Shreesh-Coder's picture
Initial commit of the model
288325c verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.5559105431309903,
"eval_steps": 500,
"global_step": 400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 2.666666666666667e-06,
"loss": 3.7352,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 5.333333333333334e-06,
"loss": 3.7667,
"step": 2
},
{
"epoch": 0.02,
"learning_rate": 8.000000000000001e-06,
"loss": 3.7797,
"step": 3
},
{
"epoch": 0.03,
"learning_rate": 1.0666666666666667e-05,
"loss": 3.6592,
"step": 4
},
{
"epoch": 0.03,
"learning_rate": 1.3333333333333333e-05,
"loss": 3.6968,
"step": 5
},
{
"epoch": 0.04,
"learning_rate": 1.6000000000000003e-05,
"loss": 3.7754,
"step": 6
},
{
"epoch": 0.04,
"learning_rate": 1.866666666666667e-05,
"loss": 3.4281,
"step": 7
},
{
"epoch": 0.05,
"learning_rate": 2.1333333333333335e-05,
"loss": 3.6136,
"step": 8
},
{
"epoch": 0.06,
"learning_rate": 2.4e-05,
"loss": 3.0098,
"step": 9
},
{
"epoch": 0.06,
"learning_rate": 2.6666666666666667e-05,
"loss": 2.7693,
"step": 10
},
{
"epoch": 0.07,
"learning_rate": 2.9333333333333333e-05,
"loss": 2.2559,
"step": 11
},
{
"epoch": 0.08,
"learning_rate": 3.2000000000000005e-05,
"loss": 2.1567,
"step": 12
},
{
"epoch": 0.08,
"learning_rate": 3.466666666666667e-05,
"loss": 1.7343,
"step": 13
},
{
"epoch": 0.09,
"learning_rate": 3.733333333333334e-05,
"loss": 1.677,
"step": 14
},
{
"epoch": 0.1,
"learning_rate": 4e-05,
"loss": 1.2854,
"step": 15
},
{
"epoch": 0.1,
"learning_rate": 3.999951904814875e-05,
"loss": 1.1381,
"step": 16
},
{
"epoch": 0.11,
"learning_rate": 3.999807621572648e-05,
"loss": 0.7843,
"step": 17
},
{
"epoch": 0.12,
"learning_rate": 3.999567157212646e-05,
"loss": 0.7415,
"step": 18
},
{
"epoch": 0.12,
"learning_rate": 3.999230523300049e-05,
"loss": 0.6684,
"step": 19
},
{
"epoch": 0.13,
"learning_rate": 3.998797736025326e-05,
"loss": 0.4267,
"step": 20
},
{
"epoch": 0.13,
"learning_rate": 3.9982688162034624e-05,
"loss": 0.5993,
"step": 21
},
{
"epoch": 0.14,
"learning_rate": 3.997643789272954e-05,
"loss": 0.555,
"step": 22
},
{
"epoch": 0.15,
"learning_rate": 3.996922685294587e-05,
"loss": 0.5142,
"step": 23
},
{
"epoch": 0.15,
"learning_rate": 3.9961055389499904e-05,
"loss": 0.3841,
"step": 24
},
{
"epoch": 0.16,
"learning_rate": 3.9951923895399696e-05,
"loss": 0.3057,
"step": 25
},
{
"epoch": 0.17,
"learning_rate": 3.9941832809826136e-05,
"loss": 0.4327,
"step": 26
},
{
"epoch": 0.17,
"learning_rate": 3.993078261811186e-05,
"loss": 0.3207,
"step": 27
},
{
"epoch": 0.18,
"learning_rate": 3.991877385171789e-05,
"loss": 0.454,
"step": 28
},
{
"epoch": 0.19,
"learning_rate": 3.990580708820805e-05,
"loss": 0.2997,
"step": 29
},
{
"epoch": 0.19,
"learning_rate": 3.9891882951221246e-05,
"loss": 0.3583,
"step": 30
},
{
"epoch": 0.2,
"learning_rate": 3.9877002110441424e-05,
"loss": 0.2823,
"step": 31
},
{
"epoch": 0.2,
"learning_rate": 3.986116528156537e-05,
"loss": 0.2659,
"step": 32
},
{
"epoch": 0.21,
"learning_rate": 3.9844373226268305e-05,
"loss": 0.3158,
"step": 33
},
{
"epoch": 0.22,
"learning_rate": 3.982662675216723e-05,
"loss": 0.2704,
"step": 34
},
{
"epoch": 0.22,
"learning_rate": 3.9807926712782115e-05,
"loss": 0.3642,
"step": 35
},
{
"epoch": 0.23,
"learning_rate": 3.978827400749481e-05,
"loss": 0.3058,
"step": 36
},
{
"epoch": 0.24,
"learning_rate": 3.976766958150581e-05,
"loss": 0.3343,
"step": 37
},
{
"epoch": 0.24,
"learning_rate": 3.97461144257888e-05,
"loss": 0.2987,
"step": 38
},
{
"epoch": 0.25,
"learning_rate": 3.972360957704298e-05,
"loss": 0.2469,
"step": 39
},
{
"epoch": 0.26,
"learning_rate": 3.970015611764323e-05,
"loss": 0.359,
"step": 40
},
{
"epoch": 0.26,
"learning_rate": 3.9675755175588006e-05,
"loss": 0.2903,
"step": 41
},
{
"epoch": 0.27,
"learning_rate": 3.9650407924445147e-05,
"loss": 0.3881,
"step": 42
},
{
"epoch": 0.27,
"learning_rate": 3.9624115583295375e-05,
"loss": 0.2877,
"step": 43
},
{
"epoch": 0.28,
"learning_rate": 3.959687941667372e-05,
"loss": 0.2628,
"step": 44
},
{
"epoch": 0.29,
"learning_rate": 3.9568700734508645e-05,
"loss": 0.2789,
"step": 45
},
{
"epoch": 0.29,
"learning_rate": 3.9539580892059086e-05,
"loss": 0.2164,
"step": 46
},
{
"epoch": 0.3,
"learning_rate": 3.950952128984927e-05,
"loss": 0.217,
"step": 47
},
{
"epoch": 0.31,
"learning_rate": 3.9478523373601325e-05,
"loss": 0.2196,
"step": 48
},
{
"epoch": 0.31,
"learning_rate": 3.944658863416575e-05,
"loss": 0.3036,
"step": 49
},
{
"epoch": 0.32,
