classification / trainer_state.json
devesh-2002's picture
Upload 8 files
526cc65 verified
raw
history blame
435 kB
Invalid JSON: Unexpected token 'N', ..."ad_norm": NaN, "... is not valid JSON
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2504,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00039936102236421724,
"grad_norm": NaN,
"learning_rate": 0.0001,
"loss": 3.4859,
"step": 1
},
{
"epoch": 0.0007987220447284345,
"grad_norm": NaN,
"learning_rate": 0.0001,
"loss": 2.8061,
"step": 2
},
{
"epoch": 0.0011980830670926517,
"grad_norm": Infinity,
"learning_rate": 0.0001,
"loss": 3.1149,
"step": 3
},
{
"epoch": 0.001597444089456869,
"grad_norm": Infinity,
"learning_rate": 0.0001,
"loss": 4.6655,
"step": 4
},
{
"epoch": 0.001996805111821086,
"grad_norm": Infinity,
"learning_rate": 0.0001,
"loss": 4.2207,
"step": 5
},
{
"epoch": 0.0023961661341853034,
"grad_norm": 109.66879272460938,
"learning_rate": 9.996006389776358e-05,
"loss": 3.4093,
"step": 6
},
{
"epoch": 0.0027955271565495207,
"grad_norm": 97.42423248291016,
"learning_rate": 9.992012779552716e-05,
"loss": 2.9816,
"step": 7
},
{
"epoch": 0.003194888178913738,
"grad_norm": 103.10030364990234,
"learning_rate": 9.988019169329074e-05,
"loss": 3.1748,
"step": 8
},
{
"epoch": 0.003594249201277955,
"grad_norm": 38.374732971191406,
"learning_rate": 9.984025559105432e-05,
"loss": 2.4536,
"step": 9
},
{
"epoch": 0.003993610223642172,
"grad_norm": 60.42717742919922,
"learning_rate": 9.980031948881789e-05,
"loss": 4.1217,
"step": 10
},
{
"epoch": 0.00439297124600639,
"grad_norm": 53.247196197509766,
"learning_rate": 9.976038338658148e-05,
"loss": 2.2694,
"step": 11
},
{
"epoch": 0.004792332268370607,
"grad_norm": 44.086368560791016,
"learning_rate": 9.972044728434505e-05,
"loss": 1.4825,
"step": 12
},
{
"epoch": 0.005191693290734824,
"grad_norm": 42.027191162109375,
"learning_rate": 9.968051118210863e-05,
"loss": 1.6461,
"step": 13
},
{
"epoch": 0.005591054313099041,
"grad_norm": 58.92544937133789,
"learning_rate": 9.96405750798722e-05,
"loss": 1.4784,
"step": 14
},
{
"epoch": 0.0059904153354632585,
"grad_norm": 47.151397705078125,
"learning_rate": 9.960063897763579e-05,
"loss": 2.0721,
"step": 15
},
{
"epoch": 0.006389776357827476,
"grad_norm": 57.48983383178711,
"learning_rate": 9.956070287539936e-05,
"loss": 2.1784,
"step": 16
},
{
"epoch": 0.006789137380191693,
"grad_norm": 42.118770599365234,
"learning_rate": 9.952076677316294e-05,
"loss": 0.7669,
"step": 17
},
{
"epoch": 0.00718849840255591,
"grad_norm": 36.975154876708984,
"learning_rate": 9.948083067092651e-05,
"loss": 1.8424,
"step": 18
},
{
"epoch": 0.0075878594249201275,
"grad_norm": 35.267333984375,
"learning_rate": 9.94408945686901e-05,
"loss": 1.5486,
"step": 19
},
{
"epoch": 0.007987220447284345,
"grad_norm": 66.77239990234375,
"learning_rate": 9.940095846645367e-05,
"loss": 0.9224,
"step": 20
},
{
"epoch": 0.008386581469648562,
"grad_norm": 96.58090209960938,
"learning_rate": 9.936102236421726e-05,
"loss": 1.9261,
"step": 21
},
{
"epoch": 0.00878594249201278,
"grad_norm": 55.3765754699707,
"learning_rate": 9.932108626198083e-05,
"loss": 2.8516,
"step": 22
},
{
"epoch": 0.009185303514376996,
"grad_norm": 33.431434631347656,
"learning_rate": 9.928115015974442e-05,
"loss": 0.9708,
"step": 23
},
{
"epoch": 0.009584664536741214,
"grad_norm": 59.41115951538086,
"learning_rate": 9.924121405750799e-05,
"loss": 1.681,
"step": 24
},
{
"epoch": 0.009984025559105431,
"grad_norm": 38.433231353759766,
"learning_rate": 9.920127795527157e-05,
"loss": 1.6056,
"step": 25
},
{
"epoch": 0.010383386581469648,
"grad_norm": 38.285858154296875,
"learning_rate": 9.916134185303515e-05,
"loss": 1.6935,
"step": 26
},
{
"epoch": 0.010782747603833865,
"grad_norm": 34.39680862426758,
"learning_rate": 9.912140575079873e-05,
"loss": 0.9844,
"step": 27
},
{
"epoch": 0.011182108626198083,
"grad_norm": 70.65652465820312,
"learning_rate": 9.90814696485623e-05,
"loss": 2.0065,
"step": 28
},
{
"epoch": 0.0115814696485623,
"grad_norm": 52.474952697753906,
"learning_rate": 9.904153354632588e-05,
"loss": 1.9903,
"step": 29
},
{
"epoch": 0.011980830670926517,
"grad_norm": 46.195247650146484,
"learning_rate": 9.900159744408946e-05,
"loss": 1.2172,
"step": 30
},
{
"epoch": 0.012380191693290734,
"grad_norm": 73.06243133544922,
"learning_rate": 9.896166134185304e-05,
"loss": 1.6751,
"step": 31
},
{
"epoch": 0.012779552715654952,
"grad_norm": 40.061119079589844,
"learning_rate": 9.892172523961661e-05,
"loss": 1.0806,
"step": 32
},
{
"epoch": 0.013178913738019169,
"grad_norm": 77.18601989746094,
"learning_rate": 9.88817891373802e-05,
"loss": 1.6097,
"step": 33
},
{
"epoch": 0.013578274760383386,
"grad_norm": 51.238807678222656,
"learning_rate": 9.884185303514377e-05,
"loss": 1.1296,
"step": 34
},
{
"epoch": 0.013977635782747603,
"grad_norm": 62.421268463134766,
"learning_rate": 9.880191693290736e-05,
"loss": 1.0097,
"step": 35
},
{
"epoch": 0.01437699680511182,
"grad_norm": 37.779502868652344,
"learning_rate": 9.876198083067093e-05,
"loss": 1.7957,
"step": 36
},
{
"epoch": 0.014776357827476038,
"grad_norm": 43.0893440246582,
"learning_rate": 9.87220447284345e-05,
"loss": 1.2923,
"step": 37
},
{
"epoch": 0.015175718849840255,
"grad_norm": 64.80990600585938,
"learning_rate": 9.868210862619809e-05,
"loss": 2.0914,
"step": 38
},
{
"epoch": 0.015575079872204472,
"grad_norm": 69.37845611572266,
"learning_rate": 9.864217252396166e-05,
"loss": 2.0845,
"step": 39
},
{
"epoch": 0.01597444089456869,
"grad_norm": 53.514888763427734,
"learning_rate": 9.860223642172524e-05,
"loss": 1.4393,
"step": 40
},
{
"epoch": 0.01637380191693291,
"grad_norm": 67.11495208740234,
"learning_rate": 9.856230031948882e-05,
"loss": 1.4004,
"step": 41
},
{
"epoch": 0.016773162939297124,
"grad_norm": 48.59361267089844,
"learning_rate": 9.85223642172524e-05,
"loss": 2.3624,
"step": 42
},
{
"epoch": 0.017172523961661343,
"grad_norm": 31.531797409057617,
"learning_rate": 9.848242811501598e-05,
"loss": 1.5319,
"step": 43
},
{
"epoch": 0.01757188498402556,
"grad_norm": 39.093631744384766,
"learning_rate": 9.844249201277955e-05,
"loss": 0.962,
"step": 44
},
{
"epoch": 0.017971246006389777,
"grad_norm": 19.76811408996582,
"learning_rate": 9.840255591054314e-05,
"loss": 0.2916,
"step": 45
},
{
"epoch": 0.018370607028753993,
"grad_norm": 73.54254150390625,
"learning_rate": 9.836261980830671e-05,
"loss": 1.5235,
"step": 46
},
{
"epoch": 0.018769968051118212,
"grad_norm": 50.31962966918945,
"learning_rate": 9.83226837060703e-05,
"loss": 1.1073,
"step": 47
},
{
"epoch": 0.019169329073482427,
"grad_norm": 36.65618896484375,
"learning_rate": 9.828274760383387e-05,
"loss": 1.7755,
"step": 48
},
{
"epoch": 0.019568690095846646,
"grad_norm": 28.4465389251709,
"learning_rate": 9.824281150159746e-05,
"loss": 1.3736,
"step": 49
},
{
"epoch": 0.019968051118210862,
"grad_norm": 33.03194046020508,
"learning_rate": 9.820287539936102e-05,
"loss": 0.8124,
"step": 50
},
{
"epoch": 0.02036741214057508,
"grad_norm": 39.046512603759766,
"learning_rate": 9.81629392971246e-05,
"loss": 0.8382,
"step": 51
},
{
"epoch": 0.020766773162939296,
"grad_norm": 54.7781982421875,
"learning_rate": 9.812300319488818e-05,
"loss": 2.7593,
"step": 52
},
{
"epoch": 0.021166134185303515,
"grad_norm": 49.990142822265625,
"learning_rate": 9.808306709265176e-05,
"loss": 1.3508,
"step": 53
},
{
"epoch": 0.02156549520766773,
"grad_norm": 26.498811721801758,
"learning_rate": 9.804313099041534e-05,
"loss": 0.524,
"step": 54
},
{
"epoch": 0.02196485623003195,
"grad_norm": 34.408504486083984,
"learning_rate": 9.800319488817892e-05,
"loss": 1.3278,
"step": 55
},
{
"epoch": 0.022364217252396165,
"grad_norm": 81.0146713256836,
"learning_rate": 9.79632587859425e-05,
"loss": 1.4644,
"step": 56
},
{
"epoch": 0.022763578274760384,
"grad_norm": 71.7741928100586,
"learning_rate": 9.792332268370608e-05,
"loss": 1.3252,
"step": 57
},
{
"epoch": 0.0231629392971246,
"grad_norm": 52.50370788574219,
"learning_rate": 9.788338658146965e-05,
"loss": 1.7473,
"step": 58
},
{
"epoch": 0.02356230031948882,
"grad_norm": 27.227245330810547,
"learning_rate": 9.784345047923324e-05,
"loss": 0.61,
"step": 59
},
{
"epoch": 0.023961661341853034,
"grad_norm": 39.41140365600586,
"learning_rate": 9.78035143769968e-05,
"loss": 1.3143,
"step": 60
},
{
"epoch": 0.024361022364217253,
"grad_norm": 39.11650848388672,
"learning_rate": 9.77635782747604e-05,
"loss": 0.6553,
"step": 61
},
{
"epoch": 0.02476038338658147,
"grad_norm": 103.85781860351562,
"learning_rate": 9.772364217252396e-05,
"loss": 1.7586,
"step": 62
},
{
"epoch": 0.025159744408945688,
"grad_norm": 30.291616439819336,
"learning_rate": 9.768370607028754e-05,
"loss": 1.8224,
"step": 63
},
{
"epoch": 0.025559105431309903,
"grad_norm": 38.669376373291016,
"learning_rate": 9.764376996805112e-05,
"loss": 1.2526,
"step": 64
},
{
"epoch": 0.025958466453674122,
"grad_norm": 29.225732803344727,
"learning_rate": 9.76038338658147e-05,
"loss": 0.576,
"step": 65
},
{
"epoch": 0.026357827476038338,
"grad_norm": 40.88968276977539,
"learning_rate": 9.756389776357828e-05,
"loss": 0.7861,
"step": 66
},
{
"epoch": 0.026757188498402557,
"grad_norm": 47.62251281738281,
"learning_rate": 9.752396166134186e-05,
"loss": 0.8853,
"step": 67
},
{
"epoch": 0.027156549520766772,
"grad_norm": 38.485469818115234,
"learning_rate": 9.748402555910544e-05,
"loss": 0.9984,
"step": 68
},
{
"epoch": 0.02755591054313099,
"grad_norm": 40.052818298339844,
"learning_rate": 9.744408945686902e-05,
"loss": 0.7452,
"step": 69
},
{
"epoch": 0.027955271565495207,
"grad_norm": 50.749229431152344,
"learning_rate": 9.740415335463259e-05,
"loss": 1.174,
"step": 70
},
{
"epoch": 0.028354632587859425,
"grad_norm": 57.11379623413086,
"learning_rate": 9.736421725239618e-05,
"loss": 0.9815,
"step": 71
},
{
"epoch": 0.02875399361022364,
"grad_norm": 93.78250122070312,
"learning_rate": 9.732428115015974e-05,
"loss": 1.8635,
"step": 72
},
{
"epoch": 0.02915335463258786,
"grad_norm": 38.072288513183594,
"learning_rate": 9.728434504792332e-05,
"loss": 0.9367,
"step": 73
},
{
"epoch": 0.029552715654952075,
"grad_norm": 70.73219299316406,
"learning_rate": 9.72444089456869e-05,
"loss": 1.5005,
"step": 74
},
{
"epoch": 0.029952076677316294,
"grad_norm": 28.157442092895508,
"learning_rate": 9.720447284345048e-05,
"loss": 0.5373,
"step": 75
},
{
"epoch": 0.03035143769968051,
"grad_norm": 42.3161506652832,
"learning_rate": 9.716453674121406e-05,
"loss": 1.2944,
"step": 76
},
{
"epoch": 0.03075079872204473,
"grad_norm": 25.9643497467041,
"learning_rate": 9.712460063897764e-05,
"loss": 0.7232,
"step": 77
},
{
"epoch": 0.031150159744408944,
"grad_norm": 40.28514099121094,
"learning_rate": 9.708466453674122e-05,
"loss": 1.131,
"step": 78
},
{
"epoch": 0.03154952076677316,
"grad_norm": 32.37431335449219,
"learning_rate": 9.70447284345048e-05,
"loss": 0.705,
"step": 79
},
{
"epoch": 0.03194888178913738,
"grad_norm": 56.19640350341797,
"learning_rate": 9.700479233226838e-05,
"loss": 1.5797,
"step": 80
},
{
"epoch": 0.032348242811501594,
"grad_norm": 60.04330062866211,
"learning_rate": 9.696485623003195e-05,
"loss": 1.2522,
"step": 81
},
{
"epoch": 0.03274760383386582,
"grad_norm": 39.540382385253906,
"learning_rate": 9.692492012779553e-05,
"loss": 0.8495,
"step": 82
},
{
"epoch": 0.03314696485623003,
"grad_norm": 22.74665069580078,
"learning_rate": 9.68849840255591e-05,
"loss": 0.3997,
"step": 83
},
{
"epoch": 0.03354632587859425,
"grad_norm": 61.46257400512695,
"learning_rate": 9.684504792332268e-05,
"loss": 1.1259,
"step": 84
},
{
"epoch": 0.03394568690095846,
"grad_norm": 72.10306549072266,
"learning_rate": 9.680511182108626e-05,
"loss": 1.1599,
"step": 85
},
{
"epoch": 0.034345047923322686,
"grad_norm": 70.86416625976562,
"learning_rate": 9.676517571884984e-05,
"loss": 1.1732,
"step": 86
},
{
"epoch": 0.0347444089456869,
"grad_norm": 61.40153884887695,
"learning_rate": 9.672523961661342e-05,
"loss": 1.5169,
"step": 87
},
{
"epoch": 0.03514376996805112,
"grad_norm": 62.27842330932617,
"learning_rate": 9.6685303514377e-05,
"loss": 1.0809,
"step": 88
},
{
"epoch": 0.03554313099041533,
"grad_norm": 36.0828971862793,
"learning_rate": 9.664536741214058e-05,
"loss": 0.6815,
"step": 89
},
{
"epoch": 0.035942492012779555,
"grad_norm": 30.376644134521484,
"learning_rate": 9.660543130990416e-05,
"loss": 0.5125,
"step": 90
},
{
"epoch": 0.03634185303514377,
"grad_norm": 50.5821418762207,
"learning_rate": 9.656549520766773e-05,
"loss": 0.8696,
"step": 91
},
{
"epoch": 0.036741214057507986,
"grad_norm": 49.501522064208984,
"learning_rate": 9.652555910543132e-05,
"loss": 1.0955,
"step": 92
},
{
"epoch": 0.0371405750798722,
"grad_norm": 26.80974006652832,
"learning_rate": 9.648562300319489e-05,
"loss": 0.5328,
"step": 93
},
{
"epoch": 0.037539936102236424,
"grad_norm": 48.32997512817383,
"learning_rate": 9.644568690095848e-05,
"loss": 1.3279,
"step": 94
},
{
"epoch": 0.03793929712460064,
"grad_norm": 37.20825958251953,
"learning_rate": 9.640575079872204e-05,
"loss": 0.8648,
"step": 95
},
{
"epoch": 0.038338658146964855,
"grad_norm": 27.375743865966797,
"learning_rate": 9.636581469648562e-05,
"loss": 1.0722,
"step": 96
},
{
"epoch": 0.03873801916932907,
"grad_norm": 33.725399017333984,
"learning_rate": 9.63258785942492e-05,
"loss": 0.4325,
"step": 97
},
{
"epoch": 0.03913738019169329,
"grad_norm": 59.94070053100586,
"learning_rate": 9.628594249201278e-05,
"loss": 1.0783,
"step": 98
},
{
"epoch": 0.03953674121405751,
"grad_norm": 35.72795486450195,
"learning_rate": 9.624600638977636e-05,
"loss": 1.2355,
"step": 99
},
{
"epoch": 0.039936102236421724,
"grad_norm": 37.44981002807617,
"learning_rate": 9.620607028753994e-05,
"loss": 0.819,
"step": 100
},
{
"epoch": 0.04033546325878594,
"grad_norm": 57.733917236328125,
"learning_rate": 9.616613418530351e-05,
"loss": 1.6044,
"step": 101
},
{
"epoch": 0.04073482428115016,
"grad_norm": 26.211715698242188,
"learning_rate": 9.61261980830671e-05,
"loss": 0.4462,
"step": 102
},
{
"epoch": 0.04113418530351438,
"grad_norm": 29.176176071166992,
"learning_rate": 9.608626198083067e-05,
"loss": 0.8821,
"step": 103
},
{
"epoch": 0.04153354632587859,
"grad_norm": 36.04392623901367,
"learning_rate": 9.604632587859426e-05,
"loss": 0.731,
"step": 104
},
{
"epoch": 0.041932907348242815,
"grad_norm": 25.222917556762695,
"learning_rate": 9.600638977635783e-05,
"loss": 0.6763,
"step": 105
},
{
"epoch": 0.04233226837060703,
"grad_norm": 27.276464462280273,
"learning_rate": 9.596645367412142e-05,
"loss": 0.6231,
"step": 106
},
{
"epoch": 0.042731629392971246,
"grad_norm": 32.2773323059082,
"learning_rate": 9.592651757188498e-05,
"loss": 0.5824,
"step": 107
},
{
"epoch": 0.04313099041533546,
"grad_norm": 21.84107208251953,
"learning_rate": 9.588658146964856e-05,
"loss": 1.1112,
"step": 108
},
{
"epoch": 0.043530351437699684,
"grad_norm": 40.790313720703125,
"learning_rate": 9.584664536741214e-05,
"loss": 0.6375,
"step": 109
},
{
"epoch": 0.0439297124600639,
"grad_norm": 39.70672607421875,
"learning_rate": 9.580670926517572e-05,
"loss": 0.5693,
"step": 110
},
{
"epoch": 0.044329073482428115,
"grad_norm": 38.590675354003906,
"learning_rate": 9.57667731629393e-05,
"loss": 0.5717,
"step": 111
},
{
"epoch": 0.04472843450479233,
"grad_norm": 26.32607078552246,
"learning_rate": 9.572683706070288e-05,
"loss": 0.9105,
"step": 112
},
{
"epoch": 0.04512779552715655,
"grad_norm": 45.79674530029297,
"learning_rate": 9.568690095846646e-05,
"loss": 0.9666,
"step": 113
},
{
"epoch": 0.04552715654952077,
"grad_norm": 23.451921463012695,
"learning_rate": 9.564696485623004e-05,
"loss": 0.9209,
"step": 114
},
{
"epoch": 0.045926517571884984,
"grad_norm": 21.571102142333984,
"learning_rate": 9.56070287539936e-05,
"loss": 0.8542,
"step": 115
},
{
"epoch": 0.0463258785942492,
"grad_norm": 28.323394775390625,
"learning_rate": 9.55670926517572e-05,
"loss": 0.6957,
"step": 116
},
{
"epoch": 0.04672523961661342,
"grad_norm": 47.98064422607422,
"learning_rate": 9.552715654952076e-05,
"loss": 1.1866,
"step": 117
},
{
"epoch": 0.04712460063897764,
"grad_norm": 44.72260665893555,
"learning_rate": 9.548722044728436e-05,
"loss": 0.7961,
"step": 118
},
{
"epoch": 0.04752396166134185,
"grad_norm": 49.99367904663086,
"learning_rate": 9.544728434504792e-05,
"loss": 0.6705,
"step": 119
},
{
"epoch": 0.04792332268370607,
"grad_norm": 52.86721420288086,
"learning_rate": 9.54073482428115e-05,
"loss": 1.0457,
"step": 120
},
{
"epoch": 0.04832268370607029,
"grad_norm": 40.06069564819336,
"learning_rate": 9.536741214057508e-05,
"loss": 0.5486,
"step": 121
},
{
"epoch": 0.048722044728434506,
"grad_norm": 41.819583892822266,
"learning_rate": 9.532747603833866e-05,
"loss": 0.9501,
"step": 122
},
{
"epoch": 0.04912140575079872,
"grad_norm": 42.055259704589844,
"learning_rate": 9.528753993610224e-05,
"loss": 1.8274,
"step": 123
},
{
"epoch": 0.04952076677316294,
"grad_norm": 60.77870559692383,
"learning_rate": 9.524760383386582e-05,
"loss": 1.0684,
"step": 124
},
{
"epoch": 0.04992012779552716,
"grad_norm": 39.735389709472656,
"learning_rate": 9.52076677316294e-05,
"loss": 0.7728,
"step": 125
},
{
"epoch": 0.050319488817891375,
"grad_norm": 34.534427642822266,
"learning_rate": 9.516773162939298e-05,
"loss": 1.0616,
"step": 126
},
{
"epoch": 0.05071884984025559,
"grad_norm": 45.00529479980469,
"learning_rate": 9.512779552715655e-05,
"loss": 0.9166,
"step": 127
},
{
"epoch": 0.051118210862619806,
"grad_norm": 32.73037338256836,
"learning_rate": 9.508785942492014e-05,
"loss": 0.5121,
"step": 128
},
{
"epoch": 0.05151757188498403,
"grad_norm": 55.523197174072266,
"learning_rate": 9.50479233226837e-05,
"loss": 0.8675,
"step": 129
},
{
"epoch": 0.051916932907348244,
"grad_norm": 32.37752914428711,
"learning_rate": 9.50079872204473e-05,
"loss": 0.7844,
"step": 130
},
{
"epoch": 0.05231629392971246,
"grad_norm": 50.899105072021484,
"learning_rate": 9.496805111821086e-05,
"loss": 1.0615,
"step": 131
},
{
"epoch": 0.052715654952076675,
"grad_norm": 97.9913101196289,
"learning_rate": 9.492811501597446e-05,
"loss": 2.4206,
"step": 132
},
{
"epoch": 0.0531150159744409,
"grad_norm": 18.92017936706543,
"learning_rate": 9.488817891373802e-05,
"loss": 0.3898,
"step": 133
},
{
"epoch": 0.05351437699680511,
"grad_norm": 16.557079315185547,
"learning_rate": 9.48482428115016e-05,
"loss": 0.1862,
"step": 134
},
{
"epoch": 0.05391373801916933,
"grad_norm": 30.00052833557129,
"learning_rate": 9.480830670926518e-05,
"loss": 1.1984,
"step": 135
},
{
"epoch": 0.054313099041533544,
"grad_norm": 45.07896041870117,
"learning_rate": 9.476837060702876e-05,
"loss": 1.7819,
"step": 136
},
{
"epoch": 0.054712460063897767,
"grad_norm": 31.61899185180664,
"learning_rate": 9.472843450479234e-05,
"loss": 0.5174,
"step": 137
},
{
"epoch": 0.05511182108626198,
"grad_norm": 90.64305114746094,
"learning_rate": 9.468849840255592e-05,
"loss": 1.6201,
"step": 138
},
{
"epoch": 0.0555111821086262,
"grad_norm": 31.74137306213379,
"learning_rate": 9.46485623003195e-05,
"loss": 1.368,
"step": 139
},
{
"epoch": 0.05591054313099041,
"grad_norm": 31.773054122924805,
"learning_rate": 9.460862619808308e-05,
"loss": 0.7843,
"step": 140
},
{
"epoch": 0.056309904153354635,
"grad_norm": 37.22177505493164,
"learning_rate": 9.456869009584664e-05,
"loss": 0.7204,
"step": 141
},
{
"epoch": 0.05670926517571885,
"grad_norm": 47.952701568603516,
"learning_rate": 9.452875399361024e-05,
"loss": 0.9739,
"step": 142
},
{
"epoch": 0.057108626198083066,
"grad_norm": 62.743370056152344,
"learning_rate": 9.44888178913738e-05,
"loss": 1.0897,
"step": 143
},
{
"epoch": 0.05750798722044728,
"grad_norm": 47.22040557861328,
"learning_rate": 9.444888178913738e-05,
"loss": 0.7103,
"step": 144
},
{
"epoch": 0.057907348242811504,
"grad_norm": 26.049118041992188,
"learning_rate": 9.440894568690096e-05,
"loss": 0.6913,
"step": 145
},
{
"epoch": 0.05830670926517572,
"grad_norm": 14.487537384033203,
"learning_rate": 9.436900958466454e-05,
"loss": 0.6847,
"step": 146
},
{
"epoch": 0.058706070287539935,
"grad_norm": 48.72547149658203,
"learning_rate": 9.432907348242812e-05,
"loss": 1.2914,
"step": 147
},
{
"epoch": 0.05910543130990415,
"grad_norm": 20.216861724853516,
"learning_rate": 9.42891373801917e-05,
"loss": 0.5133,
"step": 148
},
{
"epoch": 0.05950479233226837,
"grad_norm": 51.411521911621094,
"learning_rate": 9.424920127795528e-05,
"loss": 0.756,
"step": 149
},
{
"epoch": 0.05990415335463259,
"grad_norm": 19.46356773376465,
"learning_rate": 9.420926517571886e-05,
"loss": 0.636,
"step": 150
},
{
"epoch": 0.060303514376996804,
"grad_norm": 23.898357391357422,
"learning_rate": 9.416932907348244e-05,
"loss": 0.7019,
"step": 151
},
{
"epoch": 0.06070287539936102,
"grad_norm": 23.91347885131836,
"learning_rate": 9.412939297124602e-05,
"loss": 0.3579,
"step": 152
},
{
"epoch": 0.06110223642172524,
"grad_norm": 49.5450439453125,
"learning_rate": 9.408945686900958e-05,
"loss": 0.5192,
"step": 153
},
{
"epoch": 0.06150159744408946,
"grad_norm": 25.88799476623535,
"learning_rate": 9.404952076677316e-05,
"loss": 1.108,
"step": 154
},
{
"epoch": 0.06190095846645367,
"grad_norm": 15.553446769714355,
"learning_rate": 9.400958466453674e-05,
"loss": 0.329,
"step": 155
},
{
"epoch": 0.06230031948881789,
"grad_norm": 34.04236602783203,
"learning_rate": 9.396964856230032e-05,
"loss": 1.0397,
"step": 156
},
{
"epoch": 0.0626996805111821,
"grad_norm": 42.78719711303711,
"learning_rate": 9.39297124600639e-05,
"loss": 0.9486,
"step": 157
},
{
"epoch": 0.06309904153354633,
"grad_norm": 23.40775489807129,
"learning_rate": 9.388977635782748e-05,
"loss": 0.3714,
"step": 158
},
{
"epoch": 0.06349840255591055,
"grad_norm": 34.90858840942383,
"learning_rate": 9.384984025559106e-05,
"loss": 1.1231,
"step": 159
},
{
"epoch": 0.06389776357827476,
"grad_norm": 19.596099853515625,
"learning_rate": 9.380990415335464e-05,
"loss": 0.6115,
"step": 160
},
{
"epoch": 0.06429712460063898,
"grad_norm": 70.66490936279297,
"learning_rate": 9.376996805111822e-05,
"loss": 1.1846,
"step": 161
},
{
"epoch": 0.06469648562300319,
"grad_norm": 29.335140228271484,
"learning_rate": 9.373003194888178e-05,
"loss": 0.872,
"step": 162
},
{
"epoch": 0.06509584664536741,
"grad_norm": 20.113040924072266,
"learning_rate": 9.369009584664538e-05,
"loss": 0.5227,
"step": 163
},
{
"epoch": 0.06549520766773163,
"grad_norm": 32.45014953613281,
"learning_rate": 9.365015974440894e-05,
"loss": 0.3531,
"step": 164
},
{
"epoch": 0.06589456869009584,
"grad_norm": 36.6536979675293,
"learning_rate": 9.361022364217252e-05,
"loss": 0.3629,
"step": 165
},
{
"epoch": 0.06629392971246006,
"grad_norm": 69.90906524658203,
"learning_rate": 9.35702875399361e-05,
"loss": 1.0631,
"step": 166
},
{
"epoch": 0.06669329073482429,
"grad_norm": 64.00281524658203,
"learning_rate": 9.353035143769968e-05,
"loss": 1.2728,
"step": 167
},
{
"epoch": 0.0670926517571885,
"grad_norm": 40.394866943359375,
"learning_rate": 9.349041533546326e-05,
"loss": 0.6223,
"step": 168
},
{
"epoch": 0.06749201277955272,
"grad_norm": 19.014266967773438,
"learning_rate": 9.345047923322684e-05,
"loss": 0.2314,
"step": 169
},
{
"epoch": 0.06789137380191693,
"grad_norm": 32.70185089111328,
"learning_rate": 9.341054313099042e-05,
"loss": 0.9078,
"step": 170
},
{
"epoch": 0.06829073482428115,
"grad_norm": 64.28238677978516,
"learning_rate": 9.3370607028754e-05,
"loss": 1.0597,
"step": 171
},
{
"epoch": 0.06869009584664537,
"grad_norm": 13.45530891418457,
"learning_rate": 9.333067092651757e-05,
"loss": 0.3852,
"step": 172
},
{
"epoch": 0.06908945686900958,
"grad_norm": 57.906803131103516,
"learning_rate": 9.329073482428116e-05,
"loss": 1.2639,
"step": 173
},
{
"epoch": 0.0694888178913738,
"grad_norm": 23.327110290527344,
"learning_rate": 9.325079872204472e-05,
"loss": 0.8687,
"step": 174
},
{
"epoch": 0.06988817891373802,
"grad_norm": 37.181861877441406,
"learning_rate": 9.321086261980832e-05,
"loss": 0.4881,
"step": 175
},
{
"epoch": 0.07028753993610223,
"grad_norm": 29.89912223815918,
"learning_rate": 9.317092651757188e-05,
"loss": 0.6178,
"step": 176
},
{
"epoch": 0.07068690095846646,
"grad_norm": 42.71738815307617,
"learning_rate": 9.313099041533548e-05,
"loss": 0.7043,
"step": 177
},
{
"epoch": 0.07108626198083066,
"grad_norm": 39.89459991455078,
"learning_rate": 9.309105431309904e-05,
"loss": 0.8651,
"step": 178
},
{
"epoch": 0.07148562300319489,
"grad_norm": 22.305482864379883,
"learning_rate": 9.305111821086262e-05,
"loss": 0.3094,
"step": 179
},
{
"epoch": 0.07188498402555911,
"grad_norm": 52.93552017211914,
"learning_rate": 9.30111821086262e-05,
"loss": 0.919,
"step": 180
},
{
"epoch": 0.07228434504792332,
"grad_norm": 43.70652389526367,
"learning_rate": 9.297124600638978e-05,
"loss": 0.5332,
"step": 181
},
{
"epoch": 0.07268370607028754,
"grad_norm": 40.12041473388672,
"learning_rate": 9.293130990415336e-05,
"loss": 0.5661,
"step": 182
},
{
"epoch": 0.07308306709265176,
"grad_norm": 28.017053604125977,
"learning_rate": 9.289137380191694e-05,
"loss": 0.4858,
"step": 183
},
{
"epoch": 0.07348242811501597,
"grad_norm": 22.31307601928711,
"learning_rate": 9.28514376996805e-05,
"loss": 1.0273,
"step": 184
},
{
"epoch": 0.0738817891373802,
"grad_norm": 41.90707778930664,
"learning_rate": 9.28115015974441e-05,
"loss": 0.6705,
"step": 185
},
{
"epoch": 0.0742811501597444,
"grad_norm": 10.241188049316406,
"learning_rate": 9.277156549520766e-05,
"loss": 0.2117,
"step": 186
},
{
"epoch": 0.07468051118210862,
"grad_norm": 11.249414443969727,
"learning_rate": 9.273162939297126e-05,
"loss": 0.1067,
"step": 187
},
{
"epoch": 0.07507987220447285,
"grad_norm": 43.71814727783203,
"learning_rate": 9.269169329073482e-05,
"loss": 0.6949,
"step": 188
},
{
"epoch": 0.07547923322683706,
"grad_norm": 48.83254623413086,
"learning_rate": 9.265175718849842e-05,
"loss": 1.0912,
"step": 189
},
{
"epoch": 0.07587859424920128,
"grad_norm": 49.087135314941406,
"learning_rate": 9.261182108626198e-05,
"loss": 1.1075,
"step": 190
},
{
"epoch": 0.0762779552715655,
"grad_norm": 53.81547164916992,
"learning_rate": 9.257188498402556e-05,
"loss": 0.8856,
"step": 191
},
{
"epoch": 0.07667731629392971,
"grad_norm": 13.247990608215332,
"learning_rate": 9.253194888178914e-05,
"loss": 0.2104,
"step": 192
},
{
"epoch": 0.07707667731629393,
"grad_norm": 46.484886169433594,
"learning_rate": 9.249201277955272e-05,
"loss": 0.7165,
"step": 193
},
{
"epoch": 0.07747603833865814,
"grad_norm": 78.6562271118164,
"learning_rate": 9.24520766773163e-05,
"loss": 1.3969,
"step": 194
},
{
"epoch": 0.07787539936102236,
"grad_norm": 41.01932144165039,
"learning_rate": 9.241214057507988e-05,
"loss": 1.0471,
"step": 195
},
{
"epoch": 0.07827476038338659,
"grad_norm": 18.839237213134766,
"learning_rate": 9.237220447284346e-05,
"loss": 0.142,
"step": 196
},
{
"epoch": 0.0786741214057508,
"grad_norm": 33.116668701171875,
"learning_rate": 9.233226837060704e-05,
"loss": 1.4162,
"step": 197
},
{
"epoch": 0.07907348242811502,
"grad_norm": 60.7697639465332,
"learning_rate": 9.22923322683706e-05,
"loss": 1.1555,
"step": 198
},
{
"epoch": 0.07947284345047924,
"grad_norm": 11.003256797790527,
"learning_rate": 9.22523961661342e-05,
"loss": 0.1036,
"step": 199
},
{
"epoch": 0.07987220447284345,
"grad_norm": 22.006113052368164,
"learning_rate": 9.221246006389776e-05,
"loss": 0.3028,
"step": 200
},
{
"epoch": 0.08027156549520767,
"grad_norm": 30.231555938720703,
"learning_rate": 9.217252396166136e-05,
"loss": 0.8591,
"step": 201
},
{
"epoch": 0.08067092651757188,
"grad_norm": 39.03870391845703,
"learning_rate": 9.213258785942492e-05,
"loss": 1.6087,
"step": 202
},
{
"epoch": 0.0810702875399361,
"grad_norm": 22.63937759399414,
"learning_rate": 9.209265175718851e-05,
"loss": 0.3627,
"step": 203
},
{
"epoch": 0.08146964856230032,
"grad_norm": 44.56406784057617,
"learning_rate": 9.205271565495208e-05,
"loss": 0.5646,
"step": 204
},
{
"epoch": 0.08186900958466453,
"grad_norm": 24.724149703979492,
"learning_rate": 9.201277955271566e-05,
"loss": 0.57,
"step": 205
},
{
"epoch": 0.08226837060702875,
"grad_norm": 13.87296199798584,
"learning_rate": 9.197284345047924e-05,
"loss": 0.1987,
"step": 206
},
{
"epoch": 0.08266773162939298,
"grad_norm": 27.744592666625977,
"learning_rate": 9.193290734824282e-05,
"loss": 0.939,
"step": 207
},
{
"epoch": 0.08306709265175719,
"grad_norm": 18.747291564941406,
"learning_rate": 9.18929712460064e-05,
"loss": 0.5897,
"step": 208
},
{
"epoch": 0.08346645367412141,
"grad_norm": 18.909242630004883,
"learning_rate": 9.185303514376998e-05,
"loss": 0.2681,
"step": 209
},
{
"epoch": 0.08386581469648563,
"grad_norm": 28.48893928527832,
"learning_rate": 9.181309904153354e-05,
"loss": 0.7161,
"step": 210
},
{
"epoch": 0.08426517571884984,
"grad_norm": 26.746885299682617,
"learning_rate": 9.177316293929714e-05,
"loss": 0.7088,
"step": 211
},
{
"epoch": 0.08466453674121406,
"grad_norm": 92.04766082763672,
"learning_rate": 9.17332268370607e-05,
"loss": 1.5838,
"step": 212
},
{
"epoch": 0.08506389776357827,
"grad_norm": 82.4137191772461,
"learning_rate": 9.16932907348243e-05,
"loss": 1.7467,
"step": 213
},
{
"epoch": 0.08546325878594249,
"grad_norm": 37.091365814208984,
"learning_rate": 9.165335463258786e-05,
"loss": 0.434,
"step": 214
},
{
"epoch": 0.08586261980830671,
"grad_norm": 25.881351470947266,
"learning_rate": 9.161341853035144e-05,
"loss": 0.3934,
"step": 215
},
{
"epoch": 0.08626198083067092,
"grad_norm": 25.285507202148438,
"learning_rate": 9.157348242811502e-05,
"loss": 0.918,
"step": 216
},
{
"epoch": 0.08666134185303515,
"grad_norm": 39.832767486572266,
"learning_rate": 9.15335463258786e-05,
"loss": 0.7366,
"step": 217
},
{
"epoch": 0.08706070287539937,
"grad_norm": 32.17582321166992,
"learning_rate": 9.149361022364218e-05,
"loss": 0.8712,
"step": 218
},
{
"epoch": 0.08746006389776358,
"grad_norm": 46.2037239074707,
"learning_rate": 9.145367412140576e-05,
"loss": 0.9378,
"step": 219
},
{
"epoch": 0.0878594249201278,
"grad_norm": 27.639856338500977,
"learning_rate": 9.141373801916934e-05,
"loss": 0.4674,
"step": 220
},
{
"epoch": 0.08825878594249201,
"grad_norm": 47.99421691894531,
"learning_rate": 9.137380191693292e-05,
"loss": 0.9886,
"step": 221
},
{
"epoch": 0.08865814696485623,
"grad_norm": 40.40444564819336,
"learning_rate": 9.13338658146965e-05,
"loss": 0.4675,
"step": 222
},
{
"epoch": 0.08905750798722045,
"grad_norm": 46.891170501708984,
"learning_rate": 9.129392971246008e-05,
"loss": 0.497,
"step": 223
},
{
"epoch": 0.08945686900958466,
"grad_norm": 32.67228698730469,
"learning_rate": 9.125399361022364e-05,
"loss": 0.421,
"step": 224
},
{
"epoch": 0.08985623003194888,
"grad_norm": 21.464130401611328,
"learning_rate": 9.121405750798722e-05,
"loss": 0.7929,
"step": 225
},
{
"epoch": 0.0902555910543131,
"grad_norm": 57.36289596557617,
"learning_rate": 9.11741214057508e-05,
"loss": 0.8873,
"step": 226
},
{
"epoch": 0.09065495207667731,
"grad_norm": 32.8522834777832,
"learning_rate": 9.113418530351438e-05,
"loss": 0.4046,
"step": 227
},
{
"epoch": 0.09105431309904154,
"grad_norm": 14.597105026245117,
"learning_rate": 9.109424920127796e-05,
"loss": 0.5411,
"step": 228
},
{
"epoch": 0.09145367412140575,
"grad_norm": 8.493746757507324,
"learning_rate": 9.105431309904154e-05,
"loss": 0.1737,
"step": 229
},
{
"epoch": 0.09185303514376997,
"grad_norm": 29.6890811920166,
"learning_rate": 9.101437699680512e-05,
"loss": 0.8145,
"step": 230
},
{
"epoch": 0.09225239616613419,
"grad_norm": 27.84640884399414,
"learning_rate": 9.09744408945687e-05,
"loss": 0.2681,
"step": 231
},
{
"epoch": 0.0926517571884984,
"grad_norm": 9.199914932250977,
"learning_rate": 9.093450479233228e-05,
"loss": 0.1228,
"step": 232
},
{
"epoch": 0.09305111821086262,
"grad_norm": 30.594707489013672,
"learning_rate": 9.089456869009584e-05,
"loss": 0.3236,
"step": 233
},
{
"epoch": 0.09345047923322684,
"grad_norm": 36.86453628540039,
"learning_rate": 9.085463258785944e-05,
"loss": 0.4039,
"step": 234
},
{
"epoch": 0.09384984025559105,
"grad_norm": 17.261581420898438,
"learning_rate": 9.0814696485623e-05,
"loss": 0.2668,
"step": 235
},
{
"epoch": 0.09424920127795527,
"grad_norm": 41.31309509277344,
"learning_rate": 9.077476038338658e-05,
"loss": 0.5306,
"step": 236
},
{
"epoch": 0.09464856230031948,
"grad_norm": 87.06928253173828,
"learning_rate": 9.073482428115016e-05,
"loss": 1.5996,
"step": 237
},
{
"epoch": 0.0950479233226837,
"grad_norm": 23.009592056274414,
"learning_rate": 9.069488817891374e-05,
"loss": 0.1992,
"step": 238
},
{
"epoch": 0.09544728434504793,
"grad_norm": 63.086219787597656,
"learning_rate": 9.065495207667732e-05,
"loss": 0.8987,
"step": 239
},
{
"epoch": 0.09584664536741214,
"grad_norm": 61.294395446777344,
"learning_rate": 9.06150159744409e-05,
"loss": 0.865,
"step": 240
},
{
"epoch": 0.09624600638977636,
"grad_norm": 18.530488967895508,
"learning_rate": 9.057507987220448e-05,
"loss": 0.1606,
"step": 241
},
{
"epoch": 0.09664536741214058,
"grad_norm": 22.926959991455078,
"learning_rate": 9.053514376996806e-05,
"loss": 0.2651,
"step": 242
},
{
"epoch": 0.09704472843450479,
"grad_norm": 32.1219367980957,
"learning_rate": 9.049520766773162e-05,
"loss": 0.4109,
"step": 243
},
{
"epoch": 0.09744408945686901,
"grad_norm": 13.72384262084961,
"learning_rate": 9.045527156549522e-05,
"loss": 0.1638,
"step": 244
},
{
"epoch": 0.09784345047923322,
"grad_norm": 12.189730644226074,
"learning_rate": 9.041533546325878e-05,
"loss": 0.1649,
"step": 245
},
{
"epoch": 0.09824281150159744,
"grad_norm": 11.144944190979004,
"learning_rate": 9.037539936102238e-05,
"loss": 0.0851,
"step": 246
},
{
"epoch": 0.09864217252396167,
"grad_norm": 24.579763412475586,
"learning_rate": 9.033546325878594e-05,
"loss": 0.5565,
"step": 247
},
{
"epoch": 0.09904153354632587,
"grad_norm": 77.057861328125,
"learning_rate": 9.029552715654952e-05,
"loss": 1.7107,
"step": 248
},
{
"epoch": 0.0994408945686901,
"grad_norm": 6.41373348236084,
"learning_rate": 9.02555910543131e-05,
"loss": 0.0837,
"step": 249
},
{
"epoch": 0.09984025559105432,
"grad_norm": 49.9908447265625,
"learning_rate": 9.021565495207668e-05,
"loss": 0.8665,
"step": 250
},
{
"epoch": 0.10023961661341853,
"grad_norm": 80.27901458740234,
"learning_rate": 9.017571884984026e-05,
"loss": 1.7223,
"step": 251
},
{
"epoch": 0.10063897763578275,
"grad_norm": 16.287385940551758,
"learning_rate": 9.013578274760384e-05,
"loss": 0.1885,
"step": 252
},
{
"epoch": 0.10103833865814696,
"grad_norm": 21.324745178222656,
"learning_rate": 9.009584664536742e-05,
"loss": 0.1685,
"step": 253
},
{
"epoch": 0.10143769968051118,
"grad_norm": 29.087575912475586,
"learning_rate": 9.0055910543131e-05,
"loss": 1.2225,
"step": 254
},
{
"epoch": 0.1018370607028754,
"grad_norm": 10.620429039001465,
"learning_rate": 9.001597444089456e-05,
"loss": 0.0747,
"step": 255
},
{
"epoch": 0.10223642172523961,
"grad_norm": 25.556692123413086,
"learning_rate": 8.997603833865816e-05,
"loss": 0.9387,
"step": 256
},
{
"epoch": 0.10263578274760383,
"grad_norm": 24.554344177246094,
"learning_rate": 8.993610223642172e-05,
"loss": 0.3872,
"step": 257
},
{
"epoch": 0.10303514376996806,
"grad_norm": 14.509176254272461,
"learning_rate": 8.989616613418531e-05,
"loss": 0.1986,
"step": 258
},
{
"epoch": 0.10343450479233227,
"grad_norm": 34.95502853393555,
"learning_rate": 8.985623003194888e-05,
"loss": 0.9113,
"step": 259
},
{
"epoch": 0.10383386581469649,
"grad_norm": 11.8012113571167,
"learning_rate": 8.981629392971247e-05,
"loss": 0.1929,
"step": 260
},
{
"epoch": 0.1042332268370607,
"grad_norm": 78.53089904785156,
"learning_rate": 8.977635782747604e-05,
"loss": 1.181,
"step": 261
},
{
"epoch": 0.10463258785942492,
"grad_norm": 30.61515998840332,
"learning_rate": 8.973642172523962e-05,
"loss": 0.4117,
"step": 262
},
{
"epoch": 0.10503194888178914,
"grad_norm": 13.234383583068848,
"learning_rate": 8.96964856230032e-05,
"loss": 0.1099,
"step": 263
},
{
"epoch": 0.10543130990415335,
"grad_norm": 32.79818344116211,
"learning_rate": 8.965654952076678e-05,
"loss": 0.3107,
"step": 264
},
{
"epoch": 0.10583067092651757,
"grad_norm": 5.083640098571777,
"learning_rate": 8.961661341853036e-05,
"loss": 0.0646,
"step": 265
},
{
"epoch": 0.1062300319488818,
"grad_norm": 24.207582473754883,
"learning_rate": 8.957667731629394e-05,
"loss": 0.1831,
"step": 266
},
{
"epoch": 0.106629392971246,
"grad_norm": 3.5385963916778564,
"learning_rate": 8.95367412140575e-05,
"loss": 0.0741,
"step": 267
},
{
"epoch": 0.10702875399361023,
"grad_norm": 19.389137268066406,
"learning_rate": 8.94968051118211e-05,
"loss": 0.3853,
"step": 268
},
{
"epoch": 0.10742811501597443,
"grad_norm": 7.9426655769348145,
"learning_rate": 8.945686900958466e-05,
"loss": 0.0841,
"step": 269
},
{
"epoch": 0.10782747603833866,
"grad_norm": 43.5369987487793,
"learning_rate": 8.941693290734825e-05,
"loss": 0.6158,
"step": 270
},
{
"epoch": 0.10822683706070288,
"grad_norm": 37.19723892211914,
"learning_rate": 8.937699680511182e-05,
"loss": 0.2984,
"step": 271
},
{
"epoch": 0.10862619808306709,
"grad_norm": 45.05068588256836,
"learning_rate": 8.933706070287541e-05,
"loss": 0.7825,
"step": 272
},
{
"epoch": 0.10902555910543131,
"grad_norm": 7.495656967163086,
"learning_rate": 8.929712460063898e-05,
"loss": 0.08,
"step": 273
},
{
"epoch": 0.10942492012779553,
"grad_norm": 24.432415008544922,
"learning_rate": 8.925718849840256e-05,
"loss": 0.9135,
"step": 274
},
{
"epoch": 0.10982428115015974,
"grad_norm": 32.40925216674805,
"learning_rate": 8.921725239616614e-05,
"loss": 0.2541,
"step": 275
},
{
"epoch": 0.11022364217252396,
"grad_norm": 27.824214935302734,
"learning_rate": 8.917731629392972e-05,
"loss": 0.2873,
"step": 276
},
{
"epoch": 0.11062300319488817,
"grad_norm": 37.61885452270508,
"learning_rate": 8.91373801916933e-05,
"loss": 0.6622,
"step": 277
},
{
"epoch": 0.1110223642172524,
"grad_norm": 43.85084533691406,
"learning_rate": 8.909744408945688e-05,
"loss": 0.4723,
"step": 278
},
{
"epoch": 0.11142172523961662,
"grad_norm": 45.35409164428711,
"learning_rate": 8.905750798722046e-05,
"loss": 0.6808,
"step": 279
},
{
"epoch": 0.11182108626198083,
"grad_norm": 39.23952102661133,
"learning_rate": 8.901757188498404e-05,
"loss": 0.681,
"step": 280
},
{
"epoch": 0.11222044728434505,
"grad_norm": 45.037879943847656,
"learning_rate": 8.89776357827476e-05,
"loss": 0.6869,
"step": 281
},
{
"epoch": 0.11261980830670927,
"grad_norm": 45.63917922973633,
"learning_rate": 8.89376996805112e-05,
"loss": 1.048,
"step": 282
},
{
"epoch": 0.11301916932907348,
"grad_norm": 21.518823623657227,
"learning_rate": 8.889776357827476e-05,
"loss": 0.453,
"step": 283
},
{
"epoch": 0.1134185303514377,
"grad_norm": 20.787811279296875,
"learning_rate": 8.885782747603835e-05,
"loss": 0.9629,
"step": 284
},
{
"epoch": 0.11381789137380191,
"grad_norm": 40.3528938293457,
"learning_rate": 8.881789137380192e-05,
"loss": 0.9614,
"step": 285
},
{
"epoch": 0.11421725239616613,
"grad_norm": 28.17430877685547,
"learning_rate": 8.877795527156551e-05,
"loss": 0.4474,
"step": 286
},
{
"epoch": 0.11461661341853036,
"grad_norm": 54.16090393066406,
"learning_rate": 8.873801916932908e-05,
"loss": 0.9258,
"step": 287
},
{
"epoch": 0.11501597444089456,
"grad_norm": 13.837194442749023,
"learning_rate": 8.869808306709266e-05,
"loss": 0.3923,
"step": 288
},
{
"epoch": 0.11541533546325879,
"grad_norm": 24.029178619384766,
"learning_rate": 8.865814696485624e-05,
"loss": 0.3857,
"step": 289
},
{
"epoch": 0.11581469648562301,
"grad_norm": 43.0802001953125,
"learning_rate": 8.861821086261982e-05,
"loss": 1.0043,
"step": 290
},
{
"epoch": 0.11621405750798722,
"grad_norm": 21.064714431762695,
"learning_rate": 8.85782747603834e-05,
"loss": 0.4973,
"step": 291
},
{
"epoch": 0.11661341853035144,
"grad_norm": 19.474498748779297,
"learning_rate": 8.853833865814697e-05,
"loss": 0.4248,
"step": 292
},
{
"epoch": 0.11701277955271565,
"grad_norm": 43.035884857177734,
"learning_rate": 8.849840255591054e-05,
"loss": 0.786,
"step": 293
},
{
"epoch": 0.11741214057507987,
"grad_norm": 29.17285919189453,
"learning_rate": 8.845846645367413e-05,
"loss": 0.8711,
"step": 294
},
{
"epoch": 0.1178115015974441,
"grad_norm": 28.85057830810547,
"learning_rate": 8.84185303514377e-05,
"loss": 0.2502,
"step": 295
},
{
"epoch": 0.1182108626198083,
"grad_norm": 30.571170806884766,
"learning_rate": 8.837859424920128e-05,
"loss": 0.4638,
"step": 296
},
{
"epoch": 0.11861022364217252,
"grad_norm": 26.876842498779297,
"learning_rate": 8.833865814696486e-05,
"loss": 0.2626,
"step": 297
},
{
"epoch": 0.11900958466453675,
"grad_norm": 18.060771942138672,
"learning_rate": 8.829872204472844e-05,
"loss": 0.2458,
"step": 298
},
{
"epoch": 0.11940894568690096,
"grad_norm": 1.5710384845733643,
"learning_rate": 8.825878594249202e-05,
"loss": 0.0275,
"step": 299
},
{
"epoch": 0.11980830670926518,
"grad_norm": 34.21225357055664,
"learning_rate": 8.82188498402556e-05,
"loss": 0.8902,
"step": 300
},
{
"epoch": 0.12020766773162939,
"grad_norm": 11.219966888427734,
"learning_rate": 8.817891373801918e-05,
"loss": 0.1262,
"step": 301
},
{
"epoch": 0.12060702875399361,
"grad_norm": 49.54264450073242,
"learning_rate": 8.813897763578276e-05,
"loss": 0.6684,
"step": 302
},
{
"epoch": 0.12100638977635783,
"grad_norm": 38.01206588745117,
"learning_rate": 8.809904153354633e-05,
"loss": 0.6594,
"step": 303
},
{
"epoch": 0.12140575079872204,
"grad_norm": 6.748183727264404,
"learning_rate": 8.80591054313099e-05,
"loss": 0.09,
"step": 304
},
{
"epoch": 0.12180511182108626,
"grad_norm": 19.88102149963379,
"learning_rate": 8.80191693290735e-05,
"loss": 0.2535,
"step": 305
},
{
"epoch": 0.12220447284345048,
"grad_norm": 16.53110694885254,
"learning_rate": 8.797923322683706e-05,
"loss": 0.1283,
"step": 306
},
{
"epoch": 0.1226038338658147,
"grad_norm": 32.864200592041016,
"learning_rate": 8.793929712460064e-05,
"loss": 0.8136,
"step": 307
},
{
"epoch": 0.12300319488817892,
"grad_norm": 16.739694595336914,
"learning_rate": 8.789936102236422e-05,
"loss": 0.3403,
"step": 308
},
{
"epoch": 0.12340255591054312,
"grad_norm": 14.06915283203125,
"learning_rate": 8.78594249201278e-05,
"loss": 0.1108,
"step": 309
},
{
"epoch": 0.12380191693290735,
"grad_norm": 22.419809341430664,
"learning_rate": 8.781948881789138e-05,
"loss": 0.2557,
"step": 310
},
{
"epoch": 0.12420127795527157,
"grad_norm": 1.6914783716201782,
"learning_rate": 8.777955271565496e-05,
"loss": 0.0098,
"step": 311
},
{
"epoch": 0.12460063897763578,
"grad_norm": 24.4455623626709,
"learning_rate": 8.773961661341854e-05,
"loss": 0.4881,
"step": 312
},
{
"epoch": 0.125,
"grad_norm": 27.758831024169922,
"learning_rate": 8.769968051118212e-05,
"loss": 0.4601,
"step": 313
},
{
"epoch": 0.1253993610223642,
