pharaouk's picture
Training in progress, step 1800, checkpoint
e3693f2
raw
history blame
225 kB
{
"best_metric": 0.3296074867248535,
"best_model_checkpoint": "experts/mistralic-expert-17/checkpoint-1800",
"epoch": 0.3730763251981968,
"eval_steps": 200,
"global_step": 1800,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3764,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3983,
"step": 2
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4823,
"step": 3
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4139,
"step": 4
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.341,
"step": 5
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3615,
"step": 6
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3709,
"step": 7
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3184,
"step": 8
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3923,
"step": 9
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.5588,
"step": 10
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4881,
"step": 11
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4644,
"step": 12
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3299,
"step": 13
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.404,
"step": 14
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3933,
"step": 15
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3784,
"step": 16
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.406,
"step": 17
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4184,
"step": 18
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3953,
"step": 19
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4085,
"step": 20
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4547,
"step": 21
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4166,
"step": 22
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.4566,
"step": 23
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.3928,
"step": 24
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3639,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.386,
"step": 26
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3157,
"step": 27
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3111,
"step": 28
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.5034,
"step": 29
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4071,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6236,
"step": 31
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3294,
"step": 32
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4722,
"step": 33
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3625,
"step": 34
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.5145,
"step": 35
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.416,
"step": 36
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3519,
"step": 37
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3509,
"step": 38
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.416,
"step": 39
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4899,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3193,
"step": 41
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.5121,
"step": 42
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3911,
"step": 43
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4805,
"step": 44
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3315,
"step": 45
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3784,
"step": 46
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4312,
"step": 47
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4413,
"step": 48
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3399,
"step": 49
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3861,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4977,
"step": 51
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.515,
"step": 52
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3766,
"step": 53
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3144,
"step": 54
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3097,
"step": 55
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3277,
"step": 56
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3147,
"step": 57
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4264,
"step": 58
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3957,
"step": 59
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3379,
"step": 60
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3163,
"step": 61
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3506,
"step": 62
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3888,
"step": 63
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.347,
"step": 64
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.522,
"step": 65
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3002,
"step": 66
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.3523,
"step": 67
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6818,
"step": 68
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4075,
"step": 69
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4022,
"step": 70
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4122,
"step": 71
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4788,
"step": 72
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3144,
"step": 73
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4427,
"step": 74
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.453,
"step": 75
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3033,
"step": 76
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.475,
"step": 77
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4291,
"step": 78
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4268,
"step": 79
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3052,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.2165,
"step": 81
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5203,
"step": 82
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3173,
"step": 83
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3227,
"step": 84
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.387,
"step": 85
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3191,
"step": 86
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.2438,
"step": 87
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3296,
"step": 88
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.2889,
"step": 89
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3352,
"step": 90
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3428,
"step": 91
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3576,
"step": 92
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3154,
"step": 93
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4834,
"step": 94
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.2879,
"step": 95
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3734,
"step": 96
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4016,
"step": 97
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4421,
"step": 98
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3338,
"step": 99
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4342,
"step": 100
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3754,
"step": 101
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5515,
"step": 102
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3301,
"step": 103
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3824,
"step": 104
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3394,
"step": 105
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.2945,
"step": 106
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3857,
"step": 107
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3952,
"step": 108
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3497,
"step": 109
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4295,
"step": 110
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3641,
"step": 111
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.2941,
"step": 112
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.4429,
"step": 113
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.2982,
"step": 114
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.407,
"step": 115
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5319,
"step": 116
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3071,
"step": 117
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.353,
"step": 118
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5562,
"step": 119
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3285,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3044,
"step": 121
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3912,
"step": 122
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3579,
"step": 123
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4041,
"step": 124
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4563,
"step": 125
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3469,
"step": 126
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3381,
"step": 127
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3378,
"step": 128
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4174,
"step": 129
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3629,
"step": 130
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3426,
"step": 131
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5591,
"step": 132
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.321,
"step": 133
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3866,
"step": 134
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4362,
"step": 135
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3987,
"step": 136
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4104,
"step": 137
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3393,
"step": 138
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3664,
"step": 139
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2806,
"step": 140
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5358,
"step": 141
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5086,
"step": 142
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3179,
"step": 143
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2922,
"step": 144
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.413,
"step": 145
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3515,
"step": 146
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3137,
"step": 147
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3654,
"step": 148
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5826,
"step": 149
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4276,
"step": 150
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2985,
"step": 151
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3914,
"step": 152
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2938,
"step": 153
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2313,
"step": 154
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4092,
"step": 155
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4689,
"step": 156
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3192,
"step": 157
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3818,
"step": 158
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3851,
"step": 159
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2992,
"step": 160
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3687,
"step": 161
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3448,
"step": 162
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3684,
"step": 163
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2847,
"step": 164
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.3367,
"step": 165
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.2416,
"step": 166
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4329,
"step": 167
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.4203,
"step": 168
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.355,
"step": 169
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3465,
"step": 170
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2676,
"step": 171
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3554,
"step": 172
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2724,
"step": 173
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4885,
"step": 174
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.556,
"step": 175
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.298,
"step": 176
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.286,
"step": 177
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3485,
"step": 178
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3193,
"step": 179
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3288,
"step": 180
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2939,
"step": 181
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3689,
"step": 182
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.324,
"step": 183
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3524,
"step": 184
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3494,
"step": 185
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3723,
"step": 186
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3995,
"step": 187
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4912,
"step": 188
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3704,
"step": 189
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3676,
"step": 190
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4459,
"step": 191
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.336,
"step": 192
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.627,
"step": 193
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5047,
"step": 194
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3175,
"step": 195
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4015,
"step": 196
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3854,
"step": 197
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3298,
"step": 198
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3249,
"step": 199
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3183,
"step": 200
},
{
"epoch": 0.04,
"eval_loss": 0.36924177408218384,
"eval_runtime": 159.4711,
"eval_samples_per_second": 6.271,
"eval_steps_per_second": 3.135,
"step": 200
},
{
"epoch": 0.04,
"mmlu_eval_accuracy": 0.5940031246983312,
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
"mmlu_eval_accuracy_anatomy": 0.35714285714285715,
"mmlu_eval_accuracy_astronomy": 0.75,
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
"mmlu_eval_accuracy_college_biology": 0.5625,
"mmlu_eval_accuracy_college_chemistry": 0.25,
"mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
"mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
"mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
"mmlu_eval_accuracy_conceptual_physics": 0.5,
"mmlu_eval_accuracy_econometrics": 0.6666666666666666,
"mmlu_eval_accuracy_electrical_engineering": 0.625,
"mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
"mmlu_eval_accuracy_global_facts": 0.4,
"mmlu_eval_accuracy_high_school_biology": 0.625,
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
"mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
"mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
"mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
"mmlu_eval_accuracy_human_sexuality": 0.5,
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
"mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
"mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.88,
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
"mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
"mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
"mmlu_eval_accuracy_moral_scenarios": 0.37,
"mmlu_eval_accuracy_nutrition": 0.7878787878787878,
"mmlu_eval_accuracy_philosophy": 0.7058823529411765,
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
"mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
"mmlu_eval_accuracy_professional_law": 0.4,
"mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
"mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.7037037037037037,
"mmlu_eval_accuracy_sociology": 0.9545454545454546,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.5555555555555556,
"mmlu_eval_accuracy_world_religions": 0.8947368421052632,
"mmlu_loss": 1.4302824790888604,
"step": 200
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4524,
"step": 201
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4312,
"step": 202
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2717,
"step": 203
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3977,
"step": 204
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3864,
"step": 205
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2953,
"step": 206
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3302,
"step": 207
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4244,
"step": 208
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4016,
"step": 209
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4256,
"step": 210
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4162,
"step": 211
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.282,
"step": 212
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.3207,
"step": 213
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2938,
"step": 214
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2823,
"step": 215
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2986,
"step": 216
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.2869,
"step": 217
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.317,
"step": 218
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2975,
"step": 219
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2523,
"step": 220
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3691,
"step": 221
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.341,
"step": 222
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2603,
"step": 223
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4282,
"step": 224
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4877,
"step": 225
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2787,
"step": 226
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3901,
"step": 227
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3039,
"step": 228
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4518,
"step": 229
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.325,
"step": 230
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2634,
"step": 231
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3732,
"step": 232
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3162,
