sedrickkeh's picture
End of training
71dd12b verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9904153354632586,
"eval_steps": 500,
"global_step": 468,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006389776357827476,
"grad_norm": 5.9393923041460335,
"learning_rate": 8.510638297872341e-07,
"loss": 0.8883,
"step": 1
},
{
"epoch": 0.012779552715654952,
"grad_norm": 5.645614037221844,
"learning_rate": 1.7021276595744682e-06,
"loss": 0.8365,
"step": 2
},
{
"epoch": 0.019169329073482427,
"grad_norm": 6.260474581584071,
"learning_rate": 2.553191489361702e-06,
"loss": 0.8853,
"step": 3
},
{
"epoch": 0.025559105431309903,
"grad_norm": 5.551335163218543,
"learning_rate": 3.4042553191489363e-06,
"loss": 0.8686,
"step": 4
},
{
"epoch": 0.03194888178913738,
"grad_norm": 4.506734407013277,
"learning_rate": 4.255319148936171e-06,
"loss": 0.792,
"step": 5
},
{
"epoch": 0.038338658146964855,
"grad_norm": 2.7013711545792316,
"learning_rate": 5.106382978723404e-06,
"loss": 0.7978,
"step": 6
},
{
"epoch": 0.04472843450479233,
"grad_norm": 2.231426309167533,
"learning_rate": 5.957446808510638e-06,
"loss": 0.7656,
"step": 7
},
{
"epoch": 0.051118210862619806,
"grad_norm": 3.786605436059839,
"learning_rate": 6.808510638297873e-06,
"loss": 0.7476,
"step": 8
},
{
"epoch": 0.05750798722044728,
"grad_norm": 3.9559474965927994,
"learning_rate": 7.659574468085107e-06,
"loss": 0.751,
"step": 9
},
{
"epoch": 0.06389776357827476,
"grad_norm": 4.3050608871997476,
"learning_rate": 8.510638297872341e-06,
"loss": 0.7376,
"step": 10
},
{
"epoch": 0.07028753993610223,
"grad_norm": 3.844203590596342,
"learning_rate": 9.361702127659576e-06,
"loss": 0.6577,
"step": 11
},
{
"epoch": 0.07667731629392971,
"grad_norm": 3.7090602450962535,
"learning_rate": 1.0212765957446808e-05,
"loss": 0.7587,
"step": 12
},
{
"epoch": 0.08306709265175719,
"grad_norm": 2.6605932751385915,
"learning_rate": 1.1063829787234044e-05,
"loss": 0.7102,
"step": 13
},
{
"epoch": 0.08945686900958466,
"grad_norm": 1.841280162949687,
"learning_rate": 1.1914893617021277e-05,
"loss": 0.6686,
"step": 14
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.4654109468445635,
"learning_rate": 1.2765957446808513e-05,
"loss": 0.6976,
"step": 15
},
{
"epoch": 0.10223642172523961,
"grad_norm": 2.2718952209600842,
"learning_rate": 1.3617021276595745e-05,
"loss": 0.66,
"step": 16
},
{
"epoch": 0.10862619808306709,
"grad_norm": 1.336892858370965,
"learning_rate": 1.4468085106382981e-05,
"loss": 0.6222,
"step": 17
},
{
"epoch": 0.11501597444089456,
"grad_norm": 1.3994667991152954,
"learning_rate": 1.5319148936170214e-05,
"loss": 0.616,
"step": 18
},
{
"epoch": 0.12140575079872204,
"grad_norm": 1.3560171283396125,
"learning_rate": 1.6170212765957446e-05,
"loss": 0.6394,
"step": 19
},
{
"epoch": 0.12779552715654952,
"grad_norm": 1.1983042661201426,
"learning_rate": 1.7021276595744682e-05,
"loss": 0.5704,
"step": 20
},
{
"epoch": 0.134185303514377,
"grad_norm": 0.9862694631668227,
"learning_rate": 1.7872340425531915e-05,
"loss": 0.592,
"step": 21
},
{
"epoch": 0.14057507987220447,
"grad_norm": 1.1717058829411315,
"learning_rate": 1.872340425531915e-05,
"loss": 0.6111,
"step": 22
},
{
"epoch": 0.14696485623003194,
"grad_norm": 1.052734463937942,
"learning_rate": 1.9574468085106384e-05,
"loss": 0.6105,
"step": 23
},
{
"epoch": 0.15335463258785942,
"grad_norm": 0.9393704655813953,
"learning_rate": 2.0425531914893616e-05,
"loss": 0.6196,
"step": 24
},
{
"epoch": 0.1597444089456869,
"grad_norm": 1.064356434143729,
"learning_rate": 2.1276595744680852e-05,
"loss": 0.6022,
"step": 25
},
{
"epoch": 0.16613418530351437,
"grad_norm": 0.8527356341897169,
"learning_rate": 2.2127659574468088e-05,
"loss": 0.6149,
"step": 26
},
{
"epoch": 0.17252396166134185,
"grad_norm": 0.8856534958761616,
"learning_rate": 2.2978723404255324e-05,
"loss": 0.57,
"step": 27
},
{
"epoch": 0.17891373801916932,
"grad_norm": 0.7965378935060776,
"learning_rate": 2.3829787234042553e-05,
"loss": 0.5663,
"step": 28
},
{
"epoch": 0.1853035143769968,
"grad_norm": 0.930055407459929,
"learning_rate": 2.468085106382979e-05,
"loss": 0.5675,
"step": 29
},
{
"epoch": 0.19169329073482427,
"grad_norm": 0.9473480680946368,
"learning_rate": 2.5531914893617025e-05,
"loss": 0.5647,
"step": 30
},
{
"epoch": 0.19808306709265175,
"grad_norm": 0.9035770727022399,
"learning_rate": 2.6382978723404255e-05,
"loss": 0.5571,
"step": 31
},
{
"epoch": 0.20447284345047922,
"grad_norm": 0.9629442234127248,
"learning_rate": 2.723404255319149e-05,
"loss": 0.5675,
"step": 32
},
{
"epoch": 0.2108626198083067,
"grad_norm": 0.8898757055373155,
"learning_rate": 2.8085106382978727e-05,
"loss": 0.5684,
"step": 33
},
{
"epoch": 0.21725239616613418,
"grad_norm": 0.8818620757382937,
"learning_rate": 2.8936170212765963e-05,
"loss": 0.5394,
"step": 34
},
{
"epoch": 0.22364217252396165,
"grad_norm": 1.2556928806711947,
"learning_rate": 2.9787234042553192e-05,
"loss": 0.5912,
"step": 35
},
{
"epoch": 0.23003194888178913,
"grad_norm": 0.8995453098056544,
"learning_rate": 3.063829787234043e-05,
"loss": 0.5284,
"step": 36
},
{
"epoch": 0.2364217252396166,
"grad_norm": 1.1694120920465332,
"learning_rate": 3.1489361702127664e-05,
"loss": 0.5448,
"step": 37
},
{
"epoch": 0.24281150159744408,
"grad_norm": 0.8640750703168888,
"learning_rate": 3.234042553191489e-05,
"loss": 0.5427,
"step": 38
},
{
"epoch": 0.24920127795527156,
"grad_norm": 0.9511230435963061,
"learning_rate": 3.319148936170213e-05,
"loss": 0.5242,
"step": 39
},
{
"epoch": 0.25559105431309903,
"grad_norm": 0.9815409113958286,
"learning_rate": 3.4042553191489365e-05,
"loss": 0.5327,
"step": 40
},
{
"epoch": 0.26198083067092653,
"grad_norm": 0.9294918620323822,
"learning_rate": 3.48936170212766e-05,
"loss": 0.55,
"step": 41
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.2780033880918986,
"learning_rate": 3.574468085106383e-05,
"loss": 0.5057,
"step": 42
},
{
"epoch": 0.2747603833865815,
"grad_norm": 0.970896465496613,
"learning_rate": 3.6595744680851066e-05,
"loss": 0.563,
"step": 43
},
{
"epoch": 0.28115015974440893,
"grad_norm": 1.1743520181962854,
"learning_rate": 3.74468085106383e-05,
"loss": 0.5716,
"step": 44
},
{
"epoch": 0.28753993610223644,
"grad_norm": 0.9440517537009965,
"learning_rate": 3.829787234042554e-05,
"loss": 0.5202,
"step": 45
},
{
"epoch": 0.2939297124600639,
"grad_norm": 1.244837667755633,
"learning_rate": 3.914893617021277e-05,
"loss": 0.5333,
"step": 46
},
{
"epoch": 0.3003194888178914,
"grad_norm": 1.0156465350553556,
"learning_rate": 4e-05,
"loss": 0.5569,
"step": 47
},
{
"epoch": 0.30670926517571884,
"grad_norm": 1.0875814805232027,
"learning_rate": 3.999944315600779e-05,
"loss": 0.5412,
"step": 48
},
{
"epoch": 0.31309904153354634,
"grad_norm": 0.9841398658147307,
"learning_rate": 3.9997772655038674e-05,
"loss": 0.5493,
"step": 49
},
{
"epoch": 0.3194888178913738,
"grad_norm": 1.041702971561358,
"learning_rate": 3.99949885901135e-05,
"loss": 0.5464,
"step": 50
},
{
"epoch": 0.3258785942492013,
"grad_norm": 1.1971478020256678,
"learning_rate": 3.999109111626125e-05,
"loss": 0.5488,
"step": 51
},
{
"epoch": 0.33226837060702874,
"grad_norm": 0.7937301188136895,
"learning_rate": 3.998608045051041e-05,
"loss": 0.5406,
"step": 52
},
{
"epoch": 0.33865814696485624,
"grad_norm": 1.0893696428115565,
"learning_rate": 3.99799568718769e-05,
"loss": 0.5057,
"step": 53
},
{
"epoch": 0.3450479233226837,
"grad_norm": 1.0041397722351062,
"learning_rate": 3.9972720721348503e-05,
"loss": 0.5568,
"step": 54
},
{
"epoch": 0.3514376996805112,
"grad_norm": 1.1730368348535483,
"learning_rate": 3.996437240186593e-05,
"loss": 0.5376,
"step": 55
},
{
"epoch": 0.35782747603833864,
"grad_norm": 1.0260574947019203,
"learning_rate": 3.995491237830033e-05,
"loss": 0.5845,
"step": 56
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.9477092619910678,
"learning_rate": 3.994434117742743e-05,
"loss": 0.5861,
"step": 57
},
{
"epoch": 0.3706070287539936,
"grad_norm": 1.020580513302534,
"learning_rate": 3.9932659387898195e-05,
"loss": 0.4967,
"step": 58
},
{
"epoch": 0.3769968051118211,
