mmpe / final_w_trainer_state.json
zooblastlbz's picture
Upload final_w_trainer_state.json with huggingface_hub
4a7e43d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9998853342506593,
"eval_steps": 500,
"global_step": 2180,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00045866299736268775,
"grad_norm": 114.75894927978516,
"learning_rate": 1.5151515151515153e-05,
"loss": 8.6105,
"step": 1
},
{
"epoch": 0.0009173259947253755,
"grad_norm": 108.94357299804688,
"learning_rate": 3.0303030303030306e-05,
"loss": 8.5059,
"step": 2
},
{
"epoch": 0.0013759889920880633,
"grad_norm": 55.77202606201172,
"learning_rate": 4.545454545454546e-05,
"loss": 7.7437,
"step": 3
},
{
"epoch": 0.001834651989450751,
"grad_norm": 78.87145233154297,
"learning_rate": 6.060606060606061e-05,
"loss": 7.9607,
"step": 4
},
{
"epoch": 0.002293314986813439,
"grad_norm": 27.655637741088867,
"learning_rate": 7.575757575757576e-05,
"loss": 6.6218,
"step": 5
},
{
"epoch": 0.0027519779841761265,
"grad_norm": 21.756202697753906,
"learning_rate": 9.090909090909092e-05,
"loss": 6.2205,
"step": 6
},
{
"epoch": 0.0032106409815388145,
"grad_norm": 9.849833488464355,
"learning_rate": 0.00010606060606060606,
"loss": 5.7484,
"step": 7
},
{
"epoch": 0.003669303978901502,
"grad_norm": 7.7020955085754395,
"learning_rate": 0.00012121212121212122,
"loss": 5.284,
"step": 8
},
{
"epoch": 0.0041279669762641896,
"grad_norm": 11.210550308227539,
"learning_rate": 0.00013636363636363637,
"loss": 5.1587,
"step": 9
},
{
"epoch": 0.004586629973626878,
"grad_norm": 6.496895790100098,
"learning_rate": 0.00015151515151515152,
"loss": 4.743,
"step": 10
},
{
"epoch": 0.0050452929709895655,
"grad_norm": 4.6291728019714355,
"learning_rate": 0.00016666666666666666,
"loss": 4.5103,
"step": 11
},
{
"epoch": 0.005503955968352253,
"grad_norm": 3.8536574840545654,
"learning_rate": 0.00018181818181818183,
"loss": 4.3426,
"step": 12
},
{
"epoch": 0.0059626189657149406,
"grad_norm": 2.5668675899505615,
"learning_rate": 0.00019696969696969695,
"loss": 4.2592,
"step": 13
},
{
"epoch": 0.006421281963077629,
"grad_norm": 2.748685836791992,
"learning_rate": 0.00021212121212121213,
"loss": 4.1112,
"step": 14
},
{
"epoch": 0.0068799449604403165,
"grad_norm": 2.6287596225738525,
"learning_rate": 0.00022727272727272727,
"loss": 3.9495,
"step": 15
},
{
"epoch": 0.007338607957803004,
"grad_norm": 2.2549757957458496,
"learning_rate": 0.00024242424242424245,
"loss": 3.7381,
"step": 16
},
{
"epoch": 0.007797270955165692,
"grad_norm": 3.1407511234283447,
"learning_rate": 0.00025757575757575756,
"loss": 3.8012,
"step": 17
},
{
"epoch": 0.008255933952528379,
"grad_norm": 2.2104954719543457,
"learning_rate": 0.00027272727272727274,
"loss": 3.6311,
"step": 18
},
{
"epoch": 0.008714596949891068,
"grad_norm": 1.1581884622573853,
"learning_rate": 0.0002878787878787879,
"loss": 3.5415,
"step": 19
},
{
"epoch": 0.009173259947253756,
"grad_norm": 1.9451568126678467,
"learning_rate": 0.00030303030303030303,
"loss": 3.4077,
"step": 20
},
{
"epoch": 0.009631922944616443,
"grad_norm": 2.1019444465637207,
"learning_rate": 0.0003181818181818182,
"loss": 3.4285,
"step": 21
},
{
"epoch": 0.010090585941979131,
"grad_norm": 0.7417649030685425,
"learning_rate": 0.0003333333333333333,
"loss": 3.3311,
"step": 22
},
{
"epoch": 0.010549248939341819,
"grad_norm": 0.7884619832038879,
"learning_rate": 0.0003484848484848485,
"loss": 3.2606,
"step": 23
},
{
"epoch": 0.011007911936704506,
"grad_norm": 1.264022946357727,
"learning_rate": 0.00036363636363636367,
"loss": 3.4401,
"step": 24
},
{
"epoch": 0.011466574934067194,
"grad_norm": 0.6834126114845276,
"learning_rate": 0.0003787878787878788,
"loss": 3.3168,
"step": 25
},
{
"epoch": 0.011925237931429881,
"grad_norm": 0.7315241694450378,
"learning_rate": 0.0003939393939393939,
"loss": 3.1839,
"step": 26
},
{
"epoch": 0.01238390092879257,
"grad_norm": 0.739183247089386,
"learning_rate": 0.00040909090909090913,
"loss": 3.3526,
"step": 27
},
{
"epoch": 0.012842563926155258,
"grad_norm": 4.846123218536377,
"learning_rate": 0.00042424242424242425,
"loss": 3.3491,
"step": 28
},
{
"epoch": 0.013301226923517945,
"grad_norm": 1.3633110523223877,
"learning_rate": 0.0004393939393939394,
"loss": 3.2364,
"step": 29
},
{
"epoch": 0.013759889920880633,
"grad_norm": 1.200721263885498,
"learning_rate": 0.00045454545454545455,
"loss": 3.2942,
"step": 30
},
{
"epoch": 0.01421855291824332,
"grad_norm": 0.9348095059394836,
"learning_rate": 0.0004696969696969697,
"loss": 3.2398,
"step": 31
},
{
"epoch": 0.014677215915606008,
"grad_norm": 0.6360570192337036,
"learning_rate": 0.0004848484848484849,
"loss": 3.1595,
"step": 32
},
{
"epoch": 0.015135878912968696,
"grad_norm": 0.5648643970489502,
"learning_rate": 0.0005,
"loss": 3.1165,
"step": 33
},
{
"epoch": 0.015594541910331383,
"grad_norm": 0.5854682326316833,
"learning_rate": 0.0005151515151515151,
"loss": 3.1945,
"step": 34
},
{
"epoch": 0.01605320490769407,
"grad_norm": 0.6658660173416138,
"learning_rate": 0.0005303030303030302,
"loss": 3.0877,
"step": 35
},
{
"epoch": 0.016511867905056758,
"grad_norm": 0.7264187335968018,
"learning_rate": 0.0005454545454545455,
"loss": 3.0505,
"step": 36
},
{
"epoch": 0.016970530902419446,
"grad_norm": 0.5646975636482239,
"learning_rate": 0.0005606060606060606,
"loss": 3.1443,
"step": 37
},
{
"epoch": 0.017429193899782137,
"grad_norm": 0.7845568656921387,
"learning_rate": 0.0005757575757575758,
"loss": 3.1388,
"step": 38
},
{
"epoch": 0.017887856897144824,
"grad_norm": 0.5855001211166382,
"learning_rate": 0.0005909090909090909,
"loss": 3.1592,
"step": 39
},
{
"epoch": 0.018346519894507512,
"grad_norm": 0.6799588799476624,
"learning_rate": 0.0006060606060606061,
"loss": 3.0664,
"step": 40
},
{
"epoch": 0.0188051828918702,
"grad_norm": 0.5486564636230469,
"learning_rate": 0.0006212121212121212,
"loss": 2.894,
"step": 41
},
{
"epoch": 0.019263845889232887,
"grad_norm": 0.8732437491416931,
"learning_rate": 0.0006363636363636364,
"loss": 3.0392,
"step": 42
},
{
"epoch": 0.019722508886595574,
"grad_norm": 0.7626765370368958,
"learning_rate": 0.0006515151515151515,
"loss": 2.943,
"step": 43
},
{
"epoch": 0.020181171883958262,
"grad_norm": 0.7931069135665894,
"learning_rate": 0.0006666666666666666,
"loss": 2.8432,
"step": 44
},
{
"epoch": 0.02063983488132095,
"grad_norm": 0.8657431602478027,
"learning_rate": 0.0006818181818181818,
"loss": 2.7774,
"step": 45
},
{
"epoch": 0.021098497878683637,
"grad_norm": 0.6349948644638062,
"learning_rate": 0.000696969696969697,
"loss": 2.8877,
"step": 46
},
{
"epoch": 0.021557160876046325,
"grad_norm": 0.5843309760093689,
"learning_rate": 0.0007121212121212122,
"loss": 2.8844,
"step": 47
},
{
"epoch": 0.022015823873409012,
"grad_norm": 0.6262699365615845,
"learning_rate": 0.0007272727272727273,
"loss": 2.8671,
"step": 48
},
{
"epoch": 0.0224744868707717,
"grad_norm": 0.5009185075759888,
"learning_rate": 0.0007424242424242425,
"loss": 2.7868,
"step": 49
},
{
"epoch": 0.022933149868134387,
"grad_norm": 0.46156221628189087,
"learning_rate": 0.0007575757575757576,
"loss": 2.8813,
"step": 50
},
{
"epoch": 0.023391812865497075,
"grad_norm": 0.6038112640380859,
"learning_rate": 0.0007727272727272727,
"loss": 2.7809,
"step": 51
},
{
"epoch": 0.023850475862859762,
"grad_norm": 0.5297899842262268,
"learning_rate": 0.0007878787878787878,
"loss": 2.7979,
"step": 52
},
{
"epoch": 0.024309138860222453,
"grad_norm": 0.515750527381897,
"learning_rate": 0.000803030303030303,
"loss": 2.7394,
"step": 53
},
{
"epoch": 0.02476780185758514,
"grad_norm": 0.5053955912590027,
"learning_rate": 0.0008181818181818183,
"loss": 2.7319,
"step": 54
},
{
"epoch": 0.02522646485494783,
"grad_norm": 0.5771005153656006,
"learning_rate": 0.0008333333333333334,
"loss": 2.787,
"step": 55
},
{
"epoch": 0.025685127852310516,
"grad_norm": 0.47265496850013733,
"learning_rate": 0.0008484848484848485,
"loss": 2.7741,
"step": 56
},
{
"epoch": 0.026143790849673203,
"grad_norm": 0.48386165499687195,
"learning_rate": 0.0008636363636363636,
"loss": 2.6441,
"step": 57
},
{
"epoch": 0.02660245384703589,
"grad_norm": 0.4474497139453888,
"learning_rate": 0.0008787878787878789,
"loss": 2.6883,
"step": 58
},
{
"epoch": 0.02706111684439858,
"grad_norm": 0.45787569880485535,
"learning_rate": 0.000893939393939394,
"loss": 2.6505,
"step": 59
},
{
"epoch": 0.027519779841761266,
"grad_norm": 0.42940661311149597,
"learning_rate": 0.0009090909090909091,
"loss": 2.6752,
"step": 60
},
{
"epoch": 0.027978442839123954,
"grad_norm": 0.45745036005973816,
"learning_rate": 0.0009242424242424242,
"loss": 2.6266,
"step": 61
},
{
"epoch": 0.02843710583648664,
"grad_norm": 0.4927731454372406,
"learning_rate": 0.0009393939393939394,
"loss": 2.5905,
"step": 62
},
{
"epoch": 0.02889576883384933,
"grad_norm": 0.5060515403747559,
"learning_rate": 0.0009545454545454546,
"loss": 2.6615,
"step": 63
},
{
"epoch": 0.029354431831212016,
"grad_norm": 0.48067018389701843,
"learning_rate": 0.0009696969696969698,
"loss": 2.5592,
"step": 64
},
{
"epoch": 0.029813094828574704,
"grad_norm": 0.4308903217315674,
"learning_rate": 0.000984848484848485,
"loss": 2.6452,
"step": 65
},
{
"epoch": 0.03027175782593739,
"grad_norm": 0.45210587978363037,
"learning_rate": 0.001,
"loss": 2.5125,
"step": 66
},
{
"epoch": 0.03073042082330008,
"grad_norm": 0.4237235188484192,
"learning_rate": 0.0009999994478847943,
"loss": 2.5568,
"step": 67
},
{
"epoch": 0.031189083820662766,
"grad_norm": 0.4498521685600281,
"learning_rate": 0.0009999977915403962,
"loss": 2.4816,
"step": 68
},
{
"epoch": 0.03164774681802546,
"grad_norm": 0.47080931067466736,
"learning_rate": 0.0009999950309704639,
"loss": 2.4533,
"step": 69
},
{
"epoch": 0.03210640981538814,
"grad_norm": 0.49079686403274536,
"learning_rate": 0.000999991166181094,
"loss": 2.4833,
"step": 70
},
{
"epoch": 0.03256507281275083,
"grad_norm": 0.46766960620880127,
"learning_rate": 0.0009999861971808216,
"loss": 2.4121,
"step": 71
},
{
"epoch": 0.033023735810113516,
"grad_norm": 0.5121811032295227,
"learning_rate": 0.0009999801239806208,
"loss": 2.4709,
"step": 72
},
{
"epoch": 0.03348239880747621,
"grad_norm": 0.42303532361984253,
"learning_rate": 0.0009999729465939035,
"loss": 2.38,
"step": 73
},
{
"epoch": 0.03394106180483889,
"grad_norm": 0.4594205915927887,
"learning_rate": 0.0009999646650365212,
"loss": 2.463,
"step": 74
},
{
"epoch": 0.03439972480220158,
"grad_norm": 0.41703885793685913,
"learning_rate": 0.0009999552793267634,
"loss": 2.3982,
"step": 75
},
{
"epoch": 0.034858387799564274,
"grad_norm": 0.4080115556716919,
"learning_rate": 0.0009999447894853577,
"loss": 2.4564,
"step": 76
},
{
"epoch": 0.03531705079692696,
"grad_norm": 0.4343782961368561,
"learning_rate": 0.0009999331955354708,
"loss": 2.4348,
"step": 77
},
{
"epoch": 0.03577571379428965,
"grad_norm": 0.3965030312538147,
"learning_rate": 0.0009999204975027073,
"loss": 2.2902,
"step": 78
},
{
"epoch": 0.03623437679165233,
"grad_norm": 0.4244868755340576,
"learning_rate": 0.0009999066954151103,
"loss": 2.3428,
"step": 79
},
{
"epoch": 0.036693039789015024,
"grad_norm": 0.3948209583759308,
"learning_rate": 0.0009998917893031614,
"loss": 2.4103,
"step": 80
},
{
"epoch": 0.03715170278637771,
"grad_norm": 0.4274502098560333,
"learning_rate": 0.0009998757791997801,
"loss": 2.4451,
"step": 81
},
{
"epoch": 0.0376103657837404,
"grad_norm": 0.3707510828971863,
"learning_rate": 0.0009998586651403238,
"loss": 2.422,
"step": 82
},
{
"epoch": 0.03806902878110308,
"grad_norm": 0.34929051995277405,
"learning_rate": 0.0009998404471625885,
"loss": 2.3257,
"step": 83
},
{
"epoch": 0.038527691778465774,
"grad_norm": 0.4033382833003998,
"learning_rate": 0.0009998211253068078,
"loss": 2.3239,
"step": 84
},
{
"epoch": 0.03898635477582846,
"grad_norm": 0.4308173358440399,
"learning_rate": 0.0009998006996156535,
"loss": 2.4686,
"step": 85
},
{
"epoch": 0.03944501777319115,
"grad_norm": 0.3923931419849396,
"learning_rate": 0.0009997791701342347,
"loss": 2.2921,
"step": 86
},
{
"epoch": 0.03990368077055383,
"grad_norm": 0.36265629529953003,
"learning_rate": 0.0009997565369100983,
"loss": 2.3231,
"step": 87
},
{
"epoch": 0.040362343767916524,
"grad_norm": 0.37883153557777405,
"learning_rate": 0.0009997327999932291,
"loss": 2.2902,
"step": 88
},
{
"epoch": 0.04082100676527921,
"grad_norm": 0.35857099294662476,
"learning_rate": 0.000999707959436049,
"loss": 2.2917,
"step": 89
},
{
"epoch": 0.0412796697626419,
"grad_norm": 0.3556186556816101,
"learning_rate": 0.0009996820152934176,
"loss": 2.3093,
"step": 90
},
{
"epoch": 0.04173833276000459,
"grad_norm": 0.38860446214675903,
"learning_rate": 0.000999654967622631,
"loss": 2.2785,
"step": 91
},
{
"epoch": 0.042196995757367274,
"grad_norm": 0.36782774329185486,
"learning_rate": 0.0009996268164834238,
"loss": 2.3211,
"step": 92
},
{
"epoch": 0.042655658754729965,
"grad_norm": 0.3567314147949219,
"learning_rate": 0.000999597561937966,
"loss": 2.2993,
"step": 93
},
{
"epoch": 0.04311432175209265,
"grad_norm": 0.36403676867485046,
"learning_rate": 0.0009995672040508656,
"loss": 2.227,
"step": 94
},
{
"epoch": 0.04357298474945534,
"grad_norm": 0.37111398577690125,
"learning_rate": 0.0009995357428891662,
"loss": 2.272,
"step": 95
},
{
"epoch": 0.044031647746818024,
"grad_norm": 0.34938064217567444,
"learning_rate": 0.0009995031785223491,
"loss": 2.2402,
"step": 96
},
{
"epoch": 0.044490310744180715,
"grad_norm": 0.44558456540107727,
"learning_rate": 0.000999469511022331,
"loss": 2.4394,
"step": 97
},
{
"epoch": 0.0449489737415434,
"grad_norm": 0.33018913865089417,
"learning_rate": 0.0009994347404634657,
"loss": 2.1867,
"step": 98
},
{
"epoch": 0.04540763673890609,
"grad_norm": 0.3836520314216614,
"learning_rate": 0.0009993988669225423,
"loss": 2.3727,
"step": 99
},
{
"epoch": 0.045866299736268774,
"grad_norm": 0.3434332013130188,
"learning_rate": 0.000999361890478786,
"loss": 2.2951,
"step": 100
},
{
"epoch": 0.046324962733631465,
"grad_norm": 0.329985111951828,
"learning_rate": 0.0009993238112138583,
"loss": 2.3775,
"step": 101
},
{
"epoch": 0.04678362573099415,
"grad_norm": 0.7156427502632141,
"learning_rate": 0.0009992846292118554,
"loss": 2.2353,
"step": 102
},
{
"epoch": 0.04724228872835684,
"grad_norm": 0.36377307772636414,
"learning_rate": 0.000999244344559309,
"loss": 2.2568,
"step": 103
},
{
"epoch": 0.047700951725719525,
"grad_norm": 0.35666197538375854,
"learning_rate": 0.0009992029573451869,
"loss": 2.1644,
"step": 104
},
{
"epoch": 0.048159614723082216,
"grad_norm": 0.3300400972366333,
"learning_rate": 0.0009991604676608905,
"loss": 2.2062,
"step": 105
},
{
"epoch": 0.04861827772044491,
"grad_norm": 0.3577040135860443,
"learning_rate": 0.0009991168756002568,
"loss": 2.2576,
"step": 106
},
{
"epoch": 0.04907694071780759,
"grad_norm": 0.32270097732543945,
"learning_rate": 0.0009990721812595574,
"loss": 2.2129,
"step": 107
},
{
"epoch": 0.04953560371517028,
"grad_norm": 0.32343819737434387,
"learning_rate": 0.0009990263847374976,
"loss": 2.2346,
"step": 108
},
{
"epoch": 0.049994266712532966,
"grad_norm": 0.3173793852329254,
"learning_rate": 0.0009989794861352173,
"loss": 2.3034,
"step": 109
},
{
"epoch": 0.05045292970989566,
"grad_norm": 0.34496188163757324,
"learning_rate": 0.0009989314855562905,
"loss": 2.211,
"step": 110
},
{
"epoch": 0.05091159270725834,
"grad_norm": 0.34454038739204407,
"learning_rate": 0.0009988823831067245,
"loss": 2.1155,
"step": 111
},
{
"epoch": 0.05137025570462103,
"grad_norm": 0.3637445867061615,
"learning_rate": 0.0009988321788949597,
"loss": 2.298,
"step": 112
},
{
"epoch": 0.051828918701983716,
"grad_norm": 0.3216003179550171,
"learning_rate": 0.0009987808730318709,
"loss": 2.227,
"step": 113
},
{
"epoch": 0.05228758169934641,
"grad_norm": 0.386481910943985,
"learning_rate": 0.0009987284656307644,
"loss": 2.423,
"step": 114
},
{
"epoch": 0.05274624469670909,
"grad_norm": 0.3757198750972748,
"learning_rate": 0.0009986749568073802,
"loss": 2.2045,
"step": 115
},
{
"epoch": 0.05320490769407178,
"grad_norm": 0.3314107358455658,
"learning_rate": 0.0009986203466798905,
"loss": 2.1863,
"step": 116
},
{
"epoch": 0.053663570691434466,
"grad_norm": 0.31223729252815247,
"learning_rate": 0.0009985646353688996,
"loss": 2.2483,
"step": 117
},
{
"epoch": 0.05412223368879716,
"grad_norm": 0.33166366815567017,
"learning_rate": 0.0009985078229974437,
"loss": 2.2784,
"step": 118
},
{
"epoch": 0.05458089668615984,
"grad_norm": 0.31359270215034485,
"learning_rate": 0.0009984499096909905,
"loss": 2.3038,
"step": 119
},
{
"epoch": 0.05503955968352253,
"grad_norm": 0.32870644330978394,
"learning_rate": 0.0009983908955774397,
"loss": 2.2804,
"step": 120
},
{
"epoch": 0.05549822268088522,
"grad_norm": 0.38460472226142883,
"learning_rate": 0.0009983307807871211,
"loss": 2.1822,
"step": 121
},
{
"epoch": 0.05595688567824791,
"grad_norm": 0.32213419675827026,
"learning_rate": 0.0009982695654527965,
"loss": 2.3213,
"step": 122
},
{
"epoch": 0.0564155486756106,
"grad_norm": 0.3401404321193695,
"learning_rate": 0.0009982072497096571,
"loss": 2.3485,
"step": 123
},
{
"epoch": 0.05687421167297328,
"grad_norm": 0.2986947000026703,
"learning_rate": 0.000998143833695325,
"loss": 2.2822,
"step": 124
},
{
"epoch": 0.05733287467033597,
"grad_norm": 0.3037557303905487,
"learning_rate": 0.0009980793175498517,
"loss": 2.1775,
"step": 125
},
{
"epoch": 0.05779153766769866,
"grad_norm": 0.2960694432258606,
"learning_rate": 0.000998013701415719,
"loss": 2.223,
"step": 126
},
{
"epoch": 0.05825020066506135,
"grad_norm": 0.3112926185131073,
"learning_rate": 0.0009979469854378372,
"loss": 2.2062,
"step": 127
},
{
"epoch": 0.05870886366242403,
"grad_norm": 0.32468777894973755,
"learning_rate": 0.000997879169763546,
"loss": 2.2626,
"step": 128
},
{
"epoch": 0.05916752665978672,
"grad_norm": 0.297280877828598,
"learning_rate": 0.000997810254542614,
"loss": 2.2311,
"step": 129
},
{
"epoch": 0.05962618965714941,
"grad_norm": 0.3197387456893921,
"learning_rate": 0.0009977402399272374,
"loss": 2.2287,
"step": 130
},
{
"epoch": 0.0600848526545121,
"grad_norm": 0.35101935267448425,
"learning_rate": 0.0009976691260720407,
"loss": 2.2173,
"step": 131
},
{
"epoch": 0.06054351565187478,
"grad_norm": 0.3026212453842163,
"learning_rate": 0.0009975969131340763,
"loss": 2.2155,
"step": 132
},
{
"epoch": 0.06100217864923747,
"grad_norm": 0.34510040283203125,
"learning_rate": 0.0009975236012728236,
"loss": 2.2256,
"step": 133
},
{
"epoch": 0.06146084164660016,
"grad_norm": 0.31071922183036804,
"learning_rate": 0.0009974491906501886,
"loss": 2.2219,
"step": 134
},
{
"epoch": 0.06191950464396285,
"grad_norm": 0.3075829744338989,
"learning_rate": 0.0009973736814305049,
"loss": 2.2463,
"step": 135
},
{
"epoch": 0.06237816764132553,
"grad_norm": 0.29147791862487793,
"learning_rate": 0.0009972970737805312,
"loss": 2.2533,
"step": 136
},
{
"epoch": 0.06283683063868822,
"grad_norm": 0.3238203227519989,
"learning_rate": 0.0009972193678694525,
"loss": 2.2655,
"step": 137
},
{
"epoch": 0.06329549363605091,
"grad_norm": 0.30138716101646423,
"learning_rate": 0.0009971405638688794,
"loss": 2.2103,
"step": 138
},
{
"epoch": 0.0637541566334136,
"grad_norm": 0.31892716884613037,
"learning_rate": 0.0009970606619528475,
"loss": 2.2051,
"step": 139
},
{
"epoch": 0.06421281963077628,
"grad_norm": 0.3116203546524048,
"learning_rate": 0.000996979662297817,
"loss": 2.2006,
"step": 140
},
{
"epoch": 0.06467148262813897,
"grad_norm": 0.3188847005367279,
"learning_rate": 0.0009968975650826721,
"loss": 2.2805,
"step": 141
},
{
"epoch": 0.06513014562550166,
"grad_norm": 0.293490469455719,
"learning_rate": 0.000996814370488722,
"loss": 2.2412,
"step": 142
},
{
"epoch": 0.06558880862286436,
"grad_norm": 0.280108243227005,
"learning_rate": 0.000996730078699698,
"loss": 2.2775,
"step": 143
},
{
"epoch": 0.06604747162022703,
"grad_norm": 0.2839280366897583,
"learning_rate": 0.0009966446899017558,
"loss": 2.3075,
"step": 144
},
{
"epoch": 0.06650613461758972,
"grad_norm": 0.2796190083026886,
"learning_rate": 0.0009965582042834728,
"loss": 2.2516,
"step": 145
},
{
"epoch": 0.06696479761495241,
"grad_norm": 0.27943992614746094,
"learning_rate": 0.0009964706220358492,
"loss": 2.1755,
"step": 146
},
{
"epoch": 0.0674234606123151,
"grad_norm": 0.2933647930622101,
"learning_rate": 0.000996381943352307,
"loss": 2.1584,
"step": 147
},
{
"epoch": 0.06788212360967778,
"grad_norm": 0.34085574746131897,
"learning_rate": 0.0009962921684286896,
"loss": 2.3074,
"step": 148
},
{
"epoch": 0.06834078660704047,
"grad_norm": 0.28096017241477966,
"learning_rate": 0.0009962012974632614,
"loss": 2.221,
"step": 149
},
{
"epoch": 0.06879944960440317,
"grad_norm": 0.3065031170845032,
"learning_rate": 0.0009961093306567075,
"loss": 2.2413,
"step": 150
},
{
"epoch": 0.06925811260176586,
"grad_norm": 0.3134509325027466,
"learning_rate": 0.0009960162682121328,
"loss": 2.1774,
"step": 151
},
{
"epoch": 0.06971677559912855,
"grad_norm": 0.2973586320877075,
"learning_rate": 0.0009959221103350623,
"loss": 2.1904,
"step": 152
},
{
"epoch": 0.07017543859649122,
"grad_norm": 0.29555410146713257,
"learning_rate": 0.0009958268572334394,
"loss": 2.2793,
"step": 153
},
{
"epoch": 0.07063410159385392,
"grad_norm": 0.2969052493572235,
"learning_rate": 0.0009957305091176274,
"loss": 2.2076,
"step": 154
},
{
"epoch": 0.0710927645912166,
"grad_norm": 0.30087563395500183,
"learning_rate": 0.0009956330662004075,
"loss": 2.2273,
"step": 155
},
{
"epoch": 0.0715514275885793,
"grad_norm": 0.29606255888938904,
"learning_rate": 0.0009955345286969779,
"loss": 2.29,
"step": 156
},
{
"epoch": 0.07201009058594197,
"grad_norm": 0.322168231010437,
"learning_rate": 0.0009954348968249551,
"loss": 2.1962,
"step": 157
},
{
"epoch": 0.07246875358330467,
"grad_norm": 0.3281068205833435,
"learning_rate": 0.0009953341708043724,
"loss": 2.216,
"step": 158
},
{
"epoch": 0.07292741658066736,
"grad_norm": 0.3020949959754944,
"learning_rate": 0.0009952323508576793,
"loss": 2.1863,
"step": 159
},
{
"epoch": 0.07338607957803005,
"grad_norm": 0.3469793498516083,
"learning_rate": 0.0009951294372097406,
"loss": 2.2318,
"step": 160
},
{
"epoch": 0.07384474257539272,
"grad_norm": 0.3036288619041443,
"learning_rate": 0.0009950254300878378,
"loss": 2.1208,
"step": 161
},
{
"epoch": 0.07430340557275542,
"grad_norm": 0.2807892858982086,
"learning_rate": 0.000994920329721666,
"loss": 2.1267,
"step": 162
},
{
"epoch": 0.0747620685701181,
"grad_norm": 0.2851462662220001,
"learning_rate": 0.0009948141363433356,
"loss": 2.1157,
"step": 163
},
{
"epoch": 0.0752207315674808,
"grad_norm": 0.2624664604663849,
"learning_rate": 0.00099470685018737,
"loss": 2.1527,
"step": 164
},
{
"epoch": 0.07567939456484347,
"grad_norm": 0.3019235134124756,
"learning_rate": 0.0009945984714907073,
"loss": 2.1851,
"step": 165
},
{
"epoch": 0.07613805756220617,
"grad_norm": 0.30093374848365784,
"learning_rate": 0.000994489000492697,
"loss": 2.2436,
"step": 166
},
{
"epoch": 0.07659672055956886,
"grad_norm": 0.2693714499473572,
"learning_rate": 0.0009943784374351016,
"loss": 2.2097,
"step": 167
},
{
"epoch": 0.07705538355693155,
"grad_norm": 0.271639347076416,
"learning_rate": 0.0009942667825620951,
"loss": 2.2128,
"step": 168
},
{
"epoch": 0.07751404655429424,
"grad_norm": 0.2741166055202484,
"learning_rate": 0.0009941540361202634,
"loss": 2.2629,
"step": 169
},
{
"epoch": 0.07797270955165692,
"grad_norm": 0.2727080285549164,
"learning_rate": 0.0009940401983586022,
"loss": 2.1729,
"step": 170
},
{
"epoch": 0.0784313725490196,
"grad_norm": 0.31126195192337036,
"learning_rate": 0.000993925269528518,
"loss": 2.1706,
"step": 171
},
{
"epoch": 0.0788900355463823,
"grad_norm": 0.3058548867702484,
"learning_rate": 0.0009938092498838265,
"loss": 2.2048,
"step": 172
},
{
"epoch": 0.07934869854374499,
"grad_norm": 0.3121659755706787,
"learning_rate": 0.0009936921396807524,
"loss": 2.197,
"step": 173
},
{
"epoch": 0.07980736154110767,
"grad_norm": 0.3022741377353668,
"learning_rate": 0.0009935739391779292,
"loss": 2.1187,
"step": 174
},
{
"epoch": 0.08026602453847036,
"grad_norm": 0.2942226529121399,
"learning_rate": 0.000993454648636398,
"loss": 2.2634,
"step": 175
},
{
"epoch": 0.08072468753583305,
"grad_norm": 0.32404664158821106,
"learning_rate": 0.0009933342683196074,
"loss": 2.261,
"step": 176
},
{
"epoch": 0.08118335053319574,
"grad_norm": 0.3142852783203125,
"learning_rate": 0.0009932127984934125,
"loss": 2.1463,
"step": 177
},
{
"epoch": 0.08164201353055842,
"grad_norm": 0.2994116246700287,
"learning_rate": 0.0009930902394260745,
"loss": 2.2464,
"step": 178
},
{
"epoch": 0.08210067652792111,
"grad_norm": 0.31092846393585205,
"learning_rate": 0.0009929665913882607,
"loss": 2.179,
"step": 179
},
{
"epoch": 0.0825593395252838,
"grad_norm": 0.28189972043037415,
"learning_rate": 0.0009928418546530425,
"loss": 2.1905,
"step": 180
},
{
"epoch": 0.08301800252264649,
"grad_norm": 0.297859787940979,
"learning_rate": 0.0009927160294958964,
"loss": 2.1642,
"step": 181
},
{
"epoch": 0.08347666552000918,
"grad_norm": 0.29434502124786377,
"learning_rate": 0.000992589116194702,
"loss": 2.1873,
"step": 182
},
{
"epoch": 0.08393532851737186,
"grad_norm": 0.2727161943912506,
"learning_rate": 0.000992461115029743,
"loss": 2.2689,
"step": 183
},
{
"epoch": 0.08439399151473455,
"grad_norm": 0.28672897815704346,
"learning_rate": 0.000992332026283704,
"loss": 2.1696,
"step": 184
},
{
"epoch": 0.08485265451209724,
"grad_norm": 0.26875588297843933,
"learning_rate": 0.0009922018502416736,
"loss": 2.2052,
"step": 185
},
{
"epoch": 0.08531131750945993,
"grad_norm": 0.2772798240184784,
"learning_rate": 0.0009920705871911395,
"loss": 2.1098,
"step": 186
},
{
"epoch": 0.08576998050682261,
"grad_norm": 0.27010753750801086,
"learning_rate": 0.0009919382374219915,
"loss": 2.2374,
"step": 187
},
{
"epoch": 0.0862286435041853,
"grad_norm": 0.30180540680885315,
"learning_rate": 0.0009918048012265187,
"loss": 2.179,
"step": 188
},
{
"epoch": 0.08668730650154799,
"grad_norm": 0.2836782932281494,
"learning_rate": 0.0009916702788994097,
"loss": 2.2969,
"step": 189
},
{
"epoch": 0.08714596949891068,
"grad_norm": 0.2784883975982666,
"learning_rate": 0.0009915346707377519,
"loss": 2.1524,
"step": 190
},
{
"epoch": 0.08760463249627336,
"grad_norm": 0.2803403437137604,
"learning_rate": 0.0009913979770410305,
"loss": 2.0921,
"step": 191
},
{
"epoch": 0.08806329549363605,
"grad_norm": 0.2808901071548462,
"learning_rate": 0.0009912601981111285,
"loss": 2.1319,
"step": 192
},
{
"epoch": 0.08852195849099874,
"grad_norm": 0.26970943808555603,
"learning_rate": 0.0009911213342523248,
"loss": 2.1716,
"step": 193
},
{
"epoch": 0.08898062148836143,
"grad_norm": 0.29278749227523804,
"learning_rate": 0.000990981385771295,
"loss": 2.2402,
"step": 194
},
{
"epoch": 0.08943928448572411,
"grad_norm": 0.26599255204200745,
"learning_rate": 0.00099084035297711,
"loss": 2.1564,
"step": 195
},
{
"epoch": 0.0898979474830868,
"grad_norm": 0.30038803815841675,
"learning_rate": 0.000990698236181235,
"loss": 2.1814,
"step": 196
},
{
"epoch": 0.09035661048044949,
"grad_norm": 0.28352588415145874,
"learning_rate": 0.0009905550356975293,
"loss": 2.1633,
"step": 197
},
{
"epoch": 0.09081527347781218,
"grad_norm": 0.27703598141670227,
"learning_rate": 0.0009904107518422457,
"loss": 2.1777,
"step": 198
},
{
"epoch": 0.09127393647517487,
"grad_norm": 0.266642302274704,
"learning_rate": 0.0009902653849340295,
"loss": 2.0891,
"step": 199
},
{
"epoch": 0.09173259947253755,
"grad_norm": 0.2780592441558838,
"learning_rate": 0.0009901189352939177,
"loss": 2.1817,
"step": 200
},
{
"epoch": 0.09219126246990024,
"grad_norm": 0.2706533968448639,
"learning_rate": 0.0009899714032453387,
"loss": 2.2108,
"step": 201
},
{
"epoch": 0.09264992546726293,
"grad_norm": 0.24929527938365936,
"learning_rate": 0.000989822789114111,
"loss": 2.0834,
"step": 202
},
{
"epoch": 0.09310858846462562,
"grad_norm": 0.2574229836463928,
"learning_rate": 0.0009896730932284434,
"loss": 2.1052,
"step": 203
},
{
"epoch": 0.0935672514619883,
"grad_norm": 0.26673415303230286,
"learning_rate": 0.0009895223159189332,
"loss": 2.0967,
"step": 204
},
{
"epoch": 0.09402591445935099,
"grad_norm": 0.282674103975296,
"learning_rate": 0.0009893704575185663,
"loss": 2.1807,
"step": 205
},
{
"epoch": 0.09448457745671368,
"grad_norm": 0.2784089744091034,
"learning_rate": 0.000989217518362716,
"loss": 2.2249,
"step": 206
},
{
"epoch": 0.09494324045407637,
"grad_norm": 0.25985750555992126,
"learning_rate": 0.0009890634987891425,
"loss": 2.1209,
"step": 207
},
{
"epoch": 0.09540190345143905,
"grad_norm": 0.2884882986545563,
"learning_rate": 0.0009889083991379917,
"loss": 2.1601,
"step": 208
},
{
"epoch": 0.09586056644880174,
"grad_norm": 0.287451833486557,
"learning_rate": 0.0009887522197517954,
"loss": 2.1379,
"step": 209
},
{
"epoch": 0.09631922944616443,
"grad_norm": 0.2786431312561035,
"learning_rate": 0.0009885949609754693,
"loss": 2.1702,
"step": 210
},
{
"epoch": 0.09677789244352712,
"grad_norm": 0.299315869808197,
"learning_rate": 0.000988436623156314,
"loss": 2.2257,
"step": 211
},
{
"epoch": 0.09723655544088981,
"grad_norm": 0.2725549042224884,
"learning_rate": 0.0009882772066440114,
"loss": 2.1561,
"step": 212
},
{
"epoch": 0.09769521843825249,
"grad_norm": 0.27327272295951843,
"learning_rate": 0.0009881167117906276,
"loss": 2.1874,
"step": 213
},
{
"epoch": 0.09815388143561518,
"grad_norm": 0.2890051305294037,
"learning_rate": 0.0009879551389506084,
"loss": 2.1834,
"step": 214
},
{
"epoch": 0.09861254443297787,
"grad_norm": 0.2581946849822998,
"learning_rate": 0.0009877924884807814,
"loss": 2.106,
"step": 215
},
{
"epoch": 0.09907120743034056,
"grad_norm": 0.2918533682823181,
"learning_rate": 0.000987628760740354,
"loss": 2.1686,
"step": 216
},
{
"epoch": 0.09952987042770324,
"grad_norm": 0.3024649918079376,
"learning_rate": 0.0009874639560909118,
"loss": 2.1495,
"step": 217
},
{
"epoch": 0.09998853342506593,
"grad_norm": 0.26496902108192444,
"learning_rate": 0.0009872980748964202,
"loss": 2.2502,
"step": 218
},
{
"epoch": 0.10044719642242862,
"grad_norm": 0.25767233967781067,
"learning_rate": 0.000987131117523221,
"loss": 2.1762,
"step": 219
},
{
"epoch": 0.10090585941979131,
"grad_norm": 0.2679627537727356,
"learning_rate": 0.000986963084340033,
"loss": 2.1964,
"step": 220
},
{
"epoch": 0.10136452241715399,
"grad_norm": 0.2764521837234497,
"learning_rate": 0.0009867939757179508,
"loss": 2.1213,
"step": 221
},
{
"epoch": 0.10182318541451668,
"grad_norm": 0.29235517978668213,
"learning_rate": 0.0009866237920304443,
"loss": 2.1416,
"step": 222
},
{
"epoch": 0.10228184841187937,
"grad_norm": 0.26901116967201233,
"learning_rate": 0.0009864525336533577,
"loss": 2.2142,
"step": 223
},
{
"epoch": 0.10274051140924206,
"grad_norm": 0.2691214084625244,
"learning_rate": 0.000986280200964908,
"loss": 2.2038,
"step": 224
},
{
"epoch": 0.10319917440660474,
"grad_norm": 0.26474660634994507,
"learning_rate": 0.0009861067943456856,
"loss": 2.1406,
"step": 225
},
{
"epoch": 0.10365783740396743,
"grad_norm": 0.28170302510261536,
"learning_rate": 0.000985932314178652,
"loss": 2.1164,
"step": 226
},
{
"epoch": 0.10411650040133012,
"grad_norm": 0.28177449107170105,
"learning_rate": 0.00098575676084914,
"loss": 2.2322,
"step": 227
},
{
"epoch": 0.10457516339869281,
"grad_norm": 0.2555372416973114,
"learning_rate": 0.0009855801347448518,
"loss": 2.1738,
"step": 228
},
{
"epoch": 0.1050338263960555,
"grad_norm": 0.28172406554222107,
"learning_rate": 0.0009854024362558596,
"loss": 2.1962,
"step": 229
},
{
"epoch": 0.10549248939341818,
"grad_norm": 0.2951928973197937,
"learning_rate": 0.0009852236657746035,
"loss": 2.1303,
"step": 230
},
{
"epoch": 0.10595115239078087,
"grad_norm": 0.2907235622406006,
"learning_rate": 0.0009850438236958911,
"loss": 2.2706,
"step": 231
},
{
"epoch": 0.10640981538814356,
"grad_norm": 0.3058776259422302,
"learning_rate": 0.0009848629104168966,
"loss": 2.2238,
"step": 232
},
{
"epoch": 0.10686847838550625,
"grad_norm": 0.2716783881187439,
"learning_rate": 0.00098468092633716,
"loss": 2.1306,
"step": 233
},
{
"epoch": 0.10732714138286893,
"grad_norm": 0.3008482754230499,
"learning_rate": 0.0009844978718585855,
"loss": 2.1728,
"step": 234
},
{
"epoch": 0.10778580438023162,
"grad_norm": 0.2698744535446167,
"learning_rate": 0.0009843137473854423,
"loss": 2.1448,
"step": 235
},
{
"epoch": 0.10824446737759431,
"grad_norm": 0.24943993985652924,
"learning_rate": 0.000984128553324362,
"loss": 2.1674,
"step": 236
},
{
"epoch": 0.108703130374957,
"grad_norm": 0.23976926505565643,
"learning_rate": 0.0009839422900843383,
"loss": 2.2181,
"step": 237
},
{
"epoch": 0.10916179337231968,
"grad_norm": 0.27670568227767944,
"learning_rate": 0.0009837549580767261,
"loss": 2.2219,
"step": 238
},
{
"epoch": 0.10962045636968237,
"grad_norm": 0.251272976398468,
"learning_rate": 0.0009835665577152411,
"loss": 2.1529,
"step": 239
},
{
"epoch": 0.11007911936704506,
"grad_norm": 0.2478543072938919,
"learning_rate": 0.000983377089415958,
"loss": 2.1521,
"step": 240
},
{
"epoch": 0.11053778236440776,
"grad_norm": 0.25784021615982056,
"learning_rate": 0.0009831865535973102,
"loss": 2.0923,
"step": 241
},
{
"epoch": 0.11099644536177045,
"grad_norm": 0.2477918267250061,
"learning_rate": 0.0009829949506800885,
"loss": 2.173,
"step": 242
},
{
"epoch": 0.11145510835913312,
"grad_norm": 0.26784002780914307,
"learning_rate": 0.0009828022810874405,
"loss": 2.1819,
"step": 243
},
{
"epoch": 0.11191377135649581,
"grad_norm": 0.27960944175720215,
"learning_rate": 0.0009826085452448693,
"loss": 2.1968,
"step": 244
},
{
"epoch": 0.1123724343538585,
"grad_norm": 0.30150097608566284,
"learning_rate": 0.000982413743580233,
"loss": 2.273,
"step": 245
},
{
"epoch": 0.1128310973512212,
"grad_norm": 0.2823254466056824,
"learning_rate": 0.0009822178765237436,
"loss": 2.1208,
"step": 246
},
{
"epoch": 0.11328976034858387,
"grad_norm": 0.24425245821475983,
"learning_rate": 0.0009820209445079654,
"loss": 2.2135,
"step": 247
},
{
"epoch": 0.11374842334594656,
"grad_norm": 0.2557823956012726,
"learning_rate": 0.0009818229479678158,
"loss": 2.1678,
"step": 248
},
{
"epoch": 0.11420708634330926,
"grad_norm": 0.2726637125015259,
"learning_rate": 0.0009816238873405615,
"loss": 2.1591,
"step": 249
},
{
"epoch": 0.11466574934067195,
"grad_norm": 0.24426256120204926,
"learning_rate": 0.0009814237630658207,
"loss": 2.136,
"step": 250
},
{
"epoch": 0.11512441233803462,
"grad_norm": 0.26439622044563293,
"learning_rate": 0.00098122257558556,
"loss": 2.1589,
"step": 251
},
{
"epoch": 0.11558307533539731,
"grad_norm": 0.24074086546897888,
"learning_rate": 0.0009810203253440937,
"loss": 2.1162,
"step": 252
},
{
"epoch": 0.11604173833276,
"grad_norm": 0.27286311984062195,
"learning_rate": 0.0009808170127880837,
"loss": 2.1194,
"step": 253
},
{
"epoch": 0.1165004013301227,
"grad_norm": 0.25726890563964844,
"learning_rate": 0.000980612638366538,
"loss": 2.0884,
"step": 254
},
{
"epoch": 0.11695906432748537,
"grad_norm": 0.25794950127601624,
"learning_rate": 0.0009804072025308096,
"loss": 2.1956,
"step": 255
},
{
"epoch": 0.11741772732484806,
"grad_norm": 0.263917475938797,
"learning_rate": 0.000980200705734595,
"loss": 2.1418,
"step": 256
},
{
"epoch": 0.11787639032221076,
"grad_norm": 0.2613244652748108,
"learning_rate": 0.0009799931484339344,
"loss": 2.1684,
"step": 257
},
{
"epoch": 0.11833505331957345,
"grad_norm": 0.26328611373901367,
"learning_rate": 0.0009797845310872103,
"loss": 2.0856,
"step": 258
},
{
"epoch": 0.11879371631693614,
"grad_norm": 0.26411616802215576,
"learning_rate": 0.0009795748541551457,
"loss": 2.1892,
"step": 259
},
{
"epoch": 0.11925237931429881,
"grad_norm": 0.27116987109184265,
"learning_rate": 0.000979364118100804,
"loss": 2.0862,
"step": 260
},
{
"epoch": 0.1197110423116615,
"grad_norm": 0.2415204495191574,
"learning_rate": 0.0009791523233895875,
"loss": 2.2152,
"step": 261
},
{
"epoch": 0.1201697053090242,
"grad_norm": 0.24924740195274353,
"learning_rate": 0.0009789394704892364,
"loss": 2.1489,
"step": 262
},
{
"epoch": 0.12062836830638689,
"grad_norm": 0.2748185098171234,
"learning_rate": 0.0009787255598698282,
"loss": 2.1765,
"step": 263
},
{
"epoch": 0.12108703130374956,
"grad_norm": 0.24796909093856812,
"learning_rate": 0.0009785105920037758,
"loss": 2.2513,
"step": 264
},
{
"epoch": 0.12154569430111226,
"grad_norm": 0.26453086733818054,
"learning_rate": 0.0009782945673658275,
"loss": 2.1737,
"step": 265
},
{
"epoch": 0.12200435729847495,
"grad_norm": 0.2530960738658905,
"learning_rate": 0.0009780774864330654,
"loss": 2.1698,
"step": 266
},
{
"epoch": 0.12246302029583764,
"grad_norm": 0.24280093610286713,
"learning_rate": 0.000977859349684904,
"loss": 2.1192,
"step": 267
},
{
"epoch": 0.12292168329320032,
"grad_norm": 0.23096245527267456,
"learning_rate": 0.00097764015760309,
"loss": 2.1693,
"step": 268
},
{
"epoch": 0.123380346290563,
"grad_norm": 0.22807879745960236,
"learning_rate": 0.0009774199106717004,
"loss": 2.124,
"step": 269
},
{
"epoch": 0.1238390092879257,
"grad_norm": 0.25258147716522217,
"learning_rate": 0.0009771986093771417,
"loss": 2.0864,
"step": 270
},
{
"epoch": 0.12429767228528839,
"grad_norm": 0.2587060332298279,
"learning_rate": 0.0009769762542081496,
"loss": 2.1362,
"step": 271
},
{
"epoch": 0.12475633528265107,
"grad_norm": 0.2492268681526184,
"learning_rate": 0.000976752845655786,
"loss": 2.1656,
"step": 272
},
{
"epoch": 0.12521499828001376,
"grad_norm": 0.2765987813472748,
"learning_rate": 0.0009765283842134411,
