mistralic-expert-16 / checkpoint-600 /trainer_state.json
pharaouk's picture
Training in progress, step 600, checkpoint
5dfcbe5
raw
history blame
75 kB
{
"best_metric": 0.7626925110816956,
"best_model_checkpoint": "experts/mistralic-expert-16/checkpoint-600",
"epoch": 0.09506079930288747,
"eval_steps": 200,
"global_step": 600,
"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": 0.6637,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.8401,
"step": 2
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 1.0213,
"step": 3
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.8495,
"step": 4
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.6978,
"step": 5
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.5753,
"step": 6
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 1.0964,
"step": 7
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.7564,
"step": 8
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.7727,
"step": 9
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.628,
"step": 10
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.6582,
"step": 11
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.503,
"step": 12
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.7947,
"step": 13
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.7498,
"step": 14
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.8223,
"step": 15
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.6472,
"step": 16
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.5838,
"step": 17
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.5898,
"step": 18
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.8151,
"step": 19
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.9615,
"step": 20
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.5262,
"step": 21
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.6605,
"step": 22
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.8336,
"step": 23
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.6619,
"step": 24
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.5939,
"step": 25
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.6649,
"step": 26
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 2.0444,
"step": 27
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 1.0032,
"step": 28
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.7444,
"step": 29
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.6473,
"step": 30
},
{
"epoch": 0.0,
"learning_rate": 0.0002,
"loss": 0.559,
"step": 31
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8879,
"step": 32
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7444,
"step": 33
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 1.0212,
"step": 34
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7099,
"step": 35
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8135,
"step": 36
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7646,
"step": 37
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6463,
"step": 38
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.624,
"step": 39
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8538,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8128,
"step": 41
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6866,
"step": 42
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8001,
"step": 43
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8757,
"step": 44
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8471,
"step": 45
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7411,
"step": 46
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.9945,
"step": 47
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.9166,
"step": 48
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.9938,
"step": 49
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.5628,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 1.0387,
"step": 51
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8389,
"step": 52
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6579,
"step": 53
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8494,
"step": 54
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 1.035,
"step": 55
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 1.1137,
"step": 56
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7357,
"step": 57
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8016,
"step": 58
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7981,
"step": 59
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4606,
"step": 60
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6416,
"step": 61
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7766,
"step": 62
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6493,
"step": 63
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8368,
"step": 64
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.5657,
"step": 65
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7935,
"step": 66
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8966,
"step": 67
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6217,
"step": 68
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8815,
"step": 69
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 1.0261,
"step": 70
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8171,
"step": 71
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.9208,
"step": 72
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.808,
"step": 73
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8998,
"step": 74
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8616,
"step": 75
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.923,
"step": 76
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6681,
"step": 77
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7372,
"step": 78
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7878,
"step": 79
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.4749,
"step": 80
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7724,
"step": 81
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6716,
"step": 82
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 1.0098,
"step": 83
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.5414,
"step": 84
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8,
"step": 85
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6042,
"step": 86
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7772,
"step": 87
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7977,
"step": 88
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8397,
"step": 89
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.6203,
"step": 90
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.8338,
"step": 91
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.7896,
"step": 92
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.9499,
"step": 93
},
{
"epoch": 0.01,
"learning_rate": 0.0002,
"loss": 0.5947,
"step": 94
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.8759,
"step": 95
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6693,
"step": 96
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6887,
"step": 97
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7044,
"step": 98
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 1.0049,
"step": 99
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.795,
"step": 100
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 1.0949,
"step": 101
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.8985,
"step": 102
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5345,
"step": 103
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6864,
"step": 104
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.9974,
"step": 105
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5856,
"step": 106
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7511,
"step": 107
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5102,
"step": 108
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7603,
"step": 109
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.8859,
"step": 110
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6276,
"step": 111
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.9779,
"step": 112
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.992,
"step": 113