"learning_rate": 3.941371860744978e-05,
"loss": 0.2398,
"step": 50
},
{
"epoch": 0.33,
"learning_rate": 3.937991487434342e-05,
"loss": 0.2441,
"step": 51
},
{
"epoch": 0.33,
"learning_rate": 3.934517906064348e-05,
"loss": 0.2922,
"step": 52
},
{
"epoch": 0.34,
"learning_rate": 3.930951283697534e-05,
"loss": 0.3127,
"step": 53
},
{
"epoch": 0.35,
"learning_rate": 3.927291791871264e-05,
"loss": 0.2994,
"step": 54
},
{
"epoch": 0.35,
"learning_rate": 3.923539606589473e-05,
"loss": 0.2534,
"step": 55
},
{
"epoch": 0.36,
"learning_rate": 3.919694908314209e-05,
"loss": 0.2855,
"step": 56
},
{
"epoch": 0.36,
"learning_rate": 3.9157578819569455e-05,
"loss": 0.2889,
"step": 57
},
{
"epoch": 0.37,
"learning_rate": 3.9117287168696956e-05,
"loss": 0.3228,
"step": 58
},
{
"epoch": 0.38,
"learning_rate": 3.907607606835899e-05,
"loss": 0.2197,
"step": 59
},
{
"epoch": 0.38,
"learning_rate": 3.903394750061106e-05,
"loss": 0.3422,
"step": 60
},
{
"epoch": 0.39,
"learning_rate": 3.899090349163444e-05,
"loss": 0.2788,
"step": 61
},
{
"epoch": 0.4,
"learning_rate": 3.8946946111638696e-05,
"loss": 0.3245,
"step": 62
},
{
"epoch": 0.4,
"learning_rate": 3.8902077474762155e-05,
"loss": 0.336,
"step": 63
},
{
"epoch": 0.41,
"learning_rate": 3.8856299738970225e-05,
"loss": 0.2951,
"step": 64
},
{
"epoch": 0.42,
"learning_rate": 3.880961510595158e-05,
"loss": 0.3156,
"step": 65
},
{
"epoch": 0.42,
"learning_rate": 3.876202582101229e-05,
"loss": 0.2534,
"step": 66
},
{
"epoch": 0.43,
"learning_rate": 3.8713534172967815e-05,
"loss": 0.2775,
"step": 67
},
{
"epoch": 0.43,
"learning_rate": 3.866414249403295e-05,
"loss": 0.3381,
"step": 68
},
{
"epoch": 0.44,
"learning_rate": 3.861385315970964e-05,
"loss": 0.2593,
"step": 69
},
{
"epoch": 0.45,
"learning_rate": 3.856266858867273e-05,
"loss": 0.166,
"step": 70
},
{
"epoch": 0.45,
"learning_rate": 3.851059124265363e-05,
"loss": 0.3453,
"step": 71
},
{
"epoch": 0.46,
"learning_rate": 3.8457623626321944e-05,
"loss": 0.2238,
"step": 72
},
{
"epoch": 0.47,
"learning_rate": 3.840376828716499e-05,
"loss": 0.2773,
"step": 73
},
{
"epoch": 0.47,
"learning_rate": 3.834902781536527e-05,
"loss": 0.2777,
"step": 74
},
{
"epoch": 0.48,
"learning_rate": 3.8293404843675904e-05,
"loss": 0.3032,
"step": 75
},
{
"epoch": 0.49,
"learning_rate": 3.8236902047294015e-05,
"loss": 0.2138,
"step": 76
},
{
"epoch": 0.49,
"learning_rate": 3.817952214373206e-05,
"loss": 0.2605,
"step": 77
},
{
"epoch": 0.5,
"learning_rate": 3.812126789268712e-05,
"loss": 0.3205,
"step": 78
},
{
"epoch": 0.5,
"learning_rate": 3.806214209590819e-05,
"loss": 0.27,
"step": 79
},
{
"epoch": 0.51,
"learning_rate": 3.80021475970614e-05,
"loss": 0.2882,
"step": 80
},
{
"epoch": 0.52,
"learning_rate": 3.7941287281593284e-05,
"loss": 0.3081,
"step": 81
},
{
"epoch": 0.52,
"learning_rate": 3.787956407659198e-05,
"loss": 0.3268,
"step": 82
},
{
"epoch": 0.53,
"learning_rate": 3.781698095064647e-05,
"loss": 0.2018,
"step": 83
},
{
"epoch": 0.54,
"learning_rate": 3.775354091370376e-05,
"loss": 0.2624,
"step": 84
},
{
"epoch": 0.54,
"learning_rate": 3.7689247016924186e-05,
"loss": 0.2042,
"step": 85
},
{
"epoch": 0.55,
"learning_rate": 3.7624102352534615e-05,
"loss": 0.3193,
"step": 86
},
{
"epoch": 0.56,
"learning_rate": 3.755811005367974e-05,
"loss": 0.2487,
"step": 87
},
{
"epoch": 0.56,
"learning_rate": 3.7491273294271386e-05,
"loss": 0.1408,
"step": 88
},
{
"epoch": 0.57,
"learning_rate": 3.742359528883588e-05,
"loss": 0.276,
"step": 89
},
{
"epoch": 0.58,
"learning_rate": 3.735507929235941e-05,
"loss": 0.2293,
"step": 90
},
{
"epoch": 0.58,
"learning_rate": 3.7285728600131535e-05,
"loss": 0.2223,
"step": 91
},
{
"epoch": 0.59,
"learning_rate": 3.7215546547586596e-05,
"loss": 0.3011,
"step": 92
},
{
"epoch": 0.59,
"learning_rate": 3.7144536510143436e-05,
"loss": 0.2439,
"step": 93
},
{
"epoch": 0.6,
"learning_rate": 3.707270190304294e-05,
"loss": 0.2061,
"step": 94
},
{
"epoch": 0.61,
"learning_rate": 3.7000046181183834e-05,
"loss": 0.1888,
"step": 95
},
{
"epoch": 0.61,
"learning_rate": 3.692657283895651e-05,
"loss": 0.2698,
"step": 96
},
{
"epoch": 0.62,
"learning_rate": 3.6852285410074974e-05,
"loss": 0.1813,
"step": 97
},
{
"epoch": 0.63,
"learning_rate": 3.6777187467406857e-05,
"loss": 0.3824,
"step": 98
},
{
"epoch": 0.63,
"learning_rate": 3.6701282622801626e-05,
"loss": 0.2891,
"step": 99
},
{
"epoch": 0.64,
"learning_rate": 3.662457452691682e-05,
"loss": 0.1906,
"step": 100
},
{