"grad_norm": 1.117231150698062,
"learning_rate": 3.991986766020607e-05,
"loss": 0.5163,
"step": 59
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.9136891394466323,
"learning_rate": 3.9905966706650715e-05,
"loss": 0.5236,
"step": 60
},
{
"epoch": 0.38977635782747605,
"grad_norm": 0.9610320358708975,
"learning_rate": 3.9890957301298385e-05,
"loss": 0.5442,
"step": 61
},
{
"epoch": 0.3961661341853035,
"grad_norm": 0.9525076230315872,
"learning_rate": 3.9874840279938794e-05,
"loss": 0.5481,
"step": 62
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.9886639116884637,
"learning_rate": 3.9857616540038595e-05,
"loss": 0.587,
"step": 63
},
{
"epoch": 0.40894568690095845,
"grad_norm": 0.8789895879873663,
"learning_rate": 3.9839287040691396e-05,
"loss": 0.5405,
"step": 64
},
{
"epoch": 0.41533546325878595,
"grad_norm": 1.2919381394585123,
"learning_rate": 3.981985280256436e-05,
"loss": 0.5661,
"step": 65
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.8458840142981038,
"learning_rate": 3.979931490784136e-05,
"loss": 0.5376,
"step": 66
},
{
"epoch": 0.4281150159744409,
"grad_norm": 0.7123621144432971,
"learning_rate": 3.977767450016272e-05,
"loss": 0.5501,
"step": 67
},
{
"epoch": 0.43450479233226835,
"grad_norm": 0.9335736753943794,
"learning_rate": 3.9754932784561545e-05,
"loss": 0.5495,
"step": 68
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.7759436980865956,
"learning_rate": 3.973109102739661e-05,
"loss": 0.5468,
"step": 69
},
{
"epoch": 0.4472843450479233,
"grad_norm": 0.9317387591226213,
"learning_rate": 3.9706150556281827e-05,
"loss": 0.5578,
"step": 70
},
{
"epoch": 0.4536741214057508,
"grad_norm": 1.125415618543546,
"learning_rate": 3.968011276001236e-05,
"loss": 0.5769,
"step": 71
},
{
"epoch": 0.46006389776357826,
"grad_norm": 1.0297816336058314,
"learning_rate": 3.9652979088487246e-05,
"loss": 0.5545,
"step": 72
},
{
"epoch": 0.46645367412140576,
"grad_norm": 0.977116035792409,
"learning_rate": 3.962475105262868e-05,
"loss": 0.5443,
"step": 73
},
{
"epoch": 0.4728434504792332,
"grad_norm": 0.9591838077463976,
"learning_rate": 3.9595430224297874e-05,
"loss": 0.5597,
"step": 74
},
{
"epoch": 0.4792332268370607,
"grad_norm": 1.062273213479338,
"learning_rate": 3.9565018236207545e-05,
"loss": 0.5364,
"step": 75
},
{
"epoch": 0.48562300319488816,
"grad_norm": 1.3459560450980237,
"learning_rate": 3.953351678183099e-05,
"loss": 0.5507,
"step": 76
},
{
"epoch": 0.49201277955271566,
"grad_norm": 0.9913845916559971,
"learning_rate": 3.9500927615307745e-05,
"loss": 0.5244,
"step": 77
},
{
"epoch": 0.4984025559105431,
"grad_norm": 1.4251625906436933,
"learning_rate": 3.9467252551346e-05,
"loss": 0.5594,
"step": 78
},
{
"epoch": 0.5047923322683706,
"grad_norm": 1.0112792856876618,
"learning_rate": 3.943249346512143e-05,
"loss": 0.4883,
"step": 79
},
{
"epoch": 0.5111821086261981,
"grad_norm": 1.0989399875889763,
"learning_rate": 3.939665229217289e-05,
"loss": 0.5144,
"step": 80
},
{
"epoch": 0.5175718849840255,
"grad_norm": 1.0239637264891768,
"learning_rate": 3.935973102829456e-05,
"loss": 0.5527,
"step": 81
},
{
"epoch": 0.5239616613418531,
"grad_norm": 1.0068238967871816,
"learning_rate": 3.9321731729424826e-05,
"loss": 0.5672,
"step": 82
},
{
"epoch": 0.5303514376996805,
"grad_norm": 0.9875459555325186,
"learning_rate": 3.928265651153183e-05,
"loss": 0.529,
"step": 83
},
{
"epoch": 0.536741214057508,
"grad_norm": 1.0535928381866786,
"learning_rate": 3.92425075504956e-05,
"loss": 0.5037,
"step": 84
},
{
"epoch": 0.5431309904153354,
"grad_norm": 0.8056689145152576,
"learning_rate": 3.9201287081986905e-05,
"loss": 0.5561,
"step": 85
},
{
"epoch": 0.549520766773163,
"grad_norm": 0.9754123651205087,
"learning_rate": 3.915899740134277e-05,
"loss": 0.533,
"step": 86
},
{
"epoch": 0.5559105431309904,
"grad_norm": 1.15840636473317,
"learning_rate": 3.9115640863438666e-05,
"loss": 0.5493,
"step": 87
},
{
"epoch": 0.5623003194888179,
"grad_norm": 0.8988153421079207,
"learning_rate": 3.907121988255735e-05,
"loss": 0.5196,
"step": 88
},
{
"epoch": 0.5686900958466453,
"grad_norm": 1.1383255251414823,
"learning_rate": 3.9025736932254463e-05,
"loss": 0.527,
"step": 89
},
{
"epoch": 0.5750798722044729,
"grad_norm": 1.1283342937219425,
"learning_rate": 3.8979194545220754e-05,
"loss": 0.5295,
"step": 90
},
{
"epoch": 0.5814696485623003,
"grad_norm": 0.9227752366498948,
"learning_rate": 3.893159531314109e-05,
"loss": 0.5306,
"step": 91
},
{
"epoch": 0.5878594249201278,
"grad_norm": 1.0643565858340844,
"learning_rate": 3.888294188655012e-05,
"loss": 0.5093,
"step": 92
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.7000539222261911,
"learning_rate": 3.8833236974684666e-05,
"loss": 0.4991,
"step": 93
},
{
"epoch": 0.6006389776357828,
"grad_norm": 1.1723297653142641,
"learning_rate": 3.8782483345332876e-05,
"loss": 0.5068,
"step": 94
},
{
"epoch": 0.6070287539936102,
"grad_norm": 0.7821654004337761,
"learning_rate": 3.873068382468012e-05,
"loss": 0.4921,
"step": 95
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.9387810448789369,
"learning_rate": 3.8677841297151595e-05,
"loss": 0.5062,
"step": 96
},
{
"epoch": 0.6198083067092651,
"grad_norm": 1.0554276403586835,
"learning_rate": 3.862395870525167e-05,
"loss": 0.5563,
"step": 97
},
{
"epoch": 0.6261980830670927,
"grad_norm": 0.7778950998937124,
"learning_rate": 3.856903904940014e-05,
"loss": 0.5336,
"step": 98
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.9707505249343071,
"learning_rate": 3.851308538776501e-05,
"loss": 0.5042,
"step": 99
},
{
"epoch": 0.6389776357827476,
"grad_norm": 0.9480577698679056,
"learning_rate": 3.845610083609234e-05,
"loss": 0.5125,
"step": 100
},
{
"epoch": 0.645367412140575,
"grad_norm": 1.000490021841588,
"learning_rate": 3.839808856753264e-05,
"loss": 0.5378,
"step": 101
},
{
"epoch": 0.6517571884984026,
"grad_norm": 1.150496995903701,
"learning_rate": 3.833905181246424e-05,
"loss": 0.5201,
"step": 102
},
{
"epoch": 0.65814696485623,
"grad_norm": 1.119406744384519,
"learning_rate": 3.827899385831337e-05,
"loss": 0.5219,
"step": 103
},
{
"epoch": 0.6645367412140575,
"grad_norm": 1.0112237142528195,
"learning_rate": 3.821791804937114e-05,
"loss": 0.5298,
"step": 104
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.7034761817860105,
"learning_rate": 3.815582778660726e-05,
"loss": 0.5192,
"step": 105
},
{
"epoch": 0.6773162939297125,
"grad_norm": 1.0053867926801359,
"learning_rate": 3.809272652748072e-05,
"loss": 0.5276,
"step": 106
},
{
"epoch": 0.6837060702875399,
"grad_norm": 0.7482460221471098,
"learning_rate": 3.8028617785747224e-05,
"loss": 0.5608,
"step": 107
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.9865927820444385,
"learning_rate": 3.7963505131263534e-05,
"loss": 0.4968,
"step": 108
},
{
"epoch": 0.6964856230031949,
"grad_norm": 1.038453462373431,
"learning_rate": 3.789739218978871e-05,
"loss": 0.5693,
"step": 109
},
{
"epoch": 0.7028753993610224,
"grad_norm": 0.6920319585898572,
"learning_rate": 3.783028264278217e-05,
"loss": 0.5372,
"step": 110
},
{
"epoch": 0.7092651757188498,
"grad_norm": 1.0352100256938261,
"learning_rate": 3.7762180227198707e-05,
"loss": 0.5277,
"step": 111
},
{
"epoch": 0.7156549520766773,
"grad_norm": 0.64046404209257,
"learning_rate": 3.7693088735280435e-05,
"loss": 0.4962,
"step": 112
},
{
"epoch": 0.7220447284345048,
"grad_norm": 1.0958799819009744,
"learning_rate": 3.762301201434558e-05,
"loss": 0.5407,
"step": 113
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.7248797975531077,
"learning_rate": 3.755195396657423e-05,
"loss": 0.5282,
"step": 114
},
{
"epoch": 0.7348242811501597,
"grad_norm": 1.051488291269328,
"learning_rate": 3.74799185487911e-05,
"loss": 0.5331,
"step": 115
},
{
"epoch": 0.7412140575079872,
"grad_norm": 1.080506027543262,
"learning_rate": 3.7406909772245136e-05,
"loss": 0.5433,
"step": 116
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.8852634100172303,
"learning_rate": 3.73329317023862e-05,
"loss": 0.5128,
"step": 117
},
{
"epoch": 0.7539936102236422,
"grad_norm": 1.1931790977018395,
"learning_rate": 3.725798845863868e-05,
