9b-86 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
1f3a4bc verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 1804,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004434589800443459,
"grad_norm": 2.7257320880889893,
"learning_rate": 5.494505494505495e-08,
"loss": 1.8651859760284424,
"step": 2
},
{
"epoch": 0.008869179600886918,
"grad_norm": 6.67538595199585,
"learning_rate": 1.6483516483516484e-07,
"loss": 2.1317176818847656,
"step": 4
},
{
"epoch": 0.013303769401330377,
"grad_norm": 6.216647624969482,
"learning_rate": 2.7472527472527475e-07,
"loss": 1.904492735862732,
"step": 6
},
{
"epoch": 0.017738359201773836,
"grad_norm": 1.1432791948318481,
"learning_rate": 3.846153846153847e-07,
"loss": 1.8241561651229858,
"step": 8
},
{
"epoch": 0.022172949002217297,
"grad_norm": 2.2455904483795166,
"learning_rate": 4.945054945054946e-07,
"loss": 1.6350065469741821,
"step": 10
},
{
"epoch": 0.026607538802660754,
"grad_norm": 3.8747806549072266,
"learning_rate": 6.043956043956044e-07,
"loss": 2.1859989166259766,
"step": 12
},
{
"epoch": 0.031042128603104215,
"grad_norm": 4.237593173980713,
"learning_rate": 7.142857142857143e-07,
"loss": 1.943555235862732,
"step": 14
},
{
"epoch": 0.03547671840354767,
"grad_norm": 6.446155071258545,
"learning_rate": 8.241758241758242e-07,
"loss": 1.4925756454467773,
"step": 16
},
{
"epoch": 0.03991130820399113,
"grad_norm": 2.6784210205078125,
"learning_rate": 9.340659340659342e-07,
"loss": 1.4368367195129395,
"step": 18
},
{
"epoch": 0.04434589800443459,
"grad_norm": 4.474376201629639,
"learning_rate": 1.0439560439560442e-06,
"loss": 1.5969985723495483,
"step": 20
},
{
"epoch": 0.04878048780487805,
"grad_norm": 5.059698581695557,
"learning_rate": 1.153846153846154e-06,
"loss": 1.462158441543579,
"step": 22
},
{
"epoch": 0.05321507760532151,
"grad_norm": 1.1916331052780151,
"learning_rate": 1.2637362637362637e-06,
"loss": 1.5942579507827759,
"step": 24
},
{
"epoch": 0.057649667405764965,
"grad_norm": 1.0105680227279663,
"learning_rate": 1.3736263736263738e-06,
"loss": 1.4908994436264038,
"step": 26
},
{
"epoch": 0.06208425720620843,
"grad_norm": 1.906994104385376,
"learning_rate": 1.4835164835164837e-06,
"loss": 1.179060935974121,
"step": 28
},
{
"epoch": 0.06651884700665188,
"grad_norm": 1.2037031650543213,
"learning_rate": 1.5934065934065933e-06,
"loss": 1.3003807067871094,
"step": 30
},
{
"epoch": 0.07095343680709534,
"grad_norm": 7.113156318664551,
"learning_rate": 1.7032967032967034e-06,
"loss": 1.2082353830337524,
"step": 32
},
{
"epoch": 0.07538802660753881,
"grad_norm": 0.885746955871582,
"learning_rate": 1.8131868131868133e-06,
"loss": 1.5782675743103027,
"step": 34
},
{
"epoch": 0.07982261640798226,
"grad_norm": 3.321558713912964,
"learning_rate": 1.9230769230769234e-06,
"loss": 1.342716932296753,
"step": 36
},
{
"epoch": 0.08425720620842572,
"grad_norm": 1.9045021533966064,
"learning_rate": 2.032967032967033e-06,
"loss": 1.5886274576187134,
"step": 38
},
{
"epoch": 0.08869179600886919,
"grad_norm": 1.937488079071045,
"learning_rate": 2.1428571428571427e-06,
"loss": 0.8519856929779053,
"step": 40
},
{
"epoch": 0.09312638580931264,
"grad_norm": 1.281927227973938,
"learning_rate": 2.252747252747253e-06,
"loss": 1.4712433815002441,
"step": 42
},
{
"epoch": 0.0975609756097561,
"grad_norm": 1.415075421333313,
"learning_rate": 2.362637362637363e-06,
"loss": 1.4157054424285889,
"step": 44
},
{
"epoch": 0.10199556541019955,
"grad_norm": 1.424558162689209,
"learning_rate": 2.472527472527473e-06,
"loss": 1.428381323814392,
"step": 46
},
{
"epoch": 0.10643015521064302,
"grad_norm": 2.6216702461242676,
"learning_rate": 2.582417582417583e-06,
"loss": 1.6168309450149536,
"step": 48
},
{
"epoch": 0.11086474501108648,
"grad_norm": 1.449957013130188,
"learning_rate": 2.6923076923076923e-06,
"loss": 1.502392053604126,
"step": 50
},
{
"epoch": 0.11529933481152993,
"grad_norm": 1.0936357975006104,
"learning_rate": 2.8021978021978024e-06,
"loss": 1.4243361949920654,
"step": 52
},
{
"epoch": 0.1197339246119734,
"grad_norm": 1.5021100044250488,
"learning_rate": 2.9120879120879125e-06,
"loss": 1.4172306060791016,
"step": 54
},
{
"epoch": 0.12416851441241686,
"grad_norm": 2.1787075996398926,
"learning_rate": 3.021978021978022e-06,
"loss": 1.1664988994598389,
"step": 56
},
{
"epoch": 0.1286031042128603,
"grad_norm": 1.6643332242965698,
"learning_rate": 3.1318681318681323e-06,
"loss": 1.364538550376892,
"step": 58
},
{
"epoch": 0.13303769401330376,
"grad_norm": 0.6121675968170166,
"learning_rate": 3.2417582417582424e-06,
"loss": 1.415244221687317,
"step": 60
},
{
"epoch": 0.13747228381374724,
"grad_norm": 1.6771880388259888,
"learning_rate": 3.3516483516483516e-06,
"loss": 1.3695659637451172,
"step": 62
},
{
"epoch": 0.1419068736141907,
"grad_norm": 1.3066154718399048,
"learning_rate": 3.4615384615384617e-06,
"loss": 1.435889720916748,
"step": 64
},
{
"epoch": 0.14634146341463414,
"grad_norm": 1.6970858573913574,
"learning_rate": 3.5714285714285718e-06,
"loss": 1.4904870986938477,
"step": 66
},
{
"epoch": 0.15077605321507762,
"grad_norm": 1.193587303161621,
"learning_rate": 3.681318681318682e-06,
"loss": 1.3552013635635376,
"step": 68
},
{
"epoch": 0.15521064301552107,
"grad_norm": 1.7035346031188965,
"learning_rate": 3.7912087912087915e-06,
"loss": 1.3096868991851807,
"step": 70
},
{
"epoch": 0.15964523281596452,
"grad_norm": 1.0295628309249878,
"learning_rate": 3.901098901098901e-06,
"loss": 1.3637210130691528,
"step": 72
},
{
"epoch": 0.164079822616408,
"grad_norm": 1.0206137895584106,
"learning_rate": 4.010989010989012e-06,
"loss": 1.2611558437347412,
"step": 74
},
{
"epoch": 0.16851441241685144,
"grad_norm": 3.5194575786590576,
"learning_rate": 4.120879120879121e-06,
"loss": 1.3402124643325806,
"step": 76
},
{
"epoch": 0.1729490022172949,
"grad_norm": 1.0586298704147339,
"learning_rate": 4.230769230769231e-06,
"loss": 1.3869333267211914,
"step": 78
},
{
"epoch": 0.17738359201773837,
"grad_norm": 2.9942257404327393,
"learning_rate": 4.340659340659341e-06,
"loss": 1.3390436172485352,
"step": 80
},
{
"epoch": 0.18181818181818182,
"grad_norm": 1.4550483226776123,
"learning_rate": 4.45054945054945e-06,
"loss": 0.8444686532020569,
"step": 82
},
{
"epoch": 0.18625277161862527,
"grad_norm": 0.7003264427185059,
"learning_rate": 4.560439560439561e-06,
"loss": 1.0279695987701416,
"step": 84
},
{
"epoch": 0.19068736141906872,
"grad_norm": 4.848526954650879,
"learning_rate": 4.6703296703296706e-06,
"loss": 0.9698995351791382,
"step": 86
},
{
"epoch": 0.1951219512195122,
"grad_norm": 1.024747610092163,
"learning_rate": 4.780219780219781e-06,
"loss": 1.2624331712722778,
"step": 88
},
{
"epoch": 0.19955654101995565,
"grad_norm": 1.7162449359893799,
"learning_rate": 4.890109890109891e-06,
"loss": 1.392942190170288,
"step": 90
},
{
"epoch": 0.2039911308203991,
"grad_norm": 2.0316786766052246,
"learning_rate": 5e-06,
"loss": 1.229879379272461,
"step": 92
},
{
"epoch": 0.20842572062084258,
"grad_norm": 1.6212053298950195,
"learning_rate": 4.999984864490455e-06,
"loss": 1.2555553913116455,
"step": 94
},
{
"epoch": 0.21286031042128603,
"grad_norm": 2.102651357650757,
"learning_rate": 4.999939458165447e-06,
"loss": 1.8031316995620728,
"step": 96
},
{
"epoch": 0.21729490022172948,
"grad_norm": 1.7087279558181763,
"learning_rate": 4.999863781635863e-06,
"loss": 0.950995922088623,
"step": 98
},
{
"epoch": 0.22172949002217296,
"grad_norm": 1.2007180452346802,
"learning_rate": 4.999757835919841e-06,
"loss": 1.1569544076919556,
"step": 100
},
{
"epoch": 0.2261640798226164,
"grad_norm": 0.7960771322250366,
"learning_rate": 4.9996216224427495e-06,
"loss": 1.313308596611023,
"step": 102
},
{
"epoch": 0.23059866962305986,
"grad_norm": 1.0540348291397095,
"learning_rate": 4.999455143037178e-06,
"loss": 1.0472400188446045,
"step": 104
},
{
"epoch": 0.23503325942350334,
"grad_norm": 1.4287160634994507,
"learning_rate": 4.999258399942903e-06,
"loss": 1.2840181589126587,
"step": 106
},
{
"epoch": 0.2394678492239468,
"grad_norm": 1.540831208229065,
"learning_rate": 4.9990313958068645e-06,
"loss": 1.48462975025177,
"step": 108
},
{
"epoch": 0.24390243902439024,
"grad_norm": 1.6057487726211548,
"learning_rate": 4.998774133683127e-06,
"loss": 1.026124119758606,
"step": 110
},
{
"epoch": 0.24833702882483372,
"grad_norm": 3.355077028274536,
"learning_rate": 4.9984866170328426e-06,
"loss": 1.2828115224838257,
"step": 112
},
{
"epoch": 0.25277161862527714,
"grad_norm": 1.1793975830078125,
"learning_rate": 4.998168849724196e-06,
"loss": 0.7794591188430786,
"step": 114
},
{
"epoch": 0.2572062084257206,
"grad_norm": 1.1411471366882324,
"learning_rate": 4.997820836032363e-06,
"loss": 1.3091580867767334,
"step": 116
},
{
"epoch": 0.2616407982261641,
"grad_norm": 0.4769836962223053,
"learning_rate": 4.997442580639443e-06,
"loss": 1.0651829242706299,
"step": 118
},
{
"epoch": 0.2660753880266075,
"grad_norm": 0.9096765518188477,
"learning_rate": 4.997034088634404e-06,
"loss": 1.2504687309265137,
"step": 120
},
{
"epoch": 0.270509977827051,
"grad_norm": 1.190130591392517,
"learning_rate": 4.996595365513012e-06,
"loss": 1.1688843965530396,
"step": 122
},
{
"epoch": 0.2749445676274945,
"grad_norm": 2.0130395889282227,
"learning_rate": 4.9961264171777515e-06,
"loss": 1.5667338371276855,
"step": 124
},
{
"epoch": 0.2793791574279379,
"grad_norm": 0.7699439525604248,
"learning_rate": 4.995627249937755e-06,
"loss": 0.9175050258636475,
"step": 126
},
{
"epoch": 0.2838137472283814,
"grad_norm": 0.866606593132019,
"learning_rate": 4.995097870508711e-06,
"loss": 1.209236741065979,
"step": 128
},
{
"epoch": 0.28824833702882485,
"grad_norm": 1.518518328666687,
"learning_rate": 4.994538286012777e-06,
"loss": 0.9163856506347656,
"step": 130
},
{
"epoch": 0.2926829268292683,
"grad_norm": 1.7814369201660156,
"learning_rate": 4.993948503978484e-06,
"loss": 0.974137544631958,
"step": 132
},
{
"epoch": 0.29711751662971175,
"grad_norm": 3.4997236728668213,
"learning_rate": 4.993328532340633e-06,
"loss": 0.9840149283409119,
"step": 134
},
{
"epoch": 0.30155210643015523,
"grad_norm": 3.2781076431274414,
"learning_rate": 4.99267837944019e-06,
"loss": 1.110643982887268,
"step": 136
},
{
"epoch": 0.30598669623059865,
"grad_norm": 10.070029258728027,
"learning_rate": 4.991998054024172e-06,
"loss": 0.7842212319374084,
"step": 138
},
{
"epoch": 0.31042128603104213,
"grad_norm": 1.092695713043213,
"learning_rate": 4.991287565245534e-06,
"loss": 1.0406193733215332,
"step": 140
},
{
"epoch": 0.3148558758314856,
"grad_norm": 1.3157380819320679,
"learning_rate": 4.990546922663039e-06,
"loss": 0.9569450616836548,
"step": 142
},
{
"epoch": 0.31929046563192903,
"grad_norm": 2.488039970397949,
"learning_rate": 4.989776136241134e-06,
"loss": 0.9646241068840027,
"step": 144
},
{
"epoch": 0.3237250554323725,
"grad_norm": 1.5037761926651,
"learning_rate": 4.988975216349814e-06,
"loss": 0.9507364630699158,
"step": 146
},
{
"epoch": 0.328159645232816,
"grad_norm": 2.198122501373291,
"learning_rate": 4.988144173764486e-06,
"loss": 1.401615023612976,
"step": 148
},
{
"epoch": 0.3325942350332594,
"grad_norm": 7.834370136260986,
"learning_rate": 4.987283019665817e-06,
"loss": 1.1528998613357544,
"step": 150
},
{
"epoch": 0.3370288248337029,
"grad_norm": 3.270291566848755,
"learning_rate": 4.986391765639592e-06,
"loss": 1.060468077659607,
"step": 152
},
{
"epoch": 0.34146341463414637,
"grad_norm": 1.180214762687683,
"learning_rate": 4.985470423676551e-06,
"loss": 1.2685281038284302,
"step": 154
},
{
"epoch": 0.3458980044345898,
"grad_norm": 2.8346996307373047,
"learning_rate": 4.984519006172232e-06,
"loss": 1.2129993438720703,
"step": 156
},
{
"epoch": 0.35033259423503327,
"grad_norm": 0.9781622290611267,
"learning_rate": 4.983537525926804e-06,
"loss": 1.2925925254821777,
"step": 158
},
{
"epoch": 0.35476718403547675,
"grad_norm": 0.549372136592865,
"learning_rate": 4.982525996144891e-06,
"loss": 1.1071832180023193,
"step": 160
},
{
"epoch": 0.35920177383592017,
"grad_norm": 1.4813437461853027,
"learning_rate": 4.981484430435399e-06,
"loss": 0.8754007816314697,
"step": 162
},
{
"epoch": 0.36363636363636365,
"grad_norm": 1.7693350315093994,
"learning_rate": 4.98041284281133e-06,
"loss": 0.9003241062164307,
"step": 164
},
{
"epoch": 0.36807095343680707,
"grad_norm": 1.363362193107605,
"learning_rate": 4.979311247689596e-06,
"loss": 1.273066759109497,
"step": 166
},
{
"epoch": 0.37250554323725055,
"grad_norm": 6.419438362121582,
"learning_rate": 4.978179659890821e-06,
"loss": 1.1863045692443848,
"step": 168
},
{
"epoch": 0.376940133037694,
"grad_norm": 1.189877986907959,
"learning_rate": 4.977018094639146e-06,
"loss": 1.2364041805267334,
"step": 170
},
{
"epoch": 0.38137472283813745,
"grad_norm": 3.7446060180664062,
"learning_rate": 4.975826567562023e-06,
"loss": 0.7797529101371765,
"step": 172
},
{
"epoch": 0.3858093126385809,
"grad_norm": 4.418823719024658,
"learning_rate": 4.97460509469e-06,
"loss": 1.6287654638290405,
"step": 174
},
{
"epoch": 0.3902439024390244,
"grad_norm": 1.5152671337127686,
"learning_rate": 4.973353692456513e-06,
"loss": 1.26861572265625,
"step": 176
},
{
"epoch": 0.3946784922394678,
"grad_norm": 1.269068717956543,
"learning_rate": 4.972072377697661e-06,
"loss": 1.3224495649337769,
"step": 178
},
{
"epoch": 0.3991130820399113,
"grad_norm": 0.7425451874732971,
"learning_rate": 4.9707611676519775e-06,
"loss": 1.05917227268219,
"step": 180
},
{
"epoch": 0.4035476718403548,
"grad_norm": 1.1265835762023926,
"learning_rate": 4.969420079960203e-06,
"loss": 1.2697656154632568,
"step": 182
},
{
"epoch": 0.4079822616407982,
"grad_norm": 1.1472365856170654,
"learning_rate": 4.968049132665045e-06,
"loss": 0.884378969669342,
"step": 184
},
{
"epoch": 0.4124168514412417,
"grad_norm": 1.4855486154556274,
"learning_rate": 4.966648344210936e-06,
"loss": 0.9717956781387329,
"step": 186
},
{
"epoch": 0.41685144124168516,
"grad_norm": 1.1715413331985474,
"learning_rate": 4.965217733443782e-06,
"loss": 0.9312400817871094,
"step": 188
},
{
"epoch": 0.4212860310421286,
"grad_norm": 2.2187230587005615,
"learning_rate": 4.963757319610716e-06,
"loss": 0.9861539602279663,
"step": 190
},
{
"epoch": 0.42572062084257206,
"grad_norm": 2.1262733936309814,
"learning_rate": 4.962267122359835e-06,
"loss": 0.8997060656547546,
"step": 192
},
{
"epoch": 0.43015521064301554,
"grad_norm": 2.3965535163879395,
"learning_rate": 4.960747161739931e-06,
"loss": 1.2791428565979004,
"step": 194
},
{
"epoch": 0.43458980044345896,
"grad_norm": 1.8069835901260376,
"learning_rate": 4.9591974582002324e-06,
"loss": 1.549714207649231,
"step": 196
},
{
"epoch": 0.43902439024390244,
"grad_norm": 3.381026029586792,
"learning_rate": 4.957618032590118e-06,
"loss": 1.2765225172042847,
"step": 198
},
{
"epoch": 0.4434589800443459,
"grad_norm": 0.758112907409668,
"learning_rate": 4.956008906158842e-06,
"loss": 1.1299937963485718,
"step": 200
},
{
"epoch": 0.44789356984478934,
"grad_norm": 2.647829055786133,
"learning_rate": 4.954370100555249e-06,
"loss": 1.2492018938064575,
"step": 202
},
{
"epoch": 0.4523281596452328,
"grad_norm": 1.820816159248352,
"learning_rate": 4.952701637827476e-06,
"loss": 1.216017484664917,
"step": 204
},
{
"epoch": 0.4567627494456763,
"grad_norm": 1.9108037948608398,
"learning_rate": 4.951003540422668e-06,
"loss": 1.0710757970809937,
"step": 206
},
{
"epoch": 0.4611973392461197,
"grad_norm": 4.884266376495361,
"learning_rate": 4.949275831186663e-06,
"loss": 1.044965147972107,
"step": 208
},
{
"epoch": 0.4656319290465632,
"grad_norm": 1.51518976688385,
"learning_rate": 4.947518533363691e-06,
"loss": 0.645362377166748,
"step": 210
},
{
"epoch": 0.4700665188470067,
"grad_norm": 4.527989387512207,
"learning_rate": 4.945731670596062e-06,
"loss": 0.8511308431625366,
"step": 212
},
{
"epoch": 0.4745011086474501,
"grad_norm": 0.7997156381607056,
"learning_rate": 4.943915266923845e-06,
"loss": 1.0683618783950806,
"step": 214
},
{
"epoch": 0.4789356984478936,
"grad_norm": 1.7993850708007812,
"learning_rate": 4.942069346784547e-06,
"loss": 1.08345627784729,
"step": 216
},
{
"epoch": 0.48337028824833705,
"grad_norm": 0.7484229803085327,
"learning_rate": 4.940193935012785e-06,
"loss": 1.1189590692520142,
"step": 218
},
{
"epoch": 0.4878048780487805,
"grad_norm": 2.9010186195373535,
"learning_rate": 4.938289056839946e-06,
"loss": 1.2146114110946655,
"step": 220
},
{
"epoch": 0.49223946784922396,
"grad_norm": 1.2498693466186523,
"learning_rate": 4.936354737893854e-06,
"loss": 1.2259554862976074,
"step": 222
},
{
"epoch": 0.49667405764966743,
"grad_norm": 1.4774630069732666,
"learning_rate": 4.934391004198424e-06,
"loss": 1.2009336948394775,
"step": 224
},
{
"epoch": 0.5011086474501109,
"grad_norm": 1.1975030899047852,
