RAG-HotpotQA-PRM / trainer_state.json
RAG-Gym's picture
Upload 8 files
2e33c3e verified
Raw
History Blame Contribute Delete
423 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9997947876051714,
"eval_steps": 500,
"global_step": 2436,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0004104247896572953,
"grad_norm": 489.601318359375,
"learning_rate": 9.995894909688014e-06,
"loss": 3.7812,
"step": 1
},
{
"epoch": 0.0008208495793145906,
"grad_norm": 125.23029327392578,
"learning_rate": 9.991789819376027e-06,
"loss": 0.7441,
"step": 2
},
{
"epoch": 0.001231274368971886,
"grad_norm": 72.09346008300781,
"learning_rate": 9.98768472906404e-06,
"loss": 1.041,
"step": 3
},
{
"epoch": 0.0016416991586291811,
"grad_norm": 67.57290649414062,
"learning_rate": 9.983579638752053e-06,
"loss": 0.7207,
"step": 4
},
{
"epoch": 0.0020521239482864766,
"grad_norm": 75.5198745727539,
"learning_rate": 9.979474548440066e-06,
"loss": 0.748,
"step": 5
},
{
"epoch": 0.002462548737943772,
"grad_norm": 47.96021270751953,
"learning_rate": 9.97536945812808e-06,
"loss": 0.3174,
"step": 6
},
{
"epoch": 0.002872973527601067,
"grad_norm": 32.79164505004883,
"learning_rate": 9.971264367816092e-06,
"loss": 0.6504,
"step": 7
},
{
"epoch": 0.0032833983172583622,
"grad_norm": 30.445396423339844,
"learning_rate": 9.967159277504105e-06,
"loss": 0.396,
"step": 8
},
{
"epoch": 0.003693823106915658,
"grad_norm": 38.777469635009766,
"learning_rate": 9.96305418719212e-06,
"loss": 0.1458,
"step": 9
},
{
"epoch": 0.004104247896572953,
"grad_norm": 86.58824157714844,
"learning_rate": 9.958949096880131e-06,
"loss": 0.6182,
"step": 10
},
{
"epoch": 0.004514672686230248,
"grad_norm": 22.10052490234375,
"learning_rate": 9.954844006568145e-06,
"loss": 0.4102,
"step": 11
},
{
"epoch": 0.004925097475887544,
"grad_norm": 56.18250274658203,
"learning_rate": 9.95073891625616e-06,
"loss": 0.7969,
"step": 12
},
{
"epoch": 0.005335522265544839,
"grad_norm": 46.60348892211914,
"learning_rate": 9.946633825944172e-06,
"loss": 0.4429,
"step": 13
},
{
"epoch": 0.005745947055202134,
"grad_norm": 43.0778923034668,
"learning_rate": 9.942528735632184e-06,
"loss": 0.2646,
"step": 14
},
{
"epoch": 0.00615637184485943,
"grad_norm": 28.1630859375,
"learning_rate": 9.938423645320198e-06,
"loss": 0.439,
"step": 15
},
{
"epoch": 0.0065667966345167245,
"grad_norm": 47.220943450927734,
"learning_rate": 9.934318555008212e-06,
"loss": 0.5625,
"step": 16
},
{
"epoch": 0.00697722142417402,
"grad_norm": 46.93181610107422,
"learning_rate": 9.930213464696225e-06,
"loss": 0.6328,
"step": 17
},
{
"epoch": 0.007387646213831316,
"grad_norm": 44.84854507446289,
"learning_rate": 9.926108374384236e-06,
"loss": 0.1489,
"step": 18
},
{
"epoch": 0.007798071003488611,
"grad_norm": 42.41459655761719,
"learning_rate": 9.92200328407225e-06,
"loss": 0.5327,
"step": 19
},
{
"epoch": 0.008208495793145906,
"grad_norm": 135.06227111816406,
"learning_rate": 9.917898193760264e-06,
"loss": 1.3555,
"step": 20
},
{
"epoch": 0.008618920582803202,
"grad_norm": 57.64326858520508,
"learning_rate": 9.913793103448277e-06,
"loss": 0.8906,
"step": 21
},
{
"epoch": 0.009029345372460496,
"grad_norm": 31.03734016418457,
"learning_rate": 9.90968801313629e-06,
"loss": 0.458,
"step": 22
},
{
"epoch": 0.009439770162117792,
"grad_norm": 85.89253997802734,
"learning_rate": 9.905582922824303e-06,
"loss": 0.6211,
"step": 23
},
{
"epoch": 0.009850194951775087,
"grad_norm": 34.010196685791016,
"learning_rate": 9.901477832512316e-06,
"loss": 0.3875,
"step": 24
},
{
"epoch": 0.010260619741432383,
"grad_norm": 38.69382858276367,
"learning_rate": 9.897372742200329e-06,
"loss": 0.8496,
"step": 25
},
{
"epoch": 0.010671044531089679,
"grad_norm": 63.049400329589844,
"learning_rate": 9.893267651888342e-06,
"loss": 0.6842,
"step": 26
},
{
"epoch": 0.011081469320746972,
"grad_norm": 52.40337371826172,
"learning_rate": 9.889162561576355e-06,
"loss": 0.8984,
"step": 27
},
{
"epoch": 0.011491894110404268,
"grad_norm": 44.72108840942383,
"learning_rate": 9.885057471264368e-06,
"loss": 0.4688,
"step": 28
},
{
"epoch": 0.011902318900061564,
"grad_norm": 45.39631271362305,
"learning_rate": 9.880952380952381e-06,
"loss": 0.7891,
"step": 29
},
{
"epoch": 0.01231274368971886,
"grad_norm": 34.166866302490234,
"learning_rate": 9.876847290640394e-06,
"loss": 0.3367,
"step": 30
},
{
"epoch": 0.012723168479376155,
"grad_norm": 60.586936950683594,
"learning_rate": 9.872742200328408e-06,
"loss": 0.6582,
"step": 31
},
{
"epoch": 0.013133593269033449,
"grad_norm": 41.81749725341797,
"learning_rate": 9.86863711001642e-06,
"loss": 0.7676,
"step": 32
},
{
"epoch": 0.013544018058690745,
"grad_norm": 126.78913116455078,
"learning_rate": 9.864532019704435e-06,
"loss": 0.7656,
"step": 33
},
{
"epoch": 0.01395444284834804,
"grad_norm": 79.77574157714844,
"learning_rate": 9.860426929392447e-06,
"loss": 0.623,
"step": 34
},
{
"epoch": 0.014364867638005336,
"grad_norm": 117.93644714355469,
"learning_rate": 9.85632183908046e-06,
"loss": 0.8926,
"step": 35
},
{
"epoch": 0.014775292427662632,
"grad_norm": 192.74481201171875,
"learning_rate": 9.852216748768475e-06,
"loss": 0.5381,
"step": 36
},
{
"epoch": 0.015185717217319926,
"grad_norm": 85.22727966308594,
"learning_rate": 9.848111658456488e-06,
"loss": 0.3945,
"step": 37
},
{
"epoch": 0.015596142006977221,
"grad_norm": 53.78607177734375,
"learning_rate": 9.844006568144499e-06,
"loss": 0.8613,
"step": 38
},
{
"epoch": 0.016006566796634515,
"grad_norm": 73.2311782836914,
"learning_rate": 9.839901477832514e-06,
"loss": 0.4707,
"step": 39
},
{
"epoch": 0.016416991586291813,
"grad_norm": 92.27364349365234,
"learning_rate": 9.835796387520527e-06,
"loss": 0.6777,
"step": 40
},
{
"epoch": 0.016827416375949106,
"grad_norm": 211.47694396972656,
"learning_rate": 9.83169129720854e-06,
"loss": 0.2256,
"step": 41
},
{
"epoch": 0.017237841165606404,
"grad_norm": 74.32991790771484,
"learning_rate": 9.827586206896553e-06,
"loss": 0.271,
"step": 42
},
{
"epoch": 0.017648265955263698,
"grad_norm": 93.45068359375,
"learning_rate": 9.823481116584566e-06,
"loss": 0.6846,
"step": 43
},
{
"epoch": 0.01805869074492099,
"grad_norm": 47.72433090209961,
"learning_rate": 9.819376026272579e-06,
"loss": 0.6621,
"step": 44
},
{
"epoch": 0.01846911553457829,
"grad_norm": 50.61162185668945,
"learning_rate": 9.815270935960592e-06,
"loss": 0.3467,
"step": 45
},
{
"epoch": 0.018879540324235583,
"grad_norm": 37.95347595214844,
"learning_rate": 9.811165845648605e-06,
"loss": 0.5879,
"step": 46
},
{
"epoch": 0.01928996511389288,
"grad_norm": 79.33946990966797,
"learning_rate": 9.807060755336618e-06,
"loss": 0.3853,
"step": 47
},
{
"epoch": 0.019700389903550174,
"grad_norm": 194.1417236328125,
"learning_rate": 9.802955665024631e-06,
"loss": 0.5791,
"step": 48
},
{
"epoch": 0.02011081469320747,
"grad_norm": 56.02328109741211,
"learning_rate": 9.798850574712644e-06,
"loss": 0.4133,
"step": 49
},
{
"epoch": 0.020521239482864766,
"grad_norm": 145.2246551513672,
"learning_rate": 9.794745484400657e-06,
"loss": 0.7852,
"step": 50
},
{
"epoch": 0.02093166427252206,
"grad_norm": 257.49688720703125,
"learning_rate": 9.79064039408867e-06,
"loss": 0.5684,
"step": 51
},
{
"epoch": 0.021342089062179357,
"grad_norm": 65.12989044189453,
"learning_rate": 9.786535303776684e-06,
"loss": 0.5518,
"step": 52
},
{
"epoch": 0.02175251385183665,
"grad_norm": 80.76338195800781,
"learning_rate": 9.782430213464698e-06,
"loss": 0.6523,
"step": 53
},
{
"epoch": 0.022162938641493945,
"grad_norm": 60.46729278564453,
"learning_rate": 9.77832512315271e-06,
"loss": 0.4023,
"step": 54
},
{
"epoch": 0.022573363431151242,
"grad_norm": 112.77301025390625,
"learning_rate": 9.774220032840723e-06,
"loss": 0.9238,
"step": 55
},
{
"epoch": 0.022983788220808536,
"grad_norm": 66.21923828125,
"learning_rate": 9.770114942528738e-06,
"loss": 0.6653,
"step": 56
},
{
"epoch": 0.023394213010465834,
"grad_norm": 65.4253921508789,
"learning_rate": 9.766009852216749e-06,
"loss": 0.3171,
"step": 57
},
{
"epoch": 0.023804637800123127,
"grad_norm": 110.94827270507812,
"learning_rate": 9.761904761904762e-06,
"loss": 0.834,
"step": 58
},
{
"epoch": 0.02421506258978042,
"grad_norm": 35.9803466796875,
"learning_rate": 9.757799671592777e-06,
"loss": 0.4169,
"step": 59
},
{
"epoch": 0.02462548737943772,
"grad_norm": 111.29431915283203,
"learning_rate": 9.75369458128079e-06,
"loss": 0.8105,
"step": 60
},
{
"epoch": 0.025035912169095013,
"grad_norm": 56.25421905517578,
"learning_rate": 9.749589490968801e-06,
"loss": 0.4534,
"step": 61
},
{
"epoch": 0.02544633695875231,
"grad_norm": 89.82777404785156,
"learning_rate": 9.745484400656814e-06,
"loss": 0.5264,
"step": 62
},
{
"epoch": 0.025856761748409604,
"grad_norm": 123.6474838256836,
"learning_rate": 9.741379310344829e-06,
"loss": 0.3643,
"step": 63
},
{
"epoch": 0.026267186538066898,
"grad_norm": 140.19151306152344,
"learning_rate": 9.737274220032842e-06,
"loss": 0.4092,
"step": 64
},
{
"epoch": 0.026677611327724195,
"grad_norm": 70.18293762207031,
"learning_rate": 9.733169129720853e-06,
"loss": 0.3235,
"step": 65
},
{
"epoch": 0.02708803611738149,
"grad_norm": 99.06017303466797,
"learning_rate": 9.729064039408868e-06,
"loss": 0.7793,
"step": 66
},
{
"epoch": 0.027498460907038787,
"grad_norm": 578.2423706054688,
"learning_rate": 9.724958949096881e-06,
"loss": 0.9038,
"step": 67
},
{
"epoch": 0.02790888569669608,
"grad_norm": 67.8482437133789,
"learning_rate": 9.720853858784894e-06,
"loss": 0.4517,
"step": 68
},
{
"epoch": 0.028319310486353375,
"grad_norm": 142.90036010742188,
"learning_rate": 9.716748768472907e-06,
"loss": 0.6006,
"step": 69
},
{
"epoch": 0.028729735276010672,
"grad_norm": 24.348678588867188,
"learning_rate": 9.71264367816092e-06,
"loss": 0.1124,
"step": 70
},
{
"epoch": 0.029140160065667966,
"grad_norm": 87.90788269042969,
"learning_rate": 9.708538587848933e-06,
"loss": 0.6963,
"step": 71
},
{
"epoch": 0.029550584855325263,
"grad_norm": 64.9705810546875,
"learning_rate": 9.704433497536947e-06,
"loss": 0.7676,
"step": 72
},
{
"epoch": 0.029961009644982557,
"grad_norm": 105.4383544921875,
"learning_rate": 9.70032840722496e-06,
"loss": 0.8379,
"step": 73
},
{
"epoch": 0.03037143443463985,
"grad_norm": 87.10254669189453,
"learning_rate": 9.696223316912973e-06,
"loss": 0.2803,
"step": 74
},
{
"epoch": 0.03078185922429715,
"grad_norm": 39.44083786010742,
"learning_rate": 9.692118226600986e-06,
"loss": 0.3307,
"step": 75
},
{
"epoch": 0.031192284013954442,
"grad_norm": 43.18490219116211,
"learning_rate": 9.688013136288999e-06,
"loss": 0.4531,
"step": 76
},
{
"epoch": 0.03160270880361174,
"grad_norm": 141.7541046142578,
"learning_rate": 9.683908045977012e-06,
"loss": 1.2031,
"step": 77
},
{
"epoch": 0.03201313359326903,
"grad_norm": 71.48495483398438,
"learning_rate": 9.679802955665025e-06,
"loss": 0.6699,
"step": 78
},
{
"epoch": 0.03242355838292633,
"grad_norm": 94.1418228149414,
"learning_rate": 9.675697865353038e-06,
"loss": 0.6406,
"step": 79
},
{
"epoch": 0.032833983172583625,
"grad_norm": 94.42112731933594,
"learning_rate": 9.671592775041053e-06,
"loss": 0.4648,
"step": 80
},
{
"epoch": 0.03324440796224092,
"grad_norm": 75.63024139404297,
"learning_rate": 9.667487684729064e-06,
"loss": 0.4839,
"step": 81
},
{
"epoch": 0.03365483275189821,
"grad_norm": 111.62572479248047,
"learning_rate": 9.663382594417077e-06,
"loss": 0.9146,
"step": 82
},
{
"epoch": 0.03406525754155551,
"grad_norm": 201.1129608154297,
"learning_rate": 9.659277504105092e-06,
"loss": 0.7227,
"step": 83
},
{
"epoch": 0.03447568233121281,
"grad_norm": 164.67434692382812,
"learning_rate": 9.655172413793105e-06,
"loss": 1.0137,
"step": 84
},
{
"epoch": 0.0348861071208701,
"grad_norm": 33.38908767700195,
"learning_rate": 9.651067323481116e-06,
"loss": 0.0438,
"step": 85
},
{
"epoch": 0.035296531910527396,
"grad_norm": 80.65897369384766,
"learning_rate": 9.646962233169131e-06,
"loss": 0.5938,
"step": 86
},
{
"epoch": 0.03570695670018469,
"grad_norm": 57.92540740966797,
"learning_rate": 9.642857142857144e-06,
"loss": 0.1294,
"step": 87
},
{
"epoch": 0.03611738148984198,
"grad_norm": 71.7527084350586,
"learning_rate": 9.638752052545157e-06,
"loss": 0.2588,
"step": 88
},
{
"epoch": 0.03652780627949928,
"grad_norm": 176.32838439941406,
"learning_rate": 9.63464696223317e-06,
"loss": 0.7686,
"step": 89
},
{
"epoch": 0.03693823106915658,
"grad_norm": 22.736574172973633,
"learning_rate": 9.630541871921183e-06,
"loss": 0.0408,
"step": 90
},
{
"epoch": 0.037348655858813876,
"grad_norm": 75.41202545166016,
"learning_rate": 9.626436781609196e-06,
"loss": 0.3232,
"step": 91
},
{
"epoch": 0.037759080648471166,
"grad_norm": 99.61798095703125,
"learning_rate": 9.62233169129721e-06,
"loss": 0.8789,
"step": 92
},
{
"epoch": 0.038169505438128463,
"grad_norm": 214.0462188720703,
"learning_rate": 9.618226600985223e-06,
"loss": 0.9492,
"step": 93
},
{
"epoch": 0.03857993022778576,
"grad_norm": 45.386024475097656,
"learning_rate": 9.614121510673236e-06,
"loss": 0.4302,
"step": 94
},
{
"epoch": 0.03899035501744305,
"grad_norm": 66.32962036132812,
"learning_rate": 9.610016420361249e-06,
"loss": 0.6895,
"step": 95
},
{
"epoch": 0.03940077980710035,
"grad_norm": 25.231769561767578,
"learning_rate": 9.605911330049262e-06,
"loss": 0.2273,
"step": 96
},
{
"epoch": 0.039811204596757646,
"grad_norm": 45.79891586303711,
"learning_rate": 9.601806239737275e-06,
"loss": 0.4141,
"step": 97
},
{
"epoch": 0.04022162938641494,
"grad_norm": 64.21709442138672,
"learning_rate": 9.597701149425288e-06,
"loss": 0.6309,
"step": 98
},
{
"epoch": 0.040632054176072234,
"grad_norm": 293.3857421875,
"learning_rate": 9.593596059113301e-06,
"loss": 0.8281,
"step": 99
},
{
"epoch": 0.04104247896572953,
"grad_norm": 64.64630126953125,
"learning_rate": 9.589490968801316e-06,
"loss": 0.4404,
"step": 100
},
{
"epoch": 0.04145290375538683,
"grad_norm": 268.50079345703125,
"learning_rate": 9.585385878489327e-06,
"loss": 1.2383,
"step": 101
},
{
"epoch": 0.04186332854504412,
"grad_norm": 39.85595703125,
"learning_rate": 9.58128078817734e-06,
"loss": 0.3237,
"step": 102
},
{
"epoch": 0.04227375333470142,
"grad_norm": 25.994464874267578,
"learning_rate": 9.577175697865353e-06,
"loss": 0.4624,
"step": 103
},
{
"epoch": 0.042684178124358714,
"grad_norm": 38.49090576171875,
"learning_rate": 9.573070607553368e-06,
"loss": 0.1255,
"step": 104
},
{
"epoch": 0.043094602914016004,
"grad_norm": 37.41379165649414,
"learning_rate": 9.56896551724138e-06,
"loss": 0.5889,
"step": 105
},
{
"epoch": 0.0435050277036733,
"grad_norm": 58.176124572753906,
"learning_rate": 9.564860426929392e-06,
"loss": 0.9727,
"step": 106
},
{
"epoch": 0.0439154524933306,
"grad_norm": 25.173282623291016,
"learning_rate": 9.560755336617407e-06,
"loss": 0.1226,
"step": 107
},
{
"epoch": 0.04432587728298789,
"grad_norm": 62.66582489013672,
"learning_rate": 9.55665024630542e-06,
"loss": 0.6753,
"step": 108
},
{
"epoch": 0.04473630207264519,
"grad_norm": 21.916446685791016,
"learning_rate": 9.552545155993432e-06,
"loss": 0.288,
"step": 109
},
{
"epoch": 0.045146726862302484,
"grad_norm": 62.611412048339844,
"learning_rate": 9.548440065681446e-06,
"loss": 0.6826,
"step": 110
},
{
"epoch": 0.045557151651959775,
"grad_norm": 20.345088958740234,
"learning_rate": 9.54433497536946e-06,
"loss": 0.3005,
"step": 111
},
{
"epoch": 0.04596757644161707,
"grad_norm": 18.393754959106445,
"learning_rate": 9.54022988505747e-06,
"loss": 0.1821,
"step": 112
},
{
"epoch": 0.04637800123127437,
"grad_norm": 73.479736328125,
"learning_rate": 9.536124794745486e-06,
"loss": 0.801,
"step": 113
},
{
"epoch": 0.04678842602093167,
"grad_norm": 84.64009857177734,
"learning_rate": 9.532019704433499e-06,
"loss": 0.4033,
"step": 114
},
{
"epoch": 0.04719885081058896,
"grad_norm": 32.4448356628418,
"learning_rate": 9.527914614121512e-06,
"loss": 0.3677,
"step": 115
},
{
"epoch": 0.047609275600246255,
"grad_norm": 50.115875244140625,
"learning_rate": 9.523809523809525e-06,
"loss": 0.4727,
"step": 116
},
{
"epoch": 0.04801970038990355,
"grad_norm": 76.39530181884766,
"learning_rate": 9.519704433497538e-06,
"loss": 0.7904,
"step": 117
},
{
"epoch": 0.04843012517956084,
"grad_norm": 51.054954528808594,
"learning_rate": 9.515599343185551e-06,
"loss": 0.2456,
"step": 118
},
{
"epoch": 0.04884054996921814,
"grad_norm": 65.77157592773438,
"learning_rate": 9.511494252873564e-06,
"loss": 0.3223,
"step": 119
},
{
"epoch": 0.04925097475887544,
"grad_norm": 80.27091217041016,
"learning_rate": 9.507389162561577e-06,
"loss": 0.2534,
"step": 120
},
{
"epoch": 0.04966139954853273,
"grad_norm": 68.08970642089844,
"learning_rate": 9.50328407224959e-06,
"loss": 0.708,
"step": 121
},
{
"epoch": 0.050071824338190025,
"grad_norm": 453.7446594238281,
"learning_rate": 9.499178981937603e-06,
"loss": 0.8467,
"step": 122
},
{
"epoch": 0.05048224912784732,
"grad_norm": 159.33111572265625,
"learning_rate": 9.495073891625616e-06,
"loss": 0.7292,
"step": 123
},
{
"epoch": 0.05089267391750462,
"grad_norm": 81.7794418334961,
"learning_rate": 9.49096880131363e-06,
"loss": 0.2139,
"step": 124
},
{
"epoch": 0.05130309870716191,
"grad_norm": 12.556668281555176,
"learning_rate": 9.486863711001642e-06,
"loss": 0.052,
"step": 125
},
{
"epoch": 0.05171352349681921,
"grad_norm": 39.47209930419922,
"learning_rate": 9.482758620689655e-06,
"loss": 0.0845,
"step": 126
},
{
"epoch": 0.052123948286476506,
"grad_norm": 61.53842544555664,
"learning_rate": 9.47865353037767e-06,
"loss": 0.168,
"step": 127
},
{
"epoch": 0.052534373076133796,
"grad_norm": 32.61935806274414,
"learning_rate": 9.474548440065682e-06,
"loss": 0.2246,
"step": 128
},
{
"epoch": 0.05294479786579109,
"grad_norm": 42.09776306152344,
"learning_rate": 9.470443349753695e-06,
"loss": 0.1943,
"step": 129
},
{
"epoch": 0.05335522265544839,
"grad_norm": 438.5118713378906,
"learning_rate": 9.46633825944171e-06,
"loss": 0.5625,
"step": 130
},
{
"epoch": 0.05376564744510568,
"grad_norm": 13.452123641967773,
"learning_rate": 9.462233169129722e-06,
"loss": 0.0212,
"step": 131
},
{
"epoch": 0.05417607223476298,
"grad_norm": 29.14824867248535,
"learning_rate": 9.458128078817734e-06,
"loss": 0.2064,
"step": 132
},
{
"epoch": 0.054586497024420276,
"grad_norm": 96.0547103881836,
"learning_rate": 9.454022988505749e-06,
"loss": 0.5107,
"step": 133
},
{
"epoch": 0.05499692181407757,
"grad_norm": 249.0057373046875,
"learning_rate": 9.449917898193762e-06,
"loss": 1.4062,
"step": 134
},
{
"epoch": 0.055407346603734864,
"grad_norm": 153.24945068359375,
"learning_rate": 9.445812807881775e-06,
"loss": 0.4983,
"step": 135
},
{
"epoch": 0.05581777139339216,
"grad_norm": 140.70147705078125,
"learning_rate": 9.441707717569786e-06,
"loss": 1.3145,
"step": 136
},
{
"epoch": 0.05622819618304946,
"grad_norm": 108.32526397705078,
"learning_rate": 9.4376026272578e-06,
"loss": 0.3184,
"step": 137
},
{
"epoch": 0.05663862097270675,
"grad_norm": 161.90768432617188,
"learning_rate": 9.433497536945814e-06,
"loss": 0.6631,
"step": 138
},
{
"epoch": 0.057049045762364047,
"grad_norm": 149.4306640625,
"learning_rate": 9.429392446633827e-06,
"loss": 0.6309,
"step": 139
},
{
"epoch": 0.057459470552021344,
"grad_norm": 109.20567321777344,
"learning_rate": 9.42528735632184e-06,
"loss": 0.4824,
"step": 140
},
{
"epoch": 0.057869895341678634,
"grad_norm": 141.54905700683594,
"learning_rate": 9.421182266009853e-06,
"loss": 0.5713,
"step": 141
},
{
"epoch": 0.05828032013133593,
"grad_norm": 46.46152877807617,
"learning_rate": 9.417077175697866e-06,
"loss": 0.3579,
"step": 142
},
{
"epoch": 0.05869074492099323,
"grad_norm": 346.67413330078125,
"learning_rate": 9.41297208538588e-06,
"loss": 0.5146,
"step": 143
},
{
"epoch": 0.05910116971065053,
"grad_norm": 148.48593139648438,
"learning_rate": 9.408866995073892e-06,
"loss": 0.3967,
"step": 144
},
{
"epoch": 0.05951159450030782,
"grad_norm": 936.2588500976562,
"learning_rate": 9.404761904761905e-06,
"loss": 1.5859,
"step": 145
},
{
"epoch": 0.059922019289965114,
"grad_norm": 131.36224365234375,
"learning_rate": 9.400656814449918e-06,
"loss": 0.8154,
"step": 146
},
{
"epoch": 0.06033244407962241,
"grad_norm": 624.3028564453125,
"learning_rate": 9.396551724137931e-06,
"loss": 0.4771,
"step": 147
},
{
"epoch": 0.0607428688692797,
"grad_norm": 61.30937194824219,
"learning_rate": 9.392446633825945e-06,
"loss": 0.3545,
"step": 148
},
{
"epoch": 0.061153293658937,
"grad_norm": 549.59619140625,
"learning_rate": 9.388341543513958e-06,
"loss": 1.0154,
"step": 149
},
{
"epoch": 0.0615637184485943,
"grad_norm": 215.6182098388672,
"learning_rate": 9.38423645320197e-06,
"loss": 0.6113,
"step": 150
},
{
"epoch": 0.06197414323825159,
"grad_norm": 166.8955535888672,
"learning_rate": 9.380131362889985e-06,
"loss": 0.5508,
"step": 151
},
{
"epoch": 0.062384568027908885,
"grad_norm": 833.9037475585938,
"learning_rate": 9.376026272577997e-06,
"loss": 1.2891,
"step": 152
},
{
"epoch": 0.06279499281756618,
"grad_norm": 450.9167175292969,
"learning_rate": 9.37192118226601e-06,
"loss": 0.8157,
"step": 153
},
{
"epoch": 0.06320541760722348,
"grad_norm": 605.9237670898438,
"learning_rate": 9.367816091954025e-06,
"loss": 0.8457,
"step": 154
},
{
"epoch": 0.06361584239688077,
"grad_norm": 287.69085693359375,
"learning_rate": 9.363711001642038e-06,
"loss": 0.9141,
"step": 155
},
{
"epoch": 0.06402626718653806,
"grad_norm": 239.24017333984375,
"learning_rate": 9.359605911330049e-06,
"loss": 0.9688,
"step": 156
},
{
"epoch": 0.06443669197619536,
"grad_norm": 185.26646423339844,
"learning_rate": 9.355500821018064e-06,
"loss": 0.4805,
"step": 157
},
{
"epoch": 0.06484711676585266,
"grad_norm": 437.7055969238281,
"learning_rate": 9.351395730706077e-06,
"loss": 0.6641,
"step": 158
},
{
"epoch": 0.06525754155550996,
"grad_norm": 80.91333770751953,
"learning_rate": 9.34729064039409e-06,
"loss": 0.2568,
"step": 159
},
{
"epoch": 0.06566796634516725,
"grad_norm": 398.94134521484375,
"learning_rate": 9.343185550082103e-06,
"loss": 1.5781,
"step": 160
},
{
"epoch": 0.06607839113482454,
"grad_norm": 147.48617553710938,
"learning_rate": 9.339080459770116e-06,
"loss": 0.5632,
"step": 161
},
{
"epoch": 0.06648881592448184,
"grad_norm": 526.4559326171875,
"learning_rate": 9.334975369458129e-06,
"loss": 1.3013,
"step": 162
},
{
"epoch": 0.06689924071413914,
"grad_norm": 109.44779205322266,
"learning_rate": 9.330870279146142e-06,
"loss": 0.7461,
"step": 163
},
{
"epoch": 0.06730966550379643,
"grad_norm": 126.11822509765625,
"learning_rate": 9.326765188834155e-06,
"loss": 0.4475,
"step": 164
},
{
"epoch": 0.06772009029345373,
"grad_norm": 122.11145782470703,
"learning_rate": 9.322660098522168e-06,
"loss": 0.6948,
"step": 165
},
{
"epoch": 0.06813051508311102,
"grad_norm": 356.7264404296875,
"learning_rate": 9.318555008210181e-06,
"loss": 1.0234,
"step": 166
},
{
"epoch": 0.06854093987276831,
"grad_norm": 261.68951416015625,
"learning_rate": 9.314449917898194e-06,
"loss": 0.2727,
"step": 167
},
{
"epoch": 0.06895136466242562,
"grad_norm": 72.83009338378906,
"learning_rate": 9.310344827586207e-06,
"loss": 0.3682,
"step": 168
},
{
"epoch": 0.0693617894520829,
"grad_norm": 65.6648941040039,
"learning_rate": 9.30623973727422e-06,
"loss": 0.2,
"step": 169
},
{
"epoch": 0.0697722142417402,
"grad_norm": 114.76077270507812,
"learning_rate": 9.302134646962234e-06,
"loss": 0.3171,
"step": 170
},
{
"epoch": 0.0701826390313975,
"grad_norm": 180.42640686035156,
"learning_rate": 9.298029556650247e-06,
"loss": 0.4834,
"step": 171
},
{
"epoch": 0.07059306382105479,
"grad_norm": 119.9900894165039,
"learning_rate": 9.29392446633826e-06,
"loss": 0.4004,
"step": 172
},
{
"epoch": 0.07100348861071208,
"grad_norm": 65.80331420898438,
"learning_rate": 9.289819376026273e-06,
"loss": 0.4375,
"step": 173
},
{
"epoch": 0.07141391340036939,
"grad_norm": 31.657791137695312,
"learning_rate": 9.285714285714288e-06,
"loss": 0.2627,
"step": 174
},
{
"epoch": 0.07182433819002668,
"grad_norm": 48.153987884521484,
"learning_rate": 9.281609195402299e-06,
"loss": 0.3964,
"step": 175
},
{
"epoch": 0.07223476297968397,
"grad_norm": 51.97791290283203,
"learning_rate": 9.277504105090312e-06,
"loss": 0.502,
"step": 176
},
{
"epoch": 0.07264518776934127,
"grad_norm": 307.4492492675781,
"learning_rate": 9.273399014778327e-06,
"loss": 1.2876,
"step": 177
},
{
"epoch": 0.07305561255899856,
"grad_norm": 69.885986328125,
"learning_rate": 9.26929392446634e-06,
"loss": 0.5352,
"step": 178
},
{
"epoch": 0.07346603734865585,
"grad_norm": 49.86472702026367,
"learning_rate": 9.265188834154351e-06,
"loss": 0.1279,
"step": 179
},
{
"epoch": 0.07387646213831316,
"grad_norm": 53.40109634399414,
"learning_rate": 9.261083743842364e-06,
"loss": 0.4414,
"step": 180
},
{
"epoch": 0.07428688692797045,
"grad_norm": 41.1759033203125,
"learning_rate": 9.256978653530379e-06,
"loss": 0.4922,
"step": 181
},
{
"epoch": 0.07469731171762775,
"grad_norm": 57.673763275146484,
"learning_rate": 9.252873563218392e-06,
"loss": 0.1433,
"step": 182
},
{
"epoch": 0.07510773650728504,
"grad_norm": 112.72769165039062,
"learning_rate": 9.248768472906403e-06,
"loss": 0.6445,
"step": 183
},
{
"epoch": 0.07551816129694233,
"grad_norm": 203.80783081054688,
"learning_rate": 9.244663382594418e-06,
"loss": 0.748,
"step": 184
},
{
"epoch": 0.07592858608659964,
"grad_norm": 42.277366638183594,
"learning_rate": 9.240558292282431e-06,
"loss": 0.1908,
"step": 185
},
{
"epoch": 0.07633901087625693,
"grad_norm": 67.3951644897461,
"learning_rate": 9.236453201970444e-06,
"loss": 0.2554,
"step": 186
},
{
"epoch": 0.07674943566591422,
"grad_norm": 136.6295166015625,
"learning_rate": 9.232348111658457e-06,
"loss": 0.2017,
"step": 187
},
{
"epoch": 0.07715986045557152,
"grad_norm": 106.77710723876953,
"learning_rate": 9.22824302134647e-06,
"loss": 0.2783,
"step": 188
},
{
"epoch": 0.07757028524522881,
"grad_norm": 188.4326629638672,
"learning_rate": 9.224137931034484e-06,
"loss": 1.3008,
"step": 189
},
{
"epoch": 0.0779807100348861,
"grad_norm": 25.458778381347656,
"learning_rate": 9.220032840722497e-06,
"loss": 0.1162,
"step": 190
},
{
"epoch": 0.07839113482454341,
"grad_norm": 50.82314682006836,
"learning_rate": 9.21592775041051e-06,
"loss": 0.291,
"step": 191
},
{
"epoch": 0.0788015596142007,
"grad_norm": 107.79209899902344,
"learning_rate": 9.211822660098523e-06,
"loss": 0.793,
"step": 192
},
{
"epoch": 0.07921198440385799,
"grad_norm": 37.97771072387695,
"learning_rate": 9.207717569786536e-06,
"loss": 0.2056,
"step": 193
},
{
"epoch": 0.07962240919351529,
"grad_norm": 101.1618423461914,
"learning_rate": 9.203612479474549e-06,
"loss": 0.3672,
"step": 194
},
{
"epoch": 0.08003283398317258,
"grad_norm": 23.530296325683594,
"learning_rate": 9.199507389162562e-06,
"loss": 0.1018,
"step": 195
},
{
"epoch": 0.08044325877282987,
"grad_norm": 69.30397033691406,
"learning_rate": 9.195402298850575e-06,
"loss": 0.624,
"step": 196
},
{
"epoch": 0.08085368356248718,
"grad_norm": 26.590185165405273,
"learning_rate": 9.191297208538588e-06,
"loss": 0.0397,
"step": 197
},
{
"epoch": 0.08126410835214447,
"grad_norm": 24.015954971313477,
"learning_rate": 9.187192118226603e-06,
"loss": 0.1429,
"step": 198
},
{
"epoch": 0.08167453314180176,
"grad_norm": 76.20091247558594,
"learning_rate": 9.183087027914614e-06,
"loss": 0.8301,
"step": 199
},
{
"epoch": 0.08208495793145906,
"grad_norm": 12.56059741973877,
"learning_rate": 9.178981937602627e-06,
"loss": 0.0623,
"step": 200
},
{
"epoch": 0.08249538272111635,
"grad_norm": 25.811952590942383,
"learning_rate": 9.174876847290642e-06,
"loss": 0.0731,
"step": 201
},
{
"epoch": 0.08290580751077366,
"grad_norm": 108.41874694824219,
"learning_rate": 9.170771756978655e-06,
"loss": 0.8359,
"step": 202
},
{
"epoch": 0.08331623230043095,
"grad_norm": 51.9036750793457,
"learning_rate": 9.166666666666666e-06,
"loss": 0.4745,
"step": 203
},
{
"epoch": 0.08372665709008824,
"grad_norm": 63.12330627441406,
"learning_rate": 9.162561576354681e-06,
"loss": 0.3044,
"step": 204
},
{
"epoch": 0.08413708187974554,
"grad_norm": 113.08649444580078,
"learning_rate": 9.158456486042694e-06,
"loss": 0.7661,
"step": 205
},
{
"epoch": 0.08454750666940283,
"grad_norm": 88.81044006347656,
"learning_rate": 9.154351395730707e-06,
"loss": 0.5414,
"step": 206
},
{
"epoch": 0.08495793145906012,
"grad_norm": 30.185291290283203,
"learning_rate": 9.15024630541872e-06,
"loss": 0.0751,
"step": 207
},
{
"epoch": 0.08536835624871743,
"grad_norm": 78.1164321899414,
"learning_rate": 9.146141215106733e-06,
"loss": 0.6136,
"step": 208
},
{
"epoch": 0.08577878103837472,
"grad_norm": 146.73977661132812,
"learning_rate": 9.142036124794747e-06,
"loss": 1.3701,
"step": 209
},
{
"epoch": 0.08618920582803201,
"grad_norm": 48.88395309448242,
"learning_rate": 9.13793103448276e-06,
"loss": 0.1732,
"step": 210
},
{
"epoch": 0.08659963061768931,
"grad_norm": 55.54692459106445,
"learning_rate": 9.133825944170773e-06,
"loss": 0.4746,
"step": 211
},
{
"epoch": 0.0870100554073466,
"grad_norm": 22.834901809692383,
"learning_rate": 9.129720853858786e-06,
"loss": 0.0828,
"step": 212
},
{
"epoch": 0.0874204801970039,
"grad_norm": 39.23609924316406,
"learning_rate": 9.125615763546799e-06,
"loss": 0.1543,
"step": 213
},
{
"epoch": 0.0878309049866612,
"grad_norm": 17.21540641784668,
"learning_rate": 9.121510673234812e-06,
"loss": 0.0223,
"step": 214
},
{
"epoch": 0.08824132977631849,
"grad_norm": 17.433406829833984,
"learning_rate": 9.117405582922825e-06,
"loss": 0.0334,
"step": 215
},
{
"epoch": 0.08865175456597578,
"grad_norm": 11.346330642700195,
"learning_rate": 9.113300492610838e-06,
"loss": 0.0326,
"step": 216
},
{
"epoch": 0.08906217935563308,
"grad_norm": 46.6195182800293,
"learning_rate": 9.109195402298851e-06,
"loss": 0.4961,
"step": 217
},
{
"epoch": 0.08947260414529037,
"grad_norm": 147.48988342285156,
"learning_rate": 9.105090311986866e-06,
"loss": 0.46,
"step": 218
},
{
"epoch": 0.08988302893494766,
"grad_norm": 49.90272903442383,
"learning_rate": 9.100985221674877e-06,
"loss": 0.3223,
"step": 219
},
{
"epoch": 0.09029345372460497,
"grad_norm": 126.591064453125,
"learning_rate": 9.09688013136289e-06,
"loss": 1.0352,
"step": 220
},
{
"epoch": 0.09070387851426226,
"grad_norm": 41.19328308105469,
"learning_rate": 9.092775041050905e-06,
"loss": 0.2382,
"step": 221
},
{
"epoch": 0.09111430330391955,
"grad_norm": 31.060840606689453,
"learning_rate": 9.088669950738916e-06,
"loss": 0.1445,
"step": 222
},
{
"epoch": 0.09152472809357685,
"grad_norm": 18.65825843811035,
"learning_rate": 9.08456486042693e-06,
"loss": 0.4167,
"step": 223
},
{
"epoch": 0.09193515288323414,
"grad_norm": 24.073808670043945,
"learning_rate": 9.080459770114942e-06,
"loss": 0.0365,
"step": 224
},
{
"epoch": 0.09234557767289145,
"grad_norm": 99.49775695800781,
"learning_rate": 9.076354679802957e-06,
"loss": 0.9004,
"step": 225
},
{
"epoch": 0.09275600246254874,
"grad_norm": 25.354150772094727,
"learning_rate": 9.072249589490969e-06,
"loss": 0.0773,
"step": 226
},
{
"epoch": 0.09316642725220603,
"grad_norm": 13.476449966430664,
"learning_rate": 9.068144499178982e-06,
"loss": 0.0924,
"step": 227
},
{
"epoch": 0.09357685204186333,
"grad_norm": 38.99354553222656,
"learning_rate": 9.064039408866996e-06,
"loss": 0.2319,
"step": 228
},
{
"epoch": 0.09398727683152062,
"grad_norm": 35.506656646728516,
"learning_rate": 9.05993431855501e-06,
"loss": 0.4097,
"step": 229
},
{
"epoch": 0.09439770162117792,
"grad_norm": 78.80440521240234,
"learning_rate": 9.055829228243021e-06,
"loss": 0.8926,
"step": 230
},
{
"epoch": 0.09480812641083522,
"grad_norm": 50.49696731567383,
"learning_rate": 9.051724137931036e-06,
"loss": 0.6025,
"step": 231
},
{
"epoch": 0.09521855120049251,
"grad_norm": 36.3662109375,
"learning_rate": 9.047619047619049e-06,
"loss": 0.2991,
"step": 232
},
{
"epoch": 0.0956289759901498,
"grad_norm": 59.692203521728516,
"learning_rate": 9.043513957307062e-06,
"loss": 0.6436,
"step": 233
},
{
"epoch": 0.0960394007798071,
"grad_norm": 28.680530548095703,
"learning_rate": 9.039408866995075e-06,
"loss": 0.291,
"step": 234
},
{
"epoch": 0.0964498255694644,
"grad_norm": 190.74520874023438,
"learning_rate": 9.035303776683088e-06,
"loss": 1.2856,
"step": 235
},
{
"epoch": 0.09686025035912169,
"grad_norm": 34.94647979736328,
"learning_rate": 9.031198686371101e-06,
"loss": 0.2461,
"step": 236
},
{
"epoch": 0.09727067514877899,
"grad_norm": 25.99507713317871,
"learning_rate": 9.027093596059114e-06,
"loss": 0.2161,
"step": 237
},
{
"epoch": 0.09768109993843628,
"grad_norm": 90.48809814453125,
"learning_rate": 9.022988505747127e-06,
"loss": 1.1367,
"step": 238
},
{
"epoch": 0.09809152472809357,
"grad_norm": 93.10667419433594,
"learning_rate": 9.01888341543514e-06,
"loss": 0.2786,
"step": 239
},
{
"epoch": 0.09850194951775088,
"grad_norm": 36.001163482666016,
"learning_rate": 9.014778325123153e-06,
"loss": 0.1519,
"step": 240
},
{
"epoch": 0.09891237430740817,
"grad_norm": 126.11981964111328,
"learning_rate": 9.010673234811166e-06,
"loss": 0.9648,
"step": 241
},
{
"epoch": 0.09932279909706546,
"grad_norm": 48.246952056884766,
"learning_rate": 9.00656814449918e-06,
"loss": 0.2681,
"step": 242
},
{
"epoch": 0.09973322388672276,
"grad_norm": 21.30924415588379,
"learning_rate": 9.002463054187192e-06,
"loss": 0.0392,
"step": 243
},
{
"epoch": 0.10014364867638005,
"grad_norm": 70.95260620117188,
"learning_rate": 8.998357963875205e-06,
"loss": 0.2188,
"step": 244
},
{
"epoch": 0.10055407346603736,
"grad_norm": 8.646605491638184,
"learning_rate": 8.99425287356322e-06,
"loss": 0.0383,
"step": 245
},
{
"epoch": 0.10096449825569465,
"grad_norm": 20.226709365844727,
"learning_rate": 8.990147783251232e-06,
"loss": 0.0378,
"step": 246
},
{
"epoch": 0.10137492304535194,
"grad_norm": 15.134608268737793,
"learning_rate": 8.986042692939245e-06,
"loss": 0.0201,
"step": 247
},
{
"epoch": 0.10178534783500924,
"grad_norm": 193.9993133544922,
"learning_rate": 8.98193760262726e-06,
"loss": 2.3789,
"step": 248
},
{
"epoch": 0.10219577262466653,
"grad_norm": 92.06901550292969,
"learning_rate": 8.977832512315272e-06,
"loss": 0.3865,
"step": 249
},
{
"epoch": 0.10260619741432382,
"grad_norm": 130.63821411132812,
"learning_rate": 8.973727422003284e-06,
"loss": 0.8939,
"step": 250
},
{
"epoch": 0.10301662220398113,
"grad_norm": 30.600419998168945,
"learning_rate": 8.969622331691299e-06,
"loss": 0.0478,
"step": 251
},
{
"epoch": 0.10342704699363842,
"grad_norm": 71.38714599609375,
"learning_rate": 8.965517241379312e-06,
"loss": 0.4974,
"step": 252
},
{
"epoch": 0.1038374717832957,
"grad_norm": 68.43029022216797,
"learning_rate": 8.961412151067325e-06,
"loss": 0.4785,
"step": 253
},
{
"epoch": 0.10424789657295301,
"grad_norm": 19.519975662231445,
"learning_rate": 8.957307060755338e-06,
"loss": 0.0283,
"step": 254
},
{
"epoch": 0.1046583213626103,
"grad_norm": 14.962761878967285,
"learning_rate": 8.953201970443351e-06,
"loss": 0.0136,
"step": 255
},
{
"epoch": 0.10506874615226759,
"grad_norm": 14.830385208129883,
"learning_rate": 8.949096880131364e-06,
"loss": 0.0284,
"step": 256
},
{
"epoch": 0.1054791709419249,
"grad_norm": 47.00918960571289,
"learning_rate": 8.944991789819377e-06,
"loss": 0.1203,
"step": 257
},
{
"epoch": 0.10588959573158219,
"grad_norm": 13.37557315826416,
"learning_rate": 8.94088669950739e-06,
"loss": 0.0124,
"step": 258
},
{
"epoch": 0.10630002052123948,
"grad_norm": 304.811279296875,
"learning_rate": 8.936781609195403e-06,
"loss": 1.377,
"step": 259
},
{
"epoch": 0.10671044531089678,
"grad_norm": 23.555713653564453,
"learning_rate": 8.932676518883416e-06,
"loss": 0.0813,
"step": 260
},
{
"epoch": 0.10712087010055407,
"grad_norm": 172.98277282714844,
"learning_rate": 8.92857142857143e-06,
"loss": 0.752,
"step": 261
},
{
"epoch": 0.10753129489021136,
"grad_norm": 9.472747802734375,
"learning_rate": 8.924466338259442e-06,
"loss": 0.0272,
"step": 262
},
{
"epoch": 0.10794171967986867,
"grad_norm": 41.64617156982422,
"learning_rate": 8.920361247947455e-06,
"loss": 0.2184,
"step": 263
},
{
"epoch": 0.10835214446952596,
"grad_norm": 66.9391098022461,
"learning_rate": 8.916256157635468e-06,
"loss": 0.3513,
"step": 264
},
{
"epoch": 0.10876256925918326,
"grad_norm": 93.37043762207031,
"learning_rate": 8.912151067323481e-06,
"loss": 0.7051,
"step": 265
},
{
"epoch": 0.10917299404884055,
"grad_norm": 74.79525756835938,
"learning_rate": 8.908045977011495e-06,
"loss": 0.4552,
"step": 266
},
{
"epoch": 0.10958341883849784,
"grad_norm": 32.98771667480469,
"learning_rate": 8.903940886699508e-06,
"loss": 0.2937,
"step": 267
},
{
"epoch": 0.10999384362815515,
"grad_norm": 64.49002838134766,
"learning_rate": 8.89983579638752e-06,
"loss": 0.353,
"step": 268
},
{
"epoch": 0.11040426841781244,
"grad_norm": 75.34313201904297,
"learning_rate": 8.895730706075535e-06,
"loss": 0.4326,
"step": 269
},
{
"epoch": 0.11081469320746973,
"grad_norm": 38.482421875,
"learning_rate": 8.891625615763547e-06,
"loss": 0.4038,
"step": 270
},
{
"epoch": 0.11122511799712703,
"grad_norm": 76.3629379272461,
"learning_rate": 8.88752052545156e-06,
"loss": 0.3076,
"step": 271
},
{
"epoch": 0.11163554278678432,
"grad_norm": 85.81480407714844,
"learning_rate": 8.883415435139575e-06,
"loss": 0.4209,
"step": 272
},
{
"epoch": 0.11204596757644161,
"grad_norm": 93.22909545898438,
"learning_rate": 8.879310344827588e-06,
"loss": 0.3789,
"step": 273
},
{
"epoch": 0.11245639236609892,
"grad_norm": 71.13253021240234,
"learning_rate": 8.875205254515599e-06,
"loss": 0.2939,
"step": 274
},
{
"epoch": 0.11286681715575621,
"grad_norm": 44.35668182373047,
"learning_rate": 8.871100164203614e-06,
"loss": 0.2725,
"step": 275
},
{
"epoch": 0.1132772419454135,
"grad_norm": 192.66831970214844,
"learning_rate": 8.866995073891627e-06,
"loss": 1.199,
"step": 276
},
{
"epoch": 0.1136876667350708,
"grad_norm": 79.55447387695312,
"learning_rate": 8.862889983579638e-06,
"loss": 0.3291,
"step": 277
},
{
"epoch": 0.11409809152472809,
"grad_norm": 125.45100402832031,
"learning_rate": 8.858784893267653e-06,
"loss": 0.3916,
"step": 278
},
{
"epoch": 0.11450851631438538,
"grad_norm": 50.240142822265625,
"learning_rate": 8.854679802955666e-06,
"loss": 0.6887,
"step": 279
},
{
"epoch": 0.11491894110404269,
"grad_norm": 23.15044593811035,
"learning_rate": 8.85057471264368e-06,
"loss": 0.1133,
"step": 280
},
{
"epoch": 0.11532936589369998,
"grad_norm": 61.438507080078125,
"learning_rate": 8.846469622331692e-06,
"loss": 0.3577,
"step": 281
},
{
"epoch": 0.11573979068335727,
"grad_norm": 11.019624710083008,
"learning_rate": 8.842364532019705e-06,
"loss": 0.0336,
"step": 282
},
{
"epoch": 0.11615021547301457,
"grad_norm": 19.917442321777344,
"learning_rate": 8.838259441707718e-06,
"loss": 0.045,
"step": 283
},
{
"epoch": 0.11656064026267186,
"grad_norm": 148.9059295654297,
"learning_rate": 8.834154351395731e-06,
"loss": 0.5729,
"step": 284
},
{
"epoch": 0.11697106505232915,
"grad_norm": 335.07940673828125,
"learning_rate": 8.830049261083744e-06,
"loss": 2.0312,
"step": 285
},
{
"epoch": 0.11738148984198646,
"grad_norm": 77.18766021728516,
"learning_rate": 8.825944170771758e-06,
"loss": 0.3523,
"step": 286
},
{
"epoch": 0.11779191463164375,
"grad_norm": 316.40435791015625,
"learning_rate": 8.82183908045977e-06,
"loss": 1.6787,
"step": 287
},
{
"epoch": 0.11820233942130105,
"grad_norm": 205.8123779296875,
"learning_rate": 8.817733990147784e-06,
"loss": 0.6809,
"step": 288
},
{
"epoch": 0.11861276421095834,
"grad_norm": 138.25611877441406,
"learning_rate": 8.813628899835797e-06,
"loss": 1.1741,
"step": 289
},
{
"epoch": 0.11902318900061563,
"grad_norm": 117.10308074951172,
"learning_rate": 8.80952380952381e-06,
"loss": 0.8726,
"step": 290
},
{
"epoch": 0.11943361379027294,
"grad_norm": 69.2857437133789,
"learning_rate": 8.805418719211823e-06,
"loss": 0.1611,
"step": 291
},
{
"epoch": 0.11984403857993023,
"grad_norm": 85.6843032836914,
"learning_rate": 8.801313628899838e-06,
"loss": 0.2559,
"step": 292
},
{
"epoch": 0.12025446336958752,
"grad_norm": 250.38897705078125,
"learning_rate": 8.797208538587849e-06,
"loss": 0.7793,
"step": 293
},
{
"epoch": 0.12066488815924482,
"grad_norm": 26.297073364257812,
"learning_rate": 8.793103448275862e-06,
"loss": 0.038,
"step": 294
},
{
"epoch": 0.12107531294890211,
"grad_norm": 27.94345474243164,
"learning_rate": 8.788998357963877e-06,
"loss": 0.0886,
"step": 295
},
{
"epoch": 0.1214857377385594,
"grad_norm": 134.81576538085938,
"learning_rate": 8.78489326765189e-06,
"loss": 0.3984,
"step": 296
},
{
"epoch": 0.12189616252821671,
"grad_norm": 42.25836181640625,
"learning_rate": 8.780788177339901e-06,
"loss": 0.2509,
"step": 297
},
{
"epoch": 0.122306587317874,
"grad_norm": 23.695714950561523,
"learning_rate": 8.776683087027914e-06,
"loss": 0.0792,
"step": 298
},
{
"epoch": 0.12271701210753129,
"grad_norm": 318.5950927734375,
"learning_rate": 8.772577996715929e-06,
"loss": 0.918,
"step": 299
},
{
"epoch": 0.1231274368971886,
"grad_norm": 39.690948486328125,
"learning_rate": 8.768472906403942e-06,
"loss": 0.1516,
"step": 300
},
{
"epoch": 0.12353786168684588,
"grad_norm": 36.41111755371094,
"learning_rate": 8.764367816091954e-06,
"loss": 0.5393,
"step": 301
},
{
"epoch": 0.12394828647650317,
"grad_norm": 197.5860137939453,
"learning_rate": 8.760262725779968e-06,
"loss": 0.5239,
"step": 302
},
{
"epoch": 0.12435871126616048,
"grad_norm": 72.9620590209961,
"learning_rate": 8.756157635467981e-06,
"loss": 1.1562,
"step": 303
},
{
"epoch": 0.12476913605581777,
"grad_norm": 42.373817443847656,
"learning_rate": 8.752052545155994e-06,
"loss": 0.6348,
"step": 304
},
{