"grad_norm": 33.005638122558594,
"learning_rate": 8.765974440894568e-05,
"loss": 0.7673,
"step": 314
},
{
"epoch": 0.12579872204472844,
"grad_norm": 50.47736740112305,
"learning_rate": 8.761980830670927e-05,
"loss": 0.5843,
"step": 315
},
{
"epoch": 0.12619808306709265,
"grad_norm": 26.193117141723633,
"learning_rate": 8.757987220447284e-05,
"loss": 0.5252,
"step": 316
},
{
"epoch": 0.12659744408945686,
"grad_norm": 32.46010208129883,
"learning_rate": 8.753993610223643e-05,
"loss": 0.4976,
"step": 317
},
{
"epoch": 0.1269968051118211,
"grad_norm": 56.369178771972656,
"learning_rate": 8.75e-05,
"loss": 1.8095,
"step": 318
},
{
"epoch": 0.1273961661341853,
"grad_norm": 22.097972869873047,
"learning_rate": 8.746006389776358e-05,
"loss": 0.353,
"step": 319
},
{
"epoch": 0.12779552715654952,
"grad_norm": 53.09832763671875,
"learning_rate": 8.742012779552716e-05,
"loss": 0.7171,
"step": 320
},
{
"epoch": 0.12819488817891375,
"grad_norm": 27.46657943725586,
"learning_rate": 8.738019169329074e-05,
"loss": 0.3566,
"step": 321
},
{
"epoch": 0.12859424920127796,
"grad_norm": 34.852195739746094,
"learning_rate": 8.734025559105432e-05,
"loss": 0.4116,
"step": 322
},
{
"epoch": 0.12899361022364217,
"grad_norm": 37.44219207763672,
"learning_rate": 8.73003194888179e-05,
"loss": 0.5939,
"step": 323
},
{
"epoch": 0.12939297124600638,
"grad_norm": 51.89555358886719,
"learning_rate": 8.726038338658148e-05,
"loss": 0.6602,
"step": 324
},
{
"epoch": 0.1297923322683706,
"grad_norm": 25.05621337890625,
"learning_rate": 8.722044728434506e-05,
"loss": 0.2142,
"step": 325
},
{
"epoch": 0.13019169329073482,
"grad_norm": 24.59130096435547,
"learning_rate": 8.718051118210862e-05,
"loss": 0.854,
"step": 326
},
{
"epoch": 0.13059105431309903,
"grad_norm": 34.93574142456055,
"learning_rate": 8.714057507987221e-05,
"loss": 0.7788,
"step": 327
},
{
"epoch": 0.13099041533546327,
"grad_norm": 29.84765625,
"learning_rate": 8.710063897763578e-05,
"loss": 0.3162,
"step": 328
},
{
"epoch": 0.13138977635782748,
"grad_norm": 24.385000228881836,
"learning_rate": 8.706070287539937e-05,
"loss": 0.5139,
"step": 329
},
{
"epoch": 0.13178913738019168,
"grad_norm": 36.97303771972656,
"learning_rate": 8.702076677316294e-05,
"loss": 0.465,
"step": 330
},
{
"epoch": 0.13218849840255592,
"grad_norm": 12.05063247680664,
"learning_rate": 8.698083067092652e-05,
"loss": 0.0726,
"step": 331
},
{
"epoch": 0.13258785942492013,
"grad_norm": 26.797487258911133,
"learning_rate": 8.69408945686901e-05,
"loss": 0.3829,
"step": 332
},
{
"epoch": 0.13298722044728434,
"grad_norm": 28.903886795043945,
"learning_rate": 8.690095846645368e-05,
"loss": 0.5465,
"step": 333
},
{
"epoch": 0.13338658146964857,
"grad_norm": 28.45750617980957,
"learning_rate": 8.686102236421726e-05,
"loss": 0.5568,
"step": 334
},
{
"epoch": 0.13378594249201278,
"grad_norm": 20.86428451538086,
"learning_rate": 8.682108626198084e-05,
"loss": 0.2929,
"step": 335
},
{
"epoch": 0.134185303514377,
"grad_norm": 32.48583984375,
"learning_rate": 8.678115015974442e-05,
"loss": 0.9335,
"step": 336
},
{
"epoch": 0.13458466453674123,
"grad_norm": 50.566986083984375,
"learning_rate": 8.6741214057508e-05,
"loss": 0.837,
"step": 337
},
{
"epoch": 0.13498402555910544,
"grad_norm": 18.97594451904297,
"learning_rate": 8.670127795527156e-05,
"loss": 0.177,
"step": 338
},
{
"epoch": 0.13538338658146964,
"grad_norm": 24.457059860229492,
"learning_rate": 8.666134185303515e-05,
"loss": 0.6497,
"step": 339
},
{
"epoch": 0.13578274760383385,
"grad_norm": 47.51042175292969,
"learning_rate": 8.662140575079872e-05,
"loss": 0.9989,
"step": 340
},
{
"epoch": 0.1361821086261981,
"grad_norm": 55.363590240478516,
"learning_rate": 8.658146964856231e-05,
"loss": 0.782,
"step": 341
},
{
"epoch": 0.1365814696485623,
"grad_norm": 17.7331485748291,
"learning_rate": 8.654153354632588e-05,
"loss": 0.2987,
"step": 342
},
{
"epoch": 0.1369808306709265,
"grad_norm": 18.5281982421875,
"learning_rate": 8.650159744408947e-05,
"loss": 0.1509,
"step": 343
},
{
"epoch": 0.13738019169329074,
"grad_norm": 20.374441146850586,
"learning_rate": 8.646166134185304e-05,
"loss": 0.1878,
"step": 344
},
{
"epoch": 0.13777955271565495,
"grad_norm": 26.341569900512695,
"learning_rate": 8.642172523961662e-05,
"loss": 0.6158,
"step": 345
},
{
"epoch": 0.13817891373801916,
"grad_norm": 32.385929107666016,
"learning_rate": 8.63817891373802e-05,
"loss": 0.5448,
"step": 346
},
{
"epoch": 0.1385782747603834,
"grad_norm": 30.837196350097656,
"learning_rate": 8.634185303514378e-05,
"loss": 0.6926,
"step": 347
},
{
"epoch": 0.1389776357827476,
"grad_norm": 2.2583885192871094,
"learning_rate": 8.630191693290735e-05,
"loss": 0.014,
"step": 348
},
{
"epoch": 0.1393769968051118,
"grad_norm": 1.0454744100570679,
"learning_rate": 8.626198083067093e-05,
"loss": 0.0214,
"step": 349
},
{
"epoch": 0.13977635782747605,
"grad_norm": 20.806926727294922,
"learning_rate": 8.62220447284345e-05,
"loss": 0.6023,
"step": 350
},
{
"epoch": 0.14017571884984026,
"grad_norm": 24.15445327758789,
"learning_rate": 8.618210862619809e-05,
"loss": 0.1824,
"step": 351
},
{
"epoch": 0.14057507987220447,
"grad_norm": 14.311544418334961,
"learning_rate": 8.614217252396166e-05,
"loss": 0.1069,
"step": 352
},
{
"epoch": 0.1409744408945687,
"grad_norm": 65.05907440185547,
"learning_rate": 8.610223642172525e-05,
"loss": 0.9144,
"step": 353
},
{
"epoch": 0.1413738019169329,
"grad_norm": 18.42030143737793,
"learning_rate": 8.606230031948882e-05,
"loss": 0.123,
"step": 354
},
{
"epoch": 0.14177316293929712,
"grad_norm": 26.248994827270508,
"learning_rate": 8.602236421725241e-05,
"loss": 1.2379,
"step": 355
},
{
"epoch": 0.14217252396166133,
"grad_norm": 29.30257797241211,
"learning_rate": 8.598242811501598e-05,
"loss": 0.3357,
"step": 356
},
{
"epoch": 0.14257188498402557,
"grad_norm": 40.08498001098633,
"learning_rate": 8.594249201277956e-05,
"loss": 1.0726,
"step": 357
},
{
"epoch": 0.14297124600638977,
"grad_norm": 25.72016716003418,
"learning_rate": 8.590255591054314e-05,
"loss": 0.339,
"step": 358
},
{
"epoch": 0.14337060702875398,
"grad_norm": 21.901891708374023,
"learning_rate": 8.586261980830671e-05,
"loss": 0.4207,
"step": 359
},
{
"epoch": 0.14376996805111822,
"grad_norm": 21.876876831054688,
"learning_rate": 8.58226837060703e-05,
"loss": 0.2312,
"step": 360
},
{
"epoch": 0.14416932907348243,
"grad_norm": 28.705516815185547,
"learning_rate": 8.578274760383387e-05,
"loss": 0.2725,
"step": 361
},
{
"epoch": 0.14456869009584664,
"grad_norm": 23.67606544494629,
"learning_rate": 8.574281150159745e-05,
"loss": 0.3032,
"step": 362
},
{
"epoch": 0.14496805111821087,
"grad_norm": 56.674163818359375,
"learning_rate": 8.570287539936103e-05,
"loss": 1.2375,
"step": 363
},
{
"epoch": 0.14536741214057508,
"grad_norm": 25.14189910888672,
"learning_rate": 8.56629392971246e-05,
"loss": 0.8493,
"step": 364
},
{
"epoch": 0.1457667731629393,
"grad_norm": 8.54270076751709,
"learning_rate": 8.562300319488819e-05,
"loss": 0.0489,
"step": 365
},
{
"epoch": 0.14616613418530353,
"grad_norm": 24.299612045288086,
"learning_rate": 8.558306709265176e-05,
"loss": 0.3162,
"step": 366
},
{
"epoch": 0.14656549520766773,
"grad_norm": 31.810667037963867,
"learning_rate": 8.554313099041534e-05,
"loss": 0.6809,
"step": 367
},
{
"epoch": 0.14696485623003194,
"grad_norm": 34.37065124511719,
"learning_rate": 8.550319488817892e-05,
"loss": 0.5527,
"step": 368
},
{
"epoch": 0.14736421725239618,
"grad_norm": 37.783660888671875,
"learning_rate": 8.54632587859425e-05,
"loss": 0.6606,
"step": 369
},
{
"epoch": 0.1477635782747604,
"grad_norm": 13.607043266296387,
"learning_rate": 8.542332268370608e-05,
"loss": 0.2333,
"step": 370
},
{
"epoch": 0.1481629392971246,
"grad_norm": 14.353578567504883,
"learning_rate": 8.538338658146965e-05,
"loss": 0.0973,
"step": 371
},
{
"epoch": 0.1485623003194888,
"grad_norm": 19.128158569335938,
"learning_rate": 8.534345047923323e-05,
"loss": 0.1421,
"step": 372
},
{
"epoch": 0.14896166134185304,
"grad_norm": 38.11640167236328,
"learning_rate": 8.530351437699681e-05,
"loss": 0.6884,
"step": 373
},
{
"epoch": 0.14936102236421725,
"grad_norm": 19.969985961914062,
"learning_rate": 8.526357827476039e-05,
"loss": 0.4066,
"step": 374
},
{
"epoch": 0.14976038338658146,
"grad_norm": 38.408203125,
"learning_rate": 8.522364217252396e-05,
"loss": 1.443,
"step": 375
},
{
"epoch": 0.1501597444089457,
"grad_norm": 27.113750457763672,
"learning_rate": 8.518370607028754e-05,
"loss": 0.2402,
"step": 376
},
{
"epoch": 0.1505591054313099,
"grad_norm": 10.82494068145752,
"learning_rate": 8.514376996805112e-05,
"loss": 0.1345,
"step": 377
},
{
"epoch": 0.1509584664536741,
"grad_norm": 2.277677297592163,
"learning_rate": 8.51038338658147e-05,
"loss": 0.0459,
"step": 378
},
{
"epoch": 0.15135782747603835,
"grad_norm": 19.17725944519043,
"learning_rate": 8.506389776357828e-05,
"loss": 0.2392,
"step": 379
},
{
"epoch": 0.15175718849840256,
"grad_norm": 24.032922744750977,
"learning_rate": 8.502396166134186e-05,
"loss": 0.4149,
"step": 380
},
{
"epoch": 0.15215654952076677,
"grad_norm": 12.964612007141113,
"learning_rate": 8.498402555910544e-05,
"loss": 0.3582,
"step": 381
},
{
"epoch": 0.152555910543131,
"grad_norm": 44.06831359863281,
"learning_rate": 8.494408945686901e-05,
"loss": 0.3399,
"step": 382
},
{
"epoch": 0.1529552715654952,
"grad_norm": 21.619718551635742,
"learning_rate": 8.49041533546326e-05,
"loss": 0.1902,
"step": 383
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.2951886653900146,
"learning_rate": 8.486421725239617e-05,
"loss": 0.0369,
"step": 384
},
{
"epoch": 0.15375399361022365,
"grad_norm": 17.46122169494629,
"learning_rate": 8.482428115015974e-05,
"loss": 0.1158,
"step": 385
},
{
"epoch": 0.15415335463258786,
"grad_norm": 3.2824387550354004,
"learning_rate": 8.478434504792333e-05,
"loss": 0.0289,
"step": 386
},
{
"epoch": 0.15455271565495207,
"grad_norm": 28.71133804321289,
"learning_rate": 8.47444089456869e-05,
"loss": 0.4452,
"step": 387
},
{
"epoch": 0.15495207667731628,
"grad_norm": 20.478307723999023,
"learning_rate": 8.470447284345049e-05,
"loss": 0.1293,
"step": 388
},
{
"epoch": 0.15535143769968052,
"grad_norm": 4.031409740447998,
"learning_rate": 8.466453674121406e-05,
"loss": 0.0226,
"step": 389
},
{
"epoch": 0.15575079872204473,
"grad_norm": 40.09980010986328,
"learning_rate": 8.462460063897764e-05,
"loss": 0.3419,
"step": 390
},
{
"epoch": 0.15615015974440893,
"grad_norm": 18.734844207763672,
"learning_rate": 8.458466453674122e-05,
"loss": 0.5062,
"step": 391
},
{
"epoch": 0.15654952076677317,
"grad_norm": 33.511375427246094,
"learning_rate": 8.45447284345048e-05,
"loss": 0.6667,
"step": 392
},
{
"epoch": 0.15694888178913738,
"grad_norm": 23.9731388092041,
"learning_rate": 8.450479233226837e-05,
"loss": 0.3204,
"step": 393
},
{
"epoch": 0.1573482428115016,
"grad_norm": 21.982074737548828,
"learning_rate": 8.446485623003195e-05,
"loss": 0.6645,
"step": 394
},
{
"epoch": 0.15774760383386582,
"grad_norm": 4.137203693389893,
"learning_rate": 8.442492012779552e-05,
"loss": 0.0367,
"step": 395
},
{
"epoch": 0.15814696485623003,
"grad_norm": 27.389259338378906,
"learning_rate": 8.438498402555911e-05,
"loss": 0.5028,
"step": 396
},
{
"epoch": 0.15854632587859424,
"grad_norm": 13.590386390686035,
"learning_rate": 8.434504792332268e-05,
"loss": 0.1018,
"step": 397
},
{
"epoch": 0.15894568690095848,
"grad_norm": 9.91231918334961,
"learning_rate": 8.430511182108627e-05,
"loss": 0.1858,
"step": 398
},
{
"epoch": 0.15934504792332269,
"grad_norm": 26.712465286254883,
"learning_rate": 8.426517571884984e-05,
"loss": 0.377,
"step": 399
},
{
"epoch": 0.1597444089456869,
"grad_norm": 63.631961822509766,
"learning_rate": 8.422523961661343e-05,
"loss": 1.3597,
"step": 400
},
{
"epoch": 0.16014376996805113,
"grad_norm": 7.654021739959717,
"learning_rate": 8.4185303514377e-05,
"loss": 0.0934,
"step": 401
},
{
"epoch": 0.16054313099041534,
"grad_norm": 26.4295597076416,
"learning_rate": 8.414536741214058e-05,
"loss": 0.305,
"step": 402
},
{
"epoch": 0.16094249201277955,
"grad_norm": 18.586322784423828,
"learning_rate": 8.410543130990416e-05,
"loss": 0.1302,
"step": 403
},
{
"epoch": 0.16134185303514376,
"grad_norm": 15.485987663269043,
"learning_rate": 8.406549520766773e-05,
"loss": 0.0994,
"step": 404
},
{
"epoch": 0.161741214057508,
"grad_norm": 62.49069595336914,
"learning_rate": 8.402555910543131e-05,
"loss": 1.2294,
"step": 405
},
{
"epoch": 0.1621405750798722,
"grad_norm": 17.506107330322266,
"learning_rate": 8.39856230031949e-05,
"loss": 0.1469,
"step": 406
},
{
"epoch": 0.1625399361022364,
"grad_norm": 6.293123722076416,
"learning_rate": 8.394568690095847e-05,
"loss": 0.0483,
"step": 407
},
{
"epoch": 0.16293929712460065,
"grad_norm": 26.04009246826172,
"learning_rate": 8.390575079872205e-05,
"loss": 0.583,
"step": 408
},
{
"epoch": 0.16333865814696485,
"grad_norm": 40.334224700927734,
"learning_rate": 8.386581469648562e-05,
"loss": 0.5353,
"step": 409
},
{
"epoch": 0.16373801916932906,
"grad_norm": 29.394628524780273,
"learning_rate": 8.382587859424921e-05,
"loss": 0.2191,
"step": 410
},
{
"epoch": 0.1641373801916933,
"grad_norm": 27.580760955810547,
"learning_rate": 8.378594249201278e-05,
"loss": 0.7888,
"step": 411
},
{
"epoch": 0.1645367412140575,
"grad_norm": 72.67108917236328,
"learning_rate": 8.374600638977637e-05,
"loss": 1.735,
"step": 412
},
{
"epoch": 0.16493610223642172,
"grad_norm": 5.305516242980957,
"learning_rate": 8.370607028753994e-05,
"loss": 0.0327,
"step": 413
},
{
"epoch": 0.16533546325878595,
"grad_norm": 61.27481460571289,
"learning_rate": 8.366613418530352e-05,
"loss": 1.2937,
"step": 414
},
{
"epoch": 0.16573482428115016,
"grad_norm": 2.038816213607788,
"learning_rate": 8.36261980830671e-05,
"loss": 0.0126,
"step": 415
},
{
"epoch": 0.16613418530351437,
"grad_norm": 27.826353073120117,
"learning_rate": 8.358626198083067e-05,
"loss": 0.684,
"step": 416
},
{
"epoch": 0.1665335463258786,
"grad_norm": 48.972816467285156,
"learning_rate": 8.354632587859425e-05,
"loss": 1.7807,
"step": 417
},
{
"epoch": 0.16693290734824281,
"grad_norm": 26.85578155517578,
"learning_rate": 8.350638977635783e-05,
"loss": 0.4433,
"step": 418
},
{
"epoch": 0.16733226837060702,
"grad_norm": 28.64969253540039,
"learning_rate": 8.346645367412141e-05,
"loss": 0.7477,
"step": 419
},
{
"epoch": 0.16773162939297126,
"grad_norm": 54.42820358276367,
"learning_rate": 8.342651757188499e-05,
"loss": 1.0066,
"step": 420
},
{
"epoch": 0.16813099041533547,
"grad_norm": 27.151094436645508,
"learning_rate": 8.338658146964856e-05,
"loss": 0.5265,
"step": 421
},
{
"epoch": 0.16853035143769968,
"grad_norm": 27.435836791992188,
"learning_rate": 8.334664536741215e-05,
"loss": 0.4028,
"step": 422
},
{
"epoch": 0.16892971246006389,
"grad_norm": 27.072786331176758,
"learning_rate": 8.330670926517572e-05,
"loss": 0.7035,
"step": 423
},
{
"epoch": 0.16932907348242812,
"grad_norm": 16.292007446289062,
"learning_rate": 8.326677316293931e-05,
"loss": 0.9353,
"step": 424
},
{
"epoch": 0.16972843450479233,
"grad_norm": 23.37212562561035,
"learning_rate": 8.322683706070288e-05,
"loss": 0.2378,
"step": 425
},
{
"epoch": 0.17012779552715654,
"grad_norm": 14.183231353759766,
"learning_rate": 8.318690095846647e-05,
"loss": 0.7041,
"step": 426
},
{
"epoch": 0.17052715654952078,
"grad_norm": 18.369937896728516,
"learning_rate": 8.314696485623003e-05,
"loss": 0.1679,
"step": 427
},
{
"epoch": 0.17092651757188498,
"grad_norm": 37.56242370605469,
"learning_rate": 8.310702875399361e-05,
"loss": 0.3166,
"step": 428
},
{
"epoch": 0.1713258785942492,
"grad_norm": 11.306109428405762,
"learning_rate": 8.30670926517572e-05,
"loss": 0.4743,
"step": 429
},
{
"epoch": 0.17172523961661343,
"grad_norm": 13.266569137573242,
"learning_rate": 8.302715654952077e-05,
"loss": 0.447,
"step": 430
},
{
"epoch": 0.17212460063897764,
"grad_norm": 23.283565521240234,
"learning_rate": 8.298722044728435e-05,
"loss": 0.5382,
"step": 431
},
{
"epoch": 0.17252396166134185,
"grad_norm": 33.77288055419922,
"learning_rate": 8.294728434504793e-05,
"loss": 0.2981,
"step": 432
},
{
"epoch": 0.17292332268370608,
"grad_norm": 25.143409729003906,
"learning_rate": 8.29073482428115e-05,
"loss": 0.1996,
"step": 433
},
{
"epoch": 0.1733226837060703,
"grad_norm": 18.180683135986328,
"learning_rate": 8.286741214057509e-05,
"loss": 0.1826,
"step": 434
},
{
"epoch": 0.1737220447284345,
"grad_norm": 31.396963119506836,
"learning_rate": 8.282747603833866e-05,
"loss": 0.5022,
"step": 435
},
{
"epoch": 0.17412140575079874,
"grad_norm": 25.964462280273438,
"learning_rate": 8.278753993610225e-05,
"loss": 0.542,
"step": 436
},
{
"epoch": 0.17452076677316294,
"grad_norm": 54.64942932128906,
"learning_rate": 8.274760383386582e-05,
"loss": 1.2209,
"step": 437
},
{
"epoch": 0.17492012779552715,
"grad_norm": 18.625682830810547,
"learning_rate": 8.27076677316294e-05,
"loss": 0.3819,
"step": 438
},
{
"epoch": 0.17531948881789136,
"grad_norm": 24.026123046875,
"learning_rate": 8.266773162939297e-05,
"loss": 0.3018,
"step": 439
},
{
"epoch": 0.1757188498402556,
"grad_norm": 30.525850296020508,
"learning_rate": 8.262779552715655e-05,
"loss": 0.3266,
"step": 440
},
{
"epoch": 0.1761182108626198,
"grad_norm": 35.74408721923828,
"learning_rate": 8.258785942492013e-05,
"loss": 0.4882,
"step": 441
},
{
"epoch": 0.17651757188498401,
"grad_norm": 43.61886215209961,
"learning_rate": 8.254792332268371e-05,
"loss": 0.663,
"step": 442
},
{
"epoch": 0.17691693290734825,
"grad_norm": 29.377004623413086,
"learning_rate": 8.250798722044729e-05,
"loss": 0.5466,
"step": 443
},
{
"epoch": 0.17731629392971246,
"grad_norm": 24.392566680908203,
"learning_rate": 8.246805111821087e-05,
"loss": 0.6291,
"step": 444
},
{
"epoch": 0.17771565495207667,
"grad_norm": 23.24861717224121,
"learning_rate": 8.242811501597445e-05,
"loss": 0.162,
"step": 445
},
{
"epoch": 0.1781150159744409,
"grad_norm": 13.701735496520996,
"learning_rate": 8.238817891373803e-05,
"loss": 0.3686,
"step": 446
},
{
"epoch": 0.1785143769968051,
"grad_norm": 88.7921371459961,
"learning_rate": 8.23482428115016e-05,
"loss": 1.1888,
"step": 447
},
{
"epoch": 0.17891373801916932,
"grad_norm": 12.475123405456543,
"learning_rate": 8.230830670926518e-05,
"loss": 0.0806,
"step": 448
},
{
"epoch": 0.17931309904153356,
"grad_norm": 11.432034492492676,
"learning_rate": 8.226837060702875e-05,
"loss": 0.0839,
"step": 449
},
{
"epoch": 0.17971246006389777,
"grad_norm": 27.32512092590332,
"learning_rate": 8.222843450479233e-05,
"loss": 0.6305,
"step": 450
},
{
"epoch": 0.18011182108626198,
"grad_norm": 16.0339412689209,
"learning_rate": 8.218849840255591e-05,
"loss": 0.129,
"step": 451
},
{
"epoch": 0.1805111821086262,
"grad_norm": 38.260719299316406,
"learning_rate": 8.214856230031949e-05,
"loss": 0.4423,
"step": 452
},
{
"epoch": 0.18091054313099042,
"grad_norm": 27.91619110107422,
"learning_rate": 8.210862619808307e-05,
"loss": 0.4011,
"step": 453
},
{
"epoch": 0.18130990415335463,
"grad_norm": 15.588167190551758,
"learning_rate": 8.206869009584665e-05,
"loss": 0.3979,
"step": 454
},
{
"epoch": 0.18170926517571884,
"grad_norm": 27.18408966064453,
"learning_rate": 8.202875399361023e-05,
"loss": 0.2461,
"step": 455
},
{
"epoch": 0.18210862619808307,
"grad_norm": 3.1311802864074707,
"learning_rate": 8.19888178913738e-05,
"loss": 0.0505,
"step": 456
},
{
"epoch": 0.18250798722044728,
"grad_norm": 30.081331253051758,
"learning_rate": 8.194888178913739e-05,
"loss": 0.3859,
"step": 457
},
{
"epoch": 0.1829073482428115,
"grad_norm": 17.16853904724121,
"learning_rate": 8.190894568690096e-05,
"loss": 0.2995,
"step": 458
},
{
"epoch": 0.18330670926517573,
"grad_norm": 5.220115661621094,
"learning_rate": 8.186900958466454e-05,
"loss": 0.0301,
"step": 459
},
{
"epoch": 0.18370607028753994,
"grad_norm": 47.2158088684082,
"learning_rate": 8.182907348242811e-05,
"loss": 0.799,
"step": 460
},
{
"epoch": 0.18410543130990414,
"grad_norm": 39.21052169799805,
"learning_rate": 8.17891373801917e-05,
"loss": 0.7511,
"step": 461
},
{
"epoch": 0.18450479233226838,
"grad_norm": 20.24600601196289,
"learning_rate": 8.174920127795527e-05,
"loss": 0.3942,
"step": 462
},
{
"epoch": 0.1849041533546326,
"grad_norm": 25.908334732055664,
"learning_rate": 8.170926517571885e-05,
"loss": 0.725,
"step": 463
},
{
"epoch": 0.1853035143769968,
"grad_norm": 22.440147399902344,
"learning_rate": 8.166932907348243e-05,
"loss": 0.3098,
"step": 464
},
{
"epoch": 0.18570287539936103,
"grad_norm": 27.272539138793945,
"learning_rate": 8.162939297124601e-05,
"loss": 0.3631,
"step": 465
},
{
"epoch": 0.18610223642172524,
"grad_norm": 19.093387603759766,
"learning_rate": 8.158945686900958e-05,
"loss": 0.2873,
"step": 466
},
{
"epoch": 0.18650159744408945,
"grad_norm": 8.788177490234375,
"learning_rate": 8.154952076677317e-05,
"loss": 0.2013,
"step": 467
},
{
"epoch": 0.1869009584664537,
"grad_norm": 24.181386947631836,
"learning_rate": 8.150958466453674e-05,
"loss": 0.5086,
"step": 468
},
{
"epoch": 0.1873003194888179,
"grad_norm": 31.27284812927246,
"learning_rate": 8.146964856230033e-05,
"loss": 0.506,
"step": 469
},
{
"epoch": 0.1876996805111821,
"grad_norm": 46.10203170776367,
"learning_rate": 8.14297124600639e-05,
"loss": 0.6644,
"step": 470
},
{
"epoch": 0.1880990415335463,
"grad_norm": 19.28619384765625,
"learning_rate": 8.138977635782749e-05,
"loss": 0.1782,
"step": 471
},
{
"epoch": 0.18849840255591055,
"grad_norm": 34.94424057006836,
"learning_rate": 8.134984025559105e-05,
"loss": 0.5615,
"step": 472
},
{
"epoch": 0.18889776357827476,
"grad_norm": 8.93005084991455,
"learning_rate": 8.130990415335463e-05,
"loss": 0.1831,
"step": 473
},
{
"epoch": 0.18929712460063897,
"grad_norm": 16.664636611938477,
"learning_rate": 8.126996805111821e-05,
"loss": 0.1884,
"step": 474
},
{
"epoch": 0.1896964856230032,
"grad_norm": 33.013065338134766,
"learning_rate": 8.123003194888179e-05,
"loss": 0.6999,
"step": 475
},
{
"epoch": 0.1900958466453674,
"grad_norm": 29.782522201538086,
"learning_rate": 8.119009584664537e-05,
"loss": 0.4256,
"step": 476
},
{
"epoch": 0.19049520766773162,
"grad_norm": 13.800856590270996,
"learning_rate": 8.115015974440895e-05,
"loss": 0.4391,
"step": 477
},
{
"epoch": 0.19089456869009586,
"grad_norm": 26.930206298828125,
"learning_rate": 8.111022364217252e-05,
"loss": 0.7427,
"step": 478
},
{
"epoch": 0.19129392971246006,
"grad_norm": 19.121305465698242,
"learning_rate": 8.107028753993611e-05,
"loss": 0.2602,
"step": 479
},
{
"epoch": 0.19169329073482427,
"grad_norm": 20.222957611083984,
"learning_rate": 8.103035143769968e-05,
"loss": 0.1592,
"step": 480
},
{
"epoch": 0.1920926517571885,
"grad_norm": 6.951084613800049,
"learning_rate": 8.099041533546327e-05,
"loss": 0.0454,
"step": 481
},
{
"epoch": 0.19249201277955272,
"grad_norm": 10.007302284240723,
"learning_rate": 8.095047923322684e-05,
"loss": 0.1141,
"step": 482
},
{
"epoch": 0.19289137380191693,
"grad_norm": 38.64708709716797,
"learning_rate": 8.091054313099043e-05,
"loss": 0.8731,
"step": 483
},
{
"epoch": 0.19329073482428116,
"grad_norm": 21.756235122680664,
"learning_rate": 8.0870607028754e-05,
"loss": 0.5873,
"step": 484
},
{
"epoch": 0.19369009584664537,
"grad_norm": 10.025279998779297,
"learning_rate": 8.083067092651757e-05,
"loss": 0.0726,
"step": 485
},
{
"epoch": 0.19408945686900958,
"grad_norm": 15.280128479003906,
"learning_rate": 8.079073482428115e-05,
"loss": 0.2007,
"step": 486
},
{
"epoch": 0.1944888178913738,
"grad_norm": 4.750531196594238,
"learning_rate": 8.075079872204473e-05,
"loss": 0.0408,
"step": 487
},
{
"epoch": 0.19488817891373802,
"grad_norm": 27.46190643310547,
"learning_rate": 8.071086261980831e-05,
"loss": 0.3275,
"step": 488
},
{
"epoch": 0.19528753993610223,
"grad_norm": 43.61984634399414,
"learning_rate": 8.067092651757189e-05,
"loss": 0.6261,
"step": 489
},
{
"epoch": 0.19568690095846644,
"grad_norm": 30.51828956604004,
"learning_rate": 8.063099041533547e-05,
"loss": 0.2915,
"step": 490
},
{
"epoch": 0.19608626198083068,
"grad_norm": 23.392147064208984,
"learning_rate": 8.059105431309905e-05,
"loss": 0.5397,
"step": 491
},
{
"epoch": 0.1964856230031949,
"grad_norm": 13.042381286621094,
"learning_rate": 8.055111821086262e-05,
"loss": 0.1412,
"step": 492
},
{
"epoch": 0.1968849840255591,
"grad_norm": 47.3465461730957,
"learning_rate": 8.051118210862621e-05,
"loss": 1.0463,
"step": 493
},
{
"epoch": 0.19728434504792333,
"grad_norm": 9.875726699829102,
"learning_rate": 8.047124600638977e-05,
"loss": 0.4859,
"step": 494
},
{
"epoch": 0.19768370607028754,
"grad_norm": 14.334375381469727,
"learning_rate": 8.043130990415337e-05,
"loss": 0.5576,
"step": 495
},
{
"epoch": 0.19808306709265175,
"grad_norm": 32.280338287353516,
"learning_rate": 8.039137380191693e-05,
"loss": 0.4074,
"step": 496
},
{
"epoch": 0.19848242811501599,
"grad_norm": 20.679004669189453,
"learning_rate": 8.035143769968051e-05,
"loss": 1.1133,
"step": 497
},
{
"epoch": 0.1988817891373802,
"grad_norm": 29.116811752319336,
"learning_rate": 8.031150159744409e-05,
"loss": 0.3788,
"step": 498
},
{
"epoch": 0.1992811501597444,
"grad_norm": 29.538410186767578,
"learning_rate": 8.027156549520767e-05,
"loss": 0.9356,
"step": 499
},
{
"epoch": 0.19968051118210864,
"grad_norm": 24.485254287719727,
"learning_rate": 8.023162939297125e-05,
"loss": 0.5574,
"step": 500
},
{
"epoch": 0.20007987220447285,
"grad_norm": 20.327299118041992,
"learning_rate": 8.019169329073483e-05,
"loss": 0.3186,
"step": 501
},
{
"epoch": 0.20047923322683706,
"grad_norm": 13.071178436279297,
"learning_rate": 8.015175718849841e-05,
"loss": 0.1258,
"step": 502
},
{
"epoch": 0.20087859424920126,
"grad_norm": 5.4388909339904785,
"learning_rate": 8.011182108626199e-05,
"loss": 0.0582,
"step": 503
},
{
"epoch": 0.2012779552715655,
"grad_norm": 24.539278030395508,
"learning_rate": 8.007188498402556e-05,
"loss": 0.6277,
"step": 504
},
{
"epoch": 0.2016773162939297,
"grad_norm": 21.35367774963379,
"learning_rate": 8.003194888178915e-05,
"loss": 0.2978,
"step": 505
},
{
"epoch": 0.20207667731629392,
"grad_norm": 55.4449348449707,
"learning_rate": 7.999201277955271e-05,
"loss": 1.1812,
"step": 506
},
{
"epoch": 0.20247603833865815,
"grad_norm": 26.711278915405273,
"learning_rate": 7.995207667731631e-05,
"loss": 0.3682,
"step": 507
},
{
"epoch": 0.20287539936102236,
"grad_norm": 9.60372257232666,
"learning_rate": 7.991214057507987e-05,
"loss": 0.6127,
"step": 508
},
{
"epoch": 0.20327476038338657,
"grad_norm": 5.508291721343994,
"learning_rate": 7.987220447284345e-05,
"loss": 0.222,
"step": 509
},
{
"epoch": 0.2036741214057508,
"grad_norm": 40.95761489868164,
"learning_rate": 7.983226837060703e-05,
"loss": 0.5129,
"step": 510
},
{
"epoch": 0.20407348242811502,
"grad_norm": 42.47085189819336,
"learning_rate": 7.979233226837061e-05,
"loss": 0.8956,
"step": 511
},
{
"epoch": 0.20447284345047922,
"grad_norm": 27.247037887573242,
"learning_rate": 7.975239616613419e-05,
"loss": 0.379,
"step": 512
},
{
"epoch": 0.20487220447284346,
"grad_norm": 40.566619873046875,
"learning_rate": 7.971246006389777e-05,
"loss": 0.3145,
"step": 513
},
{
"epoch": 0.20527156549520767,
"grad_norm": 4.017962455749512,
"learning_rate": 7.967252396166135e-05,
"loss": 0.0811,
"step": 514
},
{
"epoch": 0.20567092651757188,
"grad_norm": 54.61305236816406,
"learning_rate": 7.963258785942493e-05,
"loss": 1.0618,
"step": 515
},
{
"epoch": 0.20607028753993611,
"grad_norm": 26.145051956176758,
"learning_rate": 7.959265175718851e-05,
"loss": 1.1361,
"step": 516
},
{
"epoch": 0.20646964856230032,
"grad_norm": 7.50214958190918,
"learning_rate": 7.955271565495209e-05,
"loss": 0.0511,
"step": 517
},
{
"epoch": 0.20686900958466453,
"grad_norm": 24.263626098632812,
"learning_rate": 7.951277955271565e-05,
"loss": 0.4084,
"step": 518
},
{
"epoch": 0.20726837060702874,
"grad_norm": 9.007096290588379,
"learning_rate": 7.947284345047923e-05,
"loss": 0.1599,
"step": 519
},
{
"epoch": 0.20766773162939298,
"grad_norm": 34.74939727783203,
"learning_rate": 7.943290734824281e-05,
"loss": 0.7345,
"step": 520
},
{
"epoch": 0.20806709265175719,
"grad_norm": 24.451595306396484,
"learning_rate": 7.939297124600639e-05,
"loss": 0.3272,
"step": 521
},
{
"epoch": 0.2084664536741214,
"grad_norm": 11.649699211120605,
"learning_rate": 7.935303514376997e-05,
"loss": 0.2305,
"step": 522
},
{
"epoch": 0.20886581469648563,
"grad_norm": 35.566986083984375,
"learning_rate": 7.931309904153355e-05,
"loss": 0.5498,
"step": 523
},
{
"epoch": 0.20926517571884984,
"grad_norm": 9.386147499084473,
"learning_rate": 7.927316293929713e-05,
"loss": 0.0615,
"step": 524
},
{
"epoch": 0.20966453674121405,
"grad_norm": 10.518383026123047,
"learning_rate": 7.923322683706071e-05,
"loss": 0.3447,
"step": 525
},
{
"epoch": 0.21006389776357828,
"grad_norm": 15.7105073928833,
"learning_rate": 7.919329073482429e-05,
"loss": 0.0987,
"step": 526
},
{
"epoch": 0.2104632587859425,
"grad_norm": 17.976167678833008,
"learning_rate": 7.915335463258786e-05,
"loss": 0.1674,
"step": 527
},
{
"epoch": 0.2108626198083067,
"grad_norm": 27.83255386352539,
"learning_rate": 7.911341853035145e-05,
"loss": 0.4239,
"step": 528
},
{
"epoch": 0.21126198083067094,
"grad_norm": 10.848867416381836,
"learning_rate": 7.907348242811501e-05,
"loss": 0.0698,
"step": 529
},
{
"epoch": 0.21166134185303515,
"grad_norm": 30.287899017333984,
"learning_rate": 7.90335463258786e-05,
"loss": 0.3152,
"step": 530
},
{
"epoch": 0.21206070287539935,
"grad_norm": 22.829570770263672,
"learning_rate": 7.899361022364217e-05,
"loss": 0.678,
"step": 531
},
{
"epoch": 0.2124600638977636,
"grad_norm": 32.250675201416016,
"learning_rate": 7.895367412140575e-05,
"loss": 1.0803,
"step": 532
},
{
"epoch": 0.2128594249201278,
"grad_norm": 29.182310104370117,
"learning_rate": 7.891373801916933e-05,
"loss": 0.3904,
"step": 533
},
{
"epoch": 0.213258785942492,
"grad_norm": 47.558231353759766,
"learning_rate": 7.887380191693291e-05,
"loss": 1.4202,
"step": 534
},
{
"epoch": 0.21365814696485624,
"grad_norm": 22.69037628173828,
"learning_rate": 7.883386581469649e-05,
"loss": 0.7118,
"step": 535
},
{
"epoch": 0.21405750798722045,
"grad_norm": 31.261707305908203,
"learning_rate": 7.879392971246007e-05,
"loss": 0.4615,
"step": 536
},
{
"epoch": 0.21445686900958466,
"grad_norm": 7.677802562713623,
"learning_rate": 7.875399361022364e-05,
"loss": 0.1783,
"step": 537
},
{
"epoch": 0.21485623003194887,
"grad_norm": 26.862159729003906,
"learning_rate": 7.871405750798723e-05,
"loss": 0.666,
"step": 538
},
{
"epoch": 0.2152555910543131,
"grad_norm": 22.705915451049805,
"learning_rate": 7.86741214057508e-05,
"loss": 0.2669,
"step": 539
},
{
"epoch": 0.21565495207667731,
"grad_norm": 11.319977760314941,
"learning_rate": 7.863418530351439e-05,
"loss": 0.2433,
"step": 540
},
{
"epoch": 0.21605431309904152,
"grad_norm": 38.63433837890625,
"learning_rate": 7.859424920127795e-05,
"loss": 0.5148,
"step": 541
},
{
"epoch": 0.21645367412140576,
"grad_norm": 51.30087661743164,
"learning_rate": 7.855431309904153e-05,
"loss": 0.8456,
"step": 542
},
{
"epoch": 0.21685303514376997,
"grad_norm": 10.434295654296875,
"learning_rate": 7.851437699680511e-05,
"loss": 0.2262,
"step": 543
},
{
"epoch": 0.21725239616613418,
"grad_norm": 22.641653060913086,
"learning_rate": 7.847444089456869e-05,
"loss": 0.2333,
"step": 544
},
{
"epoch": 0.2176517571884984,
"grad_norm": 8.530035972595215,
"learning_rate": 7.843450479233227e-05,
"loss": 0.0609,
"step": 545
},
{
"epoch": 0.21805111821086262,
"grad_norm": 38.13650894165039,
"learning_rate": 7.839456869009585e-05,
"loss": 0.386,
"step": 546
},
{
"epoch": 0.21845047923322683,
"grad_norm": 72.54065704345703,
"learning_rate": 7.835463258785943e-05,
"loss": 1.3895,
"step": 547
},
{
"epoch": 0.21884984025559107,
"grad_norm": 51.29950714111328,
"learning_rate": 7.831469648562301e-05,
"loss": 1.1919,
"step": 548
},
{
"epoch": 0.21924920127795527,
"grad_norm": 4.763762474060059,
"learning_rate": 7.827476038338658e-05,
"loss": 0.0375,
"step": 549
},
{
"epoch": 0.21964856230031948,
"grad_norm": 12.00156307220459,
"learning_rate": 7.823482428115017e-05,
"loss": 0.0924,
"step": 550
},
{
"epoch": 0.22004792332268372,
"grad_norm": 33.57410430908203,
"learning_rate": 7.819488817891373e-05,
"loss": 0.4545,
"step": 551
},
{
"epoch": 0.22044728434504793,
"grad_norm": 7.803763389587402,
"learning_rate": 7.815495207667733e-05,
"loss": 0.3583,
"step": 552
},
{
"epoch": 0.22084664536741214,
"grad_norm": 21.938488006591797,
"learning_rate": 7.811501597444089e-05,
"loss": 0.3672,
"step": 553
},
{
"epoch": 0.22124600638977635,
"grad_norm": 22.335405349731445,
"learning_rate": 7.807507987220449e-05,
"loss": 0.2708,
"step": 554
},
{
"epoch": 0.22164536741214058,
"grad_norm": 40.22916030883789,
"learning_rate": 7.803514376996805e-05,
"loss": 0.5552,
"step": 555
},
{
"epoch": 0.2220447284345048,
"grad_norm": 36.633914947509766,
"learning_rate": 7.799520766773163e-05,
"loss": 0.966,
"step": 556
},
{
"epoch": 0.222444089456869,
"grad_norm": 22.654062271118164,
"learning_rate": 7.795527156549521e-05,
"loss": 0.2224,
"step": 557
},
{
"epoch": 0.22284345047923323,
"grad_norm": 22.025453567504883,
"learning_rate": 7.791533546325879e-05,
"loss": 1.0247,
"step": 558
},
{
"epoch": 0.22324281150159744,
"grad_norm": 7.777953624725342,
"learning_rate": 7.787539936102237e-05,
"loss": 0.0566,
"step": 559
},
{
"epoch": 0.22364217252396165,
"grad_norm": 26.705093383789062,
"learning_rate": 7.783546325878595e-05,
"loss": 0.6082,
"step": 560
},
{
"epoch": 0.2240415335463259,
"grad_norm": 0.09895208477973938,
"learning_rate": 7.779552715654951e-05,
"loss": 0.0021,
"step": 561
},
{
"epoch": 0.2244408945686901,
"grad_norm": 16.721269607543945,
"learning_rate": 7.775559105431311e-05,
"loss": 0.3441,
"step": 562
},
{
"epoch": 0.2248402555910543,
"grad_norm": 43.88398361206055,
"learning_rate": 7.771565495207667e-05,
"loss": 0.7356,
"step": 563
},
{
"epoch": 0.22523961661341854,
"grad_norm": 1.4851919412612915,
"learning_rate": 7.767571884984027e-05,
"loss": 0.0193,
"step": 564
},
{
"epoch": 0.22563897763578275,
"grad_norm": 39.56462860107422,
"learning_rate": 7.763578274760383e-05,
"loss": 0.4662,
"step": 565
},
{
"epoch": 0.22603833865814696,
"grad_norm": 5.39362907409668,
"learning_rate": 7.759584664536743e-05,
"loss": 0.0544,
"step": 566
},
{
"epoch": 0.2264376996805112,
"grad_norm": 27.060916900634766,
"learning_rate": 7.755591054313099e-05,
"loss": 0.6322,
"step": 567
},
{
"epoch": 0.2268370607028754,
"grad_norm": 53.4051399230957,
"learning_rate": 7.751597444089457e-05,
"loss": 1.1946,
"step": 568
},
{
"epoch": 0.2272364217252396,
"grad_norm": 7.4617838859558105,
"learning_rate": 7.747603833865815e-05,
"loss": 0.0652,
"step": 569
},
{
"epoch": 0.22763578274760382,
"grad_norm": 2.287449359893799,
"learning_rate": 7.743610223642173e-05,
"loss": 0.0128,
"step": 570
},
{
"epoch": 0.22803514376996806,
"grad_norm": 8.376051902770996,
"learning_rate": 7.739616613418531e-05,
"loss": 0.0959,
"step": 571
},
{
"epoch": 0.22843450479233227,
"grad_norm": 16.224193572998047,
"learning_rate": 7.735623003194889e-05,
"loss": 0.3888,
"step": 572
},
{
"epoch": 0.22883386581469647,
"grad_norm": 53.53691482543945,
"learning_rate": 7.731629392971247e-05,
"loss": 1.6861,
"step": 573
},
{
"epoch": 0.2292332268370607,
"grad_norm": 21.933879852294922,
"learning_rate": 7.727635782747605e-05,
"loss": 0.1958,
"step": 574
},
{
"epoch": 0.22963258785942492,
"grad_norm": 18.762184143066406,
"learning_rate": 7.723642172523961e-05,
"loss": 0.1436,
"step": 575
},
{
"epoch": 0.23003194888178913,
"grad_norm": 10.35088062286377,
"learning_rate": 7.71964856230032e-05,
"loss": 0.5675,
"step": 576
},
{
"epoch": 0.23043130990415336,
"grad_norm": 24.031404495239258,
"learning_rate": 7.715654952076677e-05,
"loss": 0.2343,
"step": 577
},
{
"epoch": 0.23083067092651757,
"grad_norm": 30.824710845947266,
"learning_rate": 7.711661341853037e-05,
"loss": 1.0243,
"step": 578
},
{
"epoch": 0.23123003194888178,
"grad_norm": 27.926177978515625,
"learning_rate": 7.707667731629393e-05,
"loss": 0.5263,
"step": 579
},
{
"epoch": 0.23162939297124602,
"grad_norm": 24.238271713256836,
"learning_rate": 7.703674121405751e-05,
"loss": 0.3759,
"step": 580
},
{
"epoch": 0.23202875399361023,
"grad_norm": 25.710817337036133,
"learning_rate": 7.699680511182109e-05,
"loss": 0.2794,
"step": 581
},
{
"epoch": 0.23242811501597443,
"grad_norm": 7.0655951499938965,
"learning_rate": 7.695686900958467e-05,
"loss": 0.0432,
"step": 582
},
{
"epoch": 0.23282747603833867,
"grad_norm": 20.652233123779297,
"learning_rate": 7.691693290734825e-05,
"loss": 0.4582,
"step": 583
},
{
"epoch": 0.23322683706070288,
"grad_norm": 24.301712036132812,
"learning_rate": 7.687699680511183e-05,
"loss": 0.5028,
"step": 584
},
{
"epoch": 0.2336261980830671,
"grad_norm": 16.029043197631836,
"learning_rate": 7.683706070287541e-05,
"loss": 0.2812,
"step": 585
},
{
"epoch": 0.2340255591054313,
"grad_norm": 41.094303131103516,
"learning_rate": 7.679712460063899e-05,
"loss": 0.5828,
"step": 586
},
{
"epoch": 0.23442492012779553,
"grad_norm": 37.385894775390625,
"learning_rate": 7.675718849840255e-05,
"loss": 0.3133,
"step": 587
},
{
"epoch": 0.23482428115015974,
"grad_norm": 28.524555206298828,
"learning_rate": 7.671725239616615e-05,
"loss": 0.6478,
"step": 588
},
{
"epoch": 0.23522364217252395,
"grad_norm": 20.1973819732666,
"learning_rate": 7.667731629392971e-05,
"loss": 0.159,
"step": 589
},
{
"epoch": 0.2356230031948882,
"grad_norm": 27.267724990844727,
"learning_rate": 7.663738019169329e-05,
"loss": 1.3931,
"step": 590
},
{
"epoch": 0.2360223642172524,
"grad_norm": 53.870567321777344,
"learning_rate": 7.659744408945687e-05,
"loss": 1.2262,
"step": 591
},
{
"epoch": 0.2364217252396166,
"grad_norm": 22.93352508544922,
"learning_rate": 7.655750798722045e-05,
"loss": 0.2951,
"step": 592
},
{
"epoch": 0.23682108626198084,
"grad_norm": 25.893903732299805,
"learning_rate": 7.651757188498403e-05,
"loss": 0.5296,
"step": 593
},
{
"epoch": 0.23722044728434505,
"grad_norm": 24.8762264251709,
"learning_rate": 7.647763578274761e-05,
"loss": 0.2599,
"step": 594
},
{
"epoch": 0.23761980830670926,
"grad_norm": 1.7070298194885254,
"learning_rate": 7.643769968051119e-05,
"loss": 0.0108,
"step": 595
},
{
"epoch": 0.2380191693290735,
"grad_norm": 5.552130699157715,
"learning_rate": 7.639776357827477e-05,
"loss": 0.0417,
"step": 596
},
{
"epoch": 0.2384185303514377,
"grad_norm": 24.058055877685547,
"learning_rate": 7.635782747603835e-05,
"loss": 0.1925,
"step": 597
},
{
"epoch": 0.2388178913738019,
"grad_norm": 26.58987808227539,
"learning_rate": 7.631789137380191e-05,
"loss": 0.2317,
"step": 598
},
{
"epoch": 0.23921725239616615,
"grad_norm": 18.345590591430664,
"learning_rate": 7.62779552715655e-05,
"loss": 0.2171,
"step": 599
},
{
"epoch": 0.23961661341853036,
"grad_norm": 41.78268814086914,
"learning_rate": 7.623801916932907e-05,
"loss": 0.688,
"step": 600
},
{
"epoch": 0.24001597444089456,
"grad_norm": 6.9220805168151855,
"learning_rate": 7.619808306709265e-05,
"loss": 0.041,
"step": 601
},
{
"epoch": 0.24041533546325877,
"grad_norm": 17.554485321044922,
"learning_rate": 7.615814696485623e-05,
"loss": 0.1347,
"step": 602
},
{
"epoch": 0.240814696485623,
"grad_norm": 34.92731475830078,
"learning_rate": 7.611821086261981e-05,
"loss": 0.568,
"step": 603
},
{
"epoch": 0.24121405750798722,
"grad_norm": 23.151508331298828,
"learning_rate": 7.607827476038339e-05,
"loss": 0.5299,
"step": 604
},
{
"epoch": 0.24161341853035143,
"grad_norm": 10.962774276733398,
"learning_rate": 7.603833865814697e-05,
"loss": 0.0828,
"step": 605
},
{
"epoch": 0.24201277955271566,
"grad_norm": 1.7542988061904907,
"learning_rate": 7.599840255591055e-05,
"loss": 0.0167,
"step": 606
},
{
"epoch": 0.24241214057507987,
"grad_norm": 21.442689895629883,
"learning_rate": 7.595846645367413e-05,
"loss": 0.615,
"step": 607
},
{
"epoch": 0.24281150159744408,
"grad_norm": 41.92671585083008,
"learning_rate": 7.59185303514377e-05,
"loss": 0.5382,
"step": 608
},
{
"epoch": 0.24321086261980832,
"grad_norm": 22.601720809936523,
"learning_rate": 7.587859424920129e-05,
"loss": 0.1894,
"step": 609
},
{
"epoch": 0.24361022364217252,
"grad_norm": 5.843234539031982,
"learning_rate": 7.583865814696485e-05,
"loss": 0.0697,
"step": 610
},
{
"epoch": 0.24400958466453673,
"grad_norm": 12.49787712097168,
"learning_rate": 7.579872204472845e-05,
"loss": 0.1025,
"step": 611
},
{
"epoch": 0.24440894568690097,
"grad_norm": 17.976980209350586,
"learning_rate": 7.575878594249201e-05,
"loss": 0.3403,
"step": 612
},
{
"epoch": 0.24480830670926518,
"grad_norm": 30.415287017822266,
"learning_rate": 7.571884984025559e-05,
"loss": 0.5369,
"step": 613
},
{
"epoch": 0.2452076677316294,
"grad_norm": 31.19813346862793,
"learning_rate": 7.567891373801917e-05,
"loss": 0.2809,
"step": 614
},
{
"epoch": 0.24560702875399362,
"grad_norm": 30.63606834411621,
"learning_rate": 7.563897763578275e-05,
"loss": 0.6731,
"step": 615
},
{
"epoch": 0.24600638977635783,
"grad_norm": 16.76221466064453,
"learning_rate": 7.559904153354633e-05,
"loss": 0.157,
"step": 616
},
{
"epoch": 0.24640575079872204,
"grad_norm": 17.52460479736328,
"learning_rate": 7.555910543130991e-05,
"loss": 0.1397,
"step": 617
},
{
"epoch": 0.24680511182108625,
"grad_norm": 12.653987884521484,
"learning_rate": 7.551916932907349e-05,
"loss": 0.0763,
"step": 618
},
{
"epoch": 0.24720447284345048,
"grad_norm": 17.071046829223633,
"learning_rate": 7.547923322683707e-05,
"loss": 0.215,
"step": 619
},
{
"epoch": 0.2476038338658147,
"grad_norm": 13.926271438598633,
"learning_rate": 7.543929712460063e-05,
"loss": 0.1201,
"step": 620
},
{
"epoch": 0.2480031948881789,
"grad_norm": 43.1217041015625,
"learning_rate": 7.539936102236423e-05,
"loss": 0.3681,
"step": 621
},
{
"epoch": 0.24840255591054314,
"grad_norm": 34.11064529418945,
"learning_rate": 7.535942492012779e-05,
"loss": 0.3025,
"step": 622
},
{
"epoch": 0.24880191693290735,
"grad_norm": 45.96527862548828,
"learning_rate": 7.531948881789139e-05,
"loss": 1.019,
"step": 623
},
{
"epoch": 0.24920127795527156,
"grad_norm": 18.656265258789062,
"learning_rate": 7.527955271565495e-05,
"loss": 0.1727,
"step": 624
},
{
"epoch": 0.2496006389776358,
"grad_norm": 36.94417953491211,
"learning_rate": 7.523961661341853e-05,
"loss": 0.3873,
"step": 625
},
{
"epoch": 0.25,
"grad_norm": 3.2685413360595703,
"learning_rate": 7.519968051118211e-05,
"loss": 0.0187,
"step": 626
},
{
"epoch": 0.2503993610223642,
"grad_norm": 41.117740631103516,
"learning_rate": 7.515974440894569e-05,
"loss": 0.8145,
"step": 627
},
{
"epoch": 0.2507987220447284,
"grad_norm": 0.8575984835624695,