"loss": 0.5067,
"step": 118
},
{
"epoch": 0.7603833865814696,
"grad_norm": 0.9230539545451707,
"learning_rate": 3.718208421417208e-05,
"loss": 0.518,
"step": 119
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.9288097411893196,
"learning_rate": 3.710522319566864e-05,
"loss": 0.5025,
"step": 120
},
{
"epoch": 0.7731629392971247,
"grad_norm": 0.8269049032128745,
"learning_rate": 3.7027409683087994e-05,
"loss": 0.509,
"step": 121
},
{
"epoch": 0.7795527156549521,
"grad_norm": 1.063330879029129,
"learning_rate": 3.694864800942886e-05,
"loss": 0.5577,
"step": 122
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.7741080297489087,
"learning_rate": 3.686894256048771e-05,
"loss": 0.5132,
"step": 123
},
{
"epoch": 0.792332268370607,
"grad_norm": 0.9963198870882927,
"learning_rate": 3.678829777461457e-05,
"loss": 0.5456,
"step": 124
},
{
"epoch": 0.7987220447284346,
"grad_norm": 0.9303282772631454,
"learning_rate": 3.670671814246591e-05,
"loss": 0.51,
"step": 125
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.8193974356952273,
"learning_rate": 3.662420820675453e-05,
"loss": 0.4998,
"step": 126
},
{
"epoch": 0.8115015974440895,
"grad_norm": 0.8935114136883495,
"learning_rate": 3.654077256199662e-05,
"loss": 0.4945,
"step": 127
},
{
"epoch": 0.8178913738019169,
"grad_norm": 0.8144910990007458,
"learning_rate": 3.645641585425594e-05,
"loss": 0.5031,
"step": 128
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.8181401765209083,
"learning_rate": 3.6371142780885085e-05,
"loss": 0.4892,
"step": 129
},
{
"epoch": 0.8306709265175719,
"grad_norm": 0.8625236903783421,
"learning_rate": 3.628495809026391e-05,
"loss": 0.526,
"step": 130
},
{
"epoch": 0.8370607028753994,
"grad_norm": 0.807030171859448,
"learning_rate": 3.619786658153514e-05,
"loss": 0.5289,
"step": 131
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.8764232140015433,
"learning_rate": 3.6109873104337096e-05,
"loss": 0.5394,
"step": 132
},
{
"epoch": 0.8498402555910544,
"grad_norm": 0.6976250810888259,
"learning_rate": 3.602098255853372e-05,
"loss": 0.4852,
"step": 133
},
{
"epoch": 0.8562300319488818,
"grad_norm": 0.721487679418641,
"learning_rate": 3.593119989394163e-05,
"loss": 0.4989,
"step": 134
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.8535899066215291,
"learning_rate": 3.584053011005457e-05,
"loss": 0.5261,
"step": 135
},
{
"epoch": 0.8690095846645367,
"grad_norm": 0.8308742651526032,
"learning_rate": 3.574897825576499e-05,
"loss": 0.5336,
"step": 136
},
{
"epoch": 0.8753993610223643,
"grad_norm": 1.04154322336692,
"learning_rate": 3.565654942908289e-05,
"loss": 0.5291,
"step": 137
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.7174442743523838,
"learning_rate": 3.556324877685195e-05,
"loss": 0.5135,
"step": 138
},
{
"epoch": 0.8881789137380192,
"grad_norm": 1.0786600177139642,
"learning_rate": 3.546908149446294e-05,
"loss": 0.4844,
"step": 139
},
{
"epoch": 0.8945686900958466,
"grad_norm": 0.9288876630518276,
"learning_rate": 3.53740528255644e-05,
"loss": 0.5467,
"step": 140
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.7709318953465457,
"learning_rate": 3.527816806177068e-05,
"loss": 0.5182,
"step": 141
},
{
"epoch": 0.9073482428115016,
"grad_norm": 0.8923057265426987,
"learning_rate": 3.518143254236724e-05,
"loss": 0.5033,
"step": 142
},
{
"epoch": 0.9137380191693291,
"grad_norm": 0.6637070684857435,
"learning_rate": 3.508385165401336e-05,
"loss": 0.5127,
"step": 143
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.9230168482815491,
"learning_rate": 3.4985430830442175e-05,
"loss": 0.516,
"step": 144
},
{
"epoch": 0.9265175718849841,
"grad_norm": 0.8941920473315101,
"learning_rate": 3.4886175552158135e-05,
"loss": 0.4947,
"step": 145
},
{
"epoch": 0.9329073482428115,
"grad_norm": 0.8990821129059456,
"learning_rate": 3.478609134613176e-05,
"loss": 0.5022,
"step": 146
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.8662214281666614,
"learning_rate": 3.4685183785491956e-05,
"loss": 0.5149,
"step": 147
},
{
"epoch": 0.9456869009584664,
"grad_norm": 0.761583043458421,
"learning_rate": 3.45834584892156e-05,
"loss": 0.5484,
"step": 148
},
{
"epoch": 0.952076677316294,
"grad_norm": 0.9833756730405143,
"learning_rate": 3.448092112181469e-05,
"loss": 0.5113,
"step": 149
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.6553490250245431,
"learning_rate": 3.4377577393020955e-05,
"loss": 0.5057,
"step": 150
},
{
"epoch": 0.9648562300319489,
"grad_norm": 0.9077550016293252,
"learning_rate": 3.427343305746783e-05,
"loss": 0.4993,
"step": 151
},
{
"epoch": 0.9712460063897763,
"grad_norm": 0.7171270449880269,
"learning_rate": 3.416849391437007e-05,
"loss": 0.4784,
"step": 152
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.9933525489448933,
"learning_rate": 3.406276580720082e-05,
"loss": 0.4958,
"step": 153
},
{
"epoch": 0.9840255591054313,
"grad_norm": 0.7952422694005465,
"learning_rate": 3.3956254623366206e-05,
"loss": 0.5126,
"step": 154
},
{
"epoch": 0.9904153354632588,
"grad_norm": 0.7573030595910021,
"learning_rate": 3.384896629387751e-05,
"loss": 0.4954,
"step": 155
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.6608346407299301,
"learning_rate": 3.374090679302091e-05,
"loss": 0.5326,
"step": 156
},
{
"epoch": 1.0031948881789137,
"grad_norm": 1.221115547246655,
"learning_rate": 3.363208213802478e-05,
"loss": 0.7746,
"step": 157
},
{
"epoch": 1.0095846645367412,
"grad_norm": 0.755241193707458,
"learning_rate": 3.352249838872466e-05,
"loss": 0.4163,
"step": 158
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.5770894202417984,
"learning_rate": 3.3412161647225795e-05,
"loss": 0.3854,
"step": 159
},
{
"epoch": 1.0223642172523961,
"grad_norm": 0.6479907351500458,
"learning_rate": 3.330107805756334e-05,
"loss": 0.3674,
"step": 160
},
{
"epoch": 1.0287539936102237,
"grad_norm": 0.8588766728374346,
"learning_rate": 3.318925380536026e-05,
"loss": 0.4473,
"step": 161
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.6190221770598182,
"learning_rate": 3.307669511748285e-05,
"loss": 0.3834,
"step": 162
},
{
"epoch": 1.0415335463258786,
"grad_norm": 0.9380647190076025,
"learning_rate": 3.2963408261694026e-05,
"loss": 0.4215,
"step": 163
},
{
"epoch": 1.0479233226837061,
"grad_norm": 0.8105378167398238,
"learning_rate": 3.284939954630428e-05,
"loss": 0.4247,
"step": 164
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.7819386983119491,
"learning_rate": 3.2734675319820446e-05,
"loss": 0.4013,
"step": 165
},
{
"epoch": 1.060702875399361,
"grad_norm": 1.1818150256172972,
"learning_rate": 3.261924197059215e-05,
"loss": 0.4592,
"step": 166
},
{
"epoch": 1.0670926517571886,
"grad_norm": 0.7222613902999555,
"learning_rate": 3.25031059264561e-05,
"loss": 0.3779,
"step": 167
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.695599012715811,
"learning_rate": 3.2386273654378134e-05,
"loss": 0.3354,
"step": 168
},
{
"epoch": 1.0798722044728435,
"grad_norm": 0.6636714986165804,
"learning_rate": 3.226875166009313e-05,
"loss": 0.3658,
"step": 169
},
{
"epoch": 1.0862619808306708,
"grad_norm": 0.9119019299436991,
"learning_rate": 3.215054648774274e-05,
"loss": 0.4916,
"step": 170
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.6438001573913675,
"learning_rate": 3.2031664719510967e-05,
"loss": 0.3829,
"step": 171
},
{
"epoch": 1.099041533546326,
"grad_norm": 0.892962035381319,
"learning_rate": 3.1912112975257664e-05,
"loss": 0.4622,
"step": 172
},
{
"epoch": 1.1054313099041533,
"grad_norm": 0.7025792696391929,
"learning_rate": 3.1791897912149874e-05,
"loss": 0.3948,
"step": 173
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.659137823569491,
"learning_rate": 3.167102622429117e-05,
"loss": 0.364,
"step": 174
},
{
"epoch": 1.1182108626198084,
"grad_norm": 0.7388120415365451,
"learning_rate": 3.1549504642348856e-05,
"loss": 0.3855,
"step": 175
},
{
"epoch": 1.1246006389776357,
"grad_norm": 0.6711108216098111,
"learning_rate": 3.142733993317924e-05,
"loss": 0.3966,
"step": 176
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.6360807358108843,
"learning_rate": 3.130453889945075e-05,
