diff --git "a/checkpoint-6400/trainer_state.json" "b/checkpoint-6400/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-6400/trainer_state.json"
@@ -0,0 +1,40691 @@
+{
+ "best_metric": 0.749654233455658,
+ "best_model_checkpoint": "experts/mistralic-expert-15/checkpoint-3400",
+ "epoch": 1.0515506264119943,
+ "eval_steps": 200,
+ "global_step": 6400,
+ "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": 0.8743,
+ "step": 1
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 1.0512,
+ "step": 2
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 1.0398,
+ "step": 3
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8428,
+ "step": 4
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8539,
+ "step": 5
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6822,
+ "step": 6
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7817,
+ "step": 7
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8251,
+ "step": 8
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.641,
+ "step": 9
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.9213,
+ "step": 10
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8051,
+ "step": 11
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.9494,
+ "step": 12
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7953,
+ "step": 13
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6879,
+ "step": 14
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7866,
+ "step": 15
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7557,
+ "step": 16
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6147,
+ "step": 17
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7999,
+ "step": 18
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.9288,
+ "step": 19
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6545,
+ "step": 20
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6837,
+ "step": 21
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8564,
+ "step": 22
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6611,
+ "step": 23
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6956,
+ "step": 24
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5079,
+ "step": 25
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6142,
+ "step": 26
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7226,
+ "step": 27
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7037,
+ "step": 28
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8852,
+ "step": 29
+ },
+ {
+ "epoch": 0.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7209,
+ "step": 30
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8991,
+ "step": 31
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5839,
+ "step": 32
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6147,
+ "step": 33
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9075,
+ "step": 34
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6618,
+ "step": 35
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8367,
+ "step": 36
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7012,
+ "step": 37
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9249,
+ "step": 38
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5889,
+ "step": 39
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8997,
+ "step": 40
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7283,
+ "step": 41
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8314,
+ "step": 42
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7005,
+ "step": 43
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5693,
+ "step": 44
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8334,
+ "step": 45
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9775,
+ "step": 46
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7298,
+ "step": 47
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9334,
+ "step": 48
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7225,
+ "step": 49
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8088,
+ "step": 50
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7466,
+ "step": 51
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4508,
+ "step": 52
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4574,
+ "step": 53
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6886,
+ "step": 54
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7615,
+ "step": 55
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5989,
+ "step": 56
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 1.0519,
+ "step": 57
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9321,
+ "step": 58
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6666,
+ "step": 59
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5781,
+ "step": 60
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9561,
+ "step": 61
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5952,
+ "step": 62
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7159,
+ "step": 63
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6358,
+ "step": 64
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4376,
+ "step": 65
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6991,
+ "step": 66
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6598,
+ "step": 67
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7651,
+ "step": 68
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7757,
+ "step": 69
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7273,
+ "step": 70
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8523,
+ "step": 71
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6573,
+ "step": 72
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5618,
+ "step": 73
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6906,
+ "step": 74
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7202,
+ "step": 75
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8629,
+ "step": 76
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6166,
+ "step": 77
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8685,
+ "step": 78
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8057,
+ "step": 79
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8424,
+ "step": 80
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9356,
+ "step": 81
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8011,
+ "step": 82
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5423,
+ "step": 83
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6254,
+ "step": 84
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7161,
+ "step": 85
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6061,
+ "step": 86
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5743,
+ "step": 87
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.609,
+ "step": 88
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7292,
+ "step": 89
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9335,
+ "step": 90
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.0002,
+ "loss": 0.678,
+ "step": 91
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7919,
+ "step": 92
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6743,
+ "step": 93
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9325,
+ "step": 94
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6499,
+ "step": 95
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5631,
+ "step": 96
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8618,
+ "step": 97
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.683,
+ "step": 98
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.635,
+ "step": 99
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7615,
+ "step": 100
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8628,
+ "step": 101
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7456,
+ "step": 102
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8235,
+ "step": 103
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.916,
+ "step": 104
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 1.0372,
+ "step": 105
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9372,
+ "step": 106
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6709,
+ "step": 107
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 1.1999,
+ "step": 108
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7315,
+ "step": 109
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7147,
+ "step": 110
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6933,
+ "step": 111
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.844,
+ "step": 112
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7172,
+ "step": 113
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7477,
+ "step": 114
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.835,
+ "step": 115
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 1.0017,
+ "step": 116
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5665,
+ "step": 117
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6992,
+ "step": 118
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9693,
+ "step": 119
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.574,
+ "step": 120
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8181,
+ "step": 121
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6321,
+ "step": 122
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.85,
+ "step": 123
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6986,
+ "step": 124
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7242,
+ "step": 125
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7909,
+ "step": 126
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7143,
+ "step": 127
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5493,
+ "step": 128
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7625,
+ "step": 129
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7199,
+ "step": 130
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.3807,
+ "step": 131
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 1.322,
+ "step": 132
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.582,
+ "step": 133
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.575,
+ "step": 134
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8918,
+ "step": 135
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 1.1352,
+ "step": 136
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9177,
+ "step": 137
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9293,
+ "step": 138
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5509,
+ "step": 139
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9554,
+ "step": 140
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5094,
+ "step": 141
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5929,
+ "step": 142
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9733,
+ "step": 143
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5117,
+ "step": 144
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6024,
+ "step": 145
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6401,
+ "step": 146
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.924,
+ "step": 147
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.9399,
+ "step": 148
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.972,
+ "step": 149
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6407,
+ "step": 150
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8665,
+ "step": 151
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7225,
+ "step": 152
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6445,
+ "step": 153
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7775,
+ "step": 154
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7517,
+ "step": 155
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9464,
+ "step": 156
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5912,
+ "step": 157
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6161,
+ "step": 158
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8201,
+ "step": 159
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6348,
+ "step": 160
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 1.0841,
+ "step": 161
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8405,
+ "step": 162
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9317,
+ "step": 163
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7158,
+ "step": 164
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9286,
+ "step": 165
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9918,
+ "step": 166
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9909,
+ "step": 167
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7513,
+ "step": 168
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8818,
+ "step": 169
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8644,
+ "step": 170
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7351,
+ "step": 171
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7574,
+ "step": 172
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6465,
+ "step": 173
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7679,
+ "step": 174
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8849,
+ "step": 175
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 1.0791,
+ "step": 176
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8741,
+ "step": 177
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7822,
+ "step": 178
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8417,
+ "step": 179
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7317,
+ "step": 180
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 1.0245,
+ "step": 181
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6746,
+ "step": 182
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6059,
+ "step": 183
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9795,
+ "step": 184
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9011,
+ "step": 185
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6384,
+ "step": 186
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7801,
+ "step": 187
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.81,
+ "step": 188
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8018,
+ "step": 189
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6842,
+ "step": 190
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7468,
+ "step": 191
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8598,
+ "step": 192
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9457,
+ "step": 193
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5614,
+ "step": 194
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6668,
+ "step": 195
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8057,
+ "step": 196
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6429,
+ "step": 197
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6506,
+ "step": 198
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8127,
+ "step": 199
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9115,
+ "step": 200
+ },
+ {
+ "epoch": 0.03,
+ "eval_loss": 0.7577090859413147,
+ "eval_runtime": 131.6982,
+ "eval_samples_per_second": 7.593,
+ "eval_steps_per_second": 3.797,
+ "step": 200
+ },
+ {
+ "epoch": 0.03,
+ "mmlu_eval_accuracy": 0.5965101027344765,
+ "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": 0.6206896551724138,
+ "mmlu_eval_accuracy_college_biology": 0.5625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
+ "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": 0.5,
+ "mmlu_eval_accuracy_econometrics": 0.5,
+ "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": 0.3,
+ "mmlu_eval_accuracy_high_school_biology": 0.5625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
+ "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
+ "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
+ "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
+ "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": 1.0,
+ "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": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.34,
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
+ "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
+ "mmlu_eval_accuracy_prehistory": 0.6,
+ "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
+ "mmlu_eval_accuracy_professional_law": 0.4,
+ "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
+ "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": 0.9090909090909091,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
+ "mmlu_loss": 1.1431999449708132,
+ "step": 200
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6858,
+ "step": 201
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4792,
+ "step": 202
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9803,
+ "step": 203
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9486,
+ "step": 204
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7161,
+ "step": 205
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.635,
+ "step": 206
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4869,
+ "step": 207
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6713,
+ "step": 208
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6963,
+ "step": 209
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7726,
+ "step": 210
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 1.0783,
+ "step": 211
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.9185,
+ "step": 212
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.812,
+ "step": 213
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9379,
+ "step": 214
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9187,
+ "step": 215
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6081,
+ "step": 216
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7071,
+ "step": 217
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7934,
+ "step": 218
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7151,
+ "step": 219
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8808,
+ "step": 220
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6993,
+ "step": 221
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8453,
+ "step": 222
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7213,
+ "step": 223
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6957,
+ "step": 224
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9296,
+ "step": 225
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7856,
+ "step": 226
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5761,
+ "step": 227
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8645,
+ "step": 228
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6516,
+ "step": 229
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8441,
+ "step": 230
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.3882,
+ "step": 231
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 1.0478,
+ "step": 232
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8387,
+ "step": 233
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7188,
+ "step": 234
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9145,
+ "step": 235
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8007,
+ "step": 236
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8292,
+ "step": 237
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5326,
+ "step": 238
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 1.4012,
+ "step": 239
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9156,
+ "step": 240
