hahafofo's picture
init
1f5430c
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.962406015037594,
"eval_steps": 500,
"global_step": 330,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 0.0,
"loss": 3.6699,
"step": 1
},
{
"epoch": 0.03,
"learning_rate": 5e-06,
"loss": 3.616,
"step": 2
},
{
"epoch": 0.05,
"learning_rate": 7.924812503605782e-06,
"loss": 3.6394,
"step": 3
},
{
"epoch": 0.06,
"learning_rate": 1e-05,
"loss": 3.5342,
"step": 4
},
{
"epoch": 0.08,
"learning_rate": 1e-05,
"loss": 3.3909,
"step": 5
},
{
"epoch": 0.09,
"learning_rate": 1e-05,
"loss": 3.0278,
"step": 6
},
{
"epoch": 0.11,
"learning_rate": 1e-05,
"loss": 3.0232,
"step": 7
},
{
"epoch": 0.12,
"learning_rate": 1e-05,
"loss": 2.7886,
"step": 8
},
{
"epoch": 0.14,
"learning_rate": 1e-05,
"loss": 2.7587,
"step": 9
},
{
"epoch": 0.15,
"learning_rate": 1e-05,
"loss": 2.6985,
"step": 10
},
{
"epoch": 0.17,
"learning_rate": 1e-05,
"loss": 2.4939,
"step": 11
},
{
"epoch": 0.18,
"learning_rate": 1e-05,
"loss": 2.4019,
"step": 12
},
{
"epoch": 0.2,
"learning_rate": 1e-05,
"loss": 2.5276,
"step": 13
},
{
"epoch": 0.21,
"learning_rate": 1e-05,
"loss": 2.479,
"step": 14
},
{
"epoch": 0.23,
"learning_rate": 1e-05,
"loss": 2.2505,
"step": 15
},
{
"epoch": 0.24,
"learning_rate": 1e-05,
"loss": 2.2395,
"step": 16
},
{
"epoch": 0.26,
"learning_rate": 1e-05,
"loss": 2.0909,
"step": 17
},
{
"epoch": 0.27,
"learning_rate": 1e-05,
"loss": 2.3071,
"step": 18
},
{
"epoch": 0.29,
"learning_rate": 1e-05,
"loss": 2.222,
"step": 19
},
{
"epoch": 0.3,
"learning_rate": 1e-05,
"loss": 2.4785,
"step": 20
},
{
"epoch": 0.32,
"learning_rate": 1e-05,
"loss": 2.2537,
"step": 21
},
{
"epoch": 0.33,
"learning_rate": 1e-05,
"loss": 2.244,
"step": 22
},
{
"epoch": 0.35,
"learning_rate": 1e-05,
"loss": 2.408,
"step": 23
},
{
"epoch": 0.36,
"learning_rate": 1e-05,
"loss": 1.9875,
"step": 24
},
{
"epoch": 0.38,
"learning_rate": 1e-05,
"loss": 2.3292,
"step": 25
},
{
"epoch": 0.39,
"learning_rate": 1e-05,
"loss": 2.1417,
"step": 26
},
{
"epoch": 0.41,
"learning_rate": 1e-05,
"loss": 2.165,
"step": 27
},
{
"epoch": 0.42,
"learning_rate": 1e-05,
"loss": 2.0793,
"step": 28
},
{
"epoch": 0.44,
"learning_rate": 1e-05,
"loss": 2.3461,
"step": 29
},
{
"epoch": 0.45,
"learning_rate": 1e-05,
"loss": 2.1355,
"step": 30
},
{
"epoch": 0.47,
"learning_rate": 1e-05,
"loss": 1.8263,
"step": 31
},
{
"epoch": 0.48,
"learning_rate": 1e-05,
"loss": 2.0223,
"step": 32
},
{
"epoch": 0.5,
"learning_rate": 1e-05,
"loss": 2.2253,
"step": 33
},
{
"epoch": 0.51,
"learning_rate": 1e-05,
"loss": 2.3689,
"step": 34
},
{
"epoch": 0.53,
"learning_rate": 1e-05,
"loss": 2.0426,
"step": 35
},
{
"epoch": 0.54,
"learning_rate": 1e-05,
"loss": 2.0056,
"step": 36
},
{
"epoch": 0.56,
"learning_rate": 1e-05,
"loss": 2.0055,
"step": 37
},
{
"epoch": 0.57,
"learning_rate": 1e-05,
"loss": 2.0377,
"step": 38
},
{
"epoch": 0.59,
"learning_rate": 1e-05,
"loss": 2.1296,
"step": 39
},
{
"epoch": 0.6,
"learning_rate": 1e-05,
"loss": 1.9558,
"step": 40
},
{
"epoch": 0.62,