"loss": 2.1152,
"step": 273
},
{
"epoch": 0.12567366127737645,
"grad_norm": 0.2676598131656647,
"learning_rate": 0.0009763028703768282,
"loss": 2.1937,
"step": 274
},
{
"epoch": 0.12613232427473914,
"grad_norm": 0.24017716944217682,
"learning_rate": 0.0009760763046439862,
"loss": 2.1192,
"step": 275
},
{
"epoch": 0.12659098727210183,
"grad_norm": 0.2421436905860901,
"learning_rate": 0.0009758486875152766,
"loss": 2.1596,
"step": 276
},
{
"epoch": 0.12704965026946452,
"grad_norm": 0.2523273825645447,
"learning_rate": 0.0009756200194933829,
"loss": 2.1703,
"step": 277
},
{
"epoch": 0.1275083132668272,
"grad_norm": 0.24269650876522064,
"learning_rate": 0.0009753903010833094,
"loss": 2.2076,
"step": 278
},
{
"epoch": 0.12796697626418987,
"grad_norm": 0.23615428805351257,
"learning_rate": 0.0009751595327923803,
"loss": 2.163,
"step": 279
},
{
"epoch": 0.12842563926155257,
"grad_norm": 0.24087460339069366,
"learning_rate": 0.0009749277151302382,
"loss": 2.1789,
"step": 280
},
{
"epoch": 0.12888430225891526,
"grad_norm": 0.24662022292613983,
"learning_rate": 0.0009746948486088435,
"loss": 2.1437,
"step": 281
},
{
"epoch": 0.12934296525627795,
"grad_norm": 0.2696092128753662,
"learning_rate": 0.0009744609337424727,
"loss": 2.1002,
"step": 282
},
{
"epoch": 0.12980162825364064,
"grad_norm": 0.2578447461128235,
"learning_rate": 0.0009742259710477177,
"loss": 2.1186,
"step": 283
},
{
"epoch": 0.13026029125100333,
"grad_norm": 0.26800933480262756,
"learning_rate": 0.0009739899610434841,
"loss": 2.1163,
"step": 284
},
{
"epoch": 0.13071895424836602,
"grad_norm": 0.24647380411624908,
"learning_rate": 0.0009737529042509913,
"loss": 2.1242,
"step": 285
},
{
"epoch": 0.1311776172457287,
"grad_norm": 0.26092424988746643,
"learning_rate": 0.0009735148011937693,
"loss": 2.1852,
"step": 286
},
{
"epoch": 0.13163628024309137,
"grad_norm": 0.26144853234291077,
"learning_rate": 0.00097327565239766,
"loss": 2.1949,
"step": 287
},
{
"epoch": 0.13209494324045407,
"grad_norm": 0.24162133038043976,
"learning_rate": 0.0009730354583908136,
"loss": 2.1568,
"step": 288
},
{
"epoch": 0.13255360623781676,
"grad_norm": 0.2673940360546112,
"learning_rate": 0.0009727942197036895,
"loss": 2.0649,
"step": 289
},
{
"epoch": 0.13301226923517945,
"grad_norm": 0.24943438172340393,
"learning_rate": 0.0009725519368690539,
"loss": 2.1352,
"step": 290
},
{
"epoch": 0.13347093223254214,
"grad_norm": 0.24098846316337585,
"learning_rate": 0.0009723086104219787,
"loss": 2.0969,
"step": 291
},
{
"epoch": 0.13392959522990483,
"grad_norm": 0.2530808448791504,
"learning_rate": 0.0009720642408998409,
"loss": 2.1211,
"step": 292
},
{
"epoch": 0.13438825822726752,
"grad_norm": 0.2504442632198334,
"learning_rate": 0.0009718188288423211,
"loss": 2.0802,
"step": 293
},
{
"epoch": 0.1348469212246302,
"grad_norm": 0.2571863830089569,
"learning_rate": 0.0009715723747914022,
"loss": 2.1312,
"step": 294
},
{
"epoch": 0.1353055842219929,
"grad_norm": 0.23387186229228973,
"learning_rate": 0.0009713248792913685,
"loss": 2.1297,
"step": 295
},
{
"epoch": 0.13576424721935557,
"grad_norm": 0.23552249372005463,
"learning_rate": 0.0009710763428888037,
"loss": 2.1042,
"step": 296
},
{
"epoch": 0.13622291021671826,
"grad_norm": 0.27026769518852234,
"learning_rate": 0.0009708267661325909,
"loss": 2.1776,
"step": 297
},
{
"epoch": 0.13668157321408095,
"grad_norm": 0.2319367378950119,
"learning_rate": 0.0009705761495739107,
"loss": 2.161,
"step": 298
},
{
"epoch": 0.13714023621144364,
"grad_norm": 0.23783843219280243,
"learning_rate": 0.0009703244937662399,
"loss": 2.1241,
"step": 299
},
{
"epoch": 0.13759889920880633,
"grad_norm": 0.25627851486206055,
"learning_rate": 0.0009700717992653505,
"loss": 2.0917,
"step": 300
},
{
"epoch": 0.13805756220616902,
"grad_norm": 0.26163187623023987,
"learning_rate": 0.0009698180666293083,
"loss": 2.0972,
"step": 301
},
{
"epoch": 0.1385162252035317,
"grad_norm": 0.27848559617996216,
"learning_rate": 0.000969563296418472,
"loss": 2.1444,
"step": 302
},
{
"epoch": 0.1389748882008944,
"grad_norm": 0.2437235713005066,
"learning_rate": 0.0009693074891954914,
"loss": 2.0545,
"step": 303
},
{
"epoch": 0.1394335511982571,
"grad_norm": 0.25904133915901184,
"learning_rate": 0.0009690506455253072,
"loss": 2.1238,
"step": 304
},
{
"epoch": 0.13989221419561976,
"grad_norm": 0.24843654036521912,
"learning_rate": 0.0009687927659751481,
"loss": 2.1794,
"step": 305
},
{
"epoch": 0.14035087719298245,
"grad_norm": 0.24186407029628754,
"learning_rate": 0.0009685338511145312,
"loss": 2.0537,
"step": 306
},
{
"epoch": 0.14080954019034514,
"grad_norm": 0.25381195545196533,
"learning_rate": 0.0009682739015152598,
"loss": 2.1656,
"step": 307
},
{
"epoch": 0.14126820318770783,
"grad_norm": 0.25703132152557373,
"learning_rate": 0.0009680129177514226,
"loss": 2.1537,
"step": 308
},
{
"epoch": 0.14172686618507052,
"grad_norm": 0.2708240747451782,
"learning_rate": 0.0009677509003993915,
"loss": 2.126,
"step": 309
},
{
"epoch": 0.1421855291824332,
"grad_norm": 0.22551526129245758,
"learning_rate": 0.0009674878500378221,
"loss": 2.1549,
"step": 310
},
{
"epoch": 0.1426441921797959,
"grad_norm": 0.24933461844921112,
"learning_rate": 0.0009672237672476505,
"loss": 2.0913,
"step": 311
},
{
"epoch": 0.1431028551771586,
"grad_norm": 0.24832989275455475,
"learning_rate": 0.0009669586526120935,
"loss": 2.1006,
"step": 312
},
{
"epoch": 0.14356151817452126,
"grad_norm": 0.24595899879932404,
"learning_rate": 0.0009666925067166459,
"loss": 2.1447,
"step": 313
},
{
"epoch": 0.14402018117188395,
"grad_norm": 0.23313504457473755,
"learning_rate": 0.000966425330149081,
"loss": 2.1342,
"step": 314
},
{
"epoch": 0.14447884416924664,
"grad_norm": 0.26827338337898254,
"learning_rate": 0.0009661571234994475,
"loss": 2.1017,
"step": 315
},
{
"epoch": 0.14493750716660933,
"grad_norm": 0.24809293448925018,
"learning_rate": 0.0009658878873600691,
"loss": 2.0794,
"step": 316
},
{
"epoch": 0.14539617016397202,
"grad_norm": 0.23397238552570343,
"learning_rate": 0.0009656176223255438,
"loss": 2.1157,
"step": 317
},
{
"epoch": 0.1458548331613347,
"grad_norm": 0.22913800179958344,
"learning_rate": 0.000965346328992741,
"loss": 2.1579,
"step": 318
},
{
"epoch": 0.1463134961586974,
"grad_norm": 0.24874641001224518,
"learning_rate": 0.0009650740079608014,
"loss": 2.1918,
"step": 319
},
{
"epoch": 0.1467721591560601,
"grad_norm": 0.25332990288734436,
"learning_rate": 0.0009648006598311353,
"loss": 2.1157,
"step": 320
},
{
"epoch": 0.14723082215342279,
"grad_norm": 0.23617063462734222,
"learning_rate": 0.0009645262852074214,
"loss": 2.0508,
"step": 321
},
{
"epoch": 0.14768948515078545,
"grad_norm": 0.23993897438049316,
"learning_rate": 0.0009642508846956053,
"loss": 2.066,
"step": 322
},
{
"epoch": 0.14814814814814814,
"grad_norm": 0.2703104019165039,
"learning_rate": 0.0009639744589038983,
"loss": 2.1127,
"step": 323
},
{
"epoch": 0.14860681114551083,
"grad_norm": 0.24462592601776123,
"learning_rate": 0.0009636970084427759,
"loss": 2.0981,
"step": 324
},
{
"epoch": 0.14906547414287352,
"grad_norm": 0.2991892993450165,
"learning_rate": 0.0009634185339249766,
"loss": 2.208,
"step": 325
},
{
"epoch": 0.1495241371402362,
"grad_norm": 0.23428498208522797,
"learning_rate": 0.0009631390359655003,
"loss": 2.1462,
"step": 326
},
{
"epoch": 0.1499828001375989,
"grad_norm": 0.2386583536863327,
"learning_rate": 0.0009628585151816074,
"loss": 2.1336,
"step": 327
},
{
"epoch": 0.1504414631349616,
"grad_norm": 0.23520754277706146,
"learning_rate": 0.0009625769721928172,
"loss": 2.1685,
"step": 328
},
{
"epoch": 0.15090012613232429,
"grad_norm": 0.25803980231285095,
"learning_rate": 0.0009622944076209061,
"loss": 2.0959,
"step": 329
},
{
"epoch": 0.15135878912968695,
"grad_norm": 0.2766725718975067,
"learning_rate": 0.0009620108220899071,
"loss": 2.1592,
"step": 330
},
{
"epoch": 0.15181745212704964,
"grad_norm": 0.27849626541137695,
"learning_rate": 0.0009617262162261075,
"loss": 2.1967,
"step": 331
},
{
"epoch": 0.15227611512441233,
"grad_norm": 0.26335251331329346,
"learning_rate": 0.0009614405906580486,
"loss": 2.182,
"step": 332
},
{
"epoch": 0.15273477812177502,
"grad_norm": 0.2548454999923706,
"learning_rate": 0.000961153946016523,
"loss": 2.154,
"step": 333
},
{
"epoch": 0.1531934411191377,
"grad_norm": 0.23631909489631653,
"learning_rate": 0.000960866282934574,
"loss": 2.1279,
"step": 334
},
{
"epoch": 0.1536521041165004,
"grad_norm": 0.2573084533214569,
"learning_rate": 0.0009605776020474945,
"loss": 2.1546,
"step": 335
},
{
"epoch": 0.1541107671138631,
"grad_norm": 0.23981761932373047,
"learning_rate": 0.0009602879039928249,
"loss": 2.0919,
"step": 336
},
{
"epoch": 0.1545694301112258,
"grad_norm": 0.2648386061191559,
"learning_rate": 0.0009599971894103521,
"loss": 2.1464,
"step": 337
},
{
"epoch": 0.15502809310858848,
"grad_norm": 0.24706383049488068,
"learning_rate": 0.0009597054589421077,
"loss": 2.0898,
"step": 338
},
{
"epoch": 0.15548675610595114,
"grad_norm": 0.2729613780975342,
"learning_rate": 0.0009594127132323669,
"loss": 2.1066,
"step": 339
},
{
"epoch": 0.15594541910331383,
"grad_norm": 0.3463296890258789,
"learning_rate": 0.0009591189529276474,
"loss": 2.1875,
"step": 340
},
{
"epoch": 0.15640408210067652,
"grad_norm": 0.25584179162979126,
"learning_rate": 0.0009588241786767072,
"loss": 2.1145,
"step": 341
},
{
"epoch": 0.1568627450980392,
"grad_norm": 0.23628297448158264,
"learning_rate": 0.0009585283911305436,
"loss": 2.1152,
"step": 342
},
{
"epoch": 0.1573214080954019,
"grad_norm": 0.2776651084423065,
"learning_rate": 0.000958231590942392,
"loss": 2.1469,
"step": 343
},
{
"epoch": 0.1577800710927646,
"grad_norm": 0.22856464982032776,
"learning_rate": 0.0009579337787677238,
"loss": 2.0317,
"step": 344
},
{
"epoch": 0.1582387340901273,
"grad_norm": 0.2774614989757538,
"learning_rate": 0.0009576349552642456,
"loss": 2.2277,
"step": 345
},
{
"epoch": 0.15869739708748998,
"grad_norm": 0.30310535430908203,
"learning_rate": 0.0009573351210918975,
"loss": 2.1362,
"step": 346
},
{
"epoch": 0.15915606008485264,
"grad_norm": 0.25364595651626587,
"learning_rate": 0.0009570342769128514,
"loss": 2.2076,
"step": 347
},
{
"epoch": 0.15961472308221533,
"grad_norm": 0.2707476019859314,
"learning_rate": 0.0009567324233915099,
"loss": 2.1282,
"step": 348
},
{
"epoch": 0.16007338607957802,
"grad_norm": 0.2858078181743622,
"learning_rate": 0.0009564295611945047,
"loss": 2.1794,
"step": 349
},
{
"epoch": 0.1605320490769407,
"grad_norm": 0.23799866437911987,
"learning_rate": 0.000956125690990695,
"loss": 2.0976,
"step": 350
},
{
"epoch": 0.1609907120743034,
"grad_norm": 0.24325276911258698,
"learning_rate": 0.0009558208134511665,
"loss": 2.0986,
"step": 351
},
{
"epoch": 0.1614493750716661,
"grad_norm": 0.2673221826553345,
"learning_rate": 0.0009555149292492289,
"loss": 2.1484,
"step": 352
},
{
"epoch": 0.1619080380690288,
"grad_norm": 0.24535591900348663,
"learning_rate": 0.0009552080390604159,
"loss": 2.1511,
"step": 353
},
{
"epoch": 0.16236670106639148,
"grad_norm": 0.27296754717826843,
"learning_rate": 0.0009549001435624823,
"loss": 2.1927,
"step": 354
},
{
"epoch": 0.16282536406375417,
"grad_norm": 0.23733842372894287,
"learning_rate": 0.0009545912434354029,
"loss": 2.155,
"step": 355
},
{
"epoch": 0.16328402706111683,
"grad_norm": 0.23931311070919037,
"learning_rate": 0.0009542813393613721,
"loss": 2.1225,
"step": 356
},
{
"epoch": 0.16374269005847952,
"grad_norm": 0.24464167654514313,
"learning_rate": 0.0009539704320248006,
"loss": 2.0343,
"step": 357
},
{
"epoch": 0.16420135305584221,
"grad_norm": 0.2524767816066742,
"learning_rate": 0.0009536585221123151,
"loss": 2.1507,
"step": 358
},
{
"epoch": 0.1646600160532049,
"grad_norm": 0.27435609698295593,
"learning_rate": 0.0009533456103127565,
"loss": 2.1324,
"step": 359
},
{
"epoch": 0.1651186790505676,
"grad_norm": 0.22745080292224884,
"learning_rate": 0.000953031697317178,
"loss": 2.1831,
"step": 360
},
{
"epoch": 0.1655773420479303,
"grad_norm": 0.2494613379240036,
"learning_rate": 0.0009527167838188445,
"loss": 2.0743,
"step": 361
},
{
"epoch": 0.16603600504529298,
"grad_norm": 0.24312426149845123,
"learning_rate": 0.0009524008705132299,
"loss": 2.1174,
"step": 362
},
{
"epoch": 0.16649466804265567,
"grad_norm": 0.28243881464004517,
"learning_rate": 0.0009520839580980166,
"loss": 2.1219,
"step": 363
},
{
"epoch": 0.16695333104001836,
"grad_norm": 0.2164716124534607,
"learning_rate": 0.0009517660472730929,
"loss": 2.0759,
"step": 364
},
{
"epoch": 0.16741199403738102,
"grad_norm": 0.23837263882160187,
"learning_rate": 0.0009514471387405526,
"loss": 2.1248,
"step": 365
},
{
"epoch": 0.16787065703474371,
"grad_norm": 0.25199204683303833,
"learning_rate": 0.0009511272332046926,
"loss": 2.1362,
"step": 366
},
{
"epoch": 0.1683293200321064,
"grad_norm": 0.24965564906597137,
"learning_rate": 0.0009508063313720119,
"loss": 2.1678,
"step": 367
},
{
"epoch": 0.1687879830294691,
"grad_norm": 0.21947196125984192,
"learning_rate": 0.0009504844339512095,
"loss": 2.0702,
"step": 368
},
{
"epoch": 0.1692466460268318,
"grad_norm": 0.2504624128341675,
"learning_rate": 0.0009501615416531835,
"loss": 2.1929,
"step": 369
},
{
"epoch": 0.16970530902419448,
"grad_norm": 0.2678537368774414,
"learning_rate": 0.0009498376551910285,
"loss": 2.106,
"step": 370
},
{
"epoch": 0.17016397202155717,
"grad_norm": 0.25699949264526367,
"learning_rate": 0.0009495127752800352,
"loss": 2.0969,
"step": 371
},
{
"epoch": 0.17062263501891986,
"grad_norm": 0.23679561913013458,
"learning_rate": 0.0009491869026376882,
"loss": 2.0452,
"step": 372
},
{
"epoch": 0.17108129801628252,
"grad_norm": 0.27194786071777344,
"learning_rate": 0.0009488600379836648,
"loss": 2.1067,
"step": 373
},
{
"epoch": 0.17153996101364521,
"grad_norm": 0.26040732860565186,
"learning_rate": 0.0009485321820398321,
"loss": 2.1422,
"step": 374
},
{
"epoch": 0.1719986240110079,
"grad_norm": 0.2406633347272873,
"learning_rate": 0.0009482033355302475,
"loss": 2.255,
"step": 375
},
{
"epoch": 0.1724572870083706,
"grad_norm": 0.2553479075431824,
"learning_rate": 0.0009478734991811556,
"loss": 2.0883,
"step": 376
},
{
"epoch": 0.1729159500057333,
"grad_norm": 0.2382071614265442,
"learning_rate": 0.0009475426737209871,
"loss": 2.0833,
"step": 377
},
{
"epoch": 0.17337461300309598,
"grad_norm": 0.2471284419298172,
"learning_rate": 0.000947210859880357,
"loss": 2.1317,
"step": 378
},
{
"epoch": 0.17383327600045867,
"grad_norm": 0.24191664159297943,
"learning_rate": 0.0009468780583920631,
"loss": 2.1642,
"step": 379
},
{
"epoch": 0.17429193899782136,
"grad_norm": 0.24993795156478882,
"learning_rate": 0.0009465442699910846,
"loss": 2.0767,
"step": 380
},
{
"epoch": 0.17475060199518405,
"grad_norm": 0.2530753016471863,
"learning_rate": 0.0009462094954145801,
"loss": 2.0636,
"step": 381
},
{
"epoch": 0.17520926499254671,
"grad_norm": 0.24088294804096222,
"learning_rate": 0.0009458737354018859,
"loss": 2.1061,
"step": 382
},
{
"epoch": 0.1756679279899094,
"grad_norm": 0.2366131693124771,
"learning_rate": 0.000945536990694515,
"loss": 2.1886,
"step": 383
},
{
"epoch": 0.1761265909872721,
"grad_norm": 0.25113001465797424,
"learning_rate": 0.0009451992620361551,
"loss": 2.1663,
"step": 384
},
{
"epoch": 0.1765852539846348,
"grad_norm": 0.2510436177253723,
"learning_rate": 0.0009448605501726664,
"loss": 2.1391,
"step": 385
},
{
"epoch": 0.17704391698199748,
"grad_norm": 0.24192020297050476,
"learning_rate": 0.000944520855852081,
"loss": 2.0859,
"step": 386
},
{
"epoch": 0.17750257997936017,
"grad_norm": 0.23359414935112,
"learning_rate": 0.0009441801798246002,
"loss": 2.0095,
"step": 387
},
{
"epoch": 0.17796124297672286,
"grad_norm": 0.25025713443756104,
"learning_rate": 0.0009438385228425939,
"loss": 2.0936,
"step": 388
},
{
"epoch": 0.17841990597408555,
"grad_norm": 0.24478502571582794,
"learning_rate": 0.0009434958856605982,
"loss": 2.1557,
"step": 389
},
{
"epoch": 0.17887856897144822,
"grad_norm": 0.24249394237995148,
"learning_rate": 0.0009431522690353137,
"loss": 2.1124,
"step": 390
},
{
"epoch": 0.1793372319688109,
"grad_norm": 0.2260642796754837,
"learning_rate": 0.0009428076737256044,
"loss": 2.0645,
"step": 391
},
{
"epoch": 0.1797958949661736,
"grad_norm": 0.24059531092643738,
"learning_rate": 0.0009424621004924954,
"loss": 2.1615,
"step": 392
},
{
"epoch": 0.1802545579635363,
"grad_norm": 0.2426529973745346,
"learning_rate": 0.0009421155500991719,
"loss": 2.2244,
"step": 393
},
{
"epoch": 0.18071322096089898,
"grad_norm": 0.22897152602672577,
"learning_rate": 0.0009417680233109767,
"loss": 2.0531,
"step": 394
},
{
"epoch": 0.18117188395826167,
"grad_norm": 0.2291453331708908,
"learning_rate": 0.000941419520895409,
"loss": 2.1634,
"step": 395
},
{
"epoch": 0.18163054695562436,
"grad_norm": 0.24115078151226044,
"learning_rate": 0.0009410700436221229,
"loss": 2.1165,
"step": 396
},
{
"epoch": 0.18208920995298705,
"grad_norm": 0.2529621124267578,
"learning_rate": 0.0009407195922629252,
"loss": 2.061,
"step": 397
},
{
"epoch": 0.18254787295034974,
"grad_norm": 0.23723754286766052,
"learning_rate": 0.000940368167591774,
"loss": 2.166,
"step": 398
},
{
"epoch": 0.1830065359477124,
"grad_norm": 0.2493610680103302,
"learning_rate": 0.0009400157703847769,
"loss": 2.0952,
"step": 399
},
{
"epoch": 0.1834651989450751,
"grad_norm": 0.23279500007629395,
"learning_rate": 0.0009396624014201895,
"loss": 2.1264,
"step": 400
},
{
"epoch": 0.1839238619424378,
"grad_norm": 0.24473583698272705,
"learning_rate": 0.000939308061478413,
"loss": 2.2016,
"step": 401
},
{
"epoch": 0.18438252493980048,
"grad_norm": 0.22492888569831848,
"learning_rate": 0.0009389527513419935,
"loss": 2.0905,
"step": 402
},
{
"epoch": 0.18484118793716317,
"grad_norm": 0.22453978657722473,
"learning_rate": 0.0009385964717956195,
"loss": 2.0741,
"step": 403
},
{
"epoch": 0.18529985093452586,
"grad_norm": 0.24258217215538025,
"learning_rate": 0.0009382392236261201,
"loss": 2.1047,
"step": 404
},
{
"epoch": 0.18575851393188855,
"grad_norm": 0.25706642866134644,
"learning_rate": 0.0009378810076224644,
"loss": 2.1708,
"step": 405
},
{
"epoch": 0.18621717692925124,
"grad_norm": 0.23550614714622498,
"learning_rate": 0.0009375218245757582,
"loss": 2.1027,
"step": 406
},
{
"epoch": 0.1866758399266139,
"grad_norm": 0.23844020068645477,
"learning_rate": 0.0009371616752792432,
"loss": 2.076,
"step": 407
},
{
"epoch": 0.1871345029239766,
"grad_norm": 0.23946405947208405,
"learning_rate": 0.0009368005605282949,
"loss": 2.1885,
"step": 408
},
{
"epoch": 0.1875931659213393,
"grad_norm": 0.23271027207374573,
"learning_rate": 0.0009364384811204212,
"loss": 2.161,
"step": 409
},
{
"epoch": 0.18805182891870198,
"grad_norm": 0.22965985536575317,
"learning_rate": 0.00093607543785526,
"loss": 2.0796,
"step": 410
},
{
"epoch": 0.18851049191606467,
"grad_norm": 0.22721543908119202,
"learning_rate": 0.0009357114315345787,
"loss": 2.1007,
"step": 411
},
{
"epoch": 0.18896915491342736,
"grad_norm": 0.22923073172569275,
"learning_rate": 0.0009353464629622705,
"loss": 2.0757,
"step": 412
},
{
"epoch": 0.18942781791079005,
"grad_norm": 0.22309958934783936,
"learning_rate": 0.0009349805329443544,
"loss": 2.1459,
"step": 413
},
{
"epoch": 0.18988648090815274,
"grad_norm": 0.2316034585237503,
"learning_rate": 0.0009346136422889724,
"loss": 2.121,
"step": 414
},
{
"epoch": 0.19034514390551543,
"grad_norm": 0.23811469972133636,
"learning_rate": 0.0009342457918063882,
"loss": 2.1019,
"step": 415
},
{
"epoch": 0.1908038069028781,
"grad_norm": 0.23671261966228485,
"learning_rate": 0.0009338769823089853,
"loss": 2.1281,
"step": 416
},
{
"epoch": 0.1912624699002408,
"grad_norm": 0.2332172691822052,
"learning_rate": 0.0009335072146112648,
"loss": 2.257,
"step": 417
},
{
"epoch": 0.19172113289760348,
"grad_norm": 0.24039915204048157,
"learning_rate": 0.0009331364895298444,
"loss": 2.0223,
"step": 418
},
{
"epoch": 0.19217979589496617,
"grad_norm": 0.2453923374414444,
"learning_rate": 0.0009327648078834559,
"loss": 2.1575,
"step": 419
},
{
"epoch": 0.19263845889232886,
"grad_norm": 0.2554095685482025,
"learning_rate": 0.0009323921704929434,
"loss": 2.1138,
"step": 420
},
{
"epoch": 0.19309712188969155,
"grad_norm": 0.22456218302249908,
"learning_rate": 0.0009320185781812623,
"loss": 2.0048,
"step": 421
},
{
"epoch": 0.19355578488705424,
"grad_norm": 0.22925415635108948,
"learning_rate": 0.0009316440317734762,
"loss": 2.0441,
"step": 422
},
{
"epoch": 0.19401444788441694,
"grad_norm": 0.24674464762210846,
"learning_rate": 0.0009312685320967565,
"loss": 2.1805,
"step": 423
},
{
"epoch": 0.19447311088177963,
"grad_norm": 0.23233360052108765,
"learning_rate": 0.0009308920799803793,
"loss": 2.0655,
"step": 424
},
{
"epoch": 0.1949317738791423,
"grad_norm": 0.2464887648820877,
"learning_rate": 0.0009305146762557246,
"loss": 2.1243,
"step": 425
},
{
"epoch": 0.19539043687650498,
"grad_norm": 0.23387449979782104,
"learning_rate": 0.0009301363217562736,
"loss": 2.0985,
"step": 426
},
{
"epoch": 0.19584909987386767,
"grad_norm": 0.22677962481975555,
"learning_rate": 0.0009297570173176074,
"loss": 2.1036,
"step": 427
},
{
"epoch": 0.19630776287123036,
"grad_norm": 0.2232978492975235,
"learning_rate": 0.000929376763777405,
"loss": 2.1127,
"step": 428
},
{
"epoch": 0.19676642586859305,
"grad_norm": 0.23811952769756317,
"learning_rate": 0.0009289955619754413,
"loss": 2.1163,
"step": 429
},
{
"epoch": 0.19722508886595574,
"grad_norm": 0.20475523173809052,
"learning_rate": 0.0009286134127535859,
"loss": 2.139,
"step": 430
},
{
"epoch": 0.19768375186331844,
"grad_norm": 0.2348056435585022,
"learning_rate": 0.0009282303169558,
"loss": 2.1564,
"step": 431
},
{
"epoch": 0.19814241486068113,
"grad_norm": 0.2444457858800888,
"learning_rate": 0.0009278462754281359,
"loss": 2.108,
"step": 432
},
{
"epoch": 0.1986010778580438,
"grad_norm": 0.22553791105747223,
"learning_rate": 0.0009274612890187342,
"loss": 2.1375,
"step": 433
},
{
"epoch": 0.19905974085540648,
"grad_norm": 0.22373148798942566,
"learning_rate": 0.0009270753585778222,
"loss": 2.0714,
"step": 434
},
{
"epoch": 0.19951840385276917,
"grad_norm": 0.22777768969535828,
"learning_rate": 0.0009266884849577124,
"loss": 2.1047,
"step": 435
},
{
"epoch": 0.19997706685013186,
"grad_norm": 0.21112293004989624,
"learning_rate": 0.0009263006690127998,
"loss": 2.07,
"step": 436
},
{
"epoch": 0.20043572984749455,
"grad_norm": 0.2296253889799118,
"learning_rate": 0.0009259119115995609,
"loss": 2.0845,
"step": 437
},
{
"epoch": 0.20089439284485724,
"grad_norm": 0.22975893318653107,
"learning_rate": 0.0009255222135765511,
"loss": 2.0706,
"step": 438
},
{
"epoch": 0.20135305584221994,
"grad_norm": 0.24168571829795837,
"learning_rate": 0.0009251315758044032,
"loss": 2.1437,
"step": 439
},
{
"epoch": 0.20181171883958263,
"grad_norm": 0.23768769204616547,
"learning_rate": 0.0009247399991458255,
"loss": 2.0593,
"step": 440
},
{
"epoch": 0.20227038183694532,
"grad_norm": 0.22561083734035492,
"learning_rate": 0.0009243474844655996,
"loss": 2.1364,
"step": 441
},
{
"epoch": 0.20272904483430798,
"grad_norm": 0.22765830159187317,
"learning_rate": 0.0009239540326305791,
"loss": 2.0757,
"step": 442
},
{
"epoch": 0.20318770783167067,
"grad_norm": 0.2313157469034195,
"learning_rate": 0.0009235596445096864,
"loss": 2.1435,
"step": 443
},
{
"epoch": 0.20364637082903336,
"grad_norm": 0.23196153342723846,
"learning_rate": 0.0009231643209739127,
"loss": 2.0986,
"step": 444
},
{
"epoch": 0.20410503382639605,
"grad_norm": 0.2262774109840393,
"learning_rate": 0.0009227680628963145,
"loss": 2.0985,
"step": 445
},
{
"epoch": 0.20456369682375874,
"grad_norm": 0.25701904296875,
"learning_rate": 0.000922370871152012,
"loss": 2.1302,
"step": 446
},
{
"epoch": 0.20502235982112144,
"grad_norm": 0.23705513775348663,
"learning_rate": 0.0009219727466181877,
"loss": 2.149,
"step": 447
},
{
"epoch": 0.20548102281848413,
"grad_norm": 0.23032982647418976,
"learning_rate": 0.0009215736901740841,
"loss": 2.1079,
"step": 448
},
{
"epoch": 0.20593968581584682,
"grad_norm": 0.23615211248397827,
"learning_rate": 0.0009211737027010016,
"loss": 2.0169,
"step": 449
},
{
"epoch": 0.20639834881320948,
"grad_norm": 0.22572138905525208,
"learning_rate": 0.0009207727850822971,
"loss": 1.9792,
"step": 450
},
{
"epoch": 0.20685701181057217,
"grad_norm": 0.2595033645629883,
"learning_rate": 0.0009203709382033814,
"loss": 2.1451,
"step": 451
},
{
"epoch": 0.20731567480793486,
"grad_norm": 0.22811029851436615,
"learning_rate": 0.0009199681629517173,
"loss": 2.1257,
"step": 452
},
{
"epoch": 0.20777433780529755,
"grad_norm": 0.24860930442810059,
"learning_rate": 0.0009195644602168184,
"loss": 2.0759,
"step": 453
},
{
"epoch": 0.20823300080266025,
"grad_norm": 0.23495998978614807,
"learning_rate": 0.0009191598308902464,
"loss": 2.1248,
"step": 454
},
{
"epoch": 0.20869166380002294,
"grad_norm": 0.2244628667831421,
"learning_rate": 0.0009187542758656091,
"loss": 2.0335,
"step": 455
},
{
"epoch": 0.20915032679738563,
"grad_norm": 0.22525955736637115,
"learning_rate": 0.0009183477960385591,
"loss": 2.0326,
"step": 456
},
{
"epoch": 0.20960898979474832,
"grad_norm": 0.22283212840557098,
"learning_rate": 0.0009179403923067912,
"loss": 2.1028,
"step": 457
},
{
"epoch": 0.210067652792111,
"grad_norm": 0.23231834173202515,
"learning_rate": 0.0009175320655700406,
"loss": 2.1249,
"step": 458
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.22248585522174835,
"learning_rate": 0.0009171228167300805,
"loss": 2.1107,
"step": 459
},
{
"epoch": 0.21098497878683636,
"grad_norm": 0.24062450230121613,
"learning_rate": 0.0009167126466907215,
"loss": 2.047,
"step": 460
},
{
"epoch": 0.21144364178419905,
"grad_norm": 0.2246720790863037,
"learning_rate": 0.0009163015563578074,
"loss": 2.2001,
"step": 461
},
{
"epoch": 0.21190230478156175,
"grad_norm": 0.22428402304649353,
"learning_rate": 0.0009158895466392158,
"loss": 2.0029,
"step": 462
},
{
"epoch": 0.21236096777892444,
"grad_norm": 0.2275976687669754,
"learning_rate": 0.0009154766184448535,
"loss": 2.0648,
"step": 463
},
{
"epoch": 0.21281963077628713,
"grad_norm": 0.22484666109085083,
"learning_rate": 0.0009150627726866568,
"loss": 2.085,
"step": 464
},
{
"epoch": 0.21327829377364982,
"grad_norm": 0.22627927362918854,
"learning_rate": 0.000914648010278587,
"loss": 2.1501,
"step": 465
},
{
"epoch": 0.2137369567710125,
"grad_norm": 0.21148554980754852,
"learning_rate": 0.0009142323321366315,
"loss": 2.0994,
"step": 466
},
{
"epoch": 0.21419561976837517,
"grad_norm": 0.22367995977401733,
"learning_rate": 0.0009138157391787986,
"loss": 2.0991,
"step": 467
},
{
"epoch": 0.21465428276573786,
"grad_norm": 0.20094799995422363,
"learning_rate": 0.0009133982323251177,
"loss": 1.9903,
"step": 468
},
{
"epoch": 0.21511294576310055,
"grad_norm": 0.2096623033285141,
"learning_rate": 0.0009129798124976365,
"loss": 2.1862,
"step": 469
},
{
"epoch": 0.21557160876046325,
"grad_norm": 0.21262474358081818,
"learning_rate": 0.0009125604806204187,
"loss": 2.1134,
"step": 470
},
{
"epoch": 0.21603027175782594,
"grad_norm": 0.2981449365615845,
"learning_rate": 0.0009121402376195421,
"loss": 2.0478,
"step": 471
},
{
"epoch": 0.21648893475518863,
"grad_norm": 0.2271588295698166,
"learning_rate": 0.0009117190844230972,
"loss": 2.0298,
"step": 472
},
{
"epoch": 0.21694759775255132,
"grad_norm": 0.22686482965946198,
"learning_rate": 0.0009112970219611841,
"loss": 2.0913,
"step": 473
},
{
"epoch": 0.217406260749914,
"grad_norm": 0.24190589785575867,
"learning_rate": 0.0009108740511659115,
"loss": 2.1864,
"step": 474
},
{
"epoch": 0.2178649237472767,
"grad_norm": 0.22152303159236908,
"learning_rate": 0.0009104501729713935,
"loss": 2.0505,
"step": 475
},
{
"epoch": 0.21832358674463936,
"grad_norm": 0.22563710808753967,
"learning_rate": 0.0009100253883137488,
"loss": 2.1332,
"step": 476
},
{
"epoch": 0.21878224974200206,
"grad_norm": 0.2213224470615387,
"learning_rate": 0.0009095996981310974,
"loss": 2.0918,
"step": 477
},
{
"epoch": 0.21924091273936475,
"grad_norm": 0.21699318289756775,
"learning_rate": 0.0009091731033635596,
"loss": 2.1532,
"step": 478
},
{
"epoch": 0.21969957573672744,
"grad_norm": 0.22556233406066895,
"learning_rate": 0.0009087456049532529,
"loss": 2.0665,
"step": 479
},
{
"epoch": 0.22015823873409013,
"grad_norm": 0.21974864602088928,
"learning_rate": 0.0009083172038442914,
"loss": 2.0587,
"step": 480
},
{
"epoch": 0.22061690173145282,
"grad_norm": 0.21498467028141022,
"learning_rate": 0.0009078879009827817,
"loss": 2.0219,
"step": 481
},
{
"epoch": 0.2210755647288155,
"grad_norm": 0.2730620503425598,
"learning_rate": 0.0009074576973168223,
"loss": 2.0603,
"step": 482
},
{
"epoch": 0.2215342277261782,
"grad_norm": 0.24298104643821716,
"learning_rate": 0.0009070265937965015,
"loss": 2.1655,
"step": 483
},
{
"epoch": 0.2219928907235409,
"grad_norm": 0.21339309215545654,
"learning_rate": 0.0009065945913738942,
"loss": 2.0456,
"step": 484
},
{
"epoch": 0.22245155372090356,
"grad_norm": 0.2376120537519455,
"learning_rate": 0.0009061616910030609,
"loss": 2.074,
"step": 485
},
{
"epoch": 0.22291021671826625,
"grad_norm": 0.23291656374931335,
"learning_rate": 0.0009057278936400453,
"loss": 2.0295,
"step": 486
},
{
"epoch": 0.22336887971562894,
"grad_norm": 0.22138389945030212,
"learning_rate": 0.0009052932002428715,
"loss": 2.1996,
"step": 487
},
{
"epoch": 0.22382754271299163,
"grad_norm": 0.21928265690803528,
"learning_rate": 0.0009048576117715435,
"loss": 2.1105,
"step": 488
},
{
"epoch": 0.22428620571035432,
"grad_norm": 0.234505295753479,
"learning_rate": 0.0009044211291880407,
"loss": 2.0351,
"step": 489
},
{
"epoch": 0.224744868707717,
"grad_norm": 0.22377289831638336,
"learning_rate": 0.000903983753456318,
"loss": 2.0563,
"step": 490
},
{
"epoch": 0.2252035317050797,
"grad_norm": 0.2328329235315323,
"learning_rate": 0.0009035454855423026,
"loss": 2.093,
"step": 491
},
{
"epoch": 0.2256621947024424,
"grad_norm": 0.22616635262966156,
"learning_rate": 0.0009031063264138922,
"loss": 2.1251,
"step": 492
},
{
"epoch": 0.22612085769980506,
"grad_norm": 0.21427574753761292,
"learning_rate": 0.0009026662770409522,
"loss": 2.0714,
"step": 493
},
{
"epoch": 0.22657952069716775,
"grad_norm": 0.23495055735111237,
"learning_rate": 0.0009022253383953147,
"loss": 2.2255,
"step": 494
},
{
"epoch": 0.22703818369453044,
"grad_norm": 0.22905318439006805,
"learning_rate": 0.0009017835114507753,
"loss": 2.1183,
"step": 495
},
{
"epoch": 0.22749684669189313,
"grad_norm": 0.2222667783498764,
"learning_rate": 0.0009013407971830914,
"loss": 2.0339,
"step": 496
},
{
"epoch": 0.22795550968925582,
"grad_norm": 0.217635378241539,
"learning_rate": 0.0009008971965699801,
"loss": 2.0708,
"step": 497
},
{
"epoch": 0.2284141726866185,
"grad_norm": 0.2272225171327591,
"learning_rate": 0.0009004527105911163,
"loss": 2.0752,
"step": 498
},
{
"epoch": 0.2288728356839812,
"grad_norm": 0.23241865634918213,
"learning_rate": 0.0009000073402281295,
"loss": 2.0275,
"step": 499
},
{
"epoch": 0.2293314986813439,
"grad_norm": 0.23616348206996918,
"learning_rate": 0.0008995610864646028,
"loss": 2.1007,
"step": 500
},
{
"epoch": 0.22979016167870658,
"grad_norm": 0.21653427183628082,
"learning_rate": 0.0008991139502860703,
"loss": 2.0316,
"step": 501
},
{
"epoch": 0.23024882467606925,
"grad_norm": 0.2365088164806366,
"learning_rate": 0.0008986659326800146,
"loss": 2.0586,
"step": 502
},
{
"epoch": 0.23070748767343194,
"grad_norm": 0.21267177164554596,
"learning_rate": 0.0008982170346358651,
"loss": 2.0744,
"step": 503
},
{
"epoch": 0.23116615067079463,
"grad_norm": 0.23009279370307922,
"learning_rate": 0.0008977672571449956,
"loss": 2.0488,
"step": 504
},
{
"epoch": 0.23162481366815732,
"grad_norm": 0.21103161573410034,
"learning_rate": 0.0008973166012007217,
"loss": 2.0799,
"step": 505
},
{
"epoch": 0.23208347666552,
"grad_norm": 0.24619385600090027,
"learning_rate": 0.0008968650677982998,
"loss": 2.0945,
"step": 506
},
{
"epoch": 0.2325421396628827,
"grad_norm": 0.240698903799057,
"learning_rate": 0.0008964126579349236,
"loss": 2.022,
"step": 507
},
{
"epoch": 0.2330008026602454,
"grad_norm": 0.22663864493370056,
"learning_rate": 0.0008959593726097226,
"loss": 2.1097,
"step": 508
},
{
"epoch": 0.23345946565760808,
"grad_norm": 0.22942039370536804,
"learning_rate": 0.0008955052128237596,
"loss": 2.0979,
"step": 509
},
{
"epoch": 0.23391812865497075,
"grad_norm": 0.2181216925382614,
"learning_rate": 0.0008950501795800288,
"loss": 2.0845,
"step": 510
},
{
"epoch": 0.23437679165233344,
"grad_norm": 0.22632715106010437,
"learning_rate": 0.0008945942738834532,
"loss": 2.058,
"step": 511
},
{
"epoch": 0.23483545464969613,
"grad_norm": 0.22106988728046417,
"learning_rate": 0.0008941374967408826,
"loss": 2.0904,
"step": 512
},
{
"epoch": 0.23529411764705882,
"grad_norm": 0.21880120038986206,
"learning_rate": 0.0008936798491610916,
"loss": 2.0405,
"step": 513
},
{
"epoch": 0.2357527806444215,
"grad_norm": 0.22676599025726318,
"learning_rate": 0.0008932213321547768,
"loss": 2.1331,
"step": 514
},
{
"epoch": 0.2362114436417842,
"grad_norm": 0.22815179824829102,
"learning_rate": 0.0008927619467345554,
"loss": 2.1285,
"step": 515
},
{
"epoch": 0.2366701066391469,
"grad_norm": 0.22736455500125885,
"learning_rate": 0.0008923016939149615,
"loss": 1.973,
"step": 516
},
{
"epoch": 0.23712876963650958,
"grad_norm": 0.23418518900871277,
"learning_rate": 0.0008918405747124458,
"loss": 2.09,
"step": 517
},
{
"epoch": 0.23758743263387228,
"grad_norm": 0.23176515102386475,
"learning_rate": 0.0008913785901453721,
"loss": 2.0878,
"step": 518
},
{
"epoch": 0.23804609563123494,
"grad_norm": 0.23154568672180176,
"learning_rate": 0.000890915741234015,
"loss": 2.1737,
"step": 519
},
{
"epoch": 0.23850475862859763,
"grad_norm": 0.2533906400203705,
"learning_rate": 0.0008904520290005582,
"loss": 2.1619,
"step": 520
},
{
"epoch": 0.23896342162596032,
"grad_norm": 0.22912345826625824,
"learning_rate": 0.000889987454469092,
"loss": 2.0702,
"step": 521
},
{
"epoch": 0.239422084623323,
"grad_norm": 0.2211257666349411,
"learning_rate": 0.0008895220186656111,
"loss": 2.1308,
"step": 522
},
{
"epoch": 0.2398807476206857,
"grad_norm": 0.23078425228595734,
"learning_rate": 0.0008890557226180122,
"loss": 2.1307,
"step": 523
},
{
"epoch": 0.2403394106180484,
"grad_norm": 0.22079190611839294,
"learning_rate": 0.0008885885673560921,
"loss": 1.9881,
"step": 524
},
{
"epoch": 0.24079807361541108,
"grad_norm": 0.22202295064926147,
"learning_rate": 0.0008881205539115444,
"loss": 2.1296,
"step": 525
},
{
"epoch": 0.24125673661277378,
"grad_norm": 0.23614433407783508,
"learning_rate": 0.0008876516833179589,
"loss": 2.0957,
"step": 526
},
{
"epoch": 0.24171539961013644,
"grad_norm": 0.24915914237499237,
"learning_rate": 0.0008871819566108177,
"loss": 2.0233,
"step": 527
},
{
"epoch": 0.24217406260749913,
"grad_norm": 0.22891834378242493,
"learning_rate": 0.000886711374827494,
"loss": 2.1056,
"step": 528
},
{
"epoch": 0.24263272560486182,
"grad_norm": 0.22281162440776825,
"learning_rate": 0.0008862399390072491,
"loss": 2.1315,
"step": 529
},
{
"epoch": 0.2430913886022245,
"grad_norm": 0.2114548683166504,
"learning_rate": 0.0008857676501912305,
"loss": 2.0295,
"step": 530
},
{
"epoch": 0.2435500515995872,
"grad_norm": 0.2145853042602539,
"learning_rate": 0.0008852945094224697,
"loss": 2.072,
"step": 531
},
{
"epoch": 0.2440087145969499,
"grad_norm": 0.21046780049800873,
"learning_rate": 0.0008848205177458795,
"loss": 2.095,
"step": 532
},
{
"epoch": 0.24446737759431258,
"grad_norm": 0.22981642186641693,
"learning_rate": 0.0008843456762082518,
"loss": 2.0433,
"step": 533
},
{
"epoch": 0.24492604059167528,
"grad_norm": 0.2175978571176529,
"learning_rate": 0.0008838699858582557,
"loss": 2.0436,
"step": 534
},
{
"epoch": 0.24538470358903797,
"grad_norm": 0.21275998651981354,
"learning_rate": 0.0008833934477464347,
"loss": 2.1463,
"step": 535
},
{
"epoch": 0.24584336658640063,
"grad_norm": 0.21782918274402618,
"learning_rate": 0.0008829160629252045,
"loss": 2.0675,
"step": 536
},
{
"epoch": 0.24630202958376332,
"grad_norm": 0.2270422726869583,
"learning_rate": 0.0008824378324488509,
"loss": 2.0797,
"step": 537
},
{
"epoch": 0.246760692581126,
"grad_norm": 0.24746927618980408,
"learning_rate": 0.0008819587573735268,
"loss": 2.1096,
"step": 538
},
{
"epoch": 0.2472193555784887,
"grad_norm": 0.22298315167427063,
"learning_rate": 0.0008814788387572513,
"loss": 2.1096,
"step": 539
},
{
"epoch": 0.2476780185758514,
"grad_norm": 0.21583664417266846,
"learning_rate": 0.0008809980776599053,
"loss": 2.0956,
"step": 540
},
{
"epoch": 0.24813668157321409,
"grad_norm": 0.2309054583311081,
"learning_rate": 0.0008805164751432312,
"loss": 2.0649,
"step": 541
},
{
"epoch": 0.24859534457057678,
"grad_norm": 0.22494859993457794,
"learning_rate": 0.0008800340322708292,
"loss": 2.0597,
"step": 542
},
{
"epoch": 0.24905400756793947,
"grad_norm": 0.22071048617362976,
"learning_rate": 0.0008795507501081555,
"loss": 2.0309,
"step": 543
},
{
"epoch": 0.24951267056530213,
"grad_norm": 0.2368357926607132,
"learning_rate": 0.0008790666297225196,
"loss": 2.0637,
"step": 544
},
{
"epoch": 0.24997133356266482,
"grad_norm": 0.2137153446674347,
"learning_rate": 0.0008785816721830829,
"loss": 2.001,
"step": 545
},
{
"epoch": 0.2504299965600275,
"grad_norm": 0.22778210043907166,
"learning_rate": 0.0008780958785608546,
"loss": 2.0446,
"step": 546
},
{
"epoch": 0.2508886595573902,
"grad_norm": 0.24633993208408356,