"step": 233
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3761,
"step": 234
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2739,
"step": 235
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.344,
"step": 236
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2995,
"step": 237
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2882,
"step": 238
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2372,
"step": 239
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3331,
"step": 240
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6815,
"step": 241
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2672,
"step": 242
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4416,
"step": 243
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2813,
"step": 244
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3574,
"step": 245
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3237,
"step": 246
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2798,
"step": 247
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5538,
"step": 248
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3436,
"step": 249
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3082,
"step": 250
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5797,
"step": 251
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2435,
"step": 252
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4687,
"step": 253
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2921,
"step": 254
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3151,
"step": 255
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2787,
"step": 256
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3544,
"step": 257
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2891,
"step": 258
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4261,
"step": 259
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2829,
"step": 260
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.2342,
"step": 261
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3482,
"step": 262
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3608,
"step": 263
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4406,
"step": 264
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3543,
"step": 265
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4826,
"step": 266
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3967,
"step": 267
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3463,
"step": 268
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.2837,
"step": 269
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3827,
"step": 270
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3583,
"step": 271
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3035,
"step": 272
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3266,
"step": 273
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5502,
"step": 274
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4045,
"step": 275
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.2922,
"step": 276
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4068,
"step": 277
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.32,
"step": 278
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3039,
"step": 279
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5073,
"step": 280
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.357,
"step": 281
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.2757,
"step": 282
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3473,
"step": 283
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4409,
"step": 284
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4095,
"step": 285
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3177,
"step": 286
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5387,
"step": 287
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3647,
"step": 288
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4568,
"step": 289
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4054,
"step": 290
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3302,
"step": 291
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3426,
"step": 292
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4874,
"step": 293
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5772,
"step": 294
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.2785,
"step": 295
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3493,
"step": 296
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3909,
"step": 297
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4654,
"step": 298
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4774,
"step": 299
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.2985,
"step": 300
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4445,
"step": 301
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3078,
"step": 302
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.2923,
"step": 303
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.373,
"step": 304
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3412,
"step": 305
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.2629,
"step": 306
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.258,
"step": 307
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4849,
"step": 308
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4029,
"step": 309
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3758,
"step": 310
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4228,
"step": 311
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3841,
"step": 312
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.3038,
"step": 313
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4246,
"step": 314
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3501,
"step": 315
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3886,
"step": 316
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3705,
"step": 317
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3549,
"step": 318
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3336,
"step": 319
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3522,
"step": 320
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.2667,
"step": 321
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4979,
"step": 322
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3397,
"step": 323
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3273,
"step": 324
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3509,
"step": 325
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3015,
"step": 326
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.2906,
"step": 327
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.2315,
"step": 328
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3104,
"step": 329
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4869,
"step": 330
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4246,
"step": 331
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3485,
"step": 332
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3974,
"step": 333
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4052,
"step": 334
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4124,
"step": 335
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4349,
"step": 336
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3406,
"step": 337
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3532,
"step": 338
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4889,
"step": 339
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.5194,
"step": 340
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.514,
"step": 341
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.327,
"step": 342
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3088,
"step": 343
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3761,
"step": 344
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.2712,
"step": 345
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3488,
"step": 346
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4925,
"step": 347
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3292,
"step": 348
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.282,
"step": 349
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3427,
"step": 350
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3182,
"step": 351
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4053,
"step": 352
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3911,
"step": 353
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.2906,
"step": 354
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4012,
"step": 355
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3397,
"step": 356
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3832,
"step": 357
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3739,
"step": 358
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3106,
"step": 359
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3123,
"step": 360
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.3197,
"step": 361
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3987,
"step": 362
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2746,
"step": 363
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3073,
"step": 364
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3593,
"step": 365
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3287,
"step": 366
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.354,
"step": 367
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3877,
"step": 368
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4417,
"step": 369
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2696,
"step": 370
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3225,
"step": 371
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3691,
"step": 372
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3359,
"step": 373
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2475,
"step": 374
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3173,
"step": 375
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3585,
"step": 376
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2903,
"step": 377
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3593,
"step": 378
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3115,
"step": 379
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5,
"step": 380
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4372,
"step": 381
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.319,
"step": 382
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2573,
"step": 383
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5916,
"step": 384
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4126,
"step": 385
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3318,
"step": 386
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2875,
"step": 387
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2891,
"step": 388
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3617,
"step": 389
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3148,
"step": 390
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4457,
"step": 391
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.407,
"step": 392
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4488,
"step": 393
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4066,
"step": 394
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3808,
"step": 395
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2792,
"step": 396
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4491,
"step": 397
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3646,
"step": 398
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3868,
"step": 399
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2754,
"step": 400
},
{
"epoch": 0.08,
"eval_loss": 0.358918160200119,
"eval_runtime": 159.4642,
"eval_samples_per_second": 6.271,
"eval_steps_per_second": 3.136,
"step": 400
},
{
"epoch": 0.08,
"mmlu_eval_accuracy": 0.5999132037466034,
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
"mmlu_eval_accuracy_anatomy": 0.42857142857142855,
"mmlu_eval_accuracy_astronomy": 0.75,
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
"mmlu_eval_accuracy_college_biology": 0.625,
"mmlu_eval_accuracy_college_chemistry": 0.25,
"mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
"mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
"mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.6363636363636364,
"mmlu_eval_accuracy_conceptual_physics": 0.5,
"mmlu_eval_accuracy_econometrics": 0.6666666666666666,
"mmlu_eval_accuracy_electrical_engineering": 0.5625,
"mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
"mmlu_eval_accuracy_global_facts": 0.4,
"mmlu_eval_accuracy_high_school_biology": 0.59375,
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
"mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
"mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
"mmlu_eval_accuracy_high_school_psychology": 0.85,
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
"mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
"mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
"mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
"mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.88,
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
"mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
"mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
"mmlu_eval_accuracy_moral_scenarios": 0.33,
"mmlu_eval_accuracy_nutrition": 0.7272727272727273,
"mmlu_eval_accuracy_philosophy": 0.7058823529411765,
"mmlu_eval_accuracy_prehistory": 0.42857142857142855,
"mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
"mmlu_eval_accuracy_professional_law": 0.3941176470588235,
"mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
"mmlu_eval_accuracy_professional_psychology": 0.6376811594202898,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.6296296296296297,
"mmlu_eval_accuracy_sociology": 0.9545454545454546,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.5555555555555556,
"mmlu_eval_accuracy_world_religions": 0.8947368421052632,
"mmlu_loss": 1.4201551996774524,
"step": 400
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3131,
"step": 401
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3419,
"step": 402
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3984,
"step": 403
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3459,
"step": 404
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.4061,
"step": 405
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.2641,
"step": 406
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.332,
"step": 407
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3048,
"step": 408
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3402,
"step": 409
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3617,
"step": 410
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4099,
"step": 411
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6116,
"step": 412
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4472,
"step": 413
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2897,
"step": 414
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3139,
"step": 415
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4072,
"step": 416
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3577,
"step": 417
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4119,
"step": 418
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3019,
"step": 419
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5519,
"step": 420
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2923,
"step": 421
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4213,
"step": 422
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3394,
"step": 423
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3897,
"step": 424
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2891,
"step": 425
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.285,
"step": 426
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3758,
"step": 427
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2731,
"step": 428
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.374,
"step": 429
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3646,
"step": 430
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4202,
"step": 431
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2353,
"step": 432
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.383,
"step": 433
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3019,
"step": 434
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5666,
"step": 435
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3783,
"step": 436
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2525,
"step": 437
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3956,
"step": 438
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.395,
"step": 439
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2849,
"step": 440
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3503,
"step": 441
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3204,
"step": 442
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3204,
"step": 443
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3083,
"step": 444
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4108,
"step": 445
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2786,
"step": 446
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5366,
"step": 447
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2785,
"step": 448
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3132,
"step": 449
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4062,
"step": 450
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3812,
"step": 451
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2555,
"step": 452
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4278,
"step": 453
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.2947,
"step": 454
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3247,
"step": 455
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5202,
"step": 456
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3669,
"step": 457
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.3395,
"step": 458
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.358,
"step": 459
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2811,
"step": 460
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4165,
"step": 461
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2591,
"step": 462
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3491,
"step": 463
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.33,
"step": 464