"learning_rate": 1e-05,
"loss": 1.8363,
"step": 41
},
{
"epoch": 0.63,
"learning_rate": 1e-05,
"loss": 2.1641,
"step": 42
},
{
"epoch": 0.65,
"learning_rate": 1e-05,
"loss": 1.8467,
"step": 43
},
{
"epoch": 0.66,
"learning_rate": 1e-05,
"loss": 1.9818,
"step": 44
},
{
"epoch": 0.68,
"learning_rate": 1e-05,
"loss": 2.2552,
"step": 45
},
{
"epoch": 0.69,
"learning_rate": 1e-05,
"loss": 1.7173,
"step": 46
},
{
"epoch": 0.71,
"learning_rate": 1e-05,
"loss": 1.9606,
"step": 47
},
{
"epoch": 0.72,
"learning_rate": 1e-05,
"loss": 2.1582,
"step": 48
},
{
"epoch": 0.74,
"learning_rate": 1e-05,
"loss": 1.9778,
"step": 49
},
{
"epoch": 0.75,
"learning_rate": 1e-05,
"loss": 1.7875,
"step": 50
},
{
"epoch": 0.77,
"learning_rate": 1e-05,
"loss": 1.7636,
"step": 51
},
{
"epoch": 0.78,
"learning_rate": 1e-05,
"loss": 1.6118,
"step": 52
},
{
"epoch": 0.8,
"learning_rate": 1e-05,
"loss": 1.886,
"step": 53
},
{
"epoch": 0.81,
"learning_rate": 1e-05,
"loss": 1.699,
"step": 54
},
{
"epoch": 0.83,
"learning_rate": 1e-05,
"loss": 1.6429,
"step": 55
},
{
"epoch": 0.84,
"learning_rate": 1e-05,
"loss": 1.9658,
"step": 56
},
{
"epoch": 0.86,
"learning_rate": 1e-05,
"loss": 1.6691,
"step": 57
},
{
"epoch": 0.87,
"learning_rate": 1e-05,
"loss": 1.6262,
"step": 58
},
{
"epoch": 0.89,
"learning_rate": 1e-05,
"loss": 1.9305,
"step": 59
},
{
"epoch": 0.9,
"learning_rate": 1e-05,
"loss": 1.7902,
"step": 60
},
{
"epoch": 0.92,
"learning_rate": 1e-05,
"loss": 1.789,
"step": 61
},
{
"epoch": 0.93,
"learning_rate": 1e-05,
"loss": 1.9741,
"step": 62
},
{
"epoch": 0.95,
"learning_rate": 1e-05,
"loss": 1.7437,
"step": 63
},
{
"epoch": 0.96,
"learning_rate": 1e-05,
"loss": 1.8843,
"step": 64
},
{
"epoch": 0.98,
"learning_rate": 1e-05,
"loss": 1.852,
"step": 65
},
{
"epoch": 0.99,
"learning_rate": 1e-05,
"loss": 1.8561,
"step": 66
},
{
"epoch": 1.01,
"learning_rate": 1e-05,
"loss": 1.7507,
"step": 67
},
{
"epoch": 1.02,
"learning_rate": 1e-05,
"loss": 1.3836,
"step": 68
},
{
"epoch": 1.04,
"learning_rate": 1e-05,
"loss": 1.5647,
"step": 69
},
{
"epoch": 1.05,
"learning_rate": 1e-05,
"loss": 1.6693,
"step": 70
},
{
"epoch": 1.07,
"learning_rate": 1e-05,
"loss": 1.4477,
"step": 71
},
{
"epoch": 1.08,
"learning_rate": 1e-05,
"loss": 1.6273,
"step": 72
},
{
"epoch": 1.1,
"learning_rate": 1e-05,
"loss": 1.7846,
"step": 73
},
{
"epoch": 1.11,
"learning_rate": 1e-05,
"loss": 1.563,
"step": 74
},
{
"epoch": 1.13,
"learning_rate": 1e-05,
"loss": 1.4053,
"step": 75
},
{
"epoch": 1.14,
"learning_rate": 1e-05,
"loss": 1.7382,
"step": 76
},
{
"epoch": 1.16,
"learning_rate": 1e-05,
"loss": 1.4874,
"step": 77
},
{
"epoch": 1.17,
"learning_rate": 1e-05,
"loss": 1.5523,
"step": 78
},
{
"epoch": 1.19,
"learning_rate": 1e-05,
"loss": 1.4858,
"step": 79
},
{
"epoch": 1.2,
"learning_rate": 1e-05,
"loss": 1.705,
"step": 80
},
{
"epoch": 1.22,
"learning_rate": 1e-05,
"loss": 1.3546,
"step": 81
},
{
"epoch": 1.23,
"learning_rate": 1e-05,
"loss": 1.2475,
"step": 82
},
{
"epoch": 1.25,
"learning_rate": 1e-05,
"loss": 1.4983,