"loss": 0.4024,
"step": 177
},
{
"epoch": 1.1373801916932909,
"grad_norm": 0.6324189051210329,
"learning_rate": 3.118110837926517e-05,
"loss": 0.3553,
"step": 178
},
{
"epoch": 1.1437699680511182,
"grad_norm": 0.7964831390454978,
"learning_rate": 3.105705524577686e-05,
"loss": 0.4578,
"step": 179
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.6710667005919045,
"learning_rate": 3.093238640681003e-05,
"loss": 0.3781,
"step": 180
},
{
"epoch": 1.156549520766773,
"grad_norm": 0.6387506156404197,
"learning_rate": 3.080710880447408e-05,
"loss": 0.3887,
"step": 181
},
{
"epoch": 1.1629392971246006,
"grad_norm": 0.7478222110923342,
"learning_rate": 3.068122941477704e-05,
"loss": 0.4388,
"step": 182
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.6263047219752043,
"learning_rate": 3.05547552472371e-05,
"loss": 0.3889,
"step": 183
},
{
"epoch": 1.1757188498402555,
"grad_norm": 0.8473527998575385,
"learning_rate": 3.0427693344492278e-05,
"loss": 0.4279,
"step": 184
},
{
"epoch": 1.182108626198083,
"grad_norm": 0.5095696451206226,
"learning_rate": 3.0300050781908305e-05,
"loss": 0.3517,
"step": 185
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.7415470456207829,
"learning_rate": 3.0171834667184597e-05,
"loss": 0.3922,
"step": 186
},
{
"epoch": 1.194888178913738,
"grad_norm": 0.8177392758602474,
"learning_rate": 3.0043052139958465e-05,
"loss": 0.4332,
"step": 187
},
{
"epoch": 1.2012779552715656,
"grad_norm": 0.5193140042572805,
"learning_rate": 2.9913710371407574e-05,
"loss": 0.3973,
"step": 188
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.8215524197527592,
"learning_rate": 2.97838165638506e-05,
"loss": 0.3988,
"step": 189
},
{
"epoch": 1.2140575079872205,
"grad_norm": 0.6518458707691871,
"learning_rate": 2.965337795034617e-05,
"loss": 0.4124,
"step": 190
},
{
"epoch": 1.220447284345048,
"grad_norm": 0.7141025249008749,
"learning_rate": 2.9522401794290127e-05,
"loss": 0.478,
"step": 191
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.6807610256712961,
"learning_rate": 2.9390895389011024e-05,
"loss": 0.3373,
"step": 192
},
{
"epoch": 1.233226837060703,
"grad_norm": 0.6877212054512275,
"learning_rate": 2.9258866057364028e-05,
"loss": 0.481,
"step": 193
},
{
"epoch": 1.2396166134185305,
"grad_norm": 0.6607586787800477,
"learning_rate": 2.912632115132316e-05,
"loss": 0.3599,
"step": 194
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.6239197081486476,
"learning_rate": 2.8993268051571877e-05,
"loss": 0.4054,
"step": 195
},
{
"epoch": 1.2523961661341854,
"grad_norm": 0.6374267993723667,
"learning_rate": 2.885971416709211e-05,
"loss": 0.4582,
"step": 196
},
{
"epoch": 1.2587859424920127,
"grad_norm": 0.6083181799353529,
"learning_rate": 2.8725666934751678e-05,
"loss": 0.3904,
"step": 197
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.567022003437379,
"learning_rate": 2.859113381889018e-05,
"loss": 0.3776,
"step": 198
},
{
"epoch": 1.2715654952076676,
"grad_norm": 0.5883698301032603,
"learning_rate": 2.8456122310903342e-05,
"loss": 0.3869,
"step": 199
},
{
"epoch": 1.2779552715654952,
"grad_norm": 0.6219746159300882,
"learning_rate": 2.8320639928825885e-05,
"loss": 0.4434,
"step": 200
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.5913066862304653,
"learning_rate": 2.818469421691285e-05,
"loss": 0.4034,
"step": 201
},
{
"epoch": 1.29073482428115,
"grad_norm": 0.7406263370585998,
"learning_rate": 2.804829274521954e-05,
"loss": 0.4592,
"step": 202
},
{
"epoch": 1.2971246006389776,
"grad_norm": 0.6535050713890558,
"learning_rate": 2.7911443109179958e-05,
"loss": 0.4018,
"step": 203
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.6660994073473191,
"learning_rate": 2.777415292918386e-05,
"loss": 0.3987,
"step": 204
},
{
"epoch": 1.3099041533546325,
"grad_norm": 0.5770049980372955,
"learning_rate": 2.7636429850152452e-05,
"loss": 0.3931,
"step": 205
},
{
"epoch": 1.31629392971246,
"grad_norm": 0.7664470442831808,
"learning_rate": 2.7498281541112645e-05,
"loss": 0.406,
"step": 206
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.5845799672630331,
"learning_rate": 2.735971569477002e-05,
"loss": 0.3969,
"step": 207
},
{
"epoch": 1.329073482428115,
"grad_norm": 0.7256834317235751,
"learning_rate": 2.72207400270805e-05,
"loss": 0.4243,
"step": 208
},
{
"epoch": 1.3354632587859425,
"grad_norm": 0.5160971519592359,
"learning_rate": 2.7081362276820634e-05,
"loss": 0.3745,
"step": 209
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.8175757140515724,
"learning_rate": 2.6941590205156715e-05,
"loss": 0.4768,
"step": 210
},
{
"epoch": 1.3482428115015974,
"grad_norm": 0.47324065852496966,
"learning_rate": 2.6801431595212585e-05,
"loss": 0.3741,
"step": 211
},
{
"epoch": 1.354632587859425,
"grad_norm": 0.6620362519919619,
"learning_rate": 2.666089425163623e-05,
"loss": 0.4173,
"step": 212
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.5902652547868394,
"learning_rate": 2.6519986000165197e-05,
"loss": 0.3851,
"step": 213
},
{
"epoch": 1.3674121405750799,
"grad_norm": 0.5996280221322619,
"learning_rate": 2.637871468719081e-05,
"loss": 0.3976,
"step": 214
},
{
"epoch": 1.3738019169329074,
"grad_norm": 0.5952050484113727,
"learning_rate": 2.623708817932127e-05,
"loss": 0.3557,
"step": 215
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.5629505435064166,
"learning_rate": 2.6095114362943567e-05,
"loss": 0.3675,
"step": 216
},
{
"epoch": 1.3865814696485623,
"grad_norm": 0.7284878464385949,
"learning_rate": 2.5952801143784377e-05,
"loss": 0.4192,
"step": 217
},
{
"epoch": 1.3929712460063897,
"grad_norm": 0.5611770400782802,
"learning_rate": 2.5810156446469815e-05,
"loss": 0.3929,
"step": 218
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.7474665760149969,
"learning_rate": 2.5667188214084156e-05,
"loss": 0.4085,
"step": 219
},
{
"epoch": 1.4057507987220448,
"grad_norm": 0.6711723162682844,
"learning_rate": 2.5523904407727518e-05,
"loss": 0.4069,
"step": 220
},
{
"epoch": 1.4121405750798721,
"grad_norm": 0.6895348539751915,
"learning_rate": 2.5380313006072576e-05,
"loss": 0.3844,
"step": 221
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.6675508521520543,
"learning_rate": 2.523642200492027e-05,
"loss": 0.4525,
"step": 222
},
{
"epoch": 1.4249201277955272,
"grad_norm": 0.6837935351740067,
"learning_rate": 2.5092239416754545e-05,
"loss": 0.3858,
"step": 223
},
{
"epoch": 1.4313099041533546,
"grad_norm": 0.7055234490813711,
"learning_rate": 2.4947773270296218e-05,
"loss": 0.4076,
"step": 224
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.642041869833729,
"learning_rate": 2.4803031610055847e-05,
"loss": 0.4116,
"step": 225
},
{
"epoch": 1.4440894568690097,
"grad_norm": 0.5973792017921002,
"learning_rate": 2.465802249588583e-05,
"loss": 0.3701,
"step": 226
},
{
"epoch": 1.450479233226837,
"grad_norm": 0.6561472014273614,
"learning_rate": 2.451275400253157e-05,
"loss": 0.3497,
"step": 227
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.7636693842376161,
"learning_rate": 2.436723421918185e-05,
"loss": 0.4835,
"step": 228
},
{
"epoch": 1.4632587859424921,
"grad_norm": 0.78261861893272,
"learning_rate": 2.4221471249018368e-05,
"loss": 0.373,
"step": 229
},
{
"epoch": 1.4696485623003195,
"grad_norm": 0.7179566840944903,
"learning_rate": 2.4075473208764558e-05,
"loss": 0.3923,
"step": 230
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.5427966480029829,
"learning_rate": 2.392924822823357e-05,
"loss": 0.3575,
"step": 231
},
{
"epoch": 1.4824281150159744,
"grad_norm": 0.8958851181734642,
"learning_rate": 2.378280444987561e-05,
"loss": 0.4385,
"step": 232
},
{
"epoch": 1.488817891373802,
"grad_norm": 0.5170480560665358,
"learning_rate": 2.3636150028324486e-05,
"loss": 0.3372,
"step": 233
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.7554046281590651,
"learning_rate": 2.3489293129943555e-05,
"loss": 0.4001,
"step": 234
},
{
"epoch": 1.5015974440894568,
"grad_norm": 0.7092746883912118,
"learning_rate": 2.334224193237097e-05,
"loss": 0.4067,
"step": 235
},
{
"epoch": 1.5079872204472844,
"grad_norm": 0.7165299708968617,