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4262,
"step": 465
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3111,
"step": 466
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4079,
"step": 467
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4417,
"step": 468
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.261,
"step": 469
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2752,
"step": 470
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2719,
"step": 471
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3703,
"step": 472
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2798,
"step": 473
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3553,
"step": 474
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3205,
"step": 475
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3406,
"step": 476
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3759,
"step": 477
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3276,
"step": 478
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3335,
"step": 479
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.5614,
"step": 480
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2922,
"step": 481
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3785,
"step": 482
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3085,
"step": 483
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4565,
"step": 484
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4444,
"step": 485
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2936,
"step": 486
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3858,
"step": 487
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3834,
"step": 488
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3273,
"step": 489
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3334,
"step": 490
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3236,
"step": 491
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2944,
"step": 492
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.5452,
"step": 493
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3507,
"step": 494
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3539,
"step": 495
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2818,
"step": 496
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4009,
"step": 497
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3196,
"step": 498
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.4456,
"step": 499
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.2854,
"step": 500
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.359,
"step": 501
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3794,
"step": 502
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.581,
"step": 503
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3004,
"step": 504
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3796,
"step": 505
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 0.3091,
"step": 506
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4228,
"step": 507
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3734,
"step": 508
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2583,
"step": 509
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3844,
"step": 510
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3349,
"step": 511
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.347,
"step": 512
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2649,
"step": 513
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3511,
"step": 514
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.1902,
"step": 515
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2628,
"step": 516
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.258,
"step": 517
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4467,
"step": 518
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3565,
"step": 519
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3713,
"step": 520
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2867,
"step": 521
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2861,
"step": 522
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2694,
"step": 523
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4202,
"step": 524
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4187,
"step": 525
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4243,
"step": 526
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4473,
"step": 527
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4199,
"step": 528
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3368,
"step": 529
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3606,
"step": 530
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.6544,
"step": 531
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3448,
"step": 532
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3672,
"step": 533
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3245,
"step": 534
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3069,
"step": 535
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.47,
"step": 536
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3799,
"step": 537
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3455,
"step": 538
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4227,
"step": 539
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3065,
"step": 540
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3319,
"step": 541
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.5573,
"step": 542
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3221,
"step": 543
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2999,
"step": 544
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4778,
"step": 545
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.4276,
"step": 546
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3707,
"step": 547
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3303,
"step": 548
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.5229,
"step": 549
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3393,
"step": 550
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3118,
"step": 551
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.2864,
"step": 552
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3364,
"step": 553
},
{
"epoch": 0.11,
"learning_rate": 0.0002,
"loss": 0.3086,
"step": 554
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3338,
"step": 555
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2945,
"step": 556
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3808,
"step": 557
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2725,
"step": 558
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2995,
"step": 559
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3137,
"step": 560
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3209,
"step": 561
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2393,
"step": 562
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3013,
"step": 563
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2315,
"step": 564
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3284,
"step": 565
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3422,
"step": 566
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.4019,
"step": 567
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.5752,
"step": 568
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.317,
"step": 569
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3557,
"step": 570
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3452,
"step": 571
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3375,
"step": 572
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2105,
"step": 573
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.4988,
"step": 574
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2949,
"step": 575
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3991,
"step": 576
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3326,
"step": 577
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3207,
"step": 578
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3678,
"step": 579
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2938,
"step": 580
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3795,
"step": 581
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.4644,
"step": 582
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.4052,
"step": 583
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2996,
"step": 584
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.455,
"step": 585
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.4427,
"step": 586
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3916,
"step": 587
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3169,
"step": 588
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.576,
"step": 589
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.4151,
"step": 590
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2811,
"step": 591
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3179,
"step": 592
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2883,
"step": 593
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.5965,
"step": 594
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3987,
"step": 595
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.4434,
"step": 596
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.5114,
"step": 597
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.2867,
"step": 598
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.255,
"step": 599
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.33,
"step": 600
},
{
"epoch": 0.12,
"eval_loss": 0.3515583574771881,
"eval_runtime": 159.3504,
"eval_samples_per_second": 6.275,
"eval_steps_per_second": 3.138,
"step": 600
},
{
"epoch": 0.12,
"mmlu_eval_accuracy": 0.6095696896582704,
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
"mmlu_eval_accuracy_anatomy": 0.5,
"mmlu_eval_accuracy_astronomy": 0.8125,
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
"mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
"mmlu_eval_accuracy_college_biology": 0.625,
"mmlu_eval_accuracy_college_chemistry": 0.375,
"mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
"mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
"mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
"mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
"mmlu_eval_accuracy_econometrics": 0.6666666666666666,
"mmlu_eval_accuracy_electrical_engineering": 0.5625,
"mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415,
"mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
"mmlu_eval_accuracy_global_facts": 0.4,
"mmlu_eval_accuracy_high_school_biology": 0.625,
"mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
"mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
"mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
"mmlu_eval_accuracy_high_school_psychology": 0.85,
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
"mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
"mmlu_eval_accuracy_human_sexuality": 0.5,
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
"mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
"mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
"mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.92,
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
"mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
"mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
"mmlu_eval_accuracy_moral_scenarios": 0.36,
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
"mmlu_eval_accuracy_philosophy": 0.6764705882352942,
"mmlu_eval_accuracy_prehistory": 0.5428571428571428,
"mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
"mmlu_eval_accuracy_professional_law": 0.4,
"mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
"mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.6666666666666666,
"mmlu_eval_accuracy_sociology": 0.9090909090909091,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.5555555555555556,
"mmlu_eval_accuracy_world_religions": 0.7894736842105263,
"mmlu_loss": 1.383797744780234,
"step": 600
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3336,
"step": 601
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3879,
"step": 602
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 0.3637,
"step": 603
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3342,
"step": 604
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2545,
"step": 605
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.478,
"step": 606
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3672,
"step": 607
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3338,
"step": 608
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2637,
"step": 609
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.5447,
"step": 610
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2509,
"step": 611
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3597,
"step": 612
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3357,
"step": 613
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2906,
"step": 614
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3108,
"step": 615
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4441,
"step": 616
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4973,
"step": 617
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4386,
"step": 618
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2435,
"step": 619
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3177,
"step": 620
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2406,
"step": 621
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.344,
"step": 622
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.519,
"step": 623
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4641,
"step": 624
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3593,
"step": 625
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.442,
"step": 626
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2441,
"step": 627
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3301,
"step": 628
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3552,
"step": 629
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3342,
"step": 630
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3023,
"step": 631
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4696,
"step": 632
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3121,
"step": 633
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3117,
"step": 634
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3381,
"step": 635
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4275,
"step": 636
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.5997,
"step": 637
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3677,
"step": 638
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3019,
"step": 639
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3838,
"step": 640
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.5191,
"step": 641
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2501,
"step": 642
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2363,
"step": 643
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3212,
"step": 644
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4166,
"step": 645
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3708,
"step": 646
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3446,
"step": 647
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.2667,
"step": 648
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3117,
"step": 649
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.4283,
"step": 650
},
{
"epoch": 0.13,
"learning_rate": 0.0002,
"loss": 0.3815,
"step": 651
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3552,
"step": 652
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2827,
"step": 653
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.5081,
"step": 654
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3505,
"step": 655
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3409,
"step": 656
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.4335,
"step": 657
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3517,
"step": 658
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.4015,
"step": 659
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2506,
"step": 660
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2683,
"step": 661
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3261,
"step": 662
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3973,
"step": 663
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3333,
"step": 664
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2956,
"step": 665
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.4357,
"step": 666
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3456,
"step": 667
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2607,
"step": 668
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2917,
"step": 669
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.5605,
"step": 670
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2892,
"step": 671
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.4714,
"step": 672
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3037,
"step": 673
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.4018,
"step": 674
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.29,
"step": 675
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3111,
"step": 676
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3386,
"step": 677
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2688,
"step": 678
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2314,
"step": 679
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2978,
"step": 680
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.4678,
"step": 681
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2993,
"step": 682
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3585,
"step": 683
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2585,
"step": 684
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3749,
"step": 685
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.5372,
"step": 686
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.6269,
"step": 687
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3462,
"step": 688
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3386,
"step": 689
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3011,
"step": 690
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3638,
"step": 691
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.4762,
"step": 692
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.2494,
"step": 693