"learning_rate": 0.0008776092499286912,
"loss": 2.0725,
"step": 547
},
{
"epoch": 0.2513473225547529,
"grad_norm": 0.21789027750492096,
"learning_rate": 0.0008771217873612929,
"loss": 1.9756,
"step": 548
},
{
"epoch": 0.2518059855521156,
"grad_norm": 0.22623218595981598,
"learning_rate": 0.0008766334919352017,
"loss": 2.1072,
"step": 549
},
{
"epoch": 0.2522646485494783,
"grad_norm": 0.21800540387630463,
"learning_rate": 0.0008761443647287987,
"loss": 2.131,
"step": 550
},
{
"epoch": 0.25272331154684097,
"grad_norm": 0.2291291356086731,
"learning_rate": 0.0008756544068223026,
"loss": 2.1606,
"step": 551
},
{
"epoch": 0.25318197454420366,
"grad_norm": 0.21903391182422638,
"learning_rate": 0.0008751636192977659,
"loss": 2.1539,
"step": 552
},
{
"epoch": 0.25364063754156635,
"grad_norm": 0.2146795392036438,
"learning_rate": 0.0008746720032390737,
"loss": 2.0298,
"step": 553
},
{
"epoch": 0.25409930053892904,
"grad_norm": 0.20781736075878143,
"learning_rate": 0.0008741795597319408,
"loss": 2.0222,
"step": 554
},
{
"epoch": 0.25455796353629173,
"grad_norm": 0.21557386219501495,
"learning_rate": 0.0008736862898639095,
"loss": 2.021,
"step": 555
},
{
"epoch": 0.2550166265336544,
"grad_norm": 0.22761087119579315,
"learning_rate": 0.0008731921947243468,
"loss": 2.106,
"step": 556
},
{
"epoch": 0.25547528953101706,
"grad_norm": 0.23075616359710693,
"learning_rate": 0.0008726972754044427,
"loss": 2.095,
"step": 557
},
{
"epoch": 0.25593395252837975,
"grad_norm": 0.22966612875461578,
"learning_rate": 0.0008722015329972069,
"loss": 2.1291,
"step": 558
},
{
"epoch": 0.25639261552574244,
"grad_norm": 0.22476652264595032,
"learning_rate": 0.0008717049685974672,
"loss": 2.1237,
"step": 559
},
{
"epoch": 0.25685127852310513,
"grad_norm": 0.24779102206230164,
"learning_rate": 0.0008712075833018665,
"loss": 2.1545,
"step": 560
},
{
"epoch": 0.2573099415204678,
"grad_norm": 0.23312637209892273,
"learning_rate": 0.0008707093782088608,
"loss": 2.036,
"step": 561
},
{
"epoch": 0.2577686045178305,
"grad_norm": 0.23286572098731995,
"learning_rate": 0.0008702103544187167,
"loss": 2.0775,
"step": 562
},
{
"epoch": 0.2582272675151932,
"grad_norm": 0.21835166215896606,
"learning_rate": 0.0008697105130335085,
"loss": 2.0578,
"step": 563
},
{
"epoch": 0.2586859305125559,
"grad_norm": 0.23246073722839355,
"learning_rate": 0.0008692098551571164,
"loss": 2.1338,
"step": 564
},
{
"epoch": 0.2591445935099186,
"grad_norm": 0.2143106460571289,
"learning_rate": 0.0008687083818952235,
"loss": 2.103,
"step": 565
},
{
"epoch": 0.2596032565072813,
"grad_norm": 0.2300279587507248,
"learning_rate": 0.0008682060943553143,
"loss": 2.0455,
"step": 566
},
{
"epoch": 0.26006191950464397,
"grad_norm": 0.20937678217887878,
"learning_rate": 0.0008677029936466707,
"loss": 2.1348,
"step": 567
},
{
"epoch": 0.26052058250200666,
"grad_norm": 0.23574326932430267,
"learning_rate": 0.0008671990808803711,
"loss": 2.2181,
"step": 568
},
{
"epoch": 0.26097924549936935,
"grad_norm": 0.21991772949695587,
"learning_rate": 0.0008666943571692871,
"loss": 2.0943,
"step": 569
},
{
"epoch": 0.26143790849673204,
"grad_norm": 0.2087254673242569,
"learning_rate": 0.0008661888236280813,
"loss": 2.1008,
"step": 570
},
{
"epoch": 0.26189657149409473,
"grad_norm": 0.2515924274921417,
"learning_rate": 0.0008656824813732045,
"loss": 2.0643,
"step": 571
},
{
"epoch": 0.2623552344914574,
"grad_norm": 0.21847841143608093,
"learning_rate": 0.000865175331522894,
"loss": 2.0865,
"step": 572
},
{
"epoch": 0.2628138974888201,
"grad_norm": 0.21849699318408966,
"learning_rate": 0.0008646673751971703,
"loss": 2.1189,
"step": 573
},
{
"epoch": 0.26327256048618275,
"grad_norm": 0.22039243578910828,
"learning_rate": 0.000864158613517835,
"loss": 2.0121,
"step": 574
},
{
"epoch": 0.26373122348354544,
"grad_norm": 0.23070789873600006,
"learning_rate": 0.0008636490476084681,
"loss": 2.1115,
"step": 575
},
{
"epoch": 0.26418988648090813,
"grad_norm": 0.2162591814994812,
"learning_rate": 0.0008631386785944264,
"loss": 2.0596,
"step": 576
},
{
"epoch": 0.2646485494782708,
"grad_norm": 0.21255749464035034,
"learning_rate": 0.0008626275076028397,
"loss": 2.1233,
"step": 577
},
{
"epoch": 0.2651072124756335,
"grad_norm": 0.20645633339881897,
"learning_rate": 0.0008621155357626091,
"loss": 2.067,
"step": 578
},
{
"epoch": 0.2655658754729962,
"grad_norm": 0.24657191336154938,
"learning_rate": 0.0008616027642044042,
"loss": 2.109,
"step": 579
},
{
"epoch": 0.2660245384703589,
"grad_norm": 0.23502419888973236,
"learning_rate": 0.000861089194060661,
"loss": 2.0383,
"step": 580
},
{
"epoch": 0.2664832014677216,
"grad_norm": 0.22578121721744537,
"learning_rate": 0.000860574826465579,
"loss": 2.0788,
"step": 581
},
{
"epoch": 0.2669418644650843,
"grad_norm": 0.2112644910812378,
"learning_rate": 0.0008600596625551191,
"loss": 2.1882,
"step": 582
},
{
"epoch": 0.26740052746244697,
"grad_norm": 0.2199236899614334,
"learning_rate": 0.0008595437034670006,
"loss": 2.0461,
"step": 583
},
{
"epoch": 0.26785919045980966,
"grad_norm": 0.2161349654197693,
"learning_rate": 0.0008590269503406985,
"loss": 2.0656,
"step": 584
},
{
"epoch": 0.26831785345717235,
"grad_norm": 0.23057715594768524,
"learning_rate": 0.0008585094043174423,
"loss": 2.0316,
"step": 585
},
{
"epoch": 0.26877651645453504,
"grad_norm": 0.2260865718126297,
"learning_rate": 0.0008579910665402118,
"loss": 2.0954,
"step": 586
},
{
"epoch": 0.26923517945189773,
"grad_norm": 0.22231514751911163,
"learning_rate": 0.000857471938153736,
"loss": 2.0515,
"step": 587
},
{
"epoch": 0.2696938424492604,
"grad_norm": 0.22088243067264557,
"learning_rate": 0.0008569520203044892,
"loss": 2.0195,
"step": 588
},
{
"epoch": 0.2701525054466231,
"grad_norm": 0.21649569272994995,
"learning_rate": 0.0008564313141406901,
"loss": 2.0086,
"step": 589
},
{
"epoch": 0.2706111684439858,
"grad_norm": 0.209726482629776,
"learning_rate": 0.0008559098208122973,
"loss": 2.0144,
"step": 590
},
{
"epoch": 0.2710698314413485,
"grad_norm": 0.22525621950626373,
"learning_rate": 0.0008553875414710089,
"loss": 2.0824,
"step": 591
},
{
"epoch": 0.27152849443871113,
"grad_norm": 0.23110899329185486,
"learning_rate": 0.0008548644772702579,
"loss": 2.0443,
"step": 592
},
{
"epoch": 0.2719871574360738,
"grad_norm": 0.22233633697032928,
"learning_rate": 0.0008543406293652116,
"loss": 2.0794,
"step": 593
},
{
"epoch": 0.2724458204334365,
"grad_norm": 0.21259281039237976,
"learning_rate": 0.0008538159989127671,
"loss": 2.0711,
"step": 594
},
{
"epoch": 0.2729044834307992,
"grad_norm": 0.22887736558914185,
"learning_rate": 0.0008532905870715505,
"loss": 2.0875,
"step": 595
},
{
"epoch": 0.2733631464281619,
"grad_norm": 0.2196713536977768,
"learning_rate": 0.0008527643950019131,
"loss": 2.0406,
"step": 596
},
{
"epoch": 0.2738218094255246,
"grad_norm": 0.2098720520734787,
"learning_rate": 0.0008522374238659296,
"loss": 2.0537,
"step": 597
},
{
"epoch": 0.2742804724228873,
"grad_norm": 0.21671085059642792,
"learning_rate": 0.0008517096748273951,
"loss": 2.1635,
"step": 598
},
{
"epoch": 0.27473913542024997,
"grad_norm": 0.21734505891799927,
"learning_rate": 0.0008511811490518227,
"loss": 2.1503,
"step": 599
},
{
"epoch": 0.27519779841761266,
"grad_norm": 0.21793398261070251,
"learning_rate": 0.0008506518477064405,
"loss": 2.1468,
"step": 600
},
{
"epoch": 0.27565646141497535,
"grad_norm": 0.20423895120620728,
"learning_rate": 0.0008501217719601903,
"loss": 2.1147,
"step": 601
},
{
"epoch": 0.27611512441233804,
"grad_norm": 0.21685178577899933,
"learning_rate": 0.0008495909229837233,
"loss": 2.0797,
"step": 602
},
{
"epoch": 0.27657378740970073,
"grad_norm": 0.21056288480758667,
"learning_rate": 0.000849059301949399,
"loss": 1.9997,
"step": 603
},
{
"epoch": 0.2770324504070634,
"grad_norm": 0.22453755140304565,
"learning_rate": 0.0008485269100312812,
"loss": 2.0348,
"step": 604
},
{
"epoch": 0.2774911134044261,
"grad_norm": 0.21479322016239166,
"learning_rate": 0.0008479937484051368,
"loss": 2.0845,
"step": 605
},
{
"epoch": 0.2779497764017888,
"grad_norm": 0.2206791341304779,
"learning_rate": 0.0008474598182484323,
"loss": 2.051,
"step": 606
},
{
"epoch": 0.2784084393991515,
"grad_norm": 0.2189275324344635,
"learning_rate": 0.0008469251207403317,
"loss": 2.0858,
"step": 607
},
{
"epoch": 0.2788671023965142,
"grad_norm": 0.22278164327144623,
"learning_rate": 0.0008463896570616934,
"loss": 2.0998,
"step": 608
},
{
"epoch": 0.2793257653938768,
"grad_norm": 0.20976191759109497,
"learning_rate": 0.0008458534283950678,
"loss": 2.0574,
"step": 609
},
{
"epoch": 0.2797844283912395,
"grad_norm": 0.21160988509655,
"learning_rate": 0.0008453164359246952,
"loss": 2.1606,
"step": 610
},
{
"epoch": 0.2802430913886022,
"grad_norm": 0.2131117582321167,
"learning_rate": 0.0008447786808365022,
"loss": 2.0613,
"step": 611
},
{
"epoch": 0.2807017543859649,
"grad_norm": 0.2139628529548645,
"learning_rate": 0.0008442401643181,
"loss": 2.0876,
"step": 612
},
{
"epoch": 0.2811604173833276,
"grad_norm": 0.20437483489513397,
"learning_rate": 0.0008437008875587811,
"loss": 2.075,
"step": 613
},
{
"epoch": 0.2816190803806903,
"grad_norm": 0.21888871490955353,
"learning_rate": 0.0008431608517495171,
"loss": 2.1206,
"step": 614
},
{
"epoch": 0.28207774337805297,
"grad_norm": 0.20792806148529053,
"learning_rate": 0.0008426200580829561,
"loss": 2.006,
"step": 615
},
{
"epoch": 0.28253640637541566,
"grad_norm": 0.2103278934955597,
"learning_rate": 0.0008420785077534195,
"loss": 2.1148,
"step": 616
},
{
"epoch": 0.28299506937277835,
"grad_norm": 0.20034784078598022,
"learning_rate": 0.0008415362019569001,
"loss": 1.9729,
"step": 617
},
{
"epoch": 0.28345373237014104,
"grad_norm": 0.2146630585193634,
"learning_rate": 0.0008409931418910591,
"loss": 2.0945,
"step": 618
},
{
"epoch": 0.28391239536750373,
"grad_norm": 0.21206672489643097,
"learning_rate": 0.0008404493287552232,
"loss": 2.094,
"step": 619
},
{
"epoch": 0.2843710583648664,
"grad_norm": 0.2032850980758667,
"learning_rate": 0.0008399047637503825,
"loss": 2.0791,
"step": 620
},
{
"epoch": 0.2848297213622291,
"grad_norm": 0.19734740257263184,
"learning_rate": 0.0008393594480791875,
"loss": 2.0973,
"step": 621
},
{
"epoch": 0.2852883843595918,
"grad_norm": 0.20670683681964874,
"learning_rate": 0.0008388133829459463,
"loss": 2.0851,
"step": 622
},
{
"epoch": 0.2857470473569545,
"grad_norm": 0.22077637910842896,
"learning_rate": 0.0008382665695566227,
"loss": 2.0701,
"step": 623
},
{
"epoch": 0.2862057103543172,
"grad_norm": 0.21550698578357697,
"learning_rate": 0.0008377190091188324,
"loss": 2.0538,
"step": 624
},
{
"epoch": 0.2866643733516799,
"grad_norm": 0.22278448939323425,
"learning_rate": 0.0008371707028418413,
"loss": 2.1157,
"step": 625
},
{
"epoch": 0.2871230363490425,
"grad_norm": 0.22074009478092194,
"learning_rate": 0.0008366216519365621,
"loss": 2.1108,
"step": 626
},
{
"epoch": 0.2875816993464052,
"grad_norm": 0.2037685066461563,
"learning_rate": 0.0008360718576155525,
"loss": 2.0862,
"step": 627
},
{
"epoch": 0.2880403623437679,
"grad_norm": 0.20543797314167023,
"learning_rate": 0.0008355213210930118,
"loss": 2.0878,
"step": 628
},
{
"epoch": 0.2884990253411306,
"grad_norm": 0.20188874006271362,
"learning_rate": 0.0008349700435847778,
"loss": 2.1042,
"step": 629
},
{
"epoch": 0.2889576883384933,
"grad_norm": 0.20872662961483002,
"learning_rate": 0.0008344180263083256,
"loss": 2.1041,
"step": 630
},
{
"epoch": 0.28941635133585597,
"grad_norm": 0.19785374402999878,
"learning_rate": 0.000833865270482764,
"loss": 2.1197,
"step": 631
},
{
"epoch": 0.28987501433321866,
"grad_norm": 0.22294625639915466,
"learning_rate": 0.0008333117773288324,
"loss": 2.0132,
"step": 632
},
{
"epoch": 0.29033367733058135,
"grad_norm": 0.19978433847427368,
"learning_rate": 0.0008327575480688985,
"loss": 1.9985,
"step": 633
},
{
"epoch": 0.29079234032794404,
"grad_norm": 0.2072392851114273,
"learning_rate": 0.000832202583926956,
"loss": 2.1242,
"step": 634
},
{
"epoch": 0.29125100332530673,
"grad_norm": 0.21918737888336182,
"learning_rate": 0.0008316468861286217,
"loss": 2.112,
"step": 635
},
{
"epoch": 0.2917096663226694,
"grad_norm": 0.22510859370231628,
"learning_rate": 0.0008310904559011323,
"loss": 2.0863,
"step": 636
},
{
"epoch": 0.2921683293200321,
"grad_norm": 0.2136051505804062,
"learning_rate": 0.0008305332944733419,
"loss": 2.1353,
"step": 637
},
{
"epoch": 0.2926269923173948,
"grad_norm": 0.21353985369205475,
"learning_rate": 0.0008299754030757202,
"loss": 2.0573,
"step": 638
},
{
"epoch": 0.2930856553147575,
"grad_norm": 0.213229700922966,
"learning_rate": 0.0008294167829403481,
"loss": 2.0287,
"step": 639
},
{
"epoch": 0.2935443183121202,
"grad_norm": 0.21581120789051056,
"learning_rate": 0.0008288574353009164,
"loss": 2.1136,
"step": 640
},
{
"epoch": 0.2940029813094829,
"grad_norm": 0.22093065083026886,
"learning_rate": 0.0008282973613927225,
"loss": 2.0348,
"step": 641
},
{
"epoch": 0.29446164430684557,
"grad_norm": 0.20405928790569305,
"learning_rate": 0.0008277365624526675,
"loss": 2.0047,
"step": 642
},
{
"epoch": 0.2949203073042082,
"grad_norm": 0.21157243847846985,
"learning_rate": 0.0008271750397192541,
"loss": 2.1302,
"step": 643
},
{
"epoch": 0.2953789703015709,
"grad_norm": 0.21956898272037506,
"learning_rate": 0.0008266127944325832,
"loss": 1.9298,
"step": 644
},
{
"epoch": 0.2958376332989336,
"grad_norm": 0.23177148401737213,
"learning_rate": 0.0008260498278343513,
"loss": 2.0903,
"step": 645
},
{
"epoch": 0.2962962962962963,
"grad_norm": 0.2142215520143509,
"learning_rate": 0.0008254861411678485,
"loss": 2.0803,
"step": 646
},
{
"epoch": 0.29675495929365897,
"grad_norm": 0.23605458438396454,
"learning_rate": 0.0008249217356779544,
"loss": 2.1226,
"step": 647
},
{
"epoch": 0.29721362229102166,
"grad_norm": 0.23986950516700745,
"learning_rate": 0.0008243566126111363,
"loss": 2.0831,
"step": 648
},
{
"epoch": 0.29767228528838435,
"grad_norm": 0.2019045203924179,
"learning_rate": 0.0008237907732154466,
"loss": 2.1024,
"step": 649
},
{
"epoch": 0.29813094828574704,
"grad_norm": 0.20799516141414642,
"learning_rate": 0.0008232242187405194,
"loss": 2.0262,
"step": 650
},
{
"epoch": 0.29858961128310973,
"grad_norm": 0.20103059709072113,
"learning_rate": 0.000822656950437568,
"loss": 2.0956,
"step": 651
},
{
"epoch": 0.2990482742804724,
"grad_norm": 0.21725745499134064,
"learning_rate": 0.0008220889695593823,
"loss": 1.9961,
"step": 652
},
{
"epoch": 0.2995069372778351,
"grad_norm": 0.22156177461147308,
"learning_rate": 0.0008215202773603259,
"loss": 2.0478,
"step": 653
},
{
"epoch": 0.2999656002751978,
"grad_norm": 0.21392515301704407,
"learning_rate": 0.0008209508750963328,
"loss": 2.0243,
"step": 654
},
{
"epoch": 0.3004242632725605,
"grad_norm": 0.22471068799495697,
"learning_rate": 0.0008203807640249062,
"loss": 2.0813,
"step": 655
},
{
"epoch": 0.3008829262699232,
"grad_norm": 0.2244964987039566,
"learning_rate": 0.0008198099454051136,
"loss": 2.1023,
"step": 656
},
{
"epoch": 0.3013415892672859,
"grad_norm": 0.21745680272579193,
"learning_rate": 0.0008192384204975857,
"loss": 2.0786,
"step": 657
},
{
"epoch": 0.30180025226464857,
"grad_norm": 0.2229241579771042,
"learning_rate": 0.000818666190564513,
"loss": 2.0743,
"step": 658
},
{
"epoch": 0.30225891526201126,
"grad_norm": 0.2115909904241562,
"learning_rate": 0.0008180932568696426,
"loss": 2.0094,
"step": 659
},
{
"epoch": 0.3027175782593739,
"grad_norm": 0.2031831294298172,
"learning_rate": 0.0008175196206782764,
"loss": 2.0184,
"step": 660
},
{
"epoch": 0.3031762412567366,
"grad_norm": 0.2229335755109787,
"learning_rate": 0.0008169452832572675,
"loss": 2.1423,
"step": 661
},
{
"epoch": 0.3036349042540993,
"grad_norm": 0.21565304696559906,
"learning_rate": 0.0008163702458750173,
"loss": 2.1203,
"step": 662
},
{
"epoch": 0.30409356725146197,
"grad_norm": 0.21220865845680237,
"learning_rate": 0.0008157945098014734,
"loss": 2.0156,
"step": 663
},
{
"epoch": 0.30455223024882466,
"grad_norm": 0.2078232765197754,
"learning_rate": 0.0008152180763081267,
"loss": 2.0312,
"step": 664
},
{
"epoch": 0.30501089324618735,
"grad_norm": 0.22738665342330933,
"learning_rate": 0.0008146409466680076,
"loss": 2.0005,
"step": 665
},
{
"epoch": 0.30546955624355004,
"grad_norm": 0.21662425994873047,
"learning_rate": 0.0008140631221556845,
"loss": 2.0713,
"step": 666
},
{
"epoch": 0.30592821924091274,
"grad_norm": 0.21719245612621307,
"learning_rate": 0.0008134846040472599,
"loss": 2.1149,
"step": 667
},
{
"epoch": 0.3063868822382754,
"grad_norm": 0.22838954627513885,
"learning_rate": 0.0008129053936203688,
"loss": 2.0798,
"step": 668
},
{
"epoch": 0.3068455452356381,
"grad_norm": 0.22218266129493713,
"learning_rate": 0.0008123254921541745,
"loss": 2.05,
"step": 669
},
{
"epoch": 0.3073042082330008,
"grad_norm": 0.2189200520515442,
"learning_rate": 0.0008117449009293668,
"loss": 1.9582,
"step": 670
},
{
"epoch": 0.3077628712303635,
"grad_norm": 0.21056202054023743,
"learning_rate": 0.0008111636212281586,
"loss": 2.0575,
"step": 671
},
{
"epoch": 0.3082215342277262,
"grad_norm": 0.203217014670372,
"learning_rate": 0.0008105816543342833,
"loss": 2.1569,
"step": 672
},
{
"epoch": 0.3086801972250889,
"grad_norm": 0.21148984134197235,
"learning_rate": 0.0008099990015329919,
"loss": 2.0709,
"step": 673
},
{
"epoch": 0.3091388602224516,
"grad_norm": 0.2196563482284546,
"learning_rate": 0.0008094156641110504,
"loss": 2.0562,
"step": 674
},
{
"epoch": 0.30959752321981426,
"grad_norm": 0.2040826976299286,
"learning_rate": 0.0008088316433567369,
"loss": 2.1202,
"step": 675
},
{
"epoch": 0.31005618621717695,
"grad_norm": 0.2094750553369522,
"learning_rate": 0.0008082469405598378,
"loss": 2.0964,
"step": 676
},
{
"epoch": 0.3105148492145396,
"grad_norm": 0.22648915648460388,
"learning_rate": 0.0008076615570116468,
"loss": 2.103,
"step": 677
},
{
"epoch": 0.3109735122119023,
"grad_norm": 0.19523710012435913,
"learning_rate": 0.0008070754940049603,
"loss": 1.9951,
"step": 678
},
{
"epoch": 0.31143217520926497,
"grad_norm": 0.2192288190126419,
"learning_rate": 0.0008064887528340756,
"loss": 2.1476,
"step": 679
},
{
"epoch": 0.31189083820662766,
"grad_norm": 0.2514795958995819,
"learning_rate": 0.0008059013347947874,
"loss": 2.0782,
"step": 680
},
{
"epoch": 0.31234950120399035,
"grad_norm": 0.21343016624450684,
"learning_rate": 0.0008053132411843857,
"loss": 2.1463,
"step": 681
},
{
"epoch": 0.31280816420135305,
"grad_norm": 0.20519982278347015,
"learning_rate": 0.0008047244733016521,
"loss": 2.0681,
"step": 682
},
{
"epoch": 0.31326682719871574,
"grad_norm": 0.2167380154132843,
"learning_rate": 0.0008041350324468573,
"loss": 2.0253,
"step": 683
},
{
"epoch": 0.3137254901960784,
"grad_norm": 0.2132306694984436,
"learning_rate": 0.0008035449199217583,
"loss": 2.1157,
"step": 684
},
{
"epoch": 0.3141841531934411,
"grad_norm": 0.22044776380062103,
"learning_rate": 0.0008029541370295957,
"loss": 1.9841,
"step": 685
},
{
"epoch": 0.3146428161908038,
"grad_norm": 0.2135344296693802,
"learning_rate": 0.0008023626850750903,
"loss": 2.0137,
"step": 686
},
{
"epoch": 0.3151014791881665,
"grad_norm": 0.23021115362644196,
"learning_rate": 0.0008017705653644406,
"loss": 2.1351,
"step": 687
},
{
"epoch": 0.3155601421855292,
"grad_norm": 0.2178305685520172,
"learning_rate": 0.0008011777792053195,
"loss": 2.0981,
"step": 688
},
{
"epoch": 0.3160188051828919,
"grad_norm": 0.21275338530540466,
"learning_rate": 0.0008005843279068725,
"loss": 1.996,
"step": 689
},
{
"epoch": 0.3164774681802546,
"grad_norm": 0.21468037366867065,
"learning_rate": 0.000799990212779713,
"loss": 2.1606,
"step": 690
},
{
"epoch": 0.31693613117761726,
"grad_norm": 0.22198495268821716,
"learning_rate": 0.0007993954351359214,
"loss": 2.1147,
"step": 691
},
{
"epoch": 0.31739479417497996,
"grad_norm": 0.2034171223640442,
"learning_rate": 0.0007987999962890406,
"loss": 2.0711,
"step": 692
},
{
"epoch": 0.31785345717234265,
"grad_norm": 0.2128974348306656,
"learning_rate": 0.0007982038975540742,
"loss": 2.0427,
"step": 693
},
{
"epoch": 0.3183121201697053,
"grad_norm": 0.21687395870685577,
"learning_rate": 0.0007976071402474826,
"loss": 2.1693,
"step": 694
},
{
"epoch": 0.318770783167068,
"grad_norm": 0.2238493412733078,
"learning_rate": 0.0007970097256871811,
"loss": 2.1498,
"step": 695
},
{
"epoch": 0.31922944616443066,
"grad_norm": 0.208769291639328,
"learning_rate": 0.0007964116551925364,
"loss": 2.0682,
"step": 696
},
{
"epoch": 0.31968810916179335,
"grad_norm": 0.20230168104171753,
"learning_rate": 0.0007958129300843637,
"loss": 2.0107,
"step": 697
},
{
"epoch": 0.32014677215915605,
"grad_norm": 0.20471221208572388,
"learning_rate": 0.0007952135516849239,
"loss": 2.0368,
"step": 698
},
{
"epoch": 0.32060543515651874,
"grad_norm": 0.2189210057258606,
"learning_rate": 0.0007946135213179207,
"loss": 2.112,
"step": 699
},
{
"epoch": 0.3210640981538814,
"grad_norm": 0.21764183044433594,
"learning_rate": 0.0007940128403084977,
"loss": 2.1456,
"step": 700
},
{
"epoch": 0.3215227611512441,
"grad_norm": 0.2249792218208313,
"learning_rate": 0.0007934115099832355,
"loss": 2.0944,
"step": 701
},
{
"epoch": 0.3219814241486068,
"grad_norm": 0.2205788940191269,
"learning_rate": 0.0007928095316701483,
"loss": 2.0837,
"step": 702
},
{
"epoch": 0.3224400871459695,
"grad_norm": 0.28880539536476135,
"learning_rate": 0.0007922069066986819,
"loss": 1.9559,
"step": 703
},
{
"epoch": 0.3228987501433322,
"grad_norm": 0.20694275200366974,
"learning_rate": 0.0007916036363997097,
"loss": 2.1091,
"step": 704
},
{
"epoch": 0.3233574131406949,
"grad_norm": 0.2093353420495987,
"learning_rate": 0.0007909997221055308,
"loss": 1.966,
"step": 705
},
{
"epoch": 0.3238160761380576,
"grad_norm": 0.20323126018047333,
"learning_rate": 0.0007903951651498658,
"loss": 1.9747,
"step": 706
},
{
"epoch": 0.32427473913542026,
"grad_norm": 0.20614881813526154,
"learning_rate": 0.0007897899668678557,
"loss": 2.0793,
"step": 707
},
{
"epoch": 0.32473340213278296,
"grad_norm": 0.21958185732364655,
"learning_rate": 0.0007891841285960566,
"loss": 2.0623,
"step": 708
},
{
"epoch": 0.32519206513014565,
"grad_norm": 0.2067558765411377,
"learning_rate": 0.0007885776516724388,
"loss": 2.0762,
"step": 709
},
{
"epoch": 0.32565072812750834,
"grad_norm": 0.19834484159946442,
"learning_rate": 0.0007879705374363831,
"loss": 2.125,
"step": 710
},
{
"epoch": 0.32610939112487103,
"grad_norm": 0.21600794792175293,
"learning_rate": 0.000787362787228677,
"loss": 2.0976,
"step": 711
},
{
"epoch": 0.32656805412223366,
"grad_norm": 0.21103863418102264,
"learning_rate": 0.0007867544023915134,
"loss": 2.0676,
"step": 712
},
{
"epoch": 0.32702671711959636,
"grad_norm": 0.21780982613563538,
"learning_rate": 0.0007861453842684861,
"loss": 2.0635,
"step": 713
},
{
"epoch": 0.32748538011695905,
"grad_norm": 0.21895761787891388,
"learning_rate": 0.0007855357342045882,
"loss": 1.99,
"step": 714
},
{
"epoch": 0.32794404311432174,
"grad_norm": 0.2169560045003891,
"learning_rate": 0.0007849254535462074,
"loss": 2.1773,
"step": 715
},
{
"epoch": 0.32840270611168443,
"grad_norm": 0.2168380618095398,
"learning_rate": 0.0007843145436411252,
"loss": 2.0372,
"step": 716
},
{
"epoch": 0.3288613691090471,
"grad_norm": 0.22844718396663666,
"learning_rate": 0.0007837030058385117,
"loss": 2.0085,
"step": 717
},
{
"epoch": 0.3293200321064098,
"grad_norm": 0.22492541372776031,
"learning_rate": 0.0007830908414889246,
"loss": 2.1127,
"step": 718
},
{
"epoch": 0.3297786951037725,
"grad_norm": 0.20620620250701904,
"learning_rate": 0.0007824780519443046,
"loss": 2.0889,
"step": 719
},
{
"epoch": 0.3302373581011352,
"grad_norm": 0.20316390693187714,
"learning_rate": 0.0007818646385579735,
"loss": 2.0407,
"step": 720
},
{
"epoch": 0.3306960210984979,
"grad_norm": 0.21376098692417145,
"learning_rate": 0.0007812506026846307,
"loss": 1.9942,
"step": 721
},
{
"epoch": 0.3311546840958606,
"grad_norm": 0.20882846415042877,
"learning_rate": 0.0007806359456803504,
"loss": 2.1205,
"step": 722
},
{
"epoch": 0.33161334709322327,
"grad_norm": 0.2178703099489212,
"learning_rate": 0.0007800206689025785,
"loss": 2.0506,
"step": 723
},
{
"epoch": 0.33207201009058596,
"grad_norm": 0.21789012849330902,
"learning_rate": 0.0007794047737101297,
"loss": 2.0181,
"step": 724
},
{
"epoch": 0.33253067308794865,
"grad_norm": 0.2194446325302124,
"learning_rate": 0.0007787882614631843,
"loss": 2.0095,
"step": 725
},
{
"epoch": 0.33298933608531134,
"grad_norm": 0.2217778116464615,
"learning_rate": 0.0007781711335232856,
"loss": 2.0852,
"step": 726
},
{
"epoch": 0.33344799908267403,
"grad_norm": 0.2384095937013626,
"learning_rate": 0.0007775533912533363,
"loss": 2.059,
"step": 727
},
{
"epoch": 0.3339066620800367,
"grad_norm": 0.21221087872982025,
"learning_rate": 0.0007769350360175962,
"loss": 2.1288,
"step": 728
},
{
"epoch": 0.33436532507739936,
"grad_norm": 0.2151600867509842,
"learning_rate": 0.0007763160691816784,
"loss": 2.0792,
"step": 729
},
{
"epoch": 0.33482398807476205,
"grad_norm": 0.20686650276184082,
"learning_rate": 0.000775696492112547,
"loss": 2.0733,
"step": 730
},
{
"epoch": 0.33528265107212474,
"grad_norm": 0.21812154352664948,
"learning_rate": 0.0007750763061785137,
"loss": 2.1117,
"step": 731
},
{
"epoch": 0.33574131406948743,
"grad_norm": 0.21360473334789276,
"learning_rate": 0.000774455512749235,
"loss": 1.969,
"step": 732
},
{
"epoch": 0.3361999770668501,
"grad_norm": 0.20529839396476746,
"learning_rate": 0.0007738341131957085,
"loss": 2.0568,
"step": 733
},
{
"epoch": 0.3366586400642128,
"grad_norm": 0.22770942747592926,
"learning_rate": 0.000773212108890271,
"loss": 2.0424,
"step": 734
},
{
"epoch": 0.3371173030615755,
"grad_norm": 0.21731652319431305,
"learning_rate": 0.0007725895012065947,
"loss": 2.0715,
"step": 735
},
{
"epoch": 0.3375759660589382,
"grad_norm": 0.21372638642787933,
"learning_rate": 0.0007719662915196844,
"loss": 2.0018,
"step": 736
},
{
"epoch": 0.3380346290563009,
"grad_norm": 0.21397438645362854,
"learning_rate": 0.0007713424812058736,
"loss": 2.0194,
"step": 737
},
{
"epoch": 0.3384932920536636,
"grad_norm": 0.22247309982776642,
"learning_rate": 0.0007707180716428237,
"loss": 2.0392,
"step": 738
},
{
"epoch": 0.33895195505102627,
"grad_norm": 0.2267957627773285,
"learning_rate": 0.0007700930642095184,
"loss": 2.0822,
"step": 739
},
{
"epoch": 0.33941061804838896,
"grad_norm": 0.22300532460212708,
"learning_rate": 0.0007694674602862621,
"loss": 2.0967,
"step": 740
},
{
"epoch": 0.33986928104575165,
"grad_norm": 0.21655309200286865,
"learning_rate": 0.0007688412612546769,
"loss": 2.1384,
"step": 741
},
{
"epoch": 0.34032794404311434,
"grad_norm": 0.21486511826515198,
"learning_rate": 0.0007682144684976983,
"loss": 2.0623,
"step": 742
},
{
"epoch": 0.34078660704047703,
"grad_norm": 0.21019886434078217,
"learning_rate": 0.0007675870833995739,
"loss": 2.1098,
"step": 743
},
{
"epoch": 0.3412452700378397,
"grad_norm": 0.2292262315750122,
"learning_rate": 0.0007669591073458592,
"loss": 2.0928,
"step": 744
},
{
"epoch": 0.3417039330352024,
"grad_norm": 0.21345004439353943,
"learning_rate": 0.0007663305417234146,
"loss": 2.093,
"step": 745
},
{
"epoch": 0.34216259603256505,
"grad_norm": 0.20633019506931305,
"learning_rate": 0.0007657013879204022,
"loss": 2.0468,
"step": 746
},
{
"epoch": 0.34262125902992774,
"grad_norm": 0.224202960729599,
"learning_rate": 0.0007650716473262842,
"loss": 1.9905,
"step": 747
},
{
"epoch": 0.34307992202729043,
"grad_norm": 0.21652525663375854,
"learning_rate": 0.0007644413213318177,
"loss": 2.0764,
"step": 748
},
{
"epoch": 0.3435385850246531,
"grad_norm": 0.22413066029548645,
"learning_rate": 0.0007638104113290531,
"loss": 2.0521,
"step": 749
},
{
"epoch": 0.3439972480220158,
"grad_norm": 0.24490875005722046,
"learning_rate": 0.0007631789187113303,
"loss": 2.0217,
"step": 750
},
{
"epoch": 0.3444559110193785,
"grad_norm": 0.21686027944087982,
"learning_rate": 0.000762546844873276,
"loss": 2.0156,
"step": 751
},
{
"epoch": 0.3449145740167412,
"grad_norm": 0.2213289439678192,
"learning_rate": 0.0007619141912108007,
"loss": 2.1116,
"step": 752
},
{
"epoch": 0.3453732370141039,
"grad_norm": 0.21875837445259094,
"learning_rate": 0.000761280959121095,
"loss": 1.9902,
"step": 753
},
{
"epoch": 0.3458319000114666,
"grad_norm": 0.20345953106880188,
"learning_rate": 0.0007606471500026273,
"loss": 2.094,
"step": 754
},
{
"epoch": 0.34629056300882927,
"grad_norm": 0.20488396286964417,
"learning_rate": 0.0007600127652551401,
"loss": 2.0627,
"step": 755
},
{
"epoch": 0.34674922600619196,
"grad_norm": 0.224142923951149,
"learning_rate": 0.0007593778062796472,
"loss": 2.0499,
"step": 756
},
{
"epoch": 0.34720788900355465,
"grad_norm": 0.20819023251533508,
"learning_rate": 0.000758742274478431,
"loss": 2.0338,
"step": 757
},
{
"epoch": 0.34766655200091734,
"grad_norm": 0.2060965895652771,
"learning_rate": 0.0007581061712550381,
"loss": 2.1245,
"step": 758
},
{
"epoch": 0.34812521499828003,
"grad_norm": 0.20614653825759888,
"learning_rate": 0.0007574694980142779,
"loss": 2.0757,
"step": 759
},
{
"epoch": 0.3485838779956427,
"grad_norm": 0.203657329082489,
"learning_rate": 0.0007568322561622183,
"loss": 2.0935,
"step": 760
},
{
"epoch": 0.3490425409930054,
"grad_norm": 0.2097507119178772,
"learning_rate": 0.0007561944471061826,
"loss": 2.07,
"step": 761
},
{
"epoch": 0.3495012039903681,
"grad_norm": 0.22039324045181274,
"learning_rate": 0.0007555560722547475,
"loss": 2.1215,
"step": 762
},
{
"epoch": 0.34995986698773074,
"grad_norm": 0.20728997886180878,
"learning_rate": 0.0007549171330177387,
"loss": 1.9742,
"step": 763
},
{
"epoch": 0.35041852998509343,
"grad_norm": 0.2151811420917511,
"learning_rate": 0.0007542776308062285,
"loss": 2.1571,
"step": 764
},
{
"epoch": 0.3508771929824561,
"grad_norm": 0.22903107106685638,
"learning_rate": 0.0007536375670325325,
"loss": 2.0917,
"step": 765
},
{
"epoch": 0.3513358559798188,
"grad_norm": 0.2107861042022705,
"learning_rate": 0.0007529969431102063,
"loss": 2.0371,
"step": 766
},
{
"epoch": 0.3517945189771815,
"grad_norm": 0.21278630197048187,
"learning_rate": 0.000752355760454043,
"loss": 2.0565,
"step": 767
},
{
"epoch": 0.3522531819745442,
"grad_norm": 0.22686250507831573,
"learning_rate": 0.0007517140204800693,
"loss": 2.0396,
"step": 768
},
{
"epoch": 0.3527118449719069,
"grad_norm": 0.21308445930480957,
"learning_rate": 0.0007510717246055425,
"loss": 2.0489,
"step": 769
},
{
"epoch": 0.3531705079692696,
"grad_norm": 0.22111448645591736,
"learning_rate": 0.0007504288742489482,
"loss": 2.0487,
"step": 770
},
{
"epoch": 0.35362917096663227,
"grad_norm": 0.2148803174495697,
"learning_rate": 0.0007497854708299963,
"loss": 2.1156,
"step": 771
},
{
"epoch": 0.35408783396399496,
"grad_norm": 0.21086914837360382,
"learning_rate": 0.0007491415157696178,
"loss": 2.0567,
"step": 772
},
{
"epoch": 0.35454649696135765,
"grad_norm": 0.22268664836883545,
"learning_rate": 0.0007484970104899623,
"loss": 2.0697,
"step": 773
},
{
"epoch": 0.35500515995872034,
"grad_norm": 0.21766768395900726,
"learning_rate": 0.0007478519564143945,
"loss": 2.1321,
"step": 774
},
{
"epoch": 0.35546382295608303,
"grad_norm": 0.21559031307697296,
"learning_rate": 0.000747206354967491,
"loss": 2.0872,
"step": 775
},
{
"epoch": 0.3559224859534457,
"grad_norm": 0.22769607603549957,
"learning_rate": 0.0007465602075750373,
"loss": 2.1619,
"step": 776
},
{
"epoch": 0.3563811489508084,
"grad_norm": 0.20463575422763824,
"learning_rate": 0.0007459135156640247,
"loss": 2.0063,
"step": 777
},
{
"epoch": 0.3568398119481711,
"grad_norm": 0.21552564203739166,
"learning_rate": 0.0007452662806626468,
"loss": 2.055,
"step": 778
},
{
"epoch": 0.3572984749455338,
"grad_norm": 0.21106886863708496,
"learning_rate": 0.0007446185040002967,
"loss": 2.0111,
"step": 779
},
{
"epoch": 0.35775713794289643,
"grad_norm": 0.21511626243591309,
"learning_rate": 0.0007439701871075642,
"loss": 2.1055,
"step": 780
},
{
"epoch": 0.3582158009402591,
"grad_norm": 0.22827257215976715,
"learning_rate": 0.0007433213314162313,
"loss": 2.0779,
"step": 781
},
{
"epoch": 0.3586744639376218,
"grad_norm": 0.1980343759059906,
"learning_rate": 0.0007426719383592705,
"loss": 2.0521,
"step": 782
},
{
"epoch": 0.3591331269349845,
"grad_norm": 0.21226345002651215,
"learning_rate": 0.000742022009370841,
"loss": 2.0817,
"step": 783
},
{
"epoch": 0.3595917899323472,
"grad_norm": 0.23105488717556,
"learning_rate": 0.0007413715458862855,
"loss": 2.0321,
"step": 784
},
{
"epoch": 0.3600504529297099,
"grad_norm": 0.21000002324581146,
"learning_rate": 0.0007407205493421272,
"loss": 2.0462,
"step": 785
},
{
"epoch": 0.3605091159270726,
"grad_norm": 0.21854713559150696,
"learning_rate": 0.0007400690211760661,
"loss": 2.0385,
"step": 786
},
{
"epoch": 0.36096777892443527,
"grad_norm": 0.2056826800107956,
"learning_rate": 0.0007394169628269771,
"loss": 2.0275,
"step": 787
},
{
"epoch": 0.36142644192179796,
"grad_norm": 0.20605027675628662,
"learning_rate": 0.0007387643757349051,
"loss": 2.0256,
"step": 788
},
{
"epoch": 0.36188510491916065,
"grad_norm": 0.2052639275789261,
"learning_rate": 0.0007381112613410635,
"loss": 2.1067,
"step": 789
},
{
"epoch": 0.36234376791652334,
"grad_norm": 0.20139415562152863,
"learning_rate": 0.0007374576210878298,
"loss": 1.9975,
"step": 790
},
{
"epoch": 0.36280243091388603,
"grad_norm": 0.2176741361618042,
"learning_rate": 0.0007368034564187425,
"loss": 2.0906,
"step": 791
},
{
"epoch": 0.3632610939112487,
"grad_norm": 0.20262950658798218,
"learning_rate": 0.0007361487687784989,
"loss": 2.0571,
"step": 792
},
{
"epoch": 0.3637197569086114,
"grad_norm": 0.2004884034395218,
"learning_rate": 0.0007354935596129513,
"loss": 1.9983,
"step": 793
},
{
"epoch": 0.3641784199059741,
"grad_norm": 0.2075381726026535,
"learning_rate": 0.000734837830369103,
"loss": 2.0662,
"step": 794
},
{
"epoch": 0.3646370829033368,
"grad_norm": 0.20061127841472626,
"learning_rate": 0.0007341815824951066,
"loss": 2.0444,
"step": 795
},
{
"epoch": 0.3650957459006995,
"grad_norm": 0.20215682685375214,
"learning_rate": 0.0007335248174402597,
"loss": 2.0953,
"step": 796
},
{
"epoch": 0.3655544088980621,
"grad_norm": 0.19612139463424683,
"learning_rate": 0.0007328675366550023,
"loss": 2.0749,
"step": 797
},
{
"epoch": 0.3660130718954248,
"grad_norm": 0.21514500677585602,
"learning_rate": 0.0007322097415909134,
"loss": 1.9971,
"step": 798
},
{
"epoch": 0.3664717348927875,
"grad_norm": 0.21557655930519104,
"learning_rate": 0.0007315514337007071,
"loss": 2.1079,
"step": 799
},
{
"epoch": 0.3669303978901502,
"grad_norm": 0.21025098860263824,
"learning_rate": 0.0007308926144382312,
"loss": 2.0261,
"step": 800
},
{
"epoch": 0.3673890608875129,
"grad_norm": 0.2244543731212616,
"learning_rate": 0.0007302332852584619,
"loss": 2.0474,
"step": 801
},
{
"epoch": 0.3678477238848756,
"grad_norm": 0.21591056883335114,
"learning_rate": 0.0007295734476175018,
"loss": 2.1229,
"step": 802
},
{
"epoch": 0.36830638688223827,
"grad_norm": 0.210074782371521,
"learning_rate": 0.0007289131029725768,
"loss": 2.0539,
"step": 803
},
{
"epoch": 0.36876504987960096,
"grad_norm": 0.20620512962341309,
"learning_rate": 0.0007282522527820319,
"loss": 2.1053,
"step": 804
},
{
"epoch": 0.36922371287696365,
"grad_norm": 0.23144273459911346,
"learning_rate": 0.000727590898505329,
"loss": 2.0246,
"step": 805
},
{
"epoch": 0.36968237587432634,
"grad_norm": 0.20549125969409943,
"learning_rate": 0.0007269290416030429,
"loss": 1.9085,
"step": 806
},
{
"epoch": 0.37014103887168903,
"grad_norm": 0.26246872544288635,
"learning_rate": 0.000726266683536859,
"loss": 1.9778,
"step": 807
},
{
"epoch": 0.3705997018690517,
"grad_norm": 0.19751723110675812,
"learning_rate": 0.0007256038257695687,
"loss": 2.0446,
"step": 808
},
{
"epoch": 0.3710583648664144,
"grad_norm": 0.20792733132839203,
"learning_rate": 0.0007249404697650678,
"loss": 2.0962,
"step": 809
},
{
"epoch": 0.3715170278637771,
"grad_norm": 0.20400507748126984,
"learning_rate": 0.0007242766169883518,
"loss": 2.0847,
"step": 810
},
{
"epoch": 0.3719756908611398,
"grad_norm": 0.21465177834033966,
"learning_rate": 0.0007236122689055138,
"loss": 2.1232,
"step": 811
},
{
"epoch": 0.3724343538585025,
"grad_norm": 0.2103392481803894,
"learning_rate": 0.0007229474269837401,
"loss": 2.0738,
"step": 812
},
{
"epoch": 0.3728930168558652,
"grad_norm": 0.2060329169034958,
"learning_rate": 0.0007222820926913085,
"loss": 2.0119,
"step": 813
},
{
"epoch": 0.3733516798532278,
"grad_norm": 0.21740378439426422,
"learning_rate": 0.0007216162674975833,
"loss": 2.0631,
"step": 814
},
{
"epoch": 0.3738103428505905,
"grad_norm": 0.21104048192501068,
"learning_rate": 0.0007209499528730138,
"loss": 2.0623,
"step": 815
},
{
"epoch": 0.3742690058479532,
"grad_norm": 0.21209484338760376,
"learning_rate": 0.0007202831502891294,
"loss": 2.007,
"step": 816
},
{
"epoch": 0.3747276688453159,
"grad_norm": 0.22087699174880981,
"learning_rate": 0.0007196158612185375,
"loss": 2.0867,
"step": 817
},
{
"epoch": 0.3751863318426786,
"grad_norm": 0.2000526785850525,
"learning_rate": 0.0007189480871349201,
"loss": 1.9772,
"step": 818
},
{
"epoch": 0.37564499484004127,
"grad_norm": 0.20850495994091034,
"learning_rate": 0.0007182798295130299,
"loss": 1.962,
"step": 819
},
{
"epoch": 0.37610365783740396,
"grad_norm": 0.2149035483598709,
"learning_rate": 0.0007176110898286878,
"loss": 1.9543,
"step": 820
},
{
"epoch": 0.37656232083476665,
"grad_norm": 0.21748314797878265,
"learning_rate": 0.0007169418695587791,