"learning_rate": 7.511980830670927e-05,
"loss": 0.0131,
"step": 628
},
{
"epoch": 0.2511980830670926,
"grad_norm": 1.7656009197235107,
"learning_rate": 7.507987220447285e-05,
"loss": 0.045,
"step": 629
},
{
"epoch": 0.2515974440894569,
"grad_norm": 16.53418731689453,
"learning_rate": 7.503993610223643e-05,
"loss": 0.179,
"step": 630
},
{
"epoch": 0.2519968051118211,
"grad_norm": 3.626039505004883,
"learning_rate": 7.500000000000001e-05,
"loss": 0.0256,
"step": 631
},
{
"epoch": 0.2523961661341853,
"grad_norm": 51.482669830322266,
"learning_rate": 7.496006389776357e-05,
"loss": 0.5458,
"step": 632
},
{
"epoch": 0.2527955271565495,
"grad_norm": 7.335470676422119,
"learning_rate": 7.492012779552717e-05,
"loss": 0.0433,
"step": 633
},
{
"epoch": 0.2531948881789137,
"grad_norm": 17.852930068969727,
"learning_rate": 7.488019169329073e-05,
"loss": 0.1231,
"step": 634
},
{
"epoch": 0.25359424920127793,
"grad_norm": 2.2110884189605713,
"learning_rate": 7.484025559105432e-05,
"loss": 0.0145,
"step": 635
},
{
"epoch": 0.2539936102236422,
"grad_norm": 44.878562927246094,
"learning_rate": 7.480031948881789e-05,
"loss": 1.6522,
"step": 636
},
{
"epoch": 0.2543929712460064,
"grad_norm": 19.247100830078125,
"learning_rate": 7.476038338658148e-05,
"loss": 0.6079,
"step": 637
},
{
"epoch": 0.2547923322683706,
"grad_norm": 17.943695068359375,
"learning_rate": 7.472044728434505e-05,
"loss": 0.4105,
"step": 638
},
{
"epoch": 0.2551916932907348,
"grad_norm": 4.7492289543151855,
"learning_rate": 7.468051118210863e-05,
"loss": 0.0438,
"step": 639
},
{
"epoch": 0.25559105431309903,
"grad_norm": 17.054141998291016,
"learning_rate": 7.464057507987221e-05,
"loss": 0.1531,
"step": 640
},
{
"epoch": 0.25599041533546324,
"grad_norm": 34.81682205200195,
"learning_rate": 7.460063897763579e-05,
"loss": 0.52,
"step": 641
},
{
"epoch": 0.2563897763578275,
"grad_norm": 0.6189583539962769,
"learning_rate": 7.456070287539937e-05,
"loss": 0.0032,
"step": 642
},
{
"epoch": 0.2567891373801917,
"grad_norm": 30.111778259277344,
"learning_rate": 7.452076677316295e-05,
"loss": 0.437,
"step": 643
},
{
"epoch": 0.2571884984025559,
"grad_norm": 0.9646283388137817,
"learning_rate": 7.448083067092651e-05,
"loss": 0.0059,
"step": 644
},
{
"epoch": 0.25758785942492013,
"grad_norm": 14.961357116699219,
"learning_rate": 7.44408945686901e-05,
"loss": 0.1385,
"step": 645
},
{
"epoch": 0.25798722044728434,
"grad_norm": 0.29765433073043823,
"learning_rate": 7.440095846645367e-05,
"loss": 0.0028,
"step": 646
},
{
"epoch": 0.25838658146964855,
"grad_norm": 21.831445693969727,
"learning_rate": 7.436102236421726e-05,
"loss": 0.4391,
"step": 647
},
{
"epoch": 0.25878594249201275,
"grad_norm": 38.631805419921875,
"learning_rate": 7.432108626198083e-05,
"loss": 0.651,
"step": 648
},
{
"epoch": 0.259185303514377,
"grad_norm": 35.46656036376953,
"learning_rate": 7.428115015974442e-05,
"loss": 0.7,
"step": 649
},
{
"epoch": 0.2595846645367412,
"grad_norm": 6.481640815734863,
"learning_rate": 7.424121405750799e-05,
"loss": 0.0363,
"step": 650
},
{
"epoch": 0.25998402555910544,
"grad_norm": 15.473103523254395,
"learning_rate": 7.420127795527157e-05,
"loss": 1.1334,
"step": 651
},
{
"epoch": 0.26038338658146964,
"grad_norm": 0.6283447742462158,
"learning_rate": 7.416134185303515e-05,
"loss": 0.0072,
"step": 652
},
{
"epoch": 0.26078274760383385,
"grad_norm": 12.525262832641602,
"learning_rate": 7.412140575079873e-05,
"loss": 0.0786,
"step": 653
},
{
"epoch": 0.26118210862619806,
"grad_norm": 24.875539779663086,
"learning_rate": 7.40814696485623e-05,
"loss": 0.2071,
"step": 654
},
{
"epoch": 0.2615814696485623,
"grad_norm": 0.9434861540794373,
"learning_rate": 7.404153354632589e-05,
"loss": 0.006,
"step": 655
},
{
"epoch": 0.26198083067092653,
"grad_norm": 18.7222843170166,
"learning_rate": 7.400159744408947e-05,
"loss": 0.9633,
"step": 656
},
{
"epoch": 0.26238019169329074,
"grad_norm": 45.57905578613281,
"learning_rate": 7.396166134185304e-05,
"loss": 1.1777,
"step": 657
},
{
"epoch": 0.26277955271565495,
"grad_norm": 45.27323532104492,
"learning_rate": 7.392172523961661e-05,
"loss": 0.3544,
"step": 658
},
{
"epoch": 0.26317891373801916,
"grad_norm": 29.411115646362305,
"learning_rate": 7.38817891373802e-05,
"loss": 0.6253,
"step": 659
},
{
"epoch": 0.26357827476038337,
"grad_norm": 25.291521072387695,
"learning_rate": 7.384185303514377e-05,
"loss": 0.2443,
"step": 660
},
{
"epoch": 0.2639776357827476,
"grad_norm": 1.3430376052856445,
"learning_rate": 7.380191693290735e-05,
"loss": 0.0066,
"step": 661
},
{
"epoch": 0.26437699680511184,
"grad_norm": 14.392104148864746,
"learning_rate": 7.376198083067093e-05,
"loss": 0.1737,
"step": 662
},
{
"epoch": 0.26477635782747605,
"grad_norm": 14.25992202758789,
"learning_rate": 7.372204472843451e-05,
"loss": 0.1226,
"step": 663
},
{
"epoch": 0.26517571884984026,
"grad_norm": 19.428668975830078,
"learning_rate": 7.368210862619809e-05,
"loss": 0.1848,
"step": 664
},
{
"epoch": 0.26557507987220447,
"grad_norm": 26.346725463867188,
"learning_rate": 7.364217252396167e-05,
"loss": 0.4809,
"step": 665
},
{
"epoch": 0.2659744408945687,
"grad_norm": 1.0311084985733032,
"learning_rate": 7.360223642172525e-05,
"loss": 0.0074,
"step": 666
},
{
"epoch": 0.2663738019169329,
"grad_norm": 23.361312866210938,
"learning_rate": 7.356230031948883e-05,
"loss": 0.3303,
"step": 667
},
{
"epoch": 0.26677316293929715,
"grad_norm": 9.232175827026367,
"learning_rate": 7.35223642172524e-05,
"loss": 0.0726,
"step": 668
},
{
"epoch": 0.26717252396166136,
"grad_norm": 1.495181679725647,
"learning_rate": 7.348242811501597e-05,
"loss": 0.0075,
"step": 669
},
{
"epoch": 0.26757188498402557,
"grad_norm": 0.8175781965255737,
"learning_rate": 7.344249201277955e-05,
"loss": 0.0114,
"step": 670
},
{
"epoch": 0.2679712460063898,
"grad_norm": 29.016218185424805,
"learning_rate": 7.340255591054313e-05,
"loss": 0.4842,
"step": 671
},
{
"epoch": 0.268370607028754,
"grad_norm": 5.658761501312256,
"learning_rate": 7.336261980830671e-05,
"loss": 0.0322,
"step": 672
},
{
"epoch": 0.2687699680511182,
"grad_norm": 3.9513018131256104,
"learning_rate": 7.332268370607029e-05,
"loss": 0.0197,
"step": 673
},
{
"epoch": 0.26916932907348246,
"grad_norm": 8.827807426452637,
"learning_rate": 7.328274760383387e-05,
"loss": 0.0504,
"step": 674
},
{
"epoch": 0.26956869009584666,
"grad_norm": 9.16645622253418,
"learning_rate": 7.324281150159745e-05,
"loss": 0.0524,
"step": 675
},
{
"epoch": 0.26996805111821087,
"grad_norm": 23.435909271240234,
"learning_rate": 7.320287539936103e-05,
"loss": 0.1973,
"step": 676
},
{
"epoch": 0.2703674121405751,
"grad_norm": 21.846742630004883,
"learning_rate": 7.31629392971246e-05,
"loss": 1.061,
"step": 677
},
{
"epoch": 0.2707667731629393,
"grad_norm": 30.339181900024414,
"learning_rate": 7.312300319488819e-05,
"loss": 0.7305,
"step": 678
},
{
"epoch": 0.2711661341853035,
"grad_norm": 18.960880279541016,
"learning_rate": 7.308306709265175e-05,
"loss": 0.1489,
"step": 679
},
{
"epoch": 0.2715654952076677,
"grad_norm": 9.09797477722168,
"learning_rate": 7.304313099041534e-05,
"loss": 0.0506,
"step": 680
},
{
"epoch": 0.27196485623003197,
"grad_norm": 23.953859329223633,
"learning_rate": 7.300319488817891e-05,
"loss": 0.5519,
"step": 681
},
{
"epoch": 0.2723642172523962,
"grad_norm": 32.019649505615234,
"learning_rate": 7.29632587859425e-05,
"loss": 0.3863,
"step": 682
},
{
"epoch": 0.2727635782747604,
"grad_norm": 25.83568000793457,
"learning_rate": 7.292332268370607e-05,
"loss": 0.46,
"step": 683
},
{
"epoch": 0.2731629392971246,
"grad_norm": 8.165145874023438,
"learning_rate": 7.288338658146965e-05,
"loss": 0.0548,
"step": 684
},
{
"epoch": 0.2735623003194888,
"grad_norm": 10.49545669555664,
"learning_rate": 7.284345047923323e-05,
"loss": 0.191,
"step": 685
},
{
"epoch": 0.273961661341853,
"grad_norm": 38.300132751464844,
"learning_rate": 7.280351437699681e-05,
"loss": 0.8497,
"step": 686
},
{
"epoch": 0.2743610223642173,
"grad_norm": 49.964298248291016,
"learning_rate": 7.276357827476039e-05,
"loss": 0.7205,
"step": 687
},
{
"epoch": 0.2747603833865815,
"grad_norm": 28.185935974121094,
"learning_rate": 7.272364217252397e-05,
"loss": 0.8405,
"step": 688
},
{
"epoch": 0.2751597444089457,
"grad_norm": 39.50478744506836,
"learning_rate": 7.268370607028753e-05,
"loss": 0.4983,
"step": 689
},
{
"epoch": 0.2755591054313099,
"grad_norm": 8.873536109924316,
"learning_rate": 7.264376996805113e-05,
"loss": 0.0537,
"step": 690
},
{
"epoch": 0.2759584664536741,
"grad_norm": 27.06121063232422,
"learning_rate": 7.260383386581469e-05,
"loss": 0.3861,
"step": 691
},
{
"epoch": 0.2763578274760383,
"grad_norm": 26.800996780395508,
"learning_rate": 7.256389776357828e-05,
"loss": 0.4465,
"step": 692
},
{
"epoch": 0.27675718849840253,
"grad_norm": 12.005255699157715,
"learning_rate": 7.252396166134185e-05,
"loss": 0.8693,
"step": 693
},
{
"epoch": 0.2771565495207668,
"grad_norm": 11.912272453308105,
"learning_rate": 7.248402555910544e-05,
"loss": 0.1108,
"step": 694
},
{
"epoch": 0.277555910543131,
"grad_norm": 13.568373680114746,
"learning_rate": 7.244408945686901e-05,
"loss": 0.0864,
"step": 695
},
{
"epoch": 0.2779552715654952,
"grad_norm": 17.61863899230957,
"learning_rate": 7.240415335463259e-05,
"loss": 0.1173,
"step": 696
},
{
"epoch": 0.2783546325878594,
"grad_norm": 25.594635009765625,
"learning_rate": 7.236421725239617e-05,
"loss": 0.4978,
"step": 697
},
{
"epoch": 0.2787539936102236,
"grad_norm": 99.4169692993164,
"learning_rate": 7.232428115015975e-05,
"loss": 2.2276,
"step": 698
},
{
"epoch": 0.27915335463258784,
"grad_norm": 34.09563064575195,
"learning_rate": 7.228434504792333e-05,
"loss": 0.79,
"step": 699
},
{
"epoch": 0.2795527156549521,
"grad_norm": 24.11678695678711,
"learning_rate": 7.22444089456869e-05,
"loss": 0.2965,
"step": 700
},
{
"epoch": 0.2799520766773163,
"grad_norm": 70.73790740966797,
"learning_rate": 7.220447284345049e-05,
"loss": 1.4163,
"step": 701
},
{
"epoch": 0.2803514376996805,
"grad_norm": 27.099266052246094,
"learning_rate": 7.216453674121406e-05,
"loss": 0.6492,
"step": 702
},
{
"epoch": 0.2807507987220447,
"grad_norm": 25.678932189941406,
"learning_rate": 7.212460063897763e-05,
"loss": 0.6024,
"step": 703
},
{
"epoch": 0.28115015974440893,
"grad_norm": 14.871476173400879,
"learning_rate": 7.208466453674122e-05,
"loss": 0.0994,
"step": 704
},
{
"epoch": 0.28154952076677314,
"grad_norm": 32.67781448364258,
"learning_rate": 7.204472843450479e-05,
"loss": 0.2615,
"step": 705
},
{
"epoch": 0.2819488817891374,
"grad_norm": 26.908720016479492,
"learning_rate": 7.200479233226838e-05,
"loss": 0.7124,
"step": 706
},
{
"epoch": 0.2823482428115016,
"grad_norm": 45.03997802734375,
"learning_rate": 7.196485623003195e-05,
"loss": 0.6694,
"step": 707
},
{
"epoch": 0.2827476038338658,
"grad_norm": 15.015970230102539,
"learning_rate": 7.192492012779553e-05,
"loss": 0.1835,
"step": 708
},
{
"epoch": 0.28314696485623003,
"grad_norm": 35.212764739990234,
"learning_rate": 7.188498402555911e-05,
"loss": 0.6155,
"step": 709
},
{
"epoch": 0.28354632587859424,
"grad_norm": 43.714996337890625,
"learning_rate": 7.184504792332269e-05,
"loss": 1.1994,
"step": 710
},
{
"epoch": 0.28394568690095845,
"grad_norm": 8.840459823608398,
"learning_rate": 7.180511182108627e-05,
"loss": 0.0859,
"step": 711
},
{
"epoch": 0.28434504792332266,
"grad_norm": 15.506361961364746,
"learning_rate": 7.176517571884985e-05,
"loss": 0.155,
"step": 712
},
{
"epoch": 0.2847444089456869,
"grad_norm": 13.753129959106445,
"learning_rate": 7.172523961661343e-05,
"loss": 0.1102,
"step": 713
},
{
"epoch": 0.28514376996805113,
"grad_norm": 36.447872161865234,
"learning_rate": 7.1685303514377e-05,
"loss": 0.4419,
"step": 714
},
{
"epoch": 0.28554313099041534,
"grad_norm": 12.366829872131348,
"learning_rate": 7.164536741214057e-05,
"loss": 0.813,
"step": 715
},
{
"epoch": 0.28594249201277955,
"grad_norm": 19.380538940429688,
"learning_rate": 7.160543130990416e-05,
"loss": 0.1209,
"step": 716
},
{
"epoch": 0.28634185303514376,
"grad_norm": 34.88032531738281,
"learning_rate": 7.156549520766773e-05,
"loss": 0.5646,
"step": 717
},
{
"epoch": 0.28674121405750796,
"grad_norm": 20.40744400024414,
"learning_rate": 7.152555910543132e-05,
"loss": 0.3531,
"step": 718
},
{
"epoch": 0.28714057507987223,
"grad_norm": 20.563879013061523,
"learning_rate": 7.148562300319489e-05,
"loss": 0.6965,
"step": 719
},
{
"epoch": 0.28753993610223644,
"grad_norm": 17.817777633666992,
"learning_rate": 7.144568690095848e-05,
"loss": 0.2009,
"step": 720
},
{
"epoch": 0.28793929712460065,
"grad_norm": 2.7587890625,
"learning_rate": 7.140575079872205e-05,
"loss": 0.0163,
"step": 721
},
{
"epoch": 0.28833865814696485,
"grad_norm": 11.524543762207031,
"learning_rate": 7.136581469648563e-05,
"loss": 0.2912,
"step": 722
},
{
"epoch": 0.28873801916932906,
"grad_norm": 11.787809371948242,
"learning_rate": 7.13258785942492e-05,
"loss": 0.0677,
"step": 723
},
{
"epoch": 0.28913738019169327,
"grad_norm": 29.86106300354004,
"learning_rate": 7.128594249201279e-05,
"loss": 0.234,
"step": 724
},
{
"epoch": 0.2895367412140575,
"grad_norm": 23.616262435913086,
"learning_rate": 7.124600638977636e-05,
"loss": 0.4363,
"step": 725
},
{
"epoch": 0.28993610223642174,
"grad_norm": 45.970428466796875,
"learning_rate": 7.120607028753994e-05,
"loss": 0.6776,
"step": 726
},
{
"epoch": 0.29033546325878595,
"grad_norm": 27.366239547729492,
"learning_rate": 7.116613418530351e-05,
"loss": 0.8386,
"step": 727
},
{
"epoch": 0.29073482428115016,
"grad_norm": 25.344173431396484,
"learning_rate": 7.11261980830671e-05,
"loss": 0.3375,
"step": 728
},
{
"epoch": 0.29113418530351437,
"grad_norm": 26.718671798706055,
"learning_rate": 7.108626198083067e-05,
"loss": 0.2541,
"step": 729
},
{
"epoch": 0.2915335463258786,
"grad_norm": 14.82393741607666,
"learning_rate": 7.104632587859426e-05,
"loss": 0.877,
"step": 730
},
{
"epoch": 0.2919329073482428,
"grad_norm": 7.714087009429932,
"learning_rate": 7.100638977635783e-05,
"loss": 0.0554,
"step": 731
},
{
"epoch": 0.29233226837060705,
"grad_norm": 3.5570201873779297,
"learning_rate": 7.096645367412141e-05,
"loss": 0.0434,
"step": 732
},
{
"epoch": 0.29273162939297126,
"grad_norm": 24.86942481994629,
"learning_rate": 7.092651757188499e-05,
"loss": 0.204,
"step": 733
},
{
"epoch": 0.29313099041533547,
"grad_norm": 13.815103530883789,
"learning_rate": 7.088658146964857e-05,
"loss": 0.0911,
"step": 734
},
{
"epoch": 0.2935303514376997,
"grad_norm": 30.901763916015625,
"learning_rate": 7.084664536741215e-05,
"loss": 0.6401,
"step": 735
},
{
"epoch": 0.2939297124600639,
"grad_norm": 2.162092685699463,
"learning_rate": 7.080670926517572e-05,
"loss": 0.0128,
"step": 736
},
{
"epoch": 0.2943290734824281,
"grad_norm": 23.243850708007812,
"learning_rate": 7.07667731629393e-05,
"loss": 0.3641,
"step": 737
},
{
"epoch": 0.29472843450479236,
"grad_norm": 12.715150833129883,
"learning_rate": 7.072683706070288e-05,
"loss": 0.1512,
"step": 738
},
{
"epoch": 0.29512779552715657,
"grad_norm": 3.769099235534668,
"learning_rate": 7.068690095846646e-05,
"loss": 0.0248,
"step": 739
},
{
"epoch": 0.2955271565495208,
"grad_norm": 60.900997161865234,
"learning_rate": 7.064696485623004e-05,
"loss": 0.7913,
"step": 740
},
{
"epoch": 0.295926517571885,
"grad_norm": 18.209022521972656,
"learning_rate": 7.060702875399361e-05,
"loss": 0.1983,
"step": 741
},
{
"epoch": 0.2963258785942492,
"grad_norm": 3.9102420806884766,
"learning_rate": 7.056709265175719e-05,
"loss": 0.066,
"step": 742
},
{
"epoch": 0.2967252396166134,
"grad_norm": 30.27365493774414,
"learning_rate": 7.052715654952077e-05,
"loss": 0.5253,
"step": 743
},
{
"epoch": 0.2971246006389776,
"grad_norm": 9.9003324508667,
"learning_rate": 7.048722044728435e-05,
"loss": 0.1388,
"step": 744
},
{
"epoch": 0.2975239616613419,
"grad_norm": 46.466983795166016,
"learning_rate": 7.044728434504793e-05,
"loss": 0.559,
"step": 745
},
{
"epoch": 0.2979233226837061,
"grad_norm": 18.77151870727539,
"learning_rate": 7.04073482428115e-05,
"loss": 0.1553,
"step": 746
},
{
"epoch": 0.2983226837060703,
"grad_norm": 5.184072017669678,
"learning_rate": 7.036741214057508e-05,
"loss": 0.1052,
"step": 747
},
{
"epoch": 0.2987220447284345,
"grad_norm": 51.81536102294922,
"learning_rate": 7.032747603833866e-05,
"loss": 0.8547,
"step": 748
},
{
"epoch": 0.2991214057507987,
"grad_norm": 44.04814147949219,
"learning_rate": 7.028753993610224e-05,
"loss": 0.8835,
"step": 749
},
{
"epoch": 0.2995207667731629,
"grad_norm": 6.117720603942871,
"learning_rate": 7.024760383386581e-05,
"loss": 0.12,
"step": 750
},
{
"epoch": 0.2999201277955272,
"grad_norm": 21.6702938079834,
"learning_rate": 7.02076677316294e-05,
"loss": 0.5312,
"step": 751
},
{
"epoch": 0.3003194888178914,
"grad_norm": 1.6615736484527588,
"learning_rate": 7.016773162939297e-05,
"loss": 0.0096,
"step": 752
},
{
"epoch": 0.3007188498402556,
"grad_norm": 29.80148696899414,
"learning_rate": 7.012779552715655e-05,
"loss": 0.6572,
"step": 753
},
{
"epoch": 0.3011182108626198,
"grad_norm": 23.477785110473633,
"learning_rate": 7.008785942492013e-05,
"loss": 0.1896,
"step": 754
},
{
"epoch": 0.301517571884984,
"grad_norm": 3.659334659576416,
"learning_rate": 7.00479233226837e-05,
"loss": 0.0235,
"step": 755
},
{
"epoch": 0.3019169329073482,
"grad_norm": 3.076768159866333,
"learning_rate": 7.000798722044729e-05,
"loss": 0.0307,
"step": 756
},
{
"epoch": 0.3023162939297125,
"grad_norm": 12.378639221191406,
"learning_rate": 6.996805111821087e-05,
"loss": 0.085,
"step": 757
},
{
"epoch": 0.3027156549520767,
"grad_norm": 27.674785614013672,
"learning_rate": 6.992811501597444e-05,
"loss": 0.4586,
"step": 758
},
{
"epoch": 0.3031150159744409,
"grad_norm": 16.843034744262695,
"learning_rate": 6.988817891373802e-05,
"loss": 0.1497,
"step": 759
},
{
"epoch": 0.3035143769968051,
"grad_norm": 37.000736236572266,
"learning_rate": 6.984824281150159e-05,
"loss": 1.283,
"step": 760
},
{
"epoch": 0.3039137380191693,
"grad_norm": 36.426578521728516,
"learning_rate": 6.980830670926518e-05,
"loss": 0.5122,
"step": 761
},
{
"epoch": 0.30431309904153353,
"grad_norm": 44.887718200683594,
"learning_rate": 6.976837060702875e-05,
"loss": 0.8198,
"step": 762
},
{
"epoch": 0.30471246006389774,
"grad_norm": 17.247053146362305,
"learning_rate": 6.972843450479234e-05,
"loss": 0.1203,
"step": 763
},
{
"epoch": 0.305111821086262,
"grad_norm": 13.967123985290527,
"learning_rate": 6.968849840255591e-05,
"loss": 0.0777,
"step": 764
},
{
"epoch": 0.3055111821086262,
"grad_norm": 0.6255306601524353,
"learning_rate": 6.96485623003195e-05,
"loss": 0.004,
"step": 765
},
{
"epoch": 0.3059105431309904,
"grad_norm": 1.6210542917251587,
"learning_rate": 6.960862619808307e-05,
"loss": 0.0079,
"step": 766
},
{
"epoch": 0.30630990415335463,
"grad_norm": 59.98427963256836,
"learning_rate": 6.956869009584665e-05,
"loss": 1.1066,
"step": 767
},
{
"epoch": 0.30670926517571884,
"grad_norm": 34.064205169677734,
"learning_rate": 6.952875399361023e-05,
"loss": 0.5284,
"step": 768
},
{
"epoch": 0.30710862619808305,
"grad_norm": 14.88455581665039,
"learning_rate": 6.94888178913738e-05,
"loss": 0.0916,
"step": 769
},
{
"epoch": 0.3075079872204473,
"grad_norm": 4.375553607940674,
"learning_rate": 6.944888178913738e-05,
"loss": 0.0239,
"step": 770
},
{
"epoch": 0.3079073482428115,
"grad_norm": 46.4830207824707,
"learning_rate": 6.940894568690096e-05,
"loss": 1.6677,
"step": 771
},
{
"epoch": 0.3083067092651757,
"grad_norm": 16.678213119506836,
"learning_rate": 6.936900958466453e-05,
"loss": 0.122,
"step": 772
},
{
"epoch": 0.30870607028753994,
"grad_norm": 22.385345458984375,
"learning_rate": 6.932907348242812e-05,
"loss": 0.1895,
"step": 773
},
{
"epoch": 0.30910543130990414,
"grad_norm": 54.07867431640625,
"learning_rate": 6.928913738019169e-05,
"loss": 0.8509,
"step": 774
},
{
"epoch": 0.30950479233226835,
"grad_norm": 35.7958869934082,
"learning_rate": 6.924920127795528e-05,
"loss": 0.6439,
"step": 775
},
{
"epoch": 0.30990415335463256,
"grad_norm": 41.08598327636719,
"learning_rate": 6.920926517571885e-05,
"loss": 0.879,
"step": 776
},
{
"epoch": 0.3103035143769968,
"grad_norm": 0.7453806400299072,
"learning_rate": 6.916932907348244e-05,
"loss": 0.0059,
"step": 777
},
{
"epoch": 0.31070287539936103,
"grad_norm": 21.283592224121094,
"learning_rate": 6.9129392971246e-05,
"loss": 0.2006,
"step": 778
},
{
"epoch": 0.31110223642172524,
"grad_norm": 29.745861053466797,
"learning_rate": 6.908945686900959e-05,
"loss": 0.4349,
"step": 779
},
{
"epoch": 0.31150159744408945,
"grad_norm": 20.53271484375,
"learning_rate": 6.904952076677317e-05,
"loss": 0.2545,
"step": 780
},
{
"epoch": 0.31190095846645366,
"grad_norm": 13.69855785369873,
"learning_rate": 6.900958466453674e-05,
"loss": 0.0981,
"step": 781
},
{
"epoch": 0.31230031948881787,
"grad_norm": 34.14914321899414,
"learning_rate": 6.896964856230032e-05,
"loss": 0.5867,
"step": 782
},
{
"epoch": 0.31269968051118213,
"grad_norm": 55.345645904541016,
"learning_rate": 6.89297124600639e-05,
"loss": 1.0247,
"step": 783
},
{
"epoch": 0.31309904153354634,
"grad_norm": 37.841670989990234,
"learning_rate": 6.888977635782748e-05,
"loss": 0.3203,
"step": 784
},
{
"epoch": 0.31349840255591055,
"grad_norm": 30.82229232788086,
"learning_rate": 6.884984025559106e-05,
"loss": 0.7663,
"step": 785
},
{
"epoch": 0.31389776357827476,
"grad_norm": 91.28028869628906,
"learning_rate": 6.880990415335463e-05,
"loss": 1.6352,
"step": 786
},
{
"epoch": 0.31429712460063897,
"grad_norm": 45.83837890625,
"learning_rate": 6.876996805111822e-05,
"loss": 0.9093,
"step": 787
},
{
"epoch": 0.3146964856230032,
"grad_norm": 26.87258529663086,
"learning_rate": 6.873003194888179e-05,
"loss": 0.2196,
"step": 788
},
{
"epoch": 0.31509584664536744,
"grad_norm": 30.470144271850586,
"learning_rate": 6.869009584664538e-05,
"loss": 1.2535,
"step": 789
},
{
"epoch": 0.31549520766773165,
"grad_norm": 34.989227294921875,
"learning_rate": 6.865015974440895e-05,
"loss": 0.373,
"step": 790
},
{
"epoch": 0.31589456869009586,
"grad_norm": 9.699175834655762,
"learning_rate": 6.861022364217253e-05,
"loss": 0.0592,
"step": 791
},
{
"epoch": 0.31629392971246006,
"grad_norm": 46.248863220214844,
"learning_rate": 6.85702875399361e-05,
"loss": 2.1207,
"step": 792
},
{
"epoch": 0.3166932907348243,
"grad_norm": 14.385348320007324,
"learning_rate": 6.853035143769968e-05,
"loss": 0.1276,
"step": 793
},
{
"epoch": 0.3170926517571885,
"grad_norm": 40.777610778808594,
"learning_rate": 6.849041533546326e-05,
"loss": 0.6517,
"step": 794
},
{
"epoch": 0.3174920127795527,
"grad_norm": 6.109501838684082,
"learning_rate": 6.845047923322684e-05,
"loss": 0.0405,
"step": 795
},
{
"epoch": 0.31789137380191695,
"grad_norm": 4.209787845611572,
"learning_rate": 6.841054313099042e-05,
"loss": 0.0237,
"step": 796
},
{
"epoch": 0.31829073482428116,
"grad_norm": 0.8030423521995544,
"learning_rate": 6.8370607028754e-05,
"loss": 0.005,
"step": 797
},
{
"epoch": 0.31869009584664537,
"grad_norm": 22.89716911315918,
"learning_rate": 6.833067092651757e-05,
"loss": 0.4143,
"step": 798
},
{
"epoch": 0.3190894568690096,
"grad_norm": 52.660980224609375,
"learning_rate": 6.829073482428116e-05,
"loss": 1.188,
"step": 799
},
{
"epoch": 0.3194888178913738,
"grad_norm": 24.719175338745117,
"learning_rate": 6.825079872204473e-05,
"loss": 0.4337,
"step": 800
},
{
"epoch": 0.319888178913738,
"grad_norm": 12.101390838623047,
"learning_rate": 6.821086261980832e-05,
"loss": 0.4631,
"step": 801
},
{
"epoch": 0.32028753993610226,
"grad_norm": 32.87879180908203,
"learning_rate": 6.817092651757189e-05,
"loss": 0.7483,
"step": 802
},
{
"epoch": 0.32068690095846647,
"grad_norm": 34.370155334472656,
"learning_rate": 6.813099041533546e-05,
"loss": 0.3557,
"step": 803
},
{
"epoch": 0.3210862619808307,
"grad_norm": 0.024700459092855453,
"learning_rate": 6.809105431309904e-05,
"loss": 0.0002,
"step": 804
},
{
"epoch": 0.3214856230031949,
"grad_norm": 25.349103927612305,
"learning_rate": 6.805111821086262e-05,
"loss": 0.1662,
"step": 805
},
{
"epoch": 0.3218849840255591,
"grad_norm": 13.597650527954102,
"learning_rate": 6.80111821086262e-05,
"loss": 0.197,
"step": 806
},
{
"epoch": 0.3222843450479233,
"grad_norm": 42.23225784301758,
"learning_rate": 6.797124600638978e-05,
"loss": 0.8985,
"step": 807
},
{
"epoch": 0.3226837060702875,
"grad_norm": 16.85897445678711,
"learning_rate": 6.793130990415336e-05,
"loss": 0.1926,
"step": 808
},
{
"epoch": 0.3230830670926518,
"grad_norm": 19.262500762939453,
"learning_rate": 6.789137380191694e-05,
"loss": 0.2559,
"step": 809
},
{
"epoch": 0.323482428115016,
"grad_norm": 30.815053939819336,
"learning_rate": 6.785143769968051e-05,
"loss": 0.7303,
"step": 810
},
{
"epoch": 0.3238817891373802,
"grad_norm": 15.041943550109863,
"learning_rate": 6.78115015974441e-05,
"loss": 0.1099,
"step": 811
},
{
"epoch": 0.3242811501597444,
"grad_norm": 31.67312240600586,
"learning_rate": 6.777156549520767e-05,
"loss": 0.7074,
"step": 812
},
{
"epoch": 0.3246805111821086,
"grad_norm": 46.31482696533203,
"learning_rate": 6.773162939297125e-05,
"loss": 0.7878,
"step": 813
},
{
"epoch": 0.3250798722044728,
"grad_norm": 30.529653549194336,
"learning_rate": 6.769169329073483e-05,
"loss": 0.3623,
"step": 814
},
{
"epoch": 0.3254792332268371,
"grad_norm": 43.74906921386719,
"learning_rate": 6.76517571884984e-05,
"loss": 1.1502,
"step": 815
},
{
"epoch": 0.3258785942492013,
"grad_norm": 43.4406623840332,
"learning_rate": 6.761182108626198e-05,
"loss": 0.3694,
"step": 816
},
{
"epoch": 0.3262779552715655,
"grad_norm": 28.809701919555664,
"learning_rate": 6.757188498402556e-05,
"loss": 0.5617,
"step": 817
},
{
"epoch": 0.3266773162939297,
"grad_norm": 2.9242475032806396,
"learning_rate": 6.753194888178914e-05,
"loss": 0.0298,
"step": 818
},
{
"epoch": 0.3270766773162939,
"grad_norm": 0.9818545579910278,
"learning_rate": 6.749201277955272e-05,
"loss": 0.0065,
"step": 819
},
{
"epoch": 0.3274760383386581,
"grad_norm": 20.72620391845703,
"learning_rate": 6.74520766773163e-05,
"loss": 0.2494,
"step": 820
},
{
"epoch": 0.3278753993610224,
"grad_norm": 32.65104675292969,
"learning_rate": 6.741214057507987e-05,
"loss": 0.9293,
"step": 821
},
{
"epoch": 0.3282747603833866,
"grad_norm": 27.434093475341797,
"learning_rate": 6.737220447284346e-05,
"loss": 0.668,
"step": 822
},
{
"epoch": 0.3286741214057508,
"grad_norm": 11.194473266601562,
"learning_rate": 6.733226837060703e-05,
"loss": 0.0811,
"step": 823
},
{
"epoch": 0.329073482428115,
"grad_norm": 0.05329805612564087,
"learning_rate": 6.72923322683706e-05,
"loss": 0.0012,
"step": 824
},
{
"epoch": 0.3294728434504792,
"grad_norm": 21.5410099029541,
"learning_rate": 6.725239616613419e-05,
"loss": 0.4517,
"step": 825
},
{
"epoch": 0.32987220447284343,
"grad_norm": 17.017112731933594,
"learning_rate": 6.721246006389776e-05,
"loss": 0.2488,
"step": 826
},
{
"epoch": 0.33027156549520764,
"grad_norm": 33.19927215576172,
"learning_rate": 6.717252396166134e-05,
"loss": 0.3105,
"step": 827
},
{
"epoch": 0.3306709265175719,
"grad_norm": 21.428646087646484,
"learning_rate": 6.713258785942492e-05,
"loss": 0.2835,
"step": 828
},
{
"epoch": 0.3310702875399361,
"grad_norm": 16.02735137939453,
"learning_rate": 6.70926517571885e-05,
"loss": 0.1037,
"step": 829
},
{
"epoch": 0.3314696485623003,
"grad_norm": 28.90508460998535,
"learning_rate": 6.705271565495208e-05,
"loss": 0.4812,
"step": 830
},
{
"epoch": 0.33186900958466453,
"grad_norm": 40.19026565551758,
"learning_rate": 6.701277955271565e-05,
"loss": 0.6623,
"step": 831
},
{
"epoch": 0.33226837060702874,
"grad_norm": 30.44375991821289,
"learning_rate": 6.697284345047924e-05,
"loss": 0.2675,
"step": 832
},
{
"epoch": 0.33266773162939295,
"grad_norm": 29.03067970275879,
"learning_rate": 6.693290734824281e-05,
"loss": 0.6558,
"step": 833
},
{
"epoch": 0.3330670926517572,
"grad_norm": 23.072460174560547,
"learning_rate": 6.68929712460064e-05,
"loss": 0.2401,
"step": 834
},
{
"epoch": 0.3334664536741214,
"grad_norm": 11.488616943359375,
"learning_rate": 6.685303514376997e-05,
"loss": 0.1437,
"step": 835
},
{
"epoch": 0.33386581469648563,
"grad_norm": 15.62552261352539,
"learning_rate": 6.681309904153355e-05,
"loss": 0.4715,
"step": 836
},
{
"epoch": 0.33426517571884984,
"grad_norm": 11.214410781860352,
"learning_rate": 6.677316293929712e-05,
"loss": 0.0742,
"step": 837
},
{
"epoch": 0.33466453674121405,
"grad_norm": 13.75998306274414,
"learning_rate": 6.67332268370607e-05,
"loss": 0.0883,
"step": 838
},
{
"epoch": 0.33506389776357826,
"grad_norm": 5.7025227546691895,
"learning_rate": 6.669329073482428e-05,
"loss": 0.0749,
"step": 839
},
{
"epoch": 0.3354632587859425,
"grad_norm": 29.054548263549805,
"learning_rate": 6.665335463258786e-05,
"loss": 0.3072,
"step": 840
},
{
"epoch": 0.33586261980830673,
"grad_norm": 11.108121871948242,
"learning_rate": 6.661341853035144e-05,
"loss": 0.4479,
"step": 841
},
{
"epoch": 0.33626198083067094,
"grad_norm": 22.145505905151367,
"learning_rate": 6.657348242811502e-05,
"loss": 0.4016,
"step": 842
},
{
"epoch": 0.33666134185303515,
"grad_norm": 23.753738403320312,
"learning_rate": 6.653354632587859e-05,
"loss": 0.4608,
"step": 843
},
{
"epoch": 0.33706070287539935,
"grad_norm": 15.93416690826416,
"learning_rate": 6.649361022364218e-05,
"loss": 0.1395,
"step": 844
},
{
"epoch": 0.33746006389776356,
"grad_norm": 27.130826950073242,
"learning_rate": 6.645367412140575e-05,
"loss": 0.4245,
"step": 845
},
{
"epoch": 0.33785942492012777,
"grad_norm": 24.687938690185547,
"learning_rate": 6.641373801916934e-05,
"loss": 0.4287,
"step": 846
},
{
"epoch": 0.33825878594249204,
"grad_norm": 26.228866577148438,
"learning_rate": 6.63738019169329e-05,
"loss": 0.6083,
"step": 847
},
{
"epoch": 0.33865814696485624,
"grad_norm": 40.05247116088867,
"learning_rate": 6.63338658146965e-05,
"loss": 0.6715,
"step": 848
},
{
"epoch": 0.33905750798722045,
"grad_norm": 6.516374588012695,
"learning_rate": 6.629392971246006e-05,
"loss": 0.082,
"step": 849
},
{
"epoch": 0.33945686900958466,
"grad_norm": 23.915313720703125,
"learning_rate": 6.625399361022364e-05,
"loss": 0.3277,
"step": 850
},
{
"epoch": 0.33985623003194887,
"grad_norm": 27.337392807006836,
"learning_rate": 6.621405750798722e-05,
"loss": 0.5381,
"step": 851
},
{
"epoch": 0.3402555910543131,
"grad_norm": 13.057764053344727,
"learning_rate": 6.61741214057508e-05,
"loss": 0.2238,
"step": 852
},
{
"epoch": 0.34065495207667734,
"grad_norm": 24.886032104492188,
"learning_rate": 6.613418530351438e-05,
"loss": 0.3278,
"step": 853
},
{
"epoch": 0.34105431309904155,
"grad_norm": 22.214048385620117,
"learning_rate": 6.609424920127796e-05,
"loss": 0.2234,
"step": 854
},
{
"epoch": 0.34145367412140576,
"grad_norm": 28.066293716430664,
"learning_rate": 6.605431309904153e-05,
"loss": 0.5266,
"step": 855
},
{
"epoch": 0.34185303514376997,
"grad_norm": 30.03852653503418,
"learning_rate": 6.601437699680512e-05,
"loss": 0.3203,
"step": 856
},
{
"epoch": 0.3422523961661342,
"grad_norm": 26.966028213500977,
"learning_rate": 6.597444089456869e-05,
"loss": 0.4146,
"step": 857
},
{
"epoch": 0.3426517571884984,
"grad_norm": 35.2926025390625,
"learning_rate": 6.593450479233228e-05,
"loss": 0.3128,
"step": 858
},
{
"epoch": 0.3430511182108626,
"grad_norm": 30.514089584350586,
"learning_rate": 6.589456869009584e-05,
"loss": 0.6723,
"step": 859
},
{
"epoch": 0.34345047923322686,
"grad_norm": 28.382482528686523,
"learning_rate": 6.585463258785944e-05,
"loss": 1.094,
"step": 860
},
{
"epoch": 0.34384984025559107,
"grad_norm": 27.529977798461914,
"learning_rate": 6.5814696485623e-05,
"loss": 0.8853,
"step": 861
},
{
"epoch": 0.3442492012779553,
"grad_norm": 0.4846610724925995,
"learning_rate": 6.577476038338658e-05,
"loss": 0.0055,
"step": 862
},
{
"epoch": 0.3446485623003195,
"grad_norm": 10.081826210021973,
"learning_rate": 6.573482428115016e-05,
"loss": 0.1476,
"step": 863
},
{
"epoch": 0.3450479233226837,
"grad_norm": 12.221709251403809,
"learning_rate": 6.569488817891374e-05,
"loss": 0.1966,
"step": 864
},
{
"epoch": 0.3454472843450479,
"grad_norm": 0.5540055632591248,
"learning_rate": 6.565495207667732e-05,
"loss": 0.0035,
"step": 865
},
{
"epoch": 0.34584664536741216,
"grad_norm": 20.68918228149414,
"learning_rate": 6.56150159744409e-05,
"loss": 0.1422,
"step": 866
},
{
"epoch": 0.3462460063897764,
"grad_norm": 37.76836395263672,
"learning_rate": 6.557507987220448e-05,
"loss": 0.548,
"step": 867
},
{
"epoch": 0.3466453674121406,
"grad_norm": 48.62034225463867,
"learning_rate": 6.553514376996806e-05,
"loss": 0.5136,
"step": 868
},
{
"epoch": 0.3470447284345048,
"grad_norm": 62.92634963989258,
"learning_rate": 6.549520766773163e-05,
"loss": 0.7791,
"step": 869
},
{
"epoch": 0.347444089456869,
"grad_norm": 28.95008659362793,
"learning_rate": 6.545527156549522e-05,
"loss": 0.5263,
"step": 870
},
{
"epoch": 0.3478434504792332,
"grad_norm": 20.801164627075195,
"learning_rate": 6.541533546325878e-05,
"loss": 0.2912,
"step": 871
},
{
"epoch": 0.34824281150159747,
"grad_norm": 10.849574089050293,
"learning_rate": 6.537539936102238e-05,
"loss": 0.1583,
"step": 872
},
{
"epoch": 0.3486421725239617,
"grad_norm": 32.836708068847656,
"learning_rate": 6.533546325878594e-05,
"loss": 0.3054,
"step": 873
},
{
"epoch": 0.3490415335463259,
"grad_norm": 21.003063201904297,
"learning_rate": 6.529552715654952e-05,
"loss": 0.4865,
"step": 874
},
{
"epoch": 0.3494408945686901,
"grad_norm": 19.00481414794922,
"learning_rate": 6.52555910543131e-05,
"loss": 0.2119,
"step": 875
},
{
"epoch": 0.3498402555910543,
"grad_norm": 2.4746034145355225,
"learning_rate": 6.521565495207668e-05,
"loss": 0.0353,
"step": 876
},
{
"epoch": 0.3502396166134185,
"grad_norm": 9.271681785583496,
"learning_rate": 6.517571884984026e-05,
"loss": 0.064,
"step": 877
},
{
"epoch": 0.3506389776357827,
"grad_norm": 28.518909454345703,
"learning_rate": 6.513578274760384e-05,
"loss": 0.6002,
"step": 878
},
{
"epoch": 0.351038338658147,
"grad_norm": 25.42881202697754,
"learning_rate": 6.509584664536742e-05,
"loss": 0.7884,
"step": 879
},
{
"epoch": 0.3514376996805112,
"grad_norm": 33.3294792175293,
"learning_rate": 6.5055910543131e-05,
"loss": 0.6697,
"step": 880
},
{
"epoch": 0.3518370607028754,
"grad_norm": 19.26186180114746,
"learning_rate": 6.501597444089457e-05,
"loss": 0.2156,
"step": 881
},
{
"epoch": 0.3522364217252396,
"grad_norm": 27.477293014526367,
"learning_rate": 6.497603833865816e-05,
"loss": 0.2596,
"step": 882
},
{
"epoch": 0.3526357827476038,
"grad_norm": 45.442237854003906,
"learning_rate": 6.493610223642172e-05,
"loss": 0.6658,
"step": 883
},
{
"epoch": 0.35303514376996803,
"grad_norm": 24.96969223022461,
"learning_rate": 6.48961661341853e-05,
"loss": 0.6028,
"step": 884
},
{
"epoch": 0.3534345047923323,
"grad_norm": 22.0533447265625,
"learning_rate": 6.485623003194888e-05,
"loss": 0.397,
"step": 885
},
{
"epoch": 0.3538338658146965,
"grad_norm": 4.383850574493408,
"learning_rate": 6.481629392971246e-05,
"loss": 0.045,
"step": 886
},
{
"epoch": 0.3542332268370607,
"grad_norm": 31.18520164489746,
"learning_rate": 6.477635782747604e-05,
"loss": 0.2883,
"step": 887
},
{
"epoch": 0.3546325878594249,
"grad_norm": 4.698118686676025,
"learning_rate": 6.473642172523962e-05,
"loss": 0.0276,
"step": 888
},
{
"epoch": 0.35503194888178913,
"grad_norm": 2.144284963607788,
"learning_rate": 6.46964856230032e-05,
"loss": 0.0445,
"step": 889
},
{
"epoch": 0.35543130990415334,
"grad_norm": 23.248645782470703,
"learning_rate": 6.465654952076678e-05,
"loss": 0.2374,
"step": 890
},
{
"epoch": 0.35583067092651754,
"grad_norm": 18.654600143432617,
"learning_rate": 6.461661341853036e-05,
"loss": 0.1502,
"step": 891
},
{
"epoch": 0.3562300319488818,
"grad_norm": 49.13082504272461,
"learning_rate": 6.457667731629393e-05,
"loss": 1.1763,
"step": 892
},
{
"epoch": 0.356629392971246,
"grad_norm": 25.24797821044922,
"learning_rate": 6.45367412140575e-05,
"loss": 0.4516,
"step": 893
},
{
"epoch": 0.3570287539936102,
"grad_norm": 50.32395553588867,
"learning_rate": 6.449680511182108e-05,
"loss": 0.9372,
"step": 894
},
{
"epoch": 0.35742811501597443,
"grad_norm": 18.263736724853516,
"learning_rate": 6.445686900958466e-05,
"loss": 0.1361,
"step": 895
},
{
"epoch": 0.35782747603833864,
"grad_norm": 27.254375457763672,
"learning_rate": 6.441693290734824e-05,
"loss": 0.5457,
"step": 896
},
{
"epoch": 0.35822683706070285,
"grad_norm": 31.043920516967773,
"learning_rate": 6.437699680511182e-05,
"loss": 0.5477,
"step": 897
},
{
"epoch": 0.3586261980830671,
"grad_norm": 28.927886962890625,
"learning_rate": 6.43370607028754e-05,
"loss": 0.4984,
"step": 898
},
{
"epoch": 0.3590255591054313,
"grad_norm": 55.93524932861328,
"learning_rate": 6.429712460063898e-05,
"loss": 0.9904,
"step": 899
},
{
"epoch": 0.35942492012779553,
"grad_norm": 2.2842910289764404,
"learning_rate": 6.425718849840256e-05,
"loss": 0.0209,
"step": 900
},
{
"epoch": 0.35982428115015974,
"grad_norm": 8.854307174682617,
"learning_rate": 6.421725239616614e-05,
"loss": 0.1357,
"step": 901
},
{
"epoch": 0.36022364217252395,
"grad_norm": 12.463196754455566,
"learning_rate": 6.41773162939297e-05,
"loss": 0.2566,
"step": 902
},
{
"epoch": 0.36062300319488816,
"grad_norm": 28.88583755493164,
"learning_rate": 6.41373801916933e-05,
"loss": 0.5733,
"step": 903
},
{
"epoch": 0.3610223642172524,
"grad_norm": 16.302480697631836,
"learning_rate": 6.409744408945686e-05,
"loss": 0.1354,
"step": 904
},
{
"epoch": 0.36142172523961663,
"grad_norm": 3.5970258712768555,
"learning_rate": 6.405750798722046e-05,
"loss": 0.027,
"step": 905
},
{
"epoch": 0.36182108626198084,
"grad_norm": 27.432641983032227,
"learning_rate": 6.401757188498402e-05,
"loss": 0.5346,
"step": 906
},
{
"epoch": 0.36222044728434505,
"grad_norm": 44.062068939208984,
"learning_rate": 6.39776357827476e-05,
"loss": 0.9281,
"step": 907
},
{
"epoch": 0.36261980830670926,
"grad_norm": 24.087953567504883,
"learning_rate": 6.393769968051118e-05,
"loss": 0.2653,
"step": 908
},
{
"epoch": 0.36301916932907347,
"grad_norm": 9.621813774108887,
"learning_rate": 6.389776357827476e-05,
"loss": 0.0626,
"step": 909
},
{
"epoch": 0.3634185303514377,
"grad_norm": 16.891395568847656,
"learning_rate": 6.385782747603834e-05,
"loss": 0.1213,
"step": 910
},
{
"epoch": 0.36381789137380194,
"grad_norm": 47.794002532958984,
"learning_rate": 6.381789137380192e-05,
"loss": 0.9086,
"step": 911
},
{
"epoch": 0.36421725239616615,
"grad_norm": 7.4638671875,
"learning_rate": 6.37779552715655e-05,
"loss": 0.0489,
"step": 912
},
{
"epoch": 0.36461661341853036,
"grad_norm": 28.758914947509766,
"learning_rate": 6.373801916932908e-05,
"loss": 0.8361,
"step": 913
},
{
"epoch": 0.36501597444089456,
"grad_norm": 13.326181411743164,
"learning_rate": 6.369808306709265e-05,
"loss": 0.1093,
"step": 914
},
{
"epoch": 0.3654153354632588,
"grad_norm": 22.788196563720703,
"learning_rate": 6.365814696485624e-05,
"loss": 0.256,
"step": 915
},
{
"epoch": 0.365814696485623,
"grad_norm": 10.716828346252441,
"learning_rate": 6.36182108626198e-05,
"loss": 0.496,
"step": 916
},
{
"epoch": 0.36621405750798725,
"grad_norm": 17.397903442382812,
"learning_rate": 6.35782747603834e-05,
"loss": 0.1704,
"step": 917
},
{
"epoch": 0.36661341853035145,
"grad_norm": 11.86805248260498,
"learning_rate": 6.353833865814696e-05,
"loss": 0.2657,
"step": 918
},
{
"epoch": 0.36701277955271566,
"grad_norm": 17.720869064331055,
"learning_rate": 6.349840255591054e-05,
"loss": 0.461,
"step": 919
},
{
"epoch": 0.36741214057507987,
"grad_norm": 4.142230033874512,
"learning_rate": 6.345846645367412e-05,
"loss": 0.0741,
"step": 920
},
{
"epoch": 0.3678115015974441,
"grad_norm": 0.821068525314331,
"learning_rate": 6.34185303514377e-05,
"loss": 0.0058,
"step": 921
},
{
"epoch": 0.3682108626198083,
"grad_norm": 12.029718399047852,
"learning_rate": 6.337859424920128e-05,
"loss": 0.0948,
"step": 922
},
{
"epoch": 0.3686102236421725,
"grad_norm": 31.931339263916016,
"learning_rate": 6.333865814696486e-05,
"loss": 0.3049,
"step": 923
},
{
"epoch": 0.36900958466453676,
"grad_norm": 23.627349853515625,
"learning_rate": 6.329872204472844e-05,
"loss": 0.2509,
"step": 924
},
{
"epoch": 0.36940894568690097,
"grad_norm": 24.630802154541016,
"learning_rate": 6.325878594249202e-05,
"loss": 0.5153,
"step": 925
},
{
"epoch": 0.3698083067092652,
"grad_norm": 20.570104598999023,
"learning_rate": 6.321884984025559e-05,
"loss": 0.6206,
"step": 926
},
{
"epoch": 0.3702076677316294,
"grad_norm": 7.3514933586120605,
"learning_rate": 6.317891373801918e-05,
"loss": 0.0454,
"step": 927
},
{
"epoch": 0.3706070287539936,
"grad_norm": 8.437217712402344,
"learning_rate": 6.313897763578274e-05,
"loss": 0.1083,
"step": 928
},
{
"epoch": 0.3710063897763578,
"grad_norm": 59.57499694824219,
"learning_rate": 6.309904153354634e-05,
"loss": 0.6763,
"step": 929
},
{
"epoch": 0.37140575079872207,
"grad_norm": 27.216222763061523,
"learning_rate": 6.30591054313099e-05,
"loss": 0.725,
"step": 930
},
{
"epoch": 0.3718051118210863,
"grad_norm": 16.471351623535156,
"learning_rate": 6.30191693290735e-05,
"loss": 0.1631,
"step": 931
},
{
"epoch": 0.3722044728434505,
"grad_norm": 1.3753671646118164,
"learning_rate": 6.297923322683706e-05,
"loss": 0.0071,
"step": 932
},
{
"epoch": 0.3726038338658147,
"grad_norm": 24.514118194580078,
"learning_rate": 6.293929712460064e-05,
"loss": 0.9195,
"step": 933
},
{
"epoch": 0.3730031948881789,
"grad_norm": 15.225299835205078,
"learning_rate": 6.289936102236422e-05,
"loss": 0.1015,
"step": 934
},
{
"epoch": 0.3734025559105431,
"grad_norm": 32.29588317871094,
"learning_rate": 6.28594249201278e-05,
"loss": 0.4072,
"step": 935
},
{
"epoch": 0.3738019169329074,
"grad_norm": 29.265844345092773,
"learning_rate": 6.281948881789138e-05,
"loss": 0.6789,
"step": 936
},
{
"epoch": 0.3742012779552716,
"grad_norm": 23.81279945373535,
"learning_rate": 6.277955271565496e-05,
"loss": 0.2501,
"step": 937
},
{
"epoch": 0.3746006389776358,
"grad_norm": 17.615232467651367,
"learning_rate": 6.273961661341852e-05,
"loss": 0.1668,
"step": 938
},
{
"epoch": 0.375,
"grad_norm": 56.72080612182617,
"learning_rate": 6.269968051118212e-05,
"loss": 0.9121,
"step": 939
},
{
"epoch": 0.3753993610223642,
"grad_norm": 55.553306579589844,
"learning_rate": 6.265974440894568e-05,
"loss": 0.8228,
"step": 940
},
{
"epoch": 0.3757987220447284,
"grad_norm": 28.670480728149414,
"learning_rate": 6.261980830670928e-05,
"loss": 0.4781,
"step": 941
},
{
"epoch": 0.3761980830670926,
"grad_norm": 43.180328369140625,
"learning_rate": 6.257987220447284e-05,
"loss": 0.5523,
"step": 942
},
{
"epoch": 0.3765974440894569,
"grad_norm": 8.305335998535156,
"learning_rate": 6.253993610223644e-05,