"epoch": 0.12517956084547507,
"grad_norm": 16.23139190673828,
"learning_rate": 8.747947454844007e-06,
"loss": 0.0551,
"step": 305
},
{
"epoch": 0.12558998563513235,
"grad_norm": 121.77295684814453,
"learning_rate": 8.74384236453202e-06,
"loss": 0.5718,
"step": 306
},
{
"epoch": 0.12600041042478966,
"grad_norm": 26.034849166870117,
"learning_rate": 8.739737274220034e-06,
"loss": 0.4464,
"step": 307
},
{
"epoch": 0.12641083521444696,
"grad_norm": 56.735450744628906,
"learning_rate": 8.735632183908047e-06,
"loss": 0.1248,
"step": 308
},
{
"epoch": 0.12682126000410424,
"grad_norm": 25.753250122070312,
"learning_rate": 8.73152709359606e-06,
"loss": 0.2026,
"step": 309
},
{
"epoch": 0.12723168479376154,
"grad_norm": 46.3178825378418,
"learning_rate": 8.727422003284073e-06,
"loss": 0.1462,
"step": 310
},
{
"epoch": 0.12764210958341884,
"grad_norm": 110.8947982788086,
"learning_rate": 8.723316912972086e-06,
"loss": 0.6357,
"step": 311
},
{
"epoch": 0.12805253437307612,
"grad_norm": 29.856884002685547,
"learning_rate": 8.719211822660099e-06,
"loss": 0.0525,
"step": 312
},
{
"epoch": 0.12846295916273343,
"grad_norm": 94.4979019165039,
"learning_rate": 8.715106732348112e-06,
"loss": 0.3315,
"step": 313
},
{
"epoch": 0.12887338395239073,
"grad_norm": 18.520299911499023,
"learning_rate": 8.711001642036125e-06,
"loss": 0.0715,
"step": 314
},
{
"epoch": 0.129283808742048,
"grad_norm": 74.54774475097656,
"learning_rate": 8.706896551724138e-06,
"loss": 0.3248,
"step": 315
},
{
"epoch": 0.1296942335317053,
"grad_norm": 39.153018951416016,
"learning_rate": 8.702791461412153e-06,
"loss": 0.1953,
"step": 316
},
{
"epoch": 0.13010465832136262,
"grad_norm": 90.47451782226562,
"learning_rate": 8.698686371100164e-06,
"loss": 0.3271,
"step": 317
},
{
"epoch": 0.13051508311101992,
"grad_norm": 37.40937042236328,
"learning_rate": 8.694581280788177e-06,
"loss": 0.5723,
"step": 318
},
{
"epoch": 0.1309255079006772,
"grad_norm": 33.7971305847168,
"learning_rate": 8.690476190476192e-06,
"loss": 0.1489,
"step": 319
},
{
"epoch": 0.1313359326903345,
"grad_norm": 92.07005310058594,
"learning_rate": 8.686371100164205e-06,
"loss": 0.8672,
"step": 320
},
{
"epoch": 0.1317463574799918,
"grad_norm": 91.0810546875,
"learning_rate": 8.682266009852216e-06,
"loss": 0.7148,
"step": 321
},
{
"epoch": 0.13215678226964908,
"grad_norm": 41.58576202392578,
"learning_rate": 8.678160919540231e-06,
"loss": 0.3192,
"step": 322
},
{
"epoch": 0.13256720705930639,
"grad_norm": 29.5936222076416,
"learning_rate": 8.674055829228244e-06,
"loss": 0.0693,
"step": 323
},
{
"epoch": 0.1329776318489637,
"grad_norm": 24.272443771362305,
"learning_rate": 8.669950738916257e-06,
"loss": 0.0497,
"step": 324
},
{
"epoch": 0.13338805663862097,
"grad_norm": 198.68040466308594,
"learning_rate": 8.66584564860427e-06,
"loss": 0.4736,
"step": 325
},
{
"epoch": 0.13379848142827827,
"grad_norm": 21.794588088989258,
"learning_rate": 8.661740558292284e-06,
"loss": 0.1947,
"step": 326
},
{
"epoch": 0.13420890621793558,
"grad_norm": 11.848313331604004,
"learning_rate": 8.657635467980297e-06,
"loss": 0.0253,
"step": 327
},
{
"epoch": 0.13461933100759285,
"grad_norm": 18.124849319458008,
"learning_rate": 8.65353037766831e-06,
"loss": 0.1462,
"step": 328
},
{
"epoch": 0.13502975579725016,
"grad_norm": 55.430057525634766,
"learning_rate": 8.649425287356323e-06,
"loss": 0.1768,
"step": 329
},
{
"epoch": 0.13544018058690746,
"grad_norm": 35.88666534423828,
"learning_rate": 8.645320197044336e-06,
"loss": 0.1951,
"step": 330
},
{
"epoch": 0.13585060537656474,
"grad_norm": 23.222196578979492,
"learning_rate": 8.641215106732349e-06,
"loss": 0.0994,
"step": 331
},
{
"epoch": 0.13626103016622204,
"grad_norm": 127.41805267333984,
"learning_rate": 8.637110016420362e-06,
"loss": 0.9129,
"step": 332
},
{
"epoch": 0.13667145495587935,
"grad_norm": 55.741912841796875,
"learning_rate": 8.633004926108375e-06,
"loss": 0.2053,
"step": 333
},
{
"epoch": 0.13708187974553662,
"grad_norm": 37.40507507324219,
"learning_rate": 8.628899835796388e-06,
"loss": 0.0863,
"step": 334
},
{
"epoch": 0.13749230453519393,
"grad_norm": 204.22210693359375,
"learning_rate": 8.624794745484401e-06,
"loss": 2.8906,
"step": 335
},
{
"epoch": 0.13790272932485123,
"grad_norm": 34.04326248168945,
"learning_rate": 8.620689655172414e-06,
"loss": 0.3315,
"step": 336
},
{
"epoch": 0.1383131541145085,
"grad_norm": 27.861135482788086,
"learning_rate": 8.616584564860427e-06,
"loss": 0.095,
"step": 337
},
{
"epoch": 0.1387235789041658,
"grad_norm": 77.86309051513672,
"learning_rate": 8.61247947454844e-06,
"loss": 0.2264,
"step": 338
},
{
"epoch": 0.13913400369382312,
"grad_norm": 32.8078498840332,
"learning_rate": 8.608374384236455e-06,
"loss": 0.0867,
"step": 339
},
{
"epoch": 0.1395444284834804,
"grad_norm": 8.150931358337402,
"learning_rate": 8.604269293924466e-06,
"loss": 0.0156,
"step": 340
},
{
"epoch": 0.1399548532731377,
"grad_norm": 59.5302848815918,
"learning_rate": 8.60016420361248e-06,
"loss": 0.1979,
"step": 341
},
{
"epoch": 0.140365278062795,
"grad_norm": 211.95187377929688,
"learning_rate": 8.596059113300493e-06,
"loss": 1.5059,
"step": 342
},
{
"epoch": 0.14077570285245228,
"grad_norm": 9.38349437713623,
"learning_rate": 8.591954022988507e-06,
"loss": 0.0392,
"step": 343
},
{
"epoch": 0.14118612764210958,
"grad_norm": 73.36874389648438,
"learning_rate": 8.587848932676519e-06,
"loss": 0.2478,
"step": 344
},
{
"epoch": 0.1415965524317669,
"grad_norm": 10.98017692565918,
"learning_rate": 8.583743842364532e-06,
"loss": 0.0168,
"step": 345
},
{
"epoch": 0.14200697722142416,
"grad_norm": 182.12167358398438,
"learning_rate": 8.579638752052546e-06,
"loss": 1.0938,
"step": 346
},
{
"epoch": 0.14241740201108147,
"grad_norm": 95.05304718017578,
"learning_rate": 8.57553366174056e-06,
"loss": 0.2582,
"step": 347
},
{
"epoch": 0.14282782680073877,
"grad_norm": 156.24925231933594,
"learning_rate": 8.571428571428571e-06,
"loss": 0.863,
"step": 348
},
{
"epoch": 0.14323825159039605,
"grad_norm": 26.966121673583984,
"learning_rate": 8.567323481116586e-06,
"loss": 0.0516,
"step": 349
},
{
"epoch": 0.14364867638005335,
"grad_norm": 31.235546112060547,
"learning_rate": 8.563218390804599e-06,
"loss": 0.0623,
"step": 350
},
{
"epoch": 0.14405910116971066,
"grad_norm": 13.845409393310547,
"learning_rate": 8.559113300492612e-06,
"loss": 0.0333,
"step": 351
},
{
"epoch": 0.14446952595936793,
"grad_norm": 40.16856384277344,
"learning_rate": 8.555008210180625e-06,
"loss": 0.1445,
"step": 352
},
{
"epoch": 0.14487995074902524,
"grad_norm": 51.43763732910156,
"learning_rate": 8.550903119868638e-06,
"loss": 0.2444,
"step": 353
},
{
"epoch": 0.14529037553868254,
"grad_norm": 54.58685302734375,
"learning_rate": 8.546798029556651e-06,
"loss": 0.4394,
"step": 354
},
{
"epoch": 0.14570080032833982,
"grad_norm": 38.69440841674805,
"learning_rate": 8.542692939244664e-06,
"loss": 0.0983,
"step": 355
},
{
"epoch": 0.14611122511799712,
"grad_norm": 75.47493743896484,
"learning_rate": 8.538587848932677e-06,
"loss": 0.3687,
"step": 356
},
{
"epoch": 0.14652164990765443,
"grad_norm": 22.57318878173828,
"learning_rate": 8.53448275862069e-06,
"loss": 0.0773,
"step": 357
},
{
"epoch": 0.1469320746973117,
"grad_norm": 66.54450225830078,
"learning_rate": 8.530377668308703e-06,
"loss": 0.3232,
"step": 358
},
{
"epoch": 0.147342499486969,
"grad_norm": 63.69831848144531,
"learning_rate": 8.526272577996716e-06,
"loss": 0.1807,
"step": 359
},
{
"epoch": 0.1477529242766263,
"grad_norm": 57.714141845703125,
"learning_rate": 8.52216748768473e-06,
"loss": 0.2985,
"step": 360
},
{
"epoch": 0.14816334906628362,
"grad_norm": 74.93372344970703,
"learning_rate": 8.518062397372742e-06,
"loss": 0.2715,
"step": 361
},
{
"epoch": 0.1485737738559409,
"grad_norm": 58.63027572631836,
"learning_rate": 8.513957307060756e-06,
"loss": 0.3561,
"step": 362
},
{
"epoch": 0.1489841986455982,
"grad_norm": 50.723365783691406,
"learning_rate": 8.50985221674877e-06,
"loss": 0.2914,
"step": 363
},
{
"epoch": 0.1493946234352555,
"grad_norm": 49.11933517456055,
"learning_rate": 8.505747126436782e-06,
"loss": 0.0783,
"step": 364
},
{
"epoch": 0.14980504822491278,
"grad_norm": 48.15220260620117,
"learning_rate": 8.501642036124795e-06,
"loss": 0.1317,
"step": 365
},
{
"epoch": 0.15021547301457008,
"grad_norm": 152.74365234375,
"learning_rate": 8.49753694581281e-06,
"loss": 0.4211,
"step": 366
},
{
"epoch": 0.1506258978042274,
"grad_norm": 39.953941345214844,
"learning_rate": 8.493431855500823e-06,
"loss": 0.3148,
"step": 367
},
{
"epoch": 0.15103632259388466,
"grad_norm": 34.75541305541992,
"learning_rate": 8.489326765188834e-06,
"loss": 0.1536,
"step": 368
},
{
"epoch": 0.15144674738354197,
"grad_norm": 60.90129852294922,
"learning_rate": 8.485221674876849e-06,
"loss": 0.2306,
"step": 369
},
{
"epoch": 0.15185717217319927,
"grad_norm": 360.6073303222656,
"learning_rate": 8.481116584564862e-06,
"loss": 0.8448,
"step": 370
},
{
"epoch": 0.15226759696285655,
"grad_norm": 38.120269775390625,
"learning_rate": 8.477011494252875e-06,
"loss": 0.263,
"step": 371
},
{
"epoch": 0.15267802175251385,
"grad_norm": 41.20004653930664,
"learning_rate": 8.472906403940888e-06,
"loss": 0.1062,
"step": 372
},
{
"epoch": 0.15308844654217116,
"grad_norm": 64.84685516357422,
"learning_rate": 8.468801313628901e-06,
"loss": 0.375,
"step": 373
},
{
"epoch": 0.15349887133182843,
"grad_norm": 130.30555725097656,
"learning_rate": 8.464696223316914e-06,
"loss": 0.9854,
"step": 374
},
{
"epoch": 0.15390929612148574,
"grad_norm": 28.048683166503906,
"learning_rate": 8.460591133004927e-06,
"loss": 0.0734,
"step": 375
},
{
"epoch": 0.15431972091114304,
"grad_norm": 75.84153747558594,
"learning_rate": 8.45648604269294e-06,
"loss": 0.4297,
"step": 376
},
{
"epoch": 0.15473014570080032,
"grad_norm": 37.49334716796875,
"learning_rate": 8.452380952380953e-06,
"loss": 0.2374,
"step": 377
},
{
"epoch": 0.15514057049045762,
"grad_norm": 79.43096923828125,
"learning_rate": 8.448275862068966e-06,
"loss": 0.543,
"step": 378
},
{
"epoch": 0.15555099528011493,
"grad_norm": 102.65121459960938,
"learning_rate": 8.44417077175698e-06,
"loss": 1.0674,
"step": 379
},
{
"epoch": 0.1559614200697722,
"grad_norm": 172.36798095703125,
"learning_rate": 8.440065681444992e-06,
"loss": 0.5454,
"step": 380
},
{
"epoch": 0.1563718448594295,
"grad_norm": 21.131412506103516,
"learning_rate": 8.435960591133005e-06,
"loss": 0.1038,
"step": 381
},
{
"epoch": 0.15678226964908681,
"grad_norm": 89.00126647949219,
"learning_rate": 8.431855500821018e-06,
"loss": 0.4153,
"step": 382
},
{
"epoch": 0.1571926944387441,
"grad_norm": 20.64133644104004,
"learning_rate": 8.427750410509033e-06,
"loss": 0.0324,
"step": 383
},
{
"epoch": 0.1576031192284014,
"grad_norm": 77.31906127929688,
"learning_rate": 8.423645320197045e-06,
"loss": 0.4261,
"step": 384
},
{
"epoch": 0.1580135440180587,
"grad_norm": 9.301074028015137,
"learning_rate": 8.419540229885058e-06,
"loss": 0.0348,
"step": 385
},
{
"epoch": 0.15842396880771598,
"grad_norm": 60.01475143432617,
"learning_rate": 8.41543513957307e-06,
"loss": 0.2443,
"step": 386
},
{
"epoch": 0.15883439359737328,
"grad_norm": 75.8565902709961,
"learning_rate": 8.411330049261086e-06,
"loss": 0.2668,
"step": 387
},
{
"epoch": 0.15924481838703058,
"grad_norm": 126.0774154663086,
"learning_rate": 8.407224958949097e-06,
"loss": 0.5652,
"step": 388
},
{
"epoch": 0.15965524317668786,
"grad_norm": 78.51138305664062,
"learning_rate": 8.40311986863711e-06,
"loss": 0.4297,
"step": 389
},
{
"epoch": 0.16006566796634517,
"grad_norm": 61.736148834228516,
"learning_rate": 8.399014778325125e-06,
"loss": 0.2526,
"step": 390
},
{
"epoch": 0.16047609275600247,
"grad_norm": 27.032468795776367,
"learning_rate": 8.394909688013136e-06,
"loss": 0.0446,
"step": 391
},
{
"epoch": 0.16088651754565975,
"grad_norm": 57.427452087402344,
"learning_rate": 8.390804597701149e-06,
"loss": 0.3901,
"step": 392
},
{
"epoch": 0.16129694233531705,
"grad_norm": 42.8073616027832,
"learning_rate": 8.386699507389164e-06,
"loss": 0.1353,
"step": 393
},
{
"epoch": 0.16170736712497435,
"grad_norm": 146.55259704589844,
"learning_rate": 8.382594417077177e-06,
"loss": 0.709,
"step": 394
},
{
"epoch": 0.16211779191463163,
"grad_norm": 42.785194396972656,
"learning_rate": 8.378489326765188e-06,
"loss": 0.1416,
"step": 395
},
{
"epoch": 0.16252821670428894,
"grad_norm": 42.01571273803711,
"learning_rate": 8.374384236453203e-06,
"loss": 0.1772,
"step": 396
},
{
"epoch": 0.16293864149394624,
"grad_norm": 126.4945068359375,
"learning_rate": 8.370279146141216e-06,
"loss": 0.8301,
"step": 397
},
{
"epoch": 0.16334906628360352,
"grad_norm": 31.53195571899414,
"learning_rate": 8.36617405582923e-06,
"loss": 0.1182,
"step": 398
},
{
"epoch": 0.16375949107326082,
"grad_norm": 45.67702865600586,
"learning_rate": 8.362068965517242e-06,
"loss": 0.1952,
"step": 399
},
{
"epoch": 0.16416991586291813,
"grad_norm": 62.60844039916992,
"learning_rate": 8.357963875205255e-06,
"loss": 0.1721,
"step": 400
},
{
"epoch": 0.1645803406525754,
"grad_norm": 94.92082214355469,
"learning_rate": 8.353858784893268e-06,
"loss": 0.2183,
"step": 401
},
{
"epoch": 0.1649907654422327,
"grad_norm": 45.898292541503906,
"learning_rate": 8.349753694581281e-06,
"loss": 0.0959,
"step": 402
},
{
"epoch": 0.16540119023189,
"grad_norm": 72.65533447265625,
"learning_rate": 8.345648604269295e-06,
"loss": 0.146,
"step": 403
},
{
"epoch": 0.16581161502154732,
"grad_norm": 34.08324432373047,
"learning_rate": 8.341543513957308e-06,
"loss": 0.0786,
"step": 404
},
{
"epoch": 0.1662220398112046,
"grad_norm": 66.44620513916016,
"learning_rate": 8.33743842364532e-06,
"loss": 0.5679,
"step": 405
},
{
"epoch": 0.1666324646008619,
"grad_norm": 10.144757270812988,
"learning_rate": 8.333333333333334e-06,
"loss": 0.0215,
"step": 406
},
{
"epoch": 0.1670428893905192,
"grad_norm": 117.08031463623047,
"learning_rate": 8.329228243021347e-06,
"loss": 0.5674,
"step": 407
},
{
"epoch": 0.16745331418017648,
"grad_norm": 54.263816833496094,
"learning_rate": 8.32512315270936e-06,
"loss": 0.1321,
"step": 408
},
{
"epoch": 0.16786373896983378,
"grad_norm": 335.7568054199219,
"learning_rate": 8.321018062397373e-06,
"loss": 4.1875,
"step": 409
},
{
"epoch": 0.16827416375949109,
"grad_norm": 112.59076690673828,
"learning_rate": 8.316912972085388e-06,
"loss": 0.6958,
"step": 410
},
{
"epoch": 0.16868458854914836,
"grad_norm": 49.904415130615234,
"learning_rate": 8.312807881773399e-06,
"loss": 0.3442,
"step": 411
},
{
"epoch": 0.16909501333880567,
"grad_norm": 19.457067489624023,
"learning_rate": 8.308702791461412e-06,
"loss": 0.0302,
"step": 412
},
{
"epoch": 0.16950543812846297,
"grad_norm": 12.766022682189941,
"learning_rate": 8.304597701149427e-06,
"loss": 0.0455,
"step": 413
},
{
"epoch": 0.16991586291812025,
"grad_norm": 10.48725700378418,
"learning_rate": 8.30049261083744e-06,
"loss": 0.0169,
"step": 414
},
{
"epoch": 0.17032628770777755,
"grad_norm": 19.15460777282715,
"learning_rate": 8.296387520525451e-06,
"loss": 0.0352,
"step": 415
},
{
"epoch": 0.17073671249743486,
"grad_norm": 54.21415328979492,
"learning_rate": 8.292282430213466e-06,
"loss": 0.1484,
"step": 416
},
{
"epoch": 0.17114713728709213,
"grad_norm": 8.854345321655273,
"learning_rate": 8.288177339901479e-06,
"loss": 0.0314,
"step": 417
},
{
"epoch": 0.17155756207674944,
"grad_norm": 58.263553619384766,
"learning_rate": 8.284072249589492e-06,
"loss": 0.3568,
"step": 418
},
{
"epoch": 0.17196798686640674,
"grad_norm": 173.80926513671875,
"learning_rate": 8.279967159277504e-06,
"loss": 0.5963,
"step": 419
},
{
"epoch": 0.17237841165606402,
"grad_norm": 130.8844451904297,
"learning_rate": 8.275862068965518e-06,
"loss": 0.5659,
"step": 420
},
{
"epoch": 0.17278883644572132,
"grad_norm": 124.97215270996094,
"learning_rate": 8.271756978653531e-06,
"loss": 0.5352,
"step": 421
},
{
"epoch": 0.17319926123537863,
"grad_norm": 88.1253890991211,
"learning_rate": 8.267651888341544e-06,
"loss": 0.5269,
"step": 422
},
{
"epoch": 0.1736096860250359,
"grad_norm": 186.74684143066406,
"learning_rate": 8.263546798029558e-06,
"loss": 1.1895,
"step": 423
},
{
"epoch": 0.1740201108146932,
"grad_norm": 103.20738220214844,
"learning_rate": 8.25944170771757e-06,
"loss": 0.5605,
"step": 424
},
{
"epoch": 0.1744305356043505,
"grad_norm": 29.222522735595703,
"learning_rate": 8.255336617405584e-06,
"loss": 0.0427,
"step": 425
},
{
"epoch": 0.1748409603940078,
"grad_norm": 95.60066223144531,
"learning_rate": 8.251231527093597e-06,
"loss": 0.5278,
"step": 426
},
{
"epoch": 0.1752513851836651,
"grad_norm": 15.819268226623535,
"learning_rate": 8.24712643678161e-06,
"loss": 0.024,
"step": 427
},
{
"epoch": 0.1756618099733224,
"grad_norm": 44.277103424072266,
"learning_rate": 8.243021346469623e-06,
"loss": 0.4824,
"step": 428
},
{
"epoch": 0.17607223476297967,
"grad_norm": 40.34841537475586,
"learning_rate": 8.238916256157636e-06,
"loss": 0.1033,
"step": 429
},
{
"epoch": 0.17648265955263698,
"grad_norm": 79.78141784667969,
"learning_rate": 8.234811165845649e-06,
"loss": 0.1045,
"step": 430
},
{
"epoch": 0.17689308434229428,
"grad_norm": 69.09398651123047,
"learning_rate": 8.230706075533662e-06,
"loss": 0.2681,
"step": 431
},
{
"epoch": 0.17730350913195156,
"grad_norm": 44.11439514160156,
"learning_rate": 8.226600985221675e-06,
"loss": 0.1709,
"step": 432
},
{
"epoch": 0.17771393392160886,
"grad_norm": 21.752050399780273,
"learning_rate": 8.222495894909688e-06,
"loss": 0.0448,
"step": 433
},
{
"epoch": 0.17812435871126617,
"grad_norm": 132.4776611328125,
"learning_rate": 8.218390804597703e-06,
"loss": 0.8618,
"step": 434
},
{
"epoch": 0.17853478350092344,
"grad_norm": 60.86586380004883,
"learning_rate": 8.214285714285714e-06,
"loss": 0.2817,
"step": 435
},
{
"epoch": 0.17894520829058075,
"grad_norm": 11.644339561462402,
"learning_rate": 8.210180623973727e-06,
"loss": 0.0226,
"step": 436
},
{
"epoch": 0.17935563308023805,
"grad_norm": 22.99939727783203,
"learning_rate": 8.206075533661742e-06,
"loss": 0.0367,
"step": 437
},
{
"epoch": 0.17976605786989533,
"grad_norm": 51.97944641113281,
"learning_rate": 8.201970443349755e-06,
"loss": 0.4072,
"step": 438
},
{
"epoch": 0.18017648265955263,
"grad_norm": 14.462024688720703,
"learning_rate": 8.197865353037767e-06,
"loss": 0.0471,
"step": 439
},
{
"epoch": 0.18058690744920994,
"grad_norm": 10.458850860595703,
"learning_rate": 8.193760262725781e-06,
"loss": 0.0366,
"step": 440
},
{
"epoch": 0.18099733223886721,
"grad_norm": 28.052602767944336,
"learning_rate": 8.189655172413794e-06,
"loss": 0.0454,
"step": 441
},
{
"epoch": 0.18140775702852452,
"grad_norm": 7.607790470123291,
"learning_rate": 8.185550082101807e-06,
"loss": 0.0199,
"step": 442
},
{
"epoch": 0.18181818181818182,
"grad_norm": 45.58821487426758,
"learning_rate": 8.18144499178982e-06,
"loss": 0.175,
"step": 443
},
{
"epoch": 0.1822286066078391,
"grad_norm": 169.98419189453125,
"learning_rate": 8.177339901477834e-06,
"loss": 1.1104,
"step": 444
},
{
"epoch": 0.1826390313974964,
"grad_norm": 14.712911605834961,
"learning_rate": 8.173234811165847e-06,
"loss": 0.025,
"step": 445
},
{
"epoch": 0.1830494561871537,
"grad_norm": 13.16358470916748,
"learning_rate": 8.16912972085386e-06,
"loss": 0.0316,
"step": 446
},
{
"epoch": 0.183459880976811,
"grad_norm": 13.904580116271973,
"learning_rate": 8.165024630541873e-06,
"loss": 0.0204,
"step": 447
},
{
"epoch": 0.1838703057664683,
"grad_norm": 73.77775573730469,
"learning_rate": 8.160919540229886e-06,
"loss": 0.5806,
"step": 448
},
{
"epoch": 0.1842807305561256,
"grad_norm": 65.99251556396484,
"learning_rate": 8.156814449917899e-06,
"loss": 0.5235,
"step": 449
},
{
"epoch": 0.1846911553457829,
"grad_norm": 77.89680480957031,
"learning_rate": 8.152709359605912e-06,
"loss": 0.5205,
"step": 450
},
{
"epoch": 0.18510158013544017,
"grad_norm": 44.63160705566406,
"learning_rate": 8.148604269293925e-06,
"loss": 0.1556,
"step": 451
},
{
"epoch": 0.18551200492509748,
"grad_norm": 23.031660079956055,
"learning_rate": 8.144499178981938e-06,
"loss": 0.0911,
"step": 452
},
{
"epoch": 0.18592242971475478,
"grad_norm": 57.4281005859375,
"learning_rate": 8.140394088669951e-06,
"loss": 0.1738,
"step": 453
},
{
"epoch": 0.18633285450441206,
"grad_norm": 33.927734375,
"learning_rate": 8.136288998357964e-06,
"loss": 0.0957,
"step": 454
},
{
"epoch": 0.18674327929406936,
"grad_norm": 84.82231140136719,
"learning_rate": 8.132183908045977e-06,
"loss": 0.3438,
"step": 455
},
{
"epoch": 0.18715370408372667,
"grad_norm": 42.61494827270508,
"learning_rate": 8.12807881773399e-06,
"loss": 0.1455,
"step": 456
},
{
"epoch": 0.18756412887338395,
"grad_norm": 46.76593017578125,
"learning_rate": 8.123973727422005e-06,
"loss": 0.1316,
"step": 457
},
{
"epoch": 0.18797455366304125,
"grad_norm": 37.79883575439453,
"learning_rate": 8.119868637110016e-06,
"loss": 0.0967,
"step": 458
},
{
"epoch": 0.18838497845269855,
"grad_norm": 28.960237503051758,
"learning_rate": 8.11576354679803e-06,
"loss": 0.0554,
"step": 459
},
{
"epoch": 0.18879540324235583,
"grad_norm": 109.664794921875,
"learning_rate": 8.111658456486044e-06,
"loss": 0.5121,
"step": 460
},
{
"epoch": 0.18920582803201313,
"grad_norm": 46.965755462646484,
"learning_rate": 8.107553366174057e-06,
"loss": 0.25,
"step": 461
},
{
"epoch": 0.18961625282167044,
"grad_norm": 49.25603103637695,
"learning_rate": 8.103448275862069e-06,
"loss": 0.334,
"step": 462
},
{
"epoch": 0.19002667761132772,
"grad_norm": 50.01898193359375,
"learning_rate": 8.099343185550082e-06,
"loss": 0.2037,
"step": 463
},
{
"epoch": 0.19043710240098502,
"grad_norm": 7.691561222076416,
"learning_rate": 8.095238095238097e-06,
"loss": 0.0186,
"step": 464
},
{
"epoch": 0.19084752719064232,
"grad_norm": 24.228275299072266,
"learning_rate": 8.09113300492611e-06,
"loss": 0.0399,
"step": 465
},
{
"epoch": 0.1912579519802996,
"grad_norm": 65.93159484863281,
"learning_rate": 8.087027914614121e-06,
"loss": 0.2711,
"step": 466
},
{
"epoch": 0.1916683767699569,
"grad_norm": 4.968407154083252,
"learning_rate": 8.082922824302136e-06,
"loss": 0.0068,
"step": 467
},
{
"epoch": 0.1920788015596142,
"grad_norm": 42.073089599609375,
"learning_rate": 8.078817733990149e-06,
"loss": 0.0985,
"step": 468
},
{
"epoch": 0.19248922634927149,
"grad_norm": 28.897235870361328,
"learning_rate": 8.074712643678162e-06,
"loss": 0.1503,
"step": 469
},
{
"epoch": 0.1928996511389288,
"grad_norm": 30.285858154296875,
"learning_rate": 8.070607553366175e-06,
"loss": 0.0618,
"step": 470
},
{
"epoch": 0.1933100759285861,
"grad_norm": 40.072391510009766,
"learning_rate": 8.066502463054188e-06,
"loss": 0.1243,
"step": 471
},
{
"epoch": 0.19372050071824337,
"grad_norm": 43.57062530517578,
"learning_rate": 8.062397372742201e-06,
"loss": 0.1252,
"step": 472
},
{
"epoch": 0.19413092550790068,
"grad_norm": 25.83587074279785,
"learning_rate": 8.058292282430214e-06,
"loss": 0.0374,
"step": 473
},
{
"epoch": 0.19454135029755798,
"grad_norm": 5.446057319641113,
"learning_rate": 8.054187192118227e-06,
"loss": 0.0086,
"step": 474
},
{
"epoch": 0.19495177508721526,
"grad_norm": 116.27681732177734,
"learning_rate": 8.05008210180624e-06,
"loss": 1.1704,
"step": 475
},
{
"epoch": 0.19536219987687256,
"grad_norm": 34.303043365478516,
"learning_rate": 8.045977011494253e-06,
"loss": 0.0803,
"step": 476
},
{
"epoch": 0.19577262466652987,
"grad_norm": 25.756929397583008,
"learning_rate": 8.041871921182266e-06,
"loss": 0.0576,
"step": 477
},
{
"epoch": 0.19618304945618714,
"grad_norm": 62.485111236572266,
"learning_rate": 8.03776683087028e-06,
"loss": 0.3117,
"step": 478
},
{
"epoch": 0.19659347424584445,
"grad_norm": 22.258893966674805,
"learning_rate": 8.033661740558293e-06,
"loss": 0.1149,
"step": 479
},
{
"epoch": 0.19700389903550175,
"grad_norm": 47.862003326416016,
"learning_rate": 8.029556650246306e-06,
"loss": 0.4291,
"step": 480
},
{
"epoch": 0.19741432382515903,
"grad_norm": 34.473609924316406,
"learning_rate": 8.02545155993432e-06,
"loss": 0.1428,
"step": 481
},
{
"epoch": 0.19782474861481633,
"grad_norm": 9.425722122192383,
"learning_rate": 8.021346469622332e-06,
"loss": 0.0169,
"step": 482
},
{
"epoch": 0.19823517340447364,
"grad_norm": 40.93657684326172,
"learning_rate": 8.017241379310345e-06,
"loss": 0.2173,
"step": 483
},
{
"epoch": 0.1986455981941309,
"grad_norm": 11.333935737609863,
"learning_rate": 8.01313628899836e-06,
"loss": 0.0181,
"step": 484
},
{
"epoch": 0.19905602298378822,
"grad_norm": 26.066720962524414,
"learning_rate": 8.009031198686373e-06,
"loss": 0.0768,
"step": 485
},
{
"epoch": 0.19946644777344552,
"grad_norm": 7.945657253265381,
"learning_rate": 8.004926108374384e-06,
"loss": 0.022,
"step": 486
},
{
"epoch": 0.1998768725631028,
"grad_norm": 67.08000183105469,
"learning_rate": 8.000821018062399e-06,
"loss": 0.8615,
"step": 487
},
{
"epoch": 0.2002872973527601,
"grad_norm": 11.36076831817627,
"learning_rate": 7.996715927750412e-06,
"loss": 0.025,
"step": 488
},
{
"epoch": 0.2006977221424174,
"grad_norm": 81.91326904296875,
"learning_rate": 7.992610837438425e-06,
"loss": 0.3485,
"step": 489
},
{
"epoch": 0.2011081469320747,
"grad_norm": 2.548905372619629,
"learning_rate": 7.988505747126438e-06,
"loss": 0.0102,
"step": 490
},
{
"epoch": 0.201518571721732,
"grad_norm": 13.790837287902832,
"learning_rate": 7.984400656814451e-06,
"loss": 0.0525,
"step": 491
},
{
"epoch": 0.2019289965113893,
"grad_norm": 43.6253547668457,
"learning_rate": 7.980295566502464e-06,
"loss": 0.1584,
"step": 492
},
{
"epoch": 0.2023394213010466,
"grad_norm": 260.183837890625,
"learning_rate": 7.976190476190477e-06,
"loss": 2.625,
"step": 493
},
{
"epoch": 0.20274984609070387,
"grad_norm": 102.46743774414062,
"learning_rate": 7.97208538587849e-06,
"loss": 0.3428,
"step": 494
},
{
"epoch": 0.20316027088036118,
"grad_norm": 115.53473663330078,
"learning_rate": 7.967980295566503e-06,
"loss": 0.6837,
"step": 495
},
{
"epoch": 0.20357069567001848,
"grad_norm": 76.64769744873047,
"learning_rate": 7.963875205254516e-06,
"loss": 0.2628,
"step": 496
},
{
"epoch": 0.20398112045967576,
"grad_norm": 17.87407875061035,
"learning_rate": 7.95977011494253e-06,
"loss": 0.0735,
"step": 497
},
{
"epoch": 0.20439154524933306,
"grad_norm": 6.387015342712402,
"learning_rate": 7.955665024630542e-06,
"loss": 0.0118,
"step": 498
},
{
"epoch": 0.20480197003899037,
"grad_norm": 14.431010246276855,
"learning_rate": 7.951559934318555e-06,
"loss": 0.0197,
"step": 499
},
{
"epoch": 0.20521239482864764,
"grad_norm": 16.66281509399414,
"learning_rate": 7.947454844006569e-06,
"loss": 0.0237,
"step": 500
},
{
"epoch": 0.20562281961830495,
"grad_norm": 28.19228172302246,
"learning_rate": 7.943349753694582e-06,
"loss": 0.0873,
"step": 501
},
{
"epoch": 0.20603324440796225,
"grad_norm": 38.8900260925293,
"learning_rate": 7.939244663382595e-06,
"loss": 0.1798,
"step": 502
},
{
"epoch": 0.20644366919761953,
"grad_norm": 49.26362609863281,
"learning_rate": 7.935139573070608e-06,
"loss": 0.4037,
"step": 503
},
{
"epoch": 0.20685409398727683,
"grad_norm": 32.535423278808594,
"learning_rate": 7.93103448275862e-06,
"loss": 0.0616,
"step": 504
},
{
"epoch": 0.20726451877693414,
"grad_norm": 39.97826385498047,
"learning_rate": 7.926929392446634e-06,
"loss": 0.0979,
"step": 505
},
{
"epoch": 0.2076749435665914,
"grad_norm": 6.96060848236084,
"learning_rate": 7.922824302134647e-06,
"loss": 0.0131,
"step": 506
},
{
"epoch": 0.20808536835624872,
"grad_norm": 54.14331817626953,
"learning_rate": 7.91871921182266e-06,
"loss": 0.1836,
"step": 507
},
{
"epoch": 0.20849579314590602,
"grad_norm": 120.11117553710938,
"learning_rate": 7.914614121510675e-06,
"loss": 0.3152,
"step": 508
},
{
"epoch": 0.2089062179355633,
"grad_norm": 16.145240783691406,
"learning_rate": 7.910509031198686e-06,
"loss": 0.0255,
"step": 509
},
{
"epoch": 0.2093166427252206,
"grad_norm": 135.63153076171875,
"learning_rate": 7.9064039408867e-06,
"loss": 0.5583,
"step": 510
},
{
"epoch": 0.2097270675148779,
"grad_norm": 42.38526916503906,
"learning_rate": 7.902298850574714e-06,
"loss": 0.0941,
"step": 511
},
{
"epoch": 0.21013749230453518,
"grad_norm": 72.15447998046875,
"learning_rate": 7.898193760262727e-06,
"loss": 0.4187,
"step": 512
},
{
"epoch": 0.2105479170941925,
"grad_norm": 23.431835174560547,
"learning_rate": 7.894088669950738e-06,
"loss": 0.0486,
"step": 513
},
{
"epoch": 0.2109583418838498,
"grad_norm": 29.62925148010254,
"learning_rate": 7.889983579638753e-06,
"loss": 0.0457,
"step": 514
},
{
"epoch": 0.21136876667350707,
"grad_norm": 113.70598602294922,
"learning_rate": 7.885878489326766e-06,
"loss": 0.5449,
"step": 515
},
{
"epoch": 0.21177919146316437,
"grad_norm": 126.87225341796875,
"learning_rate": 7.88177339901478e-06,
"loss": 0.8242,
"step": 516
},
{
"epoch": 0.21218961625282168,
"grad_norm": 6.505963325500488,
"learning_rate": 7.877668308702792e-06,
"loss": 0.0135,
"step": 517
},
{
"epoch": 0.21260004104247895,
"grad_norm": 105.76856231689453,
"learning_rate": 7.873563218390805e-06,
"loss": 0.2495,
"step": 518
},
{
"epoch": 0.21301046583213626,
"grad_norm": 98.30302429199219,
"learning_rate": 7.869458128078818e-06,
"loss": 0.614,
"step": 519
},
{
"epoch": 0.21342089062179356,
"grad_norm": 34.3499755859375,
"learning_rate": 7.865353037766832e-06,
"loss": 0.1575,
"step": 520
},
{
"epoch": 0.21383131541145084,
"grad_norm": 30.03947639465332,
"learning_rate": 7.861247947454845e-06,
"loss": 0.052,
"step": 521
},
{
"epoch": 0.21424174020110814,
"grad_norm": 88.15850830078125,
"learning_rate": 7.857142857142858e-06,
"loss": 0.4331,
"step": 522
},
{
"epoch": 0.21465216499076545,
"grad_norm": 19.21493148803711,
"learning_rate": 7.85303776683087e-06,
"loss": 0.0582,
"step": 523
},
{
"epoch": 0.21506258978042272,
"grad_norm": 262.0315246582031,
"learning_rate": 7.848932676518884e-06,
"loss": 0.9939,
"step": 524
},
{
"epoch": 0.21547301457008003,
"grad_norm": 11.621757507324219,
"learning_rate": 7.844827586206897e-06,
"loss": 0.02,
"step": 525
},
{
"epoch": 0.21588343935973733,
"grad_norm": 14.428788185119629,
"learning_rate": 7.84072249589491e-06,
"loss": 0.0151,
"step": 526
},
{
"epoch": 0.2162938641493946,
"grad_norm": 73.62873077392578,
"learning_rate": 7.836617405582923e-06,
"loss": 0.2086,
"step": 527
},
{
"epoch": 0.21670428893905191,
"grad_norm": 138.35716247558594,
"learning_rate": 7.832512315270938e-06,
"loss": 0.6875,
"step": 528
},
{
"epoch": 0.21711471372870922,
"grad_norm": 183.34466552734375,
"learning_rate": 7.828407224958949e-06,
"loss": 1.5985,
"step": 529
},
{
"epoch": 0.21752513851836652,
"grad_norm": 84.54265594482422,
"learning_rate": 7.824302134646962e-06,
"loss": 0.2461,
"step": 530
},
{
"epoch": 0.2179355633080238,
"grad_norm": 67.30107879638672,
"learning_rate": 7.820197044334977e-06,
"loss": 0.1846,
"step": 531
},
{
"epoch": 0.2183459880976811,
"grad_norm": 38.21574401855469,
"learning_rate": 7.81609195402299e-06,
"loss": 0.071,
"step": 532
},
{
"epoch": 0.2187564128873384,
"grad_norm": 57.21788024902344,
"learning_rate": 7.811986863711001e-06,
"loss": 0.2034,
"step": 533
},
{
"epoch": 0.21916683767699568,
"grad_norm": 18.905309677124023,
"learning_rate": 7.807881773399016e-06,
"loss": 0.0206,
"step": 534
},
{
"epoch": 0.219577262466653,
"grad_norm": 11.245811462402344,
"learning_rate": 7.80377668308703e-06,
"loss": 0.0173,
"step": 535
},
{
"epoch": 0.2199876872563103,
"grad_norm": 43.90504837036133,
"learning_rate": 7.799671592775042e-06,
"loss": 0.108,
"step": 536
},
{
"epoch": 0.22039811204596757,
"grad_norm": 2.201463222503662,
"learning_rate": 7.795566502463054e-06,
"loss": 0.0038,
"step": 537
},
{
"epoch": 0.22080853683562487,
"grad_norm": 119.30667877197266,
"learning_rate": 7.791461412151068e-06,
"loss": 0.9964,
"step": 538
},
{
"epoch": 0.22121896162528218,
"grad_norm": 1.9806146621704102,
"learning_rate": 7.787356321839081e-06,
"loss": 0.0007,
"step": 539
},
{
"epoch": 0.22162938641493946,
"grad_norm": 21.337013244628906,
"learning_rate": 7.783251231527095e-06,
"loss": 0.0308,
"step": 540
},
{
"epoch": 0.22203981120459676,
"grad_norm": 11.702774047851562,
"learning_rate": 7.779146141215108e-06,
"loss": 0.011,
"step": 541
},
{
"epoch": 0.22245023599425406,
"grad_norm": 73.08387756347656,
"learning_rate": 7.77504105090312e-06,
"loss": 0.3843,
"step": 542
},
{
"epoch": 0.22286066078391134,
"grad_norm": 25.180776596069336,
"learning_rate": 7.770935960591134e-06,
"loss": 0.0533,
"step": 543
},
{
"epoch": 0.22327108557356865,
"grad_norm": 21.163410186767578,
"learning_rate": 7.766830870279147e-06,
"loss": 0.0316,
"step": 544
},
{
"epoch": 0.22368151036322595,
"grad_norm": 23.108901977539062,
"learning_rate": 7.76272577996716e-06,
"loss": 0.0481,
"step": 545
},
{
"epoch": 0.22409193515288323,
"grad_norm": 1.0687220096588135,
"learning_rate": 7.758620689655173e-06,
"loss": 0.0031,
"step": 546
},
{
"epoch": 0.22450235994254053,
"grad_norm": 36.70586395263672,
"learning_rate": 7.754515599343186e-06,
"loss": 0.1301,
"step": 547
},
{
"epoch": 0.22491278473219783,
"grad_norm": 11.507721900939941,
"learning_rate": 7.750410509031199e-06,
"loss": 0.0354,
"step": 548
},
{
"epoch": 0.2253232095218551,
"grad_norm": 49.49629211425781,
"learning_rate": 7.746305418719212e-06,
"loss": 0.521,
"step": 549
},
{
"epoch": 0.22573363431151242,
"grad_norm": 20.4556827545166,
"learning_rate": 7.742200328407225e-06,
"loss": 0.0753,
"step": 550
},
{
"epoch": 0.22614405910116972,
"grad_norm": 53.68857192993164,
"learning_rate": 7.738095238095238e-06,
"loss": 0.171,
"step": 551
},
{
"epoch": 0.226554483890827,
"grad_norm": 83.83357238769531,
"learning_rate": 7.733990147783253e-06,
"loss": 0.3899,
"step": 552
},
{
"epoch": 0.2269649086804843,
"grad_norm": 66.64608764648438,
"learning_rate": 7.729885057471264e-06,
"loss": 0.5232,
"step": 553
},
{
"epoch": 0.2273753334701416,
"grad_norm": 80.47610473632812,
"learning_rate": 7.725779967159277e-06,
"loss": 0.3174,
"step": 554
},
{
"epoch": 0.22778575825979888,
"grad_norm": 300.7563781738281,
"learning_rate": 7.721674876847292e-06,
"loss": 1.5869,
"step": 555
},
{
"epoch": 0.22819618304945619,
"grad_norm": 65.69540405273438,
"learning_rate": 7.717569786535304e-06,
"loss": 0.4116,
"step": 556
},
{
"epoch": 0.2286066078391135,
"grad_norm": 52.14624786376953,
"learning_rate": 7.713464696223317e-06,
"loss": 0.1665,
"step": 557
},
{
"epoch": 0.22901703262877077,
"grad_norm": 30.145658493041992,
"learning_rate": 7.709359605911331e-06,
"loss": 0.0811,
"step": 558
},
{
"epoch": 0.22942745741842807,
"grad_norm": 73.54645538330078,
"learning_rate": 7.705254515599344e-06,
"loss": 0.4346,
"step": 559
},
{
"epoch": 0.22983788220808538,
"grad_norm": 39.47355270385742,
"learning_rate": 7.701149425287356e-06,
"loss": 0.4482,
"step": 560
},
{
"epoch": 0.23024830699774265,
"grad_norm": 36.63752365112305,
"learning_rate": 7.69704433497537e-06,
"loss": 0.1175,
"step": 561
},
{
"epoch": 0.23065873178739996,
"grad_norm": 21.137189865112305,
"learning_rate": 7.692939244663384e-06,
"loss": 0.0691,
"step": 562
},
{
"epoch": 0.23106915657705726,
"grad_norm": 197.9019012451172,
"learning_rate": 7.688834154351397e-06,
"loss": 0.6442,
"step": 563
},
{
"epoch": 0.23147958136671454,
"grad_norm": 38.83695983886719,
"learning_rate": 7.68472906403941e-06,
"loss": 0.4424,
"step": 564
},
{
"epoch": 0.23189000615637184,
"grad_norm": 10.990493774414062,
"learning_rate": 7.680623973727423e-06,
"loss": 0.0228,
"step": 565
},
{
"epoch": 0.23230043094602915,
"grad_norm": 34.56005096435547,
"learning_rate": 7.676518883415436e-06,
"loss": 0.1112,
"step": 566
},
{
"epoch": 0.23271085573568642,
"grad_norm": 27.980424880981445,
"learning_rate": 7.672413793103449e-06,
"loss": 0.1287,
"step": 567
},
{
"epoch": 0.23312128052534373,
"grad_norm": 13.176775932312012,
"learning_rate": 7.668308702791462e-06,
"loss": 0.014,
"step": 568
},
{
"epoch": 0.23353170531500103,
"grad_norm": 8.472054481506348,
"learning_rate": 7.664203612479475e-06,
"loss": 0.0165,
"step": 569
},
{
"epoch": 0.2339421301046583,
"grad_norm": 121.81222534179688,
"learning_rate": 7.660098522167488e-06,
"loss": 0.498,
"step": 570
},
{
"epoch": 0.2343525548943156,
"grad_norm": 229.95872497558594,
"learning_rate": 7.655993431855501e-06,
"loss": 2.0641,
"step": 571
},
{
"epoch": 0.23476297968397292,
"grad_norm": 11.226500511169434,
"learning_rate": 7.651888341543514e-06,
"loss": 0.0136,
"step": 572
},
{
"epoch": 0.23517340447363022,
"grad_norm": 16.58848762512207,
"learning_rate": 7.647783251231527e-06,
"loss": 0.0287,
"step": 573
},
{
"epoch": 0.2355838292632875,
"grad_norm": 66.94095611572266,
"learning_rate": 7.64367816091954e-06,
"loss": 0.207,
"step": 574
},
{
"epoch": 0.2359942540529448,
"grad_norm": 23.188255310058594,
"learning_rate": 7.639573070607555e-06,
"loss": 0.039,
"step": 575
},
{
"epoch": 0.2364046788426021,
"grad_norm": 7.6452789306640625,
"learning_rate": 7.635467980295567e-06,
"loss": 0.0248,
"step": 576
},
{
"epoch": 0.23681510363225938,
"grad_norm": 79.18104553222656,
"learning_rate": 7.63136288998358e-06,
"loss": 0.5977,
"step": 577
},
{
"epoch": 0.2372255284219167,
"grad_norm": 8.282818794250488,
"learning_rate": 7.6272577996715935e-06,
"loss": 0.0076,
"step": 578
},
{
"epoch": 0.237635953211574,
"grad_norm": 51.02013397216797,
"learning_rate": 7.6231527093596066e-06,
"loss": 0.4205,
"step": 579
},
{
"epoch": 0.23804637800123127,
"grad_norm": 18.808774948120117,
"learning_rate": 7.61904761904762e-06,
"loss": 0.03,
"step": 580
},
{
"epoch": 0.23845680279088857,
"grad_norm": 464.4184265136719,
"learning_rate": 7.614942528735633e-06,
"loss": 1.1716,
"step": 581
},
{
"epoch": 0.23886722758054588,
"grad_norm": 19.74298667907715,
"learning_rate": 7.610837438423646e-06,
"loss": 0.0691,
"step": 582
},
{
"epoch": 0.23927765237020315,
"grad_norm": 63.0794677734375,
"learning_rate": 7.60673234811166e-06,
"loss": 0.1589,
"step": 583
},
{
"epoch": 0.23968807715986046,
"grad_norm": 6.858935832977295,
"learning_rate": 7.602627257799672e-06,
"loss": 0.0126,
"step": 584
},
{
"epoch": 0.24009850194951776,
"grad_norm": 7.570657253265381,
"learning_rate": 7.598522167487685e-06,
"loss": 0.0296,
"step": 585
},
{
"epoch": 0.24050892673917504,
"grad_norm": 40.57377243041992,
"learning_rate": 7.594417077175699e-06,
"loss": 0.0751,
"step": 586
},
{
"epoch": 0.24091935152883234,
"grad_norm": 8.346830368041992,
"learning_rate": 7.590311986863712e-06,
"loss": 0.0121,
"step": 587
},
{
"epoch": 0.24132977631848965,
"grad_norm": 20.54159164428711,
"learning_rate": 7.586206896551724e-06,
"loss": 0.0313,
"step": 588
},
{
"epoch": 0.24174020110814692,
"grad_norm": 12.790925979614258,
"learning_rate": 7.582101806239738e-06,
"loss": 0.0162,
"step": 589
},
{
"epoch": 0.24215062589780423,
"grad_norm": 31.540048599243164,
"learning_rate": 7.577996715927751e-06,
"loss": 0.1089,
"step": 590
},
{
"epoch": 0.24256105068746153,
"grad_norm": 41.38343811035156,
"learning_rate": 7.573891625615765e-06,
"loss": 0.5767,
"step": 591
},
{
"epoch": 0.2429714754771188,
"grad_norm": 67.2613754272461,
"learning_rate": 7.569786535303777e-06,
"loss": 0.2905,
"step": 592
},
{
"epoch": 0.2433819002667761,
"grad_norm": 34.034027099609375,
"learning_rate": 7.56568144499179e-06,
"loss": 0.0745,
"step": 593
},
{
"epoch": 0.24379232505643342,
"grad_norm": 30.42747688293457,
"learning_rate": 7.561576354679803e-06,
"loss": 0.0768,
"step": 594
},
{
"epoch": 0.2442027498460907,
"grad_norm": 74.77898406982422,
"learning_rate": 7.557471264367817e-06,
"loss": 0.3327,
"step": 595
},
{
"epoch": 0.244613174635748,
"grad_norm": 15.784136772155762,
"learning_rate": 7.5533661740558295e-06,
"loss": 0.0228,
"step": 596
},
{
"epoch": 0.2450235994254053,
"grad_norm": 85.14934539794922,
"learning_rate": 7.5492610837438426e-06,
"loss": 0.24,
"step": 597
},
{
"epoch": 0.24543402421506258,
"grad_norm": 370.8543395996094,
"learning_rate": 7.5451559934318565e-06,
"loss": 1.5907,
"step": 598
},
{
"epoch": 0.24584444900471988,
"grad_norm": 83.6479721069336,
"learning_rate": 7.5410509031198695e-06,
"loss": 0.3147,
"step": 599
},
{
"epoch": 0.2462548737943772,
"grad_norm": 10.754281997680664,
"learning_rate": 7.536945812807882e-06,
"loss": 0.0172,
"step": 600
},
{
"epoch": 0.24666529858403446,
"grad_norm": 9.650517463684082,
"learning_rate": 7.532840722495896e-06,
"loss": 0.02,
"step": 601
},
{
"epoch": 0.24707572337369177,
"grad_norm": 61.178829193115234,
"learning_rate": 7.528735632183909e-06,
"loss": 0.5327,
"step": 602
},
{
"epoch": 0.24748614816334907,
"grad_norm": 7.723118305206299,
"learning_rate": 7.524630541871923e-06,
"loss": 0.014,
"step": 603
},
{
"epoch": 0.24789657295300635,
"grad_norm": 91.85173797607422,
"learning_rate": 7.520525451559935e-06,
"loss": 0.2372,
"step": 604
},
{
"epoch": 0.24830699774266365,
"grad_norm": 22.273181915283203,
"learning_rate": 7.516420361247948e-06,
"loss": 0.1672,
"step": 605
},
{
"epoch": 0.24871742253232096,
"grad_norm": 55.78474044799805,
"learning_rate": 7.512315270935962e-06,
"loss": 0.3253,
"step": 606
},
{
"epoch": 0.24912784732197824,
"grad_norm": 36.62397766113281,
"learning_rate": 7.508210180623975e-06,
"loss": 0.2165,
"step": 607
},
{
"epoch": 0.24953827211163554,
"grad_norm": 24.468923568725586,
"learning_rate": 7.504105090311987e-06,
"loss": 0.0444,
"step": 608
},
{
"epoch": 0.24994869690129284,
"grad_norm": 50.5117301940918,
"learning_rate": 7.500000000000001e-06,
"loss": 0.2144,
"step": 609
},
{
"epoch": 0.25035912169095015,
"grad_norm": 42.80532455444336,
"learning_rate": 7.495894909688014e-06,
"loss": 0.2501,
"step": 610
},
{
"epoch": 0.25076954648060745,