"loss": 1.9662,
"step": 821
},
{
"epoch": 0.37702098383212934,
"grad_norm": 0.22076661884784698,
"learning_rate": 0.0007162721701812506,
"loss": 2.0565,
"step": 822
},
{
"epoch": 0.37747964682949203,
"grad_norm": 0.2225821167230606,
"learning_rate": 0.0007156019931751072,
"loss": 1.9728,
"step": 823
},
{
"epoch": 0.3779383098268547,
"grad_norm": 0.2167477011680603,
"learning_rate": 0.0007149313400204082,
"loss": 2.0288,
"step": 824
},
{
"epoch": 0.3783969728242174,
"grad_norm": 0.21818512678146362,
"learning_rate": 0.0007142602121982653,
"loss": 2.0103,
"step": 825
},
{
"epoch": 0.3788556358215801,
"grad_norm": 0.20744219422340393,
"learning_rate": 0.0007135886111908379,
"loss": 2.0973,
"step": 826
},
{
"epoch": 0.3793142988189428,
"grad_norm": 0.21205294132232666,
"learning_rate": 0.0007129165384813303,
"loss": 2.0725,
"step": 827
},
{
"epoch": 0.3797729618163055,
"grad_norm": 0.19800640642642975,
"learning_rate": 0.0007122439955539888,
"loss": 2.0897,
"step": 828
},
{
"epoch": 0.3802316248136682,
"grad_norm": 0.23023304343223572,
"learning_rate": 0.0007115709838940983,
"loss": 2.0818,
"step": 829
},
{
"epoch": 0.38069028781103087,
"grad_norm": 0.2049965262413025,
"learning_rate": 0.0007108975049879785,
"loss": 2.0317,
"step": 830
},
{
"epoch": 0.3811489508083935,
"grad_norm": 0.19689057767391205,
"learning_rate": 0.0007102235603229814,
"loss": 1.9914,
"step": 831
},
{
"epoch": 0.3816076138057562,
"grad_norm": 0.2079332172870636,
"learning_rate": 0.000709549151387487,
"loss": 2.0252,
"step": 832
},
{
"epoch": 0.3820662768031189,
"grad_norm": 0.20631088316440582,
"learning_rate": 0.0007088742796709013,
"loss": 2.0423,
"step": 833
},
{
"epoch": 0.3825249398004816,
"grad_norm": 0.20826935768127441,
"learning_rate": 0.000708198946663652,
"loss": 2.0725,
"step": 834
},
{
"epoch": 0.38298360279784427,
"grad_norm": 0.21292831003665924,
"learning_rate": 0.0007075231538571856,
"loss": 2.0698,
"step": 835
},
{
"epoch": 0.38344226579520696,
"grad_norm": 0.20294208824634552,
"learning_rate": 0.0007068469027439641,
"loss": 2.1463,
"step": 836
},
{
"epoch": 0.38390092879256965,
"grad_norm": 0.21025563776493073,
"learning_rate": 0.0007061701948174613,
"loss": 2.0997,
"step": 837
},
{
"epoch": 0.38435959178993234,
"grad_norm": 0.20970845222473145,
"learning_rate": 0.0007054930315721606,
"loss": 2.0256,
"step": 838
},
{
"epoch": 0.38481825478729503,
"grad_norm": 0.20193617045879364,
"learning_rate": 0.0007048154145035501,
"loss": 2.0395,
"step": 839
},
{
"epoch": 0.3852769177846577,
"grad_norm": 0.2004946917295456,
"learning_rate": 0.0007041373451081207,
"loss": 2.0753,
"step": 840
},
{
"epoch": 0.3857355807820204,
"grad_norm": 0.20870424807071686,
"learning_rate": 0.0007034588248833621,
"loss": 2.1321,
"step": 841
},
{
"epoch": 0.3861942437793831,
"grad_norm": 0.2045329362154007,
"learning_rate": 0.0007027798553277595,
"loss": 2.023,
"step": 842
},
{
"epoch": 0.3866529067767458,
"grad_norm": 0.21000954508781433,
"learning_rate": 0.0007021004379407909,
"loss": 2.0148,
"step": 843
},
{
"epoch": 0.3871115697741085,
"grad_norm": 0.21233925223350525,
"learning_rate": 0.0007014205742229227,
"loss": 2.1163,
"step": 844
},
{
"epoch": 0.3875702327714712,
"grad_norm": 0.21720074117183685,
"learning_rate": 0.0007007402656756072,
"loss": 2.0576,
"step": 845
},
{
"epoch": 0.38802889576883387,
"grad_norm": 0.204544797539711,
"learning_rate": 0.0007000595138012797,
"loss": 2.088,
"step": 846
},
{
"epoch": 0.38848755876619656,
"grad_norm": 0.20782947540283203,
"learning_rate": 0.0006993783201033535,
"loss": 1.9916,
"step": 847
},
{
"epoch": 0.38894622176355925,
"grad_norm": 0.21160821616649628,
"learning_rate": 0.0006986966860862182,
"loss": 2.0836,
"step": 848
},
{
"epoch": 0.3894048847609219,
"grad_norm": 0.2038404643535614,
"learning_rate": 0.000698014613255236,
"loss": 1.951,
"step": 849
},
{
"epoch": 0.3898635477582846,
"grad_norm": 0.20343975722789764,
"learning_rate": 0.0006973321031167382,
"loss": 2.0908,
"step": 850
},
{
"epoch": 0.39032221075564727,
"grad_norm": 0.2122567743062973,
"learning_rate": 0.0006966491571780216,
"loss": 2.1029,
"step": 851
},
{
"epoch": 0.39078087375300996,
"grad_norm": 0.20677269995212555,
"learning_rate": 0.0006959657769473453,
"loss": 1.9867,
"step": 852
},
{
"epoch": 0.39123953675037265,
"grad_norm": 0.21802811324596405,
"learning_rate": 0.000695281963933928,
"loss": 2.0419,
"step": 853
},
{
"epoch": 0.39169819974773534,
"grad_norm": 0.20267446339130402,
"learning_rate": 0.0006945977196479438,
"loss": 1.9752,
"step": 854
},
{
"epoch": 0.39215686274509803,
"grad_norm": 0.2083546221256256,
"learning_rate": 0.0006939130456005196,
"loss": 2.0687,
"step": 855
},
{
"epoch": 0.3926155257424607,
"grad_norm": 0.21336708962917328,
"learning_rate": 0.0006932279433037311,
"loss": 2.0363,
"step": 856
},
{
"epoch": 0.3930741887398234,
"grad_norm": 0.21802544593811035,
"learning_rate": 0.0006925424142705997,
"loss": 2.0415,
"step": 857
},
{
"epoch": 0.3935328517371861,
"grad_norm": 0.20592783391475677,
"learning_rate": 0.0006918564600150896,
"loss": 2.0128,
"step": 858
},
{
"epoch": 0.3939915147345488,
"grad_norm": 0.2097194492816925,
"learning_rate": 0.0006911700820521042,
"loss": 2.0275,
"step": 859
},
{
"epoch": 0.3944501777319115,
"grad_norm": 0.2024749219417572,
"learning_rate": 0.0006904832818974818,
"loss": 1.9577,
"step": 860
},
{
"epoch": 0.3949088407292742,
"grad_norm": 0.2207859307527542,
"learning_rate": 0.0006897960610679939,
"loss": 2.0489,
"step": 861
},
{
"epoch": 0.39536750372663687,
"grad_norm": 0.1989334374666214,
"learning_rate": 0.0006891084210813407,
"loss": 1.8995,
"step": 862
},
{
"epoch": 0.39582616672399956,
"grad_norm": 0.2036294937133789,
"learning_rate": 0.0006884203634561483,
"loss": 1.9404,
"step": 863
},
{
"epoch": 0.39628482972136225,
"grad_norm": 0.20011119544506073,
"learning_rate": 0.0006877318897119651,
"loss": 2.0217,
"step": 864
},
{
"epoch": 0.39674349271872494,
"grad_norm": 0.20727403461933136,
"learning_rate": 0.0006870430013692579,
"loss": 2.0865,
"step": 865
},
{
"epoch": 0.3972021557160876,
"grad_norm": 0.20020446181297302,
"learning_rate": 0.0006863536999494101,
"loss": 2.0042,
"step": 866
},
{
"epoch": 0.39766081871345027,
"grad_norm": 0.20029814541339874,
"learning_rate": 0.0006856639869747167,
"loss": 2.0621,
"step": 867
},
{
"epoch": 0.39811948171081296,
"grad_norm": 0.21974340081214905,
"learning_rate": 0.0006849738639683818,
"loss": 2.0228,
"step": 868
},
{
"epoch": 0.39857814470817565,
"grad_norm": 0.2062513679265976,
"learning_rate": 0.000684283332454515,
"loss": 2.0446,
"step": 869
},
{
"epoch": 0.39903680770553834,
"grad_norm": 0.20677417516708374,
"learning_rate": 0.0006835923939581281,
"loss": 2.0688,
"step": 870
},
{
"epoch": 0.39949547070290103,
"grad_norm": 0.20623227953910828,
"learning_rate": 0.0006829010500051318,
"loss": 1.9709,
"step": 871
},
{
"epoch": 0.3999541337002637,
"grad_norm": 0.22134390473365784,
"learning_rate": 0.0006822093021223321,
"loss": 1.9793,
"step": 872
},
{
"epoch": 0.4004127966976264,
"grad_norm": 0.20972129702568054,
"learning_rate": 0.0006815171518374268,
"loss": 2.0029,
"step": 873
},
{
"epoch": 0.4008714596949891,
"grad_norm": 0.20987388491630554,
"learning_rate": 0.0006808246006790031,
"loss": 2.0873,
"step": 874
},
{
"epoch": 0.4013301226923518,
"grad_norm": 0.2427993267774582,
"learning_rate": 0.0006801316501765329,
"loss": 2.0749,
"step": 875
},
{
"epoch": 0.4017887856897145,
"grad_norm": 0.21702928841114044,
"learning_rate": 0.0006794383018603704,
"loss": 2.0127,
"step": 876
},
{
"epoch": 0.4022474486870772,
"grad_norm": 0.20609933137893677,
"learning_rate": 0.0006787445572617481,
"loss": 2.0927,
"step": 877
},
{
"epoch": 0.40270611168443987,
"grad_norm": 0.2104141265153885,
"learning_rate": 0.0006780504179127734,
"loss": 2.0111,
"step": 878
},
{
"epoch": 0.40316477468180256,
"grad_norm": 0.22867022454738617,
"learning_rate": 0.0006773558853464265,
"loss": 2.0669,
"step": 879
},
{
"epoch": 0.40362343767916525,
"grad_norm": 0.19389353692531586,
"learning_rate": 0.000676660961096555,
"loss": 2.0541,
"step": 880
},
{
"epoch": 0.40408210067652794,
"grad_norm": 0.22168780863285065,
"learning_rate": 0.000675965646697872,
"loss": 2.0047,
"step": 881
},
{
"epoch": 0.40454076367389064,
"grad_norm": 0.20727968215942383,
"learning_rate": 0.0006752699436859519,
"loss": 1.973,
"step": 882
},
{
"epoch": 0.40499942667125327,
"grad_norm": 0.20576947927474976,
"learning_rate": 0.0006745738535972279,
"loss": 1.9368,
"step": 883
},
{
"epoch": 0.40545808966861596,
"grad_norm": 0.2393384575843811,
"learning_rate": 0.0006738773779689874,
"loss": 2.0789,
"step": 884
},
{
"epoch": 0.40591675266597865,
"grad_norm": 0.23270809650421143,
"learning_rate": 0.0006731805183393696,
"loss": 2.0411,
"step": 885
},
{
"epoch": 0.40637541566334134,
"grad_norm": 0.22750818729400635,
"learning_rate": 0.0006724832762473618,
"loss": 2.107,
"step": 886
},
{
"epoch": 0.40683407866070403,
"grad_norm": 0.20982596278190613,
"learning_rate": 0.0006717856532327956,
"loss": 2.0357,
"step": 887
},
{
"epoch": 0.4072927416580667,
"grad_norm": 0.19714857637882233,
"learning_rate": 0.0006710876508363444,
"loss": 2.0167,
"step": 888
},
{
"epoch": 0.4077514046554294,
"grad_norm": 0.22397594153881073,
"learning_rate": 0.0006703892705995189,
"loss": 2.0003,
"step": 889
},
{
"epoch": 0.4082100676527921,
"grad_norm": 0.22221064567565918,
"learning_rate": 0.0006696905140646647,
"loss": 2.0943,
"step": 890
},
{
"epoch": 0.4086687306501548,
"grad_norm": 0.2069602757692337,
"learning_rate": 0.0006689913827749581,
"loss": 2.0623,
"step": 891
},
{
"epoch": 0.4091273936475175,
"grad_norm": 0.23294809460639954,
"learning_rate": 0.0006682918782744032,
"loss": 2.0493,
"step": 892
},
{
"epoch": 0.4095860566448802,
"grad_norm": 0.2209988832473755,
"learning_rate": 0.0006675920021078282,
"loss": 2.0827,
"step": 893
},
{
"epoch": 0.41004471964224287,
"grad_norm": 0.2189313918352127,
"learning_rate": 0.0006668917558208823,
"loss": 2.0625,
"step": 894
},
{
"epoch": 0.41050338263960556,
"grad_norm": 0.2222999781370163,
"learning_rate": 0.0006661911409600321,
"loss": 2.1023,
"step": 895
},
{
"epoch": 0.41096204563696825,
"grad_norm": 0.21648626029491425,
"learning_rate": 0.0006654901590725577,
"loss": 2.095,
"step": 896
},
{
"epoch": 0.41142070863433094,
"grad_norm": 0.20311670005321503,
"learning_rate": 0.0006647888117065507,
"loss": 2.0059,
"step": 897
},
{
"epoch": 0.41187937163169364,
"grad_norm": 0.20792511105537415,
"learning_rate": 0.0006640871004109086,
"loss": 2.0479,
"step": 898
},
{
"epoch": 0.4123380346290563,
"grad_norm": 0.21453820168972015,
"learning_rate": 0.000663385026735334,
"loss": 2.0957,
"step": 899
},
{
"epoch": 0.41279669762641896,
"grad_norm": 0.23444972932338715,
"learning_rate": 0.0006626825922303285,
"loss": 2.0973,
"step": 900
},
{
"epoch": 0.41325536062378165,
"grad_norm": 0.2053755819797516,
"learning_rate": 0.0006619797984471915,
"loss": 1.9788,
"step": 901
},
{
"epoch": 0.41371402362114434,
"grad_norm": 0.2135566622018814,
"learning_rate": 0.0006612766469380158,
"loss": 2.009,
"step": 902
},
{
"epoch": 0.41417268661850704,
"grad_norm": 0.21280092000961304,
"learning_rate": 0.0006605731392556833,
"loss": 2.0295,
"step": 903
},
{
"epoch": 0.4146313496158697,
"grad_norm": 0.2071743905544281,
"learning_rate": 0.0006598692769538637,
"loss": 2.0896,
"step": 904
},
{
"epoch": 0.4150900126132324,
"grad_norm": 0.20563572645187378,
"learning_rate": 0.0006591650615870091,
"loss": 1.9705,
"step": 905
},
{
"epoch": 0.4155486756105951,
"grad_norm": 0.20979006588459015,
"learning_rate": 0.0006584604947103514,
"loss": 1.9412,
"step": 906
},
{
"epoch": 0.4160073386079578,
"grad_norm": 0.20554529130458832,
"learning_rate": 0.0006577555778798993,
"loss": 2.0496,
"step": 907
},
{
"epoch": 0.4164660016053205,
"grad_norm": 0.20870766043663025,
"learning_rate": 0.0006570503126524336,
"loss": 2.0575,
"step": 908
},
{
"epoch": 0.4169246646026832,
"grad_norm": 0.2070256471633911,
"learning_rate": 0.0006563447005855054,
"loss": 2.0517,
"step": 909
},
{
"epoch": 0.4173833276000459,
"grad_norm": 0.2138548046350479,
"learning_rate": 0.000655638743237431,
"loss": 2.043,
"step": 910
},
{
"epoch": 0.41784199059740856,
"grad_norm": 0.21714003384113312,
"learning_rate": 0.0006549324421672894,
"loss": 2.0863,
"step": 911
},
{
"epoch": 0.41830065359477125,
"grad_norm": 0.2060667872428894,
"learning_rate": 0.0006542257989349194,
"loss": 1.9269,
"step": 912
},
{
"epoch": 0.41875931659213395,
"grad_norm": 0.20156466960906982,
"learning_rate": 0.0006535188151009142,
"loss": 1.9235,
"step": 913
},
{
"epoch": 0.41921797958949664,
"grad_norm": 0.21848171949386597,
"learning_rate": 0.0006528114922266204,
"loss": 2.044,
"step": 914
},
{
"epoch": 0.4196766425868593,
"grad_norm": 0.20595327019691467,
"learning_rate": 0.0006521038318741327,
"loss": 2.0879,
"step": 915
},
{
"epoch": 0.420135305584222,
"grad_norm": 0.20734086632728577,
"learning_rate": 0.0006513958356062912,
"loss": 2.0528,
"step": 916
},
{
"epoch": 0.42059396858158465,
"grad_norm": 0.19747161865234375,
"learning_rate": 0.0006506875049866781,
"loss": 2.1392,
"step": 917
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.1931842863559723,
"learning_rate": 0.0006499788415796137,
"loss": 2.1672,
"step": 918
},
{
"epoch": 0.42151129457631004,
"grad_norm": 0.1959591805934906,
"learning_rate": 0.0006492698469501532,
"loss": 1.9969,
"step": 919
},
{
"epoch": 0.4219699575736727,
"grad_norm": 0.20635941624641418,
"learning_rate": 0.0006485605226640837,
"loss": 2.0645,
"step": 920
},
{
"epoch": 0.4224286205710354,
"grad_norm": 0.19862346351146698,
"learning_rate": 0.00064785087028792,
"loss": 2.0925,
"step": 921
},
{
"epoch": 0.4228872835683981,
"grad_norm": 0.19620586931705475,
"learning_rate": 0.0006471408913889019,
"loss": 1.9763,
"step": 922
},
{
"epoch": 0.4233459465657608,
"grad_norm": 0.19678597152233124,
"learning_rate": 0.0006464305875349892,
"loss": 2.045,
"step": 923
},
{
"epoch": 0.4238046095631235,
"grad_norm": 0.2001389116048813,
"learning_rate": 0.000645719960294861,
"loss": 2.0573,
"step": 924
},
{
"epoch": 0.4242632725604862,
"grad_norm": 0.199999138712883,
"learning_rate": 0.0006450090112379092,
"loss": 2.0152,
"step": 925
},
{
"epoch": 0.4247219355578489,
"grad_norm": 0.21809956431388855,
"learning_rate": 0.0006442977419342371,
"loss": 2.1242,
"step": 926
},
{
"epoch": 0.42518059855521156,
"grad_norm": 0.21406897902488708,
"learning_rate": 0.000643586153954655,
"loss": 2.0192,
"step": 927
},
{
"epoch": 0.42563926155257426,
"grad_norm": 0.20598581433296204,
"learning_rate": 0.0006428742488706772,
"loss": 2.025,
"step": 928
},
{
"epoch": 0.42609792454993695,
"grad_norm": 0.20392043888568878,
"learning_rate": 0.0006421620282545182,
"loss": 2.0729,
"step": 929
},
{
"epoch": 0.42655658754729964,
"grad_norm": 0.20378302037715912,
"learning_rate": 0.0006414494936790892,
"loss": 1.9795,
"step": 930
},
{
"epoch": 0.42701525054466233,
"grad_norm": 0.2105528712272644,
"learning_rate": 0.0006407366467179951,
"loss": 1.901,
"step": 931
},
{
"epoch": 0.427473913542025,
"grad_norm": 0.22872476279735565,
"learning_rate": 0.0006400234889455301,
"loss": 2.0744,
"step": 932
},
{
"epoch": 0.4279325765393877,
"grad_norm": 0.20878195762634277,
"learning_rate": 0.0006393100219366755,
"loss": 2.0282,
"step": 933
},
{
"epoch": 0.42839123953675035,
"grad_norm": 0.21447831392288208,
"learning_rate": 0.0006385962472670953,
"loss": 2.0045,
"step": 934
},
{
"epoch": 0.42884990253411304,
"grad_norm": 0.5684143304824829,
"learning_rate": 0.0006378821665131328,
"loss": 2.1334,
"step": 935
},
{
"epoch": 0.4293085655314757,
"grad_norm": 0.21263346076011658,
"learning_rate": 0.0006371677812518072,
"loss": 2.0547,
"step": 936
},
{
"epoch": 0.4297672285288384,
"grad_norm": 0.21270139515399933,
"learning_rate": 0.0006364530930608107,
"loss": 2.0148,
"step": 937
},
{
"epoch": 0.4302258915262011,
"grad_norm": 0.20758093893527985,
"learning_rate": 0.0006357381035185038,
"loss": 1.956,
"step": 938
},
{
"epoch": 0.4306845545235638,
"grad_norm": 0.20801101624965668,
"learning_rate": 0.0006350228142039131,
"loss": 1.9817,
"step": 939
},
{
"epoch": 0.4311432175209265,
"grad_norm": 0.21466778218746185,
"learning_rate": 0.000634307226696727,
"loss": 2.0606,
"step": 940
},
{
"epoch": 0.4316018805182892,
"grad_norm": 0.21379855275154114,
"learning_rate": 0.0006335913425772926,
"loss": 2.0125,
"step": 941
},
{
"epoch": 0.4320605435156519,
"grad_norm": 0.20247600972652435,
"learning_rate": 0.0006328751634266117,
"loss": 2.0628,
"step": 942
},
{
"epoch": 0.43251920651301456,
"grad_norm": 0.19737817347049713,
"learning_rate": 0.0006321586908263382,
"loss": 1.9532,
"step": 943
},
{
"epoch": 0.43297786951037726,
"grad_norm": 0.2062937319278717,
"learning_rate": 0.0006314419263587732,
"loss": 2.136,
"step": 944
},
{
"epoch": 0.43343653250773995,
"grad_norm": 0.21703611314296722,
"learning_rate": 0.0006307248716068637,
"loss": 1.9833,
"step": 945
},
{
"epoch": 0.43389519550510264,
"grad_norm": 0.20312544703483582,
"learning_rate": 0.0006300075281541964,
"loss": 2.035,
"step": 946
},
{
"epoch": 0.43435385850246533,
"grad_norm": 0.2081185132265091,
"learning_rate": 0.0006292898975849966,
"loss": 2.0236,
"step": 947
},
{
"epoch": 0.434812521499828,
"grad_norm": 0.21818359196186066,
"learning_rate": 0.000628571981484123,
"loss": 1.9796,
"step": 948
},
{
"epoch": 0.4352711844971907,
"grad_norm": 0.2088465541601181,
"learning_rate": 0.0006278537814370654,
"loss": 2.0348,
"step": 949
},
{
"epoch": 0.4357298474945534,
"grad_norm": 0.22719798982143402,
"learning_rate": 0.0006271352990299406,
"loss": 2.071,
"step": 950
},
{
"epoch": 0.43618851049191604,
"grad_norm": 0.22375385463237762,
"learning_rate": 0.0006264165358494885,
"loss": 2.0575,
"step": 951
},
{
"epoch": 0.43664717348927873,
"grad_norm": 0.20833942294120789,
"learning_rate": 0.0006256974934830694,
"loss": 2.11,
"step": 952
},
{
"epoch": 0.4371058364866414,
"grad_norm": 0.2153061181306839,
"learning_rate": 0.0006249781735186606,
"loss": 2.0542,
"step": 953
},
{
"epoch": 0.4375644994840041,
"grad_norm": 0.20888163149356842,
"learning_rate": 0.0006242585775448518,
"loss": 1.9912,
"step": 954
},
{
"epoch": 0.4380231624813668,
"grad_norm": 0.2155771106481552,
"learning_rate": 0.0006235387071508427,
"loss": 1.9887,
"step": 955
},
{
"epoch": 0.4384818254787295,
"grad_norm": 0.20684239268302917,
"learning_rate": 0.0006228185639264384,
"loss": 2.0983,
"step": 956
},
{
"epoch": 0.4389404884760922,
"grad_norm": 0.21265797317028046,
"learning_rate": 0.0006220981494620475,
"loss": 2.0428,
"step": 957
},
{
"epoch": 0.4393991514734549,
"grad_norm": 0.20863701403141022,
"learning_rate": 0.000621377465348677,
"loss": 2.0962,
"step": 958
},
{
"epoch": 0.43985781447081757,
"grad_norm": 0.20519472658634186,
"learning_rate": 0.0006206565131779293,
"loss": 2.0237,
"step": 959
},
{
"epoch": 0.44031647746818026,
"grad_norm": 0.2195596545934677,
"learning_rate": 0.0006199352945419994,
"loss": 2.0947,
"step": 960
},
{
"epoch": 0.44077514046554295,
"grad_norm": 0.20423394441604614,
"learning_rate": 0.00061921381103367,
"loss": 2.0128,
"step": 961
},
{
"epoch": 0.44123380346290564,
"grad_norm": 0.21093270182609558,
"learning_rate": 0.0006184920642463094,
"loss": 2.0903,
"step": 962
},
{
"epoch": 0.44169246646026833,
"grad_norm": 0.21820101141929626,
"learning_rate": 0.0006177700557738672,
"loss": 2.1077,
"step": 963
},
{
"epoch": 0.442151129457631,
"grad_norm": 0.18811312317848206,
"learning_rate": 0.0006170477872108706,
"loss": 1.8996,
"step": 964
},
{
"epoch": 0.4426097924549937,
"grad_norm": 0.20191539824008942,
"learning_rate": 0.0006163252601524216,
"loss": 2.1241,
"step": 965
},
{
"epoch": 0.4430684554523564,
"grad_norm": 0.2188512533903122,
"learning_rate": 0.0006156024761941925,
"loss": 2.0492,
"step": 966
},
{
"epoch": 0.4435271184497191,
"grad_norm": 0.2196035087108612,
"learning_rate": 0.000614879436932424,
"loss": 2.0739,
"step": 967
},
{
"epoch": 0.4439857814470818,
"grad_norm": 0.20328226685523987,
"learning_rate": 0.0006141561439639196,
"loss": 2.0987,
"step": 968
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.1986251175403595,
"learning_rate": 0.0006134325988860433,
"loss": 2.0369,
"step": 969
},
{
"epoch": 0.4449031074418071,
"grad_norm": 0.20789046585559845,
"learning_rate": 0.0006127088032967165,
"loss": 2.1431,
"step": 970
},
{
"epoch": 0.4453617704391698,
"grad_norm": 0.19903430342674255,
"learning_rate": 0.0006119847587944131,
"loss": 2.0446,
"step": 971
},
{
"epoch": 0.4458204334365325,
"grad_norm": 0.19700679183006287,
"learning_rate": 0.0006112604669781572,
"loss": 2.0074,
"step": 972
},
{
"epoch": 0.4462790964338952,
"grad_norm": 0.19748254120349884,
"learning_rate": 0.0006105359294475188,
"loss": 2.0321,
"step": 973
},
{
"epoch": 0.4467377594312579,
"grad_norm": 0.1988946795463562,
"learning_rate": 0.0006098111478026107,
"loss": 2.0674,
"step": 974
},
{
"epoch": 0.44719642242862057,
"grad_norm": 0.2279505580663681,
"learning_rate": 0.0006090861236440848,
"loss": 2.0193,
"step": 975
},
{
"epoch": 0.44765508542598326,
"grad_norm": 0.20271766185760498,
"learning_rate": 0.0006083608585731282,
"loss": 1.9943,
"step": 976
},
{
"epoch": 0.44811374842334595,
"grad_norm": 0.21280042827129364,
"learning_rate": 0.0006076353541914609,
"loss": 1.9912,
"step": 977
},
{
"epoch": 0.44857241142070864,
"grad_norm": 0.21517567336559296,
"learning_rate": 0.0006069096121013307,
"loss": 2.0036,
"step": 978
},
{
"epoch": 0.44903107441807133,
"grad_norm": 0.19836685061454773,
"learning_rate": 0.0006061836339055105,
"loss": 1.9575,
"step": 979
},
{
"epoch": 0.449489737415434,
"grad_norm": 0.2218286395072937,
"learning_rate": 0.0006054574212072948,
"loss": 1.9794,
"step": 980
},
{
"epoch": 0.4499484004127967,
"grad_norm": 0.22965092957019806,
"learning_rate": 0.0006047309756104958,
"loss": 2.0386,
"step": 981
},
{
"epoch": 0.4504070634101594,
"grad_norm": 0.22065341472625732,
"learning_rate": 0.00060400429871944,
"loss": 2.0025,
"step": 982
},
{
"epoch": 0.4508657264075221,
"grad_norm": 0.21281006932258606,
"learning_rate": 0.0006032773921389654,
"loss": 2.1142,
"step": 983
},
{
"epoch": 0.4513243894048848,
"grad_norm": 0.23297744989395142,
"learning_rate": 0.0006025502574744162,
"loss": 2.1015,
"step": 984
},
{
"epoch": 0.4517830524022475,
"grad_norm": 0.20511138439178467,
"learning_rate": 0.000601822896331641,
"loss": 1.9965,
"step": 985
},
{
"epoch": 0.4522417153996101,
"grad_norm": 0.21362292766571045,
"learning_rate": 0.0006010953103169883,
"loss": 1.9653,
"step": 986
},
{
"epoch": 0.4527003783969728,
"grad_norm": 0.20162364840507507,
"learning_rate": 0.0006003675010373034,
"loss": 2.0416,
"step": 987
},
{
"epoch": 0.4531590413943355,
"grad_norm": 0.20575660467147827,
"learning_rate": 0.0005996394700999246,
"loss": 1.9822,
"step": 988
},
{
"epoch": 0.4536177043916982,
"grad_norm": 0.20947544276714325,
"learning_rate": 0.0005989112191126794,
"loss": 2.1067,
"step": 989
},
{
"epoch": 0.4540763673890609,
"grad_norm": 0.21259558200836182,
"learning_rate": 0.0005981827496838822,
"loss": 2.0942,
"step": 990
},
{
"epoch": 0.45453503038642357,
"grad_norm": 0.20960339903831482,
"learning_rate": 0.0005974540634223286,
"loss": 2.025,
"step": 991
},
{
"epoch": 0.45499369338378626,
"grad_norm": 0.21487383544445038,
"learning_rate": 0.0005967251619372939,
"loss": 2.058,
"step": 992
},
{
"epoch": 0.45545235638114895,
"grad_norm": 0.20597831904888153,
"learning_rate": 0.0005959960468385284,
"loss": 2.0312,
"step": 993
},
{
"epoch": 0.45591101937851164,
"grad_norm": 0.20461833477020264,
"learning_rate": 0.0005952667197362542,
"loss": 1.9837,
"step": 994
},
{
"epoch": 0.45636968237587433,
"grad_norm": 0.20901718735694885,
"learning_rate": 0.0005945371822411621,
"loss": 2.0466,
"step": 995
},
{
"epoch": 0.456828345373237,
"grad_norm": 0.2080875188112259,
"learning_rate": 0.0005938074359644063,
"loss": 2.0162,
"step": 996
},
{
"epoch": 0.4572870083705997,
"grad_norm": 0.20954827964305878,
"learning_rate": 0.0005930774825176034,
"loss": 2.0346,
"step": 997
},
{
"epoch": 0.4577456713679624,
"grad_norm": 0.19821038842201233,
"learning_rate": 0.0005923473235128268,
"loss": 1.9293,
"step": 998
},
{
"epoch": 0.4582043343653251,
"grad_norm": 0.19981344044208527,
"learning_rate": 0.0005916169605626042,
"loss": 2.0217,
"step": 999
},
{
"epoch": 0.4586629973626878,
"grad_norm": 0.19854122400283813,
"learning_rate": 0.0005908863952799134,
"loss": 1.9895,
"step": 1000
},
{
"epoch": 0.4591216603600505,
"grad_norm": 0.2204698771238327,
"learning_rate": 0.0005901556292781793,
"loss": 2.0672,
"step": 1001
},
{
"epoch": 0.45958032335741317,
"grad_norm": 0.1977786123752594,
"learning_rate": 0.0005894246641712698,
"loss": 2.0213,
"step": 1002
},
{
"epoch": 0.4600389863547758,
"grad_norm": 0.2120916247367859,
"learning_rate": 0.0005886935015734931,
"loss": 2.0726,
"step": 1003
},
{
"epoch": 0.4604976493521385,
"grad_norm": 0.21250134706497192,
"learning_rate": 0.0005879621430995928,
"loss": 1.9938,
"step": 1004
},
{
"epoch": 0.4609563123495012,
"grad_norm": 0.21415621042251587,
"learning_rate": 0.0005872305903647455,
"loss": 2.0026,
"step": 1005
},
{
"epoch": 0.4614149753468639,
"grad_norm": 0.20980094373226166,
"learning_rate": 0.0005864988449845569,
"loss": 2.0776,
"step": 1006
},
{
"epoch": 0.46187363834422657,
"grad_norm": 0.2050776183605194,
"learning_rate": 0.0005857669085750578,
"loss": 2.0003,
"step": 1007
},
{
"epoch": 0.46233230134158926,
"grad_norm": 0.20367972552776337,
"learning_rate": 0.0005850347827527013,
"loss": 1.9709,
"step": 1008
},
{
"epoch": 0.46279096433895195,
"grad_norm": 0.21815037727355957,
"learning_rate": 0.0005843024691343584,
"loss": 2.0327,
"step": 1009
},
{
"epoch": 0.46324962733631464,
"grad_norm": 0.2235909253358841,
"learning_rate": 0.000583569969337315,
"loss": 2.1054,
"step": 1010
},
{
"epoch": 0.46370829033367733,
"grad_norm": 0.20548036694526672,
"learning_rate": 0.0005828372849792686,
"loss": 2.0473,
"step": 1011
},
{
"epoch": 0.46416695333104,
"grad_norm": 0.20795919001102448,
"learning_rate": 0.0005821044176783234,
"loss": 2.1039,
"step": 1012
},
{
"epoch": 0.4646256163284027,
"grad_norm": 0.19424349069595337,
"learning_rate": 0.0005813713690529886,
"loss": 2.0572,
"step": 1013
},
{
"epoch": 0.4650842793257654,
"grad_norm": 0.21041779220104218,
"learning_rate": 0.0005806381407221729,
"loss": 2.0071,
"step": 1014
},
{
"epoch": 0.4655429423231281,
"grad_norm": 0.21895766258239746,
"learning_rate": 0.0005799047343051826,
"loss": 1.996,
"step": 1015
},
{
"epoch": 0.4660016053204908,
"grad_norm": 0.19554315507411957,
"learning_rate": 0.0005791711514217171,
"loss": 2.0148,
"step": 1016
},
{
"epoch": 0.4664602683178535,
"grad_norm": 0.1981644630432129,
"learning_rate": 0.0005784373936918654,
"loss": 2.0046,
"step": 1017
},
{
"epoch": 0.46691893131521617,
"grad_norm": 0.20414015650749207,
"learning_rate": 0.0005777034627361025,
"loss": 2.0339,
"step": 1018
},
{
"epoch": 0.46737759431257886,
"grad_norm": 0.23540878295898438,
"learning_rate": 0.0005769693601752864,
"loss": 2.0677,
"step": 1019
},
{
"epoch": 0.4678362573099415,
"grad_norm": 0.213117316365242,
"learning_rate": 0.0005762350876306537,
"loss": 2.0308,
"step": 1020
},
{
"epoch": 0.4682949203073042,
"grad_norm": 0.21114347875118256,
"learning_rate": 0.0005755006467238168,
"loss": 2.0989,
"step": 1021
},
{
"epoch": 0.4687535833046669,
"grad_norm": 0.200186625123024,
"learning_rate": 0.0005747660390767593,
"loss": 2.0521,
"step": 1022
},
{
"epoch": 0.46921224630202957,
"grad_norm": 0.21845105290412903,
"learning_rate": 0.0005740312663118338,
"loss": 2.0594,
"step": 1023
},
{
"epoch": 0.46967090929939226,
"grad_norm": 0.21826617419719696,
"learning_rate": 0.0005732963300517568,
"loss": 1.9888,
"step": 1024
},
{
"epoch": 0.47012957229675495,
"grad_norm": 0.20372392237186432,
"learning_rate": 0.0005725612319196064,
"loss": 2.0452,
"step": 1025
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.20367003977298737,
"learning_rate": 0.000571825973538818,
"loss": 2.0831,
"step": 1026
},
{
"epoch": 0.47104689829148033,
"grad_norm": 0.1997164487838745,
"learning_rate": 0.0005710905565331811,
"loss": 2.0308,
"step": 1027
},
{
"epoch": 0.471505561288843,
"grad_norm": 0.19941207766532898,
"learning_rate": 0.0005703549825268353,
"loss": 2.0698,
"step": 1028
},
{
"epoch": 0.4719642242862057,
"grad_norm": 0.22999364137649536,
"learning_rate": 0.0005696192531442667,
"loss": 1.9619,
"step": 1029
},
{
"epoch": 0.4724228872835684,
"grad_norm": 0.22323592007160187,
"learning_rate": 0.000568883370010305,
"loss": 2.0986,
"step": 1030
},
{
"epoch": 0.4728815502809311,
"grad_norm": 0.19384577870368958,
"learning_rate": 0.0005681473347501192,
"loss": 1.9721,
"step": 1031
},
{
"epoch": 0.4733402132782938,
"grad_norm": 0.19675423204898834,
"learning_rate": 0.0005674111489892144,
"loss": 1.9694,
"step": 1032
},
{
"epoch": 0.4737988762756565,
"grad_norm": 0.20552603900432587,
"learning_rate": 0.0005666748143534282,
"loss": 1.9504,
"step": 1033
},
{
"epoch": 0.47425753927301917,
"grad_norm": 0.2119583636522293,
"learning_rate": 0.0005659383324689266,
"loss": 2.0328,
"step": 1034
},
{
"epoch": 0.47471620227038186,
"grad_norm": 0.2085297852754593,
"learning_rate": 0.0005652017049622007,
"loss": 2.0445,
"step": 1035
},
{
"epoch": 0.47517486526774455,
"grad_norm": 0.21672020852565765,
"learning_rate": 0.0005644649334600641,
"loss": 2.0212,
"step": 1036
},
{
"epoch": 0.4756335282651072,
"grad_norm": 0.19705256819725037,
"learning_rate": 0.0005637280195896474,
"loss": 2.0352,
"step": 1037
},
{
"epoch": 0.4760921912624699,
"grad_norm": 0.21350912749767303,
"learning_rate": 0.0005629909649783961,
"loss": 2.0098,
"step": 1038
},
{
"epoch": 0.47655085425983257,
"grad_norm": 0.19826769828796387,
"learning_rate": 0.0005622537712540664,
"loss": 2.0428,
"step": 1039
},
{
"epoch": 0.47700951725719526,
"grad_norm": 0.20029094815254211,
"learning_rate": 0.0005615164400447218,
"loss": 1.9363,
"step": 1040
},
{
"epoch": 0.47746818025455795,
"grad_norm": 0.2050243616104126,
"learning_rate": 0.0005607789729787294,
"loss": 2.054,
"step": 1041
},
{
"epoch": 0.47792684325192064,
"grad_norm": 0.21555177867412567,
"learning_rate": 0.0005600413716847564,
"loss": 2.0425,
"step": 1042
},
{
"epoch": 0.47838550624928333,
"grad_norm": 0.22350367903709412,
"learning_rate": 0.000559303637791766,
"loss": 2.0914,
"step": 1043
},
{
"epoch": 0.478844169246646,
"grad_norm": 0.2171928733587265,
"learning_rate": 0.0005585657729290151,
"loss": 2.012,
"step": 1044
},
{
"epoch": 0.4793028322440087,
"grad_norm": 0.21663489937782288,
"learning_rate": 0.000557827778726049,
"loss": 2.0375,
"step": 1045
},
{
"epoch": 0.4797614952413714,
"grad_norm": 0.20526684820652008,
"learning_rate": 0.0005570896568126993,
"loss": 2.0327,
"step": 1046
},
{
"epoch": 0.4802201582387341,
"grad_norm": 0.20396284759044647,
"learning_rate": 0.0005563514088190788,
"loss": 1.9856,
"step": 1047
},
{
"epoch": 0.4806788212360968,
"grad_norm": 0.20162101089954376,
"learning_rate": 0.0005556130363755798,
"loss": 2.0684,
"step": 1048
},
{
"epoch": 0.4811374842334595,
"grad_norm": 0.19466564059257507,
"learning_rate": 0.0005548745411128688,
"loss": 1.9256,
"step": 1049
},
{
"epoch": 0.48159614723082217,
"grad_norm": 0.21262109279632568,
"learning_rate": 0.0005541359246618835,
"loss": 2.0396,
"step": 1050
},
{
"epoch": 0.48205481022818486,
"grad_norm": 0.2002846896648407,
"learning_rate": 0.0005533971886538293,
"loss": 2.0645,
"step": 1051
},
{
"epoch": 0.48251347322554755,
"grad_norm": 0.2098289281129837,
"learning_rate": 0.000552658334720176,
"loss": 2.0202,
"step": 1052
},
{
"epoch": 0.48297213622291024,
"grad_norm": 0.2060837596654892,
"learning_rate": 0.0005519193644926535,
"loss": 1.9925,
"step": 1053
},
{
"epoch": 0.4834307992202729,
"grad_norm": 0.20691309869289398,
"learning_rate": 0.0005511802796032485,
"loss": 2.0584,
"step": 1054
},
{
"epoch": 0.48388946221763557,
"grad_norm": 0.2044489085674286,
"learning_rate": 0.0005504410816842009,
"loss": 2.0342,
"step": 1055
},
{
"epoch": 0.48434812521499826,
"grad_norm": 0.20181076228618622,
"learning_rate": 0.0005497017723680009,
"loss": 2.0224,
"step": 1056
},
{
"epoch": 0.48480678821236095,
"grad_norm": 0.19456301629543304,
"learning_rate": 0.0005489623532873836,
"loss": 2.016,
"step": 1057
},
{
"epoch": 0.48526545120972364,
"grad_norm": 0.20180925726890564,
"learning_rate": 0.0005482228260753273,
"loss": 2.0261,
"step": 1058
},
{
"epoch": 0.48572411420708633,
"grad_norm": 0.1942896693944931,
"learning_rate": 0.0005474831923650488,
"loss": 2.0611,
"step": 1059
},
{
"epoch": 0.486182777204449,
"grad_norm": 0.2099979966878891,
"learning_rate": 0.00054674345379,
"loss": 1.9939,
"step": 1060
},
{
"epoch": 0.4866414402018117,
"grad_norm": 0.2163979411125183,
"learning_rate": 0.000546003611983865,
"loss": 2.044,
"step": 1061
},
{
"epoch": 0.4871001031991744,
"grad_norm": 0.21241512894630432,
"learning_rate": 0.0005452636685805552,
"loss": 2.0553,
"step": 1062
},
{
"epoch": 0.4875587661965371,
"grad_norm": 0.2004908174276352,
"learning_rate": 0.0005445236252142066,
"loss": 1.9537,
"step": 1063
},
{
"epoch": 0.4880174291938998,
"grad_norm": 0.20942574739456177,
"learning_rate": 0.000543783483519176,
"loss": 2.02,
"step": 1064
},
{
"epoch": 0.4884760921912625,
"grad_norm": 0.20669794082641602,
"learning_rate": 0.0005430432451300374,
"loss": 2.0725,
"step": 1065
},
{
"epoch": 0.48893475518862517,
"grad_norm": 0.2212553322315216,
"learning_rate": 0.0005423029116815781,
"loss": 2.0184,
"step": 1066
},
{
"epoch": 0.48939341818598786,
"grad_norm": 0.2208898961544037,
"learning_rate": 0.0005415624848087959,
"loss": 1.9851,
"step": 1067
},
{
"epoch": 0.48985208118335055,
"grad_norm": 0.21011777222156525,
"learning_rate": 0.000540821966146894,
"loss": 2.0944,
"step": 1068
},
{
"epoch": 0.49031074418071324,
"grad_norm": 0.20529507100582123,
"learning_rate": 0.0005400813573312793,
"loss": 2.0075,
"step": 1069
},
{
"epoch": 0.49076940717807593,
"grad_norm": 0.2063722163438797,
"learning_rate": 0.0005393406599975572,
"loss": 1.987,
"step": 1070
},
{
"epoch": 0.49122807017543857,
"grad_norm": 0.20710569620132446,
"learning_rate": 0.0005385998757815287,
"loss": 2.0623,
"step": 1071
},
{
"epoch": 0.49168673317280126,
"grad_norm": 0.1980239748954773,
"learning_rate": 0.0005378590063191867,
"loss": 1.9984,
"step": 1072
},
{
"epoch": 0.49214539617016395,
"grad_norm": 0.19994203746318817,
"learning_rate": 0.0005371180532467124,
"loss": 1.973,
"step": 1073
},
{
"epoch": 0.49260405916752664,
"grad_norm": 0.20142818987369537,
"learning_rate": 0.000536377018200472,
"loss": 1.9992,
"step": 1074
},
{
"epoch": 0.49306272216488933,
"grad_norm": 0.20577600598335266,
"learning_rate": 0.0005356359028170118,
"loss": 2.0192,
"step": 1075
},
{
"epoch": 0.493521385162252,
"grad_norm": 0.22016586363315582,
"learning_rate": 0.0005348947087330564,
"loss": 2.0591,
"step": 1076
},
{
"epoch": 0.4939800481596147,
"grad_norm": 0.21496334671974182,
"learning_rate": 0.0005341534375855037,
"loss": 2.0895,
"step": 1077
},
{
"epoch": 0.4944387111569774,
"grad_norm": 0.20091217756271362,
"learning_rate": 0.0005334120910114222,
"loss": 1.9661,
"step": 1078
},
{
"epoch": 0.4948973741543401,
"grad_norm": 0.1933366060256958,
"learning_rate": 0.0005326706706480467,
"loss": 2.0331,
"step": 1079
},
{
"epoch": 0.4953560371517028,
"grad_norm": 0.2161436676979065,
"learning_rate": 0.0005319291781327749,
"loss": 2.021,
"step": 1080
},
{
"epoch": 0.4958147001490655,
"grad_norm": 0.22165721654891968,
"learning_rate": 0.0005311876151031642,
"loss": 2.0168,
"step": 1081
},
{
"epoch": 0.49627336314642817,
"grad_norm": 0.2062569260597229,
"learning_rate": 0.0005304459831969274,
"loss": 1.9795,
"step": 1082
},
{
"epoch": 0.49673202614379086,
"grad_norm": 0.19997736811637878,
"learning_rate": 0.0005297042840519294,
"loss": 2.0233,
"step": 1083
},
{
"epoch": 0.49719068914115355,
"grad_norm": 0.2274743616580963,
"learning_rate": 0.0005289625193061838,
"loss": 2.0486,
"step": 1084
},
{
"epoch": 0.49764935213851624,
"grad_norm": 0.20121794939041138,
"learning_rate": 0.0005282206905978489,
"loss": 2.0245,
"step": 1085
},
{
"epoch": 0.49810801513587893,
"grad_norm": 0.2135496884584427,
"learning_rate": 0.0005274787995652246,
"loss": 2.0668,
"step": 1086
},
{
"epoch": 0.4985666781332416,
"grad_norm": 0.2224859744310379,
"learning_rate": 0.000526736847846748,
"loss": 2.0063,
"step": 1087
},
{
"epoch": 0.49902534113060426,
"grad_norm": 0.21013419330120087,
"learning_rate": 0.0005259948370809901,
"loss": 1.9775,
"step": 1088
},
{
"epoch": 0.49948400412796695,
"grad_norm": 0.20106887817382812,
"learning_rate": 0.0005252527689066533,
"loss": 2.0107,
"step": 1089
},
{
"epoch": 0.49994266712532964,
"grad_norm": 0.20846565067768097,
"learning_rate": 0.0005245106449625654,
"loss": 2.0656,
"step": 1090
},
{
"epoch": 0.5004013301226924,
"grad_norm": 0.20039407908916473,
"learning_rate": 0.0005237684668876785,
"loss": 2.0373,
"step": 1091
},
{
"epoch": 0.500859993120055,
"grad_norm": 0.21051070094108582,
"learning_rate": 0.0005230262363210637,
"loss": 1.9537,
"step": 1092
},
{
"epoch": 0.5013186561174178,
"grad_norm": 0.20225124061107635,
"learning_rate": 0.0005222839549019079,
"loss": 1.9981,
"step": 1093
},
{
"epoch": 0.5017773191147804,
"grad_norm": 0.2149350494146347,
"learning_rate": 0.0005215416242695108,
"loss": 1.9725,
"step": 1094
},
{
"epoch": 0.5022359821121432,