"learning_rate": 2.319500462406434e-05,
"loss": 0.3842,
"step": 236
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.6445134623150068,
"learning_rate": 2.3047589403844702e-05,
"loss": 0.4108,
"step": 237
},
{
"epoch": 1.5207667731629393,
"grad_norm": 0.4954710255941146,
"learning_rate": 2.290000448044004e-05,
"loss": 0.3344,
"step": 238
},
{
"epoch": 1.5271565495207668,
"grad_norm": 0.6207344756135259,
"learning_rate": 2.2752258072028145e-05,
"loss": 0.4079,
"step": 239
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.4812466788237613,
"learning_rate": 2.260435840577901e-05,
"loss": 0.3711,
"step": 240
},
{
"epoch": 1.5399361022364217,
"grad_norm": 0.7722622559224537,
"learning_rate": 2.245631371739669e-05,
"loss": 0.4798,
"step": 241
},
{
"epoch": 1.5463258785942493,
"grad_norm": 0.514025904239645,
"learning_rate": 2.2308132250660717e-05,
"loss": 0.3586,
"step": 242
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.6211683701769444,
"learning_rate": 2.215982225696704e-05,
"loss": 0.341,
"step": 243
},
{
"epoch": 1.5591054313099042,
"grad_norm": 0.5633667247279115,
"learning_rate": 2.201139199486856e-05,
"loss": 0.3927,
"step": 244
},
{
"epoch": 1.5654952076677318,
"grad_norm": 0.5601679935599658,
"learning_rate": 2.186284972961525e-05,
"loss": 0.4049,
"step": 245
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.5442755218693436,
"learning_rate": 2.1714203732693905e-05,
"loss": 0.3493,
"step": 246
},
{
"epoch": 1.5782747603833864,
"grad_norm": 0.7005155470203877,
"learning_rate": 2.1565462281367572e-05,
"loss": 0.446,
"step": 247
},
{
"epoch": 1.5846645367412142,
"grad_norm": 0.5005677905989587,
"learning_rate": 2.141663365821459e-05,
"loss": 0.3761,
"step": 248
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.598933880956211,
"learning_rate": 2.1267726150667426e-05,
"loss": 0.4227,
"step": 249
},
{
"epoch": 1.5974440894568689,
"grad_norm": 0.5823697199633178,
"learning_rate": 2.1118748050551194e-05,
"loss": 0.4185,
"step": 250
},
{
"epoch": 1.6038338658146964,
"grad_norm": 0.4661817463909431,
"learning_rate": 2.0969707653621884e-05,
"loss": 0.335,
"step": 251
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.534066282750192,
"learning_rate": 2.0820613259104462e-05,
"loss": 0.4249,
"step": 252
},
{
"epoch": 1.6166134185303513,
"grad_norm": 0.5288717205900548,
"learning_rate": 2.0671473169230715e-05,
"loss": 0.4078,
"step": 253
},
{
"epoch": 1.623003194888179,
"grad_norm": 0.5850676124543679,
"learning_rate": 2.0522295688776945e-05,
"loss": 0.3872,
"step": 254
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.46385901267321095,
"learning_rate": 2.037308912460153e-05,
"loss": 0.3415,
"step": 255
},
{
"epoch": 1.6357827476038338,
"grad_norm": 0.6386876033156802,
"learning_rate": 2.022386178518236e-05,
"loss": 0.417,
"step": 256
},
{
"epoch": 1.6421725239616614,
"grad_norm": 0.4805587222138907,
"learning_rate": 2.0074621980154167e-05,
"loss": 0.3882,
"step": 257
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.5746353853534201,
"learning_rate": 1.9925378019845837e-05,
"loss": 0.4104,
"step": 258
},
{
"epoch": 1.6549520766773163,
"grad_norm": 0.501859313648766,
"learning_rate": 1.9776138214817648e-05,
"loss": 0.3774,
"step": 259
},
{
"epoch": 1.6613418530351438,
"grad_norm": 0.5820108106526879,
"learning_rate": 1.9626910875398475e-05,
"loss": 0.3576,
"step": 260
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.5511522872814458,
"learning_rate": 1.9477704311223062e-05,
"loss": 0.4259,
"step": 261
},
{
"epoch": 1.6741214057507987,
"grad_norm": 0.5192707555977589,
"learning_rate": 1.932852683076929e-05,
"loss": 0.4159,
"step": 262
},
{
"epoch": 1.680511182108626,
"grad_norm": 0.5414396653048178,
"learning_rate": 1.9179386740895538e-05,
"loss": 0.3925,
"step": 263
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.5148601689863671,
"learning_rate": 1.9030292346378116e-05,
"loss": 0.3598,
"step": 264
},
{
"epoch": 1.6932907348242812,
"grad_norm": 0.5119362335475469,
"learning_rate": 1.8881251949448817e-05,
"loss": 0.4128,
"step": 265
},
{
"epoch": 1.6996805111821085,
"grad_norm": 0.5906313617691342,
"learning_rate": 1.8732273849332577e-05,
"loss": 0.3892,
"step": 266
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.5726182137360729,
"learning_rate": 1.8583366341785424e-05,
"loss": 0.4135,
"step": 267
},
{
"epoch": 1.7124600638977636,
"grad_norm": 0.47501236190964014,
"learning_rate": 1.8434537718632438e-05,
"loss": 0.3966,
"step": 268
},
{
"epoch": 1.718849840255591,
"grad_norm": 0.5163914225737597,
"learning_rate": 1.8285796267306098e-05,
"loss": 0.3321,
"step": 269
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.5440727291416668,
"learning_rate": 1.8137150270384755e-05,
"loss": 0.4162,
"step": 270
},
{
"epoch": 1.731629392971246,
"grad_norm": 0.4932582814102065,
"learning_rate": 1.7988608005131445e-05,
"loss": 0.4086,
"step": 271
},
{
"epoch": 1.7380191693290734,
"grad_norm": 0.533011592071942,
"learning_rate": 1.7840177743032964e-05,
"loss": 0.414,
"step": 272
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.49102076107513076,
"learning_rate": 1.769186774933929e-05,
"loss": 0.4176,
"step": 273
},
{
"epoch": 1.7507987220447285,
"grad_norm": 0.5771966331162738,
"learning_rate": 1.7543686282603318e-05,
"loss": 0.4432,
"step": 274
},
{
"epoch": 1.7571884984025559,
"grad_norm": 0.5461736300422602,
"learning_rate": 1.7395641594220998e-05,
"loss": 0.3829,
"step": 275
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.6577548367990437,
"learning_rate": 1.724774192797186e-05,
"loss": 0.382,
"step": 276
},
{
"epoch": 1.769968051118211,
"grad_norm": 0.43878563775490903,
"learning_rate": 1.7099995519559963e-05,
"loss": 0.3575,
"step": 277
},
{
"epoch": 1.7763578274760383,
"grad_norm": 0.5120401210743035,
"learning_rate": 1.69524105961553e-05,
"loss": 0.4286,
"step": 278
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5653816201782214,
"learning_rate": 1.6804995375935674e-05,
"loss": 0.3742,
"step": 279
},
{
"epoch": 1.7891373801916934,
"grad_norm": 0.5726713802278968,
"learning_rate": 1.665775806762903e-05,
"loss": 0.3685,
"step": 280
},
{
"epoch": 1.7955271565495208,
"grad_norm": 0.5229354326981167,
"learning_rate": 1.6510706870056455e-05,
"loss": 0.408,
"step": 281
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.4954167189576208,
"learning_rate": 1.6363849971675524e-05,
"loss": 0.3659,
"step": 282
},
{
"epoch": 1.8083067092651757,
"grad_norm": 0.5076708644451216,
"learning_rate": 1.6217195550124397e-05,
"loss": 0.3686,
"step": 283
},
{
"epoch": 1.8146964856230032,
"grad_norm": 0.5824051094209408,
"learning_rate": 1.6070751771766434e-05,
"loss": 0.4467,
"step": 284
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5285624925239529,
"learning_rate": 1.592452679123545e-05,
"loss": 0.3776,
"step": 285
},
{
"epoch": 1.8274760383386581,
"grad_norm": 0.5376124276180292,
"learning_rate": 1.577852875098164e-05,
"loss": 0.4504,
"step": 286
},
{
"epoch": 1.8338658146964857,
"grad_norm": 0.5373102243429786,
"learning_rate": 1.563276578081816e-05,
"loss": 0.3501,
"step": 287
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.5462928026195687,
"learning_rate": 1.5487245997468434e-05,
"loss": 0.4209,
"step": 288
},
{
"epoch": 1.8466453674121406,
"grad_norm": 0.4909580614215068,
"learning_rate": 1.5341977504114176e-05,
"loss": 0.3647,
"step": 289
},
{
"epoch": 1.8530351437699681,
"grad_norm": 0.5421878755182835,
"learning_rate": 1.5196968389944153e-05,
"loss": 0.4069,
"step": 290
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.5053862903315562,
"learning_rate": 1.5052226729703782e-05,
"loss": 0.3763,
"step": 291
},
{
"epoch": 1.865814696485623,
"grad_norm": 0.43565640748900136,
"learning_rate": 1.4907760583245451e-05,
"loss": 0.367,
"step": 292
},
{
"epoch": 1.8722044728434506,
"grad_norm": 0.6034830380300681,
"learning_rate": 1.4763577995079738e-05,
"loss": 0.3964,
"step": 293
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.5420427837665357,
"learning_rate": 1.461968699392743e-05,
"loss": 0.3378,
"step": 294
},
{
"epoch": 1.8849840255591053,