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6651,
+ "step": 241
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9193,
+ "step": 242
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7638,
+ "step": 243
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6298,
+ "step": 244
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7544,
+ "step": 245
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7319,
+ "step": 246
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.807,
+ "step": 247
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7104,
+ "step": 248
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8149,
+ "step": 249
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.534,
+ "step": 250
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7246,
+ "step": 251
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7361,
+ "step": 252
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7693,
+ "step": 253
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8553,
+ "step": 254
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6042,
+ "step": 255
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6617,
+ "step": 256
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7081,
+ "step": 257
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7198,
+ "step": 258
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8238,
+ "step": 259
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7848,
+ "step": 260
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8787,
+ "step": 261
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8322,
+ "step": 262
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 1.0992,
+ "step": 263
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7046,
+ "step": 264
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4757,
+ "step": 265
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4978,
+ "step": 266
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8656,
+ "step": 267
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8534,
+ "step": 268
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 1.0157,
+ "step": 269
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6066,
+ "step": 270
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8877,
+ "step": 271
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8856,
+ "step": 272
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8631,
+ "step": 273
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 1.0294,
+ "step": 274
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.596,
+ "step": 275
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8029,
+ "step": 276
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6784,
+ "step": 277
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8277,
+ "step": 278
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9855,
+ "step": 279
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7293,
+ "step": 280
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8036,
+ "step": 281
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8581,
+ "step": 282
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6941,
+ "step": 283
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9493,
+ "step": 284
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9682,
+ "step": 285
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7554,
+ "step": 286
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8585,
+ "step": 287
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5798,
+ "step": 288
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.814,
+ "step": 289
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7303,
+ "step": 290
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7244,
+ "step": 291
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7552,
+ "step": 292
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 1.0223,
+ "step": 293
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7623,
+ "step": 294
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7383,
+ "step": 295
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5089,
+ "step": 296
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.4504,
+ "step": 297
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 1.2205,
+ "step": 298
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9076,
+ "step": 299
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9371,
+ "step": 300
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 1.1473,
+ "step": 301
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8726,
+ "step": 302
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7765,
+ "step": 303
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7015,
+ "step": 304
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7357,
+ "step": 305
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 1.1318,
+ "step": 306
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7922,
+ "step": 307
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7434,
+ "step": 308
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9389,
+ "step": 309
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 1.0817,
+ "step": 310
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6492,
+ "step": 311
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8057,
+ "step": 312
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8286,
+ "step": 313
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7106,
+ "step": 314
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9558,
+ "step": 315
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7318,
+ "step": 316
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 1.0106,
+ "step": 317
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.923,
+ "step": 318
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8891,
+ "step": 319
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7447,
+ "step": 320
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7532,
+ "step": 321
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7922,
+ "step": 322
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6687,
+ "step": 323
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6815,
+ "step": 324
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9777,
+ "step": 325
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7946,
+ "step": 326
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6075,
+ "step": 327
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.8379,
+ "step": 328
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6908,
+ "step": 329
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6723,
+ "step": 330
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6847,
+ "step": 331
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9873,
+ "step": 332
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6398,
+ "step": 333
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6382,
+ "step": 334
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.5199,
+ "step": 335
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6147,
+ "step": 336
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6788,
+ "step": 337
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7803,
+ "step": 338
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8861,
+ "step": 339
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6744,
+ "step": 340
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6291,
+ "step": 341
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8696,
+ "step": 342
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8301,
+ "step": 343
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7515,
+ "step": 344
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6998,
+ "step": 345
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6413,
+ "step": 346
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6048,
+ "step": 347
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.9201,
+ "step": 348
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7271,
+ "step": 349
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8123,
+ "step": 350
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7396,
+ "step": 351
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6787,
+ "step": 352
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7128,
+ "step": 353
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7682,
+ "step": 354
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.4681,
+ "step": 355
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7411,
+ "step": 356
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6903,
+ "step": 357
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 1.054,
+ "step": 358
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 1.0388,
+ "step": 359
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.5839,
+ "step": 360
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.9394,
+ "step": 361
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8511,
+ "step": 362
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7703,
+ "step": 363
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8453,
+ "step": 364
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8057,
+ "step": 365
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7747,
+ "step": 366
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7674,
+ "step": 367
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7861,
+ "step": 368
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7925,
+ "step": 369
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.951,
+ "step": 370
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.9278,
+ "step": 371
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.781,
+ "step": 372
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6763,
+ "step": 373
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.5281,
+ "step": 374
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.5816,
+ "step": 375
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.9666,
+ "step": 376
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.5588,
+ "step": 377
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 1.0995,
+ "step": 378
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 1.2138,
+ "step": 379
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8365,
+ "step": 380
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6347,
+ "step": 381
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7131,
+ "step": 382
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.7783,
+ "step": 383
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.5475,
+ "step": 384
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6053,
+ "step": 385
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.9546,
+ "step": 386
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.9732,
+ "step": 387
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.621,
+ "step": 388
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6031,
+ "step": 389
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6867,
+ "step": 390
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.8615,
+ "step": 391
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.9171,
+ "step": 392
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.678,
+ "step": 393
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.881,
+ "step": 394
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.435,
+ "step": 395
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7515,
+ "step": 396
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7088,
+ "step": 397
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8921,
+ "step": 398
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7842,
+ "step": 399
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7468,
+ "step": 400
+ },
+ {
+ "epoch": 0.07,
+ "eval_loss": 0.7549907565116882,
+ "eval_runtime": 131.6241,
+ "eval_samples_per_second": 7.597,
+ "eval_steps_per_second": 3.799,
+ "step": 400
+ },
+ {
+ "epoch": 0.07,
+ "mmlu_eval_accuracy": 0.5905395038007227,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.8125,
+ "mmlu_eval_accuracy_business_ethics": 0.8181818181818182,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
+ "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": 0.6363636363636364,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.6875,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
+ "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.59375,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
+ "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": 0.7833333333333333,
+ "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
+ "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
+ "mmlu_eval_accuracy_international_law": 1.0,
+ "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
+ "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": 0.7674418604651163,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.34,
+ "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
+ "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
+ "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
+ "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
+ "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
+ "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
+ "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": 0.8421052631578947,
+ "mmlu_loss": 1.1786961650412637,
+ "step": 400
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8926,
+ "step": 401
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8226,
+ "step": 402
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7972,
+ "step": 403
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9732,
+ "step": 404
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8672,
+ "step": 405
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.3649,
+ "step": 406
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9201,
+ "step": 407
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9357,
+ "step": 408
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8635,
+ "step": 409
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6937,
+ "step": 410
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 1.3983,
+ "step": 411
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7475,
+ "step": 412
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6109,
+ "step": 413
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6539,
+ "step": 414
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6245,
+ "step": 415
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8199,
+ "step": 416
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9231,
+ "step": 417
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.739,
+ "step": 418
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9367,
+ "step": 419
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.71,
+ "step": 420
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7856,
+ "step": 421
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6392,
+ "step": 422
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9306,
+ "step": 423
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7466,
+ "step": 424
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7127,
+ "step": 425
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6587,
+ "step": 426
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6164,
+ "step": 427
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7736,
+ "step": 428
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 1.016,
+ "step": 429
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7706,
+ "step": 430
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9474,
+ "step": 431
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7483,
+ "step": 432
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.811,
+ "step": 433
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7163,
+ "step": 434
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9745,
+ "step": 435
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7279,
+ "step": 436
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9218,
+ "step": 437
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8183,
+ "step": 438
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8229,
+ "step": 439
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7518,
+ "step": 440
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7858,
+ "step": 441
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 1.0076,
+ "step": 442
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.9241,
+ "step": 443
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8994,
+ "step": 444
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.438,
+ "step": 445
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7214,
+ "step": 446
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.5063,
+ "step": 447
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6436,
+ "step": 448
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6496,
+ "step": 449
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7387,
+ "step": 450
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.824,
+ "step": 451
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.5313,
+ "step": 452
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7908,
+ "step": 453
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8242,
+ "step": 454
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.8876,
+ "step": 455
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.7456,
+ "step": 456
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.752,
+ "step": 457
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.4918,
+ "step": 458
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.855,
+ "step": 459
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8322,
+ "step": 460
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7973,
+ "step": 461
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9598,
+ "step": 462
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.791,
+ "step": 463
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.5249,
+ "step": 464
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6198,
+ "step": 465
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6606,
+ "step": 466
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7283,
+ "step": 467
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8863,
+ "step": 468
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6965,
+ "step": 469
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7057,
+ "step": 470
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8994,
+ "step": 471
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6775,
+ "step": 472
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.5412,
+ "step": 473
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9663,
+ "step": 474
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9682,
+ "step": 475
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9535,
+ "step": 476
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7155,
+ "step": 477
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8014,
+ "step": 478
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8144,
+ "step": 479
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.713,
+ "step": 480
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6225,
+ "step": 481
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7206,
+ "step": 482
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.5586,