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.25,
"step": 694
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3209,
"step": 695
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3363,
"step": 696
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.3057,
"step": 697
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.29,
"step": 698
},
{
"epoch": 0.14,
"learning_rate": 0.0002,
"loss": 0.279,
"step": 699
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4144,
"step": 700
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4035,
"step": 701
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4001,
"step": 702
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3917,
"step": 703
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4514,
"step": 704
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.275,
"step": 705
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4256,
"step": 706
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3787,
"step": 707
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4171,
"step": 708
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3542,
"step": 709
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2872,
"step": 710
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3524,
"step": 711
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4246,
"step": 712
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2009,
"step": 713
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3558,
"step": 714
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4853,
"step": 715
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3248,
"step": 716
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3283,
"step": 717
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3036,
"step": 718
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2898,
"step": 719
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2929,
"step": 720
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3195,
"step": 721
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3062,
"step": 722
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3071,
"step": 723
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3382,
"step": 724
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3026,
"step": 725
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2771,
"step": 726
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3467,
"step": 727
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3839,
"step": 728
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2969,
"step": 729
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2924,
"step": 730
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4699,
"step": 731
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.2693,
"step": 732
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3428,
"step": 733
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3306,
"step": 734
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4782,
"step": 735
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4112,
"step": 736
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3568,
"step": 737
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3454,
"step": 738
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3465,
"step": 739
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3941,
"step": 740
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4345,
"step": 741
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.234,
"step": 742
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3432,
"step": 743
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3649,
"step": 744
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.4051,
"step": 745
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3145,
"step": 746
},
{
"epoch": 0.15,
"learning_rate": 0.0002,
"loss": 0.3366,
"step": 747
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2769,
"step": 748
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.438,
"step": 749
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.4599,
"step": 750
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.465,
"step": 751
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2447,
"step": 752
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3375,
"step": 753
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.356,
"step": 754
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3919,
"step": 755
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.4425,
"step": 756
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2936,
"step": 757
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3871,
"step": 758
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3868,
"step": 759
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3073,
"step": 760
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3297,
"step": 761
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3452,
"step": 762
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2743,
"step": 763
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.4263,
"step": 764
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.263,
"step": 765
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.249,
"step": 766
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3263,
"step": 767
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3864,
"step": 768
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3744,
"step": 769
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3379,
"step": 770
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2812,
"step": 771
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2239,
"step": 772
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2832,
"step": 773
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3457,
"step": 774
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3176,
"step": 775
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3036,
"step": 776
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3842,
"step": 777
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2456,
"step": 778
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.373,
"step": 779
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.411,
"step": 780
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2447,
"step": 781
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.262,
"step": 782
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.331,
"step": 783
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2911,
"step": 784
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2906,
"step": 785
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3722,
"step": 786
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2862,
"step": 787
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.4915,
"step": 788
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3048,
"step": 789
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3548,
"step": 790
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2886,
"step": 791
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2657,
"step": 792
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.4349,
"step": 793
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.3265,
"step": 794
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2918,
"step": 795
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 0.2429,
"step": 796
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3048,
"step": 797
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3035,
"step": 798
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3287,
"step": 799
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2949,
"step": 800
},
{
"epoch": 0.17,
"eval_loss": 0.3455464243888855,
"eval_runtime": 159.257,
"eval_samples_per_second": 6.279,
"eval_steps_per_second": 3.14,
"step": 800
},
{
"epoch": 0.17,
"mmlu_eval_accuracy": 0.5986589042634419,
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
"mmlu_eval_accuracy_anatomy": 0.42857142857142855,
"mmlu_eval_accuracy_astronomy": 0.6875,
"mmlu_eval_accuracy_business_ethics": 0.7272727272727273,
"mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
"mmlu_eval_accuracy_college_biology": 0.5625,
"mmlu_eval_accuracy_college_chemistry": 0.25,
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
"mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
"mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
"mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
"mmlu_eval_accuracy_econometrics": 0.5833333333333334,
"mmlu_eval_accuracy_electrical_engineering": 0.75,
"mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415,
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
"mmlu_eval_accuracy_global_facts": 0.4,
"mmlu_eval_accuracy_high_school_biology": 0.5625,
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
"mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
"mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
"mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
"mmlu_eval_accuracy_human_sexuality": 0.5,
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
"mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
"mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.92,
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
"mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
"mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
"mmlu_eval_accuracy_moral_scenarios": 0.36,
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
"mmlu_eval_accuracy_philosophy": 0.7058823529411765,
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
"mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
"mmlu_eval_accuracy_professional_law": 0.4,
"mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
"mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.6666666666666666,
"mmlu_eval_accuracy_sociology": 0.8636363636363636,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.6111111111111112,
"mmlu_eval_accuracy_world_religions": 0.7894736842105263,
"mmlu_loss": 1.5759579020287287,
"step": 800
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4054,
"step": 801
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4718,
"step": 802
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2695,
"step": 803
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3716,
"step": 804
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4109,
"step": 805
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3176,
"step": 806
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4368,
"step": 807
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3341,
"step": 808
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3081,
"step": 809
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2677,
"step": 810
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3201,
"step": 811
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2704,
"step": 812
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3967,
"step": 813
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.8647,
"step": 814
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3318,
"step": 815
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2868,
"step": 816
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4796,
"step": 817
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3111,
"step": 818
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2656,
"step": 819
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2656,
"step": 820
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2475,
"step": 821
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.388,
"step": 822
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2936,
"step": 823
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2784,
"step": 824
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3155,
"step": 825
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.1861,
"step": 826
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2374,
"step": 827
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2553,
"step": 828
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2571,
"step": 829
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3295,
"step": 830
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4078,
"step": 831
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2489,
"step": 832
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2886,
"step": 833
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2767,
"step": 834
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.6366,
"step": 835
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4288,
"step": 836
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2506,
"step": 837
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.4165,
"step": 838
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.2654,
"step": 839
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3846,
"step": 840
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3165,
"step": 841
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3183,
"step": 842
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.329,
"step": 843
},
{
"epoch": 0.17,
"learning_rate": 0.0002,
"loss": 0.3764,
"step": 844
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3789,
"step": 845
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2707,
"step": 846
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.4665,
"step": 847
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3743,
"step": 848
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3822,
"step": 849
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2279,
"step": 850
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3113,
"step": 851
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3191,
"step": 852
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3549,
"step": 853
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3279,
"step": 854
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3831,
"step": 855
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2911,
"step": 856
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.5018,
"step": 857
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2493,
"step": 858
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3465,
"step": 859
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.4005,
"step": 860
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3605,
"step": 861
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.4068,
"step": 862
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3948,
"step": 863
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3638,
"step": 864
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3076,
"step": 865
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3566,
"step": 866
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.4456,
"step": 867
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3633,
"step": 868
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.5587,
"step": 869
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.5333,
"step": 870
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3908,
"step": 871
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2926,
"step": 872
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3232,
"step": 873
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.4943,
"step": 874
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3418,
"step": 875
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3492,
"step": 876
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3138,
"step": 877
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3047,
"step": 878
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2521,
"step": 879
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2705,
"step": 880
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3355,
"step": 881
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3358,
"step": 882
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3186,
"step": 883
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3052,
"step": 884
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2069,
"step": 885
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3336,
"step": 886
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2978,
"step": 887
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2614,
"step": 888
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.2852,
"step": 889
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.324,
"step": 890
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.4048,
"step": 891
},
{
"epoch": 0.18,
"learning_rate": 0.0002,
"loss": 0.3176,
"step": 892
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4053,
"step": 893
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2959,
"step": 894
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3648,
"step": 895
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3507,
"step": 896
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3496,
"step": 897
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2629,
"step": 898
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3048,
"step": 899
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3152,
"step": 900
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3095,
"step": 901
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3185,
"step": 902
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3165,
"step": 903
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3354,
"step": 904
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3616,
"step": 905
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4569,
"step": 906
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2608,
"step": 907
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2492,
"step": 908
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2194,
"step": 909
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3576,
"step": 910
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3949,
"step": 911
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4273,
"step": 912
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3187,
"step": 913
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4265,
"step": 914
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.375,
"step": 915
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4456,
"step": 916
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2682,
"step": 917
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3754,
"step": 918
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3388,
"step": 919
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4208,
"step": 920
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2531,
"step": 921
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3969,
"step": 922
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2425,
"step": 923
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2948,
"step": 924
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3525,
"step": 925