"epoch": 0.65,
"learning_rate": 3.6547066869042524e-05,
"loss": 0.2127,
"step": 101
},
{
"epoch": 0.65,
"learning_rate": 3.6468763376923886e-05,
"loss": 0.3177,
"step": 102
},
{
"epoch": 0.66,
"learning_rate": 3.638966781658187e-05,
"loss": 0.1738,
"step": 103
},
{
"epoch": 0.66,
"learning_rate": 3.630978399213206e-05,
"loss": 0.1801,
"step": 104
},
{
"epoch": 0.67,
"learning_rate": 3.622911574560181e-05,
"loss": 0.2335,
"step": 105
},
{
"epoch": 0.68,
"learning_rate": 3.6147666956745364e-05,
"loss": 0.2521,
"step": 106
},
{
"epoch": 0.68,
"learning_rate": 3.60654415428573e-05,
"loss": 0.204,
"step": 107
},
{
"epoch": 0.69,
"learning_rate": 3.598244345858412e-05,
"loss": 0.2199,
"step": 108
},
{
"epoch": 0.7,
"learning_rate": 3.589867669573404e-05,
"loss": 0.322,
"step": 109
},
{
"epoch": 0.7,
"learning_rate": 3.5814145283085055e-05,
"loss": 0.2733,
"step": 110
},
{
"epoch": 0.71,
"learning_rate": 3.5728853286191075e-05,
"loss": 0.3102,
"step": 111
},
{
"epoch": 0.72,
"learning_rate": 3.56428048071865e-05,
"loss": 0.2365,
"step": 112
},
{
"epoch": 0.72,
"learning_rate": 3.555600398458885e-05,
"loss": 0.1718,
"step": 113
},
{
"epoch": 0.73,
"learning_rate": 3.546845499309976e-05,
"loss": 0.198,
"step": 114
},
{
"epoch": 0.73,
"learning_rate": 3.538016204340418e-05,
"loss": 0.2557,
"step": 115
},
{
"epoch": 0.74,
"learning_rate": 3.529112938196787e-05,
"loss": 0.1825,
"step": 116
},
{
"epoch": 0.75,
"learning_rate": 3.5201361290833165e-05,
"loss": 0.3099,
"step": 117
},
{
"epoch": 0.75,
"learning_rate": 3.511086208741303e-05,
"loss": 0.3273,
"step": 118
},
{
"epoch": 0.76,
"learning_rate": 3.501963612428341e-05,
"loss": 0.4337,
"step": 119
},
{
"epoch": 0.77,
"learning_rate": 3.492768778897388e-05,
"loss": 0.2754,
"step": 120
},
{
"epoch": 0.77,
"learning_rate": 3.483502150375665e-05,
"loss": 0.2226,
"step": 121
},
{
"epoch": 0.78,
"learning_rate": 3.474164172543386e-05,
"loss": 0.2698,
"step": 122
},
{
"epoch": 0.79,
"learning_rate": 3.464755294512325e-05,
"loss": 0.2,
"step": 123
},
{
"epoch": 0.79,
"learning_rate": 3.455275968804212e-05,
"loss": 0.1865,
"step": 124
},
{
"epoch": 0.8,
"learning_rate": 3.445726651328971e-05,
"loss": 0.2586,
"step": 125
},
{
"epoch": 0.81,
"learning_rate": 3.4361078013627945e-05,
"loss": 0.3064,
"step": 126
},
{
"epoch": 0.81,
"learning_rate": 3.426419881526052e-05,
"loss": 0.21,
"step": 127
},
{
"epoch": 0.82,
"learning_rate": 3.4166633577610425e-05,
"loss": 0.1701,
"step": 128
},
{
"epoch": 0.82,
"learning_rate": 3.4068386993095806e-05,
"loss": 0.2909,
"step": 129
},
{
"epoch": 0.83,
"learning_rate": 3.396946378690435e-05,
"loss": 0.2558,
"step": 130
},
{
"epoch": 0.84,
"learning_rate": 3.386986871676597e-05,
"loss": 0.1833,
"step": 131
},
{
"epoch": 0.84,
"learning_rate": 3.3769606572724e-05,
"loss": 0.184,
"step": 132
},
{
"epoch": 0.85,
"learning_rate": 3.366868217690482e-05,
"loss": 0.2481,
"step": 133
},
{
"epoch": 0.86,
"learning_rate": 3.3567100383285925e-05,
"loss": 0.1628,
"step": 134
},
{
"epoch": 0.86,
"learning_rate": 3.346486607746249e-05,
"loss": 0.1711,
"step": 135
},
{
"epoch": 0.87,
"learning_rate": 3.336198417641238e-05,
"loss": 0.2293,
"step": 136
},
{
"epoch": 0.88,
"learning_rate": 3.325845962825966e-05,
"loss": 0.2515,
"step": 137
},
{
"epoch": 0.88,
"learning_rate": 3.315429741203666e-05,
"loss": 0.1741,
"step": 138
},
{
"epoch": 0.89,
"learning_rate": 3.304950253744443e-05,
"loss": 0.1595,
"step": 139
},
{
"epoch": 0.89,
"learning_rate": 3.294408004461188e-05,
"loss": 0.1889,
"step": 140
},
{
"epoch": 0.9,
"learning_rate": 3.283803500385332e-05,
"loss": 0.2864,
"step": 141
},
{
"epoch": 0.91,
"learning_rate": 3.27313725154246e-05,
"loss": 0.1941,
"step": 142
},
{
"epoch": 0.91,
"learning_rate": 3.2624097709277855e-05,
"loss": 0.2526,
"step": 143
},
{
"epoch": 0.92,
"learning_rate": 3.251621574481475e-05,
"loss": 0.1784,
"step": 144
},
{
"epoch": 0.93,
"learning_rate": 3.240773181063834e-05,
"loss": 0.3217,
"step": 145
},
{
"epoch": 0.93,
"learning_rate": 3.229865112430352e-05,
"loss": 0.2438,
"step": 146
},
{
"epoch": 0.94,
"learning_rate": 3.218897893206608e-05,
"loss": 0.2797,
"step": 147
},
{
"epoch": 0.95,
"learning_rate": 3.2078720508630427e-05,
"loss": 0.2358,
"step": 148
},
{
"epoch": 0.95,
"learning_rate": 3.196788115689584e-05,
"loss": 0.2403,
"step": 149
},
{
"epoch": 0.96,
"learning_rate": 3.185646620770146e-05,
"loss": 0.2295,
"step": 150
},
{
"epoch": 0.96,
"learning_rate": 3.1744481019569885e-05,