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5208,
"step": 114
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7567,
"step": 115
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5836,
"step": 116
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6989,
"step": 117
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.598,
"step": 118
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7416,
"step": 119
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6019,
"step": 120
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.9242,
"step": 121
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7589,
"step": 122
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.3751,
"step": 123
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6616,
"step": 124
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6024,
"step": 125
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7027,
"step": 126
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.9824,
"step": 127
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5769,
"step": 128
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6134,
"step": 129
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5904,
"step": 130
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7515,
"step": 131
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6977,
"step": 132
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6833,
"step": 133
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7729,
"step": 134
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7972,
"step": 135
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7763,
"step": 136
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.8101,
"step": 137
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.976,
"step": 138
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6232,
"step": 139
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 1.5322,
"step": 140
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.8825,
"step": 141
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7715,
"step": 142
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7375,
"step": 143
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.9355,
"step": 144
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 1.147,
"step": 145
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.8728,
"step": 146
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.8857,
"step": 147
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.5422,
"step": 148
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.6288,
"step": 149
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7528,
"step": 150
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7872,
"step": 151
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7407,
"step": 152
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.9563,
"step": 153
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 1.0431,
"step": 154
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 1.035,
"step": 155
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.7889,
"step": 156
},
{
"epoch": 0.02,
"learning_rate": 0.0002,
"loss": 0.9344,
"step": 157
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.796,
"step": 158
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7938,
"step": 159
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5437,
"step": 160
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.9359,
"step": 161
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.87,
"step": 162
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6749,
"step": 163
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8073,
"step": 164
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8534,
"step": 165
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8366,
"step": 166
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 1.0977,
"step": 167
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6232,
"step": 168
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8216,
"step": 169
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5867,
"step": 170
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 1.0379,
"step": 171
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8962,
"step": 172
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7714,
"step": 173
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7251,
"step": 174
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6247,
"step": 175
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5575,
"step": 176
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7877,
"step": 177
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 1.0638,
"step": 178
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8963,
"step": 179
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6579,
"step": 180
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7132,
"step": 181
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7348,
"step": 182
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5733,
"step": 183
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7421,
"step": 184
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8896,
"step": 185
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8471,
"step": 186
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.9158,
"step": 187
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5107,
"step": 188
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.9068,
"step": 189
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 1.0322,
"step": 190
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8791,
"step": 191
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5369,
"step": 192
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5376,
"step": 193
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.9083,
"step": 194
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8257,
"step": 195
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8235,
"step": 196
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5283,
"step": 197
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.656,
"step": 198
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7968,
"step": 199
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.536,
"step": 200
},
{
"epoch": 0.03,
"eval_loss": 0.7664799094200134,
"eval_runtime": 154.3432,
"eval_samples_per_second": 6.479,
"eval_steps_per_second": 3.24,
"step": 200
},
{
"epoch": 0.03,
"mmlu_eval_accuracy": 0.5963691469709991,
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
"mmlu_eval_accuracy_anatomy": 0.5,
"mmlu_eval_accuracy_astronomy": 0.75,
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
"mmlu_eval_accuracy_college_biology": 0.5625,
"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": 0.5,
"mmlu_eval_accuracy_electrical_engineering": 0.5625,
"mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
"mmlu_eval_accuracy_global_facts": 0.3,
"mmlu_eval_accuracy_high_school_biology": 0.5,
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
"mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
"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": 0.5,
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
"mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
"mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
"mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
"mmlu_eval_accuracy_management": 0.9090909090909091,
"mmlu_eval_accuracy_marketing": 0.92,
"mmlu_eval_accuracy_medical_genetics": 1.0,
"mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
"mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
"mmlu_eval_accuracy_moral_scenarios": 0.31,
"mmlu_eval_accuracy_nutrition": 0.7878787878787878,
"mmlu_eval_accuracy_philosophy": 0.6764705882352942,
"mmlu_eval_accuracy_prehistory": 0.6,
"mmlu_eval_accuracy_professional_accounting": 0.6774193548387096,
"mmlu_eval_accuracy_professional_law": 0.37058823529411766,
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
"mmlu_eval_accuracy_professional_psychology": 0.6231884057971014,
"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": 0.4444444444444444,
"mmlu_eval_accuracy_world_religions": 0.8421052631578947,
"mmlu_loss": 1.2411628218839437,
"step": 200
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.9439,
"step": 201
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8041,