"learning_rate": 4.932397882173307e-06,
"loss": 1.2059407234191895,
"step": 226
},
{
"epoch": 0.5055432372505543,
"grad_norm": 0.9775891900062561,
"learning_rate": 4.930375398633543e-06,
"loss": 1.2553625106811523,
"step": 228
},
{
"epoch": 0.5099778270509978,
"grad_norm": 1.1985933780670166,
"learning_rate": 4.928323580789192e-06,
"loss": 1.8459392786026,
"step": 230
},
{
"epoch": 0.5144124168514412,
"grad_norm": 2.5975992679595947,
"learning_rate": 4.926242456244973e-06,
"loss": 0.8564022779464722,
"step": 232
},
{
"epoch": 0.5188470066518847,
"grad_norm": 1.2151294946670532,
"learning_rate": 4.924132052999892e-06,
"loss": 1.2610244750976562,
"step": 234
},
{
"epoch": 0.5232815964523282,
"grad_norm": 1.6694788932800293,
"learning_rate": 4.921992399446861e-06,
"loss": 0.9529590010643005,
"step": 236
},
{
"epoch": 0.5277161862527716,
"grad_norm": 1.3985271453857422,
"learning_rate": 4.919823524372323e-06,
"loss": 0.9686543941497803,
"step": 238
},
{
"epoch": 0.532150776053215,
"grad_norm": 0.8372143507003784,
"learning_rate": 4.91762545695586e-06,
"loss": 1.258904218673706,
"step": 240
},
{
"epoch": 0.5365853658536586,
"grad_norm": 1.2533243894577026,
"learning_rate": 4.9153982267698e-06,
"loss": 1.3248710632324219,
"step": 242
},
{
"epoch": 0.541019955654102,
"grad_norm": 7.963499069213867,
"learning_rate": 4.913141863778822e-06,
"loss": 0.9705762267112732,
"step": 244
},
{
"epoch": 0.5454545454545454,
"grad_norm": 1.3693453073501587,
"learning_rate": 4.910856398339553e-06,
"loss": 1.2580170631408691,
"step": 246
},
{
"epoch": 0.549889135254989,
"grad_norm": 1.1870553493499756,
"learning_rate": 4.9085418612001545e-06,
"loss": 1.4803242683410645,
"step": 248
},
{
"epoch": 0.5543237250554324,
"grad_norm": 2.50886607170105,
"learning_rate": 4.906198283499916e-06,
"loss": 1.2252423763275146,
"step": 250
},
{
"epoch": 0.5587583148558758,
"grad_norm": 3.0687942504882812,
"learning_rate": 4.903825696768829e-06,
"loss": 0.6616644263267517,
"step": 252
},
{
"epoch": 0.5631929046563193,
"grad_norm": 2.0669586658477783,
"learning_rate": 4.901424132927172e-06,
"loss": 1.474183440208435,
"step": 254
},
{
"epoch": 0.5676274944567627,
"grad_norm": 4.0111403465271,
"learning_rate": 4.898993624285069e-06,
"loss": 1.316019892692566,
"step": 256
},
{
"epoch": 0.5720620842572062,
"grad_norm": 0.6927316784858704,
"learning_rate": 4.896534203542062e-06,
"loss": 1.2566733360290527,
"step": 258
},
{
"epoch": 0.5764966740576497,
"grad_norm": 1.2319213151931763,
"learning_rate": 4.894045903786675e-06,
"loss": 1.2865486145019531,
"step": 260
},
{
"epoch": 0.5809312638580931,
"grad_norm": 2.781428337097168,
"learning_rate": 4.891528758495961e-06,
"loss": 0.7123095989227295,
"step": 262
},
{
"epoch": 0.5853658536585366,
"grad_norm": 1.0490000247955322,
"learning_rate": 4.888982801535053e-06,
"loss": 1.360097050666809,
"step": 264
},
{
"epoch": 0.5898004434589801,
"grad_norm": 1.8262178897857666,
"learning_rate": 4.886408067156712e-06,
"loss": 1.0533033609390259,
"step": 266
},
{
"epoch": 0.5942350332594235,
"grad_norm": 1.2792811393737793,
"learning_rate": 4.883804590000865e-06,
"loss": 1.550133228302002,
"step": 268
},
{
"epoch": 0.5986696230598669,
"grad_norm": 2.9849724769592285,
"learning_rate": 4.881172405094138e-06,
"loss": 1.1676595211029053,
"step": 270
},
{
"epoch": 0.6031042128603105,
"grad_norm": 0.7342039942741394,
"learning_rate": 4.878511547849383e-06,
"loss": 1.2335383892059326,
"step": 272
},
{
"epoch": 0.6075388026607539,
"grad_norm": 1.2436589002609253,
"learning_rate": 4.875822054065203e-06,
"loss": 1.2187210321426392,
"step": 274
},
{
"epoch": 0.6119733924611973,
"grad_norm": 1.587964653968811,
"learning_rate": 4.8731039599254754e-06,
"loss": 1.226876974105835,
"step": 276
},
{
"epoch": 0.6164079822616408,
"grad_norm": 4.160574913024902,
"learning_rate": 4.870357301998856e-06,
"loss": 1.205104112625122,
"step": 278
},
{
"epoch": 0.6208425720620843,
"grad_norm": 1.1125067472457886,
"learning_rate": 4.867582117238294e-06,
"loss": 1.287103295326233,
"step": 280
},
{
"epoch": 0.6252771618625277,
"grad_norm": 2.7902486324310303,
"learning_rate": 4.864778442980532e-06,
"loss": 0.8597289323806763,
"step": 282
},
{
"epoch": 0.6297117516629712,
"grad_norm": 1.8220659494400024,
"learning_rate": 4.861946316945605e-06,
"loss": 1.2538930177688599,
"step": 284
},
{
"epoch": 0.6341463414634146,
"grad_norm": 1.2487821578979492,
"learning_rate": 4.859085777236331e-06,
"loss": 1.330883264541626,
"step": 286
},
{
"epoch": 0.6385809312638581,
"grad_norm": 1.1677597761154175,
"learning_rate": 4.8561968623377985e-06,
"loss": 1.252233862876892,
"step": 288
},
{
"epoch": 0.6430155210643016,
"grad_norm": 1.1160072088241577,
"learning_rate": 4.853279611116852e-06,
"loss": 1.2359025478363037,
"step": 290
},
{
"epoch": 0.647450110864745,
"grad_norm": 1.4154568910598755,
"learning_rate": 4.850334062821566e-06,
"loss": 1.3491101264953613,
"step": 292
},
{
"epoch": 0.6518847006651884,
"grad_norm": 1.4015207290649414,
"learning_rate": 4.8473602570807185e-06,
"loss": 0.9202826619148254,
"step": 294
},
{
"epoch": 0.656319290465632,
"grad_norm": 1.124402403831482,
"learning_rate": 4.844358233903254e-06,
"loss": 0.9892662167549133,
"step": 296
},
{
"epoch": 0.6607538802660754,
"grad_norm": 1.2850724458694458,
"learning_rate": 4.841328033677753e-06,
"loss": 1.2050740718841553,
"step": 298
},
{
"epoch": 0.6651884700665188,
"grad_norm": 1.2479948997497559,
"learning_rate": 4.83826969717188e-06,
"loss": 1.2547260522842407,
"step": 300
},
{
"epoch": 0.6696230598669624,
"grad_norm": 1.376888394355774,
"learning_rate": 4.835183265531843e-06,
"loss": 1.1995564699172974,
"step": 302
},
{
"epoch": 0.6740576496674058,
"grad_norm": 1.6199829578399658,
"learning_rate": 4.832068780281831e-06,
"loss": 1.2650580406188965,
"step": 304
},
{
"epoch": 0.6784922394678492,
"grad_norm": 2.1889870166778564,
"learning_rate": 4.828926283323464e-06,
"loss": 1.2397799491882324,
"step": 306
},
{
"epoch": 0.6829268292682927,
"grad_norm": 1.4469552040100098,
"learning_rate": 4.8257558169352254e-06,
"loss": 0.8677200675010681,
"step": 308
},
{
"epoch": 0.6873614190687362,
"grad_norm": 1.1219711303710938,
"learning_rate": 4.8225574237718906e-06,
"loss": 1.2440086603164673,
"step": 310
},
{
"epoch": 0.6917960088691796,
"grad_norm": 0.9012990593910217,
"learning_rate": 4.819331146863958e-06,
"loss": 1.2062432765960693,
"step": 312
},
{
"epoch": 0.6962305986696231,
"grad_norm": 1.1953213214874268,
"learning_rate": 4.8160770296170685e-06,
"loss": 1.2283505201339722,
"step": 314
},
{
"epoch": 0.7006651884700665,
"grad_norm": 1.6990729570388794,
"learning_rate": 4.812795115811419e-06,
"loss": 1.3305617570877075,
"step": 316
},
{
"epoch": 0.70509977827051,
"grad_norm": 0.8295086622238159,
"learning_rate": 4.809485449601177e-06,
"loss": 0.9975929856300354,
"step": 318
},
{
"epoch": 0.7095343680709535,
"grad_norm": 1.4529755115509033,
"learning_rate": 4.806148075513883e-06,
"loss": 0.9766805768013,
"step": 320
},
{
"epoch": 0.7139689578713969,
"grad_norm": 8.476192474365234,
"learning_rate": 4.802783038449857e-06,
"loss": 0.9863343834877014,
"step": 322
},
{
"epoch": 0.7184035476718403,
"grad_norm": 1.1818149089813232,
"learning_rate": 4.799390383681587e-06,
"loss": 1.0997380018234253,
"step": 324
},
{
"epoch": 0.7228381374722838,
"grad_norm": 2.4012951850891113,
"learning_rate": 4.795970156853124e-06,
"loss": 1.077536702156067,
"step": 326
},
{
"epoch": 0.7272727272727273,
"grad_norm": 0.42787617444992065,
"learning_rate": 4.792522403979471e-06,
"loss": 0.8034789562225342,
"step": 328
},
{
"epoch": 0.7317073170731707,
"grad_norm": 1.517409324645996,
"learning_rate": 4.789047171445957e-06,
"loss": 0.8247131109237671,
"step": 330
},
{
"epoch": 0.7361419068736141,
"grad_norm": 3.9636194705963135,
"learning_rate": 4.785544506007619e-06,
"loss": 1.3770612478256226,
"step": 332
},
{
"epoch": 0.7405764966740577,
"grad_norm": 3.6474459171295166,
"learning_rate": 4.782014454788566e-06,
"loss": 1.1880348920822144,
"step": 334
},
{
"epoch": 0.7450110864745011,
"grad_norm": 1.1807231903076172,
"learning_rate": 4.778457065281355e-06,
"loss": 1.2278774976730347,
"step": 336
},
{
"epoch": 0.7494456762749445,
"grad_norm": 3.716444969177246,
"learning_rate": 4.774872385346345e-06,
"loss": 0.6877051591873169,
"step": 338
},
{
"epoch": 0.753880266075388,
"grad_norm": 10.607100486755371,
"learning_rate": 4.7712604632110524e-06,
"loss": 0.3212733566761017,
"step": 340
},
{
"epoch": 0.7583148558758315,
"grad_norm": 2.057530164718628,
"learning_rate": 4.767621347469506e-06,
"loss": 0.8571130633354187,
"step": 342
},
{
"epoch": 0.7627494456762749,
"grad_norm": 0.9771236181259155,
"learning_rate": 4.7639550870815895e-06,
"loss": 1.2813206911087036,
"step": 344
},
{
"epoch": 0.7671840354767184,
"grad_norm": 0.8891447186470032,
"learning_rate": 4.760261731372388e-06,
"loss": 1.2372010946273804,
"step": 346
},
{
"epoch": 0.7716186252771619,
"grad_norm": 2.889232873916626,
"learning_rate": 4.75654133003152e-06,
"loss": 1.0866163969039917,
"step": 348
},
{
"epoch": 0.7760532150776053,
"grad_norm": 1.4878243207931519,
"learning_rate": 4.752793933112469e-06,
"loss": 1.2248578071594238,
"step": 350
},
{
"epoch": 0.7804878048780488,
"grad_norm": 8.06610107421875,
"learning_rate": 4.749019591031914e-06,
"loss": 1.144696831703186,
"step": 352
},
{
"epoch": 0.7849223946784922,
"grad_norm": 1.4932175874710083,
"learning_rate": 4.745218354569045e-06,
"loss": 1.0152348279953003,
"step": 354
},
{
"epoch": 0.7893569844789357,
"grad_norm": 2.44124698638916,
"learning_rate": 4.741390274864885e-06,
"loss": 1.1802353858947754,
"step": 356
},
{
"epoch": 0.7937915742793792,
"grad_norm": 4.262493133544922,
"learning_rate": 4.737535403421601e-06,
"loss": 1.2564811706542969,
"step": 358
},
{
"epoch": 0.7982261640798226,
"grad_norm": 10.588669776916504,
"learning_rate": 4.733653792101809e-06,
"loss": 1.2239693403244019,
"step": 360
},
{
"epoch": 0.802660753880266,
"grad_norm": 1.7806600332260132,
"learning_rate": 4.729745493127878e-06,
"loss": 0.5834329724311829,
"step": 362
},
{
"epoch": 0.8070953436807096,
"grad_norm": 1.3935792446136475,
"learning_rate": 4.725810559081227e-06,
"loss": 1.324857234954834,
"step": 364
},
{
"epoch": 0.811529933481153,
"grad_norm": 1.9065128564834595,
"learning_rate": 4.7218490429016175e-06,
"loss": 1.2007935047149658,
"step": 366
},
{
"epoch": 0.8159645232815964,
"grad_norm": 2.5401406288146973,
"learning_rate": 4.717860997886442e-06,
"loss": 0.9981905817985535,
"step": 368
},
{
"epoch": 0.8203991130820399,
"grad_norm": 3.2703709602355957,
"learning_rate": 4.713846477690005e-06,
"loss": 0.8222334384918213,
"step": 370
},
{
"epoch": 0.8248337028824834,
"grad_norm": 1.1709415912628174,
"learning_rate": 4.709805536322804e-06,
"loss": 1.22462797164917,
"step": 372
},
{
"epoch": 0.8292682926829268,
"grad_norm": 1.1292668581008911,
"learning_rate": 4.7057382281508e-06,
"loss": 1.2273211479187012,
"step": 374
},
{
"epoch": 0.8337028824833703,
"grad_norm": 0.9493663311004639,
"learning_rate": 4.701644607894687e-06,
"loss": 1.1893556118011475,
"step": 376
},
{
"epoch": 0.8381374722838137,
"grad_norm": 1.0127923488616943,
"learning_rate": 4.697524730629159e-06,
"loss": 1.215989589691162,
"step": 378
},
{
"epoch": 0.8425720620842572,
"grad_norm": 8.141073226928711,
"learning_rate": 4.693378651782162e-06,
"loss": 0.7287262082099915,
"step": 380
},
{
"epoch": 0.8470066518847007,
"grad_norm": 1.4302774667739868,
"learning_rate": 4.689206427134155e-06,
"loss": 1.2984516620635986,
"step": 382
},
{
"epoch": 0.8514412416851441,
"grad_norm": 0.8617311120033264,
"learning_rate": 4.6850081128173595e-06,
"loss": 1.1311615705490112,
"step": 384
},
{
"epoch": 0.8558758314855875,
"grad_norm": 0.833918571472168,
"learning_rate": 4.680783765314994e-06,
"loss": 1.246683955192566,
"step": 386
},
{
"epoch": 0.8603104212860311,
"grad_norm": 0.5764380693435669,
"learning_rate": 4.6765334414605315e-06,
"loss": 1.1455856561660767,
"step": 388
},
{
"epoch": 0.8647450110864745,
"grad_norm": 2.1335196495056152,
"learning_rate": 4.672257198436918e-06,
"loss": 1.2500605583190918,
"step": 390
},
{
"epoch": 0.8691796008869179,
"grad_norm": 3.1419546604156494,
"learning_rate": 4.667955093775814e-06,
"loss": 0.9110448360443115,
"step": 392
},
{
"epoch": 0.8736141906873615,
"grad_norm": 0.8658338189125061,
"learning_rate": 4.663627185356818e-06,
"loss": 1.2182717323303223,
"step": 394
},
{
"epoch": 0.8780487804878049,
"grad_norm": 1.184532880783081,
"learning_rate": 4.65927353140668e-06,
"loss": 1.2174220085144043,
"step": 396
},
{
"epoch": 0.8824833702882483,
"grad_norm": 1.8402904272079468,
"learning_rate": 4.654894190498534e-06,
"loss": 1.210001826286316,
"step": 398
},
{
"epoch": 0.8869179600886918,
"grad_norm": 1.187842607498169,
"learning_rate": 4.650489221551095e-06,
"loss": 0.45290517807006836,
"step": 400
},
{
"epoch": 0.8913525498891353,
"grad_norm": 0.9917973279953003,
"learning_rate": 4.646058683827874e-06,
"loss": 1.045676350593567,
"step": 402
},
{
"epoch": 0.8957871396895787,
"grad_norm": 0.36595824360847473,
"learning_rate": 4.641602636936378e-06,
"loss": 0.9281713366508484,
"step": 404
},
{
"epoch": 0.9002217294900222,
"grad_norm": 0.9975008964538574,
"learning_rate": 4.637121140827311e-06,
"loss": 1.2655534744262695,
"step": 406
},
{
"epoch": 0.9046563192904656,
"grad_norm": 3.180144786834717,
"learning_rate": 4.632614255793762e-06,
"loss": 1.1325139999389648,
"step": 408
},
{
"epoch": 0.9090909090909091,
"grad_norm": 5.177031517028809,
"learning_rate": 4.6280820424704e-06,
"loss": 1.069734811782837,
"step": 410
},
{
"epoch": 0.9135254988913526,
"grad_norm": 1.0717765092849731,
"learning_rate": 4.623524561832653e-06,
"loss": 1.2320454120635986,
"step": 412
},
{
"epoch": 0.917960088691796,
"grad_norm": 1.0891166925430298,
"learning_rate": 4.618941875195893e-06,
"loss": 1.2557101249694824,
"step": 414
},
{
"epoch": 0.9223946784922394,
"grad_norm": 6.645715713500977,
"learning_rate": 4.614334044214606e-06,
"loss": 0.9693298935890198,
"step": 416
},
{
"epoch": 0.926829268292683,
"grad_norm": 0.7747200131416321,
"learning_rate": 4.6097011308815645e-06,
"loss": 1.2573974132537842,
"step": 418
},
{
"epoch": 0.9312638580931264,
"grad_norm": 14.614143371582031,
"learning_rate": 4.605043197526996e-06,
"loss": 0.7741899490356445,
"step": 420
},
{
"epoch": 0.9356984478935698,
"grad_norm": 11.045886993408203,
"learning_rate": 4.600360306817738e-06,
"loss": 1.45621657371521,
"step": 422
},
{
"epoch": 0.9401330376940134,
"grad_norm": 4.732778549194336,
"learning_rate": 4.595652521756403e-06,
"loss": 1.0499638319015503,
"step": 424
},
{
"epoch": 0.9445676274944568,
"grad_norm": 0.7621893882751465,
"learning_rate": 4.590919905680524e-06,
"loss": 1.202383041381836,
"step": 426
},
{
"epoch": 0.9490022172949002,
"grad_norm": 0.6728458404541016,
"learning_rate": 4.5861625222617065e-06,
"loss": 1.0747066736221313,
"step": 428
},
{
"epoch": 0.9534368070953437,
"grad_norm": 4.075941562652588,
"learning_rate": 4.58138043550477e-06,
"loss": 0.6646397709846497,
"step": 430
},
{
"epoch": 0.9578713968957872,
"grad_norm": 1.059487223625183,
"learning_rate": 4.576573709746887e-06,
"loss": 1.2203632593154907,
"step": 432
},
{
"epoch": 0.9623059866962306,
"grad_norm": 2.331674575805664,
"learning_rate": 4.5717424096567205e-06,
"loss": 1.0556832551956177,
"step": 434
},
{
"epoch": 0.9667405764966741,
"grad_norm": 1.2610398530960083,
"learning_rate": 4.566886600233547e-06,
"loss": 1.2815701961517334,
"step": 436
},
{
"epoch": 0.9711751662971175,
"grad_norm": 1.1328259706497192,
"learning_rate": 4.56200634680639e-06,
"loss": 1.233439564704895,
"step": 438
},
{
"epoch": 0.975609756097561,
"grad_norm": 2.2309317588806152,
"learning_rate": 4.557101715033136e-06,
"loss": 0.7945879697799683,
"step": 440
},
{
"epoch": 0.9800443458980045,
"grad_norm": 1.2745883464813232,
"learning_rate": 4.552172770899652e-06,
"loss": 1.0334054231643677,
"step": 442
},
{
"epoch": 0.9844789356984479,
"grad_norm": 0.977016806602478,
"learning_rate": 4.547219580718899e-06,
"loss": 1.2736181020736694,
"step": 444
},
{
"epoch": 0.9889135254988913,
"grad_norm": 2.0068366527557373,
"learning_rate": 4.542242211130039e-06,
"loss": 1.2092833518981934,
"step": 446
},
{
"epoch": 0.9933481152993349,
"grad_norm": 1.753902792930603,
"learning_rate": 4.537240729097539e-06,
"loss": 1.2434141635894775,
"step": 448
},
{
"epoch": 0.9977827050997783,
"grad_norm": 1.0139368772506714,
"learning_rate": 4.532215201910269e-06,
"loss": 1.0325958728790283,
"step": 450
},
{
"epoch": 1.0022172949002217,
"grad_norm": 2.0576000213623047,
"learning_rate": 4.527165697180598e-06,
"loss": 1.0579339265823364,
"step": 452
},
{
"epoch": 1.0066518847006651,