"grad_norm": 0.5951922897755679,
"learning_rate": 1.4476095592272494e-05,
"loss": 0.4326,
"step": 295
},
{
"epoch": 1.891373801916933,
"grad_norm": 0.4840728145584482,
"learning_rate": 1.4332811785915852e-05,
"loss": 0.369,
"step": 296
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.5037450720887865,
"learning_rate": 1.418984355353019e-05,
"loss": 0.3897,
"step": 297
},
{
"epoch": 1.9041533546325877,
"grad_norm": 0.6081578605737592,
"learning_rate": 1.404719885621563e-05,
"loss": 0.4251,
"step": 298
},
{
"epoch": 1.9105431309904153,
"grad_norm": 0.46944637013816987,
"learning_rate": 1.390488563705644e-05,
"loss": 0.3646,
"step": 299
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.5228222176493802,
"learning_rate": 1.3762911820678739e-05,
"loss": 0.4235,
"step": 300
},
{
"epoch": 1.9233226837060702,
"grad_norm": 0.5899737215877533,
"learning_rate": 1.3621285312809193e-05,
"loss": 0.3794,
"step": 301
},
{
"epoch": 1.9297124600638977,
"grad_norm": 0.5572929497186375,
"learning_rate": 1.3480013999834811e-05,
"loss": 0.4011,
"step": 302
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.5632588672503747,
"learning_rate": 1.3339105748363776e-05,
"loss": 0.3981,
"step": 303
},
{
"epoch": 1.9424920127795526,
"grad_norm": 0.5276160699091037,
"learning_rate": 1.3198568404787418e-05,
"loss": 0.3767,
"step": 304
},
{
"epoch": 1.9488817891373802,
"grad_norm": 0.6024773811499252,
"learning_rate": 1.3058409794843288e-05,
"loss": 0.4415,
"step": 305
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.5259393060101312,
"learning_rate": 1.2918637723179371e-05,
"loss": 0.3879,
"step": 306
},
{
"epoch": 1.961661341853035,
"grad_norm": 0.5317965049616244,
"learning_rate": 1.2779259972919512e-05,
"loss": 0.4269,
"step": 307
},
{
"epoch": 1.9680511182108626,
"grad_norm": 0.45743435082366124,
"learning_rate": 1.2640284305229989e-05,
"loss": 0.3205,
"step": 308
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.5754797561825906,
"learning_rate": 1.2501718458887369e-05,
"loss": 0.395,
"step": 309
},
{
"epoch": 1.9808306709265175,
"grad_norm": 0.5096694579354405,
"learning_rate": 1.2363570149847556e-05,
"loss": 0.3931,
"step": 310
},
{
"epoch": 1.9872204472843449,
"grad_norm": 0.5649164808169996,
"learning_rate": 1.2225847070816148e-05,
"loss": 0.3755,
"step": 311
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.4969808916941929,
"learning_rate": 1.208855689082005e-05,
"loss": 0.4243,
"step": 312
},
{
"epoch": 2.0,
"grad_norm": 0.7954635925002451,
"learning_rate": 1.1951707254780465e-05,
"loss": 0.4858,
"step": 313
},
{
"epoch": 2.0063897763578273,
"grad_norm": 0.6662795810886528,
"learning_rate": 1.1815305783087153e-05,
"loss": 0.2806,
"step": 314
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.5137310238156836,
"learning_rate": 1.1679360071174121e-05,
"loss": 0.295,
"step": 315
},
{
"epoch": 2.0191693290734825,
"grad_norm": 0.815847533118561,
"learning_rate": 1.1543877689096661e-05,
"loss": 0.2896,
"step": 316
},
{
"epoch": 2.02555910543131,
"grad_norm": 0.5069824950228584,
"learning_rate": 1.1408866181109824e-05,
"loss": 0.2837,
"step": 317
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.5850695127338842,
"learning_rate": 1.1274333065248327e-05,
"loss": 0.2856,
"step": 318
},
{
"epoch": 2.038338658146965,
"grad_norm": 0.5779513618203883,
"learning_rate": 1.1140285832907888e-05,
"loss": 0.2885,
"step": 319
},
{
"epoch": 2.0447284345047922,
"grad_norm": 0.5351383370576208,
"learning_rate": 1.1006731948428118e-05,
"loss": 0.3018,
"step": 320
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.4884202390865607,
"learning_rate": 1.0873678848676852e-05,
"loss": 0.2815,
"step": 321
},
{
"epoch": 2.0575079872204474,
"grad_norm": 0.5191469140904564,
"learning_rate": 1.074113394263598e-05,
"loss": 0.2685,
"step": 322
},
{
"epoch": 2.0638977635782747,
"grad_norm": 0.541141996904648,
"learning_rate": 1.0609104610988986e-05,
"loss": 0.2809,
"step": 323
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.5063236911612786,
"learning_rate": 1.0477598205709882e-05,
"loss": 0.2979,
"step": 324
},
{
"epoch": 2.07667731629393,
"grad_norm": 0.48822674540291033,
"learning_rate": 1.0346622049653832e-05,
"loss": 0.2493,
"step": 325
},
{
"epoch": 2.083067092651757,
"grad_norm": 0.5172260293987238,
"learning_rate": 1.0216183436149412e-05,
"loss": 0.3062,
"step": 326
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.5077833533693695,
"learning_rate": 1.0086289628592429e-05,
"loss": 0.2817,
"step": 327
},
{
"epoch": 2.0958466453674123,
"grad_norm": 0.47094358503601513,
"learning_rate": 9.956947860041535e-06,
"loss": 0.2633,
"step": 328
},
{
"epoch": 2.1022364217252396,
"grad_norm": 0.4678422611989536,
"learning_rate": 9.82816533281541e-06,
"loss": 0.2519,
"step": 329
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.4961545238058503,
"learning_rate": 9.699949218091698e-06,
"loss": 0.2806,
"step": 330
},
{
"epoch": 2.1150159744408947,
"grad_norm": 0.45400755049188124,
"learning_rate": 9.57230665550773e-06,
"loss": 0.2845,
"step": 331
},
{
"epoch": 2.121405750798722,
"grad_norm": 0.4087039835019054,
"learning_rate": 9.44524475276291e-06,
"loss": 0.2712,
"step": 332
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.45141929812295156,
"learning_rate": 9.318770585222959e-06,
"loss": 0.2751,
"step": 333
},
{
"epoch": 2.134185303514377,
"grad_norm": 0.4285243933890445,
"learning_rate": 9.19289119552592e-06,
"loss": 0.2715,
"step": 334
},
{
"epoch": 2.1405750798722045,
"grad_norm": 0.4298003352353432,
"learning_rate": 9.067613593189978e-06,
"loss": 0.2806,
"step": 335
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.45672324827943345,
"learning_rate": 8.942944754223148e-06,
"loss": 0.2973,
"step": 336
},
{
"epoch": 2.1533546325878596,
"grad_norm": 0.4370811146410844,
"learning_rate": 8.818891620734843e-06,
"loss": 0.3125,
"step": 337
},
{
"epoch": 2.159744408945687,
"grad_norm": 0.43681333400547245,
"learning_rate": 8.695461100549257e-06,
"loss": 0.2928,
"step": 338
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.4180585221813101,
"learning_rate": 8.572660066820768e-06,
"loss": 0.2434,
"step": 339
},
{
"epoch": 2.1725239616613417,
"grad_norm": 0.42261573045223133,
"learning_rate": 8.450495357651147e-06,
"loss": 0.2675,
"step": 340
},
{
"epoch": 2.1789137380191694,
"grad_norm": 0.4068510209903175,
"learning_rate": 8.328973775708848e-06,
"loss": 0.2813,
"step": 341
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.388040708146111,
"learning_rate": 8.208102087850136e-06,
"loss": 0.2878,
"step": 342
},
{
"epoch": 2.191693290734824,
"grad_norm": 0.43053220862581665,
"learning_rate": 8.087887024742343e-06,
"loss": 0.2825,
"step": 343
},
{
"epoch": 2.198083067092652,
"grad_norm": 0.41793950177347855,
"learning_rate": 7.968335280489037e-06,
"loss": 0.2986,
"step": 344
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.40031758724937355,
"learning_rate": 7.849453512257264e-06,
"loss": 0.3,
"step": 345
},
{
"epoch": 2.2108626198083066,
"grad_norm": 0.38059364454447125,
"learning_rate": 7.731248339906879e-06,
"loss": 0.2727,
"step": 346
},
{
"epoch": 2.2172523961661343,
"grad_norm": 0.41990913510389755,
"learning_rate": 7.613726345621873e-06,
"loss": 0.3046,
"step": 347
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.43424406475522437,
"learning_rate": 7.496894073543899e-06,
"loss": 0.2936,
"step": 348
},
{
"epoch": 2.230031948881789,
"grad_norm": 0.3750088497925214,
"learning_rate": 7.380758029407857e-06,
"loss": 0.2664,
"step": 349
},
{
"epoch": 2.236421725239617,
"grad_norm": 0.418532871938177,
"learning_rate": 7.2653246801795594e-06,
"loss": 0.2891,
"step": 350
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.4028035425262737,
"learning_rate": 7.150600453695726e-06,
"loss": 0.2734,
"step": 351
},
{
"epoch": 2.2492012779552715,
"grad_norm": 0.4466135219208513,
"learning_rate": 7.036591738305989e-06,
"loss": 0.2815,
"step": 352
},
{
"epoch": 2.255591054313099,
"grad_norm": 0.3867166130148223,
"learning_rate": 6.923304882517159e-06,
"loss": 0.296,
"step": 353