+ "step": 483
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8996,
+ "step": 484
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7745,
+ "step": 485
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9241,
+ "step": 486
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7356,
+ "step": 487
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6449,
+ "step": 488
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7228,
+ "step": 489
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6462,
+ "step": 490
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8052,
+ "step": 491
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9312,
+ "step": 492
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7832,
+ "step": 493
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8995,
+ "step": 494
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7184,
+ "step": 495
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7811,
+ "step": 496
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8659,
+ "step": 497
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7348,
+ "step": 498
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.7742,
+ "step": 499
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8397,
+ "step": 500
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6825,
+ "step": 501
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8492,
+ "step": 502
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 1.0731,
+ "step": 503
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.669,
+ "step": 504
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6464,
+ "step": 505
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9044,
+ "step": 506
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 1.174,
+ "step": 507
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9606,
+ "step": 508
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 1.0218,
+ "step": 509
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.5801,
+ "step": 510
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 1.0433,
+ "step": 511
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.8027,
+ "step": 512
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 1.0204,
+ "step": 513
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.9669,
+ "step": 514
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.5325,
+ "step": 515
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 1.0595,
+ "step": 516
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 1.002,
+ "step": 517
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8445,
+ "step": 518
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8647,
+ "step": 519
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7309,
+ "step": 520
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.5321,
+ "step": 521
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7862,
+ "step": 522
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8507,
+ "step": 523
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.9764,
+ "step": 524
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.5889,
+ "step": 525
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6066,
+ "step": 526
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7368,
+ "step": 527
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 1.0567,
+ "step": 528
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8779,
+ "step": 529
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 1.0253,
+ "step": 530
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8322,
+ "step": 531
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6848,
+ "step": 532
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.5222,
+ "step": 533
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7199,
+ "step": 534
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8765,
+ "step": 535
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8593,
+ "step": 536
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8867,
+ "step": 537
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6727,
+ "step": 538
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6709,
+ "step": 539
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8352,
+ "step": 540
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 1.0596,
+ "step": 541
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7518,
+ "step": 542
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.57,
+ "step": 543
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7398,
+ "step": 544
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.9338,
+ "step": 545
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7498,
+ "step": 546
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6023,
+ "step": 547
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8713,
+ "step": 548
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8121,
+ "step": 549
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 1.2485,
+ "step": 550
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8921,
+ "step": 551
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.9471,
+ "step": 552
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.4399,
+ "step": 553
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7338,
+ "step": 554
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8261,
+ "step": 555
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7494,
+ "step": 556
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7305,
+ "step": 557
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.9084,
+ "step": 558
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7184,
+ "step": 559
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7275,
+ "step": 560
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8692,
+ "step": 561
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8769,
+ "step": 562
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7692,
+ "step": 563
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.9968,
+ "step": 564
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6536,
+ "step": 565
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7829,
+ "step": 566
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7614,
+ "step": 567
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8699,
+ "step": 568
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8991,
+ "step": 569
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6783,
+ "step": 570
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.7599,
+ "step": 571
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6944,
+ "step": 572
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.748,
+ "step": 573
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6174,
+ "step": 574
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6199,
+ "step": 575
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.8565,
+ "step": 576
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.9175,
+ "step": 577
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.578,
+ "step": 578
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7151,
+ "step": 579
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7445,
+ "step": 580
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6315,
+ "step": 581
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7391,
+ "step": 582
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8364,
+ "step": 583
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.5496,
+ "step": 584
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7928,
+ "step": 585
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8136,
+ "step": 586
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.5404,
+ "step": 587
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6769,
+ "step": 588
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6331,
+ "step": 589
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.5495,
+ "step": 590
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7055,
+ "step": 591
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 1.0395,
+ "step": 592
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8604,
+ "step": 593
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8628,
+ "step": 594
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.4613,
+ "step": 595
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7412,
+ "step": 596
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.9055,
+ "step": 597
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.5592,
+ "step": 598
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 1.0462,
+ "step": 599
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7479,
+ "step": 600
+ },
+ {
+ "epoch": 0.1,
+ "eval_loss": 0.7511235475540161,
+ "eval_runtime": 131.6743,
+ "eval_samples_per_second": 7.594,
+ "eval_steps_per_second": 3.797,
+ "step": 600
+ },
+ {
+ "epoch": 0.1,
+ "mmlu_eval_accuracy": 0.5854937013814044,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.75,
+ "mmlu_eval_accuracy_business_ethics": 0.8181818181818182,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
+ "mmlu_eval_accuracy_college_biology": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
+ "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": 0.4146341463414634,
+ "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": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
+ "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
+ "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": 0.45454545454545453,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
+ "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.92,
+ "mmlu_eval_accuracy_medical_genetics": 1.0,
+ "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.24,
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
+ "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
+ "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
+ "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
+ "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": 0.9090909090909091,
+ "mmlu_eval_accuracy_virology": 0.6111111111111112,
+ "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
+ "mmlu_loss": 0.8422373449697351,
+ "step": 600
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7688,
+ "step": 601
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8277,
+ "step": 602
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.9892,
+ "step": 603
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.882,
+ "step": 604
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.5368,
+ "step": 605
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6269,
+ "step": 606
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.5438,
+ "step": 607
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.5921,
+ "step": 608
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8068,
+ "step": 609
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8916,
+ "step": 610
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7217,
+ "step": 611
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.9221,
+ "step": 612
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8893,
+ "step": 613
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6156,
+ "step": 614
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8421,
+ "step": 615
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7934,
+ "step": 616
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.9781,
+ "step": 617
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8513,
+ "step": 618
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7201,
+ "step": 619
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.9233,
+ "step": 620
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6561,
+ "step": 621
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7198,
+ "step": 622
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8383,
+ "step": 623
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8131,
+ "step": 624
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7621,
+ "step": 625
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.9067,
+ "step": 626
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 1.0717,
+ "step": 627
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7355,
+ "step": 628
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 1.0598,
+ "step": 629
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8181,
+ "step": 630
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8185,
+ "step": 631
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.8387,
+ "step": 632
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6738,
+ "step": 633
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 1.2803,
+ "step": 634
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7685,
+ "step": 635
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7091,
+ "step": 636
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.779,
+ "step": 637
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6472,
+ "step": 638
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.7769,
+ "step": 639
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6087,
+ "step": 640
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8182,
+ "step": 641
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.5228,
+ "step": 642
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.3634,
+ "step": 643
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8811,
+ "step": 644
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.9722,
+ "step": 645
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7217,
+ "step": 646
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7172,
+ "step": 647
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 1.0123,
+ "step": 648
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 1.171,
+ "step": 649
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8549,
+ "step": 650
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.9442,
+ "step": 651
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8082,
+ "step": 652
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.5462,
+ "step": 653
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7786,
+ "step": 654
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.5764,
+ "step": 655
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.5766,
+ "step": 656
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 1.2254,
+ "step": 657
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7426,
+ "step": 658
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7231,
+ "step": 659
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.9842,
+ "step": 660
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8422,
+ "step": 661
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.785,
+ "step": 662
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7621,
+ "step": 663
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8803,
+ "step": 664
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8239,
+ "step": 665
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8327,
+ "step": 666
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.9772,
+ "step": 667
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8029,
+ "step": 668
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.5865,
+ "step": 669
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8794,
+ "step": 670
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8821,
+ "step": 671
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7357,
+ "step": 672
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6092,
+ "step": 673
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7127,
+ "step": 674
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7141,
+ "step": 675
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7687,
+ "step": 676
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.9527,
+ "step": 677
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8083,
+ "step": 678
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.4753,
+ "step": 679
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7073,
+ "step": 680
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.5191,
+ "step": 681
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6506,
+ "step": 682
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6917,
+ "step": 683
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6707,
+ "step": 684
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6223,
+ "step": 685
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7281,
+ "step": 686
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8363,
+ "step": 687
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8984,
+ "step": 688
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 1.0086,
+ "step": 689
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8176,
+ "step": 690
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6743,
+ "step": 691
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.7653,
+ "step": 692
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.908,
+ "step": 693
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8402,
+ "step": 694
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8425,
+ "step": 695
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.767,
+ "step": 696
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.8191,
+ "step": 697
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.5398,
+ "step": 698
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0002,
+ "loss": 0.6728,
+ "step": 699
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8391,
+ "step": 700
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7766,
+ "step": 701
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6116,
+ "step": 702
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9081,
+ "step": 703
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6774,
+ "step": 704
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8251,
+ "step": 705
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9524,
+ "step": 706
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.535,
+ "step": 707
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6306,
+ "step": 708
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9171,
+ "step": 709
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9689,
+ "step": 710
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9917,
+ "step": 711
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8348,
+ "step": 712
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.587,
+ "step": 713
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.5918,
+ "step": 714
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7032,
+ "step": 715
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8926,
+ "step": 716
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6812,
+ "step": 717
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.828,
+ "step": 718
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7452,
+ "step": 719
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6744,
+ "step": 720
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8305,
+ "step": 721
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6471,
+ "step": 722
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.5812,
+ "step": 723
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.5891,
+ "step": 724
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.94,
+ "step": 725
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7092,