"grad_norm": 1.966776967048645,
"learning_rate": 4.522092282843481e-06,
"loss": 1.3362075090408325,
"step": 454
},
{
"epoch": 1.0110864745011086,
"grad_norm": 1.2726391553878784,
"learning_rate": 4.516995027155554e-06,
"loss": 1.2156652212142944,
"step": 456
},
{
"epoch": 1.0155210643015522,
"grad_norm": 0.7326069474220276,
"learning_rate": 4.511873998694204e-06,
"loss": 0.8079202175140381,
"step": 458
},
{
"epoch": 1.0199556541019956,
"grad_norm": 0.47465115785598755,
"learning_rate": 4.506729266356651e-06,
"loss": 0.8377529978752136,
"step": 460
},
{
"epoch": 1.024390243902439,
"grad_norm": 1.9195665121078491,
"learning_rate": 4.5015608993590276e-06,
"loss": 0.5483433604240417,
"step": 462
},
{
"epoch": 1.0288248337028825,
"grad_norm": 1.7749394178390503,
"learning_rate": 4.4963689672354375e-06,
"loss": 0.9224250316619873,
"step": 464
},
{
"epoch": 1.033259423503326,
"grad_norm": 1.9494274854660034,
"learning_rate": 4.491153539837026e-06,
"loss": 0.7165282964706421,
"step": 466
},
{
"epoch": 1.0376940133037693,
"grad_norm": 2.415039300918579,
"learning_rate": 4.4859146873310375e-06,
"loss": 1.0422555208206177,
"step": 468
},
{
"epoch": 1.042128603104213,
"grad_norm": 1.4911917448043823,
"learning_rate": 4.480652480199873e-06,
"loss": 0.6359959244728088,
"step": 470
},
{
"epoch": 1.0465631929046564,
"grad_norm": 1.3522758483886719,
"learning_rate": 4.475366989240147e-06,
"loss": 0.9994376301765442,
"step": 472
},
{
"epoch": 1.0509977827050998,
"grad_norm": 1.266831636428833,
"learning_rate": 4.470058285561721e-06,
"loss": 1.0376862287521362,
"step": 474
},
{
"epoch": 1.0554323725055432,
"grad_norm": 1.151628851890564,
"learning_rate": 4.464726440586761e-06,
"loss": 1.076725959777832,
"step": 476
},
{
"epoch": 1.0598669623059866,
"grad_norm": 1.4442075490951538,
"learning_rate": 4.45937152604877e-06,
"loss": 1.1006007194519043,
"step": 478
},
{
"epoch": 1.06430155210643,
"grad_norm": 1.1244226694107056,
"learning_rate": 4.453993613991622e-06,
"loss": 0.6436704397201538,
"step": 480
},
{
"epoch": 1.0687361419068737,
"grad_norm": 1.8516688346862793,
"learning_rate": 4.4485927767685995e-06,
"loss": 1.153225064277649,
"step": 482
},
{
"epoch": 1.0731707317073171,
"grad_norm": 0.4945599138736725,
"learning_rate": 4.443169087041409e-06,
"loss": 0.9155857563018799,
"step": 484
},
{
"epoch": 1.0776053215077606,
"grad_norm": 1.006088137626648,
"learning_rate": 4.4377226177792145e-06,
"loss": 1.0506926774978638,
"step": 486
},
{
"epoch": 1.082039911308204,
"grad_norm": 1.9565138816833496,
"learning_rate": 4.432253442257649e-06,
"loss": 0.7985799312591553,
"step": 488
},
{
"epoch": 1.0864745011086474,
"grad_norm": 0.8965359926223755,
"learning_rate": 4.426761634057831e-06,
"loss": 0.9610664248466492,
"step": 490
},
{
"epoch": 1.0909090909090908,
"grad_norm": 1.760244607925415,
"learning_rate": 4.421247267065375e-06,
"loss": 1.1698113679885864,
"step": 492
},
{
"epoch": 1.0953436807095343,
"grad_norm": 5.320736885070801,
"learning_rate": 4.415710415469394e-06,
"loss": 1.0748765468597412,
"step": 494
},
{
"epoch": 1.099778270509978,
"grad_norm": 2.7736940383911133,
"learning_rate": 4.410151153761506e-06,
"loss": 0.8963067531585693,
"step": 496
},
{
"epoch": 1.1042128603104213,
"grad_norm": 0.9353604912757874,
"learning_rate": 4.404569556734832e-06,
"loss": 1.023295521736145,
"step": 498
},
{
"epoch": 1.1086474501108647,
"grad_norm": 4.738027572631836,
"learning_rate": 4.398965699482984e-06,
"loss": 0.9625403881072998,
"step": 500
},
{
"epoch": 1.1130820399113082,
"grad_norm": 0.9695823788642883,
"learning_rate": 4.39333965739906e-06,
"loss": 1.0802518129348755,
"step": 502
},
{
"epoch": 1.1175166297117516,
"grad_norm": 2.8423125743865967,
"learning_rate": 4.3876915061746275e-06,
"loss": 0.9319751262664795,
"step": 504
},
{
"epoch": 1.1219512195121952,
"grad_norm": 1.8870809078216553,
"learning_rate": 4.382021321798707e-06,
"loss": 1.0706431865692139,
"step": 506
},
{
"epoch": 1.1263858093126387,
"grad_norm": 2.326402425765991,
"learning_rate": 4.376329180556745e-06,
"loss": 0.790678858757019,
"step": 508
},
{
"epoch": 1.130820399113082,
"grad_norm": 1.3830646276474,
"learning_rate": 4.370615159029594e-06,
"loss": 1.1257532835006714,
"step": 510
},
{
"epoch": 1.1352549889135255,
"grad_norm": 1.6302467584609985,
"learning_rate": 4.36487933409248e-06,
"loss": 0.5582272410392761,
"step": 512
},
{
"epoch": 1.139689578713969,
"grad_norm": 1.8311710357666016,
"learning_rate": 4.359121782913964e-06,
"loss": 0.541851282119751,
"step": 514
},
{
"epoch": 1.1441241685144123,
"grad_norm": 1.120679497718811,
"learning_rate": 4.3533425829549085e-06,
"loss": 1.0328795909881592,
"step": 516
},
{
"epoch": 1.1485587583148558,
"grad_norm": 1.37759530544281,
"learning_rate": 4.347541811967436e-06,
"loss": 1.163663625717163,
"step": 518
},
{
"epoch": 1.1529933481152994,
"grad_norm": 0.981916606426239,
"learning_rate": 4.341719547993879e-06,
"loss": 1.1469664573669434,
"step": 520
},
{
"epoch": 1.1574279379157428,
"grad_norm": 1.2698646783828735,
"learning_rate": 4.335875869365732e-06,
"loss": 0.5811082124710083,
"step": 522
},
{
"epoch": 1.1618625277161863,
"grad_norm": 1.2500545978546143,
"learning_rate": 4.330010854702598e-06,
"loss": 1.0397084951400757,
"step": 524
},
{
"epoch": 1.1662971175166297,
"grad_norm": 1.7220888137817383,
"learning_rate": 4.3241245829111324e-06,
"loss": 1.1979455947875977,
"step": 526
},
{
"epoch": 1.170731707317073,
"grad_norm": 1.460283875465393,
"learning_rate": 4.318217133183978e-06,
"loss": 0.6363497376441956,
"step": 528
},
{
"epoch": 1.1751662971175167,
"grad_norm": 1.075685977935791,
"learning_rate": 4.312288584998697e-06,
"loss": 0.8744889497756958,
"step": 530
},
{
"epoch": 1.1796008869179602,
"grad_norm": 0.8398450016975403,
"learning_rate": 4.306339018116714e-06,
"loss": 0.96029132604599,
"step": 532
},
{
"epoch": 1.1840354767184036,
"grad_norm": 1.3492926359176636,
"learning_rate": 4.300368512582227e-06,
"loss": 1.0959749221801758,
"step": 534
},
{
"epoch": 1.188470066518847,
"grad_norm": 0.3663475215435028,
"learning_rate": 4.294377148721144e-06,
"loss": 0.82485431432724,
"step": 536
},
{
"epoch": 1.1929046563192904,
"grad_norm": 1.7144395112991333,
"learning_rate": 4.288365007139991e-06,
"loss": 1.1424366235733032,
"step": 538
},
{
"epoch": 1.1973392461197339,
"grad_norm": 2.5541539192199707,
"learning_rate": 4.2823321687248386e-06,
"loss": 0.6629378795623779,
"step": 540
},
{
"epoch": 1.2017738359201773,
"grad_norm": 2.608015775680542,
"learning_rate": 4.276278714640203e-06,
"loss": 0.6427868008613586,
"step": 542
},
{
"epoch": 1.206208425720621,
"grad_norm": 17.38459587097168,
"learning_rate": 4.270204726327963e-06,
"loss": 1.1916182041168213,
"step": 544
},
{
"epoch": 1.2106430155210643,
"grad_norm": 0.5953897833824158,
"learning_rate": 4.264110285506259e-06,
"loss": 0.8375392556190491,
"step": 546
},
{
"epoch": 1.2150776053215078,
"grad_norm": 1.1584091186523438,
"learning_rate": 4.257995474168395e-06,
"loss": 1.216051459312439,
"step": 548
},
{
"epoch": 1.2195121951219512,
"grad_norm": 2.22684645652771,
"learning_rate": 4.251860374581736e-06,
"loss": 0.598793625831604,
"step": 550
},
{
"epoch": 1.2239467849223946,
"grad_norm": 1.4314756393432617,
"learning_rate": 4.245705069286601e-06,
"loss": 1.076664686203003,
"step": 552
},
{
"epoch": 1.2283813747228383,
"grad_norm": 1.2437039613723755,
"learning_rate": 4.239529641095149e-06,
"loss": 0.7446794509887695,
"step": 554
},
{
"epoch": 1.2328159645232817,
"grad_norm": 0.4004473090171814,
"learning_rate": 4.233334173090274e-06,
"loss": 0.7689218521118164,
"step": 556
},
{
"epoch": 1.237250554323725,
"grad_norm": 3.483046293258667,
"learning_rate": 4.227118748624478e-06,
"loss": 0.6857209205627441,
"step": 558
},
{
"epoch": 1.2416851441241685,
"grad_norm": 4.495988845825195,
"learning_rate": 4.220883451318753e-06,
"loss": 1.3058849573135376,
"step": 560
},
{
"epoch": 1.246119733924612,
"grad_norm": 1.3580222129821777,
"learning_rate": 4.2146283650614545e-06,
"loss": 0.9943916201591492,
"step": 562
},
{
"epoch": 1.2505543237250554,
"grad_norm": 0.6850365400314331,
"learning_rate": 4.208353574007179e-06,
"loss": 0.6252534985542297,
"step": 564
},
{
"epoch": 1.2549889135254988,
"grad_norm": 1.7630574703216553,
"learning_rate": 4.202059162575622e-06,
"loss": 0.9947891235351562,
"step": 566
},
{
"epoch": 1.2594235033259422,
"grad_norm": 1.371643304824829,
"learning_rate": 4.195745215450451e-06,
"loss": 0.9397783279418945,
"step": 568
},
{
"epoch": 1.2638580931263859,
"grad_norm": 4.08069372177124,
"learning_rate": 4.189411817578159e-06,
"loss": 1.136919379234314,
"step": 570
},
{
"epoch": 1.2682926829268293,
"grad_norm": 0.8603072166442871,
"learning_rate": 4.1830590541669304e-06,
"loss": 0.7842409610748291,
"step": 572
},
{
"epoch": 1.2727272727272727,
"grad_norm": 2.9320180416107178,
"learning_rate": 4.176687010685484e-06,
"loss": 1.3783859014511108,
"step": 574
},
{
"epoch": 1.2771618625277161,
"grad_norm": 1.506885290145874,
"learning_rate": 4.170295772861931e-06,
"loss": 1.0433826446533203,
"step": 576
},
{
"epoch": 1.2815964523281598,
"grad_norm": 1.3161765336990356,
"learning_rate": 4.163885426682619e-06,
"loss": 1.1605374813079834,
"step": 578
},
{
"epoch": 1.2860310421286032,
"grad_norm": 1.0546302795410156,
"learning_rate": 4.157456058390977e-06,
"loss": 1.121392011642456,
"step": 580
},
{
"epoch": 1.2904656319290466,
"grad_norm": 2.199747323989868,
"learning_rate": 4.151007754486351e-06,
"loss": 1.3379415273666382,
"step": 582
},
{
"epoch": 1.29490022172949,
"grad_norm": 2.361356258392334,
"learning_rate": 4.144540601722843e-06,
"loss": 0.7737810015678406,
"step": 584
},
{
"epoch": 1.2993348115299335,
"grad_norm": 1.0490304231643677,
"learning_rate": 4.138054687108143e-06,
"loss": 0.8205963373184204,
"step": 586
},
{
"epoch": 1.3037694013303769,
"grad_norm": 3.6670305728912354,
"learning_rate": 4.131550097902361e-06,
"loss": 0.47316789627075195,
"step": 588
},
{
"epoch": 1.3082039911308203,
"grad_norm": 4.267838478088379,
"learning_rate": 4.125026921616852e-06,
"loss": 0.9959248900413513,
"step": 590
},
{
"epoch": 1.3126385809312637,
"grad_norm": 3.6597208976745605,
"learning_rate": 4.118485246013031e-06,
"loss": 1.1072802543640137,
"step": 592
},
{
"epoch": 1.3170731707317074,
"grad_norm": 0.8594958782196045,
"learning_rate": 4.111925159101208e-06,
"loss": 1.0438408851623535,
"step": 594
},
{
"epoch": 1.3215077605321508,
"grad_norm": 1.6975743770599365,
"learning_rate": 4.1053467491393864e-06,
"loss": 1.0767079591751099,
"step": 596
},
{
"epoch": 1.3259423503325942,
"grad_norm": 2.8765268325805664,
"learning_rate": 4.098750104632091e-06,
"loss": 1.160161018371582,
"step": 598
},
{
"epoch": 1.3303769401330376,
"grad_norm": 1.8514677286148071,
"learning_rate": 4.092135314329165e-06,
"loss": 0.47440510988235474,
"step": 600
},
{
"epoch": 1.3348115299334813,
"grad_norm": 2.5123469829559326,
"learning_rate": 4.085502467224583e-06,
"loss": 1.0348572731018066,
"step": 602
},
{
"epoch": 1.3392461197339247,
"grad_norm": 2.3508543968200684,
"learning_rate": 4.078851652555254e-06,
"loss": 0.8366844654083252,
"step": 604
},
{
"epoch": 1.3436807095343681,
"grad_norm": 0.6843059659004211,
"learning_rate": 4.072182959799816e-06,
"loss": 0.782626211643219,
"step": 606
},
{
"epoch": 1.3481152993348116,
"grad_norm": 1.0305436849594116,
"learning_rate": 4.065496478677436e-06,
"loss": 1.113935112953186,
"step": 608
},
{
"epoch": 1.352549889135255,
"grad_norm": 6.971372127532959,
"learning_rate": 4.058792299146602e-06,
"loss": 1.1273419857025146,
"step": 610
},
{
"epoch": 1.3569844789356984,
"grad_norm": 9.397238731384277,
"learning_rate": 4.052070511403912e-06,
"loss": 0.7277010679244995,
"step": 612
},
{
"epoch": 1.3614190687361418,
"grad_norm": 1.1997414827346802,
"learning_rate": 4.045331205882863e-06,
"loss": 1.1359970569610596,
"step": 614
},
{
"epoch": 1.3658536585365852,
"grad_norm": 1.7059991359710693,
"learning_rate": 4.038574473252629e-06,
"loss": 0.7449517846107483,
"step": 616
},
{
"epoch": 1.370288248337029,
"grad_norm": 1.1675894260406494,
"learning_rate": 4.031800404416849e-06,
"loss": 1.0706074237823486,
"step": 618
},
{
"epoch": 1.3747228381374723,
"grad_norm": 1.6913033723831177,
"learning_rate": 4.025009090512394e-06,
"loss": 0.09774535149335861,
"step": 620
},
{
"epoch": 1.3791574279379157,
"grad_norm": 0.9009557962417603,
"learning_rate": 4.018200622908153e-06,
"loss": 1.0203994512557983,
"step": 622
},
{
"epoch": 1.3835920177383592,
"grad_norm": 1.3641494512557983,
"learning_rate": 4.011375093203793e-06,
"loss": 0.949788510799408,
"step": 624
},
{
"epoch": 1.3880266075388026,
"grad_norm": 1.6519279479980469,
"learning_rate": 4.004532593228531e-06,
"loss": 0.9569694995880127,
"step": 626
},
{
"epoch": 1.3924611973392462,
"grad_norm": 1.1832157373428345,
"learning_rate": 3.997673215039899e-06,
"loss": 1.0896062850952148,
"step": 628
},
{
"epoch": 1.3968957871396896,
"grad_norm": 1.915313959121704,
"learning_rate": 3.990797050922506e-06,
"loss": 1.084737777709961,
"step": 630
},
{
"epoch": 1.401330376940133,
"grad_norm": 2.5589466094970703,
"learning_rate": 3.9839041933867954e-06,
"loss": 0.8085731863975525,
"step": 632
},
{
"epoch": 1.4057649667405765,
"grad_norm": 2.6854422092437744,
"learning_rate": 3.976994735167796e-06,
"loss": 1.0447653532028198,
"step": 634
},
{
"epoch": 1.41019955654102,
"grad_norm": 1.3558317422866821,
"learning_rate": 3.970068769223884e-06,
"loss": 1.110194206237793,
"step": 636
},
{
"epoch": 1.4146341463414633,
"grad_norm": 2.323361873626709,
"learning_rate": 3.963126388735525e-06,
"loss": 0.8597381711006165,
"step": 638
},
{
"epoch": 1.4190687361419068,
"grad_norm": 0.8953458070755005,
"learning_rate": 3.956167687104021e-06,
"loss": 0.7458611130714417,
"step": 640
},
{
"epoch": 1.4235033259423504,
"grad_norm": 1.6884546279907227,
"learning_rate": 3.9491927579502584e-06,
"loss": 0.8080941438674927,
"step": 642
},
{
"epoch": 1.4279379157427938,
"grad_norm": 1.2389963865280151,
"learning_rate": 3.9422016951134415e-06,
"loss": 0.6925735473632812,
"step": 644
},
{
"epoch": 1.4323725055432373,
"grad_norm": 1.243801236152649,
"learning_rate": 3.935194592649836e-06,
"loss": 1.2162237167358398,
"step": 646
},
{
"epoch": 1.4368070953436807,
"grad_norm": 0.7290199995040894,
"learning_rate": 3.928171544831501e-06,
"loss": 1.1060457229614258,
"step": 648
},
{
"epoch": 1.441241685144124,
"grad_norm": 0.8302409052848816,
"learning_rate": 3.921132646145019e-06,
"loss": 1.1656242609024048,
"step": 650
},
{
"epoch": 1.4456762749445677,
"grad_norm": 3.586696147918701,
"learning_rate": 3.914077991290232e-06,
"loss": 0.9107663035392761,
"step": 652
},
{
"epoch": 1.4501108647450112,
"grad_norm": 2.574885606765747,
"learning_rate": 3.907007675178956e-06,
"loss": 1.0581393241882324,
"step": 654
},
{
"epoch": 1.4545454545454546,
"grad_norm": 4.263864517211914,
"learning_rate": 3.899921792933713e-06,
"loss": 0.8163521885871887,
"step": 656
},
{
"epoch": 1.458980044345898,
"grad_norm": 0.9238935708999634,
"learning_rate": 3.892820439886448e-06,
"loss": 0.8494032621383667,
"step": 658
},
{
"epoch": 1.4634146341463414,
"grad_norm": 1.1233266592025757,
"learning_rate": 3.885703711577249e-06,
"loss": 1.0738056898117065,
"step": 660
},
{
"epoch": 1.4678492239467849,
"grad_norm": 0.7168223261833191,
"learning_rate": 3.8785717037530555e-06,
"loss": 1.078277587890625,
"step": 662
},
{
"epoch": 1.4722838137472283,
"grad_norm": 0.8631088137626648,
"learning_rate": 3.871424512366377e-06,
"loss": 0.8074089288711548,
"step": 664
},
{
"epoch": 1.476718403547672,
"grad_norm": 0.7211161255836487,
"learning_rate": 3.864262233574e-06,
"loss": 0.5039446949958801,
"step": 666
},
{
"epoch": 1.4811529933481153,
"grad_norm": 2.315062999725342,
"learning_rate": 3.857084963735689e-06,
"loss": 0.6357030868530273,
"step": 668
},
{
"epoch": 1.4855875831485588,
"grad_norm": 4.405363082885742,
"learning_rate": 3.849892799412902e-06,
"loss": 1.0553703308105469,
"step": 670
},
{
"epoch": 1.4900221729490022,
"grad_norm": 1.0128716230392456,
"learning_rate": 3.84268583736748e-06,
"loss": 1.1253726482391357,
"step": 672
},
{
"epoch": 1.4944567627494456,
"grad_norm": 1.074711799621582,
"learning_rate": 3.835464174560349e-06,
"loss": 0.6612215042114258,
"step": 674
},
{
"epoch": 1.4988913525498893,
"grad_norm": 0.7255991101264954,
"learning_rate": 3.828227908150217e-06,
"loss": 0.753847599029541,
"step": 676
},
{
"epoch": 1.5033259423503327,
"grad_norm": 0.9374713897705078,
"learning_rate": 3.820977135492266e-06,
"loss": 1.095574140548706,
"step": 678
},
{
"epoch": 1.507760532150776,
"grad_norm": 0.7554945349693298,
"learning_rate": 3.8137119541368415e-06,
"loss": 0.6039291620254517,
"step": 680
},
