learning_rate": 6.935483870967743e-06,
"loss": 2.767058849334717,
"step": 44
},
{
"epoch": 0.04411224114691827,
"grad_norm": 3.961268663406372,
"learning_rate": 7.096774193548388e-06,
"loss": 2.359992027282715,
"step": 45
},
{
"epoch": 0.04509251317240534,
"grad_norm": 2.7393815517425537,
"learning_rate": 7.258064516129033e-06,
"loss": 2.6845054626464844,
"step": 46
},
{
"epoch": 0.046072785197892416,
"grad_norm": 2.4179811477661133,
"learning_rate": 7.4193548387096784e-06,
"loss": 2.4481725692749023,
"step": 47
},
{
"epoch": 0.04705305722337949,
"grad_norm": 3.267688035964966,
"learning_rate": 7.580645161290323e-06,
"loss": 2.3821351528167725,
"step": 48
},
{
"epoch": 0.04803332924886656,
"grad_norm": 1.7597341537475586,
"learning_rate": 7.741935483870968e-06,
"loss": 2.6481516361236572,
"step": 49
},
{
"epoch": 0.049013601274353634,
"grad_norm": 3.275454044342041,
"learning_rate": 7.903225806451613e-06,
"loss": 2.4091439247131348,
"step": 50
},
{
"epoch": 0.04999387329984071,
"grad_norm": 2.407912015914917,
"learning_rate": 8.064516129032258e-06,
"loss": 2.8670058250427246,
"step": 51
},
{
"epoch": 0.05097414532532778,
"grad_norm": 7.854158401489258,
"learning_rate": 8.225806451612904e-06,
"loss": 2.0484232902526855,
"step": 52
},
{
"epoch": 0.05195441735081485,
"grad_norm": 2.7430412769317627,
"learning_rate": 8.387096774193549e-06,
"loss": 2.5054028034210205,
"step": 53
},
{
"epoch": 0.05293468937630193,
"grad_norm": 2.8946540355682373,
"learning_rate": 8.548387096774194e-06,
"loss": 2.100900650024414,
"step": 54
},
{
"epoch": 0.053914961401788995,
"grad_norm": 2.498194932937622,
"learning_rate": 8.70967741935484e-06,
"loss": 2.1596410274505615,
"step": 55
},
{
"epoch": 0.05489523342727607,
"grad_norm": 1.4977469444274902,
"learning_rate": 8.870967741935484e-06,
"loss": 1.960092544555664,
"step": 56
},
{
"epoch": 0.05587550545276314,
"grad_norm": 2.928142547607422,
"learning_rate": 9.03225806451613e-06,
"loss": 2.1279821395874023,
"step": 57
},
{
"epoch": 0.05685577747825021,
"grad_norm": 3.1068716049194336,
"learning_rate": 9.193548387096775e-06,
"loss": 2.2416481971740723,
"step": 58
},
{
"epoch": 0.05783604950373729,
"grad_norm": 1.7385265827178955,
"learning_rate": 9.35483870967742e-06,
"loss": 1.8707771301269531,
"step": 59
},
{
"epoch": 0.05881632152922436,
"grad_norm": 2.592339038848877,
"learning_rate": 9.516129032258065e-06,
"loss": 1.6862797737121582,
"step": 60
},
{
"epoch": 0.05979659355471143,
"grad_norm": 2.848275661468506,
"learning_rate": 9.67741935483871e-06,
"loss": 2.10313081741333,
"step": 61
},
{
"epoch": 0.06077686558019851,
"grad_norm": 2.2939419746398926,
"learning_rate": 9.838709677419356e-06,
"loss": 2.1317005157470703,
"step": 62
},
{
"epoch": 0.061757137605685575,
"grad_norm": 2.7191245555877686,
"learning_rate": 1e-05,
"loss": 2.080993890762329,
"step": 63
},
{
"epoch": 0.06273740963117265,
"grad_norm": 1.6630226373672485,
"learning_rate": 9.999993706252978e-06,
"loss": 1.8356528282165527,
"step": 64
},
{
"epoch": 0.06371768165665972,
"grad_norm": 2.314754009246826,
"learning_rate": 9.999974825027756e-06,
"loss": 1.8618428707122803,
"step": 65
},
{
"epoch": 0.0646979536821468,
"grad_norm": 1.8380980491638184,
"learning_rate": 9.999943356371867e-06,
"loss": 1.679537057876587,
"step": 66
},
{
"epoch": 0.06567822570763387,
"grad_norm": 1.429089903831482,
"learning_rate": 9.999899300364534e-06,
"loss": 1.4057856798171997,
"step": 67
},
{
"epoch": 0.06665849773312094,
"grad_norm": 1.2200080156326294,
"learning_rate": 9.999842657116667e-06,
"loss": 1.5852755308151245,
"step": 68
},
{
"epoch": 0.06763876975860801,
"grad_norm": 1.3480346202850342,
"learning_rate": 9.999773426770864e-06,
"loss": 1.5105476379394531,
"step": 69
},
{
"epoch": 0.06861904178409509,
"grad_norm": 1.778369426727295,
"learning_rate": 9.999691609501418e-06,
"loss": 1.6858341693878174,
"step": 70
},
{
"epoch": 0.06959931380958216,
"grad_norm": 1.771379828453064,
"learning_rate": 9.999597205514298e-06,
"loss": 1.545698881149292,
"step": 71
},
{
"epoch": 0.07057958583506924,
"grad_norm": 1.4503899812698364,
"learning_rate": 9.999490215047167e-06,
"loss": 1.4640675783157349,
"step": 72
},
{
"epoch": 0.07155985786055631,
"grad_norm": 1.4041215181350708,
"learning_rate": 9.999370638369377e-06,
"loss": 1.7454248666763306,
"step": 73
},
{
"epoch": 0.07254012988604337,
"grad_norm": 1.3449410200119019,
"learning_rate": 9.999238475781957e-06,
"loss": 1.6297705173492432,
"step": 74
},
{
"epoch": 0.07352040191153045,
"grad_norm": 1.683989405632019,
"learning_rate": 9.99909372761763e-06,
"loss": 1.419502854347229,
"step": 75
},
{
"epoch": 0.07450067393701752,
"grad_norm": 1.4980436563491821,
"learning_rate": 9.998936394240798e-06,
"loss": 1.7188737392425537,
"step": 76
},
{
"epoch": 0.0754809459625046,
"grad_norm": 1.3428096771240234,
"learning_rate": 9.998766476047546e-06,
"loss": 1.3533999919891357,
"step": 77
},
{
"epoch": 0.07646121798799167,
"grad_norm": 1.919137716293335,
"learning_rate": 9.998583973465647e-06,
"loss": 1.4043164253234863,
"step": 78
},
{
"epoch": 0.07744149001347873,
"grad_norm": 2.547887086868286,
"learning_rate": 9.998388886954546e-06,
"loss": 1.3833211660385132,
"step": 79
},
{
"epoch": 0.07842176203896581,
"grad_norm": 2.176612615585327,
"learning_rate": 9.998181217005377e-06,
"loss": 1.4076578617095947,
"step": 80
},
{
"epoch": 0.07940203406445288,
"grad_norm": 2.519155979156494,
"learning_rate": 9.997960964140946e-06,
"loss": 1.5103759765625,
"step": 81
},
{
"epoch": 0.08038230608993996,
"grad_norm": 7.831232070922852,
"learning_rate": 9.997728128915742e-06,
"loss": 1.31194007396698,
"step": 82
},
{
"epoch": 0.08136257811542703,
"grad_norm": 1.2860954999923706,
"learning_rate": 9.997482711915926e-06,
"loss": 1.396899938583374,
"step": 83
},
{
"epoch": 0.08234285014091411,
"grad_norm": 1.4675216674804688,
"learning_rate": 9.997224713759335e-06,
"loss": 1.6464993953704834,
"step": 84
},
{
"epoch": 0.08332312216640117,
"grad_norm": 2.369067668914795,
"learning_rate": 9.99695413509548e-06,
"loss": 1.4915156364440918,
"step": 85
},
{
"epoch": 0.08430339419188825,
"grad_norm": 4.398408889770508,
"learning_rate": 9.99667097660554e-06,
"loss": 1.3953996896743774,
"step": 86
},
{
"epoch": 0.08528366621737532,
"grad_norm": 2.028582811355591,
"learning_rate": 9.99637523900237e-06,
"loss": 1.5002321004867554,
"step": 87
},
{
"epoch": 0.0862639382428624,
"grad_norm": 2.0428531169891357,
"learning_rate": 9.996066923030484e-06,
"loss": 1.279488205909729,
"step": 88
},
{
"epoch": 0.08724421026834947,
"grad_norm": 1.6132436990737915,
"learning_rate": 9.995746029466071e-06,
"loss": 1.2155799865722656,
"step": 89
},
{
"epoch": 0.08822448229383655,
"grad_norm": 1.4193308353424072,
"learning_rate": 9.995412559116979e-06,
"loss": 1.148223638534546,
"step": 90
},
{
"epoch": 0.0892047543193236,
"grad_norm": 2.1326708793640137,
"learning_rate": 9.99506651282272e-06,
"loss": 1.2982627153396606,
"step": 91
},
{
"epoch": 0.09018502634481068,
"grad_norm": 4.710076332092285,
"learning_rate": 9.994707891454466e-06,
"loss": 1.173756718635559,
"step": 92
},
{
"epoch": 0.09116529837029776,
"grad_norm": 1.604074239730835,
"learning_rate": 9.994336695915041e-06,
"loss": 1.3507356643676758,
"step": 93
},
{
"epoch": 0.09214557039578483,
"grad_norm": 3.051056385040283,
"learning_rate": 9.993952927138933e-06,
"loss": 1.3666877746582031,
"step": 94
},
{
"epoch": 0.0931258424212719,
"grad_norm": 1.6368945837020874,
"learning_rate": 9.993556586092281e-06,
"loss": 1.1468019485473633,
"step": 95
},
{
"epoch": 0.09410611444675898,
"grad_norm": 1.2669826745986938,
"learning_rate": 9.993147673772869e-06,
"loss": 1.390321969985962,
"step": 96
},
{
"epoch": 0.09508638647224604,
"grad_norm": 1.360478401184082,
"learning_rate": 9.992726191210139e-06,
"loss": 1.1032416820526123,
"step": 97
},
{
"epoch": 0.09606665849773312,
"grad_norm": 1.1637705564498901,
"learning_rate": 9.992292139465166e-06,
"loss": 1.0845814943313599,
"step": 98
},
{
"epoch": 0.09704693052322019,
"grad_norm": 1.9289820194244385,
"learning_rate": 9.991845519630679e-06,
"loss": 1.3338316679000854,
"step": 99
},
{
"epoch": 0.09802720254870727,
"grad_norm": 1.466807246208191,
"learning_rate": 9.991386332831042e-06,
"loss": 1.1557691097259521,
"step": 100
},
{
"epoch": 0.09900747457419434,
"grad_norm": 1.2937833070755005,
"learning_rate": 9.990914580222258e-06,
"loss": 1.0668269395828247,
"step": 101
},
{
"epoch": 0.09998774659968142,
"grad_norm": 2.05177903175354,
"learning_rate": 9.990430262991962e-06,
"loss": 1.3129385709762573,
"step": 102
},
{
"epoch": 0.10096801862516848,
"grad_norm": 1.105375051498413,
"learning_rate": 9.989933382359423e-06,
"loss": 0.9598901271820068,
"step": 103
},
{
"epoch": 0.10194829065065555,
"grad_norm": 1.6122556924819946,
"learning_rate": 9.989423939575538e-06,
"loss": 1.4148848056793213,
"step": 104
},
{
"epoch": 0.10292856267614263,
"grad_norm": 1.4959112405776978,
"learning_rate": 9.988901935922826e-06,
"loss": 0.9532449841499329,
"step": 105
},
{
"epoch": 0.1039088347016297,
"grad_norm": 6.660800457000732,
"learning_rate": 9.988367372715434e-06,
"loss": 1.2004680633544922,
"step": 106
},
{
"epoch": 0.10488910672711678,
"grad_norm": 1.4263871908187866,
"learning_rate": 9.987820251299121e-06,
"loss": 0.993876576423645,
"step": 107
},
{
"epoch": 0.10586937875260385,
"grad_norm": 1.862610936164856,
"learning_rate": 9.987260573051268e-06,
"loss": 1.169829249382019,
"step": 108
},
{
"epoch": 0.10684965077809092,
"grad_norm": 2.746614456176758,
"learning_rate": 9.986688339380863e-06,
"loss": 0.9846648573875427,
"step": 109
},
{
"epoch": 0.10782992280357799,
"grad_norm": 1.68449866771698,
"learning_rate": 9.9861035517285e-06,
"loss": 1.0470050573349,
"step": 110
},
{
"epoch": 0.10881019482906507,
"grad_norm": 1.4181978702545166,
"learning_rate": 9.985506211566388e-06,
"loss": 1.0499745607376099,
"step": 111
},
{
"epoch": 0.10979046685455214,
"grad_norm": 1.6402755975723267,
"learning_rate": 9.984896320398326e-06,
"loss": 1.083099365234375,
"step": 112
},
{
"epoch": 0.11077073888003922,
"grad_norm": 1.070939064025879,
"learning_rate": 9.984273879759713e-06,
"loss": 0.8868157863616943,
"step": 113
},
{
"epoch": 0.11175101090552628,
"grad_norm": 1.2195574045181274,
"learning_rate": 9.983638891217544e-06,
"loss": 1.2996704578399658,
"step": 114
},
{
"epoch": 0.11273128293101335,
"grad_norm": 2.2340879440307617,
"learning_rate": 9.982991356370404e-06,
"loss": 1.1774934530258179,
"step": 115
},
{
"epoch": 0.11371155495650043,
"grad_norm": 1.3596181869506836,
"learning_rate": 9.982331276848457e-06,
"loss": 1.3012547492980957,
"step": 116
},
{
"epoch": 0.1146918269819875,
"grad_norm": 1.1323665380477905,
"learning_rate": 9.981658654313458e-06,
"loss": 0.863355278968811,
"step": 117
},
{
"epoch": 0.11567209900747458,
"grad_norm": 1.1545711755752563,
"learning_rate": 9.980973490458728e-06,
"loss": 1.004765510559082,
"step": 118
},
{
"epoch": 0.11665237103296165,
"grad_norm": 1.2341439723968506,
"learning_rate": 9.98027578700917e-06,
"loss": 1.063584327697754,
"step": 119
},
{
"epoch": 0.11763264305844871,
"grad_norm": 1.0848900079727173,
"learning_rate": 9.979565545721249e-06,
"loss": 0.9332760572433472,
"step": 120
},
{
"epoch": 0.11861291508393579,
"grad_norm": 1.110190510749817,
"learning_rate": 9.978842768382999e-06,
"loss": 1.0000088214874268,
"step": 121
},
{
"epoch": 0.11959318710942286,
"grad_norm": 1.5142709016799927,
"learning_rate": 9.978107456814007e-06,
"loss": 1.0536956787109375,
"step": 122
},
{
"epoch": 0.12057345913490994,
"grad_norm": 1.5360547304153442,
"learning_rate": 9.977359612865424e-06,
"loss": 1.0528912544250488,
"step": 123
},
{
"epoch": 0.12155373116039701,
"grad_norm": 2.2734928131103516,
"learning_rate": 9.976599238419943e-06,
"loss": 1.0330657958984375,
"step": 124
},
{
"epoch": 0.12253400318588409,
"grad_norm": 1.6001646518707275,
"learning_rate": 9.975826335391808e-06,
"loss": 1.391056776046753,
"step": 125
},
{
"epoch": 0.12351427521137115,
"grad_norm": 1.7746310234069824,
"learning_rate": 9.975040905726799e-06,
"loss": 0.9952625036239624,
"step": 126
},
{
"epoch": 0.12449454723685822,
"grad_norm": 1.1504384279251099,
"learning_rate": 9.974242951402236e-06,
"loss": 0.9756612777709961,
"step": 127
},
{
"epoch": 0.1254748192623453,
"grad_norm": 6.513915538787842,
"learning_rate": 9.973432474426968e-06,
"loss": 0.9526469111442566,
"step": 128
},
{
"epoch": 0.12645509128783236,
"grad_norm": 2.440303087234497,
"learning_rate": 9.972609476841368e-06,
"loss": 1.037710189819336,
"step": 129
},
{
"epoch": 0.12743536331331945,
"grad_norm": 2.486783504486084,
"learning_rate": 9.971773960717333e-06,
"loss": 0.8668553233146667,
"step": 130
},
{
"epoch": 0.1284156353388065,
"grad_norm": 1.0355393886566162,
"learning_rate": 9.970925928158275e-06,
"loss": 0.9324941635131836,
"step": 131
},
{
"epoch": 0.1293959073642936,
"grad_norm": 4.5477824211120605,
"learning_rate": 9.970065381299112e-06,
"loss": 0.8806881904602051,
"step": 132
},
{
"epoch": 0.13037617938978066,
"grad_norm": 1.1980022192001343,
"learning_rate": 9.969192322306271e-06,
"loss": 0.9554859399795532,
"step": 133
},
{
"epoch": 0.13135645141526775,
"grad_norm": 0.868634045124054,
"learning_rate": 9.968306753377679e-06,
"loss": 0.9407200813293457,
"step": 134
},
{
"epoch": 0.1323367234407548,
"grad_norm": 1.3445113897323608,
"learning_rate": 9.96740867674275e-06,
"loss": 1.195878505706787,
"step": 135
},
{
"epoch": 0.13331699546624187,
"grad_norm": 1.7127512693405151,
"learning_rate": 9.966498094662398e-06,
"loss": 0.9498565196990967,
"step": 136
},
{
"epoch": 0.13429726749172896,
"grad_norm": 1.2315754890441895,
"learning_rate": 9.965575009429006e-06,
"loss": 0.8030359745025635,
"step": 137
},
{
"epoch": 0.13527753951721602,
"grad_norm": 1.4911632537841797,
"learning_rate": 9.964639423366442e-06,
"loss": 0.8708802461624146,
"step": 138
},
{
"epoch": 0.1362578115427031,
"grad_norm": 1.2748379707336426,
"learning_rate": 9.963691338830045e-06,
"loss": 0.9382426738739014,
"step": 139
},
{
"epoch": 0.13723808356819017,
"grad_norm": 1.0648022890090942,
"learning_rate": 9.962730758206612e-06,
"loss": 0.9130150079727173,
"step": 140
},
{
"epoch": 0.13821835559367723,
"grad_norm": 6.634443283081055,
"learning_rate": 9.961757683914406e-06,
"loss": 1.0724031925201416,
"step": 141
},
{
"epoch": 0.13919862761916432,
"grad_norm": 1.1606115102767944,
"learning_rate": 9.960772118403142e-06,
"loss": 0.9133974313735962,
"step": 142
},
{
"epoch": 0.14017889964465138,
"grad_norm": 1.0151580572128296,
"learning_rate": 9.959774064153977e-06,
"loss": 0.7543078660964966,
"step": 143
},
{
"epoch": 0.14115917167013847,
"grad_norm": 10.965326309204102,
"learning_rate": 9.958763523679515e-06,
"loss": 1.021949291229248,
"step": 144
},
{
"epoch": 0.14213944369562553,
"grad_norm": 1.297864317893982,
"learning_rate": 9.957740499523787e-06,
"loss": 0.8699567317962646,
"step": 145
},
{
"epoch": 0.14311971572111262,
"grad_norm": 2.5568370819091797,
"learning_rate": 9.956704994262256e-06,
"loss": 1.032677412033081,
"step": 146
},
{
"epoch": 0.14409998774659968,
"grad_norm": 1.0308430194854736,
"learning_rate": 9.955657010501807e-06,
"loss": 0.9547523260116577,
"step": 147
},
{
"epoch": 0.14508025977208674,
"grad_norm": 1.0656805038452148,
"learning_rate": 9.954596550880735e-06,
"loss": 0.8700910806655884,
"step": 148
},
{
"epoch": 0.14606053179757383,
"grad_norm": 0.9584664702415466,
"learning_rate": 9.95352361806875e-06,
"loss": 0.6975476741790771,
"step": 149
},
{
"epoch": 0.1470408038230609,
"grad_norm": 1.381972074508667,
"learning_rate": 9.952438214766956e-06,
"loss": 0.901016354560852,
"step": 150
},
{
"epoch": 0.14802107584854798,
"grad_norm": 5.689754009246826,
"learning_rate": 9.951340343707852e-06,
"loss": 1.059537649154663,
"step": 151
},
{
"epoch": 0.14900134787403504,
"grad_norm": 1.0822038650512695,
"learning_rate": 9.950230007655333e-06,
"loss": 0.6861962080001831,
"step": 152
},
{
"epoch": 0.1499816198995221,
"grad_norm": 1.072089672088623,
"learning_rate": 9.949107209404664e-06,
"loss": 1.0077919960021973,
"step": 153
},
{
"epoch": 0.1509618919250092,
"grad_norm": 1.4591175317764282,
"learning_rate": 9.94797195178249e-06,
"loss": 1.0163085460662842,
"step": 154
},
{
"epoch": 0.15194216395049626,
"grad_norm": 1.2471067905426025,
"learning_rate": 9.946824237646823e-06,
"loss": 1.0169732570648193,
"step": 155
},
{
"epoch": 0.15292243597598335,
"grad_norm": 1.0896052122116089,
"learning_rate": 9.945664069887028e-06,
"loss": 1.0910099744796753,
"step": 156
},
{
"epoch": 0.1539027080014704,
"grad_norm": 1.0520861148834229,
"learning_rate": 9.944491451423829e-06,
"loss": 0.80839604139328,
"step": 157
},
{
"epoch": 0.15488298002695747,
"grad_norm": 1.1490066051483154,
"learning_rate": 9.94330638520929e-06,
"loss": 0.9133569598197937,
"step": 158
},
{
"epoch": 0.15586325205244456,
"grad_norm": 1.1544691324234009,
"learning_rate": 9.942108874226812e-06,
"loss": 0.8751540184020996,
"step": 159
},
{
"epoch": 0.15684352407793162,
"grad_norm": 0.9796386957168579,
"learning_rate": 9.940898921491132e-06,
"loss": 0.7868462800979614,
"step": 160
},
{
"epoch": 0.1578237961034187,
"grad_norm": 0.879181981086731,
"learning_rate": 9.9396765300483e-06,
"loss": 0.785386860370636,
"step": 161
},
{
"epoch": 0.15880406812890577,
"grad_norm": 1.499031662940979,
"learning_rate": 9.938441702975689e-06,
"loss": 0.9795531034469604,
"step": 162
},
{
"epoch": 0.15978434015439286,
"grad_norm": 1.3540782928466797,
"learning_rate": 9.937194443381972e-06,
"loss": 1.0751852989196777,
"step": 163
},
{
"epoch": 0.16076461217987992,
"grad_norm": 1.082590103149414,
"learning_rate": 9.935934754407125e-06,
"loss": 0.8491311073303223,
"step": 164
},
{
"epoch": 0.16174488420536698,
"grad_norm": 4.6183247566223145,
"learning_rate": 9.934662639222412e-06,
"loss": 0.8475311398506165,
"step": 165
},
{
"epoch": 0.16272515623085407,
"grad_norm": 2.0616343021392822,
"learning_rate": 9.933378101030382e-06,
"loss": 0.7665926814079285,
"step": 166
},
{