"loss": 0.0527,
"step": 943
},
{
"epoch": 0.3769968051118211,
"grad_norm": 2.3244333267211914,
"learning_rate": 6.25e-05,
"loss": 0.0155,
"step": 944
},
{
"epoch": 0.3773961661341853,
"grad_norm": 0.44165247678756714,
"learning_rate": 6.246006389776358e-05,
"loss": 0.0027,
"step": 945
},
{
"epoch": 0.3777955271565495,
"grad_norm": 3.0560715198516846,
"learning_rate": 6.242012779552716e-05,
"loss": 0.0238,
"step": 946
},
{
"epoch": 0.3781948881789137,
"grad_norm": 23.684154510498047,
"learning_rate": 6.238019169329074e-05,
"loss": 1.1671,
"step": 947
},
{
"epoch": 0.37859424920127793,
"grad_norm": 0.487845778465271,
"learning_rate": 6.234025559105432e-05,
"loss": 0.0031,
"step": 948
},
{
"epoch": 0.3789936102236422,
"grad_norm": 33.09947204589844,
"learning_rate": 6.23003194888179e-05,
"loss": 0.5681,
"step": 949
},
{
"epoch": 0.3793929712460064,
"grad_norm": 0.9794586896896362,
"learning_rate": 6.226038338658148e-05,
"loss": 0.0053,
"step": 950
},
{
"epoch": 0.3797923322683706,
"grad_norm": 29.1749324798584,
"learning_rate": 6.222044728434506e-05,
"loss": 0.2495,
"step": 951
},
{
"epoch": 0.3801916932907348,
"grad_norm": 24.242185592651367,
"learning_rate": 6.218051118210862e-05,
"loss": 0.2951,
"step": 952
},
{
"epoch": 0.38059105431309903,
"grad_norm": 0.9559796452522278,
"learning_rate": 6.214057507987222e-05,
"loss": 0.006,
"step": 953
},
{
"epoch": 0.38099041533546324,
"grad_norm": 23.533430099487305,
"learning_rate": 6.210063897763578e-05,
"loss": 0.2662,
"step": 954
},
{
"epoch": 0.3813897763578275,
"grad_norm": 15.041793823242188,
"learning_rate": 6.206070287539936e-05,
"loss": 0.1787,
"step": 955
},
{
"epoch": 0.3817891373801917,
"grad_norm": 17.662010192871094,
"learning_rate": 6.202076677316294e-05,
"loss": 0.9676,
"step": 956
},
{
"epoch": 0.3821884984025559,
"grad_norm": 44.16839599609375,
"learning_rate": 6.198083067092652e-05,
"loss": 0.4599,
"step": 957
},
{
"epoch": 0.38258785942492013,
"grad_norm": 53.7128791809082,
"learning_rate": 6.19408945686901e-05,
"loss": 2.106,
"step": 958
},
{
"epoch": 0.38298722044728434,
"grad_norm": 3.4108924865722656,
"learning_rate": 6.190095846645368e-05,
"loss": 0.0288,
"step": 959
},
{
"epoch": 0.38338658146964855,
"grad_norm": 39.22096252441406,
"learning_rate": 6.186102236421726e-05,
"loss": 0.3506,
"step": 960
},
{
"epoch": 0.38378594249201275,
"grad_norm": 35.553226470947266,
"learning_rate": 6.182108626198084e-05,
"loss": 0.5498,
"step": 961
},
{
"epoch": 0.384185303514377,
"grad_norm": 25.94477081298828,
"learning_rate": 6.178115015974442e-05,
"loss": 0.2505,
"step": 962
},
{
"epoch": 0.3845846645367412,
"grad_norm": 8.241161346435547,
"learning_rate": 6.174121405750798e-05,
"loss": 0.0993,
"step": 963
},
{
"epoch": 0.38498402555910544,
"grad_norm": 27.373329162597656,
"learning_rate": 6.170127795527156e-05,
"loss": 0.4476,
"step": 964
},
{
"epoch": 0.38538338658146964,
"grad_norm": 26.48349952697754,
"learning_rate": 6.166134185303514e-05,
"loss": 0.3128,
"step": 965
},
{
"epoch": 0.38578274760383385,
"grad_norm": 2.881577253341675,
"learning_rate": 6.162140575079872e-05,
"loss": 0.0189,
"step": 966
},
{
"epoch": 0.38618210862619806,
"grad_norm": 42.797611236572266,
"learning_rate": 6.15814696485623e-05,
"loss": 0.7362,
"step": 967
},
{
"epoch": 0.3865814696485623,
"grad_norm": 15.259772300720215,
"learning_rate": 6.154153354632588e-05,
"loss": 0.1008,
"step": 968
},
{
"epoch": 0.38698083067092653,
"grad_norm": 18.857995986938477,
"learning_rate": 6.150159744408946e-05,
"loss": 0.1512,
"step": 969
},
{
"epoch": 0.38738019169329074,
"grad_norm": 18.7788028717041,
"learning_rate": 6.146166134185304e-05,
"loss": 0.1356,
"step": 970
},
{
"epoch": 0.38777955271565495,
"grad_norm": 24.912708282470703,
"learning_rate": 6.142172523961662e-05,
"loss": 0.3176,
"step": 971
},
{
"epoch": 0.38817891373801916,
"grad_norm": 0.36943647265434265,
"learning_rate": 6.13817891373802e-05,
"loss": 0.0027,
"step": 972
},
{
"epoch": 0.38857827476038337,
"grad_norm": 58.03767013549805,
"learning_rate": 6.134185303514376e-05,
"loss": 0.6924,
"step": 973
},
{
"epoch": 0.3889776357827476,
"grad_norm": 11.039884567260742,
"learning_rate": 6.130191693290736e-05,
"loss": 0.0632,
"step": 974
},
{
"epoch": 0.38937699680511184,
"grad_norm": 5.058833599090576,
"learning_rate": 6.126198083067092e-05,
"loss": 0.0263,
"step": 975
},
{
"epoch": 0.38977635782747605,
"grad_norm": 36.77410888671875,
"learning_rate": 6.12220447284345e-05,
"loss": 0.2944,
"step": 976
},
{
"epoch": 0.39017571884984026,
"grad_norm": 7.9986138343811035,
"learning_rate": 6.118210862619808e-05,
"loss": 0.0818,
"step": 977
},
{
"epoch": 0.39057507987220447,
"grad_norm": 22.34180450439453,
"learning_rate": 6.114217252396166e-05,
"loss": 0.3566,
"step": 978
},
{
"epoch": 0.3909744408945687,
"grad_norm": 26.381834030151367,
"learning_rate": 6.110223642172524e-05,
"loss": 0.2288,
"step": 979
},
{
"epoch": 0.3913738019169329,
"grad_norm": 27.5614013671875,
"learning_rate": 6.106230031948882e-05,
"loss": 0.8035,
"step": 980
},
{
"epoch": 0.39177316293929715,
"grad_norm": 23.947765350341797,
"learning_rate": 6.1022364217252406e-05,
"loss": 0.4025,
"step": 981
},
{
"epoch": 0.39217252396166136,
"grad_norm": 2.639420747756958,
"learning_rate": 6.098242811501598e-05,
"loss": 0.0149,
"step": 982
},
{
"epoch": 0.39257188498402557,
"grad_norm": 21.815532684326172,
"learning_rate": 6.094249201277955e-05,
"loss": 0.2147,
"step": 983
},
{
"epoch": 0.3929712460063898,
"grad_norm": 6.211671352386475,
"learning_rate": 6.090255591054314e-05,
"loss": 0.055,
"step": 984
},
{
"epoch": 0.393370607028754,
"grad_norm": 6.597772121429443,
"learning_rate": 6.086261980830671e-05,
"loss": 0.041,
"step": 985
},
{
"epoch": 0.3937699680511182,
"grad_norm": 23.386430740356445,
"learning_rate": 6.0822683706070297e-05,
"loss": 0.235,
"step": 986
},
{
"epoch": 0.39416932907348246,
"grad_norm": 24.943769454956055,
"learning_rate": 6.078274760383387e-05,
"loss": 0.2159,
"step": 987
},
{
"epoch": 0.39456869009584666,
"grad_norm": 16.671537399291992,
"learning_rate": 6.074281150159745e-05,
"loss": 0.7048,
"step": 988
},
{
"epoch": 0.39496805111821087,
"grad_norm": 4.688577175140381,
"learning_rate": 6.070287539936103e-05,
"loss": 0.0253,
"step": 989
},
{
"epoch": 0.3953674121405751,
"grad_norm": 9.834781646728516,
"learning_rate": 6.06629392971246e-05,
"loss": 0.0737,
"step": 990
},
{
"epoch": 0.3957667731629393,
"grad_norm": 16.71487808227539,
"learning_rate": 6.062300319488819e-05,
"loss": 0.7812,
"step": 991
},
{
"epoch": 0.3961661341853035,
"grad_norm": 34.22722244262695,
"learning_rate": 6.058306709265176e-05,
"loss": 0.5974,
"step": 992
},
{
"epoch": 0.3965654952076677,
"grad_norm": 30.841941833496094,
"learning_rate": 6.054313099041534e-05,
"loss": 0.3198,
"step": 993
},
{
"epoch": 0.39696485623003197,
"grad_norm": 12.327808380126953,
"learning_rate": 6.050319488817892e-05,
"loss": 0.0731,
"step": 994
},
{
"epoch": 0.3973642172523962,
"grad_norm": 6.317902088165283,
"learning_rate": 6.04632587859425e-05,
"loss": 0.0391,
"step": 995
},
{
"epoch": 0.3977635782747604,
"grad_norm": 10.387664794921875,
"learning_rate": 6.042332268370608e-05,
"loss": 0.0911,
"step": 996
},
{
"epoch": 0.3981629392971246,
"grad_norm": 23.886455535888672,
"learning_rate": 6.038338658146965e-05,
"loss": 0.3118,
"step": 997
},
{
"epoch": 0.3985623003194888,
"grad_norm": 11.50920295715332,
"learning_rate": 6.034345047923323e-05,
"loss": 0.0997,
"step": 998
},
{
"epoch": 0.398961661341853,
"grad_norm": 16.370258331298828,
"learning_rate": 6.030351437699681e-05,
"loss": 0.1511,
"step": 999
},
{
"epoch": 0.3993610223642173,
"grad_norm": 9.770106315612793,
"learning_rate": 6.026357827476039e-05,
"loss": 0.1796,
"step": 1000
},
{
"epoch": 0.3997603833865815,
"grad_norm": 5.190211772918701,
"learning_rate": 6.022364217252396e-05,
"loss": 0.101,
"step": 1001
},
{
"epoch": 0.4001597444089457,
"grad_norm": 14.707173347473145,
"learning_rate": 6.018370607028754e-05,
"loss": 0.7566,
"step": 1002
},
{
"epoch": 0.4005591054313099,
"grad_norm": 29.430030822753906,
"learning_rate": 6.014376996805112e-05,
"loss": 0.2316,
"step": 1003
},
{
"epoch": 0.4009584664536741,
"grad_norm": 11.539054870605469,
"learning_rate": 6.01038338658147e-05,
"loss": 0.1546,
"step": 1004
},
{
"epoch": 0.4013578274760383,
"grad_norm": 28.7519588470459,
"learning_rate": 6.006389776357828e-05,
"loss": 0.3881,
"step": 1005
},
{
"epoch": 0.40175718849840253,
"grad_norm": 44.26334762573242,
"learning_rate": 6.002396166134185e-05,
"loss": 1.1358,
"step": 1006
},
{
"epoch": 0.4021565495207668,
"grad_norm": 6.154173374176025,
"learning_rate": 5.998402555910544e-05,
"loss": 0.0402,
"step": 1007
},
{
"epoch": 0.402555910543131,
"grad_norm": 22.761491775512695,
"learning_rate": 5.994408945686901e-05,
"loss": 0.3816,
"step": 1008
},
{
"epoch": 0.4029552715654952,
"grad_norm": 20.903339385986328,
"learning_rate": 5.990415335463259e-05,
"loss": 0.1483,
"step": 1009
},
{
"epoch": 0.4033546325878594,
"grad_norm": 0.31397679448127747,
"learning_rate": 5.986421725239617e-05,
"loss": 0.0016,
"step": 1010
},
{
"epoch": 0.4037539936102236,
"grad_norm": 6.6242356300354,
"learning_rate": 5.982428115015974e-05,
"loss": 0.0388,
"step": 1011
},
{
"epoch": 0.40415335463258784,
"grad_norm": 22.339797973632812,
"learning_rate": 5.978434504792333e-05,
"loss": 0.1735,
"step": 1012
},
{
"epoch": 0.4045527156549521,
"grad_norm": 34.82060241699219,
"learning_rate": 5.97444089456869e-05,
"loss": 0.4274,
"step": 1013
},
{
"epoch": 0.4049520766773163,
"grad_norm": 37.66435623168945,
"learning_rate": 5.9704472843450487e-05,
"loss": 0.4666,
"step": 1014
},
{
"epoch": 0.4053514376996805,
"grad_norm": 20.97816276550293,
"learning_rate": 5.966453674121406e-05,
"loss": 0.7392,
"step": 1015
},
{
"epoch": 0.4057507987220447,
"grad_norm": 22.15311050415039,
"learning_rate": 5.962460063897763e-05,
"loss": 0.1926,
"step": 1016
},
{
"epoch": 0.40615015974440893,
"grad_norm": 27.606813430786133,
"learning_rate": 5.958466453674122e-05,
"loss": 0.7211,
"step": 1017
},
{
"epoch": 0.40654952076677314,
"grad_norm": 14.699033737182617,
"learning_rate": 5.954472843450479e-05,
"loss": 0.0981,
"step": 1018
},
{
"epoch": 0.4069488817891374,
"grad_norm": 10.961593627929688,
"learning_rate": 5.950479233226838e-05,
"loss": 0.1127,
"step": 1019
},
{
"epoch": 0.4073482428115016,
"grad_norm": 36.90821075439453,
"learning_rate": 5.946485623003195e-05,
"loss": 0.4078,
"step": 1020
},
{
"epoch": 0.4077476038338658,
"grad_norm": 7.762786865234375,
"learning_rate": 5.942492012779552e-05,
"loss": 0.1259,
"step": 1021
},
{
"epoch": 0.40814696485623003,
"grad_norm": 2.961153507232666,
"learning_rate": 5.938498402555911e-05,
"loss": 0.0291,
"step": 1022
},
{
"epoch": 0.40854632587859424,
"grad_norm": 75.8663558959961,
"learning_rate": 5.934504792332268e-05,
"loss": 1.4784,
"step": 1023
},
{
"epoch": 0.40894568690095845,
"grad_norm": 12.422547340393066,
"learning_rate": 5.930511182108627e-05,
"loss": 0.0844,
"step": 1024
},
{
"epoch": 0.40934504792332266,
"grad_norm": 49.54899978637695,
"learning_rate": 5.926517571884984e-05,
"loss": 1.3156,
"step": 1025
},
{
"epoch": 0.4097444089456869,
"grad_norm": 27.506105422973633,
"learning_rate": 5.9225239616613426e-05,
"loss": 0.2729,
"step": 1026
},
{
"epoch": 0.41014376996805113,
"grad_norm": 14.856189727783203,
"learning_rate": 5.9185303514377e-05,
"loss": 0.1233,
"step": 1027
},
{
"epoch": 0.41054313099041534,
"grad_norm": 33.993736267089844,
"learning_rate": 5.914536741214057e-05,
"loss": 0.4253,
"step": 1028
},
{
"epoch": 0.41094249201277955,
"grad_norm": 21.506208419799805,
"learning_rate": 5.910543130990416e-05,
"loss": 0.39,
"step": 1029
},
{
"epoch": 0.41134185303514376,
"grad_norm": 19.703210830688477,
"learning_rate": 5.906549520766773e-05,
"loss": 0.2023,
"step": 1030
},
{
"epoch": 0.41174121405750796,
"grad_norm": 1.2864934206008911,
"learning_rate": 5.9025559105431316e-05,
"loss": 0.0078,
"step": 1031
},
{
"epoch": 0.41214057507987223,
"grad_norm": 32.17695236206055,
"learning_rate": 5.898562300319489e-05,
"loss": 0.7999,
"step": 1032
},
{
"epoch": 0.41253993610223644,
"grad_norm": 13.878713607788086,
"learning_rate": 5.8945686900958475e-05,
"loss": 0.0955,
"step": 1033
},
{
"epoch": 0.41293929712460065,
"grad_norm": 25.551305770874023,
"learning_rate": 5.890575079872205e-05,
"loss": 0.281,
"step": 1034
},
{
"epoch": 0.41333865814696485,
"grad_norm": 7.618232250213623,
"learning_rate": 5.886581469648562e-05,
"loss": 0.2218,
"step": 1035
},
{
"epoch": 0.41373801916932906,
"grad_norm": 0.4094751477241516,
"learning_rate": 5.882587859424921e-05,
"loss": 0.0083,
"step": 1036
},
{
"epoch": 0.41413738019169327,
"grad_norm": 42.133121490478516,
"learning_rate": 5.878594249201278e-05,
"loss": 0.9781,
"step": 1037
},
{
"epoch": 0.4145367412140575,
"grad_norm": 28.13552474975586,
"learning_rate": 5.8746006389776366e-05,
"loss": 0.7221,
"step": 1038
},
{
"epoch": 0.41493610223642174,
"grad_norm": 19.163135528564453,
"learning_rate": 5.870607028753994e-05,
"loss": 0.2427,
"step": 1039
},
{
"epoch": 0.41533546325878595,
"grad_norm": 51.25397872924805,
"learning_rate": 5.866613418530351e-05,
"loss": 0.9769,
"step": 1040
},
{
"epoch": 0.41573482428115016,
"grad_norm": 39.05422592163086,
"learning_rate": 5.86261980830671e-05,
"loss": 0.4741,
"step": 1041
},
{
"epoch": 0.41613418530351437,
"grad_norm": 12.507061004638672,
"learning_rate": 5.858626198083067e-05,
"loss": 0.1722,
"step": 1042
},
{
"epoch": 0.4165335463258786,
"grad_norm": 26.537723541259766,
"learning_rate": 5.8546325878594256e-05,
"loss": 0.4927,
"step": 1043
},
{
"epoch": 0.4169329073482428,
"grad_norm": 24.323801040649414,
"learning_rate": 5.850638977635783e-05,
"loss": 0.4252,
"step": 1044
},
{
"epoch": 0.41733226837060705,
"grad_norm": 31.838563919067383,
"learning_rate": 5.8466453674121415e-05,
"loss": 0.6378,
"step": 1045
},
{
"epoch": 0.41773162939297126,
"grad_norm": 15.417181968688965,
"learning_rate": 5.842651757188499e-05,
"loss": 0.0941,
"step": 1046
},
{
"epoch": 0.41813099041533547,
"grad_norm": 21.96835708618164,
"learning_rate": 5.838658146964856e-05,
"loss": 0.1907,
"step": 1047
},
{
"epoch": 0.4185303514376997,
"grad_norm": 12.413500785827637,
"learning_rate": 5.8346645367412146e-05,
"loss": 0.1162,
"step": 1048
},
{
"epoch": 0.4189297124600639,
"grad_norm": 0.48411163687705994,
"learning_rate": 5.830670926517572e-05,
"loss": 0.0068,
"step": 1049
},
{
"epoch": 0.4193290734824281,
"grad_norm": 0.9858969449996948,
"learning_rate": 5.8266773162939305e-05,
"loss": 0.0102,
"step": 1050
},
{
"epoch": 0.41972843450479236,
"grad_norm": 0.2809281647205353,
"learning_rate": 5.822683706070288e-05,
"loss": 0.003,
"step": 1051
},
{
"epoch": 0.42012779552715657,
"grad_norm": 12.993391036987305,
"learning_rate": 5.8186900958466464e-05,
"loss": 0.088,
"step": 1052
},
{
"epoch": 0.4205271565495208,
"grad_norm": 19.031373977661133,
"learning_rate": 5.814696485623004e-05,
"loss": 1.1547,
"step": 1053
},
{
"epoch": 0.420926517571885,
"grad_norm": 44.09419250488281,
"learning_rate": 5.810702875399361e-05,
"loss": 0.4629,
"step": 1054
},
{
"epoch": 0.4213258785942492,
"grad_norm": 19.086896896362305,
"learning_rate": 5.8067092651757196e-05,
"loss": 0.136,
"step": 1055
},
{
"epoch": 0.4217252396166134,
"grad_norm": 18.201818466186523,
"learning_rate": 5.802715654952077e-05,
"loss": 0.1385,
"step": 1056
},
{
"epoch": 0.4221246006389776,
"grad_norm": 31.258581161499023,
"learning_rate": 5.7987220447284354e-05,
"loss": 0.3112,
"step": 1057
},
{
"epoch": 0.4225239616613419,
"grad_norm": 66.86408996582031,
"learning_rate": 5.794728434504793e-05,
"loss": 1.0315,
"step": 1058
},
{
"epoch": 0.4229233226837061,
"grad_norm": 13.118147850036621,
"learning_rate": 5.79073482428115e-05,
"loss": 0.1021,
"step": 1059
},
{
"epoch": 0.4233226837060703,
"grad_norm": 37.817874908447266,
"learning_rate": 5.7867412140575086e-05,
"loss": 0.4453,
"step": 1060
},
{
"epoch": 0.4237220447284345,
"grad_norm": 24.467548370361328,
"learning_rate": 5.782747603833866e-05,
"loss": 0.2814,
"step": 1061
},
{
"epoch": 0.4241214057507987,
"grad_norm": 7.4809417724609375,
"learning_rate": 5.7787539936102245e-05,
"loss": 0.0937,
"step": 1062
},
{
"epoch": 0.4245207667731629,
"grad_norm": 40.4588623046875,
"learning_rate": 5.774760383386582e-05,
"loss": 0.9463,
"step": 1063
},
{
"epoch": 0.4249201277955272,
"grad_norm": 43.77117156982422,
"learning_rate": 5.77076677316294e-05,
"loss": 0.7832,
"step": 1064
},
{
"epoch": 0.4253194888178914,
"grad_norm": 1.0945245027542114,
"learning_rate": 5.7667731629392976e-05,
"loss": 0.0057,
"step": 1065
},
{
"epoch": 0.4257188498402556,
"grad_norm": 35.09583282470703,
"learning_rate": 5.762779552715655e-05,
"loss": 0.5078,
"step": 1066
},
{
"epoch": 0.4261182108626198,
"grad_norm": 21.034931182861328,
"learning_rate": 5.7587859424920135e-05,
"loss": 0.3198,
"step": 1067
},
{
"epoch": 0.426517571884984,
"grad_norm": 5.2050933837890625,
"learning_rate": 5.754792332268371e-05,
"loss": 0.032,
"step": 1068
},
{
"epoch": 0.4269169329073482,
"grad_norm": 26.07735824584961,
"learning_rate": 5.750798722044729e-05,
"loss": 0.3936,
"step": 1069
},
{
"epoch": 0.4273162939297125,
"grad_norm": 9.418916702270508,
"learning_rate": 5.746805111821087e-05,
"loss": 0.0545,
"step": 1070
},
{
"epoch": 0.4277156549520767,
"grad_norm": 4.307847499847412,
"learning_rate": 5.7428115015974446e-05,
"loss": 0.0462,
"step": 1071
},
{
"epoch": 0.4281150159744409,
"grad_norm": 30.118999481201172,
"learning_rate": 5.7388178913738025e-05,
"loss": 0.9659,
"step": 1072
},
{
"epoch": 0.4285143769968051,
"grad_norm": 1.0838748216629028,
"learning_rate": 5.73482428115016e-05,
"loss": 0.0073,
"step": 1073
},
{
"epoch": 0.4289137380191693,
"grad_norm": 23.359113693237305,
"learning_rate": 5.730830670926518e-05,
"loss": 0.7647,
"step": 1074
},
{
"epoch": 0.42931309904153353,
"grad_norm": 22.036245346069336,
"learning_rate": 5.726837060702876e-05,
"loss": 0.1437,
"step": 1075
},
{
"epoch": 0.42971246006389774,
"grad_norm": 0.9789434671401978,
"learning_rate": 5.7228434504792336e-05,
"loss": 0.0061,
"step": 1076
},
{
"epoch": 0.430111821086262,
"grad_norm": 17.20298194885254,
"learning_rate": 5.718849840255591e-05,
"loss": 0.1251,
"step": 1077
},
{
"epoch": 0.4305111821086262,
"grad_norm": 41.5269889831543,
"learning_rate": 5.7148562300319495e-05,
"loss": 0.9007,
"step": 1078
},
{
"epoch": 0.4309105431309904,
"grad_norm": 6.071673393249512,
"learning_rate": 5.710862619808307e-05,
"loss": 0.0322,
"step": 1079
},
{
"epoch": 0.43130990415335463,
"grad_norm": 13.532049179077148,
"learning_rate": 5.706869009584665e-05,
"loss": 2.1545,
"step": 1080
},
{
"epoch": 0.43170926517571884,
"grad_norm": 20.228145599365234,
"learning_rate": 5.702875399361023e-05,
"loss": 0.2366,
"step": 1081
},
{
"epoch": 0.43210862619808305,
"grad_norm": 21.690715789794922,
"learning_rate": 5.69888178913738e-05,
"loss": 0.3223,
"step": 1082
},
{
"epoch": 0.4325079872204473,
"grad_norm": 43.59482955932617,
"learning_rate": 5.6948881789137386e-05,
"loss": 0.9141,
"step": 1083
},
{
"epoch": 0.4329073482428115,
"grad_norm": 5.669691562652588,
"learning_rate": 5.690894568690096e-05,
"loss": 0.0691,
"step": 1084
},
{
"epoch": 0.4333067092651757,
"grad_norm": 2.9354300498962402,
"learning_rate": 5.686900958466454e-05,
"loss": 0.0557,
"step": 1085
},
{
"epoch": 0.43370607028753994,
"grad_norm": 33.87556838989258,
"learning_rate": 5.682907348242812e-05,
"loss": 0.8002,
"step": 1086
},
{
"epoch": 0.43410543130990414,
"grad_norm": 9.663198471069336,
"learning_rate": 5.678913738019169e-05,
"loss": 0.0576,
"step": 1087
},
{
"epoch": 0.43450479233226835,
"grad_norm": 12.978826522827148,
"learning_rate": 5.6749201277955276e-05,
"loss": 0.1052,
"step": 1088
},
{
"epoch": 0.43490415335463256,
"grad_norm": 31.079490661621094,
"learning_rate": 5.670926517571885e-05,
"loss": 0.3794,
"step": 1089
},
{
"epoch": 0.4353035143769968,
"grad_norm": 16.41422462463379,
"learning_rate": 5.6669329073482435e-05,
"loss": 0.1354,
"step": 1090
},
{
"epoch": 0.43570287539936103,
"grad_norm": 18.472978591918945,
"learning_rate": 5.662939297124601e-05,
"loss": 0.2629,
"step": 1091
},
{
"epoch": 0.43610223642172524,
"grad_norm": 25.17333984375,
"learning_rate": 5.658945686900958e-05,
"loss": 0.3046,
"step": 1092
},
{
"epoch": 0.43650159744408945,
"grad_norm": 27.284175872802734,
"learning_rate": 5.6549520766773166e-05,
"loss": 0.9188,
"step": 1093
},
{
"epoch": 0.43690095846645366,
"grad_norm": 8.584308624267578,
"learning_rate": 5.650958466453674e-05,
"loss": 0.0475,
"step": 1094
},
{
"epoch": 0.43730031948881787,
"grad_norm": 13.85356330871582,
"learning_rate": 5.6469648562300325e-05,
"loss": 0.1756,
"step": 1095
},
{
"epoch": 0.43769968051118213,
"grad_norm": 8.63464641571045,
"learning_rate": 5.64297124600639e-05,
"loss": 0.073,
"step": 1096
},
{
"epoch": 0.43809904153354634,
"grad_norm": 21.48573875427246,
"learning_rate": 5.6389776357827484e-05,
"loss": 0.2519,
"step": 1097
},
{
"epoch": 0.43849840255591055,
"grad_norm": 5.334835529327393,
"learning_rate": 5.634984025559106e-05,
"loss": 0.0315,
"step": 1098
},
{
"epoch": 0.43889776357827476,
"grad_norm": 2.8289856910705566,
"learning_rate": 5.630990415335463e-05,
"loss": 0.016,
"step": 1099
},
{
"epoch": 0.43929712460063897,
"grad_norm": 0.9615386724472046,
"learning_rate": 5.6269968051118216e-05,
"loss": 0.0055,
"step": 1100
},
{
"epoch": 0.4396964856230032,
"grad_norm": 8.775245666503906,
"learning_rate": 5.623003194888179e-05,
"loss": 0.0555,
"step": 1101
},
{
"epoch": 0.44009584664536744,
"grad_norm": 29.180891036987305,
"learning_rate": 5.6190095846645374e-05,
"loss": 0.7973,
"step": 1102
},
{
"epoch": 0.44049520766773165,
"grad_norm": 7.3456902503967285,
"learning_rate": 5.615015974440895e-05,
"loss": 0.053,
"step": 1103
},
{
"epoch": 0.44089456869009586,
"grad_norm": 17.207468032836914,
"learning_rate": 5.611022364217252e-05,
"loss": 0.3822,
"step": 1104
},
{
"epoch": 0.44129392971246006,
"grad_norm": 19.893810272216797,
"learning_rate": 5.6070287539936106e-05,
"loss": 0.1535,
"step": 1105
},
{
"epoch": 0.4416932907348243,
"grad_norm": 28.053556442260742,
"learning_rate": 5.603035143769968e-05,
"loss": 0.4174,
"step": 1106
},
{
"epoch": 0.4420926517571885,
"grad_norm": 25.922704696655273,
"learning_rate": 5.5990415335463265e-05,
"loss": 0.5495,
"step": 1107
},
{
"epoch": 0.4424920127795527,
"grad_norm": 2.2229700088500977,
"learning_rate": 5.595047923322684e-05,
"loss": 0.0145,
"step": 1108
},
{
"epoch": 0.44289137380191695,
"grad_norm": 14.818575859069824,
"learning_rate": 5.5910543130990424e-05,
"loss": 0.1394,
"step": 1109
},
{
"epoch": 0.44329073482428116,
"grad_norm": 11.022860527038574,
"learning_rate": 5.5870607028753996e-05,
"loss": 0.0805,
"step": 1110
},
{
"epoch": 0.44369009584664537,
"grad_norm": 24.370302200317383,
"learning_rate": 5.583067092651757e-05,
"loss": 0.166,
"step": 1111
},
{
"epoch": 0.4440894568690096,
"grad_norm": 23.27841567993164,
"learning_rate": 5.5790734824281155e-05,
"loss": 0.2872,
"step": 1112
},
{
"epoch": 0.4444888178913738,
"grad_norm": 1.2480567693710327,
"learning_rate": 5.575079872204473e-05,
"loss": 0.0097,
"step": 1113
},
{
"epoch": 0.444888178913738,
"grad_norm": 24.022905349731445,
"learning_rate": 5.5710862619808314e-05,
"loss": 0.4029,
"step": 1114
},
{
"epoch": 0.44528753993610226,
"grad_norm": 3.0102906227111816,
"learning_rate": 5.5670926517571887e-05,
"loss": 0.0174,
"step": 1115
},
{
"epoch": 0.44568690095846647,
"grad_norm": 8.141077041625977,
"learning_rate": 5.563099041533547e-05,
"loss": 0.0588,
"step": 1116
},
{
"epoch": 0.4460862619808307,
"grad_norm": 30.080520629882812,
"learning_rate": 5.5591054313099045e-05,
"loss": 0.4781,
"step": 1117
},
{
"epoch": 0.4464856230031949,
"grad_norm": 20.281770706176758,
"learning_rate": 5.555111821086262e-05,
"loss": 0.18,
"step": 1118
},
{
"epoch": 0.4468849840255591,
"grad_norm": 5.243573188781738,
"learning_rate": 5.5511182108626204e-05,
"loss": 0.0377,
"step": 1119
},
{
"epoch": 0.4472843450479233,
"grad_norm": 25.744407653808594,
"learning_rate": 5.547124600638978e-05,
"loss": 0.3661,
"step": 1120
},
{
"epoch": 0.4476837060702875,
"grad_norm": 13.405241012573242,
"learning_rate": 5.543130990415336e-05,
"loss": 0.1154,
"step": 1121
},
{
"epoch": 0.4480830670926518,
"grad_norm": 3.886735439300537,
"learning_rate": 5.5391373801916936e-05,
"loss": 0.0241,
"step": 1122
},
{
"epoch": 0.448482428115016,
"grad_norm": 20.6676025390625,
"learning_rate": 5.535143769968051e-05,
"loss": 0.1465,
"step": 1123
},
{
"epoch": 0.4488817891373802,
"grad_norm": 25.80036735534668,
"learning_rate": 5.5311501597444095e-05,
"loss": 0.2694,
"step": 1124
},
{
"epoch": 0.4492811501597444,
"grad_norm": 0.5709608197212219,
"learning_rate": 5.527156549520767e-05,
"loss": 0.0048,
"step": 1125
},
{
"epoch": 0.4496805111821086,
"grad_norm": 20.979080200195312,
"learning_rate": 5.5231629392971253e-05,
"loss": 0.1765,
"step": 1126
},
{
"epoch": 0.4500798722044728,
"grad_norm": 22.06410789489746,
"learning_rate": 5.5191693290734826e-05,
"loss": 0.5323,
"step": 1127
},
{
"epoch": 0.4504792332268371,
"grad_norm": 2.254361629486084,
"learning_rate": 5.515175718849841e-05,
"loss": 0.0137,
"step": 1128
},
{
"epoch": 0.4508785942492013,
"grad_norm": 14.292792320251465,
"learning_rate": 5.5111821086261985e-05,
"loss": 0.0837,
"step": 1129
},
{
"epoch": 0.4512779552715655,
"grad_norm": 12.933573722839355,
"learning_rate": 5.507188498402556e-05,
"loss": 0.0925,
"step": 1130
},
{
"epoch": 0.4516773162939297,
"grad_norm": 7.385883808135986,
"learning_rate": 5.5031948881789144e-05,
"loss": 0.1603,
"step": 1131
},
{
"epoch": 0.4520766773162939,
"grad_norm": 28.928096771240234,
"learning_rate": 5.4992012779552716e-05,
"loss": 0.3731,
"step": 1132
},
{
"epoch": 0.4524760383386581,
"grad_norm": 8.219690322875977,
"learning_rate": 5.49520766773163e-05,
"loss": 0.0663,
"step": 1133
},
{
"epoch": 0.4528753993610224,
"grad_norm": 39.685455322265625,
"learning_rate": 5.4912140575079875e-05,
"loss": 0.5576,
"step": 1134
},
{
"epoch": 0.4532747603833866,
"grad_norm": 12.337695121765137,
"learning_rate": 5.4872204472843455e-05,
"loss": 0.1061,
"step": 1135
},
{
"epoch": 0.4536741214057508,
"grad_norm": 12.87541389465332,
"learning_rate": 5.4832268370607034e-05,
"loss": 0.1211,
"step": 1136
},
{
"epoch": 0.454073482428115,
"grad_norm": 51.56081771850586,
"learning_rate": 5.479233226837061e-05,
"loss": 1.1267,
"step": 1137
},
{
"epoch": 0.4544728434504792,
"grad_norm": 31.57260513305664,
"learning_rate": 5.475239616613419e-05,
"loss": 0.4815,
"step": 1138
},
{
"epoch": 0.45487220447284343,
"grad_norm": 23.455827713012695,
"learning_rate": 5.4712460063897766e-05,
"loss": 0.4063,
"step": 1139
},
{
"epoch": 0.45527156549520764,
"grad_norm": 1.7229710817337036,
"learning_rate": 5.4672523961661345e-05,
"loss": 0.0109,
"step": 1140
},
{
"epoch": 0.4556709265175719,
"grad_norm": 15.947394371032715,
"learning_rate": 5.4632587859424925e-05,
"loss": 0.1165,
"step": 1141
},
{
"epoch": 0.4560702875399361,
"grad_norm": 33.28199768066406,
"learning_rate": 5.4592651757188504e-05,
"loss": 0.9746,
"step": 1142
},
{
"epoch": 0.4564696485623003,
"grad_norm": 50.163944244384766,
"learning_rate": 5.4552715654952083e-05,
"loss": 1.167,
"step": 1143
},
{
"epoch": 0.45686900958466453,
"grad_norm": 30.324359893798828,
"learning_rate": 5.4512779552715656e-05,
"loss": 0.6862,
"step": 1144
},
{
"epoch": 0.45726837060702874,
"grad_norm": 15.528347969055176,
"learning_rate": 5.4472843450479235e-05,
"loss": 0.9782,
"step": 1145
},
{
"epoch": 0.45766773162939295,
"grad_norm": 0.8262531757354736,
"learning_rate": 5.4432907348242815e-05,
"loss": 0.0048,
"step": 1146
},
{
"epoch": 0.4580670926517572,
"grad_norm": 5.808155059814453,
"learning_rate": 5.4392971246006394e-05,
"loss": 0.0816,
"step": 1147
},
{
"epoch": 0.4584664536741214,
"grad_norm": 0.8603410124778748,
"learning_rate": 5.435303514376997e-05,
"loss": 0.0051,
"step": 1148
},
{
"epoch": 0.45886581469648563,
"grad_norm": 0.9969199299812317,
"learning_rate": 5.4313099041533546e-05,
"loss": 0.009,
"step": 1149
},
{
"epoch": 0.45926517571884984,
"grad_norm": 24.603368759155273,
"learning_rate": 5.4273162939297126e-05,
"loss": 0.279,
"step": 1150
},
{
"epoch": 0.45966453674121405,
"grad_norm": 7.651378631591797,
"learning_rate": 5.4233226837060705e-05,
"loss": 0.0443,
"step": 1151
},
{
"epoch": 0.46006389776357826,
"grad_norm": 26.8240909576416,
"learning_rate": 5.4193290734824285e-05,
"loss": 0.2147,
"step": 1152
},
{
"epoch": 0.4604632587859425,
"grad_norm": 3.6380791664123535,
"learning_rate": 5.415335463258786e-05,
"loss": 0.0183,
"step": 1153
},
{
"epoch": 0.46086261980830673,
"grad_norm": 17.245548248291016,
"learning_rate": 5.4113418530351444e-05,
"loss": 0.1325,
"step": 1154
},
{
"epoch": 0.46126198083067094,
"grad_norm": 23.670785903930664,
"learning_rate": 5.4073482428115016e-05,
"loss": 0.2147,
"step": 1155
},
{
"epoch": 0.46166134185303515,
"grad_norm": 9.27513313293457,
"learning_rate": 5.4033546325878596e-05,
"loss": 0.1621,
"step": 1156
},
{
"epoch": 0.46206070287539935,
"grad_norm": 46.31245040893555,
"learning_rate": 5.3993610223642175e-05,
"loss": 0.5352,
"step": 1157
},
{
"epoch": 0.46246006389776356,
"grad_norm": 27.19325065612793,
"learning_rate": 5.395367412140575e-05,
"loss": 1.0256,
"step": 1158
},
{
"epoch": 0.46285942492012777,
"grad_norm": 29.766176223754883,
"learning_rate": 5.3913738019169334e-05,
"loss": 0.2659,
"step": 1159
},
{
"epoch": 0.46325878594249204,
"grad_norm": 26.083826065063477,
"learning_rate": 5.3873801916932907e-05,
"loss": 0.548,
"step": 1160
},
{
"epoch": 0.46365814696485624,
"grad_norm": 5.670355319976807,
"learning_rate": 5.383386581469649e-05,
"loss": 0.0296,
"step": 1161
},
{
"epoch": 0.46405750798722045,
"grad_norm": 2.303006887435913,
"learning_rate": 5.3793929712460065e-05,
"loss": 0.0166,
"step": 1162
},
{
"epoch": 0.46445686900958466,
"grad_norm": 28.55370330810547,
"learning_rate": 5.375399361022364e-05,
"loss": 0.4261,
"step": 1163
},
{
"epoch": 0.46485623003194887,
"grad_norm": 23.90802764892578,
"learning_rate": 5.3714057507987224e-05,
"loss": 0.3888,
"step": 1164
},
{
"epoch": 0.4652555910543131,
"grad_norm": 21.106891632080078,
"learning_rate": 5.36741214057508e-05,
"loss": 0.1692,
"step": 1165
},
{
"epoch": 0.46565495207667734,
"grad_norm": 30.30702018737793,
"learning_rate": 5.363418530351438e-05,
"loss": 0.3731,
"step": 1166
},
{
"epoch": 0.46605431309904155,
"grad_norm": 6.6064982414245605,
"learning_rate": 5.3594249201277956e-05,
"loss": 0.06,
"step": 1167
},
{
"epoch": 0.46645367412140576,
"grad_norm": 7.251408100128174,
"learning_rate": 5.355431309904153e-05,
"loss": 0.0397,
"step": 1168
},
{
"epoch": 0.46685303514376997,
"grad_norm": 26.655858993530273,
"learning_rate": 5.3514376996805115e-05,
"loss": 1.0242,
"step": 1169
},
{
"epoch": 0.4672523961661342,
"grad_norm": 0.22087743878364563,
"learning_rate": 5.347444089456869e-05,
"loss": 0.0047,
"step": 1170
},
{
"epoch": 0.4676517571884984,
"grad_norm": 10.675420761108398,
"learning_rate": 5.3434504792332273e-05,
"loss": 0.074,
"step": 1171
},
{
"epoch": 0.4680511182108626,
"grad_norm": 38.15542984008789,
"learning_rate": 5.3394568690095846e-05,
"loss": 0.5718,
"step": 1172
},
{
"epoch": 0.46845047923322686,
"grad_norm": 2.7904064655303955,
"learning_rate": 5.335463258785943e-05,
"loss": 0.0136,
"step": 1173
},
{
"epoch": 0.46884984025559107,
"grad_norm": 64.55099487304688,
"learning_rate": 5.3314696485623005e-05,
"loss": 1.1965,
"step": 1174
},
{
"epoch": 0.4692492012779553,
"grad_norm": 0.6093473434448242,
"learning_rate": 5.327476038338658e-05,
"loss": 0.0028,
"step": 1175
},
{
"epoch": 0.4696485623003195,
"grad_norm": 24.370859146118164,
"learning_rate": 5.3234824281150164e-05,
"loss": 0.2881,
"step": 1176
},
{
"epoch": 0.4700479233226837,
"grad_norm": 24.490575790405273,
"learning_rate": 5.3194888178913736e-05,
"loss": 0.386,
"step": 1177
},
{
"epoch": 0.4704472843450479,
"grad_norm": 5.813706398010254,
"learning_rate": 5.315495207667732e-05,
"loss": 0.0359,
"step": 1178
},
{
"epoch": 0.47084664536741216,
"grad_norm": 26.41362190246582,
"learning_rate": 5.3115015974440895e-05,
"loss": 0.4844,
"step": 1179
},
{
"epoch": 0.4712460063897764,
"grad_norm": 41.8812141418457,
"learning_rate": 5.307507987220448e-05,
"loss": 0.597,
"step": 1180
},
{
"epoch": 0.4716453674121406,
"grad_norm": 17.681791305541992,
"learning_rate": 5.3035143769968054e-05,
"loss": 0.181,
"step": 1181
},
{
"epoch": 0.4720447284345048,
"grad_norm": 44.31705856323242,
"learning_rate": 5.299520766773163e-05,
"loss": 0.9357,
"step": 1182
},
{
"epoch": 0.472444089456869,
"grad_norm": 55.083824157714844,
"learning_rate": 5.295527156549521e-05,
"loss": 0.661,
"step": 1183
},
{
"epoch": 0.4728434504792332,
"grad_norm": 75.14667510986328,
"learning_rate": 5.2915335463258786e-05,
"loss": 2.0269,
"step": 1184
},
{
"epoch": 0.47324281150159747,
"grad_norm": 10.664979934692383,
"learning_rate": 5.287539936102237e-05,
"loss": 0.0632,
"step": 1185
},
{
"epoch": 0.4736421725239617,
"grad_norm": 6.919378757476807,
"learning_rate": 5.2835463258785944e-05,
"loss": 0.1278,
"step": 1186
},
{
"epoch": 0.4740415335463259,
"grad_norm": 43.13159942626953,
"learning_rate": 5.279552715654952e-05,
"loss": 1.4941,
"step": 1187
},
{
"epoch": 0.4744408945686901,
"grad_norm": 14.374273300170898,
"learning_rate": 5.27555910543131e-05,
"loss": 0.1144,
"step": 1188
},
{
"epoch": 0.4748402555910543,
"grad_norm": 7.736657619476318,
"learning_rate": 5.2715654952076676e-05,
"loss": 0.0721,
"step": 1189
},
{
"epoch": 0.4752396166134185,
"grad_norm": 27.39414405822754,
"learning_rate": 5.267571884984026e-05,
"loss": 0.4196,
"step": 1190
},
{
"epoch": 0.4756389776357827,
"grad_norm": 15.483850479125977,
"learning_rate": 5.2635782747603835e-05,
"loss": 0.102,
"step": 1191
},
{
"epoch": 0.476038338658147,
"grad_norm": 39.54878616333008,
"learning_rate": 5.259584664536742e-05,
"loss": 0.7476,
"step": 1192
},
{
"epoch": 0.4764376996805112,
"grad_norm": 1.824077844619751,
"learning_rate": 5.2555910543130994e-05,
"loss": 0.0207,
"step": 1193
},
{
"epoch": 0.4768370607028754,
"grad_norm": 4.33188009262085,
"learning_rate": 5.2515974440894566e-05,
"loss": 0.0465,
"step": 1194
},
{
"epoch": 0.4772364217252396,
"grad_norm": 34.92250442504883,
"learning_rate": 5.247603833865815e-05,
"loss": 0.6133,
"step": 1195
},
{
"epoch": 0.4776357827476038,
"grad_norm": 20.9636287689209,
"learning_rate": 5.2436102236421725e-05,
"loss": 0.166,
"step": 1196
},
{
"epoch": 0.47803514376996803,
"grad_norm": 14.764833450317383,
"learning_rate": 5.239616613418531e-05,
"loss": 0.116,
"step": 1197
},
{
"epoch": 0.4784345047923323,
"grad_norm": 8.113991737365723,
"learning_rate": 5.2356230031948884e-05,
"loss": 0.0442,
"step": 1198
},
{
"epoch": 0.4788338658146965,
"grad_norm": 0.6276927590370178,
"learning_rate": 5.231629392971247e-05,
"loss": 0.0031,
"step": 1199
},
{
"epoch": 0.4792332268370607,
"grad_norm": 11.987149238586426,
"learning_rate": 5.227635782747604e-05,
"loss": 0.0733,
"step": 1200
},
{
"epoch": 0.4796325878594249,
"grad_norm": 27.02328872680664,
"learning_rate": 5.2236421725239616e-05,
"loss": 0.4267,
"step": 1201
},
{
"epoch": 0.48003194888178913,
"grad_norm": 39.99783706665039,
"learning_rate": 5.21964856230032e-05,
"loss": 1.764,
"step": 1202
},
{
"epoch": 0.48043130990415334,
"grad_norm": 25.729215621948242,
"learning_rate": 5.2156549520766774e-05,
"loss": 0.4457,
"step": 1203
},
{
"epoch": 0.48083067092651754,
"grad_norm": 13.255203247070312,
"learning_rate": 5.211661341853036e-05,
"loss": 0.1588,
"step": 1204
},
{
"epoch": 0.4812300319488818,
"grad_norm": 54.511749267578125,
"learning_rate": 5.207667731629393e-05,
"loss": 1.9963,
"step": 1205
},
{
"epoch": 0.481629392971246,
"grad_norm": 25.611818313598633,
"learning_rate": 5.2036741214057506e-05,
"loss": 0.4328,
"step": 1206
},
{
"epoch": 0.4820287539936102,
"grad_norm": 11.872817993164062,
"learning_rate": 5.199680511182109e-05,
"loss": 0.0989,
"step": 1207
},
{
"epoch": 0.48242811501597443,
"grad_norm": 26.54136085510254,
"learning_rate": 5.1956869009584665e-05,
"loss": 0.8146,
"step": 1208
},
{
"epoch": 0.48282747603833864,
"grad_norm": 0.2588488757610321,
"learning_rate": 5.191693290734825e-05,
"loss": 0.0021,
"step": 1209
},
{
"epoch": 0.48322683706070285,
"grad_norm": 3.1697306632995605,
"learning_rate": 5.1876996805111824e-05,
"loss": 0.0219,
"step": 1210
},
{
"epoch": 0.4836261980830671,
"grad_norm": 4.142856121063232,
"learning_rate": 5.18370607028754e-05,
"loss": 0.0397,
"step": 1211
},
{
"epoch": 0.4840255591054313,
"grad_norm": 51.31745147705078,
"learning_rate": 5.179712460063898e-05,
"loss": 0.7108,
"step": 1212
},
{
"epoch": 0.48442492012779553,
"grad_norm": 23.48676300048828,
"learning_rate": 5.1757188498402555e-05,
"loss": 0.1673,
"step": 1213
},
{
"epoch": 0.48482428115015974,
"grad_norm": 0.08952261507511139,
"learning_rate": 5.171725239616614e-05,
"loss": 0.0008,
"step": 1214
},
{
"epoch": 0.48522364217252395,
"grad_norm": 15.882736206054688,
"learning_rate": 5.1677316293929714e-05,
"loss": 0.1506,
"step": 1215
},
{
"epoch": 0.48562300319488816,
"grad_norm": 19.235591888427734,
"learning_rate": 5.163738019169329e-05,
"loss": 0.2317,
"step": 1216
},
{
"epoch": 0.4860223642172524,
"grad_norm": 27.37101936340332,
"learning_rate": 5.159744408945687e-05,
"loss": 0.3499,
"step": 1217
},
{
"epoch": 0.48642172523961663,
"grad_norm": 49.29214096069336,
"learning_rate": 5.155750798722045e-05,
"loss": 1.5364,
"step": 1218
},
{
"epoch": 0.48682108626198084,
"grad_norm": 19.736608505249023,
"learning_rate": 5.151757188498403e-05,
"loss": 0.2865,
"step": 1219
},
{
"epoch": 0.48722044728434505,
"grad_norm": 0.11779716610908508,
"learning_rate": 5.1477635782747604e-05,
"loss": 0.0012,
"step": 1220
},
{
"epoch": 0.48761980830670926,
"grad_norm": 33.880489349365234,
"learning_rate": 5.1437699680511184e-05,
"loss": 1.0576,
"step": 1221
},
{
"epoch": 0.48801916932907347,
"grad_norm": 25.510852813720703,
"learning_rate": 5.139776357827476e-05,
"loss": 0.3004,
"step": 1222
},
{
"epoch": 0.4884185303514377,
"grad_norm": 39.90135955810547,
"learning_rate": 5.135782747603834e-05,
"loss": 0.4905,
"step": 1223
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.22687333822250366,
"learning_rate": 5.1317891373801915e-05,
"loss": 0.0013,
"step": 1224
},
{
"epoch": 0.48921725239616615,
"grad_norm": 39.25656509399414,
"learning_rate": 5.12779552715655e-05,
"loss": 0.9249,
"step": 1225
},
{
"epoch": 0.48961661341853036,
"grad_norm": 39.97446060180664,
"learning_rate": 5.1238019169329074e-05,
"loss": 1.3806,
"step": 1226
},
{
"epoch": 0.49001597444089456,
"grad_norm": 12.657002449035645,
"learning_rate": 5.1198083067092654e-05,
"loss": 0.4297,
"step": 1227
},
{
"epoch": 0.4904153354632588,
"grad_norm": 25.82174301147461,
"learning_rate": 5.115814696485623e-05,
"loss": 0.5713,
"step": 1228
},
{
"epoch": 0.490814696485623,
"grad_norm": 14.285473823547363,
"learning_rate": 5.1118210862619806e-05,
"loss": 0.0959,
"step": 1229
},
{
"epoch": 0.49121405750798725,
"grad_norm": 37.94523620605469,
"learning_rate": 5.107827476038339e-05,
"loss": 0.7305,
"step": 1230
},
{
"epoch": 0.49161341853035145,
"grad_norm": 4.337116718292236,
"learning_rate": 5.1038338658146964e-05,
"loss": 0.0889,
"step": 1231
},
{
"epoch": 0.49201277955271566,
"grad_norm": 24.311548233032227,
"learning_rate": 5.0998402555910544e-05,
"loss": 0.4258,
"step": 1232
},
{
"epoch": 0.49241214057507987,
"grad_norm": 28.399940490722656,
"learning_rate": 5.095846645367412e-05,
"loss": 0.5502,
"step": 1233
},
{
"epoch": 0.4928115015974441,
"grad_norm": 13.463953018188477,
"learning_rate": 5.0918530351437696e-05,
"loss": 0.1069,
"step": 1234
},
{
"epoch": 0.4932108626198083,
"grad_norm": 38.6500129699707,
"learning_rate": 5.087859424920128e-05,
"loss": 0.7697,
"step": 1235
},
{
"epoch": 0.4936102236421725,
"grad_norm": 21.22831916809082,
"learning_rate": 5.0838658146964855e-05,
"loss": 0.1619,
"step": 1236
},
{
"epoch": 0.49400958466453676,
"grad_norm": 37.205448150634766,
"learning_rate": 5.079872204472844e-05,
"loss": 0.9225,
"step": 1237
},
{
"epoch": 0.49440894568690097,
"grad_norm": 3.6579792499542236,
"learning_rate": 5.0758785942492014e-05,
"loss": 0.0224,
"step": 1238
},
{
"epoch": 0.4948083067092652,
"grad_norm": 14.0233736038208,
"learning_rate": 5.0718849840255586e-05,
"loss": 0.1082,
"step": 1239
},
{
"epoch": 0.4952076677316294,
"grad_norm": 26.735082626342773,
"learning_rate": 5.067891373801917e-05,
"loss": 0.3731,
"step": 1240
},
{
"epoch": 0.4956070287539936,
"grad_norm": 26.13973045349121,
"learning_rate": 5.0638977635782745e-05,
"loss": 0.5212,
"step": 1241
},
{
"epoch": 0.4960063897763578,
"grad_norm": 22.958791732788086,
"learning_rate": 5.059904153354633e-05,
"loss": 0.3404,
"step": 1242
},
{
"epoch": 0.49640575079872207,
"grad_norm": 17.071779251098633,
"learning_rate": 5.0559105431309904e-05,
"loss": 0.1727,
"step": 1243
},
{
"epoch": 0.4968051118210863,
"grad_norm": 21.934764862060547,
"learning_rate": 5.051916932907349e-05,
"loss": 0.2531,
"step": 1244
},
{
"epoch": 0.4972044728434505,
"grad_norm": 31.47645378112793,
"learning_rate": 5.047923322683706e-05,
"loss": 0.626,
"step": 1245
},
{
"epoch": 0.4976038338658147,
"grad_norm": 19.027381896972656,
"learning_rate": 5.0439297124600635e-05,
"loss": 0.1991,
"step": 1246
},
{
"epoch": 0.4980031948881789,
"grad_norm": 11.361639976501465,
"learning_rate": 5.039936102236422e-05,
"loss": 0.5715,
"step": 1247
},
{
"epoch": 0.4984025559105431,
"grad_norm": 25.184778213500977,
"learning_rate": 5.0359424920127794e-05,
"loss": 0.6916,
"step": 1248
},
{
"epoch": 0.4988019169329074,
"grad_norm": 24.847150802612305,
"learning_rate": 5.031948881789138e-05,
"loss": 0.4774,
"step": 1249
},
{
"epoch": 0.4992012779552716,
"grad_norm": 27.49690055847168,
"learning_rate": 5.027955271565495e-05,
"loss": 0.4375,
"step": 1250
},
{
"epoch": 0.4996006389776358,
"grad_norm": 22.39604949951172,
"learning_rate": 5.0239616613418526e-05,
"loss": 0.2191,
"step": 1251
},
{
"epoch": 0.5,
"grad_norm": 27.56192970275879,
"learning_rate": 5.019968051118211e-05,
"loss": 0.3634,
"step": 1252
},
{
"epoch": 0.5003993610223643,
"grad_norm": 40.96473693847656,
"learning_rate": 5.0159744408945685e-05,
"loss": 0.6317,
"step": 1253
},
{
"epoch": 0.5007987220447284,
"grad_norm": 5.964803218841553,
"learning_rate": 5.011980830670927e-05,
"loss": 0.0347,
"step": 1254
},
{
"epoch": 0.5011980830670927,
"grad_norm": 1.308618426322937,
"learning_rate": 5.0079872204472844e-05,