+ "step": 726
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7654,
+ "step": 727
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.828,
+ "step": 728
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7061,
+ "step": 729
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.5904,
+ "step": 730
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8591,
+ "step": 731
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9372,
+ "step": 732
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8514,
+ "step": 733
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8774,
+ "step": 734
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9201,
+ "step": 735
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8356,
+ "step": 736
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8392,
+ "step": 737
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7399,
+ "step": 738
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.5375,
+ "step": 739
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8915,
+ "step": 740
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7966,
+ "step": 741
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6087,
+ "step": 742
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6648,
+ "step": 743
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 1.0216,
+ "step": 744
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9432,
+ "step": 745
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6548,
+ "step": 746
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8015,
+ "step": 747
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7979,
+ "step": 748
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6144,
+ "step": 749
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6494,
+ "step": 750
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8614,
+ "step": 751
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.948,
+ "step": 752
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7197,
+ "step": 753
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8588,
+ "step": 754
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9157,
+ "step": 755
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.8928,
+ "step": 756
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.7218,
+ "step": 757
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.9789,
+ "step": 758
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6592,
+ "step": 759
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6305,
+ "step": 760
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.866,
+ "step": 761
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7438,
+ "step": 762
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6438,
+ "step": 763
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 1.0191,
+ "step": 764
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7692,
+ "step": 765
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6723,
+ "step": 766
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8188,
+ "step": 767
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6793,
+ "step": 768
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8244,
+ "step": 769
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.9137,
+ "step": 770
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7096,
+ "step": 771
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8164,
+ "step": 772
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7262,
+ "step": 773
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7389,
+ "step": 774
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 1.2896,
+ "step": 775
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8491,
+ "step": 776
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8356,
+ "step": 777
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8435,
+ "step": 778
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7388,
+ "step": 779
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.941,
+ "step": 780
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8939,
+ "step": 781
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7314,
+ "step": 782
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7922,
+ "step": 783
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6946,
+ "step": 784
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.752,
+ "step": 785
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.5049,
+ "step": 786
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7308,
+ "step": 787
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7585,
+ "step": 788
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8932,
+ "step": 789
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 1.0651,
+ "step": 790
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.736,
+ "step": 791
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6819,
+ "step": 792
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6963,
+ "step": 793
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.9918,
+ "step": 794
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.9083,
+ "step": 795
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.9174,
+ "step": 796
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8295,
+ "step": 797
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 1.004,
+ "step": 798
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8248,
+ "step": 799
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.5399,
+ "step": 800
+ },
+ {
+ "epoch": 0.13,
+ "eval_loss": 0.7548895478248596,
+ "eval_runtime": 131.5977,
+ "eval_samples_per_second": 7.599,
+ "eval_steps_per_second": 3.799,
+ "step": 800
+ },
+ {
+ "epoch": 0.13,
+ "mmlu_eval_accuracy": 0.5966805174146648,
+ "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.625,
+ "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": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "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": 0.3902439024390244,
+ "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.5625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
+ "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.85,
+ "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": 0.6956521739130435,
+ "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": 0.2727272727272727,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.96,
+ "mmlu_eval_accuracy_medical_genetics": 1.0,
+ "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
+ "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
+ "mmlu_eval_accuracy_moral_scenarios": 0.32,
+ "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
+ "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
+ "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516,
+ "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
+ "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": 0.7407407407407407,
+ "mmlu_eval_accuracy_sociology": 0.8636363636363636,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.9473684210526315,
+ "mmlu_loss": 0.9437568160868189,
+ "step": 800
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8248,
+ "step": 801
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7918,
+ "step": 802
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.9041,
+ "step": 803
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8476,
+ "step": 804
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7063,
+ "step": 805
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8846,
+ "step": 806
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7896,
+ "step": 807
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 1.1473,
+ "step": 808
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6126,
+ "step": 809
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7253,
+ "step": 810
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.8065,
+ "step": 811
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.9831,
+ "step": 812
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7944,
+ "step": 813
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6154,
+ "step": 814
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7495,
+ "step": 815
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7081,
+ "step": 816
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6591,
+ "step": 817
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.823,
+ "step": 818
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.7436,
+ "step": 819
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6991,
+ "step": 820
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 1.0781,
+ "step": 821
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6746,
+ "step": 822
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9573,
+ "step": 823
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.5698,
+ "step": 824
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9445,
+ "step": 825
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8332,
+ "step": 826
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7154,
+ "step": 827
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.676,
+ "step": 828
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7679,
+ "step": 829
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.5332,
+ "step": 830
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6207,
+ "step": 831
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9614,
+ "step": 832
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8742,
+ "step": 833
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9492,
+ "step": 834
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7094,
+ "step": 835
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.889,
+ "step": 836
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7355,
+ "step": 837
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 1.2038,
+ "step": 838
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9353,
+ "step": 839
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.864,
+ "step": 840
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8414,
+ "step": 841
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7492,
+ "step": 842
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8535,
+ "step": 843
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7975,
+ "step": 844
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8976,
+ "step": 845
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9244,
+ "step": 846
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.797,
+ "step": 847
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.4536,
+ "step": 848
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8585,
+ "step": 849
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.457,
+ "step": 850
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.5783,
+ "step": 851
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7445,
+ "step": 852
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6791,
+ "step": 853
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7751,
+ "step": 854
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 1.2038,
+ "step": 855
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6359,
+ "step": 856
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 1.2393,
+ "step": 857
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6282,
+ "step": 858
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6458,
+ "step": 859
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8575,
+ "step": 860
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7906,
+ "step": 861
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6737,
+ "step": 862
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8032,
+ "step": 863
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.5549,
+ "step": 864
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6854,
+ "step": 865
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.879,
+ "step": 866
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.5739,
+ "step": 867
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9231,
+ "step": 868
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8842,
+ "step": 869
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7444,
+ "step": 870
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6666,
+ "step": 871
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.697,
+ "step": 872
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7969,
+ "step": 873
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9802,
+ "step": 874
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.8818,
+ "step": 875
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6988,
+ "step": 876
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.5309,
+ "step": 877
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.7392,
+ "step": 878
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9443,
+ "step": 879
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9115,
+ "step": 880
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.5311,
+ "step": 881
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.9616,
+ "step": 882
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7202,
+ "step": 883
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.9842,
+ "step": 884
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8423,
+ "step": 885
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.9754,
+ "step": 886
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6544,
+ "step": 887
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7321,
+ "step": 888
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6975,
+ "step": 889
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8439,
+ "step": 890
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7876,
+ "step": 891
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 1.0582,
+ "step": 892
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7032,
+ "step": 893
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6425,
+ "step": 894
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.815,
+ "step": 895
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.4249,
+ "step": 896
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6071,
+ "step": 897
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6565,
+ "step": 898
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8692,
+ "step": 899
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8633,
+ "step": 900
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.4776,
+ "step": 901
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6176,
+ "step": 902
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.5023,
+ "step": 903
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.683,
+ "step": 904
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6307,
+ "step": 905
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7335,
+ "step": 906
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6745,
+ "step": 907
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8103,
+ "step": 908
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8412,
+ "step": 909
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.814,
+ "step": 910
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7424,
+ "step": 911
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7319,
+ "step": 912
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8703,
+ "step": 913
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.9498,
+ "step": 914
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.954,
+ "step": 915
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.5388,
+ "step": 916
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7777,
+ "step": 917
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8466,
+ "step": 918
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8867,
+ "step": 919
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6023,
+ "step": 920
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 1.0075,
+ "step": 921
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.5475,
+ "step": 922
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.692,
+ "step": 923
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8253,
+ "step": 924
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8064,
+ "step": 925
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 1.1761,
+ "step": 926
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.9338,
+ "step": 927
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8504,
+ "step": 928
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.5643,
+ "step": 929
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.5888,
+ "step": 930
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.9313,
+ "step": 931
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6123,
+ "step": 932
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8887,
+ "step": 933
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8548,
+ "step": 934
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 1.3875,
+ "step": 935
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6125,
+ "step": 936
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7241,
+ "step": 937
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.7575,
+ "step": 938
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8203,
+ "step": 939
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.6237,
+ "step": 940
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.652,
+ "step": 941
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.8225,
+ "step": 942
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.5296,
+ "step": 943
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6233,
+ "step": 944
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6494,
+ "step": 945
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6741,
+ "step": 946
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6992,
+ "step": 947
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 1.0143,
+ "step": 948
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 1.2206,
+ "step": 949
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6951,
+ "step": 950
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9564,
+ "step": 951
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7238,
+ "step": 952
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8592,
+ "step": 953
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7821,
+ "step": 954
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7797,
+ "step": 955
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7822,
+ "step": 956
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7729,
+ "step": 957
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.688,
+ "step": 958
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9142,
+ "step": 959
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.651,
+ "step": 960
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6945,
+ "step": 961
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6274,
+ "step": 962
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.651,
+ "step": 963
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8954,
+ "step": 964
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.878,
+ "step": 965
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6039,
+ "step": 966
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6027,
+ "step": 967
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.5063,
+ "step": 968
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8569,