"grad_norm": 0.20242740213871002,
"learning_rate": 0.0005207992460632804,
"loss": 1.9495,
"step": 1095
},
{
"epoch": 0.5026946451095058,
"grad_norm": 0.21469691395759583,
"learning_rate": 0.0005200568219227299,
"loss": 2.0253,
"step": 1096
},
{
"epoch": 0.5031533081068684,
"grad_norm": 0.21568438410758972,
"learning_rate": 0.000519314353487474,
"loss": 2.1262,
"step": 1097
},
{
"epoch": 0.5036119711042312,
"grad_norm": 0.20869800448417664,
"learning_rate": 0.0005185718423972251,
"loss": 2.0083,
"step": 1098
},
{
"epoch": 0.5040706341015938,
"grad_norm": 0.21569029986858368,
"learning_rate": 0.0005178292902917898,
"loss": 2.0074,
"step": 1099
},
{
"epoch": 0.5045292970989566,
"grad_norm": 0.2032281756401062,
"learning_rate": 0.0005170866988110656,
"loss": 1.9521,
"step": 1100
},
{
"epoch": 0.5049879600963192,
"grad_norm": 0.2032438963651657,
"learning_rate": 0.0005163440695950362,
"loss": 2.0638,
"step": 1101
},
{
"epoch": 0.5054466230936819,
"grad_norm": 0.2015044242143631,
"learning_rate": 0.0005156014042837695,
"loss": 1.9825,
"step": 1102
},
{
"epoch": 0.5059052860910446,
"grad_norm": 0.19384567439556122,
"learning_rate": 0.0005148587045174128,
"loss": 1.9959,
"step": 1103
},
{
"epoch": 0.5063639490884073,
"grad_norm": 0.19851745665073395,
"learning_rate": 0.0005141159719361891,
"loss": 2.0189,
"step": 1104
},
{
"epoch": 0.50682261208577,
"grad_norm": 0.20046968758106232,
"learning_rate": 0.0005133732081803945,
"loss": 1.9387,
"step": 1105
},
{
"epoch": 0.5072812750831327,
"grad_norm": 0.20742172002792358,
"learning_rate": 0.0005126304148903936,
"loss": 1.9977,
"step": 1106
},
{
"epoch": 0.5077399380804953,
"grad_norm": 0.21619586646556854,
"learning_rate": 0.0005118875937066161,
"loss": 2.0159,
"step": 1107
},
{
"epoch": 0.5081986010778581,
"grad_norm": 0.2112969160079956,
"learning_rate": 0.0005111447462695537,
"loss": 1.9757,
"step": 1108
},
{
"epoch": 0.5086572640752207,
"grad_norm": 0.21187366545200348,
"learning_rate": 0.0005104018742197557,
"loss": 1.95,
"step": 1109
},
{
"epoch": 0.5091159270725835,
"grad_norm": 0.20676307380199432,
"learning_rate": 0.0005096589791978261,
"loss": 1.9401,
"step": 1110
},
{
"epoch": 0.5095745900699461,
"grad_norm": 0.2058505266904831,
"learning_rate": 0.0005089160628444192,
"loss": 2.0099,
"step": 1111
},
{
"epoch": 0.5100332530673088,
"grad_norm": 0.20735453069210052,
"learning_rate": 0.0005081731268002371,
"loss": 1.9987,
"step": 1112
},
{
"epoch": 0.5104919160646715,
"grad_norm": 0.2034836709499359,
"learning_rate": 0.0005074301727060243,
"loss": 2.0591,
"step": 1113
},
{
"epoch": 0.5109505790620341,
"grad_norm": 0.21008765697479248,
"learning_rate": 0.0005066872022025663,
"loss": 2.0182,
"step": 1114
},
{
"epoch": 0.5114092420593969,
"grad_norm": 0.21602506935596466,
"learning_rate": 0.0005059442169306844,
"loss": 1.9987,
"step": 1115
},
{
"epoch": 0.5118679050567595,
"grad_norm": 0.207993745803833,
"learning_rate": 0.0005052012185312321,
"loss": 1.9227,
"step": 1116
},
{
"epoch": 0.5123265680541222,
"grad_norm": 0.20328985154628754,
"learning_rate": 0.0005044582086450925,
"loss": 2.0078,
"step": 1117
},
{
"epoch": 0.5127852310514849,
"grad_norm": 0.20380139350891113,
"learning_rate": 0.0005037151889131737,
"loss": 1.9571,
"step": 1118
},
{
"epoch": 0.5132438940488476,
"grad_norm": 0.21732039749622345,
"learning_rate": 0.0005029721609764059,
"loss": 2.0089,
"step": 1119
},
{
"epoch": 0.5137025570462103,
"grad_norm": 0.21195349097251892,
"learning_rate": 0.000502229126475737,
"loss": 2.0683,
"step": 1120
},
{
"epoch": 0.514161220043573,
"grad_norm": 0.21738679707050323,
"learning_rate": 0.0005014860870521293,
"loss": 1.9849,
"step": 1121
},
{
"epoch": 0.5146198830409356,
"grad_norm": 0.2020874172449112,
"learning_rate": 0.0005007430443465569,
"loss": 1.9579,
"step": 1122
},
{
"epoch": 0.5150785460382984,
"grad_norm": 0.21757736802101135,
"learning_rate": 0.0005,
"loss": 1.9968,
"step": 1123
},
{
"epoch": 0.515537209035661,
"grad_norm": 0.21239666640758514,
"learning_rate": 0.0004992569556534432,
"loss": 2.0172,
"step": 1124
},
{
"epoch": 0.5159958720330238,
"grad_norm": 0.22115150094032288,
"learning_rate": 0.0004985139129478707,
"loss": 1.9859,
"step": 1125
},
{
"epoch": 0.5164545350303864,
"grad_norm": 0.20106226205825806,
"learning_rate": 0.0004977708735242633,
"loss": 2.0537,
"step": 1126
},
{
"epoch": 0.5169131980277492,
"grad_norm": 0.2069558948278427,
"learning_rate": 0.0004970278390235942,
"loss": 1.982,
"step": 1127
},
{
"epoch": 0.5173718610251118,
"grad_norm": 0.22248685359954834,
"learning_rate": 0.0004962848110868262,
"loss": 1.943,
"step": 1128
},
{
"epoch": 0.5178305240224745,
"grad_norm": 0.22086036205291748,
"learning_rate": 0.0004955417913549074,
"loss": 2.0415,
"step": 1129
},
{
"epoch": 0.5182891870198372,
"grad_norm": 0.20855140686035156,
"learning_rate": 0.0004947987814687679,
"loss": 1.9992,
"step": 1130
},
{
"epoch": 0.5187478500171998,
"grad_norm": 0.21061989665031433,
"learning_rate": 0.0004940557830693157,
"loss": 2.0295,
"step": 1131
},
{
"epoch": 0.5192065130145626,
"grad_norm": 0.21608461439609528,
"learning_rate": 0.0004933127977974338,
"loss": 2.0713,
"step": 1132
},
{
"epoch": 0.5196651760119252,
"grad_norm": 0.2012651264667511,
"learning_rate": 0.0004925698272939757,
"loss": 1.986,
"step": 1133
},
{
"epoch": 0.5201238390092879,
"grad_norm": 0.19969522953033447,
"learning_rate": 0.0004918268731997632,
"loss": 1.9692,
"step": 1134
},
{
"epoch": 0.5205825020066506,
"grad_norm": 0.20453834533691406,
"learning_rate": 0.0004910839371555809,
"loss": 2.0523,
"step": 1135
},
{
"epoch": 0.5210411650040133,
"grad_norm": 0.217530757188797,
"learning_rate": 0.0004903410208021739,
"loss": 2.0396,
"step": 1136
},
{
"epoch": 0.521499828001376,
"grad_norm": 0.20996949076652527,
"learning_rate": 0.0004895981257802443,
"loss": 2.0312,
"step": 1137
},
{
"epoch": 0.5219584909987387,
"grad_norm": 0.20881879329681396,
"learning_rate": 0.0004888552537304463,
"loss": 2.0498,
"step": 1138
},
{
"epoch": 0.5224171539961013,
"grad_norm": 0.20794060826301575,
"learning_rate": 0.00048811240629338394,
"loss": 2.0014,
"step": 1139
},
{
"epoch": 0.5228758169934641,
"grad_norm": 0.20206283032894135,
"learning_rate": 0.00048736958510960663,
"loss": 2.0821,
"step": 1140
},
{
"epoch": 0.5233344799908267,
"grad_norm": 0.21406972408294678,
"learning_rate": 0.00048662679181960564,
"loss": 1.8903,
"step": 1141
},
{
"epoch": 0.5237931429881895,
"grad_norm": 0.20857201516628265,
"learning_rate": 0.00048588402806381094,
"loss": 2.0284,
"step": 1142
},
{
"epoch": 0.5242518059855521,
"grad_norm": 0.22076819837093353,
"learning_rate": 0.0004851412954825874,
"loss": 2.0306,
"step": 1143
},
{
"epoch": 0.5247104689829148,
"grad_norm": 0.2162686586380005,
"learning_rate": 0.00048439859571623034,
"loss": 2.0171,
"step": 1144
},
{
"epoch": 0.5251691319802775,
"grad_norm": 0.21557502448558807,
"learning_rate": 0.00048365593040496373,
"loss": 1.9996,
"step": 1145
},
{
"epoch": 0.5256277949776402,
"grad_norm": 0.2067699134349823,
"learning_rate": 0.00048291330118893443,
"loss": 2.0231,
"step": 1146
},
{
"epoch": 0.5260864579750029,
"grad_norm": 0.21602368354797363,
"learning_rate": 0.0004821707097082102,
"loss": 2.01,
"step": 1147
},
{
"epoch": 0.5265451209723655,
"grad_norm": 0.21548214554786682,
"learning_rate": 0.0004814281576027749,
"loss": 2.0379,
"step": 1148
},
{
"epoch": 0.5270037839697282,
"grad_norm": 0.21128974854946136,
"learning_rate": 0.000480685646512526,
"loss": 1.9763,
"step": 1149
},
{
"epoch": 0.5274624469670909,
"grad_norm": 0.19809399545192719,
"learning_rate": 0.00047994317807727025,
"loss": 1.9911,
"step": 1150
},
{
"epoch": 0.5279211099644536,
"grad_norm": 0.19243529438972473,
"learning_rate": 0.00047920075393671974,
"loss": 1.9954,
"step": 1151
},
{
"epoch": 0.5283797729618163,
"grad_norm": 0.20856811106204987,
"learning_rate": 0.0004784583757304893,
"loss": 2.0613,
"step": 1152
},
{
"epoch": 0.528838435959179,
"grad_norm": 0.20869861543178558,
"learning_rate": 0.00047771604509809214,
"loss": 2.0753,
"step": 1153
},
{
"epoch": 0.5292970989565416,
"grad_norm": 0.208896666765213,
"learning_rate": 0.0004769737636789364,
"loss": 2.0241,
"step": 1154
},
{
"epoch": 0.5297557619539044,
"grad_norm": 0.20907920598983765,
"learning_rate": 0.00047623153311232157,
"loss": 2.0662,
"step": 1155
},
{
"epoch": 0.530214424951267,
"grad_norm": 0.2107512652873993,
"learning_rate": 0.0004754893550374346,
"loss": 1.9947,
"step": 1156
},
{
"epoch": 0.5306730879486298,
"grad_norm": 0.2072741538286209,
"learning_rate": 0.00047474723109334685,
"loss": 1.9573,
"step": 1157
},
{
"epoch": 0.5311317509459924,
"grad_norm": 0.21090155839920044,
"learning_rate": 0.00047400516291900993,
"loss": 1.9681,
"step": 1158
},
{
"epoch": 0.5315904139433552,
"grad_norm": 0.22389593720436096,
"learning_rate": 0.0004732631521532522,
"loss": 1.9966,
"step": 1159
},
{
"epoch": 0.5320490769407178,
"grad_norm": 0.20400306582450867,
"learning_rate": 0.0004725212004347755,
"loss": 1.9504,
"step": 1160
},
{
"epoch": 0.5325077399380805,
"grad_norm": 0.20563088357448578,
"learning_rate": 0.00047177930940215095,
"loss": 1.9387,
"step": 1161
},
{
"epoch": 0.5329664029354432,
"grad_norm": 0.20989938080310822,
"learning_rate": 0.00047103748069381624,
"loss": 1.928,
"step": 1162
},
{
"epoch": 0.5334250659328059,
"grad_norm": 0.19796213507652283,
"learning_rate": 0.0004702957159480707,
"loss": 1.9708,
"step": 1163
},
{
"epoch": 0.5338837289301686,
"grad_norm": 0.20376820862293243,
"learning_rate": 0.00046955401680307267,
"loss": 2.004,
"step": 1164
},
{
"epoch": 0.5343423919275313,
"grad_norm": 0.20358070731163025,
"learning_rate": 0.0004688123848968359,
"loss": 2.0005,
"step": 1165
},
{
"epoch": 0.5348010549248939,
"grad_norm": 0.2075379192829132,
"learning_rate": 0.00046807082186722516,
"loss": 1.9737,
"step": 1166
},
{
"epoch": 0.5352597179222566,
"grad_norm": 0.20849154889583588,
"learning_rate": 0.0004673293293519535,
"loss": 1.955,
"step": 1167
},
{
"epoch": 0.5357183809196193,
"grad_norm": 0.20184233784675598,
"learning_rate": 0.00046658790898857806,
"loss": 2.0566,
"step": 1168
},
{
"epoch": 0.536177043916982,
"grad_norm": 0.20917584002017975,
"learning_rate": 0.0004658465624144963,
"loss": 1.9512,
"step": 1169
},
{
"epoch": 0.5366357069143447,
"grad_norm": 0.21120724081993103,
"learning_rate": 0.0004651052912669438,
"loss": 2.0463,
"step": 1170
},
{
"epoch": 0.5370943699117073,
"grad_norm": 0.19984586536884308,
"learning_rate": 0.0004643640971829883,
"loss": 2.1199,
"step": 1171
},
{
"epoch": 0.5375530329090701,
"grad_norm": 0.19981712102890015,
"learning_rate": 0.0004636229817995281,
"loss": 1.9034,
"step": 1172
},
{
"epoch": 0.5380116959064327,
"grad_norm": 0.1934569925069809,
"learning_rate": 0.0004628819467532876,
"loss": 1.9219,
"step": 1173
},
{
"epoch": 0.5384703589037955,
"grad_norm": 0.20771817862987518,
"learning_rate": 0.00046214099368081335,
"loss": 2.0264,
"step": 1174
},
{
"epoch": 0.5389290219011581,
"grad_norm": 0.20519842207431793,
"learning_rate": 0.0004614001242184714,
"loss": 1.9854,
"step": 1175
},
{
"epoch": 0.5393876848985208,
"grad_norm": 0.21878182888031006,
"learning_rate": 0.000460659340002443,
"loss": 1.9949,
"step": 1176
},
{
"epoch": 0.5398463478958835,
"grad_norm": 0.20158037543296814,
"learning_rate": 0.00045991864266872073,
"loss": 2.0082,
"step": 1177
},
{
"epoch": 0.5403050108932462,
"grad_norm": 0.20829997956752777,
"learning_rate": 0.00045917803385310595,
"loss": 2.0391,
"step": 1178
},
{
"epoch": 0.5407636738906089,
"grad_norm": 0.19806355237960815,
"learning_rate": 0.00045843751519120417,
"loss": 2.0045,
"step": 1179
},
{
"epoch": 0.5412223368879716,
"grad_norm": 0.2147803157567978,
"learning_rate": 0.00045769708831842193,
"loss": 2.0747,
"step": 1180
},
{
"epoch": 0.5416809998853342,
"grad_norm": 0.21004463732242584,
"learning_rate": 0.00045695675486996266,
"loss": 1.981,
"step": 1181
},
{
"epoch": 0.542139662882697,
"grad_norm": 0.21967051923274994,
"learning_rate": 0.00045621651648082405,
"loss": 2.013,
"step": 1182
},
{
"epoch": 0.5425983258800596,
"grad_norm": 0.2070906013250351,
"learning_rate": 0.00045547637478579356,
"loss": 1.9597,
"step": 1183
},
{
"epoch": 0.5430569888774223,
"grad_norm": 0.20285610854625702,
"learning_rate": 0.0004547363314194449,
"loss": 1.9927,
"step": 1184
},
{
"epoch": 0.543515651874785,
"grad_norm": 0.197243794798851,
"learning_rate": 0.000453996388016135,
"loss": 2.0423,
"step": 1185
},
{
"epoch": 0.5439743148721476,
"grad_norm": 0.19798541069030762,
"learning_rate": 0.0004532565462099999,
"loss": 1.9869,
"step": 1186
},
{
"epoch": 0.5444329778695104,
"grad_norm": 0.21351701021194458,
"learning_rate": 0.0004525168076349513,
"loss": 2.0077,
"step": 1187
},
{
"epoch": 0.544891640866873,
"grad_norm": 0.21316753327846527,
"learning_rate": 0.0004517771739246729,
"loss": 2.053,
"step": 1188
},
{
"epoch": 0.5453503038642358,
"grad_norm": 0.2068125307559967,
"learning_rate": 0.0004510376467126165,
"loss": 2.0333,
"step": 1189
},
{
"epoch": 0.5458089668615984,
"grad_norm": 0.2003355324268341,
"learning_rate": 0.0004502982276319992,
"loss": 2.0192,
"step": 1190
},
{
"epoch": 0.5462676298589612,
"grad_norm": 0.20878131687641144,
"learning_rate": 0.0004495589183157991,
"loss": 2.0102,
"step": 1191
},
{
"epoch": 0.5467262928563238,
"grad_norm": 0.21817845106124878,
"learning_rate": 0.0004488197203967517,
"loss": 1.9792,
"step": 1192
},
{
"epoch": 0.5471849558536865,
"grad_norm": 0.19285914301872253,
"learning_rate": 0.0004480806355073467,
"loss": 2.0412,
"step": 1193
},
{
"epoch": 0.5476436188510492,
"grad_norm": 0.20974747836589813,
"learning_rate": 0.000447341665279824,
"loss": 2.0189,
"step": 1194
},
{
"epoch": 0.5481022818484119,
"grad_norm": 0.2169572114944458,
"learning_rate": 0.0004466028113461708,
"loss": 2.0331,
"step": 1195
},
{
"epoch": 0.5485609448457746,
"grad_norm": 0.19933925569057465,
"learning_rate": 0.0004458640753381167,
"loss": 2.1035,
"step": 1196
},
{
"epoch": 0.5490196078431373,
"grad_norm": 0.19525569677352905,
"learning_rate": 0.0004451254588871313,
"loss": 1.9714,
"step": 1197
},
{
"epoch": 0.5494782708404999,
"grad_norm": 0.20419354736804962,
"learning_rate": 0.0004443869636244203,
"loss": 2.0221,
"step": 1198
},
{
"epoch": 0.5499369338378627,
"grad_norm": 0.2147618532180786,
"learning_rate": 0.0004436485911809212,
"loss": 2.0126,
"step": 1199
},
{
"epoch": 0.5503955968352253,
"grad_norm": 0.213330939412117,
"learning_rate": 0.00044291034318730087,
"loss": 1.9941,
"step": 1200
},
{
"epoch": 0.550854259832588,
"grad_norm": 0.21510273218154907,
"learning_rate": 0.0004421722212739511,
"loss": 2.0446,
"step": 1201
},
{
"epoch": 0.5513129228299507,
"grad_norm": 0.203706294298172,
"learning_rate": 0.0004414342270709848,
"loss": 1.9948,
"step": 1202
},
{
"epoch": 0.5517715858273133,
"grad_norm": 0.22854863107204437,
"learning_rate": 0.00044069636220823397,
"loss": 2.0027,
"step": 1203
},
{
"epoch": 0.5522302488246761,
"grad_norm": 0.1990957260131836,
"learning_rate": 0.0004399586283152437,
"loss": 2.0111,
"step": 1204
},
{
"epoch": 0.5526889118220387,
"grad_norm": 0.20054443180561066,
"learning_rate": 0.0004392210270212706,
"loss": 2.0476,
"step": 1205
},
{
"epoch": 0.5531475748194015,
"grad_norm": 0.19955506920814514,
"learning_rate": 0.00043848355995527825,
"loss": 1.9695,
"step": 1206
},
{
"epoch": 0.5536062378167641,
"grad_norm": 0.1930820345878601,
"learning_rate": 0.00043774622874593374,
"loss": 1.9651,
"step": 1207
},
{
"epoch": 0.5540649008141268,
"grad_norm": 0.19619160890579224,
"learning_rate": 0.000437009035021604,
"loss": 2.0851,
"step": 1208
},
{
"epoch": 0.5545235638114895,
"grad_norm": 0.21294547617435455,
"learning_rate": 0.00043627198041035274,
"loss": 2.0088,
"step": 1209
},
{
"epoch": 0.5549822268088522,
"grad_norm": 0.20454221963882446,
"learning_rate": 0.00043553506653993597,
"loss": 1.9708,
"step": 1210
},
{
"epoch": 0.5554408898062149,
"grad_norm": 0.2073518931865692,
"learning_rate": 0.0004347982950377992,
"loss": 1.9987,
"step": 1211
},
{
"epoch": 0.5558995528035776,
"grad_norm": 0.21025371551513672,
"learning_rate": 0.0004340616675310735,
"loss": 2.0562,
"step": 1212
},
{
"epoch": 0.5563582158009402,
"grad_norm": 0.21760322153568268,
"learning_rate": 0.00043332518564657193,
"loss": 1.968,
"step": 1213
},
{
"epoch": 0.556816878798303,
"grad_norm": 0.22314049303531647,
"learning_rate": 0.0004325888510107856,
"loss": 1.9412,
"step": 1214
},
{
"epoch": 0.5572755417956656,
"grad_norm": 0.21836397051811218,
"learning_rate": 0.0004318526652498809,
"loss": 2.0537,
"step": 1215
},
{
"epoch": 0.5577342047930284,
"grad_norm": 0.21008555591106415,
"learning_rate": 0.00043111662998969523,
"loss": 2.0117,
"step": 1216
},
{
"epoch": 0.558192867790391,
"grad_norm": 0.20476774871349335,
"learning_rate": 0.0004303807468557335,
"loss": 2.0566,
"step": 1217
},
{
"epoch": 0.5586515307877536,
"grad_norm": 0.2015518993139267,
"learning_rate": 0.0004296450174731648,
"loss": 1.9727,
"step": 1218
},
{
"epoch": 0.5591101937851164,
"grad_norm": 0.2051452398300171,
"learning_rate": 0.0004289094434668188,
"loss": 1.9774,
"step": 1219
},
{
"epoch": 0.559568856782479,
"grad_norm": 0.20334461331367493,
"learning_rate": 0.00042817402646118185,
"loss": 2.009,
"step": 1220
},
{
"epoch": 0.5600275197798418,
"grad_norm": 0.2029939591884613,
"learning_rate": 0.0004274387680803936,
"loss": 1.9631,
"step": 1221
},
{
"epoch": 0.5604861827772044,
"grad_norm": 0.2030264288187027,
"learning_rate": 0.00042670366994824327,
"loss": 1.9059,
"step": 1222
},
{
"epoch": 0.5609448457745672,
"grad_norm": 0.21108238399028778,
"learning_rate": 0.0004259687336881663,
"loss": 2.072,
"step": 1223
},
{
"epoch": 0.5614035087719298,
"grad_norm": 0.2084043025970459,
"learning_rate": 0.0004252339609232408,
"loss": 1.9922,
"step": 1224
},
{
"epoch": 0.5618621717692925,
"grad_norm": 0.21951663494110107,
"learning_rate": 0.0004244993532761834,
"loss": 1.9411,
"step": 1225
},
{
"epoch": 0.5623208347666552,
"grad_norm": 0.20454728603363037,
"learning_rate": 0.00042376491236934634,
"loss": 1.9346,
"step": 1226
},
{
"epoch": 0.5627794977640179,
"grad_norm": 0.21056565642356873,
"learning_rate": 0.0004230306398247136,
"loss": 2.044,
"step": 1227
},
{
"epoch": 0.5632381607613806,
"grad_norm": 0.21455536782741547,
"learning_rate": 0.0004222965372638976,
"loss": 2.0178,
"step": 1228
},
{
"epoch": 0.5636968237587433,
"grad_norm": 0.2017064392566681,
"learning_rate": 0.0004215626063081348,
"loss": 2.0616,
"step": 1229
},
{
"epoch": 0.5641554867561059,
"grad_norm": 0.19614450633525848,
"learning_rate": 0.000420828848578283,
"loss": 1.9758,
"step": 1230
},
{
"epoch": 0.5646141497534687,
"grad_norm": 0.20560809969902039,
"learning_rate": 0.0004200952656948175,
"loss": 1.9763,
"step": 1231
},
{
"epoch": 0.5650728127508313,
"grad_norm": 0.20582322776317596,
"learning_rate": 0.0004193618592778272,
"loss": 2.054,
"step": 1232
},
{
"epoch": 0.5655314757481941,
"grad_norm": 0.2015935629606247,
"learning_rate": 0.0004186286309470116,
"loss": 2.0523,
"step": 1233
},
{
"epoch": 0.5659901387455567,
"grad_norm": 0.20896977186203003,
"learning_rate": 0.0004178955823216767,
"loss": 2.0635,
"step": 1234
},
{
"epoch": 0.5664488017429193,
"grad_norm": 0.19686485826969147,
"learning_rate": 0.00041716271502073137,
"loss": 1.9839,
"step": 1235
},
{
"epoch": 0.5669074647402821,
"grad_norm": 0.20565646886825562,
"learning_rate": 0.000416430030662685,
"loss": 2.0052,
"step": 1236
},
{
"epoch": 0.5673661277376447,
"grad_norm": 0.19780460000038147,
"learning_rate": 0.00041569753086564173,
"loss": 2.0155,
"step": 1237
},
{
"epoch": 0.5678247907350075,
"grad_norm": 0.20365889370441437,
"learning_rate": 0.0004149652172472988,
"loss": 2.0083,
"step": 1238
},
{
"epoch": 0.5682834537323701,
"grad_norm": 0.19893182814121246,
"learning_rate": 0.00041423309142494234,
"loss": 1.9327,
"step": 1239
},
{
"epoch": 0.5687421167297328,
"grad_norm": 0.2157251536846161,
"learning_rate": 0.0004135011550154433,
"loss": 2.1228,
"step": 1240
},
{
"epoch": 0.5692007797270955,
"grad_norm": 0.2132796347141266,
"learning_rate": 0.0004127694096352546,
"loss": 2.0092,
"step": 1241
},
{
"epoch": 0.5696594427244582,
"grad_norm": 0.20251400768756866,
"learning_rate": 0.00041203785690040743,
"loss": 1.9592,
"step": 1242
},
{
"epoch": 0.5701181057218209,
"grad_norm": 0.20353442430496216,
"learning_rate": 0.00041130649842650694,
"loss": 2.0426,
"step": 1243
},
{
"epoch": 0.5705767687191836,
"grad_norm": 0.20936520397663116,
"learning_rate": 0.00041057533582873016,
"loss": 2.066,
"step": 1244
},
{
"epoch": 0.5710354317165462,
"grad_norm": 0.21438148617744446,
"learning_rate": 0.0004098443707218208,
"loss": 1.9942,
"step": 1245
},
{
"epoch": 0.571494094713909,
"grad_norm": 0.20946115255355835,
"learning_rate": 0.00040911360472008673,
"loss": 1.9166,
"step": 1246
},
{
"epoch": 0.5719527577112716,
"grad_norm": 0.2164328545331955,
"learning_rate": 0.0004083830394373959,
"loss": 2.0581,
"step": 1247
},
{
"epoch": 0.5724114207086344,
"grad_norm": 0.2132098227739334,
"learning_rate": 0.00040765267648717324,
"loss": 2.0299,
"step": 1248
},
{
"epoch": 0.572870083705997,
"grad_norm": 0.20624376833438873,
"learning_rate": 0.00040692251748239677,
"loss": 1.9672,
"step": 1249
},
{
"epoch": 0.5733287467033598,
"grad_norm": 0.21116051077842712,
"learning_rate": 0.00040619256403559383,
"loss": 2.0382,
"step": 1250
},
{
"epoch": 0.5737874097007224,
"grad_norm": 0.22244922816753387,
"learning_rate": 0.000405462817758838,
"loss": 2.0746,
"step": 1251
},
{
"epoch": 0.574246072698085,
"grad_norm": 0.20676471292972565,
"learning_rate": 0.0004047332802637457,
"loss": 1.9516,
"step": 1252
},
{
"epoch": 0.5747047356954478,
"grad_norm": 0.2048780620098114,
"learning_rate": 0.00040400395316147157,
"loss": 1.9481,
"step": 1253
},
{
"epoch": 0.5751633986928104,
"grad_norm": 0.20597441494464874,
"learning_rate": 0.00040327483806270627,
"loss": 1.9264,
"step": 1254
},
{
"epoch": 0.5756220616901732,
"grad_norm": 0.20193727314472198,
"learning_rate": 0.0004025459365776715,
"loss": 2.0503,
"step": 1255
},
{
"epoch": 0.5760807246875358,
"grad_norm": 0.2186981439590454,
"learning_rate": 0.00040181725031611794,
"loss": 2.0186,
"step": 1256
},
{
"epoch": 0.5765393876848985,
"grad_norm": 0.20516179502010345,
"learning_rate": 0.0004010887808873206,
"loss": 2.0233,
"step": 1257
},
{
"epoch": 0.5769980506822612,
"grad_norm": 0.20787815749645233,
"learning_rate": 0.00040036052990007553,
"loss": 1.9524,
"step": 1258
},
{
"epoch": 0.5774567136796239,
"grad_norm": 0.2047126293182373,
"learning_rate": 0.0003996324989626967,
"loss": 2.0247,
"step": 1259
},
{
"epoch": 0.5779153766769866,
"grad_norm": 0.22039343416690826,
"learning_rate": 0.00039890468968301166,
"loss": 1.9596,
"step": 1260
},
{
"epoch": 0.5783740396743493,
"grad_norm": 0.21957522630691528,
"learning_rate": 0.0003981771036683591,
"loss": 2.0455,
"step": 1261
},
{
"epoch": 0.5788327026717119,
"grad_norm": 0.20623785257339478,
"learning_rate": 0.00039744974252558385,
"loss": 1.9086,
"step": 1262
},
{
"epoch": 0.5792913656690747,
"grad_norm": 0.2081158608198166,
"learning_rate": 0.00039672260786103463,
"loss": 1.9761,
"step": 1263
},
{
"epoch": 0.5797500286664373,
"grad_norm": 0.20578722655773163,
"learning_rate": 0.00039599570128055994,
"loss": 1.9275,
"step": 1264
},
{
"epoch": 0.5802086916638001,
"grad_norm": 0.21054768562316895,
"learning_rate": 0.0003952690243895044,
"loss": 2.0512,
"step": 1265
},
{
"epoch": 0.5806673546611627,
"grad_norm": 0.2041298896074295,
"learning_rate": 0.0003945425787927054,
"loss": 1.982,
"step": 1266
},
{
"epoch": 0.5811260176585255,
"grad_norm": 0.21112172305583954,
"learning_rate": 0.00039381636609448975,
"loss": 2.0493,
"step": 1267
},
{
"epoch": 0.5815846806558881,
"grad_norm": 0.20825034379959106,
"learning_rate": 0.0003930903878986693,
"loss": 2.0099,
"step": 1268
},
{
"epoch": 0.5820433436532507,
"grad_norm": 0.204656183719635,
"learning_rate": 0.00039236464580853916,
"loss": 1.9653,
"step": 1269
},
{
"epoch": 0.5825020066506135,
"grad_norm": 0.2098570466041565,
"learning_rate": 0.0003916391414268718,
"loss": 2.0455,
"step": 1270
},
{
"epoch": 0.5829606696479761,
"grad_norm": 0.215850368142128,
"learning_rate": 0.00039091387635591536,
"loss": 2.0703,
"step": 1271
},
{
"epoch": 0.5834193326453389,
"grad_norm": 0.2082929015159607,
"learning_rate": 0.0003901888521973894,
"loss": 2.0137,
"step": 1272
},
{
"epoch": 0.5838779956427015,
"grad_norm": 0.209946870803833,
"learning_rate": 0.0003894640705524813,
"loss": 2.0057,
"step": 1273
},
{
"epoch": 0.5843366586400642,
"grad_norm": 0.1959473192691803,
"learning_rate": 0.00038873953302184284,
"loss": 1.9935,
"step": 1274
},
{
"epoch": 0.5847953216374269,
"grad_norm": 0.21814335882663727,
"learning_rate": 0.000388015241205587,
"loss": 1.9946,
"step": 1275
},
{
"epoch": 0.5852539846347896,
"grad_norm": 0.20848964154720306,
"learning_rate": 0.00038729119670328355,
"loss": 2.0484,
"step": 1276
},
{
"epoch": 0.5857126476321523,
"grad_norm": 0.2035483419895172,
"learning_rate": 0.00038656740111395665,
"loss": 1.9429,
"step": 1277
},
{
"epoch": 0.586171310629515,
"grad_norm": 0.2058047205209732,
"learning_rate": 0.00038584385603608053,
"loss": 2.1297,
"step": 1278
},
{
"epoch": 0.5866299736268776,
"grad_norm": 0.234492227435112,
"learning_rate": 0.00038512056306757615,
"loss": 2.038,
"step": 1279
},
{
"epoch": 0.5870886366242404,
"grad_norm": 0.2273351401090622,
"learning_rate": 0.0003843975238058075,
"loss": 2.0934,
"step": 1280
},
{
"epoch": 0.587547299621603,
"grad_norm": 0.20575948059558868,
"learning_rate": 0.00038367473984757863,
"loss": 1.9727,
"step": 1281
},
{
"epoch": 0.5880059626189658,
"grad_norm": 0.20239883661270142,
"learning_rate": 0.0003829522127891296,
"loss": 1.9211,
"step": 1282
},
{
"epoch": 0.5884646256163284,
"grad_norm": 0.2091120034456253,
"learning_rate": 0.0003822299442261329,
"loss": 2.0209,
"step": 1283
},
{
"epoch": 0.5889232886136911,
"grad_norm": 0.22309300303459167,
"learning_rate": 0.00038150793575369063,
"loss": 1.9744,
"step": 1284
},
{
"epoch": 0.5893819516110538,
"grad_norm": 0.21739032864570618,
"learning_rate": 0.0003807861889663299,
"loss": 1.9757,
"step": 1285
},
{
"epoch": 0.5898406146084164,
"grad_norm": 0.2251415252685547,
"learning_rate": 0.0003800647054580006,
"loss": 2.0466,
"step": 1286
},
{
"epoch": 0.5902992776057792,
"grad_norm": 0.22180496156215668,
"learning_rate": 0.00037934348682207064,
"loss": 1.9468,
"step": 1287
},
{
"epoch": 0.5907579406031418,
"grad_norm": 0.20281916856765747,
"learning_rate": 0.00037862253465132306,
"loss": 2.1013,
"step": 1288
},
{
"epoch": 0.5912166036005045,
"grad_norm": 0.19946052134037018,
"learning_rate": 0.00037790185053795245,
"loss": 1.9095,
"step": 1289
},
{
"epoch": 0.5916752665978672,
"grad_norm": 0.21237879991531372,
"learning_rate": 0.0003771814360735616,
"loss": 1.9122,
"step": 1290
},
{
"epoch": 0.5921339295952299,
"grad_norm": 0.21257384121418,
"learning_rate": 0.00037646129284915755,
"loss": 2.0552,
"step": 1291
},
{
"epoch": 0.5925925925925926,
"grad_norm": 0.20783714950084686,
"learning_rate": 0.00037574142245514825,
"loss": 2.013,
"step": 1292
},
{
"epoch": 0.5930512555899553,
"grad_norm": 0.2123468667268753,
"learning_rate": 0.0003750218264813393,
"loss": 2.0708,
"step": 1293
},
{
"epoch": 0.5935099185873179,
"grad_norm": 0.21564380824565887,
"learning_rate": 0.0003743025065169305,
"loss": 1.9075,
"step": 1294
},
{
"epoch": 0.5939685815846807,
"grad_norm": 0.23723827302455902,
"learning_rate": 0.0003735834641505116,
"loss": 2.0699,
"step": 1295
},
{
"epoch": 0.5944272445820433,
"grad_norm": 0.20259948074817657,
"learning_rate": 0.00037286470097005954,
"loss": 1.9668,
"step": 1296
},
{
"epoch": 0.5948859075794061,
"grad_norm": 0.2082609236240387,
"learning_rate": 0.0003721462185629347,
"loss": 1.9614,
"step": 1297
},
{
"epoch": 0.5953445705767687,
"grad_norm": 0.21516072750091553,
"learning_rate": 0.00037142801851587707,
"loss": 2.0535,
"step": 1298
},
{
"epoch": 0.5958032335741315,
"grad_norm": 0.22151893377304077,
"learning_rate": 0.00037071010241500357,
"loss": 1.9932,
"step": 1299
},
{
"epoch": 0.5962618965714941,
"grad_norm": 0.2197350561618805,
"learning_rate": 0.00036999247184580383,
"loss": 2.0044,
"step": 1300
},
{
"epoch": 0.5967205595688568,
"grad_norm": 0.22681637108325958,
"learning_rate": 0.00036927512839313636,
"loss": 1.9769,
"step": 1301
},
{
"epoch": 0.5971792225662195,
"grad_norm": 0.207859605550766,
"learning_rate": 0.0003685580736412268,
"loss": 1.9961,
"step": 1302
},
{
"epoch": 0.5976378855635821,
"grad_norm": 0.21766677498817444,
"learning_rate": 0.000367841309173662,
"loss": 1.9342,
"step": 1303
},
{
"epoch": 0.5980965485609449,
"grad_norm": 0.23692168295383453,
"learning_rate": 0.0003671248365733883,
"loss": 2.0689,
"step": 1304
},
{
"epoch": 0.5985552115583075,
"grad_norm": 0.22314250469207764,
"learning_rate": 0.0003664086574227075,
"loss": 1.9476,
"step": 1305
},
{
"epoch": 0.5990138745556702,
"grad_norm": 0.20415183901786804,
"learning_rate": 0.000365692773303273,
"loss": 2.0207,
"step": 1306
},
{
"epoch": 0.5994725375530329,
"grad_norm": 0.19884324073791504,
"learning_rate": 0.00036497718579608696,
"loss": 2.0386,
"step": 1307
},
{
"epoch": 0.5999312005503956,
"grad_norm": 0.20280729234218597,
"learning_rate": 0.0003642618964814964,
"loss": 1.9397,
"step": 1308
},
{
"epoch": 0.6003898635477583,
"grad_norm": 0.196305513381958,
"learning_rate": 0.00036354690693918946,
"loss": 2.0551,
"step": 1309
},
{
"epoch": 0.600848526545121,
"grad_norm": 0.20293492078781128,
"learning_rate": 0.00036283221874819284,
"loss": 1.9167,
"step": 1310
},
{
"epoch": 0.6013071895424836,
"grad_norm": 0.20085373520851135,
"learning_rate": 0.0003621178334868672,
"loss": 2.0232,
"step": 1311
},
{
"epoch": 0.6017658525398464,
"grad_norm": 0.2101745903491974,
"learning_rate": 0.00036140375273290476,
"loss": 2.0336,
"step": 1312
},
{
"epoch": 0.602224515537209,
"grad_norm": 0.21659429371356964,
"learning_rate": 0.0003606899780633245,
"loss": 2.0068,
"step": 1313
},
{
"epoch": 0.6026831785345718,
"grad_norm": 0.20987701416015625,
"learning_rate": 0.0003599765110544699,
"loss": 2.0457,
"step": 1314
},
{
"epoch": 0.6031418415319344,
"grad_norm": 0.20808376371860504,
"learning_rate": 0.0003592633532820052,
"loss": 1.9887,
"step": 1315
},
{
"epoch": 0.6036005045292971,
"grad_norm": 0.20558227598667145,
"learning_rate": 0.0003585505063209109,
"loss": 1.9766,
"step": 1316
},
{
"epoch": 0.6040591675266598,
"grad_norm": 0.2228499799966812,
"learning_rate": 0.00035783797174548194,
"loss": 2.0712,
"step": 1317
},
{
"epoch": 0.6045178305240225,
"grad_norm": 0.21406672894954681,
"learning_rate": 0.00035712575112932277,
"loss": 2.0115,
"step": 1318
},
{
"epoch": 0.6049764935213852,
"grad_norm": 0.20450539886951447,
"learning_rate": 0.000356413846045345,
"loss": 1.9665,
"step": 1319
},
{
"epoch": 0.6054351565187478,
"grad_norm": 0.20020528137683868,
"learning_rate": 0.000355702258065763,
"loss": 1.9358,
"step": 1320
},
{
"epoch": 0.6058938195161105,
"grad_norm": 0.21218223869800568,
"learning_rate": 0.0003549909887620909,
"loss": 2.0107,
"step": 1321
},
{
"epoch": 0.6063524825134732,
"grad_norm": 0.20684915781021118,
"learning_rate": 0.00035428003970513914,
"loss": 1.9115,
"step": 1322
},
{
"epoch": 0.6068111455108359,
"grad_norm": 0.2070438116788864,
"learning_rate": 0.00035356941246501085,
"loss": 1.9877,
"step": 1323
},
{
"epoch": 0.6072698085081986,
"grad_norm": 0.21211375296115875,
"learning_rate": 0.0003528591086110984,
"loss": 1.9505,
"step": 1324
},
{
"epoch": 0.6077284715055613,
"grad_norm": 0.20663602650165558,
"learning_rate": 0.00035214912971208,
"loss": 2.042,
"step": 1325
},
{
"epoch": 0.6081871345029239,
"grad_norm": 0.20612460374832153,
"learning_rate": 0.0003514394773359163,
"loss": 1.9961,
"step": 1326
},
{
"epoch": 0.6086457975002867,
"grad_norm": 0.2156996726989746,
"learning_rate": 0.0003507301530498469,
"loss": 2.067,
"step": 1327
},
{
"epoch": 0.6091044604976493,
"grad_norm": 0.20843425393104553,
"learning_rate": 0.00035002115842038646,
"loss": 2.0231,
"step": 1328
},
{
"epoch": 0.6095631234950121,
"grad_norm": 0.2046700268983841,
"learning_rate": 0.00034931249501332195,
"loss": 1.9964,
"step": 1329
},
{
"epoch": 0.6100217864923747,
"grad_norm": 0.20572437345981598,
"learning_rate": 0.00034860416439370885,
"loss": 2.0475,
"step": 1330
},
{
"epoch": 0.6104804494897375,
"grad_norm": 0.20704710483551025,
"learning_rate": 0.0003478961681258674,
"loss": 2.045,
"step": 1331
},
{
"epoch": 0.6109391124871001,
"grad_norm": 0.212172269821167,
"learning_rate": 0.0003471885077733796,
"loss": 1.9708,
"step": 1332
},
{
"epoch": 0.6113977754844628,
"grad_norm": 0.21105238795280457,
"learning_rate": 0.0003464811848990859,
"loss": 2.0149,
"step": 1333
},
{
"epoch": 0.6118564384818255,
"grad_norm": 0.2110895812511444,
"learning_rate": 0.00034577420106508063,
"loss": 1.9961,
"step": 1334
},
{
"epoch": 0.6123151014791882,
"grad_norm": 0.1892816573381424,
"learning_rate": 0.0003450675578327105,
"loss": 1.9233,
"step": 1335
},
{
"epoch": 0.6127737644765509,
"grad_norm": 0.20585507154464722,
"learning_rate": 0.000344361256762569,
"loss": 2.0056,
"step": 1336
},
{
"epoch": 0.6132324274739135,
"grad_norm": 0.22940734028816223,
"learning_rate": 0.00034365529941449456,
"loss": 1.9095,
"step": 1337
},
{
"epoch": 0.6136910904712762,
"grad_norm": 0.20214460790157318,
"learning_rate": 0.0003429496873475664,
"loss": 1.9904,
"step": 1338
},
{
"epoch": 0.6141497534686389,
"grad_norm": 0.21174807846546173,
"learning_rate": 0.0003422444221201009,
"loss": 1.9884,
"step": 1339
},
{
"epoch": 0.6146084164660016,
"grad_norm": 0.21725213527679443,
"learning_rate": 0.0003415395052896487,
"loss": 2.016,
"step": 1340
},
{
"epoch": 0.6150670794633643,
"grad_norm": 0.20423197746276855,
"learning_rate": 0.0003408349384129912,
"loss": 1.9148,
"step": 1341
},
{
"epoch": 0.615525742460727,
"grad_norm": 0.2118552327156067,
"learning_rate": 0.00034013072304613643,
"loss": 2.0211,
"step": 1342
},
{
"epoch": 0.6159844054580896,
"grad_norm": 0.21375763416290283,
"learning_rate": 0.00033942686074431674,
"loss": 1.9824,
"step": 1343
},
{
"epoch": 0.6164430684554524,
"grad_norm": 0.2107381373643875,
"learning_rate": 0.0003387233530619843,
"loss": 2.0416,
"step": 1344
},
{
"epoch": 0.616901731452815,
"grad_norm": 0.2076815664768219,
"learning_rate": 0.0003380202015528084,
"loss": 1.9886,
"step": 1345
},
{
"epoch": 0.6173603944501778,
"grad_norm": 0.2079714983701706,
"learning_rate": 0.0003373174077696715,
"loss": 1.9565,
"step": 1346
},
{
"epoch": 0.6178190574475404,
"grad_norm": 0.20392373204231262,
"learning_rate": 0.0003366149732646661,
"loss": 1.9806,
"step": 1347
},
{
"epoch": 0.6182777204449031,
"grad_norm": 0.20384877920150757,
"learning_rate": 0.00033591289958909143,
"loss": 1.9867,
"step": 1348
},
{
"epoch": 0.6187363834422658,
"grad_norm": 0.2002488374710083,
"learning_rate": 0.00033521118829344954,
"loss": 1.9396,
"step": 1349
},
{
"epoch": 0.6191950464396285,
"grad_norm": 0.19303640723228455,
"learning_rate": 0.0003345098409274423,
"loss": 1.9384,
"step": 1350
},
{
"epoch": 0.6196537094369912,
"grad_norm": 0.20531976222991943,
"learning_rate": 0.00033380885903996796,
"loss": 2.0224,
"step": 1351
},
{
"epoch": 0.6201123724343539,
"grad_norm": 0.20235009491443634,
"learning_rate": 0.00033310824417911766,
"loss": 2.0201,
"step": 1352
},
{
"epoch": 0.6205710354317165,
"grad_norm": 0.20315061509609222,
"learning_rate": 0.00033240799789217184,
"loss": 1.9595,
"step": 1353
},
{
"epoch": 0.6210296984290792,
"grad_norm": 0.19728994369506836,
"learning_rate": 0.00033170812172559694,
"loss": 1.9671,
"step": 1354
},
{
"epoch": 0.6214883614264419,
"grad_norm": 0.2093075066804886,
"learning_rate": 0.000331008617225042,
"loss": 1.9899,
"step": 1355
},
{
"epoch": 0.6219470244238046,
"grad_norm": 0.20084190368652344,
"learning_rate": 0.0003303094859353355,
"loss": 2.0674,
"step": 1356
},
{
"epoch": 0.6224056874211673,
"grad_norm": 0.19820593297481537,
"learning_rate": 0.0003296107294004812,
"loss": 1.917,
"step": 1357
},
{
"epoch": 0.6228643504185299,
"grad_norm": 0.21474476158618927,
"learning_rate": 0.0003289123491636559,
"loss": 1.949,
"step": 1358
},
{
"epoch": 0.6233230134158927,
"grad_norm": 0.20670942962169647,
"learning_rate": 0.00032821434676720443,
"loss": 1.964,
"step": 1359
},
{
"epoch": 0.6237816764132553,
"grad_norm": 0.2037562131881714,
"learning_rate": 0.00032751672375263836,
"loss": 1.9465,
"step": 1360
},
{
"epoch": 0.6242403394106181,
"grad_norm": 0.21639136970043182,
"learning_rate": 0.0003268194816606305,
"loss": 1.9944,
"step": 1361
},
{
"epoch": 0.6246990024079807,
"grad_norm": 0.21622318029403687,
"learning_rate": 0.00032612262203101267,
"loss": 2.0263,
"step": 1362
},
{
"epoch": 0.6251576654053435,
"grad_norm": 0.20640547573566437,
"learning_rate": 0.00032542614640277225,
"loss": 2.0265,
"step": 1363
},
{
"epoch": 0.6256163284027061,
"grad_norm": 0.2101866602897644,
"learning_rate": 0.0003247300563140481,
"loss": 1.9631,
"step": 1364
},
{
"epoch": 0.6260749914000688,
"grad_norm": 0.20869901776313782,
"learning_rate": 0.00032403435330212807,
"loss": 1.9652,
"step": 1365
},
{
"epoch": 0.6265336543974315,
"grad_norm": 0.21136793494224548,
"learning_rate": 0.00032333903890344515,
"loss": 2.1648,
"step": 1366
},
{
"epoch": 0.6269923173947942,
"grad_norm": 0.21378524601459503,
"learning_rate": 0.00032264411465357333,