"step": 83
},
{
"epoch": 1.26,
"learning_rate": 1e-05,
"loss": 1.4156,
"step": 84
},
{
"epoch": 1.28,
"learning_rate": 1e-05,
"loss": 1.3414,
"step": 85
},
{
"epoch": 1.29,
"learning_rate": 1e-05,
"loss": 1.5126,
"step": 86
},
{
"epoch": 1.31,
"learning_rate": 1e-05,
"loss": 1.3543,
"step": 87
},
{
"epoch": 1.32,
"learning_rate": 1e-05,
"loss": 1.5546,
"step": 88
},
{
"epoch": 1.34,
"learning_rate": 1e-05,
"loss": 1.5876,
"step": 89
},
{
"epoch": 1.35,
"learning_rate": 1e-05,
"loss": 1.4072,
"step": 90
},
{
"epoch": 1.37,
"learning_rate": 1e-05,
"loss": 1.3553,
"step": 91
},
{
"epoch": 1.38,
"learning_rate": 1e-05,
"loss": 1.5311,
"step": 92
},
{
"epoch": 1.4,
"learning_rate": 1e-05,
"loss": 1.454,
"step": 93
},
{
"epoch": 1.41,
"learning_rate": 1e-05,
"loss": 1.39,
"step": 94
},
{
"epoch": 1.43,
"learning_rate": 1e-05,
"loss": 1.4627,
"step": 95
},
{
"epoch": 1.44,
"learning_rate": 1e-05,
"loss": 1.5155,
"step": 96
},
{
"epoch": 1.46,
"learning_rate": 1e-05,
"loss": 1.3862,
"step": 97
},
{
"epoch": 1.47,
"learning_rate": 1e-05,
"loss": 1.1501,
"step": 98
},
{
"epoch": 1.49,
"learning_rate": 1e-05,
"loss": 1.3959,
"step": 99
},
{
"epoch": 1.5,
"learning_rate": 1e-05,
"loss": 1.6432,
"step": 100
},
{
"epoch": 1.52,
"learning_rate": 1e-05,
"loss": 1.4406,
"step": 101
},
{
"epoch": 1.53,
"learning_rate": 1e-05,
"loss": 1.52,
"step": 102
},
{
"epoch": 1.55,
"learning_rate": 1e-05,
"loss": 1.4255,
"step": 103
},
{
"epoch": 1.56,
"learning_rate": 1e-05,
"loss": 1.2903,
"step": 104
},
{
"epoch": 1.58,
"learning_rate": 1e-05,
"loss": 1.5818,
"step": 105
},
{
"epoch": 1.59,
"learning_rate": 1e-05,
"loss": 1.386,
"step": 106
},
{
"epoch": 1.61,
"learning_rate": 1e-05,
"loss": 1.283,
"step": 107
},
{
"epoch": 1.62,
"learning_rate": 1e-05,
"loss": 1.3798,
"step": 108
},
{
"epoch": 1.64,
"learning_rate": 1e-05,
"loss": 1.4538,
"step": 109
},
{
"epoch": 1.65,
"learning_rate": 1e-05,
"loss": 1.2159,
"step": 110
},
{
"epoch": 1.67,
"learning_rate": 1e-05,
"loss": 1.4725,
"step": 111
},
{
"epoch": 1.68,
"learning_rate": 1e-05,
"loss": 1.381,
"step": 112
},
{
"epoch": 1.7,
"learning_rate": 1e-05,
"loss": 1.252,
"step": 113
},
{
"epoch": 1.71,
"learning_rate": 1e-05,
"loss": 1.341,
"step": 114
},
{
"epoch": 1.73,
"learning_rate": 1e-05,
"loss": 1.5052,
"step": 115
},
{
"epoch": 1.74,
"learning_rate": 1e-05,
"loss": 1.1877,
"step": 116
},
{
"epoch": 1.76,
"learning_rate": 1e-05,
"loss": 1.1438,
"step": 117
},
{
"epoch": 1.77,
"learning_rate": 1e-05,
"loss": 1.3111,
"step": 118
},
{
"epoch": 1.79,
"learning_rate": 1e-05,
"loss": 1.136,
"step": 119
},
{
"epoch": 1.8,
"learning_rate": 1e-05,
"loss": 1.0645,
"step": 120
},
{
"epoch": 1.82,
"learning_rate": 1e-05,
"loss": 1.2853,
"step": 121
},
{
"epoch": 1.83,
"learning_rate": 1e-05,
"loss": 1.1764,
"step": 122
},
{
"epoch": 1.85,
"learning_rate": 1e-05,
"loss": 1.2228,
"step": 123
},
{
"epoch": 1.86,
"learning_rate": 1e-05,
"loss": 0.9635,
"step": 124
},
{
"epoch": 1.88,
"learning_rate": 1e-05,
"loss": 1.3114,
"step": 125
},
{
