fengyao1909's picture
Upload folder using huggingface_hub
8359d2e verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 939,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003194888178913738,
"grad_norm": 16.717142207453772,
"learning_rate": 0.0,
"loss": 1.805,
"step": 1
},
{
"epoch": 0.006389776357827476,
"grad_norm": 14.023251259992671,
"learning_rate": 1.0638297872340426e-07,
"loss": 1.7437,
"step": 2
},
{
"epoch": 0.009584664536741214,
"grad_norm": 20.233063373741423,
"learning_rate": 2.1276595744680852e-07,
"loss": 1.7713,
"step": 3
},
{
"epoch": 0.012779552715654952,
"grad_norm": 32.40631329434045,
"learning_rate": 3.1914893617021275e-07,
"loss": 1.8664,
"step": 4
},
{
"epoch": 0.01597444089456869,
"grad_norm": 11.151298510792111,
"learning_rate": 4.2553191489361704e-07,
"loss": 1.7119,
"step": 5
},
{
"epoch": 0.019169329073482427,
"grad_norm": 16.97709612304111,
"learning_rate": 5.319148936170213e-07,
"loss": 1.8852,
"step": 6
},
{
"epoch": 0.022364217252396165,
"grad_norm": 16.479785020928333,
"learning_rate": 6.382978723404255e-07,
"loss": 1.7909,
"step": 7
},
{
"epoch": 0.025559105431309903,
"grad_norm": 31.984915894604853,
"learning_rate": 7.446808510638298e-07,
"loss": 1.5985,
"step": 8
},
{
"epoch": 0.02875399361022364,
"grad_norm": 18.405770787032864,
"learning_rate": 8.510638297872341e-07,
"loss": 1.6649,
"step": 9
},
{
"epoch": 0.03194888178913738,
"grad_norm": 11.568457035620957,
"learning_rate": 9.574468085106384e-07,
"loss": 1.6343,
"step": 10
},
{
"epoch": 0.03514376996805112,
"grad_norm": 11.99369798309996,
"learning_rate": 1.0638297872340427e-06,
"loss": 1.7131,
"step": 11
},
{
"epoch": 0.038338658146964855,
"grad_norm": 16.579277724503285,
"learning_rate": 1.170212765957447e-06,
"loss": 1.6781,
"step": 12
},
{
"epoch": 0.04153354632587859,
"grad_norm": 23.280241560149864,
"learning_rate": 1.276595744680851e-06,
"loss": 1.9946,
"step": 13
},
{
"epoch": 0.04472843450479233,
"grad_norm": 11.069508575430309,
"learning_rate": 1.3829787234042555e-06,
"loss": 1.7037,
"step": 14
},
{
"epoch": 0.04792332268370607,
"grad_norm": 11.222580048610657,
"learning_rate": 1.4893617021276596e-06,
"loss": 1.5045,
"step": 15
},
{
"epoch": 0.051118210862619806,
"grad_norm": 9.262248481698279,
"learning_rate": 1.595744680851064e-06,
"loss": 1.6869,
"step": 16
},
{
"epoch": 0.054313099041533544,
"grad_norm": 8.164029447660837,
"learning_rate": 1.7021276595744682e-06,
"loss": 1.6781,
"step": 17
},
{
"epoch": 0.05750798722044728,
"grad_norm": 11.070212410234445,
"learning_rate": 1.8085106382978727e-06,
"loss": 1.5084,
"step": 18
},
{
"epoch": 0.06070287539936102,
"grad_norm": 18.24579269482639,
"learning_rate": 1.9148936170212767e-06,
"loss": 1.6718,
"step": 19
},
{
"epoch": 0.06389776357827476,
"grad_norm": 7.459594379769327,
"learning_rate": 2.021276595744681e-06,
"loss": 1.4803,
"step": 20
},
{
"epoch": 0.0670926517571885,
"grad_norm": 10.804667596623585,
"learning_rate": 2.1276595744680853e-06,
"loss": 1.451,
"step": 21
},
{
"epoch": 0.07028753993610223,
"grad_norm": 7.21964134256131,
"learning_rate": 2.2340425531914894e-06,
"loss": 1.522,
"step": 22
},
{
"epoch": 0.07348242811501597,
"grad_norm": 10.067355095577458,
"learning_rate": 2.340425531914894e-06,
"loss": 1.4243,
"step": 23
},
{
"epoch": 0.07667731629392971,
"grad_norm": 16.1259819071595,
"learning_rate": 2.446808510638298e-06,
"loss": 1.3842,
"step": 24
},
{
"epoch": 0.07987220447284345,
"grad_norm": 10.371859175411412,
"learning_rate": 2.553191489361702e-06,
"loss": 1.383,
"step": 25
},
{
"epoch": 0.08306709265175719,
"grad_norm": 8.964914217789572,
"learning_rate": 2.6595744680851065e-06,
"loss": 1.5774,
"step": 26
},
{
"epoch": 0.08626198083067092,
"grad_norm": 7.746583868291803,
"learning_rate": 2.765957446808511e-06,
"loss": 1.5022,
"step": 27
},
{
"epoch": 0.08945686900958466,
"grad_norm": 5.89473714557576,
"learning_rate": 2.8723404255319155e-06,
"loss": 1.3396,
"step": 28
},
{
"epoch": 0.0926517571884984,
"grad_norm": 5.02699347301002,
"learning_rate": 2.978723404255319e-06,
"loss": 1.3837,
"step": 29
},
{
"epoch": 0.09584664536741214,
"grad_norm": 9.371837428436915,
"learning_rate": 3.0851063829787237e-06,
"loss": 1.5456,
"step": 30
},
{
"epoch": 0.09904153354632587,
"grad_norm": 11.06547207040692,
"learning_rate": 3.191489361702128e-06,
"loss": 1.4454,
"step": 31
},
{
"epoch": 0.10223642172523961,
"grad_norm": 9.221564102628951,
"learning_rate": 3.297872340425532e-06,
"loss": 1.2007,
"step": 32
},
{
"epoch": 0.10543130990415335,
"grad_norm": 8.929546733679098,
"learning_rate": 3.4042553191489363e-06,
"loss": 1.4135,
"step": 33
},
{
"epoch": 0.10862619808306709,
"grad_norm": 9.211874077709819,
"learning_rate": 3.510638297872341e-06,
"loss": 1.2891,
"step": 34
},
{
"epoch": 0.11182108626198083,
"grad_norm": 7.000308016735031,
"learning_rate": 3.6170212765957453e-06,
"loss": 1.2162,
"step": 35
},
{
"epoch": 0.11501597444089456,
"grad_norm": 4.7532035690557475,
"learning_rate": 3.723404255319149e-06,
"loss": 1.2186,
"step": 36
},
{
"epoch": 0.1182108626198083,
"grad_norm": 6.27451384888466,
"learning_rate": 3.8297872340425535e-06,
"loss": 1.22,
"step": 37
},
{
"epoch": 0.12140575079872204,
"grad_norm": 5.883795747493129,
"learning_rate": 3.936170212765958e-06,
"loss": 1.1862,
"step": 38
},
{
"epoch": 0.12460063897763578,
"grad_norm": 5.585568962155399,
"learning_rate": 4.042553191489362e-06,
"loss": 1.2616,
"step": 39
},
{
"epoch": 0.12779552715654952,
"grad_norm": 6.088301652676503,
"learning_rate": 4.148936170212766e-06,
"loss": 1.2963,
"step": 40
},
{
"epoch": 0.13099041533546327,
"grad_norm": 6.236302292423119,
"learning_rate": 4.255319148936171e-06,
"loss": 1.3596,
"step": 41
},
{
"epoch": 0.134185303514377,
"grad_norm": 7.277813253034166,
"learning_rate": 4.361702127659575e-06,
"loss": 1.355,
"step": 42
},
{
"epoch": 0.13738019169329074,
"grad_norm": 6.9466699695085925,
"learning_rate": 4.468085106382979e-06,
"loss": 1.2406,
"step": 43
},
{
"epoch": 0.14057507987220447,
"grad_norm": 5.400956662876495,
"learning_rate": 4.574468085106383e-06,
"loss": 1.174,
"step": 44
},
{
"epoch": 0.14376996805111822,
"grad_norm": 4.824827492352657,
"learning_rate": 4.680851063829788e-06,
"loss": 1.1444,
"step": 45
},
{
"epoch": 0.14696485623003194,
"grad_norm": 4.13139301261386,
"learning_rate": 4.787234042553192e-06,
"loss": 0.9866,
"step": 46
},
{
"epoch": 0.1501597444089457,
"grad_norm": 4.48141844546404,
"learning_rate": 4.893617021276596e-06,
"loss": 1.2958,
"step": 47
},
{
"epoch": 0.15335463258785942,
"grad_norm": 5.194477479016162,
"learning_rate": 5e-06,
"loss": 1.279,
"step": 48
},
{
"epoch": 0.15654952076677317,
"grad_norm": 42.09662909433009,
"learning_rate": 5.106382978723404e-06,
"loss": 1.0737,
"step": 49
},
{
"epoch": 0.1597444089456869,
"grad_norm": 6.67181650955879,
"learning_rate": 5.212765957446809e-06,
"loss": 1.174,
"step": 50
},
{
"epoch": 0.16293929712460065,
"grad_norm": 4.305530439030815,
"learning_rate": 5.319148936170213e-06,
"loss": 1.1243,
"step": 51
},
{
"epoch": 0.16613418530351437,
"grad_norm": 5.8404661477440545,
"learning_rate": 5.425531914893617e-06,
"loss": 1.2639,
"step": 52
},
{
"epoch": 0.16932907348242812,
"grad_norm": 5.257506385774717,
"learning_rate": 5.531914893617022e-06,
"loss": 1.0899,
"step": 53
},
{
"epoch": 0.17252396166134185,
"grad_norm": 4.775620240966863,
"learning_rate": 5.638297872340426e-06,
"loss": 1.0457,
"step": 54
},
{
"epoch": 0.1757188498402556,
"grad_norm": 5.764147717959987,
"learning_rate": 5.744680851063831e-06,
"loss": 1.1835,
"step": 55
},
{
"epoch": 0.17891373801916932,
"grad_norm": 4.098061876130292,
"learning_rate": 5.851063829787235e-06,
"loss": 0.9886,
"step": 56
},
{
"epoch": 0.18210862619808307,
"grad_norm": 4.179637345679237,
"learning_rate": 5.957446808510638e-06,
"loss": 1.3009,
"step": 57
},
{
"epoch": 0.1853035143769968,
"grad_norm": 3.9920034272044003,
"learning_rate": 6.063829787234044e-06,
"loss": 1.1074,
"step": 58
},
{
"epoch": 0.18849840255591055,
"grad_norm": 4.408578981594532,
"learning_rate": 6.170212765957447e-06,
"loss": 1.1697,
"step": 59
},
{
"epoch": 0.19169329073482427,
"grad_norm": 4.325540588859466,
"learning_rate": 6.276595744680851e-06,
"loss": 1.0138,
"step": 60
},
{
"epoch": 0.19488817891373802,
"grad_norm": 5.837072539573206,
"learning_rate": 6.382978723404256e-06,
"loss": 1.1069,
"step": 61
},
{
"epoch": 0.19808306709265175,
"grad_norm": 3.4143360478794906,
"learning_rate": 6.48936170212766e-06,
"loss": 1.0702,
"step": 62
},
{
"epoch": 0.2012779552715655,
"grad_norm": 5.18488681162557,
"learning_rate": 6.595744680851064e-06,
"loss": 1.2569,
"step": 63
},
{
"epoch": 0.20447284345047922,
"grad_norm": 3.9554222708423223,
"learning_rate": 6.702127659574469e-06,
"loss": 1.0897,
"step": 64
},
{
"epoch": 0.20766773162939298,
"grad_norm": 3.6286361737781143,
"learning_rate": 6.808510638297873e-06,
"loss": 1.0742,
"step": 65
},
{
"epoch": 0.2108626198083067,
"grad_norm": 3.958289362639308,
"learning_rate": 6.914893617021278e-06,
"loss": 1.2868,
"step": 66
},
{
"epoch": 0.21405750798722045,
"grad_norm": 4.150110579938386,
"learning_rate": 7.021276595744682e-06,
"loss": 1.0947,
"step": 67
},
{
"epoch": 0.21725239616613418,
"grad_norm": 3.3116634215622263,
"learning_rate": 7.127659574468085e-06,
"loss": 1.1115,
"step": 68
},
{
"epoch": 0.22044728434504793,
"grad_norm": 4.6469898720363965,
"learning_rate": 7.234042553191491e-06,
"loss": 1.0326,
"step": 69
},
{
"epoch": 0.22364217252396165,
"grad_norm": 4.0281172273912995,
"learning_rate": 7.340425531914894e-06,
"loss": 1.1844,
"step": 70
},
{
"epoch": 0.2268370607028754,
"grad_norm": 4.062077614294185,
"learning_rate": 7.446808510638298e-06,
"loss": 1.2275,
"step": 71
},
{
"epoch": 0.23003194888178913,
"grad_norm": 3.749443353016541,
"learning_rate": 7.553191489361703e-06,
"loss": 1.1601,
"step": 72
},
{
"epoch": 0.23322683706070288,
"grad_norm": 4.506630758336906,
"learning_rate": 7.659574468085107e-06,
"loss": 1.181,
"step": 73
},
{
"epoch": 0.2364217252396166,
"grad_norm": 3.679699351741113,
"learning_rate": 7.765957446808511e-06,
"loss": 1.2492,
"step": 74
},
{
"epoch": 0.23961661341853036,
"grad_norm": 4.41567402339775,
"learning_rate": 7.872340425531916e-06,
"loss": 1.265,
"step": 75
},
{
"epoch": 0.24281150159744408,
"grad_norm": 3.951290133921461,
"learning_rate": 7.97872340425532e-06,
"loss": 1.1003,
"step": 76
},
{
"epoch": 0.24600638977635783,
"grad_norm": 8.461275376215076,
"learning_rate": 8.085106382978723e-06,
"loss": 0.9229,
"step": 77
},
{
"epoch": 0.24920127795527156,
"grad_norm": 4.528540314230216,
"learning_rate": 8.191489361702128e-06,
"loss": 1.2059,
"step": 78
},
{
"epoch": 0.2523961661341853,
"grad_norm": 3.826798452811914,
"learning_rate": 8.297872340425532e-06,
"loss": 1.1811,
"step": 79
},
{
"epoch": 0.25559105431309903,
"grad_norm": 4.011312797753363,
"learning_rate": 8.404255319148937e-06,
"loss": 1.233,
"step": 80
},
{
"epoch": 0.25878594249201275,
"grad_norm": 4.496947309535494,
"learning_rate": 8.510638297872341e-06,
"loss": 1.1934,
"step": 81
},
{
"epoch": 0.26198083067092653,
"grad_norm": 4.12985941671228,
"learning_rate": 8.617021276595746e-06,
"loss": 1.1893,
"step": 82
},
{
"epoch": 0.26517571884984026,
"grad_norm": 4.133181789749206,
"learning_rate": 8.72340425531915e-06,
"loss": 1.2122,
"step": 83
},
{
"epoch": 0.268370607028754,
"grad_norm": 4.397287646608956,
"learning_rate": 8.829787234042555e-06,
"loss": 1.2728,
"step": 84
},
{
"epoch": 0.2715654952076677,
"grad_norm": 3.9021342321758747,
"learning_rate": 8.936170212765958e-06,
"loss": 1.1245,
"step": 85
},
{
"epoch": 0.2747603833865815,
"grad_norm": 3.900295916232368,
"learning_rate": 9.042553191489362e-06,
"loss": 1.0495,
"step": 86
},
{
"epoch": 0.2779552715654952,
"grad_norm": 4.391512053334418,
"learning_rate": 9.148936170212767e-06,
"loss": 1.1569,
"step": 87
},
{
"epoch": 0.28115015974440893,
"grad_norm": 3.569622297990687,
"learning_rate": 9.255319148936171e-06,
"loss": 1.0791,
"step": 88
},
{
"epoch": 0.28434504792332266,
"grad_norm": 4.239762675196147,
"learning_rate": 9.361702127659576e-06,
"loss": 1.0326,
"step": 89
},
{
"epoch": 0.28753993610223644,
"grad_norm": 3.643271220308426,
"learning_rate": 9.46808510638298e-06,
"loss": 1.0463,
"step": 90
},
{
"epoch": 0.29073482428115016,
"grad_norm": 3.8312114090333904,
"learning_rate": 9.574468085106385e-06,
"loss": 1.2471,
"step": 91
},
{
"epoch": 0.2939297124600639,
"grad_norm": 3.789181561886315,
"learning_rate": 9.680851063829787e-06,
"loss": 1.0575,
"step": 92
},
{
"epoch": 0.2971246006389776,
"grad_norm": 4.258663565413159,
"learning_rate": 9.787234042553192e-06,
"loss": 0.9664,
"step": 93
},
{
"epoch": 0.3003194888178914,
"grad_norm": 3.9894692965210354,
"learning_rate": 9.893617021276596e-06,
"loss": 1.1386,
"step": 94
},
{
"epoch": 0.3035143769968051,
"grad_norm": 3.9831887232305476,
"learning_rate": 1e-05,
"loss": 1.1514,
"step": 95
},
{
"epoch": 0.30670926517571884,
"grad_norm": 4.958061793339305,
"learning_rate": 9.999965443811378e-06,
"loss": 1.0876,
"step": 96
},
{
"epoch": 0.30990415335463256,
"grad_norm": 3.6440501254714697,
"learning_rate": 9.999861775723162e-06,
"loss": 1.0402,
"step": 97
},
{
"epoch": 0.31309904153354634,
"grad_norm": 3.8161236632180127,
"learning_rate": 9.999688997168301e-06,
"loss": 0.7872,
"step": 98
},
{
"epoch": 0.31629392971246006,
"grad_norm": 4.376751849770681,
"learning_rate": 9.999447110535026e-06,
"loss": 1.0524,
"step": 99
},
{
"epoch": 0.3194888178913738,
"grad_norm": 4.427888818606079,
"learning_rate": 9.999136119166803e-06,
"loss": 1.1416,
"step": 100
},
{
"epoch": 0.3226837060702875,
"grad_norm": 4.295006073567966,
"learning_rate": 9.998756027362308e-06,
"loss": 1.0189,
"step": 101
},
{
"epoch": 0.3258785942492013,
"grad_norm": 3.717219342315633,
"learning_rate": 9.99830684037535e-06,
"loss": 1.1725,
"step": 102
},
{
"epoch": 0.329073482428115,
"grad_norm": 3.66920943595835,
"learning_rate": 9.9977885644148e-06,
"loss": 1.1265,
"step": 103
},
{
"epoch": 0.33226837060702874,
"grad_norm": 3.8472273117866034,
"learning_rate": 9.997201206644522e-06,
"loss": 1.0991,
"step": 104
},
{
"epoch": 0.3354632587859425,
"grad_norm": 4.601449936195102,
"learning_rate": 9.99654477518325e-06,
"loss": 1.1958,
"step": 105
},
{
"epoch": 0.33865814696485624,
"grad_norm": 4.21713584274273,
"learning_rate": 9.995819279104494e-06,
"loss": 1.1656,
"step": 106
},
{
"epoch": 0.34185303514376997,
"grad_norm": 4.682764581823147,
"learning_rate": 9.995024728436402e-06,
"loss": 1.3324,
"step": 107
},
{
"epoch": 0.3450479233226837,
"grad_norm": 4.365236517611888,
"learning_rate": 9.994161134161635e-06,
"loss": 1.0721,
"step": 108
},
{
"epoch": 0.34824281150159747,
"grad_norm": 4.509106719765009,
"learning_rate": 9.993228508217201e-06,
"loss": 1.2674,
"step": 109
},
{
"epoch": 0.3514376996805112,
"grad_norm": 3.947735866465357,
"learning_rate": 9.9922268634943e-06,
"loss": 0.931,
"step": 110
},
{
"epoch": 0.3546325878594249,
"grad_norm": 4.43142608253394,
"learning_rate": 9.991156213838143e-06,
"loss": 0.9717,
"step": 111
},
{
"epoch": 0.35782747603833864,
"grad_norm": 4.129661974265571,
"learning_rate": 9.990016574047757e-06,
"loss": 1.2133,
"step": 112
},
{
"epoch": 0.3610223642172524,
"grad_norm": 3.96529852403178,
"learning_rate": 9.988807959875785e-06,
"loss": 1.0823,
"step": 113
},
{
"epoch": 0.36421725239616615,
"grad_norm": 3.360371849956031,
"learning_rate": 9.987530388028269e-06,
"loss": 1.2586,
"step": 114
},
{
"epoch": 0.36741214057507987,
"grad_norm": 5.997628745903946,
"learning_rate": 9.986183876164412e-06,
"loss": 1.121,
"step": 115
},
{
"epoch": 0.3706070287539936,
"grad_norm": 3.42531462621357,
"learning_rate": 9.984768442896342e-06,
"loss": 1.1515,
"step": 116
},
{
"epoch": 0.3738019169329074,
"grad_norm": 4.259430537744425,
"learning_rate": 9.983284107788852e-06,
"loss": 1.1667,