{
"epoch": 1.5121951219512195,
"grad_norm": 1.8996697664260864,
"learning_rate": 3.80643246182814e-06,
"loss": 0.5733712315559387,
"step": 682
},
{
"epoch": 1.516629711751663,
"grad_norm": 1.0919080972671509,
"learning_rate": 3.7991387565028963e-06,
"loss": 1.0893672704696655,
"step": 684
},
{
"epoch": 1.5210643015521064,
"grad_norm": 0.780432939529419,
"learning_rate": 3.791830936289062e-06,
"loss": 1.0852909088134766,
"step": 686
},
{
"epoch": 1.5254988913525498,
"grad_norm": 91.54979705810547,
"learning_rate": 3.784509099504488e-06,
"loss": 0.6314648985862732,
"step": 688
},
{
"epoch": 1.5299334811529932,
"grad_norm": 1.411130666732788,
"learning_rate": 3.7771733446556025e-06,
"loss": 0.3787440061569214,
"step": 690
},
{
"epoch": 1.5343680709534369,
"grad_norm": 1.2953332662582397,
"learning_rate": 3.7698237704360826e-06,
"loss": 0.8831952214241028,
"step": 692
},
{
"epoch": 1.5388026607538803,
"grad_norm": 2.0927493572235107,
"learning_rate": 3.7624604757255297e-06,
"loss": 0.9063498377799988,
"step": 694
},
{
"epoch": 1.5432372505543237,
"grad_norm": 8.049708366394043,
"learning_rate": 3.7550835595881365e-06,
"loss": 0.6573284268379211,
"step": 696
},
{
"epoch": 1.5476718403547673,
"grad_norm": 1.286287784576416,
"learning_rate": 3.747693121271355e-06,
"loss": 1.1525388956069946,
"step": 698
},
{
"epoch": 1.5521064301552108,
"grad_norm": 0.8270326256752014,
"learning_rate": 3.740289260204565e-06,
"loss": 1.0669758319854736,
"step": 700
},
{
"epoch": 1.5565410199556542,
"grad_norm": 0.873605489730835,
"learning_rate": 3.732872075997729e-06,
"loss": 1.1033045053482056,
"step": 702
},
{
"epoch": 1.5609756097560976,
"grad_norm": 1.9848657846450806,
"learning_rate": 3.725441668440058e-06,
"loss": 0.8427386283874512,
"step": 704
},
{
"epoch": 1.565410199556541,
"grad_norm": 2.9728550910949707,
"learning_rate": 3.7179981374986683e-06,
"loss": 0.4659649431705475,
"step": 706
},
{
"epoch": 1.5698447893569845,
"grad_norm": 0.9899294376373291,
"learning_rate": 3.710541583317233e-06,
"loss": 1.1070295572280884,
"step": 708
},
{
"epoch": 1.5742793791574279,
"grad_norm": 1.7865556478500366,
"learning_rate": 3.70307210621464e-06,
"loss": 0.8662485480308533,
"step": 710
},
{
"epoch": 1.5787139689578713,
"grad_norm": 1.5620512962341309,
"learning_rate": 3.695589806683636e-06,
"loss": 0.6354061365127563,
"step": 712
},
{
"epoch": 1.5831485587583147,
"grad_norm": 1.4959702491760254,
"learning_rate": 3.68809478538948e-06,
"loss": 1.0189735889434814,
"step": 714
},
{
"epoch": 1.5875831485587582,
"grad_norm": 1.0991613864898682,
"learning_rate": 3.6805871431685875e-06,
"loss": 1.1094708442687988,
"step": 716
},
{
"epoch": 1.5920177383592018,
"grad_norm": 1.5194610357284546,
"learning_rate": 3.6730669810271707e-06,
"loss": 1.3121333122253418,
"step": 718
},
{
"epoch": 1.5964523281596452,
"grad_norm": 1.7078560590744019,
"learning_rate": 3.665534400139885e-06,
"loss": 0.5672956109046936,
"step": 720
},
{
"epoch": 1.6008869179600886,
"grad_norm": 0.32701054215431213,
"learning_rate": 3.6579895018484635e-06,
"loss": 0.5884336233139038,
"step": 722
},
{
"epoch": 1.6053215077605323,
"grad_norm": 0.7404405474662781,
"learning_rate": 3.650432387660354e-06,
"loss": 1.0861414670944214,
"step": 724
},
{
"epoch": 1.6097560975609757,
"grad_norm": 1.0352939367294312,
"learning_rate": 3.6428631592473584e-06,
"loss": 1.0698777437210083,
"step": 726
},
{
"epoch": 1.6141906873614191,
"grad_norm": 1.0069420337677002,
"learning_rate": 3.6352819184442552e-06,
"loss": 0.7489967346191406,
"step": 728
},
{
"epoch": 1.6186252771618626,
"grad_norm": 1.1831481456756592,
"learning_rate": 3.6276887672474374e-06,
"loss": 1.0391978025436401,
"step": 730
},
{
"epoch": 1.623059866962306,
"grad_norm": 1.3561768531799316,
"learning_rate": 3.620083807813541e-06,
"loss": 1.090510368347168,
"step": 732
},
{
"epoch": 1.6274944567627494,
"grad_norm": 0.9160586595535278,
"learning_rate": 3.6124671424580633e-06,
"loss": 1.0251963138580322,
"step": 734
},
{
"epoch": 1.6319290465631928,
"grad_norm": 0.39837855100631714,
"learning_rate": 3.604838873653991e-06,
"loss": 0.5789291262626648,
"step": 736
},
{
"epoch": 1.6363636363636362,
"grad_norm": 0.7905840873718262,
"learning_rate": 3.597199104030424e-06,
"loss": 1.1960307359695435,
"step": 738
},
{
"epoch": 1.6407982261640797,
"grad_norm": 0.7028928995132446,
"learning_rate": 3.589547936371189e-06,
"loss": 1.0560542345046997,
"step": 740
},
{
"epoch": 1.6452328159645233,
"grad_norm": 3.9433796405792236,
"learning_rate": 3.58188547361346e-06,
"loss": 1.1990491151809692,
"step": 742
},
{
"epoch": 1.6496674057649667,
"grad_norm": 5.412987232208252,
"learning_rate": 3.574211818846374e-06,
"loss": 0.8513088226318359,
"step": 744
},
{
"epoch": 1.6541019955654102,
"grad_norm": 3.399298667907715,
"learning_rate": 3.566527075309641e-06,
"loss": 0.9072363972663879,
"step": 746
},
{
"epoch": 1.6585365853658538,
"grad_norm": 0.7723504304885864,
"learning_rate": 3.558831346392159e-06,
"loss": 0.8129432201385498,
"step": 748
},
{
"epoch": 1.6629711751662972,
"grad_norm": 1.6611526012420654,
"learning_rate": 3.5511247356306205e-06,
"loss": 1.1549206972122192,
"step": 750
},
{
"epoch": 1.6674057649667406,
"grad_norm": 0.962104082107544,
"learning_rate": 3.5434073467081183e-06,
"loss": 0.8654111623764038,
"step": 752
},
{
"epoch": 1.671840354767184,
"grad_norm": 0.9911019206047058,
"learning_rate": 3.5356792834527533e-06,
"loss": 0.5958826541900635,
"step": 754
},
{
"epoch": 1.6762749445676275,
"grad_norm": 1.1465089321136475,
"learning_rate": 3.527940649836238e-06,
"loss": 0.9339615106582642,
"step": 756
},
{
"epoch": 1.680709534368071,
"grad_norm": 1.773185133934021,
"learning_rate": 3.520191549972494e-06,
"loss": 1.0644793510437012,
"step": 758
},
{
"epoch": 1.6851441241685143,
"grad_norm": 10.430180549621582,
"learning_rate": 3.512432088116255e-06,
"loss": 0.9741522669792175,
"step": 760
},
{
"epoch": 1.6895787139689578,
"grad_norm": 1.8062833547592163,
"learning_rate": 3.5046623686616627e-06,
"loss": 1.0623693466186523,
"step": 762
},
{
"epoch": 1.6940133037694012,
"grad_norm": 1.815280795097351,
"learning_rate": 3.496882496140861e-06,
"loss": 1.0941108465194702,
"step": 764
},
{
"epoch": 1.6984478935698448,
"grad_norm": 1.8432520627975464,
"learning_rate": 3.4890925752225935e-06,
"loss": 0.9967592358589172,
"step": 766
},
{
"epoch": 1.7028824833702882,
"grad_norm": 2.3747756481170654,
"learning_rate": 3.48129271071079e-06,
"loss": 0.8642722964286804,
"step": 768
},
{
"epoch": 1.7073170731707317,
"grad_norm": 0.8133605718612671,
"learning_rate": 3.4734830075431605e-06,
"loss": 1.0309913158416748,
"step": 770
},
{
"epoch": 1.7117516629711753,
"grad_norm": 5.219134330749512,
"learning_rate": 3.4656635707897823e-06,
"loss": 1.0626434087753296,
"step": 772
},
{
"epoch": 1.7161862527716187,
"grad_norm": 1.766679286956787,
"learning_rate": 3.457834505651687e-06,
"loss": 1.0065494775772095,
"step": 774
},
{
"epoch": 1.7206208425720622,
"grad_norm": 1.107305884361267,
"learning_rate": 3.449995917459442e-06,
"loss": 1.1634471416473389,
"step": 776
},
{
"epoch": 1.7250554323725056,
"grad_norm": 0.7361903786659241,
"learning_rate": 3.4421479116717394e-06,
"loss": 1.0542564392089844,
"step": 778
},
{
"epoch": 1.729490022172949,
"grad_norm": 1.2649286985397339,
"learning_rate": 3.4342905938739707e-06,
"loss": 0.71857750415802,
"step": 780
},
{
"epoch": 1.7339246119733924,
"grad_norm": 0.9642868041992188,
"learning_rate": 3.4264240697768096e-06,
"loss": 1.039180040359497,
"step": 782
},
{
"epoch": 1.7383592017738358,
"grad_norm": 0.8986912369728088,
"learning_rate": 3.418548445214791e-06,
"loss": 0.7089306712150574,
"step": 784
},
{
"epoch": 1.7427937915742793,
"grad_norm": 0.8579238057136536,
"learning_rate": 3.410663826144884e-06,
"loss": 0.8006396293640137,
"step": 786
},
{
"epoch": 1.7472283813747227,
"grad_norm": 1.4266620874404907,
"learning_rate": 3.4027703186450672e-06,
"loss": 0.6931395530700684,
"step": 788
},
{
"epoch": 1.7516629711751663,
"grad_norm": 4.200963973999023,
"learning_rate": 3.394868028912906e-06,
"loss": 0.6739537119865417,
"step": 790
},
{
"epoch": 1.7560975609756098,
"grad_norm": 1.6932109594345093,
"learning_rate": 3.386957063264115e-06,
"loss": 0.7071681618690491,
"step": 792
},
{
"epoch": 1.7605321507760532,
"grad_norm": 1.0618504285812378,
"learning_rate": 3.3790375281311355e-06,
"loss": 0.8816713094711304,
"step": 794
},
{
"epoch": 1.7649667405764968,
"grad_norm": 6.342792987823486,
"learning_rate": 3.3711095300617015e-06,
"loss": 0.8279204368591309,
"step": 796
},
{
"epoch": 1.7694013303769403,
"grad_norm": 0.9115824699401855,
"learning_rate": 3.3631731757174048e-06,
"loss": 1.0542290210723877,
"step": 798
},
{
"epoch": 1.7738359201773837,
"grad_norm": 0.9469823837280273,
"learning_rate": 3.3552285718722593e-06,
"loss": 1.0976489782333374,
"step": 800
},
{
"epoch": 1.778270509977827,
"grad_norm": 0.9568173289299011,
"learning_rate": 3.3472758254112662e-06,
"loss": 1.0607342720031738,
"step": 802
},
{
"epoch": 1.7827050997782705,
"grad_norm": 1.9793434143066406,
"learning_rate": 3.3393150433289795e-06,
"loss": 1.1424230337142944,
"step": 804
},
{
"epoch": 1.787139689578714,
"grad_norm": 1.001698613166809,
"learning_rate": 3.3313463327280576e-06,
"loss": 0.7997353672981262,
"step": 806
},
{
"epoch": 1.7915742793791574,
"grad_norm": 1.129724383354187,
"learning_rate": 3.3233698008178306e-06,
"loss": 1.1001020669937134,
"step": 808
},
{
"epoch": 1.7960088691796008,
"grad_norm": 2.6598269939422607,
"learning_rate": 3.3153855549128537e-06,
"loss": 0.60479736328125,
"step": 810
},
{
"epoch": 1.8004434589800442,
"grad_norm": 3.824202537536621,
"learning_rate": 3.3073937024314647e-06,
"loss": 0.6236993074417114,
"step": 812
},
{
"epoch": 1.8048780487804879,
"grad_norm": 1.750210165977478,
"learning_rate": 3.2993943508943386e-06,
"loss": 1.2215160131454468,
"step": 814
},
{
"epoch": 1.8093126385809313,
"grad_norm": 2.1696269512176514,
"learning_rate": 3.291387607923041e-06,
"loss": 0.9674443006515503,
"step": 816
},
{
"epoch": 1.8137472283813747,
"grad_norm": 0.8863484263420105,
"learning_rate": 3.283373581238582e-06,
"loss": 0.49516186118125916,
"step": 818
},
{
"epoch": 1.8181818181818183,
"grad_norm": 1.207751750946045,
"learning_rate": 3.2753523786599618e-06,
"loss": 1.1064543724060059,
"step": 820
},
{
"epoch": 1.8226164079822618,
"grad_norm": 1.975991129875183,
"learning_rate": 3.2673241081027263e-06,
"loss": 1.086673617362976,
"step": 822
},
{
"epoch": 1.8270509977827052,
"grad_norm": 0.8671886920928955,
"learning_rate": 3.259288877577512e-06,
"loss": 1.1535236835479736,
"step": 824
},
{
"epoch": 1.8314855875831486,
"grad_norm": 1.9907917976379395,
"learning_rate": 3.251246795188592e-06,
"loss": 0.9606926441192627,
"step": 826
},
{
"epoch": 1.835920177383592,
"grad_norm": 2.376696825027466,
"learning_rate": 3.243197969132425e-06,
"loss": 0.8494828343391418,
"step": 828
},
{
"epoch": 1.8403547671840355,
"grad_norm": 0.9770940542221069,
"learning_rate": 3.2351425076961957e-06,
"loss": 1.0482726097106934,
"step": 830
},
{
"epoch": 1.8447893569844789,
"grad_norm": 1.570204496383667,
"learning_rate": 3.22708051925636e-06,
"loss": 1.0887196063995361,
"step": 832
},
{
"epoch": 1.8492239467849223,
"grad_norm": 0.7769954800605774,
"learning_rate": 3.219012112277189e-06,
"loss": 1.0047810077667236,
"step": 834
},
{
"epoch": 1.8536585365853657,
"grad_norm": 0.8842262029647827,
"learning_rate": 3.210937395309304e-06,
"loss": 1.151458501815796,
"step": 836
},
{
"epoch": 1.8580931263858091,
"grad_norm": 0.9991164803504944,
"learning_rate": 3.202856476988222e-06,
"loss": 1.0137677192687988,
"step": 838
},
{
"epoch": 1.8625277161862528,
"grad_norm": 2.2849833965301514,
"learning_rate": 3.1947694660328914e-06,
"loss": 1.1807574033737183,
"step": 840
},
{
"epoch": 1.8669623059866962,
"grad_norm": 0.8035255670547485,
"learning_rate": 3.1866764712442273e-06,
"loss": 0.6505411863327026,
"step": 842
},
{
"epoch": 1.8713968957871396,
"grad_norm": 1.0458487272262573,
"learning_rate": 3.1785776015036533e-06,
"loss": 0.771373987197876,
"step": 844
},
{
"epoch": 1.8758314855875833,
"grad_norm": 11.2003812789917,
"learning_rate": 3.1704729657716314e-06,
"loss": 0.8910792469978333,
"step": 846
},
{
"epoch": 1.8802660753880267,
"grad_norm": 1.6752492189407349,
"learning_rate": 3.1623626730861996e-06,
"loss": 1.0148093700408936,
"step": 848
},
{
"epoch": 1.8847006651884701,
"grad_norm": 0.9156674742698669,
"learning_rate": 3.1542468325615e-06,
"loss": 1.359726071357727,
"step": 850
},
{
"epoch": 1.8891352549889135,
"grad_norm": 1.454277515411377,
"learning_rate": 3.1461255533863183e-06,
"loss": 0.7917972207069397,
"step": 852
},
{
"epoch": 1.893569844789357,
"grad_norm": 0.9512672424316406,
"learning_rate": 3.1379989448226077e-06,
"loss": 1.073644757270813,
"step": 854
},
{
"epoch": 1.8980044345898004,
"grad_norm": 1.3854492902755737,
"learning_rate": 3.1298671162040236e-06,
"loss": 0.9803899526596069,
"step": 856
},
{
"epoch": 1.9024390243902438,
"grad_norm": 0.645820677280426,
"learning_rate": 3.1217301769344488e-06,
"loss": 0.7556897401809692,
"step": 858
},
{
"epoch": 1.9068736141906872,
"grad_norm": 4.2880635261535645,
"learning_rate": 3.1135882364865262e-06,
"loss": 0.8003832101821899,
"step": 860
},
{
"epoch": 1.9113082039911307,
"grad_norm": 1.7977222204208374,
"learning_rate": 3.105441404400183e-06,
"loss": 1.0547810792922974,
"step": 862
},
{
"epoch": 1.9157427937915743,
"grad_norm": 0.9238870739936829,
"learning_rate": 3.097289790281155e-06,
"loss": 0.8129977583885193,
"step": 864
},
{
"epoch": 1.9201773835920177,
"grad_norm": 3.8049755096435547,
"learning_rate": 3.089133503799517e-06,
"loss": 0.6351861953735352,
"step": 866
},
{
"epoch": 1.9246119733924612,
"grad_norm": 0.8284581899642944,
"learning_rate": 3.0809726546882045e-06,
"loss": 1.1457592248916626,
"step": 868
},
{
"epoch": 1.9290465631929048,
"grad_norm": 1.308618426322937,
"learning_rate": 3.0728073527415376e-06,
"loss": 1.0072696208953857,
"step": 870
},
{
"epoch": 1.9334811529933482,
"grad_norm": 0.38561350107192993,
"learning_rate": 3.0646377078137424e-06,
"loss": 0.7055673599243164,
"step": 872
},
{
"epoch": 1.9379157427937916,
"grad_norm": 4.75135612487793,
"learning_rate": 3.056463829817475e-06,
"loss": 0.6644902229309082,
"step": 874
},
{
"epoch": 1.942350332594235,
"grad_norm": 2.237863540649414,
"learning_rate": 3.048285828722345e-06,
"loss": 0.7358293533325195,
"step": 876
},
{
"epoch": 1.9467849223946785,
"grad_norm": 3.476473569869995,
"learning_rate": 3.0401038145534297e-06,
"loss": 0.8098105788230896,
"step": 878
},
{
"epoch": 1.951219512195122,
"grad_norm": 1.5437142848968506,
"learning_rate": 3.031917897389799e-06,
"loss": 0.8312487602233887,
"step": 880
},
{
"epoch": 1.9556541019955653,
"grad_norm": 3.081554412841797,
"learning_rate": 3.0237281873630335e-06,
"loss": 0.6966821551322937,
"step": 882
},
{
"epoch": 1.9600886917960088,
"grad_norm": 1.036048173904419,
"learning_rate": 3.0155347946557407e-06,
"loss": 0.8802693486213684,
"step": 884
},
{
"epoch": 1.9645232815964522,
"grad_norm": 0.683119535446167,
"learning_rate": 3.007337829500075e-06,
"loss": 1.0593414306640625,
"step": 886
},
{
"epoch": 1.9689578713968958,
"grad_norm": 0.7239016890525818,
"learning_rate": 2.999137402176255e-06,
"loss": 1.0278575420379639,
"step": 888
},
{
"epoch": 1.9733924611973392,
"grad_norm": 1.6972147226333618,
"learning_rate": 2.9909336230110747e-06,
"loss": 1.0861477851867676,
"step": 890
},
{
"epoch": 1.9778270509977827,
"grad_norm": 1.00822114944458,
"learning_rate": 2.9827266023764274e-06,
"loss": 1.1251873970031738,
"step": 892
},
{
"epoch": 1.9822616407982263,
"grad_norm": 0.9262673854827881,
"learning_rate": 2.9745164506878134e-06,
"loss": 1.1135941743850708,
"step": 894
},
{
"epoch": 1.9866962305986697,
"grad_norm": 1.560065746307373,
"learning_rate": 2.9663032784028596e-06,
"loss": 0.8289718627929688,
"step": 896
},
{
"epoch": 1.9911308203991132,
"grad_norm": 1.1300979852676392,
"learning_rate": 2.9580871960198297e-06,
"loss": 1.0270378589630127,
"step": 898
},
{
"epoch": 1.9955654101995566,
"grad_norm": 1.9531309604644775,
"learning_rate": 2.949868314076142e-06,
"loss": 0.792127251625061,
"step": 900
},
{
"epoch": 2.0,
"grad_norm": 0.6045248508453369,
"learning_rate": 2.941646743146875e-06,
"loss": 1.0664583444595337,
"step": 902
},
{
"epoch": 2.0044345898004434,
"grad_norm": 0.8053255081176758,
"learning_rate": 2.9334225938432868e-06,
"loss": 0.8489485383033752,
"step": 904
},
{
"epoch": 2.008869179600887,
"grad_norm": 1.6182339191436768,
"learning_rate": 2.925195976811326e-06,
"loss": 0.9491725564002991,
"step": 906
},
{
"epoch": 2.0133037694013303,
"grad_norm": 0.8146681189537048,