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2286,
"step": 926
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2691,
"step": 927
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2962,
"step": 928
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3169,
"step": 929
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3709,
"step": 930
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2728,
"step": 931
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4006,
"step": 932
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2902,
"step": 933
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3399,
"step": 934
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4129,
"step": 935
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3946,
"step": 936
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.3531,
"step": 937
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.5031,
"step": 938
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.2395,
"step": 939
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 0.4327,
"step": 940
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3431,
"step": 941
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3588,
"step": 942
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2331,
"step": 943
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3002,
"step": 944
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2886,
"step": 945
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3718,
"step": 946
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.4566,
"step": 947
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3216,
"step": 948
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.4049,
"step": 949
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.5468,
"step": 950
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2624,
"step": 951
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3018,
"step": 952
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.328,
"step": 953
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3149,
"step": 954
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.293,
"step": 955
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3949,
"step": 956
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2829,
"step": 957
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.5452,
"step": 958
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3265,
"step": 959
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3454,
"step": 960
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2472,
"step": 961
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3075,
"step": 962
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3953,
"step": 963
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.4237,
"step": 964
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3237,
"step": 965
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.224,
"step": 966
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3196,
"step": 967
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3051,
"step": 968
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2262,
"step": 969
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3367,
"step": 970
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2836,
"step": 971
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2292,
"step": 972
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3596,
"step": 973
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.251,
"step": 974
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3142,
"step": 975
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2804,
"step": 976
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2589,
"step": 977
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2276,
"step": 978
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.247,
"step": 979
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3212,
"step": 980
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2639,
"step": 981
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.4801,
"step": 982
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3661,
"step": 983
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2641,
"step": 984
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.4465,
"step": 985
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2491,
"step": 986
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3758,
"step": 987
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.3068,
"step": 988
},
{
"epoch": 0.2,
"learning_rate": 0.0002,
"loss": 0.2582,
"step": 989
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3422,
"step": 990
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3677,
"step": 991
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2968,
"step": 992
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2564,
"step": 993
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2396,
"step": 994
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.5021,
"step": 995
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3229,
"step": 996
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3049,
"step": 997
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3978,
"step": 998
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.5366,
"step": 999
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2309,
"step": 1000
},
{
"epoch": 0.21,
"eval_loss": 0.34240442514419556,
"eval_runtime": 159.3616,
"eval_samples_per_second": 6.275,
"eval_steps_per_second": 3.138,
"step": 1000
},
{
"epoch": 0.21,
"mmlu_eval_accuracy": 0.6036630168245939,
"mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
"mmlu_eval_accuracy_anatomy": 0.5,
"mmlu_eval_accuracy_astronomy": 0.75,
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
"mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
"mmlu_eval_accuracy_college_biology": 0.5625,
"mmlu_eval_accuracy_college_chemistry": 0.25,
"mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
"mmlu_eval_accuracy_college_mathematics": 0.6363636363636364,
"mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
"mmlu_eval_accuracy_conceptual_physics": 0.6538461538461539,
"mmlu_eval_accuracy_econometrics": 0.6666666666666666,
"mmlu_eval_accuracy_electrical_engineering": 0.625,
"mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
"mmlu_eval_accuracy_global_facts": 0.3,
"mmlu_eval_accuracy_high_school_biology": 0.5625,
"mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
"mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
"mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655,
"mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
"mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
"mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
"mmlu_eval_accuracy_human_sexuality": 0.6666666666666666,
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
"mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
"mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.92,
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
"mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
"mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
"mmlu_eval_accuracy_moral_scenarios": 0.36,
"mmlu_eval_accuracy_nutrition": 0.7272727272727273,
"mmlu_eval_accuracy_philosophy": 0.6764705882352942,
"mmlu_eval_accuracy_prehistory": 0.5714285714285714,
"mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
"mmlu_eval_accuracy_professional_law": 0.40588235294117647,
"mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
"mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.6666666666666666,
"mmlu_eval_accuracy_sociology": 0.9545454545454546,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.6111111111111112,
"mmlu_eval_accuracy_world_religions": 0.8421052631578947,
"mmlu_loss": 1.4610189728964091,
"step": 1000
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.4934,
"step": 1001
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2727,
"step": 1002
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2757,
"step": 1003
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.4271,
"step": 1004
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2999,
"step": 1005
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2883,
"step": 1006
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3823,
"step": 1007
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.4903,
"step": 1008
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3183,
"step": 1009
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.397,
"step": 1010
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3115,
"step": 1011
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.4355,
"step": 1012
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.274,
"step": 1013
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.4141,
"step": 1014
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.418,
"step": 1015
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2875,
"step": 1016
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2975,
"step": 1017
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3519,
"step": 1018
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2645,
"step": 1019
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3193,
"step": 1020
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2065,
"step": 1021
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3574,
"step": 1022
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3843,
"step": 1023
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3478,
"step": 1024
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3781,
"step": 1025
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2993,
"step": 1026
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2651,
"step": 1027
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3909,
"step": 1028
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.5299,
"step": 1029
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.4472,
"step": 1030
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.3763,
"step": 1031
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.4319,
"step": 1032
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2601,
"step": 1033
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2539,
"step": 1034
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2708,
"step": 1035
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2716,
"step": 1036
},
{
"epoch": 0.21,
"learning_rate": 0.0002,
"loss": 0.2905,
"step": 1037
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3437,
"step": 1038
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3777,
"step": 1039
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3246,
"step": 1040
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3998,
"step": 1041
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2726,
"step": 1042
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3591,
"step": 1043
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2694,
"step": 1044
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3669,
"step": 1045
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.4534,
"step": 1046
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2806,
"step": 1047
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.348,
"step": 1048
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3629,
"step": 1049
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2536,
"step": 1050
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3034,
"step": 1051
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2936,
"step": 1052
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3849,
"step": 1053
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3761,
"step": 1054
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.343,
"step": 1055
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2749,
"step": 1056
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.329,
"step": 1057
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3792,
"step": 1058
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3433,
"step": 1059
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2232,
"step": 1060
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.5412,
"step": 1061
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.453,
"step": 1062
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2917,
"step": 1063
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2818,
"step": 1064
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2746,
"step": 1065
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3564,
"step": 1066
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.298,
"step": 1067
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2601,
"step": 1068
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.4126,
"step": 1069
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3852,
"step": 1070
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2738,
"step": 1071
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3528,
"step": 1072
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3714,
"step": 1073
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3275,
"step": 1074
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.436,
"step": 1075
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.2869,
"step": 1076
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3169,
"step": 1077
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3515,
"step": 1078
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.4167,
"step": 1079
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3007,
"step": 1080
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3505,
"step": 1081
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.1987,
"step": 1082
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3311,
"step": 1083
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.3327,
"step": 1084
},
{
"epoch": 0.22,
"learning_rate": 0.0002,
"loss": 0.4333,
"step": 1085
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4852,
"step": 1086
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3692,
"step": 1087
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2804,
"step": 1088
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2757,
"step": 1089
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2836,
"step": 1090
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.36,
"step": 1091
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3251,
"step": 1092
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2676,
"step": 1093
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3695,
"step": 1094
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3958,
"step": 1095
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2812,
"step": 1096
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3103,
"step": 1097
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3681,
"step": 1098
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3071,
"step": 1099
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4287,
"step": 1100
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3983,
"step": 1101
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3019,
"step": 1102
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3475,
"step": 1103
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3823,
"step": 1104
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.371,
"step": 1105
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2103,
"step": 1106
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3499,
"step": 1107
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3728,
"step": 1108
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2749,
"step": 1109
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3678,
"step": 1110
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.444,
"step": 1111
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3431,
"step": 1112
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3941,
"step": 1113
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.335,
"step": 1114
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2307,
"step": 1115
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4715,
"step": 1116
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3797,
"step": 1117
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3209,
"step": 1118
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3698,
"step": 1119
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2383,
"step": 1120
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3897,
"step": 1121
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3363,
"step": 1122
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4683,
"step": 1123
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3432,
"step": 1124
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4679,
"step": 1125
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2691,
"step": 1126
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2982,
"step": 1127
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2467,
"step": 1128
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.2586,
"step": 1129
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.3816,
"step": 1130
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4294,
"step": 1131
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4064,
"step": 1132
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 0.4467,
"step": 1133
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2648,
"step": 1134
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2737,
"step": 1135
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2972,
"step": 1136
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3052,
"step": 1137
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2764,
"step": 1138
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2214,
"step": 1139
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2559,
"step": 1140
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2989,
"step": 1141
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3718,
"step": 1142
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3936,
"step": 1143
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3278,
"step": 1144
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2424,
"step": 1145
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.291,
"step": 1146
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2476,
"step": 1147
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3314,
"step": 1148
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.4345,
"step": 1149
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2737,
"step": 1150
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3019,
"step": 1151
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3984,
"step": 1152
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.4365,
"step": 1153