"loss": 0.2063,
"step": 151
},
{
"epoch": 0.97,
"learning_rate": 3.163193097844949e-05,
"loss": 0.2654,
"step": 152
},
{
"epoch": 0.98,
"learning_rate": 3.1518821497455326e-05,
"loss": 0.2071,
"step": 153
},
{
"epoch": 0.98,
"learning_rate": 3.1405158016608806e-05,
"loss": 0.2035,
"step": 154
},
{
"epoch": 0.99,
"learning_rate": 3.129094600257611e-05,
"loss": 0.2398,
"step": 155
},
{
"epoch": 1.0,
"learning_rate": 3.1176190948405194e-05,
"loss": 0.1623,
"step": 156
},
{
"epoch": 1.0,
"learning_rate": 3.106089837326161e-05,
"loss": 0.2313,
"step": 157
},
{
"epoch": 1.01,
"learning_rate": 3.094507382216312e-05,
"loss": 0.2687,
"step": 158
},
{
"epoch": 1.02,
"learning_rate": 3.082872286571295e-05,
"loss": 0.246,
"step": 159
},
{
"epoch": 1.02,
"learning_rate": 3.0711851099831885e-05,
"loss": 0.191,
"step": 160
},
{
"epoch": 1.03,
"learning_rate": 3.059446414548915e-05,
"loss": 0.1543,
"step": 161
},
{
"epoch": 1.04,
"learning_rate": 3.047656764843203e-05,
"loss": 0.1978,
"step": 162
},
{
"epoch": 1.04,
"learning_rate": 3.0358167278914387e-05,
"loss": 0.2698,
"step": 163
},
{
"epoch": 1.05,
"learning_rate": 3.023926873142391e-05,
"loss": 0.2351,
"step": 164
},
{
"epoch": 1.05,
"learning_rate": 3.011987772440825e-05,
"loss": 0.1925,
"step": 165
},
{
"epoch": 1.06,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.2246,
"step": 166
},
{
"epoch": 1.07,
"learning_rate": 2.9879641323740505e-05,
"loss": 0.2093,
"step": 167
},
{
"epoch": 1.07,
"learning_rate": 2.9758807484302566e-05,
"loss": 0.2332,
"step": 168
},
{
"epoch": 1.08,
"learning_rate": 2.963750429321208e-05,
"loss": 0.1983,
"step": 169
},
{
"epoch": 1.09,
"learning_rate": 2.9515737584568463e-05,
"loss": 0.2707,
"step": 170
},
{
"epoch": 1.09,
"learning_rate": 2.939351321476412e-05,
"loss": 0.2485,
"step": 171
},
{
"epoch": 1.1,
"learning_rate": 2.927083706220274e-05,
"loss": 0.1947,
"step": 172
},
{
"epoch": 1.11,
"learning_rate": 2.9147715027016593e-05,
"loss": 0.1949,
"step": 173
},
{
"epoch": 1.11,
"learning_rate": 2.902415303078275e-05,
"loss": 0.1907,
"step": 174
},
{
"epoch": 1.12,
"learning_rate": 2.8900157016238296e-05,
"loss": 0.1801,
"step": 175
},
{
"epoch": 1.12,
"learning_rate": 2.8775732946994508e-05,
"loss": 0.2489,
"step": 176
},
{
"epoch": 1.13,
"learning_rate": 2.8650886807250024e-05,
"loss": 0.1969,
"step": 177
},
{
"epoch": 1.14,
"learning_rate": 2.8525624601503055e-05,
"loss": 0.1724,
"step": 178
},
{
"epoch": 1.14,
"learning_rate": 2.8399952354262566e-05,
"loss": 0.1797,
"step": 179
},
{
"epoch": 1.15,
"learning_rate": 2.8273876109758568e-05,
"loss": 0.1341,
"step": 180
},
{
"epoch": 1.16,
"learning_rate": 2.8147401931651363e-05,
"loss": 0.2223,
"step": 181
},
{
"epoch": 1.16,
"learning_rate": 2.802053590273997e-05,
"loss": 0.2221,
"step": 182
},
{
"epoch": 1.17,
"learning_rate": 2.789328412466953e-05,
"loss": 0.2019,
"step": 183
},
{
"epoch": 1.18,
"learning_rate": 2.7765652717637873e-05,
"loss": 0.1336,
"step": 184
},
{
"epoch": 1.18,
"learning_rate": 2.763764782010116e-05,
"loss": 0.2361,
"step": 185
},
{
"epoch": 1.19,
"learning_rate": 2.7509275588478606e-05,
"loss": 0.2023,
"step": 186
},
{
"epoch": 1.19,
"learning_rate": 2.738054219685647e-05,
"loss": 0.2428,
"step": 187
},
{
"epoch": 1.2,
"learning_rate": 2.725145383669106e-05,
"loss": 0.2394,
"step": 188
},
{
"epoch": 1.21,
"learning_rate": 2.712201671651094e-05,
"loss": 0.1464,
"step": 189
},
{
"epoch": 1.21,
"learning_rate": 2.699223706161839e-05,
"loss": 0.2485,
"step": 190
},
{
"epoch": 1.22,
"learning_rate": 2.6862121113789917e-05,
"loss": 0.3039,
"step": 191
},
{
"epoch": 1.23,
"learning_rate": 2.673167513097613e-05,
"loss": 0.1387,
"step": 192
},
{
"epoch": 1.23,
"learning_rate": 2.6600905387000716e-05,
"loss": 0.2657,
"step": 193
},
{
"epoch": 1.24,
"learning_rate": 2.6469818171258723e-05,
"loss": 0.176,
"step": 194
},
{
"epoch": 1.25,
"learning_rate": 2.633841978841406e-05,
"loss": 0.1876,
"step": 195
},
{
"epoch": 1.25,
"learning_rate": 2.620671655809627e-05,
"loss": 0.1726,
"step": 196
},
{
"epoch": 1.26,
"learning_rate": 2.60747148145966e-05,
"loss": 0.2352,
"step": 197
},
{
"epoch": 1.27,
"learning_rate": 2.594242090656335e-05,
"loss": 0.254,
"step": 198
},
{
"epoch": 1.27,
"learning_rate": 2.5809841196696504e-05,
"loss": 0.2125,
"step": 199
},
{
"epoch": 1.28,
"learning_rate": 2.5676982061441763e-05,
"loss": 0.2049,
"step": 200
},
{
"epoch": 1.28,
"learning_rate": 2.5543849890683813e-05,
"loss": 0.1865,