"epoch": 1.89,
"learning_rate": 1e-05,
"loss": 1.2582,
"step": 126
},
{
"epoch": 1.91,
"learning_rate": 1e-05,
"loss": 1.0929,
"step": 127
},
{
"epoch": 1.92,
"learning_rate": 1e-05,
"loss": 1.3374,
"step": 128
},
{
"epoch": 1.94,
"learning_rate": 1e-05,
"loss": 1.2309,
"step": 129
},
{
"epoch": 1.95,
"learning_rate": 1e-05,
"loss": 1.2563,
"step": 130
},
{
"epoch": 1.97,
"learning_rate": 1e-05,
"loss": 1.2559,
"step": 131
},
{
"epoch": 1.98,
"learning_rate": 1e-05,
"loss": 1.2581,
"step": 132
},
{
"epoch": 2.0,
"learning_rate": 1e-05,
"loss": 1.2812,
"step": 133
},
{
"epoch": 2.02,
"learning_rate": 1e-05,
"loss": 1.0479,
"step": 134
},
{
"epoch": 2.03,
"learning_rate": 1e-05,
"loss": 1.0381,
"step": 135
},
{
"epoch": 2.05,
"learning_rate": 1e-05,
"loss": 1.1208,
"step": 136
},
{
"epoch": 2.06,
"learning_rate": 1e-05,
"loss": 1.0136,
"step": 137
},
{
"epoch": 2.08,
"learning_rate": 1e-05,
"loss": 1.0657,
"step": 138
},
{
"epoch": 2.09,
"learning_rate": 1e-05,
"loss": 1.1409,
"step": 139
},
{
"epoch": 2.11,
"learning_rate": 1e-05,
"loss": 1.0947,
"step": 140
},
{
"epoch": 2.12,
"learning_rate": 1e-05,
"loss": 1.1044,
"step": 141
},
{
"epoch": 2.14,
"learning_rate": 1e-05,
"loss": 0.962,
"step": 142
},
{
"epoch": 2.15,
"learning_rate": 1e-05,
"loss": 1.0793,
"step": 143
},
{
"epoch": 2.17,
"learning_rate": 1e-05,
"loss": 1.0441,
"step": 144
},
{
"epoch": 2.18,
"learning_rate": 1e-05,
"loss": 0.884,
"step": 145
},
{
"epoch": 2.2,
"learning_rate": 1e-05,
"loss": 1.1797,
"step": 146
},
{
"epoch": 2.21,
"learning_rate": 1e-05,
"loss": 1.0734,
"step": 147
},
{
"epoch": 2.23,
"learning_rate": 1e-05,
"loss": 0.8103,
"step": 148
},
{
"epoch": 2.24,
"learning_rate": 1e-05,
"loss": 0.8786,
"step": 149
},
{
"epoch": 2.26,
"learning_rate": 1e-05,
"loss": 0.8878,
"step": 150
},
{
"epoch": 2.27,
"learning_rate": 1e-05,
"loss": 0.9835,
"step": 151
},
{
"epoch": 2.29,
"learning_rate": 1e-05,
"loss": 0.832,
"step": 152
},
{
"epoch": 2.3,
"learning_rate": 1e-05,
"loss": 1.0483,
"step": 153
},
{
"epoch": 2.32,
"learning_rate": 1e-05,
"loss": 0.7673,
"step": 154
},
{
"epoch": 2.33,
"learning_rate": 1e-05,
"loss": 0.9498,
"step": 155
},
{
"epoch": 2.35,
"learning_rate": 1e-05,
"loss": 1.0992,
"step": 156
},
{
"epoch": 2.36,
"learning_rate": 1e-05,
"loss": 0.7458,
"step": 157
},
{
"epoch": 2.38,
"learning_rate": 1e-05,
"loss": 0.9574,
"step": 158
},
{
"epoch": 2.39,
"learning_rate": 1e-05,
"loss": 0.9847,
"step": 159
},
{
"epoch": 2.41,
"learning_rate": 1e-05,
"loss": 0.963,
"step": 160
},
{
"epoch": 2.42,
"learning_rate": 1e-05,
"loss": 0.8669,
"step": 161
},
{
"epoch": 2.44,
"learning_rate": 1e-05,
"loss": 1.0586,
"step": 162
},
{
"epoch": 2.45,
"learning_rate": 1e-05,
"loss": 0.859,
"step": 163
},
{
"epoch": 2.47,
"learning_rate": 1e-05,
"loss": 0.7095,
"step": 164
},
{
"epoch": 2.48,
"learning_rate": 1e-05,
"loss": 0.7627,
"step": 165
},
{
"epoch": 2.5,
"learning_rate": 1e-05,
"loss": 1.0254,
"step": 166
},
{
"epoch": 2.51,
"learning_rate": 1e-05,
"loss": 1.1962,
"step": 167
},
{