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.3917494183899583,
"learning_rate": 6.81074619463975e-06,
"loss": 0.2325,
"step": 354
},
{
"epoch": 2.268370607028754,
"grad_norm": 0.4190504448138774,
"learning_rate": 6.698921942436665e-06,
"loss": 0.2535,
"step": 355
},
{
"epoch": 2.2747603833865817,
"grad_norm": 0.41331962647091514,
"learning_rate": 6.587838352774209e-06,
"loss": 0.2587,
"step": 356
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.41804090107839076,
"learning_rate": 6.477501611275345e-06,
"loss": 0.2758,
"step": 357
},
{
"epoch": 2.2875399361022364,
"grad_norm": 0.39910602146901847,
"learning_rate": 6.367917861975224e-06,
"loss": 0.2607,
"step": 358
},
{
"epoch": 2.2939297124600637,
"grad_norm": 0.36374725362936167,
"learning_rate": 6.2590932069790965e-06,
"loss": 0.2896,
"step": 359
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.3931796715710283,
"learning_rate": 6.151033706122489e-06,
"loss": 0.2562,
"step": 360
},
{
"epoch": 2.306709265175719,
"grad_norm": 0.39607678195113055,
"learning_rate": 6.043745376633792e-06,
"loss": 0.2575,
"step": 361
},
{
"epoch": 2.313099041533546,
"grad_norm": 0.370701894040692,
"learning_rate": 5.937234192799184e-06,
"loss": 0.2476,
"step": 362
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.3922675962076736,
"learning_rate": 5.831506085629937e-06,
"loss": 0.2758,
"step": 363
},
{
"epoch": 2.3258785942492013,
"grad_norm": 0.39508969009396355,
"learning_rate": 5.7265669425321765e-06,
"loss": 0.2572,
"step": 364
},
{
"epoch": 2.3322683706070286,
"grad_norm": 0.4127067246019672,
"learning_rate": 5.622422606979055e-06,
"loss": 0.2507,
"step": 365
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.40000330806363765,
"learning_rate": 5.519078878185311e-06,
"loss": 0.2973,
"step": 366
},
{
"epoch": 2.3450479233226837,
"grad_norm": 0.3982421880351729,
"learning_rate": 5.416541510784414e-06,
"loss": 0.2778,
"step": 367
},
{
"epoch": 2.351437699680511,
"grad_norm": 0.37370414007694336,
"learning_rate": 5.314816214508054e-06,
"loss": 0.2815,
"step": 368
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.4150724253275529,
"learning_rate": 5.21390865386824e-06,
"loss": 0.2968,
"step": 369
},
{
"epoch": 2.364217252396166,
"grad_norm": 0.3761463056883756,
"learning_rate": 5.113824447841873e-06,
"loss": 0.2363,
"step": 370
},
{
"epoch": 2.3706070287539935,
"grad_norm": 0.35871951677398056,
"learning_rate": 5.014569169557826e-06,
"loss": 0.2612,
"step": 371
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.3892389571463014,
"learning_rate": 4.916148345986649e-06,
"loss": 0.2759,
"step": 372
},
{
"epoch": 2.3833865814696487,
"grad_norm": 0.37772421711614007,
"learning_rate": 4.8185674576327615e-06,
"loss": 0.2608,
"step": 373
},
{
"epoch": 2.389776357827476,
"grad_norm": 0.382653409586813,
"learning_rate": 4.721831938229315e-06,
"loss": 0.3155,
"step": 374
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.39130961949763166,
"learning_rate": 4.625947174435596e-06,
"loss": 0.292,
"step": 375
},
{
"epoch": 2.402555910543131,
"grad_norm": 0.3689264594938094,
"learning_rate": 4.530918505537061e-06,
"loss": 0.2778,
"step": 376
},
{
"epoch": 2.4089456869009584,
"grad_norm": 0.3583777664156191,
"learning_rate": 4.436751223148055e-06,
"loss": 0.2793,
"step": 377
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.3792734486077673,
"learning_rate": 4.343450570917118e-06,
"loss": 0.2987,
"step": 378
},
{
"epoch": 2.4217252396166136,
"grad_norm": 0.3559642646144205,
"learning_rate": 4.251021744235013e-06,
"loss": 0.2782,
"step": 379
},
{
"epoch": 2.428115015974441,
"grad_norm": 0.4010256232612567,
"learning_rate": 4.159469889945434e-06,
"loss": 0.2529,
"step": 380
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.3765358138333099,
"learning_rate": 4.068800106058375e-06,
"loss": 0.279,
"step": 381
},
{
"epoch": 2.440894568690096,
"grad_norm": 0.3685887074154549,
"learning_rate": 3.9790174414662844e-06,
"loss": 0.2705,
"step": 382
},
{
"epoch": 2.4472843450479234,
"grad_norm": 0.3728894980841834,
"learning_rate": 3.890126895662904e-06,
"loss": 0.2583,
"step": 383
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.36277573516201544,
"learning_rate": 3.8021334184648685e-06,
"loss": 0.2807,
"step": 384
},
{
"epoch": 2.460063897763578,
"grad_norm": 0.35706119183313567,
"learning_rate": 3.7150419097360944e-06,
"loss": 0.2898,
"step": 385
},
{
"epoch": 2.466453674121406,
"grad_norm": 0.38011267281397304,
"learning_rate": 3.628857219114916e-06,
"loss": 0.2553,
"step": 386
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.3863113162029853,
"learning_rate": 3.543584145744057e-06,
"loss": 0.2534,
"step": 387
},
{
"epoch": 2.479233226837061,
"grad_norm": 0.439587610934195,
"learning_rate": 3.459227438003383e-06,
"loss": 0.2975,
"step": 388
},
{
"epoch": 2.4856230031948883,
"grad_norm": 0.3793664447722741,
"learning_rate": 3.3757917932454732e-06,
"loss": 0.2902,
"step": 389
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.406296707834338,
"learning_rate": 3.293281857534094e-06,
"loss": 0.2969,
"step": 390
},
{
"epoch": 2.498402555910543,
"grad_norm": 0.37691162956880797,
"learning_rate": 3.2117022253854356e-06,
"loss": 0.2889,
"step": 391
},
{
"epoch": 2.5047923322683707,
"grad_norm": 0.35367935501619513,
"learning_rate": 3.1310574395122994e-06,
"loss": 0.2516,
"step": 392
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.3525236924672651,
"learning_rate": 3.0513519905711474e-06,
"loss": 0.2781,
"step": 393
},
{
"epoch": 2.5175718849840254,
"grad_norm": 0.3564008946825266,
"learning_rate": 2.9725903169120073e-06,
"loss": 0.2698,
"step": 394
},
{
"epoch": 2.523961661341853,
"grad_norm": 0.3718042292779991,
"learning_rate": 2.894776804331372e-06,
"loss": 0.2557,
"step": 395
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.3716672854028638,
"learning_rate": 2.817915785827927e-06,
"loss": 0.2579,
"step": 396
},
{
"epoch": 2.536741214057508,
"grad_norm": 0.353812866963394,
"learning_rate": 2.742011541361318e-06,
"loss": 0.2715,
"step": 397
},
{
"epoch": 2.543130990415335,
"grad_norm": 0.3844045853121929,
"learning_rate": 2.6670682976137994e-06,
"loss": 0.3141,
"step": 398
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.37146608361893524,
"learning_rate": 2.5930902277548686e-06,
"loss": 0.2504,
"step": 399
},
{
"epoch": 2.5559105431309903,
"grad_norm": 0.35978150748451004,
"learning_rate": 2.5200814512089088e-06,
"loss": 0.2868,
"step": 400
},
{
"epoch": 2.562300319488818,
"grad_norm": 0.3498821367555298,
"learning_rate": 2.4480460334257727e-06,
"loss": 0.2778,
"step": 401
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.3590299815828268,
"learning_rate": 2.376987985654422e-06,
"loss": 0.2466,
"step": 402
},
{
"epoch": 2.5750798722044728,
"grad_norm": 0.3560328836577989,
"learning_rate": 2.3069112647195647e-06,
"loss": 0.229,
"step": 403
},
{
"epoch": 2.5814696485623,
"grad_norm": 0.37314736778508073,
"learning_rate": 2.2378197728012975e-06,
"loss": 0.2704,
"step": 404
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.3763392516149028,
"learning_rate": 2.169717357217844e-06,
"loss": 0.259,
"step": 405
},
{
"epoch": 2.594249201277955,
"grad_norm": 0.3399458492987759,
"learning_rate": 2.1026078102113014e-06,
"loss": 0.2684,
"step": 406
},
{
"epoch": 2.600638977635783,
"grad_norm": 0.3578314862739145,
"learning_rate": 2.0364948687364692e-06,
"loss": 0.2664,
"step": 407
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.39093714164893223,
"learning_rate": 1.971382214252784e-06,
"loss": 0.3064,
"step": 408
},
{
"epoch": 2.6134185303514377,
"grad_norm": 0.35226742624856194,
"learning_rate": 1.9072734725192844e-06,
"loss": 0.3043,
"step": 409
},
{
"epoch": 2.619808306709265,
"grad_norm": 0.34363517197221133,
"learning_rate": 1.844172213392743e-06,
"loss": 0.2771,
"step": 410
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.35717003311156614,
"learning_rate": 1.7820819506288688e-06,
"loss": 0.2916,
"step": 411
},
{
"epoch": 2.63258785942492,
"grad_norm": 0.3397444551813988,
"learning_rate": 1.7210061416866297e-06,
"loss": 0.2802,