"grad_norm": 45.99153518676758,
"learning_rate": 7.491789819376026e-06,
"loss": 0.1654,
"step": 611
},
{
"epoch": 0.2511799712702647,
"grad_norm": 99.59512329101562,
"learning_rate": 7.487684729064039e-06,
"loss": 0.3272,
"step": 612
},
{
"epoch": 0.251590396059922,
"grad_norm": 129.72328186035156,
"learning_rate": 7.483579638752053e-06,
"loss": 0.4301,
"step": 613
},
{
"epoch": 0.2520008208495793,
"grad_norm": 17.018550872802734,
"learning_rate": 7.479474548440066e-06,
"loss": 0.039,
"step": 614
},
{
"epoch": 0.2524112456392366,
"grad_norm": 34.03882598876953,
"learning_rate": 7.4753694581280786e-06,
"loss": 0.0737,
"step": 615
},
{
"epoch": 0.2528216704288939,
"grad_norm": 30.565231323242188,
"learning_rate": 7.4712643678160925e-06,
"loss": 0.0693,
"step": 616
},
{
"epoch": 0.2532320952185512,
"grad_norm": 40.32196807861328,
"learning_rate": 7.4671592775041056e-06,
"loss": 0.1562,
"step": 617
},
{
"epoch": 0.25364252000820847,
"grad_norm": 218.28121948242188,
"learning_rate": 7.4630541871921195e-06,
"loss": 0.6901,
"step": 618
},
{
"epoch": 0.2540529447978658,
"grad_norm": 22.54781723022461,
"learning_rate": 7.458949096880132e-06,
"loss": 0.0425,
"step": 619
},
{
"epoch": 0.2544633695875231,
"grad_norm": 5.83807373046875,
"learning_rate": 7.454844006568145e-06,
"loss": 0.0059,
"step": 620
},
{
"epoch": 0.2548737943771804,
"grad_norm": 43.14492416381836,
"learning_rate": 7.450738916256159e-06,
"loss": 0.1813,
"step": 621
},
{
"epoch": 0.2552842191668377,
"grad_norm": 8.126626014709473,
"learning_rate": 7.446633825944172e-06,
"loss": 0.0164,
"step": 622
},
{
"epoch": 0.255694643956495,
"grad_norm": 13.800533294677734,
"learning_rate": 7.442528735632184e-06,
"loss": 0.0216,
"step": 623
},
{
"epoch": 0.25610506874615224,
"grad_norm": 17.93963050842285,
"learning_rate": 7.438423645320198e-06,
"loss": 0.0344,
"step": 624
},
{
"epoch": 0.25651549353580955,
"grad_norm": 46.304351806640625,
"learning_rate": 7.434318555008211e-06,
"loss": 0.1477,
"step": 625
},
{
"epoch": 0.25692591832546685,
"grad_norm": 92.92662048339844,
"learning_rate": 7.430213464696224e-06,
"loss": 0.7949,
"step": 626
},
{
"epoch": 0.25733634311512416,
"grad_norm": 28.070598602294922,
"learning_rate": 7.426108374384236e-06,
"loss": 0.0262,
"step": 627
},
{
"epoch": 0.25774676790478146,
"grad_norm": 74.28411865234375,
"learning_rate": 7.42200328407225e-06,
"loss": 0.629,
"step": 628
},
{
"epoch": 0.25815719269443876,
"grad_norm": 25.93304443359375,
"learning_rate": 7.417898193760263e-06,
"loss": 0.0489,
"step": 629
},
{
"epoch": 0.258567617484096,
"grad_norm": 337.5581970214844,
"learning_rate": 7.413793103448277e-06,
"loss": 0.7561,
"step": 630
},
{
"epoch": 0.2589780422737533,
"grad_norm": 48.15328598022461,
"learning_rate": 7.409688013136289e-06,
"loss": 0.4248,
"step": 631
},
{
"epoch": 0.2593884670634106,
"grad_norm": 110.51360321044922,
"learning_rate": 7.405582922824302e-06,
"loss": 0.239,
"step": 632
},
{
"epoch": 0.2597988918530679,
"grad_norm": 0.8408578634262085,
"learning_rate": 7.401477832512316e-06,
"loss": 0.0033,
"step": 633
},
{
"epoch": 0.26020931664272523,
"grad_norm": 16.083471298217773,
"learning_rate": 7.397372742200329e-06,
"loss": 0.019,
"step": 634
},
{
"epoch": 0.26061974143238253,
"grad_norm": 51.709442138671875,
"learning_rate": 7.3932676518883416e-06,
"loss": 0.0776,
"step": 635
},
{
"epoch": 0.26103016622203984,
"grad_norm": 11.790945053100586,
"learning_rate": 7.3891625615763555e-06,
"loss": 0.0223,
"step": 636
},
{
"epoch": 0.2614405910116971,
"grad_norm": 20.131996154785156,
"learning_rate": 7.3850574712643685e-06,
"loss": 0.0435,
"step": 637
},
{
"epoch": 0.2618510158013544,
"grad_norm": 8.276931762695312,
"learning_rate": 7.380952380952382e-06,
"loss": 0.0049,
"step": 638
},
{
"epoch": 0.2622614405910117,
"grad_norm": 69.14643096923828,
"learning_rate": 7.376847290640395e-06,
"loss": 0.4009,
"step": 639
},
{
"epoch": 0.262671865380669,
"grad_norm": 5.661315441131592,
"learning_rate": 7.372742200328408e-06,
"loss": 0.008,
"step": 640
},
{
"epoch": 0.2630822901703263,
"grad_norm": 4.361414909362793,
"learning_rate": 7.368637110016421e-06,
"loss": 0.0076,
"step": 641
},
{
"epoch": 0.2634927149599836,
"grad_norm": 95.10589599609375,
"learning_rate": 7.364532019704435e-06,
"loss": 0.1452,
"step": 642
},
{
"epoch": 0.26390313974964086,
"grad_norm": 312.26361083984375,
"learning_rate": 7.360426929392447e-06,
"loss": 1.7897,
"step": 643
},
{
"epoch": 0.26431356453929816,
"grad_norm": 18.183441162109375,
"learning_rate": 7.35632183908046e-06,
"loss": 0.0811,
"step": 644
},
{
"epoch": 0.26472398932895547,
"grad_norm": 8.144716262817383,
"learning_rate": 7.352216748768474e-06,
"loss": 0.0112,
"step": 645
},
{
"epoch": 0.26513441411861277,
"grad_norm": 8.615951538085938,
"learning_rate": 7.348111658456487e-06,
"loss": 0.0083,
"step": 646
},
{
"epoch": 0.2655448389082701,
"grad_norm": 33.38519287109375,
"learning_rate": 7.344006568144499e-06,
"loss": 0.1028,
"step": 647
},
{
"epoch": 0.2659552636979274,
"grad_norm": 52.58414840698242,
"learning_rate": 7.339901477832513e-06,
"loss": 0.1135,
"step": 648
},
{
"epoch": 0.26636568848758463,
"grad_norm": 1.827383279800415,
"learning_rate": 7.335796387520526e-06,
"loss": 0.0031,
"step": 649
},
{
"epoch": 0.26677611327724193,
"grad_norm": 19.296640396118164,
"learning_rate": 7.33169129720854e-06,
"loss": 0.0244,
"step": 650
},
{
"epoch": 0.26718653806689924,
"grad_norm": 67.79634094238281,
"learning_rate": 7.327586206896552e-06,
"loss": 0.1373,
"step": 651
},
{
"epoch": 0.26759696285655654,
"grad_norm": 50.67517852783203,
"learning_rate": 7.323481116584565e-06,
"loss": 0.0814,
"step": 652
},
{
"epoch": 0.26800738764621385,
"grad_norm": 15.43031120300293,
"learning_rate": 7.319376026272578e-06,
"loss": 0.0186,
"step": 653
},
{
"epoch": 0.26841781243587115,
"grad_norm": 140.4246826171875,
"learning_rate": 7.315270935960592e-06,
"loss": 0.8984,
"step": 654
},
{
"epoch": 0.2688282372255284,
"grad_norm": 16.3335018157959,
"learning_rate": 7.3111658456486045e-06,
"loss": 0.0193,
"step": 655
},
{
"epoch": 0.2692386620151857,
"grad_norm": 24.356098175048828,
"learning_rate": 7.307060755336618e-06,
"loss": 0.057,
"step": 656
},
{
"epoch": 0.269649086804843,
"grad_norm": 89.5280532836914,
"learning_rate": 7.3029556650246315e-06,
"loss": 0.1858,
"step": 657
},
{
"epoch": 0.2700595115945003,
"grad_norm": 6.361180782318115,
"learning_rate": 7.298850574712645e-06,
"loss": 0.004,
"step": 658
},
{
"epoch": 0.2704699363841576,
"grad_norm": 34.67497253417969,
"learning_rate": 7.294745484400657e-06,
"loss": 0.07,
"step": 659
},
{
"epoch": 0.2708803611738149,
"grad_norm": 87.53612518310547,
"learning_rate": 7.290640394088671e-06,
"loss": 0.3503,
"step": 660
},
{
"epoch": 0.27129078596347217,
"grad_norm": 21.897172927856445,
"learning_rate": 7.286535303776684e-06,
"loss": 0.0351,
"step": 661
},
{
"epoch": 0.2717012107531295,
"grad_norm": 74.89786529541016,
"learning_rate": 7.282430213464698e-06,
"loss": 0.6015,
"step": 662
},
{
"epoch": 0.2721116355427868,
"grad_norm": 23.7749080657959,
"learning_rate": 7.27832512315271e-06,
"loss": 0.0396,
"step": 663
},
{
"epoch": 0.2725220603324441,
"grad_norm": 3.7347161769866943,
"learning_rate": 7.274220032840723e-06,
"loss": 0.0088,
"step": 664
},
{
"epoch": 0.2729324851221014,
"grad_norm": 29.278732299804688,
"learning_rate": 7.270114942528737e-06,
"loss": 0.048,
"step": 665
},
{
"epoch": 0.2733429099117587,
"grad_norm": 14.297079086303711,
"learning_rate": 7.266009852216749e-06,
"loss": 0.011,
"step": 666
},
{
"epoch": 0.27375333470141594,
"grad_norm": 2.6223113536834717,
"learning_rate": 7.261904761904762e-06,
"loss": 0.0017,
"step": 667
},
{
"epoch": 0.27416375949107324,
"grad_norm": 4.069161891937256,
"learning_rate": 7.257799671592776e-06,
"loss": 0.0017,
"step": 668
},
{
"epoch": 0.27457418428073055,
"grad_norm": 44.544002532958984,
"learning_rate": 7.253694581280789e-06,
"loss": 0.0499,
"step": 669
},
{
"epoch": 0.27498460907038785,
"grad_norm": 10.024898529052734,
"learning_rate": 7.249589490968801e-06,
"loss": 0.0068,
"step": 670
},
{
"epoch": 0.27539503386004516,
"grad_norm": 68.68046569824219,
"learning_rate": 7.245484400656814e-06,
"loss": 0.0733,
"step": 671
},
{
"epoch": 0.27580545864970246,
"grad_norm": 55.8798942565918,
"learning_rate": 7.241379310344828e-06,
"loss": 0.0886,
"step": 672
},
{
"epoch": 0.2762158834393597,
"grad_norm": 30.416576385498047,
"learning_rate": 7.237274220032841e-06,
"loss": 0.0428,
"step": 673
},
{
"epoch": 0.276626308229017,
"grad_norm": 31.328144073486328,
"learning_rate": 7.233169129720854e-06,
"loss": 0.0627,
"step": 674
},
{
"epoch": 0.2770367330186743,
"grad_norm": 72.36646270751953,
"learning_rate": 7.2290640394088675e-06,
"loss": 0.4165,
"step": 675
},
{
"epoch": 0.2774471578083316,
"grad_norm": 12.475491523742676,
"learning_rate": 7.224958949096881e-06,
"loss": 0.0139,
"step": 676
},
{
"epoch": 0.27785758259798893,
"grad_norm": 7.3289079666137695,
"learning_rate": 7.2208538587848945e-06,
"loss": 0.0152,
"step": 677
},
{
"epoch": 0.27826800738764623,
"grad_norm": 17.488494873046875,
"learning_rate": 7.216748768472907e-06,
"loss": 0.0222,
"step": 678
},
{
"epoch": 0.27867843217730354,
"grad_norm": 109.88444519042969,
"learning_rate": 7.21264367816092e-06,
"loss": 0.2104,
"step": 679
},
{
"epoch": 0.2790888569669608,
"grad_norm": 22.866737365722656,
"learning_rate": 7.208538587848934e-06,
"loss": 0.0369,
"step": 680
},
{
"epoch": 0.2794992817566181,
"grad_norm": 49.05271911621094,
"learning_rate": 7.204433497536947e-06,
"loss": 0.1141,
"step": 681
},
{
"epoch": 0.2799097065462754,
"grad_norm": 36.43170928955078,
"learning_rate": 7.200328407224959e-06,
"loss": 0.0443,
"step": 682
},
{
"epoch": 0.2803201313359327,
"grad_norm": 406.8815612792969,
"learning_rate": 7.196223316912973e-06,
"loss": 2.7109,
"step": 683
},
{
"epoch": 0.28073055612559,
"grad_norm": 6.176809787750244,
"learning_rate": 7.192118226600986e-06,
"loss": 0.0101,
"step": 684
},
{
"epoch": 0.2811409809152473,
"grad_norm": 18.120920181274414,
"learning_rate": 7.188013136288999e-06,
"loss": 0.0433,
"step": 685
},
{
"epoch": 0.28155140570490456,
"grad_norm": 27.748693466186523,
"learning_rate": 7.183908045977011e-06,
"loss": 0.0546,
"step": 686
},
{
"epoch": 0.28196183049456186,
"grad_norm": 24.181371688842773,
"learning_rate": 7.179802955665025e-06,
"loss": 0.0363,
"step": 687
},
{
"epoch": 0.28237225528421916,
"grad_norm": 20.504302978515625,
"learning_rate": 7.175697865353038e-06,
"loss": 0.0302,
"step": 688
},
{
"epoch": 0.28278268007387647,
"grad_norm": 2.6363723278045654,
"learning_rate": 7.171592775041052e-06,
"loss": 0.0044,
"step": 689
},
{
"epoch": 0.2831931048635338,
"grad_norm": 4.926974773406982,
"learning_rate": 7.167487684729064e-06,
"loss": 0.004,
"step": 690
},
{
"epoch": 0.2836035296531911,
"grad_norm": 15.769619941711426,
"learning_rate": 7.163382594417077e-06,
"loss": 0.0221,
"step": 691
},
{
"epoch": 0.2840139544428483,
"grad_norm": 8.697476387023926,
"learning_rate": 7.159277504105091e-06,
"loss": 0.0063,
"step": 692
},
{
"epoch": 0.28442437923250563,
"grad_norm": 8.268402099609375,
"learning_rate": 7.155172413793104e-06,
"loss": 0.0078,
"step": 693
},
{
"epoch": 0.28483480402216294,
"grad_norm": 18.324951171875,
"learning_rate": 7.151067323481117e-06,
"loss": 0.0471,
"step": 694
},
{
"epoch": 0.28524522881182024,
"grad_norm": 2.5097670555114746,
"learning_rate": 7.1469622331691305e-06,
"loss": 0.0075,
"step": 695
},
{
"epoch": 0.28565565360147754,
"grad_norm": 21.002246856689453,
"learning_rate": 7.1428571428571436e-06,
"loss": 0.0309,
"step": 696
},
{
"epoch": 0.28606607839113485,
"grad_norm": 63.235591888427734,
"learning_rate": 7.138752052545157e-06,
"loss": 0.3777,
"step": 697
},
{
"epoch": 0.2864765031807921,
"grad_norm": 20.791799545288086,
"learning_rate": 7.13464696223317e-06,
"loss": 0.0289,
"step": 698
},
{
"epoch": 0.2868869279704494,
"grad_norm": 86.95537567138672,
"learning_rate": 7.130541871921183e-06,
"loss": 0.376,
"step": 699
},
{
"epoch": 0.2872973527601067,
"grad_norm": 5.661103248596191,
"learning_rate": 7.126436781609196e-06,
"loss": 0.0111,
"step": 700
},
{
"epoch": 0.287707777549764,
"grad_norm": 174.14515686035156,
"learning_rate": 7.12233169129721e-06,
"loss": 0.2041,
"step": 701
},
{
"epoch": 0.2881182023394213,
"grad_norm": 21.69768524169922,
"learning_rate": 7.118226600985222e-06,
"loss": 0.0316,
"step": 702
},
{
"epoch": 0.2885286271290786,
"grad_norm": 30.732656478881836,
"learning_rate": 7.114121510673235e-06,
"loss": 0.0563,
"step": 703
},
{
"epoch": 0.28893905191873587,
"grad_norm": 29.325645446777344,
"learning_rate": 7.110016420361249e-06,
"loss": 0.0616,
"step": 704
},
{
"epoch": 0.28934947670839317,
"grad_norm": 45.404701232910156,
"learning_rate": 7.105911330049262e-06,
"loss": 0.1392,
"step": 705
},
{
"epoch": 0.2897599014980505,
"grad_norm": 11.00036334991455,
"learning_rate": 7.101806239737274e-06,
"loss": 0.0199,
"step": 706
},
{
"epoch": 0.2901703262877078,
"grad_norm": 12.154747009277344,
"learning_rate": 7.097701149425288e-06,
"loss": 0.0087,
"step": 707
},
{
"epoch": 0.2905807510773651,
"grad_norm": 6.20178747177124,
"learning_rate": 7.093596059113301e-06,
"loss": 0.0067,
"step": 708
},
{
"epoch": 0.2909911758670224,
"grad_norm": 21.10610008239746,
"learning_rate": 7.089490968801315e-06,
"loss": 0.0328,
"step": 709
},
{
"epoch": 0.29140160065667964,
"grad_norm": 8.64224624633789,
"learning_rate": 7.085385878489327e-06,
"loss": 0.0061,
"step": 710
},
{
"epoch": 0.29181202544633694,
"grad_norm": 3.781222105026245,
"learning_rate": 7.08128078817734e-06,
"loss": 0.0088,
"step": 711
},
{
"epoch": 0.29222245023599425,
"grad_norm": 63.63623809814453,
"learning_rate": 7.0771756978653534e-06,
"loss": 0.1005,
"step": 712
},
{
"epoch": 0.29263287502565155,
"grad_norm": 100.4776611328125,
"learning_rate": 7.073070607553367e-06,
"loss": 0.8484,
"step": 713
},
{
"epoch": 0.29304329981530886,
"grad_norm": 214.6756591796875,
"learning_rate": 7.0689655172413796e-06,
"loss": 1.1497,
"step": 714
},
{
"epoch": 0.29345372460496616,
"grad_norm": 14.875383377075195,
"learning_rate": 7.064860426929393e-06,
"loss": 0.0264,
"step": 715
},
{
"epoch": 0.2938641493946234,
"grad_norm": 95.34712219238281,
"learning_rate": 7.0607553366174065e-06,
"loss": 0.5919,
"step": 716
},
{
"epoch": 0.2942745741842807,
"grad_norm": 67.41881561279297,
"learning_rate": 7.05665024630542e-06,
"loss": 0.1469,
"step": 717
},
{
"epoch": 0.294684998973938,
"grad_norm": 3.2724251747131348,
"learning_rate": 7.052545155993432e-06,
"loss": 0.0046,
"step": 718
},
{
"epoch": 0.2950954237635953,
"grad_norm": 23.133176803588867,
"learning_rate": 7.048440065681446e-06,
"loss": 0.0442,
"step": 719
},
{
"epoch": 0.2955058485532526,
"grad_norm": 56.77151870727539,
"learning_rate": 7.044334975369459e-06,
"loss": 0.4575,
"step": 720
},
{
"epoch": 0.29591627334290993,
"grad_norm": 5.968919277191162,
"learning_rate": 7.040229885057471e-06,
"loss": 0.0073,
"step": 721
},
{
"epoch": 0.29632669813256723,
"grad_norm": 1.766608476638794,
"learning_rate": 7.036124794745485e-06,
"loss": 0.0006,
"step": 722
},
{
"epoch": 0.2967371229222245,
"grad_norm": 13.383498191833496,
"learning_rate": 7.032019704433498e-06,
"loss": 0.0146,
"step": 723
},
{
"epoch": 0.2971475477118818,
"grad_norm": 12.853846549987793,
"learning_rate": 7.027914614121512e-06,
"loss": 0.0178,
"step": 724
},
{
"epoch": 0.2975579725015391,
"grad_norm": 71.08599853515625,
"learning_rate": 7.023809523809524e-06,
"loss": 0.1175,
"step": 725
},
{
"epoch": 0.2979683972911964,
"grad_norm": 34.29999923706055,
"learning_rate": 7.019704433497537e-06,
"loss": 0.061,
"step": 726
},
{
"epoch": 0.2983788220808537,
"grad_norm": 105.27954864501953,
"learning_rate": 7.015599343185551e-06,
"loss": 0.7615,
"step": 727
},
{
"epoch": 0.298789246870511,
"grad_norm": 20.089290618896484,
"learning_rate": 7.011494252873564e-06,
"loss": 0.0646,
"step": 728
},
{
"epoch": 0.29919967166016825,
"grad_norm": 4.953437328338623,
"learning_rate": 7.007389162561576e-06,
"loss": 0.0083,
"step": 729
},
{
"epoch": 0.29961009644982556,
"grad_norm": 417.3952331542969,
"learning_rate": 7.0032840722495894e-06,
"loss": 1.8447,
"step": 730
},
{
"epoch": 0.30002052123948286,
"grad_norm": 46.873146057128906,
"learning_rate": 6.999178981937603e-06,
"loss": 0.5369,
"step": 731
},
{
"epoch": 0.30043094602914017,
"grad_norm": 32.62269973754883,
"learning_rate": 6.995073891625616e-06,
"loss": 0.0359,
"step": 732
},
{
"epoch": 0.30084137081879747,
"grad_norm": 20.26422882080078,
"learning_rate": 6.990968801313629e-06,
"loss": 0.031,
"step": 733
},
{
"epoch": 0.3012517956084548,
"grad_norm": 13.190338134765625,
"learning_rate": 6.9868637110016425e-06,
"loss": 0.0135,
"step": 734
},
{
"epoch": 0.301662220398112,
"grad_norm": 42.02793502807617,
"learning_rate": 6.982758620689656e-06,
"loss": 0.039,
"step": 735
},
{
"epoch": 0.30207264518776933,
"grad_norm": 49.855140686035156,
"learning_rate": 6.9786535303776695e-06,
"loss": 0.486,
"step": 736
},
{
"epoch": 0.30248306997742663,
"grad_norm": 20.121145248413086,
"learning_rate": 6.974548440065682e-06,
"loss": 0.0454,
"step": 737
},
{
"epoch": 0.30289349476708394,
"grad_norm": 7.158181667327881,
"learning_rate": 6.970443349753695e-06,
"loss": 0.0131,
"step": 738
},
{
"epoch": 0.30330391955674124,
"grad_norm": 64.26904296875,
"learning_rate": 6.966338259441709e-06,
"loss": 0.0676,
"step": 739
},
{
"epoch": 0.30371434434639855,
"grad_norm": 62.31405258178711,
"learning_rate": 6.962233169129722e-06,
"loss": 0.1479,
"step": 740
},
{
"epoch": 0.3041247691360558,
"grad_norm": 23.32426643371582,
"learning_rate": 6.958128078817734e-06,
"loss": 0.0375,
"step": 741
},
{
"epoch": 0.3045351939257131,
"grad_norm": 4.091766357421875,
"learning_rate": 6.954022988505748e-06,
"loss": 0.008,
"step": 742
},
{
"epoch": 0.3049456187153704,
"grad_norm": 10.512306213378906,
"learning_rate": 6.949917898193761e-06,
"loss": 0.01,
"step": 743
},
{
"epoch": 0.3053560435050277,
"grad_norm": 9.996668815612793,
"learning_rate": 6.945812807881774e-06,
"loss": 0.0251,
"step": 744
},
{
"epoch": 0.305766468294685,
"grad_norm": 17.175065994262695,
"learning_rate": 6.941707717569786e-06,
"loss": 0.0194,
"step": 745
},
{
"epoch": 0.3061768930843423,
"grad_norm": 0.848054826259613,
"learning_rate": 6.9376026272578e-06,
"loss": 0.0013,
"step": 746
},
{
"epoch": 0.30658731787399957,
"grad_norm": 45.41935729980469,
"learning_rate": 6.933497536945813e-06,
"loss": 0.2753,
"step": 747
},
{
"epoch": 0.30699774266365687,
"grad_norm": 17.148618698120117,
"learning_rate": 6.929392446633827e-06,
"loss": 0.0253,
"step": 748
},
{
"epoch": 0.3074081674533142,
"grad_norm": 1.6488438844680786,
"learning_rate": 6.925287356321839e-06,
"loss": 0.0031,
"step": 749
},
{
"epoch": 0.3078185922429715,
"grad_norm": 14.969161987304688,
"learning_rate": 6.9211822660098524e-06,
"loss": 0.0171,
"step": 750
},
{
"epoch": 0.3082290170326288,
"grad_norm": 48.02900314331055,
"learning_rate": 6.917077175697866e-06,
"loss": 0.4184,
"step": 751
},
{
"epoch": 0.3086394418222861,
"grad_norm": 19.5977840423584,
"learning_rate": 6.912972085385879e-06,
"loss": 0.0275,
"step": 752
},
{
"epoch": 0.30904986661194334,
"grad_norm": 94.47354125976562,
"learning_rate": 6.908866995073892e-06,
"loss": 0.122,
"step": 753
},
{
"epoch": 0.30946029140160064,
"grad_norm": 6.306624889373779,
"learning_rate": 6.9047619047619055e-06,
"loss": 0.0084,
"step": 754
},
{
"epoch": 0.30987071619125794,
"grad_norm": 7.978044509887695,
"learning_rate": 6.900656814449919e-06,
"loss": 0.0131,
"step": 755
},
{
"epoch": 0.31028114098091525,
"grad_norm": 19.627099990844727,
"learning_rate": 6.896551724137932e-06,
"loss": 0.0315,
"step": 756
},
{
"epoch": 0.31069156577057255,
"grad_norm": 28.8678035736084,
"learning_rate": 6.892446633825945e-06,
"loss": 0.0452,
"step": 757
},
{
"epoch": 0.31110199056022986,
"grad_norm": 40.371551513671875,
"learning_rate": 6.888341543513958e-06,
"loss": 0.0486,
"step": 758
},
{
"epoch": 0.3115124153498871,
"grad_norm": 5.470851421356201,
"learning_rate": 6.884236453201971e-06,
"loss": 0.0044,
"step": 759
},
{
"epoch": 0.3119228401395444,
"grad_norm": 12.40237045288086,
"learning_rate": 6.880131362889985e-06,
"loss": 0.0186,
"step": 760
},
{
"epoch": 0.3123332649292017,
"grad_norm": 35.31753921508789,
"learning_rate": 6.876026272577997e-06,
"loss": 0.0684,
"step": 761
},
{
"epoch": 0.312743689718859,
"grad_norm": 5.180874824523926,
"learning_rate": 6.87192118226601e-06,
"loss": 0.0053,
"step": 762
},
{
"epoch": 0.3131541145085163,
"grad_norm": 99.68106079101562,
"learning_rate": 6.867816091954024e-06,
"loss": 0.5298,
"step": 763
},
{
"epoch": 0.31356453929817363,
"grad_norm": 12.443490028381348,
"learning_rate": 6.863711001642037e-06,
"loss": 0.0074,
"step": 764
},
{
"epoch": 0.31397496408783093,
"grad_norm": 10.037583351135254,
"learning_rate": 6.859605911330049e-06,
"loss": 0.0162,
"step": 765
},
{
"epoch": 0.3143853888774882,
"grad_norm": 9.569877624511719,
"learning_rate": 6.855500821018063e-06,
"loss": 0.0188,
"step": 766
},
{
"epoch": 0.3147958136671455,
"grad_norm": 8.829991340637207,
"learning_rate": 6.851395730706076e-06,
"loss": 0.0085,
"step": 767
},
{
"epoch": 0.3152062384568028,
"grad_norm": 8.315031051635742,
"learning_rate": 6.84729064039409e-06,
"loss": 0.0082,
"step": 768
},
{
"epoch": 0.3156166632464601,
"grad_norm": 29.638437271118164,
"learning_rate": 6.843185550082102e-06,
"loss": 0.0543,
"step": 769
},
{
"epoch": 0.3160270880361174,
"grad_norm": 50.44863510131836,
"learning_rate": 6.839080459770115e-06,
"loss": 0.0668,
"step": 770
},
{
"epoch": 0.3164375128257747,
"grad_norm": 39.62846374511719,
"learning_rate": 6.834975369458129e-06,
"loss": 0.1075,
"step": 771
},
{
"epoch": 0.31684793761543195,
"grad_norm": 33.495121002197266,
"learning_rate": 6.830870279146142e-06,
"loss": 0.0592,
"step": 772
},
{
"epoch": 0.31725836240508926,
"grad_norm": 14.802083015441895,
"learning_rate": 6.826765188834155e-06,
"loss": 0.0252,
"step": 773
},
{
"epoch": 0.31766878719474656,
"grad_norm": 8.756896018981934,
"learning_rate": 6.822660098522168e-06,
"loss": 0.0077,
"step": 774
},
{
"epoch": 0.31807921198440386,
"grad_norm": 7.4073381423950195,
"learning_rate": 6.8185550082101816e-06,
"loss": 0.0044,
"step": 775
},
{
"epoch": 0.31848963677406117,
"grad_norm": 4.850039958953857,
"learning_rate": 6.814449917898194e-06,
"loss": 0.0098,
"step": 776
},
{
"epoch": 0.3189000615637185,
"grad_norm": 17.106969833374023,
"learning_rate": 6.810344827586207e-06,
"loss": 0.0212,
"step": 777
},
{
"epoch": 0.3193104863533757,
"grad_norm": 59.88249588012695,
"learning_rate": 6.806239737274221e-06,
"loss": 0.0569,
"step": 778
},
{
"epoch": 0.319720911143033,
"grad_norm": 47.606712341308594,
"learning_rate": 6.802134646962234e-06,
"loss": 0.0925,
"step": 779
},
{
"epoch": 0.32013133593269033,
"grad_norm": 7.229979515075684,
"learning_rate": 6.798029556650246e-06,
"loss": 0.0113,
"step": 780
},
{
"epoch": 0.32054176072234764,
"grad_norm": 16.2637996673584,
"learning_rate": 6.79392446633826e-06,
"loss": 0.015,
"step": 781
},
{
"epoch": 0.32095218551200494,
"grad_norm": 12.706461906433105,
"learning_rate": 6.789819376026273e-06,
"loss": 0.0117,
"step": 782
},
{
"epoch": 0.32136261030166224,
"grad_norm": 9.572243690490723,
"learning_rate": 6.785714285714287e-06,
"loss": 0.0147,
"step": 783
},
{
"epoch": 0.3217730350913195,
"grad_norm": 78.81916046142578,
"learning_rate": 6.781609195402299e-06,
"loss": 0.0938,
"step": 784
},
{
"epoch": 0.3221834598809768,
"grad_norm": 53.23609161376953,
"learning_rate": 6.777504105090312e-06,
"loss": 0.0625,
"step": 785
},
{
"epoch": 0.3225938846706341,
"grad_norm": 30.08687400817871,
"learning_rate": 6.773399014778326e-06,
"loss": 0.0469,
"step": 786
},
{
"epoch": 0.3230043094602914,
"grad_norm": 3.9887423515319824,
"learning_rate": 6.769293924466339e-06,
"loss": 0.0018,
"step": 787
},
{
"epoch": 0.3234147342499487,
"grad_norm": 106.1269760131836,
"learning_rate": 6.765188834154351e-06,
"loss": 0.6608,
"step": 788
},
{
"epoch": 0.323825159039606,
"grad_norm": 10.100157737731934,
"learning_rate": 6.7610837438423645e-06,
"loss": 0.0306,
"step": 789
},
{
"epoch": 0.32423558382926326,
"grad_norm": 4.360620975494385,
"learning_rate": 6.756978653530378e-06,
"loss": 0.0027,
"step": 790
},
{
"epoch": 0.32464600861892057,
"grad_norm": 11.223150253295898,
"learning_rate": 6.7528735632183914e-06,
"loss": 0.0092,
"step": 791
},
{
"epoch": 0.32505643340857787,
"grad_norm": 9.799832344055176,
"learning_rate": 6.748768472906404e-06,
"loss": 0.0117,
"step": 792
},
{
"epoch": 0.3254668581982352,
"grad_norm": 1.980411410331726,
"learning_rate": 6.744663382594418e-06,
"loss": 0.0013,
"step": 793
},
{
"epoch": 0.3258772829878925,
"grad_norm": 13.49509334564209,
"learning_rate": 6.740558292282431e-06,
"loss": 0.0117,
"step": 794
},
{
"epoch": 0.3262877077775498,
"grad_norm": 28.230804443359375,
"learning_rate": 6.7364532019704446e-06,
"loss": 0.051,
"step": 795
},
{
"epoch": 0.32669813256720703,
"grad_norm": 18.88104248046875,
"learning_rate": 6.732348111658457e-06,
"loss": 0.0224,
"step": 796
},
{
"epoch": 0.32710855735686434,
"grad_norm": 47.7376823425293,
"learning_rate": 6.72824302134647e-06,
"loss": 0.265,
"step": 797
},
{
"epoch": 0.32751898214652164,
"grad_norm": 21.048124313354492,
"learning_rate": 6.724137931034484e-06,
"loss": 0.0405,
"step": 798
},
{
"epoch": 0.32792940693617895,
"grad_norm": 58.21773147583008,
"learning_rate": 6.720032840722497e-06,
"loss": 0.4154,
"step": 799
},
{
"epoch": 0.32833983172583625,
"grad_norm": 54.40596008300781,
"learning_rate": 6.715927750410509e-06,
"loss": 0.1392,
"step": 800
},
{
"epoch": 0.32875025651549356,
"grad_norm": 13.651693344116211,
"learning_rate": 6.711822660098523e-06,
"loss": 0.0154,
"step": 801
},
{
"epoch": 0.3291606813051508,
"grad_norm": 16.01358985900879,
"learning_rate": 6.707717569786536e-06,
"loss": 0.0223,
"step": 802
},
{
"epoch": 0.3295711060948081,
"grad_norm": 21.4189395904541,
"learning_rate": 6.703612479474549e-06,
"loss": 0.024,
"step": 803
},
{
"epoch": 0.3299815308844654,
"grad_norm": 18.4512939453125,
"learning_rate": 6.699507389162562e-06,
"loss": 0.021,
"step": 804
},
{
"epoch": 0.3303919556741227,
"grad_norm": 13.236135482788086,
"learning_rate": 6.695402298850575e-06,
"loss": 0.0135,
"step": 805
},
{
"epoch": 0.33080238046378,
"grad_norm": 157.6916046142578,
"learning_rate": 6.691297208538588e-06,
"loss": 0.1718,
"step": 806
},
{
"epoch": 0.3312128052534373,
"grad_norm": 13.374323844909668,
"learning_rate": 6.687192118226602e-06,
"loss": 0.0196,
"step": 807
},
{
"epoch": 0.33162323004309463,
"grad_norm": 41.80398941040039,
"learning_rate": 6.683087027914614e-06,
"loss": 0.075,
"step": 808
},
{
"epoch": 0.3320336548327519,
"grad_norm": 22.308025360107422,
"learning_rate": 6.6789819376026275e-06,
"loss": 0.0432,
"step": 809
},
{
"epoch": 0.3324440796224092,
"grad_norm": 86.02617645263672,
"learning_rate": 6.674876847290641e-06,
"loss": 0.4675,
"step": 810
},
{
"epoch": 0.3328545044120665,
"grad_norm": 160.2139434814453,
"learning_rate": 6.6707717569786544e-06,
"loss": 0.3303,
"step": 811
},
{
"epoch": 0.3332649292017238,
"grad_norm": 67.07664489746094,
"learning_rate": 6.666666666666667e-06,
"loss": 0.3909,
"step": 812
},
{
"epoch": 0.3336753539913811,
"grad_norm": 30.874765396118164,
"learning_rate": 6.6625615763546806e-06,
"loss": 0.0663,
"step": 813
},
{
"epoch": 0.3340857787810384,
"grad_norm": 41.607425689697266,
"learning_rate": 6.658456486042694e-06,
"loss": 0.0845,
"step": 814
},
{
"epoch": 0.33449620357069565,
"grad_norm": 143.08201599121094,
"learning_rate": 6.654351395730707e-06,
"loss": 0.1724,
"step": 815
},
{
"epoch": 0.33490662836035295,
"grad_norm": 150.86390686035156,
"learning_rate": 6.65024630541872e-06,
"loss": 0.1957,
"step": 816
},
{
"epoch": 0.33531705315001026,
"grad_norm": 81.38294982910156,
"learning_rate": 6.646141215106733e-06,
"loss": 0.1274,
"step": 817
},
{
"epoch": 0.33572747793966756,
"grad_norm": 20.498252868652344,
"learning_rate": 6.642036124794746e-06,
"loss": 0.0265,
"step": 818
},
{
"epoch": 0.33613790272932487,
"grad_norm": 20.45568084716797,
"learning_rate": 6.63793103448276e-06,
"loss": 0.0341,
"step": 819
},
{
"epoch": 0.33654832751898217,
"grad_norm": 591.8125,
"learning_rate": 6.633825944170772e-06,
"loss": 0.9708,
"step": 820
},
{
"epoch": 0.3369587523086394,
"grad_norm": 32.497920989990234,
"learning_rate": 6.629720853858785e-06,
"loss": 0.064,
"step": 821
},
{
"epoch": 0.3373691770982967,
"grad_norm": 101.23199462890625,
"learning_rate": 6.625615763546799e-06,
"loss": 0.4697,
"step": 822
},
{
"epoch": 0.33777960188795403,
"grad_norm": 153.53814697265625,
"learning_rate": 6.621510673234812e-06,
"loss": 0.2212,
"step": 823
},
{
"epoch": 0.33819002667761133,
"grad_norm": 18.442771911621094,
"learning_rate": 6.617405582922824e-06,
"loss": 0.0229,
"step": 824
},
{
"epoch": 0.33860045146726864,
"grad_norm": 45.4644889831543,
"learning_rate": 6.613300492610838e-06,
"loss": 0.0845,
"step": 825
},
{
"epoch": 0.33901087625692594,
"grad_norm": 6.767117500305176,
"learning_rate": 6.609195402298851e-06,
"loss": 0.0127,
"step": 826
},
{
"epoch": 0.3394213010465832,
"grad_norm": 81.11670684814453,
"learning_rate": 6.605090311986865e-06,
"loss": 0.1361,
"step": 827
},
{
"epoch": 0.3398317258362405,
"grad_norm": 6.014718055725098,
"learning_rate": 6.600985221674877e-06,
"loss": 0.0184,
"step": 828
},
{
"epoch": 0.3402421506258978,
"grad_norm": 41.17439651489258,
"learning_rate": 6.5968801313628904e-06,
"loss": 0.0542,
"step": 829
},
{
"epoch": 0.3406525754155551,
"grad_norm": 8.0233736038208,
"learning_rate": 6.592775041050904e-06,
"loss": 0.0076,
"step": 830
},
{
"epoch": 0.3410630002052124,
"grad_norm": 70.58004760742188,
"learning_rate": 6.5886699507389166e-06,
"loss": 0.2734,
"step": 831
},
{
"epoch": 0.3414734249948697,
"grad_norm": 19.44939422607422,
"learning_rate": 6.58456486042693e-06,
"loss": 0.0304,
"step": 832
},
{
"epoch": 0.34188384978452696,
"grad_norm": 197.88758850097656,
"learning_rate": 6.580459770114943e-06,
"loss": 0.522,
"step": 833
},
{
"epoch": 0.34229427457418427,
"grad_norm": 8.253340721130371,
"learning_rate": 6.576354679802957e-06,
"loss": 0.0071,
"step": 834
},
{
"epoch": 0.34270469936384157,
"grad_norm": 3.466676712036133,
"learning_rate": 6.572249589490969e-06,
"loss": 0.0081,
"step": 835
},
{
"epoch": 0.3431151241534989,
"grad_norm": 4.775365352630615,
"learning_rate": 6.568144499178982e-06,
"loss": 0.0163,
"step": 836
},
{
"epoch": 0.3435255489431562,
"grad_norm": 21.229434967041016,
"learning_rate": 6.564039408866996e-06,
"loss": 0.0238,
"step": 837
},
{
"epoch": 0.3439359737328135,
"grad_norm": 22.808441162109375,
"learning_rate": 6.559934318555009e-06,
"loss": 0.0277,
"step": 838
},
{
"epoch": 0.34434639852247073,
"grad_norm": 9.179484367370605,
"learning_rate": 6.555829228243021e-06,
"loss": 0.0073,
"step": 839
},
{
"epoch": 0.34475682331212804,
"grad_norm": 3.1651358604431152,
"learning_rate": 6.551724137931035e-06,
"loss": 0.0016,
"step": 840
},
{
"epoch": 0.34516724810178534,
"grad_norm": 2.8552443981170654,
"learning_rate": 6.547619047619048e-06,
"loss": 0.0039,
"step": 841
},
{
"epoch": 0.34557767289144264,
"grad_norm": 42.81547546386719,
"learning_rate": 6.543513957307062e-06,
"loss": 0.0994,
"step": 842
},
{
"epoch": 0.34598809768109995,
"grad_norm": 66.4178237915039,
"learning_rate": 6.539408866995074e-06,
"loss": 0.3018,
"step": 843
},
{
"epoch": 0.34639852247075725,
"grad_norm": 3.738849639892578,
"learning_rate": 6.535303776683087e-06,
"loss": 0.0057,
"step": 844
},
{
"epoch": 0.3468089472604145,
"grad_norm": 87.6531982421875,
"learning_rate": 6.531198686371101e-06,
"loss": 0.3525,
"step": 845
},
{
"epoch": 0.3472193720500718,
"grad_norm": 53.315391540527344,
"learning_rate": 6.527093596059114e-06,
"loss": 0.1649,
"step": 846
},
{
"epoch": 0.3476297968397291,
"grad_norm": 51.64055252075195,
"learning_rate": 6.5229885057471264e-06,
"loss": 0.3837,
"step": 847
},
{
"epoch": 0.3480402216293864,
"grad_norm": 119.36158752441406,
"learning_rate": 6.5188834154351395e-06,
"loss": 0.1866,
"step": 848
},
{
"epoch": 0.3484506464190437,
"grad_norm": 6.374325752258301,
"learning_rate": 6.514778325123153e-06,
"loss": 0.0093,
"step": 849
},
{
"epoch": 0.348861071208701,
"grad_norm": 35.218997955322266,
"learning_rate": 6.5106732348111665e-06,
"loss": 0.0964,
"step": 850
},
{
"epoch": 0.34927149599835833,
"grad_norm": 7.959962368011475,
"learning_rate": 6.506568144499179e-06,
"loss": 0.0112,
"step": 851
},
{
"epoch": 0.3496819207880156,
"grad_norm": 28.7159481048584,
"learning_rate": 6.502463054187193e-06,
"loss": 0.0516,
"step": 852
},
{
"epoch": 0.3500923455776729,
"grad_norm": 18.469106674194336,
"learning_rate": 6.498357963875206e-06,
"loss": 0.0731,
"step": 853
},
{
"epoch": 0.3505027703673302,
"grad_norm": 15.750736236572266,
"learning_rate": 6.49425287356322e-06,
"loss": 0.0422,
"step": 854
},
{
"epoch": 0.3509131951569875,
"grad_norm": 32.09444046020508,
"learning_rate": 6.490147783251232e-06,
"loss": 0.0634,
"step": 855
},
{
"epoch": 0.3513236199466448,
"grad_norm": 37.88212966918945,
"learning_rate": 6.486042692939245e-06,
"loss": 0.2153,
"step": 856
},
{
"epoch": 0.3517340447363021,
"grad_norm": 8.812582969665527,
"learning_rate": 6.481937602627259e-06,
"loss": 0.0078,
"step": 857
},
{
"epoch": 0.35214446952595935,
"grad_norm": 4.790536403656006,
"learning_rate": 6.477832512315272e-06,
"loss": 0.0122,
"step": 858
},
{
"epoch": 0.35255489431561665,
"grad_norm": 185.9674835205078,
"learning_rate": 6.473727422003284e-06,
"loss": 0.5133,
"step": 859
},
{
"epoch": 0.35296531910527396,
"grad_norm": 2.562713384628296,
"learning_rate": 6.469622331691298e-06,
"loss": 0.0041,
"step": 860
},
{
"epoch": 0.35337574389493126,
"grad_norm": 15.286489486694336,
"learning_rate": 6.465517241379311e-06,
"loss": 0.0268,
"step": 861
},
{
"epoch": 0.35378616868458856,
"grad_norm": 8.753963470458984,
"learning_rate": 6.461412151067324e-06,
"loss": 0.0063,
"step": 862
},
{
"epoch": 0.35419659347424587,
"grad_norm": 6.741109371185303,
"learning_rate": 6.457307060755337e-06,
"loss": 0.0164,
"step": 863
},
{
"epoch": 0.3546070182639031,
"grad_norm": 45.45419692993164,
"learning_rate": 6.45320197044335e-06,
"loss": 0.0551,
"step": 864
},
{
"epoch": 0.3550174430535604,
"grad_norm": 54.81714630126953,
"learning_rate": 6.449096880131363e-06,
"loss": 0.1372,
"step": 865
},
{
"epoch": 0.3554278678432177,
"grad_norm": 7.016499996185303,
"learning_rate": 6.444991789819377e-06,
"loss": 0.0071,
"step": 866
},
{
"epoch": 0.35583829263287503,
"grad_norm": 102.70311737060547,
"learning_rate": 6.440886699507389e-06,
"loss": 0.4046,
"step": 867
},
{
"epoch": 0.35624871742253233,
"grad_norm": 64.22552490234375,
"learning_rate": 6.4367816091954025e-06,
"loss": 0.1682,
"step": 868
},
{
"epoch": 0.35665914221218964,
"grad_norm": 6.555329322814941,
"learning_rate": 6.432676518883416e-06,
"loss": 0.0064,
"step": 869
},
{
"epoch": 0.3570695670018469,
"grad_norm": 16.056413650512695,
"learning_rate": 6.4285714285714295e-06,
"loss": 0.0216,
"step": 870
},
{
"epoch": 0.3574799917915042,
"grad_norm": 24.24204444885254,
"learning_rate": 6.424466338259442e-06,
"loss": 0.0428,
"step": 871
},
{
"epoch": 0.3578904165811615,
"grad_norm": 51.104713439941406,
"learning_rate": 6.420361247947456e-06,
"loss": 0.0795,
"step": 872
},
{
"epoch": 0.3583008413708188,
"grad_norm": 13.193655967712402,
"learning_rate": 6.416256157635469e-06,
"loss": 0.0139,
"step": 873
},
{
"epoch": 0.3587112661604761,
"grad_norm": 9.999794960021973,
"learning_rate": 6.412151067323482e-06,
"loss": 0.0233,
"step": 874
},
{
"epoch": 0.3591216909501334,
"grad_norm": 11.586943626403809,
"learning_rate": 6.408045977011495e-06,
"loss": 0.0361,
"step": 875
},
{
"epoch": 0.35953211573979066,
"grad_norm": 21.36371612548828,
"learning_rate": 6.403940886699508e-06,
"loss": 0.0462,
"step": 876
},
{
"epoch": 0.35994254052944796,
"grad_norm": 8.164166450500488,
"learning_rate": 6.399835796387521e-06,
"loss": 0.0078,
"step": 877
},
{
"epoch": 0.36035296531910527,
"grad_norm": 304.26324462890625,
"learning_rate": 6.395730706075535e-06,
"loss": 1.1082,
"step": 878
},
{
"epoch": 0.36076339010876257,
"grad_norm": 10.75214672088623,
"learning_rate": 6.391625615763547e-06,
"loss": 0.0234,
"step": 879
},
{
"epoch": 0.3611738148984199,
"grad_norm": 45.88087844848633,
"learning_rate": 6.38752052545156e-06,
"loss": 0.1073,
"step": 880
},
{
"epoch": 0.3615842396880772,
"grad_norm": 11.299022674560547,
"learning_rate": 6.383415435139574e-06,
"loss": 0.0193,
"step": 881
},
{
"epoch": 0.36199466447773443,
"grad_norm": 12.817790985107422,
"learning_rate": 6.379310344827587e-06,
"loss": 0.0164,
"step": 882
},
{
"epoch": 0.36240508926739173,
"grad_norm": 3.7720139026641846,
"learning_rate": 6.375205254515599e-06,
"loss": 0.0044,
"step": 883
},
{
"epoch": 0.36281551405704904,
"grad_norm": 22.774124145507812,
"learning_rate": 6.371100164203613e-06,
"loss": 0.0337,
"step": 884
},
{
"epoch": 0.36322593884670634,
"grad_norm": 12.13525104522705,
"learning_rate": 6.366995073891626e-06,
"loss": 0.0138,
"step": 885
},
{
"epoch": 0.36363636363636365,
"grad_norm": 226.2864227294922,
"learning_rate": 6.3628899835796385e-06,
"loss": 0.9936,
"step": 886
},
{
"epoch": 0.36404678842602095,
"grad_norm": 69.00067901611328,
"learning_rate": 6.358784893267652e-06,
"loss": 0.0531,
"step": 887
},
{
"epoch": 0.3644572132156782,
"grad_norm": 5.719005107879639,
"learning_rate": 6.3546798029556655e-06,
"loss": 0.0047,
"step": 888
},
{
"epoch": 0.3648676380053355,
"grad_norm": 50.68075942993164,
"learning_rate": 6.350574712643679e-06,
"loss": 0.3266,
"step": 889
},
{
"epoch": 0.3652780627949928,
"grad_norm": 7.143153190612793,
"learning_rate": 6.346469622331692e-06,
"loss": 0.005,
"step": 890
},
{
"epoch": 0.3656884875846501,
"grad_norm": 4.772278785705566,
"learning_rate": 6.342364532019705e-06,
"loss": 0.0064,
"step": 891
},
{
"epoch": 0.3660989123743074,
"grad_norm": 6.438401699066162,
"learning_rate": 6.338259441707718e-06,
"loss": 0.0042,
"step": 892
},
{
"epoch": 0.3665093371639647,
"grad_norm": 2.9077413082122803,
"learning_rate": 6.334154351395732e-06,
"loss": 0.0057,
"step": 893
},
{
"epoch": 0.366919761953622,
"grad_norm": 6.236939430236816,
"learning_rate": 6.330049261083744e-06,
"loss": 0.0275,
"step": 894
},
{
"epoch": 0.3673301867432793,
"grad_norm": 5.546217441558838,
"learning_rate": 6.325944170771757e-06,
"loss": 0.0053,
"step": 895
},
{
"epoch": 0.3677406115329366,
"grad_norm": 13.809030532836914,
"learning_rate": 6.321839080459771e-06,
"loss": 0.0238,
"step": 896
},
{
"epoch": 0.3681510363225939,
"grad_norm": 37.953861236572266,
"learning_rate": 6.317733990147784e-06,
"loss": 0.1685,
"step": 897
},
{
"epoch": 0.3685614611122512,
"grad_norm": 10.83439826965332,
"learning_rate": 6.313628899835796e-06,
"loss": 0.0092,
"step": 898
},
{
"epoch": 0.3689718859019085,
"grad_norm": 5.664017200469971,
"learning_rate": 6.30952380952381e-06,
"loss": 0.0091,
"step": 899
},
{
"epoch": 0.3693823106915658,
"grad_norm": 44.252723693847656,
"learning_rate": 6.305418719211823e-06,
"loss": 0.0473,
"step": 900
},
{
"epoch": 0.36979273548122304,
"grad_norm": 26.698720932006836,
"learning_rate": 6.301313628899837e-06,
"loss": 0.0416,
"step": 901
},
{
"epoch": 0.37020316027088035,
"grad_norm": 15.42609691619873,
"learning_rate": 6.297208538587849e-06,
"loss": 0.0157,
"step": 902
},
{
"epoch": 0.37061358506053765,
"grad_norm": 3.491015911102295,
"learning_rate": 6.293103448275862e-06,
"loss": 0.0035,
"step": 903
},
{
"epoch": 0.37102400985019496,
"grad_norm": 5.762197971343994,
"learning_rate": 6.288998357963876e-06,
"loss": 0.0083,
"step": 904
},
{
"epoch": 0.37143443463985226,
"grad_norm": 28.95611000061035,
"learning_rate": 6.284893267651889e-06,
"loss": 0.0302,
"step": 905
},
{
"epoch": 0.37184485942950957,
"grad_norm": 6.477351665496826,
"learning_rate": 6.2807881773399015e-06,
"loss": 0.0045,
"step": 906
},
{
"epoch": 0.3722552842191668,
"grad_norm": 63.550540924072266,
"learning_rate": 6.2766830870279145e-06,
"loss": 0.2237,
"step": 907
},
{
"epoch": 0.3726657090088241,
"grad_norm": 419.23468017578125,
"learning_rate": 6.2725779967159284e-06,
"loss": 0.5075,
"step": 908
},
{
"epoch": 0.3730761337984814,
"grad_norm": 63.17408752441406,
"learning_rate": 6.2684729064039415e-06,
"loss": 0.1409,
"step": 909
},
{
"epoch": 0.37348655858813873,
"grad_norm": 1.6836413145065308,
"learning_rate": 6.264367816091954e-06,
"loss": 0.0025,
"step": 910
},
{
"epoch": 0.37389698337779603,
"grad_norm": 78.11280822753906,
"learning_rate": 6.260262725779968e-06,
"loss": 0.5371,
"step": 911
},
{
"epoch": 0.37430740816745334,
"grad_norm": 7.896859645843506,
"learning_rate": 6.256157635467981e-06,
"loss": 0.0128,
"step": 912
},
{
"epoch": 0.3747178329571106,
"grad_norm": 12.036942481994629,
"learning_rate": 6.252052545155995e-06,
"loss": 0.0204,
"step": 913
},
{
"epoch": 0.3751282577467679,
"grad_norm": 34.609703063964844,
"learning_rate": 6.247947454844007e-06,
"loss": 0.0457,
"step": 914
},
{
"epoch": 0.3755386825364252,
"grad_norm": 21.267837524414062,
"learning_rate": 6.24384236453202e-06,
"loss": 0.0297,
"step": 915
},
{
"epoch": 0.3759491073260825,
"grad_norm": 53.314422607421875,
"learning_rate": 6.239737274220034e-06,
"loss": 0.0432,
"step": 916
},
{
"epoch": 0.3763595321157398,
"grad_norm": 9.229562759399414,
"learning_rate": 6.235632183908047e-06,