"epoch": 2.53,
"learning_rate": 1e-05,
"loss": 0.7622,
"step": 168
},
{
"epoch": 2.54,
"learning_rate": 1e-05,
"loss": 0.991,
"step": 169
},
{
"epoch": 2.56,
"learning_rate": 1e-05,
"loss": 0.9023,
"step": 170
},
{
"epoch": 2.57,
"learning_rate": 1e-05,
"loss": 0.9691,
"step": 171
},
{
"epoch": 2.59,
"learning_rate": 1e-05,
"loss": 0.9405,
"step": 172
},
{
"epoch": 2.6,
"learning_rate": 1e-05,
"loss": 0.851,
"step": 173
},
{
"epoch": 2.62,
"learning_rate": 1e-05,
"loss": 0.7875,
"step": 174
},
{
"epoch": 2.63,
"learning_rate": 1e-05,
"loss": 1.0168,
"step": 175
},
{
"epoch": 2.65,
"learning_rate": 1e-05,
"loss": 0.7197,
"step": 176
},
{
"epoch": 2.66,
"learning_rate": 1e-05,
"loss": 0.6833,
"step": 177
},
{
"epoch": 2.68,
"learning_rate": 1e-05,
"loss": 0.9674,
"step": 178
},
{
"epoch": 2.69,
"learning_rate": 1e-05,
"loss": 0.6438,
"step": 179
},
{
"epoch": 2.71,
"learning_rate": 1e-05,
"loss": 0.82,
"step": 180
},
{
"epoch": 2.72,
"learning_rate": 1e-05,
"loss": 0.9368,
"step": 181
},
{
"epoch": 2.74,
"learning_rate": 1e-05,
"loss": 0.9289,
"step": 182
},
{
"epoch": 2.75,
"learning_rate": 1e-05,
"loss": 0.6869,
"step": 183
},
{
"epoch": 2.77,
"learning_rate": 1e-05,
"loss": 0.7579,
"step": 184
},
{
"epoch": 2.78,
"learning_rate": 1e-05,
"loss": 0.6716,
"step": 185
},
{
"epoch": 2.8,
"learning_rate": 1e-05,
"loss": 0.6981,
"step": 186
},
{
"epoch": 2.81,
"learning_rate": 1e-05,
"loss": 0.6319,
"step": 187
},
{
"epoch": 2.83,
"learning_rate": 1e-05,
"loss": 0.659,
"step": 188
},
{
"epoch": 2.84,
"learning_rate": 1e-05,
"loss": 0.7576,
"step": 189
},
{
"epoch": 2.86,
"learning_rate": 1e-05,
"loss": 0.5602,
"step": 190
},
{
"epoch": 2.87,
"learning_rate": 1e-05,
"loss": 0.5778,
"step": 191
},
{
"epoch": 2.89,
"learning_rate": 1e-05,
"loss": 0.8159,
"step": 192
},
{
"epoch": 2.9,
"learning_rate": 1e-05,
"loss": 0.6415,
"step": 193
},
{
"epoch": 2.92,
"learning_rate": 1e-05,
"loss": 0.7226,
"step": 194
},
{
"epoch": 2.93,
"learning_rate": 1e-05,
"loss": 0.7825,
"step": 195
},
{
"epoch": 2.95,
"learning_rate": 1e-05,
"loss": 0.6911,
"step": 196
},
{
"epoch": 2.96,
"learning_rate": 1e-05,
"loss": 0.7701,
"step": 197
},
{
"epoch": 2.98,
"learning_rate": 1e-05,
"loss": 0.6469,
"step": 198
},
{
"epoch": 2.99,
"learning_rate": 1e-05,
"loss": 0.7353,
"step": 199
},
{
"epoch": 3.01,
"learning_rate": 1e-05,
"loss": 0.7989,
"step": 200
},
{
"epoch": 3.02,
"learning_rate": 1e-05,
"loss": 0.6068,
"step": 201
},
{
"epoch": 3.04,
"learning_rate": 1e-05,
"loss": 0.6215,
"step": 202
},
{
"epoch": 3.05,
"learning_rate": 1e-05,
"loss": 0.6768,
"step": 203
},
{
"epoch": 3.07,
"learning_rate": 1e-05,
"loss": 0.4917,
"step": 204
},
{
"epoch": 3.08,
"learning_rate": 1e-05,
"loss": 0.6764,
"step": 205
},
{
"epoch": 3.1,
"learning_rate": 1e-05,
"loss": 0.727,
"step": 206
},
{
"epoch": 3.11,
"learning_rate": 1e-05,
"loss": 0.5708,
"step": 207
},
{
"epoch": 3.13,
"learning_rate": 1e-05,
"loss": 0.4929,
"step": 208
},
{
"epoch": 3.14,
"learning_rate": 1e-05,
"loss": 0.6027,
"step": 209
},
{