"step": 117
},
{
"epoch": 0.3769968051118211,
"grad_norm": 4.5220331619833685,
"learning_rate": 9.981730891359123e-06,
"loss": 1.1715,
"step": 118
},
{
"epoch": 0.3801916932907348,
"grad_norm": 3.92086887197509,
"learning_rate": 9.980108815076456e-06,
"loss": 1.1054,
"step": 119
},
{
"epoch": 0.38338658146964855,
"grad_norm": 3.9180854105362135,
"learning_rate": 9.978417901361958e-06,
"loss": 1.1234,
"step": 120
},
{
"epoch": 0.3865814696485623,
"grad_norm": 4.821675602815471,
"learning_rate": 9.976658173588244e-06,
"loss": 1.3579,
"step": 121
},
{
"epoch": 0.38977635782747605,
"grad_norm": 3.9268713839450893,
"learning_rate": 9.974829656079106e-06,
"loss": 1.2452,
"step": 122
},
{
"epoch": 0.3929712460063898,
"grad_norm": 4.138184615063423,
"learning_rate": 9.972932374109184e-06,
"loss": 1.2322,
"step": 123
},
{
"epoch": 0.3961661341853035,
"grad_norm": 3.744611618137854,
"learning_rate": 9.97096635390361e-06,
"loss": 1.2847,
"step": 124
},
{
"epoch": 0.3993610223642173,
"grad_norm": 4.378499516464371,
"learning_rate": 9.968931622637652e-06,
"loss": 1.1174,
"step": 125
},
{
"epoch": 0.402555910543131,
"grad_norm": 4.14696690274647,
"learning_rate": 9.966828208436332e-06,
"loss": 1.2383,
"step": 126
},
{
"epoch": 0.4057507987220447,
"grad_norm": 3.819830516963947,
"learning_rate": 9.96465614037404e-06,
"loss": 1.0994,
"step": 127
},
{
"epoch": 0.40894568690095845,
"grad_norm": 3.715065492438179,
"learning_rate": 9.962415448474134e-06,
"loss": 1.2265,
"step": 128
},
{
"epoch": 0.41214057507987223,
"grad_norm": 4.0798891858967705,
"learning_rate": 9.960106163708522e-06,
"loss": 1.2507,
"step": 129
},
{
"epoch": 0.41533546325878595,
"grad_norm": 3.926712283440133,
"learning_rate": 9.95772831799724e-06,
"loss": 1.1499,
"step": 130
},
{
"epoch": 0.4185303514376997,
"grad_norm": 4.053856040897845,
"learning_rate": 9.955281944207998e-06,
"loss": 1.0657,
"step": 131
},
{
"epoch": 0.4217252396166134,
"grad_norm": 4.123586886707564,
"learning_rate": 9.95276707615574e-06,
"loss": 1.2163,
"step": 132
},
{
"epoch": 0.4249201277955272,
"grad_norm": 3.6806278803859316,
"learning_rate": 9.950183748602164e-06,
"loss": 1.124,
"step": 133
},
{
"epoch": 0.4281150159744409,
"grad_norm": 4.263198157219222,
"learning_rate": 9.947531997255256e-06,
"loss": 1.3273,
"step": 134
},
{
"epoch": 0.43130990415335463,
"grad_norm": 4.30356223939137,
"learning_rate": 9.944811858768782e-06,
"loss": 1.1945,
"step": 135
},
{
"epoch": 0.43450479233226835,
"grad_norm": 3.595940469656854,
"learning_rate": 9.94202337074179e-06,
"loss": 1.08,
"step": 136
},
{
"epoch": 0.43769968051118213,
"grad_norm": 4.576837266767333,
"learning_rate": 9.939166571718086e-06,
"loss": 1.2054,
"step": 137
},
{
"epoch": 0.44089456869009586,
"grad_norm": 4.483618367740623,
"learning_rate": 9.936241501185706e-06,
"loss": 1.1901,
"step": 138
},
{
"epoch": 0.4440894568690096,
"grad_norm": 4.336391651763086,
"learning_rate": 9.933248199576366e-06,
"loss": 1.3743,
"step": 139
},
{
"epoch": 0.4472843450479233,
"grad_norm": 4.409295671646476,
"learning_rate": 9.930186708264902e-06,
"loss": 1.0918,
"step": 140
},
{
"epoch": 0.4504792332268371,
"grad_norm": 3.9495959996762444,
"learning_rate": 9.927057069568704e-06,
"loss": 1.1489,
"step": 141
},
{
"epoch": 0.4536741214057508,
"grad_norm": 3.541718202366025,
"learning_rate": 9.923859326747125e-06,
"loss": 1.1549,
"step": 142
},
{
"epoch": 0.45686900958466453,
"grad_norm": 4.150454601010433,
"learning_rate": 9.920593524000887e-06,
"loss": 1.2059,
"step": 143
},
{
"epoch": 0.46006389776357826,
"grad_norm": 4.034114854054024,
"learning_rate": 9.917259706471469e-06,
"loss": 1.0233,
"step": 144
},
{
"epoch": 0.46325878594249204,
"grad_norm": 3.343995949761663,
"learning_rate": 9.913857920240481e-06,
"loss": 1.1326,
"step": 145
},
{
"epoch": 0.46645367412140576,
"grad_norm": 3.5039436799633807,
"learning_rate": 9.91038821232903e-06,
"loss": 1.1613,
"step": 146
},
{
"epoch": 0.4696485623003195,
"grad_norm": 3.524210670866061,
"learning_rate": 9.906850630697068e-06,
"loss": 1.2671,
"step": 147
},
{
"epoch": 0.4728434504792332,
"grad_norm": 4.022997493239621,
"learning_rate": 9.903245224242732e-06,
"loss": 1.1848,
"step": 148
},
{
"epoch": 0.476038338658147,
"grad_norm": 3.7665732907224325,
"learning_rate": 9.899572042801662e-06,
"loss": 1.0669,
"step": 149
},
{
"epoch": 0.4792332268370607,
"grad_norm": 3.6907954357736736,
"learning_rate": 9.895831137146319e-06,
"loss": 1.1884,
"step": 150
},
{
"epoch": 0.48242811501597443,
"grad_norm": 3.3930002700589608,
"learning_rate": 9.89202255898528e-06,
"loss": 1.1623,
"step": 151
},
{
"epoch": 0.48562300319488816,
"grad_norm": 3.8473272128449794,
"learning_rate": 9.888146360962523e-06,
"loss": 1.2449,
"step": 152
},
{
"epoch": 0.48881789137380194,
"grad_norm": 4.437542613948092,
"learning_rate": 9.8842025966567e-06,
"loss": 1.2139,
"step": 153
},
{
"epoch": 0.49201277955271566,
"grad_norm": 3.7262020124767767,
"learning_rate": 9.880191320580396e-06,
"loss": 1.0793,
"step": 154
},
{
"epoch": 0.4952076677316294,
"grad_norm": 3.6840608248822173,
"learning_rate": 9.876112588179378e-06,
"loss": 1.0027,
"step": 155
},
{
"epoch": 0.4984025559105431,
"grad_norm": 4.86699139564884,
"learning_rate": 9.87196645583182e-06,
"loss": 1.0692,
"step": 156
},
{
"epoch": 0.5015974440894568,
"grad_norm": 3.583521742071516,
"learning_rate": 9.86775298084754e-06,
"loss": 1.1271,
"step": 157
},
{
"epoch": 0.5047923322683706,
"grad_norm": 3.6875274203282613,
"learning_rate": 9.863472221467189e-06,
"loss": 1.1372,
"step": 158
},
{
"epoch": 0.5079872204472844,
"grad_norm": 3.9006310680068648,
"learning_rate": 9.85912423686146e-06,
"loss": 0.9988,
"step": 159
},
{
"epoch": 0.5111821086261981,
"grad_norm": 3.4245972439374337,
"learning_rate": 9.854709087130261e-06,
"loss": 1.1271,
"step": 160
},
{
"epoch": 0.5143769968051118,
"grad_norm": 3.73135489664885,
"learning_rate": 9.850226833301893e-06,
"loss": 1.1243,
"step": 161
},
{
"epoch": 0.5175718849840255,
"grad_norm": 3.841626672440205,
"learning_rate": 9.8456775373322e-06,
"loss": 1.3137,
"step": 162
},
{
"epoch": 0.5207667731629393,
"grad_norm": 3.5018138465293327,
"learning_rate": 9.841061262103713e-06,
"loss": 1.12,
"step": 163
},
{
"epoch": 0.5239616613418531,
"grad_norm": 4.3999556998319616,
"learning_rate": 9.836378071424782e-06,
"loss": 1.116,
"step": 164
},
{
"epoch": 0.5271565495207667,
"grad_norm": 3.744065463115353,
"learning_rate": 9.831628030028698e-06,
"loss": 0.9883,
"step": 165
},
{
"epoch": 0.5303514376996805,
"grad_norm": 3.4093472982635493,
"learning_rate": 9.826811203572785e-06,
"loss": 1.0762,
"step": 166
},
{
"epoch": 0.5335463258785943,
"grad_norm": 3.4565307473487024,
"learning_rate": 9.821927658637518e-06,
"loss": 1.0739,
"step": 167
},
{
"epoch": 0.536741214057508,
"grad_norm": 3.604865585999556,
"learning_rate": 9.81697746272557e-06,
"loss": 1.1226,
"step": 168
},
{
"epoch": 0.5399361022364217,
"grad_norm": 3.8477468031419675,
"learning_rate": 9.811960684260907e-06,
"loss": 1.2479,
"step": 169
},
{
"epoch": 0.5431309904153354,
"grad_norm": 3.8591542903479756,
"learning_rate": 9.80687739258782e-06,
"loss": 1.3551,
"step": 170
},
{
"epoch": 0.5463258785942492,
"grad_norm": 3.9050171697232448,
"learning_rate": 9.801727657969988e-06,
"loss": 1.1226,
"step": 171
},
{
"epoch": 0.549520766773163,
"grad_norm": 3.4041910605120216,
"learning_rate": 9.796511551589492e-06,
"loss": 1.1941,
"step": 172
},
{
"epoch": 0.5527156549520766,
"grad_norm": 3.766877733574309,
"learning_rate": 9.791229145545832e-06,
"loss": 1.055,
"step": 173
},
{
"epoch": 0.5559105431309904,
"grad_norm": 3.1732627754278595,
"learning_rate": 9.785880512854937e-06,
"loss": 1.0525,
"step": 174
},
{
"epoch": 0.5591054313099042,
"grad_norm": 4.593718811571636,
"learning_rate": 9.78046572744815e-06,
"loss": 1.1408,
"step": 175
},
{
"epoch": 0.5623003194888179,
"grad_norm": 4.179697128382144,
"learning_rate": 9.77498486417121e-06,
"loss": 1.0907,
"step": 176
},
{
"epoch": 0.5654952076677316,
"grad_norm": 3.8601601251447706,
"learning_rate": 9.769437998783216e-06,
"loss": 1.2654,
"step": 177
},
{
"epoch": 0.5686900958466453,
"grad_norm": 4.309939891784436,
"learning_rate": 9.763825207955577e-06,
"loss": 1.2503,
"step": 178
},
{
"epoch": 0.5718849840255591,
"grad_norm": 3.079242356976198,
"learning_rate": 9.758146569270957e-06,
"loss": 1.1562,
"step": 179
},
{
"epoch": 0.5750798722044729,
"grad_norm": 3.1013512719941447,
"learning_rate": 9.7524021612222e-06,
"loss": 1.0511,
"step": 180
},
{
"epoch": 0.5782747603833865,
"grad_norm": 4.4883464915821785,
"learning_rate": 9.746592063211247e-06,
"loss": 0.9574,
"step": 181
},
{
"epoch": 0.5814696485623003,
"grad_norm": 4.009545636808773,
"learning_rate": 9.74071635554803e-06,
"loss": 1.1486,
"step": 182
},
{
"epoch": 0.5846645367412141,
"grad_norm": 3.7280687100599526,
"learning_rate": 9.73477511944938e-06,
"loss": 1.0465,
"step": 183
},
{
"epoch": 0.5878594249201278,
"grad_norm": 3.4639376011854934,
"learning_rate": 9.728768437037882e-06,
"loss": 1.3021,
"step": 184
},
{
"epoch": 0.5910543130990416,
"grad_norm": 3.745998839913429,
"learning_rate": 9.722696391340762e-06,
"loss": 1.1834,
"step": 185
},
{
"epoch": 0.5942492012779552,
"grad_norm": 3.6721919646510046,
"learning_rate": 9.716559066288716e-06,
"loss": 1.0709,
"step": 186
},
{
"epoch": 0.597444089456869,
"grad_norm": 3.8661953097483366,
"learning_rate": 9.710356546714774e-06,
"loss": 1.1991,
"step": 187
},
{
"epoch": 0.6006389776357828,
"grad_norm": 3.528024996228283,
"learning_rate": 9.704088918353108e-06,
"loss": 0.9713,
"step": 188
},
{
"epoch": 0.6038338658146964,
"grad_norm": 3.460565087586548,
"learning_rate": 9.697756267837856e-06,
"loss": 1.2366,
"step": 189
},
{
"epoch": 0.6070287539936102,
"grad_norm": 4.445017669130635,
"learning_rate": 9.691358682701927e-06,
"loss": 1.0936,
"step": 190
},
{
"epoch": 0.610223642172524,
"grad_norm": 3.920662218861279,
"learning_rate": 9.684896251375784e-06,
"loss": 1.2067,
"step": 191
},
{
"epoch": 0.6134185303514377,
"grad_norm": 3.7352896533732833,
"learning_rate": 9.678369063186224e-06,
"loss": 1.185,
"step": 192
},
{
"epoch": 0.6166134185303515,
"grad_norm": 4.246203307684242,
"learning_rate": 9.671777208355146e-06,
"loss": 1.1054,
"step": 193
},
{
"epoch": 0.6198083067092651,
"grad_norm": 3.792906254800112,
"learning_rate": 9.665120777998303e-06,
"loss": 1.1175,
"step": 194
},
{
"epoch": 0.6230031948881789,
"grad_norm": 3.8136114173937936,
"learning_rate": 9.658399864124037e-06,
"loss": 1.4108,
"step": 195
},
{
"epoch": 0.6261980830670927,
"grad_norm": 3.3392586828268485,
"learning_rate": 9.65161455963202e-06,
"loss": 1.2708,
"step": 196
},
{
"epoch": 0.6293929712460063,
"grad_norm": 3.7861610422986605,
"learning_rate": 9.64476495831195e-06,
"loss": 1.3396,
"step": 197
},
{
"epoch": 0.6325878594249201,
"grad_norm": 3.346243409951804,
"learning_rate": 9.637851154842279e-06,
"loss": 1.0276,
"step": 198
},
{
"epoch": 0.6357827476038339,
"grad_norm": 4.57890383795362,
"learning_rate": 9.630873244788884e-06,
"loss": 1.2145,
"step": 199
},
{
"epoch": 0.6389776357827476,
"grad_norm": 4.001372541856091,
"learning_rate": 9.623831324603755e-06,
"loss": 1.197,
"step": 200
},
{
"epoch": 0.6421725239616614,
"grad_norm": 3.5844989464794277,
"learning_rate": 9.61672549162366e-06,
"loss": 0.9723,
"step": 201
},
{
"epoch": 0.645367412140575,
"grad_norm": 3.1764342201519256,
"learning_rate": 9.6095558440688e-06,
"loss": 1.037,
"step": 202
},
{
"epoch": 0.6485623003194888,
"grad_norm": 31.912165039553248,
"learning_rate": 9.602322481041457e-06,
"loss": 1.1477,
"step": 203
},
{
"epoch": 0.6517571884984026,
"grad_norm": 3.9661217859209343,
"learning_rate": 9.595025502524609e-06,
"loss": 1.1187,
"step": 204
},
{
"epoch": 0.6549520766773163,
"grad_norm": 4.872659949553548,
"learning_rate": 9.587665009380565e-06,
"loss": 1.1969,
"step": 205
},
{
"epoch": 0.65814696485623,
"grad_norm": 4.178000617919578,
"learning_rate": 9.580241103349562e-06,
"loss": 1.1576,
"step": 206
},
{
"epoch": 0.6613418530351438,
"grad_norm": 3.3175838649598384,
"learning_rate": 9.572753887048353e-06,
"loss": 1.3112,
"step": 207
},
{
"epoch": 0.6645367412140575,
"grad_norm": 4.180106920136841,
"learning_rate": 9.565203463968808e-06,
"loss": 1.2374,
"step": 208
},
{
"epoch": 0.6677316293929713,
"grad_norm": 4.263696376648943,
"learning_rate": 9.557589938476462e-06,
"loss": 1.177,
"step": 209
},
{
"epoch": 0.670926517571885,
"grad_norm": 3.728457720945257,
"learning_rate": 9.549913415809084e-06,
"loss": 1.1166,
"step": 210
},
{
"epoch": 0.6741214057507987,
"grad_norm": 3.716885354018328,
"learning_rate": 9.542174002075221e-06,
"loss": 1.1211,
"step": 211
},
{
"epoch": 0.6773162939297125,
"grad_norm": 3.441018098753554,
"learning_rate": 9.534371804252727e-06,
"loss": 0.9557,
"step": 212
},
{
"epoch": 0.6805111821086262,
"grad_norm": 3.756759802143899,
"learning_rate": 9.526506930187294e-06,
"loss": 1.049,
"step": 213
},
{
"epoch": 0.6837060702875399,
"grad_norm": 3.8823675044350394,
"learning_rate": 9.518579488590947e-06,
"loss": 1.0248,
"step": 214
},
{
"epoch": 0.6869009584664537,
"grad_norm": 3.4511617693540764,
"learning_rate": 9.510589589040554e-06,
"loss": 1.0499,
"step": 215
},
{
"epoch": 0.6900958466453674,
"grad_norm": 3.742245060357466,
"learning_rate": 9.502537341976305e-06,
"loss": 1.2358,
"step": 216
},
{
"epoch": 0.6932907348242812,
"grad_norm": 4.113022423699429,
"learning_rate": 9.494422858700188e-06,
"loss": 1.158,
"step": 217
},
{
"epoch": 0.6964856230031949,
"grad_norm": 4.064092320316261,
"learning_rate": 9.48624625137445e-06,
"loss": 1.1086,
"step": 218
},
{
"epoch": 0.6996805111821086,
"grad_norm": 4.593216406242684,
"learning_rate": 9.478007633020043e-06,
"loss": 1.1254,
"step": 219
},
{
"epoch": 0.7028753993610224,
"grad_norm": 3.3815419807535423,
"learning_rate": 9.469707117515068e-06,
"loss": 1.0612,
"step": 220
},
{
"epoch": 0.7060702875399361,
"grad_norm": 3.6836021833527375,
"learning_rate": 9.461344819593194e-06,
"loss": 1.1656,
"step": 221
},
{
"epoch": 0.7092651757188498,
"grad_norm": 3.3998820050869942,
"learning_rate": 9.452920854842085e-06,
"loss": 1.2333,
"step": 222
},
{
"epoch": 0.7124600638977636,
"grad_norm": 4.01284676293508,
"learning_rate": 9.44443533970178e-06,
"loss": 1.1678,
"step": 223
},
{
"epoch": 0.7156549520766773,
"grad_norm": 4.102396672543517,
"learning_rate": 9.435888391463108e-06,
"loss": 1.0636,
"step": 224
},
{
"epoch": 0.7188498402555911,
"grad_norm": 4.149794876674578,
"learning_rate": 9.427280128266049e-06,
"loss": 1.1873,
"step": 225
},
{
"epoch": 0.7220447284345048,
"grad_norm": 4.112252248723449,
"learning_rate": 9.418610669098114e-06,
"loss": 1.1098,
"step": 226
},
{
"epoch": 0.7252396166134185,
"grad_norm": 3.5274908089991013,
"learning_rate": 9.409880133792684e-06,
"loss": 1.283,
"step": 227
},
{
"epoch": 0.7284345047923323,
"grad_norm": 6.33028114914793,
"learning_rate": 9.40108864302737e-06,
"loss": 1.2498,
"step": 228
},
{
"epoch": 0.731629392971246,
"grad_norm": 3.774138442592385,
"learning_rate": 9.392236318322339e-06,
"loss": 0.9698,
"step": 229
},
{
"epoch": 0.7348242811501597,
"grad_norm": 4.581620520677188,
"learning_rate": 9.383323282038632e-06,
"loss": 1.2288,
"step": 230
},
{
"epoch": 0.7380191693290735,
"grad_norm": 4.09051981232859,
"learning_rate": 9.374349657376473e-06,
"loss": 1.2483,
"step": 231
},
{
"epoch": 0.7412140575079872,
"grad_norm": 4.329551570764206,
"learning_rate": 9.365315568373569e-06,
"loss": 1.2259,
"step": 232
},
{
"epoch": 0.744408945686901,
"grad_norm": 3.5577672099293665,
"learning_rate": 9.356221139903395e-06,
"loss": 1.099,
"step": 233
},
{
"epoch": 0.7476038338658147,
"grad_norm": 3.613277677521021,
"learning_rate": 9.347066497673462e-06,
"loss": 1.2454,
"step": 234
},
{
"epoch": 0.7507987220447284,
"grad_norm": 3.6145726973639043,