"step": 412
},
{
"epoch": 2.6389776357827475,
"grad_norm": 0.3374288264378377,
"learning_rate": 1.660948187535767e-06,
"loss": 0.2996,
"step": 413
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.348235164857282,
"learning_rate": 1.6019114324673647e-06,
"loss": 0.2643,
"step": 414
},
{
"epoch": 2.6517571884984026,
"grad_norm": 0.3728291742720239,
"learning_rate": 1.543899163907665e-06,
"loss": 0.2875,
"step": 415
},
{
"epoch": 2.65814696485623,
"grad_norm": 0.33982563684444306,
"learning_rate": 1.4869146122349932e-06,
"loss": 0.2545,
"step": 416
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.34046077697953536,
"learning_rate": 1.4309609505998689e-06,
"loss": 0.2647,
"step": 417
},
{
"epoch": 2.670926517571885,
"grad_norm": 0.3459458344871144,
"learning_rate": 1.3760412947483292e-06,
"loss": 0.2972,
"step": 418
},
{
"epoch": 2.6773162939297124,
"grad_norm": 0.36579406936806147,
"learning_rate": 1.322158702848415e-06,
"loss": 0.2695,
"step": 419
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.3707374256323378,
"learning_rate": 1.269316175319879e-06,
"loss": 0.2719,
"step": 420
},
{
"epoch": 2.6900958466453675,
"grad_norm": 0.3709423910066585,
"learning_rate": 1.217516654667128e-06,
"loss": 0.298,
"step": 421
},
{
"epoch": 2.696485623003195,
"grad_norm": 0.37469291952689926,
"learning_rate": 1.1667630253153428e-06,
"loss": 0.2612,
"step": 422
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.347832218730705,
"learning_rate": 1.1170581134498849e-06,
"loss": 0.2918,
"step": 423
},
{
"epoch": 2.70926517571885,
"grad_norm": 0.36883525953686164,
"learning_rate": 1.0684046868589103e-06,
"loss": 0.2719,
"step": 424
},
{
"epoch": 2.7156549520766773,
"grad_norm": 0.34462150376832146,
"learning_rate": 1.0208054547792478e-06,
"loss": 0.2524,
"step": 425
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.33935916043461156,
"learning_rate": 9.742630677455422e-07,
"loss": 0.2525,
"step": 426
},
{
"epoch": 2.7284345047923324,
"grad_norm": 0.35482552002227297,
"learning_rate": 9.287801174426491e-07,
"loss": 0.2701,
"step": 427
},
{
"epoch": 2.7348242811501597,
"grad_norm": 0.341771079482618,
"learning_rate": 8.843591365613324e-07,
"loss": 0.2759,
"step": 428
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.3437210997091319,
"learning_rate": 8.410025986572279e-07,
"loss": 0.2507,
"step": 429
},
{
"epoch": 2.747603833865815,
"grad_norm": 0.35387497003554297,
"learning_rate": 7.987129180130981e-07,
"loss": 0.2934,
"step": 430
},
{
"epoch": 2.753993610223642,
"grad_norm": 0.3433319826548646,
"learning_rate": 7.574924495044067e-07,
"loss": 0.2816,
"step": 431
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.3445699585899121,
"learning_rate": 7.17343488468174e-07,
"loss": 0.2938,
"step": 432
},
{
"epoch": 2.7667731629392973,
"grad_norm": 0.36179132098384625,
"learning_rate": 6.782682705751775e-07,
"loss": 0.259,
"step": 433
},
{
"epoch": 2.7731629392971247,
"grad_norm": 0.35409736012157567,
"learning_rate": 6.402689717054467e-07,
"loss": 0.2694,
"step": 434
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.3347120611379767,
"learning_rate": 6.033477078271132e-07,
"loss": 0.2791,
"step": 435
},
{
"epoch": 2.7859424920127793,
"grad_norm": 0.34362095896036443,
"learning_rate": 5.675065348785724e-07,
"loss": 0.3029,
"step": 436
},
{
"epoch": 2.792332268370607,
"grad_norm": 0.3623549484091769,
"learning_rate": 5.327474486540097e-07,
"loss": 0.2857,
"step": 437
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.3849385006693377,
"learning_rate": 4.990723846922563e-07,
"loss": 0.2437,
"step": 438
},
{
"epoch": 2.8051118210862622,
"grad_norm": 0.3456451470480186,
"learning_rate": 4.664832181690204e-07,
"loss": 0.2464,
"step": 439
},
{
"epoch": 2.8115015974440896,
"grad_norm": 0.3623322322297761,
"learning_rate": 4.3498176379245696e-07,
"loss": 0.266,
"step": 440
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.33249253792620725,
"learning_rate": 4.045697757021327e-07,
"loss": 0.2745,
"step": 441
},
{
"epoch": 2.8242811501597442,
"grad_norm": 0.35011482248093373,
"learning_rate": 3.7524894737133164e-07,
"loss": 0.2849,
"step": 442
},
{
"epoch": 2.830670926517572,
"grad_norm": 0.35867683687456126,
"learning_rate": 3.4702091151276094e-07,
"loss": 0.309,
"step": 443
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.35152581911491154,
"learning_rate": 3.1988723998764405e-07,
"loss": 0.2618,
"step": 444
},
{
"epoch": 2.8434504792332267,
"grad_norm": 0.3729635038125147,
"learning_rate": 2.938494437181727e-07,
"loss": 0.2999,
"step": 445
},
{
"epoch": 2.8498402555910545,
"grad_norm": 0.33552722155677045,
"learning_rate": 2.689089726033922e-07,
"loss": 0.2825,
"step": 446
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.3924095079184395,
"learning_rate": 2.450672154384548e-07,
"loss": 0.2687,
"step": 447
},
{
"epoch": 2.862619808306709,
"grad_norm": 0.5138097567383261,
"learning_rate": 2.2232549983728146e-07,
"loss": 0.3004,
"step": 448
},
{
"epoch": 2.8690095846645365,
"grad_norm": 0.3283039402589176,
"learning_rate": 2.006850921586434e-07,
"loss": 0.294,
"step": 449
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.37738305697605573,
"learning_rate": 1.801471974356428e-07,
"loss": 0.2692,
"step": 450
},
{
"epoch": 2.8817891373801916,
"grad_norm": 0.3418654851564619,
"learning_rate": 1.607129593086043e-07,
"loss": 0.3001,
"step": 451
},
{
"epoch": 2.8881789137380194,
"grad_norm": 0.3402266094342756,
"learning_rate": 1.4238345996140824e-07,
"loss": 0.2911,
"step": 452
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.31824348901824345,
"learning_rate": 1.251597200612098e-07,
"loss": 0.279,
"step": 453
},
{
"epoch": 2.900958466453674,
"grad_norm": 0.34531177153096265,
"learning_rate": 1.0904269870161799e-07,
"loss": 0.3028,
"step": 454
},
{
"epoch": 2.9073482428115014,
"grad_norm": 0.3568279096820218,
"learning_rate": 9.40332933492849e-08,
"loss": 0.2927,
"step": 455
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.34527750344815883,
"learning_rate": 8.013233979393242e-08,
"loss": 0.2933,
"step": 456
},
{
"epoch": 2.9201277955271565,
"grad_norm": 0.3493059274199788,
"learning_rate": 6.734061210180499e-08,
"loss": 0.252,
"step": 457
},
{
"epoch": 2.9265175718849843,
"grad_norm": 0.370444099756165,
"learning_rate": 5.565882257257515e-08,
"loss": 0.2883,
"step": 458
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.3400590434922779,
"learning_rate": 4.508762169967318e-08,
"loss": 0.2418,
"step": 459
},
{
"epoch": 2.939297124600639,
"grad_norm": 0.3380417666194957,
"learning_rate": 3.562759813407146e-08,
"loss": 0.2928,
"step": 460
},
{
"epoch": 2.9456869009584663,
"grad_norm": 0.34811024230473875,
"learning_rate": 2.7279278651497487e-08,
"loss": 0.3067,
"step": 461
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.37158100906796243,
"learning_rate": 2.0043128123108378e-08,
"loss": 0.2683,
"step": 462
},
{
"epoch": 2.9584664536741214,
"grad_norm": 0.3508788377065571,
"learning_rate": 1.3919549489593841e-08,
"loss": 0.2738,
"step": 463
},
{
"epoch": 2.9648562300319488,
"grad_norm": 0.34321071515720997,
"learning_rate": 8.908883738754093e-09,
"loss": 0.2583,
"step": 464
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.34013241840819486,
"learning_rate": 5.011409886503949e-09,
"loss": 0.2974,
"step": 465
},
{
"epoch": 2.977635782747604,
"grad_norm": 0.3266932550661067,
"learning_rate": 2.2273449613274822e-09,
"loss": 0.264,
"step": 466
},
{
"epoch": 2.984025559105431,
"grad_norm": 0.33590879102094134,
"learning_rate": 5.568439922121194e-10,
"loss": 0.2854,
"step": 467
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.36927881731237316,
"learning_rate": 0.0,
"loss": 0.2865,
"step": 468
},
{
"epoch": 2.9904153354632586,
"step": 468,
"total_flos": 3.310734730054861e+17,
"train_loss": 0.41292818673910237,
"train_runtime": 10004.6891,
"train_samples_per_second": 2.998,
"train_steps_per_second": 0.047
}
],
"logging_steps": 1.0,
"max_steps": 468,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.310734730054861e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}