"loss": 2.0023,
"step": 1367
},
{
"epoch": 0.6274509803921569,
"grad_norm": 0.2143763154745102,
"learning_rate": 0.00032194958208722654,
"loss": 1.9719,
"step": 1368
},
{
"epoch": 0.6279096433895196,
"grad_norm": 0.2064313292503357,
"learning_rate": 0.00032125544273825204,
"loss": 1.9286,
"step": 1369
},
{
"epoch": 0.6283683063868822,
"grad_norm": 0.20829255878925323,
"learning_rate": 0.0003205616981396297,
"loss": 1.9587,
"step": 1370
},
{
"epoch": 0.6288269693842449,
"grad_norm": 0.20449233055114746,
"learning_rate": 0.00031986834982346713,
"loss": 1.9072,
"step": 1371
},
{
"epoch": 0.6292856323816076,
"grad_norm": 0.2113608717918396,
"learning_rate": 0.00031917539932099694,
"loss": 2.0591,
"step": 1372
},
{
"epoch": 0.6297442953789703,
"grad_norm": 0.2123221606016159,
"learning_rate": 0.00031848284816257336,
"loss": 2.0186,
"step": 1373
},
{
"epoch": 0.630202958376333,
"grad_norm": 0.21018560230731964,
"learning_rate": 0.0003177906978776682,
"loss": 1.9494,
"step": 1374
},
{
"epoch": 0.6306616213736956,
"grad_norm": 0.21022385358810425,
"learning_rate": 0.0003170989499948683,
"loss": 1.9482,
"step": 1375
},
{
"epoch": 0.6311202843710584,
"grad_norm": 0.19912466406822205,
"learning_rate": 0.0003164076060418719,
"loss": 2.0165,
"step": 1376
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.23874573409557343,
"learning_rate": 0.000315716667545485,
"loss": 2.0425,
"step": 1377
},
{
"epoch": 0.6320376103657838,
"grad_norm": 0.2215057909488678,
"learning_rate": 0.00031502613603161836,
"loss": 1.9479,
"step": 1378
},
{
"epoch": 0.6324962733631464,
"grad_norm": 0.20776169002056122,
"learning_rate": 0.00031433601302528335,
"loss": 2.0516,
"step": 1379
},
{
"epoch": 0.6329549363605091,
"grad_norm": 0.20869579911231995,
"learning_rate": 0.00031364630005058995,
"loss": 2.0227,
"step": 1380
},
{
"epoch": 0.6334135993578718,
"grad_norm": 0.1989717036485672,
"learning_rate": 0.0003129569986307422,
"loss": 2.0049,
"step": 1381
},
{
"epoch": 0.6338722623552345,
"grad_norm": 0.20962274074554443,
"learning_rate": 0.00031226811028803515,
"loss": 2.0395,
"step": 1382
},
{
"epoch": 0.6343309253525972,
"grad_norm": 0.20178210735321045,
"learning_rate": 0.00031157963654385173,
"loss": 1.9959,
"step": 1383
},
{
"epoch": 0.6347895883499599,
"grad_norm": 0.19995252788066864,
"learning_rate": 0.0003108915789186592,
"loss": 1.9759,
"step": 1384
},
{
"epoch": 0.6352482513473225,
"grad_norm": 0.20821832120418549,
"learning_rate": 0.00031020393893200604,
"loss": 1.8725,
"step": 1385
},
{
"epoch": 0.6357069143446853,
"grad_norm": 0.20003074407577515,
"learning_rate": 0.00030951671810251823,
"loss": 1.9798,
"step": 1386
},
{
"epoch": 0.6361655773420479,
"grad_norm": 0.21108178794384003,
"learning_rate": 0.0003088299179478959,
"loss": 2.0571,
"step": 1387
},
{
"epoch": 0.6366242403394106,
"grad_norm": 0.21367321908473969,
"learning_rate": 0.0003081435399849104,
"loss": 1.9622,
"step": 1388
},
{
"epoch": 0.6370829033367733,
"grad_norm": 0.20223943889141083,
"learning_rate": 0.0003074575857294004,
"loss": 1.946,
"step": 1389
},
{
"epoch": 0.637541566334136,
"grad_norm": 0.20882382988929749,
"learning_rate": 0.0003067720566962691,
"loss": 2.0547,
"step": 1390
},
{
"epoch": 0.6380002293314987,
"grad_norm": 0.2065863460302353,
"learning_rate": 0.0003060869543994806,
"loss": 2.0283,
"step": 1391
},
{
"epoch": 0.6384588923288613,
"grad_norm": 0.21225383877754211,
"learning_rate": 0.0003054022803520562,
"loss": 1.9686,
"step": 1392
},
{
"epoch": 0.6389175553262241,
"grad_norm": 0.19566106796264648,
"learning_rate": 0.0003047180360660721,
"loss": 2.0078,
"step": 1393
},
{
"epoch": 0.6393762183235867,
"grad_norm": 0.21339420974254608,
"learning_rate": 0.00030403422305265475,
"loss": 2.0024,
"step": 1394
},
{
"epoch": 0.6398348813209495,
"grad_norm": 0.20704859495162964,
"learning_rate": 0.0003033508428219785,
"loss": 2.0687,
"step": 1395
},
{
"epoch": 0.6402935443183121,
"grad_norm": 0.22996467351913452,
"learning_rate": 0.00030266789688326184,
"loss": 1.9526,
"step": 1396
},
{
"epoch": 0.6407522073156748,
"grad_norm": 0.22861891984939575,
"learning_rate": 0.00030198538674476393,
"loss": 2.0155,
"step": 1397
},
{
"epoch": 0.6412108703130375,
"grad_norm": 0.21818692982196808,
"learning_rate": 0.00030130331391378185,
"loss": 1.9804,
"step": 1398
},
{
"epoch": 0.6416695333104002,
"grad_norm": 0.2088412046432495,
"learning_rate": 0.0003006216798966468,
"loss": 2.0008,
"step": 1399
},
{
"epoch": 0.6421281963077629,
"grad_norm": 0.20698463916778564,
"learning_rate": 0.00029994048619872034,
"loss": 1.9919,
"step": 1400
},
{
"epoch": 0.6425868593051256,
"grad_norm": 0.2094312608242035,
"learning_rate": 0.0002992597343243927,
"loss": 1.9242,
"step": 1401
},
{
"epoch": 0.6430455223024882,
"grad_norm": 0.20956437289714813,
"learning_rate": 0.0002985794257770773,
"loss": 2.0194,
"step": 1402
},
{
"epoch": 0.643504185299851,
"grad_norm": 0.20087870955467224,
"learning_rate": 0.0002978995620592092,
"loss": 1.9701,
"step": 1403
},
{
"epoch": 0.6439628482972136,
"grad_norm": 0.2229154407978058,
"learning_rate": 0.0002972201446722405,
"loss": 1.9979,
"step": 1404
},
{
"epoch": 0.6444215112945763,
"grad_norm": 0.20975051820278168,
"learning_rate": 0.00029654117511663803,
"loss": 1.9081,
"step": 1405
},
{
"epoch": 0.644880174291939,
"grad_norm": 0.2102442979812622,
"learning_rate": 0.0002958626548918795,
"loss": 2.0222,
"step": 1406
},
{
"epoch": 0.6453388372893016,
"grad_norm": 0.21198633313179016,
"learning_rate": 0.00029518458549645014,
"loss": 2.0049,
"step": 1407
},
{
"epoch": 0.6457975002866644,
"grad_norm": 0.19411787390708923,
"learning_rate": 0.00029450696842783954,
"loss": 1.9965,
"step": 1408
},
{
"epoch": 0.646256163284027,
"grad_norm": 0.210153266787529,
"learning_rate": 0.00029382980518253865,
"loss": 1.9452,
"step": 1409
},
{
"epoch": 0.6467148262813898,
"grad_norm": 0.1977371871471405,
"learning_rate": 0.00029315309725603595,
"loss": 1.9488,
"step": 1410
},
{
"epoch": 0.6471734892787524,
"grad_norm": 0.2054436355829239,
"learning_rate": 0.00029247684614281446,
"loss": 2.0227,
"step": 1411
},
{
"epoch": 0.6476321522761151,
"grad_norm": 0.21129187941551208,
"learning_rate": 0.0002918010533363481,
"loss": 2.0033,
"step": 1412
},
{
"epoch": 0.6480908152734778,
"grad_norm": 0.20427916944026947,
"learning_rate": 0.0002911257203290987,
"loss": 1.9343,
"step": 1413
},
{
"epoch": 0.6485494782708405,
"grad_norm": 0.22113391757011414,
"learning_rate": 0.00029045084861251314,
"loss": 2.0443,
"step": 1414
},
{
"epoch": 0.6490081412682032,
"grad_norm": 0.20580555498600006,
"learning_rate": 0.00028977643967701897,
"loss": 2.0204,
"step": 1415
},
{
"epoch": 0.6494668042655659,
"grad_norm": 0.2212708294391632,
"learning_rate": 0.00028910249501202156,
"loss": 1.9089,
"step": 1416
},
{
"epoch": 0.6499254672629285,
"grad_norm": 0.2184964120388031,
"learning_rate": 0.00028842901610590165,
"loss": 1.9906,
"step": 1417
},
{
"epoch": 0.6503841302602913,
"grad_norm": 0.20424173772335052,
"learning_rate": 0.00028775600444601123,
"loss": 1.9896,
"step": 1418
},
{
"epoch": 0.6508427932576539,
"grad_norm": 0.20544658601284027,
"learning_rate": 0.00028708346151866973,
"loss": 1.9475,
"step": 1419
},
{
"epoch": 0.6513014562550167,
"grad_norm": 0.2100137174129486,
"learning_rate": 0.0002864113888091622,
"loss": 1.9679,
"step": 1420
},
{
"epoch": 0.6517601192523793,
"grad_norm": 0.20152142643928528,
"learning_rate": 0.0002857397878017348,
"loss": 1.9696,
"step": 1421
},
{
"epoch": 0.6522187822497421,
"grad_norm": 0.20719367265701294,
"learning_rate": 0.00028506865997959173,
"loss": 2.1103,
"step": 1422
},
{
"epoch": 0.6526774452471047,
"grad_norm": 0.21037356555461884,
"learning_rate": 0.000284398006824893,
"loss": 2.0117,
"step": 1423
},
{
"epoch": 0.6531361082444673,
"grad_norm": 0.21327276527881622,
"learning_rate": 0.00028372782981874963,
"loss": 1.9387,
"step": 1424
},
{
"epoch": 0.6535947712418301,
"grad_norm": 0.2205706387758255,
"learning_rate": 0.00028305813044122096,
"loss": 2.0319,
"step": 1425
},
{
"epoch": 0.6540534342391927,
"grad_norm": 0.20860137045383453,
"learning_rate": 0.0002823889101713122,
"loss": 2.0317,
"step": 1426
},
{
"epoch": 0.6545120972365555,
"grad_norm": 0.2153463065624237,
"learning_rate": 0.0002817201704869701,
"loss": 1.9221,
"step": 1427
},
{
"epoch": 0.6549707602339181,
"grad_norm": 0.20236098766326904,
"learning_rate": 0.00028105191286508,
"loss": 1.9714,
"step": 1428
},
{
"epoch": 0.6554294232312808,
"grad_norm": 0.20584775507450104,
"learning_rate": 0.00028038413878146245,
"loss": 1.9906,
"step": 1429
},
{
"epoch": 0.6558880862286435,
"grad_norm": 0.2037326693534851,
"learning_rate": 0.00027971684971087073,
"loss": 1.9195,
"step": 1430
},
{
"epoch": 0.6563467492260062,
"grad_norm": 0.20691841840744019,
"learning_rate": 0.00027905004712698643,
"loss": 1.9676,
"step": 1431
},
{
"epoch": 0.6568054122233689,
"grad_norm": 0.21388956904411316,
"learning_rate": 0.0002783837325024167,
"loss": 1.9159,
"step": 1432
},
{
"epoch": 0.6572640752207316,
"grad_norm": 0.2117088884115219,
"learning_rate": 0.00027771790730869153,
"loss": 1.9329,
"step": 1433
},
{
"epoch": 0.6577227382180942,
"grad_norm": 0.20468655228614807,
"learning_rate": 0.0002770525730162599,
"loss": 2.005,
"step": 1434
},
{
"epoch": 0.658181401215457,
"grad_norm": 0.19955389201641083,
"learning_rate": 0.00027638773109448645,
"loss": 1.9315,
"step": 1435
},
{
"epoch": 0.6586400642128196,
"grad_norm": 0.20124514400959015,
"learning_rate": 0.00027572338301164824,
"loss": 1.9645,
"step": 1436
},
{
"epoch": 0.6590987272101824,
"grad_norm": 0.20607447624206543,
"learning_rate": 0.0002750595302349324,
"loss": 2.0928,
"step": 1437
},
{
"epoch": 0.659557390207545,
"grad_norm": 0.2144896239042282,
"learning_rate": 0.00027439617423043145,
"loss": 1.9425,
"step": 1438
},
{
"epoch": 0.6600160532049077,
"grad_norm": 0.21110142767429352,
"learning_rate": 0.00027373331646314114,
"loss": 1.9776,
"step": 1439
},
{
"epoch": 0.6604747162022704,
"grad_norm": 0.22140033543109894,
"learning_rate": 0.0002730709583969572,
"loss": 1.9578,
"step": 1440
},
{
"epoch": 0.660933379199633,
"grad_norm": 0.20210352540016174,
"learning_rate": 0.0002724091014946711,
"loss": 2.0333,
"step": 1441
},
{
"epoch": 0.6613920421969958,
"grad_norm": 0.22223646938800812,
"learning_rate": 0.00027174774721796824,
"loss": 1.9602,
"step": 1442
},
{
"epoch": 0.6618507051943584,
"grad_norm": 0.21496334671974182,
"learning_rate": 0.0002710868970274232,
"loss": 1.9618,
"step": 1443
},
{
"epoch": 0.6623093681917211,
"grad_norm": 0.20850813388824463,
"learning_rate": 0.0002704265523824982,
"loss": 1.9765,
"step": 1444
},
{
"epoch": 0.6627680311890838,
"grad_norm": 0.21113887429237366,
"learning_rate": 0.00026976671474153826,
"loss": 1.9433,
"step": 1445
},
{
"epoch": 0.6632266941864465,
"grad_norm": 0.21002665162086487,
"learning_rate": 0.00026910738556176886,
"loss": 2.0317,
"step": 1446
},
{
"epoch": 0.6636853571838092,
"grad_norm": 0.21181471645832062,
"learning_rate": 0.0002684485662992929,
"loss": 1.9645,
"step": 1447
},
{
"epoch": 0.6641440201811719,
"grad_norm": 0.2158689945936203,
"learning_rate": 0.0002677902584090869,
"loss": 2.0016,
"step": 1448
},
{
"epoch": 0.6646026831785345,
"grad_norm": 0.21313722431659698,
"learning_rate": 0.00026713246334499774,
"loss": 2.0057,
"step": 1449
},
{
"epoch": 0.6650613461758973,
"grad_norm": 0.20478276908397675,
"learning_rate": 0.00026647518255974023,
"loss": 2.0433,
"step": 1450
},
{
"epoch": 0.6655200091732599,
"grad_norm": 0.22127315402030945,
"learning_rate": 0.0002658184175048934,
"loss": 2.0348,
"step": 1451
},
{
"epoch": 0.6659786721706227,
"grad_norm": 0.20742307603359222,
"learning_rate": 0.00026516216963089694,
"loss": 2.0355,
"step": 1452
},
{
"epoch": 0.6664373351679853,
"grad_norm": 0.19760966300964355,
"learning_rate": 0.0002645064403870488,
"loss": 2.034,
"step": 1453
},
{
"epoch": 0.6668959981653481,
"grad_norm": 0.20178966224193573,
"learning_rate": 0.0002638512312215011,
"loss": 2.0076,
"step": 1454
},
{
"epoch": 0.6673546611627107,
"grad_norm": 0.2048628032207489,
"learning_rate": 0.0002631965435812575,
"loss": 1.918,
"step": 1455
},
{
"epoch": 0.6678133241600734,
"grad_norm": 0.20856982469558716,
"learning_rate": 0.00026254237891217046,
"loss": 1.9901,
"step": 1456
},
{
"epoch": 0.6682719871574361,
"grad_norm": 0.20218762755393982,
"learning_rate": 0.0002618887386589367,
"loss": 1.926,
"step": 1457
},
{
"epoch": 0.6687306501547987,
"grad_norm": 0.21131214499473572,
"learning_rate": 0.0002612356242650949,
"loss": 1.9699,
"step": 1458
},
{
"epoch": 0.6691893131521615,
"grad_norm": 0.22699293494224548,
"learning_rate": 0.0002605830371730229,
"loss": 1.9915,
"step": 1459
},
{
"epoch": 0.6696479761495241,
"grad_norm": 0.21783480048179626,
"learning_rate": 0.0002599309788239339,
"loss": 2.0426,
"step": 1460
},
{
"epoch": 0.6701066391468868,
"grad_norm": 0.21722686290740967,
"learning_rate": 0.00025927945065787306,
"loss": 1.9692,
"step": 1461
},
{
"epoch": 0.6705653021442495,
"grad_norm": 0.20533257722854614,
"learning_rate": 0.0002586284541137145,
"loss": 1.9217,
"step": 1462
},
{
"epoch": 0.6710239651416122,
"grad_norm": 0.20481771230697632,
"learning_rate": 0.00025797799062915905,
"loss": 1.9579,
"step": 1463
},
{
"epoch": 0.6714826281389749,
"grad_norm": 0.2245722860097885,
"learning_rate": 0.00025732806164072966,
"loss": 1.9575,
"step": 1464
},
{
"epoch": 0.6719412911363376,
"grad_norm": 0.20747876167297363,
"learning_rate": 0.00025667866858376874,
"loss": 1.961,
"step": 1465
},
{
"epoch": 0.6723999541337002,
"grad_norm": 0.21304714679718018,
"learning_rate": 0.0002560298128924358,
"loss": 1.9598,
"step": 1466
},
{
"epoch": 0.672858617131063,
"grad_norm": 0.21807615458965302,
"learning_rate": 0.0002553814959997032,
"loss": 1.9675,
"step": 1467
},
{
"epoch": 0.6733172801284256,
"grad_norm": 0.20344772934913635,
"learning_rate": 0.00025473371933735334,
"loss": 2.0148,
"step": 1468
},
{
"epoch": 0.6737759431257884,
"grad_norm": 0.21874181926250458,
"learning_rate": 0.00025408648433597534,
"loss": 1.9872,
"step": 1469
},
{
"epoch": 0.674234606123151,
"grad_norm": 0.20652060210704803,
"learning_rate": 0.00025343979242496283,
"loss": 1.9486,
"step": 1470
},
{
"epoch": 0.6746932691205138,
"grad_norm": 0.20319107174873352,
"learning_rate": 0.00025279364503250925,
"loss": 2.0131,
"step": 1471
},
{
"epoch": 0.6751519321178764,
"grad_norm": 0.2156846821308136,
"learning_rate": 0.0002521480435856056,
"loss": 1.9716,
"step": 1472
},
{
"epoch": 0.6756105951152391,
"grad_norm": 0.2131464034318924,
"learning_rate": 0.0002515029895100378,
"loss": 2.0716,
"step": 1473
},
{
"epoch": 0.6760692581126018,
"grad_norm": 0.2077646702528,
"learning_rate": 0.0002508584842303822,
"loss": 1.9963,
"step": 1474
},
{
"epoch": 0.6765279211099644,
"grad_norm": 0.2074904590845108,
"learning_rate": 0.0002502145291700038,
"loss": 2.0104,
"step": 1475
},
{
"epoch": 0.6769865841073271,
"grad_norm": 0.21982137858867645,
"learning_rate": 0.0002495711257510517,
"loss": 1.8965,
"step": 1476
},
{
"epoch": 0.6774452471046898,
"grad_norm": 0.22623798251152039,
"learning_rate": 0.0002489282753944575,
"loss": 2.0683,
"step": 1477
},
{
"epoch": 0.6779039101020525,
"grad_norm": 0.2129301279783249,
"learning_rate": 0.00024828597951993093,
"loss": 1.979,
"step": 1478
},
{
"epoch": 0.6783625730994152,
"grad_norm": 0.21392731368541718,
"learning_rate": 0.00024764423954595706,
"loss": 2.0128,
"step": 1479
},
{
"epoch": 0.6788212360967779,
"grad_norm": 0.2042301446199417,
"learning_rate": 0.0002470030568897938,
"loss": 1.9078,
"step": 1480
},
{
"epoch": 0.6792798990941405,
"grad_norm": 0.22382909059524536,
"learning_rate": 0.00024636243296746773,
"loss": 1.9219,
"step": 1481
},
{
"epoch": 0.6797385620915033,
"grad_norm": 0.21174944937229156,
"learning_rate": 0.0002457223691937716,
"loss": 1.912,
"step": 1482
},
{
"epoch": 0.6801972250888659,
"grad_norm": 0.21548572182655334,
"learning_rate": 0.0002450828669822613,
"loss": 1.9752,
"step": 1483
},
{
"epoch": 0.6806558880862287,
"grad_norm": 0.21526028215885162,
"learning_rate": 0.00024444392774525253,
"loss": 1.973,
"step": 1484
},
{
"epoch": 0.6811145510835913,
"grad_norm": 0.21514205634593964,
"learning_rate": 0.00024380555289381733,
"loss": 1.8799,
"step": 1485
},
{
"epoch": 0.6815732140809541,
"grad_norm": 0.22093833982944489,
"learning_rate": 0.00024316774383778184,
"loss": 2.1027,
"step": 1486
},
{
"epoch": 0.6820318770783167,
"grad_norm": 0.21424348652362823,
"learning_rate": 0.0002425305019857222,
"loss": 1.9582,
"step": 1487
},
{
"epoch": 0.6824905400756794,
"grad_norm": 0.21793900430202484,
"learning_rate": 0.00024189382874496184,
"loss": 2.0112,
"step": 1488
},
{
"epoch": 0.6829492030730421,
"grad_norm": 0.20741984248161316,
"learning_rate": 0.00024125772552156916,
"loss": 2.0105,
"step": 1489
},
{
"epoch": 0.6834078660704048,
"grad_norm": 0.21343663334846497,
"learning_rate": 0.00024062219372035292,
"loss": 1.9632,
"step": 1490
},
{
"epoch": 0.6838665290677675,
"grad_norm": 0.21365787088871002,
"learning_rate": 0.00023998723474486007,
"loss": 1.9961,
"step": 1491
},
{
"epoch": 0.6843251920651301,
"grad_norm": 0.22821760177612305,
"learning_rate": 0.00023935284999737272,
"loss": 1.975,
"step": 1492
},
{
"epoch": 0.6847838550624928,
"grad_norm": 0.20044344663619995,
"learning_rate": 0.00023871904087890505,
"loss": 2.0484,
"step": 1493
},
{
"epoch": 0.6852425180598555,
"grad_norm": 0.2086317539215088,
"learning_rate": 0.00023808580878919945,
"loss": 1.9618,
"step": 1494
},
{
"epoch": 0.6857011810572182,
"grad_norm": 0.2119130641222,
"learning_rate": 0.00023745315512672398,
"loss": 2.0231,
"step": 1495
},
{
"epoch": 0.6861598440545809,
"grad_norm": 0.2192685753107071,
"learning_rate": 0.0002368210812886698,
"loss": 1.9982,
"step": 1496
},
{
"epoch": 0.6866185070519436,
"grad_norm": 0.20644506812095642,
"learning_rate": 0.0002361895886709471,
"loss": 1.8792,
"step": 1497
},
{
"epoch": 0.6870771700493062,
"grad_norm": 0.23283962905406952,
"learning_rate": 0.0002355586786681823,
"loss": 1.9584,
"step": 1498
},
{
"epoch": 0.687535833046669,
"grad_norm": 0.21198643743991852,
"learning_rate": 0.00023492835267371575,
"loss": 1.9147,
"step": 1499
},
{
"epoch": 0.6879944960440316,
"grad_norm": 0.22008061408996582,
"learning_rate": 0.0002342986120795978,
"loss": 2.0103,
"step": 1500
},
{
"epoch": 0.6884531590413944,
"grad_norm": 0.21158616244792938,
"learning_rate": 0.0002336694582765857,
"loss": 1.9254,
"step": 1501
},
{
"epoch": 0.688911822038757,
"grad_norm": 0.23097968101501465,
"learning_rate": 0.00023304089265414085,
"loss": 2.0026,
"step": 1502
},
{
"epoch": 0.6893704850361198,
"grad_norm": 0.22073383629322052,
"learning_rate": 0.00023241291660042613,
"loss": 1.9656,
"step": 1503
},
{
"epoch": 0.6898291480334824,
"grad_norm": 0.21214087307453156,
"learning_rate": 0.0002317855315023018,
"loss": 1.9329,
"step": 1504
},
{
"epoch": 0.6902878110308451,
"grad_norm": 0.21400022506713867,
"learning_rate": 0.00023115873874532324,
"loss": 2.0736,
"step": 1505
},
{
"epoch": 0.6907464740282078,
"grad_norm": 0.2069285660982132,
"learning_rate": 0.00023053253971373796,
"loss": 1.9982,
"step": 1506
},
{
"epoch": 0.6912051370255705,
"grad_norm": 0.20656046271324158,
"learning_rate": 0.00022990693579048166,
"loss": 1.9466,
"step": 1507
},
{
"epoch": 0.6916638000229332,
"grad_norm": 0.1999957412481308,
"learning_rate": 0.00022928192835717644,
"loss": 1.9195,
"step": 1508
},
{
"epoch": 0.6921224630202958,
"grad_norm": 0.20628796517848969,
"learning_rate": 0.00022865751879412634,
"loss": 2.0533,
"step": 1509
},
{
"epoch": 0.6925811260176585,
"grad_norm": 0.2083522081375122,
"learning_rate": 0.00022803370848031585,
"loss": 1.9742,
"step": 1510
},
{
"epoch": 0.6930397890150212,
"grad_norm": 0.21295218169689178,
"learning_rate": 0.00022741049879340542,
"loss": 1.9741,
"step": 1511
},
{
"epoch": 0.6934984520123839,
"grad_norm": 0.20642296969890594,
"learning_rate": 0.00022678789110972897,
"loss": 1.9973,
"step": 1512
},
{
"epoch": 0.6939571150097466,
"grad_norm": 0.20705148577690125,
"learning_rate": 0.00022616588680429155,
"loss": 2.0053,
"step": 1513
},
{
"epoch": 0.6944157780071093,
"grad_norm": 0.21783660352230072,
"learning_rate": 0.00022554448725076526,
"loss": 1.9165,
"step": 1514
},
{
"epoch": 0.6948744410044719,
"grad_norm": 0.21170644462108612,
"learning_rate": 0.0002249236938214863,
"loss": 1.9222,
"step": 1515
},
{
"epoch": 0.6953331040018347,
"grad_norm": 0.23451043665409088,
"learning_rate": 0.00022430350788745296,
"loss": 2.0259,
"step": 1516
},
{
"epoch": 0.6957917669991973,
"grad_norm": 0.21386070549488068,
"learning_rate": 0.00022368393081832166,
"loss": 1.9674,
"step": 1517
},
{
"epoch": 0.6962504299965601,
"grad_norm": 0.21401633322238922,
"learning_rate": 0.00022306496398240383,
"loss": 1.9736,
"step": 1518
},
{
"epoch": 0.6967090929939227,
"grad_norm": 0.23000279068946838,
"learning_rate": 0.00022244660874666373,
"loss": 1.9404,
"step": 1519
},
{
"epoch": 0.6971677559912854,
"grad_norm": 0.22375018894672394,
"learning_rate": 0.00022182886647671452,
"loss": 2.0251,
"step": 1520
},
{
"epoch": 0.6976264189886481,
"grad_norm": 0.21398292481899261,
"learning_rate": 0.0002212117385368157,
"loss": 1.993,
"step": 1521
},
{
"epoch": 0.6980850819860108,
"grad_norm": 0.2060842514038086,
"learning_rate": 0.00022059522628987038,
"loss": 1.9629,
"step": 1522
},
{
"epoch": 0.6985437449833735,
"grad_norm": 0.20732726156711578,
"learning_rate": 0.00021997933109742162,
"loss": 1.9525,
"step": 1523
},
{
"epoch": 0.6990024079807362,
"grad_norm": 0.20848816633224487,
"learning_rate": 0.00021936405431964969,
"loss": 2.0159,
"step": 1524
},
{
"epoch": 0.6994610709780988,
"grad_norm": 0.21378262341022491,
"learning_rate": 0.00021874939731536926,
"loss": 1.9574,
"step": 1525
},
{
"epoch": 0.6999197339754615,
"grad_norm": 0.21074506640434265,
"learning_rate": 0.00021813536144202656,
"loss": 1.97,
"step": 1526
},
{
"epoch": 0.7003783969728242,
"grad_norm": 0.2044081836938858,
"learning_rate": 0.00021752194805569553,
"loss": 1.8871,
"step": 1527
},
{
"epoch": 0.7008370599701869,
"grad_norm": 0.21120300889015198,
"learning_rate": 0.0002169091585110754,
"loss": 1.9117,
"step": 1528
},
{
"epoch": 0.7012957229675496,
"grad_norm": 0.20132139325141907,
"learning_rate": 0.00021629699416148828,
"loss": 2.0338,
"step": 1529
},
{
"epoch": 0.7017543859649122,
"grad_norm": 0.21953444182872772,
"learning_rate": 0.000215685456358875,
"loss": 1.9444,
"step": 1530
},
{
"epoch": 0.702213048962275,
"grad_norm": 0.20535847544670105,
"learning_rate": 0.00021507454645379258,
"loss": 1.9807,
"step": 1531
},
{
"epoch": 0.7026717119596376,
"grad_norm": 0.22630836069583893,
"learning_rate": 0.00021446426579541184,
"loss": 2.0623,
"step": 1532
},
{
"epoch": 0.7031303749570004,
"grad_norm": 0.20011521875858307,
"learning_rate": 0.00021385461573151387,
"loss": 2.0108,
"step": 1533
},
{
"epoch": 0.703589037954363,
"grad_norm": 0.2178131490945816,
"learning_rate": 0.00021324559760848677,
"loss": 2.0792,
"step": 1534
},
{
"epoch": 0.7040477009517258,
"grad_norm": 0.20352420210838318,
"learning_rate": 0.00021263721277132303,
"loss": 2.0353,
"step": 1535
},
{
"epoch": 0.7045063639490884,
"grad_norm": 0.20945791900157928,
"learning_rate": 0.0002120294625636171,
"loss": 1.9707,
"step": 1536
},
{
"epoch": 0.7049650269464511,
"grad_norm": 0.20171557366847992,
"learning_rate": 0.0002114223483275613,
"loss": 1.863,
"step": 1537
},
{
"epoch": 0.7054236899438138,
"grad_norm": 0.20855040848255157,
"learning_rate": 0.0002108158714039435,
"loss": 1.9614,
"step": 1538
},
{
"epoch": 0.7058823529411765,
"grad_norm": 0.207330122590065,
"learning_rate": 0.00021021003313214455,
"loss": 1.981,
"step": 1539
},
{
"epoch": 0.7063410159385392,
"grad_norm": 0.21026752889156342,
"learning_rate": 0.00020960483485013432,
"loss": 1.968,
"step": 1540
},
{
"epoch": 0.7067996789359019,
"grad_norm": 0.21045763790607452,
"learning_rate": 0.0002090002778944694,
"loss": 1.9689,
"step": 1541
},
{
"epoch": 0.7072583419332645,
"grad_norm": 0.2111373394727707,
"learning_rate": 0.00020839636360029025,
"loss": 1.9479,
"step": 1542
},
{
"epoch": 0.7077170049306272,
"grad_norm": 0.2146979123353958,
"learning_rate": 0.00020779309330131818,
"loss": 2.0072,
"step": 1543
},
{
"epoch": 0.7081756679279899,
"grad_norm": 0.2147378921508789,
"learning_rate": 0.00020719046832985184,
"loss": 2.0128,
"step": 1544
},
{
"epoch": 0.7086343309253526,
"grad_norm": 0.23276519775390625,
"learning_rate": 0.0002065884900167646,
"loss": 2.0625,
"step": 1545
},
{
"epoch": 0.7090929939227153,
"grad_norm": 0.2003513127565384,
"learning_rate": 0.0002059871596915024,
"loss": 1.9266,
"step": 1546
},
{
"epoch": 0.7095516569200779,
"grad_norm": 0.21435141563415527,
"learning_rate": 0.0002053864786820795,
"loss": 1.9601,
"step": 1547
},
{
"epoch": 0.7100103199174407,
"grad_norm": 0.21388588845729828,
"learning_rate": 0.00020478644831507627,
"loss": 1.9871,
"step": 1548
},
{
"epoch": 0.7104689829148033,
"grad_norm": 0.21799156069755554,
"learning_rate": 0.00020418706991563634,
"loss": 1.9848,
"step": 1549
},
{
"epoch": 0.7109276459121661,
"grad_norm": 0.2196502685546875,
"learning_rate": 0.00020358834480746363,
"loss": 1.9979,
"step": 1550
},
{
"epoch": 0.7113863089095287,
"grad_norm": 0.21330948173999786,
"learning_rate": 0.0002029902743128188,
"loss": 2.0208,
"step": 1551
},
{
"epoch": 0.7118449719068914,
"grad_norm": 0.20505408942699432,
"learning_rate": 0.0002023928597525174,
"loss": 2.0013,
"step": 1552
},
{
"epoch": 0.7123036349042541,
"grad_norm": 0.21338357031345367,
"learning_rate": 0.00020179610244592595,
"loss": 2.0069,
"step": 1553
},
{
"epoch": 0.7127622979016168,
"grad_norm": 0.22413994371891022,
"learning_rate": 0.00020120000371095937,
"loss": 2.0348,
"step": 1554
},
{
"epoch": 0.7132209608989795,
"grad_norm": 0.22910171747207642,
"learning_rate": 0.0002006045648640787,
"loss": 1.9702,
"step": 1555
},
{
"epoch": 0.7136796238963422,
"grad_norm": 0.20753081142902374,
"learning_rate": 0.00020000978722028713,
"loss": 1.9661,
"step": 1556
},
{
"epoch": 0.7141382868937048,
"grad_norm": 0.20980650186538696,
"learning_rate": 0.00019941567209312767,
"loss": 1.9765,
"step": 1557
},
{
"epoch": 0.7145969498910676,
"grad_norm": 0.2109229862689972,
"learning_rate": 0.00019882222079468036,
"loss": 1.9839,
"step": 1558
},
{
"epoch": 0.7150556128884302,
"grad_norm": 0.211203932762146,
"learning_rate": 0.0001982294346355595,
"loss": 1.9126,
"step": 1559
},
{
"epoch": 0.7155142758857929,
"grad_norm": 0.21287044882774353,
"learning_rate": 0.00019763731492490976,
"loss": 2.0469,
"step": 1560
},
{
"epoch": 0.7159729388831556,
"grad_norm": 0.2022552639245987,
"learning_rate": 0.00019704586297040422,
"loss": 1.9507,
"step": 1561
},
{
"epoch": 0.7164316018805182,
"grad_norm": 0.20304684340953827,
"learning_rate": 0.0001964550800782417,
"loss": 1.9776,
"step": 1562
},
{
"epoch": 0.716890264877881,
"grad_norm": 0.20349667966365814,
"learning_rate": 0.00019586496755314288,
"loss": 1.9877,
"step": 1563
},
{
"epoch": 0.7173489278752436,
"grad_norm": 0.21515505015850067,
"learning_rate": 0.00019527552669834798,
"loss": 1.9697,
"step": 1564
},
{
"epoch": 0.7178075908726064,
"grad_norm": 0.2068495750427246,
"learning_rate": 0.0001946867588156142,
"loss": 2.0127,
"step": 1565
},
{
"epoch": 0.718266253869969,
"grad_norm": 0.21804773807525635,
"learning_rate": 0.00019409866520521258,
"loss": 1.9813,
"step": 1566
},
{
"epoch": 0.7187249168673318,
"grad_norm": 0.2098647803068161,
"learning_rate": 0.00019351124716592455,
"loss": 1.9016,
"step": 1567
},
{
"epoch": 0.7191835798646944,
"grad_norm": 0.21640163660049438,
"learning_rate": 0.0001929245059950397,
"loss": 1.9982,
"step": 1568
},
{
"epoch": 0.7196422428620571,
"grad_norm": 0.21454420685768127,
"learning_rate": 0.0001923384429883533,
"loss": 2.0698,
"step": 1569
},
{
"epoch": 0.7201009058594198,
"grad_norm": 0.21044176816940308,
"learning_rate": 0.00019175305944016237,
"loss": 1.9399,
"step": 1570
},
{
"epoch": 0.7205595688567825,
"grad_norm": 0.20945215225219727,
"learning_rate": 0.00019116835664326326,
"loss": 1.9993,
"step": 1571
},
{
"epoch": 0.7210182318541452,
"grad_norm": 0.2314155101776123,
"learning_rate": 0.0001905843358889497,
"loss": 1.9989,
"step": 1572
},
{
"epoch": 0.7214768948515079,
"grad_norm": 0.2103126049041748,
"learning_rate": 0.00019000099846700836,
"loss": 1.9584,
"step": 1573
},
{
"epoch": 0.7219355578488705,
"grad_norm": 0.21843360364437103,
"learning_rate": 0.00018941834566571692,
"loss": 1.9796,
"step": 1574
},
{
"epoch": 0.7223942208462333,
"grad_norm": 0.23398412764072418,
"learning_rate": 0.00018883637877184145,
"loss": 1.9987,
"step": 1575
},
{
"epoch": 0.7228528838435959,
"grad_norm": 0.21901334822177887,
"learning_rate": 0.00018825509907063325,
"loss": 2.0877,
"step": 1576
},
{
"epoch": 0.7233115468409586,
"grad_norm": 0.22042055428028107,
"learning_rate": 0.00018767450784582557,
"loss": 2.0327,
"step": 1577
},
{
"epoch": 0.7237702098383213,
"grad_norm": 0.21314440667629242,
"learning_rate": 0.00018709460637963122,
"loss": 2.017,
"step": 1578
},
{
"epoch": 0.7242288728356839,
"grad_norm": 0.21846747398376465,
"learning_rate": 0.00018651539595274013,
"loss": 1.9632,
"step": 1579
},
{
"epoch": 0.7246875358330467,
"grad_norm": 0.22109127044677734,
"learning_rate": 0.00018593687784431578,
"loss": 1.9897,
"step": 1580
},
{
"epoch": 0.7251461988304093,
"grad_norm": 0.21089881658554077,
"learning_rate": 0.00018535905333199248,
"loss": 1.8972,
"step": 1581
},
{
"epoch": 0.7256048618277721,
"grad_norm": 0.2091851532459259,
"learning_rate": 0.0001847819236918733,
"loss": 1.9635,
"step": 1582
},
{
"epoch": 0.7260635248251347,
"grad_norm": 0.20924438536167145,
"learning_rate": 0.00018420549019852655,
"loss": 1.943,
"step": 1583
},
{
"epoch": 0.7265221878224974,
"grad_norm": 0.2110396921634674,
"learning_rate": 0.00018362975412498266,
"loss": 1.9721,
"step": 1584
},
{
"epoch": 0.7269808508198601,
"grad_norm": 0.21784904599189758,
"learning_rate": 0.00018305471674273261,
"loss": 1.9575,
"step": 1585
},
{
"epoch": 0.7274395138172228,
"grad_norm": 0.21486909687519073,
"learning_rate": 0.0001824803793217237,
"loss": 2.0714,
"step": 1586
},
{
"epoch": 0.7278981768145855,
"grad_norm": 0.21856904029846191,
"learning_rate": 0.00018190674313035737,
"loss": 1.9673,
"step": 1587
},
{
"epoch": 0.7283568398119482,
"grad_norm": 0.20935751497745514,
"learning_rate": 0.00018133380943548716,
"loss": 1.9809,
"step": 1588
},
{
"epoch": 0.7288155028093108,
"grad_norm": 0.2034081369638443,
"learning_rate": 0.00018076157950241452,
"loss": 2.015,
"step": 1589
},
{
"epoch": 0.7292741658066736,
"grad_norm": 0.21784910559654236,
"learning_rate": 0.00018019005459488652,
"loss": 1.9576,
"step": 1590
},
{
"epoch": 0.7297328288040362,
"grad_norm": 0.2096559703350067,
"learning_rate": 0.00017961923597509388,
"loss": 1.9087,
"step": 1591
},
{
"epoch": 0.730191491801399,
"grad_norm": 0.2057277262210846,
"learning_rate": 0.00017904912490366722,
"loss": 1.9983,
"step": 1592
},
{
"epoch": 0.7306501547987616,
"grad_norm": 0.2038499116897583,
"learning_rate": 0.00017847972263967433,
"loss": 2.0043,
"step": 1593
},
{
"epoch": 0.7311088177961242,
"grad_norm": 0.20388565957546234,
"learning_rate": 0.0001779110304406177,
"loss": 1.9617,
"step": 1594
},
{
"epoch": 0.731567480793487,
"grad_norm": 0.2122878134250641,
"learning_rate": 0.000177343049562432,
"loss": 1.9434,
"step": 1595
},
{
"epoch": 0.7320261437908496,
"grad_norm": 0.20858648419380188,
"learning_rate": 0.0001767757812594807,
"loss": 1.965,
"step": 1596
},
{
"epoch": 0.7324848067882124,
"grad_norm": 0.2169448584318161,
"learning_rate": 0.0001762092267845534,
"loss": 2.0215,
"step": 1597
},
{
"epoch": 0.732943469785575,
"grad_norm": 0.22394829988479614,
"learning_rate": 0.00017564338738886365,
"loss": 2.0357,
"step": 1598
},
{
"epoch": 0.7334021327829378,
"grad_norm": 0.21869732439517975,
"learning_rate": 0.0001750782643220457,
"loss": 2.023,
"step": 1599
},
{
"epoch": 0.7338607957803004,
"grad_norm": 0.21256084740161896,
"learning_rate": 0.00017451385883215166,
"loss": 2.0765,
"step": 1600
},
{
"epoch": 0.7343194587776631,
"grad_norm": 0.21862387657165527,
"learning_rate": 0.00017395017216564863,
"loss": 1.967,
"step": 1601
},
{
"epoch": 0.7347781217750258,
"grad_norm": 0.21675917506217957,
"learning_rate": 0.00017338720556741687,
"loss": 1.9437,
"step": 1602
},
{
"epoch": 0.7352367847723885,
"grad_norm": 0.20436620712280273,
"learning_rate": 0.00017282496028074606,
"loss": 1.9172,
"step": 1603
},
{
"epoch": 0.7356954477697512,
"grad_norm": 0.21417811512947083,
"learning_rate": 0.00017226343754733254,
"loss": 1.9063,
"step": 1604
},
{
"epoch": 0.7361541107671139,
"grad_norm": 0.20532532036304474,
"learning_rate": 0.00017170263860727769,
"loss": 1.8955,
"step": 1605
},
{
"epoch": 0.7366127737644765,
"grad_norm": 0.2180437296628952,
"learning_rate": 0.0001711425646990838,
"loss": 1.9686,
"step": 1606
},
{
"epoch": 0.7370714367618393,
"grad_norm": 0.21358822286128998,
"learning_rate": 0.00017058321705965202,
"loss": 1.8779,
"step": 1607
},
{
"epoch": 0.7375300997592019,
"grad_norm": 0.20846621692180634,
"learning_rate": 0.0001700245969242798,
"loss": 2.0667,
"step": 1608
},
{
"epoch": 0.7379887627565647,
"grad_norm": 0.21123076975345612,
"learning_rate": 0.00016946670552665804,
"loss": 1.9807,
"step": 1609
},
{
"epoch": 0.7384474257539273,
"grad_norm": 0.20466071367263794,
"learning_rate": 0.00016890954409886795,
"loss": 1.9213,
"step": 1610
},
{
"epoch": 0.7389060887512899,
"grad_norm": 0.24498245120048523,
"learning_rate": 0.00016835311387137836,
"loss": 1.9763,
"step": 1611
},
{
"epoch": 0.7393647517486527,
"grad_norm": 0.21921217441558838,
"learning_rate": 0.0001677974160730441,
"loss": 2.0283,
"step": 1612
},
{
"epoch": 0.7398234147460153,
"grad_norm": 0.21773578226566315,
"learning_rate": 0.00016724245193110176,
"loss": 1.8425,
"step": 1613
},
{
"epoch": 0.7402820777433781,
"grad_norm": 0.21932828426361084,
"learning_rate": 0.00016668822267116784,
"loss": 1.9434,
"step": 1614
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.21156875789165497,
"learning_rate": 0.00016613472951723597,
"loss": 2.0142,
"step": 1615
},
{
"epoch": 0.7411994037381034,
"grad_norm": 0.21231050789356232,
"learning_rate": 0.00016558197369167434,
"loss": 2.0113,
"step": 1616
},
{
"epoch": 0.7416580667354661,
"grad_norm": 0.21398648619651794,
"learning_rate": 0.00016502995641522216,
"loss": 2.098,
"step": 1617
},
{
"epoch": 0.7421167297328288,
"grad_norm": 0.22223353385925293,
"learning_rate": 0.00016447867890698843,
"loss": 2.0792,
"step": 1618
},
{
"epoch": 0.7425753927301915,
"grad_norm": 0.20777761936187744,
"learning_rate": 0.00016392814238444753,
"loss": 1.8945,
"step": 1619
},
{
"epoch": 0.7430340557275542,
"grad_norm": 0.21802599728107452,
"learning_rate": 0.00016337834806343782,
"loss": 1.9669,
"step": 1620
},
{
"epoch": 0.7434927187249168,
"grad_norm": 0.21594639122486115,
"learning_rate": 0.0001628292971581588,
"loss": 1.9647,
"step": 1621
},
{
"epoch": 0.7439513817222796,
"grad_norm": 0.20650894939899445,
"learning_rate": 0.00016228099088116772,
"loss": 2.0288,
"step": 1622
},
{
"epoch": 0.7444100447196422,
"grad_norm": 0.21720433235168457,
"learning_rate": 0.00016173343044337734,
"loss": 1.9983,
"step": 1623
},
{
"epoch": 0.744868707717005,
"grad_norm": 0.2116628885269165,
"learning_rate": 0.00016118661705405356,
"loss": 1.9231,
"step": 1624
},
{
"epoch": 0.7453273707143676,
"grad_norm": 0.19845157861709595,
"learning_rate": 0.00016064055192081255,
"loss": 1.9689,
"step": 1625
},
{
"epoch": 0.7457860337117304,
"grad_norm": 0.22194702923297882,
"learning_rate": 0.00016009523624961757,
"loss": 1.9457,
"step": 1626
},
{
"epoch": 0.746244696709093,
"grad_norm": 0.21709389984607697,
"learning_rate": 0.00015955067124477678,
"loss": 2.0037,
"step": 1627
},
{
"epoch": 0.7467033597064556,
"grad_norm": 0.21728098392486572,
"learning_rate": 0.000159006858108941,
"loss": 1.9214,
"step": 1628
},
{
"epoch": 0.7471620227038184,
"grad_norm": 0.21065804362297058,
"learning_rate": 0.00015846379804310002,
"loss": 1.9232,
"step": 1629
},
{
"epoch": 0.747620685701181,
"grad_norm": 0.21014493703842163,
"learning_rate": 0.00015792149224658054,
"loss": 1.9483,
"step": 1630
},
{
"epoch": 0.7480793486985438,
"grad_norm": 0.21665003895759583,
"learning_rate": 0.00015737994191704385,
"loss": 1.9151,
"step": 1631
},
{
"epoch": 0.7485380116959064,
"grad_norm": 0.2119457870721817,
"learning_rate": 0.0001568391482504829,
"loss": 1.9048,
"step": 1632
},
{
"epoch": 0.7489966746932691,
"grad_norm": 0.20531503856182098,
"learning_rate": 0.00015629911244121903,
"loss": 1.896,
"step": 1633
},
{
"epoch": 0.7494553376906318,
"grad_norm": 0.21179962158203125,
"learning_rate": 0.0001557598356819,
"loss": 1.8886,
"step": 1634
},
{
"epoch": 0.7499140006879945,
"grad_norm": 0.21961160004138947,
"learning_rate": 0.00015522131916349786,
"loss": 1.9578,
"step": 1635
},
{
"epoch": 0.7503726636853572,
"grad_norm": 0.23272109031677246,
"learning_rate": 0.00015468356407530493,
"loss": 1.9544,
"step": 1636
},
{
"epoch": 0.7508313266827199,
"grad_norm": 0.21137219667434692,
"learning_rate": 0.00015414657160493217,
"loss": 1.9105,
"step": 1637
},
{
"epoch": 0.7512899896800825,
"grad_norm": 0.20370523631572723,
"learning_rate": 0.00015361034293830673,
"loss": 1.973,
"step": 1638
},
{
"epoch": 0.7517486526774453,
"grad_norm": 0.20508180558681488,