"learning_rate": 9.337851768223589e-06,
"loss": 1.0593,
"step": 235
},
{
"epoch": 0.7539936102236422,
"grad_norm": 3.631299813703971,
"learning_rate": 9.328577078924151e-06,
"loss": 1.2006,
"step": 236
},
{
"epoch": 0.7571884984025559,
"grad_norm": 4.357605136937369,
"learning_rate": 9.319242557974306e-06,
"loss": 1.0989,
"step": 237
},
{
"epoch": 0.7603833865814696,
"grad_norm": 3.3398014590736618,
"learning_rate": 9.309848334400247e-06,
"loss": 1.1709,
"step": 238
},
{
"epoch": 0.7635782747603834,
"grad_norm": 3.116028746474284,
"learning_rate": 9.300394538053395e-06,
"loss": 1.2525,
"step": 239
},
{
"epoch": 0.7667731629392971,
"grad_norm": 3.60023633785225,
"learning_rate": 9.29088129960862e-06,
"loss": 1.1368,
"step": 240
},
{
"epoch": 0.7699680511182109,
"grad_norm": 5.513276961177674,
"learning_rate": 9.281308750562426e-06,
"loss": 1.2557,
"step": 241
},
{
"epoch": 0.7731629392971247,
"grad_norm": 3.2648847987599723,
"learning_rate": 9.271677023231137e-06,
"loss": 1.238,
"step": 242
},
{
"epoch": 0.7763578274760383,
"grad_norm": 3.4041913237763737,
"learning_rate": 9.261986250749068e-06,
"loss": 1.159,
"step": 243
},
{
"epoch": 0.7795527156549521,
"grad_norm": 3.640892138177295,
"learning_rate": 9.252236567066686e-06,
"loss": 0.9555,
"step": 244
},
{
"epoch": 0.7827476038338658,
"grad_norm": 3.688359705830294,
"learning_rate": 9.242428106948748e-06,
"loss": 1.1076,
"step": 245
},
{
"epoch": 0.7859424920127795,
"grad_norm": 3.3621106043202285,
"learning_rate": 9.23256100597246e-06,
"loss": 1.0993,
"step": 246
},
{
"epoch": 0.7891373801916933,
"grad_norm": 3.57001474735537,
"learning_rate": 9.22263540052558e-06,
"loss": 1.1857,
"step": 247
},
{
"epoch": 0.792332268370607,
"grad_norm": 3.9410270101558074,
"learning_rate": 9.212651427804544e-06,
"loss": 1.1472,
"step": 248
},
{
"epoch": 0.7955271565495208,
"grad_norm": 4.2220039831416845,
"learning_rate": 9.202609225812572e-06,
"loss": 1.2284,
"step": 249
},
{
"epoch": 0.7987220447284346,
"grad_norm": 3.7979110924096804,
"learning_rate": 9.192508933357753e-06,
"loss": 1.0237,
"step": 250
},
{
"epoch": 0.8019169329073482,
"grad_norm": 3.714835273702889,
"learning_rate": 9.182350690051134e-06,
"loss": 1.1669,
"step": 251
},
{
"epoch": 0.805111821086262,
"grad_norm": 3.808217164775829,
"learning_rate": 9.172134636304783e-06,
"loss": 1.0839,
"step": 252
},
{
"epoch": 0.8083067092651757,
"grad_norm": 3.192933046465159,
"learning_rate": 9.16186091332985e-06,
"loss": 1.0545,
"step": 253
},
{
"epoch": 0.8115015974440895,
"grad_norm": 4.453670341813729,
"learning_rate": 9.15152966313462e-06,
"loss": 1.2245,
"step": 254
},
{
"epoch": 0.8146964856230032,
"grad_norm": 3.3575082830430896,
"learning_rate": 9.141141028522544e-06,
"loss": 1.1196,
"step": 255
},
{
"epoch": 0.8178913738019169,
"grad_norm": 3.4087716887081263,
"learning_rate": 9.130695153090272e-06,
"loss": 1.3451,
"step": 256
},
{
"epoch": 0.8210862619808307,
"grad_norm": 3.9062552919953135,
"learning_rate": 9.120192181225658e-06,
"loss": 1.1748,
"step": 257
},
{
"epoch": 0.8242811501597445,
"grad_norm": 3.55554158195335,
"learning_rate": 9.109632258105771e-06,
"loss": 1.2463,
"step": 258
},
{
"epoch": 0.8274760383386581,
"grad_norm": 3.0442803513766488,
"learning_rate": 9.099015529694894e-06,
"loss": 1.0884,
"step": 259
},
{
"epoch": 0.8306709265175719,
"grad_norm": 3.6997473431712184,
"learning_rate": 9.088342142742493e-06,
"loss": 1.2503,
"step": 260
},
{
"epoch": 0.8338658146964856,
"grad_norm": 4.021910537526358,
"learning_rate": 9.077612244781196e-06,
"loss": 1.0321,
"step": 261
},
{
"epoch": 0.8370607028753994,
"grad_norm": 3.786462152159184,
"learning_rate": 9.066825984124751e-06,
"loss": 1.1915,
"step": 262
},
{
"epoch": 0.8402555910543131,
"grad_norm": 3.657974901373734,
"learning_rate": 9.055983509865988e-06,
"loss": 1.1415,
"step": 263
},
{
"epoch": 0.8434504792332268,
"grad_norm": 3.942986772647569,
"learning_rate": 9.045084971874738e-06,
"loss": 1.1898,
"step": 264
},
{
"epoch": 0.8466453674121406,
"grad_norm": 3.028316721051469,
"learning_rate": 9.034130520795774e-06,
"loss": 0.9851,
"step": 265
},
{
"epoch": 0.8498402555910544,
"grad_norm": 3.6038066471981174,
"learning_rate": 9.023120308046726e-06,
"loss": 1.1966,
"step": 266
},
{
"epoch": 0.853035143769968,
"grad_norm": 3.6905606769008013,
"learning_rate": 9.012054485815995e-06,
"loss": 1.0909,
"step": 267
},
{
"epoch": 0.8562300319488818,
"grad_norm": 4.47724324555781,
"learning_rate": 9.00093320706063e-06,
"loss": 1.2607,
"step": 268
},
{
"epoch": 0.8594249201277955,
"grad_norm": 3.646456100746613,
"learning_rate": 8.989756625504237e-06,
"loss": 1.1078,
"step": 269
},
{
"epoch": 0.8626198083067093,
"grad_norm": 3.310665306023114,
"learning_rate": 8.978524895634842e-06,
"loss": 1.1272,
"step": 270
},
{
"epoch": 0.865814696485623,
"grad_norm": 3.561023592087352,
"learning_rate": 8.967238172702754e-06,
"loss": 1.3095,
"step": 271
},
{
"epoch": 0.8690095846645367,
"grad_norm": 3.3280650326220758,
"learning_rate": 8.95589661271842e-06,
"loss": 1.1215,
"step": 272
},
{
"epoch": 0.8722044728434505,
"grad_norm": 3.4136274546766847,
"learning_rate": 8.94450037245028e-06,
"loss": 1.2049,
"step": 273
},
{
"epoch": 0.8753993610223643,
"grad_norm": 4.572011773724013,
"learning_rate": 8.933049609422582e-06,
"loss": 1.1008,
"step": 274
},
{
"epoch": 0.8785942492012779,
"grad_norm": 3.2974165036760907,
"learning_rate": 8.921544481913218e-06,
"loss": 1.0908,
"step": 275
},
{
"epoch": 0.8817891373801917,
"grad_norm": 3.3430673564203235,
"learning_rate": 8.909985148951528e-06,
"loss": 1.0997,
"step": 276
},
{
"epoch": 0.8849840255591054,
"grad_norm": 3.995951989617882,
"learning_rate": 8.898371770316113e-06,
"loss": 1.3316,
"step": 277
},
{
"epoch": 0.8881789137380192,
"grad_norm": 3.8210610924886783,
"learning_rate": 8.886704506532611e-06,
"loss": 1.2254,
"step": 278
},
{
"epoch": 0.8913738019169329,
"grad_norm": 3.4810955317090353,
"learning_rate": 8.874983518871488e-06,
"loss": 1.1858,
"step": 279
},
{
"epoch": 0.8945686900958466,
"grad_norm": 3.7506112066819868,
"learning_rate": 8.86320896934581e-06,
"loss": 1.3234,
"step": 280
},
{
"epoch": 0.8977635782747604,
"grad_norm": 3.6609641081369833,
"learning_rate": 8.851381020709e-06,
"loss": 1.0678,
"step": 281
},
{
"epoch": 0.9009584664536742,
"grad_norm": 3.6692532818654193,
"learning_rate": 8.839499836452584e-06,
"loss": 1.0867,
"step": 282
},
{
"epoch": 0.9041533546325878,
"grad_norm": 4.25557108219826,
"learning_rate": 8.827565580803944e-06,
"loss": 1.0536,
"step": 283
},
{
"epoch": 0.9073482428115016,
"grad_norm": 3.464709889072819,
"learning_rate": 8.815578418724031e-06,
"loss": 1.2374,
"step": 284
},
{
"epoch": 0.9105431309904153,
"grad_norm": 3.541905542899119,
"learning_rate": 8.803538515905102e-06,
"loss": 1.1178,
"step": 285
},
{
"epoch": 0.9137380191693291,
"grad_norm": 3.9416940805882112,
"learning_rate": 8.791446038768416e-06,
"loss": 1.0639,
"step": 286
},
{
"epoch": 0.9169329073482428,
"grad_norm": 3.509798193817301,
"learning_rate": 8.779301154461945e-06,
"loss": 1.3332,
"step": 287
},
{
"epoch": 0.9201277955271565,
"grad_norm": 3.9728766261893185,
"learning_rate": 8.76710403085805e-06,
"loss": 1.2374,
"step": 288
},
{
"epoch": 0.9233226837060703,
"grad_norm": 3.3124789910503227,
"learning_rate": 8.754854836551174e-06,
"loss": 1.0526,
"step": 289
},
{
"epoch": 0.9265175718849841,
"grad_norm": 3.814972404250732,
"learning_rate": 8.742553740855507e-06,
"loss": 1.1395,
"step": 290
},
{
"epoch": 0.9297124600638977,
"grad_norm": 3.585116035940599,
"learning_rate": 8.730200913802638e-06,
"loss": 1.1061,
"step": 291
},
{
"epoch": 0.9329073482428115,
"grad_norm": 3.422733121363084,
"learning_rate": 8.717796526139218e-06,
"loss": 1.1189,
"step": 292
},
{
"epoch": 0.9361022364217252,
"grad_norm": 3.3415323343102634,
"learning_rate": 8.70534074932459e-06,
"loss": 1.2979,
"step": 293
},
{
"epoch": 0.939297124600639,
"grad_norm": 3.5883094898645043,
"learning_rate": 8.692833755528426e-06,
"loss": 1.2265,
"step": 294
},
{
"epoch": 0.9424920127795527,
"grad_norm": 4.100035440283119,
"learning_rate": 8.680275717628336e-06,
"loss": 1.0045,
"step": 295
},
{
"epoch": 0.9456869009584664,
"grad_norm": 3.7507898438854546,
"learning_rate": 8.667666809207495e-06,
"loss": 1.0742,
"step": 296
},
{
"epoch": 0.9488817891373802,
"grad_norm": 3.662200816366686,
"learning_rate": 8.655007204552228e-06,
"loss": 1.0408,
"step": 297
},
{
"epoch": 0.952076677316294,
"grad_norm": 3.5060679129969055,
"learning_rate": 8.64229707864961e-06,
"loss": 1.202,
"step": 298
},
{
"epoch": 0.9552715654952076,
"grad_norm": 3.349715850266676,
"learning_rate": 8.629536607185042e-06,
"loss": 1.0384,
"step": 299
},
{
"epoch": 0.9584664536741214,
"grad_norm": 3.4584142619897955,
"learning_rate": 8.616725966539831e-06,
"loss": 1.1535,
"step": 300
},
{
"epoch": 0.9616613418530351,
"grad_norm": 3.74586963964992,
"learning_rate": 8.60386533378874e-06,
"loss": 1.2205,
"step": 301
},
{
"epoch": 0.9648562300319489,
"grad_norm": 3.798760970567799,
"learning_rate": 8.590954886697554e-06,
"loss": 1.1621,
"step": 302
},
{
"epoch": 0.9680511182108626,
"grad_norm": 3.5959894698449535,
"learning_rate": 8.577994803720605e-06,
"loss": 1.2921,
"step": 303
},
{
"epoch": 0.9712460063897763,
"grad_norm": 3.6229399841896273,
"learning_rate": 8.564985263998327e-06,
"loss": 1.4373,
"step": 304
},
{
"epoch": 0.9744408945686901,
"grad_norm": 3.7669298042082255,
"learning_rate": 8.551926447354759e-06,
"loss": 1.0041,
"step": 305
},
{
"epoch": 0.9776357827476039,
"grad_norm": 4.052972820009292,
"learning_rate": 8.538818534295076e-06,
"loss": 1.1434,
"step": 306
},
{
"epoch": 0.9808306709265175,
"grad_norm": 3.614315651818945,
"learning_rate": 8.525661706003083e-06,
"loss": 1.3165,
"step": 307
},
{
"epoch": 0.9840255591054313,
"grad_norm": 3.09873915860757,
"learning_rate": 8.512456144338717e-06,
"loss": 1.2425,
"step": 308
},
{
"epoch": 0.987220447284345,
"grad_norm": 3.612487879702219,
"learning_rate": 8.499202031835532e-06,
"loss": 1.2063,
"step": 309
},
{
"epoch": 0.9904153354632588,
"grad_norm": 3.9250174100537794,
"learning_rate": 8.485899551698166e-06,
"loss": 1.0278,
"step": 310
},
{
"epoch": 0.9936102236421726,
"grad_norm": 3.3482267670475214,
"learning_rate": 8.472548887799833e-06,
"loss": 0.8995,
"step": 311
},
{
"epoch": 0.9968051118210862,
"grad_norm": 3.922256788073049,
"learning_rate": 8.45915022467975e-06,
"loss": 1.1298,
"step": 312
},
{
"epoch": 1.0,
"grad_norm": 3.6705185240927864,
"learning_rate": 8.445703747540614e-06,
"loss": 1.1195,
"step": 313
},
{
"epoch": 1.0031948881789137,
"grad_norm": 3.37800423947786,
"learning_rate": 8.43220964224602e-06,
"loss": 0.799,
"step": 314
},
{
"epoch": 1.0063897763578276,
"grad_norm": 3.212406540500856,
"learning_rate": 8.418668095317912e-06,
"loss": 0.7304,
"step": 315
},
{
"epoch": 1.0095846645367412,
"grad_norm": 3.388001090526155,
"learning_rate": 8.405079293933986e-06,
"loss": 0.6255,
"step": 316
},
{
"epoch": 1.012779552715655,
"grad_norm": 3.2573938092549235,
"learning_rate": 8.391443425925118e-06,
"loss": 0.8261,
"step": 317
},
{
"epoch": 1.0159744408945688,
"grad_norm": 3.0570537125867805,
"learning_rate": 8.37776067977276e-06,
"loss": 0.6087,
"step": 318
},
{
"epoch": 1.0191693290734825,
"grad_norm": 2.660703613808608,
"learning_rate": 8.36403124460633e-06,
"loss": 0.7153,
"step": 319
},
{
"epoch": 1.0223642172523961,
"grad_norm": 3.4986447606947157,
"learning_rate": 8.350255310200611e-06,
"loss": 0.7353,
"step": 320
},
{
"epoch": 1.0255591054313098,
"grad_norm": 2.9649416954308796,
"learning_rate": 8.336433066973122e-06,
"loss": 0.772,
"step": 321
},
{
"epoch": 1.0287539936102237,
"grad_norm": 2.9785125937278623,
"learning_rate": 8.322564705981476e-06,
"loss": 0.6936,
"step": 322
},
{
"epoch": 1.0319488817891374,
"grad_norm": 2.832349618792789,
"learning_rate": 8.308650418920751e-06,
"loss": 0.7111,
"step": 323
},
{
"epoch": 1.035143769968051,
"grad_norm": 3.1773252866733634,
"learning_rate": 8.294690398120843e-06,
"loss": 0.9085,
"step": 324
},
{
"epoch": 1.038338658146965,
"grad_norm": 3.384601614777411,
"learning_rate": 8.280684836543794e-06,
"loss": 0.8286,
"step": 325
},
{
"epoch": 1.0415335463258786,
"grad_norm": 2.8233810893434663,
"learning_rate": 8.266633927781135e-06,
"loss": 0.7082,
"step": 326
},
{
"epoch": 1.0447284345047922,
"grad_norm": 3.473937910532895,
"learning_rate": 8.25253786605121e-06,
"loss": 0.6933,
"step": 327
},
{
"epoch": 1.0479233226837061,
"grad_norm": 2.605844154701731,
"learning_rate": 8.238396846196483e-06,
"loss": 0.5575,
"step": 328
},
{
"epoch": 1.0511182108626198,
"grad_norm": 3.367592955967364,
"learning_rate": 8.224211063680854e-06,
"loss": 0.8976,
"step": 329
},
{
"epoch": 1.0543130990415335,
"grad_norm": 2.618624580723435,
"learning_rate": 8.209980714586955e-06,
"loss": 0.6622,
"step": 330
},
{
"epoch": 1.0575079872204474,
"grad_norm": 3.3282730809242174,
"learning_rate": 8.195705995613436e-06,
"loss": 0.775,
"step": 331
},
{
"epoch": 1.060702875399361,
"grad_norm": 2.8543476952348295,
"learning_rate": 8.181387104072252e-06,
"loss": 0.7296,
"step": 332
},
{
"epoch": 1.0638977635782747,
"grad_norm": 3.4077509167386633,
"learning_rate": 8.167024237885927e-06,
"loss": 0.6739,
"step": 333
},
{
"epoch": 1.0670926517571886,
"grad_norm": 2.9138463323910173,
"learning_rate": 8.152617595584827e-06,
"loss": 0.654,
"step": 334
},
{
"epoch": 1.0702875399361023,
"grad_norm": 4.330239506372152,
"learning_rate": 8.138167376304411e-06,
"loss": 0.5302,
"step": 335
},
{
"epoch": 1.073482428115016,
"grad_norm": 2.6769843205215897,
"learning_rate": 8.123673779782481e-06,
"loss": 0.6993,
"step": 336
},
{
"epoch": 1.0766773162939298,
"grad_norm": 3.2486043064592267,
"learning_rate": 8.10913700635642e-06,
"loss": 0.6562,
"step": 337
},
{
"epoch": 1.0798722044728435,
"grad_norm": 3.224653666219537,
"learning_rate": 8.094557256960419e-06,
"loss": 0.6434,
"step": 338
},
{
"epoch": 1.0830670926517572,
"grad_norm": 3.8327406854946724,
"learning_rate": 8.079934733122708e-06,
"loss": 0.5923,
"step": 339
},
{
"epoch": 1.0862619808306708,
"grad_norm": 3.017268108535111,
"learning_rate": 8.065269636962765e-06,
"loss": 0.6637,
"step": 340
},
{
"epoch": 1.0894568690095847,
"grad_norm": 3.775307256636414,
"learning_rate": 8.05056217118852e-06,
"loss": 0.6415,
"step": 341
},
{
"epoch": 1.0926517571884984,
"grad_norm": 3.4666885486639396,
"learning_rate": 8.035812539093557e-06,
"loss": 0.8204,
"step": 342
},
{
"epoch": 1.095846645367412,
"grad_norm": 3.3745035452866516,
"learning_rate": 8.021020944554305e-06,
"loss": 0.5468,
"step": 343
},
{
"epoch": 1.099041533546326,
"grad_norm": 2.837194328423608,
"learning_rate": 8.006187592027215e-06,
"loss": 0.6487,
"step": 344
},
{
"epoch": 1.1022364217252396,
"grad_norm": 3.364130012300777,
"learning_rate": 7.991312686545939e-06,
"loss": 0.7617,
"step": 345
},
{
"epoch": 1.1054313099041533,
"grad_norm": 3.4911251359165867,
"learning_rate": 7.976396433718492e-06,
"loss": 0.7319,
"step": 346
},
{
"epoch": 1.1086261980830672,
"grad_norm": 4.375337817676375,
"learning_rate": 7.961439039724413e-06,
"loss": 0.6689,
"step": 347
},
{
"epoch": 1.1118210862619808,
"grad_norm": 3.6556787392434114,
"learning_rate": 7.946440711311913e-06,
"loss": 0.5605,
"step": 348
},
{
"epoch": 1.1150159744408945,
"grad_norm": 2.864980886437467,
"learning_rate": 7.931401655795021e-06,
"loss": 0.691,
"step": 349
},
{
"epoch": 1.1182108626198084,
"grad_norm": 3.3825072087226387,
"learning_rate": 7.916322081050708e-06,
"loss": 0.6932,
"step": 350
},
{
"epoch": 1.121405750798722,
"grad_norm": 3.1541514105151247,
"learning_rate": 7.90120219551603e-06,
"loss": 0.814,
"step": 351
},
{
"epoch": 1.1246006389776357,
"grad_norm": 2.8693086928172282,
"learning_rate": 7.88604220818523e-06,
"loss": 0.6133,
"step": 352
},
{
"epoch": 1.1277955271565494,
"grad_norm": 3.250404751569032,