"learning_rate": 2.9169670027301387e-06,
"loss": 1.0454566478729248,
"step": 908
},
{
"epoch": 2.0177383592017737,
"grad_norm": 0.9472193717956543,
"learning_rate": 2.9087357823105843e-06,
"loss": 0.9005047082901001,
"step": 910
},
{
"epoch": 2.022172949002217,
"grad_norm": 1.0615944862365723,
"learning_rate": 2.9005024262937427e-06,
"loss": 0.6029907464981079,
"step": 912
},
{
"epoch": 2.0266075388026605,
"grad_norm": 1.9702504873275757,
"learning_rate": 2.8922670454494247e-06,
"loss": 0.656877875328064,
"step": 914
},
{
"epoch": 2.0310421286031044,
"grad_norm": 0.9048966765403748,
"learning_rate": 2.8840297505746843e-06,
"loss": 0.6500725746154785,
"step": 916
},
{
"epoch": 2.035476718403548,
"grad_norm": 0.9751022458076477,
"learning_rate": 2.8757906524923286e-06,
"loss": 0.8578327298164368,
"step": 918
},
{
"epoch": 2.0399113082039912,
"grad_norm": 0.8147334456443787,
"learning_rate": 2.867549862049419e-06,
"loss": 0.6973807215690613,
"step": 920
},
{
"epoch": 2.0443458980044347,
"grad_norm": 0.9924507141113281,
"learning_rate": 2.859307490115791e-06,
"loss": 0.7475817799568176,
"step": 922
},
{
"epoch": 2.048780487804878,
"grad_norm": 0.5117158889770508,
"learning_rate": 2.8510636475825533e-06,
"loss": 0.24903425574302673,
"step": 924
},
{
"epoch": 2.0532150776053215,
"grad_norm": 1.5849206447601318,
"learning_rate": 2.8428184453606027e-06,
"loss": 0.849932074546814,
"step": 926
},
{
"epoch": 2.057649667405765,
"grad_norm": 2.9684829711914062,
"learning_rate": 2.8345719943791266e-06,
"loss": 0.6786062121391296,
"step": 928
},
{
"epoch": 2.0620842572062084,
"grad_norm": 0.6647122502326965,
"learning_rate": 2.826324405584114e-06,
"loss": 0.6096989512443542,
"step": 930
},
{
"epoch": 2.066518847006652,
"grad_norm": 3.739225387573242,
"learning_rate": 2.818075789936863e-06,
"loss": 0.7550086379051208,
"step": 932
},
{
"epoch": 2.070953436807095,
"grad_norm": 1.7136473655700684,
"learning_rate": 2.8098262584124834e-06,
"loss": 0.8016495704650879,
"step": 934
},
{
"epoch": 2.0753880266075386,
"grad_norm": 0.6798487305641174,
"learning_rate": 2.801575921998411e-06,
"loss": 0.9375801682472229,
"step": 936
},
{
"epoch": 2.079822616407982,
"grad_norm": 0.6735354661941528,
"learning_rate": 2.7933248916929066e-06,
"loss": 0.6009020805358887,
"step": 938
},
{
"epoch": 2.084257206208426,
"grad_norm": 0.593377947807312,
"learning_rate": 2.7850732785035705e-06,
"loss": 0.6308031678199768,
"step": 940
},
{
"epoch": 2.0886917960088693,
"grad_norm": 1.553142786026001,
"learning_rate": 2.7768211934458417e-06,
"loss": 0.8890527486801147,
"step": 942
},
{
"epoch": 2.0931263858093128,
"grad_norm": 0.8441118001937866,
"learning_rate": 2.768568747541509e-06,
"loss": 0.5294168591499329,
"step": 944
},
{
"epoch": 2.097560975609756,
"grad_norm": 1.1496201753616333,
"learning_rate": 2.7603160518172152e-06,
"loss": 0.8766027688980103,
"step": 946
},
{
"epoch": 2.1019955654101996,
"grad_norm": 1.2092889547348022,
"learning_rate": 2.752063217302966e-06,
"loss": 0.9853121042251587,
"step": 948
},
{
"epoch": 2.106430155210643,
"grad_norm": 2.6365861892700195,
"learning_rate": 2.743810355030631e-06,
"loss": 0.6440744996070862,
"step": 950
},
{
"epoch": 2.1108647450110865,
"grad_norm": 1.4584583044052124,
"learning_rate": 2.735557576032458e-06,
"loss": 0.7647169828414917,
"step": 952
},
{
"epoch": 2.11529933481153,
"grad_norm": 3.336189031600952,
"learning_rate": 2.727304991339569e-06,
"loss": 0.8036750555038452,
"step": 954
},
{
"epoch": 2.1197339246119733,
"grad_norm": 0.6153157353401184,
"learning_rate": 2.7190527119804762e-06,
"loss": 0.8988800644874573,
"step": 956
},
{
"epoch": 2.1241685144124167,
"grad_norm": 9.374985694885254,
"learning_rate": 2.710800848979582e-06,
"loss": 0.7791066765785217,
"step": 958
},
{
"epoch": 2.12860310421286,
"grad_norm": 1.6065351963043213,
"learning_rate": 2.702549513355687e-06,
"loss": 0.7849658131599426,
"step": 960
},
{
"epoch": 2.1330376940133036,
"grad_norm": 4.621856689453125,
"learning_rate": 2.694298816120497e-06,
"loss": 0.5346379280090332,
"step": 962
},
{
"epoch": 2.1374722838137474,
"grad_norm": 1.9854819774627686,
"learning_rate": 2.6860488682771306e-06,
"loss": 1.0068566799163818,
"step": 964
},
{
"epoch": 2.141906873614191,
"grad_norm": 0.6731406450271606,
"learning_rate": 2.67779978081862e-06,
"loss": 0.9128319025039673,
"step": 966
},
{
"epoch": 2.1463414634146343,
"grad_norm": 0.8809360861778259,
"learning_rate": 2.669551664726428e-06,
"loss": 0.6138067245483398,
"step": 968
},
{
"epoch": 2.1507760532150777,
"grad_norm": 0.7070305943489075,
"learning_rate": 2.6613046309689433e-06,
"loss": 0.7260922193527222,
"step": 970
},
{
"epoch": 2.155210643015521,
"grad_norm": 0.9328792691230774,
"learning_rate": 2.6530587904999966e-06,
"loss": 0.5952677726745605,
"step": 972
},
{
"epoch": 2.1596452328159645,
"grad_norm": 1.6889206171035767,
"learning_rate": 2.6448142542573624e-06,
"loss": 0.933496356010437,
"step": 974
},
{
"epoch": 2.164079822616408,
"grad_norm": 1.7583260536193848,
"learning_rate": 2.6365711331612692e-06,
"loss": 0.906917929649353,
"step": 976
},
{
"epoch": 2.1685144124168514,
"grad_norm": 2.437592029571533,
"learning_rate": 2.6283295381129066e-06,
"loss": 0.6723505258560181,
"step": 978
},
{
"epoch": 2.172949002217295,
"grad_norm": 1.2844873666763306,
"learning_rate": 2.620089579992933e-06,
"loss": 0.9519558548927307,
"step": 980
},
{
"epoch": 2.1773835920177382,
"grad_norm": 0.7902489900588989,
"learning_rate": 2.6118513696599823e-06,
"loss": 0.5118272304534912,
"step": 982
},
{
"epoch": 2.1818181818181817,
"grad_norm": 1.085485577583313,
"learning_rate": 2.603615017949178e-06,
"loss": 0.31727081537246704,
"step": 984
},
{
"epoch": 2.186252771618625,
"grad_norm": 2.681396722793579,
"learning_rate": 2.595380635670634e-06,
"loss": 0.6958884596824646,
"step": 986
},
{
"epoch": 2.1906873614190685,
"grad_norm": 2.31773042678833,
"learning_rate": 2.5871483336079694e-06,
"loss": 0.4852255582809448,
"step": 988
},
{
"epoch": 2.1951219512195124,
"grad_norm": 0.9662159085273743,
"learning_rate": 2.578918222516818e-06,
"loss": 0.8706328868865967,
"step": 990
},
{
"epoch": 2.199556541019956,
"grad_norm": 1.4837191104888916,
"learning_rate": 2.5706904131233336e-06,
"loss": 1.0283949375152588,
"step": 992
},
{
"epoch": 2.203991130820399,
"grad_norm": 0.7699248194694519,
"learning_rate": 2.5624650161227073e-06,
"loss": 0.8138683438301086,
"step": 994
},
{
"epoch": 2.2084257206208426,
"grad_norm": 2.567706346511841,
"learning_rate": 2.5542421421776696e-06,
"loss": 0.9206845760345459,
"step": 996
},
{
"epoch": 2.212860310421286,
"grad_norm": 1.1012285947799683,
"learning_rate": 2.5460219019170097e-06,
"loss": 0.8530703783035278,
"step": 998
},
{
"epoch": 2.2172949002217295,
"grad_norm": 0.972668468952179,
"learning_rate": 2.5378044059340845e-06,
"loss": 0.7263464331626892,
"step": 1000
},
{
"epoch": 2.221729490022173,
"grad_norm": 29.759498596191406,
"learning_rate": 2.5295897647853283e-06,
"loss": 0.42510873079299927,
"step": 1002
},
{
"epoch": 2.2261640798226163,
"grad_norm": 0.734957218170166,
"learning_rate": 2.521378088988767e-06,
"loss": 0.9547646641731262,
"step": 1004
},
{
"epoch": 2.2305986696230597,
"grad_norm": 1.0839177370071411,
"learning_rate": 2.513169489022531e-06,
"loss": 0.8016040921211243,
"step": 1006
},
{
"epoch": 2.235033259423503,
"grad_norm": 0.6618872880935669,
"learning_rate": 2.5049640753233705e-06,
"loss": 0.8940417766571045,
"step": 1008
},
{
"epoch": 2.2394678492239466,
"grad_norm": 1.0671074390411377,
"learning_rate": 2.496761958285167e-06,
"loss": 0.5136449933052063,
"step": 1010
},
{
"epoch": 2.2439024390243905,
"grad_norm": 0.7268696427345276,
"learning_rate": 2.488563248257451e-06,
"loss": 0.8877573609352112,
"step": 1012
},
{
"epoch": 2.248337028824834,
"grad_norm": 4.092066764831543,
"learning_rate": 2.4803680555439136e-06,
"loss": 0.769023060798645,
"step": 1014
},
{
"epoch": 2.2527716186252773,
"grad_norm": 1.659173846244812,
"learning_rate": 2.4721764904009272e-06,
"loss": 0.8959583640098572,
"step": 1016
},
{
"epoch": 2.2572062084257207,
"grad_norm": 2.1315901279449463,
"learning_rate": 2.4639886630360574e-06,
"loss": 0.5399938821792603,
"step": 1018
},
{
"epoch": 2.261640798226164,
"grad_norm": 0.9298582673072815,
"learning_rate": 2.455804683606584e-06,
"loss": 0.9447622895240784,
"step": 1020
},
{
"epoch": 2.2660753880266076,
"grad_norm": 0.33580493927001953,
"learning_rate": 2.4476246622180174e-06,
"loss": 0.4857233166694641,
"step": 1022
},
{
"epoch": 2.270509977827051,
"grad_norm": 1.8351166248321533,
"learning_rate": 2.4394487089226158e-06,
"loss": 1.283814787864685,
"step": 1024
},
{
"epoch": 2.2749445676274944,
"grad_norm": 0.8674602508544922,
"learning_rate": 2.43127693371791e-06,
"loss": 0.6131373643875122,
"step": 1026
},
{
"epoch": 2.279379157427938,
"grad_norm": 1.49549400806427,
"learning_rate": 2.423109446545213e-06,
"loss": 0.9470140337944031,
"step": 1028
},
{
"epoch": 2.2838137472283813,
"grad_norm": 0.5943393707275391,
"learning_rate": 2.4149463572881537e-06,
"loss": 0.9135305881500244,
"step": 1030
},
{
"epoch": 2.2882483370288247,
"grad_norm": 1.793879508972168,
"learning_rate": 2.4067877757711907e-06,
"loss": 0.6614438891410828,
"step": 1032
},
{
"epoch": 2.292682926829268,
"grad_norm": 1.0091503858566284,
"learning_rate": 2.3986338117581357e-06,
"loss": 0.7243059277534485,
"step": 1034
},
{
"epoch": 2.2971175166297115,
"grad_norm": 1.3116031885147095,
"learning_rate": 2.390484574950677e-06,
"loss": 0.9390950798988342,
"step": 1036
},
{
"epoch": 2.3015521064301554,
"grad_norm": 2.461923837661743,
"learning_rate": 2.382340174986906e-06,
"loss": 0.6997823715209961,
"step": 1038
},
{
"epoch": 2.305986696230599,
"grad_norm": 1.6579372882843018,
"learning_rate": 2.374200721439837e-06,
"loss": 0.9242331385612488,
"step": 1040
},
{
"epoch": 2.3104212860310422,
"grad_norm": 1.2097886800765991,
"learning_rate": 2.3660663238159405e-06,
"loss": 1.012819766998291,
"step": 1042
},
{
"epoch": 2.3148558758314857,
"grad_norm": 1.4017579555511475,
"learning_rate": 2.357937091553662e-06,
"loss": 0.6412765383720398,
"step": 1044
},
{
"epoch": 2.319290465631929,
"grad_norm": 1.0496717691421509,
"learning_rate": 2.3498131340219554e-06,
"loss": 0.9358582496643066,
"step": 1046
},
{
"epoch": 2.3237250554323725,
"grad_norm": 0.7849573493003845,
"learning_rate": 2.341694560518809e-06,
"loss": 0.8715764880180359,
"step": 1048
},
{
"epoch": 2.328159645232816,
"grad_norm": 0.8973456621170044,
"learning_rate": 2.333581480269776e-06,
"loss": 0.9279825687408447,
"step": 1050
},
{
"epoch": 2.3325942350332594,
"grad_norm": 0.914661169052124,
"learning_rate": 2.325474002426503e-06,
"loss": 0.8819960951805115,
"step": 1052
},
{
"epoch": 2.337028824833703,
"grad_norm": 0.6689122915267944,
"learning_rate": 2.3173722360652644e-06,
"loss": 0.6645777821540833,
"step": 1054
},
{
"epoch": 2.341463414634146,
"grad_norm": 0.7918397188186646,
"learning_rate": 2.309276290185494e-06,
"loss": 0.8606789708137512,
"step": 1056
},
{
"epoch": 2.3458980044345896,
"grad_norm": 0.44345715641975403,
"learning_rate": 2.3011862737083162e-06,
"loss": 0.5443306565284729,
"step": 1058
},
{
"epoch": 2.3503325942350335,
"grad_norm": 0.6253224015235901,
"learning_rate": 2.2931022954750843e-06,
"loss": 1.009533405303955,
"step": 1060
},
{
"epoch": 2.354767184035477,
"grad_norm": 2.641712188720703,
"learning_rate": 2.285024464245912e-06,
"loss": 0.4540158808231354,
"step": 1062
},
{
"epoch": 2.3592017738359203,
"grad_norm": 0.5843250751495361,
"learning_rate": 2.2769528886982158e-06,
"loss": 0.49174901843070984,
"step": 1064
},
{
"epoch": 2.3636363636363638,
"grad_norm": 3.842958927154541,
"learning_rate": 2.268887677425248e-06,
"loss": 0.7375195026397705,
"step": 1066
},
{
"epoch": 2.368070953436807,
"grad_norm": 1.1907048225402832,
"learning_rate": 2.2608289389346362e-06,
"loss": 1.0150161981582642,
"step": 1068
},
{
"epoch": 2.3725055432372506,
"grad_norm": 0.834156334400177,
"learning_rate": 2.2527767816469263e-06,
"loss": 0.8399034142494202,
"step": 1070
},
{
"epoch": 2.376940133037694,
"grad_norm": 0.8549678325653076,
"learning_rate": 2.244731313894121e-06,
"loss": 0.87836754322052,
"step": 1072
},
{
"epoch": 2.3813747228381374,
"grad_norm": 0.5392049551010132,
"learning_rate": 2.236692643918224e-06,
"loss": 0.040761929005384445,
"step": 1074
},
{
"epoch": 2.385809312638581,
"grad_norm": 2.3668556213378906,
"learning_rate": 2.2286608798697834e-06,
"loss": 0.5826558470726013,
"step": 1076
},
{
"epoch": 2.3902439024390243,
"grad_norm": 0.8297257423400879,
"learning_rate": 2.2206361298064343e-06,
"loss": 0.29340535402297974,
"step": 1078
},
{
"epoch": 2.3946784922394677,
"grad_norm": 11.721139907836914,
"learning_rate": 2.2126185016914515e-06,
"loss": 0.8513185381889343,
"step": 1080
},
{
"epoch": 2.399113082039911,
"grad_norm": 0.919727623462677,
"learning_rate": 2.2046081033922884e-06,
"loss": 0.5718480348587036,
"step": 1082
},
{
"epoch": 2.4035476718403546,
"grad_norm": 2.098891019821167,
"learning_rate": 2.1966050426791325e-06,
"loss": 0.7035835385322571,
"step": 1084
},
{
"epoch": 2.4079822616407984,
"grad_norm": 0.9428548216819763,
"learning_rate": 2.1886094272234508e-06,
"loss": 0.906488299369812,
"step": 1086
},
{
"epoch": 2.412416851441242,
"grad_norm": 0.5806995630264282,
"learning_rate": 2.1806213645965457e-06,
"loss": 0.29125481843948364,
"step": 1088
},
{
"epoch": 2.4168514412416853,
"grad_norm": 0.8353445529937744,
"learning_rate": 2.172640962268104e-06,
"loss": 0.9137746691703796,
"step": 1090
},
{
"epoch": 2.4212860310421287,
"grad_norm": 1.1309226751327515,
"learning_rate": 2.1646683276047525e-06,
"loss": 0.9136351943016052,
"step": 1092
},
{
"epoch": 2.425720620842572,
"grad_norm": 1.4746575355529785,
"learning_rate": 2.156703567868615e-06,
"loss": 0.4814295470714569,
"step": 1094
},
{
"epoch": 2.4301552106430155,
"grad_norm": 4.6678032875061035,
"learning_rate": 2.148746790215866e-06,
"loss": 0.5724620819091797,
"step": 1096
},
{
"epoch": 2.434589800443459,
"grad_norm": 0.5765445828437805,
"learning_rate": 2.140798101695291e-06,
"loss": 0.45596760511398315,
"step": 1098
},
{
"epoch": 2.4390243902439024,
"grad_norm": 3.441075086593628,
"learning_rate": 2.1328576092468476e-06,
"loss": 1.0661569833755493,
"step": 1100
},
{
"epoch": 2.443458980044346,
"grad_norm": 1.3114399909973145,
"learning_rate": 2.124925419700223e-06,
"loss": 0.9525973796844482,
"step": 1102
},
{
"epoch": 2.4478935698447892,
"grad_norm": 1.7259883880615234,
"learning_rate": 2.1170016397734e-06,
"loss": 0.45727112889289856,
"step": 1104
},
{
"epoch": 2.4523281596452327,
"grad_norm": 0.6469590067863464,
"learning_rate": 2.109086376071221e-06,
"loss": 0.940199077129364,
"step": 1106
},
{
"epoch": 2.4567627494456765,
"grad_norm": 3.1953182220458984,
"learning_rate": 2.1011797350839513e-06,
"loss": 0.8561551570892334,
"step": 1108
},
{
"epoch": 2.4611973392461195,
"grad_norm": 1.3489811420440674,
"learning_rate": 2.093281823185848e-06,
"loss": 0.9693219661712646,
"step": 1110
},
{
"epoch": 2.4656319290465634,
"grad_norm": 0.7777767777442932,
"learning_rate": 2.0853927466337315e-06,
"loss": 0.6467586755752563,
"step": 1112
},
{
"epoch": 2.470066518847007,
"grad_norm": 1.0480725765228271,
"learning_rate": 2.077512611565551e-06,
"loss": 0.870927095413208,
"step": 1114
},
{
"epoch": 2.47450110864745,
"grad_norm": 1.7313536405563354,
"learning_rate": 2.0696415239989593e-06,
"loss": 0.325950562953949,
"step": 1116
},
{
"epoch": 2.4789356984478936,
"grad_norm": 0.7683355212211609,
"learning_rate": 2.0617795898298855e-06,
"loss": 0.8618558645248413,
"step": 1118
},
{
"epoch": 2.483370288248337,
"grad_norm": 0.7160274386405945,
"learning_rate": 2.053926914831112e-06,
"loss": 0.8484733700752258,
"step": 1120
},
{
"epoch": 2.4878048780487805,
"grad_norm": 0.9705782532691956,
"learning_rate": 2.04608360465085e-06,
"loss": 0.9692363142967224,
"step": 1122
},
{
"epoch": 2.492239467849224,
"grad_norm": 4.628700256347656,
"learning_rate": 2.038249764811318e-06,
"loss": 0.9982001781463623,
"step": 1124
},
{
"epoch": 2.4966740576496673,
"grad_norm": 0.741965115070343,
"learning_rate": 2.0304255007073227e-06,
"loss": 0.9597415328025818,
"step": 1126
},
{
"epoch": 2.5011086474501107,
"grad_norm": 1.0579674243927002,
"learning_rate": 2.022610917604842e-06,
"loss": 0.6873862147331238,
"step": 1128
},
{
"epoch": 2.505543237250554,
"grad_norm": 3.054508686065674,
"learning_rate": 2.014806120639605e-06,
"loss": 0.6469390392303467,
"step": 1130
},
{
"epoch": 2.5099778270509976,
"grad_norm": 0.8802086114883423,
"learning_rate": 2.007011214815684e-06,
"loss": 0.7718120813369751,
"step": 1132
},
{
"epoch": 2.5144124168514415,
"grad_norm": 0.6990912556648254,