"epoch": 0.16370542825634113,
"grad_norm": 1.2640901803970337,
"learning_rate": 9.93208114306486e-06,
"loss": 0.8191386461257935,
"step": 167
},
{
"epoch": 0.16468570028182822,
"grad_norm": 1.4200091361999512,
"learning_rate": 9.930771768590934e-06,
"loss": 0.8942907452583313,
"step": 168
},
{
"epoch": 0.16566597230731528,
"grad_norm": 0.8642187118530273,
"learning_rate": 9.929449980904952e-06,
"loss": 0.7328740954399109,
"step": 169
},
{
"epoch": 0.16664624433280234,
"grad_norm": 1.0739811658859253,
"learning_rate": 9.928115783334518e-06,
"loss": 0.9266120195388794,
"step": 170
},
{
"epoch": 0.16762651635828943,
"grad_norm": 0.8943596482276917,
"learning_rate": 9.926769179238467e-06,
"loss": 0.7814446687698364,
"step": 171
},
{
"epoch": 0.1686067883837765,
"grad_norm": 1.433998942375183,
"learning_rate": 9.925410172006874e-06,
"loss": 0.9688506126403809,
"step": 172
},
{
"epoch": 0.16958706040926358,
"grad_norm": 0.9713033437728882,
"learning_rate": 9.924038765061042e-06,
"loss": 0.9370806217193604,
"step": 173
},
{
"epoch": 0.17056733243475064,
"grad_norm": 5.718565940856934,
"learning_rate": 9.922654961853483e-06,
"loss": 1.002172827720642,
"step": 174
},
{
"epoch": 0.17154760446023773,
"grad_norm": 1.657820224761963,
"learning_rate": 9.921258765867919e-06,
"loss": 1.0920586585998535,
"step": 175
},
{
"epoch": 0.1725278764857248,
"grad_norm": 1.479437232017517,
"learning_rate": 9.919850180619274e-06,
"loss": 0.8967446088790894,
"step": 176
},
{
"epoch": 0.17350814851121185,
"grad_norm": 1.1827943325042725,
"learning_rate": 9.918429209653662e-06,
"loss": 0.7842462658882141,
"step": 177
},
{
"epoch": 0.17448842053669894,
"grad_norm": 0.986243486404419,
"learning_rate": 9.916995856548371e-06,
"loss": 0.9605097770690918,
"step": 178
},
{
"epoch": 0.175468692562186,
"grad_norm": 0.8720986247062683,
"learning_rate": 9.915550124911866e-06,
"loss": 0.9053399562835693,
"step": 179
},
{
"epoch": 0.1764489645876731,
"grad_norm": 1.7377837896347046,
"learning_rate": 9.914092018383778e-06,
"loss": 0.7633567452430725,
"step": 180
},
{
"epoch": 0.17742923661316015,
"grad_norm": 3.610898494720459,
"learning_rate": 9.912621540634889e-06,
"loss": 0.8012740612030029,
"step": 181
},
{
"epoch": 0.1784095086386472,
"grad_norm": 1.691759467124939,
"learning_rate": 9.91113869536712e-06,
"loss": 0.7178599238395691,
"step": 182
},
{
"epoch": 0.1793897806641343,
"grad_norm": 1.3078618049621582,
"learning_rate": 9.909643486313533e-06,
"loss": 0.8127522468566895,
"step": 183
},
{
"epoch": 0.18037005268962136,
"grad_norm": 6.31986665725708,
"learning_rate": 9.908135917238321e-06,
"loss": 1.0647804737091064,
"step": 184
},
{
"epoch": 0.18135032471510845,
"grad_norm": 1.194877028465271,
"learning_rate": 9.906615991936781e-06,
"loss": 0.7892481088638306,
"step": 185
},
{
"epoch": 0.1823305967405955,
"grad_norm": 1.8201836347579956,
"learning_rate": 9.905083714235326e-06,
"loss": 0.8923194408416748,
"step": 186
},
{
"epoch": 0.18331086876608257,
"grad_norm": 4.416145324707031,
"learning_rate": 9.903539087991462e-06,
"loss": 0.9546653032302856,
"step": 187
},
{
"epoch": 0.18429114079156966,
"grad_norm": 1.0860720872879028,
"learning_rate": 9.901982117093786e-06,
"loss": 0.9052340984344482,
"step": 188
},
{
"epoch": 0.18527141281705672,
"grad_norm": 0.8239496350288391,
"learning_rate": 9.900412805461968e-06,
"loss": 0.6534804105758667,
"step": 189
},
{
"epoch": 0.1862516848425438,
"grad_norm": 1.3619135618209839,
"learning_rate": 9.898831157046748e-06,
"loss": 0.7209652662277222,
"step": 190
},
{
"epoch": 0.18723195686803087,
"grad_norm": 1.0056313276290894,
"learning_rate": 9.897237175829927e-06,
"loss": 0.8055800795555115,
"step": 191
},
{
"epoch": 0.18821222889351796,
"grad_norm": 1.5930757522583008,
"learning_rate": 9.895630865824347e-06,
"loss": 0.7460575103759766,
"step": 192
},
{
"epoch": 0.18919250091900502,
"grad_norm": 1.4717005491256714,
"learning_rate": 9.894012231073895e-06,
"loss": 0.8408622741699219,
"step": 193
},
{
"epoch": 0.19017277294449209,
"grad_norm": 0.9883076548576355,
"learning_rate": 9.89238127565348e-06,
"loss": 0.7255659103393555,
"step": 194
},
{
"epoch": 0.19115304496997917,
"grad_norm": 0.9901779890060425,
"learning_rate": 9.890738003669029e-06,
"loss": 0.9237308502197266,
"step": 195
},
{
"epoch": 0.19213331699546624,
"grad_norm": 0.842984139919281,
"learning_rate": 9.88908241925748e-06,
"loss": 0.8075700998306274,
"step": 196
},
{
"epoch": 0.19311358902095332,
"grad_norm": 0.7866753339767456,
"learning_rate": 9.887414526586764e-06,
"loss": 0.855652928352356,
"step": 197
},
{
"epoch": 0.19409386104644039,
"grad_norm": 0.9305774569511414,
"learning_rate": 9.885734329855798e-06,
"loss": 0.7251928448677063,
"step": 198
},
{
"epoch": 0.19507413307192745,
"grad_norm": 0.792121946811676,
"learning_rate": 9.884041833294477e-06,
"loss": 0.8123219013214111,
"step": 199
},
{
"epoch": 0.19605440509741454,
"grad_norm": 0.9794050455093384,
"learning_rate": 9.882337041163656e-06,
"loss": 0.8199892044067383,
"step": 200
},
{
"epoch": 0.1970346771229016,
"grad_norm": 0.8961284160614014,
"learning_rate": 9.880619957755151e-06,
"loss": 0.8511288166046143,
"step": 201
},
{
"epoch": 0.19801494914838869,
"grad_norm": 0.8756808042526245,
"learning_rate": 9.878890587391715e-06,
"loss": 0.7968616485595703,
"step": 202
},
{
"epoch": 0.19899522117387575,
"grad_norm": 1.38584303855896,
"learning_rate": 9.877148934427037e-06,
"loss": 0.7082912921905518,
"step": 203
},
{
"epoch": 0.19997549319936284,
"grad_norm": 0.9178659319877625,
"learning_rate": 9.875395003245724e-06,
"loss": 0.6546875238418579,
"step": 204
},
{
"epoch": 0.2009557652248499,
"grad_norm": 1.5947273969650269,
"learning_rate": 9.873628798263297e-06,
"loss": 1.0217643976211548,
"step": 205
},
{
"epoch": 0.20193603725033696,
"grad_norm": 1.4036369323730469,
"learning_rate": 9.871850323926178e-06,
"loss": 0.7721039056777954,
"step": 206
},
{
"epoch": 0.20291630927582405,
"grad_norm": 1.074377417564392,
"learning_rate": 9.870059584711668e-06,
"loss": 0.7908469438552856,
"step": 207
},
{
"epoch": 0.2038965813013111,
"grad_norm": 4.75644588470459,
"learning_rate": 9.868256585127956e-06,
"loss": 0.7792013883590698,
"step": 208
},
{
"epoch": 0.2048768533267982,
"grad_norm": 1.1384018659591675,
"learning_rate": 9.86644132971409e-06,
"loss": 0.7944319844245911,
"step": 209
},
{
"epoch": 0.20585712535228526,
"grad_norm": 0.8815261721611023,
"learning_rate": 9.864613823039971e-06,
"loss": 0.8812296390533447,
"step": 210
},
{
"epoch": 0.20683739737777232,
"grad_norm": 0.9276820421218872,
"learning_rate": 9.862774069706346e-06,
"loss": 0.9711546897888184,
"step": 211
},
{
"epoch": 0.2078176694032594,
"grad_norm": 1.045776605606079,
"learning_rate": 9.860922074344794e-06,
"loss": 0.804419994354248,
"step": 212
},
{
"epoch": 0.20879794142874647,
"grad_norm": 1.400531530380249,
"learning_rate": 9.859057841617709e-06,
"loss": 0.7008233070373535,
"step": 213
},
{
"epoch": 0.20977821345423356,
"grad_norm": 2.3324949741363525,
"learning_rate": 9.857181376218295e-06,
"loss": 0.7228943109512329,
"step": 214
},
{
"epoch": 0.21075848547972062,
"grad_norm": 1.052319049835205,
"learning_rate": 9.855292682870552e-06,
"loss": 0.8040943741798401,
"step": 215
},
{
"epoch": 0.2117387575052077,
"grad_norm": 1.0283721685409546,
"learning_rate": 9.853391766329264e-06,
"loss": 0.7747700214385986,
"step": 216
},
{
"epoch": 0.21271902953069477,
"grad_norm": 0.8995811343193054,
"learning_rate": 9.851478631379982e-06,
"loss": 0.723522424697876,
"step": 217
},
{
"epoch": 0.21369930155618183,
"grad_norm": 0.8091436624526978,
"learning_rate": 9.849553282839025e-06,
"loss": 0.7437203526496887,
"step": 218
},
{
"epoch": 0.21467957358166892,
"grad_norm": 1.2023676633834839,
"learning_rate": 9.847615725553457e-06,
"loss": 0.7519738674163818,
"step": 219
},
{
"epoch": 0.21565984560715598,
"grad_norm": 2.7624218463897705,
"learning_rate": 9.845665964401072e-06,
"loss": 0.7127557992935181,
"step": 220
},
{
"epoch": 0.21664011763264307,
"grad_norm": 1.1160601377487183,
"learning_rate": 9.843704004290393e-06,
"loss": 0.7732646465301514,
"step": 221
},
{
"epoch": 0.21762038965813013,
"grad_norm": 1.6516051292419434,
"learning_rate": 9.841729850160652e-06,
"loss": 0.7416677474975586,
"step": 222
},
{
"epoch": 0.2186006616836172,
"grad_norm": 1.0229910612106323,
"learning_rate": 9.839743506981783e-06,
"loss": 0.797340989112854,
"step": 223
},
{
"epoch": 0.21958093370910428,
"grad_norm": 4.91952657699585,
"learning_rate": 9.837744979754397e-06,
"loss": 0.6655223965644836,
"step": 224
},
{
"epoch": 0.22056120573459134,
"grad_norm": 1.1878681182861328,
"learning_rate": 9.835734273509787e-06,
"loss": 0.7352027893066406,
"step": 225
},
{
"epoch": 0.22154147776007843,
"grad_norm": 0.9683135747909546,
"learning_rate": 9.833711393309902e-06,
"loss": 0.916510820388794,
"step": 226
},
{
"epoch": 0.2225217497855655,
"grad_norm": 0.7483813166618347,
"learning_rate": 9.831676344247343e-06,
"loss": 0.8279455900192261,
"step": 227
},
{
"epoch": 0.22350202181105255,
"grad_norm": 0.7321900129318237,
"learning_rate": 9.829629131445342e-06,
"loss": 0.6867122650146484,
"step": 228
},
{
"epoch": 0.22448229383653964,
"grad_norm": 0.8370916247367859,
"learning_rate": 9.827569760057755e-06,
"loss": 0.5806996822357178,
"step": 229
},
{
"epoch": 0.2254625658620267,
"grad_norm": 1.1479220390319824,
"learning_rate": 9.825498235269046e-06,
"loss": 0.82973313331604,
"step": 230
},
{
"epoch": 0.2264428378875138,
"grad_norm": 0.7834895253181458,
"learning_rate": 9.82341456229428e-06,
"loss": 0.7186822891235352,
"step": 231
},
{
"epoch": 0.22742310991300085,
"grad_norm": 0.789063572883606,
"learning_rate": 9.821318746379097e-06,
"loss": 0.8105742931365967,
"step": 232
},
{
"epoch": 0.22840338193848794,
"grad_norm": 0.9181479215621948,
"learning_rate": 9.819210792799711e-06,
"loss": 0.7336741089820862,
"step": 233
},
{
"epoch": 0.229383653963975,
"grad_norm": 0.6511960625648499,
"learning_rate": 9.817090706862895e-06,
"loss": 0.6957347393035889,
"step": 234
},
{
"epoch": 0.23036392598946207,
"grad_norm": 0.949479341506958,
"learning_rate": 9.814958493905962e-06,
"loss": 0.8622881174087524,
"step": 235
},
{
"epoch": 0.23134419801494915,
"grad_norm": 1.2978613376617432,
"learning_rate": 9.812814159296757e-06,
"loss": 0.801636278629303,
"step": 236
},
{
"epoch": 0.23232447004043622,
"grad_norm": 1.334235429763794,
"learning_rate": 9.810657708433637e-06,
"loss": 0.8749450445175171,
"step": 237
},
{
"epoch": 0.2333047420659233,
"grad_norm": 0.8588714599609375,
"learning_rate": 9.808489146745466e-06,
"loss": 0.7181082367897034,
"step": 238
},
{
"epoch": 0.23428501409141037,
"grad_norm": 0.8865885138511658,
"learning_rate": 9.806308479691595e-06,
"loss": 0.6831940412521362,
"step": 239
},
{
"epoch": 0.23526528611689743,
"grad_norm": 1.6122536659240723,
"learning_rate": 9.804115712761851e-06,
"loss": 0.9708364009857178,
"step": 240
},
{
"epoch": 0.23624555814238452,
"grad_norm": 0.9729068875312805,
"learning_rate": 9.801910851476524e-06,
"loss": 0.8349888324737549,
"step": 241
},
{
"epoch": 0.23722583016787158,
"grad_norm": 0.9253604412078857,
"learning_rate": 9.799693901386346e-06,
"loss": 0.9101108312606812,
"step": 242
},
{
"epoch": 0.23820610219335867,
"grad_norm": 0.9011872410774231,
"learning_rate": 9.797464868072489e-06,
"loss": 0.6735062599182129,
"step": 243
},
{
"epoch": 0.23918637421884573,
"grad_norm": 0.8829543590545654,
"learning_rate": 9.79522375714654e-06,
"loss": 0.693250298500061,
"step": 244
},
{
"epoch": 0.24016664624433282,
"grad_norm": 1.0875765085220337,
"learning_rate": 9.792970574250493e-06,
"loss": 0.8324264287948608,
"step": 245
},
{
"epoch": 0.24114691826981988,
"grad_norm": 1.0755997896194458,
"learning_rate": 9.790705325056735e-06,
"loss": 0.7184022665023804,
"step": 246
},
{
"epoch": 0.24212719029530694,
"grad_norm": 0.7354425191879272,
"learning_rate": 9.788428015268027e-06,
"loss": 0.6558845043182373,
"step": 247
},
{
"epoch": 0.24310746232079403,
"grad_norm": 1.9448564052581787,
"learning_rate": 9.786138650617494e-06,
"loss": 0.8444551229476929,
"step": 248
},
{
"epoch": 0.2440877343462811,
"grad_norm": 1.344329833984375,
"learning_rate": 9.78383723686861e-06,
"loss": 0.6076864004135132,
"step": 249
},
{
"epoch": 0.24506800637176818,
"grad_norm": 17.73468017578125,
"learning_rate": 9.781523779815178e-06,
"loss": 0.6780315637588501,
"step": 250
},
{
"epoch": 0.24604827839725524,
"grad_norm": 1.0934889316558838,
"learning_rate": 9.779198285281326e-06,
"loss": 0.7192176580429077,
"step": 251
},
{
"epoch": 0.2470285504227423,
"grad_norm": 1.0134140253067017,
"learning_rate": 9.776860759121485e-06,
"loss": 0.8143395185470581,
"step": 252
},
{
"epoch": 0.2480088224482294,
"grad_norm": 0.7228949666023254,
"learning_rate": 9.774511207220369e-06,
"loss": 0.7440932989120483,
"step": 253
},
{
"epoch": 0.24898909447371645,
"grad_norm": 0.8832446932792664,
"learning_rate": 9.772149635492979e-06,
"loss": 0.6749849319458008,
"step": 254
},
{
"epoch": 0.24996936649920354,
"grad_norm": 1.1411173343658447,
"learning_rate": 9.769776049884564e-06,
"loss": 0.7055342197418213,
"step": 255
},
{
"epoch": 0.2509496385246906,
"grad_norm": 1.2402961254119873,
"learning_rate": 9.767390456370625e-06,
"loss": 1.0037171840667725,
"step": 256
},
{
"epoch": 0.25192991055017766,
"grad_norm": 0.9168121218681335,
"learning_rate": 9.76499286095689e-06,
"loss": 0.749294638633728,
"step": 257
},
{
"epoch": 0.2529101825756647,
"grad_norm": 1.9439083337783813,
"learning_rate": 9.762583269679304e-06,
"loss": 0.6603206396102905,
"step": 258
},
{
"epoch": 0.25389045460115184,
"grad_norm": 1.0697510242462158,
"learning_rate": 9.760161688604008e-06,
"loss": 0.8273264765739441,
"step": 259
},
{
"epoch": 0.2548707266266389,
"grad_norm": 0.7438647150993347,
"learning_rate": 9.75772812382733e-06,
"loss": 0.6545376777648926,
"step": 260
},
{
"epoch": 0.25585099865212596,
"grad_norm": 1.1132228374481201,
"learning_rate": 9.755282581475769e-06,
"loss": 0.7165011167526245,
"step": 261
},
{
"epoch": 0.256831270677613,
"grad_norm": 0.6882447004318237,
"learning_rate": 9.75282506770597e-06,
"loss": 0.6577140092849731,
"step": 262
},
{
"epoch": 0.2578115427031001,
"grad_norm": 4.1547088623046875,
"learning_rate": 9.750355588704728e-06,
"loss": 0.6683706045150757,
"step": 263
},
{
"epoch": 0.2587918147285872,
"grad_norm": 0.6356456875801086,
"learning_rate": 9.747874150688948e-06,
"loss": 0.682666540145874,
"step": 264
},
{
"epoch": 0.25977208675407426,
"grad_norm": 0.8459199070930481,
"learning_rate": 9.745380759905648e-06,
"loss": 0.738046407699585,
"step": 265
},
{
"epoch": 0.2607523587795613,
"grad_norm": 1.2721017599105835,
"learning_rate": 9.742875422631937e-06,
"loss": 0.6385523080825806,
"step": 266
},
{
"epoch": 0.2617326308050484,
"grad_norm": 1.341133713722229,
"learning_rate": 9.740358145174999e-06,
"loss": 0.7375051975250244,
"step": 267
},
{
"epoch": 0.2627129028305355,
"grad_norm": 0.6215521693229675,
"learning_rate": 9.737828933872076e-06,
"loss": 0.6837223172187805,
"step": 268
},
{
"epoch": 0.26369317485602256,
"grad_norm": 0.8820347189903259,
"learning_rate": 9.735287795090455e-06,
"loss": 0.7571901082992554,
"step": 269
},
{
"epoch": 0.2646734468815096,
"grad_norm": 0.869671106338501,
"learning_rate": 9.73273473522745e-06,
"loss": 0.6656359434127808,
"step": 270
},
{
"epoch": 0.2656537189069967,
"grad_norm": 0.7545670866966248,
"learning_rate": 9.730169760710385e-06,
"loss": 0.7184778451919556,
"step": 271
},
{
"epoch": 0.26663399093248374,
"grad_norm": 0.7405308485031128,
"learning_rate": 9.727592877996585e-06,
"loss": 0.6783982515335083,
"step": 272
},
{
"epoch": 0.26761426295797086,
"grad_norm": 0.6823749542236328,
"learning_rate": 9.725004093573343e-06,
"loss": 0.7413240671157837,
"step": 273
},
{
"epoch": 0.2685945349834579,
"grad_norm": 1.2406173944473267,
"learning_rate": 9.722403413957924e-06,
"loss": 0.6491576433181763,
"step": 274
},
{
"epoch": 0.269574807008945,
"grad_norm": 0.6837533116340637,
"learning_rate": 9.719790845697534e-06,
"loss": 0.7499884366989136,
"step": 275
},
{
"epoch": 0.27055507903443204,
"grad_norm": 0.9879984855651855,
"learning_rate": 9.717166395369312e-06,
"loss": 0.6761235594749451,
"step": 276
},
{
"epoch": 0.2715353510599191,
"grad_norm": 0.6730490326881409,
"learning_rate": 9.71453006958031e-06,
"loss": 0.6731091141700745,
"step": 277
},
{
"epoch": 0.2725156230854062,
"grad_norm": 0.8950958847999573,
"learning_rate": 9.711881874967471e-06,
"loss": 0.5140408277511597,
"step": 278
},
{
"epoch": 0.2734958951108933,
"grad_norm": 0.7745827436447144,
"learning_rate": 9.709221818197626e-06,
"loss": 0.7368078231811523,
"step": 279
},
{
"epoch": 0.27447616713638034,
"grad_norm": 0.8786028623580933,
"learning_rate": 9.70654990596746e-06,
"loss": 0.8690959215164185,
"step": 280
},
{
"epoch": 0.2754564391618674,
"grad_norm": 0.7202008962631226,
"learning_rate": 9.703866145003512e-06,
"loss": 0.7002525925636292,
"step": 281
},
{
"epoch": 0.27643671118735447,
"grad_norm": 14.457069396972656,
"learning_rate": 9.701170542062149e-06,
"loss": 0.6674705147743225,
"step": 282
},
{
"epoch": 0.2774169832128416,
"grad_norm": 0.7733708620071411,
"learning_rate": 9.698463103929542e-06,
"loss": 0.7949401140213013,
"step": 283
},
{
"epoch": 0.27839725523832864,
"grad_norm": 1.1524943113327026,
"learning_rate": 9.69574383742167e-06,
"loss": 0.5851595401763916,
"step": 284
},
{
"epoch": 0.2793775272638157,
"grad_norm": 1.1284914016723633,
"learning_rate": 9.69301274938428e-06,
"loss": 0.7911825180053711,
"step": 285
},
{
"epoch": 0.28035779928930277,
"grad_norm": 0.8577850461006165,
"learning_rate": 9.690269846692882e-06,
"loss": 0.7429852485656738,
"step": 286
},
{
"epoch": 0.28133807131478983,
"grad_norm": 0.772161066532135,
"learning_rate": 9.687515136252732e-06,
"loss": 0.7238274812698364,
"step": 287
},
{
"epoch": 0.28231834334027694,
"grad_norm": 0.7844406366348267,
"learning_rate": 9.68474862499881e-06,
"loss": 0.6914684772491455,
"step": 288
},
{
"epoch": 0.283298615365764,