+ "step": 969
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 1.0186,
+ "step": 970
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8204,
+ "step": 971
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8562,
+ "step": 972
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9192,
+ "step": 973
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9584,
+ "step": 974
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7057,
+ "step": 975
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7363,
+ "step": 976
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7962,
+ "step": 977
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8045,
+ "step": 978
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7177,
+ "step": 979
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.796,
+ "step": 980
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6995,
+ "step": 981
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.5048,
+ "step": 982
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9541,
+ "step": 983
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.4876,
+ "step": 984
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.506,
+ "step": 985
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8758,
+ "step": 986
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6352,
+ "step": 987
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.7109,
+ "step": 988
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8874,
+ "step": 989
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.812,
+ "step": 990
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.499,
+ "step": 991
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8348,
+ "step": 992
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9514,
+ "step": 993
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9723,
+ "step": 994
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.9755,
+ "step": 995
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8997,
+ "step": 996
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.82,
+ "step": 997
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8012,
+ "step": 998
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.5464,
+ "step": 999
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6164,
+ "step": 1000
+ },
+ {
+ "epoch": 0.16,
+ "eval_loss": 0.7555892467498779,
+ "eval_runtime": 131.6346,
+ "eval_samples_per_second": 7.597,
+ "eval_steps_per_second": 3.798,
+ "step": 1000
+ },
+ {
+ "epoch": 0.16,
+ "mmlu_eval_accuracy": 0.5856789112224339,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "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": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "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": 0.4146341463414634,
+ "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "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": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.85,
+ "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": 0.6956521739130435,
+ "mmlu_eval_accuracy_human_sexuality": 0.5,
+ "mmlu_eval_accuracy_international_law": 1.0,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
+ "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.92,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
+ "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
+ "mmlu_eval_accuracy_moral_scenarios": 0.39,
+ "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
+ "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
+ "mmlu_eval_accuracy_prehistory": 0.6,
+ "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
+ "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
+ "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": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
+ "mmlu_loss": 1.176194295909635,
+ "step": 1000
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8221,
+ "step": 1001
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 1.3423,
+ "step": 1002
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6437,
+ "step": 1003
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.8111,
+ "step": 1004
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6787,
+ "step": 1005
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.5613,
+ "step": 1006
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9071,
+ "step": 1007
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8575,
+ "step": 1008
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.5898,
+ "step": 1009
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7674,
+ "step": 1010
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6666,
+ "step": 1011
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6155,
+ "step": 1012
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8348,
+ "step": 1013
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6385,
+ "step": 1014
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 1.0248,
+ "step": 1015
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7654,
+ "step": 1016
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7371,
+ "step": 1017
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6062,
+ "step": 1018
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8954,
+ "step": 1019
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9026,
+ "step": 1020
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9989,
+ "step": 1021
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7809,
+ "step": 1022
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.5978,
+ "step": 1023
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7286,
+ "step": 1024
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6645,
+ "step": 1025
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8464,
+ "step": 1026
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7771,
+ "step": 1027
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9945,
+ "step": 1028
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9833,
+ "step": 1029
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6665,
+ "step": 1030
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7047,
+ "step": 1031
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7872,
+ "step": 1032
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9812,
+ "step": 1033
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6762,
+ "step": 1034
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8424,
+ "step": 1035
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9057,
+ "step": 1036
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7369,
+ "step": 1037
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7533,
+ "step": 1038
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6887,
+ "step": 1039
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8873,
+ "step": 1040
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7545,
+ "step": 1041
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6764,
+ "step": 1042
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8972,
+ "step": 1043
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8649,
+ "step": 1044
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8409,
+ "step": 1045
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 1.0243,
+ "step": 1046
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7304,
+ "step": 1047
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9103,
+ "step": 1048
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.669,
+ "step": 1049
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6538,
+ "step": 1050
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8883,
+ "step": 1051
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9743,
+ "step": 1052
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.685,
+ "step": 1053
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6473,
+ "step": 1054
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.9056,
+ "step": 1055
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 1.1629,
+ "step": 1056
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8226,
+ "step": 1057
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.5441,
+ "step": 1058
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7121,
+ "step": 1059
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.5908,
+ "step": 1060
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.7726,
+ "step": 1061
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6605,
+ "step": 1062
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.8256,
+ "step": 1063
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.5057,
+ "step": 1064
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6002,
+ "step": 1065
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7843,
+ "step": 1066
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.9437,
+ "step": 1067
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 1.1681,
+ "step": 1068
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.5524,
+ "step": 1069
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8754,
+ "step": 1070
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7274,
+ "step": 1071
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6221,
+ "step": 1072
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.9309,
+ "step": 1073
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8359,
+ "step": 1074
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 1.0487,
+ "step": 1075
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7114,
+ "step": 1076
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7412,
+ "step": 1077
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.4848,
+ "step": 1078
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6176,
+ "step": 1079
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7986,
+ "step": 1080
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6757,
+ "step": 1081
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8007,
+ "step": 1082
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6298,
+ "step": 1083
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.932,
+ "step": 1084
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 1.0875,
+ "step": 1085
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7853,
+ "step": 1086
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8039,
+ "step": 1087
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.944,
+ "step": 1088
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6463,
+ "step": 1089
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8139,
+ "step": 1090
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.9296,
+ "step": 1091
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7712,
+ "step": 1092
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.9416,
+ "step": 1093
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6695,
+ "step": 1094
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7162,
+ "step": 1095
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.711,
+ "step": 1096
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7038,
+ "step": 1097
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7496,
+ "step": 1098
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.5564,
+ "step": 1099
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.9269,
+ "step": 1100
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.9102,
+ "step": 1101
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.628,
+ "step": 1102
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7816,
+ "step": 1103
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7092,
+ "step": 1104
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7987,
+ "step": 1105
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7363,
+ "step": 1106
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8112,
+ "step": 1107
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7628,
+ "step": 1108
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.5817,
+ "step": 1109
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7048,
+ "step": 1110
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8298,
+ "step": 1111
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7488,
+ "step": 1112
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7505,
+ "step": 1113
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7865,
+ "step": 1114
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7026,
+ "step": 1115
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.5051,
+ "step": 1116
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.9068,
+ "step": 1117
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7855,
+ "step": 1118
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6746,
+ "step": 1119
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.6051,
+ "step": 1120
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8529,
+ "step": 1121
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.7515,
+ "step": 1122
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.5733,
+ "step": 1123
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.8888,
+ "step": 1124
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0002,
+ "loss": 0.4714,
+ "step": 1125
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.911,
+ "step": 1126
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8629,
+ "step": 1127
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6223,
+ "step": 1128
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7618,
+ "step": 1129
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.9176,
+ "step": 1130
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.9618,
+ "step": 1131
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5051,
+ "step": 1132
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 1.0598,
+ "step": 1133
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.9089,
+ "step": 1134
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.657,
+ "step": 1135
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8222,
+ "step": 1136
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8729,
+ "step": 1137
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8369,
+ "step": 1138
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8575,
+ "step": 1139
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8162,
+ "step": 1140
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7896,
+ "step": 1141
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8753,
+ "step": 1142
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7976,
+ "step": 1143
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6839,
+ "step": 1144
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7471,
+ "step": 1145
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6854,
+ "step": 1146
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5067,
+ "step": 1147
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6942,
+ "step": 1148
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6953,
+ "step": 1149
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7569,
+ "step": 1150
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.9129,
+ "step": 1151
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5026,
+ "step": 1152
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8202,
+ "step": 1153
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 1.0068,
+ "step": 1154
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8909,
+ "step": 1155
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7483,
+ "step": 1156
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8111,
+ "step": 1157
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5987,
+ "step": 1158
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5368,
+ "step": 1159
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.943,
+ "step": 1160
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7594,
+ "step": 1161
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8482,
+ "step": 1162
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 1.0019,
+ "step": 1163
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7763,
+ "step": 1164
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8401,
+ "step": 1165
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 1.0795,
+ "step": 1166
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5373,
+ "step": 1167
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.4947,
+ "step": 1168
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6115,
+ "step": 1169
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5907,
+ "step": 1170
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7703,
+ "step": 1171
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6941,
+ "step": 1172
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8686,
+ "step": 1173
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8994,
+ "step": 1174
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6574,
+ "step": 1175
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 1.0021,
+ "step": 1176
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.751,
+ "step": 1177
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8691,
+ "step": 1178
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.5775,
+ "step": 1179
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7803,
+ "step": 1180
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.9986,
+ "step": 1181
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.8704,
+ "step": 1182
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6584,
+ "step": 1183
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.6776,
+ "step": 1184
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.7895,
+ "step": 1185
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.4927,
+ "step": 1186
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6107,
+ "step": 1187
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8162,
+ "step": 1188
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5878,
+ "step": 1189
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5487,
+ "step": 1190
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.897,
+ "step": 1191
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.9297,
+ "step": 1192
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7531,
+ "step": 1193
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7013,
+ "step": 1194
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8751,
+ "step": 1195
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7412,
+ "step": 1196
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.858,
+ "step": 1197
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6135,
+ "step": 1198
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 1.0381,
+ "step": 1199
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 1.0329,
+ "step": 1200
+ },
+ {
+ "epoch": 0.2,
+ "eval_loss": 0.7524652481079102,
+ "eval_runtime": 131.7121,
+ "eval_samples_per_second": 7.592,
+ "eval_steps_per_second": 3.796,
+ "step": 1200
+ },
+ {
+ "epoch": 0.2,
+ "mmlu_eval_accuracy": 0.5742141326455935,
+ "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.75,
+ "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
+ "mmlu_eval_accuracy_college_biology": 0.6875,
+ "mmlu_eval_accuracy_college_chemistry": 0.125,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5,
+ "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
+ "mmlu_eval_accuracy_electrical_engineering": 0.6875,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "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": 0.6666666666666666,