"epoch": 3.16,
"learning_rate": 1e-05,
"loss": 0.4757,
"step": 210
},
{
"epoch": 3.17,
"learning_rate": 1e-05,
"loss": 0.5414,
"step": 211
},
{
"epoch": 3.19,
"learning_rate": 1e-05,
"loss": 0.5502,
"step": 212
},
{
"epoch": 3.2,
"learning_rate": 1e-05,
"loss": 0.6625,
"step": 213
},
{
"epoch": 3.22,
"learning_rate": 1e-05,
"loss": 0.4143,
"step": 214
},
{
"epoch": 3.23,
"learning_rate": 1e-05,
"loss": 0.3563,
"step": 215
},
{
"epoch": 3.25,
"learning_rate": 1e-05,
"loss": 0.5016,
"step": 216
},
{
"epoch": 3.26,
"learning_rate": 1e-05,
"loss": 0.5261,
"step": 217
},
{
"epoch": 3.28,
"learning_rate": 1e-05,
"loss": 0.4548,
"step": 218
},
{
"epoch": 3.29,
"learning_rate": 1e-05,
"loss": 0.5206,
"step": 219
},
{
"epoch": 3.31,
"learning_rate": 1e-05,
"loss": 0.4218,
"step": 220
},
{
"epoch": 3.32,
"learning_rate": 1e-05,
"loss": 0.5183,
"step": 221
},
{
"epoch": 3.34,
"learning_rate": 1e-05,
"loss": 0.5362,
"step": 222
},
{
"epoch": 3.35,
"learning_rate": 1e-05,
"loss": 0.4651,
"step": 223
},
{
"epoch": 3.37,
"learning_rate": 1e-05,
"loss": 0.4265,
"step": 224
},
{
"epoch": 3.38,
"learning_rate": 1e-05,
"loss": 0.4581,
"step": 225
},
{
"epoch": 3.4,
"learning_rate": 1e-05,
"loss": 0.5511,
"step": 226
},
{
"epoch": 3.41,
"learning_rate": 1e-05,
"loss": 0.4235,
"step": 227
},
{
"epoch": 3.43,
"learning_rate": 1e-05,
"loss": 0.5037,
"step": 228
},
{
"epoch": 3.44,
"learning_rate": 1e-05,
"loss": 0.4882,
"step": 229
},
{
"epoch": 3.46,
"learning_rate": 1e-05,
"loss": 0.4317,
"step": 230
},
{
"epoch": 3.47,
"learning_rate": 1e-05,
"loss": 0.3565,
"step": 231
},
{
"epoch": 3.49,
"learning_rate": 1e-05,
"loss": 0.4123,
"step": 232
},
{
"epoch": 3.5,
"learning_rate": 1e-05,
"loss": 0.7997,
"step": 233
},
{
"epoch": 3.52,
"learning_rate": 1e-05,
"loss": 0.5753,
"step": 234
},
{
"epoch": 3.53,
"learning_rate": 1e-05,
"loss": 0.5398,
"step": 235
},
{
"epoch": 3.55,
"learning_rate": 1e-05,
"loss": 0.5202,
"step": 236
},
{
"epoch": 3.56,
"learning_rate": 1e-05,
"loss": 0.4584,
"step": 237
},
{
"epoch": 3.58,
"learning_rate": 1e-05,
"loss": 0.6103,
"step": 238
},
{
"epoch": 3.59,
"learning_rate": 1e-05,
"loss": 0.4005,
"step": 239
},
{
"epoch": 3.61,
"learning_rate": 1e-05,
"loss": 0.435,
"step": 240
},
{
"epoch": 3.62,
"learning_rate": 1e-05,
"loss": 0.4691,
"step": 241
},
{
"epoch": 3.64,
"learning_rate": 1e-05,
"loss": 0.4488,
"step": 242
},
{
"epoch": 3.65,
"learning_rate": 1e-05,
"loss": 0.3904,
"step": 243
},
{
"epoch": 3.67,
"learning_rate": 1e-05,
"loss": 0.4746,
"step": 244
},
{
"epoch": 3.68,
"learning_rate": 1e-05,
"loss": 0.4092,
"step": 245
},
{
"epoch": 3.7,
"learning_rate": 1e-05,
"loss": 0.3671,
"step": 246
},
{
"epoch": 3.71,
"learning_rate": 1e-05,
"loss": 0.391,
"step": 247
},
{
"epoch": 3.73,
"learning_rate": 1e-05,
"loss": 0.5338,
"step": 248
},
{
"epoch": 3.74,
"learning_rate": 1e-05,
"loss": 0.3056,
"step": 249
},
{
"epoch": 3.76,
"learning_rate": 1e-05,
"loss": 0.3119,
"step": 250
},
{
"epoch": 3.77,
"learning_rate": 1e-05,
"loss": 0.5096,
"step": 251