"loss": 0.0227,
"step": 1255
},
{
"epoch": 0.5015974440894568,
"grad_norm": 12.977692604064941,
"learning_rate": 5.003993610223643e-05,
"loss": 0.0846,
"step": 1256
},
{
"epoch": 0.5019968051118211,
"grad_norm": 24.69575309753418,
"learning_rate": 5e-05,
"loss": 0.3613,
"step": 1257
},
{
"epoch": 0.5023961661341853,
"grad_norm": 20.298664093017578,
"learning_rate": 4.996006389776358e-05,
"loss": 0.3434,
"step": 1258
},
{
"epoch": 0.5027955271565495,
"grad_norm": 3.1088058948516846,
"learning_rate": 4.992012779552716e-05,
"loss": 0.0196,
"step": 1259
},
{
"epoch": 0.5031948881789138,
"grad_norm": 22.786766052246094,
"learning_rate": 4.988019169329074e-05,
"loss": 0.2846,
"step": 1260
},
{
"epoch": 0.5035942492012779,
"grad_norm": 26.23028564453125,
"learning_rate": 4.984025559105431e-05,
"loss": 0.4117,
"step": 1261
},
{
"epoch": 0.5039936102236422,
"grad_norm": 32.440982818603516,
"learning_rate": 4.980031948881789e-05,
"loss": 0.3265,
"step": 1262
},
{
"epoch": 0.5043929712460063,
"grad_norm": 8.709019660949707,
"learning_rate": 4.976038338658147e-05,
"loss": 0.1625,
"step": 1263
},
{
"epoch": 0.5047923322683706,
"grad_norm": 17.11716651916504,
"learning_rate": 4.972044728434505e-05,
"loss": 0.1428,
"step": 1264
},
{
"epoch": 0.5051916932907349,
"grad_norm": 20.467737197875977,
"learning_rate": 4.968051118210863e-05,
"loss": 0.1691,
"step": 1265
},
{
"epoch": 0.505591054313099,
"grad_norm": 32.004390716552734,
"learning_rate": 4.964057507987221e-05,
"loss": 0.4229,
"step": 1266
},
{
"epoch": 0.5059904153354633,
"grad_norm": 18.455202102661133,
"learning_rate": 4.960063897763578e-05,
"loss": 0.5663,
"step": 1267
},
{
"epoch": 0.5063897763578274,
"grad_norm": 18.889680862426758,
"learning_rate": 4.956070287539936e-05,
"loss": 0.203,
"step": 1268
},
{
"epoch": 0.5067891373801917,
"grad_norm": 23.6242733001709,
"learning_rate": 4.952076677316294e-05,
"loss": 0.4201,
"step": 1269
},
{
"epoch": 0.5071884984025559,
"grad_norm": 36.26917266845703,
"learning_rate": 4.948083067092652e-05,
"loss": 0.5398,
"step": 1270
},
{
"epoch": 0.5075878594249201,
"grad_norm": 37.99750900268555,
"learning_rate": 4.94408945686901e-05,
"loss": 0.4277,
"step": 1271
},
{
"epoch": 0.5079872204472844,
"grad_norm": 13.64342975616455,
"learning_rate": 4.940095846645368e-05,
"loss": 0.2022,
"step": 1272
},
{
"epoch": 0.5083865814696485,
"grad_norm": 25.979265213012695,
"learning_rate": 4.936102236421725e-05,
"loss": 0.4744,
"step": 1273
},
{
"epoch": 0.5087859424920128,
"grad_norm": 28.107826232910156,
"learning_rate": 4.932108626198083e-05,
"loss": 1.0235,
"step": 1274
},
{
"epoch": 0.509185303514377,
"grad_norm": 37.00997543334961,
"learning_rate": 4.928115015974441e-05,
"loss": 1.4711,
"step": 1275
},
{
"epoch": 0.5095846645367412,
"grad_norm": 0.6261576414108276,
"learning_rate": 4.924121405750799e-05,
"loss": 0.0033,
"step": 1276
},
{
"epoch": 0.5099840255591054,
"grad_norm": 22.93016242980957,
"learning_rate": 4.920127795527157e-05,
"loss": 0.9088,
"step": 1277
},
{
"epoch": 0.5103833865814696,
"grad_norm": 1.3742833137512207,
"learning_rate": 4.916134185303515e-05,
"loss": 0.0388,
"step": 1278
},
{
"epoch": 0.5107827476038339,
"grad_norm": 30.060697555541992,
"learning_rate": 4.912140575079873e-05,
"loss": 0.285,
"step": 1279
},
{
"epoch": 0.5111821086261981,
"grad_norm": 33.09109115600586,
"learning_rate": 4.90814696485623e-05,
"loss": 2.092,
"step": 1280
},
{
"epoch": 0.5115814696485623,
"grad_norm": 57.27322769165039,
"learning_rate": 4.904153354632588e-05,
"loss": 0.8568,
"step": 1281
},
{
"epoch": 0.5119808306709265,
"grad_norm": 7.3302321434021,
"learning_rate": 4.900159744408946e-05,
"loss": 0.0496,
"step": 1282
},
{
"epoch": 0.5123801916932907,
"grad_norm": 11.084355354309082,
"learning_rate": 4.896166134185304e-05,
"loss": 0.0945,
"step": 1283
},
{
"epoch": 0.512779552715655,
"grad_norm": 16.10188102722168,
"learning_rate": 4.892172523961662e-05,
"loss": 0.4806,
"step": 1284
},
{
"epoch": 0.5131789137380192,
"grad_norm": 1.199759602546692,
"learning_rate": 4.88817891373802e-05,
"loss": 0.0069,
"step": 1285
},
{
"epoch": 0.5135782747603834,
"grad_norm": 27.08547019958496,
"learning_rate": 4.884185303514377e-05,
"loss": 0.2677,
"step": 1286
},
{
"epoch": 0.5139776357827476,
"grad_norm": 4.7502264976501465,
"learning_rate": 4.880191693290735e-05,
"loss": 0.0313,
"step": 1287
},
{
"epoch": 0.5143769968051118,
"grad_norm": 13.190296173095703,
"learning_rate": 4.876198083067093e-05,
"loss": 0.1548,
"step": 1288
},
{
"epoch": 0.514776357827476,
"grad_norm": 31.386079788208008,
"learning_rate": 4.872204472843451e-05,
"loss": 0.3913,
"step": 1289
},
{
"epoch": 0.5151757188498403,
"grad_norm": 24.783611297607422,
"learning_rate": 4.868210862619809e-05,
"loss": 0.6314,
"step": 1290
},
{
"epoch": 0.5155750798722045,
"grad_norm": 29.748727798461914,
"learning_rate": 4.864217252396166e-05,
"loss": 0.3604,
"step": 1291
},
{
"epoch": 0.5159744408945687,
"grad_norm": 11.423151969909668,
"learning_rate": 4.860223642172524e-05,
"loss": 0.198,
"step": 1292
},
{
"epoch": 0.5163738019169329,
"grad_norm": 42.55264663696289,
"learning_rate": 4.856230031948882e-05,
"loss": 0.7766,
"step": 1293
},
{
"epoch": 0.5167731629392971,
"grad_norm": 25.7020320892334,
"learning_rate": 4.85223642172524e-05,
"loss": 0.544,
"step": 1294
},
{
"epoch": 0.5171725239616614,
"grad_norm": 6.718214511871338,
"learning_rate": 4.848242811501597e-05,
"loss": 0.095,
"step": 1295
},
{
"epoch": 0.5175718849840255,
"grad_norm": 32.86326217651367,
"learning_rate": 4.844249201277955e-05,
"loss": 0.4973,
"step": 1296
},
{
"epoch": 0.5179712460063898,
"grad_norm": 10.322693824768066,
"learning_rate": 4.840255591054313e-05,
"loss": 0.0862,
"step": 1297
},
{
"epoch": 0.518370607028754,
"grad_norm": 24.411251068115234,
"learning_rate": 4.836261980830671e-05,
"loss": 0.5856,
"step": 1298
},
{
"epoch": 0.5187699680511182,
"grad_norm": 4.912768840789795,
"learning_rate": 4.832268370607029e-05,
"loss": 0.0372,
"step": 1299
},
{
"epoch": 0.5191693290734825,
"grad_norm": 32.83550262451172,
"learning_rate": 4.8282747603833863e-05,
"loss": 0.6658,
"step": 1300
},
{
"epoch": 0.5195686900958466,
"grad_norm": 4.561255931854248,
"learning_rate": 4.824281150159744e-05,
"loss": 0.0271,
"step": 1301
},
{
"epoch": 0.5199680511182109,
"grad_norm": 3.064560890197754,
"learning_rate": 4.820287539936102e-05,
"loss": 0.0332,
"step": 1302
},
{
"epoch": 0.520367412140575,
"grad_norm": 21.781513214111328,
"learning_rate": 4.81629392971246e-05,
"loss": 0.1717,
"step": 1303
},
{
"epoch": 0.5207667731629393,
"grad_norm": 24.177446365356445,
"learning_rate": 4.812300319488818e-05,
"loss": 0.4691,
"step": 1304
},
{
"epoch": 0.5211661341853036,
"grad_norm": 2.1845664978027344,
"learning_rate": 4.8083067092651754e-05,
"loss": 0.0199,
"step": 1305
},
{
"epoch": 0.5215654952076677,
"grad_norm": 1.711386799812317,
"learning_rate": 4.804313099041533e-05,
"loss": 0.0121,
"step": 1306
},
{
"epoch": 0.521964856230032,
"grad_norm": 30.307846069335938,
"learning_rate": 4.800319488817891e-05,
"loss": 0.5985,
"step": 1307
},
{
"epoch": 0.5223642172523961,
"grad_norm": 21.114858627319336,
"learning_rate": 4.796325878594249e-05,
"loss": 0.2023,
"step": 1308
},
{
"epoch": 0.5227635782747604,
"grad_norm": 44.990352630615234,
"learning_rate": 4.792332268370607e-05,
"loss": 0.696,
"step": 1309
},
{
"epoch": 0.5231629392971247,
"grad_norm": 20.87180519104004,
"learning_rate": 4.788338658146965e-05,
"loss": 0.174,
"step": 1310
},
{
"epoch": 0.5235623003194888,
"grad_norm": 23.277637481689453,
"learning_rate": 4.784345047923323e-05,
"loss": 0.373,
"step": 1311
},
{
"epoch": 0.5239616613418531,
"grad_norm": 30.381942749023438,
"learning_rate": 4.78035143769968e-05,
"loss": 0.6693,
"step": 1312
},
{
"epoch": 0.5243610223642172,
"grad_norm": 8.97925853729248,
"learning_rate": 4.776357827476038e-05,
"loss": 0.193,
"step": 1313
},
{
"epoch": 0.5247603833865815,
"grad_norm": 18.511262893676758,
"learning_rate": 4.772364217252396e-05,
"loss": 0.1555,
"step": 1314
},
{
"epoch": 0.5251597444089456,
"grad_norm": 11.169072151184082,
"learning_rate": 4.768370607028754e-05,
"loss": 0.0937,
"step": 1315
},
{
"epoch": 0.5255591054313099,
"grad_norm": 19.065555572509766,
"learning_rate": 4.764376996805112e-05,
"loss": 0.3113,
"step": 1316
},
{
"epoch": 0.5259584664536742,
"grad_norm": 42.99228286743164,
"learning_rate": 4.76038338658147e-05,
"loss": 0.8504,
"step": 1317
},
{
"epoch": 0.5263578274760383,
"grad_norm": 3.087864398956299,
"learning_rate": 4.756389776357827e-05,
"loss": 0.0171,
"step": 1318
},
{
"epoch": 0.5267571884984026,
"grad_norm": 9.671379089355469,
"learning_rate": 4.752396166134185e-05,
"loss": 0.1205,
"step": 1319
},
{
"epoch": 0.5271565495207667,
"grad_norm": 1.4320189952850342,
"learning_rate": 4.748402555910543e-05,
"loss": 0.0082,
"step": 1320
},
{
"epoch": 0.527555910543131,
"grad_norm": 30.808504104614258,
"learning_rate": 4.744408945686901e-05,
"loss": 0.3048,
"step": 1321
},
{
"epoch": 0.5279552715654952,
"grad_norm": 0.8006933927536011,
"learning_rate": 4.740415335463259e-05,
"loss": 0.0048,
"step": 1322
},
{
"epoch": 0.5283546325878594,
"grad_norm": 0.30146628618240356,
"learning_rate": 4.736421725239617e-05,
"loss": 0.0023,
"step": 1323
},
{
"epoch": 0.5287539936102237,
"grad_norm": 34.035213470458984,
"learning_rate": 4.732428115015975e-05,
"loss": 0.396,
"step": 1324
},
{
"epoch": 0.5291533546325878,
"grad_norm": 34.90943908691406,
"learning_rate": 4.728434504792332e-05,
"loss": 0.5998,
"step": 1325
},
{
"epoch": 0.5295527156549521,
"grad_norm": 10.060013771057129,
"learning_rate": 4.72444089456869e-05,
"loss": 0.0584,
"step": 1326
},
{
"epoch": 0.5299520766773163,
"grad_norm": 2.5388710498809814,
"learning_rate": 4.720447284345048e-05,
"loss": 0.0139,
"step": 1327
},
{
"epoch": 0.5303514376996805,
"grad_norm": 17.229110717773438,
"learning_rate": 4.716453674121406e-05,
"loss": 0.4454,
"step": 1328
},
{
"epoch": 0.5307507987220448,
"grad_norm": 21.065710067749023,
"learning_rate": 4.712460063897764e-05,
"loss": 0.2217,
"step": 1329
},
{
"epoch": 0.5311501597444089,
"grad_norm": 1.843264102935791,
"learning_rate": 4.708466453674122e-05,
"loss": 0.0103,
"step": 1330
},
{
"epoch": 0.5315495207667732,
"grad_norm": 6.246510028839111,
"learning_rate": 4.704472843450479e-05,
"loss": 0.046,
"step": 1331
},
{
"epoch": 0.5319488817891374,
"grad_norm": 28.564393997192383,
"learning_rate": 4.700479233226837e-05,
"loss": 0.7513,
"step": 1332
},
{
"epoch": 0.5323482428115016,
"grad_norm": 18.374895095825195,
"learning_rate": 4.696485623003195e-05,
"loss": 0.9909,
"step": 1333
},
{
"epoch": 0.5327476038338658,
"grad_norm": 1.6738381385803223,
"learning_rate": 4.692492012779553e-05,
"loss": 0.0242,
"step": 1334
},
{
"epoch": 0.53314696485623,
"grad_norm": 46.157527923583984,
"learning_rate": 4.688498402555911e-05,
"loss": 0.8141,
"step": 1335
},
{
"epoch": 0.5335463258785943,
"grad_norm": 21.654476165771484,
"learning_rate": 4.684504792332269e-05,
"loss": 0.1834,
"step": 1336
},
{
"epoch": 0.5339456869009584,
"grad_norm": 0.6640011668205261,
"learning_rate": 4.680511182108626e-05,
"loss": 0.0072,
"step": 1337
},
{
"epoch": 0.5343450479233227,
"grad_norm": 11.273137092590332,
"learning_rate": 4.676517571884984e-05,
"loss": 0.1661,
"step": 1338
},
{
"epoch": 0.5347444089456869,
"grad_norm": 35.46974563598633,
"learning_rate": 4.672523961661342e-05,
"loss": 0.8334,
"step": 1339
},
{
"epoch": 0.5351437699680511,
"grad_norm": 27.042808532714844,
"learning_rate": 4.6685303514377e-05,
"loss": 0.8772,
"step": 1340
},
{
"epoch": 0.5355431309904153,
"grad_norm": 46.287330627441406,
"learning_rate": 4.664536741214058e-05,
"loss": 0.5079,
"step": 1341
},
{
"epoch": 0.5359424920127795,
"grad_norm": 30.71518898010254,
"learning_rate": 4.660543130990416e-05,
"loss": 0.8339,
"step": 1342
},
{
"epoch": 0.5363418530351438,
"grad_norm": 20.562721252441406,
"learning_rate": 4.656549520766774e-05,
"loss": 0.2124,
"step": 1343
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.18902671337127686,
"learning_rate": 4.652555910543131e-05,
"loss": 0.001,
"step": 1344
},
{
"epoch": 0.5371405750798722,
"grad_norm": 19.37459945678711,
"learning_rate": 4.648562300319489e-05,
"loss": 0.8127,
"step": 1345
},
{
"epoch": 0.5375399361022364,
"grad_norm": 12.480849266052246,
"learning_rate": 4.644568690095847e-05,
"loss": 0.0784,
"step": 1346
},
{
"epoch": 0.5379392971246006,
"grad_norm": 0.012376879341900349,
"learning_rate": 4.640575079872205e-05,
"loss": 0.0001,
"step": 1347
},
{
"epoch": 0.5383386581469649,
"grad_norm": 0.43085968494415283,
"learning_rate": 4.636581469648563e-05,
"loss": 0.0028,
"step": 1348
},
{
"epoch": 0.5387380191693291,
"grad_norm": 23.206520080566406,
"learning_rate": 4.632587859424921e-05,
"loss": 0.3025,
"step": 1349
},
{
"epoch": 0.5391373801916933,
"grad_norm": 0.9590831398963928,
"learning_rate": 4.628594249201278e-05,
"loss": 0.0076,
"step": 1350
},
{
"epoch": 0.5395367412140575,
"grad_norm": 42.57503890991211,
"learning_rate": 4.624600638977636e-05,
"loss": 0.584,
"step": 1351
},
{
"epoch": 0.5399361022364217,
"grad_norm": 28.014028549194336,
"learning_rate": 4.620607028753994e-05,
"loss": 0.2826,
"step": 1352
},
{
"epoch": 0.5403354632587859,
"grad_norm": 15.43281078338623,
"learning_rate": 4.616613418530352e-05,
"loss": 0.3655,
"step": 1353
},
{
"epoch": 0.5407348242811502,
"grad_norm": 1.2934383153915405,
"learning_rate": 4.61261980830671e-05,
"loss": 0.0078,
"step": 1354
},
{
"epoch": 0.5411341853035144,
"grad_norm": 26.0319881439209,
"learning_rate": 4.608626198083068e-05,
"loss": 0.625,
"step": 1355
},
{
"epoch": 0.5415335463258786,
"grad_norm": 2.6775145530700684,
"learning_rate": 4.604632587859426e-05,
"loss": 0.0193,
"step": 1356
},
{
"epoch": 0.5419329073482428,
"grad_norm": 21.108510971069336,
"learning_rate": 4.600638977635783e-05,
"loss": 0.2092,
"step": 1357
},
{
"epoch": 0.542332268370607,
"grad_norm": 38.54832077026367,
"learning_rate": 4.596645367412141e-05,
"loss": 0.4039,
"step": 1358
},
{
"epoch": 0.5427316293929713,
"grad_norm": 23.956266403198242,
"learning_rate": 4.592651757188499e-05,
"loss": 0.1523,
"step": 1359
},
{
"epoch": 0.5431309904153354,
"grad_norm": 25.855113983154297,
"learning_rate": 4.588658146964857e-05,
"loss": 0.4265,
"step": 1360
},
{
"epoch": 0.5435303514376997,
"grad_norm": 28.361595153808594,
"learning_rate": 4.584664536741215e-05,
"loss": 0.5631,
"step": 1361
},
{
"epoch": 0.5439297124600639,
"grad_norm": 35.72830581665039,
"learning_rate": 4.580670926517572e-05,
"loss": 0.8676,
"step": 1362
},
{
"epoch": 0.5443290734824281,
"grad_norm": 25.712142944335938,
"learning_rate": 4.57667731629393e-05,
"loss": 0.1763,
"step": 1363
},
{
"epoch": 0.5447284345047924,
"grad_norm": 15.9559965133667,
"learning_rate": 4.572683706070288e-05,
"loss": 0.2417,
"step": 1364
},
{
"epoch": 0.5451277955271565,
"grad_norm": 7.230056285858154,
"learning_rate": 4.568690095846646e-05,
"loss": 0.0769,
"step": 1365
},
{
"epoch": 0.5455271565495208,
"grad_norm": 11.801989555358887,
"learning_rate": 4.564696485623004e-05,
"loss": 0.0724,
"step": 1366
},
{
"epoch": 0.545926517571885,
"grad_norm": 34.40201950073242,
"learning_rate": 4.560702875399361e-05,
"loss": 0.2993,
"step": 1367
},
{
"epoch": 0.5463258785942492,
"grad_norm": 21.97010040283203,
"learning_rate": 4.556709265175719e-05,
"loss": 0.1921,
"step": 1368
},
{
"epoch": 0.5467252396166135,
"grad_norm": 6.390224933624268,
"learning_rate": 4.552715654952077e-05,
"loss": 0.0335,
"step": 1369
},
{
"epoch": 0.5471246006389776,
"grad_norm": 2.4946513175964355,
"learning_rate": 4.548722044728435e-05,
"loss": 0.0209,
"step": 1370
},
{
"epoch": 0.5475239616613419,
"grad_norm": 51.2524299621582,
"learning_rate": 4.544728434504792e-05,
"loss": 0.8078,
"step": 1371
},
{
"epoch": 0.547923322683706,
"grad_norm": 48.9635009765625,
"learning_rate": 4.54073482428115e-05,
"loss": 0.7223,
"step": 1372
},
{
"epoch": 0.5483226837060703,
"grad_norm": 2.169365406036377,
"learning_rate": 4.536741214057508e-05,
"loss": 0.0125,
"step": 1373
},
{
"epoch": 0.5487220447284346,
"grad_norm": 3.219621181488037,
"learning_rate": 4.532747603833866e-05,
"loss": 0.0412,
"step": 1374
},
{
"epoch": 0.5491214057507987,
"grad_norm": 32.73312759399414,
"learning_rate": 4.528753993610224e-05,
"loss": 1.0415,
"step": 1375
},
{
"epoch": 0.549520766773163,
"grad_norm": 14.35698127746582,
"learning_rate": 4.524760383386581e-05,
"loss": 0.0917,
"step": 1376
},
{
"epoch": 0.5499201277955271,
"grad_norm": 13.264641761779785,
"learning_rate": 4.520766773162939e-05,
"loss": 0.0829,
"step": 1377
},
{
"epoch": 0.5503194888178914,
"grad_norm": 48.43290328979492,
"learning_rate": 4.516773162939297e-05,
"loss": 0.8411,
"step": 1378
},
{
"epoch": 0.5507188498402555,
"grad_norm": 22.598377227783203,
"learning_rate": 4.512779552715655e-05,
"loss": 0.335,
"step": 1379
},
{
"epoch": 0.5511182108626198,
"grad_norm": 43.748111724853516,
"learning_rate": 4.508785942492013e-05,
"loss": 0.61,
"step": 1380
},
{
"epoch": 0.5515175718849841,
"grad_norm": 25.78108024597168,
"learning_rate": 4.504792332268371e-05,
"loss": 0.224,
"step": 1381
},
{
"epoch": 0.5519169329073482,
"grad_norm": 3.2784969806671143,
"learning_rate": 4.500798722044728e-05,
"loss": 0.0186,
"step": 1382
},
{
"epoch": 0.5523162939297125,
"grad_norm": 40.26729965209961,
"learning_rate": 4.496805111821086e-05,
"loss": 0.6477,
"step": 1383
},
{
"epoch": 0.5527156549520766,
"grad_norm": 12.913558959960938,
"learning_rate": 4.492811501597444e-05,
"loss": 0.0969,
"step": 1384
},
{
"epoch": 0.5531150159744409,
"grad_norm": 0.9272196292877197,
"learning_rate": 4.488817891373802e-05,
"loss": 0.0181,
"step": 1385
},
{
"epoch": 0.5535143769968051,
"grad_norm": 0.2664625346660614,
"learning_rate": 4.48482428115016e-05,
"loss": 0.0022,
"step": 1386
},
{
"epoch": 0.5539137380191693,
"grad_norm": 11.596216201782227,
"learning_rate": 4.480830670926518e-05,
"loss": 0.0791,
"step": 1387
},
{
"epoch": 0.5543130990415336,
"grad_norm": 7.174932956695557,
"learning_rate": 4.476837060702875e-05,
"loss": 0.037,
"step": 1388
},
{
"epoch": 0.5547124600638977,
"grad_norm": 24.8929443359375,
"learning_rate": 4.472843450479233e-05,
"loss": 0.7592,
"step": 1389
},
{
"epoch": 0.555111821086262,
"grad_norm": 4.477590084075928,
"learning_rate": 4.468849840255591e-05,
"loss": 0.0261,
"step": 1390
},
{
"epoch": 0.5555111821086262,
"grad_norm": 9.795445442199707,
"learning_rate": 4.464856230031949e-05,
"loss": 0.12,
"step": 1391
},
{
"epoch": 0.5559105431309904,
"grad_norm": 12.273335456848145,
"learning_rate": 4.460862619808307e-05,
"loss": 0.4001,
"step": 1392
},
{
"epoch": 0.5563099041533547,
"grad_norm": 38.01393508911133,
"learning_rate": 4.456869009584665e-05,
"loss": 0.4045,
"step": 1393
},
{
"epoch": 0.5567092651757188,
"grad_norm": 20.879480361938477,
"learning_rate": 4.452875399361023e-05,
"loss": 0.1923,
"step": 1394
},
{
"epoch": 0.5571086261980831,
"grad_norm": 9.3219575881958,
"learning_rate": 4.44888178913738e-05,
"loss": 0.1251,
"step": 1395
},
{
"epoch": 0.5575079872204473,
"grad_norm": 19.19240951538086,
"learning_rate": 4.444888178913738e-05,
"loss": 0.2498,
"step": 1396
},
{
"epoch": 0.5579073482428115,
"grad_norm": 36.68465805053711,
"learning_rate": 4.440894568690096e-05,
"loss": 1.2347,
"step": 1397
},
{
"epoch": 0.5583067092651757,
"grad_norm": 1.7589894533157349,
"learning_rate": 4.436900958466454e-05,
"loss": 0.009,
"step": 1398
},
{
"epoch": 0.5587060702875399,
"grad_norm": 1.40755295753479,
"learning_rate": 4.432907348242812e-05,
"loss": 0.0195,
"step": 1399
},
{
"epoch": 0.5591054313099042,
"grad_norm": 27.077930450439453,
"learning_rate": 4.42891373801917e-05,
"loss": 0.5341,
"step": 1400
},
{
"epoch": 0.5595047923322684,
"grad_norm": 0.2557763457298279,
"learning_rate": 4.424920127795527e-05,
"loss": 0.0013,
"step": 1401
},
{
"epoch": 0.5599041533546326,
"grad_norm": 24.573793411254883,
"learning_rate": 4.420926517571885e-05,
"loss": 0.5151,
"step": 1402
},
{
"epoch": 0.5603035143769968,
"grad_norm": 3.0533859729766846,
"learning_rate": 4.416932907348243e-05,
"loss": 0.0164,
"step": 1403
},
{
"epoch": 0.560702875399361,
"grad_norm": 0.029383031651377678,
"learning_rate": 4.412939297124601e-05,
"loss": 0.0002,
"step": 1404
},
{
"epoch": 0.5611022364217252,
"grad_norm": 12.84772777557373,
"learning_rate": 4.408945686900959e-05,
"loss": 0.0989,
"step": 1405
},
{
"epoch": 0.5615015974440895,
"grad_norm": 33.69590759277344,
"learning_rate": 4.404952076677317e-05,
"loss": 0.7344,
"step": 1406
},
{
"epoch": 0.5619009584664537,
"grad_norm": 33.0693473815918,
"learning_rate": 4.400958466453675e-05,
"loss": 0.7767,
"step": 1407
},
{
"epoch": 0.5623003194888179,
"grad_norm": 1.4902124404907227,
"learning_rate": 4.396964856230032e-05,
"loss": 0.0086,
"step": 1408
},
{
"epoch": 0.5626996805111821,
"grad_norm": 18.264385223388672,
"learning_rate": 4.39297124600639e-05,
"loss": 0.1535,
"step": 1409
},
{
"epoch": 0.5630990415335463,
"grad_norm": 28.086566925048828,
"learning_rate": 4.388977635782748e-05,
"loss": 1.5251,
"step": 1410
},
{
"epoch": 0.5634984025559105,
"grad_norm": 16.533166885375977,
"learning_rate": 4.384984025559106e-05,
"loss": 0.1068,
"step": 1411
},
{
"epoch": 0.5638977635782748,
"grad_norm": 21.53447914123535,
"learning_rate": 4.380990415335464e-05,
"loss": 0.2215,
"step": 1412
},
{
"epoch": 0.564297124600639,
"grad_norm": 10.24766731262207,
"learning_rate": 4.376996805111822e-05,
"loss": 0.0847,
"step": 1413
},
{
"epoch": 0.5646964856230032,
"grad_norm": 4.010213375091553,
"learning_rate": 4.373003194888179e-05,
"loss": 0.0201,
"step": 1414
},
{
"epoch": 0.5650958466453674,
"grad_norm": 0.3136048913002014,
"learning_rate": 4.369009584664537e-05,
"loss": 0.0021,
"step": 1415
},
{
"epoch": 0.5654952076677316,
"grad_norm": 23.668975830078125,
"learning_rate": 4.365015974440895e-05,
"loss": 0.3916,
"step": 1416
},
{
"epoch": 0.5658945686900958,
"grad_norm": 13.748516082763672,
"learning_rate": 4.361022364217253e-05,
"loss": 0.0905,
"step": 1417
},
{
"epoch": 0.5662939297124601,
"grad_norm": 12.847017288208008,
"learning_rate": 4.357028753993611e-05,
"loss": 0.1363,
"step": 1418
},
{
"epoch": 0.5666932907348243,
"grad_norm": 12.362814903259277,
"learning_rate": 4.3530351437699686e-05,
"loss": 0.0926,
"step": 1419
},
{
"epoch": 0.5670926517571885,
"grad_norm": 15.444024085998535,
"learning_rate": 4.349041533546326e-05,
"loss": 0.888,
"step": 1420
},
{
"epoch": 0.5674920127795527,
"grad_norm": 17.38373565673828,
"learning_rate": 4.345047923322684e-05,
"loss": 0.1372,
"step": 1421
},
{
"epoch": 0.5678913738019169,
"grad_norm": 27.058515548706055,
"learning_rate": 4.341054313099042e-05,
"loss": 0.2393,
"step": 1422
},
{
"epoch": 0.5682907348242812,
"grad_norm": 8.611419677734375,
"learning_rate": 4.3370607028754e-05,
"loss": 0.079,
"step": 1423
},
{
"epoch": 0.5686900958466453,
"grad_norm": 22.011167526245117,
"learning_rate": 4.333067092651758e-05,
"loss": 0.4453,
"step": 1424
},
{
"epoch": 0.5690894568690096,
"grad_norm": 7.475895881652832,
"learning_rate": 4.3290734824281156e-05,
"loss": 0.0578,
"step": 1425
},
{
"epoch": 0.5694888178913738,
"grad_norm": 22.1317195892334,
"learning_rate": 4.3250798722044736e-05,
"loss": 0.3887,
"step": 1426
},
{
"epoch": 0.569888178913738,
"grad_norm": 0.8723806738853455,
"learning_rate": 4.321086261980831e-05,
"loss": 0.006,
"step": 1427
},
{
"epoch": 0.5702875399361023,
"grad_norm": 36.70941162109375,
"learning_rate": 4.317092651757189e-05,
"loss": 1.0958,
"step": 1428
},
{
"epoch": 0.5706869009584664,
"grad_norm": 37.33980941772461,
"learning_rate": 4.313099041533547e-05,
"loss": 0.3687,
"step": 1429
},
{
"epoch": 0.5710862619808307,
"grad_norm": 10.4573392868042,
"learning_rate": 4.3091054313099047e-05,
"loss": 0.0605,
"step": 1430
},
{
"epoch": 0.5714856230031949,
"grad_norm": 50.37876510620117,
"learning_rate": 4.3051118210862626e-05,
"loss": 1.6463,
"step": 1431
},
{
"epoch": 0.5718849840255591,
"grad_norm": 24.09528350830078,
"learning_rate": 4.3011182108626205e-05,
"loss": 0.286,
"step": 1432
},
{
"epoch": 0.5722843450479234,
"grad_norm": 0.8495061993598938,
"learning_rate": 4.297124600638978e-05,
"loss": 0.0049,
"step": 1433
},
{
"epoch": 0.5726837060702875,
"grad_norm": 2.339578866958618,
"learning_rate": 4.293130990415336e-05,
"loss": 0.0132,
"step": 1434
},
{
"epoch": 0.5730830670926518,
"grad_norm": 1.481113314628601,
"learning_rate": 4.289137380191694e-05,
"loss": 0.024,
"step": 1435
},
{
"epoch": 0.5734824281150159,
"grad_norm": 0.5454221963882446,
"learning_rate": 4.2851437699680516e-05,
"loss": 0.0031,
"step": 1436
},
{
"epoch": 0.5738817891373802,
"grad_norm": 0.9217386841773987,
"learning_rate": 4.2811501597444096e-05,
"loss": 0.0053,
"step": 1437
},
{
"epoch": 0.5742811501597445,
"grad_norm": 24.213302612304688,
"learning_rate": 4.277156549520767e-05,
"loss": 0.4671,
"step": 1438
},
{
"epoch": 0.5746805111821086,
"grad_norm": 46.556060791015625,
"learning_rate": 4.273162939297125e-05,
"loss": 1.0893,
"step": 1439
},
{
"epoch": 0.5750798722044729,
"grad_norm": NaN,
"learning_rate": 4.273162939297125e-05,
"loss": 0.0,
"step": 1440
},
{
"epoch": 0.575479233226837,
"grad_norm": 6.336025714874268,
"learning_rate": 4.269169329073483e-05,
"loss": 0.0522,
"step": 1441
},
{
"epoch": 0.5758785942492013,
"grad_norm": 26.907669067382812,
"learning_rate": 4.265175718849841e-05,
"loss": 0.6891,
"step": 1442
},
{
"epoch": 0.5762779552715654,
"grad_norm": 13.385482788085938,
"learning_rate": 4.261182108626198e-05,
"loss": 0.092,
"step": 1443
},
{
"epoch": 0.5766773162939297,
"grad_norm": 46.197654724121094,
"learning_rate": 4.257188498402556e-05,
"loss": 1.1802,
"step": 1444
},
{
"epoch": 0.577076677316294,
"grad_norm": 20.579181671142578,
"learning_rate": 4.253194888178914e-05,
"loss": 0.1479,
"step": 1445
},
{
"epoch": 0.5774760383386581,
"grad_norm": 12.165633201599121,
"learning_rate": 4.249201277955272e-05,
"loss": 0.0724,
"step": 1446
},
{
"epoch": 0.5778753993610224,
"grad_norm": 1.5199843645095825,
"learning_rate": 4.24520766773163e-05,
"loss": 0.0153,
"step": 1447
},
{
"epoch": 0.5782747603833865,
"grad_norm": 30.970157623291016,
"learning_rate": 4.241214057507987e-05,
"loss": 0.6615,
"step": 1448
},
{
"epoch": 0.5786741214057508,
"grad_norm": 20.33168601989746,
"learning_rate": 4.237220447284345e-05,
"loss": 0.1946,
"step": 1449
},
{
"epoch": 0.579073482428115,
"grad_norm": 0.12380415201187134,
"learning_rate": 4.233226837060703e-05,
"loss": 0.0008,
"step": 1450
},
{
"epoch": 0.5794728434504792,
"grad_norm": 30.688859939575195,
"learning_rate": 4.229233226837061e-05,
"loss": 0.5372,
"step": 1451
},
{
"epoch": 0.5798722044728435,
"grad_norm": 2.5717713832855225,
"learning_rate": 4.225239616613419e-05,
"loss": 0.0144,
"step": 1452
},
{
"epoch": 0.5802715654952076,
"grad_norm": 8.184505462646484,
"learning_rate": 4.221246006389776e-05,
"loss": 0.043,
"step": 1453
},
{
"epoch": 0.5806709265175719,
"grad_norm": 12.167875289916992,
"learning_rate": 4.217252396166134e-05,
"loss": 0.0756,
"step": 1454
},
{
"epoch": 0.5810702875399361,
"grad_norm": 22.49947738647461,
"learning_rate": 4.213258785942492e-05,
"loss": 0.1993,
"step": 1455
},
{
"epoch": 0.5814696485623003,
"grad_norm": 27.025365829467773,
"learning_rate": 4.20926517571885e-05,
"loss": 0.2075,
"step": 1456
},
{
"epoch": 0.5818690095846646,
"grad_norm": 49.33864974975586,
"learning_rate": 4.205271565495208e-05,
"loss": 1.5733,
"step": 1457
},
{
"epoch": 0.5822683706070287,
"grad_norm": 13.533209800720215,
"learning_rate": 4.201277955271566e-05,
"loss": 0.0911,
"step": 1458
},
{
"epoch": 0.582667731629393,
"grad_norm": 26.09443473815918,
"learning_rate": 4.1972843450479237e-05,
"loss": 0.5691,
"step": 1459
},
{
"epoch": 0.5830670926517572,
"grad_norm": 0.24550671875476837,
"learning_rate": 4.193290734824281e-05,
"loss": 0.0013,
"step": 1460
},
{
"epoch": 0.5834664536741214,
"grad_norm": 12.937807083129883,
"learning_rate": 4.189297124600639e-05,
"loss": 0.0801,
"step": 1461
},
{
"epoch": 0.5838658146964856,
"grad_norm": 11.654365539550781,
"learning_rate": 4.185303514376997e-05,
"loss": 0.0911,
"step": 1462
},
{
"epoch": 0.5842651757188498,
"grad_norm": 8.658635139465332,
"learning_rate": 4.181309904153355e-05,
"loss": 0.0539,
"step": 1463
},
{
"epoch": 0.5846645367412141,
"grad_norm": 26.34325408935547,
"learning_rate": 4.177316293929713e-05,
"loss": 0.4571,
"step": 1464
},
{
"epoch": 0.5850638977635783,
"grad_norm": 30.63438606262207,
"learning_rate": 4.1733226837060706e-05,
"loss": 0.5466,
"step": 1465
},
{
"epoch": 0.5854632587859425,
"grad_norm": 34.542110443115234,
"learning_rate": 4.169329073482428e-05,
"loss": 0.6331,
"step": 1466
},
{
"epoch": 0.5858626198083067,
"grad_norm": 5.691025257110596,
"learning_rate": 4.165335463258786e-05,
"loss": 0.0398,
"step": 1467
},
{
"epoch": 0.5862619808306709,
"grad_norm": 35.501220703125,
"learning_rate": 4.161341853035144e-05,
"loss": 0.509,
"step": 1468
},
{
"epoch": 0.5866613418530351,
"grad_norm": 15.967216491699219,
"learning_rate": 4.157348242811502e-05,
"loss": 0.1074,
"step": 1469
},
{
"epoch": 0.5870607028753994,
"grad_norm": 26.530044555664062,
"learning_rate": 4.15335463258786e-05,
"loss": 0.5196,
"step": 1470
},
{
"epoch": 0.5874600638977636,
"grad_norm": 22.76455307006836,
"learning_rate": 4.1493610223642176e-05,
"loss": 0.1818,
"step": 1471
},
{
"epoch": 0.5878594249201278,
"grad_norm": 22.889503479003906,
"learning_rate": 4.145367412140575e-05,
"loss": 0.23,
"step": 1472
},
{
"epoch": 0.588258785942492,
"grad_norm": 23.3043155670166,
"learning_rate": 4.141373801916933e-05,
"loss": 0.3825,
"step": 1473
},
{
"epoch": 0.5886581469648562,
"grad_norm": 14.3019380569458,
"learning_rate": 4.137380191693291e-05,
"loss": 0.0936,
"step": 1474
},
{
"epoch": 0.5890575079872205,
"grad_norm": 23.964679718017578,
"learning_rate": 4.133386581469649e-05,
"loss": 0.3161,
"step": 1475
},
{
"epoch": 0.5894568690095847,
"grad_norm": 18.75373649597168,
"learning_rate": 4.1293929712460066e-05,
"loss": 0.5941,
"step": 1476
},
{
"epoch": 0.5898562300319489,
"grad_norm": 40.7902946472168,
"learning_rate": 4.1253993610223646e-05,
"loss": 0.628,
"step": 1477
},
{
"epoch": 0.5902555910543131,
"grad_norm": 1.2654833793640137,
"learning_rate": 4.1214057507987225e-05,
"loss": 0.0255,
"step": 1478
},
{
"epoch": 0.5906549520766773,
"grad_norm": 28.90562629699707,
"learning_rate": 4.11741214057508e-05,
"loss": 0.5608,
"step": 1479
},
{
"epoch": 0.5910543130990416,
"grad_norm": 17.284496307373047,
"learning_rate": 4.113418530351438e-05,
"loss": 0.2156,
"step": 1480
},
{
"epoch": 0.5914536741214057,
"grad_norm": 7.415022373199463,
"learning_rate": 4.109424920127796e-05,
"loss": 0.0404,
"step": 1481
},
{
"epoch": 0.59185303514377,
"grad_norm": 9.984014511108398,
"learning_rate": 4.1054313099041536e-05,
"loss": 0.0695,
"step": 1482
},
{
"epoch": 0.5922523961661342,
"grad_norm": 65.0386962890625,
"learning_rate": 4.1014376996805116e-05,
"loss": 1.6398,
"step": 1483
},
{
"epoch": 0.5926517571884984,
"grad_norm": 18.653770446777344,
"learning_rate": 4.0974440894568695e-05,
"loss": 0.3615,
"step": 1484
},
{
"epoch": 0.5930511182108626,
"grad_norm": 8.59886646270752,
"learning_rate": 4.093450479233227e-05,
"loss": 0.0484,
"step": 1485
},
{
"epoch": 0.5934504792332268,
"grad_norm": 31.509521484375,
"learning_rate": 4.089456869009585e-05,
"loss": 1.4153,
"step": 1486
},
{
"epoch": 0.5938498402555911,
"grad_norm": 23.865476608276367,
"learning_rate": 4.0854632587859427e-05,
"loss": 0.3557,
"step": 1487
},
{
"epoch": 0.5942492012779552,
"grad_norm": 11.635843276977539,
"learning_rate": 4.0814696485623006e-05,
"loss": 0.0867,
"step": 1488
},
{
"epoch": 0.5946485623003195,
"grad_norm": 53.931034088134766,
"learning_rate": 4.0774760383386585e-05,
"loss": 1.7982,
"step": 1489
},
{
"epoch": 0.5950479233226837,
"grad_norm": 0.8742843866348267,
"learning_rate": 4.0734824281150165e-05,
"loss": 0.0053,
"step": 1490
},
{
"epoch": 0.5954472843450479,
"grad_norm": 4.345067024230957,
"learning_rate": 4.0694888178913744e-05,
"loss": 0.0353,
"step": 1491
},
{
"epoch": 0.5958466453674122,
"grad_norm": 17.150531768798828,
"learning_rate": 4.065495207667732e-05,
"loss": 0.3784,
"step": 1492
},
{
"epoch": 0.5962460063897763,
"grad_norm": 19.932376861572266,
"learning_rate": 4.0615015974440896e-05,
"loss": 0.1793,
"step": 1493
},
{
"epoch": 0.5966453674121406,
"grad_norm": 24.488161087036133,
"learning_rate": 4.0575079872204476e-05,
"loss": 0.5384,
"step": 1494
},
{
"epoch": 0.5970447284345048,
"grad_norm": 4.809638977050781,
"learning_rate": 4.0535143769968055e-05,
"loss": 0.0257,
"step": 1495
},
{
"epoch": 0.597444089456869,
"grad_norm": 30.255441665649414,
"learning_rate": 4.0495207667731635e-05,
"loss": 0.3865,
"step": 1496
},
{
"epoch": 0.5978434504792333,
"grad_norm": 5.52847957611084,
"learning_rate": 4.0455271565495214e-05,
"loss": 0.032,
"step": 1497
},
{
"epoch": 0.5982428115015974,
"grad_norm": 1.2699713706970215,
"learning_rate": 4.041533546325879e-05,
"loss": 0.0076,
"step": 1498
},
{
"epoch": 0.5986421725239617,
"grad_norm": 33.55049514770508,
"learning_rate": 4.0375399361022366e-05,
"loss": 0.4309,
"step": 1499
},
{
"epoch": 0.5990415335463258,
"grad_norm": 37.890071868896484,
"learning_rate": 4.0335463258785946e-05,
"loss": 0.6186,
"step": 1500
},
{
"epoch": 0.5994408945686901,
"grad_norm": 31.14183807373047,
"learning_rate": 4.0295527156549525e-05,
"loss": 0.4699,
"step": 1501
},
{
"epoch": 0.5998402555910544,
"grad_norm": 1.4622600078582764,
"learning_rate": 4.0255591054313104e-05,
"loss": 0.008,
"step": 1502
},
{
"epoch": 0.6002396166134185,
"grad_norm": 1.1340428590774536,
"learning_rate": 4.0215654952076684e-05,
"loss": 0.0063,
"step": 1503
},
{
"epoch": 0.6006389776357828,
"grad_norm": 26.51593017578125,
"learning_rate": 4.0175718849840257e-05,
"loss": 0.1936,
"step": 1504
},
{
"epoch": 0.6010383386581469,
"grad_norm": 7.85975456237793,
"learning_rate": 4.0135782747603836e-05,
"loss": 0.0514,
"step": 1505
},
{
"epoch": 0.6014376996805112,
"grad_norm": 45.6800537109375,
"learning_rate": 4.0095846645367415e-05,
"loss": 1.4469,
"step": 1506
},
{
"epoch": 0.6018370607028753,
"grad_norm": 40.89308166503906,
"learning_rate": 4.0055910543130995e-05,
"loss": 0.5633,
"step": 1507
},
{
"epoch": 0.6022364217252396,
"grad_norm": 62.10407638549805,
"learning_rate": 4.0015974440894574e-05,
"loss": 1.1645,
"step": 1508
},
{
"epoch": 0.6026357827476039,
"grad_norm": 64.09742736816406,
"learning_rate": 3.9976038338658154e-05,
"loss": 0.7816,
"step": 1509
},
{
"epoch": 0.603035143769968,
"grad_norm": 23.903287887573242,
"learning_rate": 3.9936102236421726e-05,
"loss": 0.2286,
"step": 1510
},
{
"epoch": 0.6034345047923323,
"grad_norm": 15.465903282165527,
"learning_rate": 3.9896166134185306e-05,
"loss": 0.6747,
"step": 1511
},
{
"epoch": 0.6038338658146964,
"grad_norm": 25.007946014404297,
"learning_rate": 3.9856230031948885e-05,
"loss": 0.2692,
"step": 1512
},
{
"epoch": 0.6042332268370607,
"grad_norm": 13.803542137145996,
"learning_rate": 3.9816293929712465e-05,
"loss": 0.4269,
"step": 1513
},
{
"epoch": 0.604632587859425,
"grad_norm": 24.897754669189453,
"learning_rate": 3.9776357827476044e-05,
"loss": 0.5073,
"step": 1514
},
{
"epoch": 0.6050319488817891,
"grad_norm": 16.646692276000977,
"learning_rate": 3.973642172523962e-05,
"loss": 0.1123,
"step": 1515
},
{
"epoch": 0.6054313099041534,
"grad_norm": 25.96765899658203,
"learning_rate": 3.9696485623003196e-05,
"loss": 0.4376,
"step": 1516
},
{
"epoch": 0.6058306709265175,
"grad_norm": 33.86484909057617,
"learning_rate": 3.9656549520766775e-05,
"loss": 1.093,
"step": 1517
},
{
"epoch": 0.6062300319488818,
"grad_norm": 22.27755355834961,
"learning_rate": 3.9616613418530355e-05,
"loss": 0.2376,
"step": 1518
},
{
"epoch": 0.606629392971246,
"grad_norm": 22.464330673217773,
"learning_rate": 3.957667731629393e-05,
"loss": 0.2449,
"step": 1519
},
{
"epoch": 0.6070287539936102,
"grad_norm": 12.433492660522461,
"learning_rate": 3.953674121405751e-05,
"loss": 0.0668,
"step": 1520
},
{
"epoch": 0.6074281150159745,
"grad_norm": 45.56315994262695,
"learning_rate": 3.9496805111821086e-05,
"loss": 0.4779,
"step": 1521
},
{
"epoch": 0.6078274760383386,
"grad_norm": 0.8806638121604919,
"learning_rate": 3.9456869009584666e-05,
"loss": 0.0081,
"step": 1522
},
{
"epoch": 0.6082268370607029,
"grad_norm": 20.422027587890625,
"learning_rate": 3.9416932907348245e-05,
"loss": 0.3354,
"step": 1523
},
{
"epoch": 0.6086261980830671,
"grad_norm": 6.575976848602295,
"learning_rate": 3.937699680511182e-05,
"loss": 0.04,
"step": 1524
},
{
"epoch": 0.6090255591054313,
"grad_norm": 6.349497318267822,
"learning_rate": 3.93370607028754e-05,
"loss": 0.0331,
"step": 1525
},
{
"epoch": 0.6094249201277955,
"grad_norm": 24.035322189331055,
"learning_rate": 3.929712460063898e-05,
"loss": 0.5404,
"step": 1526
},
{
"epoch": 0.6098242811501597,
"grad_norm": 0.13440853357315063,
"learning_rate": 3.9257188498402556e-05,
"loss": 0.0012,
"step": 1527
},
{
"epoch": 0.610223642172524,
"grad_norm": 11.109216690063477,
"learning_rate": 3.9217252396166136e-05,
"loss": 0.066,
"step": 1528
},
{
"epoch": 0.6106230031948882,
"grad_norm": 21.41535758972168,
"learning_rate": 3.9177316293929715e-05,
"loss": 0.2028,
"step": 1529
},
{
"epoch": 0.6110223642172524,
"grad_norm": 36.52550506591797,
"learning_rate": 3.913738019169329e-05,
"loss": 0.5855,
"step": 1530
},
{
"epoch": 0.6114217252396166,
"grad_norm": 14.640092849731445,
"learning_rate": 3.909744408945687e-05,
"loss": 0.0954,
"step": 1531
},
{
"epoch": 0.6118210862619808,
"grad_norm": 13.958724021911621,
"learning_rate": 3.9057507987220447e-05,
"loss": 0.0947,
"step": 1532
},
{
"epoch": 0.612220447284345,
"grad_norm": 20.32636260986328,
"learning_rate": 3.9017571884984026e-05,
"loss": 0.2305,
"step": 1533
},
{
"epoch": 0.6126198083067093,
"grad_norm": 45.86320877075195,
"learning_rate": 3.8977635782747605e-05,
"loss": 1.1186,
"step": 1534
},
{
"epoch": 0.6130191693290735,
"grad_norm": 6.9453511238098145,
"learning_rate": 3.8937699680511185e-05,
"loss": 0.0404,
"step": 1535
},
{
"epoch": 0.6134185303514377,
"grad_norm": 26.327165603637695,
"learning_rate": 3.889776357827476e-05,
"loss": 0.5602,
"step": 1536
},
{
"epoch": 0.6138178913738019,
"grad_norm": 28.690813064575195,
"learning_rate": 3.885782747603834e-05,
"loss": 0.3309,
"step": 1537
},
{
"epoch": 0.6142172523961661,
"grad_norm": 0.793572187423706,
"learning_rate": 3.8817891373801916e-05,
"loss": 0.0046,
"step": 1538
},
{
"epoch": 0.6146166134185304,
"grad_norm": 11.167415618896484,
"learning_rate": 3.8777955271565496e-05,
"loss": 0.0712,
"step": 1539
},
{
"epoch": 0.6150159744408946,
"grad_norm": 1.7750678062438965,
"learning_rate": 3.8738019169329075e-05,
"loss": 0.0155,
"step": 1540
},
{
"epoch": 0.6154153354632588,
"grad_norm": 4.840240478515625,
"learning_rate": 3.8698083067092655e-05,
"loss": 0.0874,
"step": 1541
},
{
"epoch": 0.615814696485623,
"grad_norm": 11.00268268585205,
"learning_rate": 3.8658146964856234e-05,
"loss": 0.1139,
"step": 1542
},
{
"epoch": 0.6162140575079872,
"grad_norm": 34.251869201660156,
"learning_rate": 3.861821086261981e-05,
"loss": 0.8442,
"step": 1543
},
{
"epoch": 0.6166134185303515,
"grad_norm": 20.13202476501465,
"learning_rate": 3.8578274760383386e-05,
"loss": 0.1605,
"step": 1544
},
{
"epoch": 0.6170127795527156,
"grad_norm": 35.75127410888672,
"learning_rate": 3.8538338658146966e-05,
"loss": 0.7653,
"step": 1545
},
{
"epoch": 0.6174121405750799,
"grad_norm": 13.408823013305664,
"learning_rate": 3.8498402555910545e-05,
"loss": 0.4444,
"step": 1546
},
{
"epoch": 0.6178115015974441,
"grad_norm": 12.361610412597656,
"learning_rate": 3.8458466453674124e-05,
"loss": 0.077,
"step": 1547
},
{
"epoch": 0.6182108626198083,
"grad_norm": 66.31970977783203,
"learning_rate": 3.8418530351437704e-05,
"loss": 1.1782,
"step": 1548
},
{
"epoch": 0.6186102236421726,
"grad_norm": 7.942250728607178,
"learning_rate": 3.8378594249201276e-05,
"loss": 0.0707,
"step": 1549
},
{
"epoch": 0.6190095846645367,
"grad_norm": 15.362750053405762,
"learning_rate": 3.8338658146964856e-05,
"loss": 0.1244,
"step": 1550
},
{
"epoch": 0.619408945686901,
"grad_norm": 63.679901123046875,
"learning_rate": 3.8298722044728435e-05,
"loss": 1.058,
"step": 1551
},
{
"epoch": 0.6198083067092651,
"grad_norm": 0.10918296873569489,
"learning_rate": 3.8258785942492015e-05,
"loss": 0.0006,
"step": 1552
},
{
"epoch": 0.6202076677316294,
"grad_norm": 13.151423454284668,
"learning_rate": 3.8218849840255594e-05,
"loss": 0.1287,
"step": 1553
},
{
"epoch": 0.6206070287539937,
"grad_norm": 13.7886962890625,
"learning_rate": 3.8178913738019174e-05,
"loss": 0.0877,
"step": 1554
},
{
"epoch": 0.6210063897763578,
"grad_norm": 2.290269136428833,
"learning_rate": 3.813897763578275e-05,
"loss": 0.0125,
"step": 1555
},
{
"epoch": 0.6214057507987221,
"grad_norm": 13.206582069396973,
"learning_rate": 3.8099041533546326e-05,
"loss": 0.0833,
"step": 1556
},
{
"epoch": 0.6218051118210862,
"grad_norm": 1.2925217151641846,
"learning_rate": 3.8059105431309905e-05,
"loss": 0.0078,
"step": 1557
},
{
"epoch": 0.6222044728434505,
"grad_norm": 24.776552200317383,
"learning_rate": 3.8019169329073485e-05,
"loss": 0.4143,
"step": 1558
},
{
"epoch": 0.6226038338658147,
"grad_norm": 19.237014770507812,
"learning_rate": 3.7979233226837064e-05,
"loss": 0.1736,
"step": 1559
},
{
"epoch": 0.6230031948881789,
"grad_norm": 24.245960235595703,
"learning_rate": 3.793929712460064e-05,
"loss": 0.4651,
"step": 1560
},
{
"epoch": 0.6234025559105432,
"grad_norm": 5.9789533615112305,
"learning_rate": 3.789936102236422e-05,
"loss": 0.0452,
"step": 1561
},
{
"epoch": 0.6238019169329073,
"grad_norm": 26.231645584106445,
"learning_rate": 3.7859424920127795e-05,
"loss": 0.4262,
"step": 1562
},
{
"epoch": 0.6242012779552716,
"grad_norm": 2.5016167163848877,
"learning_rate": 3.7819488817891375e-05,
"loss": 0.0124,
"step": 1563
},
{
"epoch": 0.6246006389776357,
"grad_norm": 18.809574127197266,
"learning_rate": 3.7779552715654954e-05,
"loss": 0.151,
"step": 1564
},
{
"epoch": 0.625,
"grad_norm": 0.2166312336921692,
"learning_rate": 3.7739616613418534e-05,
"loss": 0.0013,
"step": 1565
},
{
"epoch": 0.6253993610223643,
"grad_norm": 46.6550407409668,
"learning_rate": 3.769968051118211e-05,
"loss": 1.7998,
"step": 1566
},
{
"epoch": 0.6257987220447284,
"grad_norm": 4.920192718505859,
"learning_rate": 3.765974440894569e-05,
"loss": 0.0418,
"step": 1567
},
{
"epoch": 0.6261980830670927,
"grad_norm": 6.210689544677734,
"learning_rate": 3.7619808306709265e-05,
"loss": 0.034,
"step": 1568
},
{