"step": 201
},
{
"epoch": 1.29,
"learning_rate": 2.5410451087439075e-05,
"loss": 0.2203,
"step": 202
},
{
"epoch": 1.3,
"learning_rate": 2.5276792067547672e-05,
"loss": 0.2002,
"step": 203
},
{
"epoch": 1.3,
"learning_rate": 2.514287925936492e-05,
"loss": 0.1524,
"step": 204
},
{
"epoch": 1.31,
"learning_rate": 2.500871910345212e-05,
"loss": 0.2059,
"step": 205
},
{
"epoch": 1.32,
"learning_rate": 2.4874318052266794e-05,
"loss": 0.1885,
"step": 206
},
{
"epoch": 1.32,
"learning_rate": 2.473968256985238e-05,
"loss": 0.1837,
"step": 207
},
{
"epoch": 1.33,
"learning_rate": 2.460481913152734e-05,
"loss": 0.1938,
"step": 208
},
{
"epoch": 1.34,
"learning_rate": 2.4469734223573703e-05,
"loss": 0.1497,
"step": 209
},
{
"epoch": 1.34,
"learning_rate": 2.4334434342925133e-05,
"loss": 0.2644,
"step": 210
},
{
"epoch": 1.35,
"learning_rate": 2.4198925996854422e-05,
"loss": 0.1937,
"step": 211
},
{
"epoch": 1.35,
"learning_rate": 2.4063215702660564e-05,
"loss": 0.2791,
"step": 212
},
{
"epoch": 1.36,
"learning_rate": 2.392730998735529e-05,
"loss": 0.24,
"step": 213
},
{
"epoch": 1.37,
"learning_rate": 2.379121538734912e-05,
"loss": 0.1574,
"step": 214
},
{
"epoch": 1.37,
"learning_rate": 2.3654938448137062e-05,
"loss": 0.1896,
"step": 215
},
{
"epoch": 1.38,
"learning_rate": 2.351848572398371e-05,
"loss": 0.1865,
"step": 216
},
{
"epoch": 1.39,
"learning_rate": 2.338186377760811e-05,
"loss": 0.1336,
"step": 217
},
{
"epoch": 1.39,
"learning_rate": 2.3245079179868054e-05,
"loss": 0.2047,
"step": 218
},
{
"epoch": 1.4,
"learning_rate": 2.31081385094441e-05,
"loss": 0.31,
"step": 219
},
{
"epoch": 1.41,
"learning_rate": 2.297104835252314e-05,
"loss": 0.2096,
"step": 220
},
{
"epoch": 1.41,
"learning_rate": 2.283381530248165e-05,
"loss": 0.2593,
"step": 221
},
{
"epoch": 1.42,
"learning_rate": 2.2696445959568577e-05,
"loss": 0.1585,
"step": 222
},
{
"epoch": 1.42,
"learning_rate": 2.2558946930587907e-05,
"loss": 0.1781,
"step": 223
},
{
"epoch": 1.43,
"learning_rate": 2.2421324828580877e-05,
"loss": 0.1462,
"step": 224
},
{
"epoch": 1.44,
"learning_rate": 2.2283586272507975e-05,
"loss": 0.1687,
"step": 225
},
{
"epoch": 1.44,
"learning_rate": 2.214573788693054e-05,
"loss": 0.1716,
"step": 226
},
{
"epoch": 1.45,
"learning_rate": 2.2007786301692205e-05,
"loss": 0.2694,
"step": 227
},
{
"epoch": 1.46,
"learning_rate": 2.18697381516e-05,
"loss": 0.1615,
"step": 228
},
{
"epoch": 1.46,
"learning_rate": 2.1731600076105264e-05,
"loss": 0.1891,
"step": 229
},
{
"epoch": 1.47,
"learning_rate": 2.159337871898431e-05,
"loss": 0.1321,
"step": 230
},
{
"epoch": 1.48,
"learning_rate": 2.145508072801888e-05,
"loss": 0.1706,
"step": 231
},
{
"epoch": 1.48,
"learning_rate": 2.131671275467647e-05,
"loss": 0.2391,
"step": 232
},
{
"epoch": 1.49,
"learning_rate": 2.1178281453790358e-05,
"loss": 0.1462,
"step": 233
},
{
"epoch": 1.5,
"learning_rate": 2.1039793483239607e-05,
"loss": 0.1928,
"step": 234
},
{
"epoch": 1.5,
"learning_rate": 2.090125550362879e-05,
"loss": 0.3621,
"step": 235
},
{
"epoch": 1.51,
"learning_rate": 2.0762674177967676e-05,
"loss": 0.1987,
"step": 236
},
{
"epoch": 1.51,
"learning_rate": 2.0624056171350785e-05,
"loss": 0.2071,
"step": 237
},
{
"epoch": 1.52,
"learning_rate": 2.0485408150636804e-05,
"loss": 0.1817,
"step": 238
},
{
"epoch": 1.53,
"learning_rate": 2.0346736784127955e-05,
"loss": 0.1615,
"step": 239
},
{
"epoch": 1.53,
"learning_rate": 2.0208048741249288e-05,
"loss": 0.1521,
"step": 240
},
{
"epoch": 1.54,
"learning_rate": 2.006935069222789e-05,
"loss": 0.156,
"step": 241
},
{
"epoch": 1.55,
"learning_rate": 1.9930649307772114e-05,
"loss": 0.2438,
"step": 242
},
{
"epoch": 1.55,
"learning_rate": 1.979195125875072e-05,
"loss": 0.2432,
"step": 243
},
{
"epoch": 1.56,
"learning_rate": 1.9653263215872048e-05,
"loss": 0.1292,
"step": 244
},
{
"epoch": 1.57,
"learning_rate": 1.9514591849363203e-05,
"loss": 0.2293,
"step": 245
},
{
"epoch": 1.57,
"learning_rate": 1.9375943828649215e-05,
"loss": 0.1932,
"step": 246
},
{
"epoch": 1.58,
"learning_rate": 1.923732582203233e-05,
"loss": 0.1508,
"step": 247
},
{
"epoch": 1.58,
"learning_rate": 1.909874449637122e-05,
"loss": 0.2317,
"step": 248
},
{
"epoch": 1.59,
"learning_rate": 1.8960206516760396e-05,
"loss": 0.218,
"step": 249
},
{
"epoch": 1.6,
"learning_rate": 1.8821718546209646e-05,
"loss": 0.1777,
"step": 250
},
{
"epoch": 1.6,
"learning_rate": 1.8683287245323536e-05,
"loss": 0.1577,
"step": 251
},
{
"epoch": 1.61,