"learning_rate": 7.870842328606863e-06,
"loss": 0.5391,
"step": 353
},
{
"epoch": 1.1309904153354633,
"grad_norm": 3.8873422245808817,
"learning_rate": 7.85560276688089e-06,
"loss": 0.7031,
"step": 354
},
{
"epoch": 1.134185303514377,
"grad_norm": 3.114093582905189,
"learning_rate": 7.84032373365578e-06,
"loss": 0.7503,
"step": 355
},
{
"epoch": 1.1373801916932909,
"grad_norm": 3.6314953200050417,
"learning_rate": 7.825005440125595e-06,
"loss": 0.8248,
"step": 356
},
{
"epoch": 1.1405750798722045,
"grad_norm": 2.934792096451637,
"learning_rate": 7.809648098027067e-06,
"loss": 0.5856,
"step": 357
},
{
"epoch": 1.1437699680511182,
"grad_norm": 3.137125534207016,
"learning_rate": 7.794251919636687e-06,
"loss": 0.6994,
"step": 358
},
{
"epoch": 1.1469648562300319,
"grad_norm": 2.7879778074061012,
"learning_rate": 7.778817117767748e-06,
"loss": 0.7631,
"step": 359
},
{
"epoch": 1.1501597444089458,
"grad_norm": 4.110754213240846,
"learning_rate": 7.76334390576742e-06,
"loss": 0.782,
"step": 360
},
{
"epoch": 1.1533546325878594,
"grad_norm": 3.5185267700031773,
"learning_rate": 7.747832497513797e-06,
"loss": 0.6425,
"step": 361
},
{
"epoch": 1.156549520766773,
"grad_norm": 2.9208646762487684,
"learning_rate": 7.732283107412938e-06,
"loss": 0.6878,
"step": 362
},
{
"epoch": 1.159744408945687,
"grad_norm": 2.6283670320687964,
"learning_rate": 7.71669595039591e-06,
"loss": 0.5011,
"step": 363
},
{
"epoch": 1.1629392971246006,
"grad_norm": 3.030917023923999,
"learning_rate": 7.701071241915804e-06,
"loss": 0.4819,
"step": 364
},
{
"epoch": 1.1661341853035143,
"grad_norm": 2.8515455598610813,
"learning_rate": 7.685409197944768e-06,
"loss": 0.7462,
"step": 365
},
{
"epoch": 1.1693290734824282,
"grad_norm": 2.907142333143968,
"learning_rate": 7.669710034971025e-06,
"loss": 0.5291,
"step": 366
},
{
"epoch": 1.1725239616613419,
"grad_norm": 4.122531973399474,
"learning_rate": 7.653973969995866e-06,
"loss": 0.779,
"step": 367
},
{
"epoch": 1.1757188498402555,
"grad_norm": 2.8192324245130136,
"learning_rate": 7.638201220530664e-06,
"loss": 0.6558,
"step": 368
},
{
"epoch": 1.1789137380191694,
"grad_norm": 3.6374116093247695,
"learning_rate": 7.622392004593862e-06,
"loss": 0.7514,
"step": 369
},
{
"epoch": 1.182108626198083,
"grad_norm": 3.534285403348492,
"learning_rate": 7.60654654070796e-06,
"loss": 0.7463,
"step": 370
},
{
"epoch": 1.1853035143769968,
"grad_norm": 3.346398273077841,
"learning_rate": 7.59066504789649e-06,
"loss": 0.6526,
"step": 371
},
{
"epoch": 1.1884984025559104,
"grad_norm": 3.1005866977474326,
"learning_rate": 7.574747745681e-06,
"loss": 0.7355,
"step": 372
},
{
"epoch": 1.1916932907348243,
"grad_norm": 3.170219332918214,
"learning_rate": 7.558794854078006e-06,
"loss": 0.5801,
"step": 373
},
{
"epoch": 1.194888178913738,
"grad_norm": 3.2638286226616997,
"learning_rate": 7.542806593595963e-06,
"loss": 0.6382,
"step": 374
},
{
"epoch": 1.1980830670926517,
"grad_norm": 2.6820547296506505,
"learning_rate": 7.526783185232208e-06,
"loss": 0.6225,
"step": 375
},
{
"epoch": 1.2012779552715656,
"grad_norm": 3.25118296401286,
"learning_rate": 7.51072485046991e-06,
"loss": 0.5725,
"step": 376
},
{
"epoch": 1.2044728434504792,
"grad_norm": 2.8359293325624813,
"learning_rate": 7.494631811275008e-06,
"loss": 0.6745,
"step": 377
},
{
"epoch": 1.207667731629393,
"grad_norm": 2.799926492426499,
"learning_rate": 7.478504290093138e-06,
"loss": 0.6645,
"step": 378
},
{
"epoch": 1.2108626198083068,
"grad_norm": 3.220204026362489,
"learning_rate": 7.462342509846571e-06,
"loss": 0.6672,
"step": 379
},
{
"epoch": 1.2140575079872205,
"grad_norm": 3.137984017089317,
"learning_rate": 7.446146693931111e-06,
"loss": 0.5594,
"step": 380
},
{
"epoch": 1.2172523961661341,
"grad_norm": 3.5411954419366465,
"learning_rate": 7.42991706621303e-06,
"loss": 0.7462,
"step": 381
},
{
"epoch": 1.220447284345048,
"grad_norm": 3.3649992836206803,
"learning_rate": 7.413653851025959e-06,
"loss": 0.7693,
"step": 382
},
{
"epoch": 1.2236421725239617,
"grad_norm": 3.8186909851442508,
"learning_rate": 7.397357273167789e-06,
"loss": 0.7012,
"step": 383
},
{
"epoch": 1.2268370607028753,
"grad_norm": 3.052803849491951,
"learning_rate": 7.381027557897568e-06,
"loss": 0.6829,
"step": 384
},
{
"epoch": 1.230031948881789,
"grad_norm": 3.0520916591187284,
"learning_rate": 7.364664930932385e-06,
"loss": 0.869,
"step": 385
},
{
"epoch": 1.233226837060703,
"grad_norm": 3.3324108688597884,
"learning_rate": 7.348269618444248e-06,
"loss": 0.6287,
"step": 386
},
{
"epoch": 1.2364217252396166,
"grad_norm": 3.1004465014782534,
"learning_rate": 7.331841847056962e-06,
"loss": 0.5682,
"step": 387
},
{
"epoch": 1.2396166134185305,
"grad_norm": 3.499677944338085,
"learning_rate": 7.315381843842995e-06,
"loss": 0.6721,
"step": 388
},
{
"epoch": 1.2428115015974441,
"grad_norm": 3.274717064898877,
"learning_rate": 7.298889836320334e-06,
"loss": 0.7794,
"step": 389
},
{
"epoch": 1.2460063897763578,
"grad_norm": 3.753398798386701,
"learning_rate": 7.282366052449351e-06,
"loss": 0.7136,
"step": 390
},
{
"epoch": 1.2492012779552715,
"grad_norm": 3.2542525821089954,
"learning_rate": 7.265810720629643e-06,
"loss": 0.6677,
"step": 391
},
{
"epoch": 1.2523961661341854,
"grad_norm": 3.719030879521475,
"learning_rate": 7.249224069696876e-06,
"loss": 0.7505,
"step": 392
},
{
"epoch": 1.255591054313099,
"grad_norm": 3.110454563622499,
"learning_rate": 7.232606328919627e-06,
"loss": 0.6129,
"step": 393
},
{
"epoch": 1.2587859424920127,
"grad_norm": 3.3444725850201507,
"learning_rate": 7.215957727996208e-06,
"loss": 0.6196,
"step": 394
},
{
"epoch": 1.2619808306709266,
"grad_norm": 3.9648381266714052,
"learning_rate": 7.199278497051498e-06,
"loss": 0.5943,
"step": 395
},
{
"epoch": 1.2651757188498403,
"grad_norm": 3.4027694329702998,
"learning_rate": 7.182568866633757e-06,
"loss": 0.6036,
"step": 396
},
{
"epoch": 1.268370607028754,
"grad_norm": 3.0106526924818753,
"learning_rate": 7.16582906771144e-06,
"loss": 0.4707,
"step": 397
},
{
"epoch": 1.2715654952076676,
"grad_norm": 3.277499198410177,
"learning_rate": 7.149059331670009e-06,
"loss": 0.8761,
"step": 398
},
{
"epoch": 1.2747603833865815,
"grad_norm": 3.1283446610436854,
"learning_rate": 7.132259890308726e-06,
"loss": 0.6207,
"step": 399
},
{
"epoch": 1.2779552715654952,
"grad_norm": 2.8792438849397537,
"learning_rate": 7.115430975837457e-06,
"loss": 0.599,
"step": 400
},
{
"epoch": 1.281150159744409,
"grad_norm": 2.801621836848967,
"learning_rate": 7.098572820873461e-06,
"loss": 0.6326,
"step": 401
},
{
"epoch": 1.2843450479233227,
"grad_norm": 3.8269898623442913,
"learning_rate": 7.081685658438173e-06,
"loss": 0.8722,
"step": 402
},
{
"epoch": 1.2875399361022364,
"grad_norm": 2.830953218359214,
"learning_rate": 7.064769721953975e-06,
"loss": 0.657,
"step": 403
},
{
"epoch": 1.29073482428115,
"grad_norm": 3.1532951770681925,
"learning_rate": 7.047825245240989e-06,
"loss": 0.8082,
"step": 404
},
{
"epoch": 1.293929712460064,
"grad_norm": 3.458762179716139,
"learning_rate": 7.030852462513827e-06,
"loss": 0.7999,
"step": 405
},
{
"epoch": 1.2971246006389776,
"grad_norm": 3.637520450747609,
"learning_rate": 7.013851608378359e-06,
"loss": 0.6606,
"step": 406
},
{
"epoch": 1.3003194888178915,
"grad_norm": 5.419958259914624,
"learning_rate": 6.9968229178284775e-06,
"loss": 0.761,
"step": 407
},
{
"epoch": 1.3035143769968052,
"grad_norm": 3.2596588918562297,
"learning_rate": 6.979766626242839e-06,
"loss": 0.7361,
"step": 408
},
{
"epoch": 1.3067092651757188,
"grad_norm": 2.937660308357846,
"learning_rate": 6.9626829693816135e-06,
"loss": 0.6308,
"step": 409
},
{
"epoch": 1.3099041533546325,
"grad_norm": 3.0526286411927304,
"learning_rate": 6.945572183383229e-06,
"loss": 0.6958,
"step": 410
},
{
"epoch": 1.3130990415335464,
"grad_norm": 3.579656500307108,
"learning_rate": 6.928434504761106e-06,
"loss": 0.7308,
"step": 411
},
{
"epoch": 1.31629392971246,
"grad_norm": 2.9508324898774587,
"learning_rate": 6.911270170400385e-06,
"loss": 0.6988,
"step": 412
},
{
"epoch": 1.3194888178913737,
"grad_norm": 2.8135769490705287,
"learning_rate": 6.894079417554657e-06,
"loss": 0.6317,
"step": 413
},
{
"epoch": 1.3226837060702876,
"grad_norm": 3.1996485839569866,
"learning_rate": 6.8768624838426815e-06,
"loss": 0.5235,
"step": 414
},
{
"epoch": 1.3258785942492013,
"grad_norm": 3.1927676636354696,
"learning_rate": 6.859619607245102e-06,
"loss": 0.5841,
"step": 415
},
{
"epoch": 1.329073482428115,
"grad_norm": 3.1955296165750138,
"learning_rate": 6.842351026101155e-06,
"loss": 0.6608,
"step": 416
},
{
"epoch": 1.3322683706070286,
"grad_norm": 3.631363785439693,
"learning_rate": 6.825056979105382e-06,
"loss": 0.7076,
"step": 417
},
{
"epoch": 1.3354632587859425,
"grad_norm": 3.423736149201848,
"learning_rate": 6.807737705304324e-06,
"loss": 0.5832,
"step": 418
},
{
"epoch": 1.3386581469648562,
"grad_norm": 2.8520847673519674,
"learning_rate": 6.790393444093214e-06,
"loss": 0.7199,
"step": 419
},
{
"epoch": 1.34185303514377,
"grad_norm": 3.6686233711673086,
"learning_rate": 6.773024435212678e-06,
"loss": 0.7035,
"step": 420
},
{
"epoch": 1.3450479233226837,
"grad_norm": 3.1727394814071648,
"learning_rate": 6.7556309187454185e-06,
"loss": 0.7566,
"step": 421
},
{
"epoch": 1.3482428115015974,
"grad_norm": 3.187825620483754,
"learning_rate": 6.738213135112884e-06,
"loss": 0.7118,
"step": 422
},
{
"epoch": 1.351437699680511,
"grad_norm": 3.243940323976302,
"learning_rate": 6.720771325071965e-06,
"loss": 0.582,
"step": 423
},
{
"epoch": 1.354632587859425,
"grad_norm": 3.105583902088351,
"learning_rate": 6.703305729711653e-06,
"loss": 0.6959,
"step": 424
},
{
"epoch": 1.3578274760383386,
"grad_norm": 3.1920135784649726,
"learning_rate": 6.685816590449708e-06,
"loss": 0.6322,
"step": 425
},
{
"epoch": 1.3610223642172525,
"grad_norm": 3.3124906628574506,
"learning_rate": 6.668304149029331e-06,
"loss": 0.6158,
"step": 426
},
{
"epoch": 1.3642172523961662,
"grad_norm": 2.6165875286640268,
"learning_rate": 6.650768647515813e-06,
"loss": 0.6312,
"step": 427
},
{
"epoch": 1.3674121405750799,
"grad_norm": 3.255512773637248,
"learning_rate": 6.63321032829319e-06,
"loss": 0.6626,
"step": 428
},
{
"epoch": 1.3706070287539935,
"grad_norm": 3.6540450559118356,
"learning_rate": 6.615629434060903e-06,
"loss": 0.5316,
"step": 429
},
{
"epoch": 1.3738019169329074,
"grad_norm": 3.1523042906243868,
"learning_rate": 6.598026207830428e-06,
"loss": 0.563,
"step": 430
},
{
"epoch": 1.376996805111821,
"grad_norm": 3.319448369511113,
"learning_rate": 6.5804008929219284e-06,
"loss": 0.8034,
"step": 431
},
{
"epoch": 1.3801916932907348,
"grad_norm": 3.4713856779563055,
"learning_rate": 6.562753732960887e-06,
"loss": 0.7985,
"step": 432
},
{
"epoch": 1.3833865814696487,
"grad_norm": 3.676652103812925,
"learning_rate": 6.545084971874738e-06,
"loss": 0.932,
"step": 433
},
{
"epoch": 1.3865814696485623,
"grad_norm": 2.8809072042065633,
"learning_rate": 6.527394853889499e-06,
"loss": 0.6497,
"step": 434
},
{
"epoch": 1.389776357827476,
"grad_norm": 2.8206366715569984,
"learning_rate": 6.5096836235263904e-06,
"loss": 0.634,
"step": 435
},
{
"epoch": 1.3929712460063897,
"grad_norm": 3.121808199542805,
"learning_rate": 6.491951525598461e-06,
"loss": 0.5907,
"step": 436
},
{
"epoch": 1.3961661341853036,
"grad_norm": 3.217892309857476,
"learning_rate": 6.4741988052071965e-06,
"loss": 0.7149,
"step": 437
},
{
"epoch": 1.3993610223642172,
"grad_norm": 3.035032338258565,
"learning_rate": 6.45642570773914e-06,
"loss": 0.6924,
"step": 438
},
{
"epoch": 1.4025559105431311,
"grad_norm": 3.5537920483931953,
"learning_rate": 6.438632478862495e-06,
"loss": 0.6986,
"step": 439
},
{
"epoch": 1.4057507987220448,
"grad_norm": 4.164578461909941,
"learning_rate": 6.4208193645237314e-06,
"loss": 0.7119,
"step": 440
},
{
"epoch": 1.4089456869009584,
"grad_norm": 3.254292751804667,
"learning_rate": 6.402986610944183e-06,
"loss": 0.8124,
"step": 441
},
{
"epoch": 1.4121405750798721,
"grad_norm": 3.1343863322340173,
"learning_rate": 6.385134464616649e-06,
"loss": 0.8267,
"step": 442
},
{
"epoch": 1.415335463258786,
"grad_norm": 4.0741280217586615,
"learning_rate": 6.367263172301985e-06,
"loss": 0.8676,
"step": 443
},
{
"epoch": 1.4185303514376997,
"grad_norm": 3.232544130897969,
"learning_rate": 6.3493729810256895e-06,
"loss": 0.5022,
"step": 444
},
{
"epoch": 1.4217252396166133,
"grad_norm": 3.2005251534996857,
"learning_rate": 6.331464138074493e-06,
"loss": 0.7469,
"step": 445
},
{
"epoch": 1.4249201277955272,
"grad_norm": 3.2027863463285353,
"learning_rate": 6.313536890992935e-06,
"loss": 0.6082,
"step": 446
},
{
"epoch": 1.428115015974441,
"grad_norm": 3.3671894279456733,
"learning_rate": 6.29559148757995e-06,
"loss": 0.8812,
"step": 447
},
{
"epoch": 1.4313099041533546,
"grad_norm": 3.33035901999165,
"learning_rate": 6.277628175885437e-06,
"loss": 0.6958,
"step": 448
},
{
"epoch": 1.4345047923322682,
"grad_norm": 3.0528168553285036,
"learning_rate": 6.2596472042068275e-06,
"loss": 0.6377,
"step": 449
},
{
"epoch": 1.4376996805111821,
"grad_norm": 3.2318370369006333,
"learning_rate": 6.241648821085666e-06,
"loss": 0.7325,
"step": 450
},
{
"epoch": 1.4408945686900958,
"grad_norm": 2.759896482605236,
"learning_rate": 6.223633275304157e-06,
"loss": 0.6758,
"step": 451
},
{
"epoch": 1.4440894568690097,
"grad_norm": 3.0334501888536383,
"learning_rate": 6.205600815881741e-06,
"loss": 0.7571,
"step": 452
},
{
"epoch": 1.4472843450479234,
"grad_norm": 3.959112804741712,
"learning_rate": 6.187551692071648e-06,
"loss": 0.7477,
"step": 453
},
{
"epoch": 1.450479233226837,
"grad_norm": 3.6892163219784058,
"learning_rate": 6.1694861533574445e-06,
"loss": 0.6038,
"step": 454
},
{
"epoch": 1.4536741214057507,
"grad_norm": 3.394123233214594,
"learning_rate": 6.1514044494496e-06,
"loss": 0.7368,
"step": 455
},
{
"epoch": 1.4568690095846646,
"grad_norm": 3.7811664317380673,
"learning_rate": 6.133306830282021e-06,
"loss": 0.8956,
"step": 456
},
{
"epoch": 1.4600638977635783,
"grad_norm": 3.1135366433850873,
"learning_rate": 6.115193546008602e-06,
"loss": 0.7857,
"step": 457
},
{
"epoch": 1.4632587859424921,
"grad_norm": 2.988920973015636,
"learning_rate": 6.097064846999774e-06,
"loss": 0.7737,
"step": 458
},
{
"epoch": 1.4664536741214058,
"grad_norm": 3.542502474428412,
"learning_rate": 6.078920983839032e-06,
"loss": 0.7206,
"step": 459
},
{
"epoch": 1.4696485623003195,
"grad_norm": 3.308697128228699,
"learning_rate": 6.060762207319479e-06,
"loss": 0.7527,
"step": 460
},
{
"epoch": 1.4728434504792332,
"grad_norm": 3.005994083980224,
"learning_rate": 6.042588768440358e-06,
"loss": 0.7519,
"step": 461
},
{
"epoch": 1.476038338658147,
"grad_norm": 3.7095912488248874,
"learning_rate": 6.024400918403581e-06,
"loss": 0.802,
"step": 462
},
{
"epoch": 1.4792332268370607,
"grad_norm": 2.9685893810133543,
"learning_rate": 6.006198908610261e-06,
"loss": 0.6283,
"step": 463
},
{
"epoch": 1.4824281150159744,
"grad_norm": 4.140075200796993,
"learning_rate": 5.987982990657229e-06,
"loss": 0.6303,
"step": 464
},
{
"epoch": 1.4856230031948883,
"grad_norm": 3.5802439168528455,
"learning_rate": 5.9697534163335645e-06,
"loss": 0.8805,
"step": 465
},
{
"epoch": 1.488817891373802,
"grad_norm": 3.498083341788312,
"learning_rate": 5.95151043761711e-06,
"loss": 0.6836,
"step": 466
},
{
"epoch": 1.4920127795527156,
"grad_norm": 2.929589923348224,
"learning_rate": 5.933254306670995e-06,
"loss": 0.5689,
"step": 467
},
{
"epoch": 1.4952076677316293,
"grad_norm": 3.222376957252629,
"learning_rate": 5.914985275840135e-06,
"loss": 0.7332,
"step": 468
},
{
"epoch": 1.4984025559105432,
"grad_norm": 2.713080685310734,
"learning_rate": 5.896703597647765e-06,
"loss": 0.5528,
"step": 469
},
{
"epoch": 1.5015974440894568,
"grad_norm": 3.4772831320265576,
"learning_rate": 5.878409524791931e-06,
"loss": 0.7838,
"step": 470
},
{