+ "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": 0.7619047619047619,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
+ "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
+ "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": 0.8461538461538461,
+ "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.88,
+ "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
+ "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
+ "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
+ "mmlu_eval_accuracy_moral_scenarios": 0.36,
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
+ "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
+ "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
+ "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
+ "mmlu_eval_accuracy_professional_law": 0.3235294117647059,
+ "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.7037037037037037,
+ "mmlu_eval_accuracy_sociology": 0.9090909090909091,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
+ "mmlu_eval_accuracy_virology": 0.4444444444444444,
+ "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
+ "mmlu_loss": 1.2399911641996773,
+ "step": 1200
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5216,
+ "step": 1201
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.618,
+ "step": 1202
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5964,
+ "step": 1203
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8837,
+ "step": 1204
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.4266,
+ "step": 1205
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8449,
+ "step": 1206
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7134,
+ "step": 1207
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6538,
+ "step": 1208
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.9778,
+ "step": 1209
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8887,
+ "step": 1210
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5954,
+ "step": 1211
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7968,
+ "step": 1212
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.719,
+ "step": 1213
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 1.1052,
+ "step": 1214
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8811,
+ "step": 1215
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8834,
+ "step": 1216
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.807,
+ "step": 1217
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6735,
+ "step": 1218
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6044,
+ "step": 1219
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7845,
+ "step": 1220
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.9047,
+ "step": 1221
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 1.0172,
+ "step": 1222
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6626,
+ "step": 1223
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.639,
+ "step": 1224
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8858,
+ "step": 1225
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8591,
+ "step": 1226
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8587,
+ "step": 1227
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8154,
+ "step": 1228
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7013,
+ "step": 1229
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7533,
+ "step": 1230
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7474,
+ "step": 1231
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.746,
+ "step": 1232
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.931,
+ "step": 1233
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5149,
+ "step": 1234
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.9315,
+ "step": 1235
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5211,
+ "step": 1236
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6216,
+ "step": 1237
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.7627,
+ "step": 1238
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8407,
+ "step": 1239
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.9171,
+ "step": 1240
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 1.0086,
+ "step": 1241
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.759,
+ "step": 1242
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8143,
+ "step": 1243
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.8281,
+ "step": 1244
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6156,
+ "step": 1245
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.5096,
+ "step": 1246
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.671,
+ "step": 1247
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.4925,
+ "step": 1248
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8338,
+ "step": 1249
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9885,
+ "step": 1250
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7908,
+ "step": 1251
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8172,
+ "step": 1252
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6143,
+ "step": 1253
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6758,
+ "step": 1254
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7881,
+ "step": 1255
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8398,
+ "step": 1256
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7218,
+ "step": 1257
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6242,
+ "step": 1258
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6917,
+ "step": 1259
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9483,
+ "step": 1260
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7747,
+ "step": 1261
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6527,
+ "step": 1262
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8877,
+ "step": 1263
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9047,
+ "step": 1264
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9139,
+ "step": 1265
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7574,
+ "step": 1266
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8861,
+ "step": 1267
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.81,
+ "step": 1268
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7459,
+ "step": 1269
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7822,
+ "step": 1270
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7384,
+ "step": 1271
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7143,
+ "step": 1272
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.5266,
+ "step": 1273
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9644,
+ "step": 1274
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7135,
+ "step": 1275
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8658,
+ "step": 1276
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6821,
+ "step": 1277
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9459,
+ "step": 1278
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7442,
+ "step": 1279
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.781,
+ "step": 1280
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.64,
+ "step": 1281
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6969,
+ "step": 1282
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.4252,
+ "step": 1283
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8044,
+ "step": 1284
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9488,
+ "step": 1285
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 1.0964,
+ "step": 1286
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7516,
+ "step": 1287
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6294,
+ "step": 1288
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.931,
+ "step": 1289
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8273,
+ "step": 1290
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6675,
+ "step": 1291
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8139,
+ "step": 1292
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.5207,
+ "step": 1293
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6666,
+ "step": 1294
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7511,
+ "step": 1295
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6742,
+ "step": 1296
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8203,
+ "step": 1297
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7265,
+ "step": 1298
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9514,
+ "step": 1299
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.9468,
+ "step": 1300
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7451,
+ "step": 1301
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7945,
+ "step": 1302
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.825,
+ "step": 1303
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.8965,
+ "step": 1304
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 1.0714,
+ "step": 1305
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7463,
+ "step": 1306
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6405,
+ "step": 1307
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.7203,
+ "step": 1308
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7591,
+ "step": 1309
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7341,
+ "step": 1310
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7748,
+ "step": 1311
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7815,
+ "step": 1312
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8286,
+ "step": 1313
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.9297,
+ "step": 1314
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 1.029,
+ "step": 1315
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7089,
+ "step": 1316
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8273,
+ "step": 1317
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7323,
+ "step": 1318
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7651,
+ "step": 1319
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.9488,
+ "step": 1320
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6787,
+ "step": 1321
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.746,
+ "step": 1322
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8983,
+ "step": 1323
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.5873,
+ "step": 1324
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 1.0014,
+ "step": 1325
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7568,
+ "step": 1326
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 1.0171,
+ "step": 1327
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.3873,
+ "step": 1328
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7452,
+ "step": 1329
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7642,
+ "step": 1330
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.753,
+ "step": 1331
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.5854,
+ "step": 1332
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8696,
+ "step": 1333
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 1.0902,
+ "step": 1334
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8505,
+ "step": 1335
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7669,
+ "step": 1336
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8207,
+ "step": 1337
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6588,
+ "step": 1338
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8637,
+ "step": 1339
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6975,
+ "step": 1340
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8329,
+ "step": 1341
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.9193,
+ "step": 1342
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.926,
+ "step": 1343
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.5889,
+ "step": 1344
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.9468,
+ "step": 1345
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.798,
+ "step": 1346
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8462,
+ "step": 1347
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 1.0115,
+ "step": 1348
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8523,
+ "step": 1349
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6957,
+ "step": 1350
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.9172,
+ "step": 1351
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.5134,
+ "step": 1352
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6298,
+ "step": 1353
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7291,
+ "step": 1354
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.873,
+ "step": 1355
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6002,
+ "step": 1356
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8432,
+ "step": 1357
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.9966,
+ "step": 1358
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.99,
+ "step": 1359
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.68,
+ "step": 1360
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8726,
+ "step": 1361
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.5941,
+ "step": 1362
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6808,
+ "step": 1363
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6686,
+ "step": 1364
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.8719,
+ "step": 1365
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6443,
+ "step": 1366
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6231,
+ "step": 1367
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7072,
+ "step": 1368
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.7883,
+ "step": 1369
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.897,
+ "step": 1370
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7346,
+ "step": 1371
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.776,
+ "step": 1372
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6668,
+ "step": 1373
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8826,
+ "step": 1374
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.9369,
+ "step": 1375
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8684,
+ "step": 1376
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6926,
+ "step": 1377
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8416,
+ "step": 1378
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7519,
+ "step": 1379
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6369,
+ "step": 1380
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.924,
+ "step": 1381
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 1.2479,
+ "step": 1382
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.685,
+ "step": 1383
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.5376,
+ "step": 1384
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.484,
+ "step": 1385
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7056,
+ "step": 1386
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 1.0267,
+ "step": 1387
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8326,
+ "step": 1388
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7695,
+ "step": 1389
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8156,
+ "step": 1390
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7312,
+ "step": 1391
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 1.0382,
+ "step": 1392
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.9356,
+ "step": 1393
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7705,
+ "step": 1394
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.9324,
+ "step": 1395
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 1.1027,
+ "step": 1396
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7073,
+ "step": 1397
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7028,
+ "step": 1398
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8509,
+ "step": 1399
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7476,
+ "step": 1400
+ },
+ {
+ "epoch": 0.23,
+ "eval_loss": 0.7565351724624634,
+ "eval_runtime": 131.6911,
+ "eval_samples_per_second": 7.594,
+ "eval_steps_per_second": 3.797,
+ "step": 1400
+ },
+ {
+ "epoch": 0.23,
+ "mmlu_eval_accuracy": 0.5850023557597949,
+ "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.6875,
+ "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
+ "mmlu_eval_accuracy_college_biology": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.25,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
+ "mmlu_eval_accuracy_econometrics": 0.5,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
+ "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": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.85,
+ "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": 0.5833333333333334,
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
+ "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.88,
+ "mmlu_eval_accuracy_medical_genetics": 1.0,
+ "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605,
+ "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
+ "mmlu_eval_accuracy_moral_scenarios": 0.39,
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
+ "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
+ "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
+ "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
+ "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": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
+ "mmlu_loss": 1.295089167574367,
+ "step": 1400
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7247,
+ "step": 1401
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7038,
+ "step": 1402
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8196,
+ "step": 1403
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7339,
+ "step": 1404
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.625,
+ "step": 1405
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7286,
+ "step": 1406
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.9877,
+ "step": 1407
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8991,
+ "step": 1408
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6369,
+ "step": 1409
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.765,
+ "step": 1410
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 1.0762,
+ "step": 1411
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7993,
+ "step": 1412
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6695,
+ "step": 1413
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6561,
+ "step": 1414
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7605,
+ "step": 1415
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 1.1527,
+ "step": 1416
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6211,
+ "step": 1417
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6008,
+ "step": 1418
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7419,
+ "step": 1419
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.597,
+ "step": 1420
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6108,
+ "step": 1421
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.9615,
+ "step": 1422
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.9308,
+ "step": 1423
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6633,