"step": 202
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7885,
"step": 203
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.5351,
"step": 204
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6244,
"step": 205
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6268,
"step": 206
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7006,
"step": 207
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.757,
"step": 208
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7654,
"step": 209
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6594,
"step": 210
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.9099,
"step": 211
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.7043,
"step": 212
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8959,
"step": 213
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.756,
"step": 214
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8411,
"step": 215
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.567,
"step": 216
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8157,
"step": 217
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6843,
"step": 218
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.8001,
"step": 219
},
{
"epoch": 0.03,
"learning_rate": 0.0002,
"loss": 0.6045,
"step": 220
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5119,
"step": 221
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7622,
"step": 222
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7242,
"step": 223
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5367,
"step": 224
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.106,
"step": 225
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7645,
"step": 226
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7407,
"step": 227
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8071,
"step": 228
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.759,
"step": 229
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.0373,
"step": 230
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.886,
"step": 231
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5928,
"step": 232
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.663,
"step": 233
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7304,
"step": 234
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8002,
"step": 235
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.554,
"step": 236
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8972,
"step": 237
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.2181,
"step": 238
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7627,
"step": 239
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.6485,
"step": 240
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8756,
"step": 241
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5099,
"step": 242
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7772,
"step": 243
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7654,
"step": 244
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.6795,
"step": 245
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.709,
"step": 246
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.0877,
"step": 247
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7239,
"step": 248
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.946,
"step": 249
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.838,
"step": 250
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7542,
"step": 251
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8244,
"step": 252
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.0065,
"step": 253
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7358,
"step": 254
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.6672,
"step": 255
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.6232,
"step": 256
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8619,
"step": 257
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7068,
"step": 258
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8178,
"step": 259
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5931,
"step": 260
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8097,
"step": 261
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8769,
"step": 262
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.0777,
"step": 263
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.1692,
"step": 264
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5373,
"step": 265
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.229,
"step": 266
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.6988,
"step": 267
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.865,
"step": 268
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7268,
"step": 269
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.9461,
"step": 270
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8048,
"step": 271
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.6674,
"step": 272
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7445,
"step": 273
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8691,
"step": 274
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.9511,
"step": 275
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8585,
"step": 276
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.9054,
"step": 277
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.8688,
"step": 278
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 1.1202,
"step": 279
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7605,
"step": 280
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.5843,
"step": 281
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.693,
"step": 282
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.4731,
"step": 283
},
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 0.7056,
"step": 284
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8016,
"step": 285
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8471,
"step": 286
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5018,
"step": 287
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7771,
"step": 288
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5368,
"step": 289
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6716,
"step": 290
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.9226,
"step": 291
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6514,
"step": 292
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6389,
"step": 293
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 1.0019,
"step": 294
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.9027,
"step": 295
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8403,
"step": 296
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7016,
"step": 297
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5971,
"step": 298
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8033,
"step": 299
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7893,
"step": 300
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7054,
"step": 301
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.9471,
"step": 302
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8982,
"step": 303
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.9682,
"step": 304
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6024,
"step": 305
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7184,
"step": 306
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7566,
"step": 307
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7607,
"step": 308
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5993,
"step": 309
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7502,
"step": 310
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8958,
"step": 311
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8621,
"step": 312
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7835,
"step": 313
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6815,
"step": 314
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7599,
"step": 315
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.734,
"step": 316
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6543,
"step": 317
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6149,