"epoch": 1.5047923322683707,
"grad_norm": 3.1696065235629995,
"learning_rate": 5.8601033101420055e-06,
"loss": 0.7457,
"step": 471
},
{
"epoch": 1.5079872204472844,
"grad_norm": 3.6216853032028906,
"learning_rate": 5.841785206735192e-06,
"loss": 0.7815,
"step": 472
},
{
"epoch": 1.511182108626198,
"grad_norm": 2.763716376264006,
"learning_rate": 5.823455467773027e-06,
"loss": 0.7378,
"step": 473
},
{
"epoch": 1.5143769968051117,
"grad_norm": 3.1537355907344,
"learning_rate": 5.805114346617874e-06,
"loss": 0.8431,
"step": 474
},
{
"epoch": 1.5175718849840254,
"grad_norm": 3.2794472580717335,
"learning_rate": 5.786762096789431e-06,
"loss": 0.6001,
"step": 475
},
{
"epoch": 1.5207667731629393,
"grad_norm": 3.202734090764703,
"learning_rate": 5.768398971961221e-06,
"loss": 0.6778,
"step": 476
},
{
"epoch": 1.5239616613418532,
"grad_norm": 3.249800687141456,
"learning_rate": 5.750025225957086e-06,
"loss": 0.7585,
"step": 477
},
{
"epoch": 1.5271565495207668,
"grad_norm": 3.480095829427004,
"learning_rate": 5.731641112747679e-06,
"loss": 0.4924,
"step": 478
},
{
"epoch": 1.5303514376996805,
"grad_norm": 3.2205058588350983,
"learning_rate": 5.713246886446954e-06,
"loss": 0.648,
"step": 479
},
{
"epoch": 1.5335463258785942,
"grad_norm": 3.1011323347087556,
"learning_rate": 5.694842801308651e-06,
"loss": 0.7217,
"step": 480
},
{
"epoch": 1.5367412140575079,
"grad_norm": 2.98111377761598,
"learning_rate": 5.676429111722786e-06,
"loss": 0.665,
"step": 481
},
{
"epoch": 1.5399361022364217,
"grad_norm": 3.207617145035696,
"learning_rate": 5.6580060722121325e-06,
"loss": 0.667,
"step": 482
},
{
"epoch": 1.5431309904153354,
"grad_norm": 3.268810321249256,
"learning_rate": 5.639573937428699e-06,
"loss": 0.5819,
"step": 483
},
{
"epoch": 1.5463258785942493,
"grad_norm": 3.321389847041822,
"learning_rate": 5.621132962150216e-06,
"loss": 0.6214,
"step": 484
},
{
"epoch": 1.549520766773163,
"grad_norm": 3.6635640626602197,
"learning_rate": 5.6026834012766155e-06,
"loss": 0.8364,
"step": 485
},
{
"epoch": 1.5527156549520766,
"grad_norm": 3.404055732023433,
"learning_rate": 5.584225509826497e-06,
"loss": 0.6085,
"step": 486
},
{
"epoch": 1.5559105431309903,
"grad_norm": 3.1696172780759047,
"learning_rate": 5.565759542933612e-06,
"loss": 0.636,
"step": 487
},
{
"epoch": 1.5591054313099042,
"grad_norm": 4.084417655490281,
"learning_rate": 5.547285755843334e-06,
"loss": 0.6871,
"step": 488
},
{
"epoch": 1.5623003194888179,
"grad_norm": 3.123679571026527,
"learning_rate": 5.5288044039091335e-06,
"loss": 0.7758,
"step": 489
},
{
"epoch": 1.5654952076677318,
"grad_norm": 3.049865725238824,
"learning_rate": 5.510315742589042e-06,
"loss": 0.7059,
"step": 490
},
{
"epoch": 1.5686900958466454,
"grad_norm": 3.2020865187454763,
"learning_rate": 5.491820027442126e-06,
"loss": 0.7691,
"step": 491
},
{
"epoch": 1.571884984025559,
"grad_norm": 4.173223006569756,
"learning_rate": 5.473317514124958e-06,
"loss": 0.9137,
"step": 492
},
{
"epoch": 1.5750798722044728,
"grad_norm": 3.492720294865073,
"learning_rate": 5.454808458388069e-06,
"loss": 0.6593,
"step": 493
},
{
"epoch": 1.5782747603833864,
"grad_norm": 3.245633885808069,
"learning_rate": 5.436293116072431e-06,
"loss": 0.7897,
"step": 494
},
{
"epoch": 1.5814696485623003,
"grad_norm": 3.10512035286901,
"learning_rate": 5.417771743105908e-06,
"loss": 0.6376,
"step": 495
},
{
"epoch": 1.5846645367412142,
"grad_norm": 3.022511023097933,
"learning_rate": 5.399244595499721e-06,
"loss": 0.4778,
"step": 496
},
{
"epoch": 1.5878594249201279,
"grad_norm": 3.08369208760699,
"learning_rate": 5.380711929344915e-06,
"loss": 0.6479,
"step": 497
},
{
"epoch": 1.5910543130990416,
"grad_norm": 3.122422281943651,
"learning_rate": 5.362174000808813e-06,
"loss": 0.77,
"step": 498
},
{
"epoch": 1.5942492012779552,
"grad_norm": 3.1935052075345784,
"learning_rate": 5.343631066131476e-06,
"loss": 0.5334,
"step": 499
},
{
"epoch": 1.5974440894568689,
"grad_norm": 3.864259186586219,
"learning_rate": 5.325083381622165e-06,
"loss": 0.743,
"step": 500
},
{
"epoch": 1.6006389776357828,
"grad_norm": 2.970793541818657,
"learning_rate": 5.30653120365579e-06,
"loss": 0.7806,
"step": 501
},
{
"epoch": 1.6038338658146964,
"grad_norm": 3.2472425652972317,
"learning_rate": 5.28797478866938e-06,
"loss": 0.671,
"step": 502
},
{
"epoch": 1.6070287539936103,
"grad_norm": 3.591386022267759,
"learning_rate": 5.269414393158523e-06,
"loss": 0.6927,
"step": 503
},
{
"epoch": 1.610223642172524,
"grad_norm": 3.364037571286065,
"learning_rate": 5.250850273673831e-06,
"loss": 0.6775,
"step": 504
},
{
"epoch": 1.6134185303514377,
"grad_norm": 2.9547171690141383,
"learning_rate": 5.232282686817392e-06,
"loss": 0.6299,
"step": 505
},
{
"epoch": 1.6166134185303513,
"grad_norm": 2.9981837687654593,
"learning_rate": 5.213711889239214e-06,
"loss": 0.7422,
"step": 506
},
{
"epoch": 1.619808306709265,
"grad_norm": 3.0882575965185866,
"learning_rate": 5.195138137633695e-06,
"loss": 0.5666,
"step": 507
},
{
"epoch": 1.623003194888179,
"grad_norm": 3.5254034969122463,
"learning_rate": 5.17656168873606e-06,
"loss": 0.6342,
"step": 508
},
{
"epoch": 1.6261980830670928,
"grad_norm": 3.3827005361720137,
"learning_rate": 5.157982799318817e-06,
"loss": 0.8754,
"step": 509
},
{
"epoch": 1.6293929712460065,
"grad_norm": 3.0646553196819313,
"learning_rate": 5.139401726188208e-06,
"loss": 0.6372,
"step": 510
},
{
"epoch": 1.6325878594249201,
"grad_norm": 3.188140419863299,
"learning_rate": 5.120818726180662e-06,
"loss": 0.8163,
"step": 511
},
{
"epoch": 1.6357827476038338,
"grad_norm": 4.076588371131254,
"learning_rate": 5.1022340561592396e-06,
"loss": 0.7791,
"step": 512
},
{
"epoch": 1.6389776357827475,
"grad_norm": 3.1540391888481327,
"learning_rate": 5.083647973010085e-06,
"loss": 0.6848,
"step": 513
},
{
"epoch": 1.6421725239616614,
"grad_norm": 2.9509700089924413,
"learning_rate": 5.065060733638878e-06,
"loss": 0.846,
"step": 514
},
{
"epoch": 1.645367412140575,
"grad_norm": 3.211204170864709,
"learning_rate": 5.046472594967279e-06,
"loss": 0.5942,
"step": 515
},
{
"epoch": 1.648562300319489,
"grad_norm": 3.738030932514416,
"learning_rate": 5.027883813929374e-06,
"loss": 0.9047,
"step": 516
},
{
"epoch": 1.6517571884984026,
"grad_norm": 2.6770449743530125,
"learning_rate": 5.009294647468137e-06,
"loss": 0.7523,
"step": 517
},
{
"epoch": 1.6549520766773163,
"grad_norm": 2.8271218635525415,
"learning_rate": 4.990705352531864e-06,
"loss": 0.7368,
"step": 518
},
{
"epoch": 1.65814696485623,
"grad_norm": 3.2410597870311055,
"learning_rate": 4.972116186070626e-06,
"loss": 0.7366,
"step": 519
},
{
"epoch": 1.6613418530351438,
"grad_norm": 3.588293993352811,
"learning_rate": 4.953527405032723e-06,
"loss": 0.7147,
"step": 520
},
{
"epoch": 1.6645367412140575,
"grad_norm": 3.4620869110265495,
"learning_rate": 4.934939266361123e-06,
"loss": 0.6403,
"step": 521
},
{
"epoch": 1.6677316293929714,
"grad_norm": 3.375490766149462,
"learning_rate": 4.916352026989914e-06,
"loss": 0.7337,
"step": 522
},
{
"epoch": 1.670926517571885,
"grad_norm": 3.2679827038548894,
"learning_rate": 4.897765943840761e-06,
"loss": 0.6824,
"step": 523
},
{
"epoch": 1.6741214057507987,
"grad_norm": 3.0106787602470693,
"learning_rate": 4.87918127381934e-06,
"loss": 0.6017,
"step": 524
},
{
"epoch": 1.6773162939297124,
"grad_norm": 3.3611003769428827,
"learning_rate": 4.860598273811793e-06,
"loss": 0.6617,
"step": 525
},
{
"epoch": 1.680511182108626,
"grad_norm": 3.268720860496977,
"learning_rate": 4.842017200681185e-06,
"loss": 0.7991,
"step": 526
},
{
"epoch": 1.68370607028754,
"grad_norm": 3.812672771640437,
"learning_rate": 4.823438311263943e-06,
"loss": 0.7453,
"step": 527
},
{
"epoch": 1.6869009584664538,
"grad_norm": 3.8642996024300356,
"learning_rate": 4.804861862366306e-06,
"loss": 0.879,
"step": 528
},
{
"epoch": 1.6900958466453675,
"grad_norm": 3.1988450992104314,
"learning_rate": 4.786288110760787e-06,
"loss": 0.7044,
"step": 529
},
{
"epoch": 1.6932907348242812,
"grad_norm": 3.8031112432027445,
"learning_rate": 4.767717313182611e-06,
"loss": 0.7886,
"step": 530
},
{
"epoch": 1.6964856230031948,
"grad_norm": 3.341483145951611,
"learning_rate": 4.74914972632617e-06,
"loss": 0.6797,
"step": 531
},
{
"epoch": 1.6996805111821085,
"grad_norm": 2.8441545281140512,
"learning_rate": 4.730585606841479e-06,
"loss": 0.6516,
"step": 532
},
{
"epoch": 1.7028753993610224,
"grad_norm": 3.7101898376953235,
"learning_rate": 4.7120252113306216e-06,
"loss": 0.8133,
"step": 533
},
{
"epoch": 1.706070287539936,
"grad_norm": 3.91523997432471,
"learning_rate": 4.693468796344211e-06,
"loss": 0.7918,
"step": 534
},
{
"epoch": 1.70926517571885,
"grad_norm": 3.502148526430181,
"learning_rate": 4.6749166183778375e-06,
"loss": 0.6289,
"step": 535
},
{
"epoch": 1.7124600638977636,
"grad_norm": 2.827305837536753,
"learning_rate": 4.656368933868525e-06,
"loss": 0.6762,
"step": 536
},
{
"epoch": 1.7156549520766773,
"grad_norm": 3.2561245032285684,
"learning_rate": 4.637825999191189e-06,
"loss": 0.6081,
"step": 537
},
{
"epoch": 1.718849840255591,
"grad_norm": 3.7569488909115525,
"learning_rate": 4.619288070655086e-06,
"loss": 0.9933,
"step": 538
},
{
"epoch": 1.7220447284345048,
"grad_norm": 3.272922363372765,
"learning_rate": 4.600755404500281e-06,
"loss": 0.8367,
"step": 539
},
{
"epoch": 1.7252396166134185,
"grad_norm": 3.1341475281996902,
"learning_rate": 4.582228256894093e-06,
"loss": 0.7073,
"step": 540
},
{
"epoch": 1.7284345047923324,
"grad_norm": 3.459999462905938,
"learning_rate": 4.56370688392757e-06,
"loss": 0.7816,
"step": 541
},
{
"epoch": 1.731629392971246,
"grad_norm": 3.324161912003788,
"learning_rate": 4.545191541611933e-06,
"loss": 0.7615,
"step": 542
},
{
"epoch": 1.7348242811501597,
"grad_norm": 2.908313337467647,
"learning_rate": 4.526682485875044e-06,
"loss": 0.7933,
"step": 543
},
{
"epoch": 1.7380191693290734,
"grad_norm": 2.6834807493833366,
"learning_rate": 4.508179972557875e-06,
"loss": 0.7446,
"step": 544
},
{
"epoch": 1.741214057507987,
"grad_norm": 3.266037164987975,
"learning_rate": 4.489684257410959e-06,
"loss": 0.713,
"step": 545
},
{
"epoch": 1.744408945686901,
"grad_norm": 3.494662823929772,
"learning_rate": 4.471195596090867e-06,
"loss": 0.6355,
"step": 546
},
{
"epoch": 1.7476038338658149,
"grad_norm": 3.308239242533923,
"learning_rate": 4.452714244156667e-06,
"loss": 0.5427,
"step": 547
},
{
"epoch": 1.7507987220447285,
"grad_norm": 3.4869552219762467,
"learning_rate": 4.434240457066388e-06,
"loss": 0.7435,
"step": 548
},
{
"epoch": 1.7539936102236422,
"grad_norm": 2.900916005227966,
"learning_rate": 4.415774490173504e-06,
"loss": 0.8386,
"step": 549
},
{
"epoch": 1.7571884984025559,
"grad_norm": 3.686108496575093,
"learning_rate": 4.397316598723385e-06,
"loss": 0.8248,
"step": 550
},
{
"epoch": 1.7603833865814695,
"grad_norm": 3.76988534595701,
"learning_rate": 4.3788670378497836e-06,
"loss": 0.7048,
"step": 551
},
{
"epoch": 1.7635782747603834,
"grad_norm": 3.313266253788763,
"learning_rate": 4.360426062571303e-06,
"loss": 0.4702,
"step": 552
},
{
"epoch": 1.766773162939297,
"grad_norm": 3.4621782890916912,
"learning_rate": 4.341993927787871e-06,
"loss": 0.66,
"step": 553
},
{
"epoch": 1.769968051118211,
"grad_norm": 2.980008209602484,
"learning_rate": 4.323570888277215e-06,
"loss": 0.6151,
"step": 554
},
{
"epoch": 1.7731629392971247,
"grad_norm": 3.560685184874404,
"learning_rate": 4.305157198691351e-06,
"loss": 0.6692,
"step": 555
},
{
"epoch": 1.7763578274760383,
"grad_norm": 3.152491524557221,
"learning_rate": 4.286753113553049e-06,
"loss": 0.6319,
"step": 556
},
{
"epoch": 1.779552715654952,
"grad_norm": 3.474891265388844,
"learning_rate": 4.268358887252322e-06,
"loss": 0.7154,
"step": 557
},
{
"epoch": 1.7827476038338657,
"grad_norm": 3.287156139386131,
"learning_rate": 4.249974774042915e-06,
"loss": 0.7007,
"step": 558
},
{
"epoch": 1.7859424920127795,
"grad_norm": 2.6804308554958456,
"learning_rate": 4.231601028038781e-06,
"loss": 0.7664,
"step": 559
},
{
"epoch": 1.7891373801916934,
"grad_norm": 3.9878666705024592,
"learning_rate": 4.2132379032105695e-06,
"loss": 0.7483,
"step": 560
},
{
"epoch": 1.792332268370607,
"grad_norm": 3.435273688723802,
"learning_rate": 4.194885653382128e-06,
"loss": 0.5763,
"step": 561
},
{
"epoch": 1.7955271565495208,
"grad_norm": 2.880954804782811,
"learning_rate": 4.176544532226974e-06,
"loss": 0.4328,
"step": 562
},
{
"epoch": 1.7987220447284344,
"grad_norm": 3.305202220460776,
"learning_rate": 4.158214793264808e-06,
"loss": 0.6005,
"step": 563
},
{
"epoch": 1.8019169329073481,
"grad_norm": 3.4163671397148407,
"learning_rate": 4.139896689857995e-06,
"loss": 0.7597,
"step": 564
},
{
"epoch": 1.805111821086262,
"grad_norm": 3.1184016752240735,
"learning_rate": 4.121590475208071e-06,
"loss": 0.7185,
"step": 565
},
{
"epoch": 1.8083067092651757,
"grad_norm": 3.40228150298239,
"learning_rate": 4.1032964023522366e-06,
"loss": 0.612,
"step": 566
},
{
"epoch": 1.8115015974440896,
"grad_norm": 2.7149534484020377,
"learning_rate": 4.085014724159866e-06,
"loss": 0.5856,
"step": 567
},
{
"epoch": 1.8146964856230032,
"grad_norm": 3.363267946710568,
"learning_rate": 4.066745693329008e-06,
"loss": 0.6541,
"step": 568
},
{
"epoch": 1.817891373801917,
"grad_norm": 2.8919154686224067,
"learning_rate": 4.0484895623828906e-06,
"loss": 0.5948,
"step": 569
},
{
"epoch": 1.8210862619808306,
"grad_norm": 3.187373251305753,
"learning_rate": 4.030246583666437e-06,
"loss": 0.8523,
"step": 570
},
{
"epoch": 1.8242811501597445,
"grad_norm": 2.907936951863522,
"learning_rate": 4.012017009342773e-06,
"loss": 0.6585,
"step": 571
},
{
"epoch": 1.8274760383386581,
"grad_norm": 3.2971730263745314,
"learning_rate": 3.99380109138974e-06,
"loss": 0.6991,
"step": 572
},
{
"epoch": 1.830670926517572,
"grad_norm": 3.046134502565829,
"learning_rate": 3.97559908159642e-06,
"loss": 0.526,
"step": 573
},
{
"epoch": 1.8338658146964857,
"grad_norm": 2.8350337358672317,
"learning_rate": 3.9574112315596425e-06,
"loss": 0.5328,
"step": 574
},
{
"epoch": 1.8370607028753994,
"grad_norm": 3.202962781216958,
"learning_rate": 3.9392377926805226e-06,
"loss": 0.7647,
"step": 575
},
{
"epoch": 1.840255591054313,
"grad_norm": 3.4460542834034165,
"learning_rate": 3.92107901616097e-06,
"loss": 0.6653,
"step": 576
},
{
"epoch": 1.8434504792332267,
"grad_norm": 3.6649223022451607,
"learning_rate": 3.9029351530002264e-06,
"loss": 0.6888,
"step": 577
},
{
"epoch": 1.8466453674121406,
"grad_norm": 3.0702627868266177,
"learning_rate": 3.884806453991399e-06,
"loss": 0.5281,
"step": 578
},
{
"epoch": 1.8498402555910545,
"grad_norm": 3.8601689983737892,
"learning_rate": 3.866693169717982e-06,
"loss": 0.7618,
"step": 579
},
{
"epoch": 1.8530351437699681,
"grad_norm": 4.038499051063804,
"learning_rate": 3.848595550550401e-06,
"loss": 0.7512,
"step": 580
},
{
"epoch": 1.8562300319488818,
"grad_norm": 3.35579252980495,
"learning_rate": 3.830513846642556e-06,
"loss": 0.7462,
"step": 581
},
{
"epoch": 1.8594249201277955,
"grad_norm": 3.184204379689852,
"learning_rate": 3.8124483079283546e-06,
"loss": 0.6961,
"step": 582
},
{
"epoch": 1.8626198083067091,
"grad_norm": 3.771573455436531,
"learning_rate": 3.7943991841182586e-06,
"loss": 0.7593,
"step": 583
},
{
"epoch": 1.865814696485623,
"grad_norm": 3.25056308229965,
"learning_rate": 3.7763667246958447e-06,
"loss": 0.82,
"step": 584
},
{
"epoch": 1.8690095846645367,
"grad_norm": 4.076115123769542,
"learning_rate": 3.758351178914336e-06,
"loss": 0.7649,
"step": 585
},
{
"epoch": 1.8722044728434506,
"grad_norm": 3.659353933344175,
"learning_rate": 3.7403527957931716e-06,
"loss": 0.6211,
"step": 586
},
{
"epoch": 1.8753993610223643,
"grad_norm": 3.0976675838365244,
"learning_rate": 3.7223718241145646e-06,
"loss": 0.73,
"step": 587
},
{
"epoch": 1.878594249201278,
"grad_norm": 3.2466219996437955,
"learning_rate": 3.7044085124200517e-06,