+ "step": 1424
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7345,
+ "step": 1425
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.8532,
+ "step": 1426
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.5529,
+ "step": 1427
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6888,
+ "step": 1428
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6228,
+ "step": 1429
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.7561,
+ "step": 1430
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6594,
+ "step": 1431
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7793,
+ "step": 1432
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8974,
+ "step": 1433
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 1.0135,
+ "step": 1434
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6814,
+ "step": 1435
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9646,
+ "step": 1436
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6376,
+ "step": 1437
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6886,
+ "step": 1438
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.588,
+ "step": 1439
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9395,
+ "step": 1440
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9995,
+ "step": 1441
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6857,
+ "step": 1442
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8066,
+ "step": 1443
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7895,
+ "step": 1444
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8392,
+ "step": 1445
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.5989,
+ "step": 1446
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 1.0545,
+ "step": 1447
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9732,
+ "step": 1448
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6197,
+ "step": 1449
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9419,
+ "step": 1450
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7909,
+ "step": 1451
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8606,
+ "step": 1452
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6496,
+ "step": 1453
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8656,
+ "step": 1454
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.823,
+ "step": 1455
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7034,
+ "step": 1456
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8816,
+ "step": 1457
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.857,
+ "step": 1458
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8636,
+ "step": 1459
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8487,
+ "step": 1460
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8579,
+ "step": 1461
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6756,
+ "step": 1462
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8837,
+ "step": 1463
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7554,
+ "step": 1464
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.8105,
+ "step": 1465
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7906,
+ "step": 1466
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9217,
+ "step": 1467
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7679,
+ "step": 1468
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.69,
+ "step": 1469
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6041,
+ "step": 1470
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7895,
+ "step": 1471
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.6045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.835,
+ "step": 1473
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9231,
+ "step": 1474
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.5395,
+ "step": 1475
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.787,
+ "step": 1476
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9674,
+ "step": 1477
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9376,
+ "step": 1478
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9131,
+ "step": 1479
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9775,
+ "step": 1480
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.3288,
+ "step": 1481
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.593,
+ "step": 1482
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9766,
+ "step": 1483
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7524,
+ "step": 1484
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9152,
+ "step": 1485
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7092,
+ "step": 1486
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7326,
+ "step": 1487
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.7106,
+ "step": 1488
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 1.0405,
+ "step": 1489
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 1.0875,
+ "step": 1490
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.0002,
+ "loss": 0.9197,
+ "step": 1491
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.4796,
+ "step": 1492
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.9769,
+ "step": 1493
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.5788,
+ "step": 1494
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6842,
+ "step": 1495
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.674,
+ "step": 1496
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6809,
+ "step": 1497
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.4968,
+ "step": 1498
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7073,
+ "step": 1499
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7067,
+ "step": 1500
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.0425,
+ "step": 1501
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.4321,
+ "step": 1502
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7369,
+ "step": 1503
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8082,
+ "step": 1504
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.5295,
+ "step": 1505
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7709,
+ "step": 1506
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.5363,
+ "step": 1507
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7775,
+ "step": 1508
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.9121,
+ "step": 1509
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.432,
+ "step": 1510
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6316,
+ "step": 1511
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8131,
+ "step": 1512
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.97,
+ "step": 1513
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6833,
+ "step": 1514
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.4482,
+ "step": 1515
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6134,
+ "step": 1516
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7192,
+ "step": 1517
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.698,
+ "step": 1518
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7455,
+ "step": 1519
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8642,
+ "step": 1520
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.809,
+ "step": 1521
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8495,
+ "step": 1522
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8622,
+ "step": 1523
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.9561,
+ "step": 1524
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7723,
+ "step": 1525
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7609,
+ "step": 1526
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7192,
+ "step": 1527
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8696,
+ "step": 1528
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6784,
+ "step": 1529
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6138,
+ "step": 1530
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8004,
+ "step": 1531
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.1055,
+ "step": 1532
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.726,
+ "step": 1533
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6405,
+ "step": 1534
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7927,
+ "step": 1535
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.0059,
+ "step": 1536
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.4738,
+ "step": 1537
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.0319,
+ "step": 1538
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.9731,
+ "step": 1539
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8639,
+ "step": 1540
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.1349,
+ "step": 1541
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.9366,
+ "step": 1542
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7726,
+ "step": 1543
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7024,
+ "step": 1544
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.0042,
+ "step": 1545
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.0105,
+ "step": 1546
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.8245,
+ "step": 1547
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.5256,
+ "step": 1548
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 1.0531,
+ "step": 1549
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.7971,
+ "step": 1550
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0002,
+ "loss": 0.6856,
+ "step": 1551
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8739,
+ "step": 1552
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 1.0163,
+ "step": 1553
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6874,
+ "step": 1554
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.9986,
+ "step": 1555
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8165,
+ "step": 1556
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7431,
+ "step": 1557
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7756,
+ "step": 1558
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7082,
+ "step": 1559
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5777,
+ "step": 1560
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7756,
+ "step": 1561
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7613,
+ "step": 1562
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6612,
+ "step": 1563
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7807,
+ "step": 1564
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8527,
+ "step": 1565
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5728,
+ "step": 1566
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7335,
+ "step": 1567
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8667,
+ "step": 1568
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6938,
+ "step": 1569
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7762,
+ "step": 1570
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8781,
+ "step": 1571
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.9252,
+ "step": 1572
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6258,
+ "step": 1573
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5836,
+ "step": 1574
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.9178,
+ "step": 1575
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5666,
+ "step": 1576
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.4898,
+ "step": 1577
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.9138,
+ "step": 1578
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5762,
+ "step": 1579
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8416,
+ "step": 1580
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.4505,
+ "step": 1581
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5698,
+ "step": 1582
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8154,
+ "step": 1583
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7504,
+ "step": 1584
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7091,
+ "step": 1585
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5972,
+ "step": 1586
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.9147,
+ "step": 1587
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6601,
+ "step": 1588
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8604,
+ "step": 1589
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6772,
+ "step": 1590
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7868,
+ "step": 1591
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8426,
+ "step": 1592
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.9511,
+ "step": 1593
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7094,
+ "step": 1594
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 1.0559,
+ "step": 1595
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7736,
+ "step": 1596
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5268,
+ "step": 1597
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7912,
+ "step": 1598
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.764,
+ "step": 1599
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5238,
+ "step": 1600
+ },
+ {
+ "epoch": 0.26,
+ "eval_loss": 0.7527549862861633,
+ "eval_runtime": 131.6921,
+ "eval_samples_per_second": 7.593,
+ "eval_steps_per_second": 3.797,
+ "step": 1600
+ },
+ {
+ "epoch": 0.26,
+ "mmlu_eval_accuracy": 0.5841085819573677,
+ "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "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": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364,
+ "mmlu_eval_accuracy_college_medicine": 0.5,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
+ "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
+ "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
+ "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": 0.36363636363636365,
+ "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": 0.6190476190476191,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.85,
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
+ "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": 0.6111111111111112,
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "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": 0.35,
+ "mmlu_eval_accuracy_nutrition": 0.8181818181818182,
+ "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
+ "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
+ "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516,
+ "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
+ "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
+ "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": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
+ "mmlu_loss": 1.031481397175291,
+ "step": 1600
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8259,
+ "step": 1601
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.7597,
+ "step": 1602
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6521,
+ "step": 1603
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.739,
+ "step": 1604
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6859,
+ "step": 1605
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.5375,
+ "step": 1606
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8233,
+ "step": 1607
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 1.0097,
+ "step": 1608
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 1.0705,
+ "step": 1609
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.8505,
+ "step": 1610
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.4597,
+ "step": 1611
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.0002,
+ "loss": 0.6459,
+ "step": 1612
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.8404,
+ "step": 1613
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6851,
+ "step": 1614
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.5922,
+ "step": 1615
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6669,
+ "step": 1616
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 1.0592,
+ "step": 1617
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6426,
+ "step": 1618
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.8768,
+ "step": 1619
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7848,
+ "step": 1620
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7126,
+ "step": 1621
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6475,
+ "step": 1622
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.688,
+ "step": 1623
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9468,
+ "step": 1624
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9996,
+ "step": 1625
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6853,
+ "step": 1626
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7665,
+ "step": 1627
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6701,
+ "step": 1628
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.679,
+ "step": 1629
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.4117,
+ "step": 1630
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 1.0549,
+ "step": 1631
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.942,
+ "step": 1632
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6824,
+ "step": 1633
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7501,
+ "step": 1634
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9103,
+ "step": 1635
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.5898,
+ "step": 1636
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.4918,
+ "step": 1637
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6478,
+ "step": 1638
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7254,
+ "step": 1639
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.8687,
+ "step": 1640
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9022,
+ "step": 1641
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9671,
+ "step": 1642
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9609,
+ "step": 1643
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7365,
+ "step": 1644
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6662,
+ "step": 1645
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.636,
+ "step": 1646
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.5832,
+ "step": 1647
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6802,
+ "step": 1648
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.8406,
+ "step": 1649
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7152,
+ "step": 1650
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6391,
+ "step": 1651
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 1.0056,
+ "step": 1652
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.5359,
+ "step": 1653
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.5797,
+ "step": 1654
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9781,
+ "step": 1655
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6859,
+ "step": 1656
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.583,
+ "step": 1657
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.5682,
+ "step": 1658
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7168,
+ "step": 1659
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 1.0028,
+ "step": 1660
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.7273,
+ "step": 1661
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9537,
+ "step": 1662
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.5807,