},
{
"epoch": 3.79,
"learning_rate": 1e-05,
"loss": 0.3941,
"step": 252
},
{
"epoch": 3.8,
"learning_rate": 1e-05,
"loss": 0.4121,
"step": 253
},
{
"epoch": 3.82,
"learning_rate": 1e-05,
"loss": 0.4883,
"step": 254
},
{
"epoch": 3.83,
"learning_rate": 1e-05,
"loss": 0.4144,
"step": 255
},
{
"epoch": 3.85,
"learning_rate": 1e-05,
"loss": 0.3773,
"step": 256
},
{
"epoch": 3.86,
"learning_rate": 1e-05,
"loss": 0.2575,
"step": 257
},
{
"epoch": 3.88,
"learning_rate": 1e-05,
"loss": 0.412,
"step": 258
},
{
"epoch": 3.89,
"learning_rate": 1e-05,
"loss": 0.3506,
"step": 259
},
{
"epoch": 3.91,
"learning_rate": 1e-05,
"loss": 0.3076,
"step": 260
},
{
"epoch": 3.92,
"learning_rate": 1e-05,
"loss": 0.4445,
"step": 261
},
{
"epoch": 3.94,
"learning_rate": 1e-05,
"loss": 0.3698,
"step": 262
},
{
"epoch": 3.95,
"learning_rate": 1e-05,
"loss": 0.3619,
"step": 263
},
{
"epoch": 3.97,
"learning_rate": 1e-05,
"loss": 0.3515,
"step": 264
},
{
"epoch": 3.98,
"learning_rate": 1e-05,
"loss": 0.3284,
"step": 265
},
{
"epoch": 4.0,
"learning_rate": 1e-05,
"loss": 0.4578,
"step": 266
},
{
"epoch": 4.02,
"learning_rate": 1e-05,
"loss": 0.3942,
"step": 267
},
{
"epoch": 4.03,
"learning_rate": 1e-05,
"loss": 0.3376,
"step": 268
},
{
"epoch": 4.05,
"learning_rate": 1e-05,
"loss": 0.337,
"step": 269
},
{
"epoch": 4.06,
"learning_rate": 1e-05,
"loss": 0.2748,
"step": 270
},
{
"epoch": 4.08,
"learning_rate": 1e-05,
"loss": 0.3154,
"step": 271
},
{
"epoch": 4.09,
"learning_rate": 1e-05,
"loss": 0.3655,
"step": 272
},
{
"epoch": 4.11,
"learning_rate": 1e-05,
"loss": 0.3308,
"step": 273
},
{
"epoch": 4.12,
"learning_rate": 1e-05,
"loss": 0.2619,
"step": 274
},
{
"epoch": 4.14,
"learning_rate": 1e-05,
"loss": 0.2412,
"step": 275
},
{
"epoch": 4.15,
"learning_rate": 1e-05,
"loss": 0.2608,
"step": 276
},
{
"epoch": 4.17,
"learning_rate": 1e-05,
"loss": 0.2722,
"step": 277
},
{
"epoch": 4.18,
"learning_rate": 1e-05,
"loss": 0.2229,
"step": 278
},
{
"epoch": 4.2,
"learning_rate": 1e-05,
"loss": 0.3556,
"step": 279
},
{
"epoch": 4.21,
"learning_rate": 1e-05,
"loss": 0.2753,
"step": 280
},
{
"epoch": 4.23,
"learning_rate": 1e-05,
"loss": 0.1647,
"step": 281
},
{
"epoch": 4.24,
"learning_rate": 1e-05,
"loss": 0.181,
"step": 282
},
{
"epoch": 4.26,
"learning_rate": 1e-05,
"loss": 0.2137,
"step": 283
},
{
"epoch": 4.27,
"learning_rate": 1e-05,
"loss": 0.2369,
"step": 284
},
{
"epoch": 4.29,
"learning_rate": 1e-05,
"loss": 0.1902,
"step": 285
},
{
"epoch": 4.3,
"learning_rate": 1e-05,
"loss": 0.2406,
"step": 286
},
{
"epoch": 4.32,
"learning_rate": 1e-05,
"loss": 0.1385,
"step": 287
},
{
"epoch": 4.33,
"learning_rate": 1e-05,
"loss": 0.2464,
"step": 288
},
{
"epoch": 4.35,
"learning_rate": 1e-05,
"loss": 0.3149,
"step": 289
},
{
"epoch": 4.36,
"learning_rate": 1e-05,
"loss": 0.1562,
"step": 290
},
{
"epoch": 4.38,
"learning_rate": 1e-05,
"loss": 0.2204,
"step": 291
},
{
"epoch": 4.39,
"learning_rate": 1e-05,
"loss": 0.2969,
"step": 292
},
{
"epoch": 4.41,
"learning_rate": 1e-05,
"loss": 0.2168,
"step": 293