"learning_rate": 1.8544919271981125e-05,
"loss": 0.2977,
"step": 252
},
{
"epoch": 1.62,
"learning_rate": 1.84066212810157e-05,
"loss": 0.2277,
"step": 253
},
{
"epoch": 1.62,
"learning_rate": 1.8268399923894736e-05,
"loss": 0.1574,
"step": 254
},
{
"epoch": 1.63,
"learning_rate": 1.8130261848399996e-05,
"loss": 0.3187,
"step": 255
},
{
"epoch": 1.64,
"learning_rate": 1.7992213698307795e-05,
"loss": 0.1758,
"step": 256
},
{
"epoch": 1.64,
"learning_rate": 1.7854262113069468e-05,
"loss": 0.2118,
"step": 257
},
{
"epoch": 1.65,
"learning_rate": 1.7716413727492035e-05,
"loss": 0.1539,
"step": 258
},
{
"epoch": 1.65,
"learning_rate": 1.757867517141913e-05,
"loss": 0.1885,
"step": 259
},
{
"epoch": 1.66,
"learning_rate": 1.7441053069412103e-05,
"loss": 0.1915,
"step": 260
},
{
"epoch": 1.67,
"learning_rate": 1.7303554040431426e-05,
"loss": 0.1716,
"step": 261
},
{
"epoch": 1.67,
"learning_rate": 1.7166184697518352e-05,
"loss": 0.2217,
"step": 262
},
{
"epoch": 1.68,
"learning_rate": 1.7028951647476862e-05,
"loss": 0.3266,
"step": 263
},
{
"epoch": 1.69,
"learning_rate": 1.6891861490555906e-05,
"loss": 0.2631,
"step": 264
},
{
"epoch": 1.69,
"learning_rate": 1.6754920820131946e-05,
"loss": 0.2481,
"step": 265
},
{
"epoch": 1.7,
"learning_rate": 1.6618136222391893e-05,
"loss": 0.2521,
"step": 266
},
{
"epoch": 1.71,
"learning_rate": 1.6481514276016297e-05,
"loss": 0.2022,
"step": 267
},
{
"epoch": 1.71,
"learning_rate": 1.634506155186295e-05,
"loss": 0.2055,
"step": 268
},
{
"epoch": 1.72,
"learning_rate": 1.6208784612650883e-05,
"loss": 0.2257,
"step": 269
},
{
"epoch": 1.73,
"learning_rate": 1.6072690012644717e-05,
"loss": 0.2634,
"step": 270
},
{
"epoch": 1.73,
"learning_rate": 1.593678429733944e-05,
"loss": 0.2195,
"step": 271
},
{
"epoch": 1.74,
"learning_rate": 1.5801074003145585e-05,
"loss": 0.2063,
"step": 272
},
{
"epoch": 1.74,
"learning_rate": 1.5665565657074874e-05,
"loss": 0.2601,
"step": 273
},
{
"epoch": 1.75,
"learning_rate": 1.5530265776426294e-05,
"loss": 0.1633,
"step": 274
},
{
"epoch": 1.76,
"learning_rate": 1.5395180868472662e-05,
"loss": 0.2333,
"step": 275
},
{
"epoch": 1.76,
"learning_rate": 1.5260317430147627e-05,
"loss": 0.149,
"step": 276
},
{
"epoch": 1.77,
"learning_rate": 1.512568194773322e-05,
"loss": 0.2587,
"step": 277
},
{
"epoch": 1.78,
"learning_rate": 1.4991280896547893e-05,
"loss": 0.1951,
"step": 278
},
{
"epoch": 1.78,
"learning_rate": 1.4857120740635084e-05,
"loss": 0.2757,
"step": 279
},
{
"epoch": 1.79,
"learning_rate": 1.472320793245233e-05,
"loss": 0.2321,
"step": 280
},
{
"epoch": 1.8,
"learning_rate": 1.4589548912560932e-05,
"loss": 0.2177,
"step": 281
},
{
"epoch": 1.8,
"learning_rate": 1.4456150109316192e-05,
"loss": 0.1363,
"step": 282
},
{
"epoch": 1.81,
"learning_rate": 1.4323017938558245e-05,
"loss": 0.1428,
"step": 283
},
{
"epoch": 1.81,
"learning_rate": 1.4190158803303498e-05,
"loss": 0.2112,
"step": 284
},
{
"epoch": 1.82,
"learning_rate": 1.4057579093436653e-05,
"loss": 0.2269,
"step": 285
},
{
"epoch": 1.83,
"learning_rate": 1.3925285185403406e-05,
"loss": 0.1712,
"step": 286
},
{
"epoch": 1.83,
"learning_rate": 1.3793283441903737e-05,
"loss": 0.1182,
"step": 287
},
{
"epoch": 1.84,
"learning_rate": 1.3661580211585947e-05,
"loss": 0.1667,
"step": 288
},
{
"epoch": 1.85,
"learning_rate": 1.3530181828741285e-05,
"loss": 0.2887,
"step": 289
},
{
"epoch": 1.85,
"learning_rate": 1.3399094612999291e-05,
"loss": 0.1624,
"step": 290
},
{
"epoch": 1.86,
"learning_rate": 1.3268324869023878e-05,
"loss": 0.2356,
"step": 291
},
{
"epoch": 1.87,
"learning_rate": 1.313787888621009e-05,
"loss": 0.2557,
"step": 292
},
{
"epoch": 1.87,
"learning_rate": 1.3007762938381619e-05,
"loss": 0.2203,
"step": 293
},
{
"epoch": 1.88,
"learning_rate": 1.2877983283489062e-05,
"loss": 0.3142,
"step": 294
},
{
"epoch": 1.88,
"learning_rate": 1.2748546163308947e-05,
"loss": 0.16,
"step": 295
},
{
"epoch": 1.89,
"learning_rate": 1.261945780314354e-05,
"loss": 0.1647,
"step": 296
},
{
"epoch": 1.9,
"learning_rate": 1.2490724411521406e-05,
"loss": 0.2373,
"step": 297
},
{
"epoch": 1.9,
"learning_rate": 1.2362352179898855e-05,
"loss": 0.1701,
"step": 298
},
{
"epoch": 1.91,
"learning_rate": 1.2234347282362129e-05,
"loss": 0.1522,
"step": 299
},
{
"epoch": 1.92,
"learning_rate": 1.2106715875330475e-05,
"loss": 0.1613,
"step": 300
},
{
"epoch": 1.92,
"learning_rate": 1.1979464097260039e-05,
"loss": 0.1883,
"step": 301
},
{
"epoch": 1.93,