"epoch": 0.6265974440894568,
"grad_norm": 7.125272750854492,
"learning_rate": 3.7579872204472845e-05,
"loss": 0.0382,
"step": 1569
},
{
"epoch": 0.6269968051118211,
"grad_norm": 22.040767669677734,
"learning_rate": 3.7539936102236424e-05,
"loss": 0.1893,
"step": 1570
},
{
"epoch": 0.6273961661341853,
"grad_norm": 19.284626007080078,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.1594,
"step": 1571
},
{
"epoch": 0.6277955271565495,
"grad_norm": 4.47816801071167,
"learning_rate": 3.746006389776358e-05,
"loss": 0.026,
"step": 1572
},
{
"epoch": 0.6281948881789138,
"grad_norm": 22.661113739013672,
"learning_rate": 3.742012779552716e-05,
"loss": 0.6954,
"step": 1573
},
{
"epoch": 0.6285942492012779,
"grad_norm": 30.986406326293945,
"learning_rate": 3.738019169329074e-05,
"loss": 0.9723,
"step": 1574
},
{
"epoch": 0.6289936102236422,
"grad_norm": 36.61789321899414,
"learning_rate": 3.7340255591054314e-05,
"loss": 0.6714,
"step": 1575
},
{
"epoch": 0.6293929712460063,
"grad_norm": 33.37625503540039,
"learning_rate": 3.7300319488817894e-05,
"loss": 0.8274,
"step": 1576
},
{
"epoch": 0.6297923322683706,
"grad_norm": 7.138182163238525,
"learning_rate": 3.726038338658147e-05,
"loss": 0.0391,
"step": 1577
},
{
"epoch": 0.6301916932907349,
"grad_norm": 14.522740364074707,
"learning_rate": 3.722044728434505e-05,
"loss": 0.5677,
"step": 1578
},
{
"epoch": 0.630591054313099,
"grad_norm": 1.2679247856140137,
"learning_rate": 3.718051118210863e-05,
"loss": 0.0068,
"step": 1579
},
{
"epoch": 0.6309904153354633,
"grad_norm": 4.87261438369751,
"learning_rate": 3.714057507987221e-05,
"loss": 0.0431,
"step": 1580
},
{
"epoch": 0.6313897763578274,
"grad_norm": 22.896696090698242,
"learning_rate": 3.7100638977635784e-05,
"loss": 0.2088,
"step": 1581
},
{
"epoch": 0.6317891373801917,
"grad_norm": 18.34836769104004,
"learning_rate": 3.7060702875399364e-05,
"loss": 0.3843,
"step": 1582
},
{
"epoch": 0.6321884984025559,
"grad_norm": 37.93985366821289,
"learning_rate": 3.702076677316294e-05,
"loss": 0.9422,
"step": 1583
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.1080612987279892,
"learning_rate": 3.698083067092652e-05,
"loss": 0.001,
"step": 1584
},
{
"epoch": 0.6329872204472844,
"grad_norm": 20.865276336669922,
"learning_rate": 3.69408945686901e-05,
"loss": 0.1558,
"step": 1585
},
{
"epoch": 0.6333865814696485,
"grad_norm": 9.279122352600098,
"learning_rate": 3.6900958466453675e-05,
"loss": 0.0525,
"step": 1586
},
{
"epoch": 0.6337859424920128,
"grad_norm": 18.24469757080078,
"learning_rate": 3.6861022364217254e-05,
"loss": 0.1535,
"step": 1587
},
{
"epoch": 0.634185303514377,
"grad_norm": 24.877601623535156,
"learning_rate": 3.6821086261980833e-05,
"loss": 0.4628,
"step": 1588
},
{
"epoch": 0.6345846645367412,
"grad_norm": 0.6034784317016602,
"learning_rate": 3.678115015974441e-05,
"loss": 0.0056,
"step": 1589
},
{
"epoch": 0.6349840255591054,
"grad_norm": 20.00201416015625,
"learning_rate": 3.6741214057507985e-05,
"loss": 0.2139,
"step": 1590
},
{
"epoch": 0.6353833865814696,
"grad_norm": 25.50446891784668,
"learning_rate": 3.6701277955271565e-05,
"loss": 0.3511,
"step": 1591
},
{
"epoch": 0.6357827476038339,
"grad_norm": 0.49296045303344727,
"learning_rate": 3.6661341853035144e-05,
"loss": 0.0038,
"step": 1592
},
{
"epoch": 0.6361821086261981,
"grad_norm": 0.5980103611946106,
"learning_rate": 3.6621405750798724e-05,
"loss": 0.0042,
"step": 1593
},
{
"epoch": 0.6365814696485623,
"grad_norm": 13.532281875610352,
"learning_rate": 3.65814696485623e-05,
"loss": 0.0878,
"step": 1594
},
{
"epoch": 0.6369808306709265,
"grad_norm": 21.31557273864746,
"learning_rate": 3.6541533546325876e-05,
"loss": 0.4456,
"step": 1595
},
{
"epoch": 0.6373801916932907,
"grad_norm": 25.42170524597168,
"learning_rate": 3.6501597444089455e-05,
"loss": 0.7673,
"step": 1596
},
{
"epoch": 0.637779552715655,
"grad_norm": 29.405590057373047,
"learning_rate": 3.6461661341853035e-05,
"loss": 0.5188,
"step": 1597
},
{
"epoch": 0.6381789137380192,
"grad_norm": 1.2862151861190796,
"learning_rate": 3.6421725239616614e-05,
"loss": 0.0212,
"step": 1598
},
{
"epoch": 0.6385782747603834,
"grad_norm": 26.857885360717773,
"learning_rate": 3.6381789137380194e-05,
"loss": 0.4874,
"step": 1599
},
{
"epoch": 0.6389776357827476,
"grad_norm": 25.04949378967285,
"learning_rate": 3.6341853035143766e-05,
"loss": 0.2629,
"step": 1600
},
{
"epoch": 0.6393769968051118,
"grad_norm": 13.319676399230957,
"learning_rate": 3.6301916932907346e-05,
"loss": 0.145,
"step": 1601
},
{
"epoch": 0.639776357827476,
"grad_norm": 27.14652442932129,
"learning_rate": 3.6261980830670925e-05,
"loss": 0.5719,
"step": 1602
},
{
"epoch": 0.6401757188498403,
"grad_norm": 19.81570053100586,
"learning_rate": 3.6222044728434504e-05,
"loss": 0.159,
"step": 1603
},
{
"epoch": 0.6405750798722045,
"grad_norm": 31.76645851135254,
"learning_rate": 3.6182108626198084e-05,
"loss": 0.7079,
"step": 1604
},
{
"epoch": 0.6409744408945687,
"grad_norm": 16.83795166015625,
"learning_rate": 3.614217252396166e-05,
"loss": 0.1157,
"step": 1605
},
{
"epoch": 0.6413738019169329,
"grad_norm": 22.444040298461914,
"learning_rate": 3.610223642172524e-05,
"loss": 0.2312,
"step": 1606
},
{
"epoch": 0.6417731629392971,
"grad_norm": 12.5982084274292,
"learning_rate": 3.6062300319488815e-05,
"loss": 0.0993,
"step": 1607
},
{
"epoch": 0.6421725239616614,
"grad_norm": 1.7280575037002563,
"learning_rate": 3.6022364217252395e-05,
"loss": 0.0107,
"step": 1608
},
{
"epoch": 0.6425718849840255,
"grad_norm": 19.766515731811523,
"learning_rate": 3.5982428115015974e-05,
"loss": 0.7886,
"step": 1609
},
{
"epoch": 0.6429712460063898,
"grad_norm": 18.071475982666016,
"learning_rate": 3.5942492012779554e-05,
"loss": 0.1322,
"step": 1610
},
{
"epoch": 0.643370607028754,
"grad_norm": 30.4601993560791,
"learning_rate": 3.590255591054313e-05,
"loss": 0.7157,
"step": 1611
},
{
"epoch": 0.6437699680511182,
"grad_norm": 8.822271347045898,
"learning_rate": 3.586261980830671e-05,
"loss": 0.0513,
"step": 1612
},
{
"epoch": 0.6441693290734825,
"grad_norm": 29.195974349975586,
"learning_rate": 3.5822683706070285e-05,
"loss": 0.6351,
"step": 1613
},
{
"epoch": 0.6445686900958466,
"grad_norm": 27.005413055419922,
"learning_rate": 3.5782747603833865e-05,
"loss": 0.5205,
"step": 1614
},
{
"epoch": 0.6449680511182109,
"grad_norm": 0.3215716779232025,
"learning_rate": 3.5742811501597444e-05,
"loss": 0.0024,
"step": 1615
},
{
"epoch": 0.645367412140575,
"grad_norm": 34.198577880859375,
"learning_rate": 3.5702875399361023e-05,
"loss": 0.8255,
"step": 1616
},
{
"epoch": 0.6457667731629393,
"grad_norm": 6.152607440948486,
"learning_rate": 3.56629392971246e-05,
"loss": 0.0335,
"step": 1617
},
{
"epoch": 0.6461661341853036,
"grad_norm": 24.414854049682617,
"learning_rate": 3.562300319488818e-05,
"loss": 0.3862,
"step": 1618
},
{
"epoch": 0.6465654952076677,
"grad_norm": 3.27604341506958,
"learning_rate": 3.5583067092651755e-05,
"loss": 0.0536,
"step": 1619
},
{
"epoch": 0.646964856230032,
"grad_norm": 18.17131233215332,
"learning_rate": 3.5543130990415334e-05,
"loss": 0.1532,
"step": 1620
},
{
"epoch": 0.6473642172523961,
"grad_norm": 0.08511903882026672,
"learning_rate": 3.5503194888178914e-05,
"loss": 0.0007,
"step": 1621
},
{
"epoch": 0.6477635782747604,
"grad_norm": 7.779764175415039,
"learning_rate": 3.546325878594249e-05,
"loss": 0.0582,
"step": 1622
},
{
"epoch": 0.6481629392971247,
"grad_norm": 27.803491592407227,
"learning_rate": 3.542332268370607e-05,
"loss": 0.5498,
"step": 1623
},
{
"epoch": 0.6485623003194888,
"grad_norm": 24.998760223388672,
"learning_rate": 3.538338658146965e-05,
"loss": 0.4726,
"step": 1624
},
{
"epoch": 0.6489616613418531,
"grad_norm": 26.206653594970703,
"learning_rate": 3.534345047923323e-05,
"loss": 0.801,
"step": 1625
},
{
"epoch": 0.6493610223642172,
"grad_norm": 23.767261505126953,
"learning_rate": 3.5303514376996804e-05,
"loss": 0.8463,
"step": 1626
},
{
"epoch": 0.6497603833865815,
"grad_norm": 5.040603160858154,
"learning_rate": 3.5263578274760384e-05,
"loss": 0.028,
"step": 1627
},
{
"epoch": 0.6501597444089456,
"grad_norm": 14.891948699951172,
"learning_rate": 3.522364217252396e-05,
"loss": 0.1563,
"step": 1628
},
{
"epoch": 0.6505591054313099,
"grad_norm": 4.9926910400390625,
"learning_rate": 3.518370607028754e-05,
"loss": 0.0501,
"step": 1629
},
{
"epoch": 0.6509584664536742,
"grad_norm": 6.800023555755615,
"learning_rate": 3.514376996805112e-05,
"loss": 0.0535,
"step": 1630
},
{
"epoch": 0.6513578274760383,
"grad_norm": 50.537452697753906,
"learning_rate": 3.51038338658147e-05,
"loss": 1.2541,
"step": 1631
},
{
"epoch": 0.6517571884984026,
"grad_norm": 17.22194480895996,
"learning_rate": 3.5063897763578274e-05,
"loss": 0.1598,
"step": 1632
},
{
"epoch": 0.6521565495207667,
"grad_norm": 33.52593231201172,
"learning_rate": 3.502396166134185e-05,
"loss": 0.5932,
"step": 1633
},
{
"epoch": 0.652555910543131,
"grad_norm": 23.069252014160156,
"learning_rate": 3.498402555910543e-05,
"loss": 0.3979,
"step": 1634
},
{
"epoch": 0.6529552715654952,
"grad_norm": 8.714667320251465,
"learning_rate": 3.494408945686901e-05,
"loss": 0.0457,
"step": 1635
},
{
"epoch": 0.6533546325878594,
"grad_norm": 1.3112952709197998,
"learning_rate": 3.490415335463259e-05,
"loss": 0.0094,
"step": 1636
},
{
"epoch": 0.6537539936102237,
"grad_norm": 48.327239990234375,
"learning_rate": 3.486421725239617e-05,
"loss": 0.8532,
"step": 1637
},
{
"epoch": 0.6541533546325878,
"grad_norm": 34.09596633911133,
"learning_rate": 3.482428115015975e-05,
"loss": 0.5333,
"step": 1638
},
{
"epoch": 0.6545527156549521,
"grad_norm": 25.1065673828125,
"learning_rate": 3.478434504792332e-05,
"loss": 0.2841,
"step": 1639
},
{
"epoch": 0.6549520766773163,
"grad_norm": 45.087215423583984,
"learning_rate": 3.47444089456869e-05,
"loss": 0.8533,
"step": 1640
},
{
"epoch": 0.6553514376996805,
"grad_norm": 23.661073684692383,
"learning_rate": 3.470447284345048e-05,
"loss": 0.3821,
"step": 1641
},
{
"epoch": 0.6557507987220448,
"grad_norm": 8.155472755432129,
"learning_rate": 3.466453674121406e-05,
"loss": 0.1146,
"step": 1642
},
{
"epoch": 0.6561501597444089,
"grad_norm": 9.792318344116211,
"learning_rate": 3.462460063897764e-05,
"loss": 0.0778,
"step": 1643
},
{
"epoch": 0.6565495207667732,
"grad_norm": 39.94092559814453,
"learning_rate": 3.458466453674122e-05,
"loss": 1.1581,
"step": 1644
},
{
"epoch": 0.6569488817891374,
"grad_norm": 2.0859429836273193,
"learning_rate": 3.454472843450479e-05,
"loss": 0.0122,
"step": 1645
},
{
"epoch": 0.6573482428115016,
"grad_norm": 41.719669342041016,
"learning_rate": 3.450479233226837e-05,
"loss": 1.1925,
"step": 1646
},
{
"epoch": 0.6577476038338658,
"grad_norm": 20.607593536376953,
"learning_rate": 3.446485623003195e-05,
"loss": 0.3995,
"step": 1647
},
{
"epoch": 0.65814696485623,
"grad_norm": 4.18411922454834,
"learning_rate": 3.442492012779553e-05,
"loss": 0.0523,
"step": 1648
},
{
"epoch": 0.6585463258785943,
"grad_norm": 25.284082412719727,
"learning_rate": 3.438498402555911e-05,
"loss": 0.5084,
"step": 1649
},
{
"epoch": 0.6589456869009584,
"grad_norm": 3.829207420349121,
"learning_rate": 3.434504792332269e-05,
"loss": 0.0233,
"step": 1650
},
{
"epoch": 0.6593450479233227,
"grad_norm": 1.535058856010437,
"learning_rate": 3.430511182108626e-05,
"loss": 0.0144,
"step": 1651
},
{
"epoch": 0.6597444089456869,
"grad_norm": 27.17057228088379,
"learning_rate": 3.426517571884984e-05,
"loss": 0.478,
"step": 1652
},
{
"epoch": 0.6601437699680511,
"grad_norm": 24.053367614746094,
"learning_rate": 3.422523961661342e-05,
"loss": 0.2503,
"step": 1653
},
{
"epoch": 0.6605431309904153,
"grad_norm": 22.587265014648438,
"learning_rate": 3.4185303514377e-05,
"loss": 0.2572,
"step": 1654
},
{
"epoch": 0.6609424920127795,
"grad_norm": 33.512672424316406,
"learning_rate": 3.414536741214058e-05,
"loss": 2.0723,
"step": 1655
},
{
"epoch": 0.6613418530351438,
"grad_norm": 5.721658229827881,
"learning_rate": 3.410543130990416e-05,
"loss": 0.0294,
"step": 1656
},
{
"epoch": 0.661741214057508,
"grad_norm": 1.3004157543182373,
"learning_rate": 3.406549520766773e-05,
"loss": 0.0087,
"step": 1657
},
{
"epoch": 0.6621405750798722,
"grad_norm": 18.728103637695312,
"learning_rate": 3.402555910543131e-05,
"loss": 0.1727,
"step": 1658
},
{
"epoch": 0.6625399361022364,
"grad_norm": 28.70616340637207,
"learning_rate": 3.398562300319489e-05,
"loss": 0.365,
"step": 1659
},
{
"epoch": 0.6629392971246006,
"grad_norm": 7.147408962249756,
"learning_rate": 3.394568690095847e-05,
"loss": 0.077,
"step": 1660
},
{
"epoch": 0.6633386581469649,
"grad_norm": 3.3001344203948975,
"learning_rate": 3.390575079872205e-05,
"loss": 0.0176,
"step": 1661
},
{
"epoch": 0.6637380191693291,
"grad_norm": 19.7530460357666,
"learning_rate": 3.386581469648562e-05,
"loss": 0.631,
"step": 1662
},
{
"epoch": 0.6641373801916933,
"grad_norm": 18.8158016204834,
"learning_rate": 3.38258785942492e-05,
"loss": 0.214,
"step": 1663
},
{
"epoch": 0.6645367412140575,
"grad_norm": 25.574384689331055,
"learning_rate": 3.378594249201278e-05,
"loss": 0.2062,
"step": 1664
},
{
"epoch": 0.6649361022364217,
"grad_norm": 1.231232762336731,
"learning_rate": 3.374600638977636e-05,
"loss": 0.0072,
"step": 1665
},
{
"epoch": 0.6653354632587859,
"grad_norm": 20.98204231262207,
"learning_rate": 3.3706070287539934e-05,
"loss": 0.1649,
"step": 1666
},
{
"epoch": 0.6657348242811502,
"grad_norm": 38.914146423339844,
"learning_rate": 3.366613418530351e-05,
"loss": 1.0626,
"step": 1667
},
{
"epoch": 0.6661341853035144,
"grad_norm": 5.786464214324951,
"learning_rate": 3.362619808306709e-05,
"loss": 0.0339,
"step": 1668
},
{
"epoch": 0.6665335463258786,
"grad_norm": 2.259472370147705,
"learning_rate": 3.358626198083067e-05,
"loss": 0.0177,
"step": 1669
},
{
"epoch": 0.6669329073482428,
"grad_norm": 12.137011528015137,
"learning_rate": 3.354632587859425e-05,
"loss": 0.0724,
"step": 1670
},
{
"epoch": 0.667332268370607,
"grad_norm": 25.429401397705078,
"learning_rate": 3.3506389776357824e-05,
"loss": 0.6998,
"step": 1671
},
{
"epoch": 0.6677316293929713,
"grad_norm": 27.266828536987305,
"learning_rate": 3.3466453674121404e-05,
"loss": 0.538,
"step": 1672
},
{
"epoch": 0.6681309904153354,
"grad_norm": 9.027865409851074,
"learning_rate": 3.342651757188498e-05,
"loss": 0.0616,
"step": 1673
},
{
"epoch": 0.6685303514376997,
"grad_norm": 26.240299224853516,
"learning_rate": 3.338658146964856e-05,
"loss": 0.3495,
"step": 1674
},
{
"epoch": 0.6689297124600639,
"grad_norm": 22.66031837463379,
"learning_rate": 3.334664536741214e-05,
"loss": 0.1953,
"step": 1675
},
{
"epoch": 0.6693290734824281,
"grad_norm": 26.547893524169922,
"learning_rate": 3.330670926517572e-05,
"loss": 0.3896,
"step": 1676
},
{
"epoch": 0.6697284345047924,
"grad_norm": 20.041688919067383,
"learning_rate": 3.3266773162939294e-05,
"loss": 0.1729,
"step": 1677
},
{
"epoch": 0.6701277955271565,
"grad_norm": 22.736209869384766,
"learning_rate": 3.322683706070287e-05,
"loss": 0.195,
"step": 1678
},
{
"epoch": 0.6705271565495208,
"grad_norm": 6.814960956573486,
"learning_rate": 3.318690095846645e-05,
"loss": 0.0489,
"step": 1679
},
{
"epoch": 0.670926517571885,
"grad_norm": 25.638141632080078,
"learning_rate": 3.314696485623003e-05,
"loss": 0.3723,
"step": 1680
},
{
"epoch": 0.6713258785942492,
"grad_norm": 22.816814422607422,
"learning_rate": 3.310702875399361e-05,
"loss": 0.535,
"step": 1681
},
{
"epoch": 0.6717252396166135,
"grad_norm": 30.337574005126953,
"learning_rate": 3.306709265175719e-05,
"loss": 0.8897,
"step": 1682
},
{
"epoch": 0.6721246006389776,
"grad_norm": 17.132484436035156,
"learning_rate": 3.3027156549520764e-05,
"loss": 0.2494,
"step": 1683
},
{
"epoch": 0.6725239616613419,
"grad_norm": 13.495040893554688,
"learning_rate": 3.298722044728434e-05,
"loss": 0.162,
"step": 1684
},
{
"epoch": 0.672923322683706,
"grad_norm": 21.766117095947266,
"learning_rate": 3.294728434504792e-05,
"loss": 1.1163,
"step": 1685
},
{
"epoch": 0.6733226837060703,
"grad_norm": 26.330821990966797,
"learning_rate": 3.29073482428115e-05,
"loss": 0.511,
"step": 1686
},
{
"epoch": 0.6737220447284346,
"grad_norm": 0.20950870215892792,
"learning_rate": 3.286741214057508e-05,
"loss": 0.002,
"step": 1687
},
{
"epoch": 0.6741214057507987,
"grad_norm": 18.804336547851562,
"learning_rate": 3.282747603833866e-05,
"loss": 0.1577,
"step": 1688
},
{
"epoch": 0.674520766773163,
"grad_norm": 24.623777389526367,
"learning_rate": 3.278753993610224e-05,
"loss": 0.2652,
"step": 1689
},
{
"epoch": 0.6749201277955271,
"grad_norm": 2.072949171066284,
"learning_rate": 3.274760383386581e-05,
"loss": 0.0127,
"step": 1690
},
{
"epoch": 0.6753194888178914,
"grad_norm": 3.2308807373046875,
"learning_rate": 3.270766773162939e-05,
"loss": 0.0161,
"step": 1691
},
{
"epoch": 0.6757188498402555,
"grad_norm": 12.08854866027832,
"learning_rate": 3.266773162939297e-05,
"loss": 0.124,
"step": 1692
},
{
"epoch": 0.6761182108626198,
"grad_norm": 1.129089593887329,
"learning_rate": 3.262779552715655e-05,
"loss": 0.01,
"step": 1693
},
{
"epoch": 0.6765175718849841,
"grad_norm": 1.0835005044937134,
"learning_rate": 3.258785942492013e-05,
"loss": 0.0052,
"step": 1694
},
{
"epoch": 0.6769169329073482,
"grad_norm": 25.65920639038086,
"learning_rate": 3.254792332268371e-05,
"loss": 0.3507,
"step": 1695
},
{
"epoch": 0.6773162939297125,
"grad_norm": 20.962966918945312,
"learning_rate": 3.250798722044728e-05,
"loss": 0.5725,
"step": 1696
},
{
"epoch": 0.6777156549520766,
"grad_norm": 13.08654499053955,
"learning_rate": 3.246805111821086e-05,
"loss": 0.0826,
"step": 1697
},
{
"epoch": 0.6781150159744409,
"grad_norm": 9.744897842407227,
"learning_rate": 3.242811501597444e-05,
"loss": 0.057,
"step": 1698
},
{
"epoch": 0.6785143769968051,
"grad_norm": 76.88810729980469,
"learning_rate": 3.238817891373802e-05,
"loss": 1.3331,
"step": 1699
},
{
"epoch": 0.6789137380191693,
"grad_norm": 25.93117904663086,
"learning_rate": 3.23482428115016e-05,
"loss": 0.2845,
"step": 1700
},
{
"epoch": 0.6793130990415336,
"grad_norm": 26.85693359375,
"learning_rate": 3.230830670926518e-05,
"loss": 0.2354,
"step": 1701
},
{
"epoch": 0.6797124600638977,
"grad_norm": 54.21931457519531,
"learning_rate": 3.226837060702875e-05,
"loss": 0.7315,
"step": 1702
},
{
"epoch": 0.680111821086262,
"grad_norm": 5.801416397094727,
"learning_rate": 3.222843450479233e-05,
"loss": 0.1032,
"step": 1703
},
{
"epoch": 0.6805111821086262,
"grad_norm": 54.91061019897461,
"learning_rate": 3.218849840255591e-05,
"loss": 0.944,
"step": 1704
},
{
"epoch": 0.6809105431309904,
"grad_norm": 37.826622009277344,
"learning_rate": 3.214856230031949e-05,
"loss": 0.8449,
"step": 1705
},
{
"epoch": 0.6813099041533547,
"grad_norm": 16.816484451293945,
"learning_rate": 3.210862619808307e-05,
"loss": 0.1241,
"step": 1706
},
{
"epoch": 0.6817092651757188,
"grad_norm": 25.002817153930664,
"learning_rate": 3.206869009584665e-05,
"loss": 0.3249,
"step": 1707
},
{
"epoch": 0.6821086261980831,
"grad_norm": 2.9583425521850586,
"learning_rate": 3.202875399361023e-05,
"loss": 0.0436,
"step": 1708
},
{
"epoch": 0.6825079872204473,
"grad_norm": 1.1382460594177246,
"learning_rate": 3.19888178913738e-05,
"loss": 0.0143,
"step": 1709
},
{
"epoch": 0.6829073482428115,
"grad_norm": 1.5206804275512695,
"learning_rate": 3.194888178913738e-05,
"loss": 0.0107,
"step": 1710
},
{
"epoch": 0.6833067092651757,
"grad_norm": 0.9285376071929932,
"learning_rate": 3.190894568690096e-05,
"loss": 0.0045,
"step": 1711
},
{
"epoch": 0.6837060702875399,
"grad_norm": 4.590746879577637,
"learning_rate": 3.186900958466454e-05,
"loss": 0.0263,
"step": 1712
},
{
"epoch": 0.6841054313099042,
"grad_norm": 14.753111839294434,
"learning_rate": 3.182907348242812e-05,
"loss": 0.1474,
"step": 1713
},
{
"epoch": 0.6845047923322684,
"grad_norm": 1.8788858652114868,
"learning_rate": 3.17891373801917e-05,
"loss": 0.0133,
"step": 1714
},
{
"epoch": 0.6849041533546326,
"grad_norm": 23.19841766357422,
"learning_rate": 3.174920127795527e-05,
"loss": 0.3017,
"step": 1715
},
{
"epoch": 0.6853035143769968,
"grad_norm": 8.508781433105469,
"learning_rate": 3.170926517571885e-05,
"loss": 0.0581,
"step": 1716
},
{
"epoch": 0.685702875399361,
"grad_norm": 1.278319001197815,
"learning_rate": 3.166932907348243e-05,
"loss": 0.007,
"step": 1717
},
{
"epoch": 0.6861022364217252,
"grad_norm": 13.274981498718262,
"learning_rate": 3.162939297124601e-05,
"loss": 0.179,
"step": 1718
},
{
"epoch": 0.6865015974440895,
"grad_norm": 25.666845321655273,
"learning_rate": 3.158945686900959e-05,
"loss": 0.4124,
"step": 1719
},
{
"epoch": 0.6869009584664537,
"grad_norm": 21.09577751159668,
"learning_rate": 3.154952076677317e-05,
"loss": 0.1835,
"step": 1720
},
{
"epoch": 0.6873003194888179,
"grad_norm": 23.83101463317871,
"learning_rate": 3.150958466453675e-05,
"loss": 0.2809,
"step": 1721
},
{
"epoch": 0.6876996805111821,
"grad_norm": 0.9714841246604919,
"learning_rate": 3.146964856230032e-05,
"loss": 0.0072,
"step": 1722
},
{
"epoch": 0.6880990415335463,
"grad_norm": 19.704824447631836,
"learning_rate": 3.14297124600639e-05,
"loss": 0.1601,
"step": 1723
},
{
"epoch": 0.6884984025559105,
"grad_norm": 15.176055908203125,
"learning_rate": 3.138977635782748e-05,
"loss": 0.1987,
"step": 1724
},
{
"epoch": 0.6888977635782748,
"grad_norm": 9.060149192810059,
"learning_rate": 3.134984025559106e-05,
"loss": 0.0525,
"step": 1725
},
{
"epoch": 0.689297124600639,
"grad_norm": 37.687355041503906,
"learning_rate": 3.130990415335464e-05,
"loss": 0.4055,
"step": 1726
},
{
"epoch": 0.6896964856230032,
"grad_norm": 26.2545223236084,
"learning_rate": 3.126996805111822e-05,
"loss": 0.5144,
"step": 1727
},
{
"epoch": 0.6900958466453674,
"grad_norm": 18.40363883972168,
"learning_rate": 3.123003194888179e-05,
"loss": 0.1547,
"step": 1728
},
{
"epoch": 0.6904952076677316,
"grad_norm": 12.444860458374023,
"learning_rate": 3.119009584664537e-05,
"loss": 0.1268,
"step": 1729
},
{
"epoch": 0.6908945686900958,
"grad_norm": 19.316720962524414,
"learning_rate": 3.115015974440895e-05,
"loss": 0.1448,
"step": 1730
},
{
"epoch": 0.6912939297124601,
"grad_norm": 8.904155731201172,
"learning_rate": 3.111022364217253e-05,
"loss": 0.0499,
"step": 1731
},
{
"epoch": 0.6916932907348243,
"grad_norm": 9.765684127807617,
"learning_rate": 3.107028753993611e-05,
"loss": 0.056,
"step": 1732
},
{
"epoch": 0.6920926517571885,
"grad_norm": 3.605161666870117,
"learning_rate": 3.103035143769968e-05,
"loss": 0.0197,
"step": 1733
},
{
"epoch": 0.6924920127795527,
"grad_norm": 36.04047393798828,
"learning_rate": 3.099041533546326e-05,
"loss": 1.0468,
"step": 1734
},
{
"epoch": 0.6928913738019169,
"grad_norm": 25.5123233795166,
"learning_rate": 3.095047923322684e-05,
"loss": 1.0348,
"step": 1735
},
{
"epoch": 0.6932907348242812,
"grad_norm": 7.475257396697998,
"learning_rate": 3.091054313099042e-05,
"loss": 0.04,
"step": 1736
},
{
"epoch": 0.6936900958466453,
"grad_norm": 10.03475284576416,
"learning_rate": 3.087060702875399e-05,
"loss": 0.0547,
"step": 1737
},
{
"epoch": 0.6940894568690096,
"grad_norm": 19.754467010498047,
"learning_rate": 3.083067092651757e-05,
"loss": 1.7871,
"step": 1738
},
{
"epoch": 0.6944888178913738,
"grad_norm": 15.78553295135498,
"learning_rate": 3.079073482428115e-05,
"loss": 0.1044,
"step": 1739
},
{
"epoch": 0.694888178913738,
"grad_norm": 17.165983200073242,
"learning_rate": 3.075079872204473e-05,
"loss": 0.116,
"step": 1740
},
{
"epoch": 0.6952875399361023,
"grad_norm": 3.0492587089538574,
"learning_rate": 3.071086261980831e-05,
"loss": 0.0184,
"step": 1741
},
{
"epoch": 0.6956869009584664,
"grad_norm": 9.662361145019531,
"learning_rate": 3.067092651757188e-05,
"loss": 0.0637,
"step": 1742
},
{
"epoch": 0.6960862619808307,
"grad_norm": 22.180517196655273,
"learning_rate": 3.063099041533546e-05,
"loss": 0.1357,
"step": 1743
},
{
"epoch": 0.6964856230031949,
"grad_norm": 43.7991943359375,
"learning_rate": 3.059105431309904e-05,
"loss": 0.5775,
"step": 1744
},
{
"epoch": 0.6968849840255591,
"grad_norm": 3.3218820095062256,
"learning_rate": 3.055111821086262e-05,
"loss": 0.0254,
"step": 1745
},
{
"epoch": 0.6972843450479234,
"grad_norm": 18.16935920715332,
"learning_rate": 3.0511182108626203e-05,
"loss": 0.142,
"step": 1746
},
{
"epoch": 0.6976837060702875,
"grad_norm": 6.840794086456299,
"learning_rate": 3.0471246006389776e-05,
"loss": 0.0381,
"step": 1747
},
{
"epoch": 0.6980830670926518,
"grad_norm": 2.832826614379883,
"learning_rate": 3.0431309904153355e-05,
"loss": 0.0154,
"step": 1748
},
{
"epoch": 0.6984824281150159,
"grad_norm": 19.237863540649414,
"learning_rate": 3.0391373801916935e-05,
"loss": 0.8735,
"step": 1749
},
{
"epoch": 0.6988817891373802,
"grad_norm": 0.21930202841758728,
"learning_rate": 3.0351437699680514e-05,
"loss": 0.0012,
"step": 1750
},
{
"epoch": 0.6992811501597445,
"grad_norm": 37.86859130859375,
"learning_rate": 3.0311501597444093e-05,
"loss": 0.6794,
"step": 1751
},
{
"epoch": 0.6996805111821086,
"grad_norm": 3.847663164138794,
"learning_rate": 3.027156549520767e-05,
"loss": 0.0203,
"step": 1752
},
{
"epoch": 0.7000798722044729,
"grad_norm": 2.7159671783447266,
"learning_rate": 3.023162939297125e-05,
"loss": 0.0156,
"step": 1753
},
{
"epoch": 0.700479233226837,
"grad_norm": 21.46092414855957,
"learning_rate": 3.0191693290734825e-05,
"loss": 0.2851,
"step": 1754
},
{
"epoch": 0.7008785942492013,
"grad_norm": 25.011201858520508,
"learning_rate": 3.0151757188498404e-05,
"loss": 0.2134,
"step": 1755
},
{
"epoch": 0.7012779552715654,
"grad_norm": 23.424509048461914,
"learning_rate": 3.011182108626198e-05,
"loss": 0.791,
"step": 1756
},
{
"epoch": 0.7016773162939297,
"grad_norm": 23.109739303588867,
"learning_rate": 3.007188498402556e-05,
"loss": 0.3769,
"step": 1757
},
{
"epoch": 0.702076677316294,
"grad_norm": 0.2398453801870346,
"learning_rate": 3.003194888178914e-05,
"loss": 0.0013,
"step": 1758
},
{
"epoch": 0.7024760383386581,
"grad_norm": 1.180621862411499,
"learning_rate": 2.999201277955272e-05,
"loss": 0.0056,
"step": 1759
},
{
"epoch": 0.7028753993610224,
"grad_norm": 0.2165040224790573,
"learning_rate": 2.9952076677316295e-05,
"loss": 0.0012,
"step": 1760
},
{
"epoch": 0.7032747603833865,
"grad_norm": 10.526522636413574,
"learning_rate": 2.991214057507987e-05,
"loss": 0.065,
"step": 1761
},
{
"epoch": 0.7036741214057508,
"grad_norm": 23.738988876342773,
"learning_rate": 2.987220447284345e-05,
"loss": 0.521,
"step": 1762
},
{
"epoch": 0.704073482428115,
"grad_norm": 27.365737915039062,
"learning_rate": 2.983226837060703e-05,
"loss": 0.4314,
"step": 1763
},
{
"epoch": 0.7044728434504792,
"grad_norm": 0.6628259420394897,
"learning_rate": 2.979233226837061e-05,
"loss": 0.0031,
"step": 1764
},
{
"epoch": 0.7048722044728435,
"grad_norm": 3.7723774909973145,
"learning_rate": 2.975239616613419e-05,
"loss": 0.0764,
"step": 1765
},
{
"epoch": 0.7052715654952076,
"grad_norm": 17.615034103393555,
"learning_rate": 2.971246006389776e-05,
"loss": 0.1388,
"step": 1766
},
{
"epoch": 0.7056709265175719,
"grad_norm": 24.538965225219727,
"learning_rate": 2.967252396166134e-05,
"loss": 0.4985,
"step": 1767
},
{
"epoch": 0.7060702875399361,
"grad_norm": 0.37246400117874146,
"learning_rate": 2.963258785942492e-05,
"loss": 0.0024,
"step": 1768
},
{
"epoch": 0.7064696485623003,
"grad_norm": 7.174455642700195,
"learning_rate": 2.95926517571885e-05,
"loss": 0.042,
"step": 1769
},
{
"epoch": 0.7068690095846646,
"grad_norm": 27.808277130126953,
"learning_rate": 2.955271565495208e-05,
"loss": 0.4359,
"step": 1770
},
{
"epoch": 0.7072683706070287,
"grad_norm": 1.0314329862594604,
"learning_rate": 2.9512779552715658e-05,
"loss": 0.005,
"step": 1771
},
{
"epoch": 0.707667731629393,
"grad_norm": 29.471818923950195,
"learning_rate": 2.9472843450479238e-05,
"loss": 0.5137,
"step": 1772
},
{
"epoch": 0.7080670926517572,
"grad_norm": 30.447032928466797,
"learning_rate": 2.943290734824281e-05,
"loss": 1.1244,
"step": 1773
},
{
"epoch": 0.7084664536741214,
"grad_norm": 26.38878631591797,
"learning_rate": 2.939297124600639e-05,
"loss": 0.5333,
"step": 1774
},
{
"epoch": 0.7088658146964856,
"grad_norm": 0.014631543308496475,
"learning_rate": 2.935303514376997e-05,
"loss": 0.0001,
"step": 1775
},
{
"epoch": 0.7092651757188498,
"grad_norm": 26.239370346069336,
"learning_rate": 2.931309904153355e-05,
"loss": 0.3512,
"step": 1776
},
{
"epoch": 0.7096645367412141,
"grad_norm": 24.212711334228516,
"learning_rate": 2.9273162939297128e-05,
"loss": 0.2595,
"step": 1777
},
{
"epoch": 0.7100638977635783,
"grad_norm": 0.8857204914093018,
"learning_rate": 2.9233226837060707e-05,
"loss": 0.004,
"step": 1778
},
{
"epoch": 0.7104632587859425,
"grad_norm": 8.431693077087402,
"learning_rate": 2.919329073482428e-05,
"loss": 0.184,
"step": 1779
},
{
"epoch": 0.7108626198083067,
"grad_norm": 14.892561912536621,
"learning_rate": 2.915335463258786e-05,
"loss": 0.1849,
"step": 1780
},
{
"epoch": 0.7112619808306709,
"grad_norm": 8.936697006225586,
"learning_rate": 2.911341853035144e-05,
"loss": 0.1551,
"step": 1781
},
{
"epoch": 0.7116613418530351,
"grad_norm": 30.382755279541016,
"learning_rate": 2.907348242811502e-05,
"loss": 1.2419,
"step": 1782
},
{
"epoch": 0.7120607028753994,
"grad_norm": 1.0230177640914917,
"learning_rate": 2.9033546325878598e-05,
"loss": 0.0104,
"step": 1783
},
{
"epoch": 0.7124600638977636,
"grad_norm": 62.94099044799805,
"learning_rate": 2.8993610223642177e-05,
"loss": 0.7198,
"step": 1784
},
{
"epoch": 0.7128594249201278,
"grad_norm": 23.40749168395996,
"learning_rate": 2.895367412140575e-05,
"loss": 0.3401,
"step": 1785
},
{
"epoch": 0.713258785942492,
"grad_norm": 36.92448043823242,
"learning_rate": 2.891373801916933e-05,
"loss": 0.5548,
"step": 1786
},
{
"epoch": 0.7136581469648562,
"grad_norm": 28.278301239013672,
"learning_rate": 2.887380191693291e-05,
"loss": 0.5288,
"step": 1787
},
{
"epoch": 0.7140575079872205,
"grad_norm": 1.8783739805221558,
"learning_rate": 2.8833865814696488e-05,
"loss": 0.0091,
"step": 1788
},
{
"epoch": 0.7144568690095847,
"grad_norm": 1.7681828737258911,
"learning_rate": 2.8793929712460068e-05,
"loss": 0.0108,
"step": 1789
},
{
"epoch": 0.7148562300319489,
"grad_norm": 27.970117568969727,
"learning_rate": 2.8753993610223644e-05,
"loss": 1.1797,
"step": 1790
},
{
"epoch": 0.7152555910543131,
"grad_norm": 26.68086051940918,
"learning_rate": 2.8714057507987223e-05,
"loss": 0.6876,
"step": 1791
},
{
"epoch": 0.7156549520766773,
"grad_norm": 0.8822888731956482,
"learning_rate": 2.86741214057508e-05,
"loss": 0.0058,
"step": 1792
},
{
"epoch": 0.7160543130990416,
"grad_norm": 17.80529022216797,
"learning_rate": 2.863418530351438e-05,
"loss": 0.1675,
"step": 1793
},
{
"epoch": 0.7164536741214057,
"grad_norm": 2.3040823936462402,
"learning_rate": 2.8594249201277955e-05,
"loss": 0.0112,
"step": 1794
},
{
"epoch": 0.71685303514377,
"grad_norm": 0.007929012179374695,
"learning_rate": 2.8554313099041534e-05,
"loss": 0.0001,
"step": 1795
},
{
"epoch": 0.7172523961661342,
"grad_norm": 25.419710159301758,
"learning_rate": 2.8514376996805113e-05,
"loss": 0.321,
"step": 1796
},
{
"epoch": 0.7176517571884984,
"grad_norm": 23.49529266357422,
"learning_rate": 2.8474440894568693e-05,
"loss": 0.3437,
"step": 1797
},
{
"epoch": 0.7180511182108626,
"grad_norm": 10.225401878356934,
"learning_rate": 2.843450479233227e-05,
"loss": 0.0654,
"step": 1798
},
{
"epoch": 0.7184504792332268,
"grad_norm": 8.978129386901855,
"learning_rate": 2.8394568690095845e-05,
"loss": 0.051,
"step": 1799
},
{
"epoch": 0.7188498402555911,
"grad_norm": 27.620361328125,
"learning_rate": 2.8354632587859424e-05,
"loss": 0.7792,
"step": 1800
},
{
"epoch": 0.7192492012779552,
"grad_norm": 0.8732817769050598,
"learning_rate": 2.8314696485623004e-05,
"loss": 0.0062,
"step": 1801
},
{
"epoch": 0.7196485623003195,
"grad_norm": 1.4758378267288208,
"learning_rate": 2.8274760383386583e-05,
"loss": 0.0103,
"step": 1802
},
{
"epoch": 0.7200479233226837,
"grad_norm": 22.186138153076172,
"learning_rate": 2.8234824281150163e-05,
"loss": 0.2469,
"step": 1803
},
{
"epoch": 0.7204472843450479,
"grad_norm": 49.41655349731445,
"learning_rate": 2.8194888178913742e-05,
"loss": 1.1747,
"step": 1804
},
{
"epoch": 0.7208466453674122,
"grad_norm": 1.8028851747512817,
"learning_rate": 2.8154952076677315e-05,
"loss": 0.0097,
"step": 1805
},
{
"epoch": 0.7212460063897763,
"grad_norm": 0.023198001086711884,
"learning_rate": 2.8115015974440894e-05,
"loss": 0.0003,
"step": 1806
},
{
"epoch": 0.7216453674121406,
"grad_norm": 11.880631446838379,
"learning_rate": 2.8075079872204474e-05,
"loss": 0.2445,
"step": 1807
},
{
"epoch": 0.7220447284345048,
"grad_norm": 36.606746673583984,
"learning_rate": 2.8035143769968053e-05,
"loss": 0.5253,
"step": 1808
},
{
"epoch": 0.722444089456869,
"grad_norm": 14.793195724487305,
"learning_rate": 2.7995207667731632e-05,
"loss": 0.1073,
"step": 1809
},
{
"epoch": 0.7228434504792333,
"grad_norm": 27.433156967163086,
"learning_rate": 2.7955271565495212e-05,
"loss": 1.3556,
"step": 1810
},
{
"epoch": 0.7232428115015974,
"grad_norm": 9.402713775634766,
"learning_rate": 2.7915335463258784e-05,
"loss": 0.078,
"step": 1811
},
{
"epoch": 0.7236421725239617,
"grad_norm": 2.7684760093688965,
"learning_rate": 2.7875399361022364e-05,
"loss": 0.0143,
"step": 1812
},
{
"epoch": 0.7240415335463258,
"grad_norm": 9.32330322265625,
"learning_rate": 2.7835463258785943e-05,
"loss": 0.0527,
"step": 1813
},
{
"epoch": 0.7244408945686901,
"grad_norm": 0.741439163684845,
"learning_rate": 2.7795527156549523e-05,
"loss": 0.004,
"step": 1814
},
{
"epoch": 0.7248402555910544,
"grad_norm": 35.9520378112793,
"learning_rate": 2.7755591054313102e-05,
"loss": 0.4485,
"step": 1815
},
{
"epoch": 0.7252396166134185,
"grad_norm": 22.152671813964844,
"learning_rate": 2.771565495207668e-05,
"loss": 0.3186,
"step": 1816
},
{
"epoch": 0.7256389776357828,
"grad_norm": 0.04724668338894844,
"learning_rate": 2.7675718849840254e-05,
"loss": 0.0003,
"step": 1817
},
{
"epoch": 0.7260383386581469,
"grad_norm": 29.23550033569336,
"learning_rate": 2.7635782747603834e-05,
"loss": 0.6203,
"step": 1818
},
{
"epoch": 0.7264376996805112,
"grad_norm": 2.192640542984009,
"learning_rate": 2.7595846645367413e-05,
"loss": 0.0122,
"step": 1819
},
{
"epoch": 0.7268370607028753,
"grad_norm": 9.211907386779785,
"learning_rate": 2.7555910543130992e-05,
"loss": 0.0899,
"step": 1820
},
{
"epoch": 0.7272364217252396,
"grad_norm": 10.185802459716797,
"learning_rate": 2.7515974440894572e-05,
"loss": 0.0583,
"step": 1821
},
{
"epoch": 0.7276357827476039,
"grad_norm": 2.241572380065918,
"learning_rate": 2.747603833865815e-05,
"loss": 0.013,
"step": 1822
},
{
"epoch": 0.728035143769968,
"grad_norm": 35.477474212646484,
"learning_rate": 2.7436102236421727e-05,
"loss": 0.6049,
"step": 1823
},
{
"epoch": 0.7284345047923323,
"grad_norm": 7.556691646575928,
"learning_rate": 2.7396166134185303e-05,
"loss": 0.0431,
"step": 1824
},
{
"epoch": 0.7288338658146964,
"grad_norm": 15.439949035644531,
"learning_rate": 2.7356230031948883e-05,
"loss": 0.2659,
"step": 1825
},
{
"epoch": 0.7292332268370607,
"grad_norm": 11.104592323303223,
"learning_rate": 2.7316293929712462e-05,
"loss": 0.0734,
"step": 1826
},
{
"epoch": 0.729632587859425,
"grad_norm": 18.421199798583984,
"learning_rate": 2.7276357827476042e-05,
"loss": 0.163,
"step": 1827
},
{
"epoch": 0.7300319488817891,
"grad_norm": 39.135536193847656,
"learning_rate": 2.7236421725239618e-05,
"loss": 0.8346,
"step": 1828
},
{
"epoch": 0.7304313099041534,
"grad_norm": 19.551483154296875,
"learning_rate": 2.7196485623003197e-05,
"loss": 0.1574,
"step": 1829
},
{
"epoch": 0.7308306709265175,
"grad_norm": 0.8751269578933716,
"learning_rate": 2.7156549520766773e-05,
"loss": 0.0052,
"step": 1830
},
{
"epoch": 0.7312300319488818,
"grad_norm": 47.220191955566406,
"learning_rate": 2.7116613418530353e-05,
"loss": 0.9449,
"step": 1831
},
{
"epoch": 0.731629392971246,
"grad_norm": 0.35806116461753845,
"learning_rate": 2.707667731629393e-05,
"loss": 0.0022,
"step": 1832
},
{
"epoch": 0.7320287539936102,
"grad_norm": 40.39183807373047,
"learning_rate": 2.7036741214057508e-05,
"loss": 0.5222,
"step": 1833
},
{
"epoch": 0.7324281150159745,
"grad_norm": 7.567315578460693,
"learning_rate": 2.6996805111821088e-05,
"loss": 0.0655,
"step": 1834
},
{
"epoch": 0.7328274760383386,
"grad_norm": 20.103134155273438,
"learning_rate": 2.6956869009584667e-05,
"loss": 0.1644,
"step": 1835
},
{
"epoch": 0.7332268370607029,
"grad_norm": 40.51066589355469,
"learning_rate": 2.6916932907348246e-05,
"loss": 0.6058,
"step": 1836
},
{
"epoch": 0.7336261980830671,
"grad_norm": 10.805747032165527,
"learning_rate": 2.687699680511182e-05,
"loss": 0.0648,
"step": 1837
},
{
"epoch": 0.7340255591054313,
"grad_norm": 20.027856826782227,
"learning_rate": 2.68370607028754e-05,
"loss": 0.221,
"step": 1838
},
{
"epoch": 0.7344249201277955,
"grad_norm": 32.546478271484375,
"learning_rate": 2.6797124600638978e-05,
"loss": 0.4202,
"step": 1839
},
{
"epoch": 0.7348242811501597,
"grad_norm": 18.23740577697754,
"learning_rate": 2.6757188498402557e-05,
"loss": 0.2131,
"step": 1840
},
{
"epoch": 0.735223642172524,
"grad_norm": 0.28376197814941406,
"learning_rate": 2.6717252396166137e-05,
"loss": 0.0027,
"step": 1841
},
{
"epoch": 0.7356230031948882,
"grad_norm": 34.965370178222656,
"learning_rate": 2.6677316293929716e-05,
"loss": 0.6202,
"step": 1842
},
{
"epoch": 0.7360223642172524,
"grad_norm": 36.734928131103516,
"learning_rate": 2.663738019169329e-05,
"loss": 0.7715,
"step": 1843
},
{
"epoch": 0.7364217252396166,
"grad_norm": 0.027508750557899475,
"learning_rate": 2.6597444089456868e-05,
"loss": 0.0001,
"step": 1844
},
{
"epoch": 0.7368210862619808,
"grad_norm": 25.188220977783203,
"learning_rate": 2.6557507987220448e-05,
"loss": 0.6513,
"step": 1845
},
{
"epoch": 0.737220447284345,
"grad_norm": 2.8948588371276855,
"learning_rate": 2.6517571884984027e-05,
"loss": 0.0238,
"step": 1846
},
{
"epoch": 0.7376198083067093,
"grad_norm": 5.405972480773926,
"learning_rate": 2.6477635782747606e-05,
"loss": 0.032,
"step": 1847
},
{
"epoch": 0.7380191693290735,
"grad_norm": 26.83759307861328,
"learning_rate": 2.6437699680511186e-05,
"loss": 0.6591,
"step": 1848
},
{
"epoch": 0.7384185303514377,
"grad_norm": 26.6892032623291,
"learning_rate": 2.639776357827476e-05,
"loss": 0.3665,
"step": 1849
},
{
"epoch": 0.7388178913738019,
"grad_norm": 1.2713249921798706,
"learning_rate": 2.6357827476038338e-05,
"loss": 0.0113,
"step": 1850
},
{
"epoch": 0.7392172523961661,
"grad_norm": 4.662669658660889,
"learning_rate": 2.6317891373801917e-05,
"loss": 0.0269,
"step": 1851
},
{
"epoch": 0.7396166134185304,
"grad_norm": 59.468204498291016,
"learning_rate": 2.6277955271565497e-05,
"loss": 0.5923,
"step": 1852
},
{
"epoch": 0.7400159744408946,
"grad_norm": 42.279476165771484,
"learning_rate": 2.6238019169329076e-05,
"loss": 0.7792,
"step": 1853
},
{
"epoch": 0.7404153354632588,
"grad_norm": 8.715761184692383,
"learning_rate": 2.6198083067092656e-05,
"loss": 0.0478,
"step": 1854
},
{
"epoch": 0.740814696485623,
"grad_norm": 2.2495293617248535,
"learning_rate": 2.6158146964856235e-05,
"loss": 0.012,
"step": 1855
},
{
"epoch": 0.7412140575079872,
"grad_norm": 35.61282730102539,
"learning_rate": 2.6118210862619808e-05,
"loss": 0.8424,
"step": 1856
},
{
"epoch": 0.7416134185303515,
"grad_norm": 55.19141387939453,
"learning_rate": 2.6078274760383387e-05,
"loss": 0.9011,
"step": 1857
},
{
"epoch": 0.7420127795527156,
"grad_norm": 27.184951782226562,
"learning_rate": 2.6038338658146967e-05,
"loss": 0.8403,
"step": 1858
},
{
"epoch": 0.7424121405750799,
"grad_norm": 19.42462730407715,
"learning_rate": 2.5998402555910546e-05,
"loss": 0.1409,
"step": 1859
},
{
"epoch": 0.7428115015974441,
"grad_norm": 9.517576217651367,
"learning_rate": 2.5958466453674125e-05,
"loss": 0.0616,
"step": 1860
},
{
"epoch": 0.7432108626198083,
"grad_norm": 27.552711486816406,
"learning_rate": 2.59185303514377e-05,
"loss": 0.2563,
"step": 1861
},
{
"epoch": 0.7436102236421726,
"grad_norm": 27.618139266967773,
"learning_rate": 2.5878594249201278e-05,
"loss": 0.9683,
"step": 1862
},
{
"epoch": 0.7440095846645367,
"grad_norm": 0.22927550971508026,
"learning_rate": 2.5838658146964857e-05,
"loss": 0.0014,
"step": 1863
},
{
"epoch": 0.744408945686901,
"grad_norm": 1.0848472118377686,
"learning_rate": 2.5798722044728436e-05,
"loss": 0.0074,
"step": 1864
},
{
"epoch": 0.7448083067092651,
"grad_norm": 23.63319969177246,
"learning_rate": 2.5758785942492016e-05,
"loss": 0.2432,
"step": 1865
},
{
"epoch": 0.7452076677316294,
"grad_norm": 25.738426208496094,
"learning_rate": 2.5718849840255592e-05,
"loss": 0.3847,
"step": 1866
},
{
"epoch": 0.7456070287539937,
"grad_norm": 25.457054138183594,
"learning_rate": 2.567891373801917e-05,
"loss": 0.3593,
"step": 1867
},
{
"epoch": 0.7460063897763578,
"grad_norm": 24.357812881469727,
"learning_rate": 2.563897763578275e-05,
"loss": 0.1958,
"step": 1868
},
{
"epoch": 0.7464057507987221,
"grad_norm": 1.6893336772918701,
"learning_rate": 2.5599041533546327e-05,
"loss": 0.0093,
"step": 1869
},
{
"epoch": 0.7468051118210862,
"grad_norm": 1.5610909461975098,
"learning_rate": 2.5559105431309903e-05,
"loss": 0.0081,
"step": 1870
},
{
"epoch": 0.7472044728434505,
"grad_norm": 9.321490287780762,
"learning_rate": 2.5519169329073482e-05,
"loss": 0.2181,
"step": 1871
},
{
"epoch": 0.7476038338658147,
"grad_norm": 24.42435073852539,
"learning_rate": 2.547923322683706e-05,
"loss": 0.277,
"step": 1872
},
{
"epoch": 0.7480031948881789,
"grad_norm": 3.8002254962921143,
"learning_rate": 2.543929712460064e-05,
"loss": 0.0312,
"step": 1873
},
{
"epoch": 0.7484025559105432,
"grad_norm": 25.692888259887695,
"learning_rate": 2.539936102236422e-05,
"loss": 0.334,
"step": 1874
},
{
"epoch": 0.7488019169329073,
"grad_norm": 29.6540470123291,
"learning_rate": 2.5359424920127793e-05,
"loss": 0.6418,
"step": 1875
},
{
"epoch": 0.7492012779552716,
"grad_norm": 32.178226470947266,
"learning_rate": 2.5319488817891373e-05,
"loss": 0.5684,
"step": 1876
},
{
"epoch": 0.7496006389776357,
"grad_norm": 7.568628311157227,
"learning_rate": 2.5279552715654952e-05,
"loss": 0.0726,
"step": 1877
},
{
"epoch": 0.75,
"grad_norm": 49.11946487426758,
"learning_rate": 2.523961661341853e-05,
"loss": 0.786,
"step": 1878
},
{
"epoch": 0.7503993610223643,
"grad_norm": 21.5537052154541,
"learning_rate": 2.519968051118211e-05,
"loss": 0.2306,
"step": 1879
},
{
"epoch": 0.7507987220447284,
"grad_norm": 27.20477294921875,
"learning_rate": 2.515974440894569e-05,
"loss": 0.7819,
"step": 1880
},
{
"epoch": 0.7511980830670927,
"grad_norm": 17.074106216430664,