"loss": 0.0082,
"step": 917
},
{
"epoch": 0.3767699569053971,
"grad_norm": 2.253166675567627,
"learning_rate": 6.231527093596059e-06,
"loss": 0.0033,
"step": 918
},
{
"epoch": 0.37718038169505436,
"grad_norm": 269.947509765625,
"learning_rate": 6.227422003284073e-06,
"loss": 0.5709,
"step": 919
},
{
"epoch": 0.37759080648471166,
"grad_norm": 5.581409454345703,
"learning_rate": 6.223316912972086e-06,
"loss": 0.0135,
"step": 920
},
{
"epoch": 0.37800123127436897,
"grad_norm": 57.35350799560547,
"learning_rate": 6.219211822660099e-06,
"loss": 0.0741,
"step": 921
},
{
"epoch": 0.37841165606402627,
"grad_norm": 25.515399932861328,
"learning_rate": 6.215106732348112e-06,
"loss": 0.0233,
"step": 922
},
{
"epoch": 0.3788220808536836,
"grad_norm": 10.220430374145508,
"learning_rate": 6.211001642036125e-06,
"loss": 0.0128,
"step": 923
},
{
"epoch": 0.3792325056433409,
"grad_norm": 18.87088394165039,
"learning_rate": 6.206896551724138e-06,
"loss": 0.0143,
"step": 924
},
{
"epoch": 0.3796429304329981,
"grad_norm": 72.07530212402344,
"learning_rate": 6.202791461412152e-06,
"loss": 0.4562,
"step": 925
},
{
"epoch": 0.38005335522265543,
"grad_norm": 5.5519022941589355,
"learning_rate": 6.1986863711001645e-06,
"loss": 0.0157,
"step": 926
},
{
"epoch": 0.38046378001231274,
"grad_norm": 8.940093994140625,
"learning_rate": 6.1945812807881775e-06,
"loss": 0.0063,
"step": 927
},
{
"epoch": 0.38087420480197004,
"grad_norm": 7.912701606750488,
"learning_rate": 6.1904761904761914e-06,
"loss": 0.0144,
"step": 928
},
{
"epoch": 0.38128462959162734,
"grad_norm": 17.445594787597656,
"learning_rate": 6.1863711001642045e-06,
"loss": 0.013,
"step": 929
},
{
"epoch": 0.38169505438128465,
"grad_norm": 8.633885383605957,
"learning_rate": 6.182266009852217e-06,
"loss": 0.0082,
"step": 930
},
{
"epoch": 0.3821054791709419,
"grad_norm": 2.5007827281951904,
"learning_rate": 6.178160919540231e-06,
"loss": 0.002,
"step": 931
},
{
"epoch": 0.3825159039605992,
"grad_norm": 63.88785171508789,
"learning_rate": 6.174055829228244e-06,
"loss": 0.0794,
"step": 932
},
{
"epoch": 0.3829263287502565,
"grad_norm": 7.725972652435303,
"learning_rate": 6.169950738916258e-06,
"loss": 0.0071,
"step": 933
},
{
"epoch": 0.3833367535399138,
"grad_norm": 22.41608238220215,
"learning_rate": 6.16584564860427e-06,
"loss": 0.0322,
"step": 934
},
{
"epoch": 0.3837471783295711,
"grad_norm": 6.5261616706848145,
"learning_rate": 6.161740558292283e-06,
"loss": 0.0068,
"step": 935
},
{
"epoch": 0.3841576031192284,
"grad_norm": 68.64938354492188,
"learning_rate": 6.157635467980296e-06,
"loss": 0.5134,
"step": 936
},
{
"epoch": 0.3845680279088857,
"grad_norm": 57.931053161621094,
"learning_rate": 6.15353037766831e-06,
"loss": 0.2489,
"step": 937
},
{
"epoch": 0.38497845269854297,
"grad_norm": 7.4314374923706055,
"learning_rate": 6.149425287356322e-06,
"loss": 0.0122,
"step": 938
},
{
"epoch": 0.3853888774882003,
"grad_norm": 11.907365798950195,
"learning_rate": 6.145320197044335e-06,
"loss": 0.0123,
"step": 939
},
{
"epoch": 0.3857993022778576,
"grad_norm": 93.22732543945312,
"learning_rate": 6.141215106732349e-06,
"loss": 0.2217,
"step": 940
},
{
"epoch": 0.3862097270675149,
"grad_norm": 48.61363220214844,
"learning_rate": 6.137110016420362e-06,
"loss": 0.3247,
"step": 941
},
{
"epoch": 0.3866201518571722,
"grad_norm": 4.708130836486816,
"learning_rate": 6.133004926108374e-06,
"loss": 0.0074,
"step": 942
},
{
"epoch": 0.3870305766468295,
"grad_norm": 0.8834515810012817,
"learning_rate": 6.128899835796388e-06,
"loss": 0.0007,
"step": 943
},
{
"epoch": 0.38744100143648674,
"grad_norm": 7.853377342224121,
"learning_rate": 6.124794745484401e-06,
"loss": 0.0056,
"step": 944
},
{
"epoch": 0.38785142622614405,
"grad_norm": 17.863636016845703,
"learning_rate": 6.1206896551724135e-06,
"loss": 0.0229,
"step": 945
},
{
"epoch": 0.38826185101580135,
"grad_norm": 20.97165298461914,
"learning_rate": 6.1165845648604274e-06,
"loss": 0.0186,
"step": 946
},
{
"epoch": 0.38867227580545866,
"grad_norm": 2.9225423336029053,
"learning_rate": 6.1124794745484405e-06,
"loss": 0.0043,
"step": 947
},
{
"epoch": 0.38908270059511596,
"grad_norm": 39.09851837158203,
"learning_rate": 6.108374384236454e-06,
"loss": 0.3374,
"step": 948
},
{
"epoch": 0.38949312538477326,
"grad_norm": 10.05467414855957,
"learning_rate": 6.104269293924467e-06,
"loss": 0.009,
"step": 949
},
{
"epoch": 0.3899035501744305,
"grad_norm": 19.8225040435791,
"learning_rate": 6.10016420361248e-06,
"loss": 0.029,
"step": 950
},
{
"epoch": 0.3903139749640878,
"grad_norm": 37.21845626831055,
"learning_rate": 6.096059113300493e-06,
"loss": 0.0871,
"step": 951
},
{
"epoch": 0.3907243997537451,
"grad_norm": 15.355683326721191,
"learning_rate": 6.091954022988507e-06,
"loss": 0.0161,
"step": 952
},
{
"epoch": 0.3911348245434024,
"grad_norm": 11.209900856018066,
"learning_rate": 6.087848932676519e-06,
"loss": 0.0116,
"step": 953
},
{
"epoch": 0.39154524933305973,
"grad_norm": 2.2375736236572266,
"learning_rate": 6.083743842364532e-06,
"loss": 0.003,
"step": 954
},
{
"epoch": 0.39195567412271703,
"grad_norm": 5.738273620605469,
"learning_rate": 6.079638752052546e-06,
"loss": 0.0044,
"step": 955
},
{
"epoch": 0.3923660989123743,
"grad_norm": 10.774125099182129,
"learning_rate": 6.075533661740559e-06,
"loss": 0.0091,
"step": 956
},
{
"epoch": 0.3927765237020316,
"grad_norm": 57.18001174926758,
"learning_rate": 6.071428571428571e-06,
"loss": 0.3245,
"step": 957
},
{
"epoch": 0.3931869484916889,
"grad_norm": 10.118163108825684,
"learning_rate": 6.067323481116585e-06,
"loss": 0.0082,
"step": 958
},
{
"epoch": 0.3935973732813462,
"grad_norm": 35.93637466430664,
"learning_rate": 6.063218390804598e-06,
"loss": 0.0309,
"step": 959
},
{
"epoch": 0.3940077980710035,
"grad_norm": 4.547665596008301,
"learning_rate": 6.059113300492612e-06,
"loss": 0.0127,
"step": 960
},
{
"epoch": 0.3944182228606608,
"grad_norm": 0.9685713648796082,
"learning_rate": 6.055008210180624e-06,
"loss": 0.0022,
"step": 961
},
{
"epoch": 0.39482864765031805,
"grad_norm": 58.37193298339844,
"learning_rate": 6.050903119868637e-06,
"loss": 0.4875,
"step": 962
},
{
"epoch": 0.39523907243997536,
"grad_norm": 10.679574966430664,
"learning_rate": 6.046798029556651e-06,
"loss": 0.0112,
"step": 963
},
{
"epoch": 0.39564949722963266,
"grad_norm": 99.4417495727539,
"learning_rate": 6.042692939244664e-06,
"loss": 0.0592,
"step": 964
},
{
"epoch": 0.39605992201928997,
"grad_norm": 4.979918479919434,
"learning_rate": 6.0385878489326765e-06,
"loss": 0.0029,
"step": 965
},
{
"epoch": 0.39647034680894727,
"grad_norm": 8.102355003356934,
"learning_rate": 6.03448275862069e-06,
"loss": 0.0113,
"step": 966
},
{
"epoch": 0.3968807715986046,
"grad_norm": 5.101386547088623,
"learning_rate": 6.0303776683087035e-06,
"loss": 0.0064,
"step": 967
},
{
"epoch": 0.3972911963882618,
"grad_norm": 102.72512817382812,
"learning_rate": 6.0262725779967165e-06,
"loss": 0.5602,
"step": 968
},
{
"epoch": 0.39770162117791913,
"grad_norm": 5.838058948516846,
"learning_rate": 6.022167487684729e-06,
"loss": 0.0078,
"step": 969
},
{
"epoch": 0.39811204596757643,
"grad_norm": 15.141478538513184,
"learning_rate": 6.018062397372743e-06,
"loss": 0.0103,
"step": 970
},
{
"epoch": 0.39852247075723374,
"grad_norm": 44.59075164794922,
"learning_rate": 6.013957307060756e-06,
"loss": 0.2552,
"step": 971
},
{
"epoch": 0.39893289554689104,
"grad_norm": 95.1900863647461,
"learning_rate": 6.00985221674877e-06,
"loss": 0.3928,
"step": 972
},
{
"epoch": 0.39934332033654835,
"grad_norm": 19.56851577758789,
"learning_rate": 6.005747126436782e-06,
"loss": 0.0081,
"step": 973
},
{
"epoch": 0.3997537451262056,
"grad_norm": 15.437690734863281,
"learning_rate": 6.001642036124795e-06,
"loss": 0.018,
"step": 974
},
{
"epoch": 0.4001641699158629,
"grad_norm": 10.937396049499512,
"learning_rate": 5.997536945812809e-06,
"loss": 0.009,
"step": 975
},
{
"epoch": 0.4005745947055202,
"grad_norm": 94.95052337646484,
"learning_rate": 5.993431855500822e-06,
"loss": 0.2601,
"step": 976
},
{
"epoch": 0.4009850194951775,
"grad_norm": 48.627811431884766,
"learning_rate": 5.989326765188834e-06,
"loss": 0.083,
"step": 977
},
{
"epoch": 0.4013954442848348,
"grad_norm": 22.44864273071289,
"learning_rate": 5.985221674876848e-06,
"loss": 0.0375,
"step": 978
},
{
"epoch": 0.4018058690744921,
"grad_norm": 48.15469741821289,
"learning_rate": 5.981116584564861e-06,
"loss": 0.3105,
"step": 979
},
{
"epoch": 0.4022162938641494,
"grad_norm": 12.416152954101562,
"learning_rate": 5.977011494252874e-06,
"loss": 0.0143,
"step": 980
},
{
"epoch": 0.40262671865380667,
"grad_norm": 48.1343994140625,
"learning_rate": 5.972906403940887e-06,
"loss": 0.0477,
"step": 981
},
{
"epoch": 0.403037143443464,
"grad_norm": 9.80142593383789,
"learning_rate": 5.9688013136289e-06,
"loss": 0.0127,
"step": 982
},
{
"epoch": 0.4034475682331213,
"grad_norm": 6.255521774291992,
"learning_rate": 5.964696223316913e-06,
"loss": 0.0074,
"step": 983
},
{
"epoch": 0.4038579930227786,
"grad_norm": 61.690818786621094,
"learning_rate": 5.960591133004927e-06,
"loss": 0.0779,
"step": 984
},
{
"epoch": 0.4042684178124359,
"grad_norm": 41.711368560791016,
"learning_rate": 5.9564860426929395e-06,
"loss": 0.3969,
"step": 985
},
{
"epoch": 0.4046788426020932,
"grad_norm": 14.073623657226562,
"learning_rate": 5.9523809523809525e-06,
"loss": 0.0226,
"step": 986
},
{
"epoch": 0.40508926739175044,
"grad_norm": 28.2772274017334,
"learning_rate": 5.9482758620689665e-06,
"loss": 0.0135,
"step": 987
},
{
"epoch": 0.40549969218140774,
"grad_norm": 58.14522933959961,
"learning_rate": 5.9441707717569795e-06,
"loss": 0.682,
"step": 988
},
{
"epoch": 0.40591011697106505,
"grad_norm": 152.39698791503906,
"learning_rate": 5.940065681444992e-06,
"loss": 0.9219,
"step": 989
},
{
"epoch": 0.40632054176072235,
"grad_norm": 11.694584846496582,
"learning_rate": 5.935960591133006e-06,
"loss": 0.02,
"step": 990
},
{
"epoch": 0.40673096655037966,
"grad_norm": 11.670843124389648,
"learning_rate": 5.931855500821019e-06,
"loss": 0.0266,
"step": 991
},
{
"epoch": 0.40714139134003696,
"grad_norm": 15.231695175170898,
"learning_rate": 5.927750410509033e-06,
"loss": 0.0154,
"step": 992
},
{
"epoch": 0.4075518161296942,
"grad_norm": 2.232269763946533,
"learning_rate": 5.923645320197045e-06,
"loss": 0.004,
"step": 993
},
{
"epoch": 0.4079622409193515,
"grad_norm": 6.8556599617004395,
"learning_rate": 5.919540229885058e-06,
"loss": 0.012,
"step": 994
},
{
"epoch": 0.4083726657090088,
"grad_norm": 6.16105842590332,
"learning_rate": 5.915435139573071e-06,
"loss": 0.0105,
"step": 995
},
{
"epoch": 0.4087830904986661,
"grad_norm": 3.1271626949310303,
"learning_rate": 5.911330049261085e-06,
"loss": 0.0027,
"step": 996
},
{
"epoch": 0.40919351528832343,
"grad_norm": 3.8831164836883545,
"learning_rate": 5.907224958949097e-06,
"loss": 0.003,
"step": 997
},
{
"epoch": 0.40960394007798073,
"grad_norm": 69.96463775634766,
"learning_rate": 5.90311986863711e-06,
"loss": 0.5244,
"step": 998
},
{
"epoch": 0.410014364867638,
"grad_norm": 7.346024990081787,
"learning_rate": 5.899014778325124e-06,
"loss": 0.006,
"step": 999
},
{
"epoch": 0.4104247896572953,
"grad_norm": 90.3431625366211,
"learning_rate": 5.894909688013136e-06,
"loss": 0.4858,
"step": 1000
},
{
"epoch": 0.4108352144469526,
"grad_norm": 102.3450698852539,
"learning_rate": 5.890804597701149e-06,
"loss": 0.2667,
"step": 1001
},
{
"epoch": 0.4112456392366099,
"grad_norm": 22.334352493286133,
"learning_rate": 5.886699507389163e-06,
"loss": 0.0358,
"step": 1002
},
{
"epoch": 0.4116560640262672,
"grad_norm": 29.094234466552734,
"learning_rate": 5.882594417077176e-06,
"loss": 0.0403,
"step": 1003
},
{
"epoch": 0.4120664888159245,
"grad_norm": 22.555749893188477,
"learning_rate": 5.8784893267651886e-06,
"loss": 0.0331,
"step": 1004
},
{
"epoch": 0.41247691360558175,
"grad_norm": 67.39612579345703,
"learning_rate": 5.8743842364532025e-06,
"loss": 0.2654,
"step": 1005
},
{
"epoch": 0.41288733839523906,
"grad_norm": 17.151153564453125,
"learning_rate": 5.8702791461412155e-06,
"loss": 0.021,
"step": 1006
},
{
"epoch": 0.41329776318489636,
"grad_norm": 10.856535911560059,
"learning_rate": 5.8661740558292294e-06,
"loss": 0.0127,
"step": 1007
},
{
"epoch": 0.41370818797455367,
"grad_norm": 17.873050689697266,
"learning_rate": 5.862068965517242e-06,
"loss": 0.02,
"step": 1008
},
{
"epoch": 0.41411861276421097,
"grad_norm": 4.6350417137146,
"learning_rate": 5.857963875205255e-06,
"loss": 0.0098,
"step": 1009
},
{
"epoch": 0.4145290375538683,
"grad_norm": 3.230347156524658,
"learning_rate": 5.853858784893268e-06,
"loss": 0.003,
"step": 1010
},
{
"epoch": 0.4149394623435255,
"grad_norm": 5.676575183868408,
"learning_rate": 5.849753694581282e-06,
"loss": 0.0047,
"step": 1011
},
{
"epoch": 0.4153498871331828,
"grad_norm": 8.017566680908203,
"learning_rate": 5.845648604269294e-06,
"loss": 0.0105,
"step": 1012
},
{
"epoch": 0.41576031192284013,
"grad_norm": 1.5919480323791504,
"learning_rate": 5.841543513957307e-06,
"loss": 0.003,
"step": 1013
},
{
"epoch": 0.41617073671249744,
"grad_norm": 3.309227466583252,
"learning_rate": 5.837438423645321e-06,
"loss": 0.0076,
"step": 1014
},
{
"epoch": 0.41658116150215474,
"grad_norm": 9.972696304321289,
"learning_rate": 5.833333333333334e-06,
"loss": 0.0173,
"step": 1015
},
{
"epoch": 0.41699158629181204,
"grad_norm": 5.839928150177002,
"learning_rate": 5.829228243021346e-06,
"loss": 0.01,
"step": 1016
},
{
"epoch": 0.41740201108146935,
"grad_norm": 114.14157104492188,
"learning_rate": 5.82512315270936e-06,
"loss": 0.1781,
"step": 1017
},
{
"epoch": 0.4178124358711266,
"grad_norm": 3.221158266067505,
"learning_rate": 5.821018062397373e-06,
"loss": 0.0031,
"step": 1018
},
{
"epoch": 0.4182228606607839,
"grad_norm": 20.551694869995117,
"learning_rate": 5.816912972085387e-06,
"loss": 0.0283,
"step": 1019
},
{
"epoch": 0.4186332854504412,
"grad_norm": 12.58276653289795,
"learning_rate": 5.812807881773399e-06,
"loss": 0.0124,
"step": 1020
},
{
"epoch": 0.4190437102400985,
"grad_norm": 15.103583335876465,
"learning_rate": 5.808702791461412e-06,
"loss": 0.0163,
"step": 1021
},
{
"epoch": 0.4194541350297558,
"grad_norm": 17.643802642822266,
"learning_rate": 5.804597701149426e-06,
"loss": 0.0279,
"step": 1022
},
{
"epoch": 0.4198645598194131,
"grad_norm": 37.83623123168945,
"learning_rate": 5.800492610837439e-06,
"loss": 0.0632,
"step": 1023
},
{
"epoch": 0.42027498460907037,
"grad_norm": 11.32297420501709,
"learning_rate": 5.7963875205254515e-06,
"loss": 0.0182,
"step": 1024
},
{
"epoch": 0.42068540939872767,
"grad_norm": 45.07491683959961,
"learning_rate": 5.7922824302134654e-06,
"loss": 0.0662,
"step": 1025
},
{
"epoch": 0.421095834188385,
"grad_norm": 130.98736572265625,
"learning_rate": 5.7881773399014785e-06,
"loss": 0.4233,
"step": 1026
},
{
"epoch": 0.4215062589780423,
"grad_norm": 2.0141468048095703,
"learning_rate": 5.7840722495894916e-06,
"loss": 0.0021,
"step": 1027
},
{
"epoch": 0.4219166837676996,
"grad_norm": 12.804417610168457,
"learning_rate": 5.779967159277504e-06,
"loss": 0.0166,
"step": 1028
},
{
"epoch": 0.4223271085573569,
"grad_norm": 10.731762886047363,
"learning_rate": 5.775862068965518e-06,
"loss": 0.0136,
"step": 1029
},
{
"epoch": 0.42273753334701414,
"grad_norm": 13.460164070129395,
"learning_rate": 5.771756978653531e-06,
"loss": 0.0126,
"step": 1030
},
{
"epoch": 0.42314795813667144,
"grad_norm": 39.81835174560547,
"learning_rate": 5.767651888341545e-06,
"loss": 0.1086,
"step": 1031
},
{
"epoch": 0.42355838292632875,
"grad_norm": 11.312249183654785,
"learning_rate": 5.763546798029557e-06,
"loss": 0.0093,
"step": 1032
},
{
"epoch": 0.42396880771598605,
"grad_norm": 2.3691084384918213,
"learning_rate": 5.75944170771757e-06,
"loss": 0.0013,
"step": 1033
},
{
"epoch": 0.42437923250564336,
"grad_norm": 4.003438949584961,
"learning_rate": 5.755336617405584e-06,
"loss": 0.0045,
"step": 1034
},
{
"epoch": 0.42478965729530066,
"grad_norm": 29.364086151123047,
"learning_rate": 5.751231527093597e-06,
"loss": 0.1555,
"step": 1035
},
{
"epoch": 0.4252000820849579,
"grad_norm": 0.8115753531455994,
"learning_rate": 5.747126436781609e-06,
"loss": 0.0023,
"step": 1036
},
{
"epoch": 0.4256105068746152,
"grad_norm": 1.5334142446517944,
"learning_rate": 5.743021346469623e-06,
"loss": 0.0035,
"step": 1037
},
{
"epoch": 0.4260209316642725,
"grad_norm": 48.21739959716797,
"learning_rate": 5.738916256157636e-06,
"loss": 0.2865,
"step": 1038
},
{
"epoch": 0.4264313564539298,
"grad_norm": 13.993239402770996,
"learning_rate": 5.734811165845649e-06,
"loss": 0.0194,
"step": 1039
},
{
"epoch": 0.4268417812435871,
"grad_norm": 120.41979217529297,
"learning_rate": 5.730706075533662e-06,
"loss": 0.2759,
"step": 1040
},
{
"epoch": 0.42725220603324443,
"grad_norm": 13.057295799255371,
"learning_rate": 5.726600985221675e-06,
"loss": 0.013,
"step": 1041
},
{
"epoch": 0.4276626308229017,
"grad_norm": 6.338970184326172,
"learning_rate": 5.722495894909688e-06,
"loss": 0.0049,
"step": 1042
},
{
"epoch": 0.428073055612559,
"grad_norm": 25.32278823852539,
"learning_rate": 5.718390804597702e-06,
"loss": 0.0348,
"step": 1043
},
{
"epoch": 0.4284834804022163,
"grad_norm": 3.7727205753326416,
"learning_rate": 5.7142857142857145e-06,
"loss": 0.0069,
"step": 1044
},
{
"epoch": 0.4288939051918736,
"grad_norm": 13.289739608764648,
"learning_rate": 5.7101806239737276e-06,
"loss": 0.0131,
"step": 1045
},
{
"epoch": 0.4293043299815309,
"grad_norm": 3.7899727821350098,
"learning_rate": 5.7060755336617415e-06,
"loss": 0.0019,
"step": 1046
},
{
"epoch": 0.4297147547711882,
"grad_norm": 56.44416809082031,
"learning_rate": 5.7019704433497546e-06,
"loss": 0.1325,
"step": 1047
},
{
"epoch": 0.43012517956084545,
"grad_norm": 4.5755696296691895,
"learning_rate": 5.697865353037767e-06,
"loss": 0.0069,
"step": 1048
},
{
"epoch": 0.43053560435050275,
"grad_norm": 12.52660846710205,
"learning_rate": 5.693760262725781e-06,
"loss": 0.015,
"step": 1049
},
{
"epoch": 0.43094602914016006,
"grad_norm": 7.6207709312438965,
"learning_rate": 5.689655172413794e-06,
"loss": 0.0054,
"step": 1050
},
{
"epoch": 0.43135645392981736,
"grad_norm": 116.74578094482422,
"learning_rate": 5.685550082101808e-06,
"loss": 0.2038,
"step": 1051
},
{
"epoch": 0.43176687871947467,
"grad_norm": 44.4088134765625,
"learning_rate": 5.68144499178982e-06,
"loss": 0.0448,
"step": 1052
},
{
"epoch": 0.43217730350913197,
"grad_norm": 1.7474746704101562,
"learning_rate": 5.677339901477833e-06,
"loss": 0.0008,
"step": 1053
},
{
"epoch": 0.4325877282987892,
"grad_norm": 30.239830017089844,
"learning_rate": 5.673234811165846e-06,
"loss": 0.0503,
"step": 1054
},
{
"epoch": 0.4329981530884465,
"grad_norm": 71.79635620117188,
"learning_rate": 5.669129720853859e-06,
"loss": 0.0623,
"step": 1055
},
{
"epoch": 0.43340857787810383,
"grad_norm": 58.20773696899414,
"learning_rate": 5.665024630541872e-06,
"loss": 0.1812,
"step": 1056
},
{
"epoch": 0.43381900266776113,
"grad_norm": 0.5399773120880127,
"learning_rate": 5.660919540229885e-06,
"loss": 0.0011,
"step": 1057
},
{
"epoch": 0.43422942745741844,
"grad_norm": 55.22797775268555,
"learning_rate": 5.656814449917899e-06,
"loss": 0.1154,
"step": 1058
},
{
"epoch": 0.43463985224707574,
"grad_norm": 10.137975692749023,
"learning_rate": 5.652709359605911e-06,
"loss": 0.0089,
"step": 1059
},
{
"epoch": 0.43505027703673305,
"grad_norm": 12.998510360717773,
"learning_rate": 5.648604269293924e-06,
"loss": 0.0126,
"step": 1060
},
{
"epoch": 0.4354607018263903,
"grad_norm": 15.42429256439209,
"learning_rate": 5.644499178981938e-06,
"loss": 0.0228,
"step": 1061
},
{
"epoch": 0.4358711266160476,
"grad_norm": 5.051717281341553,
"learning_rate": 5.640394088669951e-06,
"loss": 0.0038,
"step": 1062
},
{
"epoch": 0.4362815514057049,
"grad_norm": 53.436485290527344,
"learning_rate": 5.636288998357964e-06,
"loss": 0.0545,
"step": 1063
},
{
"epoch": 0.4366919761953622,
"grad_norm": 14.382844924926758,
"learning_rate": 5.6321839080459775e-06,
"loss": 0.0234,
"step": 1064
},
{
"epoch": 0.4371024009850195,
"grad_norm": 4.340682029724121,
"learning_rate": 5.6280788177339906e-06,
"loss": 0.0025,
"step": 1065
},
{
"epoch": 0.4375128257746768,
"grad_norm": 15.963550567626953,
"learning_rate": 5.6239737274220045e-06,
"loss": 0.0211,
"step": 1066
},
{
"epoch": 0.43792325056433407,
"grad_norm": 7.981912612915039,
"learning_rate": 5.619868637110017e-06,
"loss": 0.0071,
"step": 1067
},
{
"epoch": 0.43833367535399137,
"grad_norm": 13.171272277832031,
"learning_rate": 5.61576354679803e-06,
"loss": 0.0139,
"step": 1068
},
{
"epoch": 0.4387441001436487,
"grad_norm": 5.715782165527344,
"learning_rate": 5.611658456486044e-06,
"loss": 0.0032,
"step": 1069
},
{
"epoch": 0.439154524933306,
"grad_norm": 7.413060188293457,
"learning_rate": 5.607553366174057e-06,
"loss": 0.0053,
"step": 1070
},
{
"epoch": 0.4395649497229633,
"grad_norm": 3.820998191833496,
"learning_rate": 5.603448275862069e-06,
"loss": 0.0049,
"step": 1071
},
{
"epoch": 0.4399753745126206,
"grad_norm": 68.07166290283203,
"learning_rate": 5.599343185550082e-06,
"loss": 0.0641,
"step": 1072
},
{
"epoch": 0.44038579930227784,
"grad_norm": 135.13821411132812,
"learning_rate": 5.595238095238096e-06,
"loss": 0.257,
"step": 1073
},
{
"epoch": 0.44079622409193514,
"grad_norm": 105.56874084472656,
"learning_rate": 5.591133004926109e-06,
"loss": 0.6738,
"step": 1074
},
{
"epoch": 0.44120664888159244,
"grad_norm": 4.185121536254883,
"learning_rate": 5.587027914614121e-06,
"loss": 0.0024,
"step": 1075
},
{
"epoch": 0.44161707367124975,
"grad_norm": 55.040584564208984,
"learning_rate": 5.582922824302135e-06,
"loss": 0.4138,
"step": 1076
},
{
"epoch": 0.44202749846090705,
"grad_norm": 12.63291072845459,
"learning_rate": 5.578817733990148e-06,
"loss": 0.0169,
"step": 1077
},
{
"epoch": 0.44243792325056436,
"grad_norm": 10.089688301086426,
"learning_rate": 5.574712643678162e-06,
"loss": 0.0076,
"step": 1078
},
{
"epoch": 0.4428483480402216,
"grad_norm": 7.332252502441406,
"learning_rate": 5.570607553366174e-06,
"loss": 0.0066,
"step": 1079
},
{
"epoch": 0.4432587728298789,
"grad_norm": 12.770564079284668,
"learning_rate": 5.566502463054187e-06,
"loss": 0.0147,
"step": 1080
},
{
"epoch": 0.4436691976195362,
"grad_norm": 19.162717819213867,
"learning_rate": 5.562397372742201e-06,
"loss": 0.0296,
"step": 1081
},
{
"epoch": 0.4440796224091935,
"grad_norm": 3.255197763442993,
"learning_rate": 5.558292282430214e-06,
"loss": 0.0049,
"step": 1082
},
{
"epoch": 0.4444900471988508,
"grad_norm": 15.771563529968262,
"learning_rate": 5.5541871921182266e-06,
"loss": 0.0229,
"step": 1083
},
{
"epoch": 0.44490047198850813,
"grad_norm": 7.867246627807617,
"learning_rate": 5.5500821018062405e-06,
"loss": 0.007,
"step": 1084
},
{
"epoch": 0.4453108967781654,
"grad_norm": 12.81164264678955,
"learning_rate": 5.5459770114942535e-06,
"loss": 0.0125,
"step": 1085
},
{
"epoch": 0.4457213215678227,
"grad_norm": 14.956745147705078,
"learning_rate": 5.541871921182267e-06,
"loss": 0.0328,
"step": 1086
},
{
"epoch": 0.44613174635748,
"grad_norm": 142.6184844970703,
"learning_rate": 5.537766830870279e-06,
"loss": 0.667,
"step": 1087
},
{
"epoch": 0.4465421711471373,
"grad_norm": 8.76194953918457,
"learning_rate": 5.533661740558293e-06,
"loss": 0.0062,
"step": 1088
},
{
"epoch": 0.4469525959367946,
"grad_norm": 9.59299087524414,
"learning_rate": 5.529556650246306e-06,
"loss": 0.014,
"step": 1089
},
{
"epoch": 0.4473630207264519,
"grad_norm": 6.725271701812744,
"learning_rate": 5.52545155993432e-06,
"loss": 0.0038,
"step": 1090
},
{
"epoch": 0.44777344551610915,
"grad_norm": 6.48527193069458,
"learning_rate": 5.521346469622332e-06,
"loss": 0.0038,
"step": 1091
},
{
"epoch": 0.44818387030576645,
"grad_norm": 395.1700134277344,
"learning_rate": 5.517241379310345e-06,
"loss": 0.7074,
"step": 1092
},
{
"epoch": 0.44859429509542376,
"grad_norm": 2.1478097438812256,
"learning_rate": 5.513136288998359e-06,
"loss": 0.0012,
"step": 1093
},
{
"epoch": 0.44900471988508106,
"grad_norm": 10.103233337402344,
"learning_rate": 5.509031198686372e-06,
"loss": 0.0096,
"step": 1094
},
{
"epoch": 0.44941514467473836,
"grad_norm": 159.85614013671875,
"learning_rate": 5.504926108374384e-06,
"loss": 0.4152,
"step": 1095
},
{
"epoch": 0.44982556946439567,
"grad_norm": 12.248689651489258,
"learning_rate": 5.500821018062398e-06,
"loss": 0.0142,
"step": 1096
},
{
"epoch": 0.4502359942540529,
"grad_norm": 14.340502738952637,
"learning_rate": 5.496715927750411e-06,
"loss": 0.0155,
"step": 1097
},
{
"epoch": 0.4506464190437102,
"grad_norm": 10.734267234802246,
"learning_rate": 5.492610837438424e-06,
"loss": 0.0083,
"step": 1098
},
{
"epoch": 0.4510568438333675,
"grad_norm": 2.9998457431793213,
"learning_rate": 5.488505747126437e-06,
"loss": 0.0021,
"step": 1099
},
{
"epoch": 0.45146726862302483,
"grad_norm": 5.331489562988281,
"learning_rate": 5.48440065681445e-06,
"loss": 0.0063,
"step": 1100
},
{
"epoch": 0.45187769341268214,
"grad_norm": 3.9620110988616943,
"learning_rate": 5.480295566502463e-06,
"loss": 0.002,
"step": 1101
},
{
"epoch": 0.45228811820233944,
"grad_norm": 56.58110046386719,
"learning_rate": 5.476190476190477e-06,
"loss": 0.4018,
"step": 1102
},
{
"epoch": 0.45269854299199674,
"grad_norm": 118.57501983642578,
"learning_rate": 5.4720853858784895e-06,
"loss": 0.3027,
"step": 1103
},
{
"epoch": 0.453108967781654,
"grad_norm": 3.6158645153045654,
"learning_rate": 5.467980295566503e-06,
"loss": 0.002,
"step": 1104
},
{
"epoch": 0.4535193925713113,
"grad_norm": 7.839849948883057,
"learning_rate": 5.4638752052545165e-06,
"loss": 0.0063,
"step": 1105
},
{
"epoch": 0.4539298173609686,
"grad_norm": 5.707149028778076,
"learning_rate": 5.45977011494253e-06,
"loss": 0.011,
"step": 1106
},
{
"epoch": 0.4543402421506259,
"grad_norm": 6.764120578765869,
"learning_rate": 5.455665024630542e-06,
"loss": 0.0072,
"step": 1107
},
{
"epoch": 0.4547506669402832,
"grad_norm": 8.733452796936035,
"learning_rate": 5.451559934318556e-06,
"loss": 0.0051,
"step": 1108
},
{
"epoch": 0.4551610917299405,
"grad_norm": 28.740921020507812,
"learning_rate": 5.447454844006569e-06,
"loss": 0.0447,
"step": 1109
},
{
"epoch": 0.45557151651959776,
"grad_norm": 13.214491844177246,
"learning_rate": 5.443349753694581e-06,
"loss": 0.0187,
"step": 1110
},
{
"epoch": 0.45598194130925507,
"grad_norm": 59.74604034423828,
"learning_rate": 5.439244663382595e-06,
"loss": 0.4866,
"step": 1111
},
{
"epoch": 0.45639236609891237,
"grad_norm": 157.1139678955078,
"learning_rate": 5.435139573070608e-06,
"loss": 0.5574,
"step": 1112
},
{
"epoch": 0.4568027908885697,
"grad_norm": 8.195821762084961,
"learning_rate": 5.431034482758621e-06,
"loss": 0.0077,
"step": 1113
},
{
"epoch": 0.457213215678227,
"grad_norm": 82.6183853149414,
"learning_rate": 5.426929392446634e-06,
"loss": 0.2476,
"step": 1114
},
{
"epoch": 0.4576236404678843,
"grad_norm": 3.9289438724517822,
"learning_rate": 5.422824302134647e-06,
"loss": 0.0024,
"step": 1115
},
{
"epoch": 0.45803406525754153,
"grad_norm": 32.800838470458984,
"learning_rate": 5.41871921182266e-06,
"loss": 0.0604,
"step": 1116
},
{
"epoch": 0.45844449004719884,
"grad_norm": 1.2474287748336792,
"learning_rate": 5.414614121510674e-06,
"loss": 0.0014,
"step": 1117
},
{
"epoch": 0.45885491483685614,
"grad_norm": 60.06214141845703,
"learning_rate": 5.410509031198686e-06,
"loss": 0.3188,
"step": 1118
},
{
"epoch": 0.45926533962651345,
"grad_norm": 14.512068748474121,
"learning_rate": 5.406403940886699e-06,
"loss": 0.0163,
"step": 1119
},
{
"epoch": 0.45967576441617075,
"grad_norm": 47.16853713989258,
"learning_rate": 5.402298850574713e-06,
"loss": 0.4529,
"step": 1120
},
{
"epoch": 0.46008618920582806,
"grad_norm": 28.543357849121094,
"learning_rate": 5.398193760262726e-06,
"loss": 0.0552,
"step": 1121
},
{
"epoch": 0.4604966139954853,
"grad_norm": 32.87594985961914,
"learning_rate": 5.394088669950739e-06,
"loss": 0.048,
"step": 1122
},
{
"epoch": 0.4609070387851426,
"grad_norm": 37.48674392700195,
"learning_rate": 5.3899835796387525e-06,
"loss": 0.1185,
"step": 1123
},
{
"epoch": 0.4613174635747999,
"grad_norm": 17.515987396240234,
"learning_rate": 5.385878489326766e-06,
"loss": 0.0238,
"step": 1124
},
{
"epoch": 0.4617278883644572,
"grad_norm": 20.690622329711914,
"learning_rate": 5.3817733990147795e-06,
"loss": 0.0356,
"step": 1125
},
{
"epoch": 0.4621383131541145,
"grad_norm": 56.95072937011719,
"learning_rate": 5.377668308702792e-06,
"loss": 0.0952,
"step": 1126
},
{
"epoch": 0.4625487379437718,
"grad_norm": 5.704539775848389,
"learning_rate": 5.373563218390805e-06,
"loss": 0.0067,
"step": 1127
},
{
"epoch": 0.4629591627334291,
"grad_norm": 3.889218807220459,
"learning_rate": 5.369458128078819e-06,
"loss": 0.0024,
"step": 1128
},
{
"epoch": 0.4633695875230864,
"grad_norm": 11.582088470458984,
"learning_rate": 5.365353037766832e-06,
"loss": 0.0123,
"step": 1129
},
{
"epoch": 0.4637800123127437,
"grad_norm": 1.7015328407287598,
"learning_rate": 5.361247947454844e-06,
"loss": 0.0032,
"step": 1130
},
{
"epoch": 0.464190437102401,
"grad_norm": 1.8252333402633667,
"learning_rate": 5.357142857142857e-06,
"loss": 0.0022,
"step": 1131
},
{
"epoch": 0.4646008618920583,
"grad_norm": 1.8149172067642212,
"learning_rate": 5.353037766830871e-06,
"loss": 0.0016,
"step": 1132
},
{
"epoch": 0.4650112866817156,
"grad_norm": 66.025634765625,
"learning_rate": 5.348932676518884e-06,
"loss": 0.2476,
"step": 1133
},
{
"epoch": 0.46542171147137285,
"grad_norm": 4.145283222198486,
"learning_rate": 5.344827586206896e-06,
"loss": 0.0068,
"step": 1134
},
{
"epoch": 0.46583213626103015,
"grad_norm": 12.98816967010498,
"learning_rate": 5.34072249589491e-06,
"loss": 0.0115,
"step": 1135
},
{
"epoch": 0.46624256105068745,
"grad_norm": 107.0401611328125,
"learning_rate": 5.336617405582923e-06,
"loss": 0.4033,
"step": 1136
},
{
"epoch": 0.46665298584034476,
"grad_norm": 12.25761604309082,
"learning_rate": 5.332512315270937e-06,
"loss": 0.0171,
"step": 1137
},
{
"epoch": 0.46706341063000206,
"grad_norm": 32.78744125366211,
"learning_rate": 5.328407224958949e-06,
"loss": 0.0321,
"step": 1138
},
{
"epoch": 0.46747383541965937,
"grad_norm": 1.2796099185943604,
"learning_rate": 5.324302134646962e-06,
"loss": 0.002,
"step": 1139
},
{
"epoch": 0.4678842602093166,
"grad_norm": 11.587821960449219,
"learning_rate": 5.320197044334976e-06,
"loss": 0.0103,
"step": 1140
},
{
"epoch": 0.4682946849989739,
"grad_norm": 3.9461026191711426,
"learning_rate": 5.316091954022989e-06,
"loss": 0.0032,
"step": 1141
},
{
"epoch": 0.4687051097886312,
"grad_norm": 10.535643577575684,
"learning_rate": 5.311986863711002e-06,
"loss": 0.0123,
"step": 1142
},
{
"epoch": 0.46911553457828853,
"grad_norm": 2.1246049404144287,
"learning_rate": 5.3078817733990155e-06,
"loss": 0.0052,
"step": 1143
},
{
"epoch": 0.46952595936794583,
"grad_norm": 6.683266639709473,
"learning_rate": 5.3037766830870286e-06,
"loss": 0.01,
"step": 1144
},
{
"epoch": 0.46993638415760314,
"grad_norm": 4.963865756988525,
"learning_rate": 5.299671592775042e-06,
"loss": 0.0042,
"step": 1145
},
{
"epoch": 0.47034680894726044,
"grad_norm": 21.160747528076172,
"learning_rate": 5.295566502463054e-06,
"loss": 0.0416,
"step": 1146
},
{
"epoch": 0.4707572337369177,
"grad_norm": 45.63813781738281,
"learning_rate": 5.291461412151068e-06,
"loss": 0.0906,
"step": 1147
},
{
"epoch": 0.471167658526575,
"grad_norm": 10.82851791381836,
"learning_rate": 5.287356321839081e-06,
"loss": 0.0155,
"step": 1148
},
{
"epoch": 0.4715780833162323,
"grad_norm": 3.084293842315674,
"learning_rate": 5.283251231527095e-06,
"loss": 0.0017,
"step": 1149
},
{
"epoch": 0.4719885081058896,
"grad_norm": 59.54617691040039,
"learning_rate": 5.279146141215107e-06,
"loss": 0.5199,
"step": 1150
},
{
"epoch": 0.4723989328955469,
"grad_norm": 6.270060062408447,
"learning_rate": 5.27504105090312e-06,
"loss": 0.0047,
"step": 1151
},
{
"epoch": 0.4728093576852042,
"grad_norm": 2.8544931411743164,
"learning_rate": 5.270935960591134e-06,
"loss": 0.0032,
"step": 1152
},
{
"epoch": 0.47321978247486146,
"grad_norm": 11.794116973876953,
"learning_rate": 5.266830870279147e-06,
"loss": 0.0107,
"step": 1153
},
{
"epoch": 0.47363020726451877,
"grad_norm": 5.781325340270996,
"learning_rate": 5.262725779967159e-06,
"loss": 0.0041,
"step": 1154
},
{
"epoch": 0.47404063205417607,
"grad_norm": 17.149524688720703,
"learning_rate": 5.258620689655173e-06,
"loss": 0.0225,
"step": 1155
},
{
"epoch": 0.4744510568438334,
"grad_norm": 12.543281555175781,
"learning_rate": 5.254515599343186e-06,
"loss": 0.0213,
"step": 1156
},
{
"epoch": 0.4748614816334907,
"grad_norm": 7.4617509841918945,
"learning_rate": 5.250410509031199e-06,
"loss": 0.0045,
"step": 1157
},
{
"epoch": 0.475271906423148,
"grad_norm": 111.67198181152344,
"learning_rate": 5.246305418719212e-06,
"loss": 0.556,
"step": 1158
},
{
"epoch": 0.47568233121280523,
"grad_norm": 3.6835389137268066,
"learning_rate": 5.242200328407225e-06,
"loss": 0.0085,
"step": 1159
},
{
"epoch": 0.47609275600246254,
"grad_norm": 2.393995523452759,
"learning_rate": 5.2380952380952384e-06,
"loss": 0.0027,
"step": 1160
},
{
"epoch": 0.47650318079211984,
"grad_norm": 9.624814987182617,
"learning_rate": 5.233990147783252e-06,
"loss": 0.022,
"step": 1161
},
{
"epoch": 0.47691360558177714,
"grad_norm": 16.023832321166992,
"learning_rate": 5.2298850574712646e-06,
"loss": 0.0194,
"step": 1162
},
{
"epoch": 0.47732403037143445,
"grad_norm": 7.5904974937438965,
"learning_rate": 5.225779967159278e-06,
"loss": 0.0057,
"step": 1163
},
{
"epoch": 0.47773445516109175,
"grad_norm": 4.885858058929443,
"learning_rate": 5.2216748768472915e-06,
"loss": 0.0051,
"step": 1164
},
{
"epoch": 0.478144879950749,
"grad_norm": 1.347660779953003,
"learning_rate": 5.217569786535304e-06,
"loss": 0.0024,
"step": 1165
},
{
"epoch": 0.4785553047404063,
"grad_norm": 17.13306427001953,
"learning_rate": 5.213464696223317e-06,
"loss": 0.0139,
"step": 1166
},
{
"epoch": 0.4789657295300636,
"grad_norm": 7.861404895782471,
"learning_rate": 5.209359605911331e-06,
"loss": 0.0074,
"step": 1167
},
{
"epoch": 0.4793761543197209,
"grad_norm": 1.0164753198623657,
"learning_rate": 5.205254515599344e-06,
"loss": 0.0016,
"step": 1168
},
{
"epoch": 0.4797865791093782,
"grad_norm": 1.228754997253418,
"learning_rate": 5.201149425287356e-06,
"loss": 0.0011,
"step": 1169
},
{
"epoch": 0.4801970038990355,
"grad_norm": 299.19171142578125,
"learning_rate": 5.19704433497537e-06,
"loss": 0.1483,
"step": 1170
},
{
"epoch": 0.4806074286886928,
"grad_norm": 80.7892837524414,
"learning_rate": 5.192939244663383e-06,
"loss": 0.1964,
"step": 1171
},
{
"epoch": 0.4810178534783501,
"grad_norm": 0.5468462705612183,
"learning_rate": 5.188834154351396e-06,
"loss": 0.0007,
"step": 1172
},
{
"epoch": 0.4814282782680074,
"grad_norm": 5.169793128967285,
"learning_rate": 5.184729064039409e-06,
"loss": 0.0022,
"step": 1173
},
{
"epoch": 0.4818387030576647,
"grad_norm": 12.510892868041992,
"learning_rate": 5.180623973727422e-06,
"loss": 0.0132,
"step": 1174
},
{
"epoch": 0.482249127847322,
"grad_norm": 1.1645554304122925,
"learning_rate": 5.176518883415435e-06,
"loss": 0.001,
"step": 1175
},
{
"epoch": 0.4826595526369793,
"grad_norm": 3.121722936630249,
"learning_rate": 5.172413793103449e-06,
"loss": 0.002,
"step": 1176
},
{
"epoch": 0.48306997742663654,
"grad_norm": 2.8475632667541504,
"learning_rate": 5.168308702791461e-06,
"loss": 0.0024,
"step": 1177
},
{
"epoch": 0.48348040221629385,
"grad_norm": 10.669745445251465,
"learning_rate": 5.1642036124794744e-06,
"loss": 0.0084,
"step": 1178
},
{
"epoch": 0.48389082700595115,
"grad_norm": 3.702817440032959,
"learning_rate": 5.160098522167488e-06,
"loss": 0.0063,
"step": 1179
},
{
"epoch": 0.48430125179560846,
"grad_norm": 4.249784469604492,
"learning_rate": 5.1559934318555014e-06,
"loss": 0.0053,
"step": 1180
},
{
"epoch": 0.48471167658526576,
"grad_norm": 0.7857574224472046,
"learning_rate": 5.151888341543514e-06,
"loss": 0.001,
"step": 1181
},
{
"epoch": 0.48512210137492306,
"grad_norm": 2.24009108543396,
"learning_rate": 5.1477832512315276e-06,
"loss": 0.0012,
"step": 1182
},
{
"epoch": 0.4855325261645803,
"grad_norm": 6.223039150238037,
"learning_rate": 5.143678160919541e-06,
"loss": 0.0127,
"step": 1183
},
{
"epoch": 0.4859429509542376,
"grad_norm": 6.224484443664551,
"learning_rate": 5.1395730706075545e-06,
"loss": 0.0029,
"step": 1184
},
{
"epoch": 0.4863533757438949,
"grad_norm": 7.365463733673096,
"learning_rate": 5.135467980295567e-06,
"loss": 0.0066,
"step": 1185
},
{
"epoch": 0.4867638005335522,
"grad_norm": 3.9487838745117188,
"learning_rate": 5.13136288998358e-06,
"loss": 0.0023,
"step": 1186
},
{
"epoch": 0.48717422532320953,
"grad_norm": 0.860590934753418,
"learning_rate": 5.127257799671594e-06,
"loss": 0.0009,
"step": 1187
},
{
"epoch": 0.48758465011286684,
"grad_norm": 9.55481243133545,
"learning_rate": 5.123152709359607e-06,
"loss": 0.012,
"step": 1188
},
{
"epoch": 0.48799507490252414,
"grad_norm": 1.8246153593063354,
"learning_rate": 5.119047619047619e-06,
"loss": 0.0038,
"step": 1189
},
{
"epoch": 0.4884054996921814,
"grad_norm": 8.474342346191406,
"learning_rate": 5.114942528735632e-06,
"loss": 0.0094,
"step": 1190
},
{
"epoch": 0.4888159244818387,
"grad_norm": 2.81972599029541,
"learning_rate": 5.110837438423646e-06,
"loss": 0.0057,
"step": 1191
},
{
"epoch": 0.489226349271496,
"grad_norm": 168.85972595214844,
"learning_rate": 5.106732348111659e-06,
"loss": 0.7816,
"step": 1192
},
{
"epoch": 0.4896367740611533,
"grad_norm": 8.733819007873535,
"learning_rate": 5.102627257799671e-06,
"loss": 0.0107,
"step": 1193
},
{
"epoch": 0.4900471988508106,
"grad_norm": 14.976457595825195,
"learning_rate": 5.098522167487685e-06,
"loss": 0.0183,
"step": 1194
},
{
"epoch": 0.4904576236404679,
"grad_norm": 117.59420013427734,
"learning_rate": 5.094417077175698e-06,
"loss": 0.3754,
"step": 1195
},
{
"epoch": 0.49086804843012516,
"grad_norm": 6.668210506439209,
"learning_rate": 5.090311986863712e-06,
"loss": 0.0034,
"step": 1196
},
{
"epoch": 0.49127847321978246,
"grad_norm": 8.601287841796875,
"learning_rate": 5.086206896551724e-06,
"loss": 0.0052,
"step": 1197
},
{
"epoch": 0.49168889800943977,
"grad_norm": 4.39802360534668,
"learning_rate": 5.0821018062397374e-06,
"loss": 0.0025,
"step": 1198
},
{
"epoch": 0.49209932279909707,
"grad_norm": 41.07631301879883,
"learning_rate": 5.077996715927751e-06,
"loss": 0.2194,
"step": 1199
},
{
"epoch": 0.4925097475887544,
"grad_norm": 9.858050346374512,
"learning_rate": 5.073891625615764e-06,
"loss": 0.0079,
"step": 1200
},
{
"epoch": 0.4929201723784117,
"grad_norm": 0.21371275186538696,
"learning_rate": 5.069786535303777e-06,
"loss": 0.0004,
"step": 1201
},
{
"epoch": 0.49333059716806893,
"grad_norm": 5.991517543792725,
"learning_rate": 5.0656814449917905e-06,
"loss": 0.0033,
"step": 1202
},
{
"epoch": 0.49374102195772623,
"grad_norm": 4.217563629150391,
"learning_rate": 5.061576354679804e-06,
"loss": 0.0061,
"step": 1203
},
{
"epoch": 0.49415144674738354,
"grad_norm": 3.594877004623413,
"learning_rate": 5.057471264367817e-06,
"loss": 0.0035,
"step": 1204
},
{
"epoch": 0.49456187153704084,
"grad_norm": 1.1519529819488525,
"learning_rate": 5.053366174055829e-06,
"loss": 0.0014,
"step": 1205
},
{
"epoch": 0.49497229632669815,
"grad_norm": 4.432569980621338,
"learning_rate": 5.049261083743843e-06,
"loss": 0.0017,
"step": 1206
},
{
"epoch": 0.49538272111635545,
"grad_norm": 1.235167145729065,
"learning_rate": 5.045155993431856e-06,
"loss": 0.0009,
"step": 1207
},
{
"epoch": 0.4957931459060127,
"grad_norm": 62.06901931762695,
"learning_rate": 5.04105090311987e-06,
"loss": 0.5126,
"step": 1208
},
{
"epoch": 0.49620357069567,
"grad_norm": 3.39192271232605,
"learning_rate": 5.036945812807882e-06,
"loss": 0.0034,
"step": 1209
},
{
"epoch": 0.4966139954853273,
"grad_norm": 56.114288330078125,
"learning_rate": 5.032840722495895e-06,
"loss": 0.5337,
"step": 1210
},
{
"epoch": 0.4970244202749846,
"grad_norm": 118.14190673828125,
"learning_rate": 5.028735632183909e-06,
"loss": 0.3707,
"step": 1211
},
{
"epoch": 0.4974348450646419,
"grad_norm": 8.5828275680542,
"learning_rate": 5.024630541871922e-06,
"loss": 0.0086,
"step": 1212
},
{
"epoch": 0.4978452698542992,
"grad_norm": 7.391578674316406,
"learning_rate": 5.020525451559934e-06,
"loss": 0.0071,
"step": 1213
},
{
"epoch": 0.49825569464395647,
"grad_norm": 8.93129825592041,
"learning_rate": 5.016420361247948e-06,
"loss": 0.0081,
"step": 1214
},
{
"epoch": 0.4986661194336138,
"grad_norm": 12.378411293029785,
"learning_rate": 5.012315270935961e-06,
"loss": 0.0104,
"step": 1215
},
{
"epoch": 0.4990765442232711,
"grad_norm": 8.039668083190918,
"learning_rate": 5.008210180623974e-06,
"loss": 0.006,
"step": 1216
},
{
"epoch": 0.4994869690129284,
"grad_norm": 6.898679256439209,
"learning_rate": 5.004105090311987e-06,
"loss": 0.0054,
"step": 1217
},
{
"epoch": 0.4998973938025857,
"grad_norm": 151.4322967529297,
"learning_rate": 5e-06,
"loss": 0.2748,
"step": 1218
},
{
"epoch": 0.5003078185922429,
"grad_norm": 97.9362564086914,
"learning_rate": 4.9958949096880135e-06,
"loss": 0.1818,
"step": 1219
},
{
"epoch": 0.5007182433819003,
"grad_norm": 4.9238128662109375,
"learning_rate": 4.9917898193760265e-06,
"loss": 0.0038,
"step": 1220
},
{
"epoch": 0.5011286681715575,
"grad_norm": 3.835169792175293,
"learning_rate": 4.98768472906404e-06,
"loss": 0.0023,
"step": 1221
},
{
"epoch": 0.5015390929612149,