"grad_norm": 0.7209534049034119,
"learning_rate": 9.681970319895804e-06,
"loss": 0.7430475354194641,
"step": 289
},
{
"epoch": 0.28427888739125107,
"grad_norm": 1.3969320058822632,
"learning_rate": 9.679180227938095e-06,
"loss": 0.7238540649414062,
"step": 290
},
{
"epoch": 0.28525915941673813,
"grad_norm": 0.9757972359657288,
"learning_rate": 9.676378356149733e-06,
"loss": 0.76640385389328,
"step": 291
},
{
"epoch": 0.28623943144222525,
"grad_norm": 0.6902493238449097,
"learning_rate": 9.67356471158443e-06,
"loss": 0.6268141269683838,
"step": 292
},
{
"epoch": 0.2872197034677123,
"grad_norm": 0.9343335628509521,
"learning_rate": 9.670739301325534e-06,
"loss": 0.7309048771858215,
"step": 293
},
{
"epoch": 0.28819997549319937,
"grad_norm": 1.1967025995254517,
"learning_rate": 9.667902132486009e-06,
"loss": 0.8633978962898254,
"step": 294
},
{
"epoch": 0.28918024751868643,
"grad_norm": 1.1634700298309326,
"learning_rate": 9.665053212208426e-06,
"loss": 0.6298658847808838,
"step": 295
},
{
"epoch": 0.2901605195441735,
"grad_norm": 0.8434868454933167,
"learning_rate": 9.662192547664938e-06,
"loss": 0.7368426322937012,
"step": 296
},
{
"epoch": 0.2911407915696606,
"grad_norm": 1.1653025150299072,
"learning_rate": 9.659320146057263e-06,
"loss": 0.702348530292511,
"step": 297
},
{
"epoch": 0.29212106359514767,
"grad_norm": 0.6269257664680481,
"learning_rate": 9.65643601461667e-06,
"loss": 0.5639951229095459,
"step": 298
},
{
"epoch": 0.29310133562063473,
"grad_norm": 0.6926528215408325,
"learning_rate": 9.653540160603956e-06,
"loss": 0.7130146622657776,
"step": 299
},
{
"epoch": 0.2940816076461218,
"grad_norm": 0.6133829951286316,
"learning_rate": 9.650632591309431e-06,
"loss": 0.7320060729980469,
"step": 300
},
{
"epoch": 0.29506187967160885,
"grad_norm": 0.7793422937393188,
"learning_rate": 9.647713314052896e-06,
"loss": 0.6421869993209839,
"step": 301
},
{
"epoch": 0.29604215169709597,
"grad_norm": 2.3321449756622314,
"learning_rate": 9.64478233618363e-06,
"loss": 0.6765047311782837,
"step": 302
},
{
"epoch": 0.29702242372258303,
"grad_norm": 2.3831698894500732,
"learning_rate": 9.641839665080363e-06,
"loss": 0.6518298387527466,
"step": 303
},
{
"epoch": 0.2980026957480701,
"grad_norm": 0.7824472188949585,
"learning_rate": 9.63888530815127e-06,
"loss": 0.5803199410438538,
"step": 304
},
{
"epoch": 0.29898296777355715,
"grad_norm": 0.8554420471191406,
"learning_rate": 9.635919272833938e-06,
"loss": 0.6380887031555176,
"step": 305
},
{
"epoch": 0.2999632397990442,
"grad_norm": 0.8283202648162842,
"learning_rate": 9.632941566595357e-06,
"loss": 0.7760581970214844,
"step": 306
},
{
"epoch": 0.30094351182453133,
"grad_norm": 1.3910760879516602,
"learning_rate": 9.629952196931902e-06,
"loss": 0.6023370027542114,
"step": 307
},
{
"epoch": 0.3019237838500184,
"grad_norm": 0.7322402596473694,
"learning_rate": 9.626951171369306e-06,
"loss": 0.6995968818664551,
"step": 308
},
{
"epoch": 0.30290405587550545,
"grad_norm": 0.9565230011940002,
"learning_rate": 9.623938497462647e-06,
"loss": 0.9693416357040405,
"step": 309
},
{
"epoch": 0.3038843279009925,
"grad_norm": 0.812511146068573,
"learning_rate": 9.620914182796327e-06,
"loss": 0.5506547689437866,
"step": 310
},
{
"epoch": 0.3048645999264796,
"grad_norm": 0.9502701759338379,
"learning_rate": 9.617878234984056e-06,
"loss": 0.6230248808860779,
"step": 311
},
{
"epoch": 0.3058448719519667,
"grad_norm": 0.9335958957672119,
"learning_rate": 9.61483066166883e-06,
"loss": 0.6191809773445129,
"step": 312
},
{
"epoch": 0.30682514397745375,
"grad_norm": 5.47977352142334,
"learning_rate": 9.611771470522908e-06,
"loss": 0.6781808137893677,
"step": 313
},
{
"epoch": 0.3078054160029408,
"grad_norm": 1.4746034145355225,
"learning_rate": 9.608700669247803e-06,
"loss": 0.6442365050315857,
"step": 314
},
{
"epoch": 0.3087856880284279,
"grad_norm": 0.9942647814750671,
"learning_rate": 9.60561826557425e-06,
"loss": 0.597833514213562,
"step": 315
},
{
"epoch": 0.30976596005391493,
"grad_norm": 1.0260273218154907,
"learning_rate": 9.602524267262202e-06,
"loss": 0.7060080766677856,
"step": 316
},
{
"epoch": 0.31074623207940205,
"grad_norm": 0.9933707118034363,
"learning_rate": 9.599418682100793e-06,
"loss": 0.6442773342132568,
"step": 317
},
{
"epoch": 0.3117265041048891,
"grad_norm": 0.7926250100135803,
"learning_rate": 9.596301517908329e-06,
"loss": 0.6088917255401611,
"step": 318
},
{
"epoch": 0.3127067761303762,
"grad_norm": 1.0485554933547974,
"learning_rate": 9.59317278253227e-06,
"loss": 0.7573684453964233,
"step": 319
},
{
"epoch": 0.31368704815586324,
"grad_norm": 0.5886617302894592,
"learning_rate": 9.5900324838492e-06,
"loss": 0.6841493844985962,
"step": 320
},
{
"epoch": 0.31466732018135035,
"grad_norm": 0.6529507040977478,
"learning_rate": 9.586880629764817e-06,
"loss": 0.6112189888954163,
"step": 321
},
{
"epoch": 0.3156475922068374,
"grad_norm": 2.5114712715148926,
"learning_rate": 9.583717228213914e-06,
"loss": 0.6881189942359924,
"step": 322
},
{
"epoch": 0.3166278642323245,
"grad_norm": 0.9349796175956726,
"learning_rate": 9.580542287160348e-06,
"loss": 0.7102172374725342,
"step": 323
},
{
"epoch": 0.31760813625781154,
"grad_norm": 0.9204770922660828,
"learning_rate": 9.577355814597031e-06,
"loss": 0.7994803190231323,
"step": 324
},
{
"epoch": 0.3185884082832986,
"grad_norm": 2.387535810470581,
"learning_rate": 9.574157818545902e-06,
"loss": 0.6115997433662415,
"step": 325
},
{
"epoch": 0.3195686803087857,
"grad_norm": 1.2952972650527954,
"learning_rate": 9.570948307057912e-06,
"loss": 0.7757495641708374,
"step": 326
},
{
"epoch": 0.3205489523342728,
"grad_norm": 0.7808994054794312,
"learning_rate": 9.567727288213005e-06,
"loss": 0.5963928699493408,
"step": 327
},
{
"epoch": 0.32152922435975984,
"grad_norm": 0.8594168424606323,
"learning_rate": 9.564494770120089e-06,
"loss": 0.6375862956047058,
"step": 328
},
{
"epoch": 0.3225094963852469,
"grad_norm": 1.2403042316436768,
"learning_rate": 9.561250760917026e-06,
"loss": 0.6881876587867737,
"step": 329
},
{
"epoch": 0.32348976841073396,
"grad_norm": 0.750239908695221,
"learning_rate": 9.557995268770609e-06,
"loss": 0.7515847682952881,
"step": 330
},
{
"epoch": 0.3244700404362211,
"grad_norm": 57.436126708984375,
"learning_rate": 9.554728301876525e-06,
"loss": 0.599770188331604,
"step": 331
},
{
"epoch": 0.32545031246170814,
"grad_norm": 0.8042021989822388,
"learning_rate": 9.551449868459371e-06,
"loss": 0.7066971063613892,
"step": 332
},
{
"epoch": 0.3264305844871952,
"grad_norm": 1.1938726902008057,
"learning_rate": 9.548159976772593e-06,
"loss": 0.6569622755050659,
"step": 333
},
{
"epoch": 0.32741085651268226,
"grad_norm": 1.2020000219345093,
"learning_rate": 9.544858635098491e-06,
"loss": 0.6445066928863525,
"step": 334
},
{
"epoch": 0.3283911285381693,
"grad_norm": 1.0184276103973389,
"learning_rate": 9.541545851748186e-06,
"loss": 0.7591553926467896,
"step": 335
},
{
"epoch": 0.32937140056365644,
"grad_norm": 3.5472099781036377,
"learning_rate": 9.538221635061611e-06,
"loss": 0.5578639507293701,
"step": 336
},
{
"epoch": 0.3303516725891435,
"grad_norm": 1.9224342107772827,
"learning_rate": 9.534885993407474e-06,
"loss": 0.6877670288085938,
"step": 337
},
{
"epoch": 0.33133194461463056,
"grad_norm": 0.8383845686912537,
"learning_rate": 9.531538935183252e-06,
"loss": 0.6924867033958435,
"step": 338
},
{
"epoch": 0.3323122166401176,
"grad_norm": 0.922551691532135,
"learning_rate": 9.528180468815155e-06,
"loss": 0.7199808359146118,
"step": 339
},
{
"epoch": 0.3332924886656047,
"grad_norm": 0.7657532691955566,
"learning_rate": 9.524810602758122e-06,
"loss": 0.6060621738433838,
"step": 340
},
{
"epoch": 0.3342727606910918,
"grad_norm": 0.821231484413147,
"learning_rate": 9.521429345495787e-06,
"loss": 0.6362730860710144,
"step": 341
},
{
"epoch": 0.33525303271657886,
"grad_norm": 1.1259968280792236,
"learning_rate": 9.518036705540459e-06,
"loss": 0.8384077548980713,
"step": 342
},
{
"epoch": 0.3362333047420659,
"grad_norm": 0.9031018018722534,
"learning_rate": 9.514632691433108e-06,
"loss": 0.6827899217605591,
"step": 343
},
{
"epoch": 0.337213576767553,
"grad_norm": 1.2601686716079712,
"learning_rate": 9.51121731174333e-06,
"loss": 0.6730697154998779,
"step": 344
},
{
"epoch": 0.33819384879304004,
"grad_norm": 1.6862537860870361,
"learning_rate": 9.507790575069347e-06,
"loss": 0.7867573499679565,
"step": 345
},
{
"epoch": 0.33917412081852716,
"grad_norm": 0.957073450088501,
"learning_rate": 9.504352490037957e-06,
"loss": 0.6127393841743469,
"step": 346
},
{
"epoch": 0.3401543928440142,
"grad_norm": 1.2228965759277344,
"learning_rate": 9.50090306530454e-06,
"loss": 0.771955132484436,
"step": 347
},
{
"epoch": 0.3411346648695013,
"grad_norm": 0.7784944176673889,
"learning_rate": 9.497442309553017e-06,
"loss": 0.5940264463424683,
"step": 348
},
{
"epoch": 0.34211493689498834,
"grad_norm": 1.5570017099380493,
"learning_rate": 9.493970231495836e-06,
"loss": 0.7897027134895325,
"step": 349
},
{
"epoch": 0.34309520892047546,
"grad_norm": 5.5942487716674805,
"learning_rate": 9.490486839873949e-06,
"loss": 0.7134181261062622,
"step": 350
},
{
"epoch": 0.3440754809459625,
"grad_norm": 0.7008062601089478,
"learning_rate": 9.486992143456792e-06,
"loss": 0.6386063694953918,
"step": 351
},
{
"epoch": 0.3450557529714496,
"grad_norm": 0.8100072741508484,
"learning_rate": 9.483486151042258e-06,
"loss": 0.6287713050842285,
"step": 352
},
{
"epoch": 0.34603602499693664,
"grad_norm": 1.141973853111267,
"learning_rate": 9.47996887145668e-06,
"loss": 0.5509628057479858,
"step": 353
},
{
"epoch": 0.3470162970224237,
"grad_norm": 0.8357543349266052,
"learning_rate": 9.476440313554803e-06,
"loss": 0.6555135250091553,
"step": 354
},
{
"epoch": 0.3479965690479108,
"grad_norm": 0.681983470916748,
"learning_rate": 9.47290048621977e-06,
"loss": 0.6392881870269775,
"step": 355
},
{
"epoch": 0.3489768410733979,
"grad_norm": 0.691612184047699,
"learning_rate": 9.469349398363088e-06,
"loss": 0.6376049518585205,
"step": 356
},
{
"epoch": 0.34995711309888494,
"grad_norm": 0.7664949297904968,
"learning_rate": 9.46578705892462e-06,
"loss": 0.6754952669143677,
"step": 357
},
{
"epoch": 0.350937385124372,
"grad_norm": 7.023342132568359,
"learning_rate": 9.46221347687255e-06,
"loss": 0.7657310962677002,
"step": 358
},
{
"epoch": 0.35191765714985906,
"grad_norm": 0.7769929766654968,
"learning_rate": 9.458628661203368e-06,
"loss": 0.6024239659309387,
"step": 359
},
{
"epoch": 0.3528979291753462,
"grad_norm": 0.7423725724220276,
"learning_rate": 9.45503262094184e-06,
"loss": 0.6729433536529541,
"step": 360
},
{
"epoch": 0.35387820120083324,
"grad_norm": 4.273207187652588,
"learning_rate": 9.451425365140997e-06,
"loss": 0.6454064846038818,
"step": 361
},
{
"epoch": 0.3548584732263203,
"grad_norm": 0.5799880623817444,
"learning_rate": 9.447806902882098e-06,
"loss": 0.5626070499420166,
"step": 362
},
{
"epoch": 0.35583874525180736,
"grad_norm": 24.709762573242188,
"learning_rate": 9.444177243274619e-06,
"loss": 0.6042275428771973,
"step": 363
},
{
"epoch": 0.3568190172772944,
"grad_norm": 0.972450852394104,
"learning_rate": 9.440536395456223e-06,
"loss": 0.6978904008865356,
"step": 364
},
{
"epoch": 0.35779928930278154,
"grad_norm": 0.8055760264396667,
"learning_rate": 9.43688436859274e-06,
"loss": 0.6959550380706787,
"step": 365
},
{
"epoch": 0.3587795613282686,
"grad_norm": 0.6148384213447571,
"learning_rate": 9.433221171878144e-06,
"loss": 0.5560431480407715,
"step": 366
},
{
"epoch": 0.35975983335375566,
"grad_norm": 0.6735830307006836,
"learning_rate": 9.429546814534528e-06,
"loss": 0.6831464171409607,
"step": 367
},
{
"epoch": 0.3607401053792427,
"grad_norm": 0.7387506365776062,
"learning_rate": 9.425861305812083e-06,
"loss": 0.8408324122428894,
"step": 368
},
{
"epoch": 0.3617203774047298,
"grad_norm": 1.61435866355896,
"learning_rate": 9.422164654989073e-06,
"loss": 0.6973490118980408,
"step": 369
},
{
"epoch": 0.3627006494302169,
"grad_norm": 0.7609229683876038,
"learning_rate": 9.41845687137181e-06,
"loss": 0.7418415546417236,
"step": 370
},
{
"epoch": 0.36368092145570396,
"grad_norm": 1.7951691150665283,
"learning_rate": 9.414737964294636e-06,
"loss": 0.594944417476654,
"step": 371
},
{
"epoch": 0.364661193481191,
"grad_norm": 0.9224826693534851,
"learning_rate": 9.411007943119895e-06,
"loss": 0.9148576259613037,
"step": 372
},
{
"epoch": 0.3656414655066781,
"grad_norm": 0.8597149848937988,
"learning_rate": 9.40726681723791e-06,
"loss": 0.6997848749160767,
"step": 373
},
{
"epoch": 0.36662173753216515,
"grad_norm": 0.6958048343658447,
"learning_rate": 9.403514596066964e-06,
"loss": 0.6196635961532593,
"step": 374
},
{
"epoch": 0.36760200955765227,
"grad_norm": 0.6815536618232727,
"learning_rate": 9.399751289053267e-06,
"loss": 0.6499385833740234,
"step": 375
},
{
"epoch": 0.3685822815831393,
"grad_norm": 0.7790945768356323,
"learning_rate": 9.39597690567094e-06,
"loss": 0.5256167650222778,
"step": 376
},
{
"epoch": 0.3695625536086264,
"grad_norm": 0.6119083762168884,
"learning_rate": 9.392191455421989e-06,
"loss": 0.5613681077957153,
"step": 377
},
{
"epoch": 0.37054282563411345,
"grad_norm": 0.7187506556510925,
"learning_rate": 9.388394947836278e-06,
"loss": 0.5893592834472656,
"step": 378
},
{
"epoch": 0.37152309765960057,
"grad_norm": 0.6685920357704163,
"learning_rate": 9.384587392471516e-06,
"loss": 0.5686390995979309,
"step": 379
},
{
"epoch": 0.3725033696850876,
"grad_norm": 0.8494288325309753,
"learning_rate": 9.380768798913213e-06,
"loss": 0.6900075674057007,
"step": 380
},
{
"epoch": 0.3734836417105747,
"grad_norm": 1.3484470844268799,
"learning_rate": 9.376939176774678e-06,
"loss": 0.6516958475112915,
"step": 381
},
{
"epoch": 0.37446391373606175,
"grad_norm": 0.5787715315818787,
"learning_rate": 9.37309853569698e-06,
"loss": 0.6007320880889893,
"step": 382
},
{
"epoch": 0.3754441857615488,
"grad_norm": 0.7211255431175232,
"learning_rate": 9.369246885348926e-06,
"loss": 0.6861116290092468,
"step": 383
},
{
"epoch": 0.3764244577870359,
"grad_norm": 0.5919385552406311,
"learning_rate": 9.365384235427042e-06,
"loss": 0.6424866914749146,
"step": 384
},
{
"epoch": 0.377404729812523,
"grad_norm": 16.4747314453125,
"learning_rate": 9.361510595655545e-06,
"loss": 0.7053197026252747,
"step": 385
},
{
"epoch": 0.37838500183801005,
"grad_norm": 0.5669404864311218,
"learning_rate": 9.357625975786318e-06,
"loss": 0.6393853425979614,
"step": 386
},
{
"epoch": 0.3793652738634971,
"grad_norm": 0.8009690046310425,
"learning_rate": 9.353730385598887e-06,
"loss": 0.6367558240890503,
"step": 387
},
{
"epoch": 0.38034554588898417,
"grad_norm": 0.7134267687797546,
"learning_rate": 9.349823834900396e-06,
"loss": 0.5795502066612244,
"step": 388
},
{
"epoch": 0.3813258179144713,
"grad_norm": 0.6554087996482849,
"learning_rate": 9.345906333525582e-06,
"loss": 0.5673925876617432,
"step": 389
},
{
"epoch": 0.38230608993995835,
"grad_norm": 0.7672719359397888,
"learning_rate": 9.341977891336748e-06,
"loss": 0.626176118850708,
"step": 390
},
{
"epoch": 0.3832863619654454,
"grad_norm": 0.8481528162956238,
"learning_rate": 9.338038518223746e-06,
"loss": 0.5982210636138916,
"step": 391
},
{
"epoch": 0.38426663399093247,
"grad_norm": 0.6634806990623474,
"learning_rate": 9.334088224103942e-06,
"loss": 0.5824027061462402,
"step": 392
},
{
"epoch": 0.38524690601641953,
"grad_norm": 0.7833440899848938,
"learning_rate": 9.330127018922195e-06,
"loss": 0.7327437400817871,
"step": 393
},
{
"epoch": 0.38622717804190665,
"grad_norm": 7.219211101531982,
"learning_rate": 9.326154912650834e-06,
"loss": 0.5804681777954102,
"step": 394
},
{
"epoch": 0.3872074500673937,
"grad_norm": 1.305361032485962,
"learning_rate": 9.322171915289635e-06,
"loss": 0.6933940052986145,
"step": 395
},
{
"epoch": 0.38818772209288077,
"grad_norm": 0.9371297359466553,
"learning_rate": 9.318178036865786e-06,
"loss": 0.5943507552146912,
"step": 396
},
{
"epoch": 0.38916799411836783,
"grad_norm": 0.9979481101036072,
"learning_rate": 9.314173287433874e-06,
"loss": 0.7934183478355408,
"step": 397
},
{
"epoch": 0.3901482661438549,
"grad_norm": 0.7355287075042725,
"learning_rate": 9.310157677075847e-06,
"loss": 0.6028023362159729,
"step": 398
},
{
"epoch": 0.391128538169342,
"grad_norm": 0.8032557368278503,
"learning_rate": 9.306131215901004e-06,
"loss": 0.6857631206512451,
"step": 399
},
{
"epoch": 0.39210881019482907,
"grad_norm": 0.5795166492462158,
"learning_rate": 9.302093914045953e-06,
"loss": 0.47602543234825134,
"step": 400
},
{
"epoch": 0.39308908222031613,
"grad_norm": 0.6684191823005676,
"learning_rate": 9.298045781674595e-06,
"loss": 0.6103488802909851,
"step": 401
},
{
"epoch": 0.3940693542458032,
"grad_norm": 2.738152265548706,
"learning_rate": 9.293986828978106e-06,
"loss": 0.6556514501571655,
"step": 402
},
{
"epoch": 0.3950496262712903,
"grad_norm": 1.0126097202301025,
"learning_rate": 9.289917066174887e-06,
"loss": 0.6506785154342651,
"step": 403
},
{
"epoch": 0.39602989829677737,
"grad_norm": 0.6366498470306396,
"learning_rate": 9.285836503510562e-06,
"loss": 0.5810170769691467,
"step": 404
},
{
"epoch": 0.39701017032226443,
"grad_norm": 0.912286102771759,
"learning_rate": 9.281745151257946e-06,
"loss": 0.6202950477600098,
"step": 405
},
{
"epoch": 0.3979904423477515,
"grad_norm": 26.518205642700195,
"learning_rate": 9.277643019717011e-06,
"loss": 0.7940681576728821,
"step": 406
},
{
"epoch": 0.39897071437323856,
"grad_norm": 1.4417810440063477,
"learning_rate": 9.273530119214868e-06,
"loss": 0.6001892685890198,
"step": 407
},
{
"epoch": 0.39995098639872567,
"grad_norm": 2.202831983566284,
"learning_rate": 9.269406460105742e-06,
"loss": 0.5752210021018982,
"step": 408
},
{
"epoch": 0.40093125842421273,
"grad_norm": 0.620313286781311,
"learning_rate": 9.265272052770936e-06,
"loss": 0.5697149634361267,
"step": 409
},
{
"epoch": 0.4019115304496998,
"grad_norm": 0.5553503632545471,
"learning_rate": 9.261126907618818e-06,
"loss": 0.560996413230896,
"step": 410
},
{
"epoch": 0.40289180247518686,