"learning_rate": 1.1852598068348642e-05,
"loss": 0.2338,
"step": 302
},
{
"epoch": 1.94,
"learning_rate": 1.1726123890241439e-05,
"loss": 0.2248,
"step": 303
},
{
"epoch": 1.94,
"learning_rate": 1.1600047645737433e-05,
"loss": 0.1811,
"step": 304
},
{
"epoch": 1.95,
"learning_rate": 1.1474375398496948e-05,
"loss": 0.1893,
"step": 305
},
{
"epoch": 1.96,
"learning_rate": 1.1349113192749986e-05,
"loss": 0.203,
"step": 306
},
{
"epoch": 1.96,
"learning_rate": 1.1224267053005504e-05,
"loss": 0.1425,
"step": 307
},
{
"epoch": 1.97,
"learning_rate": 1.1099842983761712e-05,
"loss": 0.1337,
"step": 308
},
{
"epoch": 1.97,
"learning_rate": 1.0975846969217258e-05,
"loss": 0.1684,
"step": 309
},
{
"epoch": 1.98,
"learning_rate": 1.0852284972983415e-05,
"loss": 0.165,
"step": 310
},
{
"epoch": 1.99,
"learning_rate": 1.0729162937797257e-05,
"loss": 0.2462,
"step": 311
},
{
"epoch": 1.99,
"learning_rate": 1.0606486785235879e-05,
"loss": 0.2298,
"step": 312
},
{
"epoch": 2.0,
"learning_rate": 1.0484262415431536e-05,
"loss": 0.2047,
"step": 313
},
{
"epoch": 2.01,
"learning_rate": 1.0362495706787923e-05,
"loss": 0.2365,
"step": 314
},
{
"epoch": 2.01,
"learning_rate": 1.0241192515697432e-05,
"loss": 0.1738,
"step": 315
},
{
"epoch": 2.02,
"learning_rate": 1.0120358676259508e-05,
"loss": 0.2467,
"step": 316
},
{
"epoch": 2.03,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.1459,
"step": 317
},
{
"epoch": 2.03,
"learning_rate": 9.880122275591752e-06,
"loss": 0.1594,
"step": 318
},
{
"epoch": 2.04,
"learning_rate": 9.760731268576095e-06,
"loss": 0.1145,
"step": 319
},
{
"epoch": 2.04,
"learning_rate": 9.64183272108562e-06,
"loss": 0.1485,
"step": 320
},
{
"epoch": 2.05,
"learning_rate": 9.523432351567979e-06,
"loss": 0.0932,
"step": 321
},
{
"epoch": 2.06,
"learning_rate": 9.405535854510863e-06,
"loss": 0.1819,
"step": 322
},
{
"epoch": 2.06,
"learning_rate": 9.288148900168122e-06,
"loss": 0.1534,
"step": 323
},
{
"epoch": 2.07,
"learning_rate": 9.171277134287057e-06,
"loss": 0.1875,
"step": 324
},
{
"epoch": 2.08,
"learning_rate": 9.054926177836878e-06,
"loss": 0.1416,
"step": 325
},
{
"epoch": 2.08,
"learning_rate": 8.939101626738395e-06,
"loss": 0.1717,
"step": 326
},
{
"epoch": 2.09,
"learning_rate": 8.823809051594816e-06,
"loss": 0.1975,
"step": 327
},
{
"epoch": 2.1,
"learning_rate": 8.70905399742389e-06,
"loss": 0.1706,
"step": 328
},
{
"epoch": 2.1,
"learning_rate": 8.594841983391196e-06,
"loss": 0.1869,
"step": 329
},
{
"epoch": 2.11,
"learning_rate": 8.481178502544684e-06,
"loss": 0.2313,
"step": 330
},
{
"epoch": 2.12,
"learning_rate": 8.368069021550516e-06,
"loss": 0.1843,
"step": 331
},
{
"epoch": 2.12,
"learning_rate": 8.255518980430115e-06,
"loss": 0.2359,
"step": 332
},
{
"epoch": 2.13,
"learning_rate": 8.143533792298545e-06,
"loss": 0.1243,
"step": 333
},
{
"epoch": 2.13,
"learning_rate": 8.032118843104164e-06,
"loss": 0.1325,
"step": 334
},
{
"epoch": 2.14,
"learning_rate": 7.921279491369575e-06,
"loss": 0.099,
"step": 335
},
{
"epoch": 2.15,
"learning_rate": 7.811021067933919e-06,
"loss": 0.2295,
"step": 336
},
{
"epoch": 2.15,
"learning_rate": 7.701348875696486e-06,
"loss": 0.1494,
"step": 337
},
{
"epoch": 2.16,
"learning_rate": 7.59226818936166e-06,
"loss": 0.1063,
"step": 338
},
{
"epoch": 2.17,
"learning_rate": 7.483784255185249e-06,
"loss": 0.1389,
"step": 339
},
{
"epoch": 2.17,
"learning_rate": 7.375902290722146e-06,
"loss": 0.176,
"step": 340
},
{
"epoch": 2.18,
"learning_rate": 7.268627484575406e-06,
"loss": 0.142,
"step": 341
},
{
"epoch": 2.19,
"learning_rate": 7.161964996146689e-06,
"loss": 0.2029,
"step": 342
},
{
"epoch": 2.19,
"learning_rate": 7.055919955388122e-06,
"loss": 0.1416,
"step": 343
},
{
"epoch": 2.2,
"learning_rate": 6.95049746255557e-06,
"loss": 0.1515,
"step": 344
},
{
"epoch": 2.2,
"learning_rate": 6.845702587963352e-06,
"loss": 0.1573,
"step": 345
},
{
"epoch": 2.21,
"learning_rate": 6.741540371740347e-06,
"loss": 0.2249,
"step": 346
},
{
"epoch": 2.22,
"learning_rate": 6.6380158235876335e-06,
"loss": 0.2207,
"step": 347
},
{
"epoch": 2.22,
"learning_rate": 6.535133922537513e-06,
"loss": 0.1194,
"step": 348
},
{
"epoch": 2.23,
"learning_rate": 6.4328996167140786e-06,
"loss": 0.1655,
"step": 349
},
{
"epoch": 2.24,
"learning_rate": 6.331317823095184e-06,
"loss": 0.1652,
"step": 350
},
{
"epoch": 2.24,
"learning_rate": 6.230393427276e-06,
"loss": 0.2378,
"step": 351
},
{
"epoch": 2.25,
"learning_rate": 6.130131283234031e-06,
"loss": 0.229,