"grad_norm": 5.244224548339844,
"learning_rate": 4.983579638752053e-06,
"loss": 0.0023,
"step": 1222
},
{
"epoch": 0.5019495177508722,
"grad_norm": 11.275529861450195,
"learning_rate": 4.979474548440066e-06,
"loss": 0.0091,
"step": 1223
},
{
"epoch": 0.5023599425405294,
"grad_norm": 14.805060386657715,
"learning_rate": 4.97536945812808e-06,
"loss": 0.018,
"step": 1224
},
{
"epoch": 0.5027703673301868,
"grad_norm": 136.29385375976562,
"learning_rate": 4.971264367816092e-06,
"loss": 0.6029,
"step": 1225
},
{
"epoch": 0.503180792119844,
"grad_norm": 12.416361808776855,
"learning_rate": 4.967159277504106e-06,
"loss": 0.0166,
"step": 1226
},
{
"epoch": 0.5035912169095014,
"grad_norm": 64.48587799072266,
"learning_rate": 4.963054187192118e-06,
"loss": 0.0532,
"step": 1227
},
{
"epoch": 0.5040016416991586,
"grad_norm": 9.685049057006836,
"learning_rate": 4.958949096880132e-06,
"loss": 0.012,
"step": 1228
},
{
"epoch": 0.504412066488816,
"grad_norm": 9.604591369628906,
"learning_rate": 4.954844006568145e-06,
"loss": 0.0072,
"step": 1229
},
{
"epoch": 0.5048224912784732,
"grad_norm": 9.923593521118164,
"learning_rate": 4.950738916256158e-06,
"loss": 0.0089,
"step": 1230
},
{
"epoch": 0.5052329160681305,
"grad_norm": 2.9217567443847656,
"learning_rate": 4.946633825944171e-06,
"loss": 0.0045,
"step": 1231
},
{
"epoch": 0.5056433408577878,
"grad_norm": 4.328164577484131,
"learning_rate": 4.942528735632184e-06,
"loss": 0.005,
"step": 1232
},
{
"epoch": 0.5060537656474451,
"grad_norm": 6.946155548095703,
"learning_rate": 4.938423645320197e-06,
"loss": 0.006,
"step": 1233
},
{
"epoch": 0.5064641904371024,
"grad_norm": 8.665051460266113,
"learning_rate": 4.93431855500821e-06,
"loss": 0.0062,
"step": 1234
},
{
"epoch": 0.5068746152267597,
"grad_norm": 1.1814903020858765,
"learning_rate": 4.930213464696223e-06,
"loss": 0.0007,
"step": 1235
},
{
"epoch": 0.5072850400164169,
"grad_norm": 5.667411804199219,
"learning_rate": 4.926108374384237e-06,
"loss": 0.0055,
"step": 1236
},
{
"epoch": 0.5076954648060743,
"grad_norm": 7.802727222442627,
"learning_rate": 4.9220032840722495e-06,
"loss": 0.0077,
"step": 1237
},
{
"epoch": 0.5081058895957316,
"grad_norm": 199.56512451171875,
"learning_rate": 4.917898193760263e-06,
"loss": 0.1608,
"step": 1238
},
{
"epoch": 0.5085163143853889,
"grad_norm": 6.478273391723633,
"learning_rate": 4.9137931034482765e-06,
"loss": 0.015,
"step": 1239
},
{
"epoch": 0.5089267391750462,
"grad_norm": 11.63586139678955,
"learning_rate": 4.9096880131362895e-06,
"loss": 0.0189,
"step": 1240
},
{
"epoch": 0.5093371639647035,
"grad_norm": 7.227424621582031,
"learning_rate": 4.905582922824303e-06,
"loss": 0.0035,
"step": 1241
},
{
"epoch": 0.5097475887543608,
"grad_norm": 3.1947948932647705,
"learning_rate": 4.901477832512316e-06,
"loss": 0.0026,
"step": 1242
},
{
"epoch": 0.510158013544018,
"grad_norm": 13.523475646972656,
"learning_rate": 4.897372742200329e-06,
"loss": 0.0205,
"step": 1243
},
{
"epoch": 0.5105684383336754,
"grad_norm": 215.79766845703125,
"learning_rate": 4.893267651888342e-06,
"loss": 0.6973,
"step": 1244
},
{
"epoch": 0.5109788631233326,
"grad_norm": 7.183282375335693,
"learning_rate": 4.889162561576355e-06,
"loss": 0.0044,
"step": 1245
},
{
"epoch": 0.51138928791299,
"grad_norm": 3.569234848022461,
"learning_rate": 4.885057471264369e-06,
"loss": 0.0027,
"step": 1246
},
{
"epoch": 0.5117997127026472,
"grad_norm": 153.31161499023438,
"learning_rate": 4.880952380952381e-06,
"loss": 0.4229,
"step": 1247
},
{
"epoch": 0.5122101374923045,
"grad_norm": 12.08935546875,
"learning_rate": 4.876847290640395e-06,
"loss": 0.0123,
"step": 1248
},
{
"epoch": 0.5126205622819618,
"grad_norm": 3.3031253814697266,
"learning_rate": 4.872742200328407e-06,
"loss": 0.0045,
"step": 1249
},
{
"epoch": 0.5130309870716191,
"grad_norm": 8.546093940734863,
"learning_rate": 4.868637110016421e-06,
"loss": 0.0078,
"step": 1250
},
{
"epoch": 0.5134414118612765,
"grad_norm": 9.880775451660156,
"learning_rate": 4.864532019704434e-06,
"loss": 0.0076,
"step": 1251
},
{
"epoch": 0.5138518366509337,
"grad_norm": 2.1825859546661377,
"learning_rate": 4.860426929392447e-06,
"loss": 0.0011,
"step": 1252
},
{
"epoch": 0.5142622614405911,
"grad_norm": 39.055728912353516,
"learning_rate": 4.85632183908046e-06,
"loss": 0.0313,
"step": 1253
},
{
"epoch": 0.5146726862302483,
"grad_norm": 2.0506696701049805,
"learning_rate": 4.852216748768473e-06,
"loss": 0.0023,
"step": 1254
},
{
"epoch": 0.5150831110199056,
"grad_norm": 6.520775318145752,
"learning_rate": 4.848111658456486e-06,
"loss": 0.0035,
"step": 1255
},
{
"epoch": 0.5154935358095629,
"grad_norm": 1.5494943857192993,
"learning_rate": 4.844006568144499e-06,
"loss": 0.0013,
"step": 1256
},
{
"epoch": 0.5159039605992202,
"grad_norm": 619.266357421875,
"learning_rate": 4.8399014778325125e-06,
"loss": 1.0242,
"step": 1257
},
{
"epoch": 0.5163143853888775,
"grad_norm": 151.16917419433594,
"learning_rate": 4.835796387520526e-06,
"loss": 0.2408,
"step": 1258
},
{
"epoch": 0.5167248101785348,
"grad_norm": 2.2832939624786377,
"learning_rate": 4.831691297208539e-06,
"loss": 0.0043,
"step": 1259
},
{
"epoch": 0.517135234968192,
"grad_norm": 8.788455963134766,
"learning_rate": 4.8275862068965525e-06,
"loss": 0.0093,
"step": 1260
},
{
"epoch": 0.5175456597578494,
"grad_norm": 99.35799407958984,
"learning_rate": 4.8234811165845656e-06,
"loss": 0.1958,
"step": 1261
},
{
"epoch": 0.5179560845475066,
"grad_norm": 6.032120704650879,
"learning_rate": 4.819376026272579e-06,
"loss": 0.0034,
"step": 1262
},
{
"epoch": 0.518366509337164,
"grad_norm": 5.858295917510986,
"learning_rate": 4.815270935960592e-06,
"loss": 0.0043,
"step": 1263
},
{
"epoch": 0.5187769341268212,
"grad_norm": 7.7139081954956055,
"learning_rate": 4.811165845648605e-06,
"loss": 0.0128,
"step": 1264
},
{
"epoch": 0.5191873589164786,
"grad_norm": 8.401812553405762,
"learning_rate": 4.807060755336618e-06,
"loss": 0.0061,
"step": 1265
},
{
"epoch": 0.5195977837061359,
"grad_norm": 8.46750259399414,
"learning_rate": 4.802955665024631e-06,
"loss": 0.0075,
"step": 1266
},
{
"epoch": 0.5200082084957931,
"grad_norm": 6.227928161621094,
"learning_rate": 4.798850574712644e-06,
"loss": 0.0042,
"step": 1267
},
{
"epoch": 0.5204186332854505,
"grad_norm": 65.28329467773438,
"learning_rate": 4.794745484400658e-06,
"loss": 0.5747,
"step": 1268
},
{
"epoch": 0.5208290580751077,
"grad_norm": 49.62480163574219,
"learning_rate": 4.79064039408867e-06,
"loss": 0.4586,
"step": 1269
},
{
"epoch": 0.5212394828647651,
"grad_norm": 69.17855072021484,
"learning_rate": 4.786535303776684e-06,
"loss": 0.5167,
"step": 1270
},
{
"epoch": 0.5216499076544223,
"grad_norm": 11.419864654541016,
"learning_rate": 4.782430213464696e-06,
"loss": 0.0144,
"step": 1271
},
{
"epoch": 0.5220603324440797,
"grad_norm": 12.840147972106934,
"learning_rate": 4.77832512315271e-06,
"loss": 0.0117,
"step": 1272
},
{
"epoch": 0.5224707572337369,
"grad_norm": 24.375146865844727,
"learning_rate": 4.774220032840723e-06,
"loss": 0.0743,
"step": 1273
},
{
"epoch": 0.5228811820233942,
"grad_norm": 117.50786590576172,
"learning_rate": 4.770114942528735e-06,
"loss": 0.2646,
"step": 1274
},
{
"epoch": 0.5232916068130515,
"grad_norm": 15.543289184570312,
"learning_rate": 4.766009852216749e-06,
"loss": 0.0155,
"step": 1275
},
{
"epoch": 0.5237020316027088,
"grad_norm": 4.836006164550781,
"learning_rate": 4.761904761904762e-06,
"loss": 0.0027,
"step": 1276
},
{
"epoch": 0.5241124563923661,
"grad_norm": 3.3769748210906982,
"learning_rate": 4.7577996715927754e-06,
"loss": 0.0025,
"step": 1277
},
{
"epoch": 0.5245228811820234,
"grad_norm": 1.2123010158538818,
"learning_rate": 4.7536945812807885e-06,
"loss": 0.0008,
"step": 1278
},
{
"epoch": 0.5249333059716806,
"grad_norm": 10.587141036987305,
"learning_rate": 4.7495894909688016e-06,
"loss": 0.0107,
"step": 1279
},
{
"epoch": 0.525343730761338,
"grad_norm": 0.7079842686653137,
"learning_rate": 4.745484400656815e-06,
"loss": 0.0011,
"step": 1280
},
{
"epoch": 0.5257541555509953,
"grad_norm": 6.404101371765137,
"learning_rate": 4.741379310344828e-06,
"loss": 0.0069,
"step": 1281
},
{
"epoch": 0.5261645803406526,
"grad_norm": 0.7190276384353638,
"learning_rate": 4.737274220032841e-06,
"loss": 0.0012,
"step": 1282
},
{
"epoch": 0.5265750051303099,
"grad_norm": 17.04617691040039,
"learning_rate": 4.733169129720855e-06,
"loss": 0.0109,
"step": 1283
},
{
"epoch": 0.5269854299199672,
"grad_norm": 7.678450584411621,
"learning_rate": 4.729064039408867e-06,
"loss": 0.0055,
"step": 1284
},
{
"epoch": 0.5273958547096245,
"grad_norm": 11.268597602844238,
"learning_rate": 4.724958949096881e-06,
"loss": 0.0107,
"step": 1285
},
{
"epoch": 0.5278062794992817,
"grad_norm": 15.689214706420898,
"learning_rate": 4.720853858784893e-06,
"loss": 0.0208,
"step": 1286
},
{
"epoch": 0.5282167042889391,
"grad_norm": 58.094364166259766,
"learning_rate": 4.716748768472907e-06,
"loss": 0.5902,
"step": 1287
},
{
"epoch": 0.5286271290785963,
"grad_norm": 8.045167922973633,
"learning_rate": 4.71264367816092e-06,
"loss": 0.0051,
"step": 1288
},
{
"epoch": 0.5290375538682537,
"grad_norm": 4.597346782684326,
"learning_rate": 4.708538587848933e-06,
"loss": 0.002,
"step": 1289
},
{
"epoch": 0.5294479786579109,
"grad_norm": 12.006298065185547,
"learning_rate": 4.704433497536946e-06,
"loss": 0.0114,
"step": 1290
},
{
"epoch": 0.5298584034475682,
"grad_norm": 70.83368682861328,
"learning_rate": 4.700328407224959e-06,
"loss": 0.749,
"step": 1291
},
{
"epoch": 0.5302688282372255,
"grad_norm": 8.56693172454834,
"learning_rate": 4.696223316912972e-06,
"loss": 0.0185,
"step": 1292
},
{
"epoch": 0.5306792530268828,
"grad_norm": 8.063687324523926,
"learning_rate": 4.692118226600985e-06,
"loss": 0.0083,
"step": 1293
},
{
"epoch": 0.5310896778165402,
"grad_norm": 5.791623115539551,
"learning_rate": 4.688013136288998e-06,
"loss": 0.0034,
"step": 1294
},
{
"epoch": 0.5315001026061974,
"grad_norm": 1.853096842765808,
"learning_rate": 4.683908045977012e-06,
"loss": 0.0031,
"step": 1295
},
{
"epoch": 0.5319105273958548,
"grad_norm": 4.376800537109375,
"learning_rate": 4.6798029556650245e-06,
"loss": 0.0102,
"step": 1296
},
{
"epoch": 0.532320952185512,
"grad_norm": 1.2729754447937012,
"learning_rate": 4.675697865353038e-06,
"loss": 0.0015,
"step": 1297
},
{
"epoch": 0.5327313769751693,
"grad_norm": 2.7059133052825928,
"learning_rate": 4.6715927750410515e-06,
"loss": 0.004,
"step": 1298
},
{
"epoch": 0.5331418017648266,
"grad_norm": 132.32781982421875,
"learning_rate": 4.6674876847290645e-06,
"loss": 0.1764,
"step": 1299
},
{
"epoch": 0.5335522265544839,
"grad_norm": 12.253974914550781,
"learning_rate": 4.663382594417078e-06,
"loss": 0.0149,
"step": 1300
},
{
"epoch": 0.5339626513441412,
"grad_norm": 0.4290545880794525,
"learning_rate": 4.659277504105091e-06,
"loss": 0.001,
"step": 1301
},
{
"epoch": 0.5343730761337985,
"grad_norm": 9.426305770874023,
"learning_rate": 4.655172413793104e-06,
"loss": 0.0091,
"step": 1302
},
{
"epoch": 0.5347835009234557,
"grad_norm": 2.155017375946045,
"learning_rate": 4.651067323481117e-06,
"loss": 0.0026,
"step": 1303
},
{
"epoch": 0.5351939257131131,
"grad_norm": 11.846675872802734,
"learning_rate": 4.64696223316913e-06,
"loss": 0.0201,
"step": 1304
},
{
"epoch": 0.5356043505027703,
"grad_norm": 5.391582489013672,
"learning_rate": 4.642857142857144e-06,
"loss": 0.0038,
"step": 1305
},
{
"epoch": 0.5360147752924277,
"grad_norm": 5.011373996734619,
"learning_rate": 4.638752052545156e-06,
"loss": 0.0036,
"step": 1306
},
{
"epoch": 0.5364252000820849,
"grad_norm": 1.6165841817855835,
"learning_rate": 4.63464696223317e-06,
"loss": 0.002,
"step": 1307
},
{
"epoch": 0.5368356248717423,
"grad_norm": 1.4232211112976074,
"learning_rate": 4.630541871921182e-06,
"loss": 0.0013,
"step": 1308
},
{
"epoch": 0.5372460496613995,
"grad_norm": 137.78468322753906,
"learning_rate": 4.626436781609196e-06,
"loss": 0.2918,
"step": 1309
},
{
"epoch": 0.5376564744510568,
"grad_norm": 11.368677139282227,
"learning_rate": 4.622331691297209e-06,
"loss": 0.0118,
"step": 1310
},
{
"epoch": 0.5380668992407142,
"grad_norm": 0.42143961787223816,
"learning_rate": 4.618226600985222e-06,
"loss": 0.0005,
"step": 1311
},
{
"epoch": 0.5384773240303714,
"grad_norm": 10.286303520202637,
"learning_rate": 4.614121510673235e-06,
"loss": 0.0169,
"step": 1312
},
{
"epoch": 0.5388877488200288,
"grad_norm": 7.026170253753662,
"learning_rate": 4.610016420361248e-06,
"loss": 0.0057,
"step": 1313
},
{
"epoch": 0.539298173609686,
"grad_norm": 28.311649322509766,
"learning_rate": 4.605911330049261e-06,
"loss": 0.072,
"step": 1314
},
{
"epoch": 0.5397085983993434,
"grad_norm": 22.815528869628906,
"learning_rate": 4.6018062397372744e-06,
"loss": 0.0434,
"step": 1315
},
{
"epoch": 0.5401190231890006,
"grad_norm": 2.450845241546631,
"learning_rate": 4.5977011494252875e-06,
"loss": 0.0032,
"step": 1316
},
{
"epoch": 0.5405294479786579,
"grad_norm": 6.078365802764893,
"learning_rate": 4.593596059113301e-06,
"loss": 0.0036,
"step": 1317
},
{
"epoch": 0.5409398727683152,
"grad_norm": 155.92636108398438,
"learning_rate": 4.589490968801314e-06,
"loss": 0.2072,
"step": 1318
},
{
"epoch": 0.5413502975579725,
"grad_norm": 10.98482894897461,
"learning_rate": 4.5853858784893275e-06,
"loss": 0.0125,
"step": 1319
},
{
"epoch": 0.5417607223476298,
"grad_norm": 4.186033248901367,
"learning_rate": 4.581280788177341e-06,
"loss": 0.0029,
"step": 1320
},
{
"epoch": 0.5421711471372871,
"grad_norm": 5.184717178344727,
"learning_rate": 4.577175697865354e-06,
"loss": 0.0024,
"step": 1321
},
{
"epoch": 0.5425815719269443,
"grad_norm": 27.369516372680664,
"learning_rate": 4.573070607553367e-06,
"loss": 0.0421,
"step": 1322
},
{
"epoch": 0.5429919967166017,
"grad_norm": 10.292860984802246,
"learning_rate": 4.56896551724138e-06,
"loss": 0.0091,
"step": 1323
},
{
"epoch": 0.543402421506259,
"grad_norm": 10.870683670043945,
"learning_rate": 4.564860426929393e-06,
"loss": 0.01,
"step": 1324
},
{
"epoch": 0.5438128462959163,
"grad_norm": 30.06772232055664,
"learning_rate": 4.560755336617406e-06,
"loss": 0.1249,
"step": 1325
},
{
"epoch": 0.5442232710855736,
"grad_norm": 18.459630966186523,
"learning_rate": 4.556650246305419e-06,
"loss": 0.0156,
"step": 1326
},
{
"epoch": 0.5446336958752309,
"grad_norm": 16.96495246887207,
"learning_rate": 4.552545155993433e-06,
"loss": 0.0406,
"step": 1327
},
{
"epoch": 0.5450441206648882,
"grad_norm": 17.62691307067871,
"learning_rate": 4.548440065681445e-06,
"loss": 0.0388,
"step": 1328
},
{
"epoch": 0.5454545454545454,
"grad_norm": 7.42934513092041,
"learning_rate": 4.544334975369458e-06,
"loss": 0.0073,
"step": 1329
},
{
"epoch": 0.5458649702442028,
"grad_norm": 1.3290879726409912,
"learning_rate": 4.540229885057471e-06,
"loss": 0.0013,
"step": 1330
},
{
"epoch": 0.54627539503386,
"grad_norm": 17.09844970703125,
"learning_rate": 4.536124794745484e-06,
"loss": 0.017,
"step": 1331
},
{
"epoch": 0.5466858198235174,
"grad_norm": 16.51416778564453,
"learning_rate": 4.532019704433498e-06,
"loss": 0.0212,
"step": 1332
},
{
"epoch": 0.5470962446131746,
"grad_norm": 22.115854263305664,
"learning_rate": 4.5279146141215104e-06,
"loss": 0.0382,
"step": 1333
},
{
"epoch": 0.5475066694028319,
"grad_norm": 2.7200286388397217,
"learning_rate": 4.523809523809524e-06,
"loss": 0.0035,
"step": 1334
},
{
"epoch": 0.5479170941924892,
"grad_norm": 5.75656270980835,
"learning_rate": 4.519704433497537e-06,
"loss": 0.0053,
"step": 1335
},
{
"epoch": 0.5483275189821465,
"grad_norm": 2.9613118171691895,
"learning_rate": 4.5155993431855505e-06,
"loss": 0.0017,
"step": 1336
},
{
"epoch": 0.5487379437718038,
"grad_norm": 1.270560622215271,
"learning_rate": 4.5114942528735635e-06,
"loss": 0.0007,
"step": 1337
},
{
"epoch": 0.5491483685614611,
"grad_norm": 94.896484375,
"learning_rate": 4.507389162561577e-06,
"loss": 0.1754,
"step": 1338
},
{
"epoch": 0.5495587933511185,
"grad_norm": 4.370944023132324,
"learning_rate": 4.50328407224959e-06,
"loss": 0.0054,
"step": 1339
},
{
"epoch": 0.5499692181407757,
"grad_norm": 1.269740104675293,
"learning_rate": 4.499178981937603e-06,
"loss": 0.0012,
"step": 1340
},
{
"epoch": 0.550379642930433,
"grad_norm": 1.720219373703003,
"learning_rate": 4.495073891625616e-06,
"loss": 0.0007,
"step": 1341
},
{
"epoch": 0.5507900677200903,
"grad_norm": 1.6981797218322754,
"learning_rate": 4.49096880131363e-06,
"loss": 0.0024,
"step": 1342
},
{
"epoch": 0.5512004925097476,
"grad_norm": 6.375542640686035,
"learning_rate": 4.486863711001642e-06,
"loss": 0.0033,
"step": 1343
},
{
"epoch": 0.5516109172994049,
"grad_norm": 2.7067935466766357,
"learning_rate": 4.482758620689656e-06,
"loss": 0.0028,
"step": 1344
},
{
"epoch": 0.5520213420890622,
"grad_norm": 110.53211212158203,
"learning_rate": 4.478653530377669e-06,
"loss": 0.2874,
"step": 1345
},
{
"epoch": 0.5524317668787194,
"grad_norm": 1.7498997449874878,
"learning_rate": 4.474548440065682e-06,
"loss": 0.0023,
"step": 1346
},
{
"epoch": 0.5528421916683768,
"grad_norm": 1.027415156364441,
"learning_rate": 4.470443349753695e-06,
"loss": 0.0016,
"step": 1347
},
{
"epoch": 0.553252616458034,
"grad_norm": 2.513089895248413,
"learning_rate": 4.466338259441708e-06,
"loss": 0.0015,
"step": 1348
},
{
"epoch": 0.5536630412476914,
"grad_norm": 10.082626342773438,
"learning_rate": 4.462233169129721e-06,
"loss": 0.0081,
"step": 1349
},
{
"epoch": 0.5540734660373486,
"grad_norm": 50.76285171508789,
"learning_rate": 4.458128078817734e-06,
"loss": 0.5707,
"step": 1350
},
{
"epoch": 0.554483890827006,
"grad_norm": 74.11210632324219,
"learning_rate": 4.454022988505747e-06,
"loss": 0.075,
"step": 1351
},
{
"epoch": 0.5548943156166632,
"grad_norm": 9.972468376159668,
"learning_rate": 4.44991789819376e-06,
"loss": 0.0073,
"step": 1352
},
{
"epoch": 0.5553047404063205,
"grad_norm": 4.967834949493408,
"learning_rate": 4.445812807881773e-06,
"loss": 0.0021,
"step": 1353
},
{
"epoch": 0.5557151651959779,
"grad_norm": 2.347085952758789,
"learning_rate": 4.441707717569787e-06,
"loss": 0.0009,
"step": 1354
},
{
"epoch": 0.5561255899856351,
"grad_norm": 6.57852840423584,
"learning_rate": 4.4376026272577995e-06,
"loss": 0.0035,
"step": 1355
},
{
"epoch": 0.5565360147752925,
"grad_norm": 8.229926109313965,
"learning_rate": 4.4334975369458135e-06,
"loss": 0.0048,
"step": 1356
},
{
"epoch": 0.5569464395649497,
"grad_norm": 44.55419158935547,
"learning_rate": 4.4293924466338265e-06,
"loss": 0.5341,
"step": 1357
},
{
"epoch": 0.5573568643546071,
"grad_norm": 0.6986808776855469,
"learning_rate": 4.42528735632184e-06,
"loss": 0.0014,
"step": 1358
},
{
"epoch": 0.5577672891442643,
"grad_norm": 8.169197082519531,
"learning_rate": 4.421182266009853e-06,
"loss": 0.005,
"step": 1359
},
{
"epoch": 0.5581777139339216,
"grad_norm": 3.2294626235961914,
"learning_rate": 4.417077175697866e-06,
"loss": 0.0021,
"step": 1360
},
{
"epoch": 0.5585881387235789,
"grad_norm": 7.880626201629639,
"learning_rate": 4.412972085385879e-06,
"loss": 0.0289,
"step": 1361
},
{
"epoch": 0.5589985635132362,
"grad_norm": 1.5473792552947998,
"learning_rate": 4.408866995073892e-06,
"loss": 0.0012,
"step": 1362
},
{
"epoch": 0.5594089883028935,
"grad_norm": 2.111675500869751,
"learning_rate": 4.404761904761905e-06,
"loss": 0.0014,
"step": 1363
},
{
"epoch": 0.5598194130925508,
"grad_norm": 48.12710952758789,
"learning_rate": 4.400656814449919e-06,
"loss": 0.0657,
"step": 1364
},
{
"epoch": 0.560229837882208,
"grad_norm": 3.8823342323303223,
"learning_rate": 4.396551724137931e-06,
"loss": 0.0063,
"step": 1365
},
{
"epoch": 0.5606402626718654,
"grad_norm": 8.080864906311035,
"learning_rate": 4.392446633825945e-06,
"loss": 0.0051,
"step": 1366
},
{
"epoch": 0.5610506874615226,
"grad_norm": 6.161201000213623,
"learning_rate": 4.388341543513957e-06,
"loss": 0.0033,
"step": 1367
},
{
"epoch": 0.56146111225118,
"grad_norm": 22.312198638916016,
"learning_rate": 4.384236453201971e-06,
"loss": 0.0258,
"step": 1368
},
{
"epoch": 0.5618715370408373,
"grad_norm": 1.219581961631775,
"learning_rate": 4.380131362889984e-06,
"loss": 0.0024,
"step": 1369
},
{
"epoch": 0.5622819618304946,
"grad_norm": 6.506668567657471,
"learning_rate": 4.376026272577997e-06,
"loss": 0.0065,
"step": 1370
},
{
"epoch": 0.5626923866201519,
"grad_norm": 1.4874587059020996,
"learning_rate": 4.37192118226601e-06,
"loss": 0.0022,
"step": 1371
},
{
"epoch": 0.5631028114098091,
"grad_norm": 2.1908175945281982,
"learning_rate": 4.367816091954023e-06,
"loss": 0.0015,
"step": 1372
},
{
"epoch": 0.5635132361994665,
"grad_norm": 2.006659746170044,
"learning_rate": 4.363711001642036e-06,
"loss": 0.0018,
"step": 1373
},
{
"epoch": 0.5639236609891237,
"grad_norm": 1.0635766983032227,
"learning_rate": 4.3596059113300495e-06,
"loss": 0.0008,
"step": 1374
},
{
"epoch": 0.5643340857787811,
"grad_norm": 3.217841386795044,
"learning_rate": 4.3555008210180625e-06,
"loss": 0.0062,
"step": 1375
},
{
"epoch": 0.5647445105684383,
"grad_norm": 6.82529354095459,
"learning_rate": 4.3513957307060764e-06,
"loss": 0.0037,
"step": 1376
},
{
"epoch": 0.5651549353580956,
"grad_norm": 4.836833953857422,
"learning_rate": 4.347290640394089e-06,
"loss": 0.0024,
"step": 1377
},
{
"epoch": 0.5655653601477529,
"grad_norm": 9.42879581451416,
"learning_rate": 4.3431855500821026e-06,
"loss": 0.011,
"step": 1378
},
{
"epoch": 0.5659757849374102,
"grad_norm": 130.5061492919922,
"learning_rate": 4.339080459770116e-06,
"loss": 0.5476,
"step": 1379
},
{
"epoch": 0.5663862097270675,
"grad_norm": 1.0617592334747314,
"learning_rate": 4.334975369458129e-06,
"loss": 0.0011,
"step": 1380
},
{
"epoch": 0.5667966345167248,
"grad_norm": 28.150285720825195,
"learning_rate": 4.330870279146142e-06,
"loss": 0.023,
"step": 1381
},
{
"epoch": 0.5672070593063822,
"grad_norm": 144.79393005371094,
"learning_rate": 4.326765188834155e-06,
"loss": 0.3726,
"step": 1382
},
{
"epoch": 0.5676174840960394,
"grad_norm": 7.5425825119018555,
"learning_rate": 4.322660098522168e-06,
"loss": 0.0053,
"step": 1383
},
{
"epoch": 0.5680279088856967,
"grad_norm": 63.821712493896484,
"learning_rate": 4.318555008210181e-06,
"loss": 0.6448,
"step": 1384
},
{
"epoch": 0.568438333675354,
"grad_norm": 3.385441541671753,
"learning_rate": 4.314449917898194e-06,
"loss": 0.0057,
"step": 1385
},
{
"epoch": 0.5688487584650113,
"grad_norm": 32.44557571411133,
"learning_rate": 4.310344827586207e-06,
"loss": 0.0649,
"step": 1386
},
{
"epoch": 0.5692591832546686,
"grad_norm": 4.058762073516846,
"learning_rate": 4.30623973727422e-06,
"loss": 0.0013,
"step": 1387
},
{
"epoch": 0.5696696080443259,
"grad_norm": 13.882416725158691,
"learning_rate": 4.302134646962233e-06,
"loss": 0.0134,
"step": 1388
},
{
"epoch": 0.5700800328339831,
"grad_norm": 6.785825729370117,
"learning_rate": 4.298029556650246e-06,
"loss": 0.0065,
"step": 1389
},
{
"epoch": 0.5704904576236405,
"grad_norm": 1.8979191780090332,
"learning_rate": 4.293924466338259e-06,
"loss": 0.0022,
"step": 1390
},
{
"epoch": 0.5709008824132977,
"grad_norm": 72.9226303100586,
"learning_rate": 4.289819376026273e-06,
"loss": 0.7115,
"step": 1391
},
{
"epoch": 0.5713113072029551,
"grad_norm": 14.514235496520996,
"learning_rate": 4.2857142857142855e-06,
"loss": 0.0146,
"step": 1392
},
{
"epoch": 0.5717217319926123,
"grad_norm": 101.5116958618164,
"learning_rate": 4.281609195402299e-06,
"loss": 0.1804,
"step": 1393
},
{
"epoch": 0.5721321567822697,
"grad_norm": 7.747633934020996,
"learning_rate": 4.2775041050903124e-06,
"loss": 0.007,
"step": 1394
},
{
"epoch": 0.572542581571927,
"grad_norm": 14.548973083496094,
"learning_rate": 4.2733990147783255e-06,
"loss": 0.0148,
"step": 1395
},
{
"epoch": 0.5729530063615842,
"grad_norm": 0.6806254982948303,
"learning_rate": 4.2692939244663386e-06,
"loss": 0.0015,
"step": 1396
},
{
"epoch": 0.5733634311512416,
"grad_norm": 16.218456268310547,
"learning_rate": 4.265188834154352e-06,
"loss": 0.0104,
"step": 1397
},
{
"epoch": 0.5737738559408988,
"grad_norm": 14.263364791870117,
"learning_rate": 4.261083743842365e-06,
"loss": 0.0223,
"step": 1398
},
{
"epoch": 0.5741842807305562,
"grad_norm": 23.395963668823242,
"learning_rate": 4.256978653530378e-06,
"loss": 0.0186,
"step": 1399
},
{
"epoch": 0.5745947055202134,
"grad_norm": 3.9345924854278564,
"learning_rate": 4.252873563218391e-06,
"loss": 0.0068,
"step": 1400
},
{
"epoch": 0.5750051303098708,
"grad_norm": 3.707388401031494,
"learning_rate": 4.248768472906405e-06,
"loss": 0.002,
"step": 1401
},
{
"epoch": 0.575415555099528,
"grad_norm": 3.5885872840881348,
"learning_rate": 4.244663382594417e-06,
"loss": 0.0027,
"step": 1402
},
{
"epoch": 0.5758259798891853,
"grad_norm": 3.226529359817505,
"learning_rate": 4.240558292282431e-06,
"loss": 0.0064,
"step": 1403
},
{
"epoch": 0.5762364046788426,
"grad_norm": 3.5987508296966553,
"learning_rate": 4.236453201970444e-06,
"loss": 0.0013,
"step": 1404
},
{
"epoch": 0.5766468294684999,
"grad_norm": 2.3818750381469727,
"learning_rate": 4.232348111658457e-06,
"loss": 0.0016,
"step": 1405
},
{
"epoch": 0.5770572542581572,
"grad_norm": 7.555981159210205,
"learning_rate": 4.22824302134647e-06,
"loss": 0.0044,
"step": 1406
},
{
"epoch": 0.5774676790478145,
"grad_norm": 99.77206420898438,
"learning_rate": 4.224137931034483e-06,
"loss": 0.2485,
"step": 1407
},
{
"epoch": 0.5778781038374717,
"grad_norm": 6.380500316619873,
"learning_rate": 4.220032840722496e-06,
"loss": 0.0047,
"step": 1408
},
{
"epoch": 0.5782885286271291,
"grad_norm": 10.142494201660156,
"learning_rate": 4.215927750410509e-06,
"loss": 0.01,
"step": 1409
},
{
"epoch": 0.5786989534167863,
"grad_norm": 2.5076282024383545,
"learning_rate": 4.211822660098522e-06,
"loss": 0.0019,
"step": 1410
},
{
"epoch": 0.5791093782064437,
"grad_norm": 3.4601635932922363,
"learning_rate": 4.207717569786535e-06,
"loss": 0.002,
"step": 1411
},
{
"epoch": 0.579519802996101,
"grad_norm": 3.3743066787719727,
"learning_rate": 4.2036124794745484e-06,
"loss": 0.0032,
"step": 1412
},
{
"epoch": 0.5799302277857583,
"grad_norm": 58.78200149536133,
"learning_rate": 4.199507389162562e-06,
"loss": 0.4127,
"step": 1413
},
{
"epoch": 0.5803406525754156,
"grad_norm": 2.0065393447875977,
"learning_rate": 4.1954022988505746e-06,
"loss": 0.0013,
"step": 1414
},
{
"epoch": 0.5807510773650728,
"grad_norm": 11.722634315490723,
"learning_rate": 4.1912972085385885e-06,
"loss": 0.0188,
"step": 1415
},
{
"epoch": 0.5811615021547302,
"grad_norm": 1.7201281785964966,
"learning_rate": 4.1871921182266015e-06,
"loss": 0.0009,
"step": 1416
},
{
"epoch": 0.5815719269443874,
"grad_norm": 16.89137077331543,
"learning_rate": 4.183087027914615e-06,
"loss": 0.0188,
"step": 1417
},
{
"epoch": 0.5819823517340448,
"grad_norm": 2.840872049331665,
"learning_rate": 4.178981937602628e-06,
"loss": 0.0026,
"step": 1418
},
{
"epoch": 0.582392776523702,
"grad_norm": 3.695894241333008,
"learning_rate": 4.174876847290641e-06,
"loss": 0.0021,
"step": 1419
},
{
"epoch": 0.5828032013133593,
"grad_norm": 7.897422790527344,
"learning_rate": 4.170771756978654e-06,
"loss": 0.0086,
"step": 1420
},
{
"epoch": 0.5832136261030166,
"grad_norm": 0.9823002815246582,
"learning_rate": 4.166666666666667e-06,
"loss": 0.001,
"step": 1421
},
{
"epoch": 0.5836240508926739,
"grad_norm": 200.99986267089844,
"learning_rate": 4.16256157635468e-06,
"loss": 0.3102,
"step": 1422
},
{
"epoch": 0.5840344756823312,
"grad_norm": 0.5076814293861389,
"learning_rate": 4.158456486042694e-06,
"loss": 0.0008,
"step": 1423
},
{
"epoch": 0.5844449004719885,
"grad_norm": 2.731102228164673,
"learning_rate": 4.154351395730706e-06,
"loss": 0.0032,
"step": 1424
},
{
"epoch": 0.5848553252616459,
"grad_norm": 5.57316255569458,
"learning_rate": 4.15024630541872e-06,
"loss": 0.0031,
"step": 1425
},
{
"epoch": 0.5852657500513031,
"grad_norm": 38.747642517089844,
"learning_rate": 4.146141215106733e-06,
"loss": 0.0213,
"step": 1426
},
{
"epoch": 0.5856761748409604,
"grad_norm": 8.587905883789062,
"learning_rate": 4.142036124794746e-06,
"loss": 0.0058,
"step": 1427
},
{
"epoch": 0.5860865996306177,
"grad_norm": 10.4324951171875,
"learning_rate": 4.137931034482759e-06,
"loss": 0.0099,
"step": 1428
},
{
"epoch": 0.586497024420275,
"grad_norm": 52.306480407714844,
"learning_rate": 4.133825944170772e-06,
"loss": 0.634,
"step": 1429
},
{
"epoch": 0.5869074492099323,
"grad_norm": 50.536556243896484,
"learning_rate": 4.129720853858785e-06,
"loss": 0.3931,
"step": 1430
},
{
"epoch": 0.5873178739995896,
"grad_norm": 2.0114059448242188,
"learning_rate": 4.125615763546798e-06,
"loss": 0.0009,
"step": 1431
},
{
"epoch": 0.5877282987892468,
"grad_norm": 2.925786256790161,
"learning_rate": 4.121510673234811e-06,
"loss": 0.001,
"step": 1432
},
{
"epoch": 0.5881387235789042,
"grad_norm": 1.5649899244308472,
"learning_rate": 4.1174055829228245e-06,
"loss": 0.0006,
"step": 1433
},
{
"epoch": 0.5885491483685614,
"grad_norm": 1.6353716850280762,
"learning_rate": 4.1133004926108375e-06,
"loss": 0.001,
"step": 1434
},
{
"epoch": 0.5889595731582188,
"grad_norm": 6.459379196166992,
"learning_rate": 4.1091954022988515e-06,
"loss": 0.0065,
"step": 1435
},
{
"epoch": 0.589369997947876,
"grad_norm": 86.11415100097656,
"learning_rate": 4.105090311986864e-06,
"loss": 0.5648,
"step": 1436
},
{
"epoch": 0.5897804227375334,
"grad_norm": 8.325303077697754,
"learning_rate": 4.100985221674878e-06,
"loss": 0.0058,
"step": 1437
},
{
"epoch": 0.5901908475271906,
"grad_norm": 0.8115727305412292,
"learning_rate": 4.096880131362891e-06,
"loss": 0.0011,
"step": 1438
},
{
"epoch": 0.5906012723168479,
"grad_norm": 7.185229301452637,
"learning_rate": 4.092775041050904e-06,
"loss": 0.0039,
"step": 1439
},
{
"epoch": 0.5910116971065053,
"grad_norm": 3.8980748653411865,
"learning_rate": 4.088669950738917e-06,
"loss": 0.0065,
"step": 1440
},
{
"epoch": 0.5914221218961625,
"grad_norm": 3.4298312664031982,
"learning_rate": 4.08456486042693e-06,
"loss": 0.0061,
"step": 1441
},
{
"epoch": 0.5918325466858199,
"grad_norm": 1.746060848236084,
"learning_rate": 4.080459770114943e-06,
"loss": 0.0007,
"step": 1442
},
{
"epoch": 0.5922429714754771,
"grad_norm": 8.414861679077148,
"learning_rate": 4.076354679802956e-06,
"loss": 0.0066,
"step": 1443
},
{
"epoch": 0.5926533962651345,
"grad_norm": 13.675786972045898,
"learning_rate": 4.072249589490969e-06,
"loss": 0.0148,
"step": 1444
},
{
"epoch": 0.5930638210547917,
"grad_norm": 72.5360336303711,
"learning_rate": 4.068144499178982e-06,
"loss": 0.1635,
"step": 1445
},
{
"epoch": 0.593474245844449,
"grad_norm": 1.837649941444397,
"learning_rate": 4.064039408866995e-06,
"loss": 0.0016,
"step": 1446
},
{
"epoch": 0.5938846706341063,
"grad_norm": 61.57942199707031,
"learning_rate": 4.059934318555008e-06,
"loss": 0.6326,
"step": 1447
},
{
"epoch": 0.5942950954237636,
"grad_norm": 6.053219795227051,
"learning_rate": 4.055829228243022e-06,
"loss": 0.003,
"step": 1448
},
{
"epoch": 0.5947055202134209,
"grad_norm": 1.588888168334961,
"learning_rate": 4.051724137931034e-06,
"loss": 0.0015,
"step": 1449
},
{
"epoch": 0.5951159450030782,
"grad_norm": 65.66236114501953,
"learning_rate": 4.047619047619048e-06,
"loss": 0.5505,
"step": 1450
},
{
"epoch": 0.5955263697927354,
"grad_norm": 3.839322090148926,
"learning_rate": 4.0435139573070605e-06,
"loss": 0.0019,
"step": 1451
},
{
"epoch": 0.5959367945823928,
"grad_norm": 13.249491691589355,
"learning_rate": 4.039408866995074e-06,
"loss": 0.0169,
"step": 1452
},
{
"epoch": 0.59634721937205,
"grad_norm": 8.089970588684082,
"learning_rate": 4.0353037766830875e-06,
"loss": 0.0059,
"step": 1453
},
{
"epoch": 0.5967576441617074,
"grad_norm": 10.867718696594238,
"learning_rate": 4.0311986863711005e-06,
"loss": 0.0088,
"step": 1454
},
{
"epoch": 0.5971680689513647,
"grad_norm": 10.713374137878418,
"learning_rate": 4.027093596059114e-06,
"loss": 0.0087,
"step": 1455
},
{
"epoch": 0.597578493741022,
"grad_norm": 7.401761531829834,
"learning_rate": 4.022988505747127e-06,
"loss": 0.0051,
"step": 1456
},
{
"epoch": 0.5979889185306793,
"grad_norm": 10.082868576049805,
"learning_rate": 4.01888341543514e-06,
"loss": 0.0084,
"step": 1457
},
{
"epoch": 0.5983993433203365,
"grad_norm": 3.8851568698883057,
"learning_rate": 4.014778325123153e-06,
"loss": 0.0032,
"step": 1458
},
{
"epoch": 0.5988097681099939,
"grad_norm": 48.306419372558594,
"learning_rate": 4.010673234811166e-06,
"loss": 0.5186,
"step": 1459
},
{
"epoch": 0.5992201928996511,
"grad_norm": 75.66358184814453,
"learning_rate": 4.00656814449918e-06,
"loss": 0.4076,
"step": 1460
},
{
"epoch": 0.5996306176893085,
"grad_norm": 12.211490631103516,
"learning_rate": 4.002463054187192e-06,
"loss": 0.0065,
"step": 1461
},
{
"epoch": 0.6000410424789657,
"grad_norm": 6.542180061340332,
"learning_rate": 3.998357963875206e-06,
"loss": 0.0042,
"step": 1462
},
{
"epoch": 0.600451467268623,
"grad_norm": 41.89014434814453,
"learning_rate": 3.994252873563219e-06,
"loss": 0.0437,
"step": 1463
},
{
"epoch": 0.6008618920582803,
"grad_norm": 13.929396629333496,
"learning_rate": 3.990147783251232e-06,
"loss": 0.0142,
"step": 1464
},
{
"epoch": 0.6012723168479376,
"grad_norm": 11.666210174560547,
"learning_rate": 3.986042692939245e-06,
"loss": 0.0106,
"step": 1465
},
{
"epoch": 0.6016827416375949,
"grad_norm": 1.8232412338256836,
"learning_rate": 3.981937602627258e-06,
"loss": 0.0029,
"step": 1466
},
{
"epoch": 0.6020931664272522,
"grad_norm": 30.970861434936523,
"learning_rate": 3.977832512315271e-06,
"loss": 0.0176,
"step": 1467
},
{
"epoch": 0.6025035912169096,
"grad_norm": 59.86967468261719,
"learning_rate": 3.973727422003284e-06,
"loss": 0.8323,
"step": 1468
},
{
"epoch": 0.6029140160065668,
"grad_norm": 4.856724262237549,
"learning_rate": 3.969622331691297e-06,
"loss": 0.0026,
"step": 1469
},
{
"epoch": 0.603324440796224,
"grad_norm": 4.615633964538574,
"learning_rate": 3.96551724137931e-06,
"loss": 0.003,
"step": 1470
},
{
"epoch": 0.6037348655858814,
"grad_norm": 3.7640128135681152,
"learning_rate": 3.9614121510673235e-06,
"loss": 0.0024,
"step": 1471
},
{
"epoch": 0.6041452903755387,
"grad_norm": 1.5890988111495972,
"learning_rate": 3.957307060755337e-06,
"loss": 0.0019,
"step": 1472
},
{
"epoch": 0.604555715165196,
"grad_norm": 12.737597465515137,
"learning_rate": 3.95320197044335e-06,
"loss": 0.0121,
"step": 1473
},
{
"epoch": 0.6049661399548533,
"grad_norm": 154.61598205566406,
"learning_rate": 3.9490968801313635e-06,
"loss": 0.4605,
"step": 1474
},
{
"epoch": 0.6053765647445105,
"grad_norm": 5.379822254180908,
"learning_rate": 3.9449917898193766e-06,
"loss": 0.004,
"step": 1475
},
{
"epoch": 0.6057869895341679,
"grad_norm": 5.692903995513916,
"learning_rate": 3.94088669950739e-06,
"loss": 0.0055,
"step": 1476
},
{
"epoch": 0.6061974143238251,
"grad_norm": 5.0556182861328125,
"learning_rate": 3.936781609195403e-06,
"loss": 0.0046,
"step": 1477
},
{
"epoch": 0.6066078391134825,
"grad_norm": 6.753775119781494,
"learning_rate": 3.932676518883416e-06,
"loss": 0.0162,
"step": 1478
},
{
"epoch": 0.6070182639031397,
"grad_norm": 6.950207710266113,
"learning_rate": 3.928571428571429e-06,
"loss": 0.0052,
"step": 1479
},
{
"epoch": 0.6074286886927971,
"grad_norm": 5.104891777038574,
"learning_rate": 3.924466338259442e-06,
"loss": 0.0036,
"step": 1480
},
{
"epoch": 0.6078391134824543,
"grad_norm": 28.03950309753418,
"learning_rate": 3.920361247947455e-06,
"loss": 0.032,
"step": 1481
},
{
"epoch": 0.6082495382721116,
"grad_norm": 2.254540205001831,
"learning_rate": 3.916256157635469e-06,
"loss": 0.0072,
"step": 1482
},
{
"epoch": 0.608659963061769,
"grad_norm": 0.45824363827705383,
"learning_rate": 3.912151067323481e-06,
"loss": 0.0009,
"step": 1483
},
{
"epoch": 0.6090703878514262,
"grad_norm": 0.8383445739746094,
"learning_rate": 3.908045977011495e-06,
"loss": 0.0011,
"step": 1484
},
{
"epoch": 0.6094808126410836,
"grad_norm": 80.79524993896484,
"learning_rate": 3.903940886699508e-06,
"loss": 0.5552,
"step": 1485
},
{
"epoch": 0.6098912374307408,
"grad_norm": 4.396963119506836,
"learning_rate": 3.899835796387521e-06,
"loss": 0.0031,
"step": 1486
},
{
"epoch": 0.6103016622203982,
"grad_norm": 3.1831438541412354,
"learning_rate": 3.895730706075534e-06,
"loss": 0.0027,
"step": 1487
},
{
"epoch": 0.6107120870100554,
"grad_norm": 9.1783447265625,
"learning_rate": 3.891625615763547e-06,
"loss": 0.0251,
"step": 1488
},
{
"epoch": 0.6111225117997127,
"grad_norm": 3.632807731628418,
"learning_rate": 3.88752052545156e-06,
"loss": 0.0018,
"step": 1489
},
{
"epoch": 0.61153293658937,
"grad_norm": 8.295243263244629,
"learning_rate": 3.883415435139573e-06,
"loss": 0.0071,
"step": 1490
},
{
"epoch": 0.6119433613790273,
"grad_norm": 0.8969875574111938,
"learning_rate": 3.8793103448275865e-06,
"loss": 0.0012,
"step": 1491
},
{
"epoch": 0.6123537861686846,
"grad_norm": 3.3994805812835693,
"learning_rate": 3.8752052545155995e-06,
"loss": 0.0043,
"step": 1492
},
{
"epoch": 0.6127642109583419,
"grad_norm": 73.15042114257812,
"learning_rate": 3.871100164203613e-06,
"loss": 0.5391,
"step": 1493
},
{
"epoch": 0.6131746357479991,
"grad_norm": 7.513078212738037,
"learning_rate": 3.8669950738916265e-06,
"loss": 0.0051,
"step": 1494
},
{
"epoch": 0.6135850605376565,
"grad_norm": 5.106591701507568,
"learning_rate": 3.862889983579639e-06,
"loss": 0.0026,
"step": 1495
},
{
"epoch": 0.6139954853273137,
"grad_norm": 8.418148040771484,
"learning_rate": 3.858784893267652e-06,
"loss": 0.005,
"step": 1496
},
{
"epoch": 0.6144059101169711,
"grad_norm": 5.201192855834961,
"learning_rate": 3.854679802955666e-06,
"loss": 0.0089,
"step": 1497
},
{
"epoch": 0.6148163349066283,
"grad_norm": 0.5819182991981506,
"learning_rate": 3.850574712643678e-06,
"loss": 0.0006,
"step": 1498
},
{
"epoch": 0.6152267596962857,
"grad_norm": 3.651946783065796,
"learning_rate": 3.846469622331692e-06,
"loss": 0.0014,
"step": 1499
},
{
"epoch": 0.615637184485943,
"grad_norm": 212.60975646972656,
"learning_rate": 3.842364532019705e-06,
"loss": 0.5246,
"step": 1500
},
{
"epoch": 0.6160476092756002,
"grad_norm": 1.5426466464996338,
"learning_rate": 3.838259441707718e-06,
"loss": 0.0011,
"step": 1501
},
{
"epoch": 0.6164580340652576,
"grad_norm": 8.853590965270996,
"learning_rate": 3.834154351395731e-06,
"loss": 0.0064,
"step": 1502
},
{
"epoch": 0.6168684588549148,
"grad_norm": 33.16612243652344,
"learning_rate": 3.830049261083744e-06,
"loss": 0.0346,
"step": 1503
},
{
"epoch": 0.6172788836445722,
"grad_norm": 4.655913352966309,
"learning_rate": 3.825944170771757e-06,
"loss": 0.0021,
"step": 1504
},
{
"epoch": 0.6176893084342294,
"grad_norm": 7.68121337890625,
"learning_rate": 3.82183908045977e-06,
"loss": 0.0068,
"step": 1505
},
{
"epoch": 0.6180997332238867,
"grad_norm": 3.063318967819214,
"learning_rate": 3.817733990147783e-06,
"loss": 0.0013,
"step": 1506
},
{
"epoch": 0.618510158013544,
"grad_norm": 0.15942241251468658,
"learning_rate": 3.8136288998357968e-06,
"loss": 0.0006,
"step": 1507
},
{
"epoch": 0.6189205828032013,
"grad_norm": 2.4072606563568115,
"learning_rate": 3.80952380952381e-06,
"loss": 0.0008,
"step": 1508
},
{
"epoch": 0.6193310075928586,
"grad_norm": 11.427154541015625,
"learning_rate": 3.805418719211823e-06,
"loss": 0.006,
"step": 1509
},
{
"epoch": 0.6197414323825159,
"grad_norm": 2.0546107292175293,
"learning_rate": 3.801313628899836e-06,
"loss": 0.0018,
"step": 1510
},
{
"epoch": 0.6201518571721732,
"grad_norm": 14.144604682922363,
"learning_rate": 3.7972085385878494e-06,
"loss": 0.0222,
"step": 1511
},
{
"epoch": 0.6205622819618305,
"grad_norm": 13.577574729919434,
"learning_rate": 3.793103448275862e-06,
"loss": 0.0133,
"step": 1512
},
{
"epoch": 0.6209727067514877,
"grad_norm": 13.416616439819336,
"learning_rate": 3.7889983579638756e-06,
"loss": 0.0143,
"step": 1513
},
{
"epoch": 0.6213831315411451,
"grad_norm": 19.088411331176758,
"learning_rate": 3.7848932676518886e-06,
"loss": 0.0544,
"step": 1514
},
{
"epoch": 0.6217935563308024,
"grad_norm": 8.653985977172852,
"learning_rate": 3.7807881773399017e-06,
"loss": 0.007,
"step": 1515
},
{
"epoch": 0.6222039811204597,
"grad_norm": 13.005366325378418,
"learning_rate": 3.7766830870279148e-06,
"loss": 0.0134,
"step": 1516
},
{
"epoch": 0.622614405910117,
"grad_norm": 32.579315185546875,
"learning_rate": 3.7725779967159282e-06,
"loss": 0.0327,
"step": 1517
},
{
"epoch": 0.6230248306997742,
"grad_norm": 7.84839391708374,
"learning_rate": 3.768472906403941e-06,
"loss": 0.005,
"step": 1518
},
{
"epoch": 0.6234352554894316,
"grad_norm": 0.25949326157569885,
"learning_rate": 3.7643678160919544e-06,
"loss": 0.001,
"step": 1519
},
{
"epoch": 0.6238456802790888,
"grad_norm": 2.403491735458374,
"learning_rate": 3.7602627257799674e-06,
"loss": 0.0055,
"step": 1520
},
{
"epoch": 0.6242561050687462,
"grad_norm": 6.699248790740967,
"learning_rate": 3.756157635467981e-06,
"loss": 0.0052,
"step": 1521
},
{
"epoch": 0.6246665298584034,
"grad_norm": 142.95343017578125,
"learning_rate": 3.7520525451559936e-06,
"loss": 0.8081,
"step": 1522
},
{
"epoch": 0.6250769546480608,
"grad_norm": 1.975730061531067,
"learning_rate": 3.747947454844007e-06,
"loss": 0.0013,
"step": 1523
},
{
"epoch": 0.625487379437718,
"grad_norm": 3.523585557937622,
"learning_rate": 3.7438423645320197e-06,
"loss": 0.0014,
"step": 1524
},
{
"epoch": 0.6258978042273753,
"grad_norm": 111.44734954833984,
"learning_rate": 3.739737274220033e-06,
"loss": 0.3469,
"step": 1525
},
{
"epoch": 0.6263082290170326,
"grad_norm": 9.248785018920898,
"learning_rate": 3.7356321839080462e-06,
"loss": 0.0066,
"step": 1526
},
{
"epoch": 0.6267186538066899,