"loss": 0.6108,
"step": 588
},
{
"epoch": 1.8817891373801916,
"grad_norm": 3.4463019269852864,
"learning_rate": 3.6864631090070656e-06,
"loss": 0.6814,
"step": 589
},
{
"epoch": 1.8849840255591053,
"grad_norm": 2.8369029314283574,
"learning_rate": 3.668535861925509e-06,
"loss": 0.6031,
"step": 590
},
{
"epoch": 1.8881789137380192,
"grad_norm": 5.9616680880468875,
"learning_rate": 3.650627018974312e-06,
"loss": 0.7184,
"step": 591
},
{
"epoch": 1.891373801916933,
"grad_norm": 3.5426944987299316,
"learning_rate": 3.632736827698015e-06,
"loss": 0.7967,
"step": 592
},
{
"epoch": 1.8945686900958467,
"grad_norm": 3.393418719709665,
"learning_rate": 3.6148655353833518e-06,
"loss": 0.6664,
"step": 593
},
{
"epoch": 1.8977635782747604,
"grad_norm": 2.9281147543170993,
"learning_rate": 3.5970133890558184e-06,
"loss": 0.724,
"step": 594
},
{
"epoch": 1.900958466453674,
"grad_norm": 3.981844925357407,
"learning_rate": 3.5791806354762702e-06,
"loss": 0.6622,
"step": 595
},
{
"epoch": 1.9041533546325877,
"grad_norm": 3.289172623749105,
"learning_rate": 3.5613675211375066e-06,
"loss": 0.67,
"step": 596
},
{
"epoch": 1.9073482428115016,
"grad_norm": 2.8860637003463916,
"learning_rate": 3.5435742922608618e-06,
"loss": 0.6066,
"step": 597
},
{
"epoch": 1.9105431309904153,
"grad_norm": 3.673809916922039,
"learning_rate": 3.525801194792805e-06,
"loss": 0.7079,
"step": 598
},
{
"epoch": 1.9137380191693292,
"grad_norm": 3.402310946685531,
"learning_rate": 3.508048474401541e-06,
"loss": 0.689,
"step": 599
},
{
"epoch": 1.9169329073482428,
"grad_norm": 3.3000085243720187,
"learning_rate": 3.4903163764736104e-06,
"loss": 0.7074,
"step": 600
},
{
"epoch": 1.9201277955271565,
"grad_norm": 3.6535416969880568,
"learning_rate": 3.4726051461105016e-06,
"loss": 0.7119,
"step": 601
},
{
"epoch": 1.9233226837060702,
"grad_norm": 3.136860179313549,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.551,
"step": 602
},
{
"epoch": 1.926517571884984,
"grad_norm": 3.1128124640916703,
"learning_rate": 3.437246267039115e-06,
"loss": 0.774,
"step": 603
},
{
"epoch": 1.9297124600638977,
"grad_norm": 2.7380418682334065,
"learning_rate": 3.419599107078073e-06,
"loss": 0.5846,
"step": 604
},
{
"epoch": 1.9329073482428116,
"grad_norm": 2.5934517475912653,
"learning_rate": 3.401973792169574e-06,
"loss": 0.6846,
"step": 605
},
{
"epoch": 1.9361022364217253,
"grad_norm": 3.0327464063953995,
"learning_rate": 3.384370565939098e-06,
"loss": 0.4665,
"step": 606
},
{
"epoch": 1.939297124600639,
"grad_norm": 3.432650934816916,
"learning_rate": 3.3667896717068105e-06,
"loss": 0.705,
"step": 607
},
{
"epoch": 1.9424920127795526,
"grad_norm": 3.44925873600294,
"learning_rate": 3.34923135248419e-06,
"loss": 0.5016,
"step": 608
},
{
"epoch": 1.9456869009584663,
"grad_norm": 3.9052964319528334,
"learning_rate": 3.33169585097067e-06,
"loss": 0.541,
"step": 609
},
{
"epoch": 1.9488817891373802,
"grad_norm": 2.8646242377678606,
"learning_rate": 3.314183409550293e-06,
"loss": 0.5598,
"step": 610
},
{
"epoch": 1.952076677316294,
"grad_norm": 3.5859475707816433,
"learning_rate": 3.2966942702883494e-06,
"loss": 0.7197,
"step": 611
},
{
"epoch": 1.9552715654952078,
"grad_norm": 4.742233132469254,
"learning_rate": 3.279228674928035e-06,
"loss": 0.6554,
"step": 612
},
{
"epoch": 1.9584664536741214,
"grad_norm": 2.8140338537679286,
"learning_rate": 3.261786864887117e-06,
"loss": 0.8165,
"step": 613
},
{
"epoch": 1.961661341853035,
"grad_norm": 3.3536995396201865,
"learning_rate": 3.244369081254585e-06,
"loss": 0.7105,
"step": 614
},
{
"epoch": 1.9648562300319488,
"grad_norm": 3.324725174722957,
"learning_rate": 3.226975564787322e-06,
"loss": 0.6841,
"step": 615
},
{
"epoch": 1.9680511182108626,
"grad_norm": 3.209019311998896,
"learning_rate": 3.209606555906788e-06,
"loss": 0.6919,
"step": 616
},
{
"epoch": 1.9712460063897763,
"grad_norm": 3.0157386433949243,
"learning_rate": 3.192262294695679e-06,
"loss": 0.7635,
"step": 617
},
{
"epoch": 1.9744408945686902,
"grad_norm": 2.8386190902149426,
"learning_rate": 3.174943020894618e-06,
"loss": 0.5711,
"step": 618
},
{
"epoch": 1.9776357827476039,
"grad_norm": 3.0717503655522904,
"learning_rate": 3.1576489738988457e-06,
"loss": 0.6496,
"step": 619
},
{
"epoch": 1.9808306709265175,
"grad_norm": 2.9603001146043253,
"learning_rate": 3.140380392754901e-06,
"loss": 0.8105,
"step": 620
},
{
"epoch": 1.9840255591054312,
"grad_norm": 3.3926701466325433,
"learning_rate": 3.12313751615732e-06,
"loss": 0.6593,
"step": 621
},
{
"epoch": 1.9872204472843449,
"grad_norm": 3.003724663477104,
"learning_rate": 3.1059205824453446e-06,
"loss": 0.5949,
"step": 622
},
{
"epoch": 1.9904153354632588,
"grad_norm": 3.4594498209147853,
"learning_rate": 3.0887298295996183e-06,
"loss": 0.7674,
"step": 623
},
{
"epoch": 1.9936102236421727,
"grad_norm": 3.313854798919945,
"learning_rate": 3.0715654952388957e-06,
"loss": 0.7382,
"step": 624
},
{
"epoch": 1.9968051118210863,
"grad_norm": 3.0323469008003987,
"learning_rate": 3.054427816616773e-06,
"loss": 0.7152,
"step": 625
},
{
"epoch": 2.0,
"grad_norm": 3.0518931902824598,
"learning_rate": 3.0373170306183885e-06,
"loss": 0.5414,
"step": 626
},
{
"epoch": 2.0031948881789137,
"grad_norm": 4.018067241684659,
"learning_rate": 3.020233373757162e-06,
"loss": 0.3166,
"step": 627
},
{
"epoch": 2.0063897763578273,
"grad_norm": 3.2995121301383854,
"learning_rate": 3.0031770821715233e-06,
"loss": 0.3597,
"step": 628
},
{
"epoch": 2.009584664536741,
"grad_norm": 3.4231581081414553,
"learning_rate": 2.9861483916216404e-06,
"loss": 0.3776,
"step": 629
},
{
"epoch": 2.012779552715655,
"grad_norm": 4.10101501234919,
"learning_rate": 2.969147537486175e-06,
"loss": 0.3468,
"step": 630
},
{
"epoch": 2.015974440894569,
"grad_norm": 3.4299328281237136,
"learning_rate": 2.952174754759012e-06,
"loss": 0.3064,
"step": 631
},
{
"epoch": 2.0191693290734825,
"grad_norm": 3.4124767666280533,
"learning_rate": 2.935230278046025e-06,
"loss": 0.2649,
"step": 632
},
{
"epoch": 2.022364217252396,
"grad_norm": 3.135720598552705,
"learning_rate": 2.9183143415618297e-06,
"loss": 0.4123,
"step": 633
},
{
"epoch": 2.02555910543131,
"grad_norm": 3.02784705404137,
"learning_rate": 2.9014271791265403e-06,
"loss": 0.4112,
"step": 634
},
{
"epoch": 2.0287539936102235,
"grad_norm": 2.8162469001391646,
"learning_rate": 2.8845690241625437e-06,
"loss": 0.2923,
"step": 635
},
{
"epoch": 2.0319488817891376,
"grad_norm": 4.863422449177138,
"learning_rate": 2.867740109691277e-06,
"loss": 0.4557,
"step": 636
},
{
"epoch": 2.0351437699680512,
"grad_norm": 4.769740901546116,
"learning_rate": 2.850940668329996e-06,
"loss": 0.3597,
"step": 637
},
{
"epoch": 2.038338658146965,
"grad_norm": 3.112867677918967,
"learning_rate": 2.8341709322885624e-06,
"loss": 0.3737,
"step": 638
},
{
"epoch": 2.0415335463258786,
"grad_norm": 4.76933904307883,
"learning_rate": 2.817431133366246e-06,
"loss": 0.409,
"step": 639
},
{
"epoch": 2.0447284345047922,
"grad_norm": 3.8685981352917924,
"learning_rate": 2.800721502948506e-06,
"loss": 0.302,
"step": 640
},
{
"epoch": 2.047923322683706,
"grad_norm": 3.4566511403628453,
"learning_rate": 2.7840422720037943e-06,
"loss": 0.4972,
"step": 641
},
{
"epoch": 2.0511182108626196,
"grad_norm": 4.573695473514111,
"learning_rate": 2.767393671080376e-06,
"loss": 0.3591,
"step": 642
},
{
"epoch": 2.0543130990415337,
"grad_norm": 3.720318376171526,
"learning_rate": 2.7507759303031257e-06,
"loss": 0.3567,
"step": 643
},
{
"epoch": 2.0575079872204474,
"grad_norm": 3.1469813322662645,
"learning_rate": 2.7341892793703594e-06,
"loss": 0.3602,
"step": 644
},
{
"epoch": 2.060702875399361,
"grad_norm": 3.596022751249667,
"learning_rate": 2.7176339475506515e-06,
"loss": 0.3156,
"step": 645
},
{
"epoch": 2.0638977635782747,
"grad_norm": 3.490803916160429,
"learning_rate": 2.7011101636796677e-06,
"loss": 0.2629,
"step": 646
},
{
"epoch": 2.0670926517571884,
"grad_norm": 2.999102695773987,
"learning_rate": 2.6846181561570085e-06,
"loss": 0.4541,
"step": 647
},
{
"epoch": 2.070287539936102,
"grad_norm": 2.6551429332396235,
"learning_rate": 2.668158152943039e-06,
"loss": 0.3425,
"step": 648
},
{
"epoch": 2.073482428115016,
"grad_norm": 2.7850485727984213,
"learning_rate": 2.651730381555754e-06,
"loss": 0.3166,
"step": 649
},
{
"epoch": 2.07667731629393,
"grad_norm": 2.5545436047875114,
"learning_rate": 2.635335069067617e-06,
"loss": 0.4024,
"step": 650
},
{
"epoch": 2.0798722044728435,
"grad_norm": 2.955531877352549,
"learning_rate": 2.618972442102432e-06,
"loss": 0.4445,
"step": 651
},
{
"epoch": 2.083067092651757,
"grad_norm": 2.884242616472994,
"learning_rate": 2.602642726832212e-06,
"loss": 0.3516,
"step": 652
},
{
"epoch": 2.086261980830671,
"grad_norm": 2.9860920869551566,
"learning_rate": 2.5863461489740403e-06,
"loss": 0.2822,
"step": 653
},
{
"epoch": 2.0894568690095845,
"grad_norm": 2.7356141888664007,
"learning_rate": 2.57008293378697e-06,
"loss": 0.2609,
"step": 654
},
{
"epoch": 2.0926517571884986,
"grad_norm": 3.1787705305934932,
"learning_rate": 2.553853306068888e-06,
"loss": 0.3081,
"step": 655
},
{
"epoch": 2.0958466453674123,
"grad_norm": 2.734830858362911,
"learning_rate": 2.5376574901534303e-06,
"loss": 0.3136,
"step": 656
},
{
"epoch": 2.099041533546326,
"grad_norm": 2.9107738021481437,
"learning_rate": 2.5214957099068613e-06,
"loss": 0.3248,
"step": 657
},
{
"epoch": 2.1022364217252396,
"grad_norm": 2.8714624819080052,
"learning_rate": 2.5053681887249916e-06,
"loss": 0.2794,
"step": 658
},
{
"epoch": 2.1054313099041533,
"grad_norm": 2.949311931377633,
"learning_rate": 2.4892751495300893e-06,
"loss": 0.309,
"step": 659
},
{
"epoch": 2.108626198083067,
"grad_norm": 2.5714955804911264,
"learning_rate": 2.4732168147677927e-06,
"loss": 0.3548,
"step": 660
},
{
"epoch": 2.1118210862619806,
"grad_norm": 2.9976082256678818,
"learning_rate": 2.4571934064040364e-06,
"loss": 0.2676,
"step": 661
},
{
"epoch": 2.1150159744408947,
"grad_norm": 3.7233799070741993,
"learning_rate": 2.4412051459219945e-06,
"loss": 0.3398,
"step": 662
},
{
"epoch": 2.1182108626198084,
"grad_norm": 3.407461746421168,
"learning_rate": 2.425252254319002e-06,
"loss": 0.3044,
"step": 663
},
{
"epoch": 2.121405750798722,
"grad_norm": 3.108478276490539,
"learning_rate": 2.4093349521035105e-06,
"loss": 0.3252,
"step": 664
},
{
"epoch": 2.1246006389776357,
"grad_norm": 3.4932682714684455,
"learning_rate": 2.3934534592920416e-06,
"loss": 0.3031,
"step": 665
},
{
"epoch": 2.1277955271565494,
"grad_norm": 3.396320957671799,
"learning_rate": 2.3776079954061385e-06,
"loss": 0.2578,
"step": 666
},
{
"epoch": 2.130990415335463,
"grad_norm": 3.2832514661538865,
"learning_rate": 2.3617987794693358e-06,
"loss": 0.3221,
"step": 667
},
{
"epoch": 2.134185303514377,
"grad_norm": 3.1113593494176146,
"learning_rate": 2.3460260300041355e-06,
"loss": 0.3207,
"step": 668
},
{
"epoch": 2.137380191693291,
"grad_norm": 2.4331843936222377,
"learning_rate": 2.3302899650289773e-06,
"loss": 0.2287,
"step": 669
},
{
"epoch": 2.1405750798722045,
"grad_norm": 2.468466590306324,
"learning_rate": 2.314590802055232e-06,
"loss": 0.2888,
"step": 670
},
{
"epoch": 2.143769968051118,
"grad_norm": 3.102855943085225,
"learning_rate": 2.2989287580841985e-06,
"loss": 0.243,
"step": 671
},
{
"epoch": 2.146964856230032,
"grad_norm": 2.9617214831840557,
"learning_rate": 2.2833040496040925e-06,
"loss": 0.2299,
"step": 672
},
{
"epoch": 2.1501597444089455,
"grad_norm": 4.063147642179533,
"learning_rate": 2.267716892587062e-06,
"loss": 0.3226,
"step": 673
},
{
"epoch": 2.1533546325878596,
"grad_norm": 3.4053545823888918,
"learning_rate": 2.252167502486205e-06,
"loss": 0.4821,
"step": 674
},
{
"epoch": 2.1565495207667733,
"grad_norm": 2.694386761587779,
"learning_rate": 2.2366560942325833e-06,
"loss": 0.3136,
"step": 675
},
{
"epoch": 2.159744408945687,
"grad_norm": 2.963523789409397,
"learning_rate": 2.2211828822322547e-06,
"loss": 0.3355,
"step": 676
},
{
"epoch": 2.1629392971246006,
"grad_norm": 2.783994873171231,
"learning_rate": 2.205748080363316e-06,
"loss": 0.251,
"step": 677
},
{
"epoch": 2.1661341853035143,
"grad_norm": 3.6531408626423807,
"learning_rate": 2.190351901972935e-06,
"loss": 0.3598,
"step": 678
},
{
"epoch": 2.169329073482428,
"grad_norm": 3.180879386646897,
"learning_rate": 2.1749945598744076e-06,
"loss": 0.3197,
"step": 679
},
{
"epoch": 2.1725239616613417,
"grad_norm": 3.5851334104601853,
"learning_rate": 2.159676266344222e-06,
"loss": 0.4493,
"step": 680
},
{
"epoch": 2.1757188498402558,
"grad_norm": 2.397702487817462,
"learning_rate": 2.144397233119112e-06,
"loss": 0.3004,
"step": 681
},
{
"epoch": 2.1789137380191694,
"grad_norm": 3.017326276902077,
"learning_rate": 2.1291576713931382e-06,
"loss": 0.3833,
"step": 682
},
{
"epoch": 2.182108626198083,
"grad_norm": 2.755425946091686,
"learning_rate": 2.1139577918147715e-06,
"loss": 0.4015,
"step": 683
},
{
"epoch": 2.1853035143769968,
"grad_norm": 2.9850214216473874,
"learning_rate": 2.0987978044839707e-06,
"loss": 0.4453,
"step": 684
},
{
"epoch": 2.1884984025559104,
"grad_norm": 3.6504795916537263,
"learning_rate": 2.0836779189492925e-06,
"loss": 0.37,
"step": 685
},
{
"epoch": 2.191693290734824,
"grad_norm": 3.3557757969892217,
"learning_rate": 2.068598344204981e-06,
"loss": 0.5072,
"step": 686
},
{
"epoch": 2.194888178913738,
"grad_norm": 2.2834724292254784,
"learning_rate": 2.053559288688086e-06,
"loss": 0.2147,
"step": 687
},
{
"epoch": 2.198083067092652,
"grad_norm": 3.099127322146567,
"learning_rate": 2.0385609602755878e-06,
"loss": 0.2499,
"step": 688
},
{
"epoch": 2.2012779552715656,
"grad_norm": 3.1919656442252013,
"learning_rate": 2.02360356628151e-06,
"loss": 0.3223,
"step": 689
},
{
"epoch": 2.2044728434504792,
"grad_norm": 2.9604836553944844,
"learning_rate": 2.0086873134540626e-06,
"loss": 0.2768,
"step": 690
},
{
"epoch": 2.207667731629393,
"grad_norm": 2.8967316967421115,
"learning_rate": 1.9938124079727874e-06,
"loss": 0.3213,
"step": 691
},
{
"epoch": 2.2108626198083066,
"grad_norm": 2.5411851845006375,
"learning_rate": 1.9789790554456977e-06,
"loss": 0.2873,
"step": 692
},
{
"epoch": 2.2140575079872207,
"grad_norm": 2.5731078804595118,
"learning_rate": 1.9641874609064443e-06,
"loss": 0.2166,
"step": 693
},
{
"epoch": 2.2172523961661343,
"grad_norm": 3.220551983284203,
"learning_rate": 1.9494378288114816e-06,
"loss": 0.3747,
"step": 694
},
{
"epoch": 2.220447284345048,
"grad_norm": 2.706485713834695,
"learning_rate": 1.9347303630372373e-06,
"loss": 0.2966,
"step": 695
},
{
"epoch": 2.2236421725239617,
"grad_norm": 2.756619627326469,
"learning_rate": 1.9200652668772924e-06,
"loss": 0.3026,
"step": 696
},
{
"epoch": 2.2268370607028753,
"grad_norm": 2.6262072550520004,
"learning_rate": 1.9054427430395828e-06,
"loss": 0.2118,
"step": 697
},
{
"epoch": 2.230031948881789,
"grad_norm": 2.6622767155353957,
"learning_rate": 1.890862993643583e-06,
"loss": 0.2679,
"step": 698
},
{
"epoch": 2.2332268370607027,
"grad_norm": 3.06869248444695,
"learning_rate": 1.8763262202175204e-06,
"loss": 0.2778,
"step": 699
},
{
"epoch": 2.236421725239617,
"grad_norm": 4.604722246117854,
"learning_rate": 1.8618326236955908e-06,
"loss": 0.2859,
"step": 700
},
{
"epoch": 2.2396166134185305,
"grad_norm": 3.6097052776249265,
"learning_rate": 1.8473824044151762e-06,
"loss": 0.4508,
"step": 701
},
{
"epoch": 2.242811501597444,
"grad_norm": 2.631769460311486,
"learning_rate": 1.8329757621140748e-06,
"loss": 0.2919,
"step": 702
},
{
"epoch": 2.246006389776358,
"grad_norm": 2.9000375637496623,
"learning_rate": 1.81861289592775e-06,
"loss": 0.3565,
"step": 703
},
{
"epoch": 2.2492012779552715,
"grad_norm": 3.4121040117749093,
"learning_rate": 1.8042940043865658e-06,
"loss": 0.3371,
"step": 704
},
{
"epoch": 2.252396166134185,
"grad_norm": 2.9407868911863693,
"learning_rate": 1.7900192854130465e-06,
"loss": 0.3802,
"step": 705
},
{