"learning_rate": 0.00015307487925966844,
"loss": 1.9443,
"step": 1639
},
{
"epoch": 0.7522073156748079,
"grad_norm": 0.20535223186016083,
"learning_rate": 0.00015254018175156776,
"loss": 1.9941,
"step": 1640
},
{
"epoch": 0.7526659786721707,
"grad_norm": 0.2090098261833191,
"learning_rate": 0.0001520062515948632,
"loss": 1.9328,
"step": 1641
},
{
"epoch": 0.7531246416695333,
"grad_norm": 0.2108336240053177,
"learning_rate": 0.0001514730899687189,
"loss": 1.8758,
"step": 1642
},
{
"epoch": 0.753583304666896,
"grad_norm": 0.21868261694908142,
"learning_rate": 0.00015094069805060122,
"loss": 2.0467,
"step": 1643
},
{
"epoch": 0.7540419676642587,
"grad_norm": 0.22521698474884033,
"learning_rate": 0.00015040907701627666,
"loss": 2.015,
"step": 1644
},
{
"epoch": 0.7545006306616213,
"grad_norm": 0.21842342615127563,
"learning_rate": 0.00014987822803980976,
"loss": 2.0241,
"step": 1645
},
{
"epoch": 0.7549592936589841,
"grad_norm": 0.21006381511688232,
"learning_rate": 0.00014934815229355965,
"loss": 1.9406,
"step": 1646
},
{
"epoch": 0.7554179566563467,
"grad_norm": 0.21095126867294312,
"learning_rate": 0.00014881885094817748,
"loss": 1.9072,
"step": 1647
},
{
"epoch": 0.7558766196537094,
"grad_norm": 0.20989342033863068,
"learning_rate": 0.00014829032517260488,
"loss": 2.0022,
"step": 1648
},
{
"epoch": 0.7563352826510721,
"grad_norm": 0.2133699506521225,
"learning_rate": 0.0001477625761340704,
"loss": 2.0145,
"step": 1649
},
{
"epoch": 0.7567939456484348,
"grad_norm": 0.21815617382526398,
"learning_rate": 0.0001472356049980868,
"loss": 1.99,
"step": 1650
},
{
"epoch": 0.7572526086457975,
"grad_norm": 0.21013103425502777,
"learning_rate": 0.00014670941292844954,
"loss": 1.9299,
"step": 1651
},
{
"epoch": 0.7577112716431602,
"grad_norm": 0.22692517936229706,
"learning_rate": 0.00014618400108723295,
"loss": 1.9145,
"step": 1652
},
{
"epoch": 0.7581699346405228,
"grad_norm": 0.21964749693870544,
"learning_rate": 0.00014565937063478862,
"loss": 1.9737,
"step": 1653
},
{
"epoch": 0.7586285976378856,
"grad_norm": 0.21574419736862183,
"learning_rate": 0.00014513552272974207,
"loss": 1.9325,
"step": 1654
},
{
"epoch": 0.7590872606352482,
"grad_norm": 0.2371809035539627,
"learning_rate": 0.0001446124585289913,
"loss": 1.9478,
"step": 1655
},
{
"epoch": 0.759545923632611,
"grad_norm": 0.20493346452713013,
"learning_rate": 0.00014409017918770266,
"loss": 1.9183,
"step": 1656
},
{
"epoch": 0.7600045866299736,
"grad_norm": 0.20638450980186462,
"learning_rate": 0.00014356868585930994,
"loss": 2.0113,
"step": 1657
},
{
"epoch": 0.7604632496273364,
"grad_norm": 0.2113826721906662,
"learning_rate": 0.00014304797969551077,
"loss": 1.9067,
"step": 1658
},
{
"epoch": 0.760921912624699,
"grad_norm": 0.20950795710086823,
"learning_rate": 0.00014252806184626417,
"loss": 1.9339,
"step": 1659
},
{
"epoch": 0.7613805756220617,
"grad_norm": 0.2124875783920288,
"learning_rate": 0.00014200893345978817,
"loss": 2.0331,
"step": 1660
},
{
"epoch": 0.7618392386194244,
"grad_norm": 0.20337583124637604,
"learning_rate": 0.00014149059568255778,
"loss": 1.8815,
"step": 1661
},
{
"epoch": 0.762297901616787,
"grad_norm": 0.21365618705749512,
"learning_rate": 0.00014097304965930157,
"loss": 1.9608,
"step": 1662
},
{
"epoch": 0.7627565646141498,
"grad_norm": 0.21089614927768707,
"learning_rate": 0.00014045629653299953,
"loss": 1.9246,
"step": 1663
},
{
"epoch": 0.7632152276115124,
"grad_norm": 0.21705089509487152,
"learning_rate": 0.00013994033744488076,
"loss": 1.9043,
"step": 1664
},
{
"epoch": 0.7636738906088751,
"grad_norm": 0.2139820009469986,
"learning_rate": 0.00013942517353442092,
"loss": 1.9461,
"step": 1665
},
{
"epoch": 0.7641325536062378,
"grad_norm": 0.2277866005897522,
"learning_rate": 0.0001389108059393391,
"loss": 2.051,
"step": 1666
},
{
"epoch": 0.7645912166036005,
"grad_norm": 0.20648564398288727,
"learning_rate": 0.00013839723579559581,
"loss": 1.9363,
"step": 1667
},
{
"epoch": 0.7650498796009632,
"grad_norm": 0.21434453129768372,
"learning_rate": 0.00013788446423739103,
"loss": 2.0154,
"step": 1668
},
{
"epoch": 0.7655085425983259,
"grad_norm": 0.214177206158638,
"learning_rate": 0.00013737249239716042,
"loss": 2.0584,
"step": 1669
},
{
"epoch": 0.7659672055956885,
"grad_norm": 0.21177199482917786,
"learning_rate": 0.00013686132140557355,
"loss": 1.9585,
"step": 1670
},
{
"epoch": 0.7664258685930513,
"grad_norm": 0.21180562674999237,
"learning_rate": 0.00013635095239153188,
"loss": 1.8559,
"step": 1671
},
{
"epoch": 0.7668845315904139,
"grad_norm": 0.21189041435718536,
"learning_rate": 0.00013584138648216527,
"loss": 1.9582,
"step": 1672
},
{
"epoch": 0.7673431945877767,
"grad_norm": 0.2170976847410202,
"learning_rate": 0.0001353326248028298,
"loss": 1.9539,
"step": 1673
},
{
"epoch": 0.7678018575851393,
"grad_norm": 0.21785518527030945,
"learning_rate": 0.00013482466847710594,
"loss": 1.9467,
"step": 1674
},
{
"epoch": 0.768260520582502,
"grad_norm": 0.21004082262516022,
"learning_rate": 0.00013431751862679554,
"loss": 1.9908,
"step": 1675
},
{
"epoch": 0.7687191835798647,
"grad_norm": 0.21676945686340332,
"learning_rate": 0.00013381117637191887,
"loss": 2.0125,
"step": 1676
},
{
"epoch": 0.7691778465772274,
"grad_norm": 0.21960406005382538,
"learning_rate": 0.00013330564283071293,
"loss": 1.9723,
"step": 1677
},
{
"epoch": 0.7696365095745901,
"grad_norm": 0.2181558907032013,
"learning_rate": 0.000132800919119629,
"loss": 1.9438,
"step": 1678
},
{
"epoch": 0.7700951725719528,
"grad_norm": 0.22525444626808167,
"learning_rate": 0.00013229700635332948,
"loss": 1.9927,
"step": 1679
},
{
"epoch": 0.7705538355693154,
"grad_norm": 0.21948324143886566,
"learning_rate": 0.00013179390564468585,
"loss": 2.0394,
"step": 1680
},
{
"epoch": 0.7710124985666781,
"grad_norm": 0.21652016043663025,
"learning_rate": 0.00013129161810477641,
"loss": 2.0125,
"step": 1681
},
{
"epoch": 0.7714711615640408,
"grad_norm": 0.2246413379907608,
"learning_rate": 0.0001307901448428837,
"loss": 1.997,
"step": 1682
},
{
"epoch": 0.7719298245614035,
"grad_norm": 0.22197282314300537,
"learning_rate": 0.0001302894869664916,
"loss": 1.9914,
"step": 1683
},
{
"epoch": 0.7723884875587662,
"grad_norm": 0.20929904282093048,
"learning_rate": 0.00012978964558128336,
"loss": 1.9433,
"step": 1684
},
{
"epoch": 0.7728471505561288,
"grad_norm": 0.21058063209056854,
"learning_rate": 0.00012929062179113925,
"loss": 2.0249,
"step": 1685
},
{
"epoch": 0.7733058135534916,
"grad_norm": 0.2080467939376831,
"learning_rate": 0.00012879241669813368,
"loss": 2.0126,
"step": 1686
},
{
"epoch": 0.7737644765508542,
"grad_norm": 0.2149587869644165,
"learning_rate": 0.00012829503140253295,
"loss": 1.8873,
"step": 1687
},
{
"epoch": 0.774223139548217,
"grad_norm": 0.20427700877189636,
"learning_rate": 0.0001277984670027933,
"loss": 1.9683,
"step": 1688
},
{
"epoch": 0.7746818025455796,
"grad_norm": 0.20516161620616913,
"learning_rate": 0.00012730272459555737,
"loss": 1.9958,
"step": 1689
},
{
"epoch": 0.7751404655429424,
"grad_norm": 0.20784947276115417,
"learning_rate": 0.00012680780527565312,
"loss": 1.9607,
"step": 1690
},
{
"epoch": 0.775599128540305,
"grad_norm": 0.2163698673248291,
"learning_rate": 0.0001263137101360905,
"loss": 1.9821,
"step": 1691
},
{
"epoch": 0.7760577915376677,
"grad_norm": 0.22307239472866058,
"learning_rate": 0.00012582044026805922,
"loss": 1.9843,
"step": 1692
},
{
"epoch": 0.7765164545350304,
"grad_norm": 0.20408907532691956,
"learning_rate": 0.00012532799676092627,
"loss": 1.9404,
"step": 1693
},
{
"epoch": 0.7769751175323931,
"grad_norm": 0.21838350594043732,
"learning_rate": 0.00012483638070223414,
"loss": 1.9049,
"step": 1694
},
{
"epoch": 0.7774337805297558,
"grad_norm": 0.21656344830989838,
"learning_rate": 0.00012434559317769752,
"loss": 1.9248,
"step": 1695
},
{
"epoch": 0.7778924435271185,
"grad_norm": 0.21894484758377075,
"learning_rate": 0.0001238556352712012,
"loss": 1.9147,
"step": 1696
},
{
"epoch": 0.7783511065244811,
"grad_norm": 0.2081957757472992,
"learning_rate": 0.00012336650806479827,
"loss": 1.965,
"step": 1697
},
{
"epoch": 0.7788097695218438,
"grad_norm": 0.22024701535701752,
"learning_rate": 0.00012287821263870708,
"loss": 1.9845,
"step": 1698
},
{
"epoch": 0.7792684325192065,
"grad_norm": 0.2200576514005661,
"learning_rate": 0.00012239075007130885,
"loss": 1.8755,
"step": 1699
},
{
"epoch": 0.7797270955165692,
"grad_norm": 0.21726536750793457,
"learning_rate": 0.00012190412143914536,
"loss": 1.9677,
"step": 1700
},
{
"epoch": 0.7801857585139319,
"grad_norm": 0.21251441538333893,
"learning_rate": 0.0001214183278169172,
"loss": 1.9991,
"step": 1701
},
{
"epoch": 0.7806444215112945,
"grad_norm": 0.22695939242839813,
"learning_rate": 0.00012093337027748042,
"loss": 2.04,
"step": 1702
},
{
"epoch": 0.7811030845086573,
"grad_norm": 0.2209661304950714,
"learning_rate": 0.00012044924989184459,
"loss": 2.0554,
"step": 1703
},
{
"epoch": 0.7815617475060199,
"grad_norm": 0.21679779887199402,
"learning_rate": 0.0001199659677291709,
"loss": 1.979,
"step": 1704
},
{
"epoch": 0.7820204105033827,
"grad_norm": 0.19809038937091827,
"learning_rate": 0.00011948352485676895,
"loss": 1.8864,
"step": 1705
},
{
"epoch": 0.7824790735007453,
"grad_norm": 0.20385852456092834,
"learning_rate": 0.00011900192234009477,
"loss": 1.9263,
"step": 1706
},
{
"epoch": 0.782937736498108,
"grad_norm": 0.2089887410402298,
"learning_rate": 0.00011852116124274875,
"loss": 1.9223,
"step": 1707
},
{
"epoch": 0.7833963994954707,
"grad_norm": 0.20236855745315552,
"learning_rate": 0.00011804124262647314,
"loss": 1.856,
"step": 1708
},
{
"epoch": 0.7838550624928334,
"grad_norm": 0.21958214044570923,
"learning_rate": 0.00011756216755114929,
"loss": 1.8853,
"step": 1709
},
{
"epoch": 0.7843137254901961,
"grad_norm": 0.20870545506477356,
"learning_rate": 0.00011708393707479548,
"loss": 1.9928,
"step": 1710
},
{
"epoch": 0.7847723884875588,
"grad_norm": 0.22010666131973267,
"learning_rate": 0.00011660655225356531,
"loss": 2.054,
"step": 1711
},
{
"epoch": 0.7852310514849214,
"grad_norm": 0.20709814131259918,
"learning_rate": 0.0001161300141417444,
"loss": 1.9084,
"step": 1712
},
{
"epoch": 0.7856897144822842,
"grad_norm": 0.2124544084072113,
"learning_rate": 0.00011565432379174823,
"loss": 2.0395,
"step": 1713
},
{
"epoch": 0.7861483774796468,
"grad_norm": 0.20998547971248627,
"learning_rate": 0.00011517948225412056,
"loss": 2.0042,
"step": 1714
},
{
"epoch": 0.7866070404770095,
"grad_norm": 0.20224225521087646,
"learning_rate": 0.00011470549057753032,
"loss": 2.0026,
"step": 1715
},
{
"epoch": 0.7870657034743722,
"grad_norm": 0.22083058953285217,
"learning_rate": 0.00011423234980876957,
"loss": 2.0096,
"step": 1716
},
{
"epoch": 0.7875243664717348,
"grad_norm": 0.21419985592365265,
"learning_rate": 0.00011376006099275099,
"loss": 1.992,
"step": 1717
},
{
"epoch": 0.7879830294690976,
"grad_norm": 0.21626605093479156,
"learning_rate": 0.00011328862517250609,
"loss": 1.8747,
"step": 1718
},
{
"epoch": 0.7884416924664602,
"grad_norm": 0.21133039891719818,
"learning_rate": 0.00011281804338918239,
"loss": 1.9943,
"step": 1719
},
{
"epoch": 0.788900355463823,
"grad_norm": 0.22870618104934692,
"learning_rate": 0.00011234831668204115,
"loss": 2.0126,
"step": 1720
},
{
"epoch": 0.7893590184611856,
"grad_norm": 0.21979862451553345,
"learning_rate": 0.00011187944608845569,
"loss": 1.8556,
"step": 1721
},
{
"epoch": 0.7898176814585484,
"grad_norm": 0.210460364818573,
"learning_rate": 0.00011141143264390801,
"loss": 2.0277,
"step": 1722
},
{
"epoch": 0.790276344455911,
"grad_norm": 0.21769151091575623,
"learning_rate": 0.0001109442773819877,
"loss": 1.9275,
"step": 1723
},
{
"epoch": 0.7907350074532737,
"grad_norm": 0.21359185874462128,
"learning_rate": 0.0001104779813343889,
"loss": 1.9246,
"step": 1724
},
{
"epoch": 0.7911936704506364,
"grad_norm": 0.21665528416633606,
"learning_rate": 0.00011001254553090812,
"loss": 1.9551,
"step": 1725
},
{
"epoch": 0.7916523334479991,
"grad_norm": 0.2076639086008072,
"learning_rate": 0.00010954797099944186,
"loss": 2.0043,
"step": 1726
},
{
"epoch": 0.7921109964453618,
"grad_norm": 0.21552379429340363,
"learning_rate": 0.0001090842587659851,
"loss": 1.9884,
"step": 1727
},
{
"epoch": 0.7925696594427245,
"grad_norm": 0.21690185368061066,
"learning_rate": 0.00010862140985462804,
"loss": 2.0004,
"step": 1728
},
{
"epoch": 0.7930283224400871,
"grad_norm": 0.21822625398635864,
"learning_rate": 0.00010815942528755418,
"loss": 1.9551,
"step": 1729
},
{
"epoch": 0.7934869854374499,
"grad_norm": 0.2195984125137329,
"learning_rate": 0.00010769830608503844,
"loss": 1.9363,
"step": 1730
},
{
"epoch": 0.7939456484348125,
"grad_norm": 0.22050587832927704,
"learning_rate": 0.00010723805326544473,
"loss": 1.9675,
"step": 1731
},
{
"epoch": 0.7944043114321752,
"grad_norm": 0.20689231157302856,
"learning_rate": 0.00010677866784522316,
"loss": 1.9815,
"step": 1732
},
{
"epoch": 0.7948629744295379,
"grad_norm": 0.2166183739900589,
"learning_rate": 0.00010632015083890839,
"loss": 1.9328,
"step": 1733
},
{
"epoch": 0.7953216374269005,
"grad_norm": 0.1969301551580429,
"learning_rate": 0.00010586250325911745,
"loss": 1.9925,
"step": 1734
},
{
"epoch": 0.7957803004242633,
"grad_norm": 0.21839383244514465,
"learning_rate": 0.00010540572611654697,
"loss": 2.0089,
"step": 1735
},
{
"epoch": 0.7962389634216259,
"grad_norm": 0.22961905598640442,
"learning_rate": 0.00010494982041997126,
"loss": 1.9799,
"step": 1736
},
{
"epoch": 0.7966976264189887,
"grad_norm": 0.20711341500282288,
"learning_rate": 0.0001044947871762405,
"loss": 1.8958,
"step": 1737
},
{
"epoch": 0.7971562894163513,
"grad_norm": 0.20954112708568573,
"learning_rate": 0.00010404062739027753,
"loss": 1.9235,
"step": 1738
},
{
"epoch": 0.797614952413714,
"grad_norm": 0.21551868319511414,
"learning_rate": 0.00010358734206507641,
"loss": 1.8973,
"step": 1739
},
{
"epoch": 0.7980736154110767,
"grad_norm": 0.20958000421524048,
"learning_rate": 0.00010313493220170017,
"loss": 1.9089,
"step": 1740
},
{
"epoch": 0.7985322784084394,
"grad_norm": 0.21942633390426636,
"learning_rate": 0.00010268339879927836,
"loss": 1.9675,
"step": 1741
},
{
"epoch": 0.7989909414058021,
"grad_norm": 0.21762575209140778,
"learning_rate": 0.00010223274285500466,
"loss": 1.863,
"step": 1742
},
{
"epoch": 0.7994496044031648,
"grad_norm": 0.21566879749298096,
"learning_rate": 0.00010178296536413495,
"loss": 2.0114,
"step": 1743
},
{
"epoch": 0.7999082674005275,
"grad_norm": 0.2065243422985077,
"learning_rate": 0.00010133406731998546,
"loss": 1.8854,
"step": 1744
},
{
"epoch": 0.8003669303978902,
"grad_norm": 0.2149953842163086,
"learning_rate": 0.00010088604971392979,
"loss": 1.9918,
"step": 1745
},
{
"epoch": 0.8008255933952528,
"grad_norm": 0.20899248123168945,
"learning_rate": 0.0001004389135353972,
"loss": 1.893,
"step": 1746
},
{
"epoch": 0.8012842563926156,
"grad_norm": 0.20870202779769897,
"learning_rate": 9.999265977187049e-05,
"loss": 2.0535,
"step": 1747
},
{
"epoch": 0.8017429193899782,
"grad_norm": 0.20780949294567108,
"learning_rate": 9.95472894088838e-05,
"loss": 1.9911,
"step": 1748
},
{
"epoch": 0.8022015823873409,
"grad_norm": 0.20701158046722412,
"learning_rate": 9.910280343001993e-05,
"loss": 1.9874,
"step": 1749
},
{
"epoch": 0.8026602453847036,
"grad_norm": 0.22388163208961487,
"learning_rate": 9.865920281690866e-05,
"loss": 1.9265,
"step": 1750
},
{
"epoch": 0.8031189083820662,
"grad_norm": 0.22293399274349213,
"learning_rate": 9.821648854922482e-05,
"loss": 1.948,
"step": 1751
},
{
"epoch": 0.803577571379429,
"grad_norm": 0.21864351630210876,
"learning_rate": 9.77746616046854e-05,
"loss": 1.9588,
"step": 1752
},
{
"epoch": 0.8040362343767916,
"grad_norm": 0.21940460801124573,
"learning_rate": 9.733372295904774e-05,
"loss": 2.0086,
"step": 1753
},
{
"epoch": 0.8044948973741544,
"grad_norm": 0.2046966701745987,
"learning_rate": 9.68936735861079e-05,
"loss": 1.8681,
"step": 1754
},
{
"epoch": 0.804953560371517,
"grad_norm": 0.22357942163944244,
"learning_rate": 9.645451445769737e-05,
"loss": 1.9196,
"step": 1755
},
{
"epoch": 0.8054122233688797,
"grad_norm": 0.21548981964588165,
"learning_rate": 9.601624654368196e-05,
"loss": 2.0283,
"step": 1756
},
{
"epoch": 0.8058708863662424,
"grad_norm": 0.21971726417541504,
"learning_rate": 9.557887081195938e-05,
"loss": 2.0366,
"step": 1757
},
{
"epoch": 0.8063295493636051,
"grad_norm": 0.22221064567565918,
"learning_rate": 9.514238822845667e-05,
"loss": 2.019,
"step": 1758
},
{
"epoch": 0.8067882123609678,
"grad_norm": 0.2300874888896942,
"learning_rate": 9.470679975712837e-05,
"loss": 1.8487,
"step": 1759
},
{
"epoch": 0.8072468753583305,
"grad_norm": 0.23173339664936066,
"learning_rate": 9.427210635995481e-05,
"loss": 1.9672,
"step": 1760
},
{
"epoch": 0.8077055383556931,
"grad_norm": 0.21483147144317627,
"learning_rate": 9.383830899693923e-05,
"loss": 1.883,
"step": 1761
},
{
"epoch": 0.8081642013530559,
"grad_norm": 0.21176545321941376,
"learning_rate": 9.340540862610591e-05,
"loss": 1.9692,
"step": 1762
},
{
"epoch": 0.8086228643504185,
"grad_norm": 0.21161183714866638,
"learning_rate": 9.297340620349854e-05,
"loss": 1.9803,
"step": 1763
},
{
"epoch": 0.8090815273477813,
"grad_norm": 0.21678751707077026,
"learning_rate": 9.25423026831777e-05,
"loss": 1.9335,
"step": 1764
},
{
"epoch": 0.8095401903451439,
"grad_norm": 0.22137022018432617,
"learning_rate": 9.211209901721846e-05,
"loss": 1.9599,
"step": 1765
},
{
"epoch": 0.8099988533425065,
"grad_norm": 0.21324829757213593,
"learning_rate": 9.168279615570863e-05,
"loss": 1.936,
"step": 1766
},
{
"epoch": 0.8104575163398693,
"grad_norm": 0.22642551362514496,
"learning_rate": 9.125439504674699e-05,
"loss": 2.0377,
"step": 1767
},
{
"epoch": 0.8109161793372319,
"grad_norm": 0.2122025042772293,
"learning_rate": 9.082689663644057e-05,
"loss": 1.9847,
"step": 1768
},
{
"epoch": 0.8113748423345947,
"grad_norm": 0.21190258860588074,
"learning_rate": 9.040030186890264e-05,
"loss": 1.9949,
"step": 1769
},
{
"epoch": 0.8118335053319573,
"grad_norm": 0.21017447113990784,
"learning_rate": 8.997461168625138e-05,
"loss": 1.9137,
"step": 1770
},
{
"epoch": 0.81229216832932,
"grad_norm": 0.226249560713768,
"learning_rate": 8.954982702860664e-05,
"loss": 1.8867,
"step": 1771
},
{
"epoch": 0.8127508313266827,
"grad_norm": 0.21902820467948914,
"learning_rate": 8.912594883408865e-05,
"loss": 1.9958,
"step": 1772
},
{
"epoch": 0.8132094943240454,
"grad_norm": 0.21206340193748474,
"learning_rate": 8.870297803881589e-05,
"loss": 1.9158,
"step": 1773
},
{
"epoch": 0.8136681573214081,
"grad_norm": 0.19799992442131042,
"learning_rate": 8.828091557690287e-05,
"loss": 1.9842,
"step": 1774
},
{
"epoch": 0.8141268203187708,
"grad_norm": 0.21157929301261902,
"learning_rate": 8.785976238045801e-05,
"loss": 2.081,
"step": 1775
},
{
"epoch": 0.8145854833161335,
"grad_norm": 0.2175525277853012,
"learning_rate": 8.743951937958144e-05,
"loss": 1.9553,
"step": 1776
},
{
"epoch": 0.8150441463134962,
"grad_norm": 0.2139703631401062,
"learning_rate": 8.702018750236357e-05,
"loss": 1.9289,
"step": 1777
},
{
"epoch": 0.8155028093108588,
"grad_norm": 0.21208608150482178,
"learning_rate": 8.660176767488237e-05,
"loss": 1.9366,
"step": 1778
},
{
"epoch": 0.8159614723082216,
"grad_norm": 0.2037891149520874,
"learning_rate": 8.618426082120146e-05,
"loss": 1.9365,
"step": 1779
},
{
"epoch": 0.8164201353055842,
"grad_norm": 0.21880751848220825,
"learning_rate": 8.576766786336854e-05,
"loss": 1.9774,
"step": 1780
},
{
"epoch": 0.816878798302947,
"grad_norm": 0.2151077836751938,
"learning_rate": 8.535198972141294e-05,
"loss": 2.0369,
"step": 1781
},
{
"epoch": 0.8173374613003096,
"grad_norm": 0.2105722725391388,
"learning_rate": 8.493722731334347e-05,
"loss": 1.953,
"step": 1782
},
{
"epoch": 0.8177961242976722,
"grad_norm": 0.2058856338262558,
"learning_rate": 8.452338155514644e-05,
"loss": 1.8083,
"step": 1783
},
{
"epoch": 0.818254787295035,
"grad_norm": 0.20629949867725372,
"learning_rate": 8.411045336078426e-05,
"loss": 1.9848,
"step": 1784
},
{
"epoch": 0.8187134502923976,
"grad_norm": 0.20577141642570496,
"learning_rate": 8.369844364219264e-05,
"loss": 1.9843,
"step": 1785
},
{
"epoch": 0.8191721132897604,
"grad_norm": 0.20712928473949432,
"learning_rate": 8.328735330927873e-05,
"loss": 1.9598,
"step": 1786
},
{
"epoch": 0.819630776287123,
"grad_norm": 0.21145014464855194,
"learning_rate": 8.287718326991961e-05,
"loss": 2.0169,
"step": 1787
},
{
"epoch": 0.8200894392844857,
"grad_norm": 0.21547095477581024,
"learning_rate": 8.246793442995954e-05,
"loss": 1.9437,
"step": 1788
},
{
"epoch": 0.8205481022818484,
"grad_norm": 0.2225932478904724,
"learning_rate": 8.205960769320875e-05,
"loss": 1.952,
"step": 1789
},
{
"epoch": 0.8210067652792111,
"grad_norm": 0.21980445086956024,
"learning_rate": 8.165220396144085e-05,
"loss": 1.9833,
"step": 1790
},
{
"epoch": 0.8214654282765738,
"grad_norm": 0.20865364372730255,
"learning_rate": 8.12457241343909e-05,
"loss": 1.9661,
"step": 1791
},
{
"epoch": 0.8219240912739365,
"grad_norm": 0.21497952938079834,
"learning_rate": 8.084016910975367e-05,
"loss": 1.9688,
"step": 1792
},
{
"epoch": 0.8223827542712991,
"grad_norm": 0.21235626935958862,
"learning_rate": 8.043553978318169e-05,
"loss": 1.9721,
"step": 1793
},
{
"epoch": 0.8228414172686619,
"grad_norm": 0.20945817232131958,
"learning_rate": 8.003183704828281e-05,
"loss": 2.1,
"step": 1794
},
{
"epoch": 0.8233000802660245,
"grad_norm": 0.2188664972782135,
"learning_rate": 7.962906179661872e-05,
"loss": 1.8637,
"step": 1795
},
{
"epoch": 0.8237587432633873,
"grad_norm": 0.21034635603427887,
"learning_rate": 7.922721491770296e-05,
"loss": 1.9373,
"step": 1796
},
{
"epoch": 0.8242174062607499,
"grad_norm": 0.21669074892997742,
"learning_rate": 7.882629729899832e-05,
"loss": 2.0551,
"step": 1797
},
{
"epoch": 0.8246760692581127,
"grad_norm": 0.21942898631095886,
"learning_rate": 7.842630982591598e-05,
"loss": 2.0542,
"step": 1798
},
{
"epoch": 0.8251347322554753,
"grad_norm": 0.2175815850496292,
"learning_rate": 7.802725338181232e-05,
"loss": 1.9731,
"step": 1799
},
{
"epoch": 0.8255933952528379,
"grad_norm": 0.21761317551136017,
"learning_rate": 7.762912884798812e-05,
"loss": 1.9458,
"step": 1800
},
{
"epoch": 0.8260520582502007,
"grad_norm": 0.20805209875106812,
"learning_rate": 7.723193710368564e-05,
"loss": 1.9239,
"step": 1801
},
{
"epoch": 0.8265107212475633,
"grad_norm": 0.2184089571237564,
"learning_rate": 7.683567902608729e-05,
"loss": 1.9253,
"step": 1802
},
{
"epoch": 0.826969384244926,
"grad_norm": 0.21594616770744324,
"learning_rate": 7.644035549031364e-05,
"loss": 1.8741,
"step": 1803
},
{
"epoch": 0.8274280472422887,
"grad_norm": 0.217102512717247,
"learning_rate": 7.604596736942115e-05,
"loss": 2.0182,
"step": 1804
},
{
"epoch": 0.8278867102396514,
"grad_norm": 0.23005014657974243,
"learning_rate": 7.56525155344004e-05,
"loss": 1.9765,
"step": 1805
},
{
"epoch": 0.8283453732370141,
"grad_norm": 0.22077760100364685,
"learning_rate": 7.52600008541745e-05,
"loss": 1.9795,
"step": 1806
},
{
"epoch": 0.8288040362343768,
"grad_norm": 0.21680955588817596,
"learning_rate": 7.486842419559681e-05,
"loss": 1.9761,
"step": 1807
},
{
"epoch": 0.8292626992317395,
"grad_norm": 0.22827918827533722,
"learning_rate": 7.447778642344898e-05,
"loss": 1.9791,
"step": 1808
},
{
"epoch": 0.8297213622291022,
"grad_norm": 0.22785818576812744,
"learning_rate": 7.408808840043912e-05,
"loss": 1.9541,
"step": 1809
},
{
"epoch": 0.8301800252264648,
"grad_norm": 0.22585523128509521,
"learning_rate": 7.369933098720021e-05,
"loss": 2.0336,
"step": 1810
},
{
"epoch": 0.8306386882238276,
"grad_norm": 0.21963390707969666,
"learning_rate": 7.331151504228767e-05,
"loss": 1.9425,
"step": 1811
},
{
"epoch": 0.8310973512211902,
"grad_norm": 0.21416924893856049,
"learning_rate": 7.292464142217775e-05,
"loss": 1.9774,
"step": 1812
},
{
"epoch": 0.831556014218553,
"grad_norm": 0.22111089527606964,
"learning_rate": 7.25387109812658e-05,
"loss": 2.0149,
"step": 1813
},
{
"epoch": 0.8320146772159156,
"grad_norm": 0.21024999022483826,
"learning_rate": 7.215372457186415e-05,
"loss": 1.9416,
"step": 1814
},
{
"epoch": 0.8324733402132783,
"grad_norm": 0.2175164818763733,
"learning_rate": 7.176968304420007e-05,
"loss": 2.0588,
"step": 1815
},
{
"epoch": 0.832932003210641,
"grad_norm": 0.22124679386615753,
"learning_rate": 7.138658724641417e-05,
"loss": 1.9503,
"step": 1816
},
{
"epoch": 0.8333906662080036,
"grad_norm": 0.20970623195171356,
"learning_rate": 7.10044380245587e-05,
"loss": 1.9712,
"step": 1817
},
{
"epoch": 0.8338493292053664,
"grad_norm": 0.20987161993980408,
"learning_rate": 7.062323622259515e-05,
"loss": 1.9724,
"step": 1818
},
{
"epoch": 0.834307992202729,
"grad_norm": 0.21737320721149445,
"learning_rate": 7.024298268239265e-05,
"loss": 1.9229,
"step": 1819
},
{
"epoch": 0.8347666552000917,
"grad_norm": 0.20847219228744507,
"learning_rate": 6.986367824372647e-05,
"loss": 1.9851,
"step": 1820
},
{
"epoch": 0.8352253181974544,
"grad_norm": 0.22695927321910858,
"learning_rate": 6.948532374427541e-05,
"loss": 1.9853,
"step": 1821
},
{
"epoch": 0.8356839811948171,
"grad_norm": 0.23395609855651855,
"learning_rate": 6.910792001962063e-05,
"loss": 2.0039,
"step": 1822
},
{
"epoch": 0.8361426441921798,
"grad_norm": 0.20454825460910797,
"learning_rate": 6.873146790324358e-05,
"loss": 1.953,
"step": 1823
},
{
"epoch": 0.8366013071895425,
"grad_norm": 0.20532380044460297,
"learning_rate": 6.83559682265239e-05,
"loss": 2.0127,
"step": 1824
},
{
"epoch": 0.8370599701869051,
"grad_norm": 0.20612367987632751,
"learning_rate": 6.798142181873784e-05,
"loss": 1.9187,
"step": 1825
},
{
"epoch": 0.8375186331842679,
"grad_norm": 0.2154759168624878,
"learning_rate": 6.760782950705662e-05,
"loss": 1.9408,
"step": 1826
},
{
"epoch": 0.8379772961816305,
"grad_norm": 0.21940329670906067,
"learning_rate": 6.723519211654422e-05,
"loss": 2.0246,
"step": 1827
},
{
"epoch": 0.8384359591789933,
"grad_norm": 0.2025623470544815,
"learning_rate": 6.686351047015554e-05,
"loss": 1.9592,
"step": 1828
},
{
"epoch": 0.8388946221763559,
"grad_norm": 0.2119021713733673,
"learning_rate": 6.649278538873515e-05,
"loss": 1.9258,
"step": 1829
},
{
"epoch": 0.8393532851737187,
"grad_norm": 0.21875682473182678,
"learning_rate": 6.612301769101465e-05,
"loss": 1.9592,
"step": 1830
},
{
"epoch": 0.8398119481710813,
"grad_norm": 0.21817618608474731,
"learning_rate": 6.575420819361177e-05,
"loss": 1.9494,
"step": 1831
},
{
"epoch": 0.840270611168444,
"grad_norm": 0.21169809997081757,
"learning_rate": 6.538635771102757e-05,
"loss": 1.9672,
"step": 1832
},
{
"epoch": 0.8407292741658067,
"grad_norm": 0.20909613370895386,
"learning_rate": 6.501946705564566e-05,
"loss": 2.0119,
"step": 1833
},
{
"epoch": 0.8411879371631693,
"grad_norm": 0.2298404425382614,
"learning_rate": 6.465353703772959e-05,
"loss": 2.0004,
"step": 1834
},
{
"epoch": 0.841646600160532,
"grad_norm": 0.22172114253044128,
"learning_rate": 6.428856846542136e-05,
"loss": 1.9698,
"step": 1835
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.20185887813568115,
"learning_rate": 6.392456214473996e-05,
"loss": 1.8932,
"step": 1836
},
{
"epoch": 0.8425639261552574,
"grad_norm": 0.21903829276561737,
"learning_rate": 6.3561518879579e-05,
"loss": 2.0053,
"step": 1837
},
{
"epoch": 0.8430225891526201,
"grad_norm": 0.21103151142597198,
"learning_rate": 6.31994394717052e-05,
"loss": 1.9789,
"step": 1838
},
{
"epoch": 0.8434812521499828,
"grad_norm": 0.2163507044315338,
"learning_rate": 6.283832472075685e-05,
"loss": 2.0329,
"step": 1839
},
{
"epoch": 0.8439399151473455,
"grad_norm": 0.20768873393535614,
"learning_rate": 6.247817542424178e-05,
"loss": 1.8401,
"step": 1840
},
{
"epoch": 0.8443985781447082,
"grad_norm": 0.21500693261623383,
"learning_rate": 6.211899237753559e-05,
"loss": 2.0292,
"step": 1841
},
{
"epoch": 0.8448572411420708,
"grad_norm": 0.2255236655473709,
"learning_rate": 6.176077637387984e-05,
"loss": 2.0089,
"step": 1842
},
{
"epoch": 0.8453159041394336,
"grad_norm": 0.2072249948978424,
"learning_rate": 6.140352820438066e-05,
"loss": 1.9659,
"step": 1843
},
{
"epoch": 0.8457745671367962,
"grad_norm": 0.21149484813213348,
"learning_rate": 6.104724865800665e-05,
"loss": 1.9802,
"step": 1844
},
{
"epoch": 0.846233230134159,
"grad_norm": 0.22263088822364807,
"learning_rate": 6.069193852158711e-05,
"loss": 1.9791,
"step": 1845
},
{
"epoch": 0.8466918931315216,
"grad_norm": 0.22894656658172607,
"learning_rate": 6.0337598579810584e-05,
"loss": 2.0276,
"step": 1846
},
{
"epoch": 0.8471505561288843,
"grad_norm": 0.21203044056892395,
"learning_rate": 5.9984229615223096e-05,
"loss": 1.9466,
"step": 1847
},
{
"epoch": 0.847609219126247,
"grad_norm": 0.22357547283172607,
"learning_rate": 5.963183240822606e-05,
"loss": 1.9474,
"step": 1848
},
{
"epoch": 0.8480678821236097,
"grad_norm": 0.20883163809776306,
"learning_rate": 5.9280407737074825e-05,
"loss": 1.8745,
"step": 1849
},
{
"epoch": 0.8485265451209724,
"grad_norm": 0.20360969007015228,
"learning_rate": 5.8929956377877125e-05,
"loss": 1.9975,
"step": 1850
},
{
"epoch": 0.848985208118335,
"grad_norm": 0.22608418762683868,
"learning_rate": 5.8580479104591075e-05,
"loss": 1.9358,
"step": 1851
},
{
"epoch": 0.8494438711156977,
"grad_norm": 0.21142396330833435,
"learning_rate": 5.823197668902341e-05,
"loss": 2.0163,
"step": 1852
},
{
"epoch": 0.8499025341130604,
"grad_norm": 0.21048052608966827,
"learning_rate": 5.78844499008282e-05,
"loss": 1.906,
"step": 1853
},
{
"epoch": 0.8503611971104231,
"grad_norm": 0.22009438276290894,
"learning_rate": 5.753789950750454e-05,
"loss": 1.9669,
"step": 1854
},
{
"epoch": 0.8508198601077858,
"grad_norm": 0.21856717765331268,
"learning_rate": 5.719232627439558e-05,
"loss": 1.8656,
"step": 1855
},
{
"epoch": 0.8512785231051485,
"grad_norm": 0.2113906741142273,
"learning_rate": 5.6847730964686315e-05,
"loss": 1.9565,
"step": 1856
},
{
"epoch": 0.8517371861025111,
"grad_norm": 0.21547025442123413,
"learning_rate": 5.650411433940189e-05,
"loss": 1.9999,
"step": 1857
},
{
"epoch": 0.8521958490998739,
"grad_norm": 0.2205989956855774,
"learning_rate": 5.61614771574061e-05,
"loss": 1.9514,
"step": 1858
},
{
"epoch": 0.8526545120972365,
"grad_norm": 0.20781150460243225,
"learning_rate": 5.581982017539988e-05,
"loss": 1.955,
"step": 1859
},
{
"epoch": 0.8531131750945993,
"grad_norm": 0.23161308467388153,
"learning_rate": 5.5479144147919216e-05,
"loss": 1.8636,
"step": 1860
},
{
"epoch": 0.8535718380919619,
"grad_norm": 0.21077843010425568,
"learning_rate": 5.51394498273336e-05,
"loss": 1.8628,
"step": 1861
},
{
"epoch": 0.8540305010893247,
"grad_norm": 0.22295503318309784,
"learning_rate": 5.480073796384494e-05,
"loss": 1.9196,
"step": 1862
},
{
"epoch": 0.8544891640866873,
"grad_norm": 0.2204519361257553,
"learning_rate": 5.446300930548492e-05,
"loss": 1.976,
"step": 1863
},
{
"epoch": 0.85494782708405,
"grad_norm": 0.21935133635997772,
"learning_rate": 5.412626459811415e-05,
"loss": 1.8926,
"step": 1864
},
{
"epoch": 0.8554064900814127,
"grad_norm": 0.21474871039390564,
"learning_rate": 5.3790504585419954e-05,
"loss": 1.9433,
"step": 1865
},
{
"epoch": 0.8558651530787754,
"grad_norm": 0.21786822378635406,
"learning_rate": 5.345573000891541e-05,
"loss": 1.9395,
"step": 1866
},
{
"epoch": 0.8563238160761381,
"grad_norm": 0.21752221882343292,
"learning_rate": 5.312194160793693e-05,
"loss": 1.9556,
"step": 1867
},
{
"epoch": 0.8567824790735007,
"grad_norm": 0.23508411645889282,
"learning_rate": 5.278914011964303e-05,
"loss": 1.8817,
"step": 1868
},
{
"epoch": 0.8572411420708634,
"grad_norm": 0.2097083032131195,
"learning_rate": 5.2457326279013006e-05,
"loss": 2.0136,
"step": 1869
},
{
"epoch": 0.8576998050682261,
"grad_norm": 0.23794038593769073,
"learning_rate": 5.2126500818844514e-05,
"loss": 1.9534,
"step": 1870
},
{
"epoch": 0.8581584680655888,
"grad_norm": 0.2141590267419815,
"learning_rate": 5.1796664469752566e-05,
"loss": 1.8467,
"step": 1871
},
{
"epoch": 0.8586171310629515,
"grad_norm": 0.21538317203521729,
"learning_rate": 5.1467817960167975e-05,
"loss": 1.9833,
"step": 1872
},
{
"epoch": 0.8590757940603142,
"grad_norm": 0.21154989302158356,
"learning_rate": 5.113996201633536e-05,
"loss": 1.9261,
"step": 1873
},
{
"epoch": 0.8595344570576768,
"grad_norm": 0.22566799819469452,
"learning_rate": 5.0813097362311765e-05,
"loss": 1.9581,
"step": 1874
},
{
"epoch": 0.8599931200550396,
"grad_norm": 0.21788544952869415,
"learning_rate": 5.048722471996475e-05,
"loss": 1.9572,
"step": 1875
},
{
"epoch": 0.8604517830524022,
"grad_norm": 0.21963658928871155,
"learning_rate": 5.016234480897158e-05,
"loss": 1.9655,
"step": 1876
},
{
"epoch": 0.860910446049765,
"grad_norm": 0.2191489338874817,
"learning_rate": 4.9838458346816664e-05,
"loss": 1.9048,
"step": 1877
},
{
"epoch": 0.8613691090471276,
"grad_norm": 0.21286913752555847,
"learning_rate": 4.9515566048790485e-05,
"loss": 1.9438,
"step": 1878
},
{
"epoch": 0.8618277720444903,
"grad_norm": 0.19762615859508514,
"learning_rate": 4.9193668627988074e-05,
"loss": 1.8924,
"step": 1879
},
{
"epoch": 0.862286435041853,
"grad_norm": 0.21391743421554565,
"learning_rate": 4.887276679530744e-05,
"loss": 1.9776,
"step": 1880
},
{
"epoch": 0.8627450980392157,
"grad_norm": 0.22335486114025116,
"learning_rate": 4.855286125944752e-05,
"loss": 1.9659,
"step": 1881
},
{
"epoch": 0.8632037610365784,
"grad_norm": 0.21024760603904724,
"learning_rate": 4.8233952726907224e-05,
"loss": 1.8403,
"step": 1882
},
{
"epoch": 0.8636624240339411,
"grad_norm": 0.23468610644340515,
"learning_rate": 4.7916041901983565e-05,
"loss": 1.9811,
"step": 1883
},
{
"epoch": 0.8641210870313037,
"grad_norm": 0.21268507838249207,
"learning_rate": 4.7599129486770145e-05,
"loss": 1.9579,
"step": 1884
},
{
"epoch": 0.8645797500286664,
"grad_norm": 0.2174464911222458,
"learning_rate": 4.728321618115555e-05,
"loss": 2.087,
"step": 1885
},
{
"epoch": 0.8650384130260291,
"grad_norm": 0.22840915620326996,
"learning_rate": 4.696830268282204e-05,
"loss": 1.9801,
"step": 1886
},
{
"epoch": 0.8654970760233918,
"grad_norm": 0.22013407945632935,
"learning_rate": 4.665438968724361e-05,
"loss": 1.9836,
"step": 1887
},
{
"epoch": 0.8659557390207545,
"grad_norm": 0.22799570858478546,
"learning_rate": 4.634147788768489e-05,
"loss": 1.9298,
"step": 1888
},
{
"epoch": 0.8664144020181171,
"grad_norm": 0.22210249304771423,
"learning_rate": 4.6029567975199414e-05,
"loss": 1.9866,
"step": 1889
},
{
"epoch": 0.8668730650154799,
"grad_norm": 0.21139603853225708,
"learning_rate": 4.571866063862795e-05,
"loss": 1.9141,
"step": 1890
},
{
"epoch": 0.8673317280128425,
"grad_norm": 0.21694214642047882,
"learning_rate": 4.540875656459703e-05,
"loss": 1.9206,
"step": 1891
},
{
"epoch": 0.8677903910102053,
"grad_norm": 0.22073958814144135,
"learning_rate": 4.509985643751785e-05,
"loss": 2.0198,
"step": 1892
},
{
"epoch": 0.8682490540075679,
"grad_norm": 0.21395589411258698,
"learning_rate": 4.479196093958421e-05,
"loss": 2.0859,
"step": 1893
},
{
"epoch": 0.8687077170049307,
"grad_norm": 0.22524289786815643,
"learning_rate": 4.4485070750771187e-05,
"loss": 1.9544,
"step": 1894
},
{
"epoch": 0.8691663800022933,
"grad_norm": 0.21504966914653778,
"learning_rate": 4.417918654883363e-05,
"loss": 1.9162,
"step": 1895
},
{
"epoch": 0.869625042999656,
"grad_norm": 0.2223048359155655,
"learning_rate": 4.3874309009305e-05,
"loss": 1.8844,
"step": 1896
},
{
"epoch": 0.8700837059970187,
"grad_norm": 0.22380509972572327,
"learning_rate": 4.357043880549538e-05,
"loss": 1.9807,
"step": 1897
},
{
"epoch": 0.8705423689943814,
"grad_norm": 0.21838629245758057,
"learning_rate": 4.326757660849012e-05,
"loss": 2.0224,
"step": 1898
},
{
"epoch": 0.8710010319917441,
"grad_norm": 0.21738266944885254,
"learning_rate": 4.2965723087148635e-05,
"loss": 2.0143,
"step": 1899
},
{
"epoch": 0.8714596949891068,
"grad_norm": 0.20904994010925293,
"learning_rate": 4.266487890810256e-05,
"loss": 1.9482,
"step": 1900
},
{
"epoch": 0.8719183579864694,
"grad_norm": 0.21268314123153687,
"learning_rate": 4.2365044735754365e-05,
"loss": 1.8887,
"step": 1901
},
{
"epoch": 0.8723770209838321,
"grad_norm": 0.223023921251297,
"learning_rate": 4.2066221232276266e-05,
"loss": 2.0963,
"step": 1902
},
{
"epoch": 0.8728356839811948,
"grad_norm": 0.22754767537117004,
"learning_rate": 4.176840905760815e-05,
"loss": 1.9945,
"step": 1903
},
{
"epoch": 0.8732943469785575,
"grad_norm": 0.2097444236278534,
"learning_rate": 4.1471608869456443e-05,
"loss": 1.8991,
"step": 1904
},
{
"epoch": 0.8737530099759202,
"grad_norm": 0.21096356213092804,
"learning_rate": 4.117582132329284e-05,
"loss": 2.0542,
"step": 1905
},
{
"epoch": 0.8742116729732828,
"grad_norm": 0.21985404193401337,
"learning_rate": 4.088104707235263e-05,
"loss": 2.0786,
"step": 1906
},
{
"epoch": 0.8746703359706456,
"grad_norm": 0.21271944046020508,
"learning_rate": 4.058728676763313e-05,
"loss": 1.9849,
"step": 1907
},
{
"epoch": 0.8751289989680082,
"grad_norm": 0.21331104636192322,
"learning_rate": 4.0294541057892375e-05,
"loss": 2.0167,
"step": 1908
},
{
"epoch": 0.875587661965371,
"grad_norm": 0.21177497506141663,
"learning_rate": 4.000281058964794e-05,
"loss": 1.9857,
"step": 1909
},
{
"epoch": 0.8760463249627336,
"grad_norm": 0.2141164392232895,
"learning_rate": 3.971209600717507e-05,
"loss": 1.9253,
"step": 1910
},
{
"epoch": 0.8765049879600963,
"grad_norm": 0.21916434168815613,
"learning_rate": 3.9422397952505465e-05,