"learning_rate": 2.5119808306709263e-05,
"loss": 0.1218,
"step": 1881
},
{
"epoch": 0.7515974440894568,
"grad_norm": 20.554603576660156,
"learning_rate": 2.5079872204472842e-05,
"loss": 0.1831,
"step": 1882
},
{
"epoch": 0.7519968051118211,
"grad_norm": 24.359376907348633,
"learning_rate": 2.5039936102236422e-05,
"loss": 0.5292,
"step": 1883
},
{
"epoch": 0.7523961661341853,
"grad_norm": 1.1500135660171509,
"learning_rate": 2.5e-05,
"loss": 0.0073,
"step": 1884
},
{
"epoch": 0.7527955271565495,
"grad_norm": 29.37909507751465,
"learning_rate": 2.496006389776358e-05,
"loss": 0.5698,
"step": 1885
},
{
"epoch": 0.7531948881789138,
"grad_norm": 5.348720073699951,
"learning_rate": 2.4920127795527157e-05,
"loss": 0.041,
"step": 1886
},
{
"epoch": 0.7535942492012779,
"grad_norm": 0.045818768441677094,
"learning_rate": 2.4880191693290736e-05,
"loss": 0.0002,
"step": 1887
},
{
"epoch": 0.7539936102236422,
"grad_norm": 25.749011993408203,
"learning_rate": 2.4840255591054316e-05,
"loss": 0.9744,
"step": 1888
},
{
"epoch": 0.7543929712460063,
"grad_norm": 4.104675769805908,
"learning_rate": 2.480031948881789e-05,
"loss": 0.0303,
"step": 1889
},
{
"epoch": 0.7547923322683706,
"grad_norm": 17.35829734802246,
"learning_rate": 2.476038338658147e-05,
"loss": 0.1565,
"step": 1890
},
{
"epoch": 0.7551916932907349,
"grad_norm": 5.168485641479492,
"learning_rate": 2.472044728434505e-05,
"loss": 0.026,
"step": 1891
},
{
"epoch": 0.755591054313099,
"grad_norm": 0.05468692630529404,
"learning_rate": 2.4680511182108626e-05,
"loss": 0.0005,
"step": 1892
},
{
"epoch": 0.7559904153354633,
"grad_norm": 12.524698257446289,
"learning_rate": 2.4640575079872206e-05,
"loss": 0.0813,
"step": 1893
},
{
"epoch": 0.7563897763578274,
"grad_norm": 5.245563983917236,
"learning_rate": 2.4600638977635785e-05,
"loss": 0.0269,
"step": 1894
},
{
"epoch": 0.7567891373801917,
"grad_norm": 24.55987548828125,
"learning_rate": 2.4560702875399365e-05,
"loss": 0.2655,
"step": 1895
},
{
"epoch": 0.7571884984025559,
"grad_norm": 41.911922454833984,
"learning_rate": 2.452076677316294e-05,
"loss": 0.6,
"step": 1896
},
{
"epoch": 0.7575878594249201,
"grad_norm": 34.4975700378418,
"learning_rate": 2.448083067092652e-05,
"loss": 0.6545,
"step": 1897
},
{
"epoch": 0.7579872204472844,
"grad_norm": 14.530588150024414,
"learning_rate": 2.44408945686901e-05,
"loss": 0.0917,
"step": 1898
},
{
"epoch": 0.7583865814696485,
"grad_norm": 18.229738235473633,
"learning_rate": 2.4400958466453676e-05,
"loss": 0.1238,
"step": 1899
},
{
"epoch": 0.7587859424920128,
"grad_norm": 29.947608947753906,
"learning_rate": 2.4361022364217255e-05,
"loss": 0.2858,
"step": 1900
},
{
"epoch": 0.759185303514377,
"grad_norm": 35.98379135131836,
"learning_rate": 2.432108626198083e-05,
"loss": 0.4684,
"step": 1901
},
{
"epoch": 0.7595846645367412,
"grad_norm": 16.402294158935547,
"learning_rate": 2.428115015974441e-05,
"loss": 0.1941,
"step": 1902
},
{
"epoch": 0.7599840255591054,
"grad_norm": 0.3438807725906372,
"learning_rate": 2.4241214057507987e-05,
"loss": 0.0079,
"step": 1903
},
{
"epoch": 0.7603833865814696,
"grad_norm": 9.543700218200684,
"learning_rate": 2.4201277955271566e-05,
"loss": 0.0632,
"step": 1904
},
{
"epoch": 0.7607827476038339,
"grad_norm": 28.966257095336914,
"learning_rate": 2.4161341853035145e-05,
"loss": 0.6137,
"step": 1905
},
{
"epoch": 0.7611821086261981,
"grad_norm": 5.125729084014893,
"learning_rate": 2.412140575079872e-05,
"loss": 0.0268,
"step": 1906
},
{
"epoch": 0.7615814696485623,
"grad_norm": 18.944843292236328,
"learning_rate": 2.40814696485623e-05,
"loss": 0.3934,
"step": 1907
},
{
"epoch": 0.7619808306709265,
"grad_norm": 0.2795671820640564,
"learning_rate": 2.4041533546325877e-05,
"loss": 0.0014,
"step": 1908
},
{
"epoch": 0.7623801916932907,
"grad_norm": 2.9400384426116943,
"learning_rate": 2.4001597444089456e-05,
"loss": 0.0207,
"step": 1909
},
{
"epoch": 0.762779552715655,
"grad_norm": 2.9416182041168213,
"learning_rate": 2.3961661341853036e-05,
"loss": 0.0157,
"step": 1910
},
{
"epoch": 0.7631789137380192,
"grad_norm": 14.079166412353516,
"learning_rate": 2.3921725239616615e-05,
"loss": 0.0931,
"step": 1911
},
{
"epoch": 0.7635782747603834,
"grad_norm": 38.94692611694336,
"learning_rate": 2.388178913738019e-05,
"loss": 2.1125,
"step": 1912
},
{
"epoch": 0.7639776357827476,
"grad_norm": 0.5932596921920776,
"learning_rate": 2.384185303514377e-05,
"loss": 0.0095,
"step": 1913
},
{
"epoch": 0.7643769968051118,
"grad_norm": 14.754734992980957,
"learning_rate": 2.380191693290735e-05,
"loss": 0.3401,
"step": 1914
},
{
"epoch": 0.764776357827476,
"grad_norm": 0.11272578686475754,
"learning_rate": 2.3761980830670926e-05,
"loss": 0.0007,
"step": 1915
},
{
"epoch": 0.7651757188498403,
"grad_norm": 0.26745203137397766,
"learning_rate": 2.3722044728434506e-05,
"loss": 0.0021,
"step": 1916
},
{
"epoch": 0.7655750798722045,
"grad_norm": 12.069680213928223,
"learning_rate": 2.3682108626198085e-05,
"loss": 0.7947,
"step": 1917
},
{
"epoch": 0.7659744408945687,
"grad_norm": 2.384843111038208,
"learning_rate": 2.364217252396166e-05,
"loss": 0.0115,
"step": 1918
},
{
"epoch": 0.7663738019169329,
"grad_norm": 14.328927993774414,
"learning_rate": 2.360223642172524e-05,
"loss": 0.162,
"step": 1919
},
{
"epoch": 0.7667731629392971,
"grad_norm": 11.159598350524902,
"learning_rate": 2.356230031948882e-05,
"loss": 0.5653,
"step": 1920
},
{
"epoch": 0.7671725239616614,
"grad_norm": 21.735401153564453,
"learning_rate": 2.3522364217252396e-05,
"loss": 0.5186,
"step": 1921
},
{
"epoch": 0.7675718849840255,
"grad_norm": 5.53592586517334,
"learning_rate": 2.3482428115015975e-05,
"loss": 0.0271,
"step": 1922
},
{
"epoch": 0.7679712460063898,
"grad_norm": 40.26034164428711,
"learning_rate": 2.3442492012779555e-05,
"loss": 2.138,
"step": 1923
},
{
"epoch": 0.768370607028754,
"grad_norm": 18.33621597290039,
"learning_rate": 2.340255591054313e-05,
"loss": 0.1209,
"step": 1924
},
{
"epoch": 0.7687699680511182,
"grad_norm": 4.677008628845215,
"learning_rate": 2.336261980830671e-05,
"loss": 0.0265,
"step": 1925
},
{
"epoch": 0.7691693290734825,
"grad_norm": 0.3163404166698456,
"learning_rate": 2.332268370607029e-05,
"loss": 0.0028,
"step": 1926
},
{
"epoch": 0.7695686900958466,
"grad_norm": 42.921844482421875,
"learning_rate": 2.328274760383387e-05,
"loss": 0.6476,
"step": 1927
},
{
"epoch": 0.7699680511182109,
"grad_norm": 18.02935028076172,
"learning_rate": 2.3242811501597445e-05,
"loss": 0.1317,
"step": 1928
},
{
"epoch": 0.770367412140575,
"grad_norm": 0.5707393884658813,
"learning_rate": 2.3202875399361025e-05,
"loss": 0.0031,
"step": 1929
},
{
"epoch": 0.7707667731629393,
"grad_norm": 32.925697326660156,
"learning_rate": 2.3162939297124604e-05,
"loss": 1.9073,
"step": 1930
},
{
"epoch": 0.7711661341853036,
"grad_norm": 27.82205581665039,
"learning_rate": 2.312300319488818e-05,
"loss": 0.2361,
"step": 1931
},
{
"epoch": 0.7715654952076677,
"grad_norm": 9.846780776977539,
"learning_rate": 2.308306709265176e-05,
"loss": 0.0587,
"step": 1932
},
{
"epoch": 0.771964856230032,
"grad_norm": 0.027774516493082047,
"learning_rate": 2.304313099041534e-05,
"loss": 0.0002,
"step": 1933
},
{
"epoch": 0.7723642172523961,
"grad_norm": 3.0917274951934814,
"learning_rate": 2.3003194888178915e-05,
"loss": 0.0163,
"step": 1934
},
{
"epoch": 0.7727635782747604,
"grad_norm": 0.46168580651283264,
"learning_rate": 2.2963258785942494e-05,
"loss": 0.0022,
"step": 1935
},
{
"epoch": 0.7731629392971247,
"grad_norm": 4.78501033782959,
"learning_rate": 2.2923322683706074e-05,
"loss": 0.0252,
"step": 1936
},
{
"epoch": 0.7735623003194888,
"grad_norm": 0.9523777365684509,
"learning_rate": 2.288338658146965e-05,
"loss": 0.0069,
"step": 1937
},
{
"epoch": 0.7739616613418531,
"grad_norm": 0.7553321719169617,
"learning_rate": 2.284345047923323e-05,
"loss": 0.0051,
"step": 1938
},
{
"epoch": 0.7743610223642172,
"grad_norm": 3.235398769378662,
"learning_rate": 2.2803514376996805e-05,
"loss": 0.0234,
"step": 1939
},
{
"epoch": 0.7747603833865815,
"grad_norm": 23.795608520507812,
"learning_rate": 2.2763578274760385e-05,
"loss": 0.4863,
"step": 1940
},
{
"epoch": 0.7751597444089456,
"grad_norm": 1.325252890586853,
"learning_rate": 2.272364217252396e-05,
"loss": 0.007,
"step": 1941
},
{
"epoch": 0.7755591054313099,
"grad_norm": 46.531883239746094,
"learning_rate": 2.268370607028754e-05,
"loss": 0.9373,
"step": 1942
},
{
"epoch": 0.7759584664536742,
"grad_norm": 48.34103775024414,
"learning_rate": 2.264376996805112e-05,
"loss": 0.8595,
"step": 1943
},
{
"epoch": 0.7763578274760383,
"grad_norm": 62.15392303466797,
"learning_rate": 2.2603833865814696e-05,
"loss": 1.3586,
"step": 1944
},
{
"epoch": 0.7767571884984026,
"grad_norm": 21.986440658569336,
"learning_rate": 2.2563897763578275e-05,
"loss": 0.2059,
"step": 1945
},
{
"epoch": 0.7771565495207667,
"grad_norm": 19.268985748291016,
"learning_rate": 2.2523961661341854e-05,
"loss": 0.1498,
"step": 1946
},
{
"epoch": 0.777555910543131,
"grad_norm": 0.5338484644889832,
"learning_rate": 2.248402555910543e-05,
"loss": 0.0042,
"step": 1947
},
{
"epoch": 0.7779552715654952,
"grad_norm": 22.367355346679688,
"learning_rate": 2.244408945686901e-05,
"loss": 0.3576,
"step": 1948
},
{
"epoch": 0.7783546325878594,
"grad_norm": 0.22647859156131744,
"learning_rate": 2.240415335463259e-05,
"loss": 0.0013,
"step": 1949
},
{
"epoch": 0.7787539936102237,
"grad_norm": 24.617694854736328,
"learning_rate": 2.2364217252396165e-05,
"loss": 0.3473,
"step": 1950
},
{
"epoch": 0.7791533546325878,
"grad_norm": 65.94686126708984,
"learning_rate": 2.2324281150159745e-05,
"loss": 1.5547,
"step": 1951
},
{
"epoch": 0.7795527156549521,
"grad_norm": 0.08273879438638687,
"learning_rate": 2.2284345047923324e-05,
"loss": 0.0006,
"step": 1952
},
{
"epoch": 0.7799520766773163,
"grad_norm": 25.480587005615234,
"learning_rate": 2.22444089456869e-05,
"loss": 0.4868,
"step": 1953
},
{
"epoch": 0.7803514376996805,
"grad_norm": 1.102811574935913,
"learning_rate": 2.220447284345048e-05,
"loss": 0.0055,
"step": 1954
},
{
"epoch": 0.7807507987220448,
"grad_norm": 0.03942089155316353,
"learning_rate": 2.216453674121406e-05,
"loss": 0.0003,
"step": 1955
},
{
"epoch": 0.7811501597444089,
"grad_norm": 20.63765525817871,
"learning_rate": 2.2124600638977635e-05,
"loss": 0.2025,
"step": 1956
},
{
"epoch": 0.7815495207667732,
"grad_norm": 21.358478546142578,
"learning_rate": 2.2084664536741215e-05,
"loss": 0.4334,
"step": 1957
},
{
"epoch": 0.7819488817891374,
"grad_norm": 0.4731053113937378,
"learning_rate": 2.2044728434504794e-05,
"loss": 0.0023,
"step": 1958
},
{
"epoch": 0.7823482428115016,
"grad_norm": 2.595877170562744,
"learning_rate": 2.2004792332268373e-05,
"loss": 0.0162,
"step": 1959
},
{
"epoch": 0.7827476038338658,
"grad_norm": 3.825486183166504,
"learning_rate": 2.196485623003195e-05,
"loss": 0.0248,
"step": 1960
},
{
"epoch": 0.78314696485623,
"grad_norm": 16.529983520507812,
"learning_rate": 2.192492012779553e-05,
"loss": 0.1778,
"step": 1961
},
{
"epoch": 0.7835463258785943,
"grad_norm": 0.37582963705062866,
"learning_rate": 2.188498402555911e-05,
"loss": 0.0018,
"step": 1962
},
{
"epoch": 0.7839456869009584,
"grad_norm": 9.193746566772461,
"learning_rate": 2.1845047923322684e-05,
"loss": 0.0548,
"step": 1963
},
{
"epoch": 0.7843450479233227,
"grad_norm": 4.836997985839844,
"learning_rate": 2.1805111821086264e-05,
"loss": 0.0352,
"step": 1964
},
{
"epoch": 0.7847444089456869,
"grad_norm": 6.8939714431762695,
"learning_rate": 2.1765175718849843e-05,
"loss": 0.0432,
"step": 1965
},
{
"epoch": 0.7851437699680511,
"grad_norm": 18.11557388305664,
"learning_rate": 2.172523961661342e-05,
"loss": 0.7722,
"step": 1966
},
{
"epoch": 0.7855431309904153,
"grad_norm": 10.954044342041016,
"learning_rate": 2.1685303514377e-05,
"loss": 0.0807,
"step": 1967
},
{
"epoch": 0.7859424920127795,
"grad_norm": 13.73788833618164,
"learning_rate": 2.1645367412140578e-05,
"loss": 0.0881,
"step": 1968
},
{
"epoch": 0.7863418530351438,
"grad_norm": 22.047380447387695,
"learning_rate": 2.1605431309904154e-05,
"loss": 0.7137,
"step": 1969
},
{
"epoch": 0.786741214057508,
"grad_norm": 22.64723014831543,
"learning_rate": 2.1565495207667734e-05,
"loss": 0.2437,
"step": 1970
},
{
"epoch": 0.7871405750798722,
"grad_norm": 19.854734420776367,
"learning_rate": 2.1525559105431313e-05,
"loss": 0.1775,
"step": 1971
},
{
"epoch": 0.7875399361022364,
"grad_norm": 4.277410984039307,
"learning_rate": 2.148562300319489e-05,
"loss": 0.0226,
"step": 1972
},
{
"epoch": 0.7879392971246006,
"grad_norm": 2.1209876537323,
"learning_rate": 2.144568690095847e-05,
"loss": 0.0259,
"step": 1973
},
{
"epoch": 0.7883386581469649,
"grad_norm": 3.6271777153015137,
"learning_rate": 2.1405750798722048e-05,
"loss": 0.0426,
"step": 1974
},
{
"epoch": 0.7887380191693291,
"grad_norm": 23.32756805419922,
"learning_rate": 2.1365814696485624e-05,
"loss": 0.7425,
"step": 1975
},
{
"epoch": 0.7891373801916933,
"grad_norm": 25.394142150878906,
"learning_rate": 2.1325878594249203e-05,
"loss": 0.5263,
"step": 1976
},
{
"epoch": 0.7895367412140575,
"grad_norm": 0.2824519872665405,
"learning_rate": 2.128594249201278e-05,
"loss": 0.002,
"step": 1977
},
{
"epoch": 0.7899361022364217,
"grad_norm": 0.22561533749103546,
"learning_rate": 2.124600638977636e-05,
"loss": 0.0014,
"step": 1978
},
{
"epoch": 0.7903354632587859,
"grad_norm": 4.569544315338135,
"learning_rate": 2.1206070287539935e-05,
"loss": 0.0278,
"step": 1979
},
{
"epoch": 0.7907348242811502,
"grad_norm": 29.087385177612305,
"learning_rate": 2.1166134185303514e-05,
"loss": 0.725,
"step": 1980
},
{
"epoch": 0.7911341853035144,
"grad_norm": 1.9695864915847778,
"learning_rate": 2.1126198083067094e-05,
"loss": 0.0108,
"step": 1981
},
{
"epoch": 0.7915335463258786,
"grad_norm": 0.03727534040808678,
"learning_rate": 2.108626198083067e-05,
"loss": 0.0003,
"step": 1982
},
{
"epoch": 0.7919329073482428,
"grad_norm": 32.19395065307617,
"learning_rate": 2.104632587859425e-05,
"loss": 0.5024,
"step": 1983
},
{
"epoch": 0.792332268370607,
"grad_norm": 0.2529641091823578,
"learning_rate": 2.100638977635783e-05,
"loss": 0.0012,
"step": 1984
},
{
"epoch": 0.7927316293929713,
"grad_norm": 45.88161849975586,
"learning_rate": 2.0966453674121405e-05,
"loss": 1.2488,
"step": 1985
},
{
"epoch": 0.7931309904153354,
"grad_norm": 2.647453546524048,
"learning_rate": 2.0926517571884984e-05,
"loss": 0.0139,
"step": 1986
},
{
"epoch": 0.7935303514376997,
"grad_norm": 9.707804679870605,
"learning_rate": 2.0886581469648563e-05,
"loss": 0.0882,
"step": 1987
},
{
"epoch": 0.7939297124600639,
"grad_norm": 25.47239112854004,
"learning_rate": 2.084664536741214e-05,
"loss": 0.5409,
"step": 1988
},
{
"epoch": 0.7943290734824281,
"grad_norm": 0.06127059459686279,
"learning_rate": 2.080670926517572e-05,
"loss": 0.0004,
"step": 1989
},
{
"epoch": 0.7947284345047924,
"grad_norm": 20.8686580657959,
"learning_rate": 2.07667731629393e-05,
"loss": 0.332,
"step": 1990
},
{
"epoch": 0.7951277955271565,
"grad_norm": 13.513041496276855,
"learning_rate": 2.0726837060702874e-05,
"loss": 0.1576,
"step": 1991
},
{
"epoch": 0.7955271565495208,
"grad_norm": 1.6546295881271362,
"learning_rate": 2.0686900958466454e-05,
"loss": 0.0105,
"step": 1992
},
{
"epoch": 0.795926517571885,
"grad_norm": 29.154664993286133,
"learning_rate": 2.0646964856230033e-05,
"loss": 1.0127,
"step": 1993
},
{
"epoch": 0.7963258785942492,
"grad_norm": 25.465614318847656,
"learning_rate": 2.0607028753993613e-05,
"loss": 0.6293,
"step": 1994
},
{
"epoch": 0.7967252396166135,
"grad_norm": 10.496687889099121,
"learning_rate": 2.056709265175719e-05,
"loss": 0.0605,
"step": 1995
},
{
"epoch": 0.7971246006389776,
"grad_norm": 22.776596069335938,
"learning_rate": 2.0527156549520768e-05,
"loss": 0.4882,
"step": 1996
},
{
"epoch": 0.7975239616613419,
"grad_norm": 37.33203125,
"learning_rate": 2.0487220447284348e-05,
"loss": 0.6127,
"step": 1997
},
{
"epoch": 0.797923322683706,
"grad_norm": 27.224748611450195,
"learning_rate": 2.0447284345047924e-05,
"loss": 0.3837,
"step": 1998
},
{
"epoch": 0.7983226837060703,
"grad_norm": 27.56450653076172,
"learning_rate": 2.0407348242811503e-05,
"loss": 1.1253,
"step": 1999
},
{
"epoch": 0.7987220447284346,
"grad_norm": 4.094717979431152,
"learning_rate": 2.0367412140575082e-05,
"loss": 0.0301,
"step": 2000
},
{
"epoch": 0.7991214057507987,
"grad_norm": 20.570545196533203,
"learning_rate": 2.032747603833866e-05,
"loss": 0.27,
"step": 2001
},
{
"epoch": 0.799520766773163,
"grad_norm": 27.64883804321289,
"learning_rate": 2.0287539936102238e-05,
"loss": 0.5845,
"step": 2002
},
{
"epoch": 0.7999201277955271,
"grad_norm": 13.820975303649902,
"learning_rate": 2.0247603833865817e-05,
"loss": 0.2909,
"step": 2003
},
{
"epoch": 0.8003194888178914,
"grad_norm": 16.222801208496094,
"learning_rate": 2.0207667731629393e-05,
"loss": 0.6165,
"step": 2004
},
{
"epoch": 0.8007188498402555,
"grad_norm": 3.2522950172424316,
"learning_rate": 2.0167731629392973e-05,
"loss": 0.0157,
"step": 2005
},
{
"epoch": 0.8011182108626198,
"grad_norm": 25.17967987060547,
"learning_rate": 2.0127795527156552e-05,
"loss": 0.3438,
"step": 2006
},
{
"epoch": 0.8015175718849841,
"grad_norm": 3.792173147201538,
"learning_rate": 2.0087859424920128e-05,
"loss": 0.0189,
"step": 2007
},
{
"epoch": 0.8019169329073482,
"grad_norm": 8.371594429016113,
"learning_rate": 2.0047923322683708e-05,
"loss": 0.305,
"step": 2008
},
{
"epoch": 0.8023162939297125,
"grad_norm": 16.190778732299805,
"learning_rate": 2.0007987220447287e-05,
"loss": 0.1051,
"step": 2009
},
{
"epoch": 0.8027156549520766,
"grad_norm": 25.592206954956055,
"learning_rate": 1.9968051118210863e-05,
"loss": 0.2439,
"step": 2010
},
{
"epoch": 0.8031150159744409,
"grad_norm": 27.271949768066406,
"learning_rate": 1.9928115015974443e-05,
"loss": 0.4984,
"step": 2011
},
{
"epoch": 0.8035143769968051,
"grad_norm": 30.888286590576172,
"learning_rate": 1.9888178913738022e-05,
"loss": 0.4881,
"step": 2012
},
{
"epoch": 0.8039137380191693,
"grad_norm": 40.29209899902344,
"learning_rate": 1.9848242811501598e-05,
"loss": 0.7451,
"step": 2013
},
{
"epoch": 0.8043130990415336,
"grad_norm": 9.887882232666016,
"learning_rate": 1.9808306709265177e-05,
"loss": 0.0579,
"step": 2014
},
{
"epoch": 0.8047124600638977,
"grad_norm": 4.055179595947266,
"learning_rate": 1.9768370607028753e-05,
"loss": 0.039,
"step": 2015
},
{
"epoch": 0.805111821086262,
"grad_norm": NaN,
"learning_rate": 1.9768370607028753e-05,
"loss": 0.0,
"step": 2016
},
{
"epoch": 0.8055111821086262,
"grad_norm": 0.05239780247211456,
"learning_rate": 1.9728434504792333e-05,
"loss": 0.0004,
"step": 2017
},
{
"epoch": 0.8059105431309904,
"grad_norm": 24.955101013183594,
"learning_rate": 1.968849840255591e-05,
"loss": 1.0647,
"step": 2018
},
{
"epoch": 0.8063099041533547,
"grad_norm": 0.6851922273635864,
"learning_rate": 1.964856230031949e-05,
"loss": 0.0037,
"step": 2019
},
{
"epoch": 0.8067092651757188,
"grad_norm": 43.424869537353516,
"learning_rate": 1.9608626198083068e-05,
"loss": 0.6338,
"step": 2020
},
{
"epoch": 0.8071086261980831,
"grad_norm": 8.280627250671387,
"learning_rate": 1.9568690095846644e-05,
"loss": 0.0465,
"step": 2021
},
{
"epoch": 0.8075079872204473,
"grad_norm": 10.086182594299316,
"learning_rate": 1.9528753993610223e-05,
"loss": 0.0607,
"step": 2022
},
{
"epoch": 0.8079073482428115,
"grad_norm": 6.562862396240234,
"learning_rate": 1.9488817891373803e-05,
"loss": 0.0449,
"step": 2023
},
{
"epoch": 0.8083067092651757,
"grad_norm": 25.09941864013672,
"learning_rate": 1.944888178913738e-05,
"loss": 0.2785,
"step": 2024
},
{
"epoch": 0.8087060702875399,
"grad_norm": 25.251237869262695,
"learning_rate": 1.9408945686900958e-05,
"loss": 0.3189,
"step": 2025
},
{
"epoch": 0.8091054313099042,
"grad_norm": 0.03467867150902748,
"learning_rate": 1.9369009584664538e-05,
"loss": 0.0002,
"step": 2026
},
{
"epoch": 0.8095047923322684,
"grad_norm": 26.8580322265625,
"learning_rate": 1.9329073482428117e-05,
"loss": 0.6184,
"step": 2027
},
{
"epoch": 0.8099041533546326,
"grad_norm": 27.285167694091797,
"learning_rate": 1.9289137380191693e-05,
"loss": 0.4969,
"step": 2028
},
{
"epoch": 0.8103035143769968,
"grad_norm": 26.7601318359375,
"learning_rate": 1.9249201277955272e-05,
"loss": 1.0765,
"step": 2029
},
{
"epoch": 0.810702875399361,
"grad_norm": 46.121551513671875,
"learning_rate": 1.9209265175718852e-05,
"loss": 0.6369,
"step": 2030
},
{
"epoch": 0.8111022364217252,
"grad_norm": 0.04755905643105507,
"learning_rate": 1.9169329073482428e-05,
"loss": 0.0003,
"step": 2031
},
{
"epoch": 0.8115015974440895,
"grad_norm": 0.9151645302772522,
"learning_rate": 1.9129392971246007e-05,
"loss": 0.0052,
"step": 2032
},
{
"epoch": 0.8119009584664537,
"grad_norm": 21.00222396850586,
"learning_rate": 1.9089456869009587e-05,
"loss": 0.2004,
"step": 2033
},
{
"epoch": 0.8123003194888179,
"grad_norm": 26.967365264892578,
"learning_rate": 1.9049520766773163e-05,
"loss": 0.85,
"step": 2034
},
{
"epoch": 0.8126996805111821,
"grad_norm": 1.8073612451553345,
"learning_rate": 1.9009584664536742e-05,
"loss": 0.0111,
"step": 2035
},
{
"epoch": 0.8130990415335463,
"grad_norm": 48.42094421386719,
"learning_rate": 1.896964856230032e-05,
"loss": 0.7448,
"step": 2036
},
{
"epoch": 0.8134984025559105,
"grad_norm": 4.510549545288086,
"learning_rate": 1.8929712460063898e-05,
"loss": 0.0254,
"step": 2037
},
{
"epoch": 0.8138977635782748,
"grad_norm": 27.159839630126953,
"learning_rate": 1.8889776357827477e-05,
"loss": 1.1698,
"step": 2038
},
{
"epoch": 0.814297124600639,
"grad_norm": 16.140609741210938,
"learning_rate": 1.8849840255591057e-05,
"loss": 0.176,
"step": 2039
},
{
"epoch": 0.8146964856230032,
"grad_norm": 30.619962692260742,
"learning_rate": 1.8809904153354633e-05,
"loss": 1.3897,
"step": 2040
},
{
"epoch": 0.8150958466453674,
"grad_norm": 23.231786727905273,
"learning_rate": 1.8769968051118212e-05,
"loss": 0.381,
"step": 2041
},
{
"epoch": 0.8154952076677316,
"grad_norm": 5.284379482269287,
"learning_rate": 1.873003194888179e-05,
"loss": 0.0364,
"step": 2042
},
{
"epoch": 0.8158945686900958,
"grad_norm": 28.24431610107422,
"learning_rate": 1.869009584664537e-05,
"loss": 0.618,
"step": 2043
},
{
"epoch": 0.8162939297124601,
"grad_norm": 0.7619858384132385,
"learning_rate": 1.8650159744408947e-05,
"loss": 0.0074,
"step": 2044
},
{
"epoch": 0.8166932907348243,
"grad_norm": 24.426652908325195,
"learning_rate": 1.8610223642172526e-05,
"loss": 0.4968,
"step": 2045
},
{
"epoch": 0.8170926517571885,
"grad_norm": 14.167215347290039,
"learning_rate": 1.8570287539936106e-05,
"loss": 0.1197,
"step": 2046
},
{
"epoch": 0.8174920127795527,
"grad_norm": 1.9893509149551392,
"learning_rate": 1.8530351437699682e-05,
"loss": 0.0117,
"step": 2047
},
{
"epoch": 0.8178913738019169,
"grad_norm": 21.341678619384766,
"learning_rate": 1.849041533546326e-05,
"loss": 0.1803,
"step": 2048
},
{
"epoch": 0.8182907348242812,
"grad_norm": 49.41813659667969,
"learning_rate": 1.8450479233226837e-05,
"loss": 1.4563,
"step": 2049
},
{
"epoch": 0.8186900958466453,
"grad_norm": 27.773418426513672,
"learning_rate": 1.8410543130990417e-05,
"loss": 0.5219,
"step": 2050
},
{
"epoch": 0.8190894568690096,
"grad_norm": 3.51265811920166,
"learning_rate": 1.8370607028753993e-05,
"loss": 0.0283,
"step": 2051
},
{
"epoch": 0.8194888178913738,
"grad_norm": 14.190735816955566,
"learning_rate": 1.8330670926517572e-05,
"loss": 0.09,
"step": 2052
},
{
"epoch": 0.819888178913738,
"grad_norm": 36.42401885986328,
"learning_rate": 1.829073482428115e-05,
"loss": 0.3333,
"step": 2053
},
{
"epoch": 0.8202875399361023,
"grad_norm": 0.17509286105632782,
"learning_rate": 1.8250798722044728e-05,
"loss": 0.0013,
"step": 2054
},
{
"epoch": 0.8206869009584664,
"grad_norm": 3.6796927452087402,
"learning_rate": 1.8210862619808307e-05,
"loss": 0.028,
"step": 2055
},
{
"epoch": 0.8210862619808307,
"grad_norm": 6.2489166259765625,
"learning_rate": 1.8170926517571883e-05,
"loss": 0.2223,
"step": 2056
},
{
"epoch": 0.8214856230031949,
"grad_norm": 7.9750871658325195,
"learning_rate": 1.8130990415335463e-05,
"loss": 0.0445,
"step": 2057
},
{
"epoch": 0.8218849840255591,
"grad_norm": 6.204566955566406,
"learning_rate": 1.8091054313099042e-05,
"loss": 0.0477,
"step": 2058
},
{
"epoch": 0.8222843450479234,
"grad_norm": 12.677854537963867,
"learning_rate": 1.805111821086262e-05,
"loss": 0.0795,
"step": 2059
},
{
"epoch": 0.8226837060702875,
"grad_norm": 0.11914076656103134,
"learning_rate": 1.8011182108626197e-05,
"loss": 0.0009,
"step": 2060
},
{
"epoch": 0.8230830670926518,
"grad_norm": 5.898823261260986,
"learning_rate": 1.7971246006389777e-05,
"loss": 0.0313,
"step": 2061
},
{
"epoch": 0.8234824281150159,
"grad_norm": 0.49199068546295166,
"learning_rate": 1.7931309904153356e-05,
"loss": 0.0024,
"step": 2062
},
{
"epoch": 0.8238817891373802,
"grad_norm": 17.422271728515625,
"learning_rate": 1.7891373801916932e-05,
"loss": 0.1659,
"step": 2063
},
{
"epoch": 0.8242811501597445,
"grad_norm": 18.48604965209961,
"learning_rate": 1.7851437699680512e-05,
"loss": 0.5283,
"step": 2064
},
{
"epoch": 0.8246805111821086,
"grad_norm": 9.180458068847656,
"learning_rate": 1.781150159744409e-05,
"loss": 0.0534,
"step": 2065
},
{
"epoch": 0.8250798722044729,
"grad_norm": 3.1538937091827393,
"learning_rate": 1.7771565495207667e-05,
"loss": 0.0156,
"step": 2066
},
{
"epoch": 0.825479233226837,
"grad_norm": 14.112957000732422,
"learning_rate": 1.7731629392971247e-05,
"loss": 0.1385,
"step": 2067
},
{
"epoch": 0.8258785942492013,
"grad_norm": 19.790239334106445,
"learning_rate": 1.7691693290734826e-05,
"loss": 0.1765,
"step": 2068
},
{
"epoch": 0.8262779552715654,
"grad_norm": 6.075618267059326,
"learning_rate": 1.7651757188498402e-05,
"loss": 0.0424,
"step": 2069
},
{
"epoch": 0.8266773162939297,
"grad_norm": 0.46476125717163086,
"learning_rate": 1.761182108626198e-05,
"loss": 0.0044,
"step": 2070
},
{
"epoch": 0.827076677316294,
"grad_norm": 13.75184154510498,
"learning_rate": 1.757188498402556e-05,
"loss": 0.1609,
"step": 2071
},
{
"epoch": 0.8274760383386581,
"grad_norm": 49.82762145996094,
"learning_rate": 1.7531948881789137e-05,
"loss": 1.1167,
"step": 2072
},
{
"epoch": 0.8278753993610224,
"grad_norm": 2.3235244750976562,
"learning_rate": 1.7492012779552716e-05,
"loss": 0.0137,
"step": 2073
},
{
"epoch": 0.8282747603833865,
"grad_norm": 26.00851821899414,
"learning_rate": 1.7452076677316296e-05,
"loss": 0.4016,
"step": 2074
},
{
"epoch": 0.8286741214057508,
"grad_norm": 43.63404846191406,
"learning_rate": 1.7412140575079875e-05,
"loss": 0.9574,
"step": 2075
},
{
"epoch": 0.829073482428115,
"grad_norm": 23.824705123901367,
"learning_rate": 1.737220447284345e-05,
"loss": 1.3035,
"step": 2076
},
{
"epoch": 0.8294728434504792,
"grad_norm": 10.624068260192871,
"learning_rate": 1.733226837060703e-05,
"loss": 0.0638,
"step": 2077
},
{
"epoch": 0.8298722044728435,
"grad_norm": 38.459171295166016,
"learning_rate": 1.729233226837061e-05,
"loss": 0.5098,
"step": 2078
},
{
"epoch": 0.8302715654952076,
"grad_norm": 27.814769744873047,
"learning_rate": 1.7252396166134186e-05,
"loss": 0.5435,
"step": 2079
},
{
"epoch": 0.8306709265175719,
"grad_norm": 9.73934555053711,
"learning_rate": 1.7212460063897766e-05,
"loss": 0.0595,
"step": 2080
},
{
"epoch": 0.8310702875399361,
"grad_norm": 1.2187345027923584,
"learning_rate": 1.7172523961661345e-05,
"loss": 0.008,
"step": 2081
},
{
"epoch": 0.8314696485623003,
"grad_norm": 7.023210525512695,
"learning_rate": 1.713258785942492e-05,
"loss": 0.0431,
"step": 2082
},
{
"epoch": 0.8318690095846646,
"grad_norm": 24.067867279052734,
"learning_rate": 1.70926517571885e-05,
"loss": 0.2341,
"step": 2083
},
{
"epoch": 0.8322683706070287,
"grad_norm": 3.418260097503662,
"learning_rate": 1.705271565495208e-05,
"loss": 0.0212,
"step": 2084
},
{
"epoch": 0.832667731629393,
"grad_norm": 16.629919052124023,
"learning_rate": 1.7012779552715656e-05,
"loss": 0.1497,
"step": 2085
},
{
"epoch": 0.8330670926517572,
"grad_norm": 0.7153478860855103,
"learning_rate": 1.6972843450479235e-05,
"loss": 0.0083,
"step": 2086
},
{
"epoch": 0.8334664536741214,
"grad_norm": 37.845619201660156,
"learning_rate": 1.693290734824281e-05,
"loss": 0.3354,
"step": 2087
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.2952727973461151,
"learning_rate": 1.689297124600639e-05,
"loss": 0.0022,
"step": 2088
},
{
"epoch": 0.8342651757188498,
"grad_norm": 10.644645690917969,
"learning_rate": 1.6853035143769967e-05,
"loss": 0.063,
"step": 2089
},
{
"epoch": 0.8346645367412141,
"grad_norm": 0.1481899917125702,
"learning_rate": 1.6813099041533546e-05,
"loss": 0.001,
"step": 2090
},
{
"epoch": 0.8350638977635783,
"grad_norm": 26.899490356445312,
"learning_rate": 1.6773162939297126e-05,
"loss": 0.844,
"step": 2091
},
{
"epoch": 0.8354632587859425,
"grad_norm": 34.56169509887695,
"learning_rate": 1.6733226837060702e-05,
"loss": 0.6515,
"step": 2092
},
{
"epoch": 0.8358626198083067,
"grad_norm": 23.621883392333984,
"learning_rate": 1.669329073482428e-05,
"loss": 0.7375,
"step": 2093
},
{
"epoch": 0.8362619808306709,
"grad_norm": 10.25305461883545,
"learning_rate": 1.665335463258786e-05,
"loss": 0.0625,
"step": 2094
},
{
"epoch": 0.8366613418530351,
"grad_norm": 30.474939346313477,
"learning_rate": 1.6613418530351437e-05,
"loss": 1.2193,
"step": 2095
},
{
"epoch": 0.8370607028753994,
"grad_norm": 5.237939834594727,
"learning_rate": 1.6573482428115016e-05,
"loss": 0.1217,
"step": 2096
},
{
"epoch": 0.8374600638977636,
"grad_norm": 26.669078826904297,
"learning_rate": 1.6533546325878595e-05,
"loss": 0.5011,
"step": 2097
},
{
"epoch": 0.8378594249201278,
"grad_norm": 22.223894119262695,
"learning_rate": 1.649361022364217e-05,
"loss": 0.2295,
"step": 2098
},
{
"epoch": 0.838258785942492,
"grad_norm": 14.179681777954102,
"learning_rate": 1.645367412140575e-05,
"loss": 0.1004,
"step": 2099
},
{
"epoch": 0.8386581469648562,
"grad_norm": 15.955572128295898,
"learning_rate": 1.641373801916933e-05,
"loss": 0.1441,
"step": 2100
},
{
"epoch": 0.8390575079872205,
"grad_norm": 1.5127172470092773,
"learning_rate": 1.6373801916932906e-05,
"loss": 0.0139,
"step": 2101
},
{
"epoch": 0.8394568690095847,
"grad_norm": 27.79009437561035,
"learning_rate": 1.6333865814696486e-05,
"loss": 0.9633,
"step": 2102
},
{
"epoch": 0.8398562300319489,
"grad_norm": 42.68457794189453,
"learning_rate": 1.6293929712460065e-05,
"loss": 0.9236,
"step": 2103
},
{
"epoch": 0.8402555910543131,
"grad_norm": 36.883689880371094,
"learning_rate": 1.625399361022364e-05,
"loss": 0.2844,
"step": 2104
},
{
"epoch": 0.8406549520766773,
"grad_norm": 4.79236364364624,
"learning_rate": 1.621405750798722e-05,
"loss": 0.0288,
"step": 2105
},
{
"epoch": 0.8410543130990416,
"grad_norm": 13.579639434814453,
"learning_rate": 1.61741214057508e-05,
"loss": 0.1166,
"step": 2106
},
{
"epoch": 0.8414536741214057,
"grad_norm": 33.839725494384766,
"learning_rate": 1.6134185303514376e-05,
"loss": 0.5547,
"step": 2107
},
{
"epoch": 0.84185303514377,
"grad_norm": 49.261844635009766,
"learning_rate": 1.6094249201277956e-05,
"loss": 0.5594,
"step": 2108
},
{
"epoch": 0.8422523961661342,
"grad_norm": 24.08683204650879,
"learning_rate": 1.6054313099041535e-05,
"loss": 0.4575,
"step": 2109
},
{
"epoch": 0.8426517571884984,
"grad_norm": 26.96343994140625,
"learning_rate": 1.6014376996805114e-05,
"loss": 0.8266,
"step": 2110
},
{
"epoch": 0.8430511182108626,
"grad_norm": 12.93642520904541,
"learning_rate": 1.597444089456869e-05,
"loss": 0.1324,
"step": 2111
},
{
"epoch": 0.8434504792332268,
"grad_norm": 21.444591522216797,
"learning_rate": 1.593450479233227e-05,
"loss": 0.2067,
"step": 2112
},
{
"epoch": 0.8438498402555911,
"grad_norm": 5.765625953674316,
"learning_rate": 1.589456869009585e-05,
"loss": 0.0355,
"step": 2113
},
{
"epoch": 0.8442492012779552,
"grad_norm": 7.043456554412842,
"learning_rate": 1.5854632587859425e-05,
"loss": 0.0481,
"step": 2114
},
{
"epoch": 0.8446485623003195,
"grad_norm": 44.436767578125,
"learning_rate": 1.5814696485623005e-05,
"loss": 1.1182,
"step": 2115
},
{
"epoch": 0.8450479233226837,
"grad_norm": 1.7820690870285034,
"learning_rate": 1.5774760383386584e-05,
"loss": 0.013,
"step": 2116
},
{
"epoch": 0.8454472843450479,
"grad_norm": 29.34307289123535,
"learning_rate": 1.573482428115016e-05,
"loss": 0.5128,
"step": 2117
},
{
"epoch": 0.8458466453674122,
"grad_norm": 0.3048868775367737,
"learning_rate": 1.569488817891374e-05,
"loss": 0.0033,
"step": 2118
},
{
"epoch": 0.8462460063897763,
"grad_norm": 18.553211212158203,
"learning_rate": 1.565495207667732e-05,
"loss": 0.1462,
"step": 2119
},
{
"epoch": 0.8466453674121406,
"grad_norm": 0.07028963416814804,
"learning_rate": 1.5615015974440895e-05,
"loss": 0.0006,
"step": 2120
},
{
"epoch": 0.8470447284345048,
"grad_norm": 23.51447296142578,
"learning_rate": 1.5575079872204475e-05,
"loss": 0.2442,
"step": 2121
},
{
"epoch": 0.847444089456869,
"grad_norm": 6.016641616821289,
"learning_rate": 1.5535143769968054e-05,
"loss": 0.0317,
"step": 2122
},
{
"epoch": 0.8478434504792333,
"grad_norm": 0.2717338800430298,
"learning_rate": 1.549520766773163e-05,
"loss": 0.0014,
"step": 2123
},
{
"epoch": 0.8482428115015974,
"grad_norm": 33.222747802734375,
"learning_rate": 1.545527156549521e-05,
"loss": 0.3242,
"step": 2124
},
{
"epoch": 0.8486421725239617,
"grad_norm": 15.63247299194336,
"learning_rate": 1.5415335463258786e-05,
"loss": 0.1301,
"step": 2125
},
{
"epoch": 0.8490415335463258,
"grad_norm": 15.038786888122559,
"learning_rate": 1.5375399361022365e-05,
"loss": 0.1022,
"step": 2126
},
{
"epoch": 0.8494408945686901,
"grad_norm": 26.29620933532715,
"learning_rate": 1.533546325878594e-05,
"loss": 0.4729,
"step": 2127
},
{
"epoch": 0.8498402555910544,
"grad_norm": 22.74943733215332,
"learning_rate": 1.529552715654952e-05,
"loss": 0.4069,
"step": 2128
},
{
"epoch": 0.8502396166134185,
"grad_norm": 36.124481201171875,
"learning_rate": 1.5255591054313102e-05,
"loss": 0.3412,
"step": 2129
},
{
"epoch": 0.8506389776357828,
"grad_norm": 27.65908432006836,
"learning_rate": 1.5215654952076678e-05,
"loss": 0.5846,
"step": 2130
},
{
"epoch": 0.8510383386581469,
"grad_norm": 23.590261459350586,
"learning_rate": 1.5175718849840257e-05,
"loss": 0.2546,
"step": 2131
},
{
"epoch": 0.8514376996805112,
"grad_norm": 44.088077545166016,
"learning_rate": 1.5135782747603835e-05,
"loss": 0.82,
"step": 2132
},
{
"epoch": 0.8518370607028753,
"grad_norm": 5.3508782386779785,
"learning_rate": 1.5095846645367412e-05,
"loss": 0.0281,
"step": 2133
},
{
"epoch": 0.8522364217252396,
"grad_norm": 25.344985961914062,
"learning_rate": 1.505591054313099e-05,
"loss": 0.3271,
"step": 2134
},
{
"epoch": 0.8526357827476039,
"grad_norm": 16.268213272094727,
"learning_rate": 1.501597444089457e-05,
"loss": 0.5952,
"step": 2135
},
{
"epoch": 0.853035143769968,
"grad_norm": 7.255505084991455,
"learning_rate": 1.4976038338658147e-05,
"loss": 0.0492,
"step": 2136
},
{
"epoch": 0.8534345047923323,
"grad_norm": 24.21525001525879,
"learning_rate": 1.4936102236421725e-05,
"loss": 0.3002,
"step": 2137
},
{
"epoch": 0.8538338658146964,
"grad_norm": 15.554210662841797,
"learning_rate": 1.4896166134185305e-05,
"loss": 0.1457,
"step": 2138
},
{
"epoch": 0.8542332268370607,
"grad_norm": 38.429832458496094,
"learning_rate": 1.485623003194888e-05,
"loss": 0.4262,
"step": 2139
},
{
"epoch": 0.854632587859425,
"grad_norm": 30.281091690063477,
"learning_rate": 1.481629392971246e-05,
"loss": 0.8847,
"step": 2140
},
{
"epoch": 0.8550319488817891,
"grad_norm": 26.23798179626465,
"learning_rate": 1.477635782747604e-05,
"loss": 0.2644,
"step": 2141
},
{
"epoch": 0.8554313099041534,
"grad_norm": 25.942359924316406,
"learning_rate": 1.4736421725239619e-05,
"loss": 0.4022,
"step": 2142
},
{
"epoch": 0.8558306709265175,
"grad_norm": 0.3484882116317749,
"learning_rate": 1.4696485623003195e-05,
"loss": 0.0021,
"step": 2143
},
{
"epoch": 0.8562300319488818,
"grad_norm": 27.08789825439453,
"learning_rate": 1.4656549520766774e-05,
"loss": 0.6829,
"step": 2144
},
{
"epoch": 0.856629392971246,
"grad_norm": 31.635160446166992,
"learning_rate": 1.4616613418530354e-05,
"loss": 0.5384,
"step": 2145
},
{
"epoch": 0.8570287539936102,
"grad_norm": 23.96395492553711,
"learning_rate": 1.457667731629393e-05,
"loss": 0.2897,
"step": 2146
},
{
"epoch": 0.8574281150159745,
"grad_norm": 26.04054069519043,
"learning_rate": 1.453674121405751e-05,
"loss": 0.4421,
"step": 2147
},
{
"epoch": 0.8578274760383386,
"grad_norm": 1.9363259077072144,
"learning_rate": 1.4496805111821089e-05,
"loss": 0.0185,
"step": 2148
},
{
"epoch": 0.8582268370607029,
"grad_norm": 14.575151443481445,
"learning_rate": 1.4456869009584665e-05,
"loss": 0.1802,
"step": 2149
},
{
"epoch": 0.8586261980830671,
"grad_norm": 25.385629653930664,
"learning_rate": 1.4416932907348244e-05,
"loss": 0.7835,
"step": 2150
},
{
"epoch": 0.8590255591054313,
"grad_norm": 29.131591796875,
"learning_rate": 1.4376996805111822e-05,
"loss": 0.612,
"step": 2151
},
{
"epoch": 0.8594249201277955,
"grad_norm": 25.658008575439453,
"learning_rate": 1.43370607028754e-05,
"loss": 0.7959,
"step": 2152
},
{
"epoch": 0.8598242811501597,
"grad_norm": 23.30354881286621,
"learning_rate": 1.4297124600638977e-05,
"loss": 0.5093,
"step": 2153
},
{
"epoch": 0.860223642172524,
"grad_norm": 1.4935873746871948,
"learning_rate": 1.4257188498402557e-05,
"loss": 0.0101,
"step": 2154
},
{
"epoch": 0.8606230031948882,
"grad_norm": 21.294401168823242,
"learning_rate": 1.4217252396166134e-05,
"loss": 0.2015,
"step": 2155
},
{
"epoch": 0.8610223642172524,
"grad_norm": 27.149263381958008,
"learning_rate": 1.4177316293929712e-05,
"loss": 0.4172,
"step": 2156
},
{
"epoch": 0.8614217252396166,
"grad_norm": 3.81364369392395,
"learning_rate": 1.4137380191693292e-05,
"loss": 0.0373,
"step": 2157
},
{
"epoch": 0.8618210862619808,
"grad_norm": 10.141016006469727,
"learning_rate": 1.4097444089456871e-05,
"loss": 0.0745,
"step": 2158
},
{
"epoch": 0.862220447284345,
"grad_norm": 15.155673027038574,
"learning_rate": 1.4057507987220447e-05,
"loss": 0.1311,
"step": 2159
},
{
"epoch": 0.8626198083067093,
"grad_norm": 30.608505249023438,
"learning_rate": 1.4017571884984026e-05,
"loss": 0.494,
"step": 2160
},
{
"epoch": 0.8630191693290735,
"grad_norm": 0.2576078176498413,
"learning_rate": 1.3977635782747606e-05,
"loss": 0.0015,
"step": 2161
},
{
"epoch": 0.8634185303514377,
"grad_norm": 3.3154046535491943,
"learning_rate": 1.3937699680511182e-05,
"loss": 0.0176,
"step": 2162
},
{
"epoch": 0.8638178913738019,
"grad_norm": 9.802332878112793,
"learning_rate": 1.3897763578274761e-05,
"loss": 0.0986,
"step": 2163
},
{
"epoch": 0.8642172523961661,
"grad_norm": 21.24544906616211,
"learning_rate": 1.385782747603834e-05,
"loss": 0.1902,
"step": 2164
},
{
"epoch": 0.8646166134185304,
"grad_norm": 5.243937015533447,
"learning_rate": 1.3817891373801917e-05,
"loss": 0.0306,
"step": 2165
},
{
"epoch": 0.8650159744408946,
"grad_norm": 18.579870223999023,
"learning_rate": 1.3777955271565496e-05,
"loss": 0.9808,
"step": 2166
},
{
"epoch": 0.8654153354632588,
"grad_norm": 2.484816312789917,
"learning_rate": 1.3738019169329076e-05,
"loss": 0.0139,
"step": 2167
},
{
"epoch": 0.865814696485623,
"grad_norm": 25.7633113861084,
"learning_rate": 1.3698083067092652e-05,
"loss": 0.2435,
"step": 2168
},
{
"epoch": 0.8662140575079872,
"grad_norm": 38.99945831298828,
"learning_rate": 1.3658146964856231e-05,
"loss": 0.4071,
"step": 2169
},
{
"epoch": 0.8666134185303515,
"grad_norm": 2.2768900394439697,
"learning_rate": 1.3618210862619809e-05,
"loss": 0.0139,
"step": 2170
},
{
"epoch": 0.8670127795527156,
"grad_norm": 14.990517616271973,
"learning_rate": 1.3578274760383387e-05,
"loss": 0.1164,
"step": 2171
},
{
"epoch": 0.8674121405750799,
"grad_norm": 2.5337345600128174,
"learning_rate": 1.3538338658146964e-05,
"loss": 0.0125,
"step": 2172
},
{
"epoch": 0.8678115015974441,
"grad_norm": 2.587754011154175,
"learning_rate": 1.3498402555910544e-05,
"loss": 0.0248,
"step": 2173
},
{
"epoch": 0.8682108626198083,
"grad_norm": 19.899553298950195,
"learning_rate": 1.3458466453674123e-05,
"loss": 0.1645,
"step": 2174
},
{
"epoch": 0.8686102236421726,
"grad_norm": 21.9910945892334,
"learning_rate": 1.34185303514377e-05,
"loss": 0.1841,
"step": 2175
},
{
"epoch": 0.8690095846645367,
"grad_norm": 0.20630501210689545,
"learning_rate": 1.3378594249201279e-05,
"loss": 0.0031,
"step": 2176
},
{
"epoch": 0.869408945686901,
"grad_norm": 0.5906779170036316,
"learning_rate": 1.3338658146964858e-05,
"loss": 0.0029,
"step": 2177
},
{
"epoch": 0.8698083067092651,
"grad_norm": 42.74087905883789,
"learning_rate": 1.3298722044728434e-05,
"loss": 0.6435,
"step": 2178
},
{
"epoch": 0.8702076677316294,
"grad_norm": 7.409557819366455,
"learning_rate": 1.3258785942492014e-05,
"loss": 0.0402,
"step": 2179
},
{
"epoch": 0.8706070287539937,
"grad_norm": 1.136225938796997,
"learning_rate": 1.3218849840255593e-05,
"loss": 0.0057,
"step": 2180
},
{
"epoch": 0.8710063897763578,
"grad_norm": 45.87004089355469,
"learning_rate": 1.3178913738019169e-05,
"loss": 0.5008,
"step": 2181
},
{
"epoch": 0.8714057507987221,
"grad_norm": 41.57841873168945,
"learning_rate": 1.3138977635782748e-05,
"loss": 0.6726,
"step": 2182
},
{
"epoch": 0.8718051118210862,
"grad_norm": 9.963240623474121,
"learning_rate": 1.3099041533546328e-05,
"loss": 0.2818,
"step": 2183
},
{
"epoch": 0.8722044728434505,
"grad_norm": 16.591941833496094,
"learning_rate": 1.3059105431309904e-05,
"loss": 0.1249,
"step": 2184
},
{
"epoch": 0.8726038338658147,
"grad_norm": 40.33612823486328,
"learning_rate": 1.3019169329073483e-05,
"loss": 1.2938,
"step": 2185
},
{
"epoch": 0.8730031948881789,
"grad_norm": 2.7833244800567627,
"learning_rate": 1.2979233226837063e-05,
"loss": 0.0262,
"step": 2186
},
{
"epoch": 0.8734025559105432,
"grad_norm": 25.857683181762695,
"learning_rate": 1.2939297124600639e-05,
"loss": 0.6253,
"step": 2187
},
{
"epoch": 0.8738019169329073,
"grad_norm": 0.884412407875061,
"learning_rate": 1.2899361022364218e-05,
"loss": 0.0057,
"step": 2188
},
{
"epoch": 0.8742012779552716,
"grad_norm": 24.262216567993164,
"learning_rate": 1.2859424920127796e-05,
"loss": 0.394,
"step": 2189
},
{
"epoch": 0.8746006389776357,
"grad_norm": 26.28341293334961,
"learning_rate": 1.2819488817891375e-05,
"loss": 0.5833,
"step": 2190
},
{
"epoch": 0.875,
"grad_norm": 21.53606414794922,
"learning_rate": 1.2779552715654951e-05,
"loss": 0.1844,
"step": 2191
},
{
"epoch": 0.8753993610223643,
"grad_norm": 4.769852161407471,
"learning_rate": 1.273961661341853e-05,
"loss": 0.0253,
"step": 2192
},
{
"epoch": 0.8757987220447284,
"grad_norm": 1.418129324913025,