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.4559,
"step": 1154
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.33,
"step": 1155
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2701,
"step": 1156
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3531,
"step": 1157
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.459,
"step": 1158
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.1893,
"step": 1159
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3033,
"step": 1160
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3243,
"step": 1161
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2844,
"step": 1162
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.4774,
"step": 1163
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.361,
"step": 1164
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.291,
"step": 1165
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2967,
"step": 1166
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3633,
"step": 1167
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2392,
"step": 1168
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3356,
"step": 1169
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2664,
"step": 1170
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2522,
"step": 1171
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3911,
"step": 1172
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3153,
"step": 1173
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3247,
"step": 1174
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3252,
"step": 1175
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3449,
"step": 1176
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2576,
"step": 1177
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2894,
"step": 1178
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3495,
"step": 1179
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.3257,
"step": 1180
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.251,
"step": 1181
},
{
"epoch": 0.24,
"learning_rate": 0.0002,
"loss": 0.2906,
"step": 1182
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.362,
"step": 1183
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3473,
"step": 1184
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.5011,
"step": 1185
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2469,
"step": 1186
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2893,
"step": 1187
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3803,
"step": 1188
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.274,
"step": 1189
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2162,
"step": 1190
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.311,
"step": 1191
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.4521,
"step": 1192
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3156,
"step": 1193
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2632,
"step": 1194
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3344,
"step": 1195
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.5234,
"step": 1196
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.227,
"step": 1197
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3469,
"step": 1198
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3397,
"step": 1199
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2175,
"step": 1200
},
{
"epoch": 0.25,
"eval_loss": 0.3368752598762512,
"eval_runtime": 159.297,
"eval_samples_per_second": 6.278,
"eval_steps_per_second": 3.139,
"step": 1200
},
{
"epoch": 0.25,
"mmlu_eval_accuracy": 0.604796655596056,
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
"mmlu_eval_accuracy_anatomy": 0.42857142857142855,
"mmlu_eval_accuracy_astronomy": 0.8125,
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
"mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
"mmlu_eval_accuracy_college_biology": 0.625,
"mmlu_eval_accuracy_college_chemistry": 0.375,
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
"mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
"mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.6363636363636364,
"mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
"mmlu_eval_accuracy_econometrics": 0.5833333333333334,
"mmlu_eval_accuracy_electrical_engineering": 0.75,
"mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
"mmlu_eval_accuracy_global_facts": 0.3,
"mmlu_eval_accuracy_high_school_biology": 0.625,
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
"mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186,
"mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
"mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
"mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
"mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
"mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
"mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.88,
"mmlu_eval_accuracy_medical_genetics": 1.0,
"mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
"mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
"mmlu_eval_accuracy_moral_scenarios": 0.32,
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
"mmlu_eval_accuracy_philosophy": 0.7058823529411765,
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
"mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
"mmlu_eval_accuracy_professional_law": 0.4,
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
"mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.6666666666666666,
"mmlu_eval_accuracy_sociology": 0.9545454545454546,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.5555555555555556,
"mmlu_eval_accuracy_world_religions": 0.8421052631578947,
"mmlu_loss": 1.4258578755447822,
"step": 1200
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3934,
"step": 1201
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2435,
"step": 1202
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3464,
"step": 1203
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3642,
"step": 1204
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3491,
"step": 1205
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2644,
"step": 1206
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2944,
"step": 1207
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3238,
"step": 1208
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2265,
"step": 1209
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2891,
"step": 1210
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3696,
"step": 1211
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3022,
"step": 1212
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2857,
"step": 1213
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.4168,
"step": 1214
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3136,
"step": 1215
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.363,
"step": 1216
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.345,
"step": 1217
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3945,
"step": 1218
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2759,
"step": 1219
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.297,
"step": 1220
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.237,
"step": 1221
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.6326,
"step": 1222
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2829,
"step": 1223
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.232,
"step": 1224
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.2952,
"step": 1225
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3061,
"step": 1226
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3009,
"step": 1227
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3351,
"step": 1228
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3436,
"step": 1229
},
{
"epoch": 0.25,
"learning_rate": 0.0002,
"loss": 0.3163,
"step": 1230
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3444,
"step": 1231
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2698,
"step": 1232
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.318,
"step": 1233
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2286,
"step": 1234
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.344,
"step": 1235
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3274,
"step": 1236
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2399,
"step": 1237
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.299,
"step": 1238
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.5281,
"step": 1239
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2902,
"step": 1240
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3541,
"step": 1241
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3137,
"step": 1242
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4756,
"step": 1243
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3564,
"step": 1244
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4608,
"step": 1245
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.438,
"step": 1246
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4233,
"step": 1247
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4087,
"step": 1248
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4123,
"step": 1249
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3395,
"step": 1250
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3677,
"step": 1251
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3982,
"step": 1252
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4741,
"step": 1253
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3421,
"step": 1254
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3133,
"step": 1255
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3753,
"step": 1256
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2696,
"step": 1257
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3316,
"step": 1258
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4674,
"step": 1259
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2973,
"step": 1260
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2093,
"step": 1261
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3269,
"step": 1262
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4071,
"step": 1263
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4745,
"step": 1264
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3916,
"step": 1265
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2964,
"step": 1266
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4568,
"step": 1267
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2627,
"step": 1268
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3151,
"step": 1269
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2679,
"step": 1270
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3435,
"step": 1271
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.3807,
"step": 1272
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2732,
"step": 1273
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.353,
"step": 1274
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.404,
"step": 1275
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.304,
"step": 1276
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.4008,
"step": 1277
},
{
"epoch": 0.26,
"learning_rate": 0.0002,
"loss": 0.2743,
"step": 1278
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3276,
"step": 1279
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3674,
"step": 1280
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2979,
"step": 1281
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2657,
"step": 1282
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.314,
"step": 1283
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.293,
"step": 1284
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.4689,
"step": 1285
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2089,
"step": 1286
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3878,
"step": 1287
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3381,
"step": 1288
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.377,
"step": 1289
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.319,
"step": 1290
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3792,
"step": 1291
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2516,
"step": 1292
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.4335,
"step": 1293
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.384,
"step": 1294
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3692,
"step": 1295
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2112,
"step": 1296
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3993,
"step": 1297
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.4006,
"step": 1298
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2382,
"step": 1299
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3246,
"step": 1300
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3233,
"step": 1301
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.4042,
"step": 1302
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.313,
"step": 1303
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2501,
"step": 1304
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2538,
"step": 1305
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3374,
"step": 1306
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2801,
"step": 1307
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.344,
"step": 1308
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.446,
"step": 1309
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.4066,
"step": 1310
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2646,
"step": 1311
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2941,
"step": 1312
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.4165,
"step": 1313
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2693,
"step": 1314
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3053,
"step": 1315
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3058,
"step": 1316
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2666,
"step": 1317
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3697,
"step": 1318
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3374,
"step": 1319
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.4698,
"step": 1320
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2904,
"step": 1321
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3597,
"step": 1322
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2734,
"step": 1323
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.287,
"step": 1324
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.2877,
"step": 1325
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 0.3607,
"step": 1326
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.28,
"step": 1327
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2711,
"step": 1328
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.4512,
"step": 1329
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2502,
"step": 1330
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.4236,
"step": 1331
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.48,
"step": 1332
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.4059,
"step": 1333
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.4075,
"step": 1334
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2909,
"step": 1335
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3353,
"step": 1336
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2889,
"step": 1337
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2324,
"step": 1338
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2577,
"step": 1339
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.309,
"step": 1340
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2902,
"step": 1341
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3583,
"step": 1342
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3089,
"step": 1343
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3162,
"step": 1344
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3177,
"step": 1345
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3074,
"step": 1346
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.337,
"step": 1347
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3847,
"step": 1348
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3894,
"step": 1349
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2633,
"step": 1350
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2988,
"step": 1351
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3381,
"step": 1352
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.331,
"step": 1353
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3434,
"step": 1354
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3924,
"step": 1355
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.4081,
"step": 1356
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.297,
"step": 1357
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.373,
"step": 1358
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.329,
"step": 1359
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3728,
"step": 1360
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3213,
"step": 1361
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3627,
"step": 1362
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3423,
"step": 1363
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2909,
"step": 1364
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2971,
"step": 1365
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.3209,
"step": 1366
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.336,
"step": 1367
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.4604,
"step": 1368
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.6131,
"step": 1369
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2421,
"step": 1370
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2734,
"step": 1371
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.283,
"step": 1372
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.352,
"step": 1373
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.4599,
"step": 1374
},
{
"epoch": 0.28,
"learning_rate": 0.0002,
"loss": 0.2357,
"step": 1375
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2923,
"step": 1376
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2678,
"step": 1377
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.5359,
"step": 1378
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.291,
"step": 1379
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3565,
"step": 1380
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.4448,
"step": 1381
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3475,
"step": 1382