"learning_rate": 1.9992263050040737e-06,
"loss": 0.5093148350715637,
"step": 1134
},
{
"epoch": 2.5188470066518844,
"grad_norm": 1.247750997543335,
"learning_rate": 1.991451495941289e-06,
"loss": 1.002577543258667,
"step": 1136
},
{
"epoch": 2.5232815964523283,
"grad_norm": 0.7929804921150208,
"learning_rate": 1.983686892227948e-06,
"loss": 0.7795249223709106,
"step": 1138
},
{
"epoch": 2.5277161862527717,
"grad_norm": 1.0441555976867676,
"learning_rate": 1.975932598327369e-06,
"loss": 0.9324785470962524,
"step": 1140
},
{
"epoch": 2.532150776053215,
"grad_norm": 1.305307149887085,
"learning_rate": 1.9681887185641646e-06,
"loss": 0.4589383602142334,
"step": 1142
},
{
"epoch": 2.5365853658536586,
"grad_norm": 0.3339233696460724,
"learning_rate": 1.9604553571228395e-06,
"loss": 0.5734773874282837,
"step": 1144
},
{
"epoch": 2.541019955654102,
"grad_norm": 0.8641642332077026,
"learning_rate": 1.9527326180463855e-06,
"loss": 0.8844019770622253,
"step": 1146
},
{
"epoch": 2.5454545454545454,
"grad_norm": 0.8415119647979736,
"learning_rate": 1.9450206052348823e-06,
"loss": 0.8842139840126038,
"step": 1148
},
{
"epoch": 2.549889135254989,
"grad_norm": 0.9703028798103333,
"learning_rate": 1.9373194224441028e-06,
"loss": 0.8934570550918579,
"step": 1150
},
{
"epoch": 2.5543237250554323,
"grad_norm": 1.8766459226608276,
"learning_rate": 1.929629173284114e-06,
"loss": 0.63346928358078,
"step": 1152
},
{
"epoch": 2.5587583148558757,
"grad_norm": 0.8151181936264038,
"learning_rate": 1.9219499612178836e-06,
"loss": 0.30481529235839844,
"step": 1154
},
{
"epoch": 2.5631929046563195,
"grad_norm": 0.7087282538414001,
"learning_rate": 1.9142818895598908e-06,
"loss": 0.5541834831237793,
"step": 1156
},
{
"epoch": 2.5676274944567625,
"grad_norm": 2.2988929748535156,
"learning_rate": 1.9066250614747317e-06,
"loss": 0.6360606551170349,
"step": 1158
},
{
"epoch": 2.5720620842572064,
"grad_norm": 1.320077896118164,
"learning_rate": 1.8989795799757348e-06,
"loss": 0.824614405632019,
"step": 1160
},
{
"epoch": 2.57649667405765,
"grad_norm": 0.7024682760238647,
"learning_rate": 1.8913455479235754e-06,
"loss": 0.8806930184364319,
"step": 1162
},
{
"epoch": 2.5809312638580932,
"grad_norm": 0.8074402809143066,
"learning_rate": 1.8837230680248874e-06,
"loss": 0.8695907592773438,
"step": 1164
},
{
"epoch": 2.5853658536585367,
"grad_norm": 0.8083245158195496,
"learning_rate": 1.8761122428308875e-06,
"loss": 0.6029998660087585,
"step": 1166
},
{
"epoch": 2.58980044345898,
"grad_norm": 0.8537569642066956,
"learning_rate": 1.8685131747359902e-06,
"loss": 0.9235411882400513,
"step": 1168
},
{
"epoch": 2.5942350332594235,
"grad_norm": 1.1320568323135376,
"learning_rate": 1.8609259659764345e-06,
"loss": 0.8941707611083984,
"step": 1170
},
{
"epoch": 2.598669623059867,
"grad_norm": 2.4210267066955566,
"learning_rate": 1.853350718628904e-06,
"loss": 0.3670371174812317,
"step": 1172
},
{
"epoch": 2.6031042128603104,
"grad_norm": 1.3128033876419067,
"learning_rate": 1.845787534609157e-06,
"loss": 0.6117712259292603,
"step": 1174
},
{
"epoch": 2.6075388026607538,
"grad_norm": 0.8735086917877197,
"learning_rate": 1.8382365156706566e-06,
"loss": 0.6652023196220398,
"step": 1176
},
{
"epoch": 2.611973392461197,
"grad_norm": 0.6818370223045349,
"learning_rate": 1.8306977634031976e-06,
"loss": 0.5446036458015442,
"step": 1178
},
{
"epoch": 2.6164079822616406,
"grad_norm": 3.452026844024658,
"learning_rate": 1.8231713792315403e-06,
"loss": 0.7938367128372192,
"step": 1180
},
{
"epoch": 2.6208425720620845,
"grad_norm": 0.9159925580024719,
"learning_rate": 1.8156574644140495e-06,
"loss": 0.8983339667320251,
"step": 1182
},
{
"epoch": 2.6252771618625275,
"grad_norm": 0.8673039674758911,
"learning_rate": 1.8081561200413295e-06,
"loss": 0.8714417815208435,
"step": 1184
},
{
"epoch": 2.6297117516629713,
"grad_norm": 1.4530643224716187,
"learning_rate": 1.800667447034864e-06,
"loss": 0.874497652053833,
"step": 1186
},
{
"epoch": 2.6341463414634148,
"grad_norm": 2.0185797214508057,
"learning_rate": 1.7931915461456573e-06,
"loss": 1.0262384414672852,
"step": 1188
},
{
"epoch": 2.638580931263858,
"grad_norm": 0.7627118229866028,
"learning_rate": 1.7857285179528838e-06,
"loss": 0.5378797054290771,
"step": 1190
},
{
"epoch": 2.6430155210643016,
"grad_norm": 1.8396339416503906,
"learning_rate": 1.7782784628625305e-06,
"loss": 0.7439752221107483,
"step": 1192
},
{
"epoch": 2.647450110864745,
"grad_norm": 1.5848437547683716,
"learning_rate": 1.7708414811060437e-06,
"loss": 0.5809391736984253,
"step": 1194
},
{
"epoch": 2.6518847006651884,
"grad_norm": 0.9342997074127197,
"learning_rate": 1.763417672738989e-06,
"loss": 0.7553034424781799,
"step": 1196
},
{
"epoch": 2.656319290465632,
"grad_norm": 0.5442282557487488,
"learning_rate": 1.7560071376396953e-06,
"loss": 0.25871163606643677,
"step": 1198
},
{
"epoch": 2.6607538802660753,
"grad_norm": 0.8709027767181396,
"learning_rate": 1.7486099755079197e-06,
"loss": 0.9024768471717834,
"step": 1200
},
{
"epoch": 2.6651884700665187,
"grad_norm": 0.8886547088623047,
"learning_rate": 1.7412262858634987e-06,
"loss": 0.9686548113822937,
"step": 1202
},
{
"epoch": 2.6696230598669626,
"grad_norm": 0.6681715250015259,
"learning_rate": 1.7338561680450171e-06,
"loss": 0.46400630474090576,
"step": 1204
},
{
"epoch": 2.6740576496674056,
"grad_norm": 1.2975910902023315,
"learning_rate": 1.7264997212084616e-06,
"loss": 0.7697018384933472,
"step": 1206
},
{
"epoch": 2.6784922394678494,
"grad_norm": 2.711646556854248,
"learning_rate": 1.7191570443258976e-06,
"loss": 1.056341290473938,
"step": 1208
},
{
"epoch": 2.682926829268293,
"grad_norm": 1.6570303440093994,
"learning_rate": 1.711828236184131e-06,
"loss": 0.5748533606529236,
"step": 1210
},
{
"epoch": 2.6873614190687363,
"grad_norm": 1.9034491777420044,
"learning_rate": 1.704513395383378e-06,
"loss": 0.31726494431495667,
"step": 1212
},
{
"epoch": 2.6917960088691797,
"grad_norm": 1.501547932624817,
"learning_rate": 1.6972126203359454e-06,
"loss": 0.6147481203079224,
"step": 1214
},
{
"epoch": 2.696230598669623,
"grad_norm": 3.692929744720459,
"learning_rate": 1.6899260092648995e-06,
"loss": 0.6757416129112244,
"step": 1216
},
{
"epoch": 2.7006651884700665,
"grad_norm": 1.5111517906188965,
"learning_rate": 1.6826536602027471e-06,
"loss": 0.6192297339439392,
"step": 1218
},
{
"epoch": 2.70509977827051,
"grad_norm": 1.2896053791046143,
"learning_rate": 1.6753956709901202e-06,
"loss": 1.0000026226043701,
"step": 1220
},
{
"epoch": 2.7095343680709534,
"grad_norm": 2.9973676204681396,
"learning_rate": 1.6681521392744515e-06,
"loss": 0.7289214730262756,
"step": 1222
},
{
"epoch": 2.713968957871397,
"grad_norm": 2.971303939819336,
"learning_rate": 1.660923162508671e-06,
"loss": 0.8310694694519043,
"step": 1224
},
{
"epoch": 2.7184035476718402,
"grad_norm": 2.4088239669799805,
"learning_rate": 1.6537088379498872e-06,
"loss": 0.47293317317962646,
"step": 1226
},
{
"epoch": 2.7228381374722836,
"grad_norm": 0.6797178387641907,
"learning_rate": 1.6465092626580787e-06,
"loss": 0.983069896697998,
"step": 1228
},
{
"epoch": 2.7272727272727275,
"grad_norm": 1.6718101501464844,
"learning_rate": 1.6393245334947942e-06,
"loss": 0.7170325517654419,
"step": 1230
},
{
"epoch": 2.7317073170731705,
"grad_norm": 0.7277650833129883,
"learning_rate": 1.6321547471218432e-06,
"loss": 0.882874608039856,
"step": 1232
},
{
"epoch": 2.7361419068736144,
"grad_norm": 1.7578508853912354,
"learning_rate": 1.6250000000000007e-06,
"loss": 0.5405250787734985,
"step": 1234
},
{
"epoch": 2.740576496674058,
"grad_norm": 0.9388027191162109,
"learning_rate": 1.6178603883877032e-06,
"loss": 0.8325910568237305,
"step": 1236
},
{
"epoch": 2.745011086474501,
"grad_norm": 0.7891237735748291,
"learning_rate": 1.6107360083397604e-06,
"loss": 0.5409272313117981,
"step": 1238
},
{
"epoch": 2.7494456762749446,
"grad_norm": 0.7106161713600159,
"learning_rate": 1.6036269557060594e-06,
"loss": 0.6126099824905396,
"step": 1240
},
{
"epoch": 2.753880266075388,
"grad_norm": 0.8075646162033081,
"learning_rate": 1.5965333261302735e-06,
"loss": 0.9320923089981079,
"step": 1242
},
{
"epoch": 2.7583148558758315,
"grad_norm": 0.8899918794631958,
"learning_rate": 1.5894552150485801e-06,
"loss": 1.0979969501495361,
"step": 1244
},
{
"epoch": 2.762749445676275,
"grad_norm": 3.5077619552612305,
"learning_rate": 1.5823927176883725e-06,
"loss": 0.4379834234714508,
"step": 1246
},
{
"epoch": 2.7671840354767183,
"grad_norm": 3.83512544631958,
"learning_rate": 1.5753459290669792e-06,
"loss": 0.6199414134025574,
"step": 1248
},
{
"epoch": 2.7716186252771617,
"grad_norm": 3.1204421520233154,
"learning_rate": 1.5683149439903905e-06,
"loss": 0.6263423562049866,
"step": 1250
},
{
"epoch": 2.776053215077605,
"grad_norm": 1.005743145942688,
"learning_rate": 1.5612998570519746e-06,
"loss": 0.9981138706207275,
"step": 1252
},
{
"epoch": 2.7804878048780486,
"grad_norm": 0.676094651222229,
"learning_rate": 1.5543007626312129e-06,
"loss": 1.003664255142212,
"step": 1254
},
{
"epoch": 2.7849223946784925,
"grad_norm": 0.5613422989845276,
"learning_rate": 1.5473177548924267e-06,
"loss": 0.9180096983909607,
"step": 1256
},
{
"epoch": 2.7893569844789354,
"grad_norm": 0.6209838390350342,
"learning_rate": 1.5403509277835077e-06,
"loss": 0.18954633176326752,
"step": 1258
},
{
"epoch": 2.7937915742793793,
"grad_norm": 0.7056342959403992,
"learning_rate": 1.5334003750346608e-06,
"loss": 0.8927637338638306,
"step": 1260
},
{
"epoch": 2.7982261640798227,
"grad_norm": 1.2110857963562012,
"learning_rate": 1.5264661901571349e-06,
"loss": 0.6852344870567322,
"step": 1262
},
{
"epoch": 2.802660753880266,
"grad_norm": 0.8435057997703552,
"learning_rate": 1.5195484664419732e-06,
"loss": 0.5791198015213013,
"step": 1264
},
{
"epoch": 2.8070953436807096,
"grad_norm": 0.7665271162986755,
"learning_rate": 1.5126472969587502e-06,
"loss": 1.0053340196609497,
"step": 1266
},
{
"epoch": 2.811529933481153,
"grad_norm": 1.9054136276245117,
"learning_rate": 1.5057627745543269e-06,
"loss": 0.9256460666656494,
"step": 1268
},
{
"epoch": 2.8159645232815964,
"grad_norm": 0.928837239742279,
"learning_rate": 1.4988949918515947e-06,
"loss": 0.872265100479126,
"step": 1270
},
{
"epoch": 2.82039911308204,
"grad_norm": 1.4256490468978882,
"learning_rate": 1.4920440412482345e-06,
"loss": 0.38903388381004333,
"step": 1272
},
{
"epoch": 2.8248337028824833,
"grad_norm": 1.0866613388061523,
"learning_rate": 1.485210014915473e-06,
"loss": 0.5502209663391113,
"step": 1274
},
{
"epoch": 2.8292682926829267,
"grad_norm": 0.8656668066978455,
"learning_rate": 1.4783930047968388e-06,
"loss": 0.9950301051139832,
"step": 1276
},
{
"epoch": 2.8337028824833705,
"grad_norm": 2.2869789600372314,
"learning_rate": 1.4715931026069273e-06,
"loss": 0.9035691618919373,
"step": 1278
},
{
"epoch": 2.8381374722838135,
"grad_norm": 0.892152726650238,
"learning_rate": 1.4648103998301716e-06,
"loss": 0.5569464564323425,
"step": 1280
},
{
"epoch": 2.8425720620842574,
"grad_norm": 2.0605618953704834,
"learning_rate": 1.4580449877196035e-06,
"loss": 0.6820242404937744,
"step": 1282
},
{
"epoch": 2.847006651884701,
"grad_norm": 0.8823329210281372,
"learning_rate": 1.4512969572956328e-06,
"loss": 0.6583420038223267,
"step": 1284
},
{
"epoch": 2.8514412416851442,
"grad_norm": 0.7509365677833557,
"learning_rate": 1.4445663993448173e-06,
"loss": 0.9154214262962341,
"step": 1286
},
{
"epoch": 2.8558758314855877,
"grad_norm": 2.52254319190979,
"learning_rate": 1.437853404418646e-06,
"loss": 0.31365761160850525,
"step": 1288
},
{
"epoch": 2.860310421286031,
"grad_norm": 0.8587144613265991,
"learning_rate": 1.431158062832318e-06,
"loss": 0.9376072883605957,
"step": 1290
},
{
"epoch": 2.8647450110864745,
"grad_norm": 1.4515576362609863,
"learning_rate": 1.4244804646635266e-06,
"loss": 1.0736974477767944,
"step": 1292
},
{
"epoch": 2.869179600886918,
"grad_norm": 0.9386347532272339,
"learning_rate": 1.4178206997512522e-06,
"loss": 0.8537707924842834,
"step": 1294
},
{
"epoch": 2.8736141906873613,
"grad_norm": 0.8694044947624207,
"learning_rate": 1.4111788576945467e-06,
"loss": 0.9072995781898499,
"step": 1296
},
{
"epoch": 2.8780487804878048,
"grad_norm": 1.4627509117126465,
"learning_rate": 1.4045550278513351e-06,
"loss": 0.5681540966033936,
"step": 1298
},
{
"epoch": 2.882483370288248,
"grad_norm": 1.8899142742156982,
"learning_rate": 1.3979492993372074e-06,
"loss": 0.9094551205635071,
"step": 1300
},
{
"epoch": 2.8869179600886916,
"grad_norm": 2.1445658206939697,
"learning_rate": 1.391361761024222e-06,
"loss": 0.3109537363052368,
"step": 1302
},
{
"epoch": 2.8913525498891355,
"grad_norm": 0.7595453858375549,
"learning_rate": 1.3847925015397146e-06,
"loss": 0.8516042828559875,
"step": 1304
},
{
"epoch": 2.8957871396895785,
"grad_norm": 0.9547167420387268,
"learning_rate": 1.3782416092650957e-06,
"loss": 0.7857693433761597,
"step": 1306
},
{
"epoch": 2.9002217294900223,
"grad_norm": 0.8833897113800049,
"learning_rate": 1.3717091723346699e-06,
"loss": 0.40065449476242065,
"step": 1308
},
{
"epoch": 2.9046563192904657,
"grad_norm": 0.7363049983978271,
"learning_rate": 1.3651952786344485e-06,
"loss": 0.518020749092102,
"step": 1310
},
{
"epoch": 2.909090909090909,
"grad_norm": 0.7660597562789917,
"learning_rate": 1.3587000158009638e-06,
"loss": 0.8922036290168762,
"step": 1312
},
{
"epoch": 2.9135254988913526,
"grad_norm": 0.7471197843551636,
"learning_rate": 1.3522234712200954e-06,
"loss": 0.9531146883964539,
"step": 1314
},
{
"epoch": 2.917960088691796,
"grad_norm": 2.3324310779571533,
"learning_rate": 1.3457657320258878e-06,
"loss": 0.6479524374008179,
"step": 1316
},
{
"epoch": 2.9223946784922394,
"grad_norm": 0.8263819813728333,
"learning_rate": 1.3393268850993852e-06,
"loss": 0.9447596073150635,
"step": 1318
},
{
"epoch": 2.926829268292683,
"grad_norm": 1.0561071634292603,
"learning_rate": 1.332907017067458e-06,
"loss": 0.7807310819625854,
"step": 1320
},
{
"epoch": 2.9312638580931263,
"grad_norm": 0.7978630065917969,
"learning_rate": 1.3265062143016378e-06,
"loss": 0.8912954926490784,
"step": 1322
},
{
"epoch": 2.9356984478935697,
"grad_norm": 1.006141185760498,
"learning_rate": 1.3201245629169574e-06,
"loss": 0.9309762716293335,
"step": 1324
},
{
"epoch": 2.9401330376940136,
"grad_norm": 0.6220606565475464,
"learning_rate": 1.3137621487707902e-06,
"loss": 0.8276110887527466,
"step": 1326
},
{
"epoch": 2.9445676274944566,
"grad_norm": 0.7150996923446655,
"learning_rate": 1.307419057461697e-06,
"loss": 0.9218543171882629,
"step": 1328
},
{
"epoch": 2.9490022172949004,
"grad_norm": 0.7995206713676453,
"learning_rate": 1.3010953743282724e-06,
"loss": 0.8906182050704956,
"step": 1330
},
{
"epoch": 2.953436807095344,
"grad_norm": 1.6357812881469727,
"learning_rate": 1.294791184447996e-06,
"loss": 0.907779335975647,
"step": 1332
},
{
"epoch": 2.9578713968957873,
"grad_norm": 0.705119788646698,
"learning_rate": 1.2885065726360925e-06,
"loss": 0.45520275831222534,
"step": 1334
},
{
"epoch": 2.9623059866962307,
"grad_norm": 0.8450793027877808,
"learning_rate": 1.282241623444386e-06,
"loss": 0.9082697629928589,
"step": 1336
},
{
"epoch": 2.966740576496674,
"grad_norm": 0.715003252029419,
"learning_rate": 1.2759964211601633e-06,
"loss": 0.8805263042449951,
"step": 1338
},
{
"epoch": 2.9711751662971175,
"grad_norm": 1.0061311721801758,
"learning_rate": 1.269771049805042e-06,
"loss": 0.7469978928565979,
"step": 1340
},
{
"epoch": 2.975609756097561,
"grad_norm": 2.486464262008667,
"learning_rate": 1.2635655931338364e-06,
"loss": 0.6190311312675476,
"step": 1342
},
{
"epoch": 2.9800443458980044,
"grad_norm": 0.5506015419960022,
"learning_rate": 1.2573801346334355e-06,
"loss": 0.19588389992713928,
"step": 1344
},
{
"epoch": 2.984478935698448,
"grad_norm": 0.7587252259254456,
"learning_rate": 1.251214757521675e-06,
"loss": 0.6986634731292725,
"step": 1346
},
{
"epoch": 2.988913525498891,
"grad_norm": 0.548460841178894,
"learning_rate": 1.2450695447462214e-06,
"loss": 0.5783390998840332,
"step": 1348
},
{
"epoch": 2.9933481152993346,
"grad_norm": 4.170194149017334,
"learning_rate": 1.2389445789834534e-06,
"loss": 0.6150118112564087,
"step": 1350
},
{
"epoch": 2.9977827050997785,
"grad_norm": 0.9792019128799438,
"learning_rate": 1.2328399426373511e-06,
"loss": 0.6677907109260559,
"step": 1352
},
{
"epoch": 3.002217294900222,
"grad_norm": 0.8542647361755371,
"learning_rate": 1.2267557178383886e-06,
"loss": 0.8499741554260254,
"step": 1354
},
{
"epoch": 3.0066518847006654,
"grad_norm": 0.7518671154975891,
"learning_rate": 1.220691986442424e-06,
"loss": 0.42442217469215393,
"step": 1356
},
{
"epoch": 3.011086474501109,
"grad_norm": 1.2137625217437744,
"learning_rate": 1.2146488300296047e-06,