"learning_rate": 1.269968051118211e-05,
"loss": 0.0093,
"step": 2193
},
{
"epoch": 0.8761980830670927,
"grad_norm": 19.682043075561523,
"learning_rate": 1.2659744408945686e-05,
"loss": 0.1263,
"step": 2194
},
{
"epoch": 0.8765974440894568,
"grad_norm": 0.5761043429374695,
"learning_rate": 1.2619808306709266e-05,
"loss": 0.0088,
"step": 2195
},
{
"epoch": 0.8769968051118211,
"grad_norm": 21.17721939086914,
"learning_rate": 1.2579872204472845e-05,
"loss": 0.7706,
"step": 2196
},
{
"epoch": 0.8773961661341853,
"grad_norm": 3.1339364051818848,
"learning_rate": 1.2539936102236421e-05,
"loss": 0.0315,
"step": 2197
},
{
"epoch": 0.8777955271565495,
"grad_norm": 0.9560225605964661,
"learning_rate": 1.25e-05,
"loss": 0.0076,
"step": 2198
},
{
"epoch": 0.8781948881789138,
"grad_norm": 18.006439208984375,
"learning_rate": 1.2460063897763578e-05,
"loss": 0.3629,
"step": 2199
},
{
"epoch": 0.8785942492012779,
"grad_norm": 0.5707654356956482,
"learning_rate": 1.2420127795527158e-05,
"loss": 0.0035,
"step": 2200
},
{
"epoch": 0.8789936102236422,
"grad_norm": 25.81300926208496,
"learning_rate": 1.2380191693290735e-05,
"loss": 0.2955,
"step": 2201
},
{
"epoch": 0.8793929712460063,
"grad_norm": 22.9124755859375,
"learning_rate": 1.2340255591054313e-05,
"loss": 0.2136,
"step": 2202
},
{
"epoch": 0.8797923322683706,
"grad_norm": 20.88273811340332,
"learning_rate": 1.2300319488817893e-05,
"loss": 0.1834,
"step": 2203
},
{
"epoch": 0.8801916932907349,
"grad_norm": 20.055179595947266,
"learning_rate": 1.226038338658147e-05,
"loss": 0.1465,
"step": 2204
},
{
"epoch": 0.880591054313099,
"grad_norm": 6.466705799102783,
"learning_rate": 1.222044728434505e-05,
"loss": 0.0388,
"step": 2205
},
{
"epoch": 0.8809904153354633,
"grad_norm": 2.010955333709717,
"learning_rate": 1.2180511182108628e-05,
"loss": 0.0116,
"step": 2206
},
{
"epoch": 0.8813897763578274,
"grad_norm": 17.53694725036621,
"learning_rate": 1.2140575079872205e-05,
"loss": 0.2846,
"step": 2207
},
{
"epoch": 0.8817891373801917,
"grad_norm": 28.81610870361328,
"learning_rate": 1.2100638977635783e-05,
"loss": 0.5809,
"step": 2208
},
{
"epoch": 0.8821884984025559,
"grad_norm": 52.048194885253906,
"learning_rate": 1.206070287539936e-05,
"loss": 1.469,
"step": 2209
},
{
"epoch": 0.8825878594249201,
"grad_norm": 0.7374428510665894,
"learning_rate": 1.2020766773162938e-05,
"loss": 0.0057,
"step": 2210
},
{
"epoch": 0.8829872204472844,
"grad_norm": 16.59804916381836,
"learning_rate": 1.1980830670926518e-05,
"loss": 0.1478,
"step": 2211
},
{
"epoch": 0.8833865814696485,
"grad_norm": 6.262393951416016,
"learning_rate": 1.1940894568690096e-05,
"loss": 0.038,
"step": 2212
},
{
"epoch": 0.8837859424920128,
"grad_norm": 2.658107280731201,
"learning_rate": 1.1900958466453675e-05,
"loss": 0.013,
"step": 2213
},
{
"epoch": 0.884185303514377,
"grad_norm": 7.391872882843018,
"learning_rate": 1.1861022364217253e-05,
"loss": 0.0436,
"step": 2214
},
{
"epoch": 0.8845846645367412,
"grad_norm": 21.136445999145508,
"learning_rate": 1.182108626198083e-05,
"loss": 0.2305,
"step": 2215
},
{
"epoch": 0.8849840255591054,
"grad_norm": 60.5404167175293,
"learning_rate": 1.178115015974441e-05,
"loss": 1.0031,
"step": 2216
},
{
"epoch": 0.8853833865814696,
"grad_norm": 24.526824951171875,
"learning_rate": 1.1741214057507988e-05,
"loss": 0.3,
"step": 2217
},
{
"epoch": 0.8857827476038339,
"grad_norm": 6.606674671173096,
"learning_rate": 1.1701277955271565e-05,
"loss": 0.0618,
"step": 2218
},
{
"epoch": 0.8861821086261981,
"grad_norm": 0.8497007489204407,
"learning_rate": 1.1661341853035145e-05,
"loss": 0.0049,
"step": 2219
},
{
"epoch": 0.8865814696485623,
"grad_norm": 6.738035678863525,
"learning_rate": 1.1621405750798723e-05,
"loss": 0.0357,
"step": 2220
},
{
"epoch": 0.8869808306709265,
"grad_norm": 42.1725959777832,
"learning_rate": 1.1581469648562302e-05,
"loss": 0.4446,
"step": 2221
},
{
"epoch": 0.8873801916932907,
"grad_norm": 13.766042709350586,
"learning_rate": 1.154153354632588e-05,
"loss": 0.2414,
"step": 2222
},
{
"epoch": 0.887779552715655,
"grad_norm": 5.741289138793945,
"learning_rate": 1.1501597444089457e-05,
"loss": 0.047,
"step": 2223
},
{
"epoch": 0.8881789137380192,
"grad_norm": 0.293002724647522,
"learning_rate": 1.1461661341853037e-05,
"loss": 0.0022,
"step": 2224
},
{
"epoch": 0.8885782747603834,
"grad_norm": 8.95388126373291,
"learning_rate": 1.1421725239616615e-05,
"loss": 0.054,
"step": 2225
},
{
"epoch": 0.8889776357827476,
"grad_norm": 3.570014715194702,
"learning_rate": 1.1381789137380192e-05,
"loss": 0.0206,
"step": 2226
},
{
"epoch": 0.8893769968051118,
"grad_norm": 2.6293859481811523,
"learning_rate": 1.134185303514377e-05,
"loss": 0.0145,
"step": 2227
},
{
"epoch": 0.889776357827476,
"grad_norm": 5.195387840270996,
"learning_rate": 1.1301916932907348e-05,
"loss": 0.0467,
"step": 2228
},
{
"epoch": 0.8901757188498403,
"grad_norm": 35.00132751464844,
"learning_rate": 1.1261980830670927e-05,
"loss": 1.0233,
"step": 2229
},
{
"epoch": 0.8905750798722045,
"grad_norm": 6.274228096008301,
"learning_rate": 1.1222044728434505e-05,
"loss": 0.0436,
"step": 2230
},
{
"epoch": 0.8909744408945687,
"grad_norm": 26.142656326293945,
"learning_rate": 1.1182108626198083e-05,
"loss": 0.3755,
"step": 2231
},
{
"epoch": 0.8913738019169329,
"grad_norm": 18.75430679321289,
"learning_rate": 1.1142172523961662e-05,
"loss": 0.1763,
"step": 2232
},
{
"epoch": 0.8917731629392971,
"grad_norm": 11.69100570678711,
"learning_rate": 1.110223642172524e-05,
"loss": 0.0725,
"step": 2233
},
{
"epoch": 0.8921725239616614,
"grad_norm": 26.685670852661133,
"learning_rate": 1.1062300319488818e-05,
"loss": 0.4991,
"step": 2234
},
{
"epoch": 0.8925718849840255,
"grad_norm": 26.953086853027344,
"learning_rate": 1.1022364217252397e-05,
"loss": 0.3768,
"step": 2235
},
{
"epoch": 0.8929712460063898,
"grad_norm": 19.495054244995117,
"learning_rate": 1.0982428115015975e-05,
"loss": 0.1513,
"step": 2236
},
{
"epoch": 0.893370607028754,
"grad_norm": 26.936817169189453,
"learning_rate": 1.0942492012779554e-05,
"loss": 0.5858,
"step": 2237
},
{
"epoch": 0.8937699680511182,
"grad_norm": 20.8134708404541,
"learning_rate": 1.0902555910543132e-05,
"loss": 0.2547,
"step": 2238
},
{
"epoch": 0.8941693290734825,
"grad_norm": 20.475303649902344,
"learning_rate": 1.086261980830671e-05,
"loss": 0.1579,
"step": 2239
},
{
"epoch": 0.8945686900958466,
"grad_norm": 25.860332489013672,
"learning_rate": 1.0822683706070289e-05,
"loss": 0.4831,
"step": 2240
},
{
"epoch": 0.8949680511182109,
"grad_norm": 33.197731018066406,
"learning_rate": 1.0782747603833867e-05,
"loss": 0.3158,
"step": 2241
},
{
"epoch": 0.895367412140575,
"grad_norm": 12.196266174316406,
"learning_rate": 1.0742811501597445e-05,
"loss": 0.0765,
"step": 2242
},
{
"epoch": 0.8957667731629393,
"grad_norm": 2.0627737045288086,
"learning_rate": 1.0702875399361024e-05,
"loss": 0.0115,
"step": 2243
},
{
"epoch": 0.8961661341853036,
"grad_norm": 0.044521380215883255,
"learning_rate": 1.0662939297124602e-05,
"loss": 0.0007,
"step": 2244
},
{
"epoch": 0.8965654952076677,
"grad_norm": 17.35426902770996,
"learning_rate": 1.062300319488818e-05,
"loss": 0.1667,
"step": 2245
},
{
"epoch": 0.896964856230032,
"grad_norm": 28.53548240661621,
"learning_rate": 1.0583067092651757e-05,
"loss": 0.5658,
"step": 2246
},
{
"epoch": 0.8973642172523961,
"grad_norm": 37.569332122802734,
"learning_rate": 1.0543130990415335e-05,
"loss": 1.0443,
"step": 2247
},
{
"epoch": 0.8977635782747604,
"grad_norm": 26.550031661987305,
"learning_rate": 1.0503194888178914e-05,
"loss": 0.4301,
"step": 2248
},
{
"epoch": 0.8981629392971247,
"grad_norm": 17.748807907104492,
"learning_rate": 1.0463258785942492e-05,
"loss": 0.1267,
"step": 2249
},
{
"epoch": 0.8985623003194888,
"grad_norm": 24.39689826965332,
"learning_rate": 1.042332268370607e-05,
"loss": 0.2923,
"step": 2250
},
{
"epoch": 0.8989616613418531,
"grad_norm": 15.560029983520508,
"learning_rate": 1.038338658146965e-05,
"loss": 0.2905,
"step": 2251
},
{
"epoch": 0.8993610223642172,
"grad_norm": 8.739490509033203,
"learning_rate": 1.0343450479233227e-05,
"loss": 0.1952,
"step": 2252
},
{
"epoch": 0.8997603833865815,
"grad_norm": 0.7885538339614868,
"learning_rate": 1.0303514376996806e-05,
"loss": 0.0059,
"step": 2253
},
{
"epoch": 0.9001597444089456,
"grad_norm": 37.800621032714844,
"learning_rate": 1.0263578274760384e-05,
"loss": 0.3054,
"step": 2254
},
{
"epoch": 0.9005591054313099,
"grad_norm": 5.644620418548584,
"learning_rate": 1.0223642172523962e-05,
"loss": 0.0301,
"step": 2255
},
{
"epoch": 0.9009584664536742,
"grad_norm": 3.504420518875122,
"learning_rate": 1.0183706070287541e-05,
"loss": 0.0367,
"step": 2256
},
{
"epoch": 0.9013578274760383,
"grad_norm": 24.251201629638672,
"learning_rate": 1.0143769968051119e-05,
"loss": 0.2953,
"step": 2257
},
{
"epoch": 0.9017571884984026,
"grad_norm": 0.3165093660354614,
"learning_rate": 1.0103833865814697e-05,
"loss": 0.0018,
"step": 2258
},
{
"epoch": 0.9021565495207667,
"grad_norm": 26.208641052246094,
"learning_rate": 1.0063897763578276e-05,
"loss": 0.6927,
"step": 2259
},
{
"epoch": 0.902555910543131,
"grad_norm": 29.2253360748291,
"learning_rate": 1.0023961661341854e-05,
"loss": 0.539,
"step": 2260
},
{
"epoch": 0.9029552715654952,
"grad_norm": 21.93425178527832,
"learning_rate": 9.984025559105432e-06,
"loss": 0.2574,
"step": 2261
},
{
"epoch": 0.9033546325878594,
"grad_norm": 0.29664546251296997,
"learning_rate": 9.944089456869011e-06,
"loss": 0.0019,
"step": 2262
},
{
"epoch": 0.9037539936102237,
"grad_norm": 16.4567813873291,
"learning_rate": 9.904153354632589e-06,
"loss": 0.132,
"step": 2263
},
{
"epoch": 0.9041533546325878,
"grad_norm": 29.963830947875977,
"learning_rate": 9.864217252396166e-06,
"loss": 0.2704,
"step": 2264
},
{
"epoch": 0.9045527156549521,
"grad_norm": 37.3609504699707,
"learning_rate": 9.824281150159744e-06,
"loss": 0.4519,
"step": 2265
},
{
"epoch": 0.9049520766773163,
"grad_norm": 29.024925231933594,
"learning_rate": 9.784345047923322e-06,
"loss": 0.3276,
"step": 2266
},
{
"epoch": 0.9053514376996805,
"grad_norm": 0.5080374479293823,
"learning_rate": 9.744408945686901e-06,
"loss": 0.0025,
"step": 2267
},
{
"epoch": 0.9057507987220448,
"grad_norm": 15.353757858276367,
"learning_rate": 9.704472843450479e-06,
"loss": 0.1093,
"step": 2268
},
{
"epoch": 0.9061501597444089,
"grad_norm": 31.80521011352539,
"learning_rate": 9.664536741214059e-06,
"loss": 0.661,
"step": 2269
},
{
"epoch": 0.9065495207667732,
"grad_norm": 5.236536502838135,
"learning_rate": 9.624600638977636e-06,
"loss": 0.0283,
"step": 2270
},
{
"epoch": 0.9069488817891374,
"grad_norm": 6.562463760375977,
"learning_rate": 9.584664536741214e-06,
"loss": 0.0555,
"step": 2271
},
{
"epoch": 0.9073482428115016,
"grad_norm": 22.86138343811035,
"learning_rate": 9.544728434504793e-06,
"loss": 1.8629,
"step": 2272
},
{
"epoch": 0.9077476038338658,
"grad_norm": 26.078073501586914,
"learning_rate": 9.504792332268371e-06,
"loss": 0.3995,
"step": 2273
},
{
"epoch": 0.90814696485623,
"grad_norm": 1.0606603622436523,
"learning_rate": 9.464856230031949e-06,
"loss": 0.0073,
"step": 2274
},
{
"epoch": 0.9085463258785943,
"grad_norm": 9.603179931640625,
"learning_rate": 9.424920127795528e-06,
"loss": 0.0646,
"step": 2275
},
{
"epoch": 0.9089456869009584,
"grad_norm": 31.62209129333496,
"learning_rate": 9.384984025559106e-06,
"loss": 0.7958,
"step": 2276
},
{
"epoch": 0.9093450479233227,
"grad_norm": 4.506667613983154,
"learning_rate": 9.345047923322685e-06,
"loss": 0.0312,
"step": 2277
},
{
"epoch": 0.9097444089456869,
"grad_norm": 15.974720001220703,
"learning_rate": 9.305111821086263e-06,
"loss": 0.1319,
"step": 2278
},
{
"epoch": 0.9101437699680511,
"grad_norm": 22.956083297729492,
"learning_rate": 9.265175718849841e-06,
"loss": 0.221,
"step": 2279
},
{
"epoch": 0.9105431309904153,
"grad_norm": 13.125929832458496,
"learning_rate": 9.225239616613419e-06,
"loss": 0.1019,
"step": 2280
},
{
"epoch": 0.9109424920127795,
"grad_norm": 0.08441688865423203,
"learning_rate": 9.185303514376996e-06,
"loss": 0.0005,
"step": 2281
},
{
"epoch": 0.9113418530351438,
"grad_norm": 30.863683700561523,
"learning_rate": 9.145367412140576e-06,
"loss": 0.5395,
"step": 2282
},
{
"epoch": 0.911741214057508,
"grad_norm": 12.245810508728027,
"learning_rate": 9.105431309904154e-06,
"loss": 0.1462,
"step": 2283
},
{
"epoch": 0.9121405750798722,
"grad_norm": 18.747602462768555,
"learning_rate": 9.065495207667731e-06,
"loss": 0.1267,
"step": 2284
},
{
"epoch": 0.9125399361022364,
"grad_norm": 8.89615249633789,
"learning_rate": 9.02555910543131e-06,
"loss": 0.0543,
"step": 2285
},
{
"epoch": 0.9129392971246006,
"grad_norm": 25.749221801757812,
"learning_rate": 8.985623003194888e-06,
"loss": 0.2471,
"step": 2286
},
{
"epoch": 0.9133386581469649,
"grad_norm": 1.4204703569412231,
"learning_rate": 8.945686900958466e-06,
"loss": 0.0078,
"step": 2287
},
{
"epoch": 0.9137380191693291,
"grad_norm": 14.662181854248047,
"learning_rate": 8.905750798722046e-06,
"loss": 0.0999,
"step": 2288
},
{
"epoch": 0.9141373801916933,
"grad_norm": 26.76403045654297,
"learning_rate": 8.865814696485623e-06,
"loss": 0.5261,
"step": 2289
},
{
"epoch": 0.9145367412140575,
"grad_norm": 43.350807189941406,
"learning_rate": 8.825878594249201e-06,
"loss": 0.9436,
"step": 2290
},
{
"epoch": 0.9149361022364217,
"grad_norm": 3.3935983180999756,
"learning_rate": 8.78594249201278e-06,
"loss": 0.018,
"step": 2291
},
{
"epoch": 0.9153354632587859,
"grad_norm": 9.029473304748535,
"learning_rate": 8.746006389776358e-06,
"loss": 0.0737,
"step": 2292
},
{
"epoch": 0.9157348242811502,
"grad_norm": 22.468652725219727,
"learning_rate": 8.706070287539938e-06,
"loss": 0.3739,
"step": 2293
},
{
"epoch": 0.9161341853035144,
"grad_norm": 16.91623878479004,
"learning_rate": 8.666134185303515e-06,
"loss": 0.141,
"step": 2294
},
{
"epoch": 0.9165335463258786,
"grad_norm": 22.57716941833496,
"learning_rate": 8.626198083067093e-06,
"loss": 0.1916,
"step": 2295
},
{
"epoch": 0.9169329073482428,
"grad_norm": 24.174352645874023,
"learning_rate": 8.586261980830673e-06,
"loss": 0.7076,
"step": 2296
},
{
"epoch": 0.917332268370607,
"grad_norm": 26.055498123168945,
"learning_rate": 8.54632587859425e-06,
"loss": 0.519,
"step": 2297
},
{
"epoch": 0.9177316293929713,
"grad_norm": 5.417344093322754,
"learning_rate": 8.506389776357828e-06,
"loss": 0.0345,
"step": 2298
},
{
"epoch": 0.9181309904153354,
"grad_norm": 0.5970049500465393,
"learning_rate": 8.466453674121406e-06,
"loss": 0.003,
"step": 2299
},
{
"epoch": 0.9185303514376997,
"grad_norm": 17.132591247558594,
"learning_rate": 8.426517571884983e-06,
"loss": 0.1597,
"step": 2300
},
{
"epoch": 0.9189297124600639,
"grad_norm": 14.512480735778809,
"learning_rate": 8.386581469648563e-06,
"loss": 0.1084,
"step": 2301
},
{
"epoch": 0.9193290734824281,
"grad_norm": 22.194793701171875,
"learning_rate": 8.34664536741214e-06,
"loss": 0.2467,
"step": 2302
},
{
"epoch": 0.9197284345047924,
"grad_norm": 31.158233642578125,
"learning_rate": 8.306709265175718e-06,
"loss": 0.3752,
"step": 2303
},
{
"epoch": 0.9201277955271565,
"grad_norm": 20.321792602539062,
"learning_rate": 8.266773162939298e-06,
"loss": 0.1434,
"step": 2304
},
{
"epoch": 0.9205271565495208,
"grad_norm": 18.444812774658203,
"learning_rate": 8.226837060702875e-06,
"loss": 0.2675,
"step": 2305
},
{
"epoch": 0.920926517571885,
"grad_norm": 39.47265625,
"learning_rate": 8.186900958466453e-06,
"loss": 0.9925,
"step": 2306
},
{
"epoch": 0.9213258785942492,
"grad_norm": 0.33508554100990295,
"learning_rate": 8.146964856230033e-06,
"loss": 0.0018,
"step": 2307
},
{
"epoch": 0.9217252396166135,
"grad_norm": 27.08177375793457,
"learning_rate": 8.10702875399361e-06,
"loss": 0.6421,
"step": 2308
},
{
"epoch": 0.9221246006389776,
"grad_norm": 17.82419776916504,
"learning_rate": 8.067092651757188e-06,
"loss": 0.1457,
"step": 2309
},
{
"epoch": 0.9225239616613419,
"grad_norm": 25.19672966003418,
"learning_rate": 8.027156549520768e-06,
"loss": 0.3516,
"step": 2310
},
{
"epoch": 0.922923322683706,
"grad_norm": 42.58757019042969,
"learning_rate": 7.987220447284345e-06,
"loss": 0.3494,
"step": 2311
},
{
"epoch": 0.9233226837060703,
"grad_norm": 21.062822341918945,
"learning_rate": 7.947284345047925e-06,
"loss": 0.3392,
"step": 2312
},
{
"epoch": 0.9237220447284346,
"grad_norm": 12.202067375183105,
"learning_rate": 7.907348242811502e-06,
"loss": 0.0751,
"step": 2313
},
{
"epoch": 0.9241214057507987,
"grad_norm": 29.747121810913086,
"learning_rate": 7.86741214057508e-06,
"loss": 0.5624,
"step": 2314
},
{
"epoch": 0.924520766773163,
"grad_norm": 1.432660698890686,
"learning_rate": 7.82747603833866e-06,
"loss": 0.0081,
"step": 2315
},
{
"epoch": 0.9249201277955271,
"grad_norm": 10.168231964111328,
"learning_rate": 7.787539936102237e-06,
"loss": 0.0781,
"step": 2316
},
{
"epoch": 0.9253194888178914,
"grad_norm": 33.55418014526367,
"learning_rate": 7.747603833865815e-06,
"loss": 1.3949,
"step": 2317
},
{
"epoch": 0.9257188498402555,
"grad_norm": 13.585084915161133,
"learning_rate": 7.707667731629393e-06,
"loss": 0.0811,
"step": 2318
},
{
"epoch": 0.9261182108626198,
"grad_norm": 32.28488540649414,
"learning_rate": 7.66773162939297e-06,
"loss": 0.4806,
"step": 2319
},
{
"epoch": 0.9265175718849841,
"grad_norm": 0.030924487859010696,
"learning_rate": 7.627795527156551e-06,
"loss": 0.0004,
"step": 2320
},
{
"epoch": 0.9269169329073482,
"grad_norm": 25.162763595581055,
"learning_rate": 7.5878594249201285e-06,
"loss": 0.3418,
"step": 2321
},
{
"epoch": 0.9273162939297125,
"grad_norm": 25.48050308227539,
"learning_rate": 7.547923322683706e-06,
"loss": 0.9607,
"step": 2322
},
{
"epoch": 0.9277156549520766,
"grad_norm": 4.641092300415039,
"learning_rate": 7.507987220447285e-06,
"loss": 0.0349,
"step": 2323
},
{
"epoch": 0.9281150159744409,
"grad_norm": 15.052102088928223,
"learning_rate": 7.4680511182108625e-06,
"loss": 0.1002,
"step": 2324
},
{
"epoch": 0.9285143769968051,
"grad_norm": 6.498836517333984,
"learning_rate": 7.42811501597444e-06,
"loss": 0.0352,
"step": 2325
},
{
"epoch": 0.9289137380191693,
"grad_norm": 21.226430892944336,
"learning_rate": 7.38817891373802e-06,
"loss": 0.2618,
"step": 2326
},
{
"epoch": 0.9293130990415336,
"grad_norm": 34.26365661621094,
"learning_rate": 7.3482428115015974e-06,
"loss": 0.2883,
"step": 2327
},
{
"epoch": 0.9297124600638977,
"grad_norm": 3.406686782836914,
"learning_rate": 7.308306709265177e-06,
"loss": 0.0287,
"step": 2328
},
{
"epoch": 0.930111821086262,
"grad_norm": 28.853410720825195,
"learning_rate": 7.268370607028755e-06,
"loss": 0.42,
"step": 2329
},
{
"epoch": 0.9305111821086262,
"grad_norm": 1.6478089094161987,
"learning_rate": 7.228434504792332e-06,
"loss": 0.0217,
"step": 2330
},
{
"epoch": 0.9309105431309904,
"grad_norm": 26.49935531616211,
"learning_rate": 7.188498402555911e-06,
"loss": 0.2434,
"step": 2331
},
{
"epoch": 0.9313099041533547,
"grad_norm": 6.1432576179504395,
"learning_rate": 7.148562300319489e-06,
"loss": 0.0938,
"step": 2332
},
{
"epoch": 0.9317092651757188,
"grad_norm": 22.87884521484375,
"learning_rate": 7.108626198083067e-06,
"loss": 0.3578,
"step": 2333
},
{
"epoch": 0.9321086261980831,
"grad_norm": 8.84752082824707,
"learning_rate": 7.068690095846646e-06,
"loss": 0.0893,
"step": 2334
},
{
"epoch": 0.9325079872204473,
"grad_norm": 6.084292411804199,
"learning_rate": 7.0287539936102235e-06,
"loss": 0.0436,
"step": 2335
},
{
"epoch": 0.9329073482428115,
"grad_norm": 0.5903221964836121,
"learning_rate": 6.988817891373803e-06,
"loss": 0.0073,
"step": 2336
},
{
"epoch": 0.9333067092651757,
"grad_norm": 23.68195915222168,
"learning_rate": 6.948881789137381e-06,
"loss": 0.2413,
"step": 2337
},
{
"epoch": 0.9337060702875399,
"grad_norm": 14.517341613769531,
"learning_rate": 6.908945686900958e-06,
"loss": 0.1277,
"step": 2338
},
{
"epoch": 0.9341054313099042,
"grad_norm": 26.17272186279297,
"learning_rate": 6.869009584664538e-06,
"loss": 0.3685,
"step": 2339
},
{
"epoch": 0.9345047923322684,
"grad_norm": 34.10908889770508,
"learning_rate": 6.8290734824281156e-06,
"loss": 0.9022,
"step": 2340
},
{
"epoch": 0.9349041533546326,
"grad_norm": 14.13460922241211,
"learning_rate": 6.789137380191693e-06,
"loss": 0.1037,
"step": 2341
},
{
"epoch": 0.9353035143769968,
"grad_norm": 26.47983741760254,
"learning_rate": 6.749201277955272e-06,
"loss": 0.5637,
"step": 2342
},
{
"epoch": 0.935702875399361,
"grad_norm": 25.320003509521484,
"learning_rate": 6.70926517571885e-06,
"loss": 0.604,
"step": 2343
},
{
"epoch": 0.9361022364217252,
"grad_norm": 0.400365948677063,
"learning_rate": 6.669329073482429e-06,
"loss": 0.002,
"step": 2344
},
{
"epoch": 0.9365015974440895,
"grad_norm": 18.762229919433594,
"learning_rate": 6.629392971246007e-06,
"loss": 0.1318,
"step": 2345
},
{
"epoch": 0.9369009584664537,
"grad_norm": 8.723437309265137,
"learning_rate": 6.5894568690095845e-06,
"loss": 0.0585,
"step": 2346
},
{
"epoch": 0.9373003194888179,
"grad_norm": 12.249885559082031,
"learning_rate": 6.549520766773164e-06,
"loss": 0.3686,
"step": 2347
},
{
"epoch": 0.9376996805111821,
"grad_norm": 24.543588638305664,
"learning_rate": 6.509584664536742e-06,
"loss": 0.2233,
"step": 2348
},
{
"epoch": 0.9380990415335463,
"grad_norm": 41.918235778808594,
"learning_rate": 6.469648562300319e-06,
"loss": 0.7428,
"step": 2349
},
{
"epoch": 0.9384984025559105,
"grad_norm": 40.692562103271484,
"learning_rate": 6.429712460063898e-06,
"loss": 0.6329,
"step": 2350
},
{
"epoch": 0.9388977635782748,
"grad_norm": 13.25906753540039,
"learning_rate": 6.389776357827476e-06,
"loss": 0.0881,
"step": 2351
},
{
"epoch": 0.939297124600639,
"grad_norm": 22.844648361206055,
"learning_rate": 6.349840255591055e-06,
"loss": 0.2473,
"step": 2352
},
{
"epoch": 0.9396964856230032,
"grad_norm": 22.59808921813965,
"learning_rate": 6.309904153354633e-06,
"loss": 0.2201,
"step": 2353
},
{
"epoch": 0.9400958466453674,
"grad_norm": 19.90828514099121,
"learning_rate": 6.269968051118211e-06,
"loss": 0.2458,
"step": 2354
},
{
"epoch": 0.9404952076677316,
"grad_norm": 12.037470817565918,
"learning_rate": 6.230031948881789e-06,
"loss": 0.0754,
"step": 2355
},
{
"epoch": 0.9408945686900958,
"grad_norm": 43.09394836425781,
"learning_rate": 6.190095846645368e-06,
"loss": 0.7691,
"step": 2356
},
{
"epoch": 0.9412939297124601,
"grad_norm": 14.067617416381836,
"learning_rate": 6.150159744408946e-06,
"loss": 0.7744,
"step": 2357
},
{
"epoch": 0.9416932907348243,
"grad_norm": 23.883544921875,
"learning_rate": 6.110223642172525e-06,
"loss": 0.3773,
"step": 2358
},
{
"epoch": 0.9420926517571885,
"grad_norm": 16.951818466186523,
"learning_rate": 6.070287539936103e-06,
"loss": 0.3496,
"step": 2359
},
{
"epoch": 0.9424920127795527,
"grad_norm": 28.95001220703125,
"learning_rate": 6.03035143769968e-06,
"loss": 0.3151,
"step": 2360
},
{
"epoch": 0.9428913738019169,
"grad_norm": 4.39561128616333,
"learning_rate": 5.990415335463259e-06,
"loss": 0.0255,
"step": 2361
},
{
"epoch": 0.9432907348242812,
"grad_norm": 0.003173028351739049,
"learning_rate": 5.9504792332268375e-06,
"loss": 0.0,
"step": 2362
},
{
"epoch": 0.9436900958466453,
"grad_norm": 29.000295639038086,
"learning_rate": 5.910543130990415e-06,
"loss": 0.3647,
"step": 2363
},
{
"epoch": 0.9440894568690096,
"grad_norm": 2.6554622650146484,
"learning_rate": 5.870607028753994e-06,
"loss": 0.0134,
"step": 2364
},
{
"epoch": 0.9444888178913738,
"grad_norm": 8.05947208404541,
"learning_rate": 5.830670926517572e-06,
"loss": 0.0445,
"step": 2365
},
{
"epoch": 0.944888178913738,
"grad_norm": 35.0678596496582,
"learning_rate": 5.790734824281151e-06,
"loss": 0.6887,
"step": 2366
},
{
"epoch": 0.9452875399361023,
"grad_norm": 10.419208526611328,
"learning_rate": 5.750798722044729e-06,
"loss": 0.0657,
"step": 2367
},
{
"epoch": 0.9456869009584664,
"grad_norm": 28.15788459777832,
"learning_rate": 5.710862619808307e-06,
"loss": 1.0672,
"step": 2368
},
{
"epoch": 0.9460862619808307,
"grad_norm": 15.911201477050781,
"learning_rate": 5.670926517571885e-06,
"loss": 0.3833,
"step": 2369
},
{
"epoch": 0.9464856230031949,
"grad_norm": 19.57373046875,
"learning_rate": 5.630990415335464e-06,
"loss": 0.1245,
"step": 2370
},
{
"epoch": 0.9468849840255591,
"grad_norm": 2.191161632537842,
"learning_rate": 5.591054313099041e-06,
"loss": 0.0112,
"step": 2371
},
{
"epoch": 0.9472843450479234,
"grad_norm": 11.419310569763184,
"learning_rate": 5.55111821086262e-06,
"loss": 0.7617,
"step": 2372
},
{
"epoch": 0.9476837060702875,
"grad_norm": 21.01795196533203,
"learning_rate": 5.5111821086261985e-06,
"loss": 0.223,
"step": 2373
},
{
"epoch": 0.9480830670926518,
"grad_norm": 11.465447425842285,
"learning_rate": 5.471246006389777e-06,
"loss": 0.0971,
"step": 2374
},
{
"epoch": 0.9484824281150159,
"grad_norm": 4.532099723815918,
"learning_rate": 5.431309904153355e-06,
"loss": 0.0291,
"step": 2375
},
{
"epoch": 0.9488817891373802,
"grad_norm": 1.8142410516738892,
"learning_rate": 5.391373801916933e-06,
"loss": 0.0183,
"step": 2376
},
{
"epoch": 0.9492811501597445,
"grad_norm": 11.216056823730469,
"learning_rate": 5.351437699680512e-06,
"loss": 0.0765,
"step": 2377
},
{
"epoch": 0.9496805111821086,
"grad_norm": 13.96428394317627,
"learning_rate": 5.31150159744409e-06,
"loss": 0.0873,
"step": 2378
},
{
"epoch": 0.9500798722044729,
"grad_norm": 0.8237498998641968,
"learning_rate": 5.2715654952076674e-06,
"loss": 0.005,
"step": 2379
},
{
"epoch": 0.950479233226837,
"grad_norm": 24.64752960205078,
"learning_rate": 5.231629392971246e-06,
"loss": 0.7289,
"step": 2380
},
{
"epoch": 0.9508785942492013,
"grad_norm": 30.878896713256836,
"learning_rate": 5.191693290734825e-06,
"loss": 0.3221,
"step": 2381
},
{
"epoch": 0.9512779552715654,
"grad_norm": 26.4088134765625,
"learning_rate": 5.151757188498403e-06,
"loss": 0.4023,
"step": 2382
},
{
"epoch": 0.9516773162939297,
"grad_norm": 19.35401725769043,
"learning_rate": 5.111821086261981e-06,
"loss": 0.1683,
"step": 2383
},
{
"epoch": 0.952076677316294,
"grad_norm": 29.801679611206055,
"learning_rate": 5.0718849840255595e-06,
"loss": 0.3299,
"step": 2384
},
{
"epoch": 0.9524760383386581,
"grad_norm": 32.75882339477539,
"learning_rate": 5.031948881789138e-06,
"loss": 1.4834,
"step": 2385
},
{
"epoch": 0.9528753993610224,
"grad_norm": 8.555570602416992,
"learning_rate": 4.992012779552716e-06,
"loss": 0.0758,
"step": 2386
},
{
"epoch": 0.9532747603833865,
"grad_norm": 19.17962074279785,
"learning_rate": 4.952076677316294e-06,
"loss": 0.1546,
"step": 2387
},
{
"epoch": 0.9536741214057508,
"grad_norm": 21.570669174194336,
"learning_rate": 4.912140575079872e-06,
"loss": 0.2827,
"step": 2388
},
{
"epoch": 0.954073482428115,
"grad_norm": 25.59955596923828,
"learning_rate": 4.872204472843451e-06,
"loss": 0.3286,
"step": 2389
},
{
"epoch": 0.9544728434504792,
"grad_norm": 8.170394897460938,
"learning_rate": 4.832268370607029e-06,
"loss": 0.0742,
"step": 2390
},
{
"epoch": 0.9548722044728435,
"grad_norm": 43.146968841552734,
"learning_rate": 4.792332268370607e-06,
"loss": 0.6443,
"step": 2391
},
{
"epoch": 0.9552715654952076,
"grad_norm": 23.907411575317383,
"learning_rate": 4.7523961661341856e-06,
"loss": 0.3729,
"step": 2392
},
{
"epoch": 0.9556709265175719,
"grad_norm": 28.709169387817383,
"learning_rate": 4.712460063897764e-06,
"loss": 0.2335,
"step": 2393
},
{
"epoch": 0.9560702875399361,
"grad_norm": 20.422616958618164,
"learning_rate": 4.672523961661343e-06,
"loss": 0.1856,
"step": 2394
},
{
"epoch": 0.9564696485623003,
"grad_norm": 12.558745384216309,
"learning_rate": 4.6325878594249205e-06,
"loss": 0.0856,
"step": 2395
},
{
"epoch": 0.9568690095846646,
"grad_norm": 0.76097571849823,
"learning_rate": 4.592651757188498e-06,
"loss": 0.0066,
"step": 2396
},
{
"epoch": 0.9572683706070287,
"grad_norm": 23.985401153564453,
"learning_rate": 4.552715654952077e-06,
"loss": 0.2479,
"step": 2397
},
{
"epoch": 0.957667731629393,
"grad_norm": 26.551990509033203,
"learning_rate": 4.512779552715655e-06,
"loss": 0.8312,
"step": 2398
},
{
"epoch": 0.9580670926517572,
"grad_norm": 19.035991668701172,
"learning_rate": 4.472843450479233e-06,
"loss": 0.1666,
"step": 2399
},
{
"epoch": 0.9584664536741214,
"grad_norm": 55.58621597290039,
"learning_rate": 4.432907348242812e-06,
"loss": 1.1327,
"step": 2400
},
{
"epoch": 0.9588658146964856,
"grad_norm": 27.019744873046875,
"learning_rate": 4.39297124600639e-06,
"loss": 0.9537,
"step": 2401
},
{
"epoch": 0.9592651757188498,
"grad_norm": 26.172531127929688,
"learning_rate": 4.353035143769969e-06,
"loss": 0.4087,
"step": 2402
},
{
"epoch": 0.9596645367412141,
"grad_norm": 8.895151138305664,
"learning_rate": 4.3130990415335465e-06,
"loss": 0.0502,
"step": 2403
},
{
"epoch": 0.9600638977635783,
"grad_norm": 42.020259857177734,
"learning_rate": 4.273162939297125e-06,
"loss": 0.7708,
"step": 2404
},
{
"epoch": 0.9604632587859425,
"grad_norm": 16.39902114868164,
"learning_rate": 4.233226837060703e-06,
"loss": 0.1879,
"step": 2405
},
{
"epoch": 0.9608626198083067,
"grad_norm": 34.795047760009766,
"learning_rate": 4.1932907348242814e-06,
"loss": 0.7724,
"step": 2406
},
{
"epoch": 0.9612619808306709,
"grad_norm": 5.834359169006348,
"learning_rate": 4.153354632587859e-06,
"loss": 0.0437,
"step": 2407
},
{
"epoch": 0.9616613418530351,
"grad_norm": 0.9828183054924011,
"learning_rate": 4.113418530351438e-06,
"loss": 0.0051,
"step": 2408
},
{
"epoch": 0.9620607028753994,
"grad_norm": 2.1337320804595947,
"learning_rate": 4.073482428115016e-06,
"loss": 0.0181,
"step": 2409
},
{
"epoch": 0.9624600638977636,
"grad_norm": 13.43707275390625,
"learning_rate": 4.033546325878594e-06,
"loss": 0.079,
"step": 2410
},
{
"epoch": 0.9628594249201278,
"grad_norm": 4.026017665863037,
"learning_rate": 3.993610223642173e-06,
"loss": 0.0204,
"step": 2411
},
{
"epoch": 0.963258785942492,
"grad_norm": 24.468547821044922,
"learning_rate": 3.953674121405751e-06,
"loss": 0.2563,
"step": 2412
},
{
"epoch": 0.9636581469648562,
"grad_norm": 17.157119750976562,
"learning_rate": 3.91373801916933e-06,
"loss": 0.1448,
"step": 2413
},
{
"epoch": 0.9640575079872205,
"grad_norm": 17.715129852294922,
"learning_rate": 3.8738019169329075e-06,
"loss": 0.483,
"step": 2414
},
{
"epoch": 0.9644568690095847,
"grad_norm": 12.795321464538574,
"learning_rate": 3.833865814696485e-06,
"loss": 0.2605,
"step": 2415
},
{
"epoch": 0.9648562300319489,
"grad_norm": 9.92728042602539,
"learning_rate": 3.7939297124600643e-06,
"loss": 0.0598,
"step": 2416
},
{
"epoch": 0.9652555910543131,
"grad_norm": 25.595884323120117,
"learning_rate": 3.7539936102236424e-06,
"loss": 0.4766,
"step": 2417
},
{
"epoch": 0.9656549520766773,
"grad_norm": 4.1089701652526855,
"learning_rate": 3.71405750798722e-06,
"loss": 0.0253,
"step": 2418
},
{
"epoch": 0.9660543130990416,
"grad_norm": 0.9210758805274963,
"learning_rate": 3.6741214057507987e-06,
"loss": 0.005,
"step": 2419
},
{
"epoch": 0.9664536741214057,
"grad_norm": 6.59714412689209,
"learning_rate": 3.6341853035143773e-06,
"loss": 0.0383,
"step": 2420
},
{
"epoch": 0.96685303514377,
"grad_norm": 0.061767518520355225,
"learning_rate": 3.5942492012779555e-06,
"loss": 0.0004,
"step": 2421
},
{
"epoch": 0.9672523961661342,
"grad_norm": 11.18925952911377,
"learning_rate": 3.5543130990415336e-06,
"loss": 0.7649,
"step": 2422
},
{
"epoch": 0.9676517571884984,
"grad_norm": 23.92156410217285,
"learning_rate": 3.5143769968051118e-06,
"loss": 0.5612,
"step": 2423
},
{
"epoch": 0.9680511182108626,
"grad_norm": 13.648837089538574,
"learning_rate": 3.4744408945686903e-06,
"loss": 0.5967,
"step": 2424
},
{
"epoch": 0.9684504792332268,
"grad_norm": 0.42803990840911865,
"learning_rate": 3.434504792332269e-06,
"loss": 0.0022,
"step": 2425
},
{
"epoch": 0.9688498402555911,
"grad_norm": 1.8483178615570068,
"learning_rate": 3.3945686900958466e-06,
"loss": 0.0118,
"step": 2426
},
{
"epoch": 0.9692492012779552,
"grad_norm": 3.051996946334839,
"learning_rate": 3.354632587859425e-06,
"loss": 0.0452,
"step": 2427
},
{
"epoch": 0.9696485623003195,
"grad_norm": 0.46153897047042847,
"learning_rate": 3.3146964856230034e-06,
"loss": 0.0038,
"step": 2428
},
{
"epoch": 0.9700479233226837,
"grad_norm": 10.072418212890625,
"learning_rate": 3.274760383386582e-06,
"loss": 0.0597,
"step": 2429
},
{
"epoch": 0.9704472843450479,
"grad_norm": 32.10613250732422,
"learning_rate": 3.2348242811501597e-06,
"loss": 0.377,
"step": 2430
},
{
"epoch": 0.9708466453674122,
"grad_norm": 5.294210433959961,
"learning_rate": 3.194888178913738e-06,
"loss": 0.0316,
"step": 2431
},
{
"epoch": 0.9712460063897763,
"grad_norm": 2.1650867462158203,
"learning_rate": 3.1549520766773164e-06,
"loss": 0.0118,
"step": 2432
},
{
"epoch": 0.9716453674121406,
"grad_norm": 0.9304328560829163,
"learning_rate": 3.1150159744408946e-06,
"loss": 0.0047,
"step": 2433
},
{
"epoch": 0.9720447284345048,
"grad_norm": 1.3281729221343994,
"learning_rate": 3.075079872204473e-06,
"loss": 0.0114,
"step": 2434
},
{
"epoch": 0.972444089456869,
"grad_norm": 35.94524383544922,
"learning_rate": 3.0351437699680513e-06,
"loss": 0.2908,
"step": 2435
},
{
"epoch": 0.9728434504792333,
"grad_norm": 22.210372924804688,
"learning_rate": 2.9952076677316295e-06,
"loss": 0.2426,
"step": 2436
},
{
"epoch": 0.9732428115015974,
"grad_norm": 0.4289431571960449,
"learning_rate": 2.9552715654952076e-06,
"loss": 0.0034,
"step": 2437
},
{
"epoch": 0.9736421725239617,
"grad_norm": 15.886605262756348,
"learning_rate": 2.915335463258786e-06,
"loss": 0.1603,
"step": 2438
},
{
"epoch": 0.9740415335463258,
"grad_norm": 27.158832550048828,
"learning_rate": 2.8753993610223644e-06,
"loss": 0.6368,
"step": 2439
},
{
"epoch": 0.9744408945686901,
"grad_norm": 14.720096588134766,
"learning_rate": 2.8354632587859425e-06,
"loss": 0.1249,
"step": 2440
},
{
"epoch": 0.9748402555910544,
"grad_norm": 18.753557205200195,
"learning_rate": 2.7955271565495207e-06,
"loss": 0.1431,
"step": 2441
},
{
"epoch": 0.9752396166134185,
"grad_norm": 25.568758010864258,
"learning_rate": 2.7555910543130992e-06,
"loss": 0.3295,
"step": 2442
},
{
"epoch": 0.9756389776357828,
"grad_norm": 5.250750541687012,
"learning_rate": 2.7156549520766774e-06,
"loss": 0.0296,
"step": 2443
},
{
"epoch": 0.9760383386581469,
"grad_norm": 31.85196304321289,
"learning_rate": 2.675718849840256e-06,
"loss": 0.7586,
"step": 2444
},
{
"epoch": 0.9764376996805112,
"grad_norm": 10.516316413879395,
"learning_rate": 2.6357827476038337e-06,
"loss": 0.0643,
"step": 2445
},
{
"epoch": 0.9768370607028753,
"grad_norm": 44.44916915893555,
"learning_rate": 2.5958466453674123e-06,
"loss": 0.7131,
"step": 2446
},
{
"epoch": 0.9772364217252396,
"grad_norm": 5.133357048034668,
"learning_rate": 2.5559105431309904e-06,
"loss": 0.1009,
"step": 2447
},
{
"epoch": 0.9776357827476039,
"grad_norm": 36.29154968261719,
"learning_rate": 2.515974440894569e-06,
"loss": 0.2891,
"step": 2448
},
{
"epoch": 0.978035143769968,
"grad_norm": 2.3214895725250244,
"learning_rate": 2.476038338658147e-06,
"loss": 0.0125,
"step": 2449
},
{
"epoch": 0.9784345047923323,
"grad_norm": 21.88632583618164,
"learning_rate": 2.4361022364217253e-06,
"loss": 0.1976,
"step": 2450
},
{
"epoch": 0.9788338658146964,
"grad_norm": 2.5108139514923096,
"learning_rate": 2.3961661341853035e-06,
"loss": 0.0132,
"step": 2451
},
{
"epoch": 0.9792332268370607,
"grad_norm": 17.268936157226562,
"learning_rate": 2.356230031948882e-06,
"loss": 0.3298,
"step": 2452
},
{
"epoch": 0.979632587859425,
"grad_norm": 14.017935752868652,
"learning_rate": 2.3162939297124602e-06,
"loss": 0.2557,
"step": 2453
},
{
"epoch": 0.9800319488817891,
"grad_norm": 0.1412898302078247,
"learning_rate": 2.2763578274760384e-06,
"loss": 0.0008,
"step": 2454
},
{
"epoch": 0.9804313099041534,
"grad_norm": 3.6527366638183594,
"learning_rate": 2.2364217252396165e-06,
"loss": 0.0312,
"step": 2455
},
{
"epoch": 0.9808306709265175,
"grad_norm": 47.88630294799805,
"learning_rate": 2.196485623003195e-06,
"loss": 0.9848,
"step": 2456
},
{
"epoch": 0.9812300319488818,
"grad_norm": 12.234745979309082,
"learning_rate": 2.1565495207667733e-06,
"loss": 0.0828,
"step": 2457
},
{
"epoch": 0.981629392971246,
"grad_norm": 0.517323911190033,
"learning_rate": 2.1166134185303514e-06,
"loss": 0.0034,
"step": 2458
},
{
"epoch": 0.9820287539936102,
"grad_norm": 27.938676834106445,
"learning_rate": 2.0766773162939296e-06,
"loss": 0.6197,
"step": 2459
},
{
"epoch": 0.9824281150159745,
"grad_norm": 25.854755401611328,
"learning_rate": 2.036741214057508e-06,
"loss": 0.58,
"step": 2460
},
{
"epoch": 0.9828274760383386,
"grad_norm": 39.98038101196289,
"learning_rate": 1.9968051118210863e-06,
"loss": 0.5386,
"step": 2461
},
{
"epoch": 0.9832268370607029,
"grad_norm": 14.912919998168945,
"learning_rate": 1.956869009584665e-06,
"loss": 0.1576,
"step": 2462
},
{
"epoch": 0.9836261980830671,
"grad_norm": 12.249653816223145,
"learning_rate": 1.9169329073482426e-06,
"loss": 1.1421,
"step": 2463
},
{
"epoch": 0.9840255591054313,
"grad_norm": 31.525224685668945,
"learning_rate": 1.8769968051118212e-06,
"loss": 0.4055,
"step": 2464
},
{
"epoch": 0.9844249201277955,
"grad_norm": 30.09122657775879,
"learning_rate": 1.8370607028753994e-06,
"loss": 0.4325,
"step": 2465
},
{
"epoch": 0.9848242811501597,
"grad_norm": 11.4757661819458,
"learning_rate": 1.7971246006389777e-06,
"loss": 0.0899,
"step": 2466
},
{
"epoch": 0.985223642172524,
"grad_norm": 2.731527328491211,
"learning_rate": 1.7571884984025559e-06,
"loss": 0.0161,
"step": 2467
},
{
"epoch": 0.9856230031948882,
"grad_norm": 30.222370147705078,
"learning_rate": 1.7172523961661345e-06,
"loss": 0.5428,
"step": 2468
},
{
"epoch": 0.9860223642172524,
"grad_norm": 11.862004280090332,
"learning_rate": 1.6773162939297124e-06,
"loss": 0.281,
"step": 2469
},
{
"epoch": 0.9864217252396166,
"grad_norm": 0.36124488711357117,
"learning_rate": 1.637380191693291e-06,
"loss": 0.0032,
"step": 2470
},
{
"epoch": 0.9868210862619808,
"grad_norm": 22.596059799194336,
"learning_rate": 1.597444089456869e-06,
"loss": 0.26,
"step": 2471
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.9926369786262512,
"learning_rate": 1.5575079872204473e-06,
"loss": 0.0052,
"step": 2472
},
{
"epoch": 0.9876198083067093,
"grad_norm": 7.41887092590332,
"learning_rate": 1.5175718849840257e-06,
"loss": 0.0404,
"step": 2473
},
{
"epoch": 0.9880191693290735,
"grad_norm": 15.789602279663086,
"learning_rate": 1.4776357827476038e-06,
"loss": 0.0928,
"step": 2474
},
{
"epoch": 0.9884185303514377,
"grad_norm": 23.634061813354492,
"learning_rate": 1.4376996805111822e-06,
"loss": 0.5798,
"step": 2475
},
{
"epoch": 0.9888178913738019,
"grad_norm": 0.7789590954780579,
"learning_rate": 1.3977635782747603e-06,
"loss": 0.0055,
"step": 2476
},
{
"epoch": 0.9892172523961661,
"grad_norm": 28.20090675354004,
"learning_rate": 1.3578274760383387e-06,
"loss": 0.5132,
"step": 2477
},
{
"epoch": 0.9896166134185304,
"grad_norm": 41.384700775146484,
"learning_rate": 1.3178913738019169e-06,
"loss": 1.0393,
"step": 2478
},
{
"epoch": 0.9900159744408946,
"grad_norm": 5.890468597412109,
"learning_rate": 1.2779552715654952e-06,
"loss": 0.0419,
"step": 2479
},
{
"epoch": 0.9904153354632588,
"grad_norm": 27.606475830078125,
"learning_rate": 1.2380191693290736e-06,
"loss": 0.5094,
"step": 2480
},
{
"epoch": 0.990814696485623,
"grad_norm": 0.01615994982421398,
"learning_rate": 1.1980830670926517e-06,
"loss": 0.0003,
"step": 2481
},
{
"epoch": 0.9912140575079872,
"grad_norm": 25.46713638305664,
"learning_rate": 1.1581469648562301e-06,
"loss": 0.5218,
"step": 2482
},
{
"epoch": 0.9916134185303515,
"grad_norm": 15.973341941833496,
"learning_rate": 1.1182108626198083e-06,
"loss": 0.1173,
"step": 2483
},
{
"epoch": 0.9920127795527156,
"grad_norm": 2.3560566902160645,
"learning_rate": 1.0782747603833866e-06,
"loss": 0.0166,
"step": 2484
},
{
"epoch": 0.9924121405750799,
"grad_norm": 2.720769166946411,
"learning_rate": 1.0383386581469648e-06,
"loss": 0.0141,
"step": 2485
},
{
"epoch": 0.9928115015974441,
"grad_norm": 28.12672996520996,
"learning_rate": 9.984025559105432e-07,
"loss": 0.9129,
"step": 2486
},
{
"epoch": 0.9932108626198083,
"grad_norm": 12.441367149353027,
"learning_rate": 9.584664536741213e-07,
"loss": 0.0893,
"step": 2487
},
{
"epoch": 0.9936102236421726,
"grad_norm": 29.818260192871094,
"learning_rate": 9.185303514376997e-07,
"loss": 0.9816,
"step": 2488
},
{
"epoch": 0.9940095846645367,
"grad_norm": 22.679412841796875,
"learning_rate": 8.785942492012779e-07,
"loss": 0.2136,
"step": 2489
},
{
"epoch": 0.994408945686901,
"grad_norm": 1.8009179830551147,
"learning_rate": 8.386581469648562e-07,
"loss": 0.0119,
"step": 2490
},
{
"epoch": 0.9948083067092651,
"grad_norm": 12.913415908813477,
"learning_rate": 7.987220447284345e-07,
"loss": 0.1772,
"step": 2491
},
{
"epoch": 0.9952076677316294,
"grad_norm": 19.164196014404297,
"learning_rate": 7.587859424920128e-07,
"loss": 0.1817,
"step": 2492
},
{
"epoch": 0.9956070287539937,
"grad_norm": 6.7110185623168945,
"learning_rate": 7.188498402555911e-07,
"loss": 0.0386,
"step": 2493
},
{
"epoch": 0.9960063897763578,
"grad_norm": 2.2846806049346924,
"learning_rate": 6.789137380191694e-07,
"loss": 0.015,
"step": 2494
},
{
"epoch": 0.9964057507987221,
"grad_norm": 36.19529342651367,
"learning_rate": 6.389776357827476e-07,
"loss": 0.367,
"step": 2495
},
{
"epoch": 0.9968051118210862,
"grad_norm": 14.790620803833008,
"learning_rate": 5.990415335463259e-07,
"loss": 0.2825,
"step": 2496
},
{
"epoch": 0.9972044728434505,
"grad_norm": 4.378154277801514,
"learning_rate": 5.591054313099041e-07,
"loss": 0.0254,
"step": 2497
},
{
"epoch": 0.9976038338658147,
"grad_norm": 29.516511917114258,
"learning_rate": 5.191693290734824e-07,
"loss": 0.7071,
"step": 2498
},
{
"epoch": 0.9980031948881789,
"grad_norm": 1.9763110876083374,
"learning_rate": 4.792332268370607e-07,
"loss": 0.0113,
"step": 2499
},
{
"epoch": 0.9984025559105432,
"grad_norm": 0.9840452671051025,
"learning_rate": 4.3929712460063897e-07,
"loss": 0.0048,
"step": 2500
},
{
"epoch": 0.9988019169329073,
"grad_norm": 50.83481979370117,
"learning_rate": 3.9936102236421723e-07,
"loss": 2.0288,
"step": 2501
},
{
"epoch": 0.9992012779552716,
"grad_norm": 1.1539658308029175,
"learning_rate": 3.5942492012779555e-07,
"loss": 0.0145,
"step": 2502
},
{
"epoch": 0.9996006389776357,
"grad_norm": 13.557058334350586,
"learning_rate": 3.194888178913738e-07,
"loss": 0.1185,
"step": 2503
},
{
"epoch": 1.0,
"grad_norm": 0.5868565440177917,
"learning_rate": 2.7955271565495207e-07,
"loss": 0.0027,
"step": 2504
}
],
"logging_steps": 1,
"max_steps": 2504,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.302025062875136e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}