"grad_norm": 0.9686580300331116,
"learning_rate": 9.256971035084786e-06,
"loss": 0.7202277183532715,
"step": 411
},
{
"epoch": 0.4038720745006739,
"grad_norm": 0.8185970783233643,
"learning_rate": 9.252804445631243e-06,
"loss": 0.5707693099975586,
"step": 412
},
{
"epoch": 0.40485234652616103,
"grad_norm": 0.672945499420166,
"learning_rate": 9.248627149747573e-06,
"loss": 0.7931631803512573,
"step": 413
},
{
"epoch": 0.4058326185516481,
"grad_norm": 1.389769196510315,
"learning_rate": 9.244439157950114e-06,
"loss": 0.7967973351478577,
"step": 414
},
{
"epoch": 0.40681289057713516,
"grad_norm": 0.7315179705619812,
"learning_rate": 9.24024048078213e-06,
"loss": 0.5561456680297852,
"step": 415
},
{
"epoch": 0.4077931626026222,
"grad_norm": 0.8126598596572876,
"learning_rate": 9.236031128813786e-06,
"loss": 0.6264670491218567,
"step": 416
},
{
"epoch": 0.4087734346281093,
"grad_norm": 0.8071774244308472,
"learning_rate": 9.231811112642121e-06,
"loss": 0.6467783451080322,
"step": 417
},
{
"epoch": 0.4097537066535964,
"grad_norm": 0.6707936525344849,
"learning_rate": 9.227580442891022e-06,
"loss": 0.5915917158126831,
"step": 418
},
{
"epoch": 0.41073397867908346,
"grad_norm": 0.679583728313446,
"learning_rate": 9.223339130211194e-06,
"loss": 0.6092835664749146,
"step": 419
},
{
"epoch": 0.4117142507045705,
"grad_norm": 19.181005477905273,
"learning_rate": 9.219087185280132e-06,
"loss": 0.7507760524749756,
"step": 420
},
{
"epoch": 0.4126945227300576,
"grad_norm": 0.997204601764679,
"learning_rate": 9.214824618802108e-06,
"loss": 0.7204283475875854,
"step": 421
},
{
"epoch": 0.41367479475554464,
"grad_norm": 0.9711258411407471,
"learning_rate": 9.210551441508126e-06,
"loss": 0.5186525583267212,
"step": 422
},
{
"epoch": 0.41465506678103176,
"grad_norm": 2.5584208965301514,
"learning_rate": 9.206267664155906e-06,
"loss": 0.5395511388778687,
"step": 423
},
{
"epoch": 0.4156353388065188,
"grad_norm": 0.8428700566291809,
"learning_rate": 9.201973297529852e-06,
"loss": 0.5764715671539307,
"step": 424
},
{
"epoch": 0.4166156108320059,
"grad_norm": 0.719062089920044,
"learning_rate": 9.197668352441025e-06,
"loss": 0.5695844888687134,
"step": 425
},
{
"epoch": 0.41759588285749294,
"grad_norm": 0.8574051260948181,
"learning_rate": 9.193352839727122e-06,
"loss": 0.5854992866516113,
"step": 426
},
{
"epoch": 0.41857615488298,
"grad_norm": 1.0974955558776855,
"learning_rate": 9.189026770252437e-06,
"loss": 0.5945618152618408,
"step": 427
},
{
"epoch": 0.4195564269084671,
"grad_norm": 2.0582549571990967,
"learning_rate": 9.18469015490785e-06,
"loss": 0.7335432767868042,
"step": 428
},
{
"epoch": 0.4205366989339542,
"grad_norm": 0.66349858045578,
"learning_rate": 9.18034300461078e-06,
"loss": 0.6619755029678345,
"step": 429
},
{
"epoch": 0.42151697095944124,
"grad_norm": 1.2293531894683838,
"learning_rate": 9.175985330305176e-06,
"loss": 0.6925309896469116,
"step": 430
},
{
"epoch": 0.4224972429849283,
"grad_norm": 0.5356952548027039,
"learning_rate": 9.171617142961477e-06,
"loss": 0.5922069549560547,
"step": 431
},
{
"epoch": 0.4234775150104154,
"grad_norm": 0.7660447359085083,
"learning_rate": 9.167238453576589e-06,
"loss": 0.6395033597946167,
"step": 432
},
{
"epoch": 0.4244577870359025,
"grad_norm": 16.272747039794922,
"learning_rate": 9.162849273173857e-06,
"loss": 0.6422967910766602,
"step": 433
},
{
"epoch": 0.42543805906138954,
"grad_norm": 1.514641523361206,
"learning_rate": 9.158449612803039e-06,
"loss": 0.7670032978057861,
"step": 434
},
{
"epoch": 0.4264183310868766,
"grad_norm": 0.7433525919914246,
"learning_rate": 9.154039483540273e-06,
"loss": 0.6374302506446838,
"step": 435
},
{
"epoch": 0.42739860311236366,
"grad_norm": 0.950039267539978,
"learning_rate": 9.149618896488056e-06,
"loss": 0.6156798601150513,
"step": 436
},
{
"epoch": 0.4283788751378508,
"grad_norm": 0.7084690928459167,
"learning_rate": 9.145187862775208e-06,
"loss": 0.4703499674797058,
"step": 437
},
{
"epoch": 0.42935914716333784,
"grad_norm": 0.8909004330635071,
"learning_rate": 9.140746393556853e-06,
"loss": 0.6387060880661011,
"step": 438
},
{
"epoch": 0.4303394191888249,
"grad_norm": 1.5755165815353394,
"learning_rate": 9.136294500014387e-06,
"loss": 0.6969273090362549,
"step": 439
},
{
"epoch": 0.43131969121431196,
"grad_norm": 0.86232590675354,
"learning_rate": 9.13183219335544e-06,
"loss": 0.6250169277191162,
"step": 440
},
{
"epoch": 0.432299963239799,
"grad_norm": 0.7907266616821289,
"learning_rate": 9.12735948481387e-06,
"loss": 0.640777587890625,
"step": 441
},
{
"epoch": 0.43328023526528614,
"grad_norm": 1.3714030981063843,
"learning_rate": 9.122876385649713e-06,
"loss": 0.5701188445091248,
"step": 442
},
{
"epoch": 0.4342605072907732,
"grad_norm": 0.9190168976783752,
"learning_rate": 9.118382907149164e-06,
"loss": 0.7065989971160889,
"step": 443
},
{
"epoch": 0.43524077931626026,
"grad_norm": 0.8260921239852905,
"learning_rate": 9.113879060624554e-06,
"loss": 0.6392488479614258,
"step": 444
},
{
"epoch": 0.4362210513417473,
"grad_norm": 0.6012373566627502,
"learning_rate": 9.109364857414306e-06,
"loss": 0.6170405745506287,
"step": 445
},
{
"epoch": 0.4372013233672344,
"grad_norm": 0.6749585866928101,
"learning_rate": 9.104840308882925e-06,
"loss": 0.6490533351898193,
"step": 446
},
{
"epoch": 0.4381815953927215,
"grad_norm": 1.9798930883407593,
"learning_rate": 9.100305426420957e-06,
"loss": 0.6532576084136963,
"step": 447
},
{
"epoch": 0.43916186741820856,
"grad_norm": 3.4725520610809326,
"learning_rate": 9.09576022144496e-06,
"loss": 0.5746736526489258,
"step": 448
},
{
"epoch": 0.4401421394436956,
"grad_norm": 5.0959696769714355,
"learning_rate": 9.091204705397485e-06,
"loss": 0.5824971199035645,
"step": 449
},
{
"epoch": 0.4411224114691827,
"grad_norm": 0.8858083486557007,
"learning_rate": 9.086638889747034e-06,
"loss": 0.6251646280288696,
"step": 450
},
{
"epoch": 0.44210268349466975,
"grad_norm": 0.5421618819236755,
"learning_rate": 9.08206278598805e-06,
"loss": 0.4371255040168762,
"step": 451
},
{
"epoch": 0.44308295552015686,
"grad_norm": 0.6242964863777161,
"learning_rate": 9.077476405640862e-06,
"loss": 0.7059438228607178,
"step": 452
},
{
"epoch": 0.4440632275456439,
"grad_norm": 0.7129407525062561,
"learning_rate": 9.07287976025168e-06,
"loss": 0.652492880821228,
"step": 453
},
{
"epoch": 0.445043499571131,
"grad_norm": 0.8711827397346497,
"learning_rate": 9.06827286139255e-06,
"loss": 0.6065002679824829,
"step": 454
},
{
"epoch": 0.44602377159661805,
"grad_norm": 0.812637984752655,
"learning_rate": 9.06365572066134e-06,
"loss": 0.7273547649383545,
"step": 455
},
{
"epoch": 0.4470040436221051,
"grad_norm": 0.8697981238365173,
"learning_rate": 9.059028349681693e-06,
"loss": 0.6040782928466797,
"step": 456
},
{
"epoch": 0.4479843156475922,
"grad_norm": 0.5100929737091064,
"learning_rate": 9.05439076010301e-06,
"loss": 0.5205973386764526,
"step": 457
},
{
"epoch": 0.4489645876730793,
"grad_norm": 1.4821308851242065,
"learning_rate": 9.04974296360042e-06,
"loss": 0.6293808221817017,
"step": 458
},
{
"epoch": 0.44994485969856635,
"grad_norm": 0.7332229018211365,
"learning_rate": 9.045084971874738e-06,
"loss": 0.660082221031189,
"step": 459
},
{
"epoch": 0.4509251317240534,
"grad_norm": 1.5905967950820923,
"learning_rate": 9.040416796652459e-06,
"loss": 0.6811535358428955,
"step": 460
},
{
"epoch": 0.4519054037495405,
"grad_norm": 0.7074926495552063,
"learning_rate": 9.035738449685707e-06,
"loss": 0.5867305994033813,
"step": 461
},
{
"epoch": 0.4528856757750276,
"grad_norm": 1.1056369543075562,
"learning_rate": 9.031049942752215e-06,
"loss": 0.7666922807693481,
"step": 462
},
{
"epoch": 0.45386594780051465,
"grad_norm": 0.6271103024482727,
"learning_rate": 9.026351287655294e-06,
"loss": 0.5773208141326904,
"step": 463
},
{
"epoch": 0.4548462198260017,
"grad_norm": 3.558201789855957,
"learning_rate": 9.021642496223802e-06,
"loss": 0.6142169237136841,
"step": 464
},
{
"epoch": 0.45582649185148877,
"grad_norm": 0.7741997838020325,
"learning_rate": 9.016923580312114e-06,
"loss": 0.6007839441299438,
"step": 465
},
{
"epoch": 0.4568067638769759,
"grad_norm": 0.6544026732444763,
"learning_rate": 9.012194551800098e-06,
"loss": 0.5652458667755127,
"step": 466
},
{
"epoch": 0.45778703590246295,
"grad_norm": 0.8572518825531006,
"learning_rate": 9.007455422593077e-06,
"loss": 0.699006199836731,
"step": 467
},
{
"epoch": 0.45876730792795,
"grad_norm": 0.7190974950790405,
"learning_rate": 9.002706204621802e-06,
"loss": 0.5994861125946045,
"step": 468
},
{
"epoch": 0.45974757995343707,
"grad_norm": 0.7033573985099792,
"learning_rate": 8.997946909842426e-06,
"loss": 0.5507349967956543,
"step": 469
},
{
"epoch": 0.46072785197892413,
"grad_norm": 0.8660258650779724,
"learning_rate": 8.993177550236464e-06,
"loss": 0.5761494636535645,
"step": 470
},
{
"epoch": 0.46170812400441125,
"grad_norm": 6.447513103485107,
"learning_rate": 8.988398137810778e-06,
"loss": 0.5650913715362549,
"step": 471
},
{
"epoch": 0.4626883960298983,
"grad_norm": 2.910141944885254,
"learning_rate": 8.983608684597529e-06,
"loss": 0.683720588684082,
"step": 472
},
{
"epoch": 0.46366866805538537,
"grad_norm": 0.7440175414085388,
"learning_rate": 8.978809202654161e-06,
"loss": 0.6870052814483643,
"step": 473
},
{
"epoch": 0.46464894008087243,
"grad_norm": 0.8271398544311523,
"learning_rate": 8.973999704063365e-06,
"loss": 0.6572490334510803,
"step": 474
},
{
"epoch": 0.4656292121063595,
"grad_norm": 0.6863585114479065,
"learning_rate": 8.969180200933048e-06,
"loss": 0.7248791456222534,
"step": 475
},
{
"epoch": 0.4666094841318466,
"grad_norm": 0.6197232604026794,
"learning_rate": 8.9643507053963e-06,
"loss": 0.6060189008712769,
"step": 476
},
{
"epoch": 0.46758975615733367,
"grad_norm": 0.909180223941803,
"learning_rate": 8.959511229611377e-06,
"loss": 0.8351896405220032,
"step": 477
},
{
"epoch": 0.46857002818282073,
"grad_norm": 0.5272825956344604,
"learning_rate": 8.954661785761648e-06,
"loss": 0.674493670463562,
"step": 478
},
{
"epoch": 0.4695503002083078,
"grad_norm": 0.6872929930686951,
"learning_rate": 8.949802386055582e-06,
"loss": 0.744827151298523,
"step": 479
},
{
"epoch": 0.47053057223379485,
"grad_norm": 1.0290851593017578,
"learning_rate": 8.944933042726715e-06,
"loss": 0.6393369436264038,
"step": 480
},
{
"epoch": 0.47151084425928197,
"grad_norm": 0.8363896012306213,
"learning_rate": 8.94005376803361e-06,
"loss": 0.7189200520515442,
"step": 481
},
{
"epoch": 0.47249111628476903,
"grad_norm": 0.7918285131454468,
"learning_rate": 8.935164574259838e-06,
"loss": 0.7011231780052185,
"step": 482
},
{
"epoch": 0.4734713883102561,
"grad_norm": 1.030903935432434,
"learning_rate": 8.930265473713939e-06,
"loss": 0.6436402797698975,
"step": 483
},
{
"epoch": 0.47445166033574315,
"grad_norm": 1.104646921157837,
"learning_rate": 8.925356478729389e-06,
"loss": 0.5960245132446289,
"step": 484
},
{
"epoch": 0.4754319323612302,
"grad_norm": 0.6375426650047302,
"learning_rate": 8.92043760166458e-06,
"loss": 0.6274738311767578,
"step": 485
},
{
"epoch": 0.47641220438671733,
"grad_norm": 0.9720222353935242,
"learning_rate": 8.915508854902778e-06,
"loss": 0.5411441326141357,
"step": 486
},
{
"epoch": 0.4773924764122044,
"grad_norm": 0.7098354697227478,
"learning_rate": 8.910570250852098e-06,
"loss": 0.47495734691619873,
"step": 487
},
{
"epoch": 0.47837274843769145,
"grad_norm": 0.5941671133041382,
"learning_rate": 8.905621801945467e-06,
"loss": 0.6280163526535034,
"step": 488
},
{
"epoch": 0.4793530204631785,
"grad_norm": 0.5638043880462646,
"learning_rate": 8.900663520640605e-06,
"loss": 0.6621802449226379,
"step": 489
},
{
"epoch": 0.48033329248866563,
"grad_norm": 0.9914025664329529,
"learning_rate": 8.895695419419973e-06,
"loss": 0.8763112425804138,
"step": 490
},
{
"epoch": 0.4813135645141527,
"grad_norm": 0.6684858202934265,
"learning_rate": 8.890717510790763e-06,
"loss": 0.5855538249015808,
"step": 491
},
{
"epoch": 0.48229383653963975,
"grad_norm": 0.68495112657547,
"learning_rate": 8.885729807284855e-06,
"loss": 0.6838175654411316,
"step": 492
},
{
"epoch": 0.4832741085651268,
"grad_norm": 1.0476689338684082,
"learning_rate": 8.880732321458785e-06,
"loss": 0.6090328693389893,
"step": 493
},
{
"epoch": 0.4842543805906139,
"grad_norm": 1.5346369743347168,
"learning_rate": 8.875725065893717e-06,
"loss": 0.6232689619064331,
"step": 494
},
{
"epoch": 0.485234652616101,
"grad_norm": 0.9822683930397034,
"learning_rate": 8.870708053195414e-06,
"loss": 0.673301100730896,
"step": 495
},
{
"epoch": 0.48621492464158805,
"grad_norm": 0.7001924514770508,
"learning_rate": 8.865681295994195e-06,
"loss": 0.6804222464561462,
"step": 496
},
{
"epoch": 0.4871951966670751,
"grad_norm": 0.6514866352081299,
"learning_rate": 8.860644806944917e-06,
"loss": 0.5138349533081055,
"step": 497
},
{
"epoch": 0.4881754686925622,
"grad_norm": 0.9068660736083984,
"learning_rate": 8.85559859872694e-06,
"loss": 0.679947018623352,
"step": 498
},
{
"epoch": 0.48915574071804924,
"grad_norm": 0.8489823937416077,
"learning_rate": 8.850542684044078e-06,
"loss": 0.5598639249801636,
"step": 499
},
{
"epoch": 0.49013601274353635,
"grad_norm": 0.6451029777526855,
"learning_rate": 8.845477075624598e-06,
"loss": 0.6119801998138428,
"step": 500
},
{
"epoch": 0.4911162847690234,
"grad_norm": 0.5757226943969727,
"learning_rate": 8.84040178622116e-06,
"loss": 0.6911470293998718,
"step": 501
},
{
"epoch": 0.4920965567945105,
"grad_norm": 0.9325789213180542,
"learning_rate": 8.835316828610801e-06,
"loss": 0.7041717767715454,
"step": 502
},
{
"epoch": 0.49307682881999754,
"grad_norm": 1.6333476305007935,
"learning_rate": 8.83022221559489e-06,
"loss": 0.691689670085907,
"step": 503
},
{
"epoch": 0.4940571008454846,
"grad_norm": 0.8043912053108215,
"learning_rate": 8.825117959999117e-06,
"loss": 0.527850866317749,
"step": 504
},
{
"epoch": 0.4950373728709717,
"grad_norm": 0.6235102415084839,
"learning_rate": 8.820004074673433e-06,
"loss": 0.6724036931991577,
"step": 505
},
{
"epoch": 0.4960176448964588,
"grad_norm": 0.9563659429550171,
"learning_rate": 8.814880572492043e-06,
"loss": 0.6215025186538696,
"step": 506
},
{
"epoch": 0.49699791692194584,
"grad_norm": 1.213689923286438,
"learning_rate": 8.809747466353356e-06,
"loss": 0.7895469665527344,
"step": 507
},
{
"epoch": 0.4979781889474329,
"grad_norm": 2.611093044281006,
"learning_rate": 8.804604769179958e-06,
"loss": 0.5000928044319153,
"step": 508
},
{
"epoch": 0.49895846097291996,
"grad_norm": 0.7150916457176208,
"learning_rate": 8.799452493918586e-06,
"loss": 0.7691665291786194,
"step": 509
},
{
"epoch": 0.4999387329984071,
"grad_norm": 0.7139464020729065,
"learning_rate": 8.794290653540086e-06,
"loss": 0.7239859700202942,
"step": 510
},
{
"epoch": 0.5009190050238941,
"grad_norm": 0.5817137956619263,
"learning_rate": 8.789119261039385e-06,
"loss": 0.5050749778747559,
"step": 511
},
{
"epoch": 0.5018992770493812,
"grad_norm": 1.6720318794250488,
"learning_rate": 8.783938329435458e-06,
"loss": 0.5298192501068115,
"step": 512
},
{
"epoch": 0.5028795490748683,
"grad_norm": 0.7029421925544739,
"learning_rate": 8.778747871771293e-06,
"loss": 0.5629361271858215,
"step": 513
},
{
"epoch": 0.5038598211003553,
"grad_norm": 0.6082753539085388,
"learning_rate": 8.773547901113862e-06,
"loss": 0.5543243885040283,
"step": 514
},
{
"epoch": 0.5048400931258424,
"grad_norm": 0.7907118797302246,
"learning_rate": 8.768338430554083e-06,
"loss": 0.551690399646759,
"step": 515
},
{
"epoch": 0.5058203651513294,
"grad_norm": 0.5746113657951355,
"learning_rate": 8.763119473206795e-06,
"loss": 0.6502900123596191,
"step": 516
},
{
"epoch": 0.5068006371768166,
"grad_norm": 0.6561967134475708,
"learning_rate": 8.757891042210713e-06,
"loss": 0.7445024251937866,
"step": 517
},
{
"epoch": 0.5077809092023037,
"grad_norm": 0.8462710380554199,
"learning_rate": 8.752653150728412e-06,
"loss": 0.5634571313858032,
"step": 518
},
{
"epoch": 0.5087611812277907,
"grad_norm": 0.9798592329025269,
"learning_rate": 8.747405811946272e-06,
"loss": 0.5478039383888245,
"step": 519
},
{
"epoch": 0.5097414532532778,
"grad_norm": 0.7543477416038513,
"learning_rate": 8.742149039074463e-06,
"loss": 0.5274918079376221,
"step": 520
},
{
"epoch": 0.5107217252787648,
"grad_norm": 0.6517899036407471,
"learning_rate": 8.736882845346906e-06,
"loss": 0.7066526412963867,
"step": 521
},
{
"epoch": 0.5117019973042519,
"grad_norm": 0.809022068977356,
"learning_rate": 8.731607244021236e-06,
"loss": 0.6081236004829407,
"step": 522
},
{
"epoch": 0.512682269329739,
"grad_norm": 0.7802321314811707,
"learning_rate": 8.726322248378775e-06,
"loss": 0.6819972991943359,
"step": 523
},
{
"epoch": 0.513662541355226,
"grad_norm": 0.6699939966201782,
"learning_rate": 8.72102787172449e-06,
"loss": 0.6287943124771118,
"step": 524
},
{
"epoch": 0.5146428133807132,
"grad_norm": 1.4717462062835693,
"learning_rate": 8.715724127386971e-06,
"loss": 0.6525439023971558,
"step": 525
},
{
"epoch": 0.5156230854062002,
"grad_norm": 1.0533978939056396,
"learning_rate": 8.710411028718388e-06,
"loss": 0.5400506258010864,
"step": 526
},
{
"epoch": 0.5166033574316873,
"grad_norm": 0.7796100974082947,
"learning_rate": 8.705088589094458e-06,
"loss": 0.49268490076065063,
"step": 527
},
{
"epoch": 0.5175836294571744,
"grad_norm": 0.6495805382728577,
"learning_rate": 8.69975682191442e-06,
"loss": 0.6517795324325562,
"step": 528
},
{
"epoch": 0.5185639014826614,
"grad_norm": 0.6956337690353394,
"learning_rate": 8.69441574060099e-06,
"loss": 0.5145743489265442,
"step": 529
},
{
"epoch": 0.5195441735081485,
"grad_norm": 1.0180596113204956,
"learning_rate": 8.68906535860033e-06,
"loss": 0.578504204750061,
"step": 530
},
{
"epoch": 0.5205244455336356,
"grad_norm": 0.8827187418937683,
"learning_rate": 8.683705689382025e-06,
"loss": 0.6472887992858887,
"step": 531
},
{
"epoch": 0.5215047175591226,
"grad_norm": 0.8511680960655212,
"learning_rate": 8.678336746439036e-06,
"loss": 0.7100465297698975,
"step": 532
},
{
"epoch": 0.5224849895846098,
"grad_norm": 1.7726173400878906,
"learning_rate": 8.672958543287666e-06,