"grad_norm": 6.064499855041504,
"learning_rate": 3.7315270935960597e-06,
"loss": 0.0032,
"step": 1527
},
{
"epoch": 0.6271290785963473,
"grad_norm": 9.146154403686523,
"learning_rate": 3.7274220032840724e-06,
"loss": 0.0068,
"step": 1528
},
{
"epoch": 0.6275395033860045,
"grad_norm": 77.05956268310547,
"learning_rate": 3.723316912972086e-06,
"loss": 0.6102,
"step": 1529
},
{
"epoch": 0.6279499281756619,
"grad_norm": 74.25747680664062,
"learning_rate": 3.719211822660099e-06,
"loss": 0.7334,
"step": 1530
},
{
"epoch": 0.6283603529653191,
"grad_norm": 32.94594955444336,
"learning_rate": 3.715106732348112e-06,
"loss": 0.0138,
"step": 1531
},
{
"epoch": 0.6287707777549764,
"grad_norm": 17.942737579345703,
"learning_rate": 3.711001642036125e-06,
"loss": 0.0228,
"step": 1532
},
{
"epoch": 0.6291812025446337,
"grad_norm": 8.167811393737793,
"learning_rate": 3.7068965517241385e-06,
"loss": 0.0052,
"step": 1533
},
{
"epoch": 0.629591627334291,
"grad_norm": 1.070863962173462,
"learning_rate": 3.702791461412151e-06,
"loss": 0.0011,
"step": 1534
},
{
"epoch": 0.6300020521239483,
"grad_norm": 1.1257433891296387,
"learning_rate": 3.6986863711001647e-06,
"loss": 0.0011,
"step": 1535
},
{
"epoch": 0.6304124769136056,
"grad_norm": 5.018160343170166,
"learning_rate": 3.6945812807881777e-06,
"loss": 0.0023,
"step": 1536
},
{
"epoch": 0.6308229017032628,
"grad_norm": 0.9321138262748718,
"learning_rate": 3.690476190476191e-06,
"loss": 0.0006,
"step": 1537
},
{
"epoch": 0.6312333264929202,
"grad_norm": 6.610182285308838,
"learning_rate": 3.686371100164204e-06,
"loss": 0.005,
"step": 1538
},
{
"epoch": 0.6316437512825774,
"grad_norm": 6.210333824157715,
"learning_rate": 3.6822660098522174e-06,
"loss": 0.008,
"step": 1539
},
{
"epoch": 0.6320541760722348,
"grad_norm": 151.78269958496094,
"learning_rate": 3.67816091954023e-06,
"loss": 0.1126,
"step": 1540
},
{
"epoch": 0.632464600861892,
"grad_norm": 9.836950302124023,
"learning_rate": 3.6740558292282435e-06,
"loss": 0.0102,
"step": 1541
},
{
"epoch": 0.6328750256515494,
"grad_norm": 8.94073486328125,
"learning_rate": 3.6699507389162565e-06,
"loss": 0.0065,
"step": 1542
},
{
"epoch": 0.6332854504412067,
"grad_norm": 11.193456649780273,
"learning_rate": 3.66584564860427e-06,
"loss": 0.0098,
"step": 1543
},
{
"epoch": 0.6336958752308639,
"grad_norm": 0.8447343707084656,
"learning_rate": 3.6617405582922827e-06,
"loss": 0.0005,
"step": 1544
},
{
"epoch": 0.6341063000205213,
"grad_norm": 4.402817249298096,
"learning_rate": 3.657635467980296e-06,
"loss": 0.0019,
"step": 1545
},
{
"epoch": 0.6345167248101785,
"grad_norm": 1.503555417060852,
"learning_rate": 3.653530377668309e-06,
"loss": 0.0017,
"step": 1546
},
{
"epoch": 0.6349271495998359,
"grad_norm": 8.301360130310059,
"learning_rate": 3.6494252873563223e-06,
"loss": 0.0068,
"step": 1547
},
{
"epoch": 0.6353375743894931,
"grad_norm": 3.4098973274230957,
"learning_rate": 3.6453201970443354e-06,
"loss": 0.0024,
"step": 1548
},
{
"epoch": 0.6357479991791504,
"grad_norm": 5.081045150756836,
"learning_rate": 3.641215106732349e-06,
"loss": 0.0048,
"step": 1549
},
{
"epoch": 0.6361584239688077,
"grad_norm": 1.038743257522583,
"learning_rate": 3.6371100164203615e-06,
"loss": 0.0013,
"step": 1550
},
{
"epoch": 0.636568848758465,
"grad_norm": 4.059690475463867,
"learning_rate": 3.6330049261083745e-06,
"loss": 0.0049,
"step": 1551
},
{
"epoch": 0.6369792735481223,
"grad_norm": 3.365403175354004,
"learning_rate": 3.628899835796388e-06,
"loss": 0.0015,
"step": 1552
},
{
"epoch": 0.6373896983377796,
"grad_norm": 4.091079235076904,
"learning_rate": 3.6247947454844007e-06,
"loss": 0.002,
"step": 1553
},
{
"epoch": 0.637800123127437,
"grad_norm": 2.699700117111206,
"learning_rate": 3.620689655172414e-06,
"loss": 0.0014,
"step": 1554
},
{
"epoch": 0.6382105479170942,
"grad_norm": 6.829959869384766,
"learning_rate": 3.616584564860427e-06,
"loss": 0.0036,
"step": 1555
},
{
"epoch": 0.6386209727067514,
"grad_norm": 2.9897313117980957,
"learning_rate": 3.6124794745484403e-06,
"loss": 0.0035,
"step": 1556
},
{
"epoch": 0.6390313974964088,
"grad_norm": 15.34519100189209,
"learning_rate": 3.6083743842364534e-06,
"loss": 0.0229,
"step": 1557
},
{
"epoch": 0.639441822286066,
"grad_norm": 1.1519958972930908,
"learning_rate": 3.604269293924467e-06,
"loss": 0.0005,
"step": 1558
},
{
"epoch": 0.6398522470757234,
"grad_norm": 885.9804077148438,
"learning_rate": 3.6001642036124795e-06,
"loss": 1.243,
"step": 1559
},
{
"epoch": 0.6402626718653807,
"grad_norm": 4.302264213562012,
"learning_rate": 3.596059113300493e-06,
"loss": 0.0028,
"step": 1560
},
{
"epoch": 0.6406730966550379,
"grad_norm": 7.761496067047119,
"learning_rate": 3.5919540229885056e-06,
"loss": 0.0057,
"step": 1561
},
{
"epoch": 0.6410835214446953,
"grad_norm": 1.9562568664550781,
"learning_rate": 3.587848932676519e-06,
"loss": 0.0029,
"step": 1562
},
{
"epoch": 0.6414939462343525,
"grad_norm": 4.528267860412598,
"learning_rate": 3.583743842364532e-06,
"loss": 0.0022,
"step": 1563
},
{
"epoch": 0.6419043710240099,
"grad_norm": 117.30449676513672,
"learning_rate": 3.5796387520525457e-06,
"loss": 0.0723,
"step": 1564
},
{
"epoch": 0.6423147958136671,
"grad_norm": 11.791247367858887,
"learning_rate": 3.5755336617405583e-06,
"loss": 0.0129,
"step": 1565
},
{
"epoch": 0.6427252206033245,
"grad_norm": 15.02712345123291,
"learning_rate": 3.5714285714285718e-06,
"loss": 0.0264,
"step": 1566
},
{
"epoch": 0.6431356453929817,
"grad_norm": 13.489358901977539,
"learning_rate": 3.567323481116585e-06,
"loss": 0.0137,
"step": 1567
},
{
"epoch": 0.643546070182639,
"grad_norm": 10.104757308959961,
"learning_rate": 3.563218390804598e-06,
"loss": 0.009,
"step": 1568
},
{
"epoch": 0.6439564949722963,
"grad_norm": 0.3910963535308838,
"learning_rate": 3.559113300492611e-06,
"loss": 0.0009,
"step": 1569
},
{
"epoch": 0.6443669197619536,
"grad_norm": 13.20108699798584,
"learning_rate": 3.5550082101806245e-06,
"loss": 0.0063,
"step": 1570
},
{
"epoch": 0.644777344551611,
"grad_norm": 1.4573471546173096,
"learning_rate": 3.550903119868637e-06,
"loss": 0.0013,
"step": 1571
},
{
"epoch": 0.6451877693412682,
"grad_norm": 5.820040702819824,
"learning_rate": 3.5467980295566506e-06,
"loss": 0.0036,
"step": 1572
},
{
"epoch": 0.6455981941309256,
"grad_norm": 0.754395067691803,
"learning_rate": 3.5426929392446637e-06,
"loss": 0.0008,
"step": 1573
},
{
"epoch": 0.6460086189205828,
"grad_norm": 3.895562171936035,
"learning_rate": 3.5385878489326767e-06,
"loss": 0.0038,
"step": 1574
},
{
"epoch": 0.6464190437102401,
"grad_norm": 376.50030517578125,
"learning_rate": 3.5344827586206898e-06,
"loss": 0.1744,
"step": 1575
},
{
"epoch": 0.6468294684998974,
"grad_norm": 2.4993691444396973,
"learning_rate": 3.5303776683087033e-06,
"loss": 0.0009,
"step": 1576
},
{
"epoch": 0.6472398932895547,
"grad_norm": 206.059814453125,
"learning_rate": 3.526272577996716e-06,
"loss": 0.2356,
"step": 1577
},
{
"epoch": 0.647650318079212,
"grad_norm": 3.327181339263916,
"learning_rate": 3.5221674876847294e-06,
"loss": 0.0011,
"step": 1578
},
{
"epoch": 0.6480607428688693,
"grad_norm": 3.4286413192749023,
"learning_rate": 3.5180623973727425e-06,
"loss": 0.0071,
"step": 1579
},
{
"epoch": 0.6484711676585265,
"grad_norm": 13.062957763671875,
"learning_rate": 3.513957307060756e-06,
"loss": 0.0107,
"step": 1580
},
{
"epoch": 0.6488815924481839,
"grad_norm": 3.391874074935913,
"learning_rate": 3.5098522167487686e-06,
"loss": 0.0016,
"step": 1581
},
{
"epoch": 0.6492920172378411,
"grad_norm": 1.732949137687683,
"learning_rate": 3.505747126436782e-06,
"loss": 0.0018,
"step": 1582
},
{
"epoch": 0.6497024420274985,
"grad_norm": 4.822823524475098,
"learning_rate": 3.5016420361247947e-06,
"loss": 0.0032,
"step": 1583
},
{
"epoch": 0.6501128668171557,
"grad_norm": 0.9874751567840576,
"learning_rate": 3.497536945812808e-06,
"loss": 0.0005,
"step": 1584
},
{
"epoch": 0.6505232916068131,
"grad_norm": 11.67338752746582,
"learning_rate": 3.4934318555008213e-06,
"loss": 0.0091,
"step": 1585
},
{
"epoch": 0.6509337163964704,
"grad_norm": 63.813560485839844,
"learning_rate": 3.4893267651888348e-06,
"loss": 0.5627,
"step": 1586
},
{
"epoch": 0.6513441411861276,
"grad_norm": 4.260919570922852,
"learning_rate": 3.4852216748768474e-06,
"loss": 0.002,
"step": 1587
},
{
"epoch": 0.651754565975785,
"grad_norm": 3.3686842918395996,
"learning_rate": 3.481116584564861e-06,
"loss": 0.0015,
"step": 1588
},
{
"epoch": 0.6521649907654422,
"grad_norm": 1.8096188306808472,
"learning_rate": 3.477011494252874e-06,
"loss": 0.0027,
"step": 1589
},
{
"epoch": 0.6525754155550996,
"grad_norm": 5.045438766479492,
"learning_rate": 3.472906403940887e-06,
"loss": 0.0022,
"step": 1590
},
{
"epoch": 0.6529858403447568,
"grad_norm": 2.2811977863311768,
"learning_rate": 3.4688013136289e-06,
"loss": 0.0019,
"step": 1591
},
{
"epoch": 0.6533962651344141,
"grad_norm": 167.1295623779297,
"learning_rate": 3.4646962233169136e-06,
"loss": 0.141,
"step": 1592
},
{
"epoch": 0.6538066899240714,
"grad_norm": 1.6545850038528442,
"learning_rate": 3.4605911330049262e-06,
"loss": 0.0026,
"step": 1593
},
{
"epoch": 0.6542171147137287,
"grad_norm": 2.2303104400634766,
"learning_rate": 3.4564860426929397e-06,
"loss": 0.001,
"step": 1594
},
{
"epoch": 0.654627539503386,
"grad_norm": 8.218694686889648,
"learning_rate": 3.4523809523809528e-06,
"loss": 0.0055,
"step": 1595
},
{
"epoch": 0.6550379642930433,
"grad_norm": 6.807530403137207,
"learning_rate": 3.448275862068966e-06,
"loss": 0.0039,
"step": 1596
},
{
"epoch": 0.6554483890827006,
"grad_norm": 22.862268447875977,
"learning_rate": 3.444170771756979e-06,
"loss": 0.0242,
"step": 1597
},
{
"epoch": 0.6558588138723579,
"grad_norm": 123.71745300292969,
"learning_rate": 3.4400656814449924e-06,
"loss": 0.045,
"step": 1598
},
{
"epoch": 0.6562692386620151,
"grad_norm": 5.188385963439941,
"learning_rate": 3.435960591133005e-06,
"loss": 0.0031,
"step": 1599
},
{
"epoch": 0.6566796634516725,
"grad_norm": 1.0113394260406494,
"learning_rate": 3.4318555008210185e-06,
"loss": 0.0024,
"step": 1600
},
{
"epoch": 0.6570900882413298,
"grad_norm": 2.260026216506958,
"learning_rate": 3.4277504105090316e-06,
"loss": 0.0022,
"step": 1601
},
{
"epoch": 0.6575005130309871,
"grad_norm": 27.6624698638916,
"learning_rate": 3.423645320197045e-06,
"loss": 0.0237,
"step": 1602
},
{
"epoch": 0.6579109378206444,
"grad_norm": 1.9735910892486572,
"learning_rate": 3.4195402298850577e-06,
"loss": 0.0012,
"step": 1603
},
{
"epoch": 0.6583213626103016,
"grad_norm": 16.901208877563477,
"learning_rate": 3.415435139573071e-06,
"loss": 0.0369,
"step": 1604
},
{
"epoch": 0.658731787399959,
"grad_norm": 3.2877731323242188,
"learning_rate": 3.411330049261084e-06,
"loss": 0.0017,
"step": 1605
},
{
"epoch": 0.6591422121896162,
"grad_norm": 4.226407527923584,
"learning_rate": 3.407224958949097e-06,
"loss": 0.0039,
"step": 1606
},
{
"epoch": 0.6595526369792736,
"grad_norm": 99.34418487548828,
"learning_rate": 3.4031198686371104e-06,
"loss": 0.5745,
"step": 1607
},
{
"epoch": 0.6599630617689308,
"grad_norm": 6.999695777893066,
"learning_rate": 3.399014778325123e-06,
"loss": 0.0076,
"step": 1608
},
{
"epoch": 0.6603734865585882,
"grad_norm": 6.130064487457275,
"learning_rate": 3.3949096880131365e-06,
"loss": 0.0036,
"step": 1609
},
{
"epoch": 0.6607839113482454,
"grad_norm": 2.5607962608337402,
"learning_rate": 3.3908045977011496e-06,
"loss": 0.0012,
"step": 1610
},
{
"epoch": 0.6611943361379027,
"grad_norm": 0.20795013010501862,
"learning_rate": 3.386699507389163e-06,
"loss": 0.0002,
"step": 1611
},
{
"epoch": 0.66160476092756,
"grad_norm": 8.671700477600098,
"learning_rate": 3.3825944170771757e-06,
"loss": 0.0056,
"step": 1612
},
{
"epoch": 0.6620151857172173,
"grad_norm": 9.685857772827148,
"learning_rate": 3.378489326765189e-06,
"loss": 0.0069,
"step": 1613
},
{
"epoch": 0.6624256105068747,
"grad_norm": 2.5081799030303955,
"learning_rate": 3.374384236453202e-06,
"loss": 0.0019,
"step": 1614
},
{
"epoch": 0.6628360352965319,
"grad_norm": 45.03568649291992,
"learning_rate": 3.3702791461412153e-06,
"loss": 0.0313,
"step": 1615
},
{
"epoch": 0.6632464600861893,
"grad_norm": 151.89785766601562,
"learning_rate": 3.3661740558292284e-06,
"loss": 0.1809,
"step": 1616
},
{
"epoch": 0.6636568848758465,
"grad_norm": 8.448513984680176,
"learning_rate": 3.362068965517242e-06,
"loss": 0.0053,
"step": 1617
},
{
"epoch": 0.6640673096655038,
"grad_norm": 1.4444835186004639,
"learning_rate": 3.3579638752052545e-06,
"loss": 0.0009,
"step": 1618
},
{
"epoch": 0.6644777344551611,
"grad_norm": 77.56280517578125,
"learning_rate": 3.353858784893268e-06,
"loss": 0.5004,
"step": 1619
},
{
"epoch": 0.6648881592448184,
"grad_norm": 2.4265871047973633,
"learning_rate": 3.349753694581281e-06,
"loss": 0.001,
"step": 1620
},
{
"epoch": 0.6652985840344757,
"grad_norm": 56.640724182128906,
"learning_rate": 3.345648604269294e-06,
"loss": 0.5946,
"step": 1621
},
{
"epoch": 0.665709008824133,
"grad_norm": 0.6659132242202759,
"learning_rate": 3.341543513957307e-06,
"loss": 0.0005,
"step": 1622
},
{
"epoch": 0.6661194336137902,
"grad_norm": 1.2735685110092163,
"learning_rate": 3.3374384236453207e-06,
"loss": 0.0005,
"step": 1623
},
{
"epoch": 0.6665298584034476,
"grad_norm": 2.022310495376587,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.0009,
"step": 1624
},
{
"epoch": 0.6669402831931048,
"grad_norm": 118.7616195678711,
"learning_rate": 3.329228243021347e-06,
"loss": 0.1807,
"step": 1625
},
{
"epoch": 0.6673507079827622,
"grad_norm": 1.9879318475723267,
"learning_rate": 3.32512315270936e-06,
"loss": 0.0008,
"step": 1626
},
{
"epoch": 0.6677611327724194,
"grad_norm": 7.494212627410889,
"learning_rate": 3.321018062397373e-06,
"loss": 0.0044,
"step": 1627
},
{
"epoch": 0.6681715575620768,
"grad_norm": 0.6684659719467163,
"learning_rate": 3.316912972085386e-06,
"loss": 0.0008,
"step": 1628
},
{
"epoch": 0.668581982351734,
"grad_norm": 1.3548862934112549,
"learning_rate": 3.3128078817733995e-06,
"loss": 0.0008,
"step": 1629
},
{
"epoch": 0.6689924071413913,
"grad_norm": 42.32175827026367,
"learning_rate": 3.308702791461412e-06,
"loss": 0.0248,
"step": 1630
},
{
"epoch": 0.6694028319310487,
"grad_norm": 70.48522186279297,
"learning_rate": 3.3045977011494256e-06,
"loss": 0.6018,
"step": 1631
},
{
"epoch": 0.6698132567207059,
"grad_norm": 2.5127832889556885,
"learning_rate": 3.3004926108374387e-06,
"loss": 0.0039,
"step": 1632
},
{
"epoch": 0.6702236815103633,
"grad_norm": 3.474381923675537,
"learning_rate": 3.296387520525452e-06,
"loss": 0.0022,
"step": 1633
},
{
"epoch": 0.6706341063000205,
"grad_norm": 8.46904468536377,
"learning_rate": 3.292282430213465e-06,
"loss": 0.0057,
"step": 1634
},
{
"epoch": 0.6710445310896778,
"grad_norm": 8.803572654724121,
"learning_rate": 3.2881773399014783e-06,
"loss": 0.0058,
"step": 1635
},
{
"epoch": 0.6714549558793351,
"grad_norm": 4.0813493728637695,
"learning_rate": 3.284072249589491e-06,
"loss": 0.0014,
"step": 1636
},
{
"epoch": 0.6718653806689924,
"grad_norm": 2.0117602348327637,
"learning_rate": 3.2799671592775044e-06,
"loss": 0.0012,
"step": 1637
},
{
"epoch": 0.6722758054586497,
"grad_norm": 9.080547332763672,
"learning_rate": 3.2758620689655175e-06,
"loss": 0.0063,
"step": 1638
},
{
"epoch": 0.672686230248307,
"grad_norm": 0.1470031589269638,
"learning_rate": 3.271756978653531e-06,
"loss": 0.0003,
"step": 1639
},
{
"epoch": 0.6730966550379643,
"grad_norm": 1.5364474058151245,
"learning_rate": 3.2676518883415436e-06,
"loss": 0.0013,
"step": 1640
},
{
"epoch": 0.6735070798276216,
"grad_norm": 1.20888352394104,
"learning_rate": 3.263546798029557e-06,
"loss": 0.0011,
"step": 1641
},
{
"epoch": 0.6739175046172788,
"grad_norm": 2.1010286808013916,
"learning_rate": 3.2594417077175698e-06,
"loss": 0.0012,
"step": 1642
},
{
"epoch": 0.6743279294069362,
"grad_norm": 6.265618324279785,
"learning_rate": 3.2553366174055832e-06,
"loss": 0.0035,
"step": 1643
},
{
"epoch": 0.6747383541965934,
"grad_norm": 6.477845668792725,
"learning_rate": 3.2512315270935963e-06,
"loss": 0.0062,
"step": 1644
},
{
"epoch": 0.6751487789862508,
"grad_norm": 21.314041137695312,
"learning_rate": 3.24712643678161e-06,
"loss": 0.0227,
"step": 1645
},
{
"epoch": 0.6755592037759081,
"grad_norm": 8.308874130249023,
"learning_rate": 3.2430213464696224e-06,
"loss": 0.0078,
"step": 1646
},
{
"epoch": 0.6759696285655653,
"grad_norm": 13.464950561523438,
"learning_rate": 3.238916256157636e-06,
"loss": 0.0104,
"step": 1647
},
{
"epoch": 0.6763800533552227,
"grad_norm": 0.6306696534156799,
"learning_rate": 3.234811165845649e-06,
"loss": 0.0005,
"step": 1648
},
{
"epoch": 0.6767904781448799,
"grad_norm": 9.468406677246094,
"learning_rate": 3.230706075533662e-06,
"loss": 0.0071,
"step": 1649
},
{
"epoch": 0.6772009029345373,
"grad_norm": 1.350114107131958,
"learning_rate": 3.226600985221675e-06,
"loss": 0.0006,
"step": 1650
},
{
"epoch": 0.6776113277241945,
"grad_norm": 6.334753513336182,
"learning_rate": 3.2224958949096886e-06,
"loss": 0.0031,
"step": 1651
},
{
"epoch": 0.6780217525138519,
"grad_norm": 2.9790263175964355,
"learning_rate": 3.2183908045977012e-06,
"loss": 0.0044,
"step": 1652
},
{
"epoch": 0.6784321773035091,
"grad_norm": 3.698438882827759,
"learning_rate": 3.2142857142857147e-06,
"loss": 0.0025,
"step": 1653
},
{
"epoch": 0.6788426020931664,
"grad_norm": 9.005804061889648,
"learning_rate": 3.210180623973728e-06,
"loss": 0.0061,
"step": 1654
},
{
"epoch": 0.6792530268828237,
"grad_norm": 1.5982589721679688,
"learning_rate": 3.206075533661741e-06,
"loss": 0.0005,
"step": 1655
},
{
"epoch": 0.679663451672481,
"grad_norm": 3.5061275959014893,
"learning_rate": 3.201970443349754e-06,
"loss": 0.0015,
"step": 1656
},
{
"epoch": 0.6800738764621383,
"grad_norm": 0.9127114415168762,
"learning_rate": 3.1978653530377674e-06,
"loss": 0.0017,
"step": 1657
},
{
"epoch": 0.6804843012517956,
"grad_norm": 75.44326782226562,
"learning_rate": 3.19376026272578e-06,
"loss": 0.0391,
"step": 1658
},
{
"epoch": 0.680894726041453,
"grad_norm": 8.253225326538086,
"learning_rate": 3.1896551724137935e-06,
"loss": 0.0064,
"step": 1659
},
{
"epoch": 0.6813051508311102,
"grad_norm": 8.943428039550781,
"learning_rate": 3.1855500821018066e-06,
"loss": 0.0065,
"step": 1660
},
{
"epoch": 0.6817155756207675,
"grad_norm": 2.8385157585144043,
"learning_rate": 3.1814449917898192e-06,
"loss": 0.0017,
"step": 1661
},
{
"epoch": 0.6821260004104248,
"grad_norm": 3.8180720806121826,
"learning_rate": 3.1773399014778327e-06,
"loss": 0.0015,
"step": 1662
},
{
"epoch": 0.6825364252000821,
"grad_norm": 2.483686923980713,
"learning_rate": 3.173234811165846e-06,
"loss": 0.0015,
"step": 1663
},
{
"epoch": 0.6829468499897394,
"grad_norm": 4.4074907302856445,
"learning_rate": 3.169129720853859e-06,
"loss": 0.0038,
"step": 1664
},
{
"epoch": 0.6833572747793967,
"grad_norm": 4.459950923919678,
"learning_rate": 3.165024630541872e-06,
"loss": 0.0018,
"step": 1665
},
{
"epoch": 0.6837676995690539,
"grad_norm": 108.27052307128906,
"learning_rate": 3.1609195402298854e-06,
"loss": 0.0959,
"step": 1666
},
{
"epoch": 0.6841781243587113,
"grad_norm": 1.2607488632202148,
"learning_rate": 3.156814449917898e-06,
"loss": 0.0014,
"step": 1667
},
{
"epoch": 0.6845885491483685,
"grad_norm": 17.275903701782227,
"learning_rate": 3.1527093596059115e-06,
"loss": 0.0072,
"step": 1668
},
{
"epoch": 0.6849989739380259,
"grad_norm": 4.489612102508545,
"learning_rate": 3.1486042692939246e-06,
"loss": 0.0024,
"step": 1669
},
{
"epoch": 0.6854093987276831,
"grad_norm": 1.5423874855041504,
"learning_rate": 3.144499178981938e-06,
"loss": 0.0019,
"step": 1670
},
{
"epoch": 0.6858198235173405,
"grad_norm": 3.3824098110198975,
"learning_rate": 3.1403940886699507e-06,
"loss": 0.0018,
"step": 1671
},
{
"epoch": 0.6862302483069977,
"grad_norm": 0.9304296374320984,
"learning_rate": 3.1362889983579642e-06,
"loss": 0.0004,
"step": 1672
},
{
"epoch": 0.686640673096655,
"grad_norm": 8.90660285949707,
"learning_rate": 3.132183908045977e-06,
"loss": 0.0061,
"step": 1673
},
{
"epoch": 0.6870510978863124,
"grad_norm": 2.4211432933807373,
"learning_rate": 3.1280788177339904e-06,
"loss": 0.0007,
"step": 1674
},
{
"epoch": 0.6874615226759696,
"grad_norm": 4.040661811828613,
"learning_rate": 3.1239737274220034e-06,
"loss": 0.0041,
"step": 1675
},
{
"epoch": 0.687871947465627,
"grad_norm": 2.5999324321746826,
"learning_rate": 3.119868637110017e-06,
"loss": 0.0013,
"step": 1676
},
{
"epoch": 0.6882823722552842,
"grad_norm": 303.9701232910156,
"learning_rate": 3.1157635467980295e-06,
"loss": 0.344,
"step": 1677
},
{
"epoch": 0.6886927970449415,
"grad_norm": 1.5300837755203247,
"learning_rate": 3.111658456486043e-06,
"loss": 0.0005,
"step": 1678
},
{
"epoch": 0.6891032218345988,
"grad_norm": 6.11021614074707,
"learning_rate": 3.107553366174056e-06,
"loss": 0.003,
"step": 1679
},
{
"epoch": 0.6895136466242561,
"grad_norm": 0.2761654853820801,
"learning_rate": 3.103448275862069e-06,
"loss": 0.0003,
"step": 1680
},
{
"epoch": 0.6899240714139134,
"grad_norm": 58.930152893066406,
"learning_rate": 3.0993431855500822e-06,
"loss": 0.5758,
"step": 1681
},
{
"epoch": 0.6903344962035707,
"grad_norm": 2.949009418487549,
"learning_rate": 3.0952380952380957e-06,
"loss": 0.0009,
"step": 1682
},
{
"epoch": 0.690744920993228,
"grad_norm": 1.7862677574157715,
"learning_rate": 3.0911330049261084e-06,
"loss": 0.0009,
"step": 1683
},
{
"epoch": 0.6911553457828853,
"grad_norm": 118.8187026977539,
"learning_rate": 3.087027914614122e-06,
"loss": 0.4376,
"step": 1684
},
{
"epoch": 0.6915657705725425,
"grad_norm": 5.356272220611572,
"learning_rate": 3.082922824302135e-06,
"loss": 0.0051,
"step": 1685
},
{
"epoch": 0.6919761953621999,
"grad_norm": 1.970411777496338,
"learning_rate": 3.078817733990148e-06,
"loss": 0.0011,
"step": 1686
},
{
"epoch": 0.6923866201518571,
"grad_norm": 10.418769836425781,
"learning_rate": 3.074712643678161e-06,
"loss": 0.0101,
"step": 1687
},
{
"epoch": 0.6927970449415145,
"grad_norm": 0.9992920160293579,
"learning_rate": 3.0706075533661745e-06,
"loss": 0.0007,
"step": 1688
},
{
"epoch": 0.6932074697311718,
"grad_norm": 5.811168193817139,
"learning_rate": 3.066502463054187e-06,
"loss": 0.0032,
"step": 1689
},
{
"epoch": 0.693617894520829,
"grad_norm": 4.979766368865967,
"learning_rate": 3.0623973727422007e-06,
"loss": 0.0022,
"step": 1690
},
{
"epoch": 0.6940283193104864,
"grad_norm": 2.8830883502960205,
"learning_rate": 3.0582922824302137e-06,
"loss": 0.0013,
"step": 1691
},
{
"epoch": 0.6944387441001436,
"grad_norm": 4.688591957092285,
"learning_rate": 3.054187192118227e-06,
"loss": 0.0019,
"step": 1692
},
{
"epoch": 0.694849168889801,
"grad_norm": 54.9259033203125,
"learning_rate": 3.05008210180624e-06,
"loss": 0.54,
"step": 1693
},
{
"epoch": 0.6952595936794582,
"grad_norm": 1.3389537334442139,
"learning_rate": 3.0459770114942533e-06,
"loss": 0.0006,
"step": 1694
},
{
"epoch": 0.6956700184691156,
"grad_norm": 5.892660617828369,
"learning_rate": 3.041871921182266e-06,
"loss": 0.0053,
"step": 1695
},
{
"epoch": 0.6960804432587728,
"grad_norm": 119.73632049560547,
"learning_rate": 3.0377668308702795e-06,
"loss": 0.1063,
"step": 1696
},
{
"epoch": 0.6964908680484301,
"grad_norm": 16.33275604248047,
"learning_rate": 3.0336617405582925e-06,
"loss": 0.043,
"step": 1697
},
{
"epoch": 0.6969012928380874,
"grad_norm": 5.878701210021973,
"learning_rate": 3.029556650246306e-06,
"loss": 0.0027,
"step": 1698
},
{
"epoch": 0.6973117176277447,
"grad_norm": 1.3243129253387451,
"learning_rate": 3.0254515599343187e-06,
"loss": 0.0009,
"step": 1699
},
{
"epoch": 0.697722142417402,
"grad_norm": 2.2633652687072754,
"learning_rate": 3.021346469622332e-06,
"loss": 0.0034,
"step": 1700
},
{
"epoch": 0.6981325672070593,
"grad_norm": 2.6514108180999756,
"learning_rate": 3.017241379310345e-06,
"loss": 0.001,
"step": 1701
},
{
"epoch": 0.6985429919967167,
"grad_norm": 3.867926597595215,
"learning_rate": 3.0131362889983583e-06,
"loss": 0.0016,
"step": 1702
},
{
"epoch": 0.6989534167863739,
"grad_norm": 4.302265644073486,
"learning_rate": 3.0090311986863713e-06,
"loss": 0.0015,
"step": 1703
},
{
"epoch": 0.6993638415760312,
"grad_norm": 6.881789207458496,
"learning_rate": 3.004926108374385e-06,
"loss": 0.0042,
"step": 1704
},
{
"epoch": 0.6997742663656885,
"grad_norm": 2.5011143684387207,
"learning_rate": 3.0008210180623975e-06,
"loss": 0.0012,
"step": 1705
},
{
"epoch": 0.7001846911553458,
"grad_norm": 1.0712230205535889,
"learning_rate": 2.996715927750411e-06,
"loss": 0.0013,
"step": 1706
},
{
"epoch": 0.7005951159450031,
"grad_norm": 1.7363107204437256,
"learning_rate": 2.992610837438424e-06,
"loss": 0.0004,
"step": 1707
},
{
"epoch": 0.7010055407346604,
"grad_norm": 2.6714284420013428,
"learning_rate": 2.988505747126437e-06,
"loss": 0.0008,
"step": 1708
},
{
"epoch": 0.7014159655243176,
"grad_norm": 11.298062324523926,
"learning_rate": 2.98440065681445e-06,
"loss": 0.0206,
"step": 1709
},
{
"epoch": 0.701826390313975,
"grad_norm": 140.8682098388672,
"learning_rate": 2.9802955665024636e-06,
"loss": 0.4993,
"step": 1710
},
{
"epoch": 0.7022368151036322,
"grad_norm": 3.890449285507202,
"learning_rate": 2.9761904761904763e-06,
"loss": 0.0032,
"step": 1711
},
{
"epoch": 0.7026472398932896,
"grad_norm": 4.849581241607666,
"learning_rate": 2.9720853858784898e-06,
"loss": 0.0021,
"step": 1712
},
{
"epoch": 0.7030576646829468,
"grad_norm": 5.9226202964782715,
"learning_rate": 2.967980295566503e-06,
"loss": 0.0027,
"step": 1713
},
{
"epoch": 0.7034680894726042,
"grad_norm": 8.299067497253418,
"learning_rate": 2.9638752052545163e-06,
"loss": 0.0052,
"step": 1714
},
{
"epoch": 0.7038785142622614,
"grad_norm": 8.285329818725586,
"learning_rate": 2.959770114942529e-06,
"loss": 0.0051,
"step": 1715
},
{
"epoch": 0.7042889390519187,
"grad_norm": 4.277763843536377,
"learning_rate": 2.9556650246305424e-06,
"loss": 0.002,
"step": 1716
},
{
"epoch": 0.704699363841576,
"grad_norm": 5.09735107421875,
"learning_rate": 2.951559934318555e-06,
"loss": 0.0026,
"step": 1717
},
{
"epoch": 0.7051097886312333,
"grad_norm": 0.7356064915657043,
"learning_rate": 2.947454844006568e-06,
"loss": 0.0008,
"step": 1718
},
{
"epoch": 0.7055202134208907,
"grad_norm": 0.8859161734580994,
"learning_rate": 2.9433497536945816e-06,
"loss": 0.0009,
"step": 1719
},
{
"epoch": 0.7059306382105479,
"grad_norm": 3.1467137336730957,
"learning_rate": 2.9392446633825943e-06,
"loss": 0.001,
"step": 1720
},
{
"epoch": 0.7063410630002052,
"grad_norm": 6.474218845367432,
"learning_rate": 2.9351395730706078e-06,
"loss": 0.0037,
"step": 1721
},
{
"epoch": 0.7067514877898625,
"grad_norm": 4.863564491271973,
"learning_rate": 2.931034482758621e-06,
"loss": 0.0021,
"step": 1722
},
{
"epoch": 0.7071619125795198,
"grad_norm": 1.6868122816085815,
"learning_rate": 2.926929392446634e-06,
"loss": 0.0009,
"step": 1723
},
{
"epoch": 0.7075723373691771,
"grad_norm": 7.0117011070251465,
"learning_rate": 2.922824302134647e-06,
"loss": 0.0038,
"step": 1724
},
{
"epoch": 0.7079827621588344,
"grad_norm": 1.8728100061416626,
"learning_rate": 2.9187192118226604e-06,
"loss": 0.0008,
"step": 1725
},
{
"epoch": 0.7083931869484917,
"grad_norm": 1.6929377317428589,
"learning_rate": 2.914614121510673e-06,
"loss": 0.0021,
"step": 1726
},
{
"epoch": 0.708803611738149,
"grad_norm": 7.063464164733887,
"learning_rate": 2.9105090311986866e-06,
"loss": 0.0056,
"step": 1727
},
{
"epoch": 0.7092140365278062,
"grad_norm": 3.620776414871216,
"learning_rate": 2.9064039408866996e-06,
"loss": 0.0019,
"step": 1728
},
{
"epoch": 0.7096244613174636,
"grad_norm": 1.7781239748001099,
"learning_rate": 2.902298850574713e-06,
"loss": 0.0023,
"step": 1729
},
{
"epoch": 0.7100348861071208,
"grad_norm": 146.65328979492188,
"learning_rate": 2.8981937602627258e-06,
"loss": 0.0616,
"step": 1730
},
{
"epoch": 0.7104453108967782,
"grad_norm": 4.017334938049316,
"learning_rate": 2.8940886699507393e-06,
"loss": 0.0033,
"step": 1731
},
{
"epoch": 0.7108557356864355,
"grad_norm": 117.36515808105469,
"learning_rate": 2.889983579638752e-06,
"loss": 0.5622,
"step": 1732
},
{
"epoch": 0.7112661604760927,
"grad_norm": 0.2897828221321106,
"learning_rate": 2.8858784893267654e-06,
"loss": 0.0003,
"step": 1733
},
{
"epoch": 0.7116765852657501,
"grad_norm": 1.81857168674469,
"learning_rate": 2.8817733990147784e-06,
"loss": 0.0007,
"step": 1734
},
{
"epoch": 0.7120870100554073,
"grad_norm": 10.188179016113281,
"learning_rate": 2.877668308702792e-06,
"loss": 0.008,
"step": 1735
},
{
"epoch": 0.7124974348450647,
"grad_norm": 3.2443203926086426,
"learning_rate": 2.8735632183908046e-06,
"loss": 0.0016,
"step": 1736
},
{
"epoch": 0.7129078596347219,
"grad_norm": 8.915731430053711,
"learning_rate": 2.869458128078818e-06,
"loss": 0.0038,
"step": 1737
},
{
"epoch": 0.7133182844243793,
"grad_norm": 3.4013304710388184,
"learning_rate": 2.865353037766831e-06,
"loss": 0.0037,
"step": 1738
},
{
"epoch": 0.7137287092140365,
"grad_norm": 0.20401613414287567,
"learning_rate": 2.861247947454844e-06,
"loss": 0.0004,
"step": 1739
},
{
"epoch": 0.7141391340036938,
"grad_norm": 0.4457375407218933,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.0006,
"step": 1740
},
{
"epoch": 0.7145495587933511,
"grad_norm": 8.259719848632812,
"learning_rate": 2.8530377668308707e-06,
"loss": 0.0051,
"step": 1741
},
{
"epoch": 0.7149599835830084,
"grad_norm": 386.2672424316406,
"learning_rate": 2.8489326765188834e-06,
"loss": 3.5315,
"step": 1742
},
{
"epoch": 0.7153704083726657,
"grad_norm": 11.39659309387207,
"learning_rate": 2.844827586206897e-06,
"loss": 0.0131,
"step": 1743
},
{
"epoch": 0.715780833162323,
"grad_norm": 1.0108168125152588,
"learning_rate": 2.84072249589491e-06,
"loss": 0.0007,
"step": 1744
},
{
"epoch": 0.7161912579519804,
"grad_norm": 1.7869186401367188,
"learning_rate": 2.836617405582923e-06,
"loss": 0.001,
"step": 1745
},
{
"epoch": 0.7166016827416376,
"grad_norm": 5.351038932800293,
"learning_rate": 2.832512315270936e-06,
"loss": 0.0051,
"step": 1746
},
{
"epoch": 0.7170121075312949,
"grad_norm": 2.936776638031006,
"learning_rate": 2.8284072249589496e-06,
"loss": 0.0009,
"step": 1747
},
{
"epoch": 0.7174225323209522,
"grad_norm": 3.854748487472534,
"learning_rate": 2.824302134646962e-06,
"loss": 0.002,
"step": 1748
},
{
"epoch": 0.7178329571106095,
"grad_norm": 2.4170145988464355,
"learning_rate": 2.8201970443349757e-06,
"loss": 0.0006,
"step": 1749
},
{
"epoch": 0.7182433819002668,
"grad_norm": 1.5164121389389038,
"learning_rate": 2.8160919540229887e-06,
"loss": 0.0013,
"step": 1750
},
{
"epoch": 0.7186538066899241,
"grad_norm": 9.010221481323242,
"learning_rate": 2.8119868637110022e-06,
"loss": 0.0078,
"step": 1751
},
{
"epoch": 0.7190642314795813,
"grad_norm": 3.3969566822052,
"learning_rate": 2.807881773399015e-06,
"loss": 0.0019,
"step": 1752
},
{
"epoch": 0.7194746562692387,
"grad_norm": 19.021427154541016,
"learning_rate": 2.8037766830870284e-06,
"loss": 0.0255,
"step": 1753
},
{
"epoch": 0.7198850810588959,
"grad_norm": 1.3397949934005737,
"learning_rate": 2.799671592775041e-06,
"loss": 0.0007,
"step": 1754
},
{
"epoch": 0.7202955058485533,
"grad_norm": 3.1240394115448,
"learning_rate": 2.7955665024630545e-06,
"loss": 0.0039,
"step": 1755
},
{
"epoch": 0.7207059306382105,
"grad_norm": 4.0964860916137695,
"learning_rate": 2.7914614121510676e-06,
"loss": 0.0037,
"step": 1756
},
{
"epoch": 0.7211163554278679,
"grad_norm": 24.345102310180664,
"learning_rate": 2.787356321839081e-06,
"loss": 0.0202,
"step": 1757
},
{
"epoch": 0.7215267802175251,
"grad_norm": 2.8003315925598145,
"learning_rate": 2.7832512315270937e-06,
"loss": 0.0011,
"step": 1758
},
{
"epoch": 0.7219372050071824,
"grad_norm": 60.72071838378906,
"learning_rate": 2.779146141215107e-06,
"loss": 0.7305,
"step": 1759
},
{
"epoch": 0.7223476297968398,
"grad_norm": 5.591928958892822,
"learning_rate": 2.7750410509031202e-06,
"loss": 0.0029,
"step": 1760
},
{
"epoch": 0.722758054586497,
"grad_norm": 4.695021629333496,
"learning_rate": 2.7709359605911333e-06,
"loss": 0.002,
"step": 1761
},
{
"epoch": 0.7231684793761544,
"grad_norm": 2.218022108078003,
"learning_rate": 2.7668308702791464e-06,
"loss": 0.0023,
"step": 1762
},
{
"epoch": 0.7235789041658116,
"grad_norm": 8.411977767944336,
"learning_rate": 2.76272577996716e-06,
"loss": 0.0077,
"step": 1763
},
{
"epoch": 0.7239893289554689,
"grad_norm": 4.89019250869751,
"learning_rate": 2.7586206896551725e-06,
"loss": 0.0023,
"step": 1764
},
{
"epoch": 0.7243997537451262,
"grad_norm": 5.954493045806885,
"learning_rate": 2.754515599343186e-06,
"loss": 0.0043,
"step": 1765
},
{
"epoch": 0.7248101785347835,
"grad_norm": 9.40628433227539,
"learning_rate": 2.750410509031199e-06,
"loss": 0.0086,
"step": 1766
},
{
"epoch": 0.7252206033244408,
"grad_norm": 7.449375152587891,
"learning_rate": 2.746305418719212e-06,
"loss": 0.0045,
"step": 1767
},
{
"epoch": 0.7256310281140981,
"grad_norm": 1.727170467376709,
"learning_rate": 2.742200328407225e-06,
"loss": 0.0009,
"step": 1768
},
{
"epoch": 0.7260414529037554,
"grad_norm": 0.9619573950767517,
"learning_rate": 2.7380952380952387e-06,
"loss": 0.0004,
"step": 1769
},
{
"epoch": 0.7264518776934127,
"grad_norm": 87.88870239257812,
"learning_rate": 2.7339901477832513e-06,
"loss": 0.2281,
"step": 1770
},
{
"epoch": 0.7268623024830699,
"grad_norm": 140.26438903808594,
"learning_rate": 2.729885057471265e-06,
"loss": 0.2228,
"step": 1771
},
{
"epoch": 0.7272727272727273,
"grad_norm": 20.6065673828125,
"learning_rate": 2.725779967159278e-06,
"loss": 0.0067,
"step": 1772
},
{
"epoch": 0.7276831520623845,
"grad_norm": 1.1657085418701172,
"learning_rate": 2.7216748768472905e-06,
"loss": 0.0016,
"step": 1773
},
{
"epoch": 0.7280935768520419,
"grad_norm": 5.506007194519043,
"learning_rate": 2.717569786535304e-06,
"loss": 0.0026,
"step": 1774
},
{
"epoch": 0.7285040016416992,
"grad_norm": 1.0102894306182861,
"learning_rate": 2.713464696223317e-06,
"loss": 0.0007,
"step": 1775
},
{
"epoch": 0.7289144264313564,
"grad_norm": 1.9247766733169556,
"learning_rate": 2.70935960591133e-06,
"loss": 0.0011,
"step": 1776
},
{
"epoch": 0.7293248512210138,
"grad_norm": 55.23438262939453,
"learning_rate": 2.705254515599343e-06,
"loss": 0.7587,
"step": 1777
},
{
"epoch": 0.729735276010671,
"grad_norm": 2.3482375144958496,
"learning_rate": 2.7011494252873567e-06,
"loss": 0.0011,
"step": 1778
},
{
"epoch": 0.7301457008003284,
"grad_norm": 2.4818296432495117,
"learning_rate": 2.6970443349753693e-06,
"loss": 0.0013,
"step": 1779
},
{
"epoch": 0.7305561255899856,
"grad_norm": 54.68598556518555,
"learning_rate": 2.692939244663383e-06,
"loss": 0.6279,
"step": 1780
},
{
"epoch": 0.730966550379643,
"grad_norm": 3.7740962505340576,
"learning_rate": 2.688834154351396e-06,
"loss": 0.0012,
"step": 1781
},
{
"epoch": 0.7313769751693002,
"grad_norm": 4.0678582191467285,
"learning_rate": 2.6847290640394093e-06,
"loss": 0.0022,
"step": 1782
},
{
"epoch": 0.7317873999589575,
"grad_norm": 1.2385218143463135,
"learning_rate": 2.680623973727422e-06,
"loss": 0.0006,
"step": 1783
},
{
"epoch": 0.7321978247486148,
"grad_norm": 4.705441951751709,
"learning_rate": 2.6765188834154355e-06,
"loss": 0.0023,
"step": 1784
},
{
"epoch": 0.7326082495382721,
"grad_norm": 1.6500025987625122,
"learning_rate": 2.672413793103448e-06,
"loss": 0.0011,
"step": 1785
},
{
"epoch": 0.7330186743279294,
"grad_norm": 3.948617935180664,
"learning_rate": 2.6683087027914616e-06,
"loss": 0.0017,
"step": 1786
},
{
"epoch": 0.7334290991175867,
"grad_norm": 57.763973236083984,
"learning_rate": 2.6642036124794747e-06,
"loss": 0.0493,
"step": 1787
},
{
"epoch": 0.733839523907244,
"grad_norm": 4.830066680908203,
"learning_rate": 2.660098522167488e-06,
"loss": 0.0023,
"step": 1788
},
{
"epoch": 0.7342499486969013,
"grad_norm": 5.771871566772461,
"learning_rate": 2.655993431855501e-06,
"loss": 0.0031,
"step": 1789
},
{
"epoch": 0.7346603734865585,
"grad_norm": 3.3376963138580322,
"learning_rate": 2.6518883415435143e-06,
"loss": 0.0026,
"step": 1790
},
{
"epoch": 0.7350707982762159,
"grad_norm": 1.079437017440796,
"learning_rate": 2.647783251231527e-06,
"loss": 0.0015,
"step": 1791
},
{
"epoch": 0.7354812230658732,
"grad_norm": 3.7522876262664795,
"learning_rate": 2.6436781609195404e-06,
"loss": 0.0059,
"step": 1792
},
{
"epoch": 0.7358916478555305,
"grad_norm": 9.526302337646484,
"learning_rate": 2.6395730706075535e-06,
"loss": 0.0083,
"step": 1793
},
{
"epoch": 0.7363020726451878,
"grad_norm": 8.065803527832031,
"learning_rate": 2.635467980295567e-06,
"loss": 0.006,
"step": 1794
},
{
"epoch": 0.736712497434845,
"grad_norm": 10.088679313659668,
"learning_rate": 2.6313628899835796e-06,
"loss": 0.0094,
"step": 1795
},
{
"epoch": 0.7371229222245024,
"grad_norm": 1.3803825378417969,
"learning_rate": 2.627257799671593e-06,
"loss": 0.0007,
"step": 1796
},
{
"epoch": 0.7375333470141596,
"grad_norm": 1.0153740644454956,
"learning_rate": 2.623152709359606e-06,
"loss": 0.0005,
"step": 1797
},
{
"epoch": 0.737943771803817,
"grad_norm": 223.10459899902344,
"learning_rate": 2.6190476190476192e-06,
"loss": 0.2084,
"step": 1798
},
{
"epoch": 0.7383541965934742,
"grad_norm": 36.70653533935547,
"learning_rate": 2.6149425287356323e-06,
"loss": 0.0583,
"step": 1799
},
{
"epoch": 0.7387646213831316,
"grad_norm": 1.4940204620361328,
"learning_rate": 2.6108374384236458e-06,
"loss": 0.0004,
"step": 1800
},
{
"epoch": 0.7391750461727888,
"grad_norm": 4.37270975112915,
"learning_rate": 2.6067323481116584e-06,
"loss": 0.0022,
"step": 1801
},
{
"epoch": 0.7395854709624461,
"grad_norm": 61.442535400390625,
"learning_rate": 2.602627257799672e-06,
"loss": 0.0288,
"step": 1802
},
{
"epoch": 0.7399958957521034,
"grad_norm": 2.4429192543029785,
"learning_rate": 2.598522167487685e-06,
"loss": 0.0024,
"step": 1803
},
{
"epoch": 0.7404063205417607,
"grad_norm": 2.01938533782959,
"learning_rate": 2.594417077175698e-06,
"loss": 0.0019,
"step": 1804
},
{
"epoch": 0.7408167453314181,
"grad_norm": 1.1255134344100952,
"learning_rate": 2.590311986863711e-06,
"loss": 0.0008,
"step": 1805
},
{
"epoch": 0.7412271701210753,
"grad_norm": 3.1225945949554443,
"learning_rate": 2.5862068965517246e-06,
"loss": 0.0009,
"step": 1806
},
{
"epoch": 0.7416375949107326,
"grad_norm": 1.957938313484192,
"learning_rate": 2.5821018062397372e-06,
"loss": 0.0006,
"step": 1807
},
{
"epoch": 0.7420480197003899,
"grad_norm": 5.877543926239014,
"learning_rate": 2.5779967159277507e-06,
"loss": 0.0028,
"step": 1808
},
{
"epoch": 0.7424584444900472,
"grad_norm": 4.319574356079102,
"learning_rate": 2.5738916256157638e-06,
"loss": 0.0031,
"step": 1809
},
{
"epoch": 0.7428688692797045,
"grad_norm": 60.285797119140625,
"learning_rate": 2.5697865353037773e-06,
"loss": 0.5277,
"step": 1810
},
{
"epoch": 0.7432792940693618,
"grad_norm": 3.361220359802246,
"learning_rate": 2.56568144499179e-06,
"loss": 0.0012,
"step": 1811
},
{
"epoch": 0.7436897188590191,
"grad_norm": 3.710512638092041,
"learning_rate": 2.5615763546798034e-06,
"loss": 0.0014,
"step": 1812
},
{
"epoch": 0.7441001436486764,
"grad_norm": 7.4387664794921875,
"learning_rate": 2.557471264367816e-06,
"loss": 0.0071,
"step": 1813
},
{
"epoch": 0.7445105684383336,
"grad_norm": 8.903318405151367,
"learning_rate": 2.5533661740558295e-06,
"loss": 0.0036,
"step": 1814
},
{
"epoch": 0.744920993227991,
"grad_norm": 44.64363479614258,
"learning_rate": 2.5492610837438426e-06,
"loss": 0.0194,
"step": 1815
},
{
"epoch": 0.7453314180176482,
"grad_norm": 6.1879754066467285,
"learning_rate": 2.545155993431856e-06,
"loss": 0.0033,
"step": 1816
},
{
"epoch": 0.7457418428073056,
"grad_norm": 5.424530029296875,
"learning_rate": 2.5410509031198687e-06,
"loss": 0.0025,
"step": 1817
},
{
"epoch": 0.7461522675969628,
"grad_norm": 9.184375762939453,
"learning_rate": 2.536945812807882e-06,
"loss": 0.0081,
"step": 1818
},
{
"epoch": 0.7465626923866201,
"grad_norm": 14.942904472351074,
"learning_rate": 2.5328407224958953e-06,
"loss": 0.0393,
"step": 1819
},
{
"epoch": 0.7469731171762775,
"grad_norm": 3.2511539459228516,
"learning_rate": 2.5287356321839083e-06,
"loss": 0.0018,
"step": 1820
},
{
"epoch": 0.7473835419659347,
"grad_norm": 2.477485418319702,
"learning_rate": 2.5246305418719214e-06,
"loss": 0.0021,
"step": 1821
},
{
"epoch": 0.7477939667555921,
"grad_norm": 7.637425422668457,
"learning_rate": 2.520525451559935e-06,
"loss": 0.0041,
"step": 1822
},
{
"epoch": 0.7482043915452493,
"grad_norm": 1.6476895809173584,
"learning_rate": 2.5164203612479475e-06,
"loss": 0.0024,
"step": 1823
},
{
"epoch": 0.7486148163349067,
"grad_norm": 4.551702976226807,
"learning_rate": 2.512315270935961e-06,
"loss": 0.0028,
"step": 1824
},
{
"epoch": 0.7490252411245639,
"grad_norm": 3.9219398498535156,
"learning_rate": 2.508210180623974e-06,
"loss": 0.0015,
"step": 1825
},
{
"epoch": 0.7494356659142212,
"grad_norm": 2.38871693611145,
"learning_rate": 2.504105090311987e-06,
"loss": 0.0008,
"step": 1826
},
{
"epoch": 0.7498460907038785,
"grad_norm": 0.8471001982688904,
"learning_rate": 2.5e-06,
"loss": 0.0005,
"step": 1827
},
{
"epoch": 0.7502565154935358,
"grad_norm": 0.46941304206848145,
"learning_rate": 2.4958949096880133e-06,
"loss": 0.0003,
"step": 1828
},
{
"epoch": 0.7506669402831931,
"grad_norm": 127.58653259277344,
"learning_rate": 2.4917898193760263e-06,
"loss": 0.079,
"step": 1829
},
{
"epoch": 0.7510773650728504,
"grad_norm": 0.6094871759414673,
"learning_rate": 2.48768472906404e-06,
"loss": 0.0005,
"step": 1830
},
{
"epoch": 0.7514877898625077,
"grad_norm": 1.7991559505462646,