"epoch": 2.255591054313099,
"grad_norm": 2.783410790821283,
"learning_rate": 1.7757889363191484e-06,
"loss": 0.3138,
"step": 706
},
{
"epoch": 2.258785942492013,
"grad_norm": 2.4734552416317364,
"learning_rate": 1.7616031538035189e-06,
"loss": 0.1405,
"step": 707
},
{
"epoch": 2.2619808306709266,
"grad_norm": 3.7533482858181784,
"learning_rate": 1.7474621339487925e-06,
"loss": 0.3993,
"step": 708
},
{
"epoch": 2.2651757188498403,
"grad_norm": 2.9538179511877938,
"learning_rate": 1.7333660722188667e-06,
"loss": 0.3017,
"step": 709
},
{
"epoch": 2.268370607028754,
"grad_norm": 2.940994484285432,
"learning_rate": 1.7193151634562071e-06,
"loss": 0.3392,
"step": 710
},
{
"epoch": 2.2715654952076676,
"grad_norm": 3.463993799417318,
"learning_rate": 1.7053096018791588e-06,
"loss": 0.3097,
"step": 711
},
{
"epoch": 2.2747603833865817,
"grad_norm": 3.0511431874294663,
"learning_rate": 1.691349581079249e-06,
"loss": 0.3028,
"step": 712
},
{
"epoch": 2.2779552715654954,
"grad_norm": 2.3176661224961563,
"learning_rate": 1.6774352940185269e-06,
"loss": 0.3125,
"step": 713
},
{
"epoch": 2.281150159744409,
"grad_norm": 3.26988435034849,
"learning_rate": 1.663566933026879e-06,
"loss": 0.333,
"step": 714
},
{
"epoch": 2.2843450479233227,
"grad_norm": 2.7413112867776137,
"learning_rate": 1.6497446897993885e-06,
"loss": 0.328,
"step": 715
},
{
"epoch": 2.2875399361022364,
"grad_norm": 2.417372041158517,
"learning_rate": 1.6359687553936714e-06,
"loss": 0.3829,
"step": 716
},
{
"epoch": 2.29073482428115,
"grad_norm": 4.187728952425172,
"learning_rate": 1.6222393202272414e-06,
"loss": 0.4602,
"step": 717
},
{
"epoch": 2.2939297124600637,
"grad_norm": 2.625299501575419,
"learning_rate": 1.6085565740748825e-06,
"loss": 0.3721,
"step": 718
},
{
"epoch": 2.297124600638978,
"grad_norm": 2.815408725842671,
"learning_rate": 1.5949207060660138e-06,
"loss": 0.2484,
"step": 719
},
{
"epoch": 2.3003194888178915,
"grad_norm": 2.5170189230921234,
"learning_rate": 1.581331904682089e-06,
"loss": 0.2311,
"step": 720
},
{
"epoch": 2.303514376996805,
"grad_norm": 2.718368866346991,
"learning_rate": 1.5677903577539806e-06,
"loss": 0.2719,
"step": 721
},
{
"epoch": 2.306709265175719,
"grad_norm": 3.0220498785865226,
"learning_rate": 1.5542962524593869e-06,
"loss": 0.3678,
"step": 722
},
{
"epoch": 2.3099041533546325,
"grad_norm": 3.416525611995376,
"learning_rate": 1.54084977532025e-06,
"loss": 0.2472,
"step": 723
},
{
"epoch": 2.313099041533546,
"grad_norm": 3.3867246086342218,
"learning_rate": 1.5274511122001684e-06,
"loss": 0.3916,
"step": 724
},
{
"epoch": 2.31629392971246,
"grad_norm": 2.995590403735797,
"learning_rate": 1.5141004483018323e-06,
"loss": 0.36,
"step": 725
},
{
"epoch": 2.319488817891374,
"grad_norm": 2.9313335606489654,
"learning_rate": 1.5007979681644696e-06,
"loss": 0.1954,
"step": 726
},
{
"epoch": 2.3226837060702876,
"grad_norm": 3.011231479223783,
"learning_rate": 1.4875438556612836e-06,
"loss": 0.4096,
"step": 727
},
{
"epoch": 2.3258785942492013,
"grad_norm": 3.3287090648863193,
"learning_rate": 1.474338293996917e-06,
"loss": 0.3368,
"step": 728
},
{
"epoch": 2.329073482428115,
"grad_norm": 2.6174433452016252,
"learning_rate": 1.4611814657049257e-06,
"loss": 0.3136,
"step": 729
},
{
"epoch": 2.3322683706070286,
"grad_norm": 3.132278374293764,
"learning_rate": 1.4480735526452427e-06,
"loss": 0.2718,
"step": 730
},
{
"epoch": 2.3354632587859427,
"grad_norm": 3.4589776435548893,
"learning_rate": 1.4350147360016743e-06,
"loss": 0.3643,
"step": 731
},
{
"epoch": 2.3386581469648564,
"grad_norm": 3.3729419852344598,
"learning_rate": 1.4220051962793952e-06,
"loss": 0.3607,
"step": 732
},
{
"epoch": 2.34185303514377,
"grad_norm": 3.5972107044310486,
"learning_rate": 1.4090451133024473e-06,
"loss": 0.3519,
"step": 733
},
{
"epoch": 2.3450479233226837,
"grad_norm": 2.7814896874597905,
"learning_rate": 1.3961346662112585e-06,
"loss": 0.327,
"step": 734
},
{
"epoch": 2.3482428115015974,
"grad_norm": 2.6065002379372584,
"learning_rate": 1.3832740334601692e-06,
"loss": 0.3197,
"step": 735
},
{
"epoch": 2.351437699680511,
"grad_norm": 3.094873499174958,
"learning_rate": 1.3704633928149575e-06,
"loss": 0.3582,
"step": 736
},
{
"epoch": 2.3546325878594248,
"grad_norm": 2.886238022213933,
"learning_rate": 1.3577029213503911e-06,
"loss": 0.3468,
"step": 737
},
{
"epoch": 2.357827476038339,
"grad_norm": 3.4761358031449885,
"learning_rate": 1.3449927954477732e-06,
"loss": 0.416,
"step": 738
},
{
"epoch": 2.3610223642172525,
"grad_norm": 2.6183934639150594,
"learning_rate": 1.3323331907925046e-06,
"loss": 0.3121,
"step": 739
},
{
"epoch": 2.364217252396166,
"grad_norm": 2.5154067747569835,
"learning_rate": 1.319724282371664e-06,
"loss": 0.2456,
"step": 740
},
{
"epoch": 2.36741214057508,
"grad_norm": 2.9974249110962092,
"learning_rate": 1.307166244471576e-06,
"loss": 0.3251,
"step": 741
},
{
"epoch": 2.3706070287539935,
"grad_norm": 2.367595142844569,
"learning_rate": 1.2946592506754097e-06,
"loss": 0.3342,
"step": 742
},
{
"epoch": 2.373801916932907,
"grad_norm": 2.9826360190143473,
"learning_rate": 1.282203473860783e-06,
"loss": 0.3789,
"step": 743
},
{
"epoch": 2.376996805111821,
"grad_norm": 2.637693696990813,
"learning_rate": 1.2697990861973635e-06,
"loss": 0.2268,
"step": 744
},
{
"epoch": 2.380191693290735,
"grad_norm": 2.430870377730051,
"learning_rate": 1.257446259144494e-06,
"loss": 0.2173,
"step": 745
},
{
"epoch": 2.3833865814696487,
"grad_norm": 2.855827562015474,
"learning_rate": 1.2451451634488264e-06,
"loss": 0.27,
"step": 746
},
{
"epoch": 2.3865814696485623,
"grad_norm": 2.542241964311071,
"learning_rate": 1.2328959691419517e-06,
"loss": 0.3735,
"step": 747
},
{
"epoch": 2.389776357827476,
"grad_norm": 2.9054934145328515,
"learning_rate": 1.2206988455380558e-06,
"loss": 0.4531,
"step": 748
},
{
"epoch": 2.3929712460063897,
"grad_norm": 3.0258569317247503,
"learning_rate": 1.2085539612315844e-06,
"loss": 0.4056,
"step": 749
},
{
"epoch": 2.3961661341853033,
"grad_norm": 2.5262532974452307,
"learning_rate": 1.1964614840949002e-06,
"loss": 0.294,
"step": 750
},
{
"epoch": 2.3993610223642174,
"grad_norm": 3.0744294255504396,
"learning_rate": 1.1844215812759708e-06,
"loss": 0.2597,
"step": 751
},
{
"epoch": 2.402555910543131,
"grad_norm": 2.487260546188676,
"learning_rate": 1.1724344191960591e-06,
"loss": 0.4637,
"step": 752
},
{
"epoch": 2.405750798722045,
"grad_norm": 2.675609262310502,
"learning_rate": 1.1605001635474183e-06,
"loss": 0.2503,
"step": 753
},
{
"epoch": 2.4089456869009584,
"grad_norm": 3.0510765463937872,
"learning_rate": 1.1486189792910024e-06,
"loss": 0.3575,
"step": 754
},
{
"epoch": 2.412140575079872,
"grad_norm": 2.758019598261454,
"learning_rate": 1.1367910306541918e-06,
"loss": 0.2761,
"step": 755
},
{
"epoch": 2.415335463258786,
"grad_norm": 3.9953750426679853,
"learning_rate": 1.1250164811285148e-06,
"loss": 0.2697,
"step": 756
},
{
"epoch": 2.4185303514377,
"grad_norm": 2.5830665024684683,
"learning_rate": 1.1132954934673911e-06,
"loss": 0.231,
"step": 757
},
{
"epoch": 2.4217252396166136,
"grad_norm": 3.7519166889704323,
"learning_rate": 1.1016282296838887e-06,
"loss": 0.3503,
"step": 758
},
{
"epoch": 2.4249201277955272,
"grad_norm": 2.1720140538958312,
"learning_rate": 1.090014851048473e-06,
"loss": 0.2282,
"step": 759
},
{
"epoch": 2.428115015974441,
"grad_norm": 2.723215788543118,
"learning_rate": 1.078455518086784e-06,
"loss": 0.2454,
"step": 760
},
{
"epoch": 2.4313099041533546,
"grad_norm": 3.9675636337969458,
"learning_rate": 1.0669503905774198e-06,
"loss": 0.2829,
"step": 761
},
{
"epoch": 2.4345047923322682,
"grad_norm": 3.431034393719208,
"learning_rate": 1.055499627549722e-06,
"loss": 0.3508,
"step": 762
},
{
"epoch": 2.437699680511182,
"grad_norm": 2.62642260750986,
"learning_rate": 1.0441033872815804e-06,
"loss": 0.2761,
"step": 763
},
{
"epoch": 2.440894568690096,
"grad_norm": 3.590850078999076,
"learning_rate": 1.0327618272972484e-06,
"loss": 0.3186,
"step": 764
},
{
"epoch": 2.4440894568690097,
"grad_norm": 3.4343573037291253,
"learning_rate": 1.0214751043651582e-06,
"loss": 0.3002,
"step": 765
},
{
"epoch": 2.4472843450479234,
"grad_norm": 3.0802440223738836,
"learning_rate": 1.010243374495763e-06,
"loss": 0.3932,
"step": 766
},
{
"epoch": 2.450479233226837,
"grad_norm": 2.585942576218794,
"learning_rate": 9.990667929393715e-07,
"loss": 0.2977,
"step": 767
},
{
"epoch": 2.4536741214057507,
"grad_norm": 2.9346121393398286,
"learning_rate": 9.879455141840067e-07,
"loss": 0.4482,
"step": 768
},
{
"epoch": 2.4568690095846644,
"grad_norm": 3.1860138881407516,
"learning_rate": 9.768796919532742e-07,
"loss": 0.2683,
"step": 769
},
{
"epoch": 2.460063897763578,
"grad_norm": 2.6572160868591714,
"learning_rate": 9.658694792042284e-07,
"loss": 0.4466,
"step": 770
},
{
"epoch": 2.463258785942492,
"grad_norm": 3.1649450840924422,
"learning_rate": 9.549150281252633e-07,
"loss": 0.3764,
"step": 771
},
{
"epoch": 2.466453674121406,
"grad_norm": 3.0697507852185213,
"learning_rate": 9.440164901340127e-07,
"loss": 0.36,
"step": 772
},
{
"epoch": 2.4696485623003195,
"grad_norm": 2.8539861649357188,
"learning_rate": 9.331740158752495e-07,
"loss": 0.2323,
"step": 773
},
{
"epoch": 2.472843450479233,
"grad_norm": 2.369740078463609,
"learning_rate": 9.223877552188065e-07,
"loss": 0.2973,
"step": 774
},
{
"epoch": 2.476038338658147,
"grad_norm": 2.3381953841998264,
"learning_rate": 9.116578572575091e-07,
"loss": 0.2938,
"step": 775
},
{
"epoch": 2.479233226837061,
"grad_norm": 3.017784356239987,
"learning_rate": 9.009844703051063e-07,
"loss": 0.3202,
"step": 776
},
{
"epoch": 2.4824281150159746,
"grad_norm": 3.5982320235537726,
"learning_rate": 8.903677418942292e-07,
"loss": 0.3561,
"step": 777
},
{
"epoch": 2.4856230031948883,
"grad_norm": 2.505443232448581,
"learning_rate": 8.79807818774343e-07,
"loss": 0.344,
"step": 778
},
{
"epoch": 2.488817891373802,
"grad_norm": 2.658189569939545,
"learning_rate": 8.693048469097293e-07,
"loss": 0.2371,
"step": 779
},
{
"epoch": 2.4920127795527156,
"grad_norm": 2.567219244463206,
"learning_rate": 8.58858971477457e-07,
"loss": 0.2563,
"step": 780
},
{
"epoch": 2.4952076677316293,
"grad_norm": 3.557533902661021,
"learning_rate": 8.484703368653812e-07,
"loss": 0.3296,
"step": 781
},
{
"epoch": 2.498402555910543,
"grad_norm": 2.6168648309504334,
"learning_rate": 8.381390866701517e-07,
"loss": 0.2818,
"step": 782
},
{
"epoch": 2.501597444089457,
"grad_norm": 2.912793055699558,
"learning_rate": 8.278653636952177e-07,
"loss": 0.3501,
"step": 783
},
{
"epoch": 2.5047923322683707,
"grad_norm": 2.887681441459192,
"learning_rate": 8.176493099488664e-07,
"loss": 0.2816,
"step": 784
},
{
"epoch": 2.5079872204472844,
"grad_norm": 2.9770638448778803,
"learning_rate": 8.074910666422475e-07,
"loss": 0.3606,
"step": 785
},
{
"epoch": 2.511182108626198,
"grad_norm": 2.6401805900066027,
"learning_rate": 7.973907741874287e-07,
"loss": 0.3024,
"step": 786
},
{
"epoch": 2.5143769968051117,
"grad_norm": 2.8189124957754523,
"learning_rate": 7.873485721954572e-07,
"loss": 0.2791,
"step": 787
},
{
"epoch": 2.5175718849840254,
"grad_norm": 3.5129637543439602,
"learning_rate": 7.773645994744222e-07,
"loss": 0.2695,
"step": 788
},
{
"epoch": 2.520766773162939,
"grad_norm": 2.8774341352846537,
"learning_rate": 7.674389940275406e-07,
"loss": 0.3719,
"step": 789
},
{
"epoch": 2.523961661341853,
"grad_norm": 2.4633663554384664,
"learning_rate": 7.575718930512516e-07,
"loss": 0.1755,
"step": 790
},
{
"epoch": 2.527156549520767,
"grad_norm": 2.7359732901485017,
"learning_rate": 7.47763432933315e-07,
"loss": 0.2788,
"step": 791
},
{
"epoch": 2.5303514376996805,
"grad_norm": 2.9657323006652483,
"learning_rate": 7.380137492509309e-07,
"loss": 0.2845,
"step": 792
},
{
"epoch": 2.533546325878594,
"grad_norm": 2.9736194397751112,
"learning_rate": 7.283229767688627e-07,
"loss": 0.2812,
"step": 793
},
{
"epoch": 2.536741214057508,
"grad_norm": 2.9826318116418644,
"learning_rate": 7.186912494375736e-07,
"loss": 0.3175,
"step": 794
},
{
"epoch": 2.539936102236422,
"grad_norm": 2.847774304408052,
"learning_rate": 7.091187003913802e-07,
"loss": 0.3365,
"step": 795
},
{
"epoch": 2.543130990415335,
"grad_norm": 2.51922039668547,
"learning_rate": 6.996054619466053e-07,
"loss": 0.3202,
"step": 796
},
{
"epoch": 2.5463258785942493,
"grad_norm": 2.8009781132245273,
"learning_rate": 6.901516655997536e-07,
"loss": 0.3112,
"step": 797
},
{
"epoch": 2.549520766773163,
"grad_norm": 2.721473315611517,
"learning_rate": 6.80757442025694e-07,
"loss": 0.1705,
"step": 798
},
{
"epoch": 2.5527156549520766,
"grad_norm": 2.7940031328916044,
"learning_rate": 6.714229210758516e-07,
"loss": 0.2786,
"step": 799
},
{
"epoch": 2.5559105431309903,
"grad_norm": 3.498939539988549,
"learning_rate": 6.621482317764105e-07,
"loss": 0.3267,
"step": 800
},
{
"epoch": 2.559105431309904,
"grad_norm": 2.6122264925136247,
"learning_rate": 6.529335023265387e-07,
"loss": 0.3636,
"step": 801
},
{
"epoch": 2.562300319488818,
"grad_norm": 3.092200799066715,
"learning_rate": 6.437788600966066e-07,
"loss": 0.2955,
"step": 802
},
{
"epoch": 2.5654952076677318,
"grad_norm": 2.971868036987963,
"learning_rate": 6.346844316264312e-07,
"loss": 0.2871,
"step": 803
},
{
"epoch": 2.5686900958466454,
"grad_norm": 2.6456671089009305,
"learning_rate": 6.256503426235277e-07,
"loss": 0.2391,
"step": 804
},
{
"epoch": 2.571884984025559,
"grad_norm": 3.475578069082645,
"learning_rate": 6.166767179613691e-07,
"loss": 0.355,
"step": 805
},
{
"epoch": 2.5750798722044728,
"grad_norm": 2.749882226514787,
"learning_rate": 6.077636816776611e-07,
"loss": 0.2539,
"step": 806
},
{
"epoch": 2.5782747603833864,
"grad_norm": 2.9833966827404748,
"learning_rate": 5.989113569726312e-07,
"loss": 0.3373,
"step": 807
},
{
"epoch": 2.5814696485623,
"grad_norm": 2.870613864126515,
"learning_rate": 5.901198662073188e-07,
"loss": 0.2726,
"step": 808
},
{
"epoch": 2.584664536741214,
"grad_norm": 3.609955370838127,
"learning_rate": 5.813893309018881e-07,
"loss": 0.3427,
"step": 809
},
{
"epoch": 2.587859424920128,
"grad_norm": 2.3844785021764405,
"learning_rate": 5.727198717339511e-07,
"loss": 0.1803,
"step": 810
},
{
"epoch": 2.5910543130990416,
"grad_norm": 3.1465403649757526,
"learning_rate": 5.641116085368931e-07,
"loss": 0.2405,
"step": 811
},
{
"epoch": 2.594249201277955,
"grad_norm": 2.569934846209212,
"learning_rate": 5.555646602982207e-07,
"loss": 0.3006,
"step": 812
},
{
"epoch": 2.597444089456869,
"grad_norm": 2.599286683697681,
"learning_rate": 5.470791451579172e-07,
"loss": 0.359,
"step": 813
},
{
"epoch": 2.600638977635783,
"grad_norm": 3.271540032015256,
"learning_rate": 5.386551804068063e-07,
"loss": 0.351,
"step": 814
},
{
"epoch": 2.6038338658146962,
"grad_norm": 2.8687489094374903,
"learning_rate": 5.302928824849335e-07,
"loss": 0.2988,
"step": 815
},
{
"epoch": 2.6070287539936103,
"grad_norm": 2.9257565634865923,
"learning_rate": 5.219923669799587e-07,
"loss": 0.2114,
"step": 816
},
{
"epoch": 2.610223642172524,
"grad_norm": 3.3654420629780475,
"learning_rate": 5.137537486255517e-07,
"loss": 0.3157,
"step": 817
},
{
"epoch": 2.6134185303514377,
"grad_norm": 3.7694516760455534,
"learning_rate": 5.055771412998122e-07,
"loss": 0.48,
"step": 818
},
{
"epoch": 2.6166134185303513,
"grad_norm": 2.6976707955356116,
"learning_rate": 4.974626580236957e-07,
"loss": 0.2468,
"step": 819
},
{
"epoch": 2.619808306709265,
"grad_norm": 3.651100971656411,
"learning_rate": 4.894104109594466e-07,
"loss": 0.3924,
"step": 820
},
{
"epoch": 2.623003194888179,
"grad_norm": 3.3478565697486227,
"learning_rate": 4.814205114090543e-07,
"loss": 0.2496,
"step": 821
},
{
"epoch": 2.626198083067093,
"grad_norm": 2.6690732604831644,
"learning_rate": 4.734930698127077e-07,
"loss": 0.2393,
"step": 822
},
{
"epoch": 2.6293929712460065,
"grad_norm": 2.6502291854870546,
"learning_rate": 4.6562819574727304e-07,