},
{
"epoch": 4.42,
"learning_rate": 1e-05,
"loss": 0.2189,
"step": 294
},
{
"epoch": 4.44,
"learning_rate": 1e-05,
"loss": 0.254,
"step": 295
},
{
"epoch": 4.45,
"learning_rate": 1e-05,
"loss": 0.1854,
"step": 296
},
{
"epoch": 4.47,
"learning_rate": 1e-05,
"loss": 0.182,
"step": 297
},
{
"epoch": 4.48,
"learning_rate": 1e-05,
"loss": 0.1822,
"step": 298
},
{
"epoch": 4.5,
"learning_rate": 1e-05,
"loss": 0.3223,
"step": 299
},
{
"epoch": 4.51,
"learning_rate": 1e-05,
"loss": 0.409,
"step": 300
},
{
"epoch": 4.53,
"learning_rate": 1e-05,
"loss": 0.1711,
"step": 301
},
{
"epoch": 4.54,
"learning_rate": 1e-05,
"loss": 0.2508,
"step": 302
},
{
"epoch": 4.56,
"learning_rate": 1e-05,
"loss": 0.2287,
"step": 303
},
{
"epoch": 4.57,
"learning_rate": 1e-05,
"loss": 0.2624,
"step": 304
},
{
"epoch": 4.59,
"learning_rate": 1e-05,
"loss": 0.2577,
"step": 305
},
{
"epoch": 4.6,
"learning_rate": 1e-05,
"loss": 0.2291,
"step": 306
},
{
"epoch": 4.62,
"learning_rate": 1e-05,
"loss": 0.2088,
"step": 307
},
{
"epoch": 4.63,
"learning_rate": 1e-05,
"loss": 0.282,
"step": 308
},
{
"epoch": 4.65,
"learning_rate": 1e-05,
"loss": 0.2005,
"step": 309
},
{
"epoch": 4.66,
"learning_rate": 1e-05,
"loss": 0.1633,
"step": 310
},
{
"epoch": 4.68,
"learning_rate": 1e-05,
"loss": 0.2802,
"step": 311
},
{
"epoch": 4.69,
"learning_rate": 1e-05,
"loss": 0.1332,
"step": 312
},
{
"epoch": 4.71,
"learning_rate": 1e-05,
"loss": 0.2053,
"step": 313
},
{
"epoch": 4.72,
"learning_rate": 1e-05,
"loss": 0.2388,
"step": 314
},
{
"epoch": 4.74,
"learning_rate": 1e-05,
"loss": 0.2342,
"step": 315
},
{
"epoch": 4.75,
"learning_rate": 1e-05,
"loss": 0.1563,
"step": 316
},
{
"epoch": 4.77,
"learning_rate": 1e-05,
"loss": 0.2479,
"step": 317
},
{
"epoch": 4.78,
"learning_rate": 1e-05,
"loss": 0.2365,
"step": 318
},
{
"epoch": 4.8,
"learning_rate": 1e-05,
"loss": 0.2418,
"step": 319
},
{
"epoch": 4.81,
"learning_rate": 1e-05,
"loss": 0.2198,
"step": 320
},
{
"epoch": 4.83,
"learning_rate": 1e-05,
"loss": 0.193,
"step": 321
},
{
"epoch": 4.84,
"learning_rate": 1e-05,
"loss": 0.2195,
"step": 322
},
{
"epoch": 4.86,
"learning_rate": 1e-05,
"loss": 0.1273,
"step": 323
},
{
"epoch": 4.87,
"learning_rate": 1e-05,
"loss": 0.1448,
"step": 324
},
{
"epoch": 4.89,
"learning_rate": 1e-05,
"loss": 0.1977,
"step": 325
},
{
"epoch": 4.9,
"learning_rate": 1e-05,
"loss": 0.149,
"step": 326
},
{
"epoch": 4.92,
"learning_rate": 1e-05,
"loss": 0.2235,
"step": 327
},
{
"epoch": 4.93,
"learning_rate": 1e-05,
"loss": 0.2269,
"step": 328
},
{
"epoch": 4.95,
"learning_rate": 1e-05,
"loss": 0.1771,
"step": 329
},
{
"epoch": 4.96,
"learning_rate": 1e-05,
"loss": 0.2369,
"step": 330
},
{
"epoch": 4.96,
"step": 330,
"total_flos": 8037448286208.0,
"train_loss": 1.0460747798283896,
"train_runtime": 22298.7817,
"train_samples_per_second": 0.477,
"train_steps_per_second": 0.015
}
],
"logging_steps": 1.0,
"max_steps": 330,
"num_train_epochs": 5,
"save_steps": 1000,
"total_flos": 8037448286208.0,
"trial_name": null,
"trial_params": null
}