"loss": 0.6354161500930786,
"step": 533
},
{
"epoch": 0.5234652616100968,
"grad_norm": 0.7771535515785217,
"learning_rate": 8.667571093467542e-06,
"loss": 0.5966746211051941,
"step": 534
},
{
"epoch": 0.5244455336355839,
"grad_norm": 1.7584549188613892,
"learning_rate": 8.662174410541556e-06,
"loss": 0.6464035511016846,
"step": 535
},
{
"epoch": 0.525425805661071,
"grad_norm": 0.5571742653846741,
"learning_rate": 8.656768508095853e-06,
"loss": 0.6089268922805786,
"step": 536
},
{
"epoch": 0.526406077686558,
"grad_norm": 0.6896767020225525,
"learning_rate": 8.651353399739787e-06,
"loss": 0.6475738286972046,
"step": 537
},
{
"epoch": 0.5273863497120451,
"grad_norm": 0.7882434725761414,
"learning_rate": 8.645929099105886e-06,
"loss": 0.5963877439498901,
"step": 538
},
{
"epoch": 0.5283666217375321,
"grad_norm": 0.7716541290283203,
"learning_rate": 8.640495619849821e-06,
"loss": 0.6715061664581299,
"step": 539
},
{
"epoch": 0.5293468937630192,
"grad_norm": 0.7195318937301636,
"learning_rate": 8.63505297565037e-06,
"loss": 0.5596070289611816,
"step": 540
},
{
"epoch": 0.5303271657885064,
"grad_norm": 0.6509541869163513,
"learning_rate": 8.629601180209382e-06,
"loss": 0.6290733814239502,
"step": 541
},
{
"epoch": 0.5313074378139934,
"grad_norm": 3.264305830001831,
"learning_rate": 8.624140247251745e-06,
"loss": 0.5861960649490356,
"step": 542
},
{
"epoch": 0.5322877098394805,
"grad_norm": 0.6421966552734375,
"learning_rate": 8.61867019052535e-06,
"loss": 0.5340760946273804,
"step": 543
},
{
"epoch": 0.5332679818649675,
"grad_norm": 0.6535043120384216,
"learning_rate": 8.613191023801065e-06,
"loss": 0.5379872918128967,
"step": 544
},
{
"epoch": 0.5342482538904546,
"grad_norm": 0.5565196871757507,
"learning_rate": 8.607702760872679e-06,
"loss": 0.5976905226707458,
"step": 545
},
{
"epoch": 0.5352285259159417,
"grad_norm": 0.595465362071991,
"learning_rate": 8.602205415556889e-06,
"loss": 0.5598548650741577,
"step": 546
},
{
"epoch": 0.5362087979414287,
"grad_norm": 0.508159875869751,
"learning_rate": 8.596699001693257e-06,
"loss": 0.5352834463119507,
"step": 547
},
{
"epoch": 0.5371890699669158,
"grad_norm": 0.6390469074249268,
"learning_rate": 8.591183533144172e-06,
"loss": 0.5889360904693604,
"step": 548
},
{
"epoch": 0.5381693419924028,
"grad_norm": 0.5292559862136841,
"learning_rate": 8.585659023794818e-06,
"loss": 0.516267716884613,
"step": 549
},
{
"epoch": 0.53914961401789,
"grad_norm": 1.288053035736084,
"learning_rate": 8.580125487553143e-06,
"loss": 0.5956067442893982,
"step": 550
},
{
"epoch": 0.5401298860433771,
"grad_norm": 0.7991668581962585,
"learning_rate": 8.574582938349818e-06,
"loss": 0.5806759595870972,
"step": 551
},
{
"epoch": 0.5411101580688641,
"grad_norm": 1.8865896463394165,
"learning_rate": 8.569031390138202e-06,
"loss": 0.691784143447876,
"step": 552
},
{
"epoch": 0.5420904300943512,
"grad_norm": 1.6936720609664917,
"learning_rate": 8.563470856894316e-06,
"loss": 0.5519540905952454,
"step": 553
},
{
"epoch": 0.5430707021198382,
"grad_norm": 0.7170090675354004,
"learning_rate": 8.557901352616788e-06,
"loss": 0.6600730419158936,
"step": 554
},
{
"epoch": 0.5440509741453253,
"grad_norm": 0.5595294237136841,
"learning_rate": 8.552322891326846e-06,
"loss": 0.5864146947860718,
"step": 555
},
{
"epoch": 0.5450312461708124,
"grad_norm": 0.6426265239715576,
"learning_rate": 8.546735487068253e-06,
"loss": 0.6375794410705566,
"step": 556
},
{
"epoch": 0.5460115181962995,
"grad_norm": 0.6871543526649475,
"learning_rate": 8.541139153907296e-06,
"loss": 0.5888508558273315,
"step": 557
},
{
"epoch": 0.5469917902217866,
"grad_norm": 0.5912147760391235,
"learning_rate": 8.535533905932739e-06,
"loss": 0.5039655566215515,
"step": 558
},
{
"epoch": 0.5479720622472736,
"grad_norm": 0.52177494764328,
"learning_rate": 8.529919757255783e-06,
"loss": 0.5552338361740112,
"step": 559
},
{
"epoch": 0.5489523342727607,
"grad_norm": 0.5878831744194031,
"learning_rate": 8.524296722010042e-06,
"loss": 0.617256224155426,
"step": 560
},
{
"epoch": 0.5499326062982478,
"grad_norm": 0.5522481203079224,
"learning_rate": 8.518664814351502e-06,
"loss": 0.5995144844055176,
"step": 561
},
{
"epoch": 0.5509128783237348,
"grad_norm": 0.8053575158119202,
"learning_rate": 8.513024048458484e-06,
"loss": 0.5527684688568115,
"step": 562
},
{
"epoch": 0.5518931503492219,
"grad_norm": 0.6572372913360596,
"learning_rate": 8.507374438531606e-06,
"loss": 0.6319049596786499,
"step": 563
},
{
"epoch": 0.5528734223747089,
"grad_norm": 0.6177840232849121,
"learning_rate": 8.501715998793758e-06,
"loss": 0.5257951021194458,
"step": 564
},
{
"epoch": 0.553853694400196,
"grad_norm": 0.7017096281051636,
"learning_rate": 8.496048743490053e-06,
"loss": 0.5938905477523804,
"step": 565
},
{
"epoch": 0.5548339664256832,
"grad_norm": 0.6734871864318848,
"learning_rate": 8.490372686887802e-06,
"loss": 0.6027377843856812,
"step": 566
},
{
"epoch": 0.5558142384511702,
"grad_norm": 0.5894632339477539,
"learning_rate": 8.48468784327647e-06,
"loss": 0.6919211149215698,
"step": 567
},
{
"epoch": 0.5567945104766573,
"grad_norm": 1.7311468124389648,
"learning_rate": 8.478994226967638e-06,
"loss": 0.6514636874198914,
"step": 568
},
{
"epoch": 0.5577747825021443,
"grad_norm": 0.7870670557022095,
"learning_rate": 8.473291852294986e-06,
"loss": 0.6501897573471069,
"step": 569
},
{
"epoch": 0.5587550545276314,
"grad_norm": 0.6551438570022583,
"learning_rate": 8.467580733614233e-06,
"loss": 0.6422196626663208,
"step": 570
},
{
"epoch": 0.5597353265531185,
"grad_norm": 1.5118303298950195,
"learning_rate": 8.461860885303116e-06,
"loss": 0.5647146105766296,
"step": 571
},
{
"epoch": 0.5607155985786055,
"grad_norm": 0.6467048525810242,
"learning_rate": 8.45613232176134e-06,
"loss": 0.547932505607605,
"step": 572
},
{
"epoch": 0.5616958706040927,
"grad_norm": 20.77583122253418,
"learning_rate": 8.450395057410561e-06,
"loss": 0.6273069381713867,
"step": 573
},
{
"epoch": 0.5626761426295797,
"grad_norm": 0.635107159614563,
"learning_rate": 8.444649106694335e-06,
"loss": 0.650834321975708,
"step": 574
},
{
"epoch": 0.5636564146550668,
"grad_norm": 0.7083423733711243,
"learning_rate": 8.438894484078086e-06,
"loss": 0.5039582848548889,
"step": 575
},
{
"epoch": 0.5646366866805539,
"grad_norm": 16.103914260864258,
"learning_rate": 8.433131204049067e-06,
"loss": 0.6886800527572632,
"step": 576
},
{
"epoch": 0.5656169587060409,
"grad_norm": 0.545829713344574,
"learning_rate": 8.427359281116335e-06,
"loss": 0.5469794273376465,
"step": 577
},
{
"epoch": 0.566597230731528,
"grad_norm": 1.1052147150039673,
"learning_rate": 8.421578729810693e-06,
"loss": 0.6619667410850525,
"step": 578
},
{
"epoch": 0.567577502757015,
"grad_norm": 1.4017364978790283,
"learning_rate": 8.415789564684673e-06,
"loss": 0.7482278347015381,
"step": 579
},
{
"epoch": 0.5685577747825021,
"grad_norm": 2.920403480529785,
"learning_rate": 8.409991800312493e-06,
"loss": 0.5967468023300171,
"step": 580
},
{
"epoch": 0.5695380468079893,
"grad_norm": 0.6070059537887573,
"learning_rate": 8.404185451290017e-06,
"loss": 0.6574193239212036,
"step": 581
},
{
"epoch": 0.5705183188334763,
"grad_norm": 0.7417165637016296,
"learning_rate": 8.398370532234723e-06,
"loss": 0.7415742874145508,
"step": 582
},
{
"epoch": 0.5714985908589634,
"grad_norm": 0.6569010019302368,
"learning_rate": 8.392547057785662e-06,
"loss": 0.4218091070652008,
"step": 583
},
{
"epoch": 0.5724788628844505,
"grad_norm": 0.6027299165725708,
"learning_rate": 8.386715042603424e-06,
"loss": 0.5996318459510803,
"step": 584
},
{
"epoch": 0.5734591349099375,
"grad_norm": 0.7054585218429565,
"learning_rate": 8.380874501370098e-06,
"loss": 0.5750961303710938,
"step": 585
},
{
"epoch": 0.5744394069354246,
"grad_norm": 0.5906835198402405,
"learning_rate": 8.375025448789242e-06,
"loss": 0.5237380266189575,
"step": 586
},
{
"epoch": 0.5754196789609116,
"grad_norm": 0.8131166696548462,
"learning_rate": 8.36916789958584e-06,
"loss": 0.5778658986091614,
"step": 587
},
{
"epoch": 0.5763999509863987,
"grad_norm": 0.8021810054779053,
"learning_rate": 8.363301868506264e-06,
"loss": 0.6412160396575928,
"step": 588
},
{
"epoch": 0.5773802230118859,
"grad_norm": 0.5896162390708923,
"learning_rate": 8.357427370318239e-06,
"loss": 0.558003306388855,
"step": 589
},
{
"epoch": 0.5783604950373729,
"grad_norm": 0.5071391463279724,
"learning_rate": 8.351544419810808e-06,
"loss": 0.529178261756897,
"step": 590
},
{
"epoch": 0.57934076706286,
"grad_norm": 0.7972590923309326,
"learning_rate": 8.345653031794292e-06,
"loss": 0.5976440906524658,
"step": 591
},
{
"epoch": 0.580321039088347,
"grad_norm": 0.5813724994659424,
"learning_rate": 8.339753221100254e-06,
"loss": 0.6044202446937561,
"step": 592
},
{
"epoch": 0.5813013111138341,
"grad_norm": 0.780792236328125,
"learning_rate": 8.33384500258146e-06,
"loss": 0.7718487977981567,
"step": 593
},
{
"epoch": 0.5822815831393212,
"grad_norm": 1.2205802202224731,
"learning_rate": 8.327928391111841e-06,
"loss": 0.5783860683441162,
"step": 594
},
{
"epoch": 0.5832618551648082,
"grad_norm": 0.8029574155807495,
"learning_rate": 8.322003401586463e-06,
"loss": 0.6617282032966614,
"step": 595
},
{
"epoch": 0.5842421271902953,
"grad_norm": 0.7669466733932495,
"learning_rate": 8.316070048921477e-06,
"loss": 0.5271584391593933,
"step": 596
},
{
"epoch": 0.5852223992157823,
"grad_norm": 1.1002589464187622,
"learning_rate": 8.310128348054093e-06,
"loss": 0.46577566862106323,
"step": 597
},
{
"epoch": 0.5862026712412695,
"grad_norm": 0.6689755320549011,
"learning_rate": 8.304178313942536e-06,
"loss": 0.5580577254295349,
"step": 598
},
{
"epoch": 0.5871829432667566,
"grad_norm": 0.6669601798057556,
"learning_rate": 8.298219961566008e-06,
"loss": 0.5831491947174072,
"step": 599
},
{
"epoch": 0.5881632152922436,
"grad_norm": 3.1721479892730713,
"learning_rate": 8.292253305924656e-06,
"loss": 0.6172327995300293,
"step": 600
},
{
"epoch": 0.5891434873177307,
"grad_norm": 0.9356026649475098,
"learning_rate": 8.286278362039527e-06,
"loss": 0.6560968160629272,
"step": 601
},
{
"epoch": 0.5901237593432177,
"grad_norm": 0.6569904685020447,
"learning_rate": 8.280295144952537e-06,
"loss": 0.659455418586731,
"step": 602
},
{
"epoch": 0.5911040313687048,
"grad_norm": 1.4484615325927734,
"learning_rate": 8.274303669726427e-06,
"loss": 0.5224281549453735,
"step": 603
},
{
"epoch": 0.5920843033941919,
"grad_norm": 0.5144608616828918,
"learning_rate": 8.268303951444727e-06,
"loss": 0.5236537456512451,
"step": 604
},
{
"epoch": 0.5930645754196789,
"grad_norm": 0.6344110369682312,
"learning_rate": 8.262296005211722e-06,
"loss": 0.497511088848114,
"step": 605
},
{
"epoch": 0.5940448474451661,
"grad_norm": 9.12228775024414,
"learning_rate": 8.25627984615241e-06,
"loss": 0.6992365121841431,
"step": 606
},
{
"epoch": 0.5950251194706531,
"grad_norm": 1.136747121810913,
"learning_rate": 8.250255489412464e-06,
"loss": 0.5866008996963501,
"step": 607
},
{
"epoch": 0.5960053914961402,
"grad_norm": 0.5450854301452637,
"learning_rate": 8.244222950158194e-06,
"loss": 0.6557418704032898,
"step": 608
},
{
"epoch": 0.5969856635216273,
"grad_norm": 0.633201003074646,
"learning_rate": 8.238182243576512e-06,
"loss": 0.48425519466400146,
"step": 609
},
{
"epoch": 0.5979659355471143,
"grad_norm": 0.5643640756607056,
"learning_rate": 8.232133384874887e-06,
"loss": 0.5908252000808716,
"step": 610
},
{
"epoch": 0.5989462075726014,
"grad_norm": 0.5243382453918457,
"learning_rate": 8.226076389281316e-06,
"loss": 0.6378910541534424,
"step": 611
},
{
"epoch": 0.5999264795980884,
"grad_norm": 4.544177532196045,
"learning_rate": 8.220011272044276e-06,
"loss": 0.6232719421386719,
"step": 612
},
{
"epoch": 0.6009067516235755,
"grad_norm": 0.5729827284812927,
"learning_rate": 8.213938048432697e-06,
"loss": 0.5917630791664124,
"step": 613
},
{
"epoch": 0.6018870236490627,
"grad_norm": 2.7956576347351074,
"learning_rate": 8.207856733735908e-06,
"loss": 0.540428638458252,
"step": 614
},
{
"epoch": 0.6028672956745497,
"grad_norm": 0.5147668719291687,
"learning_rate": 8.201767343263612e-06,
"loss": 0.5766816139221191,
"step": 615
},
{
"epoch": 0.6038475677000368,
"grad_norm": 0.5646007657051086,
"learning_rate": 8.195669892345842e-06,
"loss": 0.5441969633102417,
"step": 616
},
{
"epoch": 0.6048278397255238,
"grad_norm": 0.8327797651290894,
"learning_rate": 8.189564396332927e-06,
"loss": 0.5784144997596741,
"step": 617
},
{
"epoch": 0.6058081117510109,
"grad_norm": 4.997214317321777,
"learning_rate": 8.183450870595443e-06,
"loss": 0.5072758793830872,
"step": 618
},
{
"epoch": 0.606788383776498,
"grad_norm": 0.7011016011238098,
"learning_rate": 8.177329330524182e-06,
"loss": 0.560646653175354,
"step": 619
},
{
"epoch": 0.607768655801985,
"grad_norm": 0.6436876058578491,
"learning_rate": 8.171199791530117e-06,
"loss": 0.5915713310241699,
"step": 620
},
{
"epoch": 0.6087489278274721,
"grad_norm": 1.0767780542373657,
"learning_rate": 8.165062269044353e-06,
"loss": 0.5400211811065674,
"step": 621
},
{
"epoch": 0.6097291998529591,
"grad_norm": 2.081508159637451,
"learning_rate": 8.158916778518097e-06,
"loss": 0.6144669055938721,
"step": 622
},
{
"epoch": 0.6107094718784463,
"grad_norm": 0.7203176617622375,
"learning_rate": 8.152763335422612e-06,
"loss": 0.5864992141723633,
"step": 623
},
{
"epoch": 0.6116897439039334,
"grad_norm": 0.673382580280304,
"learning_rate": 8.146601955249187e-06,
"loss": 0.5887113809585571,
"step": 624
},
{
"epoch": 0.6126700159294204,
"grad_norm": 0.6209499835968018,
"learning_rate": 8.140432653509089e-06,
"loss": 0.5816879868507385,
"step": 625
},
{
"epoch": 0.6136502879549075,
"grad_norm": 0.8314803242683411,
"learning_rate": 8.134255445733524e-06,
"loss": 0.6866248846054077,
"step": 626
},
{
"epoch": 0.6146305599803945,
"grad_norm": 0.45962321758270264,
"learning_rate": 8.128070347473609e-06,
"loss": 0.5957669019699097,
"step": 627
},
{
"epoch": 0.6156108320058816,
"grad_norm": 0.5579507350921631,
"learning_rate": 8.121877374300318e-06,
"loss": 0.5307108163833618,
"step": 628
},
{
"epoch": 0.6165911040313687,
"grad_norm": 0.5649368166923523,
"learning_rate": 8.115676541804456e-06,
"loss": 0.5324884653091431,
"step": 629
},
{
"epoch": 0.6175713760568557,
"grad_norm": 0.9094960689544678,
"learning_rate": 8.109467865596612e-06,
"loss": 0.5494385957717896,
"step": 630
},
{
"epoch": 0.6185516480823429,
"grad_norm": 0.7991915941238403,
"learning_rate": 8.10325136130712e-06,
"loss": 0.5825810432434082,
"step": 631
},
{
"epoch": 0.6195319201078299,
"grad_norm": 0.5570130944252014,
"learning_rate": 8.097027044586021e-06,
"loss": 0.7477834224700928,
"step": 632
},
{
"epoch": 0.620512192133317,
"grad_norm": 1.9130502939224243,
"learning_rate": 8.090794931103026e-06,
"loss": 0.552212655544281,
"step": 633
},
{
"epoch": 0.6214924641588041,
"grad_norm": 0.7020754814147949,
"learning_rate": 8.084555036547475e-06,
"loss": 0.6400367021560669,
"step": 634
},
{
"epoch": 0.6224727361842911,
"grad_norm": 0.7942973375320435,
"learning_rate": 8.078307376628292e-06,
"loss": 0.4825086295604706,
"step": 635
},
{
"epoch": 0.6234530082097782,
"grad_norm": 0.7500221133232117,
"learning_rate": 8.072051967073955e-06,
"loss": 0.574561595916748,
"step": 636
},
{
"epoch": 0.6244332802352652,
"grad_norm": 2.0203747749328613,
"learning_rate": 8.065788823632451e-06,
"loss": 0.4386993944644928,
"step": 637
},
{
"epoch": 0.6254135522607523,
"grad_norm": 1.349569320678711,
"learning_rate": 8.059517962071234e-06,
"loss": 0.47828274965286255,
"step": 638
},
{
"epoch": 0.6263938242862395,
"grad_norm": 49.53786087036133,
"learning_rate": 8.053239398177191e-06,
"loss": 0.5649581551551819,
"step": 639
},
{
"epoch": 0.6273740963117265,
"grad_norm": 0.5078736543655396,
"learning_rate": 8.046953147756601e-06,
"loss": 0.6091820001602173,
"step": 640
},
{
"epoch": 0.6283543683372136,
"grad_norm": 0.5322301387786865,
"learning_rate": 8.04065922663509e-06,
"loss": 0.49557721614837646,
"step": 641
},
{
"epoch": 0.6293346403627007,
"grad_norm": 0.6344025731086731,
"learning_rate": 8.034357650657597e-06,
"loss": 0.7918206453323364,
"step": 642
},
{
"epoch": 0.6303149123881877,
"grad_norm": 4.8755621910095215,
"learning_rate": 8.028048435688333e-06,
"loss": 0.5421925187110901,
"step": 643
},
{
"epoch": 0.6312951844136748,
"grad_norm": 6.257164001464844,
"learning_rate": 8.02173159761074e-06,
"loss": 0.6071158647537231,
"step": 644
},
{
"epoch": 0.6322754564391618,
"grad_norm": 0.6894906163215637,
"learning_rate": 8.015407152327448e-06,
"loss": 0.6801676750183105,
"step": 645
},
{
"epoch": 0.633255728464649,
"grad_norm": 1.5630754232406616,
"learning_rate": 8.009075115760243e-06,
"loss": 0.7003681063652039,
"step": 646
},
{
"epoch": 0.6342360004901361,
"grad_norm": 0.5957940220832825,
"learning_rate": 8.002735503850016e-06,
"loss": 0.5458991527557373,
"step": 647
},
{
"epoch": 0.6352162725156231,
"grad_norm": 0.6229478716850281,
"learning_rate": 7.996388332556735e-06,
"loss": 0.657683253288269,
"step": 648
},
{
"epoch": 0.6361965445411102,
"grad_norm": 1.3712888956069946,
"learning_rate": 7.990033617859396e-06,
"loss": 0.5221937894821167,
"step": 649
},
{
"epoch": 0.6371768165665972,
"grad_norm": 0.6126247048377991,
"learning_rate": 7.983671375755986e-06,
"loss": 0.5948827862739563,
"step": 650
},
{
"epoch": 0.6381570885920843,
"grad_norm": 0.7009912133216858,
"learning_rate": 7.97730162226344e-06,
"loss": 0.6552862524986267,
"step": 651
},
{
"epoch": 0.6391373606175714,
"grad_norm": 0.5230313539505005,
"learning_rate": 7.970924373417608e-06,
"loss": 0.6374645829200745,
"step": 652
},
{
"epoch": 0.6401176326430584,
"grad_norm": 0.5560128688812256,
"learning_rate": 7.964539645273204e-06,
"loss": 0.5496177077293396,
"step": 653
},
{
"epoch": 0.6410979046685455,
"grad_norm": 0.410652756690979,
"learning_rate": 7.958147453903773e-06,
"loss": 0.47082626819610596,
"step": 654
},
{
"epoch": 0.6420781766940326,
"grad_norm": 1.357462763786316,
"learning_rate": 7.951747815401651e-06,
"loss": 0.5593617558479309,
"step": 655
},
{
"epoch": 0.6430584487195197,