"learning_rate": 2.483579638752053e-06,
"loss": 0.0008,
"step": 1831
},
{
"epoch": 0.751898214652165,
"grad_norm": 4.724639892578125,
"learning_rate": 2.479474548440066e-06,
"loss": 0.0021,
"step": 1832
},
{
"epoch": 0.7523086394418222,
"grad_norm": 65.07522583007812,
"learning_rate": 2.475369458128079e-06,
"loss": 0.543,
"step": 1833
},
{
"epoch": 0.7527190642314796,
"grad_norm": 5.241346836090088,
"learning_rate": 2.471264367816092e-06,
"loss": 0.004,
"step": 1834
},
{
"epoch": 0.7531294890211369,
"grad_norm": 3.883795976638794,
"learning_rate": 2.467159277504105e-06,
"loss": 0.0017,
"step": 1835
},
{
"epoch": 0.7535399138107942,
"grad_norm": 8.09839153289795,
"learning_rate": 2.4630541871921186e-06,
"loss": 0.0053,
"step": 1836
},
{
"epoch": 0.7539503386004515,
"grad_norm": 1.7221275568008423,
"learning_rate": 2.4589490968801317e-06,
"loss": 0.0004,
"step": 1837
},
{
"epoch": 0.7543607633901087,
"grad_norm": 3.316983699798584,
"learning_rate": 2.4548440065681448e-06,
"loss": 0.0019,
"step": 1838
},
{
"epoch": 0.7547711881797661,
"grad_norm": 0.6631819605827332,
"learning_rate": 2.450738916256158e-06,
"loss": 0.0006,
"step": 1839
},
{
"epoch": 0.7551816129694233,
"grad_norm": 6.961943626403809,
"learning_rate": 2.446633825944171e-06,
"loss": 0.0048,
"step": 1840
},
{
"epoch": 0.7555920377590807,
"grad_norm": 54.20858383178711,
"learning_rate": 2.4425287356321844e-06,
"loss": 0.6021,
"step": 1841
},
{
"epoch": 0.7560024625487379,
"grad_norm": 2.483574628829956,
"learning_rate": 2.4384236453201974e-06,
"loss": 0.0016,
"step": 1842
},
{
"epoch": 0.7564128873383953,
"grad_norm": 1.5103100538253784,
"learning_rate": 2.4343185550082105e-06,
"loss": 0.0012,
"step": 1843
},
{
"epoch": 0.7568233121280525,
"grad_norm": 11.679078102111816,
"learning_rate": 2.4302134646962236e-06,
"loss": 0.0183,
"step": 1844
},
{
"epoch": 0.7572337369177098,
"grad_norm": 6.317850589752197,
"learning_rate": 2.4261083743842366e-06,
"loss": 0.0034,
"step": 1845
},
{
"epoch": 0.7576441617073671,
"grad_norm": 3.4411873817443848,
"learning_rate": 2.4220032840722497e-06,
"loss": 0.0009,
"step": 1846
},
{
"epoch": 0.7580545864970244,
"grad_norm": 5.234242916107178,
"learning_rate": 2.417898193760263e-06,
"loss": 0.0036,
"step": 1847
},
{
"epoch": 0.7584650112866818,
"grad_norm": 4.564059257507324,
"learning_rate": 2.4137931034482762e-06,
"loss": 0.0019,
"step": 1848
},
{
"epoch": 0.758875436076339,
"grad_norm": 2.06823992729187,
"learning_rate": 2.4096880131362893e-06,
"loss": 0.001,
"step": 1849
},
{
"epoch": 0.7592858608659963,
"grad_norm": 0.25391143560409546,
"learning_rate": 2.4055829228243024e-06,
"loss": 0.0006,
"step": 1850
},
{
"epoch": 0.7596962856556536,
"grad_norm": 59.320377349853516,
"learning_rate": 2.4014778325123154e-06,
"loss": 0.6416,
"step": 1851
},
{
"epoch": 0.7601067104453109,
"grad_norm": 5.715360164642334,
"learning_rate": 2.397372742200329e-06,
"loss": 0.0123,
"step": 1852
},
{
"epoch": 0.7605171352349682,
"grad_norm": 0.5541730523109436,
"learning_rate": 2.393267651888342e-06,
"loss": 0.001,
"step": 1853
},
{
"epoch": 0.7609275600246255,
"grad_norm": 2.531996488571167,
"learning_rate": 2.389162561576355e-06,
"loss": 0.0009,
"step": 1854
},
{
"epoch": 0.7613379848142828,
"grad_norm": 70.10574340820312,
"learning_rate": 2.3850574712643677e-06,
"loss": 0.7349,
"step": 1855
},
{
"epoch": 0.7617484096039401,
"grad_norm": 10.588346481323242,
"learning_rate": 2.380952380952381e-06,
"loss": 0.0103,
"step": 1856
},
{
"epoch": 0.7621588343935973,
"grad_norm": 6.1468987464904785,
"learning_rate": 2.3768472906403943e-06,
"loss": 0.0028,
"step": 1857
},
{
"epoch": 0.7625692591832547,
"grad_norm": 3.401718854904175,
"learning_rate": 2.3727422003284073e-06,
"loss": 0.0023,
"step": 1858
},
{
"epoch": 0.7629796839729119,
"grad_norm": 4.722829341888428,
"learning_rate": 2.3686371100164204e-06,
"loss": 0.0036,
"step": 1859
},
{
"epoch": 0.7633901087625693,
"grad_norm": 11.005879402160645,
"learning_rate": 2.3645320197044334e-06,
"loss": 0.0222,
"step": 1860
},
{
"epoch": 0.7638005335522265,
"grad_norm": 5.0078887939453125,
"learning_rate": 2.3604269293924465e-06,
"loss": 0.0036,
"step": 1861
},
{
"epoch": 0.7642109583418838,
"grad_norm": 3.654003620147705,
"learning_rate": 2.35632183908046e-06,
"loss": 0.002,
"step": 1862
},
{
"epoch": 0.7646213831315412,
"grad_norm": 4.98245096206665,
"learning_rate": 2.352216748768473e-06,
"loss": 0.0054,
"step": 1863
},
{
"epoch": 0.7650318079211984,
"grad_norm": 2.5934438705444336,
"learning_rate": 2.348111658456486e-06,
"loss": 0.0013,
"step": 1864
},
{
"epoch": 0.7654422327108558,
"grad_norm": 3.363818645477295,
"learning_rate": 2.344006568144499e-06,
"loss": 0.0027,
"step": 1865
},
{
"epoch": 0.765852657500513,
"grad_norm": 1.1064995527267456,
"learning_rate": 2.3399014778325123e-06,
"loss": 0.0007,
"step": 1866
},
{
"epoch": 0.7662630822901704,
"grad_norm": 4.390050411224365,
"learning_rate": 2.3357963875205257e-06,
"loss": 0.0018,
"step": 1867
},
{
"epoch": 0.7666735070798276,
"grad_norm": 1.8092371225357056,
"learning_rate": 2.331691297208539e-06,
"loss": 0.0013,
"step": 1868
},
{
"epoch": 0.7670839318694849,
"grad_norm": 9.223669052124023,
"learning_rate": 2.327586206896552e-06,
"loss": 0.0075,
"step": 1869
},
{
"epoch": 0.7674943566591422,
"grad_norm": 1.3455674648284912,
"learning_rate": 2.323481116584565e-06,
"loss": 0.0008,
"step": 1870
},
{
"epoch": 0.7679047814487995,
"grad_norm": 5.38093900680542,
"learning_rate": 2.319376026272578e-06,
"loss": 0.003,
"step": 1871
},
{
"epoch": 0.7683152062384568,
"grad_norm": 3.4971182346343994,
"learning_rate": 2.315270935960591e-06,
"loss": 0.002,
"step": 1872
},
{
"epoch": 0.7687256310281141,
"grad_norm": 1.138192892074585,
"learning_rate": 2.3111658456486046e-06,
"loss": 0.0008,
"step": 1873
},
{
"epoch": 0.7691360558177714,
"grad_norm": 4.474194526672363,
"learning_rate": 2.3070607553366176e-06,
"loss": 0.0018,
"step": 1874
},
{
"epoch": 0.7695464806074287,
"grad_norm": 7.659038066864014,
"learning_rate": 2.3029556650246307e-06,
"loss": 0.005,
"step": 1875
},
{
"epoch": 0.7699569053970859,
"grad_norm": 96.67131805419922,
"learning_rate": 2.2988505747126437e-06,
"loss": 1.1445,
"step": 1876
},
{
"epoch": 0.7703673301867433,
"grad_norm": 0.3868589699268341,
"learning_rate": 2.294745484400657e-06,
"loss": 0.0006,
"step": 1877
},
{
"epoch": 0.7707777549764006,
"grad_norm": 12.768189430236816,
"learning_rate": 2.2906403940886703e-06,
"loss": 0.0136,
"step": 1878
},
{
"epoch": 0.7711881797660579,
"grad_norm": 5.7690510749816895,
"learning_rate": 2.2865353037766834e-06,
"loss": 0.0039,
"step": 1879
},
{
"epoch": 0.7715986045557152,
"grad_norm": 4.802737712860107,
"learning_rate": 2.2824302134646964e-06,
"loss": 0.0051,
"step": 1880
},
{
"epoch": 0.7720090293453724,
"grad_norm": 1.3911082744598389,
"learning_rate": 2.2783251231527095e-06,
"loss": 0.0009,
"step": 1881
},
{
"epoch": 0.7724194541350298,
"grad_norm": 1.4041454792022705,
"learning_rate": 2.2742200328407226e-06,
"loss": 0.0005,
"step": 1882
},
{
"epoch": 0.772829878924687,
"grad_norm": 527.201171875,
"learning_rate": 2.2701149425287356e-06,
"loss": 0.954,
"step": 1883
},
{
"epoch": 0.7732403037143444,
"grad_norm": 2.9974329471588135,
"learning_rate": 2.266009852216749e-06,
"loss": 0.001,
"step": 1884
},
{
"epoch": 0.7736507285040016,
"grad_norm": 1.6187399625778198,
"learning_rate": 2.261904761904762e-06,
"loss": 0.0022,
"step": 1885
},
{
"epoch": 0.774061153293659,
"grad_norm": 3.705063819885254,
"learning_rate": 2.2577996715927752e-06,
"loss": 0.0026,
"step": 1886
},
{
"epoch": 0.7744715780833162,
"grad_norm": 1.4835983514785767,
"learning_rate": 2.2536945812807883e-06,
"loss": 0.0006,
"step": 1887
},
{
"epoch": 0.7748820028729735,
"grad_norm": 4.038225173950195,
"learning_rate": 2.2495894909688014e-06,
"loss": 0.0041,
"step": 1888
},
{
"epoch": 0.7752924276626308,
"grad_norm": 1.6489189863204956,
"learning_rate": 2.245484400656815e-06,
"loss": 0.0014,
"step": 1889
},
{
"epoch": 0.7757028524522881,
"grad_norm": 1.9301745891571045,
"learning_rate": 2.241379310344828e-06,
"loss": 0.0008,
"step": 1890
},
{
"epoch": 0.7761132772419455,
"grad_norm": 2.65002179145813,
"learning_rate": 2.237274220032841e-06,
"loss": 0.0023,
"step": 1891
},
{
"epoch": 0.7765237020316027,
"grad_norm": 9.618758201599121,
"learning_rate": 2.233169129720854e-06,
"loss": 0.0125,
"step": 1892
},
{
"epoch": 0.77693412682126,
"grad_norm": 93.1954574584961,
"learning_rate": 2.229064039408867e-06,
"loss": 0.0764,
"step": 1893
},
{
"epoch": 0.7773445516109173,
"grad_norm": 6.456679344177246,
"learning_rate": 2.22495894909688e-06,
"loss": 0.0036,
"step": 1894
},
{
"epoch": 0.7777549764005746,
"grad_norm": 6.486172676086426,
"learning_rate": 2.2208538587848937e-06,
"loss": 0.0058,
"step": 1895
},
{
"epoch": 0.7781654011902319,
"grad_norm": 3.803011417388916,
"learning_rate": 2.2167487684729067e-06,
"loss": 0.0076,
"step": 1896
},
{
"epoch": 0.7785758259798892,
"grad_norm": 0.2337176501750946,
"learning_rate": 2.21264367816092e-06,
"loss": 0.0005,
"step": 1897
},
{
"epoch": 0.7789862507695465,
"grad_norm": 11.978750228881836,
"learning_rate": 2.208538587848933e-06,
"loss": 0.0176,
"step": 1898
},
{
"epoch": 0.7793966755592038,
"grad_norm": 3.1180338859558105,
"learning_rate": 2.204433497536946e-06,
"loss": 0.0014,
"step": 1899
},
{
"epoch": 0.779807100348861,
"grad_norm": 0.7417744994163513,
"learning_rate": 2.2003284072249594e-06,
"loss": 0.0019,
"step": 1900
},
{
"epoch": 0.7802175251385184,
"grad_norm": 3.067805528640747,
"learning_rate": 2.1962233169129725e-06,
"loss": 0.0013,
"step": 1901
},
{
"epoch": 0.7806279499281756,
"grad_norm": 3.004901647567749,
"learning_rate": 2.1921182266009855e-06,
"loss": 0.0047,
"step": 1902
},
{
"epoch": 0.781038374717833,
"grad_norm": 1.791553258895874,
"learning_rate": 2.1880131362889986e-06,
"loss": 0.0023,
"step": 1903
},
{
"epoch": 0.7814487995074902,
"grad_norm": 1.4738259315490723,
"learning_rate": 2.1839080459770117e-06,
"loss": 0.0012,
"step": 1904
},
{
"epoch": 0.7818592242971475,
"grad_norm": 1.6063909530639648,
"learning_rate": 2.1798029556650247e-06,
"loss": 0.0007,
"step": 1905
},
{
"epoch": 0.7822696490868049,
"grad_norm": 0.10639221966266632,
"learning_rate": 2.1756978653530382e-06,
"loss": 0.0003,
"step": 1906
},
{
"epoch": 0.7826800738764621,
"grad_norm": 1.543922781944275,
"learning_rate": 2.1715927750410513e-06,
"loss": 0.0005,
"step": 1907
},
{
"epoch": 0.7830904986661195,
"grad_norm": 39.642173767089844,
"learning_rate": 2.1674876847290643e-06,
"loss": 0.0326,
"step": 1908
},
{
"epoch": 0.7835009234557767,
"grad_norm": 2.116955041885376,
"learning_rate": 2.1633825944170774e-06,
"loss": 0.0011,
"step": 1909
},
{
"epoch": 0.7839113482454341,
"grad_norm": 0.9525993466377258,
"learning_rate": 2.1592775041050905e-06,
"loss": 0.0009,
"step": 1910
},
{
"epoch": 0.7843217730350913,
"grad_norm": 3.7287814617156982,
"learning_rate": 2.1551724137931035e-06,
"loss": 0.0013,
"step": 1911
},
{
"epoch": 0.7847321978247486,
"grad_norm": 4.359703540802002,
"learning_rate": 2.1510673234811166e-06,
"loss": 0.0015,
"step": 1912
},
{
"epoch": 0.7851426226144059,
"grad_norm": 3.5863497257232666,
"learning_rate": 2.1469622331691297e-06,
"loss": 0.0015,
"step": 1913
},
{
"epoch": 0.7855530474040632,
"grad_norm": 0.9854378700256348,
"learning_rate": 2.1428571428571427e-06,
"loss": 0.0005,
"step": 1914
},
{
"epoch": 0.7859634721937205,
"grad_norm": 0.5441867113113403,
"learning_rate": 2.1387520525451562e-06,
"loss": 0.0007,
"step": 1915
},
{
"epoch": 0.7863738969833778,
"grad_norm": 2.84483003616333,
"learning_rate": 2.1346469622331693e-06,
"loss": 0.0011,
"step": 1916
},
{
"epoch": 0.7867843217730351,
"grad_norm": 1.3996257781982422,
"learning_rate": 2.1305418719211823e-06,
"loss": 0.0006,
"step": 1917
},
{
"epoch": 0.7871947465626924,
"grad_norm": 8.8582763671875,
"learning_rate": 2.1264367816091954e-06,
"loss": 0.006,
"step": 1918
},
{
"epoch": 0.7876051713523496,
"grad_norm": 31.40988540649414,
"learning_rate": 2.1223316912972085e-06,
"loss": 0.0211,
"step": 1919
},
{
"epoch": 0.788015596142007,
"grad_norm": 0.830303430557251,
"learning_rate": 2.118226600985222e-06,
"loss": 0.0008,
"step": 1920
},
{
"epoch": 0.7884260209316643,
"grad_norm": 1.0683906078338623,
"learning_rate": 2.114121510673235e-06,
"loss": 0.0007,
"step": 1921
},
{
"epoch": 0.7888364457213216,
"grad_norm": 3.9102957248687744,
"learning_rate": 2.110016420361248e-06,
"loss": 0.003,
"step": 1922
},
{
"epoch": 0.7892468705109789,
"grad_norm": 1.3141251802444458,
"learning_rate": 2.105911330049261e-06,
"loss": 0.0003,
"step": 1923
},
{
"epoch": 0.7896572953006361,
"grad_norm": 3.7058238983154297,
"learning_rate": 2.1018062397372742e-06,
"loss": 0.0017,
"step": 1924
},
{
"epoch": 0.7900677200902935,
"grad_norm": 173.9900665283203,
"learning_rate": 2.0977011494252873e-06,
"loss": 0.1517,
"step": 1925
},
{
"epoch": 0.7904781448799507,
"grad_norm": 6.058109760284424,
"learning_rate": 2.0935960591133008e-06,
"loss": 0.0039,
"step": 1926
},
{
"epoch": 0.7908885696696081,
"grad_norm": 0.7982239723205566,
"learning_rate": 2.089490968801314e-06,
"loss": 0.0004,
"step": 1927
},
{
"epoch": 0.7912989944592653,
"grad_norm": 3.12764835357666,
"learning_rate": 2.085385878489327e-06,
"loss": 0.0009,
"step": 1928
},
{
"epoch": 0.7917094192489227,
"grad_norm": 4.371438026428223,
"learning_rate": 2.08128078817734e-06,
"loss": 0.0019,
"step": 1929
},
{
"epoch": 0.7921198440385799,
"grad_norm": 3.8917617797851562,
"learning_rate": 2.077175697865353e-06,
"loss": 0.0014,
"step": 1930
},
{
"epoch": 0.7925302688282372,
"grad_norm": 3.2977142333984375,
"learning_rate": 2.0730706075533665e-06,
"loss": 0.0015,
"step": 1931
},
{
"epoch": 0.7929406936178945,
"grad_norm": 1.835898756980896,
"learning_rate": 2.0689655172413796e-06,
"loss": 0.003,
"step": 1932
},
{
"epoch": 0.7933511184075518,
"grad_norm": 57.17356491088867,
"learning_rate": 2.0648604269293926e-06,
"loss": 0.5791,
"step": 1933
},
{
"epoch": 0.7937615431972092,
"grad_norm": 9.054430961608887,
"learning_rate": 2.0607553366174057e-06,
"loss": 0.005,
"step": 1934
},
{
"epoch": 0.7941719679868664,
"grad_norm": 4.325497150421143,
"learning_rate": 2.0566502463054188e-06,
"loss": 0.0022,
"step": 1935
},
{
"epoch": 0.7945823927765236,
"grad_norm": 11.953618049621582,
"learning_rate": 2.052545155993432e-06,
"loss": 0.0117,
"step": 1936
},
{
"epoch": 0.794992817566181,
"grad_norm": 6.052687168121338,
"learning_rate": 2.0484400656814453e-06,
"loss": 0.003,
"step": 1937
},
{
"epoch": 0.7954032423558383,
"grad_norm": 5.179054260253906,
"learning_rate": 2.0443349753694584e-06,
"loss": 0.0023,
"step": 1938
},
{
"epoch": 0.7958136671454956,
"grad_norm": 2.1290230751037598,
"learning_rate": 2.0402298850574715e-06,
"loss": 0.0011,
"step": 1939
},
{
"epoch": 0.7962240919351529,
"grad_norm": 3.5093226432800293,
"learning_rate": 2.0361247947454845e-06,
"loss": 0.0011,
"step": 1940
},
{
"epoch": 0.7966345167248102,
"grad_norm": 3.4184255599975586,
"learning_rate": 2.0320197044334976e-06,
"loss": 0.0012,
"step": 1941
},
{
"epoch": 0.7970449415144675,
"grad_norm": 3.2808680534362793,
"learning_rate": 2.027914614121511e-06,
"loss": 0.001,
"step": 1942
},
{
"epoch": 0.7974553663041247,
"grad_norm": 23.93288230895996,
"learning_rate": 2.023809523809524e-06,
"loss": 0.0213,
"step": 1943
},
{
"epoch": 0.7978657910937821,
"grad_norm": 6.950771808624268,
"learning_rate": 2.019704433497537e-06,
"loss": 0.0037,
"step": 1944
},
{
"epoch": 0.7982762158834393,
"grad_norm": 6.473175048828125,
"learning_rate": 2.0155993431855503e-06,
"loss": 0.0046,
"step": 1945
},
{
"epoch": 0.7986866406730967,
"grad_norm": 12.730740547180176,
"learning_rate": 2.0114942528735633e-06,
"loss": 0.0073,
"step": 1946
},
{
"epoch": 0.7990970654627539,
"grad_norm": 10.024772644042969,
"learning_rate": 2.0073891625615764e-06,
"loss": 0.0051,
"step": 1947
},
{
"epoch": 0.7995074902524112,
"grad_norm": 7.000028133392334,
"learning_rate": 2.00328407224959e-06,
"loss": 0.0044,
"step": 1948
},
{
"epoch": 0.7999179150420686,
"grad_norm": 1.7929660081863403,
"learning_rate": 1.999178981937603e-06,
"loss": 0.0005,
"step": 1949
},
{
"epoch": 0.8003283398317258,
"grad_norm": 1.4799710512161255,
"learning_rate": 1.995073891625616e-06,
"loss": 0.0013,
"step": 1950
},
{
"epoch": 0.8007387646213832,
"grad_norm": 2.5182385444641113,
"learning_rate": 1.990968801313629e-06,
"loss": 0.0024,
"step": 1951
},
{
"epoch": 0.8011491894110404,
"grad_norm": 3.422564744949341,
"learning_rate": 1.986863711001642e-06,
"loss": 0.0015,
"step": 1952
},
{
"epoch": 0.8015596142006978,
"grad_norm": 2.3426549434661865,
"learning_rate": 1.982758620689655e-06,
"loss": 0.0012,
"step": 1953
},
{
"epoch": 0.801970038990355,
"grad_norm": 49.27020263671875,
"learning_rate": 1.9786535303776687e-06,
"loss": 0.5788,
"step": 1954
},
{
"epoch": 0.8023804637800123,
"grad_norm": 8.774833679199219,
"learning_rate": 1.9745484400656818e-06,
"loss": 0.0056,
"step": 1955
},
{
"epoch": 0.8027908885696696,
"grad_norm": 2.258565664291382,
"learning_rate": 1.970443349753695e-06,
"loss": 0.0025,
"step": 1956
},
{
"epoch": 0.8032013133593269,
"grad_norm": 4.170324802398682,
"learning_rate": 1.966338259441708e-06,
"loss": 0.0014,
"step": 1957
},
{
"epoch": 0.8036117381489842,
"grad_norm": 18.955150604248047,
"learning_rate": 1.962233169129721e-06,
"loss": 0.0069,
"step": 1958
},
{
"epoch": 0.8040221629386415,
"grad_norm": 48.227088928222656,
"learning_rate": 1.9581280788177344e-06,
"loss": 0.591,
"step": 1959
},
{
"epoch": 0.8044325877282988,
"grad_norm": 237.68084716796875,
"learning_rate": 1.9540229885057475e-06,
"loss": 0.3067,
"step": 1960
},
{
"epoch": 0.8048430125179561,
"grad_norm": 1.7197836637496948,
"learning_rate": 1.9499178981937606e-06,
"loss": 0.0005,
"step": 1961
},
{
"epoch": 0.8052534373076133,
"grad_norm": 9.634839057922363,
"learning_rate": 1.9458128078817736e-06,
"loss": 0.0067,
"step": 1962
},
{
"epoch": 0.8056638620972707,
"grad_norm": 7.381905555725098,
"learning_rate": 1.9417077175697867e-06,
"loss": 0.0047,
"step": 1963
},
{
"epoch": 0.806074286886928,
"grad_norm": 14.642587661743164,
"learning_rate": 1.9376026272577998e-06,
"loss": 0.0345,
"step": 1964
},
{
"epoch": 0.8064847116765853,
"grad_norm": 7.100318431854248,
"learning_rate": 1.9334975369458132e-06,
"loss": 0.0046,
"step": 1965
},
{
"epoch": 0.8068951364662426,
"grad_norm": 7.1860032081604,
"learning_rate": 1.929392446633826e-06,
"loss": 0.0127,
"step": 1966
},
{
"epoch": 0.8073055612558998,
"grad_norm": 0.6282884478569031,
"learning_rate": 1.925287356321839e-06,
"loss": 0.0006,
"step": 1967
},
{
"epoch": 0.8077159860455572,
"grad_norm": 2.5628530979156494,
"learning_rate": 1.9211822660098524e-06,
"loss": 0.0009,
"step": 1968
},
{
"epoch": 0.8081264108352144,
"grad_norm": 57.84944152832031,
"learning_rate": 1.9170771756978655e-06,
"loss": 0.6016,
"step": 1969
},
{
"epoch": 0.8085368356248718,
"grad_norm": 14.582354545593262,
"learning_rate": 1.9129720853858786e-06,
"loss": 0.0143,
"step": 1970
},
{
"epoch": 0.808947260414529,
"grad_norm": 8.436103820800781,
"learning_rate": 1.9088669950738916e-06,
"loss": 0.0055,
"step": 1971
},
{
"epoch": 0.8093576852041864,
"grad_norm": 1.6803200244903564,
"learning_rate": 1.904761904761905e-06,
"loss": 0.0007,
"step": 1972
},
{
"epoch": 0.8097681099938436,
"grad_norm": 4.183961868286133,
"learning_rate": 1.900656814449918e-06,
"loss": 0.0016,
"step": 1973
},
{
"epoch": 0.8101785347835009,
"grad_norm": 42.9221076965332,
"learning_rate": 1.896551724137931e-06,
"loss": 0.0135,
"step": 1974
},
{
"epoch": 0.8105889595731582,
"grad_norm": 0.8136171698570251,
"learning_rate": 1.8924466338259443e-06,
"loss": 0.0004,
"step": 1975
},
{
"epoch": 0.8109993843628155,
"grad_norm": 3.321906566619873,
"learning_rate": 1.8883415435139574e-06,
"loss": 0.0011,
"step": 1976
},
{
"epoch": 0.8114098091524728,
"grad_norm": 7.887020587921143,
"learning_rate": 1.8842364532019704e-06,
"loss": 0.0053,
"step": 1977
},
{
"epoch": 0.8118202339421301,
"grad_norm": 6.269182205200195,
"learning_rate": 1.8801313628899837e-06,
"loss": 0.0032,
"step": 1978
},
{
"epoch": 0.8122306587317873,
"grad_norm": 3.4489238262176514,
"learning_rate": 1.8760262725779968e-06,
"loss": 0.0028,
"step": 1979
},
{
"epoch": 0.8126410835214447,
"grad_norm": 1.9782854318618774,
"learning_rate": 1.8719211822660098e-06,
"loss": 0.0007,
"step": 1980
},
{
"epoch": 0.813051508311102,
"grad_norm": 1.2954609394073486,
"learning_rate": 1.8678160919540231e-06,
"loss": 0.0012,
"step": 1981
},
{
"epoch": 0.8134619331007593,
"grad_norm": 4.862510681152344,
"learning_rate": 1.8637110016420362e-06,
"loss": 0.0023,
"step": 1982
},
{
"epoch": 0.8138723578904166,
"grad_norm": 5.918731689453125,
"learning_rate": 1.8596059113300495e-06,
"loss": 0.0028,
"step": 1983
},
{
"epoch": 0.8142827826800739,
"grad_norm": 1.4226924180984497,
"learning_rate": 1.8555008210180625e-06,
"loss": 0.0008,
"step": 1984
},
{
"epoch": 0.8146932074697312,
"grad_norm": 2.211376190185547,
"learning_rate": 1.8513957307060756e-06,
"loss": 0.004,
"step": 1985
},
{
"epoch": 0.8151036322593884,
"grad_norm": 57.22856140136719,
"learning_rate": 1.8472906403940889e-06,
"loss": 0.5118,
"step": 1986
},
{
"epoch": 0.8155140570490458,
"grad_norm": 6.255296230316162,
"learning_rate": 1.843185550082102e-06,
"loss": 0.0031,
"step": 1987
},
{
"epoch": 0.815924481838703,
"grad_norm": 0.7972679138183594,
"learning_rate": 1.839080459770115e-06,
"loss": 0.0004,
"step": 1988
},
{
"epoch": 0.8163349066283604,
"grad_norm": 0.2298550009727478,
"learning_rate": 1.8349753694581283e-06,
"loss": 0.0004,
"step": 1989
},
{
"epoch": 0.8167453314180176,
"grad_norm": 2.974949359893799,
"learning_rate": 1.8308702791461413e-06,
"loss": 0.0048,
"step": 1990
},
{
"epoch": 0.8171557562076749,
"grad_norm": 2.8365256786346436,
"learning_rate": 1.8267651888341544e-06,
"loss": 0.0009,
"step": 1991
},
{
"epoch": 0.8175661809973322,
"grad_norm": 1.8940539360046387,
"learning_rate": 1.8226600985221677e-06,
"loss": 0.0006,
"step": 1992
},
{
"epoch": 0.8179766057869895,
"grad_norm": 1.8099260330200195,
"learning_rate": 1.8185550082101807e-06,
"loss": 0.0008,
"step": 1993
},
{
"epoch": 0.8183870305766469,
"grad_norm": 0.22202126681804657,
"learning_rate": 1.814449917898194e-06,
"loss": 0.0004,
"step": 1994
},
{
"epoch": 0.8187974553663041,
"grad_norm": 7.072625160217285,
"learning_rate": 1.810344827586207e-06,
"loss": 0.004,
"step": 1995
},
{
"epoch": 0.8192078801559615,
"grad_norm": 6.058025360107422,
"learning_rate": 1.8062397372742201e-06,
"loss": 0.0032,
"step": 1996
},
{
"epoch": 0.8196183049456187,
"grad_norm": 7.244338035583496,
"learning_rate": 1.8021346469622334e-06,
"loss": 0.0041,
"step": 1997
},
{
"epoch": 0.820028729735276,
"grad_norm": 3.7517342567443848,
"learning_rate": 1.7980295566502465e-06,
"loss": 0.0014,
"step": 1998
},
{
"epoch": 0.8204391545249333,
"grad_norm": 0.5613964200019836,
"learning_rate": 1.7939244663382596e-06,
"loss": 0.0003,
"step": 1999
},
{
"epoch": 0.8208495793145906,
"grad_norm": 1.5299192667007446,
"learning_rate": 1.7898193760262728e-06,
"loss": 0.0014,
"step": 2000
},
{
"epoch": 0.8212600041042479,
"grad_norm": 0.9930475354194641,
"learning_rate": 1.7857142857142859e-06,
"loss": 0.0014,
"step": 2001
},
{
"epoch": 0.8216704288939052,
"grad_norm": 5.368375301361084,
"learning_rate": 1.781609195402299e-06,
"loss": 0.0048,
"step": 2002
},
{
"epoch": 0.8220808536835625,
"grad_norm": 0.853631317615509,
"learning_rate": 1.7775041050903122e-06,
"loss": 0.0006,
"step": 2003
},
{
"epoch": 0.8224912784732198,
"grad_norm": 5.655002593994141,
"learning_rate": 1.7733990147783253e-06,
"loss": 0.0026,
"step": 2004
},
{
"epoch": 0.822901703262877,
"grad_norm": 4.9236674308776855,
"learning_rate": 1.7692939244663384e-06,
"loss": 0.0035,
"step": 2005
},
{
"epoch": 0.8233121280525344,
"grad_norm": 3.6850600242614746,
"learning_rate": 1.7651888341543516e-06,
"loss": 0.0037,
"step": 2006
},
{
"epoch": 0.8237225528421916,
"grad_norm": 2.7300682067871094,
"learning_rate": 1.7610837438423647e-06,
"loss": 0.0013,
"step": 2007
},
{
"epoch": 0.824132977631849,
"grad_norm": 4.209996700286865,
"learning_rate": 1.756978653530378e-06,
"loss": 0.002,
"step": 2008
},
{
"epoch": 0.8245434024215063,
"grad_norm": 2.1867096424102783,
"learning_rate": 1.752873563218391e-06,
"loss": 0.0008,
"step": 2009
},
{
"epoch": 0.8249538272111635,
"grad_norm": 4.019346714019775,
"learning_rate": 1.748768472906404e-06,
"loss": 0.0015,
"step": 2010
},
{
"epoch": 0.8253642520008209,
"grad_norm": 67.91747283935547,
"learning_rate": 1.7446633825944174e-06,
"loss": 0.6019,
"step": 2011
},
{
"epoch": 0.8257746767904781,
"grad_norm": 5.092018127441406,
"learning_rate": 1.7405582922824304e-06,
"loss": 0.0026,
"step": 2012
},
{
"epoch": 0.8261851015801355,
"grad_norm": 1.3320807218551636,
"learning_rate": 1.7364532019704435e-06,
"loss": 0.0013,
"step": 2013
},
{
"epoch": 0.8265955263697927,
"grad_norm": 5.210352897644043,
"learning_rate": 1.7323481116584568e-06,
"loss": 0.0023,
"step": 2014
},
{
"epoch": 0.8270059511594501,
"grad_norm": 3.1786768436431885,
"learning_rate": 1.7282430213464698e-06,
"loss": 0.002,
"step": 2015
},
{
"epoch": 0.8274163759491073,
"grad_norm": 4.868100643157959,
"learning_rate": 1.724137931034483e-06,
"loss": 0.0024,
"step": 2016
},
{
"epoch": 0.8278268007387646,
"grad_norm": 4.6474385261535645,
"learning_rate": 1.7200328407224962e-06,
"loss": 0.0036,
"step": 2017
},
{
"epoch": 0.8282372255284219,
"grad_norm": 84.08039093017578,
"learning_rate": 1.7159277504105093e-06,
"loss": 0.0422,
"step": 2018
},
{
"epoch": 0.8286476503180792,
"grad_norm": 32.934852600097656,
"learning_rate": 1.7118226600985225e-06,
"loss": 0.0103,
"step": 2019
},
{
"epoch": 0.8290580751077365,
"grad_norm": 2.267946720123291,
"learning_rate": 1.7077175697865356e-06,
"loss": 0.0039,
"step": 2020
},
{
"epoch": 0.8294684998973938,
"grad_norm": 59.974853515625,
"learning_rate": 1.7036124794745484e-06,
"loss": 0.4265,
"step": 2021
},
{
"epoch": 0.829878924687051,
"grad_norm": 0.9429296851158142,
"learning_rate": 1.6995073891625615e-06,
"loss": 0.0019,
"step": 2022
},
{
"epoch": 0.8302893494767084,
"grad_norm": 28.2806396484375,
"learning_rate": 1.6954022988505748e-06,
"loss": 0.0182,
"step": 2023
},
{
"epoch": 0.8306997742663657,
"grad_norm": 1.9367765188217163,
"learning_rate": 1.6912972085385879e-06,
"loss": 0.0015,
"step": 2024
},
{
"epoch": 0.831110199056023,
"grad_norm": 4.499584197998047,
"learning_rate": 1.687192118226601e-06,
"loss": 0.0036,
"step": 2025
},
{
"epoch": 0.8315206238456803,
"grad_norm": 2.8841772079467773,
"learning_rate": 1.6830870279146142e-06,
"loss": 0.0009,
"step": 2026
},
{
"epoch": 0.8319310486353376,
"grad_norm": 3.6246395111083984,
"learning_rate": 1.6789819376026273e-06,
"loss": 0.0017,
"step": 2027
},
{
"epoch": 0.8323414734249949,
"grad_norm": 0.5069646239280701,
"learning_rate": 1.6748768472906405e-06,
"loss": 0.0007,
"step": 2028
},
{
"epoch": 0.8327518982146521,
"grad_norm": 163.3522491455078,
"learning_rate": 1.6707717569786536e-06,
"loss": 0.2638,
"step": 2029
},
{
"epoch": 0.8331623230043095,
"grad_norm": 60.53173828125,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.5823,
"step": 2030
},
{
"epoch": 0.8335727477939667,
"grad_norm": 1.5301424264907837,
"learning_rate": 1.66256157635468e-06,
"loss": 0.0008,
"step": 2031
},
{
"epoch": 0.8339831725836241,
"grad_norm": 5.511871814727783,
"learning_rate": 1.658456486042693e-06,
"loss": 0.0033,
"step": 2032
},
{
"epoch": 0.8343935973732813,
"grad_norm": 3.6529266834259033,
"learning_rate": 1.654351395730706e-06,
"loss": 0.0032,
"step": 2033
},
{
"epoch": 0.8348040221629387,
"grad_norm": 5.143014430999756,
"learning_rate": 1.6502463054187193e-06,
"loss": 0.0021,
"step": 2034
},
{
"epoch": 0.835214446952596,
"grad_norm": 73.11016082763672,
"learning_rate": 1.6461412151067324e-06,
"loss": 0.0449,
"step": 2035
},
{
"epoch": 0.8356248717422532,
"grad_norm": 3.6816155910491943,
"learning_rate": 1.6420361247947455e-06,
"loss": 0.0014,
"step": 2036
},
{
"epoch": 0.8360352965319106,
"grad_norm": 0.718214213848114,
"learning_rate": 1.6379310344827587e-06,
"loss": 0.0007,
"step": 2037
},
{
"epoch": 0.8364457213215678,
"grad_norm": 1.9869035482406616,
"learning_rate": 1.6338259441707718e-06,
"loss": 0.0011,
"step": 2038
},
{
"epoch": 0.8368561461112252,
"grad_norm": 2.402268648147583,
"learning_rate": 1.6297208538587849e-06,
"loss": 0.001,
"step": 2039
},
{
"epoch": 0.8372665709008824,
"grad_norm": 1.4185056686401367,
"learning_rate": 1.6256157635467982e-06,
"loss": 0.0005,
"step": 2040
},
{
"epoch": 0.8376769956905397,
"grad_norm": 3.280837059020996,
"learning_rate": 1.6215106732348112e-06,
"loss": 0.0021,
"step": 2041
},
{
"epoch": 0.838087420480197,
"grad_norm": 7.002508163452148,
"learning_rate": 1.6174055829228245e-06,
"loss": 0.0057,
"step": 2042
},
{
"epoch": 0.8384978452698543,
"grad_norm": 1.321762204170227,
"learning_rate": 1.6133004926108376e-06,
"loss": 0.0009,
"step": 2043
},
{
"epoch": 0.8389082700595116,
"grad_norm": 6.491580009460449,
"learning_rate": 1.6091954022988506e-06,
"loss": 0.0033,
"step": 2044
},
{
"epoch": 0.8393186948491689,
"grad_norm": 0.9028833508491516,
"learning_rate": 1.605090311986864e-06,
"loss": 0.0017,
"step": 2045
},
{
"epoch": 0.8397291196388262,
"grad_norm": 9.430718421936035,
"learning_rate": 1.600985221674877e-06,
"loss": 0.0085,
"step": 2046
},
{
"epoch": 0.8401395444284835,
"grad_norm": 16.14491844177246,
"learning_rate": 1.59688013136289e-06,
"loss": 0.0061,
"step": 2047
},
{
"epoch": 0.8405499692181407,
"grad_norm": 6.151805400848389,
"learning_rate": 1.5927750410509033e-06,
"loss": 0.0036,
"step": 2048
},
{
"epoch": 0.8409603940077981,
"grad_norm": 2.237708330154419,
"learning_rate": 1.5886699507389164e-06,
"loss": 0.0006,
"step": 2049
},
{
"epoch": 0.8413708187974553,
"grad_norm": 1.5920524597167969,
"learning_rate": 1.5845648604269294e-06,
"loss": 0.0007,
"step": 2050
},
{
"epoch": 0.8417812435871127,
"grad_norm": 5.258532524108887,
"learning_rate": 1.5804597701149427e-06,
"loss": 0.0025,
"step": 2051
},
{
"epoch": 0.84219166837677,
"grad_norm": 54.998939514160156,
"learning_rate": 1.5763546798029558e-06,
"loss": 0.6615,
"step": 2052
},
{
"epoch": 0.8426020931664272,
"grad_norm": 0.4841395616531372,
"learning_rate": 1.572249589490969e-06,
"loss": 0.0003,
"step": 2053
},
{
"epoch": 0.8430125179560846,
"grad_norm": 5.585233688354492,
"learning_rate": 1.5681444991789821e-06,
"loss": 0.0055,
"step": 2054
},
{
"epoch": 0.8434229427457418,
"grad_norm": 1.2184664011001587,
"learning_rate": 1.5640394088669952e-06,
"loss": 0.0014,
"step": 2055
},
{
"epoch": 0.8438333675353992,
"grad_norm": 5.7302350997924805,
"learning_rate": 1.5599343185550085e-06,
"loss": 0.0112,
"step": 2056
},
{
"epoch": 0.8442437923250564,
"grad_norm": 7.946174621582031,
"learning_rate": 1.5558292282430215e-06,
"loss": 0.005,
"step": 2057
},
{
"epoch": 0.8446542171147138,
"grad_norm": 57.734317779541016,
"learning_rate": 1.5517241379310346e-06,
"loss": 0.0389,
"step": 2058
},
{
"epoch": 0.845064641904371,
"grad_norm": 4.1938347816467285,
"learning_rate": 1.5476190476190479e-06,
"loss": 0.0018,
"step": 2059
},
{
"epoch": 0.8454750666940283,
"grad_norm": 0.6884797811508179,
"learning_rate": 1.543513957307061e-06,
"loss": 0.0005,
"step": 2060
},
{
"epoch": 0.8458854914836856,
"grad_norm": 57.44015121459961,
"learning_rate": 1.539408866995074e-06,
"loss": 0.6062,
"step": 2061
},
{
"epoch": 0.8462959162733429,
"grad_norm": 5.821646690368652,
"learning_rate": 1.5353037766830873e-06,
"loss": 0.002,
"step": 2062
},
{
"epoch": 0.8467063410630002,
"grad_norm": 2.7798759937286377,
"learning_rate": 1.5311986863711003e-06,
"loss": 0.0013,
"step": 2063
},
{
"epoch": 0.8471167658526575,
"grad_norm": 2.8673994541168213,
"learning_rate": 1.5270935960591136e-06,
"loss": 0.0017,
"step": 2064
},
{
"epoch": 0.8475271906423147,
"grad_norm": 1.2268073558807373,
"learning_rate": 1.5229885057471267e-06,
"loss": 0.0006,
"step": 2065
},
{
"epoch": 0.8479376154319721,
"grad_norm": 3.339931011199951,
"learning_rate": 1.5188834154351397e-06,
"loss": 0.0012,
"step": 2066
},
{
"epoch": 0.8483480402216294,
"grad_norm": 4.370004177093506,
"learning_rate": 1.514778325123153e-06,
"loss": 0.0017,
"step": 2067
},
{
"epoch": 0.8487584650112867,
"grad_norm": 75.70170593261719,
"learning_rate": 1.510673234811166e-06,
"loss": 0.07,
"step": 2068
},
{
"epoch": 0.849168889800944,
"grad_norm": 1.0300614833831787,
"learning_rate": 1.5065681444991791e-06,
"loss": 0.0019,
"step": 2069
},
{
"epoch": 0.8495793145906013,
"grad_norm": 24.59672737121582,
"learning_rate": 1.5024630541871924e-06,
"loss": 0.0126,
"step": 2070
},
{
"epoch": 0.8499897393802586,
"grad_norm": 4.18646764755249,
"learning_rate": 1.4983579638752055e-06,
"loss": 0.0027,
"step": 2071
},
{
"epoch": 0.8504001641699158,
"grad_norm": 1.2641215324401855,
"learning_rate": 1.4942528735632185e-06,
"loss": 0.0015,
"step": 2072
},
{
"epoch": 0.8508105889595732,
"grad_norm": 4.257776737213135,
"learning_rate": 1.4901477832512318e-06,
"loss": 0.0023,
"step": 2073
},
{
"epoch": 0.8512210137492304,
"grad_norm": 6.014791488647461,
"learning_rate": 1.4860426929392449e-06,
"loss": 0.0073,
"step": 2074
},
{
"epoch": 0.8516314385388878,
"grad_norm": 1.5175026655197144,
"learning_rate": 1.4819376026272582e-06,
"loss": 0.0016,
"step": 2075
},
{
"epoch": 0.852041863328545,
"grad_norm": 0.8277934789657593,
"learning_rate": 1.4778325123152712e-06,
"loss": 0.0015,
"step": 2076
},
{
"epoch": 0.8524522881182024,
"grad_norm": 4.112128734588623,
"learning_rate": 1.473727422003284e-06,
"loss": 0.0016,
"step": 2077
},
{
"epoch": 0.8528627129078596,
"grad_norm": 0.8551503419876099,
"learning_rate": 1.4696223316912971e-06,
"loss": 0.0005,
"step": 2078
},
{
"epoch": 0.8532731376975169,
"grad_norm": 1.2928259372711182,
"learning_rate": 1.4655172413793104e-06,
"loss": 0.0004,
"step": 2079
},
{
"epoch": 0.8536835624871743,
"grad_norm": 7.004717826843262,
"learning_rate": 1.4614121510673235e-06,
"loss": 0.0068,
"step": 2080
},
{
"epoch": 0.8540939872768315,
"grad_norm": 0.9702602624893188,
"learning_rate": 1.4573070607553365e-06,
"loss": 0.0008,
"step": 2081
},
{
"epoch": 0.8545044120664889,
"grad_norm": 182.40318298339844,
"learning_rate": 1.4532019704433498e-06,
"loss": 0.7324,
"step": 2082
},
{
"epoch": 0.8549148368561461,
"grad_norm": 64.67376708984375,
"learning_rate": 1.4490968801313629e-06,
"loss": 0.6951,
"step": 2083
},
{
"epoch": 0.8553252616458034,
"grad_norm": 2.4038000106811523,
"learning_rate": 1.444991789819376e-06,
"loss": 0.0058,
"step": 2084
},
{
"epoch": 0.8557356864354607,
"grad_norm": 127.4338607788086,
"learning_rate": 1.4408866995073892e-06,
"loss": 0.159,
"step": 2085
},
{
"epoch": 0.856146111225118,
"grad_norm": 2.2899816036224365,
"learning_rate": 1.4367816091954023e-06,
"loss": 0.0007,
"step": 2086
},
{
"epoch": 0.8565565360147753,
"grad_norm": 2.042107105255127,
"learning_rate": 1.4326765188834156e-06,
"loss": 0.001,
"step": 2087
},
{
"epoch": 0.8569669608044326,
"grad_norm": 0.29681575298309326,
"learning_rate": 1.4285714285714286e-06,
"loss": 0.0007,
"step": 2088
},
{
"epoch": 0.8573773855940899,
"grad_norm": 38.84279251098633,
"learning_rate": 1.4244663382594417e-06,
"loss": 0.0273,
"step": 2089
},
{
"epoch": 0.8577878103837472,
"grad_norm": 1.8666824102401733,
"learning_rate": 1.420361247947455e-06,
"loss": 0.0026,
"step": 2090
},
{
"epoch": 0.8581982351734044,
"grad_norm": 1.8169337511062622,
"learning_rate": 1.416256157635468e-06,
"loss": 0.0005,
"step": 2091
},
{
"epoch": 0.8586086599630618,
"grad_norm": 168.6756134033203,
"learning_rate": 1.412151067323481e-06,
"loss": 0.8945,
"step": 2092
},
{
"epoch": 0.859019084752719,
"grad_norm": 5.650978088378906,
"learning_rate": 1.4080459770114944e-06,
"loss": 0.0095,
"step": 2093
},
{
"epoch": 0.8594295095423764,
"grad_norm": 4.385357856750488,
"learning_rate": 1.4039408866995074e-06,
"loss": 0.0024,
"step": 2094
},
{
"epoch": 0.8598399343320337,
"grad_norm": 2.1580545902252197,
"learning_rate": 1.3998357963875205e-06,
"loss": 0.002,
"step": 2095
},
{
"epoch": 0.8602503591216909,
"grad_norm": 0.6650879383087158,
"learning_rate": 1.3957307060755338e-06,
"loss": 0.0006,
"step": 2096
},
{
"epoch": 0.8606607839113483,
"grad_norm": 3.4510574340820312,
"learning_rate": 1.3916256157635468e-06,
"loss": 0.0012,
"step": 2097
},
{
"epoch": 0.8610712087010055,
"grad_norm": 2.040491819381714,
"learning_rate": 1.3875205254515601e-06,
"loss": 0.0007,
"step": 2098
},
{
"epoch": 0.8614816334906629,
"grad_norm": 1.3360214233398438,
"learning_rate": 1.3834154351395732e-06,
"loss": 0.0012,
"step": 2099
},
{
"epoch": 0.8618920582803201,
"grad_norm": 3.590639352798462,
"learning_rate": 1.3793103448275862e-06,
"loss": 0.0012,
"step": 2100
},
{
"epoch": 0.8623024830699775,
"grad_norm": 1.7434356212615967,
"learning_rate": 1.3752052545155995e-06,
"loss": 0.0007,
"step": 2101
},
{
"epoch": 0.8627129078596347,
"grad_norm": 2.298342227935791,
"learning_rate": 1.3711001642036126e-06,
"loss": 0.0015,
"step": 2102
},
{
"epoch": 0.863123332649292,
"grad_norm": 5.2816057205200195,
"learning_rate": 1.3669950738916257e-06,
"loss": 0.0024,
"step": 2103
},
{
"epoch": 0.8635337574389493,
"grad_norm": 103.97406005859375,
"learning_rate": 1.362889983579639e-06,
"loss": 0.0846,
"step": 2104
},
{
"epoch": 0.8639441822286066,
"grad_norm": 1.2924036979675293,
"learning_rate": 1.358784893267652e-06,
"loss": 0.0009,
"step": 2105
},
{
"epoch": 0.8643546070182639,
"grad_norm": 7.957437038421631,
"learning_rate": 1.354679802955665e-06,
"loss": 0.005,
"step": 2106
},
{
"epoch": 0.8647650318079212,
"grad_norm": 2.229443073272705,
"learning_rate": 1.3505747126436783e-06,
"loss": 0.0016,
"step": 2107
},
{
"epoch": 0.8651754565975784,
"grad_norm": 27.475372314453125,
"learning_rate": 1.3464696223316914e-06,
"loss": 0.0151,
"step": 2108
},
{
"epoch": 0.8655858813872358,
"grad_norm": 0.6112897992134094,
"learning_rate": 1.3423645320197047e-06,
"loss": 0.0004,
"step": 2109
},
{
"epoch": 0.865996306176893,
"grad_norm": 8.505683898925781,
"learning_rate": 1.3382594417077177e-06,
"loss": 0.006,
"step": 2110
},
{
"epoch": 0.8664067309665504,
"grad_norm": 6.502181053161621,
"learning_rate": 1.3341543513957308e-06,
"loss": 0.0035,
"step": 2111
},
{
"epoch": 0.8668171557562077,
"grad_norm": 18.60784912109375,
"learning_rate": 1.330049261083744e-06,
"loss": 0.0067,
"step": 2112
},
{
"epoch": 0.867227580545865,
"grad_norm": 3.8766109943389893,
"learning_rate": 1.3259441707717571e-06,
"loss": 0.0013,
"step": 2113
},
{
"epoch": 0.8676380053355223,
"grad_norm": 132.60948181152344,
"learning_rate": 1.3218390804597702e-06,
"loss": 0.5472,
"step": 2114
},
{
"epoch": 0.8680484301251795,
"grad_norm": 0.7578341960906982,
"learning_rate": 1.3177339901477835e-06,
"loss": 0.0007,
"step": 2115
},
{
"epoch": 0.8684588549148369,
"grad_norm": 3.6058578491210938,
"learning_rate": 1.3136288998357965e-06,
"loss": 0.0012,
"step": 2116
},
{
"epoch": 0.8688692797044941,
"grad_norm": 2.2759504318237305,
"learning_rate": 1.3095238095238096e-06,
"loss": 0.0006,
"step": 2117
},
{
"epoch": 0.8692797044941515,
"grad_norm": 5.895321846008301,
"learning_rate": 1.3054187192118229e-06,
"loss": 0.0028,
"step": 2118
},
{
"epoch": 0.8696901292838087,
"grad_norm": 3.7929725646972656,
"learning_rate": 1.301313628899836e-06,
"loss": 0.0013,
"step": 2119
},
{
"epoch": 0.8701005540734661,
"grad_norm": 1.7934080362319946,
"learning_rate": 1.297208538587849e-06,
"loss": 0.001,
"step": 2120
},
{
"epoch": 0.8705109788631233,
"grad_norm": 0.6949759125709534,
"learning_rate": 1.2931034482758623e-06,
"loss": 0.0007,
"step": 2121
},
{
"epoch": 0.8709214036527806,
"grad_norm": 0.25476956367492676,
"learning_rate": 1.2889983579638754e-06,
"loss": 0.0005,
"step": 2122
},
{
"epoch": 0.871331828442438,
"grad_norm": 4.590909004211426,
"learning_rate": 1.2848932676518886e-06,
"loss": 0.0018,
"step": 2123
},
{
"epoch": 0.8717422532320952,
"grad_norm": 19.81357192993164,
"learning_rate": 1.2807881773399017e-06,
"loss": 0.0102,
"step": 2124
},
{
"epoch": 0.8721526780217526,
"grad_norm": 0.9998217225074768,
"learning_rate": 1.2766830870279148e-06,
"loss": 0.001,
"step": 2125
},
{
"epoch": 0.8725631028114098,
"grad_norm": 60.47079849243164,
"learning_rate": 1.272577996715928e-06,
"loss": 0.0264,
"step": 2126
},
{
"epoch": 0.8729735276010671,
"grad_norm": 2.567127227783203,
"learning_rate": 1.268472906403941e-06,
"loss": 0.0026,
"step": 2127
},
{
"epoch": 0.8733839523907244,
"grad_norm": 107.20476531982422,
"learning_rate": 1.2643678160919542e-06,
"loss": 0.6096,
"step": 2128
},
{
"epoch": 0.8737943771803817,
"grad_norm": 3.633134603500366,
"learning_rate": 1.2602627257799674e-06,
"loss": 0.0024,
"step": 2129
},
{
"epoch": 0.874204801970039,
"grad_norm": 1.4011846780776978,
"learning_rate": 1.2561576354679805e-06,
"loss": 0.0008,
"step": 2130
},
{
"epoch": 0.8746152267596963,
"grad_norm": 3.047912359237671,
"learning_rate": 1.2520525451559936e-06,
"loss": 0.0011,
"step": 2131
},
{
"epoch": 0.8750256515493536,
"grad_norm": 1.1926379203796387,
"learning_rate": 1.2479474548440066e-06,
"loss": 0.0006,
"step": 2132
},
{
"epoch": 0.8754360763390109,
"grad_norm": 2.291133403778076,
"learning_rate": 1.24384236453202e-06,
"loss": 0.0008,
"step": 2133
},
{
"epoch": 0.8758465011286681,
"grad_norm": 1.1550277471542358,
"learning_rate": 1.239737274220033e-06,
"loss": 0.0004,
"step": 2134