"loss": 0.1855,
"step": 823
},
{
"epoch": 2.63258785942492,
"grad_norm": 3.3774094875631406,
"learning_rate": 4.578259979247801e-07,
"loss": 0.3486,
"step": 824
},
{
"epoch": 2.635782747603834,
"grad_norm": 4.319551142114074,
"learning_rate": 4.500865841909169e-07,
"loss": 0.25,
"step": 825
},
{
"epoch": 2.6389776357827475,
"grad_norm": 3.1750507234466707,
"learning_rate": 4.4241006152353885e-07,
"loss": 0.4404,
"step": 826
},
{
"epoch": 2.642172523961661,
"grad_norm": 3.3586534242565986,
"learning_rate": 4.3479653603119287e-07,
"loss": 0.2602,
"step": 827
},
{
"epoch": 2.6453674121405752,
"grad_norm": 2.5122893506603408,
"learning_rate": 4.2724611295164755e-07,
"loss": 0.3324,
"step": 828
},
{
"epoch": 2.648562300319489,
"grad_norm": 3.6843703751019765,
"learning_rate": 4.197588966504401e-07,
"loss": 0.2901,
"step": 829
},
{
"epoch": 2.6517571884984026,
"grad_norm": 4.255069159327654,
"learning_rate": 4.123349906194357e-07,
"loss": 0.3766,
"step": 830
},
{
"epoch": 2.6549520766773163,
"grad_norm": 2.7133598522708633,
"learning_rate": 4.0497449747539217e-07,
"loss": 0.4902,
"step": 831
},
{
"epoch": 2.65814696485623,
"grad_norm": 2.9940175911104996,
"learning_rate": 3.9767751895854467e-07,
"loss": 0.2893,
"step": 832
},
{
"epoch": 2.661341853035144,
"grad_norm": 3.1177334834961687,
"learning_rate": 3.904441559312006e-07,
"loss": 0.3359,
"step": 833
},
{
"epoch": 2.6645367412140573,
"grad_norm": 3.1161220354619705,
"learning_rate": 3.8327450837634284e-07,
"loss": 0.3193,
"step": 834
},
{
"epoch": 2.6677316293929714,
"grad_norm": 2.7840704526127413,
"learning_rate": 3.7616867539624733e-07,
"loss": 0.2791,
"step": 835
},
{
"epoch": 2.670926517571885,
"grad_norm": 2.9702379695614662,
"learning_rate": 3.691267552111183e-07,
"loss": 0.3842,
"step": 836
},
{
"epoch": 2.6741214057507987,
"grad_norm": 2.758794631297366,
"learning_rate": 3.621488451577221e-07,
"loss": 0.383,
"step": 837
},
{
"epoch": 2.6773162939297124,
"grad_norm": 3.1621851299387846,
"learning_rate": 3.552350416880507e-07,
"loss": 0.323,
"step": 838
},
{
"epoch": 2.680511182108626,
"grad_norm": 2.664639636322113,
"learning_rate": 3.483854403679832e-07,
"loss": 0.265,
"step": 839
},
{
"epoch": 2.68370607028754,
"grad_norm": 3.098568213381224,
"learning_rate": 3.416001358759635e-07,
"loss": 0.3719,
"step": 840
},
{
"epoch": 2.686900958466454,
"grad_norm": 2.504591716379194,
"learning_rate": 3.3487922200169944e-07,
"loss": 0.2845,
"step": 841
},
{
"epoch": 2.6900958466453675,
"grad_norm": 2.875695040385189,
"learning_rate": 3.2822279164485494e-07,
"loss": 0.2776,
"step": 842
},
{
"epoch": 2.693290734824281,
"grad_norm": 3.2608638772030893,
"learning_rate": 3.2163093681377765e-07,
"loss": 0.4259,
"step": 843
},
{
"epoch": 2.696485623003195,
"grad_norm": 4.257484816923254,
"learning_rate": 3.151037486242181e-07,
"loss": 0.2993,
"step": 844
},
{
"epoch": 2.6996805111821085,
"grad_norm": 2.797483350721647,
"learning_rate": 3.08641317298074e-07,
"loss": 0.3329,
"step": 845
},
{
"epoch": 2.702875399361022,
"grad_norm": 2.312130288437099,
"learning_rate": 3.022437321621452e-07,
"loss": 0.2265,
"step": 846
},
{
"epoch": 2.7060702875399363,
"grad_norm": 3.146010391510561,
"learning_rate": 2.959110816468935e-07,
"loss": 0.3277,
"step": 847
},
{
"epoch": 2.70926517571885,
"grad_norm": 4.30884225073185,
"learning_rate": 2.896434532852277e-07,
"loss": 0.3595,
"step": 848
},
{
"epoch": 2.7124600638977636,
"grad_norm": 3.8804523171458833,
"learning_rate": 2.834409337112842e-07,
"loss": 0.3428,
"step": 849
},
{
"epoch": 2.7156549520766773,
"grad_norm": 3.1949144821081314,
"learning_rate": 2.7730360865923954e-07,
"loss": 0.3248,
"step": 850
},
{
"epoch": 2.718849840255591,
"grad_norm": 3.129061239292728,
"learning_rate": 2.712315629621176e-07,
"loss": 0.2865,
"step": 851
},
{
"epoch": 2.722044728434505,
"grad_norm": 2.924072515274875,
"learning_rate": 2.6522488055062076e-07,
"loss": 0.3724,
"step": 852
},
{
"epoch": 2.7252396166134183,
"grad_norm": 2.241976067536373,
"learning_rate": 2.5928364445196975e-07,
"loss": 0.2777,
"step": 853
},
{
"epoch": 2.7284345047923324,
"grad_norm": 3.1048253231227867,
"learning_rate": 2.534079367887549e-07,
"loss": 0.3616,
"step": 854
},
{
"epoch": 2.731629392971246,
"grad_norm": 2.892814852274967,
"learning_rate": 2.475978387778e-07,
"loss": 0.2713,
"step": 855
},
{
"epoch": 2.7348242811501597,
"grad_norm": 2.9703976899571813,
"learning_rate": 2.4185343072904376e-07,
"loss": 0.3921,
"step": 856
},
{
"epoch": 2.7380191693290734,
"grad_norm": 3.8218550995564806,
"learning_rate": 2.3617479204442462e-07,
"loss": 0.3447,
"step": 857
},
{
"epoch": 2.741214057507987,
"grad_norm": 2.922629450726275,
"learning_rate": 2.305620012167853e-07,
"loss": 0.2682,
"step": 858
},
{
"epoch": 2.744408945686901,
"grad_norm": 2.665354460957857,
"learning_rate": 2.2501513582879108e-07,
"loss": 0.2921,
"step": 859
},
{
"epoch": 2.747603833865815,
"grad_norm": 2.577484902850229,
"learning_rate": 2.1953427255185122e-07,
"loss": 0.2785,
"step": 860
},
{
"epoch": 2.7507987220447285,
"grad_norm": 3.4360845970357876,
"learning_rate": 2.1411948714506414e-07,
"loss": 0.4059,
"step": 861
},
{
"epoch": 2.753993610223642,
"grad_norm": 2.2212502975356005,
"learning_rate": 2.0877085445416889e-07,
"loss": 0.2315,
"step": 862
},
{
"epoch": 2.757188498402556,
"grad_norm": 3.143208207313535,
"learning_rate": 2.034884484105093e-07,
"loss": 0.273,
"step": 863
},
{
"epoch": 2.7603833865814695,
"grad_norm": 2.3684334404460725,
"learning_rate": 1.98272342030012e-07,
"loss": 0.3056,
"step": 864
},
{
"epoch": 2.763578274760383,
"grad_norm": 3.3761581614827167,
"learning_rate": 1.9312260741218114e-07,
"loss": 0.3205,
"step": 865
},
{
"epoch": 2.7667731629392973,
"grad_norm": 3.1413600066831493,
"learning_rate": 1.8803931573909584e-07,
"loss": 0.2972,
"step": 866
},
{
"epoch": 2.769968051118211,
"grad_norm": 2.649199570865497,
"learning_rate": 1.8302253727443041e-07,
"loss": 0.3368,
"step": 867
},
{
"epoch": 2.7731629392971247,
"grad_norm": 3.27637254522789,
"learning_rate": 1.7807234136248296e-07,
"loss": 0.412,
"step": 868
},
{
"epoch": 2.7763578274760383,
"grad_norm": 2.6259650405573804,
"learning_rate": 1.731887964272144e-07,
"loss": 0.28,
"step": 869
},
{
"epoch": 2.779552715654952,
"grad_norm": 2.8722729196579393,
"learning_rate": 1.6837196997130434e-07,
"loss": 0.4333,
"step": 870
},
{
"epoch": 2.7827476038338657,
"grad_norm": 3.3953663943131094,
"learning_rate": 1.6362192857521942e-07,
"loss": 0.2619,
"step": 871
},
{
"epoch": 2.7859424920127793,
"grad_norm": 2.605633307366119,
"learning_rate": 1.5893873789628812e-07,
"loss": 0.3658,
"step": 872
},
{
"epoch": 2.7891373801916934,
"grad_norm": 2.768996824426186,
"learning_rate": 1.5432246266780083e-07,
"loss": 0.3452,
"step": 873
},
{
"epoch": 2.792332268370607,
"grad_norm": 3.0057713366690755,
"learning_rate": 1.4977316669810782e-07,
"loss": 0.2186,
"step": 874
},
{
"epoch": 2.7955271565495208,
"grad_norm": 3.028794377895449,
"learning_rate": 1.4529091286973994e-07,
"loss": 0.4345,
"step": 875
},
{
"epoch": 2.7987220447284344,
"grad_norm": 3.2822428241138377,
"learning_rate": 1.4087576313854212e-07,
"loss": 0.4219,
"step": 876
},
{
"epoch": 2.801916932907348,
"grad_norm": 2.933367752904408,
"learning_rate": 1.365277785328123e-07,
"loss": 0.3129,
"step": 877
},
{
"epoch": 2.8051118210862622,
"grad_norm": 3.672776465498361,
"learning_rate": 1.3224701915246053e-07,
"loss": 0.3546,
"step": 878
},
{
"epoch": 2.8083067092651754,
"grad_norm": 2.619657875425601,
"learning_rate": 1.280335441681796e-07,
"loss": 0.2597,
"step": 879
},
{
"epoch": 2.8115015974440896,
"grad_norm": 2.6039965900627706,
"learning_rate": 1.2388741182062348e-07,
"loss": 0.277,
"step": 880
},
{
"epoch": 2.8146964856230032,
"grad_norm": 3.053141999428792,
"learning_rate": 1.198086794196035e-07,
"loss": 0.2656,
"step": 881
},
{
"epoch": 2.817891373801917,
"grad_norm": 2.7225299473833546,
"learning_rate": 1.1579740334330014e-07,
"loss": 0.3681,
"step": 882
},
{
"epoch": 2.8210862619808306,
"grad_norm": 2.848521799304417,
"learning_rate": 1.1185363903747748e-07,
"loss": 0.3353,
"step": 883
},
{
"epoch": 2.8242811501597442,
"grad_norm": 3.4094536523848795,
"learning_rate": 1.0797744101472052e-07,
"loss": 0.3722,
"step": 884
},
{
"epoch": 2.8274760383386583,
"grad_norm": 2.5070762776644506,
"learning_rate": 1.0416886285368188e-07,
"loss": 0.2167,
"step": 885
},
{
"epoch": 2.830670926517572,
"grad_norm": 2.5431157579318238,
"learning_rate": 1.0042795719833964e-07,
"loss": 0.2288,
"step": 886
},
{
"epoch": 2.8338658146964857,
"grad_norm": 3.151237618241689,
"learning_rate": 9.675477575726954e-08,
"loss": 0.4222,
"step": 887
},
{
"epoch": 2.8370607028753994,
"grad_norm": 2.899674471339342,
"learning_rate": 9.314936930293283e-08,
"loss": 0.2864,
"step": 888
},
{
"epoch": 2.840255591054313,
"grad_norm": 3.095154064912419,
"learning_rate": 8.961178767097178e-08,
"loss": 0.4139,
"step": 889
},
{
"epoch": 2.8434504792332267,
"grad_norm": 3.3800960055515312,
"learning_rate": 8.614207975952083e-08,
"loss": 0.3053,
"step": 890
},
{
"epoch": 2.8466453674121404,
"grad_norm": 2.8830001557718923,
"learning_rate": 8.274029352853264e-08,
"loss": 0.2999,
"step": 891
},
{
"epoch": 2.8498402555910545,
"grad_norm": 2.636128099104962,
"learning_rate": 7.940647599911477e-08,
"loss": 0.2676,
"step": 892
},
{
"epoch": 2.853035143769968,
"grad_norm": 2.8286496678696884,
"learning_rate": 7.614067325287632e-08,
"loss": 0.389,
"step": 893
},
{
"epoch": 2.856230031948882,
"grad_norm": 3.6429247825702467,
"learning_rate": 7.294293043129785e-08,
"loss": 0.3162,
"step": 894
},
{
"epoch": 2.8594249201277955,
"grad_norm": 3.0374792728785334,
"learning_rate": 6.981329173509909e-08,
"loss": 0.3574,
"step": 895
},
{
"epoch": 2.862619808306709,
"grad_norm": 2.913384721791684,
"learning_rate": 6.675180042363505e-08,
"loss": 0.3179,
"step": 896
},
{
"epoch": 2.8658146964856233,
"grad_norm": 3.2917746837754085,
"learning_rate": 6.375849881429418e-08,
"loss": 0.3153,
"step": 897
},
{
"epoch": 2.8690095846645365,
"grad_norm": 2.5435846782257445,
"learning_rate": 6.083342828191453e-08,
"loss": 0.2843,
"step": 898
},
{
"epoch": 2.8722044728434506,
"grad_norm": 2.6905025881821834,
"learning_rate": 5.797662925821068e-08,
"loss": 0.3436,
"step": 899
},
{
"epoch": 2.8753993610223643,
"grad_norm": 2.932441153849384,
"learning_rate": 5.518814123121885e-08,
"loss": 0.2586,
"step": 900
},
{
"epoch": 2.878594249201278,
"grad_norm": 2.7277012813339447,
"learning_rate": 5.246800274474439e-08,
"loss": 0.2966,
"step": 901
},
{
"epoch": 2.8817891373801916,
"grad_norm": 2.7485325951177844,
"learning_rate": 4.981625139783619e-08,
"loss": 0.2542,
"step": 902
},
{
"epoch": 2.8849840255591053,
"grad_norm": 2.5810864799806836,
"learning_rate": 4.723292384426203e-08,
"loss": 0.2645,
"step": 903
},
{
"epoch": 2.8881789137380194,
"grad_norm": 2.8926196010082976,
"learning_rate": 4.471805579200239e-08,
"loss": 0.229,
"step": 904
},
{
"epoch": 2.891373801916933,
"grad_norm": 2.745451547190662,
"learning_rate": 4.227168200276077e-08,
"loss": 0.2825,
"step": 905
},
{
"epoch": 2.8945686900958467,
"grad_norm": 2.7787276082308323,
"learning_rate": 3.989383629147747e-08,
"loss": 0.3049,
"step": 906
},
{
"epoch": 2.8977635782747604,
"grad_norm": 2.3995249725648486,
"learning_rate": 3.758455152586715e-08,
"loss": 0.2789,
"step": 907
},
{
"epoch": 2.900958466453674,
"grad_norm": 2.9731570350090784,
"learning_rate": 3.534385962596143e-08,
"loss": 0.2847,
"step": 908
},
{
"epoch": 2.9041533546325877,
"grad_norm": 3.5722561710741876,
"learning_rate": 3.3171791563669785e-08,
"loss": 0.3594,
"step": 909
},
{
"epoch": 2.9073482428115014,
"grad_norm": 2.999377271575578,
"learning_rate": 3.10683773623488e-08,
"loss": 0.3422,
"step": 910
},
{
"epoch": 2.9105431309904155,
"grad_norm": 2.3390763680789908,
"learning_rate": 2.9033646096390255e-08,
"loss": 0.1714,
"step": 911
},
{
"epoch": 2.913738019169329,
"grad_norm": 2.79440551009593,
"learning_rate": 2.706762589081646e-08,
"loss": 0.2889,
"step": 912
},
{
"epoch": 2.916932907348243,
"grad_norm": 2.8334198141616898,
"learning_rate": 2.517034392089446e-08,
"loss": 0.304,
"step": 913
},
{
"epoch": 2.9201277955271565,
"grad_norm": 2.6492127951842988,
"learning_rate": 2.3341826411756863e-08,
"loss": 0.2743,
"step": 914
},
{
"epoch": 2.92332268370607,
"grad_norm": 2.5926056634489485,
"learning_rate": 2.158209863804217e-08,
"loss": 0.2808,
"step": 915
},
{
"epoch": 2.9265175718849843,
"grad_norm": 2.929343815942257,
"learning_rate": 1.9891184923544472e-08,
"loss": 0.3783,
"step": 916
},
{
"epoch": 2.9297124600638975,
"grad_norm": 3.0556363036567316,
"learning_rate": 1.826910864087761e-08,
"loss": 0.4242,
"step": 917
},
{
"epoch": 2.9329073482428116,
"grad_norm": 2.872033778289878,
"learning_rate": 1.6715892211150442e-08,
"loss": 0.3351,
"step": 918
},
{
"epoch": 2.9361022364217253,
"grad_norm": 3.3905203642444595,
"learning_rate": 1.5231557103658755e-08,
"loss": 0.4391,
"step": 919
},
{
"epoch": 2.939297124600639,
"grad_norm": 3.107371753494745,
"learning_rate": 1.3816123835588835e-08,
"loss": 0.3644,
"step": 920
},
{
"epoch": 2.9424920127795526,
"grad_norm": 2.5089417337734523,
"learning_rate": 1.2469611971731576e-08,
"loss": 0.2737,
"step": 921
},
{
"epoch": 2.9456869009584663,
"grad_norm": 2.166519253558388,
"learning_rate": 1.1192040124214931e-08,
"loss": 0.2395,
"step": 922
},
{
"epoch": 2.9488817891373804,
"grad_norm": 3.150150171798085,
"learning_rate": 9.983425952243552e-09,
"loss": 0.3938,
"step": 923
},
{
"epoch": 2.952076677316294,
"grad_norm": 2.780055165115101,
"learning_rate": 8.84378616185788e-09,
"loss": 0.3215,
"step": 924
},
{
"epoch": 2.9552715654952078,
"grad_norm": 3.6204049570837222,
"learning_rate": 7.773136505700995e-09,
"loss": 0.34,
"step": 925
},
{
"epoch": 2.9584664536741214,
"grad_norm": 2.9102079544741084,
"learning_rate": 6.7714917828004545e-09,
"loss": 0.3072,
"step": 926
},
{
"epoch": 2.961661341853035,
"grad_norm": 2.780088335958556,
"learning_rate": 5.838865838366792e-09,
"loss": 0.4582,
"step": 927
},
{
"epoch": 2.9648562300319488,
"grad_norm": 3.519543279375193,
"learning_rate": 4.975271563599227e-09,
"loss": 0.3057,
"step": 928
},
{
"epoch": 2.9680511182108624,
"grad_norm": 2.361253950476948,
"learning_rate": 4.180720895508028e-09,
"loss": 0.3041,
"step": 929
},
{
"epoch": 2.9712460063897765,
"grad_norm": 2.826713840533136,
"learning_rate": 3.4552248167507576e-09,
"loss": 0.3904,
"step": 930
},
{
"epoch": 2.97444089456869,
"grad_norm": 2.7066284920444383,
"learning_rate": 2.798793355478502e-09,
"loss": 0.2921,
"step": 931
},
{
"epoch": 2.977635782747604,
"grad_norm": 3.387115303059883,
"learning_rate": 2.2114355851993175e-09,
"loss": 0.3266,
"step": 932
},
{
"epoch": 2.9808306709265175,
"grad_norm": 2.315715217607924,
"learning_rate": 1.6931596246516636e-09,
"loss": 0.2233,
"step": 933
},
{
"epoch": 2.984025559105431,
"grad_norm": 3.115995399493926,
"learning_rate": 1.24397263769227e-09,
"loss": 0.2986,
"step": 934
},
{
"epoch": 2.987220447284345,
"grad_norm": 2.646328389939148,
"learning_rate": 8.638808331973281e-10,
"loss": 0.3112,
"step": 935
},
{
"epoch": 2.9904153354632586,
"grad_norm": 2.8345372122138768,
"learning_rate": 5.528894649758921e-10,
"loss": 0.387,
"step": 936
},
{
"epoch": 2.9936102236421727,
"grad_norm": 2.6136431958590505,
"learning_rate": 3.1100283169938074e-10,
"loss": 0.3357,
"step": 937
},
{
"epoch": 2.9968051118210863,
"grad_norm": 3.325928206994385,
"learning_rate": 1.3822427683884975e-10,
"loss": 0.3684,
"step": 938
},
{
"epoch": 3.0,
"grad_norm": 2.5101085856367975,
"learning_rate": 3.4556188622802964e-11,
"loss": 0.1622,
"step": 939
}
],
"logging_steps": 1,
"max_steps": 939,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5691721383936.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}