"grad_norm": 0.8234784603118896,
"learning_rate": 7.945340745877919e-06,
"loss": 0.6077075600624084,
"step": 656
},
{
"epoch": 0.6440387207450068,
"grad_norm": 0.8372496366500854,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6533162593841553,
"step": 657
},
{
"epoch": 0.6450189927704938,
"grad_norm": 0.6667627096176147,
"learning_rate": 7.932504378303452e-06,
"loss": 0.7186424732208252,
"step": 658
},
{
"epoch": 0.6459992647959809,
"grad_norm": 0.49903690814971924,
"learning_rate": 7.92607511256826e-06,
"loss": 0.5159540176391602,
"step": 659
},
{
"epoch": 0.6469795368214679,
"grad_norm": 0.5543757081031799,
"learning_rate": 7.919638480442453e-06,
"loss": 0.582042932510376,
"step": 660
},
{
"epoch": 0.647959808846955,
"grad_norm": 0.559472918510437,
"learning_rate": 7.913194498130252e-06,
"loss": 0.5199068784713745,
"step": 661
},
{
"epoch": 0.6489400808724421,
"grad_norm": 0.5725347995758057,
"learning_rate": 7.90674318185437e-06,
"loss": 0.6477447748184204,
"step": 662
},
{
"epoch": 0.6499203528979292,
"grad_norm": 0.5415868759155273,
"learning_rate": 7.900284547855992e-06,
"loss": 0.626210629940033,
"step": 663
},
{
"epoch": 0.6509006249234163,
"grad_norm": 0.536611795425415,
"learning_rate": 7.893818612394717e-06,
"loss": 0.5620501041412354,
"step": 664
},
{
"epoch": 0.6518808969489033,
"grad_norm": 9.495330810546875,
"learning_rate": 7.887345391748533e-06,
"loss": 0.557387113571167,
"step": 665
},
{
"epoch": 0.6528611689743904,
"grad_norm": 0.8832949995994568,
"learning_rate": 7.880864902213765e-06,
"loss": 0.707818865776062,
"step": 666
},
{
"epoch": 0.6538414409998775,
"grad_norm": 0.591690182685852,
"learning_rate": 7.874377160105037e-06,
"loss": 0.5649204254150391,
"step": 667
},
{
"epoch": 0.6548217130253645,
"grad_norm": 0.7255027294158936,
"learning_rate": 7.86788218175523e-06,
"loss": 0.6381409764289856,
"step": 668
},
{
"epoch": 0.6558019850508516,
"grad_norm": 0.6787216067314148,
"learning_rate": 7.861379983515449e-06,
"loss": 0.6573998928070068,
"step": 669
},
{
"epoch": 0.6567822570763386,
"grad_norm": 0.6130008101463318,
"learning_rate": 7.854870581754967e-06,
"loss": 0.5367627143859863,
"step": 670
},
{
"epoch": 0.6577625291018258,
"grad_norm": 0.6828269362449646,
"learning_rate": 7.848353992861195e-06,
"loss": 0.6018369197845459,
"step": 671
},
{
"epoch": 0.6587428011273129,
"grad_norm": 0.48248061537742615,
"learning_rate": 7.841830233239638e-06,
"loss": 0.5254629850387573,
"step": 672
},
{
"epoch": 0.6597230731527999,
"grad_norm": 0.6929543614387512,
"learning_rate": 7.835299319313854e-06,
"loss": 0.540311336517334,
"step": 673
},
{
"epoch": 0.660703345178287,
"grad_norm": 0.5000416040420532,
"learning_rate": 7.82876126752541e-06,
"loss": 0.5304262638092041,
"step": 674
},
{
"epoch": 0.661683617203774,
"grad_norm": 0.4851613938808441,
"learning_rate": 7.822216094333847e-06,
"loss": 0.4816325902938843,
"step": 675
},
{
"epoch": 0.6626638892292611,
"grad_norm": 0.6733199954032898,
"learning_rate": 7.815663816216626e-06,
"loss": 0.6401560306549072,
"step": 676
},
{
"epoch": 0.6636441612547482,
"grad_norm": 0.6005511283874512,
"learning_rate": 7.8091044496691e-06,
"loss": 0.5745944976806641,
"step": 677
},
{
"epoch": 0.6646244332802352,
"grad_norm": 0.4921104609966278,
"learning_rate": 7.80253801120447e-06,
"loss": 0.5637980699539185,
"step": 678
},
{
"epoch": 0.6656047053057224,
"grad_norm": 2.315213918685913,
"learning_rate": 7.795964517353734e-06,
"loss": 0.5256866216659546,
"step": 679
},
{
"epoch": 0.6665849773312094,
"grad_norm": 0.7630632519721985,
"learning_rate": 7.789383984665658e-06,
"loss": 0.5266314148902893,
"step": 680
},
{
"epoch": 0.6675652493566965,
"grad_norm": 0.5020211935043335,
"learning_rate": 7.782796429706721e-06,
"loss": 0.5310854911804199,
"step": 681
},
{
"epoch": 0.6685455213821836,
"grad_norm": 0.615732729434967,
"learning_rate": 7.776201869061086e-06,
"loss": 0.6091412305831909,
"step": 682
},
{
"epoch": 0.6695257934076706,
"grad_norm": 0.4945887625217438,
"learning_rate": 7.769600319330553e-06,
"loss": 0.524612545967102,
"step": 683
},
{
"epoch": 0.6705060654331577,
"grad_norm": 0.5611762404441833,
"learning_rate": 7.762991797134513e-06,
"loss": 0.5760433077812195,
"step": 684
},
{
"epoch": 0.6714863374586447,
"grad_norm": 0.5890234708786011,
"learning_rate": 7.756376319109917e-06,
"loss": 0.527816653251648,
"step": 685
},
{
"epoch": 0.6724666094841318,
"grad_norm": 1.0880018472671509,
"learning_rate": 7.74975390191122e-06,
"loss": 0.5751053094863892,
"step": 686
},
{
"epoch": 0.673446881509619,
"grad_norm": 0.538910448551178,
"learning_rate": 7.743124562210351e-06,
"loss": 0.7098783254623413,
"step": 687
},
{
"epoch": 0.674427153535106,
"grad_norm": 0.7028256058692932,
"learning_rate": 7.736488316696663e-06,
"loss": 0.45180726051330566,
"step": 688
},
{
"epoch": 0.6754074255605931,
"grad_norm": 0.6184514760971069,
"learning_rate": 7.729845182076896e-06,
"loss": 0.6585104465484619,
"step": 689
},
{
"epoch": 0.6763876975860801,
"grad_norm": 8.588869094848633,
"learning_rate": 7.723195175075136e-06,
"loss": 0.5983972549438477,
"step": 690
},
{
"epoch": 0.6773679696115672,
"grad_norm": 1.356168270111084,
"learning_rate": 7.716538312432767e-06,
"loss": 0.60518878698349,
"step": 691
},
{
"epoch": 0.6783482416370543,
"grad_norm": 3.287382125854492,
"learning_rate": 7.70987461090843e-06,
"loss": 0.6142452955245972,
"step": 692
},
{
"epoch": 0.6793285136625413,
"grad_norm": 0.5793871879577637,
"learning_rate": 7.703204087277989e-06,
"loss": 0.6304627060890198,
"step": 693
},
{
"epoch": 0.6803087856880284,
"grad_norm": 1.0979608297348022,
"learning_rate": 7.696526758334477e-06,
"loss": 0.6076656579971313,
"step": 694
},
{
"epoch": 0.6812890577135156,
"grad_norm": 0.43790462613105774,
"learning_rate": 7.689842640888063e-06,
"loss": 0.472328245639801,
"step": 695
},
{
"epoch": 0.6822693297390026,
"grad_norm": 0.5489725470542908,
"learning_rate": 7.683151751766005e-06,
"loss": 0.530362069606781,
"step": 696
},
{
"epoch": 0.6832496017644897,
"grad_norm": 0.600641667842865,
"learning_rate": 7.676454107812608e-06,
"loss": 0.7364434003829956,
"step": 697
},
{
"epoch": 0.6842298737899767,
"grad_norm": 0.5122668743133545,
"learning_rate": 7.669749725889182e-06,
"loss": 0.5904144644737244,
"step": 698
},
{
"epoch": 0.6852101458154638,
"grad_norm": 0.6352488398551941,
"learning_rate": 7.663038622873999e-06,
"loss": 0.6036946773529053,
"step": 699
},
{
"epoch": 0.6861904178409509,
"grad_norm": 1.1342604160308838,
"learning_rate": 7.656320815662257e-06,
"loss": 0.6565285325050354,
"step": 700
},
{
"epoch": 0.6871706898664379,
"grad_norm": 0.5251612663269043,
"learning_rate": 7.649596321166024e-06,
"loss": 0.5592460632324219,
"step": 701
},
{
"epoch": 0.688150961891925,
"grad_norm": 0.5946106910705566,
"learning_rate": 7.64286515631421e-06,
"loss": 0.514103889465332,
"step": 702
},
{
"epoch": 0.689131233917412,
"grad_norm": 0.5876750349998474,
"learning_rate": 7.636127338052513e-06,
"loss": 0.622397780418396,
"step": 703
},
{
"epoch": 0.6901115059428992,
"grad_norm": 0.9107208251953125,
"learning_rate": 7.629382883343381e-06,
"loss": 0.5825748443603516,
"step": 704
},
{
"epoch": 0.6910917779683863,
"grad_norm": 0.365422785282135,
"learning_rate": 7.622631809165972e-06,
"loss": 0.41773760318756104,
"step": 705
},
{
"epoch": 0.6920720499938733,
"grad_norm": 0.514562726020813,
"learning_rate": 7.615874132516108e-06,
"loss": 0.4052053689956665,
"step": 706
},
{
"epoch": 0.6930523220193604,
"grad_norm": 2.9211266040802,
"learning_rate": 7.60910987040623e-06,
"loss": 0.5205483436584473,
"step": 707
},
{
"epoch": 0.6940325940448474,
"grad_norm": 0.4625905156135559,
"learning_rate": 7.602339039865362e-06,
"loss": 0.5304774045944214,
"step": 708
},
{
"epoch": 0.6950128660703345,
"grad_norm": 0.5147523283958435,
"learning_rate": 7.595561657939061e-06,
"loss": 0.5491712093353271,
"step": 709
},
{
"epoch": 0.6959931380958216,
"grad_norm": 0.6532384157180786,
"learning_rate": 7.588777741689376e-06,
"loss": 0.6249186396598816,
"step": 710
},
{
"epoch": 0.6969734101213086,
"grad_norm": 0.5328655242919922,
"learning_rate": 7.5819873081948105e-06,
"loss": 0.5749078392982483,
"step": 711
},
{
"epoch": 0.6979536821467958,
"grad_norm": 0.5245689153671265,
"learning_rate": 7.575190374550272e-06,
"loss": 0.5168697834014893,
"step": 712
},
{
"epoch": 0.6989339541722828,
"grad_norm": 0.7289692163467407,
"learning_rate": 7.568386957867033e-06,
"loss": 0.5824940204620361,
"step": 713
},
{
"epoch": 0.6999142261977699,
"grad_norm": 1.814070463180542,
"learning_rate": 7.561577075272686e-06,
"loss": 0.6069308519363403,
"step": 714
},
{
"epoch": 0.700894498223257,
"grad_norm": 0.6956872940063477,
"learning_rate": 7.554760743911104e-06,
"loss": 0.5050805807113647,
"step": 715
},
{
"epoch": 0.701874770248744,
"grad_norm": 1.2698864936828613,
"learning_rate": 7.5479379809423905e-06,
"loss": 0.6088472008705139,
"step": 716
},
{
"epoch": 0.7028550422742311,
"grad_norm": 0.5213562250137329,
"learning_rate": 7.541108803542846e-06,
"loss": 0.5583027005195618,
"step": 717
},
{
"epoch": 0.7038353142997181,
"grad_norm": 0.5847312211990356,
"learning_rate": 7.534273228904916e-06,
"loss": 0.5814040899276733,
"step": 718
},
{
"epoch": 0.7048155863252052,
"grad_norm": 1.5454168319702148,
"learning_rate": 7.527431274237149e-06,
"loss": 0.45609384775161743,
"step": 719
},
{
"epoch": 0.7057958583506924,
"grad_norm": 0.5802167057991028,
"learning_rate": 7.52058295676416e-06,
"loss": 0.5678255558013916,
"step": 720
},
{
"epoch": 0.7067761303761794,
"grad_norm": 0.7184268236160278,
"learning_rate": 7.5137282937265796e-06,
"loss": 0.6155892610549927,
"step": 721
},
{
"epoch": 0.7077564024016665,
"grad_norm": 0.4999910593032837,
"learning_rate": 7.506867302381016e-06,
"loss": 0.5121070146560669,
"step": 722
},
{
"epoch": 0.7087366744271535,
"grad_norm": 0.4655957520008087,
"learning_rate": 7.500000000000001e-06,
"loss": 0.4872013032436371,
"step": 723
},
{
"epoch": 0.7097169464526406,
"grad_norm": 0.4562497138977051,
"learning_rate": 7.493126403871964e-06,
"loss": 0.46803396940231323,
"step": 724
},
{
"epoch": 0.7106972184781277,
"grad_norm": 0.6512648463249207,
"learning_rate": 7.486246531301178e-06,
"loss": 0.668323814868927,
"step": 725
},
{
"epoch": 0.7116774905036147,
"grad_norm": 0.7226558923721313,
"learning_rate": 7.479360399607707e-06,
"loss": 0.5372772216796875,
"step": 726
},
{
"epoch": 0.7126577625291018,
"grad_norm": 0.5560422539710999,
"learning_rate": 7.472468026127385e-06,
"loss": 0.5733708143234253,
"step": 727
},
{
"epoch": 0.7136380345545889,
"grad_norm": 0.6203676462173462,
"learning_rate": 7.465569428211752e-06,
"loss": 0.6157099604606628,
"step": 728
},
{
"epoch": 0.714618306580076,
"grad_norm": 0.4851949214935303,
"learning_rate": 7.45866462322802e-06,
"loss": 0.5322341918945312,
"step": 729
},
{
"epoch": 0.7155985786055631,
"grad_norm": 0.6843154430389404,
"learning_rate": 7.451753628559027e-06,
"loss": 0.47740745544433594,
"step": 730
},
{
"epoch": 0.7165788506310501,
"grad_norm": 0.7879361510276794,
"learning_rate": 7.444836461603195e-06,
"loss": 0.6443947553634644,
"step": 731
},
{
"epoch": 0.7175591226565372,
"grad_norm": 5.579281330108643,
"learning_rate": 7.437913139774482e-06,
"loss": 0.4795237183570862,
"step": 732
},
{
"epoch": 0.7185393946820242,
"grad_norm": 0.4848901927471161,
"learning_rate": 7.430983680502344e-06,
"loss": 0.4812300205230713,
"step": 733
},
{
"epoch": 0.7195196667075113,
"grad_norm": 0.5133047699928284,
"learning_rate": 7.424048101231687e-06,
"loss": 0.5148070454597473,
"step": 734
},
{
"epoch": 0.7204999387329984,
"grad_norm": 0.8202506899833679,
"learning_rate": 7.4171064194228196e-06,
"loss": 0.55904221534729,
"step": 735
},
{
"epoch": 0.7214802107584855,
"grad_norm": 0.8866989612579346,
"learning_rate": 7.410158652551421e-06,
"loss": 0.5958048701286316,
"step": 736
},
{
"epoch": 0.7224604827839726,
"grad_norm": 0.610833466053009,
"learning_rate": 7.403204818108487e-06,
"loss": 0.6387548446655273,
"step": 737
},
{
"epoch": 0.7234407548094596,
"grad_norm": 0.5252264738082886,
"learning_rate": 7.396244933600285e-06,
"loss": 0.6145691275596619,
"step": 738
},
{
"epoch": 0.7244210268349467,
"grad_norm": 0.8728129863739014,
"learning_rate": 7.3892790165483164e-06,
"loss": 0.5701698064804077,
"step": 739
},
{
"epoch": 0.7254012988604338,
"grad_norm": 5.201029300689697,
"learning_rate": 7.382307084489269e-06,
"loss": 0.5565284490585327,
"step": 740
},
{
"epoch": 0.7263815708859208,
"grad_norm": 0.5779921412467957,
"learning_rate": 7.3753291549749764e-06,
"loss": 0.5796201229095459,
"step": 741
},
{
"epoch": 0.7273618429114079,
"grad_norm": 0.560639500617981,
"learning_rate": 7.368345245572362e-06,
"loss": 0.45622915029525757,
"step": 742
},
{
"epoch": 0.7283421149368949,
"grad_norm": 0.5316780805587769,
"learning_rate": 7.361355373863415e-06,
"loss": 0.49733832478523254,
"step": 743
},
{
"epoch": 0.729322386962382,
"grad_norm": 0.49601390957832336,
"learning_rate": 7.354359557445125e-06,
"loss": 0.624125599861145,
"step": 744
},
{
"epoch": 0.7303026589878692,
"grad_norm": 0.5751706957817078,
"learning_rate": 7.347357813929455e-06,
"loss": 0.7379841804504395,
"step": 745
},
{
"epoch": 0.7312829310133562,
"grad_norm": 3.8335304260253906,
"learning_rate": 7.3403501609432835e-06,
"loss": 0.5610790848731995,
"step": 746
},
{
"epoch": 0.7322632030388433,
"grad_norm": 1.0456550121307373,
"learning_rate": 7.333336616128369e-06,
"loss": 0.5720086693763733,
"step": 747
},
{
"epoch": 0.7332434750643303,
"grad_norm": 0.5482881665229797,
"learning_rate": 7.326317197141304e-06,
"loss": 0.6755410432815552,
"step": 748
},
{
"epoch": 0.7342237470898174,
"grad_norm": 0.9965558648109436,
"learning_rate": 7.319291921653464e-06,
"loss": 0.6127687692642212,
"step": 749
},
{
"epoch": 0.7352040191153045,
"grad_norm": 0.8069666028022766,
"learning_rate": 7.312260807350975e-06,
"loss": 0.5598824620246887,
"step": 750
},
{
"epoch": 0.7361842911407915,
"grad_norm": 3.977120876312256,
"learning_rate": 7.305223871934657e-06,
"loss": 0.572356104850769,
"step": 751
},
{
"epoch": 0.7371645631662787,
"grad_norm": 0.563526451587677,
"learning_rate": 7.298181133119987e-06,
"loss": 0.5057710409164429,
"step": 752
},
{
"epoch": 0.7381448351917658,
"grad_norm": 0.5309803485870361,
"learning_rate": 7.291132608637053e-06,
"loss": 0.6270614862442017,
"step": 753
},
{
"epoch": 0.7391251072172528,
"grad_norm": 0.6150522828102112,
"learning_rate": 7.284078316230504e-06,
"loss": 0.5108097791671753,
"step": 754
},
{
"epoch": 0.7401053792427399,
"grad_norm": 0.4957369565963745,
"learning_rate": 7.2770182736595164e-06,
"loss": 0.5090512037277222,
"step": 755
},
{
"epoch": 0.7410856512682269,
"grad_norm": 0.6274080276489258,
"learning_rate": 7.269952498697734e-06,
"loss": 0.5452395081520081,
"step": 756
},
{
"epoch": 0.742065923293714,
"grad_norm": 0.8962954878807068,
"learning_rate": 7.262881009133242e-06,
"loss": 0.6746000647544861,
"step": 757
},
{
"epoch": 0.7430461953192011,
"grad_norm": 0.622025728225708,
"learning_rate": 7.255803822768504e-06,
"loss": 0.5007013082504272,
"step": 758
},
{
"epoch": 0.7440264673446881,
"grad_norm": 0.7198691368103027,
"learning_rate": 7.24872095742033e-06,
"loss": 0.5582835674285889,
"step": 759
},
{
"epoch": 0.7450067393701753,
"grad_norm": 0.7098339200019836,
"learning_rate": 7.241632430919823e-06,
"loss": 0.5809189677238464,
"step": 760
},
{
"epoch": 0.7459870113956623,
"grad_norm": 0.4255586564540863,
"learning_rate": 7.234538261112342e-06,
"loss": 0.4454438388347626,
"step": 761
},
{
"epoch": 0.7469672834211494,
"grad_norm": 0.5514088869094849,
"learning_rate": 7.227438465857448e-06,
"loss": 0.6373841762542725,
"step": 762
},
{
"epoch": 0.7479475554466365,
"grad_norm": 0.5340459942817688,
"learning_rate": 7.2203330630288714e-06,
"loss": 0.47046729922294617,
"step": 763
},
{
"epoch": 0.7489278274721235,
"grad_norm": 0.7507835030555725,
"learning_rate": 7.213222070514453e-06,
"loss": 0.758367657661438,
"step": 764
},
{
"epoch": 0.7499080994976106,
"grad_norm": 0.5694777965545654,
"learning_rate": 7.206105506216107e-06,
"loss": 0.5257217884063721,
"step": 765
},
{
"epoch": 0.7508883715230976,
"grad_norm": 0.6181869506835938,
"learning_rate": 7.198983388049779e-06,
"loss": 0.5684465169906616,
"step": 766
},
{
"epoch": 0.7518686435485847,
"grad_norm": 0.5873339772224426,
"learning_rate": 7.191855733945388e-06,
"loss": 0.5533320903778076,
"step": 767
},
{
"epoch": 0.7528489155740719,
"grad_norm": 0.4421772062778473,
"learning_rate": 7.1847225618467975e-06,
"loss": 0.4341365396976471,
"step": 768
},
{
"epoch": 0.7538291875995589,
"grad_norm": 0.5524192452430725,
"learning_rate": 7.177583889711763e-06,
"loss": 0.5365884304046631,
"step": 769
},
{
"epoch": 0.754809459625046,
"grad_norm": 0.7932672500610352,
"learning_rate": 7.170439735511876e-06,
"loss": 0.6242421269416809,
"step": 770
},
{
"epoch": 0.755789731650533,
"grad_norm": 0.4954005181789398,
"learning_rate": 7.163290117232542e-06,
"loss": 0.5581051707267761,
"step": 771
},
{
"epoch": 0.7567700036760201,
"grad_norm": 5.256842136383057,
"learning_rate": 7.1561350528729145e-06,
"loss": 0.5975151658058167,
"step": 772
},
{
"epoch": 0.7577502757015072,
"grad_norm": 0.432738721370697,
"learning_rate": 7.148974560445859e-06,
"loss": 0.4351341724395752,
"step": 773
},
{
"epoch": 0.7587305477269942,
"grad_norm": 0.5628976821899414,
"learning_rate": 7.1418086579779075e-06,
"loss": 0.5010636448860168,
"step": 774
},
{
"epoch": 0.7597108197524813,
"grad_norm": 0.5239070057868958,
"learning_rate": 7.1346373635092095e-06,
"loss": 0.5849080085754395,
"step": 775
},
{
"epoch": 0.7606910917779683,
"grad_norm": 0.6855177879333496,
"learning_rate": 7.127460695093493e-06,
"loss": 0.6382731795310974,
"step": 776
},
{
"epoch": 0.7616713638034555,
"grad_norm": 0.579451858997345,
"learning_rate": 7.12027867079801e-06,
"loss": 0.5916203260421753,
"step": 777
},
{
"epoch": 0.7626516358289426,
"grad_norm": 5.167208671569824,
"learning_rate": 7.113091308703498e-06,