"loss": 0.6506487727165222,
"step": 1358
},
{
"epoch": 3.015521064301552,
"grad_norm": 0.703611433506012,
"learning_rate": 1.2086263299032652e-06,
"loss": 0.8044725656509399,
"step": 1360
},
{
"epoch": 3.0199556541019956,
"grad_norm": 0.8881542682647705,
"learning_rate": 1.2026245670888343e-06,
"loss": 0.7103544473648071,
"step": 1362
},
{
"epoch": 3.024390243902439,
"grad_norm": 0.7568899393081665,
"learning_rate": 1.196643622332747e-06,
"loss": 0.8767702579498291,
"step": 1364
},
{
"epoch": 3.0288248337028825,
"grad_norm": 0.25662195682525635,
"learning_rate": 1.1906835761013547e-06,
"loss": 0.30692797899246216,
"step": 1366
},
{
"epoch": 3.033259423503326,
"grad_norm": 0.7841421961784363,
"learning_rate": 1.184744508579846e-06,
"loss": 0.46566513180732727,
"step": 1368
},
{
"epoch": 3.0376940133037693,
"grad_norm": 1.7681467533111572,
"learning_rate": 1.178826499671167e-06,
"loss": 0.5476133227348328,
"step": 1370
},
{
"epoch": 3.0421286031042127,
"grad_norm": 0.914473295211792,
"learning_rate": 1.172929628994943e-06,
"loss": 0.7700155377388,
"step": 1372
},
{
"epoch": 3.046563192904656,
"grad_norm": 1.8283075094223022,
"learning_rate": 1.167053975886413e-06,
"loss": 0.29538294672966003,
"step": 1374
},
{
"epoch": 3.0509977827050996,
"grad_norm": 2.4396283626556396,
"learning_rate": 1.1611996193953569e-06,
"loss": 0.4983513057231903,
"step": 1376
},
{
"epoch": 3.0554323725055434,
"grad_norm": 1.0929163694381714,
"learning_rate": 1.1553666382850366e-06,
"loss": 0.3278832733631134,
"step": 1378
},
{
"epoch": 3.059866962305987,
"grad_norm": 0.1607833206653595,
"learning_rate": 1.1495551110311324e-06,
"loss": 0.319016695022583,
"step": 1380
},
{
"epoch": 3.0643015521064303,
"grad_norm": 2.2611818313598633,
"learning_rate": 1.1437651158206904e-06,
"loss": 0.2882533073425293,
"step": 1382
},
{
"epoch": 3.0687361419068737,
"grad_norm": 1.358695387840271,
"learning_rate": 1.137996730551069e-06,
"loss": 0.5356056690216064,
"step": 1384
},
{
"epoch": 3.073170731707317,
"grad_norm": 0.34475207328796387,
"learning_rate": 1.1322500328288897e-06,
"loss": 0.31114962697029114,
"step": 1386
},
{
"epoch": 3.0776053215077606,
"grad_norm": 0.14836685359477997,
"learning_rate": 1.1265250999689966e-06,
"loss": 0.4476943016052246,
"step": 1388
},
{
"epoch": 3.082039911308204,
"grad_norm": 0.9479687809944153,
"learning_rate": 1.1208220089934118e-06,
"loss": 0.6670686602592468,
"step": 1390
},
{
"epoch": 3.0864745011086474,
"grad_norm": 2.1511411666870117,
"learning_rate": 1.1151408366303024e-06,
"loss": 0.31225770711898804,
"step": 1392
},
{
"epoch": 3.090909090909091,
"grad_norm": 3.655069589614868,
"learning_rate": 1.1094816593129475e-06,
"loss": 0.513544499874115,
"step": 1394
},
{
"epoch": 3.0953436807095343,
"grad_norm": 0.8022751212120056,
"learning_rate": 1.1038445531787083e-06,
"loss": 0.7358729243278503,
"step": 1396
},
{
"epoch": 3.0997782705099777,
"grad_norm": 0.9091879725456238,
"learning_rate": 1.098229594068007e-06,
"loss": 0.739376425743103,
"step": 1398
},
{
"epoch": 3.104212860310421,
"grad_norm": 0.7697157859802246,
"learning_rate": 1.0926368575233032e-06,
"loss": 0.562745213508606,
"step": 1400
},
{
"epoch": 3.1086474501108645,
"grad_norm": 0.6193404197692871,
"learning_rate": 1.087066418788078e-06,
"loss": 0.2242104858160019,
"step": 1402
},
{
"epoch": 3.1130820399113084,
"grad_norm": 1.587528944015503,
"learning_rate": 1.0815183528058248e-06,
"loss": 0.47392910718917847,
"step": 1404
},
{
"epoch": 3.117516629711752,
"grad_norm": 0.8710333108901978,
"learning_rate": 1.0759927342190362e-06,
"loss": 0.6808566451072693,
"step": 1406
},
{
"epoch": 3.1219512195121952,
"grad_norm": 0.8722860217094421,
"learning_rate": 1.0704896373682052e-06,
"loss": 0.8002766966819763,
"step": 1408
},
{
"epoch": 3.1263858093126387,
"grad_norm": 0.8693475723266602,
"learning_rate": 1.0650091362908189e-06,
"loss": 0.6717104315757751,
"step": 1410
},
{
"epoch": 3.130820399113082,
"grad_norm": 0.8810287714004517,
"learning_rate": 1.0595513047203693e-06,
"loss": 0.6879040598869324,
"step": 1412
},
{
"epoch": 3.1352549889135255,
"grad_norm": 1.0683605670928955,
"learning_rate": 1.0541162160853538e-06,
"loss": 0.4185694754123688,
"step": 1414
},
{
"epoch": 3.139689578713969,
"grad_norm": 0.7451454401016235,
"learning_rate": 1.0487039435082941e-06,
"loss": 0.057561103254556656,
"step": 1416
},
{
"epoch": 3.1441241685144123,
"grad_norm": 1.1728911399841309,
"learning_rate": 1.0433145598047495e-06,
"loss": 0.6490657925605774,
"step": 1418
},
{
"epoch": 3.1485587583148558,
"grad_norm": 0.90044105052948,
"learning_rate": 1.0379481374823358e-06,
"loss": 0.7569578886032104,
"step": 1420
},
{
"epoch": 3.152993348115299,
"grad_norm": 1.4448527097702026,
"learning_rate": 1.032604748739751e-06,
"loss": 0.6902734041213989,
"step": 1422
},
{
"epoch": 3.1574279379157426,
"grad_norm": 0.5912182927131653,
"learning_rate": 1.0272844654658069e-06,
"loss": 0.12579640746116638,
"step": 1424
},
{
"epoch": 3.1618625277161865,
"grad_norm": 1.5718578100204468,
"learning_rate": 1.0219873592384556e-06,
"loss": 0.6934728026390076,
"step": 1426
},
{
"epoch": 3.16629711751663,
"grad_norm": 1.5502285957336426,
"learning_rate": 1.016713501323834e-06,
"loss": 0.7339826822280884,
"step": 1428
},
{
"epoch": 3.1707317073170733,
"grad_norm": 3.2946419715881348,
"learning_rate": 1.0114629626752973e-06,
"loss": 0.4983486831188202,
"step": 1430
},
{
"epoch": 3.1751662971175167,
"grad_norm": 1.9185384511947632,
"learning_rate": 1.0062358139324715e-06,
"loss": 0.39363744854927063,
"step": 1432
},
{
"epoch": 3.17960088691796,
"grad_norm": 0.8971115946769714,
"learning_rate": 1.0010321254202992e-06,
"loss": 0.5175668597221375,
"step": 1434
},
{
"epoch": 3.1840354767184036,
"grad_norm": 2.1288814544677734,
"learning_rate": 9.958519671480919e-07,
"loss": 0.17975324392318726,
"step": 1436
},
{
"epoch": 3.188470066518847,
"grad_norm": 0.8228841423988342,
"learning_rate": 9.906954088085929e-07,
"loss": 0.42570653557777405,
"step": 1438
},
{
"epoch": 3.1929046563192904,
"grad_norm": 2.6906468868255615,
"learning_rate": 9.85562519777035e-07,
"loss": 0.585996150970459,
"step": 1440
},
{
"epoch": 3.197339246119734,
"grad_norm": 1.0107629299163818,
"learning_rate": 9.804533691102112e-07,
"loss": 0.7266984581947327,
"step": 1442
},
{
"epoch": 3.2017738359201773,
"grad_norm": 0.8374084830284119,
"learning_rate": 9.75368025545542e-07,
"loss": 0.7630691528320312,
"step": 1444
},
{
"epoch": 3.2062084257206207,
"grad_norm": 0.8851264119148254,
"learning_rate": 9.703065575001518e-07,
"loss": 0.6669173240661621,
"step": 1446
},
{
"epoch": 3.210643015521064,
"grad_norm": 4.757336139678955,
"learning_rate": 9.65269033069952e-07,
"loss": 0.6446712613105774,
"step": 1448
},
{
"epoch": 3.2150776053215075,
"grad_norm": 0.8732156753540039,
"learning_rate": 9.602555200287184e-07,
"loss": 0.8526778817176819,
"step": 1450
},
{
"epoch": 3.2195121951219514,
"grad_norm": 1.501761794090271,
"learning_rate": 9.552660858271835e-07,
"loss": 0.4225665330886841,
"step": 1452
},
{
"epoch": 3.223946784922395,
"grad_norm": 0.790913999080658,
"learning_rate": 9.503007975921294e-07,
"loss": 0.5034030675888062,
"step": 1454
},
{
"epoch": 3.2283813747228383,
"grad_norm": 2.5785648822784424,
"learning_rate": 9.453597221254821e-07,
"loss": 0.7649686336517334,
"step": 1456
},
{
"epoch": 3.2328159645232817,
"grad_norm": 5.038004398345947,
"learning_rate": 9.404429259034156e-07,
"loss": 0.45995619893074036,
"step": 1458
},
{
"epoch": 3.237250554323725,
"grad_norm": 0.8205391764640808,
"learning_rate": 9.355504750754543e-07,
"loss": 0.8184725046157837,
"step": 1460
},
{
"epoch": 3.2416851441241685,
"grad_norm": 0.9232129454612732,
"learning_rate": 9.306824354635866e-07,
"loss": 0.5478826761245728,
"step": 1462
},
{
"epoch": 3.246119733924612,
"grad_norm": 2.364551067352295,
"learning_rate": 9.258388725613776e-07,
"loss": 0.69716876745224,
"step": 1464
},
{
"epoch": 3.2505543237250554,
"grad_norm": 1.0339198112487793,
"learning_rate": 9.21019851533086e-07,
"loss": 0.8049119710922241,
"step": 1466
},
{
"epoch": 3.254988913525499,
"grad_norm": 0.22390811145305634,
"learning_rate": 9.162254372127921e-07,
"loss": 0.12469253689050674,
"step": 1468
},
{
"epoch": 3.259423503325942,
"grad_norm": 1.6547586917877197,
"learning_rate": 9.114556941035199e-07,
"loss": 0.8042660355567932,
"step": 1470
},
{
"epoch": 3.2638580931263856,
"grad_norm": 1.0059466361999512,
"learning_rate": 9.067106863763752e-07,
"loss": 0.8942850828170776,
"step": 1472
},
{
"epoch": 3.2682926829268295,
"grad_norm": 3.251643657684326,
"learning_rate": 9.01990477869677e-07,
"loss": 0.3387027978897095,
"step": 1474
},
{
"epoch": 3.2727272727272725,
"grad_norm": 0.6991317272186279,
"learning_rate": 8.972951320881014e-07,
"loss": 0.5067039728164673,
"step": 1476
},
{
"epoch": 3.2771618625277164,
"grad_norm": 0.7981709241867065,
"learning_rate": 8.92624712201827e-07,
"loss": 0.4391624331474304,
"step": 1478
},
{
"epoch": 3.2815964523281598,
"grad_norm": 0.6799890398979187,
"learning_rate": 8.879792810456861e-07,
"loss": 0.5848779678344727,
"step": 1480
},
{
"epoch": 3.286031042128603,
"grad_norm": 0.9311388731002808,
"learning_rate": 8.833589011183147e-07,
"loss": 0.6908966898918152,
"step": 1482
},
{
"epoch": 3.2904656319290466,
"grad_norm": 0.7512919306755066,
"learning_rate": 8.78763634581318e-07,
"loss": 0.541054904460907,
"step": 1484
},
{
"epoch": 3.29490022172949,
"grad_norm": 0.7206840515136719,
"learning_rate": 8.741935432584292e-07,
"loss": 0.4378011226654053,
"step": 1486
},
{
"epoch": 3.2993348115299335,
"grad_norm": 26.240053176879883,
"learning_rate": 8.696486886346805e-07,
"loss": 0.4610789716243744,
"step": 1488
},
{
"epoch": 3.303769401330377,
"grad_norm": 2.8554205894470215,
"learning_rate": 8.651291318555745e-07,
"loss": 0.26280251145362854,
"step": 1490
},
{
"epoch": 3.3082039911308203,
"grad_norm": 0.8796905875205994,
"learning_rate": 8.606349337262623e-07,
"loss": 0.7376017570495605,
"step": 1492
},
{
"epoch": 3.3126385809312637,
"grad_norm": 1.916146993637085,
"learning_rate": 8.561661547107243e-07,
"loss": 0.5012311935424805,
"step": 1494
},
{
"epoch": 3.317073170731707,
"grad_norm": 2.0767107009887695,
"learning_rate": 8.517228549309588e-07,
"loss": 0.5897710919380188,
"step": 1496
},
{
"epoch": 3.3215077605321506,
"grad_norm": 2.7314648628234863,
"learning_rate": 8.473050941661717e-07,
"loss": 0.7489433288574219,
"step": 1498
},
{
"epoch": 3.3259423503325944,
"grad_norm": 0.9693230986595154,
"learning_rate": 8.429129318519711e-07,
"loss": 0.40758612751960754,
"step": 1500
},
{
"epoch": 3.330376940133038,
"grad_norm": 1.1836127042770386,
"learning_rate": 8.38546427079571e-07,
"loss": 0.749980092048645,
"step": 1502
},
{
"epoch": 3.3348115299334813,
"grad_norm": 1.2712533473968506,
"learning_rate": 8.342056385949929e-07,
"loss": 0.7348231673240662,
"step": 1504
},
{
"epoch": 3.3392461197339247,
"grad_norm": 0.25616952776908875,
"learning_rate": 8.298906247982768e-07,
"loss": 0.32258349657058716,
"step": 1506
},
{
"epoch": 3.343680709534368,
"grad_norm": 1.4267375469207764,
"learning_rate": 8.25601443742697e-07,
"loss": 0.4655570089817047,
"step": 1508
},
{
"epoch": 3.3481152993348116,
"grad_norm": 1.8292971849441528,
"learning_rate": 8.213381531339776e-07,
"loss": 0.8381154537200928,
"step": 1510
},
{
"epoch": 3.352549889135255,
"grad_norm": 0.254131019115448,
"learning_rate": 8.1710081032952e-07,
"loss": 0.40903714299201965,
"step": 1512
},
{
"epoch": 3.3569844789356984,
"grad_norm": 1.1773041486740112,
"learning_rate": 8.128894723376285e-07,
"loss": 0.78590327501297,
"step": 1514
},
{
"epoch": 3.361419068736142,
"grad_norm": 2.625889539718628,
"learning_rate": 8.087041958167438e-07,
"loss": 0.5213326215744019,
"step": 1516
},
{
"epoch": 3.3658536585365852,
"grad_norm": 0.7647179365158081,
"learning_rate": 8.04545037074683e-07,
"loss": 0.5064011812210083,
"step": 1518
},
{
"epoch": 3.3702882483370287,
"grad_norm": 0.8953445553779602,
"learning_rate": 8.004120520678768e-07,
"loss": 0.7650377154350281,
"step": 1520
},
{
"epoch": 3.374722838137472,
"grad_norm": 0.39387643337249756,
"learning_rate": 7.963052964006243e-07,
"loss": 0.45021387934684753,
"step": 1522
},
{
"epoch": 3.3791574279379155,
"grad_norm": 3.076669692993164,
"learning_rate": 7.922248253243367e-07,
"loss": 0.6630456447601318,
"step": 1524
},
{
"epoch": 3.3835920177383594,
"grad_norm": 1.8935463428497314,
"learning_rate": 7.881706937368005e-07,
"loss": 0.6750819683074951,
"step": 1526
},
{
"epoch": 3.388026607538803,
"grad_norm": 0.7205215692520142,
"learning_rate": 7.84142956181436e-07,
"loss": 0.37881627678871155,
"step": 1528
},
{
"epoch": 3.3924611973392462,
"grad_norm": 0.5228025317192078,
"learning_rate": 7.801416668465621e-07,
"loss": 0.3390671908855438,
"step": 1530
},
{
"epoch": 3.3968957871396896,
"grad_norm": 2.084331512451172,
"learning_rate": 7.76166879564672e-07,
"loss": 0.2297561913728714,
"step": 1532
},
{
"epoch": 3.401330376940133,
"grad_norm": 0.8670969605445862,
"learning_rate": 7.722186478117031e-07,
"loss": 0.6209254860877991,
"step": 1534
},
{
"epoch": 3.4057649667405765,
"grad_norm": 0.9295784831047058,
"learning_rate": 7.682970247063212e-07,
"loss": 0.6978744864463806,
"step": 1536
},
{
"epoch": 3.41019955654102,
"grad_norm": 1.0750006437301636,
"learning_rate": 7.644020630092066e-07,
"loss": 0.6786344647407532,
"step": 1538
},
{
"epoch": 3.4146341463414633,
"grad_norm": 1.2284328937530518,
"learning_rate": 7.605338151223401e-07,
"loss": 0.7001453042030334,
"step": 1540
},
{
"epoch": 3.4190687361419068,
"grad_norm": 1.4736789464950562,
"learning_rate": 7.566923330883029e-07,
"loss": 0.5472738742828369,
"step": 1542
},
{
"epoch": 3.42350332594235,
"grad_norm": 1.1253695487976074,
"learning_rate": 7.528776685895731e-07,
"loss": 0.6272318363189697,
"step": 1544
},
{
"epoch": 3.4279379157427936,
"grad_norm": 3.8921940326690674,
"learning_rate": 7.490898729478312e-07,
"loss": 0.25309649109840393,
"step": 1546
},
{
"epoch": 3.4323725055432375,
"grad_norm": 0.1467740684747696,
"learning_rate": 7.45328997123271e-07,
"loss": 0.1823056936264038,
"step": 1548
},
{
"epoch": 3.436807095343681,
"grad_norm": 1.1179344654083252,
"learning_rate": 7.415950917139106e-07,
"loss": 0.8248109221458435,
"step": 1550
},
{
"epoch": 3.4412416851441243,
"grad_norm": 0.9201335906982422,
"learning_rate": 7.378882069549166e-07,
"loss": 0.6940004229545593,
"step": 1552
},
{
"epoch": 3.4456762749445677,
"grad_norm": 0.1357531100511551,
"learning_rate": 7.342083927179235e-07,
"loss": 0.3549342155456543,
"step": 1554
},
{
"epoch": 3.450110864745011,
"grad_norm": 1.1405051946640015,
"learning_rate": 7.30555698510366e-07,
"loss": 0.808836817741394,
"step": 1556
},
{
"epoch": 3.4545454545454546,
"grad_norm": 2.6760106086730957,
"learning_rate": 7.269301734748107e-07,
"loss": 0.7992438077926636,
"step": 1558
},
{
"epoch": 3.458980044345898,
"grad_norm": 0.6596611738204956,
"learning_rate": 7.233318663882968e-07,
"loss": 0.7546770572662354,
"step": 1560
},
{
"epoch": 3.4634146341463414,
"grad_norm": 2.3677260875701904,
"learning_rate": 7.197608256616792e-07,
"loss": 0.2441236525774002,
"step": 1562
},
{
"epoch": 3.467849223946785,
"grad_norm": 2.9253904819488525,
"learning_rate": 7.162170993389763e-07,
"loss": 0.72310870885849,
"step": 1564
},
{
"epoch": 3.4722838137472283,
"grad_norm": 0.8974405527114868,
"learning_rate": 7.127007350967241e-07,
"loss": 0.38693636655807495,
"step": 1566
},
{
"epoch": 3.4767184035476717,
"grad_norm": 1.2987812757492065,
"learning_rate": 7.092117802433362e-07,
"loss": 0.6902230381965637,
"step": 1568
},
{
"epoch": 3.481152993348115,
"grad_norm": 0.820105254650116,
"learning_rate": 7.057502817184648e-07,
"loss": 0.4683529734611511,
"step": 1570
},
{
"epoch": 3.4855875831485585,
"grad_norm": 0.34536460041999817,
"learning_rate": 7.023162860923722e-07,
"loss": 0.4866012930870056,
"step": 1572
},
{
"epoch": 3.4900221729490024,
"grad_norm": 0.8184029459953308,
"learning_rate": 6.989098395653005e-07,
"loss": 0.878727912902832,
"step": 1574
},
{
"epoch": 3.494456762749446,
"grad_norm": 2.2575831413269043,
"learning_rate": 6.955309879668537e-07,
"loss": 0.4023507833480835,
"step": 1576
},
{
"epoch": 3.4988913525498893,
"grad_norm": 0.8604305982589722,
"learning_rate": 6.921797767553794e-07,
"loss": 0.47066283226013184,
"step": 1578
},
{
"epoch": 3.5033259423503327,
"grad_norm": 0.9628890752792358,
"learning_rate": 6.88856251017356e-07,
"loss": 0.7100802659988403,
"step": 1580
},
{
"epoch": 3.507760532150776,
"grad_norm": 0.574712336063385,
"learning_rate": 6.855604554667897e-07,
"loss": 0.8129547834396362,
"step": 1582
},
{
"epoch": 3.5121951219512195,
"grad_norm": 0.8234847784042358,