"step": 318
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7239,
"step": 319
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.3942,
"step": 320
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7239,
"step": 321
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6636,
"step": 322
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.4779,
"step": 323
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5454,
"step": 324
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8962,
"step": 325
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6651,
"step": 326
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6797,
"step": 327
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7515,
"step": 328
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7732,
"step": 329
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6828,
"step": 330
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.8203,
"step": 331
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.9032,
"step": 332
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6609,
"step": 333
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6706,
"step": 334
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.555,
"step": 335
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6961,
"step": 336
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7812,
"step": 337
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7667,
"step": 338
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6431,
"step": 339
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6958,
"step": 340
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7055,
"step": 341
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.5383,
"step": 342
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6951,
"step": 343
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.624,
"step": 344
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.7244,
"step": 345
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6788,
"step": 346
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 1.0403,
"step": 347
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8656,
"step": 348
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7682,
"step": 349
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6556,
"step": 350
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9285,
"step": 351
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8778,
"step": 352
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 1.3931,
"step": 353
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7759,
"step": 354
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5455,
"step": 355
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9343,
"step": 356
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.627,
"step": 357
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7332,
"step": 358
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4948,
"step": 359
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6612,
"step": 360
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8086,
"step": 361
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7435,
"step": 362
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9533,
"step": 363
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6942,
"step": 364
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7618,
"step": 365
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7144,
"step": 366
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6153,
"step": 367
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6408,
"step": 368
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 1.1895,
"step": 369
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.949,
"step": 370
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.4565,
"step": 371
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6903,
"step": 372
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6956,
"step": 373
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6568,
"step": 374
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6129,
"step": 375
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6205,
"step": 376
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8534,
"step": 377
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6967,
"step": 378
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7698,
"step": 379
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9859,
"step": 380
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 1.0283,
"step": 381
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7195,
"step": 382
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5506,
"step": 383
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7118,
"step": 384
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7334,
"step": 385
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8707,
"step": 386
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6609,
"step": 387
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8106,
"step": 388
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9443,
"step": 389
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.558,
"step": 390
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9331,
"step": 391
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7413,
"step": 392
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9114,
"step": 393
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 1.1099,
"step": 394
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8798,
"step": 395
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9745,
"step": 396
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7468,
"step": 397
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5748,
"step": 398
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9129,
"step": 399
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.8696,
"step": 400
},
{
"epoch": 0.06,
"eval_loss": 0.7634292244911194,
"eval_runtime": 154.2679,
"eval_samples_per_second": 6.482,
"eval_steps_per_second": 3.241,
"step": 400
},
{
"epoch": 0.06,
"mmlu_eval_accuracy": 0.582014575618387,
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
"mmlu_eval_accuracy_anatomy": 0.42857142857142855,
"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": 0.5,
"mmlu_eval_accuracy_college_chemistry": 0.25,
"mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
"mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
"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": 0.5,
"mmlu_eval_accuracy_electrical_engineering": 0.4375,
"mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
"mmlu_eval_accuracy_global_facts": 0.6,
"mmlu_eval_accuracy_high_school_biology": 0.53125,
"mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
"mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
"mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186,
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
"mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
"mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
"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": 0.5454545454545454,
"mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
"mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
"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": 0.5526315789473685,
"mmlu_eval_accuracy_moral_scenarios": 0.32,
"mmlu_eval_accuracy_nutrition": 0.7575757575757576,
"mmlu_eval_accuracy_philosophy": 0.7647058823529411,
"mmlu_eval_accuracy_prehistory": 0.6,
"mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
"mmlu_eval_accuracy_professional_law": 0.36470588235294116,
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
"mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.6296296296296297,
"mmlu_eval_accuracy_sociology": 0.9090909090909091,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.4444444444444444,
"mmlu_eval_accuracy_world_religions": 0.8421052631578947,
"mmlu_loss": 1.1038525643376706,
"step": 400
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9033,
"step": 401
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7314,
"step": 402
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.5781,
"step": 403
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6335,
"step": 404
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7935,