"loss": 0.5930073261260986,
"step": 778
},
{
"epoch": 0.7636319078544296,
"grad_norm": 0.5476218461990356,
"learning_rate": 7.105898626904134e-06,
"loss": 0.5630462169647217,
"step": 779
},
{
"epoch": 0.7646121798799167,
"grad_norm": 0.5525639057159424,
"learning_rate": 7.098700643507485e-06,
"loss": 0.6280099153518677,
"step": 780
},
{
"epoch": 0.7655924519054037,
"grad_norm": 0.48628807067871094,
"learning_rate": 7.0914973766344645e-06,
"loss": 0.4585922062397003,
"step": 781
},
{
"epoch": 0.7665727239308908,
"grad_norm": 6.107976913452148,
"learning_rate": 7.0842888444192905e-06,
"loss": 0.7055581212043762,
"step": 782
},
{
"epoch": 0.7675529959563779,
"grad_norm": 1.1519224643707275,
"learning_rate": 7.0770750650094335e-06,
"loss": 0.633791446685791,
"step": 783
},
{
"epoch": 0.7685332679818649,
"grad_norm": 0.8850412368774414,
"learning_rate": 7.069856056565573e-06,
"loss": 0.5128419995307922,
"step": 784
},
{
"epoch": 0.7695135400073521,
"grad_norm": 0.8855915665626526,
"learning_rate": 7.062631837261556e-06,
"loss": 0.5462392568588257,
"step": 785
},
{
"epoch": 0.7704938120328391,
"grad_norm": 0.6897853016853333,
"learning_rate": 7.055402425284346e-06,
"loss": 0.48411428928375244,
"step": 786
},
{
"epoch": 0.7714740840583262,
"grad_norm": 0.6594675779342651,
"learning_rate": 7.048167838833977e-06,
"loss": 0.6231911778450012,
"step": 787
},
{
"epoch": 0.7724543560838133,
"grad_norm": 0.810693621635437,
"learning_rate": 7.040928096123516e-06,
"loss": 0.553763210773468,
"step": 788
},
{
"epoch": 0.7734346281093003,
"grad_norm": 0.5512497425079346,
"learning_rate": 7.033683215379002e-06,
"loss": 0.5670645236968994,
"step": 789
},
{
"epoch": 0.7744149001347874,
"grad_norm": 0.552123486995697,
"learning_rate": 7.026433214839416e-06,
"loss": 0.5537929534912109,
"step": 790
},
{
"epoch": 0.7753951721602744,
"grad_norm": 0.5684515833854675,
"learning_rate": 7.019178112756625e-06,
"loss": 0.461880087852478,
"step": 791
},
{
"epoch": 0.7763754441857615,
"grad_norm": 1.1695958375930786,
"learning_rate": 7.011917927395341e-06,
"loss": 0.5209773778915405,
"step": 792
},
{
"epoch": 0.7773557162112487,
"grad_norm": 0.976492166519165,
"learning_rate": 7.004652677033069e-06,
"loss": 0.5506386756896973,
"step": 793
},
{
"epoch": 0.7783359882367357,
"grad_norm": 0.7950183153152466,
"learning_rate": 6.997382379960072e-06,
"loss": 0.5871502757072449,
"step": 794
},
{
"epoch": 0.7793162602622228,
"grad_norm": 0.5692216753959656,
"learning_rate": 6.990107054479313e-06,
"loss": 0.5101208686828613,
"step": 795
},
{
"epoch": 0.7802965322877098,
"grad_norm": 0.4804418385028839,
"learning_rate": 6.9828267189064125e-06,
"loss": 0.546038031578064,
"step": 796
},
{
"epoch": 0.7812768043131969,
"grad_norm": 0.5595743656158447,
"learning_rate": 6.9755413915696105e-06,
"loss": 0.6369494199752808,
"step": 797
},
{
"epoch": 0.782257076338684,
"grad_norm": 0.9475952982902527,
"learning_rate": 6.968251090809708e-06,
"loss": 0.4539196193218231,
"step": 798
},
{
"epoch": 0.783237348364171,
"grad_norm": 0.5939152836799622,
"learning_rate": 6.960955834980028e-06,
"loss": 0.5116217732429504,
"step": 799
},
{
"epoch": 0.7842176203896581,
"grad_norm": 0.5029829740524292,
"learning_rate": 6.953655642446368e-06,
"loss": 0.6185356378555298,
"step": 800
},
{
"epoch": 0.7851978924151451,
"grad_norm": 0.5817827582359314,
"learning_rate": 6.946350531586959e-06,
"loss": 0.4881250262260437,
"step": 801
},
{
"epoch": 0.7861781644406323,
"grad_norm": 0.5651492476463318,
"learning_rate": 6.9390405207924014e-06,
"loss": 0.5443364381790161,
"step": 802
},
{
"epoch": 0.7871584364661194,
"grad_norm": 0.6898699998855591,
"learning_rate": 6.931725628465643e-06,
"loss": 0.5040018558502197,
"step": 803
},
{
"epoch": 0.7881387084916064,
"grad_norm": 0.8682623505592346,
"learning_rate": 6.9244058730219176e-06,
"loss": 0.6212958097457886,
"step": 804
},
{
"epoch": 0.7891189805170935,
"grad_norm": 0.3396937847137451,
"learning_rate": 6.917081272888697e-06,
"loss": 0.37508147954940796,
"step": 805
},
{
"epoch": 0.7900992525425806,
"grad_norm": 0.6038400530815125,
"learning_rate": 6.909751846505657e-06,
"loss": 0.4781612157821655,
"step": 806
},
{
"epoch": 0.7910795245680676,
"grad_norm": 0.5450059771537781,
"learning_rate": 6.902417612324615e-06,
"loss": 0.5534794330596924,
"step": 807
},
{
"epoch": 0.7920597965935547,
"grad_norm": 0.5613927245140076,
"learning_rate": 6.895078588809503e-06,
"loss": 0.5415093898773193,
"step": 808
},
{
"epoch": 0.7930400686190417,
"grad_norm": 0.824486255645752,
"learning_rate": 6.887734794436301e-06,
"loss": 0.4346888065338135,
"step": 809
},
{
"epoch": 0.7940203406445289,
"grad_norm": 0.48579537868499756,
"learning_rate": 6.880386247692999e-06,
"loss": 0.4570724070072174,
"step": 810
},
{
"epoch": 0.795000612670016,
"grad_norm": 0.4951340854167938,
"learning_rate": 6.873032967079562e-06,
"loss": 0.49875521659851074,
"step": 811
},
{
"epoch": 0.795980884695503,
"grad_norm": 0.5246482491493225,
"learning_rate": 6.865674971107859e-06,
"loss": 0.5968499779701233,
"step": 812
},
{
"epoch": 0.7969611567209901,
"grad_norm": 0.4873504936695099,
"learning_rate": 6.858312278301638e-06,
"loss": 0.5942188501358032,
"step": 813
},
{
"epoch": 0.7979414287464771,
"grad_norm": 0.46564146876335144,
"learning_rate": 6.850944907196472e-06,
"loss": 0.5106335878372192,
"step": 814
},
{
"epoch": 0.7989217007719642,
"grad_norm": 0.461176335811615,
"learning_rate": 6.8435728763397045e-06,
"loss": 0.4938886761665344,
"step": 815
},
{
"epoch": 0.7999019727974513,
"grad_norm": 0.5260149240493774,
"learning_rate": 6.836196204290417e-06,
"loss": 0.5463055372238159,
"step": 816
},
{
"epoch": 0.8008822448229383,
"grad_norm": 0.511102020740509,
"learning_rate": 6.828814909619374e-06,
"loss": 0.5764646530151367,
"step": 817
},
{
"epoch": 0.8018625168484255,
"grad_norm": 0.5147871375083923,
"learning_rate": 6.821429010908972e-06,
"loss": 0.5110339522361755,
"step": 818
},
{
"epoch": 0.8028427888739125,
"grad_norm": 0.456109881401062,
"learning_rate": 6.814038526753205e-06,
"loss": 0.5640659332275391,
"step": 819
},
{
"epoch": 0.8038230608993996,
"grad_norm": 4.816317558288574,
"learning_rate": 6.806643475757608e-06,
"loss": 0.5095980167388916,
"step": 820
},
{
"epoch": 0.8048033329248867,
"grad_norm": 0.618434727191925,
"learning_rate": 6.799243876539213e-06,
"loss": 0.5504512190818787,
"step": 821
},
{
"epoch": 0.8057836049503737,
"grad_norm": 0.4873816967010498,
"learning_rate": 6.7918397477265e-06,
"loss": 0.6287258863449097,
"step": 822
},
{
"epoch": 0.8067638769758608,
"grad_norm": 0.5904303789138794,
"learning_rate": 6.78443110795936e-06,
"loss": 0.5399340391159058,
"step": 823
},
{
"epoch": 0.8077441490013478,
"grad_norm": 0.4016742706298828,
"learning_rate": 6.77701797588903e-06,
"loss": 0.4996564984321594,
"step": 824
},
{
"epoch": 0.808724421026835,
"grad_norm": 0.5517071485519409,
"learning_rate": 6.76960037017806e-06,
"loss": 0.4603460431098938,
"step": 825
},
{
"epoch": 0.8097046930523221,
"grad_norm": 0.7395501136779785,
"learning_rate": 6.762178309500267e-06,
"loss": 0.6746194958686829,
"step": 826
},
{
"epoch": 0.8106849650778091,
"grad_norm": 0.5671557784080505,
"learning_rate": 6.75475181254068e-06,
"loss": 0.6013075113296509,
"step": 827
},
{
"epoch": 0.8116652371032962,
"grad_norm": 0.42417994141578674,
"learning_rate": 6.747320897995493e-06,
"loss": 0.5811848044395447,
"step": 828
},
{
"epoch": 0.8126455091287832,
"grad_norm": 0.48666852712631226,
"learning_rate": 6.739885584572026e-06,
"loss": 0.505912184715271,
"step": 829
},
{
"epoch": 0.8136257811542703,
"grad_norm": 0.485988050699234,
"learning_rate": 6.732445890988675e-06,
"loss": 0.6918061971664429,
"step": 830
},
{
"epoch": 0.8146060531797574,
"grad_norm": 0.6492186784744263,
"learning_rate": 6.725001835974854e-06,
"loss": 0.5441348552703857,
"step": 831
},
{
"epoch": 0.8155863252052444,
"grad_norm": 0.5266467928886414,
"learning_rate": 6.717553438270966e-06,
"loss": 0.4809187054634094,
"step": 832
},
{
"epoch": 0.8165665972307316,
"grad_norm": 1.7357603311538696,
"learning_rate": 6.710100716628345e-06,
"loss": 0.5574184060096741,
"step": 833
},
{
"epoch": 0.8175468692562186,
"grad_norm": 0.5953986048698425,
"learning_rate": 6.702643689809205e-06,
"loss": 0.43936827778816223,
"step": 834
},
{
"epoch": 0.8185271412817057,
"grad_norm": 0.549314558506012,
"learning_rate": 6.695182376586603e-06,
"loss": 0.502187967300415,
"step": 835
},
{
"epoch": 0.8195074133071928,
"grad_norm": 0.7117368578910828,
"learning_rate": 6.68771679574439e-06,
"loss": 0.6332178115844727,
"step": 836
},
{
"epoch": 0.8204876853326798,
"grad_norm": 0.5685935020446777,
"learning_rate": 6.680246966077151e-06,
"loss": 0.534398078918457,
"step": 837
},
{
"epoch": 0.8214679573581669,
"grad_norm": 5.580088138580322,
"learning_rate": 6.672772906390177e-06,
"loss": 0.4974400997161865,
"step": 838
},
{
"epoch": 0.8224482293836539,
"grad_norm": 0.5502187013626099,
"learning_rate": 6.665294635499404e-06,
"loss": 0.4470885097980499,
"step": 839
},
{
"epoch": 0.823428501409141,
"grad_norm": 0.7231400012969971,
"learning_rate": 6.65781217223137e-06,
"loss": 0.572614312171936,
"step": 840
},
{
"epoch": 0.8244087734346282,
"grad_norm": 0.4648838937282562,
"learning_rate": 6.650325535423166e-06,
"loss": 0.6567956209182739,
"step": 841
},
{
"epoch": 0.8253890454601152,
"grad_norm": 0.5262501239776611,
"learning_rate": 6.642834743922393e-06,
"loss": 0.7384023666381836,
"step": 842
},
{
"epoch": 0.8263693174856023,
"grad_norm": 0.45658308267593384,
"learning_rate": 6.635339816587109e-06,
"loss": 0.5313633680343628,
"step": 843
},
{
"epoch": 0.8273495895110893,
"grad_norm": 0.48368290066719055,
"learning_rate": 6.627840772285784e-06,
"loss": 0.5441750884056091,
"step": 844
},
{
"epoch": 0.8283298615365764,
"grad_norm": 0.5507257580757141,
"learning_rate": 6.6203376298972535e-06,
"loss": 0.5555803775787354,
"step": 845
},
{
"epoch": 0.8293101335620635,
"grad_norm": 0.6254823803901672,
"learning_rate": 6.612830408310671e-06,
"loss": 0.6455385684967041,
"step": 846
},
{
"epoch": 0.8302904055875505,
"grad_norm": 0.5825659036636353,
"learning_rate": 6.605319126425455e-06,
"loss": 0.5643920302391052,
"step": 847
},
{
"epoch": 0.8312706776130376,
"grad_norm": 0.5851500630378723,
"learning_rate": 6.5978038031512505e-06,
"loss": 0.4288056790828705,
"step": 848
},
{
"epoch": 0.8322509496385246,
"grad_norm": 0.6298871636390686,
"learning_rate": 6.590284457407876e-06,
"loss": 0.6632612943649292,
"step": 849
},
{
"epoch": 0.8332312216640118,
"grad_norm": 0.5183656811714172,
"learning_rate": 6.582761108125272e-06,
"loss": 0.4780927002429962,
"step": 850
},
{
"epoch": 0.8342114936894989,
"grad_norm": 0.4236246943473816,
"learning_rate": 6.5752337742434644e-06,
"loss": 0.43156588077545166,
"step": 851
},
{
"epoch": 0.8351917657149859,
"grad_norm": 7.04653787612915,
"learning_rate": 6.567702474712507e-06,
"loss": 0.5891600847244263,
"step": 852
},
{
"epoch": 0.836172037740473,
"grad_norm": 0.4767302870750427,
"learning_rate": 6.560167228492436e-06,
"loss": 0.49769872426986694,
"step": 853
},
{
"epoch": 0.83715230976596,
"grad_norm": 1.685430884361267,
"learning_rate": 6.552628054553225e-06,
"loss": 0.5213395357131958,
"step": 854
},
{
"epoch": 0.8381325817914471,
"grad_norm": 0.4784300923347473,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6092575788497925,
"step": 855
},
{
"epoch": 0.8391128538169342,
"grad_norm": 0.6260440945625305,
"learning_rate": 6.537537999446673e-06,
"loss": 0.543560266494751,
"step": 856
},
{
"epoch": 0.8400931258424212,
"grad_norm": 0.5237056612968445,
"learning_rate": 6.529987156268527e-06,
"loss": 0.5183249711990356,
"step": 857
},
{
"epoch": 0.8410733978679084,
"grad_norm": 0.5505848526954651,
"learning_rate": 6.522432461349536e-06,
"loss": 0.5145560503005981,
"step": 858
},
{
"epoch": 0.8420536698933954,
"grad_norm": 1.1515392065048218,
"learning_rate": 6.514873933708637e-06,
"loss": 0.6804383993148804,
"step": 859
},
{
"epoch": 0.8430339419188825,
"grad_norm": 0.506171703338623,
"learning_rate": 6.507311592374415e-06,
"loss": 0.5647867918014526,
"step": 860
},
{
"epoch": 0.8440142139443696,
"grad_norm": 0.6942906975746155,
"learning_rate": 6.499745456385054e-06,
"loss": 0.631658673286438,
"step": 861
},
{
"epoch": 0.8449944859698566,
"grad_norm": 0.6451912522315979,
"learning_rate": 6.4921755447882926e-06,
"loss": 0.51536625623703,
"step": 862
},
{
"epoch": 0.8459747579953437,
"grad_norm": 0.5695760846138,
"learning_rate": 6.484601876641375e-06,
"loss": 0.5741226673126221,
"step": 863
},
{
"epoch": 0.8469550300208308,
"grad_norm": 0.5603156685829163,
"learning_rate": 6.477024471011001e-06,
"loss": 0.5038372278213501,
"step": 864
},
{
"epoch": 0.8479353020463178,
"grad_norm": 0.6673778295516968,
"learning_rate": 6.469443346973281e-06,
"loss": 0.5298438668251038,
"step": 865
},
{
"epoch": 0.848915574071805,
"grad_norm": 0.5815482139587402,
"learning_rate": 6.461858523613684e-06,
"loss": 0.5317597985267639,
"step": 866
},
{
"epoch": 0.849895846097292,
"grad_norm": 0.5633111000061035,
"learning_rate": 6.454270020026996e-06,
"loss": 0.6263190507888794,
"step": 867
},
{
"epoch": 0.8508761181227791,
"grad_norm": 0.43612876534461975,
"learning_rate": 6.446677855317265e-06,
"loss": 0.4790484309196472,
"step": 868
},
{
"epoch": 0.8518563901482662,
"grad_norm": 0.5164619088172913,
"learning_rate": 6.439082048597755e-06,
"loss": 0.5105284452438354,
"step": 869
},
{
"epoch": 0.8528366621737532,
"grad_norm": 0.5535115003585815,
"learning_rate": 6.431482618990902e-06,
"loss": 0.5017833113670349,
"step": 870
},
{
"epoch": 0.8538169341992403,
"grad_norm": 1.2556020021438599,
"learning_rate": 6.423879585628262e-06,
"loss": 0.6469460725784302,
"step": 871
},
{
"epoch": 0.8547972062247273,
"grad_norm": 0.7312003970146179,
"learning_rate": 6.41627296765046e-06,
"loss": 0.6473737955093384,
"step": 872
},
{
"epoch": 0.8557774782502144,
"grad_norm": 8.008946418762207,
"learning_rate": 6.408662784207149e-06,
"loss": 0.5481289625167847,
"step": 873
},
{
"epoch": 0.8567577502757016,
"grad_norm": 0.5909143090248108,
"learning_rate": 6.401049054456957e-06,
"loss": 0.6918421983718872,
"step": 874
},
{
"epoch": 0.8577380223011886,
"grad_norm": 0.5711172223091125,
"learning_rate": 6.39343179756744e-06,
"loss": 0.46269646286964417,
"step": 875
},
{
"epoch": 0.8587182943266757,
"grad_norm": 0.6003888845443726,
"learning_rate": 6.385811032715031e-06,
"loss": 0.49102991819381714,
"step": 876
},
{
"epoch": 0.8596985663521627,
"grad_norm": 2.252908229827881,
"learning_rate": 6.378186779084996e-06,
"loss": 0.562686562538147,
"step": 877
},
{
"epoch": 0.8606788383776498,
"grad_norm": 0.5845773220062256,
"learning_rate": 6.370559055871389e-06,
"loss": 0.6050096154212952,
"step": 878
},
{
"epoch": 0.8616591104031369,
"grad_norm": 0.712838888168335,
"learning_rate": 6.362927882276991e-06,
"loss": 0.7435606718063354,
"step": 879
},
{
"epoch": 0.8626393824286239,
"grad_norm": 0.638928234577179,
"learning_rate": 6.355293277513271e-06,
"loss": 0.5028635263442993,
"step": 880
},
{
"epoch": 0.863619654454111,
"grad_norm": 0.9236575365066528,
"learning_rate": 6.34765526080034e-06,
"loss": 0.4573128819465637,
"step": 881
},
{
"epoch": 0.864599926479598,
"grad_norm": 0.841889500617981,
"learning_rate": 6.3400138513668955e-06,
"loss": 0.4685365557670593,
"step": 882
},
{
"epoch": 0.8655801985050852,
"grad_norm": 0.4550904929637909,
"learning_rate": 6.332369068450175e-06,
"loss": 0.5830044746398926,
"step": 883
},
{
"epoch": 0.8665604705305723,
"grad_norm": 0.42788735032081604,
"learning_rate": 6.324720931295913e-06,
"loss": 0.39731931686401367,
"step": 884
},
{
"epoch": 0.8675407425560593,
"grad_norm": 0.9267769455909729,
"learning_rate": 6.317069459158284e-06,
"loss": 0.620535135269165,
"step": 885
},
{
"epoch": 0.8685210145815464,
"grad_norm": 1.656376838684082,
"learning_rate": 6.3094146712998595e-06,
"loss": 0.5635170936584473,
"step": 886
},
{
"epoch": 0.8695012866070334,
"grad_norm": 0.5082754492759705,
"learning_rate": 6.301756586991561e-06,
"loss": 0.5327299237251282,
"step": 887
},
{
"epoch": 0.8704815586325205,
"grad_norm": 0.49824175238609314,
"learning_rate": 6.294095225512604e-06,
"loss": 0.4931308925151825,
"step": 888
},
{
"epoch": 0.8714618306580076,
"grad_norm": 0.5918230414390564,
"learning_rate": 6.286430606150458e-06,
"loss": 0.5891330242156982,
"step": 889
},
{
"epoch": 0.8724421026834946,
"grad_norm": 0.39389893412590027,
"learning_rate": 6.278762748200795e-06,
"loss": 0.46725448966026306,
"step": 890
},
{
"epoch": 0.8734223747089818,
"grad_norm": 0.5401540398597717,
"learning_rate": 6.271091670967437e-06,
"loss": 0.55974942445755,
"step": 891
},
{
"epoch": 0.8744026467344688,
"grad_norm": 0.45587074756622314,
"learning_rate": 6.2634173937623075e-06,
"loss": 0.49288398027420044,
"step": 892
},
{
"epoch": 0.8753829187599559,
"grad_norm": 1.2449387311935425,
"learning_rate": 6.255739935905396e-06,
"loss": 0.49451929330825806,
"step": 893
},
{
"epoch": 0.876363190785443,
"grad_norm": 0.4955168068408966,
"learning_rate": 6.248059316724693e-06,
"loss": 0.4612429141998291,
"step": 894
},
{
"epoch": 0.87734346281093,
"grad_norm": 0.4262677729129791,
"learning_rate": 6.240375555556145e-06,
"loss": 0.6077908277511597,
"step": 895
},
{
"epoch": 0.8783237348364171,
"grad_norm": 0.44123634696006775,
"learning_rate": 6.232688671743613e-06,
"loss": 0.48461630940437317,
"step": 896
},
{
"epoch": 0.8793040068619041,
"grad_norm": 1.8144631385803223,
"learning_rate": 6.22499868463882e-06,
"loss": 0.5334204435348511,
"step": 897
},
{
"epoch": 0.8802842788873912,
"grad_norm": 0.5446246266365051,
"learning_rate": 6.217305613601296e-06,
"loss": 0.6494865417480469,
"step": 898
},
{
"epoch": 0.8812645509128784,
"grad_norm": 0.6631665825843811,
"learning_rate": 6.209609477998339e-06,
"loss": 0.6008027791976929,
"step": 899
},
{
"epoch": 0.8822448229383654,
"grad_norm": 0.5320142507553101,
"learning_rate": 6.201910297204963e-06,
"loss": 0.48704683780670166,
"step": 900