},
{
"epoch": 0.8762569259183255,
"grad_norm": 1.5513360500335693,
"learning_rate": 1.235632183908046e-06,
"loss": 0.001,
"step": 2135
},
{
"epoch": 0.8766673507079827,
"grad_norm": 103.6968002319336,
"learning_rate": 1.2315270935960593e-06,
"loss": 0.0699,
"step": 2136
},
{
"epoch": 0.8770777754976401,
"grad_norm": 2.844360828399658,
"learning_rate": 1.2274220032840724e-06,
"loss": 0.0008,
"step": 2137
},
{
"epoch": 0.8774882002872973,
"grad_norm": 0.605464518070221,
"learning_rate": 1.2233169129720854e-06,
"loss": 0.0006,
"step": 2138
},
{
"epoch": 0.8778986250769546,
"grad_norm": 387.9739990234375,
"learning_rate": 1.2192118226600987e-06,
"loss": 0.3758,
"step": 2139
},
{
"epoch": 0.878309049866612,
"grad_norm": 2.190920114517212,
"learning_rate": 1.2151067323481118e-06,
"loss": 0.0023,
"step": 2140
},
{
"epoch": 0.8787194746562692,
"grad_norm": 4.451165199279785,
"learning_rate": 1.2110016420361248e-06,
"loss": 0.0018,
"step": 2141
},
{
"epoch": 0.8791298994459266,
"grad_norm": 0.46043872833251953,
"learning_rate": 1.2068965517241381e-06,
"loss": 0.0003,
"step": 2142
},
{
"epoch": 0.8795403242355838,
"grad_norm": 2.7850561141967773,
"learning_rate": 1.2027914614121512e-06,
"loss": 0.0011,
"step": 2143
},
{
"epoch": 0.8799507490252412,
"grad_norm": 0.5206378698348999,
"learning_rate": 1.1986863711001645e-06,
"loss": 0.0004,
"step": 2144
},
{
"epoch": 0.8803611738148984,
"grad_norm": 2.808671236038208,
"learning_rate": 1.1945812807881775e-06,
"loss": 0.002,
"step": 2145
},
{
"epoch": 0.8807715986045557,
"grad_norm": 2.3434700965881348,
"learning_rate": 1.1904761904761906e-06,
"loss": 0.0009,
"step": 2146
},
{
"epoch": 0.881182023394213,
"grad_norm": 4.329005718231201,
"learning_rate": 1.1863711001642037e-06,
"loss": 0.0016,
"step": 2147
},
{
"epoch": 0.8815924481838703,
"grad_norm": 1.0745855569839478,
"learning_rate": 1.1822660098522167e-06,
"loss": 0.0017,
"step": 2148
},
{
"epoch": 0.8820028729735276,
"grad_norm": 4.068753719329834,
"learning_rate": 1.17816091954023e-06,
"loss": 0.0022,
"step": 2149
},
{
"epoch": 0.8824132977631849,
"grad_norm": 2.9023873805999756,
"learning_rate": 1.174055829228243e-06,
"loss": 0.0009,
"step": 2150
},
{
"epoch": 0.8828237225528421,
"grad_norm": 2.935148000717163,
"learning_rate": 1.1699507389162561e-06,
"loss": 0.0015,
"step": 2151
},
{
"epoch": 0.8832341473424995,
"grad_norm": 103.91587829589844,
"learning_rate": 1.1658456486042694e-06,
"loss": 0.4845,
"step": 2152
},
{
"epoch": 0.8836445721321567,
"grad_norm": 4.398167133331299,
"learning_rate": 1.1617405582922825e-06,
"loss": 0.0016,
"step": 2153
},
{
"epoch": 0.8840549969218141,
"grad_norm": 6.654697418212891,
"learning_rate": 1.1576354679802955e-06,
"loss": 0.0039,
"step": 2154
},
{
"epoch": 0.8844654217114714,
"grad_norm": 6.510483264923096,
"learning_rate": 1.1535303776683088e-06,
"loss": 0.0077,
"step": 2155
},
{
"epoch": 0.8848758465011287,
"grad_norm": 4.1658830642700195,
"learning_rate": 1.1494252873563219e-06,
"loss": 0.0024,
"step": 2156
},
{
"epoch": 0.885286271290786,
"grad_norm": 63.34073257446289,
"learning_rate": 1.1453201970443351e-06,
"loss": 0.0332,
"step": 2157
},
{
"epoch": 0.8856966960804432,
"grad_norm": 182.70494079589844,
"learning_rate": 1.1412151067323482e-06,
"loss": 0.1947,
"step": 2158
},
{
"epoch": 0.8861071208701006,
"grad_norm": 6.552812099456787,
"learning_rate": 1.1371100164203613e-06,
"loss": 0.0051,
"step": 2159
},
{
"epoch": 0.8865175456597578,
"grad_norm": 0.42228230834007263,
"learning_rate": 1.1330049261083746e-06,
"loss": 0.0003,
"step": 2160
},
{
"epoch": 0.8869279704494152,
"grad_norm": 3.6216838359832764,
"learning_rate": 1.1288998357963876e-06,
"loss": 0.0014,
"step": 2161
},
{
"epoch": 0.8873383952390724,
"grad_norm": 13.326390266418457,
"learning_rate": 1.1247947454844007e-06,
"loss": 0.0098,
"step": 2162
},
{
"epoch": 0.8877488200287298,
"grad_norm": 0.9341579079627991,
"learning_rate": 1.120689655172414e-06,
"loss": 0.0009,
"step": 2163
},
{
"epoch": 0.888159244818387,
"grad_norm": 8.089537620544434,
"learning_rate": 1.116584564860427e-06,
"loss": 0.006,
"step": 2164
},
{
"epoch": 0.8885696696080443,
"grad_norm": 1.2521154880523682,
"learning_rate": 1.11247947454844e-06,
"loss": 0.0004,
"step": 2165
},
{
"epoch": 0.8889800943977016,
"grad_norm": 57.11647415161133,
"learning_rate": 1.1083743842364534e-06,
"loss": 0.495,
"step": 2166
},
{
"epoch": 0.8893905191873589,
"grad_norm": 2.9028825759887695,
"learning_rate": 1.1042692939244664e-06,
"loss": 0.0021,
"step": 2167
},
{
"epoch": 0.8898009439770163,
"grad_norm": 3.974114179611206,
"learning_rate": 1.1001642036124797e-06,
"loss": 0.002,
"step": 2168
},
{
"epoch": 0.8902113687666735,
"grad_norm": 19.403711318969727,
"learning_rate": 1.0960591133004928e-06,
"loss": 0.0121,
"step": 2169
},
{
"epoch": 0.8906217935563308,
"grad_norm": 1.8327593803405762,
"learning_rate": 1.0919540229885058e-06,
"loss": 0.0006,
"step": 2170
},
{
"epoch": 0.8910322183459881,
"grad_norm": 2.5765860080718994,
"learning_rate": 1.0878489326765191e-06,
"loss": 0.0011,
"step": 2171
},
{
"epoch": 0.8914426431356454,
"grad_norm": 29.263242721557617,
"learning_rate": 1.0837438423645322e-06,
"loss": 0.067,
"step": 2172
},
{
"epoch": 0.8918530679253027,
"grad_norm": 6.581966400146484,
"learning_rate": 1.0796387520525452e-06,
"loss": 0.0052,
"step": 2173
},
{
"epoch": 0.89226349271496,
"grad_norm": 1.646833896636963,
"learning_rate": 1.0755336617405583e-06,
"loss": 0.0013,
"step": 2174
},
{
"epoch": 0.8926739175046173,
"grad_norm": 5.351321220397949,
"learning_rate": 1.0714285714285714e-06,
"loss": 0.0068,
"step": 2175
},
{
"epoch": 0.8930843422942746,
"grad_norm": 7.479823112487793,
"learning_rate": 1.0673234811165846e-06,
"loss": 0.0056,
"step": 2176
},
{
"epoch": 0.8934947670839318,
"grad_norm": 2.8000588417053223,
"learning_rate": 1.0632183908045977e-06,
"loss": 0.0011,
"step": 2177
},
{
"epoch": 0.8939051918735892,
"grad_norm": 8.061798095703125,
"learning_rate": 1.059113300492611e-06,
"loss": 0.0054,
"step": 2178
},
{
"epoch": 0.8943156166632464,
"grad_norm": 1.5384347438812256,
"learning_rate": 1.055008210180624e-06,
"loss": 0.0005,
"step": 2179
},
{
"epoch": 0.8947260414529038,
"grad_norm": 4.42774772644043,
"learning_rate": 1.0509031198686371e-06,
"loss": 0.0017,
"step": 2180
},
{
"epoch": 0.895136466242561,
"grad_norm": 1.9187287092208862,
"learning_rate": 1.0467980295566504e-06,
"loss": 0.0008,
"step": 2181
},
{
"epoch": 0.8955468910322183,
"grad_norm": 1.6684490442276,
"learning_rate": 1.0426929392446635e-06,
"loss": 0.0005,
"step": 2182
},
{
"epoch": 0.8959573158218757,
"grad_norm": 3.399998426437378,
"learning_rate": 1.0385878489326765e-06,
"loss": 0.0012,
"step": 2183
},
{
"epoch": 0.8963677406115329,
"grad_norm": 1.5626261234283447,
"learning_rate": 1.0344827586206898e-06,
"loss": 0.0007,
"step": 2184
},
{
"epoch": 0.8967781654011903,
"grad_norm": 2.4970765113830566,
"learning_rate": 1.0303776683087029e-06,
"loss": 0.002,
"step": 2185
},
{
"epoch": 0.8971885901908475,
"grad_norm": 6.217938423156738,
"learning_rate": 1.026272577996716e-06,
"loss": 0.0035,
"step": 2186
},
{
"epoch": 0.8975990149805049,
"grad_norm": 2.9275400638580322,
"learning_rate": 1.0221674876847292e-06,
"loss": 0.0021,
"step": 2187
},
{
"epoch": 0.8980094397701621,
"grad_norm": 2.5872483253479004,
"learning_rate": 1.0180623973727423e-06,
"loss": 0.0013,
"step": 2188
},
{
"epoch": 0.8984198645598194,
"grad_norm": 4.4500508308410645,
"learning_rate": 1.0139573070607555e-06,
"loss": 0.0016,
"step": 2189
},
{
"epoch": 0.8988302893494767,
"grad_norm": 1.9133249521255493,
"learning_rate": 1.0098522167487686e-06,
"loss": 0.0008,
"step": 2190
},
{
"epoch": 0.899240714139134,
"grad_norm": 2.75472354888916,
"learning_rate": 1.0057471264367817e-06,
"loss": 0.003,
"step": 2191
},
{
"epoch": 0.8996511389287913,
"grad_norm": 3.0145866870880127,
"learning_rate": 1.001642036124795e-06,
"loss": 0.0022,
"step": 2192
},
{
"epoch": 0.9000615637184486,
"grad_norm": 1.6914114952087402,
"learning_rate": 9.97536945812808e-07,
"loss": 0.0005,
"step": 2193
},
{
"epoch": 0.9004719885081058,
"grad_norm": 2.3173418045043945,
"learning_rate": 9.93431855500821e-07,
"loss": 0.0007,
"step": 2194
},
{
"epoch": 0.9008824132977632,
"grad_norm": 123.37835693359375,
"learning_rate": 9.893267651888343e-07,
"loss": 0.3529,
"step": 2195
},
{
"epoch": 0.9012928380874204,
"grad_norm": 1.5026416778564453,
"learning_rate": 9.852216748768474e-07,
"loss": 0.0009,
"step": 2196
},
{
"epoch": 0.9017032628770778,
"grad_norm": 1.3729249238967896,
"learning_rate": 9.811165845648605e-07,
"loss": 0.0008,
"step": 2197
},
{
"epoch": 0.902113687666735,
"grad_norm": 3.0635335445404053,
"learning_rate": 9.770114942528738e-07,
"loss": 0.001,
"step": 2198
},
{
"epoch": 0.9025241124563924,
"grad_norm": 5.324329376220703,
"learning_rate": 9.729064039408868e-07,
"loss": 0.0029,
"step": 2199
},
{
"epoch": 0.9029345372460497,
"grad_norm": 4.149939060211182,
"learning_rate": 9.688013136288999e-07,
"loss": 0.0017,
"step": 2200
},
{
"epoch": 0.9033449620357069,
"grad_norm": 2.8204097747802734,
"learning_rate": 9.64696223316913e-07,
"loss": 0.0042,
"step": 2201
},
{
"epoch": 0.9037553868253643,
"grad_norm": 4.358647346496582,
"learning_rate": 9.605911330049262e-07,
"loss": 0.0017,
"step": 2202
},
{
"epoch": 0.9041658116150215,
"grad_norm": 5.836299419403076,
"learning_rate": 9.564860426929393e-07,
"loss": 0.0026,
"step": 2203
},
{
"epoch": 0.9045762364046789,
"grad_norm": 2.5344083309173584,
"learning_rate": 9.523809523809525e-07,
"loss": 0.0008,
"step": 2204
},
{
"epoch": 0.9049866611943361,
"grad_norm": 2.2534592151641846,
"learning_rate": 9.482758620689655e-07,
"loss": 0.0006,
"step": 2205
},
{
"epoch": 0.9053970859839935,
"grad_norm": 2.0328423976898193,
"learning_rate": 9.441707717569787e-07,
"loss": 0.0031,
"step": 2206
},
{
"epoch": 0.9058075107736507,
"grad_norm": 0.9508869051933289,
"learning_rate": 9.400656814449919e-07,
"loss": 0.0004,
"step": 2207
},
{
"epoch": 0.906217935563308,
"grad_norm": 1.8879873752593994,
"learning_rate": 9.359605911330049e-07,
"loss": 0.0006,
"step": 2208
},
{
"epoch": 0.9066283603529653,
"grad_norm": 6.979659557342529,
"learning_rate": 9.318555008210181e-07,
"loss": 0.0042,
"step": 2209
},
{
"epoch": 0.9070387851426226,
"grad_norm": 16.006826400756836,
"learning_rate": 9.277504105090313e-07,
"loss": 0.0242,
"step": 2210
},
{
"epoch": 0.90744920993228,
"grad_norm": 63.03201675415039,
"learning_rate": 9.236453201970444e-07,
"loss": 0.0406,
"step": 2211
},
{
"epoch": 0.9078596347219372,
"grad_norm": 5.355207443237305,
"learning_rate": 9.195402298850575e-07,
"loss": 0.0026,
"step": 2212
},
{
"epoch": 0.9082700595115945,
"grad_norm": 2.559473752975464,
"learning_rate": 9.154351395730707e-07,
"loss": 0.0018,
"step": 2213
},
{
"epoch": 0.9086804843012518,
"grad_norm": 3.1311941146850586,
"learning_rate": 9.113300492610838e-07,
"loss": 0.001,
"step": 2214
},
{
"epoch": 0.9090909090909091,
"grad_norm": 2.9037139415740967,
"learning_rate": 9.07224958949097e-07,
"loss": 0.0009,
"step": 2215
},
{
"epoch": 0.9095013338805664,
"grad_norm": 2.399092435836792,
"learning_rate": 9.031198686371101e-07,
"loss": 0.0024,
"step": 2216
},
{
"epoch": 0.9099117586702237,
"grad_norm": 5.6226372718811035,
"learning_rate": 8.990147783251232e-07,
"loss": 0.0024,
"step": 2217
},
{
"epoch": 0.910322183459881,
"grad_norm": 315.07080078125,
"learning_rate": 8.949096880131364e-07,
"loss": 0.2169,
"step": 2218
},
{
"epoch": 0.9107326082495383,
"grad_norm": 58.41863250732422,
"learning_rate": 8.908045977011495e-07,
"loss": 0.5552,
"step": 2219
},
{
"epoch": 0.9111430330391955,
"grad_norm": 0.7334727048873901,
"learning_rate": 8.866995073891626e-07,
"loss": 0.0005,
"step": 2220
},
{
"epoch": 0.9115534578288529,
"grad_norm": 10.305622100830078,
"learning_rate": 8.825944170771758e-07,
"loss": 0.0045,
"step": 2221
},
{
"epoch": 0.9119638826185101,
"grad_norm": 6.364269256591797,
"learning_rate": 8.78489326765189e-07,
"loss": 0.003,
"step": 2222
},
{
"epoch": 0.9123743074081675,
"grad_norm": 5.827622413635254,
"learning_rate": 8.74384236453202e-07,
"loss": 0.0035,
"step": 2223
},
{
"epoch": 0.9127847321978247,
"grad_norm": 2.402320623397827,
"learning_rate": 8.702791461412152e-07,
"loss": 0.004,
"step": 2224
},
{
"epoch": 0.913195156987482,
"grad_norm": 4.273941516876221,
"learning_rate": 8.661740558292284e-07,
"loss": 0.0016,
"step": 2225
},
{
"epoch": 0.9136055817771394,
"grad_norm": 145.0655517578125,
"learning_rate": 8.620689655172415e-07,
"loss": 0.1477,
"step": 2226
},
{
"epoch": 0.9140160065667966,
"grad_norm": 1.109365463256836,
"learning_rate": 8.579638752052546e-07,
"loss": 0.0007,
"step": 2227
},
{
"epoch": 0.914426431356454,
"grad_norm": 3.5715272426605225,
"learning_rate": 8.538587848932678e-07,
"loss": 0.0034,
"step": 2228
},
{
"epoch": 0.9148368561461112,
"grad_norm": 7.471683025360107,
"learning_rate": 8.497536945812808e-07,
"loss": 0.0042,
"step": 2229
},
{
"epoch": 0.9152472809357686,
"grad_norm": 1.8124293088912964,
"learning_rate": 8.456486042692939e-07,
"loss": 0.0018,
"step": 2230
},
{
"epoch": 0.9156577057254258,
"grad_norm": 4.016838550567627,
"learning_rate": 8.415435139573071e-07,
"loss": 0.0013,
"step": 2231
},
{
"epoch": 0.9160681305150831,
"grad_norm": 11.869513511657715,
"learning_rate": 8.374384236453203e-07,
"loss": 0.0043,
"step": 2232
},
{
"epoch": 0.9164785553047404,
"grad_norm": 4.943122386932373,
"learning_rate": 8.333333333333333e-07,
"loss": 0.0023,
"step": 2233
},
{
"epoch": 0.9168889800943977,
"grad_norm": 4.305215358734131,
"learning_rate": 8.292282430213465e-07,
"loss": 0.0019,
"step": 2234
},
{
"epoch": 0.917299404884055,
"grad_norm": 3.149911642074585,
"learning_rate": 8.251231527093597e-07,
"loss": 0.0011,
"step": 2235
},
{
"epoch": 0.9177098296737123,
"grad_norm": 1.3957020044326782,
"learning_rate": 8.210180623973727e-07,
"loss": 0.0004,
"step": 2236
},
{
"epoch": 0.9181202544633695,
"grad_norm": 1.6041291952133179,
"learning_rate": 8.169129720853859e-07,
"loss": 0.0004,
"step": 2237
},
{
"epoch": 0.9185306792530269,
"grad_norm": 3.8794071674346924,
"learning_rate": 8.128078817733991e-07,
"loss": 0.0014,
"step": 2238
},
{
"epoch": 0.9189411040426841,
"grad_norm": 0.6249457001686096,
"learning_rate": 8.087027914614122e-07,
"loss": 0.0004,
"step": 2239
},
{
"epoch": 0.9193515288323415,
"grad_norm": 2.0174367427825928,
"learning_rate": 8.045977011494253e-07,
"loss": 0.001,
"step": 2240
},
{
"epoch": 0.9197619536219988,
"grad_norm": 56.651004791259766,
"learning_rate": 8.004926108374385e-07,
"loss": 0.5049,
"step": 2241
},
{
"epoch": 0.9201723784116561,
"grad_norm": 4.478601455688477,
"learning_rate": 7.963875205254517e-07,
"loss": 0.0026,
"step": 2242
},
{
"epoch": 0.9205828032013134,
"grad_norm": 4.857123851776123,
"learning_rate": 7.922824302134647e-07,
"loss": 0.002,
"step": 2243
},
{
"epoch": 0.9209932279909706,
"grad_norm": 7.383138179779053,
"learning_rate": 7.881773399014779e-07,
"loss": 0.0035,
"step": 2244
},
{
"epoch": 0.921403652780628,
"grad_norm": 1.0620945692062378,
"learning_rate": 7.840722495894911e-07,
"loss": 0.0004,
"step": 2245
},
{
"epoch": 0.9218140775702852,
"grad_norm": 1.0185391902923584,
"learning_rate": 7.799671592775042e-07,
"loss": 0.0009,
"step": 2246
},
{
"epoch": 0.9222245023599426,
"grad_norm": 0.77149498462677,
"learning_rate": 7.758620689655173e-07,
"loss": 0.0003,
"step": 2247
},
{
"epoch": 0.9226349271495998,
"grad_norm": 6.385399341583252,
"learning_rate": 7.717569786535305e-07,
"loss": 0.0035,
"step": 2248
},
{
"epoch": 0.9230453519392572,
"grad_norm": 4.066201686859131,
"learning_rate": 7.676518883415436e-07,
"loss": 0.0018,
"step": 2249
},
{
"epoch": 0.9234557767289144,
"grad_norm": 0.5383831858634949,
"learning_rate": 7.635467980295568e-07,
"loss": 0.0004,
"step": 2250
},
{
"epoch": 0.9238662015185717,
"grad_norm": 1.1348515748977661,
"learning_rate": 7.594417077175699e-07,
"loss": 0.0004,
"step": 2251
},
{
"epoch": 0.924276626308229,
"grad_norm": 3.3495075702667236,
"learning_rate": 7.55336617405583e-07,
"loss": 0.001,
"step": 2252
},
{
"epoch": 0.9246870510978863,
"grad_norm": 3.414954423904419,
"learning_rate": 7.512315270935962e-07,
"loss": 0.0012,
"step": 2253
},
{
"epoch": 0.9250974758875437,
"grad_norm": 3.6016907691955566,
"learning_rate": 7.471264367816093e-07,
"loss": 0.0012,
"step": 2254
},
{
"epoch": 0.9255079006772009,
"grad_norm": 4.806239128112793,
"learning_rate": 7.430213464696224e-07,
"loss": 0.0048,
"step": 2255
},
{
"epoch": 0.9259183254668581,
"grad_norm": 2.2628512382507324,
"learning_rate": 7.389162561576356e-07,
"loss": 0.0014,
"step": 2256
},
{
"epoch": 0.9263287502565155,
"grad_norm": 2.4400737285614014,
"learning_rate": 7.348111658456486e-07,
"loss": 0.0009,
"step": 2257
},
{
"epoch": 0.9267391750461728,
"grad_norm": 2.300161838531494,
"learning_rate": 7.307060755336617e-07,
"loss": 0.0007,
"step": 2258
},
{
"epoch": 0.9271495998358301,
"grad_norm": 1.025620698928833,
"learning_rate": 7.266009852216749e-07,
"loss": 0.0005,
"step": 2259
},
{
"epoch": 0.9275600246254874,
"grad_norm": 1.3631598949432373,
"learning_rate": 7.22495894909688e-07,
"loss": 0.0014,
"step": 2260
},
{
"epoch": 0.9279704494151447,
"grad_norm": 0.40626609325408936,
"learning_rate": 7.183908045977011e-07,
"loss": 0.0003,
"step": 2261
},
{
"epoch": 0.928380874204802,
"grad_norm": 172.8359375,
"learning_rate": 7.142857142857143e-07,
"loss": 0.2776,
"step": 2262
},
{
"epoch": 0.9287912989944592,
"grad_norm": 3.122516393661499,
"learning_rate": 7.101806239737275e-07,
"loss": 0.0026,
"step": 2263
},
{
"epoch": 0.9292017237841166,
"grad_norm": 2.4915544986724854,
"learning_rate": 7.060755336617405e-07,
"loss": 0.0016,
"step": 2264
},
{
"epoch": 0.9296121485737738,
"grad_norm": 6.385788917541504,
"learning_rate": 7.019704433497537e-07,
"loss": 0.0044,
"step": 2265
},
{
"epoch": 0.9300225733634312,
"grad_norm": 0.6820462346076965,
"learning_rate": 6.978653530377669e-07,
"loss": 0.001,
"step": 2266
},
{
"epoch": 0.9304329981530884,
"grad_norm": 1.7900935411453247,
"learning_rate": 6.937602627257801e-07,
"loss": 0.0006,
"step": 2267
},
{
"epoch": 0.9308434229427457,
"grad_norm": 101.79829406738281,
"learning_rate": 6.896551724137931e-07,
"loss": 0.0687,
"step": 2268
},
{
"epoch": 0.931253847732403,
"grad_norm": 3.4876976013183594,
"learning_rate": 6.855500821018063e-07,
"loss": 0.0012,
"step": 2269
},
{
"epoch": 0.9316642725220603,
"grad_norm": 0.470086008310318,
"learning_rate": 6.814449917898195e-07,
"loss": 0.0008,
"step": 2270
},
{
"epoch": 0.9320746973117177,
"grad_norm": 3.5489258766174316,
"learning_rate": 6.773399014778325e-07,
"loss": 0.0042,
"step": 2271
},
{
"epoch": 0.9324851221013749,
"grad_norm": 9.691227912902832,
"learning_rate": 6.732348111658457e-07,
"loss": 0.0065,
"step": 2272
},
{
"epoch": 0.9328955468910323,
"grad_norm": 6.473835468292236,
"learning_rate": 6.691297208538589e-07,
"loss": 0.0052,
"step": 2273
},
{
"epoch": 0.9333059716806895,
"grad_norm": 6.724939346313477,
"learning_rate": 6.65024630541872e-07,
"loss": 0.0078,
"step": 2274
},
{
"epoch": 0.9337163964703468,
"grad_norm": 3.4430534839630127,
"learning_rate": 6.609195402298851e-07,
"loss": 0.0018,
"step": 2275
},
{
"epoch": 0.9341268212600041,
"grad_norm": 1.2493078708648682,
"learning_rate": 6.568144499178983e-07,
"loss": 0.0017,
"step": 2276
},
{
"epoch": 0.9345372460496614,
"grad_norm": 2.2264821529388428,
"learning_rate": 6.527093596059114e-07,
"loss": 0.003,
"step": 2277
},
{
"epoch": 0.9349476708393187,
"grad_norm": 192.00787353515625,
"learning_rate": 6.486042692939245e-07,
"loss": 0.1877,
"step": 2278
},
{
"epoch": 0.935358095628976,
"grad_norm": 1.2581437826156616,
"learning_rate": 6.444991789819377e-07,
"loss": 0.0007,
"step": 2279
},
{
"epoch": 0.9357685204186332,
"grad_norm": 96.22144317626953,
"learning_rate": 6.403940886699508e-07,
"loss": 0.04,
"step": 2280
},
{
"epoch": 0.9361789452082906,
"grad_norm": 2.0211939811706543,
"learning_rate": 6.36288998357964e-07,
"loss": 0.0011,
"step": 2281
},
{
"epoch": 0.9365893699979478,
"grad_norm": 4.270722389221191,
"learning_rate": 6.321839080459771e-07,
"loss": 0.0017,
"step": 2282
},
{
"epoch": 0.9369997947876052,
"grad_norm": 0.9350568056106567,
"learning_rate": 6.280788177339903e-07,
"loss": 0.0012,
"step": 2283
},
{
"epoch": 0.9374102195772624,
"grad_norm": 4.065085411071777,
"learning_rate": 6.239737274220033e-07,
"loss": 0.0018,
"step": 2284
},
{
"epoch": 0.9378206443669198,
"grad_norm": 3.125955104827881,
"learning_rate": 6.198686371100165e-07,
"loss": 0.0024,
"step": 2285
},
{
"epoch": 0.9382310691565771,
"grad_norm": 73.26425170898438,
"learning_rate": 6.157635467980297e-07,
"loss": 0.0811,
"step": 2286
},
{
"epoch": 0.9386414939462343,
"grad_norm": 0.6261829137802124,
"learning_rate": 6.116584564860427e-07,
"loss": 0.001,
"step": 2287
},
{
"epoch": 0.9390519187358917,
"grad_norm": 85.94723510742188,
"learning_rate": 6.075533661740559e-07,
"loss": 0.4689,
"step": 2288
},
{
"epoch": 0.9394623435255489,
"grad_norm": 1.3829810619354248,
"learning_rate": 6.034482758620691e-07,
"loss": 0.0011,
"step": 2289
},
{
"epoch": 0.9398727683152063,
"grad_norm": 157.90957641601562,
"learning_rate": 5.993431855500822e-07,
"loss": 0.7148,
"step": 2290
},
{
"epoch": 0.9402831931048635,
"grad_norm": 1045.1341552734375,
"learning_rate": 5.952380952380953e-07,
"loss": 0.6993,
"step": 2291
},
{
"epoch": 0.9406936178945209,
"grad_norm": 4.626864433288574,
"learning_rate": 5.911330049261084e-07,
"loss": 0.0018,
"step": 2292
},
{
"epoch": 0.9411040426841781,
"grad_norm": 2.9818336963653564,
"learning_rate": 5.870279146141215e-07,
"loss": 0.0017,
"step": 2293
},
{
"epoch": 0.9415144674738354,
"grad_norm": 0.04159677401185036,
"learning_rate": 5.829228243021347e-07,
"loss": 0.0002,
"step": 2294
},
{
"epoch": 0.9419248922634927,
"grad_norm": 1.4131793975830078,
"learning_rate": 5.788177339901478e-07,
"loss": 0.0007,
"step": 2295
},
{
"epoch": 0.94233531705315,
"grad_norm": 98.79226684570312,
"learning_rate": 5.747126436781609e-07,
"loss": 0.4337,
"step": 2296
},
{
"epoch": 0.9427457418428073,
"grad_norm": 3.148480176925659,
"learning_rate": 5.706075533661741e-07,
"loss": 0.0009,
"step": 2297
},
{
"epoch": 0.9431561666324646,
"grad_norm": 0.8846041560173035,
"learning_rate": 5.665024630541873e-07,
"loss": 0.0006,
"step": 2298
},
{
"epoch": 0.9435665914221218,
"grad_norm": 1.0956591367721558,
"learning_rate": 5.623973727422003e-07,
"loss": 0.0008,
"step": 2299
},
{
"epoch": 0.9439770162117792,
"grad_norm": 2.240388870239258,
"learning_rate": 5.582922824302135e-07,
"loss": 0.0008,
"step": 2300
},
{
"epoch": 0.9443874410014365,
"grad_norm": 1.7029458284378052,
"learning_rate": 5.541871921182267e-07,
"loss": 0.0012,
"step": 2301
},
{
"epoch": 0.9447978657910938,
"grad_norm": 88.86080169677734,
"learning_rate": 5.500821018062399e-07,
"loss": 0.3713,
"step": 2302
},
{
"epoch": 0.9452082905807511,
"grad_norm": 0.8428985476493835,
"learning_rate": 5.459770114942529e-07,
"loss": 0.0013,
"step": 2303
},
{
"epoch": 0.9456187153704084,
"grad_norm": 1.917651891708374,
"learning_rate": 5.418719211822661e-07,
"loss": 0.0015,
"step": 2304
},
{
"epoch": 0.9460291401600657,
"grad_norm": 4.175112247467041,
"learning_rate": 5.377668308702792e-07,
"loss": 0.0033,
"step": 2305
},
{
"epoch": 0.9464395649497229,
"grad_norm": 0.2951417863368988,
"learning_rate": 5.336617405582923e-07,
"loss": 0.0003,
"step": 2306
},
{
"epoch": 0.9468499897393803,
"grad_norm": 0.38347041606903076,
"learning_rate": 5.295566502463055e-07,
"loss": 0.0002,
"step": 2307
},
{
"epoch": 0.9472604145290375,
"grad_norm": 2.6731433868408203,
"learning_rate": 5.254515599343186e-07,
"loss": 0.0008,
"step": 2308
},
{
"epoch": 0.9476708393186949,
"grad_norm": 0.9371820092201233,
"learning_rate": 5.213464696223317e-07,
"loss": 0.0003,
"step": 2309
},
{
"epoch": 0.9480812641083521,
"grad_norm": 7.747935771942139,
"learning_rate": 5.172413793103449e-07,
"loss": 0.0034,
"step": 2310
},
{
"epoch": 0.9484916888980094,
"grad_norm": 1.9723658561706543,
"learning_rate": 5.13136288998358e-07,
"loss": 0.0013,
"step": 2311
},
{
"epoch": 0.9489021136876667,
"grad_norm": 4.726565361022949,
"learning_rate": 5.090311986863711e-07,
"loss": 0.0018,
"step": 2312
},
{
"epoch": 0.949312538477324,
"grad_norm": 1.9389468431472778,
"learning_rate": 5.049261083743843e-07,
"loss": 0.0022,
"step": 2313
},
{
"epoch": 0.9497229632669814,
"grad_norm": 3.7789430618286133,
"learning_rate": 5.008210180623975e-07,
"loss": 0.002,
"step": 2314
},
{
"epoch": 0.9501333880566386,
"grad_norm": 0.4461725652217865,
"learning_rate": 4.967159277504105e-07,
"loss": 0.0004,
"step": 2315
},
{
"epoch": 0.950543812846296,
"grad_norm": 3.646289348602295,
"learning_rate": 4.926108374384237e-07,
"loss": 0.0013,
"step": 2316
},
{
"epoch": 0.9509542376359532,
"grad_norm": 85.94480895996094,
"learning_rate": 4.885057471264369e-07,
"loss": 0.0974,
"step": 2317
},
{
"epoch": 0.9513646624256105,
"grad_norm": 2.2480366230010986,
"learning_rate": 4.844006568144499e-07,
"loss": 0.0007,
"step": 2318
},
{
"epoch": 0.9517750872152678,
"grad_norm": 1.4306501150131226,
"learning_rate": 4.802955665024631e-07,
"loss": 0.0007,
"step": 2319
},
{
"epoch": 0.9521855120049251,
"grad_norm": 3.0637645721435547,
"learning_rate": 4.7619047619047623e-07,
"loss": 0.0009,
"step": 2320
},
{
"epoch": 0.9525959367945824,
"grad_norm": 3.191483974456787,
"learning_rate": 4.7208538587848934e-07,
"loss": 0.0009,
"step": 2321
},
{
"epoch": 0.9530063615842397,
"grad_norm": 1.9184027910232544,
"learning_rate": 4.6798029556650246e-07,
"loss": 0.0006,
"step": 2322
},
{
"epoch": 0.9534167863738969,
"grad_norm": 3.060690402984619,
"learning_rate": 4.6387520525451563e-07,
"loss": 0.0011,
"step": 2323
},
{
"epoch": 0.9538272111635543,
"grad_norm": 6.263881206512451,
"learning_rate": 4.5977011494252875e-07,
"loss": 0.0038,
"step": 2324
},
{
"epoch": 0.9542376359532115,
"grad_norm": 4.2969255447387695,
"learning_rate": 4.556650246305419e-07,
"loss": 0.0025,
"step": 2325
},
{
"epoch": 0.9546480607428689,
"grad_norm": 1.9044766426086426,
"learning_rate": 4.5155993431855504e-07,
"loss": 0.0007,
"step": 2326
},
{
"epoch": 0.9550584855325261,
"grad_norm": 0.7727172374725342,
"learning_rate": 4.474548440065682e-07,
"loss": 0.0003,
"step": 2327
},
{
"epoch": 0.9554689103221835,
"grad_norm": 7.0687055587768555,
"learning_rate": 4.433497536945813e-07,
"loss": 0.0039,
"step": 2328
},
{
"epoch": 0.9558793351118408,
"grad_norm": 2.839017629623413,
"learning_rate": 4.392446633825945e-07,
"loss": 0.0009,
"step": 2329
},
{
"epoch": 0.956289759901498,
"grad_norm": 3.28006911277771,
"learning_rate": 4.351395730706076e-07,
"loss": 0.0021,
"step": 2330
},
{
"epoch": 0.9567001846911554,
"grad_norm": 1.1960148811340332,
"learning_rate": 4.3103448275862073e-07,
"loss": 0.0007,
"step": 2331
},
{
"epoch": 0.9571106094808126,
"grad_norm": 9.051663398742676,
"learning_rate": 4.269293924466339e-07,
"loss": 0.0097,
"step": 2332
},
{
"epoch": 0.95752103427047,
"grad_norm": 3.090595006942749,
"learning_rate": 4.2282430213464696e-07,
"loss": 0.002,
"step": 2333
},
{
"epoch": 0.9579314590601272,
"grad_norm": 1.5346806049346924,
"learning_rate": 4.1871921182266013e-07,
"loss": 0.0006,
"step": 2334
},
{
"epoch": 0.9583418838497846,
"grad_norm": 2.3552331924438477,
"learning_rate": 4.1461412151067325e-07,
"loss": 0.0006,
"step": 2335
},
{
"epoch": 0.9587523086394418,
"grad_norm": 1.153477668762207,
"learning_rate": 4.1050903119868637e-07,
"loss": 0.0005,
"step": 2336
},
{
"epoch": 0.9591627334290991,
"grad_norm": 2.9100239276885986,
"learning_rate": 4.0640394088669954e-07,
"loss": 0.0011,
"step": 2337
},
{
"epoch": 0.9595731582187564,
"grad_norm": 1.8488872051239014,
"learning_rate": 4.0229885057471266e-07,
"loss": 0.0012,
"step": 2338
},
{
"epoch": 0.9599835830084137,
"grad_norm": 0.7377915382385254,
"learning_rate": 3.981937602627258e-07,
"loss": 0.0007,
"step": 2339
},
{
"epoch": 0.960394007798071,
"grad_norm": 3.259815216064453,
"learning_rate": 3.9408866995073894e-07,
"loss": 0.0011,
"step": 2340
},
{
"epoch": 0.9608044325877283,
"grad_norm": 21.21051597595215,
"learning_rate": 3.899835796387521e-07,
"loss": 0.012,
"step": 2341
},
{
"epoch": 0.9612148573773855,
"grad_norm": 0.8501923680305481,
"learning_rate": 3.8587848932676523e-07,
"loss": 0.0009,
"step": 2342
},
{
"epoch": 0.9616252821670429,
"grad_norm": 4.353545188903809,
"learning_rate": 3.817733990147784e-07,
"loss": 0.0022,
"step": 2343
},
{
"epoch": 0.9620357069567002,
"grad_norm": 133.92620849609375,
"learning_rate": 3.776683087027915e-07,
"loss": 0.3951,
"step": 2344
},
{
"epoch": 0.9624461317463575,
"grad_norm": 7.8884968757629395,
"learning_rate": 3.7356321839080463e-07,
"loss": 0.0046,
"step": 2345
},
{
"epoch": 0.9628565565360148,
"grad_norm": 1.040542483329773,
"learning_rate": 3.694581280788178e-07,
"loss": 0.0012,
"step": 2346
},
{
"epoch": 0.9632669813256721,
"grad_norm": 0.9905815720558167,
"learning_rate": 3.6535303776683087e-07,
"loss": 0.0005,
"step": 2347
},
{
"epoch": 0.9636774061153294,
"grad_norm": 1.3919775485992432,
"learning_rate": 3.61247947454844e-07,
"loss": 0.0017,
"step": 2348
},
{
"epoch": 0.9640878309049866,
"grad_norm": 1.7385010719299316,
"learning_rate": 3.5714285714285716e-07,
"loss": 0.0008,
"step": 2349
},
{
"epoch": 0.964498255694644,
"grad_norm": 3.3651723861694336,
"learning_rate": 3.530377668308703e-07,
"loss": 0.0043,
"step": 2350
},
{
"epoch": 0.9649086804843012,
"grad_norm": 4.083765506744385,
"learning_rate": 3.4893267651888344e-07,
"loss": 0.0015,
"step": 2351
},
{
"epoch": 0.9653191052739586,
"grad_norm": 5.205965042114258,
"learning_rate": 3.4482758620689656e-07,
"loss": 0.0024,
"step": 2352
},
{
"epoch": 0.9657295300636158,
"grad_norm": 1.1334319114685059,
"learning_rate": 3.4072249589490973e-07,
"loss": 0.0009,
"step": 2353
},
{
"epoch": 0.9661399548532731,
"grad_norm": 4.812343120574951,
"learning_rate": 3.3661740558292285e-07,
"loss": 0.002,
"step": 2354
},
{
"epoch": 0.9665503796429304,
"grad_norm": 1.7188268899917603,
"learning_rate": 3.32512315270936e-07,
"loss": 0.0005,
"step": 2355
},
{
"epoch": 0.9669608044325877,
"grad_norm": 0.82108074426651,
"learning_rate": 3.2840722495894914e-07,
"loss": 0.0002,
"step": 2356
},
{
"epoch": 0.967371229222245,
"grad_norm": 3.3514091968536377,
"learning_rate": 3.2430213464696225e-07,
"loss": 0.001,
"step": 2357
},
{
"epoch": 0.9677816540119023,
"grad_norm": 1.0918655395507812,
"learning_rate": 3.201970443349754e-07,
"loss": 0.0006,
"step": 2358
},
{
"epoch": 0.9681920788015597,
"grad_norm": 4.499138832092285,
"learning_rate": 3.1609195402298854e-07,
"loss": 0.0017,
"step": 2359
},
{
"epoch": 0.9686025035912169,
"grad_norm": 0.37328487634658813,
"learning_rate": 3.1198686371100166e-07,
"loss": 0.0002,
"step": 2360
},
{
"epoch": 0.9690129283808742,
"grad_norm": 0.759497880935669,
"learning_rate": 3.0788177339901483e-07,
"loss": 0.0002,
"step": 2361
},
{
"epoch": 0.9694233531705315,
"grad_norm": 297.21331787109375,
"learning_rate": 3.0377668308702795e-07,
"loss": 0.6185,
"step": 2362
},
{
"epoch": 0.9698337779601888,
"grad_norm": 6.752994060516357,
"learning_rate": 2.996715927750411e-07,
"loss": 0.004,
"step": 2363
},
{
"epoch": 0.9702442027498461,
"grad_norm": 0.5635434985160828,
"learning_rate": 2.955665024630542e-07,
"loss": 0.0002,
"step": 2364
},
{
"epoch": 0.9706546275395034,
"grad_norm": 0.26415130496025085,
"learning_rate": 2.9146141215106735e-07,
"loss": 0.0004,
"step": 2365
},
{
"epoch": 0.9710650523291606,
"grad_norm": 9.474302291870117,
"learning_rate": 2.8735632183908047e-07,
"loss": 0.0055,
"step": 2366
},
{
"epoch": 0.971475477118818,
"grad_norm": 0.971859872341156,
"learning_rate": 2.8325123152709364e-07,
"loss": 0.0012,
"step": 2367
},
{
"epoch": 0.9718859019084752,
"grad_norm": 7.848605632781982,
"learning_rate": 2.7914614121510676e-07,
"loss": 0.0074,
"step": 2368
},
{
"epoch": 0.9722963266981326,
"grad_norm": 3.5056517124176025,
"learning_rate": 2.750410509031199e-07,
"loss": 0.0026,
"step": 2369
},
{
"epoch": 0.9727067514877898,
"grad_norm": 0.788481593132019,
"learning_rate": 2.7093596059113304e-07,
"loss": 0.0004,
"step": 2370
},
{
"epoch": 0.9731171762774472,
"grad_norm": 1.0786170959472656,
"learning_rate": 2.6683087027914616e-07,
"loss": 0.0006,
"step": 2371
},
{
"epoch": 0.9735276010671045,
"grad_norm": 4.725919723510742,
"learning_rate": 2.627257799671593e-07,
"loss": 0.0019,
"step": 2372
},
{
"epoch": 0.9739380258567617,
"grad_norm": 4.775391101837158,
"learning_rate": 2.5862068965517245e-07,
"loss": 0.0025,
"step": 2373
},
{
"epoch": 0.9743484506464191,
"grad_norm": 1.2527748346328735,
"learning_rate": 2.5451559934318556e-07,
"loss": 0.0009,
"step": 2374
},
{
"epoch": 0.9747588754360763,
"grad_norm": 9.04389762878418,
"learning_rate": 2.5041050903119874e-07,
"loss": 0.0061,
"step": 2375
},
{
"epoch": 0.9751693002257337,
"grad_norm": 0.8954558372497559,
"learning_rate": 2.4630541871921185e-07,
"loss": 0.0008,
"step": 2376
},
{
"epoch": 0.9755797250153909,
"grad_norm": 0.9494397640228271,
"learning_rate": 2.4220032840722497e-07,
"loss": 0.0008,
"step": 2377
},
{
"epoch": 0.9759901498050483,
"grad_norm": 2.322842836380005,
"learning_rate": 2.3809523809523811e-07,
"loss": 0.0009,
"step": 2378
},
{
"epoch": 0.9764005745947055,
"grad_norm": 1.768543004989624,
"learning_rate": 2.3399014778325123e-07,
"loss": 0.0019,
"step": 2379
},
{
"epoch": 0.9768109993843628,
"grad_norm": 0.8863083124160767,
"learning_rate": 2.2988505747126437e-07,
"loss": 0.0005,
"step": 2380
},
{
"epoch": 0.9772214241740201,
"grad_norm": 2.6160225868225098,
"learning_rate": 2.2577996715927752e-07,
"loss": 0.0007,
"step": 2381
},
{
"epoch": 0.9776318489636774,
"grad_norm": 2.446171283721924,
"learning_rate": 2.2167487684729066e-07,
"loss": 0.0012,
"step": 2382
},
{
"epoch": 0.9780422737533347,
"grad_norm": 17.26587677001953,
"learning_rate": 2.175697865353038e-07,
"loss": 0.0087,
"step": 2383
},
{
"epoch": 0.978452698542992,
"grad_norm": 1.4119954109191895,
"learning_rate": 2.1346469622331695e-07,
"loss": 0.0005,
"step": 2384
},
{
"epoch": 0.9788631233326492,
"grad_norm": 13.542911529541016,
"learning_rate": 2.0935960591133007e-07,
"loss": 0.0126,
"step": 2385
},
{
"epoch": 0.9792735481223066,
"grad_norm": 4.336092472076416,
"learning_rate": 2.0525451559934318e-07,
"loss": 0.0021,
"step": 2386
},
{
"epoch": 0.9796839729119639,
"grad_norm": 3.5591189861297607,
"learning_rate": 2.0114942528735633e-07,
"loss": 0.0013,
"step": 2387
},
{
"epoch": 0.9800943977016212,
"grad_norm": 1.2056372165679932,
"learning_rate": 1.9704433497536947e-07,
"loss": 0.001,
"step": 2388
},
{
"epoch": 0.9805048224912785,
"grad_norm": 1.6742585897445679,
"learning_rate": 1.9293924466338262e-07,
"loss": 0.0005,
"step": 2389
},
{
"epoch": 0.9809152472809358,
"grad_norm": 2.1341464519500732,
"learning_rate": 1.8883415435139576e-07,
"loss": 0.0011,
"step": 2390
},
{
"epoch": 0.9813256720705931,
"grad_norm": 3.184566020965576,
"learning_rate": 1.847290640394089e-07,
"loss": 0.0017,
"step": 2391
},
{
"epoch": 0.9817360968602503,
"grad_norm": 3.60213565826416,
"learning_rate": 1.80623973727422e-07,
"loss": 0.0011,
"step": 2392
},
{
"epoch": 0.9821465216499077,
"grad_norm": 2.4714314937591553,
"learning_rate": 1.7651888341543514e-07,
"loss": 0.0006,
"step": 2393
},
{
"epoch": 0.9825569464395649,
"grad_norm": 1.7221025228500366,
"learning_rate": 1.7241379310344828e-07,
"loss": 0.0006,
"step": 2394
},
{
"epoch": 0.9829673712292223,
"grad_norm": 1.4830291271209717,
"learning_rate": 1.6830870279146142e-07,
"loss": 0.0007,
"step": 2395
},
{
"epoch": 0.9833777960188795,
"grad_norm": 7.292273998260498,
"learning_rate": 1.6420361247947457e-07,
"loss": 0.0032,
"step": 2396
},
{
"epoch": 0.9837882208085368,
"grad_norm": 2.007511615753174,
"learning_rate": 1.600985221674877e-07,
"loss": 0.0012,
"step": 2397
},
{
"epoch": 0.9841986455981941,
"grad_norm": 0.7533149719238281,
"learning_rate": 1.5599343185550083e-07,
"loss": 0.0004,
"step": 2398
},
{
"epoch": 0.9846090703878514,
"grad_norm": 1.2285100221633911,
"learning_rate": 1.5188834154351397e-07,
"loss": 0.0018,
"step": 2399
},
{
"epoch": 0.9850194951775088,
"grad_norm": 20.67816734313965,
"learning_rate": 1.477832512315271e-07,
"loss": 0.0174,
"step": 2400
},
{
"epoch": 0.985429919967166,
"grad_norm": 0.7094917297363281,
"learning_rate": 1.4367816091954023e-07,
"loss": 0.0006,
"step": 2401
},
{
"epoch": 0.9858403447568234,
"grad_norm": 3.2195661067962646,
"learning_rate": 1.3957307060755338e-07,
"loss": 0.0022,
"step": 2402
},
{
"epoch": 0.9862507695464806,
"grad_norm": 1.73604154586792,
"learning_rate": 1.3546798029556652e-07,
"loss": 0.0019,
"step": 2403
},
{
"epoch": 0.9866611943361379,
"grad_norm": 1.900382399559021,
"learning_rate": 1.3136288998357964e-07,
"loss": 0.0025,
"step": 2404
},
{
"epoch": 0.9870716191257952,
"grad_norm": 2.500910520553589,
"learning_rate": 1.2725779967159278e-07,
"loss": 0.0006,
"step": 2405
},
{
"epoch": 0.9874820439154525,
"grad_norm": 0.6686133146286011,
"learning_rate": 1.2315270935960593e-07,
"loss": 0.0005,
"step": 2406
},
{
"epoch": 0.9878924687051098,
"grad_norm": 2.1467971801757812,
"learning_rate": 1.1904761904761906e-07,
"loss": 0.001,
"step": 2407
},
{
"epoch": 0.9883028934947671,
"grad_norm": 1.3372756242752075,
"learning_rate": 1.1494252873563219e-07,
"loss": 0.001,
"step": 2408
},
{
"epoch": 0.9887133182844243,
"grad_norm": 0.9334439039230347,
"learning_rate": 1.1083743842364533e-07,
"loss": 0.0003,
"step": 2409
},
{
"epoch": 0.9891237430740817,
"grad_norm": 5.272099494934082,
"learning_rate": 1.0673234811165847e-07,
"loss": 0.0027,
"step": 2410
},
{
"epoch": 0.9895341678637389,
"grad_norm": 0.4123227000236511,
"learning_rate": 1.0262725779967159e-07,
"loss": 0.0009,
"step": 2411
},
{
"epoch": 0.9899445926533963,
"grad_norm": 1.5040390491485596,
"learning_rate": 9.852216748768474e-08,
"loss": 0.0012,
"step": 2412
},
{
"epoch": 0.9903550174430535,
"grad_norm": 3.656888961791992,
"learning_rate": 9.441707717569788e-08,
"loss": 0.0011,
"step": 2413
},
{
"epoch": 0.9907654422327109,
"grad_norm": 3.3520798683166504,
"learning_rate": 9.0311986863711e-08,
"loss": 0.001,
"step": 2414
},
{
"epoch": 0.9911758670223682,
"grad_norm": 3.941927433013916,
"learning_rate": 8.620689655172414e-08,
"loss": 0.0043,
"step": 2415
},
{
"epoch": 0.9915862918120254,
"grad_norm": 0.6502656936645508,
"learning_rate": 8.210180623973728e-08,
"loss": 0.0005,
"step": 2416
},
{
"epoch": 0.9919967166016828,
"grad_norm": 3.5289177894592285,
"learning_rate": 7.799671592775041e-08,
"loss": 0.0017,
"step": 2417
},
{
"epoch": 0.99240714139134,
"grad_norm": 0.9630202651023865,
"learning_rate": 7.389162561576355e-08,
"loss": 0.0003,
"step": 2418
},
{
"epoch": 0.9928175661809974,
"grad_norm": 1.088210940361023,
"learning_rate": 6.978653530377669e-08,
"loss": 0.0013,
"step": 2419
},
{
"epoch": 0.9932279909706546,
"grad_norm": 4.486795425415039,
"learning_rate": 6.568144499178982e-08,
"loss": 0.0025,
"step": 2420
},
{
"epoch": 0.993638415760312,
"grad_norm": 0.32438138127326965,
"learning_rate": 6.157635467980296e-08,
"loss": 0.0005,
"step": 2421
},
{
"epoch": 0.9940488405499692,
"grad_norm": 1.0020209550857544,
"learning_rate": 5.7471264367816094e-08,
"loss": 0.0004,
"step": 2422
},
{
"epoch": 0.9944592653396265,
"grad_norm": 2.0896785259246826,
"learning_rate": 5.336617405582924e-08,
"loss": 0.0014,
"step": 2423
},
{
"epoch": 0.9948696901292838,
"grad_norm": 0.6274513602256775,
"learning_rate": 4.926108374384237e-08,
"loss": 0.0008,
"step": 2424
},
{
"epoch": 0.9952801149189411,
"grad_norm": 2.2850875854492188,
"learning_rate": 4.51559934318555e-08,
"loss": 0.0039,
"step": 2425
},
{
"epoch": 0.9956905397085984,
"grad_norm": 2.633758783340454,
"learning_rate": 4.105090311986864e-08,
"loss": 0.0009,
"step": 2426
},
{
"epoch": 0.9961009644982557,
"grad_norm": 1.6427994966506958,
"learning_rate": 3.694581280788177e-08,
"loss": 0.0005,
"step": 2427
},
{
"epoch": 0.9965113892879129,
"grad_norm": 0.3915906846523285,
"learning_rate": 3.284072249589491e-08,
"loss": 0.0003,
"step": 2428
},
{
"epoch": 0.9969218140775703,
"grad_norm": 0.8069103360176086,
"learning_rate": 2.8735632183908047e-08,
"loss": 0.0008,
"step": 2429
},
{
"epoch": 0.9973322388672275,
"grad_norm": 2.8513858318328857,
"learning_rate": 2.4630541871921184e-08,
"loss": 0.0009,
"step": 2430
},
{
"epoch": 0.9977426636568849,
"grad_norm": 3.9494268894195557,
"learning_rate": 2.052545155993432e-08,
"loss": 0.0029,
"step": 2431
},
{
"epoch": 0.9981530884465422,
"grad_norm": 0.6888309121131897,
"learning_rate": 1.6420361247947455e-08,
"loss": 0.0003,
"step": 2432
},
{
"epoch": 0.9985635132361995,
"grad_norm": 3.6870369911193848,
"learning_rate": 1.2315270935960592e-08,
"loss": 0.0014,
"step": 2433
},
{
"epoch": 0.9989739380258568,
"grad_norm": 0.6045457720756531,
"learning_rate": 8.210180623973727e-09,
"loss": 0.0003,
"step": 2434
},
{
"epoch": 0.999384362815514,
"grad_norm": 2.450587749481201,
"learning_rate": 4.105090311986864e-09,
"loss": 0.0007,
"step": 2435
},
{
"epoch": 0.9997947876051714,
"grad_norm": 6.319553852081299,
"learning_rate": 0.0,
"loss": 0.0032,
"step": 2436
},
{
"epoch": 0.9997947876051714,
"eval_accuracy": 0.8535469107551488,
"eval_loss": 0.6232770681381226,
"eval_runtime": 347.7,
"eval_samples_per_second": 1.26,
"eval_steps_per_second": 1.26,
"step": 2436
}
],
"logging_steps": 1,
"max_steps": 2436,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}