"loss": 1.8727,
"step": 1911
},
{
"epoch": 0.876963650957459,
"grad_norm": 0.2155229151248932,
"learning_rate": 3.913371706542596e-05,
"loss": 1.9827,
"step": 1912
},
{
"epoch": 0.8774223139548217,
"grad_norm": 0.2108934074640274,
"learning_rate": 3.884605398347707e-05,
"loss": 2.0235,
"step": 1913
},
{
"epoch": 0.8778809769521844,
"grad_norm": 0.2142520397901535,
"learning_rate": 3.8559409341951456e-05,
"loss": 1.9938,
"step": 1914
},
{
"epoch": 0.8783396399495471,
"grad_norm": 0.2252727448940277,
"learning_rate": 3.8273783773892404e-05,
"loss": 1.9512,
"step": 1915
},
{
"epoch": 0.8787983029469097,
"grad_norm": 0.21326278150081635,
"learning_rate": 3.798917791009293e-05,
"loss": 1.988,
"step": 1916
},
{
"epoch": 0.8792569659442725,
"grad_norm": 0.22047008574008942,
"learning_rate": 3.770559237909393e-05,
"loss": 2.0689,
"step": 1917
},
{
"epoch": 0.8797156289416351,
"grad_norm": 0.22180549800395966,
"learning_rate": 3.742302780718288e-05,
"loss": 1.8787,
"step": 1918
},
{
"epoch": 0.8801742919389978,
"grad_norm": 0.22128057479858398,
"learning_rate": 3.7141484818392635e-05,
"loss": 2.0022,
"step": 1919
},
{
"epoch": 0.8806329549363605,
"grad_norm": 0.2186572402715683,
"learning_rate": 3.686096403449973e-05,
"loss": 1.8856,
"step": 1920
},
{
"epoch": 0.8810916179337231,
"grad_norm": 0.21626652777194977,
"learning_rate": 3.658146607502344e-05,
"loss": 1.9652,
"step": 1921
},
{
"epoch": 0.8815502809310859,
"grad_norm": 0.21364837884902954,
"learning_rate": 3.630299155722411e-05,
"loss": 1.9602,
"step": 1922
},
{
"epoch": 0.8820089439284485,
"grad_norm": 0.20267778635025024,
"learning_rate": 3.6025541096101676e-05,
"loss": 1.9561,
"step": 1923
},
{
"epoch": 0.8824676069258113,
"grad_norm": 0.2140640765428543,
"learning_rate": 3.574911530439473e-05,
"loss": 1.9656,
"step": 1924
},
{
"epoch": 0.8829262699231739,
"grad_norm": 0.2372148036956787,
"learning_rate": 3.5473714792578606e-05,
"loss": 2.0418,
"step": 1925
},
{
"epoch": 0.8833849329205367,
"grad_norm": 0.22110199928283691,
"learning_rate": 3.519934016886478e-05,
"loss": 1.9337,
"step": 1926
},
{
"epoch": 0.8838435959178993,
"grad_norm": 0.20307524502277374,
"learning_rate": 3.4925992039198776e-05,
"loss": 1.9183,
"step": 1927
},
{
"epoch": 0.884302258915262,
"grad_norm": 0.21610182523727417,
"learning_rate": 3.465367100725908e-05,
"loss": 1.9351,
"step": 1928
},
{
"epoch": 0.8847609219126247,
"grad_norm": 0.21621868014335632,
"learning_rate": 3.438237767445618e-05,
"loss": 1.938,
"step": 1929
},
{
"epoch": 0.8852195849099874,
"grad_norm": 0.21566170454025269,
"learning_rate": 3.4112112639930804e-05,
"loss": 2.0859,
"step": 1930
},
{
"epoch": 0.8856782479073501,
"grad_norm": 0.21650229394435883,
"learning_rate": 3.3842876500552564e-05,
"loss": 1.9329,
"step": 1931
},
{
"epoch": 0.8861369109047128,
"grad_norm": 0.2132548689842224,
"learning_rate": 3.357466985091906e-05,
"loss": 1.9457,
"step": 1932
},
{
"epoch": 0.8865955739020754,
"grad_norm": 0.2157929241657257,
"learning_rate": 3.330749328335414e-05,
"loss": 1.9869,
"step": 1933
},
{
"epoch": 0.8870542368994382,
"grad_norm": 0.22266636788845062,
"learning_rate": 3.304134738790659e-05,
"loss": 1.9345,
"step": 1934
},
{
"epoch": 0.8875128998968008,
"grad_norm": 0.21370664238929749,
"learning_rate": 3.277623275234953e-05,
"loss": 1.886,
"step": 1935
},
{
"epoch": 0.8879715628941636,
"grad_norm": 0.21012216806411743,
"learning_rate": 3.2512149962177994e-05,
"loss": 1.9103,
"step": 1936
},
{
"epoch": 0.8884302258915262,
"grad_norm": 0.20435811579227448,
"learning_rate": 3.224909960060851e-05,
"loss": 1.9953,
"step": 1937
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.22254841029644012,
"learning_rate": 3.198708224857755e-05,
"loss": 2.0026,
"step": 1938
},
{
"epoch": 0.8893475518862516,
"grad_norm": 0.22595800459384918,
"learning_rate": 3.172609848474023e-05,
"loss": 1.9727,
"step": 1939
},
{
"epoch": 0.8898062148836142,
"grad_norm": 0.2171122282743454,
"learning_rate": 3.1466148885468895e-05,
"loss": 1.9,
"step": 1940
},
{
"epoch": 0.890264877880977,
"grad_norm": 0.22012001276016235,
"learning_rate": 3.120723402485198e-05,
"loss": 1.8445,
"step": 1941
},
{
"epoch": 0.8907235408783396,
"grad_norm": 0.21668146550655365,
"learning_rate": 3.094935447469294e-05,
"loss": 1.915,
"step": 1942
},
{
"epoch": 0.8911822038757024,
"grad_norm": 0.21970802545547485,
"learning_rate": 3.069251080450863e-05,
"loss": 1.9733,
"step": 1943
},
{
"epoch": 0.891640866873065,
"grad_norm": 0.22544094920158386,
"learning_rate": 3.0436703581528113e-05,
"loss": 1.9367,
"step": 1944
},
{
"epoch": 0.8920995298704277,
"grad_norm": 0.22278785705566406,
"learning_rate": 3.0181933370691694e-05,
"loss": 1.9593,
"step": 1945
},
{
"epoch": 0.8925581928677904,
"grad_norm": 0.21345330774784088,
"learning_rate": 2.9928200734649523e-05,
"loss": 1.9914,
"step": 1946
},
{
"epoch": 0.8930168558651531,
"grad_norm": 0.21392430365085602,
"learning_rate": 2.9675506233760142e-05,
"loss": 1.87,
"step": 1947
},
{
"epoch": 0.8934755188625157,
"grad_norm": 0.22131407260894775,
"learning_rate": 2.942385042608925e-05,
"loss": 1.9424,
"step": 1948
},
{
"epoch": 0.8939341818598785,
"grad_norm": 0.2129591405391693,
"learning_rate": 2.9173233867409054e-05,
"loss": 1.9021,
"step": 1949
},
{
"epoch": 0.8943928448572411,
"grad_norm": 0.222428098320961,
"learning_rate": 2.892365711119638e-05,
"loss": 1.9239,
"step": 1950
},
{
"epoch": 0.8948515078546039,
"grad_norm": 0.21448694169521332,
"learning_rate": 2.8675120708631596e-05,
"loss": 2.0576,
"step": 1951
},
{
"epoch": 0.8953101708519665,
"grad_norm": 0.22179824113845825,
"learning_rate": 2.8427625208597764e-05,
"loss": 1.9528,
"step": 1952
},
{
"epoch": 0.8957688338493293,
"grad_norm": 0.2197989523410797,
"learning_rate": 2.8181171157678874e-05,
"loss": 1.8644,
"step": 1953
},
{
"epoch": 0.8962274968466919,
"grad_norm": 0.21610473096370697,
"learning_rate": 2.7935759100159053e-05,
"loss": 1.9953,
"step": 1954
},
{
"epoch": 0.8966861598440545,
"grad_norm": 0.22994394600391388,
"learning_rate": 2.7691389578021365e-05,
"loss": 1.9462,
"step": 1955
},
{
"epoch": 0.8971448228414173,
"grad_norm": 0.21879933774471283,
"learning_rate": 2.7448063130946223e-05,
"loss": 1.9432,
"step": 1956
},
{
"epoch": 0.8976034858387799,
"grad_norm": 0.2237020879983902,
"learning_rate": 2.7205780296310544e-05,
"loss": 1.9035,
"step": 1957
},
{
"epoch": 0.8980621488361427,
"grad_norm": 0.2197963148355484,
"learning_rate": 2.6964541609186378e-05,
"loss": 2.0552,
"step": 1958
},
{
"epoch": 0.8985208118335053,
"grad_norm": 0.21400946378707886,
"learning_rate": 2.6724347602340104e-05,
"loss": 2.0119,
"step": 1959
},
{
"epoch": 0.898979474830868,
"grad_norm": 0.2101636379957199,
"learning_rate": 2.6485198806230682e-05,
"loss": 1.9704,
"step": 1960
},
{
"epoch": 0.8994381378282307,
"grad_norm": 0.2181745320558548,
"learning_rate": 2.6247095749008797e-05,
"loss": 1.9794,
"step": 1961
},
{
"epoch": 0.8998968008255934,
"grad_norm": 0.22878077626228333,
"learning_rate": 2.6010038956515826e-05,
"loss": 1.9887,
"step": 1962
},
{
"epoch": 0.9003554638229561,
"grad_norm": 0.22518637776374817,
"learning_rate": 2.5774028952282423e-05,
"loss": 1.9926,
"step": 1963
},
{
"epoch": 0.9008141268203188,
"grad_norm": 0.2122706174850464,
"learning_rate": 2.5539066257527277e-05,
"loss": 1.969,
"step": 1964
},
{
"epoch": 0.9012727898176814,
"grad_norm": 0.21740639209747314,
"learning_rate": 2.530515139115652e-05,
"loss": 1.9705,
"step": 1965
},
{
"epoch": 0.9017314528150442,
"grad_norm": 0.20997415482997894,
"learning_rate": 2.5072284869761874e-05,
"loss": 1.9578,
"step": 1966
},
{
"epoch": 0.9021901158124068,
"grad_norm": 0.21009652316570282,
"learning_rate": 2.4840467207619786e-05,
"loss": 1.8642,
"step": 1967
},
{
"epoch": 0.9026487788097696,
"grad_norm": 0.20143096148967743,
"learning_rate": 2.460969891669068e-05,
"loss": 1.9569,
"step": 1968
},
{
"epoch": 0.9031074418071322,
"grad_norm": 0.2204340696334839,
"learning_rate": 2.4379980506617272e-05,
"loss": 1.9089,
"step": 1969
},
{
"epoch": 0.903566104804495,
"grad_norm": 0.21802832186222076,
"learning_rate": 2.4151312484723464e-05,
"loss": 1.9209,
"step": 1970
},
{
"epoch": 0.9040247678018576,
"grad_norm": 0.20921742916107178,
"learning_rate": 2.3923695356013798e-05,
"loss": 1.942,
"step": 1971
},
{
"epoch": 0.9044834307992202,
"grad_norm": 0.21560555696487427,
"learning_rate": 2.3697129623171833e-05,
"loss": 1.855,
"step": 1972
},
{
"epoch": 0.904942093796583,
"grad_norm": 0.21814152598381042,
"learning_rate": 2.3471615786559042e-05,
"loss": 1.9651,
"step": 1973
},
{
"epoch": 0.9054007567939456,
"grad_norm": 0.21140599250793457,
"learning_rate": 2.3247154344213818e-05,
"loss": 2.0128,
"step": 1974
},
{
"epoch": 0.9058594197913084,
"grad_norm": 0.2205861210823059,
"learning_rate": 2.3023745791850625e-05,
"loss": 2.0581,
"step": 1975
},
{
"epoch": 0.906318082788671,
"grad_norm": 0.215336412191391,
"learning_rate": 2.2801390622858354e-05,
"loss": 1.9826,
"step": 1976
},
{
"epoch": 0.9067767457860337,
"grad_norm": 0.21261458098888397,
"learning_rate": 2.2580089328299746e-05,
"loss": 2.0285,
"step": 1977
},
{
"epoch": 0.9072354087833964,
"grad_norm": 0.21328283846378326,
"learning_rate": 2.235984239690997e-05,
"loss": 1.9586,
"step": 1978
},
{
"epoch": 0.9076940717807591,
"grad_norm": 0.22091887891292572,
"learning_rate": 2.2140650315095934e-05,
"loss": 1.8523,
"step": 1979
},
{
"epoch": 0.9081527347781218,
"grad_norm": 0.2180681824684143,
"learning_rate": 2.192251356693459e-05,
"loss": 1.9441,
"step": 1980
},
{
"epoch": 0.9086113977754845,
"grad_norm": 0.21652057766914368,
"learning_rate": 2.170543263417246e-05,
"loss": 1.922,
"step": 1981
},
{
"epoch": 0.9090700607728471,
"grad_norm": 0.21362507343292236,
"learning_rate": 2.1489407996224286e-05,
"loss": 1.9832,
"step": 1982
},
{
"epoch": 0.9095287237702099,
"grad_norm": 0.21520759165287018,
"learning_rate": 2.127444013017199e-05,
"loss": 2.0227,
"step": 1983
},
{
"epoch": 0.9099873867675725,
"grad_norm": 0.21869130432605743,
"learning_rate": 2.1060529510763648e-05,
"loss": 1.8989,
"step": 1984
},
{
"epoch": 0.9104460497649353,
"grad_norm": 0.22385789453983307,
"learning_rate": 2.084767661041259e-05,
"loss": 1.9403,
"step": 1985
},
{
"epoch": 0.9109047127622979,
"grad_norm": 0.21046306192874908,
"learning_rate": 2.063588189919596e-05,
"loss": 1.9581,
"step": 1986
},
{
"epoch": 0.9113633757596606,
"grad_norm": 0.2088364064693451,
"learning_rate": 2.0425145844854275e-05,
"loss": 1.9338,
"step": 1987
},
{
"epoch": 0.9118220387570233,
"grad_norm": 0.2254353016614914,
"learning_rate": 2.0215468912789693e-05,
"loss": 1.9923,
"step": 1988
},
{
"epoch": 0.9122807017543859,
"grad_norm": 0.21805663406848907,
"learning_rate": 2.0006851566065575e-05,
"loss": 1.9905,
"step": 1989
},
{
"epoch": 0.9127393647517487,
"grad_norm": 0.22075405716896057,
"learning_rate": 1.9799294265405166e-05,
"loss": 1.8899,
"step": 1990
},
{
"epoch": 0.9131980277491113,
"grad_norm": 0.21199987828731537,
"learning_rate": 1.9592797469190572e-05,
"loss": 2.0004,
"step": 1991
},
{
"epoch": 0.913656690746474,
"grad_norm": 0.23190243542194366,
"learning_rate": 1.938736163346194e-05,
"loss": 1.9152,
"step": 1992
},
{
"epoch": 0.9141153537438367,
"grad_norm": 0.2083907276391983,
"learning_rate": 1.9182987211916246e-05,
"loss": 2.0083,
"step": 1993
},
{
"epoch": 0.9145740167411994,
"grad_norm": 0.21573933959007263,
"learning_rate": 1.8979674655906332e-05,
"loss": 1.9931,
"step": 1994
},
{
"epoch": 0.9150326797385621,
"grad_norm": 0.21702571213245392,
"learning_rate": 1.8777424414440024e-05,
"loss": 1.985,
"step": 1995
},
{
"epoch": 0.9154913427359248,
"grad_norm": 0.21441228687763214,
"learning_rate": 1.8576236934179202e-05,
"loss": 1.9859,
"step": 1996
},
{
"epoch": 0.9159500057332874,
"grad_norm": 0.20592324435710907,
"learning_rate": 1.8376112659438393e-05,
"loss": 1.9083,
"step": 1997
},
{
"epoch": 0.9164086687306502,
"grad_norm": 0.2111009806394577,
"learning_rate": 1.8177052032184282e-05,
"loss": 1.8968,
"step": 1998
},
{
"epoch": 0.9168673317280128,
"grad_norm": 0.22085466980934143,
"learning_rate": 1.7979055492034435e-05,
"loss": 1.913,
"step": 1999
},
{
"epoch": 0.9173259947253756,
"grad_norm": 0.22350527346134186,
"learning_rate": 1.7782123476256407e-05,
"loss": 1.953,
"step": 2000
},
{
"epoch": 0.9177846577227382,
"grad_norm": 0.21731773018836975,
"learning_rate": 1.7586256419766965e-05,
"loss": 1.8946,
"step": 2001
},
{
"epoch": 0.918243320720101,
"grad_norm": 0.21339870989322662,
"learning_rate": 1.7391454755130766e-05,
"loss": 1.8475,
"step": 2002
},
{
"epoch": 0.9187019837174636,
"grad_norm": 0.21597647666931152,
"learning_rate": 1.7197718912559557e-05,
"loss": 1.9351,
"step": 2003
},
{
"epoch": 0.9191606467148263,
"grad_norm": 0.21743234992027283,
"learning_rate": 1.700504931991148e-05,
"loss": 1.9221,
"step": 2004
},
{
"epoch": 0.919619309712189,
"grad_norm": 0.21817663311958313,
"learning_rate": 1.681344640268978e-05,
"loss": 1.9185,
"step": 2005
},
{
"epoch": 0.9200779727095516,
"grad_norm": 0.21743422746658325,
"learning_rate": 1.6622910584041974e-05,
"loss": 1.9284,
"step": 2006
},
{
"epoch": 0.9205366357069144,
"grad_norm": 0.21452870965003967,
"learning_rate": 1.6433442284758903e-05,
"loss": 1.8983,
"step": 2007
},
{
"epoch": 0.920995298704277,
"grad_norm": 0.20577944815158844,
"learning_rate": 1.624504192327392e-05,
"loss": 1.989,
"step": 2008
},
{
"epoch": 0.9214539617016397,
"grad_norm": 0.21474146842956543,
"learning_rate": 1.6057709915661856e-05,
"loss": 1.8506,
"step": 2009
},
{
"epoch": 0.9219126246990024,
"grad_norm": 0.22081056237220764,
"learning_rate": 1.5871446675638057e-05,
"loss": 1.9627,
"step": 2010
},
{
"epoch": 0.9223712876963651,
"grad_norm": 0.22188882529735565,
"learning_rate": 1.5686252614557638e-05,
"loss": 1.9486,
"step": 2011
},
{
"epoch": 0.9228299506937278,
"grad_norm": 0.2160676270723343,
"learning_rate": 1.5502128141414497e-05,
"loss": 1.793,
"step": 2012
},
{
"epoch": 0.9232886136910905,
"grad_norm": 0.22649309039115906,
"learning_rate": 1.5319073662840188e-05,
"loss": 2.0279,
"step": 2013
},
{
"epoch": 0.9237472766884531,
"grad_norm": 0.20843824744224548,
"learning_rate": 1.5137089583103391e-05,
"loss": 1.9181,
"step": 2014
},
{
"epoch": 0.9242059396858159,
"grad_norm": 0.22371330857276917,
"learning_rate": 1.4956176304108893e-05,
"loss": 1.9333,
"step": 2015
},
{
"epoch": 0.9246646026831785,
"grad_norm": 0.21973450481891632,
"learning_rate": 1.4776334225396481e-05,
"loss": 1.9454,
"step": 2016
},
{
"epoch": 0.9251232656805413,
"grad_norm": 0.216129332780838,
"learning_rate": 1.4597563744140397e-05,
"loss": 1.9719,
"step": 2017
},
{
"epoch": 0.9255819286779039,
"grad_norm": 0.22348949313163757,
"learning_rate": 1.4419865255148269e-05,
"loss": 2.0441,
"step": 2018
},
{
"epoch": 0.9260405916752666,
"grad_norm": 0.2120838314294815,
"learning_rate": 1.4243239150860122e-05,
"loss": 1.9895,
"step": 2019
},
{
"epoch": 0.9264992546726293,
"grad_norm": 0.21578241884708405,
"learning_rate": 1.4067685821347932e-05,
"loss": 1.9861,
"step": 2020
},
{
"epoch": 0.926957917669992,
"grad_norm": 0.21596373617649078,
"learning_rate": 1.389320565431429e-05,
"loss": 1.9582,
"step": 2021
},
{
"epoch": 0.9274165806673547,
"grad_norm": 0.2289682924747467,
"learning_rate": 1.3719799035091851e-05,
"loss": 1.9728,
"step": 2022
},
{
"epoch": 0.9278752436647173,
"grad_norm": 0.22146721184253693,
"learning_rate": 1.3547466346642278e-05,
"loss": 1.9003,
"step": 2023
},
{
"epoch": 0.92833390666208,
"grad_norm": 0.2133312225341797,
"learning_rate": 1.3376207969555577e-05,
"loss": 1.9437,
"step": 2024
},
{
"epoch": 0.9287925696594427,
"grad_norm": 0.21331818401813507,
"learning_rate": 1.32060242820492e-05,
"loss": 1.9157,
"step": 2025
},
{
"epoch": 0.9292512326568054,
"grad_norm": 0.2094801664352417,
"learning_rate": 1.3036915659967118e-05,
"loss": 1.9674,
"step": 2026
},
{
"epoch": 0.9297098956541681,
"grad_norm": 0.2158496379852295,
"learning_rate": 1.2868882476779087e-05,
"loss": 1.9467,
"step": 2027
},
{
"epoch": 0.9301685586515308,
"grad_norm": 0.2160826474428177,
"learning_rate": 1.2701925103579815e-05,
"loss": 1.8989,
"step": 2028
},
{
"epoch": 0.9306272216488934,
"grad_norm": 0.21136772632598877,
"learning_rate": 1.2536043909088191e-05,
"loss": 1.9129,
"step": 2029
},
{
"epoch": 0.9310858846462562,
"grad_norm": 0.22172369062900543,
"learning_rate": 1.2371239259646228e-05,
"loss": 1.8711,
"step": 2030
},
{
"epoch": 0.9315445476436188,
"grad_norm": 0.2180812656879425,
"learning_rate": 1.2207511519218672e-05,
"loss": 1.9473,
"step": 2031
},
{
"epoch": 0.9320032106409816,
"grad_norm": 0.21696338057518005,
"learning_rate": 1.2044861049391676e-05,
"loss": 1.9055,
"step": 2032
},
{
"epoch": 0.9324618736383442,
"grad_norm": 0.2245176136493683,
"learning_rate": 1.1883288209372512e-05,
"loss": 2.0057,
"step": 2033
},
{
"epoch": 0.932920536635707,
"grad_norm": 0.22167469561100006,
"learning_rate": 1.1722793355988471e-05,
"loss": 1.9383,
"step": 2034
},
{
"epoch": 0.9333791996330696,
"grad_norm": 0.21972328424453735,
"learning_rate": 1.1563376843686135e-05,
"loss": 1.9263,
"step": 2035
},
{
"epoch": 0.9338378626304323,
"grad_norm": 0.21612678468227386,
"learning_rate": 1.140503902453055e-05,
"loss": 1.8956,
"step": 2036
},
{
"epoch": 0.934296525627795,
"grad_norm": 0.2241501808166504,
"learning_rate": 1.1247780248204665e-05,
"loss": 2.0174,
"step": 2037
},
{
"epoch": 0.9347551886251577,
"grad_norm": 0.21281364560127258,
"learning_rate": 1.1091600862008333e-05,
"loss": 1.918,
"step": 2038
},
{
"epoch": 0.9352138516225204,
"grad_norm": 0.22636374831199646,
"learning_rate": 1.0936501210857652e-05,
"loss": 1.9614,
"step": 2039
},
{
"epoch": 0.935672514619883,
"grad_norm": 0.2182752639055252,
"learning_rate": 1.0782481637284013e-05,
"loss": 1.9413,
"step": 2040
},
{
"epoch": 0.9361311776172457,
"grad_norm": 0.21929451823234558,
"learning_rate": 1.0629542481433663e-05,
"loss": 1.9297,
"step": 2041
},
{
"epoch": 0.9365898406146084,
"grad_norm": 0.22284549474716187,
"learning_rate": 1.0477684081066751e-05,
"loss": 1.9133,
"step": 2042
},
{
"epoch": 0.9370485036119711,
"grad_norm": 0.22120410203933716,
"learning_rate": 1.0326906771556566e-05,
"loss": 1.9116,
"step": 2043
},
{
"epoch": 0.9375071666093338,
"grad_norm": 0.21954967081546783,
"learning_rate": 1.017721088588891e-05,
"loss": 1.9769,
"step": 2044
},
{
"epoch": 0.9379658296066965,
"grad_norm": 0.21843045949935913,
"learning_rate": 1.0028596754661334e-05,
"loss": 1.9741,
"step": 2045
},
{
"epoch": 0.9384244926040591,
"grad_norm": 0.2236836701631546,
"learning_rate": 9.881064706082298e-06,
"loss": 1.947,
"step": 2046
},
{
"epoch": 0.9388831556014219,
"grad_norm": 0.22054068744182587,
"learning_rate": 9.734615065970454e-06,
"loss": 1.9567,
"step": 2047
},
{
"epoch": 0.9393418185987845,
"grad_norm": 0.21976153552532196,
"learning_rate": 9.58924815775425e-06,
"loss": 1.8958,
"step": 2048
},
{
"epoch": 0.9398004815961473,
"grad_norm": 0.2242991030216217,
"learning_rate": 9.444964302470715e-06,
"loss": 2.0594,
"step": 2049
},
{
"epoch": 0.9402591445935099,
"grad_norm": 0.2180870771408081,
"learning_rate": 9.301763818765018e-06,
"loss": 1.8928,
"step": 2050
},
{
"epoch": 0.9407178075908726,
"grad_norm": 0.21730895340442657,
"learning_rate": 9.15964702288996e-06,
"loss": 1.8825,
"step": 2051
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.2222636193037033,
"learning_rate": 9.018614228704925e-06,
"loss": 2.0385,
"step": 2052
},
{
"epoch": 0.941635133585598,
"grad_norm": 0.22359733283519745,
"learning_rate": 8.878665747675152e-06,
"loss": 1.9102,
"step": 2053
},
{
"epoch": 0.9420937965829607,
"grad_norm": 0.21196125447750092,
"learning_rate": 8.739801888871469e-06,
"loss": 1.9919,
"step": 2054
},
{
"epoch": 0.9425524595803234,
"grad_norm": 0.21883940696716309,
"learning_rate": 8.602022958969336e-06,
"loss": 1.9638,
"step": 2055
},
{
"epoch": 0.943011122577686,
"grad_norm": 0.2190505415201187,
"learning_rate": 8.465329262248078e-06,
"loss": 1.9317,
"step": 2056
},
{
"epoch": 0.9434697855750487,
"grad_norm": 0.23451970517635345,
"learning_rate": 8.32972110059027e-06,
"loss": 1.9867,
"step": 2057
},
{
"epoch": 0.9439284485724114,
"grad_norm": 0.20750227570533752,
"learning_rate": 8.195198773481406e-06,
"loss": 1.9247,
"step": 2058
},
{
"epoch": 0.9443871115697741,
"grad_norm": 0.20929892361164093,
"learning_rate": 8.061762578008613e-06,
"loss": 2.0159,
"step": 2059
},
{
"epoch": 0.9448457745671368,
"grad_norm": 0.2086939662694931,
"learning_rate": 7.929412808860559e-06,
"loss": 1.9219,
"step": 2060
},
{
"epoch": 0.9453044375644994,
"grad_norm": 0.22294549643993378,
"learning_rate": 7.79814975832649e-06,
"loss": 2.0034,
"step": 2061
},
{
"epoch": 0.9457631005618622,
"grad_norm": 0.22373966872692108,
"learning_rate": 7.667973716295851e-06,
"loss": 1.8978,
"step": 2062
},
{
"epoch": 0.9462217635592248,
"grad_norm": 0.22192026674747467,
"learning_rate": 7.5388849702571205e-06,
"loss": 1.9996,
"step": 2063
},
{
"epoch": 0.9466804265565876,
"grad_norm": 0.21932195127010345,
"learning_rate": 7.4108838052979185e-06,
"loss": 1.8824,
"step": 2064
},
{
"epoch": 0.9471390895539502,
"grad_norm": 0.20995944738388062,
"learning_rate": 7.283970504103732e-06,
"loss": 1.9474,
"step": 2065
},
{
"epoch": 0.947597752551313,
"grad_norm": 0.2205418348312378,
"learning_rate": 7.1581453469575785e-06,
"loss": 1.9614,
"step": 2066
},
{
"epoch": 0.9480564155486756,
"grad_norm": 0.21611550450325012,
"learning_rate": 7.033408611739456e-06,
"loss": 1.9028,
"step": 2067
},
{
"epoch": 0.9485150785460383,
"grad_norm": 0.22107021510601044,
"learning_rate": 6.909760573925561e-06,
"loss": 1.9518,
"step": 2068
},
{
"epoch": 0.948973741543401,
"grad_norm": 0.21450047194957733,
"learning_rate": 6.787201506587626e-06,
"loss": 1.8755,
"step": 2069
},
{
"epoch": 0.9494324045407637,
"grad_norm": 0.21897783875465393,
"learning_rate": 6.66573168039264e-06,
"loss": 1.901,
"step": 2070
},
{
"epoch": 0.9498910675381264,
"grad_norm": 0.22471900284290314,
"learning_rate": 6.545351363601959e-06,
"loss": 2.094,
"step": 2071
},
{
"epoch": 0.9503497305354891,
"grad_norm": 0.2023703008890152,
"learning_rate": 6.426060822070812e-06,
"loss": 1.9648,
"step": 2072
},
{
"epoch": 0.9508083935328517,
"grad_norm": 0.217071071267128,
"learning_rate": 6.3078603192475716e-06,
"loss": 1.9706,
"step": 2073
},
{
"epoch": 0.9512670565302144,
"grad_norm": 0.21377383172512054,
"learning_rate": 6.1907501161735934e-06,
"loss": 1.9644,
"step": 2074
},
{
"epoch": 0.9517257195275771,
"grad_norm": 0.20817036926746368,
"learning_rate": 6.074730471482049e-06,
"loss": 1.9386,
"step": 2075
},
{
"epoch": 0.9521843825249398,
"grad_norm": 0.21762020885944366,
"learning_rate": 5.959801641397755e-06,
"loss": 1.9583,
"step": 2076
},
{
"epoch": 0.9526430455223025,
"grad_norm": 0.2194538712501526,
"learning_rate": 5.845963879736627e-06,
"loss": 1.9969,
"step": 2077
},
{
"epoch": 0.9531017085196651,
"grad_norm": 0.2142002135515213,
"learning_rate": 5.733217437904892e-06,
"loss": 2.0112,
"step": 2078
},
{
"epoch": 0.9535603715170279,
"grad_norm": 0.22616985440254211,
"learning_rate": 5.621562564898597e-06,
"loss": 1.9143,
"step": 2079
},
{
"epoch": 0.9540190345143905,
"grad_norm": 0.21155084669589996,
"learning_rate": 5.51099950730316e-06,
"loss": 1.9227,
"step": 2080
},
{
"epoch": 0.9544776975117533,
"grad_norm": 0.2239096462726593,
"learning_rate": 5.401528509292763e-06,
"loss": 2.0205,
"step": 2081
},
{
"epoch": 0.9549363605091159,
"grad_norm": 0.21380391716957092,
"learning_rate": 5.2931498126298495e-06,
"loss": 1.9527,
"step": 2082
},
{
"epoch": 0.9553950235064786,
"grad_norm": 0.22131973505020142,
"learning_rate": 5.1858636566645135e-06,
"loss": 2.0284,
"step": 2083
},
{
"epoch": 0.9558536865038413,
"grad_norm": 0.20455728471279144,
"learning_rate": 5.0796702783340035e-06,
"loss": 1.9919,
"step": 2084
},
{
"epoch": 0.956312349501204,
"grad_norm": 0.2129702866077423,
"learning_rate": 4.97456991216233e-06,
"loss": 1.9491,
"step": 2085
},
{
"epoch": 0.9567710124985667,
"grad_norm": 0.2284303605556488,
"learning_rate": 4.870562790259325e-06,
"loss": 1.9946,
"step": 2086
},
{
"epoch": 0.9572296754959294,
"grad_norm": 0.21666593849658966,
"learning_rate": 4.7676491423208625e-06,
"loss": 1.8502,
"step": 2087
},
{
"epoch": 0.957688338493292,
"grad_norm": 0.2121717780828476,
"learning_rate": 4.66582919562758e-06,
"loss": 1.9672,
"step": 2088
},
{
"epoch": 0.9581470014906548,
"grad_norm": 0.20628033578395844,
"learning_rate": 4.5651031750448825e-06,
"loss": 1.8744,
"step": 2089
},
{
"epoch": 0.9586056644880174,
"grad_norm": 0.2210129350423813,
"learning_rate": 4.465471303022217e-06,
"loss": 1.9728,
"step": 2090
},
{
"epoch": 0.9590643274853801,
"grad_norm": 0.22552601993083954,
"learning_rate": 4.366933799592743e-06,
"loss": 2.0238,
"step": 2091
},
{
"epoch": 0.9595229904827428,
"grad_norm": 0.21505334973335266,
"learning_rate": 4.269490882372551e-06,
"loss": 1.9944,
"step": 2092
},
{
"epoch": 0.9599816534801054,
"grad_norm": 0.21251259744167328,
"learning_rate": 4.1731427665606115e-06,
"loss": 1.9343,
"step": 2093
},
{
"epoch": 0.9604403164774682,
"grad_norm": 0.22040241956710815,
"learning_rate": 4.077889664937884e-06,
"loss": 1.958,
"step": 2094
},
{
"epoch": 0.9608989794748308,
"grad_norm": 0.21440507471561432,
"learning_rate": 3.983731787867207e-06,
"loss": 1.9481,
"step": 2095
},
{
"epoch": 0.9613576424721936,
"grad_norm": 0.22844387590885162,
"learning_rate": 3.890669343292464e-06,
"loss": 2.0238,
"step": 2096
},
{
"epoch": 0.9618163054695562,
"grad_norm": 0.22273798286914825,
"learning_rate": 3.7987025367384743e-06,
"loss": 1.9049,
"step": 2097
},
{
"epoch": 0.962274968466919,
"grad_norm": 0.2164827287197113,
"learning_rate": 3.707831571310327e-06,
"loss": 2.0388,
"step": 2098
},
{
"epoch": 0.9627336314642816,
"grad_norm": 0.21262814104557037,
"learning_rate": 3.6180566476929912e-06,
"loss": 1.8726,
"step": 2099
},
{
"epoch": 0.9631922944616443,
"grad_norm": 0.21005943417549133,
"learning_rate": 3.529377964150815e-06,
"loss": 1.9358,
"step": 2100
},
{
"epoch": 0.963650957459007,
"grad_norm": 0.2163417488336563,
"learning_rate": 3.441795716527307e-06,
"loss": 2.0304,
"step": 2101
},
{
"epoch": 0.9641096204563697,
"grad_norm": 0.22290202975273132,
"learning_rate": 3.355310098244302e-06,
"loss": 1.9277,
"step": 2102
},
{
"epoch": 0.9645682834537324,
"grad_norm": 0.2204234004020691,
"learning_rate": 3.269921300301959e-06,
"loss": 1.9891,
"step": 2103
},
{
"epoch": 0.9650269464510951,
"grad_norm": 0.21776074171066284,
"learning_rate": 3.1856295112780988e-06,
"loss": 1.9383,
"step": 2104
},
{
"epoch": 0.9654856094484577,
"grad_norm": 0.21141210198402405,
"learning_rate": 3.102434917327812e-06,
"loss": 1.9128,
"step": 2105
},
{
"epoch": 0.9659442724458205,
"grad_norm": 0.2165384739637375,
"learning_rate": 3.0203377021831292e-06,
"loss": 1.9352,
"step": 2106
},
{
"epoch": 0.9664029354431831,
"grad_norm": 0.221245676279068,
"learning_rate": 2.939338047152573e-06,
"loss": 1.8948,
"step": 2107
},
{
"epoch": 0.9668615984405458,
"grad_norm": 0.22557511925697327,
"learning_rate": 2.8594361311206073e-06,
"loss": 1.9769,
"step": 2108
},
{
"epoch": 0.9673202614379085,
"grad_norm": 0.2159567028284073,
"learning_rate": 2.7806321305475225e-06,
"loss": 1.9128,
"step": 2109
},
{
"epoch": 0.9677789244352711,
"grad_norm": 0.2295222133398056,
"learning_rate": 2.7029262194688818e-06,
"loss": 1.9346,
"step": 2110
},
{
"epoch": 0.9682375874326339,
"grad_norm": 0.2073698192834854,
"learning_rate": 2.626318569495134e-06,
"loss": 1.907,
"step": 2111
},
{
"epoch": 0.9686962504299965,
"grad_norm": 0.2221342921257019,
"learning_rate": 2.550809349811334e-06,
"loss": 1.9156,
"step": 2112
},
{
"epoch": 0.9691549134273593,
"grad_norm": 0.21161620318889618,
"learning_rate": 2.476398727176532e-06,
"loss": 1.867,
"step": 2113
},
{
"epoch": 0.9696135764247219,
"grad_norm": 0.2093185931444168,
"learning_rate": 2.4030868659237204e-06,
"loss": 1.8279,
"step": 2114
},
{
"epoch": 0.9700722394220846,
"grad_norm": 0.21819551289081573,
"learning_rate": 2.3308739279593317e-06,
"loss": 1.9217,
"step": 2115
},
{
"epoch": 0.9705309024194473,
"grad_norm": 0.21790070831775665,
"learning_rate": 2.2597600727626845e-06,
"loss": 1.9149,
"step": 2116
},
{
"epoch": 0.97098956541681,
"grad_norm": 0.22319115698337555,
"learning_rate": 2.1897454573860387e-06,
"loss": 1.8727,
"step": 2117
},
{
"epoch": 0.9714482284141727,
"grad_norm": 0.2239585518836975,
"learning_rate": 2.1208302364538746e-06,
"loss": 1.9274,
"step": 2118
},
{
"epoch": 0.9719068914115354,
"grad_norm": 0.2205500453710556,
"learning_rate": 2.0530145621627804e-06,
"loss": 1.8964,
"step": 2119
},
{
"epoch": 0.972365554408898,
"grad_norm": 0.21439099311828613,
"learning_rate": 1.9862985842810653e-06,
"loss": 1.9715,
"step": 2120
},
{
"epoch": 0.9728242174062608,
"grad_norm": 0.23177304863929749,
"learning_rate": 1.920682450148259e-06,
"loss": 1.9536,
"step": 2121
},
{
"epoch": 0.9732828804036234,
"grad_norm": 0.21910591423511505,
"learning_rate": 1.856166304675111e-06,
"loss": 1.9063,
"step": 2122
},
{
"epoch": 0.9737415434009862,
"grad_norm": 0.2140384018421173,
"learning_rate": 1.792750290342926e-06,
"loss": 1.9972,
"step": 2123
},
{
"epoch": 0.9742002063983488,
"grad_norm": 0.21132169663906097,
"learning_rate": 1.7304345472035632e-06,
"loss": 1.9432,
"step": 2124
},
{
"epoch": 0.9746588693957114,
"grad_norm": 0.22471146285533905,
"learning_rate": 1.6692192128788253e-06,
"loss": 1.935,
"step": 2125
},
{
"epoch": 0.9751175323930742,
"grad_norm": 0.21662886440753937,
"learning_rate": 1.6091044225604035e-06,
"loss": 1.9631,
"step": 2126
},
{
"epoch": 0.9755761953904368,
"grad_norm": 0.22543148696422577,
"learning_rate": 1.5500903090094888e-06,
"loss": 1.9457,
"step": 2127
},
{
"epoch": 0.9760348583877996,
"grad_norm": 0.2169279307126999,
"learning_rate": 1.492177002556383e-06,
"loss": 1.9594,
"step": 2128
},
{
"epoch": 0.9764935213851622,
"grad_norm": 0.2178993672132492,
"learning_rate": 1.4353646311004443e-06,
"loss": 1.9606,
"step": 2129
},
{
"epoch": 0.976952184382525,
"grad_norm": 0.20791199803352356,
"learning_rate": 1.3796533201094752e-06,
"loss": 1.8573,
"step": 2130
},
{
"epoch": 0.9774108473798876,
"grad_norm": 0.2268703728914261,
"learning_rate": 1.3250431926197793e-06,
"loss": 1.9247,
"step": 2131
},
{
"epoch": 0.9778695103772503,
"grad_norm": 0.22626438736915588,
"learning_rate": 1.2715343692356607e-06,
"loss": 2.01,
"step": 2132
},
{
"epoch": 0.978328173374613,
"grad_norm": 0.2251403033733368,
"learning_rate": 1.2191269681292582e-06,
"loss": 1.945,
"step": 2133
},
{
"epoch": 0.9787868363719757,
"grad_norm": 0.21841411292552948,
"learning_rate": 1.1678211050402676e-06,
"loss": 1.9939,
"step": 2134
},
{
"epoch": 0.9792454993693384,
"grad_norm": 0.226179301738739,
"learning_rate": 1.117616893275719e-06,
"loss": 1.978,
"step": 2135
},
{
"epoch": 0.9797041623667011,
"grad_norm": 0.21425890922546387,
"learning_rate": 1.068514443709534e-06,
"loss": 1.9982,
"step": 2136
},
{
"epoch": 0.9801628253640637,
"grad_norm": 0.21500264108181,
"learning_rate": 1.0205138647826905e-06,
"loss": 1.9962,
"step": 2137
},
{
"epoch": 0.9806214883614265,
"grad_norm": 0.21788109838962555,
"learning_rate": 9.73615262502503e-07,
"loss": 1.8926,
"step": 2138
},
{
"epoch": 0.9810801513587891,
"grad_norm": 0.22228911519050598,
"learning_rate": 9.278187404426763e-07,
"loss": 1.9624,
"step": 2139
},
{
"epoch": 0.9815388143561519,
"grad_norm": 0.224578395485878,
"learning_rate": 8.831243997431404e-07,
"loss": 1.8979,
"step": 2140
},
{
"epoch": 0.9819974773535145,
"grad_norm": 0.217121422290802,
"learning_rate": 8.395323391094944e-07,
"loss": 2.0203,
"step": 2141
},
{
"epoch": 0.9824561403508771,
"grad_norm": 0.21875330805778503,
"learning_rate": 7.970426548131183e-07,
"loss": 1.9446,
"step": 2142
},
{
"epoch": 0.9829148033482399,
"grad_norm": 0.21179907023906708,
"learning_rate": 7.556554406908389e-07,
"loss": 1.9295,
"step": 2143
},
{
"epoch": 0.9833734663456025,
"grad_norm": 0.21734905242919922,
"learning_rate": 7.153707881446536e-07,
"loss": 2.0076,
"step": 2144
},
{
"epoch": 0.9838321293429653,
"grad_norm": 0.21071693301200867,
"learning_rate": 6.761887861417293e-07,
"loss": 1.9438,
"step": 2145
},
{
"epoch": 0.9842907923403279,
"grad_norm": 0.21275602281093597,
"learning_rate": 6.381095212139032e-07,
"loss": 1.9978,
"step": 2146
},
{
"epoch": 0.9847494553376906,
"grad_norm": 0.21144671738147736,
"learning_rate": 6.011330774577384e-07,
"loss": 1.9733,
"step": 2147
},
{
"epoch": 0.9852081183350533,
"grad_norm": 0.21650995314121246,
"learning_rate": 5.652595365343016e-07,
"loss": 2.0474,
"step": 2148
},
{
"epoch": 0.985666781332416,
"grad_norm": 0.2181413322687149,
"learning_rate": 5.304889776688859e-07,
"loss": 1.923,
"step": 2149
},
{
"epoch": 0.9861254443297787,
"grad_norm": 0.21923984587192535,
"learning_rate": 4.968214776508994e-07,
"loss": 1.9043,
"step": 2150
},
{
"epoch": 0.9865841073271414,
"grad_norm": 0.21624214947223663,
"learning_rate": 4.6425711083375454e-07,
"loss": 1.9147,
"step": 2151
},
{
"epoch": 0.987042770324504,
"grad_norm": 0.22860468924045563,
"learning_rate": 4.3279594913447906e-07,
"loss": 1.9623,
"step": 2152
},
{
"epoch": 0.9875014333218668,
"grad_norm": 0.2104010432958603,
"learning_rate": 4.02438062033883e-07,
"loss": 1.9367,
"step": 2153
},
{
"epoch": 0.9879600963192294,
"grad_norm": 0.21578551828861237,
"learning_rate": 3.7318351657616987e-07,
"loss": 1.921,
"step": 2154
},
{
"epoch": 0.9884187593165922,
"grad_norm": 0.20753897726535797,
"learning_rate": 3.4503237736882573e-07,
"loss": 1.9321,
"step": 2155
},
{
"epoch": 0.9888774223139548,
"grad_norm": 0.21836236119270325,
"learning_rate": 3.179847065825081e-07,
"loss": 2.0036,
"step": 2156
},
{
"epoch": 0.9893360853113176,
"grad_norm": 0.21465463936328888,
"learning_rate": 2.9204056395104594e-07,
"loss": 1.8222,
"step": 2157
},
{
"epoch": 0.9897947483086802,
"grad_norm": 0.2177664041519165,
"learning_rate": 2.672000067709956e-07,
"loss": 1.8761,
"step": 2158
},
{
"epoch": 0.9902534113060428,
"grad_norm": 0.2041962891817093,
"learning_rate": 2.4346308990175204e-07,
"loss": 1.9776,
"step": 2159
},
{
"epoch": 0.9907120743034056,
"grad_norm": 0.2094731330871582,
"learning_rate": 2.208298657653818e-07,
"loss": 1.9808,
"step": 2160
},
{
"epoch": 0.9911707373007682,
"grad_norm": 0.22576749324798584,
"learning_rate": 1.9930038434645692e-07,
"loss": 2.0357,
"step": 2161
},
{
"epoch": 0.991629400298131,
"grad_norm": 0.22563716769218445,
"learning_rate": 1.7887469319205484e-07,
"loss": 1.9667,
"step": 2162
},
{
"epoch": 0.9920880632954936,
"grad_norm": 0.2230522483587265,
"learning_rate": 1.5955283741142523e-07,
"loss": 1.9285,
"step": 2163
},
{
"epoch": 0.9925467262928563,
"grad_norm": 0.21589314937591553,
"learning_rate": 1.4133485967615655e-07,
"loss": 1.9981,
"step": 2164
},
{
"epoch": 0.993005389290219,
"grad_norm": 0.23776021599769592,
"learning_rate": 1.2422080021995407e-07,
"loss": 2.028,
"step": 2165
},
{
"epoch": 0.9934640522875817,
"grad_norm": 0.2170460820198059,
"learning_rate": 1.082106968385288e-07,
"loss": 1.9329,
"step": 2166
},
{
"epoch": 0.9939227152849444,
"grad_norm": 0.21792776882648468,
"learning_rate": 9.330458488959748e-08,
"loss": 2.0203,
"step": 2167
},
{
"epoch": 0.9943813782823071,
"grad_norm": 0.21459391713142395,
"learning_rate": 7.950249729271608e-08,
"loss": 1.9956,
"step": 2168
},
{
"epoch": 0.9948400412796697,
"grad_norm": 0.21822136640548706,
"learning_rate": 6.680446452922429e-08,
"loss": 1.8864,
"step": 2169
},
{
"epoch": 0.9952987042770325,
"grad_norm": 0.22067728638648987,
"learning_rate": 5.521051464230098e-08,
"loss": 1.8525,
"step": 2170
},
{
"epoch": 0.9957573672743951,
"grad_norm": 0.22874440252780914,
"learning_rate": 4.4720673236631206e-08,
"loss": 1.9855,
"step": 2171
},
{
"epoch": 0.9962160302717579,
"grad_norm": 0.20219753682613373,
"learning_rate": 3.53349634786837e-08,
"loss": 1.8585,
"step": 2172
},
{
"epoch": 0.9966746932691205,
"grad_norm": 0.21162958443164825,
"learning_rate": 2.7053406096433365e-08,
"loss": 1.8842,
"step": 2173
},
{
"epoch": 0.9971333562664833,
"grad_norm": 0.2165650874376297,
"learning_rate": 1.987601937930572e-08,
"loss": 1.9235,
"step": 2174
},
{
"epoch": 0.9975920192638459,
"grad_norm": 0.22486677765846252,
"learning_rate": 1.3802819178398984e-08,
"loss": 2.0349,
"step": 2175
},
{
"epoch": 0.9980506822612085,
"grad_norm": 0.21611721813678741,
"learning_rate": 8.833818906039959e-09,
"loss": 1.9273,
"step": 2176
},
{
"epoch": 0.9985093452585713,
"grad_norm": 0.21843695640563965,
"learning_rate": 4.969029536061598e-09,
"loss": 2.0224,
"step": 2177
},
{
"epoch": 0.9989680082559339,
"grad_norm": 0.21215592324733734,
"learning_rate": 2.2084596038030037e-09,
"loss": 1.8675,
"step": 2178
},
{
"epoch": 0.9994266712532967,
"grad_norm": 0.2176358699798584,
"learning_rate": 5.521152057763601e-10,
"loss": 1.9198,
"step": 2179
},
{
"epoch": 0.9998853342506593,
"grad_norm": 0.22994180023670197,
"learning_rate": 0.0,
"loss": 2.0607,
"step": 2180
},
{
"epoch": 0.9998853342506593,
"step": 2180,
"total_flos": 6.163095853926973e+17,
"train_loss": 2.085200167850617,
"train_runtime": 33029.3352,
"train_samples_per_second": 16.898,
"train_steps_per_second": 0.066
}
],
"logging_steps": 1.0,
"max_steps": 2180,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 24000,
"total_flos": 6.163095853926973e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}