"step": 405
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7954,
"step": 406
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.7832,
"step": 407
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.911,
"step": 408
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.9981,
"step": 409
},
{
"epoch": 0.06,
"learning_rate": 0.0002,
"loss": 0.6957,
"step": 410
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 1.0077,
"step": 411
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7651,
"step": 412
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9796,
"step": 413
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8033,
"step": 414
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7299,
"step": 415
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8103,
"step": 416
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.6264,
"step": 417
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9188,
"step": 418
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.675,
"step": 419
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9005,
"step": 420
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9,
"step": 421
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8124,
"step": 422
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.878,
"step": 423
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 1.012,
"step": 424
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9317,
"step": 425
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9145,
"step": 426
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.551,
"step": 427
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7277,
"step": 428
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8815,
"step": 429
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.5955,
"step": 430
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7337,
"step": 431
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8115,
"step": 432
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.4938,
"step": 433
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7597,
"step": 434
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7131,
"step": 435
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8459,
"step": 436
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.682,
"step": 437
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7539,
"step": 438
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.5832,
"step": 439
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 1.027,
"step": 440
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8823,
"step": 441
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8111,
"step": 442
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7896,
"step": 443
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8611,
"step": 444
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7889,
"step": 445
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.842,
"step": 446
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 1.1133,
"step": 447
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8547,
"step": 448
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.635,
"step": 449
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.819,
"step": 450
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.6771,
"step": 451
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9835,
"step": 452
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7865,
"step": 453
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7251,
"step": 454
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7252,
"step": 455
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.6285,
"step": 456
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8226,
"step": 457
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9227,
"step": 458
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8963,
"step": 459
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.9053,
"step": 460
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8643,
"step": 461
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8897,
"step": 462
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8069,
"step": 463
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.6344,
"step": 464
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8596,
"step": 465
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 1.0561,
"step": 466
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.7554,
"step": 467
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8137,
"step": 468
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.922,
"step": 469
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8109,
"step": 470
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8298,
"step": 471
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.5106,
"step": 472
},
{
"epoch": 0.07,
"learning_rate": 0.0002,
"loss": 0.8882,
"step": 473
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6041,
"step": 474
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8929,
"step": 475
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7443,
"step": 476
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 1.0332,
"step": 477
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6776,
"step": 478
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6819,
"step": 479
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7143,
"step": 480
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9434,
"step": 481
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 1.0725,
"step": 482
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5402,
"step": 483
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8952,
"step": 484
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7315,
"step": 485
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 1.1633,
"step": 486
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6815,
"step": 487
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5161,
"step": 488
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6741,
"step": 489
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7827,
"step": 490
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5693,
"step": 491
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5521,
"step": 492
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5783,
"step": 493
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9696,
"step": 494
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8316,
"step": 495
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8936,
"step": 496
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7704,
"step": 497
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9177,
"step": 498
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6919,
"step": 499
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5591,
"step": 500
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.793,
"step": 501
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6148,
"step": 502
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8616,
"step": 503
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7995,
"step": 504
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9026,
"step": 505
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8013,
"step": 506
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7315,
"step": 507
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9467,
"step": 508
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9412,
"step": 509
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.3872,
"step": 510
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.5475,
"step": 511
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8641,
"step": 512
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8478,
"step": 513
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9471,
"step": 514
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6307,
"step": 515
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8896,
"step": 516
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.9323,
"step": 517
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 1.1377,
"step": 518
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6785,
"step": 519
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7828,
"step": 520
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7498,
"step": 521