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3464,
"step": 1383
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3087,
"step": 1384
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2698,
"step": 1385
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3618,
"step": 1386
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2829,
"step": 1387
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.272,
"step": 1388
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.305,
"step": 1389
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3789,
"step": 1390
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2664,
"step": 1391
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2509,
"step": 1392
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2936,
"step": 1393
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3184,
"step": 1394
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3389,
"step": 1395
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2824,
"step": 1396
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.343,
"step": 1397
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3114,
"step": 1398
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.4532,
"step": 1399
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3418,
"step": 1400
},
{
"epoch": 0.29,
"eval_loss": 0.33663323521614075,
"eval_runtime": 159.4231,
"eval_samples_per_second": 6.273,
"eval_steps_per_second": 3.136,
"step": 1400
},
{
"epoch": 0.29,
"mmlu_eval_accuracy": 0.5988431625993802,
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
"mmlu_eval_accuracy_anatomy": 0.2857142857142857,
"mmlu_eval_accuracy_astronomy": 0.75,
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
"mmlu_eval_accuracy_college_biology": 0.625,
"mmlu_eval_accuracy_college_chemistry": 0.375,
"mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
"mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
"mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
"mmlu_eval_accuracy_conceptual_physics": 0.6538461538461539,
"mmlu_eval_accuracy_econometrics": 0.6666666666666666,
"mmlu_eval_accuracy_electrical_engineering": 0.5625,
"mmlu_eval_accuracy_elementary_mathematics": 0.5365853658536586,
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
"mmlu_eval_accuracy_global_facts": 0.4,
"mmlu_eval_accuracy_high_school_biology": 0.53125,
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
"mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
"mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
"mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
"mmlu_eval_accuracy_high_school_psychology": 0.85,
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
"mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
"mmlu_eval_accuracy_human_sexuality": 0.5,
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
"mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
"mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.92,
"mmlu_eval_accuracy_medical_genetics": 1.0,
"mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
"mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
"mmlu_eval_accuracy_moral_scenarios": 0.34,
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
"mmlu_eval_accuracy_philosophy": 0.7058823529411765,
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
"mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
"mmlu_eval_accuracy_professional_law": 0.4,
"mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
"mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.5925925925925926,
"mmlu_eval_accuracy_sociology": 0.9090909090909091,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.6111111111111112,
"mmlu_eval_accuracy_world_religions": 0.8421052631578947,
"mmlu_loss": 1.4995008165079993,
"step": 1400
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2955,
"step": 1401
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2544,
"step": 1402
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.4088,
"step": 1403
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2959,
"step": 1404
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.4989,
"step": 1405
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.6567,
"step": 1406
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2983,
"step": 1407
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2839,
"step": 1408
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3097,
"step": 1409
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.4868,
"step": 1410
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.4287,
"step": 1411
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2474,
"step": 1412
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.335,
"step": 1413
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2997,
"step": 1414
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3079,
"step": 1415
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2379,
"step": 1416
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2006,
"step": 1417
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3114,
"step": 1418
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3807,
"step": 1419
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3207,
"step": 1420
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.3256,
"step": 1421
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.2834,
"step": 1422
},
{
"epoch": 0.29,
"learning_rate": 0.0002,
"loss": 0.6884,
"step": 1423
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2604,
"step": 1424
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3474,
"step": 1425
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3066,
"step": 1426
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.4209,
"step": 1427
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.4044,
"step": 1428
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.5416,
"step": 1429
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2346,
"step": 1430
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.4765,
"step": 1431
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.363,
"step": 1432
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3189,
"step": 1433
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2185,
"step": 1434
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3477,
"step": 1435
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2465,
"step": 1436
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.276,
"step": 1437
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3196,
"step": 1438
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.5027,
"step": 1439
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3729,
"step": 1440
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3676,
"step": 1441
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2836,
"step": 1442
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2014,
"step": 1443
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3293,
"step": 1444
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.6591,
"step": 1445
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2323,
"step": 1446
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3745,
"step": 1447
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.7213,
"step": 1448
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3208,
"step": 1449
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2419,
"step": 1450
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3249,
"step": 1451
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2996,
"step": 1452
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3522,
"step": 1453
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3228,
"step": 1454
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3358,
"step": 1455
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.4661,
"step": 1456
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.361,
"step": 1457
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3583,
"step": 1458
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.301,
"step": 1459
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3003,
"step": 1460
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2422,
"step": 1461
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.5315,
"step": 1462
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3418,
"step": 1463
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3769,
"step": 1464
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2861,
"step": 1465
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.2097,
"step": 1466
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.336,
"step": 1467
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3407,
"step": 1468
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3419,
"step": 1469
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3095,
"step": 1470
},
{
"epoch": 0.3,
"learning_rate": 0.0002,
"loss": 0.3398,
"step": 1471
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.283,
"step": 1472
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4463,
"step": 1473
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2983,
"step": 1474
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2905,
"step": 1475
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2503,
"step": 1476
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4403,
"step": 1477
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3506,
"step": 1478
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3436,
"step": 1479
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3225,
"step": 1480
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2525,
"step": 1481
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2677,
"step": 1482
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4132,
"step": 1483
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2935,
"step": 1484
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3177,
"step": 1485
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4628,
"step": 1486
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3345,
"step": 1487
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2949,
"step": 1488
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2881,
"step": 1489
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3309,
"step": 1490
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.358,
"step": 1491
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.291,
"step": 1492
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2234,
"step": 1493
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3209,
"step": 1494
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3667,
"step": 1495
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4149,
"step": 1496
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3423,
"step": 1497
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3355,
"step": 1498
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.326,
"step": 1499
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.301,
"step": 1500
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4943,
"step": 1501
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3528,
"step": 1502
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2953,
"step": 1503
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3267,
"step": 1504
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2319,
"step": 1505
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2877,
"step": 1506
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2859,
"step": 1507
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4401,
"step": 1508
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.4634,
"step": 1509
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.247,
"step": 1510
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.377,
"step": 1511
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2854,
"step": 1512
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.308,
"step": 1513
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3239,
"step": 1514
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3237,
"step": 1515
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3177,
"step": 1516
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2864,
"step": 1517
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.2372,
"step": 1518
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 0.3515,
"step": 1519
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2747,
"step": 1520
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2489,
"step": 1521
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4183,
"step": 1522
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4211,
"step": 1523
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3316,
"step": 1524
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3223,
"step": 1525
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2908,
"step": 1526
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.5003,
"step": 1527
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3043,
"step": 1528
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4197,
"step": 1529
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3874,
"step": 1530
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3501,
"step": 1531
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3674,
"step": 1532
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2134,
"step": 1533
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.364,
"step": 1534
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2458,
"step": 1535
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2149,
"step": 1536
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3037,
"step": 1537
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3675,
"step": 1538
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2494,
"step": 1539
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4273,
"step": 1540
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3962,
"step": 1541
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3332,
"step": 1542
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3336,
"step": 1543
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3131,
"step": 1544
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3025,
"step": 1545
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2949,
"step": 1546
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4005,
"step": 1547
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.258,
"step": 1548
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2922,
"step": 1549
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2556,
"step": 1550
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2101,
"step": 1551
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2162,
"step": 1552
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.337,
"step": 1553
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3086,
"step": 1554
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4988,
"step": 1555
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3734,
"step": 1556
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4444,
"step": 1557
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.4434,
"step": 1558
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2584,
"step": 1559
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.1929,
"step": 1560
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3251,
"step": 1561
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.209,
"step": 1562
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2978,
"step": 1563
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2986,
"step": 1564
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.41,
"step": 1565
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.394,
"step": 1566
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.2986,
"step": 1567
},
{
"epoch": 0.32,
"learning_rate": 0.0002,
"loss": 0.3496,
"step": 1568
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3498,
"step": 1569
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3109,
"step": 1570
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2075,
"step": 1571
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3613,
"step": 1572
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2628,
"step": 1573
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3275,
"step": 1574
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2889,
"step": 1575
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3382,
"step": 1576
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2797,
"step": 1577
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.5451,
"step": 1578
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3645,
"step": 1579
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.264,
"step": 1580
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3261,
"step": 1581
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2957,
"step": 1582
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2267,
"step": 1583
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2626,
"step": 1584
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3648,
"step": 1585
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.5535,
"step": 1586
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3266,
"step": 1587
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3017,
"step": 1588
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2288,
"step": 1589
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3663,
"step": 1590
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2932,
"step": 1591
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3949,
"step": 1592
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2377,
"step": 1593
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3316,
"step": 1594
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3649,
"step": 1595
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3073,
"step": 1596
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3401,
"step": 1597
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3836,
"step": 1598
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2566,
"step": 1599
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.4909,
"step": 1600
},
{
"epoch": 0.33,
"eval_loss": 0.3317873179912567,
"eval_runtime": 159.3941,
"eval_samples_per_second": 6.274,
"eval_steps_per_second": 3.137,
"step": 1600
},
{
"epoch": 0.33,
"mmlu_eval_accuracy": 0.604104053743682,
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
"mmlu_eval_accuracy_anatomy": 0.42857142857142855,
"mmlu_eval_accuracy_astronomy": 0.8125,
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
"mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
"mmlu_eval_accuracy_college_biology": 0.5625,
"mmlu_eval_accuracy_college_chemistry": 0.375,
"mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
"mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
"mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
"mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
"mmlu_eval_accuracy_econometrics": 0.5,
"mmlu_eval_accuracy_electrical_engineering": 0.6875,
"mmlu_eval_accuracy_elementary_mathematics": 0.5609756097560976,
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
"mmlu_eval_accuracy_global_facts": 0.4,
"mmlu_eval_accuracy_high_school_biology": 0.53125,
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
"mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186,
"mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
"mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
"mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
"mmlu_eval_accuracy_human_sexuality": 0.5,
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
"mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
"mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.84,
"mmlu_eval_accuracy_medical_genetics": 1.0,
"mmlu_eval_accuracy_miscellaneous": 0.7790697674418605,
"mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
"mmlu_eval_accuracy_moral_scenarios": 0.33,
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
"mmlu_eval_accuracy_philosophy": 0.7352941176470589,
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
"mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
"mmlu_eval_accuracy_professional_law": 0.40588235294117647,
"mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
"mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.5925925925925926,
"mmlu_eval_accuracy_sociology": 0.9545454545454546,
"mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
"mmlu_eval_accuracy_virology": 0.6666666666666666,
"mmlu_eval_accuracy_world_religions": 0.8421052631578947,
"mmlu_loss": 1.4852339224622393,
"step": 1600
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3334,
"step": 1601
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.4838,
"step": 1602
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3516,
"step": 1603
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.25,
"step": 1604
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.49,
"step": 1605
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2733,
"step": 1606
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2783,
"step": 1607
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2539,
"step": 1608
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3256,
"step": 1609
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3972,
"step": 1610
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.4074,
"step": 1611
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.379,
"step": 1612
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3495,
"step": 1613
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.3087,
"step": 1614
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2345,
"step": 1615
},
{
"epoch": 0.33,
"learning_rate": 0.0002,
"loss": 0.2718,
"step": 1616
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3027,
"step": 1617
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2855,
"step": 1618
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3173,
"step": 1619
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2171,
"step": 1620
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2817,
"step": 1621
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3155,
"step": 1622
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.4401,
"step": 1623
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.4339,
"step": 1624
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3175,
"step": 1625
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2457,
"step": 1626
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.269,
"step": 1627
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3708,
"step": 1628
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2432,
"step": 1629
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2381,
"step": 1630
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3507,
"step": 1631
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2882,
"step": 1632
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2381,
"step": 1633
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3296,
"step": 1634
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3083,
"step": 1635
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.4051,
"step": 1636
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3996,
"step": 1637
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2851,
"step": 1638
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3441,
"step": 1639
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.4172,
"step": 1640
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3062,
"step": 1641
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3192,
"step": 1642
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2048,
"step": 1643
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3372,
"step": 1644
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2578,
"step": 1645
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3007,
"step": 1646
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3709,
"step": 1647
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3193,
"step": 1648
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3146,
"step": 1649
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2601,
"step": 1650
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.4825,
"step": 1651
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2891,
"step": 1652
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2325,
"step": 1653
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3162,
"step": 1654
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2901,
"step": 1655
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.4135,
"step": 1656
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3579,
"step": 1657
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2688,
"step": 1658
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2458,
"step": 1659
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.4185,
"step": 1660
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3543,
"step": 1661
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3159,
"step": 1662
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.2966,
"step": 1663
},
{
"epoch": 0.34,
"learning_rate": 0.0002,
"loss": 0.3614,
"step": 1664
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2745,
"step": 1665
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3058,
"step": 1666
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.4681,
"step": 1667
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.328,
"step": 1668
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2062,
"step": 1669
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.4278,
"step": 1670
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.4948,
"step": 1671
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.278,
"step": 1672
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3436,
"step": 1673
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3287,
"step": 1674
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3692,
"step": 1675
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.5099,
"step": 1676
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3628,
"step": 1677
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2621,
"step": 1678
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3419,
"step": 1679
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.32,
"step": 1680
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.233,
"step": 1681
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2551,
"step": 1682
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2684,
"step": 1683
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2937,
"step": 1684
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3269,
"step": 1685
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2774,
"step": 1686
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.345,
"step": 1687
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3396,
"step": 1688
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2436,
"step": 1689
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.329,
"step": 1690
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2881,
"step": 1691
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.4038,
"step": 1692
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2867,
"step": 1693
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2699,
"step": 1694
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.5154,
"step": 1695
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3967,
"step": 1696
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2934,
"step": 1697
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2944,
"step": 1698
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2675,
"step": 1699
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2601,
"step": 1700
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.347,
"step": 1701
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.4217,
"step": 1702
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2517,
"step": 1703
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.4265,
"step": 1704
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.4762,
"step": 1705
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.442,
"step": 1706
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3528,
"step": 1707
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2042,
"step": 1708
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3589,
"step": 1709
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3871,
"step": 1710
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.3317,
"step": 1711
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 0.2544,
"step": 1712
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.4393,
"step": 1713
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2329,
"step": 1714
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.5222,
"step": 1715
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.388,
"step": 1716
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3759,
"step": 1717
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.4002,
"step": 1718
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3111,
"step": 1719
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.4526,
"step": 1720
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3783,
"step": 1721
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.4324,
"step": 1722
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3174,
"step": 1723
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2132,
"step": 1724
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2844,
"step": 1725
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2983,
"step": 1726
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2927,
"step": 1727
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2667,
"step": 1728
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2995,
"step": 1729
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3512,
"step": 1730
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2147,
"step": 1731
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2851,
"step": 1732
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2828,
"step": 1733
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3263,
"step": 1734
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3379,
"step": 1735
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3372,
"step": 1736
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2135,
"step": 1737
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3782,
"step": 1738
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3953,
"step": 1739
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2299,
"step": 1740
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2322,
"step": 1741
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3818,
"step": 1742
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3403,
"step": 1743
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3873,
"step": 1744
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2827,
"step": 1745
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2546,
"step": 1746
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3017,
"step": 1747
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.4125,
"step": 1748
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2431,
"step": 1749
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3023,
"step": 1750
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.5773,
"step": 1751
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3329,
"step": 1752
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3149,
"step": 1753
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.257,
"step": 1754
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2559,
"step": 1755
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2261,
"step": 1756
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.285,
"step": 1757
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2734,
"step": 1758
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.296,
"step": 1759
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.2676,
"step": 1760
},
{
"epoch": 0.36,
"learning_rate": 0.0002,
"loss": 0.3249,
"step": 1761
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2878,
"step": 1762
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2918,
"step": 1763
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2361,
"step": 1764
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2812,
"step": 1765
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.4694,
"step": 1766
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3372,
"step": 1767
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.371,
"step": 1768
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.4062,
"step": 1769
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.4353,
"step": 1770
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.4231,
"step": 1771
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3154,
"step": 1772
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3907,
"step": 1773
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3659,
"step": 1774
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2288,
"step": 1775
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3506,
"step": 1776
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2445,
"step": 1777
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3384,
"step": 1778
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3252,
"step": 1779
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3783,
"step": 1780
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2369,
"step": 1781
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2717,
"step": 1782
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2256,
"step": 1783
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2701,
"step": 1784
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2927,
"step": 1785
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3099,
"step": 1786
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3354,
"step": 1787
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.6063,
"step": 1788
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3382,
"step": 1789
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.375,
"step": 1790
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2494,
"step": 1791
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3243,
"step": 1792
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3112,
"step": 1793
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3974,
"step": 1794
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3529,
"step": 1795
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.2518,
"step": 1796
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.6142,
"step": 1797
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.266,
"step": 1798
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.3405,
"step": 1799
},
{
"epoch": 0.37,
"learning_rate": 0.0002,
"loss": 0.4709,
"step": 1800
},
{
"epoch": 0.37,
"eval_loss": 0.3296074867248535,
"eval_runtime": 159.4124,
"eval_samples_per_second": 6.273,
"eval_steps_per_second": 3.137,
"step": 1800
},
{
"epoch": 0.37,
"mmlu_eval_accuracy": 0.5929397997737007,
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
"mmlu_eval_accuracy_anatomy": 0.5,
"mmlu_eval_accuracy_astronomy": 0.6875,
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
"mmlu_eval_accuracy_college_biology": 0.5,
"mmlu_eval_accuracy_college_chemistry": 0.375,
"mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
"mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
"mmlu_eval_accuracy_college_medicine": 0.7272727272727273,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
"mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
"mmlu_eval_accuracy_econometrics": 0.5833333333333334,
"mmlu_eval_accuracy_electrical_engineering": 0.6875,
"mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
"mmlu_eval_accuracy_global_facts": 0.3,
"mmlu_eval_accuracy_high_school_biology": 0.53125,
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
"mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
"mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
"mmlu_eval_accuracy_high_school_physics": 0.058823529411764705,
"mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
"mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
"mmlu_eval_accuracy_human_sexuality": 0.5,
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
"mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
"mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.88,
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
"mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
"mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
"mmlu_eval_accuracy_moral_scenarios": 0.37,
"mmlu_eval_accuracy_nutrition": 0.7272727272727273,
"mmlu_eval_accuracy_philosophy": 0.7352941176470589,
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
"mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
"mmlu_eval_accuracy_professional_law": 0.40588235294117647,
"mmlu_eval_accuracy_professional_medicine": 0.7096774193548387,
"mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.5925925925925926,
"mmlu_eval_accuracy_sociology": 0.8636363636363636,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.5,
"mmlu_eval_accuracy_world_religions": 0.8947368421052632,
"mmlu_loss": 1.290066049864647,
"step": 1800
}
],
"logging_steps": 1,
"max_steps": 14472,
"num_train_epochs": 3,
"save_steps": 200,
"total_flos": 4.478916039396557e+17,
"trial_name": null,
"trial_params": null
}