"step": 352
},
{
"epoch": 2.26,
"learning_rate": 6.0305362130956504e-06,
"loss": 0.1397,
"step": 353
},
{
"epoch": 2.26,
"learning_rate": 5.931613006904196e-06,
"loss": 0.1139,
"step": 354
},
{
"epoch": 2.27,
"learning_rate": 5.8333664223895906e-06,
"loss": 0.1611,
"step": 355
},
{
"epoch": 2.27,
"learning_rate": 5.735801184739489e-06,
"loss": 0.1784,
"step": 356
},
{
"epoch": 2.28,
"learning_rate": 5.638921986372064e-06,
"loss": 0.2004,
"step": 357
},
{
"epoch": 2.29,
"learning_rate": 5.542733486710299e-06,
"loss": 0.1553,
"step": 358
},
{
"epoch": 2.29,
"learning_rate": 5.447240311957891e-06,
"loss": 0.1787,
"step": 359
},
{
"epoch": 2.3,
"learning_rate": 5.352447054876755e-06,
"loss": 0.1856,
"step": 360
},
{
"epoch": 2.31,
"learning_rate": 5.258358274566142e-06,
"loss": 0.2069,
"step": 361
},
{
"epoch": 2.31,
"learning_rate": 5.164978496243354e-06,
"loss": 0.2023,
"step": 362
},
{
"epoch": 2.32,
"learning_rate": 5.072312211026125e-06,
"loss": 0.1453,
"step": 363
},
{
"epoch": 2.33,
"learning_rate": 4.980363875716592e-06,
"loss": 0.1377,
"step": 364
},
{
"epoch": 2.33,
"learning_rate": 4.889137912586972e-06,
"loss": 0.1887,
"step": 365
},
{
"epoch": 2.34,
"learning_rate": 4.7986387091668365e-06,
"loss": 0.0932,
"step": 366
},
{
"epoch": 2.35,
"learning_rate": 4.708870618032133e-06,
"loss": 0.1506,
"step": 367
},
{
"epoch": 2.35,
"learning_rate": 4.619837956595825e-06,
"loss": 0.1293,
"step": 368
},
{
"epoch": 2.36,
"learning_rate": 4.531545006900244e-06,
"loss": 0.1505,
"step": 369
},
{
"epoch": 2.36,
"learning_rate": 4.443996015411151e-06,
"loss": 0.1176,
"step": 370
},
{
"epoch": 2.37,
"learning_rate": 4.357195192813504e-06,
"loss": 0.2187,
"step": 371
},
{
"epoch": 2.38,
"learning_rate": 4.271146713808927e-06,
"loss": 0.1538,
"step": 372
},
{
"epoch": 2.38,
"learning_rate": 4.185854716914952e-06,
"loss": 0.2165,
"step": 373
},
{
"epoch": 2.39,
"learning_rate": 4.1013233042659606e-06,
"loss": 0.1662,
"step": 374
},
{
"epoch": 2.4,
"learning_rate": 4.017556541415888e-06,
"loss": 0.143,
"step": 375
},
{
"epoch": 2.4,
"learning_rate": 3.9345584571427055e-06,
"loss": 0.1883,
"step": 376
},
{
"epoch": 2.41,
"learning_rate": 3.852333043254639e-06,
"loss": 0.1624,
"step": 377
},
{
"epoch": 2.42,
"learning_rate": 3.7708842543981928e-06,
"loss": 0.1824,
"step": 378
},
{
"epoch": 2.42,
"learning_rate": 3.690216007867944e-06,
"loss": 0.1933,
"step": 379
},
{
"epoch": 2.43,
"learning_rate": 3.6103321834181437e-06,
"loss": 0.1557,
"step": 380
},
{
"epoch": 2.43,
"learning_rate": 3.5312366230761154e-06,
"loss": 0.143,
"step": 381
},
{
"epoch": 2.44,
"learning_rate": 3.452933130957481e-06,
"loss": 0.1588,
"step": 382
},
{
"epoch": 2.45,
"learning_rate": 3.375425473083185e-06,
"loss": 0.1964,
"step": 383
},
{
"epoch": 2.45,
"learning_rate": 3.2987173771983816e-06,
"loss": 0.1688,
"step": 384
},
{
"epoch": 2.46,
"learning_rate": 3.2228125325931514e-06,
"loss": 0.1837,
"step": 385
},
{
"epoch": 2.47,
"learning_rate": 3.1477145899250326e-06,
"loss": 0.1463,
"step": 386
},
{
"epoch": 2.47,
"learning_rate": 3.073427161043492e-06,
"loss": 0.2133,
"step": 387
},
{
"epoch": 2.48,
"learning_rate": 2.9999538188161705e-06,
"loss": 0.1345,
"step": 388
},
{
"epoch": 2.49,
"learning_rate": 2.927298096957063e-06,
"loss": 0.2226,
"step": 389
},
{
"epoch": 2.49,
"learning_rate": 2.8554634898565668e-06,
"loss": 0.1664,
"step": 390
},
{
"epoch": 2.5,
"learning_rate": 2.784453452413405e-06,
"loss": 0.1388,
"step": 391
},
{
"epoch": 2.5,
"learning_rate": 2.714271399868473e-06,
"loss": 0.2065,
"step": 392
},
{
"epoch": 2.51,
"learning_rate": 2.6449207076405857e-06,
"loss": 0.1353,
"step": 393
},
{
"epoch": 2.52,
"learning_rate": 2.57640471116412e-06,
"loss": 0.1601,
"step": 394
},
{
"epoch": 2.52,
"learning_rate": 2.508726705728617e-06,
"loss": 0.2478,
"step": 395
},
{
"epoch": 2.53,
"learning_rate": 2.441889946320266e-06,
"loss": 0.1515,
"step": 396
},
{
"epoch": 2.54,
"learning_rate": 2.3758976474653904e-06,
"loss": 0.11,
"step": 397
},
{
"epoch": 2.54,
"learning_rate": 2.310752983075819e-06,
"loss": 0.179,
"step": 398
},
{
"epoch": 2.55,
"learning_rate": 2.2464590862962443e-06,
"loss": 0.1746,
"step": 399
},
{
"epoch": 2.56,
"learning_rate": 2.1830190493535385e-06,
"loss": 0.1632,
"step": 400
}
],
"logging_steps": 1.0,
"max_steps": 468,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 200,
"total_flos": 3.6382398814964637e+21,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}