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8784,
"step": 522
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8723,
"step": 523
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6768,
"step": 524
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.57,
"step": 525
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6339,
"step": 526
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6221,
"step": 527
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8042,
"step": 528
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6164,
"step": 529
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6869,
"step": 530
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7531,
"step": 531
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6863,
"step": 532
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.6329,
"step": 533
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.8289,
"step": 534
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7399,
"step": 535
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 0.7921,
"step": 536
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6602,
"step": 537
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8025,
"step": 538
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7965,
"step": 539
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9087,
"step": 540
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7666,
"step": 541
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5291,
"step": 542
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 1.0275,
"step": 543
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7976,
"step": 544
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9139,
"step": 545
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6894,
"step": 546
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.815,
"step": 547
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7209,
"step": 548
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5423,
"step": 549
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.903,
"step": 550
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6087,
"step": 551
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8032,
"step": 552
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9988,
"step": 553
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 1.0055,
"step": 554
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7856,
"step": 555
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8602,
"step": 556
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7981,
"step": 557
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8529,
"step": 558
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9516,
"step": 559
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5964,
"step": 560
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6579,
"step": 561
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6909,
"step": 562
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7012,
"step": 563
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8099,
"step": 564
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8283,
"step": 565
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6402,
"step": 566
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6288,
"step": 567
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6598,
"step": 568
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9796,
"step": 569
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.614,
"step": 570
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9603,
"step": 571
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 1.0403,
"step": 572
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7975,
"step": 573
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6631,
"step": 574
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.589,
"step": 575
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8022,
"step": 576
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8301,
"step": 577
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 1.3195,
"step": 578
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8354,
"step": 579
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.994,
"step": 580
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7999,
"step": 581
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7224,
"step": 582
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5131,
"step": 583
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.8155,
"step": 584
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7564,
"step": 585
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7305,
"step": 586
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9284,
"step": 587
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.5877,
"step": 588
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.4468,
"step": 589
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9086,
"step": 590
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7976,
"step": 591
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7157,
"step": 592
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6894,
"step": 593
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.9683,
"step": 594
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7784,
"step": 595
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.7333,
"step": 596
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6739,
"step": 597
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6285,
"step": 598
},
{
"epoch": 0.09,
"learning_rate": 0.0002,
"loss": 0.6718,
"step": 599
},
{
"epoch": 0.1,
"learning_rate": 0.0002,
"loss": 1.0186,
"step": 600
},
{
"epoch": 0.1,
"eval_loss": 0.7626925110816956,
"eval_runtime": 154.285,
"eval_samples_per_second": 6.482,
"eval_steps_per_second": 3.241,
"step": 600
},
{
"epoch": 0.1,
"mmlu_eval_accuracy": 0.5739124443829915,
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
"mmlu_eval_accuracy_anatomy": 0.5,
"mmlu_eval_accuracy_astronomy": 0.75,
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
"mmlu_eval_accuracy_college_biology": 0.5625,
"mmlu_eval_accuracy_college_chemistry": 0.25,
"mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
"mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
"mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
"mmlu_eval_accuracy_computer_security": 0.7272727272727273,
"mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
"mmlu_eval_accuracy_econometrics": 0.5,
"mmlu_eval_accuracy_electrical_engineering": 0.4375,
"mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
"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": 0.5555555555555556,
"mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
"mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
"mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
"mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
"mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
"mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
"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": 0.6818181818181818,
"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": 1.0,
"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": 0.88,
"mmlu_eval_accuracy_medical_genetics": 1.0,
"mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
"mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
"mmlu_eval_accuracy_moral_scenarios": 0.33,
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
"mmlu_eval_accuracy_philosophy": 0.7352941176470589,
"mmlu_eval_accuracy_prehistory": 0.6571428571428571,
"mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
"mmlu_eval_accuracy_professional_law": 0.36470588235294116,
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
"mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
"mmlu_eval_accuracy_public_relations": 0.5,
"mmlu_eval_accuracy_security_studies": 0.6296296296296297,
"mmlu_eval_accuracy_sociology": 0.9090909090909091,
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
"mmlu_eval_accuracy_virology": 0.4444444444444444,
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
"mmlu_loss": 1.3611424469605438,
"step": 600
}
],
"logging_steps": 1,
"max_steps": 18933,
"num_train_epochs": 3,
"save_steps": 200,
"total_flos": 1.41004811216683e+17,
"trial_name": null,
"trial_params": null
}