"learning_rate": 6.822924344446081e-07,
"loss": 0.45433831214904785,
"step": 1584
},
{
"epoch": 3.516629711751663,
"grad_norm": 0.8851921558380127,
"learning_rate": 6.790522319180687e-07,
"loss": 0.38872238993644714,
"step": 1586
},
{
"epoch": 3.5210643015521064,
"grad_norm": 0.7976731657981873,
"learning_rate": 6.758398914801628e-07,
"loss": 0.7368452548980713,
"step": 1588
},
{
"epoch": 3.52549889135255,
"grad_norm": 1.8044689893722534,
"learning_rate": 6.726554563490321e-07,
"loss": 0.48450690507888794,
"step": 1590
},
{
"epoch": 3.529933481152993,
"grad_norm": 9.239014625549316,
"learning_rate": 6.694989693673872e-07,
"loss": 0.4795666038990021,
"step": 1592
},
{
"epoch": 3.5343680709534366,
"grad_norm": 1.6922059059143066,
"learning_rate": 6.663704730019285e-07,
"loss": 0.8343867063522339,
"step": 1594
},
{
"epoch": 3.5388026607538805,
"grad_norm": 3.8835480213165283,
"learning_rate": 6.632700093427774e-07,
"loss": 0.2447872906923294,
"step": 1596
},
{
"epoch": 3.5432372505543235,
"grad_norm": 1.5197457075119019,
"learning_rate": 6.601976201029095e-07,
"loss": 0.4911421537399292,
"step": 1598
},
{
"epoch": 3.5476718403547673,
"grad_norm": 0.9553492665290833,
"learning_rate": 6.571533466175928e-07,
"loss": 0.5422607064247131,
"step": 1600
},
{
"epoch": 3.5521064301552108,
"grad_norm": 0.8694108724594116,
"learning_rate": 6.541372298438325e-07,
"loss": 0.7197349667549133,
"step": 1602
},
{
"epoch": 3.556541019955654,
"grad_norm": 0.8564309477806091,
"learning_rate": 6.511493103598184e-07,
"loss": 0.8826733231544495,
"step": 1604
},
{
"epoch": 3.5609756097560976,
"grad_norm": 0.7103037238121033,
"learning_rate": 6.481896283643808e-07,
"loss": 0.7101774215698242,
"step": 1606
},
{
"epoch": 3.565410199556541,
"grad_norm": 0.16576789319515228,
"learning_rate": 6.452582236764495e-07,
"loss": 0.0969938412308693,
"step": 1608
},
{
"epoch": 3.5698447893569845,
"grad_norm": 3.15248966217041,
"learning_rate": 6.423551357345154e-07,
"loss": 0.5980420112609863,
"step": 1610
},
{
"epoch": 3.574279379157428,
"grad_norm": 1.84687077999115,
"learning_rate": 6.394804035961038e-07,
"loss": 0.15914049744606018,
"step": 1612
},
{
"epoch": 3.5787139689578713,
"grad_norm": 3.980046272277832,
"learning_rate": 6.366340659372462e-07,
"loss": 0.5454177260398865,
"step": 1614
},
{
"epoch": 3.5831485587583147,
"grad_norm": 0.8702554702758789,
"learning_rate": 6.338161610519618e-07,
"loss": 0.7745500206947327,
"step": 1616
},
{
"epoch": 3.587583148558758,
"grad_norm": 0.7443963885307312,
"learning_rate": 6.310267268517397e-07,
"loss": 0.46495673060417175,
"step": 1618
},
{
"epoch": 3.5920177383592016,
"grad_norm": 9.520201683044434,
"learning_rate": 6.282658008650318e-07,
"loss": 0.6363322138786316,
"step": 1620
},
{
"epoch": 3.5964523281596454,
"grad_norm": 1.0178289413452148,
"learning_rate": 6.255334202367462e-07,
"loss": 0.7026889324188232,
"step": 1622
},
{
"epoch": 3.6008869179600884,
"grad_norm": 1.213256597518921,
"learning_rate": 6.228296217277481e-07,
"loss": 0.6059472560882568,
"step": 1624
},
{
"epoch": 3.6053215077605323,
"grad_norm": 0.5287774801254272,
"learning_rate": 6.201544417143641e-07,
"loss": 0.23650091886520386,
"step": 1626
},
{
"epoch": 3.6097560975609757,
"grad_norm": 0.9942798614501953,
"learning_rate": 6.175079161878951e-07,
"loss": 0.7636827826499939,
"step": 1628
},
{
"epoch": 3.614190687361419,
"grad_norm": 3.0106701850891113,
"learning_rate": 6.148900807541295e-07,
"loss": 0.6264622807502747,
"step": 1630
},
{
"epoch": 3.6186252771618626,
"grad_norm": 0.984876275062561,
"learning_rate": 6.123009706328659e-07,
"loss": 0.6594371795654297,
"step": 1632
},
{
"epoch": 3.623059866962306,
"grad_norm": 5.279505729675293,
"learning_rate": 6.097406206574378e-07,
"loss": 0.7096071839332581,
"step": 1634
},
{
"epoch": 3.6274944567627494,
"grad_norm": 0.8422771096229553,
"learning_rate": 6.072090652742475e-07,
"loss": 0.4414654076099396,
"step": 1636
},
{
"epoch": 3.631929046563193,
"grad_norm": 6.341533184051514,
"learning_rate": 6.047063385422993e-07,
"loss": 0.6763387322425842,
"step": 1638
},
{
"epoch": 3.6363636363636362,
"grad_norm": 0.9789008498191833,
"learning_rate": 6.022324741327438e-07,
"loss": 0.7263614535331726,
"step": 1640
},
{
"epoch": 3.6407982261640797,
"grad_norm": 1.6151984930038452,
"learning_rate": 5.997875053284248e-07,
"loss": 0.3981097936630249,
"step": 1642
},
{
"epoch": 3.6452328159645235,
"grad_norm": 1.091586947441101,
"learning_rate": 5.973714650234287e-07,
"loss": 0.5544517636299133,
"step": 1644
},
{
"epoch": 3.6496674057649665,
"grad_norm": 0.3249668776988983,
"learning_rate": 5.949843857226466e-07,
"loss": 0.33355870842933655,
"step": 1646
},
{
"epoch": 3.6541019955654104,
"grad_norm": 1.1351978778839111,
"learning_rate": 5.926262995413329e-07,
"loss": 0.3144383728504181,
"step": 1648
},
{
"epoch": 3.658536585365854,
"grad_norm": 1.1642546653747559,
"learning_rate": 5.902972382046742e-07,
"loss": 0.47315874695777893,
"step": 1650
},
{
"epoch": 3.662971175166297,
"grad_norm": 1.5506455898284912,
"learning_rate": 5.879972330473651e-07,
"loss": 0.6661707758903503,
"step": 1652
},
{
"epoch": 3.6674057649667406,
"grad_norm": 0.875584065914154,
"learning_rate": 5.857263150131825e-07,
"loss": 0.3526417315006256,
"step": 1654
},
{
"epoch": 3.671840354767184,
"grad_norm": 0.9616864323616028,
"learning_rate": 5.834845146545726e-07,
"loss": 0.787284791469574,
"step": 1656
},
{
"epoch": 3.6762749445676275,
"grad_norm": 1.849548101425171,
"learning_rate": 5.812718621322386e-07,
"loss": 0.5197221636772156,
"step": 1658
},
{
"epoch": 3.680709534368071,
"grad_norm": 1.266353726387024,
"learning_rate": 5.790883872147341e-07,
"loss": 0.4882799983024597,
"step": 1660
},
{
"epoch": 3.6851441241685143,
"grad_norm": 0.2810562551021576,
"learning_rate": 5.769341192780643e-07,
"loss": 0.38302525877952576,
"step": 1662
},
{
"epoch": 3.6895787139689578,
"grad_norm": 0.9332227110862732,
"learning_rate": 5.748090873052892e-07,
"loss": 0.3155737817287445,
"step": 1664
},
{
"epoch": 3.694013303769401,
"grad_norm": 1.3223484754562378,
"learning_rate": 5.727133198861353e-07,
"loss": 0.6590555906295776,
"step": 1666
},
{
"epoch": 3.6984478935698446,
"grad_norm": 0.7587100863456726,
"learning_rate": 5.706468452166091e-07,
"loss": 0.6262103915214539,
"step": 1668
},
{
"epoch": 3.7028824833702885,
"grad_norm": 0.962223470211029,
"learning_rate": 5.686096910986189e-07,
"loss": 0.6349048018455505,
"step": 1670
},
{
"epoch": 3.7073170731707314,
"grad_norm": 1.7819786071777344,
"learning_rate": 5.666018849396016e-07,
"loss": 0.6529886722564697,
"step": 1672
},
{
"epoch": 3.7117516629711753,
"grad_norm": 0.7751228213310242,
"learning_rate": 5.646234537521513e-07,
"loss": 0.7602441310882568,
"step": 1674
},
{
"epoch": 3.7161862527716187,
"grad_norm": 0.8239582777023315,
"learning_rate": 5.626744241536589e-07,
"loss": 0.7868402004241943,
"step": 1676
},
{
"epoch": 3.720620842572062,
"grad_norm": 1.011643648147583,
"learning_rate": 5.607548223659519e-07,
"loss": 0.7444002628326416,
"step": 1678
},
{
"epoch": 3.7250554323725056,
"grad_norm": 1.968907117843628,
"learning_rate": 5.58864674214942e-07,
"loss": 0.3951123058795929,
"step": 1680
},
{
"epoch": 3.729490022172949,
"grad_norm": 0.7832499742507935,
"learning_rate": 5.57004005130279e-07,
"loss": 0.5796975493431091,
"step": 1682
},
{
"epoch": 3.7339246119733924,
"grad_norm": 0.818016529083252,
"learning_rate": 5.551728401450067e-07,
"loss": 0.43335816264152527,
"step": 1684
},
{
"epoch": 3.738359201773836,
"grad_norm": 2.979412078857422,
"learning_rate": 5.533712038952278e-07,
"loss": 0.6772918105125427,
"step": 1686
},
{
"epoch": 3.7427937915742793,
"grad_norm": 0.39508089423179626,
"learning_rate": 5.51599120619771e-07,
"loss": 0.2403266280889511,
"step": 1688
},
{
"epoch": 3.7472283813747227,
"grad_norm": 1.1755974292755127,
"learning_rate": 5.498566141598662e-07,
"loss": 0.7278658747673035,
"step": 1690
},
{
"epoch": 3.7516629711751666,
"grad_norm": 0.7855948209762573,
"learning_rate": 5.481437079588227e-07,
"loss": 0.6658391952514648,
"step": 1692
},
{
"epoch": 3.7560975609756095,
"grad_norm": 0.7915768027305603,
"learning_rate": 5.464604250617143e-07,
"loss": 0.8541386127471924,
"step": 1694
},
{
"epoch": 3.7605321507760534,
"grad_norm": 0.7992685437202454,
"learning_rate": 5.448067881150697e-07,
"loss": 0.8131424784660339,
"step": 1696
},
{
"epoch": 3.764966740576497,
"grad_norm": 0.9431101679801941,
"learning_rate": 5.431828193665664e-07,
"loss": 0.4288075268268585,
"step": 1698
},
{
"epoch": 3.7694013303769403,
"grad_norm": 2.6646132469177246,
"learning_rate": 5.415885406647334e-07,
"loss": 0.5178676843643188,
"step": 1700
},
{
"epoch": 3.7738359201773837,
"grad_norm": 1.1520713567733765,
"learning_rate": 5.400239734586551e-07,
"loss": 0.912034273147583,
"step": 1702
},
{
"epoch": 3.778270509977827,
"grad_norm": 0.39294496178627014,
"learning_rate": 5.384891387976845e-07,
"loss": 0.11222898960113525,
"step": 1704
},
{
"epoch": 3.7827050997782705,
"grad_norm": 0.22877201437950134,
"learning_rate": 5.369840573311593e-07,
"loss": 0.4198029041290283,
"step": 1706
},
{
"epoch": 3.787139689578714,
"grad_norm": 0.7760717272758484,
"learning_rate": 5.355087493081236e-07,
"loss": 0.5717388391494751,
"step": 1708
},
{
"epoch": 3.7915742793791574,
"grad_norm": 0.8549593687057495,
"learning_rate": 5.340632345770564e-07,
"loss": 0.8026604652404785,
"step": 1710
},
{
"epoch": 3.796008869179601,
"grad_norm": 1.970601201057434,
"learning_rate": 5.326475325856036e-07,
"loss": 0.5300987958908081,
"step": 1712
},
{
"epoch": 3.800443458980044,
"grad_norm": 8.506375312805176,
"learning_rate": 5.312616623803174e-07,
"loss": 0.5400223731994629,
"step": 1714
},
{
"epoch": 3.8048780487804876,
"grad_norm": 1.6792744398117065,
"learning_rate": 5.299056426063995e-07,
"loss": 0.676804780960083,
"step": 1716
},
{
"epoch": 3.8093126385809315,
"grad_norm": 1.1426763534545898,
"learning_rate": 5.2857949150745e-07,
"loss": 0.5867292881011963,
"step": 1718
},
{
"epoch": 3.8137472283813745,
"grad_norm": 0.8977690935134888,
"learning_rate": 5.27283226925222e-07,
"loss": 0.5383524894714355,
"step": 1720
},
{
"epoch": 3.8181818181818183,
"grad_norm": 0.8802960515022278,
"learning_rate": 5.260168662993824e-07,
"loss": 0.7805699110031128,
"step": 1722
},
{
"epoch": 3.8226164079822618,
"grad_norm": 1.9600553512573242,
"learning_rate": 5.247804266672765e-07,
"loss": 0.6948915719985962,
"step": 1724
},
{
"epoch": 3.827050997782705,
"grad_norm": 0.7845137119293213,
"learning_rate": 5.235739246636988e-07,
"loss": 0.6090778708457947,
"step": 1726
},
{
"epoch": 3.8314855875831486,
"grad_norm": 0.8858228921890259,
"learning_rate": 5.223973765206694e-07,
"loss": 0.38666534423828125,
"step": 1728
},
{
"epoch": 3.835920177383592,
"grad_norm": 2.2224628925323486,
"learning_rate": 5.212507980672155e-07,
"loss": 0.5477333664894104,
"step": 1730
},
{
"epoch": 3.8403547671840355,
"grad_norm": 0.6720737814903259,
"learning_rate": 5.201342047291587e-07,
"loss": 0.7528340816497803,
"step": 1732
},
{
"epoch": 3.844789356984479,
"grad_norm": 9.535623550415039,
"learning_rate": 5.190476115289063e-07,
"loss": 0.7895328402519226,
"step": 1734
},
{
"epoch": 3.8492239467849223,
"grad_norm": 0.94792240858078,
"learning_rate": 5.179910330852521e-07,
"loss": 0.8134070634841919,
"step": 1736
},
{
"epoch": 3.8536585365853657,
"grad_norm": 1.6452993154525757,
"learning_rate": 5.169644836131759e-07,
"loss": 0.8293890953063965,
"step": 1738
},
{
"epoch": 3.858093126385809,
"grad_norm": 0.7658231258392334,
"learning_rate": 5.159679769236553e-07,
"loss": 0.4353466331958771,
"step": 1740
},
{
"epoch": 3.8625277161862526,
"grad_norm": 1.3875316381454468,
"learning_rate": 5.150015264234782e-07,
"loss": 0.42465826869010925,
"step": 1742
},
{
"epoch": 3.8669623059866964,
"grad_norm": 0.9564653038978577,
"learning_rate": 5.140651451150627e-07,
"loss": 0.7305734157562256,
"step": 1744
},
{
"epoch": 3.8713968957871394,
"grad_norm": 0.9205288290977478,
"learning_rate": 5.131588455962835e-07,
"loss": 0.6675954461097717,
"step": 1746
},
{
"epoch": 3.8758314855875833,
"grad_norm": 1.4623075723648071,
"learning_rate": 5.122826400602999e-07,
"loss": 0.43095389008522034,
"step": 1748
},
{
"epoch": 3.8802660753880267,
"grad_norm": 0.8612291812896729,
"learning_rate": 5.114365402953946e-07,
"loss": 0.5411372184753418,
"step": 1750
},
{
"epoch": 3.88470066518847,
"grad_norm": 1.1240525245666504,
"learning_rate": 5.106205576848123e-07,
"loss": 0.6075332760810852,
"step": 1752
},
{
"epoch": 3.8891352549889135,
"grad_norm": 1.0948525667190552,
"learning_rate": 5.09834703206609e-07,
"loss": 0.7073659896850586,
"step": 1754
},
{
"epoch": 3.893569844789357,
"grad_norm": 6.487936496734619,
"learning_rate": 5.090789874335027e-07,
"loss": 0.5417055487632751,
"step": 1756
},
{
"epoch": 3.8980044345898004,
"grad_norm": 0.8754919171333313,
"learning_rate": 5.083534205327321e-07,
"loss": 0.796424150466919,
"step": 1758
},
{
"epoch": 3.902439024390244,
"grad_norm": 0.22185245156288147,
"learning_rate": 5.076580122659192e-07,
"loss": 0.006255284883081913,
"step": 1760
},
{
"epoch": 3.9068736141906872,
"grad_norm": 3.0637013912200928,
"learning_rate": 5.069927719889383e-07,
"loss": 0.36486220359802246,
"step": 1762
},
{
"epoch": 3.9113082039911307,
"grad_norm": 0.33201995491981506,
"learning_rate": 5.063577086517894e-07,
"loss": 0.3559632897377014,
"step": 1764
},
{
"epoch": 3.9157427937915745,
"grad_norm": 1.5193932056427002,
"learning_rate": 5.057528307984792e-07,
"loss": 0.538555920124054,
"step": 1766
},
{
"epoch": 3.9201773835920175,
"grad_norm": 1.2443639039993286,
"learning_rate": 5.051781465669053e-07,
"loss": 0.7477619647979736,
"step": 1768
},
{
"epoch": 3.9246119733924614,
"grad_norm": 0.8854790925979614,
"learning_rate": 5.04633663688746e-07,
"loss": 0.7172592878341675,
"step": 1770
},
{
"epoch": 3.929046563192905,
"grad_norm": 1.6672402620315552,
"learning_rate": 5.04119389489358e-07,
"loss": 0.27976056933403015,
"step": 1772
},
{
"epoch": 3.933481152993348,
"grad_norm": 0.8740100264549255,
"learning_rate": 5.036353308876764e-07,
"loss": 0.6978716254234314,
"step": 1774
},
{
"epoch": 3.9379157427937916,
"grad_norm": 0.8391640186309814,
"learning_rate": 5.031814943961221e-07,
"loss": 0.8046401143074036,
"step": 1776
},
{
"epoch": 3.942350332594235,
"grad_norm": 0.3492567241191864,
"learning_rate": 5.027578861205139e-07,
"loss": 0.13979971408843994,
"step": 1778
},
{
"epoch": 3.9467849223946785,
"grad_norm": 2.362764596939087,
"learning_rate": 5.023645117599877e-07,
"loss": 0.48992088437080383,
"step": 1780
},
{
"epoch": 3.951219512195122,
"grad_norm": 0.279731422662735,
"learning_rate": 5.020013766069176e-07,
"loss": 0.4223584532737732,
"step": 1782
},
{
"epoch": 3.9556541019955653,
"grad_norm": 0.77850341796875,
"learning_rate": 5.016684855468464e-07,
"loss": 0.477926105260849,
"step": 1784
},
{
"epoch": 3.9600886917960088,
"grad_norm": 1.2909936904907227,
"learning_rate": 5.013658430584194e-07,
"loss": 0.7137855887413025,
"step": 1786
},
{
"epoch": 3.964523281596452,
"grad_norm": 0.8325835466384888,
"learning_rate": 5.010934532133236e-07,
"loss": 0.4540127217769623,
"step": 1788
},
{
"epoch": 3.9689578713968956,
"grad_norm": 0.9248032569885254,
"learning_rate": 5.008513196762342e-07,
"loss": 0.5620056390762329,
"step": 1790
},
{
"epoch": 3.9733924611973395,
"grad_norm": 0.9865884780883789,
"learning_rate": 5.006394457047638e-07,
"loss": 0.6748104095458984,
"step": 1792
},
{
"epoch": 3.9778270509977824,
"grad_norm": 1.5635875463485718,
"learning_rate": 5.004578341494197e-07,
"loss": 0.38833579421043396,
"step": 1794
},
{
"epoch": 3.9822616407982263,
"grad_norm": 0.9720349907875061,
"learning_rate": 5.003064874535649e-07,
"loss": 0.680509626865387,
"step": 1796
},
{
"epoch": 3.9866962305986697,
"grad_norm": 1.0985702276229858,
"learning_rate": 5.00185407653385e-07,
"loss": 0.7808913588523865,
"step": 1798
},
{
"epoch": 3.991130820399113,
"grad_norm": 4.035973072052002,
"learning_rate": 5.000945963778627e-07,
"loss": 0.6849638223648071,
"step": 1800
},
{
"epoch": 3.9955654101995566,
"grad_norm": 0.8696609735488892,
"learning_rate": 5.000340548487528e-07,
"loss": 0.7774142622947693,
"step": 1802
},
{
"epoch": 4.0,
"grad_norm": 1.2884798049926758,
"learning_rate": 5.000037838805682e-07,
"loss": 0.4334436058998108,
"step": 1804
},
{
"epoch": 4.0,
"step": 1804,
"total_flos": 3.4175049861232067e+18,
"train_loss": 0.8579487466323551,
"train_runtime": 9192.0887,
"train_samples_per_second": 5.888,
"train_steps_per_second": 0.196
}
],
"logging_steps": 2,
"max_steps": 1804,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.4175049861232067e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}