},
{
"epoch": 0.8832250949638525,
"grad_norm": 1.1154892444610596,
"learning_rate": 6.194208090603845e-06,
"loss": 0.5158573389053345,
"step": 901
},
{
"epoch": 0.8842053669893395,
"grad_norm": 0.5630819201469421,
"learning_rate": 6.186502877585281e-06,
"loss": 0.5137242078781128,
"step": 902
},
{
"epoch": 0.8851856390148266,
"grad_norm": 0.555842936038971,
"learning_rate": 6.178794677547138e-06,
"loss": 0.5206568837165833,
"step": 903
},
{
"epoch": 0.8861659110403137,
"grad_norm": 0.47663378715515137,
"learning_rate": 6.171083509894795e-06,
"loss": 0.4792681336402893,
"step": 904
},
{
"epoch": 0.8871461830658007,
"grad_norm": 0.8589124083518982,
"learning_rate": 6.163369394041112e-06,
"loss": 0.49368637800216675,
"step": 905
},
{
"epoch": 0.8881264550912878,
"grad_norm": 0.8252849578857422,
"learning_rate": 6.155652349406366e-06,
"loss": 0.5167207717895508,
"step": 906
},
{
"epoch": 0.8891067271167749,
"grad_norm": 0.4597589671611786,
"learning_rate": 6.1479323954182055e-06,
"loss": 0.5020819902420044,
"step": 907
},
{
"epoch": 0.890086999142262,
"grad_norm": 0.49710938334465027,
"learning_rate": 6.140209551511609e-06,
"loss": 0.590738832950592,
"step": 908
},
{
"epoch": 0.8910672711677491,
"grad_norm": 1.4514707326889038,
"learning_rate": 6.132483837128823e-06,
"loss": 0.4894905388355255,
"step": 909
},
{
"epoch": 0.8920475431932361,
"grad_norm": 0.6761613488197327,
"learning_rate": 6.124755271719326e-06,
"loss": 0.769707977771759,
"step": 910
},
{
"epoch": 0.8930278152187232,
"grad_norm": 0.7030767798423767,
"learning_rate": 6.1170238747397715e-06,
"loss": 0.6154923439025879,
"step": 911
},
{
"epoch": 0.8940080872442102,
"grad_norm": 0.7982624769210815,
"learning_rate": 6.109289665653944e-06,
"loss": 0.5044841170310974,
"step": 912
},
{
"epoch": 0.8949883592696973,
"grad_norm": 0.7993609309196472,
"learning_rate": 6.101552663932704e-06,
"loss": 0.5166581869125366,
"step": 913
},
{
"epoch": 0.8959686312951844,
"grad_norm": 3.4683406352996826,
"learning_rate": 6.093812889053945e-06,
"loss": 0.585026741027832,
"step": 914
},
{
"epoch": 0.8969489033206715,
"grad_norm": 3.066880226135254,
"learning_rate": 6.08607036050254e-06,
"loss": 0.5106961131095886,
"step": 915
},
{
"epoch": 0.8979291753461586,
"grad_norm": 0.5521878600120544,
"learning_rate": 6.078325097770296e-06,
"loss": 0.5308998823165894,
"step": 916
},
{
"epoch": 0.8989094473716457,
"grad_norm": 0.9254027009010315,
"learning_rate": 6.070577120355903e-06,
"loss": 0.7255176305770874,
"step": 917
},
{
"epoch": 0.8998897193971327,
"grad_norm": 0.5461670756340027,
"learning_rate": 6.062826447764883e-06,
"loss": 0.5172099471092224,
"step": 918
},
{
"epoch": 0.9008699914226198,
"grad_norm": 1.2190165519714355,
"learning_rate": 6.055073099509549e-06,
"loss": 0.5512736439704895,
"step": 919
},
{
"epoch": 0.9018502634481068,
"grad_norm": 0.49891889095306396,
"learning_rate": 6.047317095108943e-06,
"loss": 0.6239607334136963,
"step": 920
},
{
"epoch": 0.9028305354735939,
"grad_norm": 0.41971200704574585,
"learning_rate": 6.039558454088796e-06,
"loss": 0.41604092717170715,
"step": 921
},
{
"epoch": 0.903810807499081,
"grad_norm": 0.7603155374526978,
"learning_rate": 6.031797195981481e-06,
"loss": 0.5952428579330444,
"step": 922
},
{
"epoch": 0.904791079524568,
"grad_norm": 0.48261648416519165,
"learning_rate": 6.024033340325954e-06,
"loss": 0.4906275272369385,
"step": 923
},
{
"epoch": 0.9057713515500552,
"grad_norm": 0.6290880441665649,
"learning_rate": 6.016266906667711e-06,
"loss": 0.5181742310523987,
"step": 924
},
{
"epoch": 0.9067516235755422,
"grad_norm": 0.9149678349494934,
"learning_rate": 6.0084979145587444e-06,
"loss": 0.522479236125946,
"step": 925
},
{
"epoch": 0.9077318956010293,
"grad_norm": 0.5393350124359131,
"learning_rate": 6.000726383557476e-06,
"loss": 0.5145924091339111,
"step": 926
},
{
"epoch": 0.9087121676265164,
"grad_norm": 1.5026754140853882,
"learning_rate": 5.9929523332287275e-06,
"loss": 0.5173094272613525,
"step": 927
},
{
"epoch": 0.9096924396520034,
"grad_norm": 0.5697621703147888,
"learning_rate": 5.985175783143666e-06,
"loss": 0.529559850692749,
"step": 928
},
{
"epoch": 0.9106727116774905,
"grad_norm": 0.5383663773536682,
"learning_rate": 5.977396752879742e-06,
"loss": 0.6133013963699341,
"step": 929
},
{
"epoch": 0.9116529837029775,
"grad_norm": 0.4323938488960266,
"learning_rate": 5.9696152620206575e-06,
"loss": 0.43817365169525146,
"step": 930
},
{
"epoch": 0.9126332557284647,
"grad_norm": 0.5100679993629456,
"learning_rate": 5.961831330156306e-06,
"loss": 0.6532332301139832,
"step": 931
},
{
"epoch": 0.9136135277539518,
"grad_norm": 0.6687671542167664,
"learning_rate": 5.954044976882725e-06,
"loss": 0.5830169916152954,
"step": 932
},
{
"epoch": 0.9145937997794388,
"grad_norm": 0.4441485106945038,
"learning_rate": 5.946256221802052e-06,
"loss": 0.5336486101150513,
"step": 933
},
{
"epoch": 0.9155740718049259,
"grad_norm": 0.5613795518875122,
"learning_rate": 5.938465084522468e-06,
"loss": 0.6328088045120239,
"step": 934
},
{
"epoch": 0.9165543438304129,
"grad_norm": 0.7554731369018555,
"learning_rate": 5.930671584658151e-06,
"loss": 0.4782413840293884,
"step": 935
},
{
"epoch": 0.9175346158559,
"grad_norm": 0.4980623126029968,
"learning_rate": 5.922875741829227e-06,
"loss": 0.5756092071533203,
"step": 936
},
{
"epoch": 0.9185148878813871,
"grad_norm": 0.5682260394096375,
"learning_rate": 5.915077575661723e-06,
"loss": 0.6436828970909119,
"step": 937
},
{
"epoch": 0.9194951599068741,
"grad_norm": 1.7436925172805786,
"learning_rate": 5.907277105787513e-06,
"loss": 0.4476461410522461,
"step": 938
},
{
"epoch": 0.9204754319323613,
"grad_norm": 0.45005002617836,
"learning_rate": 5.89947435184427e-06,
"loss": 0.4117186963558197,
"step": 939
},
{
"epoch": 0.9214557039578483,
"grad_norm": 0.5202357172966003,
"learning_rate": 5.891669333475416e-06,
"loss": 0.6808879375457764,
"step": 940
},
{
"epoch": 0.9224359759833354,
"grad_norm": 0.6051023006439209,
"learning_rate": 5.883862070330079e-06,
"loss": 0.6317353248596191,
"step": 941
},
{
"epoch": 0.9234162480088225,
"grad_norm": 0.4564392566680908,
"learning_rate": 5.876052582063031e-06,
"loss": 0.5340703725814819,
"step": 942
},
{
"epoch": 0.9243965200343095,
"grad_norm": 0.3938648998737335,
"learning_rate": 5.8682408883346535e-06,
"loss": 0.44789746403694153,
"step": 943
},
{
"epoch": 0.9253767920597966,
"grad_norm": 1.5355894565582275,
"learning_rate": 5.860427008810872e-06,
"loss": 0.5019229650497437,
"step": 944
},
{
"epoch": 0.9263570640852836,
"grad_norm": 0.4937191605567932,
"learning_rate": 5.85261096316312e-06,
"loss": 0.5558509230613708,
"step": 945
},
{
"epoch": 0.9273373361107707,
"grad_norm": 0.49368053674697876,
"learning_rate": 5.844792771068284e-06,
"loss": 0.5843484401702881,
"step": 946
},
{
"epoch": 0.9283176081362579,
"grad_norm": 0.4684690833091736,
"learning_rate": 5.8369724522086545e-06,
"loss": 0.529102087020874,
"step": 947
},
{
"epoch": 0.9292978801617449,
"grad_norm": 0.4931308329105377,
"learning_rate": 5.829150026271871e-06,
"loss": 0.6499549746513367,
"step": 948
},
{
"epoch": 0.930278152187232,
"grad_norm": 0.44310131669044495,
"learning_rate": 5.821325512950886e-06,
"loss": 0.5058345794677734,
"step": 949
},
{
"epoch": 0.931258424212719,
"grad_norm": 0.8803094029426575,
"learning_rate": 5.8134989319439004e-06,
"loss": 0.6572579741477966,
"step": 950
},
{
"epoch": 0.9322386962382061,
"grad_norm": 0.5434937477111816,
"learning_rate": 5.805670302954322e-06,
"loss": 0.5048413276672363,
"step": 951
},
{
"epoch": 0.9332189682636932,
"grad_norm": 1.510877013206482,
"learning_rate": 5.797839645690715e-06,
"loss": 0.5112013816833496,
"step": 952
},
{
"epoch": 0.9341992402891802,
"grad_norm": 0.5750903487205505,
"learning_rate": 5.79000697986675e-06,
"loss": 0.5129085779190063,
"step": 953
},
{
"epoch": 0.9351795123146673,
"grad_norm": 1.1053482294082642,
"learning_rate": 5.782172325201155e-06,
"loss": 0.5052158236503601,
"step": 954
},
{
"epoch": 0.9361597843401543,
"grad_norm": 0.6035041809082031,
"learning_rate": 5.774335701417662e-06,
"loss": 0.47337424755096436,
"step": 955
},
{
"epoch": 0.9371400563656415,
"grad_norm": 2.741614818572998,
"learning_rate": 5.766497128244963e-06,
"loss": 0.5770152807235718,
"step": 956
},
{
"epoch": 0.9381203283911286,
"grad_norm": 0.6729233264923096,
"learning_rate": 5.758656625416659e-06,
"loss": 0.5259652733802795,
"step": 957
},
{
"epoch": 0.9391006004166156,
"grad_norm": 0.720126211643219,
"learning_rate": 5.750814212671202e-06,
"loss": 0.610198974609375,
"step": 958
},
{
"epoch": 0.9400808724421027,
"grad_norm": 0.553177535533905,
"learning_rate": 5.7429699097518585e-06,
"loss": 0.5532019734382629,
"step": 959
},
{
"epoch": 0.9410611444675897,
"grad_norm": 0.7339922189712524,
"learning_rate": 5.735123736406655e-06,
"loss": 0.5227667093276978,
"step": 960
},
{
"epoch": 0.9420414164930768,
"grad_norm": 0.8617972731590271,
"learning_rate": 5.727275712388318e-06,
"loss": 0.5203966498374939,
"step": 961
},
{
"epoch": 0.9430216885185639,
"grad_norm": 0.5477783679962158,
"learning_rate": 5.719425857454242e-06,
"loss": 0.7109853029251099,
"step": 962
},
{
"epoch": 0.9440019605440509,
"grad_norm": 1.2344744205474854,
"learning_rate": 5.711574191366427e-06,
"loss": 0.4629364013671875,
"step": 963
},
{
"epoch": 0.9449822325695381,
"grad_norm": 2.135744571685791,
"learning_rate": 5.7037207338914326e-06,
"loss": 0.539191484451294,
"step": 964
},
{
"epoch": 0.9459625045950251,
"grad_norm": 3.0576016902923584,
"learning_rate": 5.695865504800328e-06,
"loss": 0.618647575378418,
"step": 965
},
{
"epoch": 0.9469427766205122,
"grad_norm": 0.9277212619781494,
"learning_rate": 5.688008523868646e-06,
"loss": 0.5556693077087402,
"step": 966
},
{
"epoch": 0.9479230486459993,
"grad_norm": 0.6881096959114075,
"learning_rate": 5.680149810876322e-06,
"loss": 0.6218986511230469,
"step": 967
},
{
"epoch": 0.9489033206714863,
"grad_norm": 0.5277224183082581,
"learning_rate": 5.6722893856076596e-06,
"loss": 0.4958314299583435,
"step": 968
},
{
"epoch": 0.9498835926969734,
"grad_norm": 0.6968176364898682,
"learning_rate": 5.664427267851271e-06,
"loss": 0.5367750525474548,
"step": 969
},
{
"epoch": 0.9508638647224604,
"grad_norm": 0.5265281200408936,
"learning_rate": 5.656563477400026e-06,
"loss": 0.5782604813575745,
"step": 970
},
{
"epoch": 0.9518441367479475,
"grad_norm": 0.5750949382781982,
"learning_rate": 5.648698034051009e-06,
"loss": 0.5680809020996094,
"step": 971
},
{
"epoch": 0.9528244087734347,
"grad_norm": 1.0558995008468628,
"learning_rate": 5.640830957605466e-06,
"loss": 0.5180156230926514,
"step": 972
},
{
"epoch": 0.9538046807989217,
"grad_norm": 0.5897397994995117,
"learning_rate": 5.632962267868747e-06,
"loss": 0.49604538083076477,
"step": 973
},
{
"epoch": 0.9547849528244088,
"grad_norm": 0.5551328063011169,
"learning_rate": 5.625091984650273e-06,
"loss": 0.5703722238540649,
"step": 974
},
{
"epoch": 0.9557652248498959,
"grad_norm": 0.6641311645507812,
"learning_rate": 5.617220127763474e-06,
"loss": 0.4712502360343933,
"step": 975
},
{
"epoch": 0.9567454968753829,
"grad_norm": 0.5024252533912659,
"learning_rate": 5.609346717025738e-06,
"loss": 0.5163110494613647,
"step": 976
},
{
"epoch": 0.95772576890087,
"grad_norm": 1.2409754991531372,
"learning_rate": 5.601471772258368e-06,
"loss": 0.5272206664085388,
"step": 977
},
{
"epoch": 0.958706040926357,
"grad_norm": 0.4743415117263794,
"learning_rate": 5.593595313286526e-06,
"loss": 0.4263104498386383,
"step": 978
},
{
"epoch": 0.9596863129518441,
"grad_norm": 0.4841744899749756,
"learning_rate": 5.585717359939192e-06,
"loss": 0.48904165625572205,
"step": 979
},
{
"epoch": 0.9606665849773313,
"grad_norm": 0.49298396706581116,
"learning_rate": 5.5778379320491e-06,
"loss": 0.4394080340862274,
"step": 980
},
{
"epoch": 0.9616468570028183,
"grad_norm": 0.5323792695999146,
"learning_rate": 5.569957049452703e-06,
"loss": 0.6521806120872498,
"step": 981
},
{
"epoch": 0.9626271290283054,
"grad_norm": 0.6238925457000732,
"learning_rate": 5.562074731990115e-06,
"loss": 0.4940195381641388,
"step": 982
},
{
"epoch": 0.9636074010537924,
"grad_norm": 0.46199095249176025,
"learning_rate": 5.5541909995050554e-06,
"loss": 0.5050048232078552,
"step": 983
},
{
"epoch": 0.9645876730792795,
"grad_norm": 0.7361982464790344,
"learning_rate": 5.546305871844816e-06,
"loss": 0.40546220541000366,
"step": 984
},
{
"epoch": 0.9655679451047666,
"grad_norm": 2.979095220565796,
"learning_rate": 5.538419368860196e-06,
"loss": 0.44479626417160034,
"step": 985
},
{
"epoch": 0.9665482171302536,
"grad_norm": 0.4303815960884094,
"learning_rate": 5.530531510405454e-06,
"loss": 0.4572027921676636,
"step": 986
},
{
"epoch": 0.9675284891557407,
"grad_norm": 0.5292936563491821,
"learning_rate": 5.522642316338268e-06,
"loss": 0.47948259115219116,
"step": 987
},
{
"epoch": 0.9685087611812278,
"grad_norm": 0.5669738054275513,
"learning_rate": 5.514751806519673e-06,
"loss": 0.5793241262435913,
"step": 988
},
{
"epoch": 0.9694890332067149,
"grad_norm": 0.6330426335334778,
"learning_rate": 5.506860000814017e-06,
"loss": 0.44303247332572937,
"step": 989
},
{
"epoch": 0.970469305232202,
"grad_norm": 0.45050469040870667,
"learning_rate": 5.498966919088914e-06,
"loss": 0.5273669958114624,
"step": 990
},
{
"epoch": 0.971449577257689,
"grad_norm": 0.4541574716567993,
"learning_rate": 5.491072581215186e-06,
"loss": 0.532174825668335,
"step": 991
},
{
"epoch": 0.9724298492831761,
"grad_norm": 0.48249125480651855,
"learning_rate": 5.48317700706682e-06,
"loss": 0.49193787574768066,
"step": 992
},
{
"epoch": 0.9734101213086631,
"grad_norm": 0.41582420468330383,
"learning_rate": 5.475280216520913e-06,
"loss": 0.4485260248184204,
"step": 993
},
{
"epoch": 0.9743903933341502,
"grad_norm": 0.41106855869293213,
"learning_rate": 5.467382229457629e-06,
"loss": 0.5209757685661316,
"step": 994
},
{
"epoch": 0.9753706653596373,
"grad_norm": 0.4153555929660797,
"learning_rate": 5.459483065760138e-06,
"loss": 0.5320965051651001,
"step": 995
},
{
"epoch": 0.9763509373851244,
"grad_norm": 0.44608160853385925,
"learning_rate": 5.451582745314576e-06,
"loss": 0.5319690704345703,
"step": 996
},
{
"epoch": 0.9773312094106115,
"grad_norm": 2.1951260566711426,
"learning_rate": 5.443681288009991e-06,
"loss": 0.5222268104553223,
"step": 997
},
{
"epoch": 0.9783114814360985,
"grad_norm": 0.689635694026947,
"learning_rate": 5.435778713738292e-06,
"loss": 0.5071982145309448,
"step": 998
},
{
"epoch": 0.9792917534615856,
"grad_norm": 0.5424144864082336,
"learning_rate": 5.4278750423942e-06,
"loss": 0.5717045068740845,
"step": 999
},
{
"epoch": 0.9802720254870727,
"grad_norm": 0.43961331248283386,
"learning_rate": 5.419970293875197e-06,
"loss": 0.44293683767318726,
"step": 1000
},
{
"epoch": 0.9812522975125597,
"grad_norm": 0.7518972754478455,
"learning_rate": 5.412064488081482e-06,
"loss": 0.533876895904541,
"step": 1001
},
{
"epoch": 0.9822325695380468,
"grad_norm": 0.4143977761268616,
"learning_rate": 5.404157644915907e-06,
"loss": 0.5120701193809509,
"step": 1002
},
{
"epoch": 0.9832128415635338,
"grad_norm": 0.5535058975219727,
"learning_rate": 5.396249784283943e-06,
"loss": 0.5277990102767944,
"step": 1003
},
{
"epoch": 0.984193113589021,
"grad_norm": 0.45615097880363464,
"learning_rate": 5.388340926093619e-06,
"loss": 0.4987753629684448,
"step": 1004
},
{
"epoch": 0.9851733856145081,
"grad_norm": 0.5017585754394531,
"learning_rate": 5.380431090255475e-06,
"loss": 0.5158082246780396,
"step": 1005
},
{
"epoch": 0.9861536576399951,
"grad_norm": 0.4945541322231293,
"learning_rate": 5.372520296682516e-06,
"loss": 0.6469739675521851,
"step": 1006
},
{
"epoch": 0.9871339296654822,
"grad_norm": 0.5786065459251404,
"learning_rate": 5.364608565290154e-06,
"loss": 0.45730066299438477,
"step": 1007
},
{
"epoch": 0.9881142016909692,
"grad_norm": 0.5092796683311462,
"learning_rate": 5.356695915996162e-06,
"loss": 0.5266056060791016,
"step": 1008
},
{
"epoch": 0.9890944737164563,
"grad_norm": 0.45691388845443726,
"learning_rate": 5.348782368720627e-06,
"loss": 0.5465075969696045,
"step": 1009
},
{
"epoch": 0.9900747457419434,
"grad_norm": 0.43501365184783936,
"learning_rate": 5.340867943385895e-06,
"loss": 0.5065740346908569,
"step": 1010
},
{
"epoch": 0.9910550177674304,
"grad_norm": 1.1877309083938599,
"learning_rate": 5.33295265991652e-06,
"loss": 0.5574820637702942,
"step": 1011
},
{
"epoch": 0.9920352897929176,
"grad_norm": 0.40967103838920593,
"learning_rate": 5.325036538239221e-06,
"loss": 0.470042884349823,
"step": 1012
},
{
"epoch": 0.9930155618184046,
"grad_norm": 0.4669896066188812,
"learning_rate": 5.317119598282823e-06,
"loss": 0.4594653248786926,
"step": 1013
},
{
"epoch": 0.9939958338438917,
"grad_norm": 0.7284072637557983,
"learning_rate": 5.3092018599782155e-06,
"loss": 0.571753203868866,
"step": 1014
},
{
"epoch": 0.9949761058693788,
"grad_norm": 1.7042508125305176,
"learning_rate": 5.301283343258293e-06,
"loss": 0.5708970427513123,
"step": 1015
},
{
"epoch": 0.9959563778948658,
"grad_norm": 1.265141487121582,
"learning_rate": 5.293364068057912e-06,
"loss": 0.5755418539047241,
"step": 1016
},
{
"epoch": 0.9969366499203529,
"grad_norm": 0.4056626856327057,
"learning_rate": 5.285444054313841e-06,
"loss": 0.3639076352119446,
"step": 1017
},
{
"epoch": 0.9979169219458399,
"grad_norm": 0.5877487659454346,
"learning_rate": 5.277523321964701e-06,
"loss": 0.6229694485664368,
"step": 1018
},
{
"epoch": 0.998897193971327,
"grad_norm": 0.41981130838394165,
"learning_rate": 5.26960189095093e-06,
"loss": 0.4316851496696472,
"step": 1019
},
{
"epoch": 0.9998774659968142,
"grad_norm": 0.6003679037094116,
"learning_rate": 5.2616797812147205e-06,
"loss": 0.4310376048088074,
"step": 1020
},
{
"epoch": 1.0,
"grad_norm": 1.5014209747314453,
"learning_rate": 5.253757012699972e-06,
"loss": 0.6824495792388916,
"step": 1021
}
],
"logging_steps": 1,
"max_steps": 2042,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.937962266327777e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}