+ "step": 1663
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.716,
+ "step": 1664
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.699,
+ "step": 1665
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6075,
+ "step": 1666
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6855,
+ "step": 1667
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6207,
+ "step": 1668
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9699,
+ "step": 1669
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.9061,
+ "step": 1670
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6173,
+ "step": 1671
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.6367,
+ "step": 1672
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.0002,
+ "loss": 0.8444,
+ "step": 1673
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8603,
+ "step": 1674
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8036,
+ "step": 1675
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9018,
+ "step": 1676
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.5705,
+ "step": 1677
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6724,
+ "step": 1678
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9261,
+ "step": 1679
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6321,
+ "step": 1680
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7732,
+ "step": 1681
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.545,
+ "step": 1682
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.759,
+ "step": 1683
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7698,
+ "step": 1684
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8507,
+ "step": 1685
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9669,
+ "step": 1686
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 1.0176,
+ "step": 1687
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7758,
+ "step": 1688
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9836,
+ "step": 1689
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6254,
+ "step": 1690
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.5343,
+ "step": 1691
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9225,
+ "step": 1692
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8519,
+ "step": 1693
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.5515,
+ "step": 1694
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7881,
+ "step": 1695
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6856,
+ "step": 1696
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.5237,
+ "step": 1697
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6892,
+ "step": 1698
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8218,
+ "step": 1699
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.5861,
+ "step": 1700
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6668,
+ "step": 1701
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6571,
+ "step": 1702
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8541,
+ "step": 1703
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9976,
+ "step": 1704
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9278,
+ "step": 1705
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6639,
+ "step": 1706
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6281,
+ "step": 1707
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7273,
+ "step": 1708
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6752,
+ "step": 1709
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6093,
+ "step": 1710
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 1.1475,
+ "step": 1711
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.705,
+ "step": 1712
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7031,
+ "step": 1713
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6636,
+ "step": 1714
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7738,
+ "step": 1715
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.5624,
+ "step": 1716
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9593,
+ "step": 1717
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7804,
+ "step": 1718
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9136,
+ "step": 1719
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 1.094,
+ "step": 1720
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8778,
+ "step": 1721
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6492,
+ "step": 1722
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6876,
+ "step": 1723
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7212,
+ "step": 1724
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.5282,
+ "step": 1725
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7868,
+ "step": 1726
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9316,
+ "step": 1727
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9305,
+ "step": 1728
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.9418,
+ "step": 1729
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7478,
+ "step": 1730
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.6085,
+ "step": 1731
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7648,
+ "step": 1732
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.7998,
+ "step": 1733
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0002,
+ "loss": 0.8439,
+ "step": 1734
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.651,
+ "step": 1735
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8026,
+ "step": 1736
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8653,
+ "step": 1737
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8301,
+ "step": 1738
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.5024,
+ "step": 1739
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8239,
+ "step": 1740
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8542,
+ "step": 1741
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 1.0579,
+ "step": 1742
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8412,
+ "step": 1743
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.5013,
+ "step": 1744
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.804,
+ "step": 1745
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8203,
+ "step": 1746
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 1.0144,
+ "step": 1747
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.947,
+ "step": 1748
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.5192,
+ "step": 1749
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.9724,
+ "step": 1750
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.6316,
+ "step": 1751
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 1.0797,
+ "step": 1752
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7627,
+ "step": 1753
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8389,
+ "step": 1754
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8919,
+ "step": 1755
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.6576,
+ "step": 1756
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8673,
+ "step": 1757
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8281,
+ "step": 1758
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.831,
+ "step": 1759
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7995,
+ "step": 1760
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7144,
+ "step": 1761
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 1.0493,
+ "step": 1762
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8622,
+ "step": 1763
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7824,
+ "step": 1764
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.762,
+ "step": 1765
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8588,
+ "step": 1766
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.5117,
+ "step": 1767
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7963,
+ "step": 1768
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.797,
+ "step": 1769
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7286,
+ "step": 1770
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.6041,
+ "step": 1771
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.9522,
+ "step": 1772
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.6672,
+ "step": 1773
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7562,
+ "step": 1774
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8356,
+ "step": 1775
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.6668,
+ "step": 1776
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.6884,
+ "step": 1777
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7675,
+ "step": 1778
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.5017,
+ "step": 1779
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7173,
+ "step": 1780
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.9622,
+ "step": 1781
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.5934,
+ "step": 1782
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.9902,
+ "step": 1783
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.68,
+ "step": 1784
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.7687,
+ "step": 1785
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8256,
+ "step": 1786
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 1.0602,
+ "step": 1787
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.9092,
+ "step": 1788
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 1.0037,
+ "step": 1789
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 1.0462,
+ "step": 1790
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8461,
+ "step": 1791
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.8195,
+ "step": 1792
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.885,
+ "step": 1793
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.4779,
+ "step": 1794
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.0002,
+ "loss": 0.9366,
+ "step": 1795
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8306,
+ "step": 1796
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7913,
+ "step": 1797
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8551,
+ "step": 1798
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7799,
+ "step": 1799
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.9489,
+ "step": 1800
+ },
+ {
+ "epoch": 0.3,
+ "eval_loss": 0.7564910650253296,
+ "eval_runtime": 131.6705,
+ "eval_samples_per_second": 7.595,
+ "eval_steps_per_second": 3.797,
+ "step": 1800
+ },
+ {
+ "epoch": 0.3,
+ "mmlu_eval_accuracy": 0.5870383762445152,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
+ "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": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364,
+ "mmlu_eval_accuracy_college_medicine": 0.5,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
+ "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5,
+ "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": 0.6875,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.5454545454545454,
+ "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": 0.7619047619047619,
+ "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": 0.17647058823529413,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
+ "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": 0.6363636363636364,
+ "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": 0.7674418604651163,
+ "mmlu_eval_accuracy_moral_disputes": 0.631578947368421,
+ "mmlu_eval_accuracy_moral_scenarios": 0.42,
+ "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
+ "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
+ "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
+ "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
+ "mmlu_eval_accuracy_sociology": 0.8181818181818182,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.9473684210526315,
+ "mmlu_loss": 1.0342699467707863,
+ "step": 1800
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8336,
+ "step": 1801
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.9158,
+ "step": 1802
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6537,
+ "step": 1803
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7154,
+ "step": 1804
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.4424,
+ "step": 1805
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8845,
+ "step": 1806
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 1.0182,
+ "step": 1807
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7668,
+ "step": 1808
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8308,
+ "step": 1809
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6782,
+ "step": 1810
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8811,
+ "step": 1811
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8927,
+ "step": 1812
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6867,
+ "step": 1813
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8708,
+ "step": 1814
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.5932,
+ "step": 1815
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.763,
+ "step": 1816
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.3666,
+ "step": 1817
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6287,
+ "step": 1818
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 1.1761,
+ "step": 1819
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6282,
+ "step": 1820
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.9538,
+ "step": 1821
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6732,
+ "step": 1822
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7191,
+ "step": 1823
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7516,
+ "step": 1824
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7324,
+ "step": 1825
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8245,
+ "step": 1826
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7211,
+ "step": 1827
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8983,
+ "step": 1828
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.9165,
+ "step": 1829
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8816,
+ "step": 1830
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6977,
+ "step": 1831
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.5939,
+ "step": 1832
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8262,
+ "step": 1833
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.5445,
+ "step": 1834
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.814,
+ "step": 1835
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.83,
+ "step": 1836
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 1.0397,
+ "step": 1837
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.5553,
+ "step": 1838
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7688,
+ "step": 1839
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8618,
+ "step": 1840
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8449,
+ "step": 1841
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8293,
+ "step": 1842
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.7413,
+ "step": 1843
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8758,
+ "step": 1844
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8831,
+ "step": 1845
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.578,
+ "step": 1846
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8054,
+ "step": 1847
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.9829,
+ "step": 1848
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8205,
+ "step": 1849
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.5944,
+ "step": 1850
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 1.0171,
+ "step": 1851
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8148,
+ "step": 1852
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.5911,
+ "step": 1853
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.9356,
+ "step": 1854
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.6723,
+ "step": 1855
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.0002,
+ "loss": 0.8174,
+ "step": 1856
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9787,
+ "step": 1857
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 1.0051,
+ "step": 1858
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.4867,
+ "step": 1859
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 1.1325,
+ "step": 1860
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 1.1187,
+ "step": 1861
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8395,
+ "step": 1862
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9261,
+ "step": 1863
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8604,
+ "step": 1864
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.718,
+ "step": 1865
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6281,
+ "step": 1866
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 1.2542,
+ "step": 1867
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6471,
+ "step": 1868
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6661,
+ "step": 1869
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7216,
+ "step": 1870
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7274,
+ "step": 1871
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7226,
+ "step": 1872
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6299,
+ "step": 1873
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8366,
+ "step": 1874
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7208,
+ "step": 1875
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7434,
+ "step": 1876
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.5952,
+ "step": 1877
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8474,
+ "step": 1878
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8277,
+ "step": 1879
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7267,
+ "step": 1880
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.856,
+ "step": 1881
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.985,
+ "step": 1882
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8828,
+ "step": 1883
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7644,
+ "step": 1884
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7251,
+ "step": 1885
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6503,
+ "step": 1886
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7388,
+ "step": 1887
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6078,
+ "step": 1888
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8609,
+ "step": 1889
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.739,
+ "step": 1890
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6842,
+ "step": 1891
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.3554,
+ "step": 1892
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9325,
+ "step": 1893
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.608,
+ "step": 1894
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7599,
+ "step": 1895
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7637,
+ "step": 1896
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7881,
+ "step": 1897
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9611,
+ "step": 1898
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.792,
+ "step": 1899
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8939,
+ "step": 1900
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8405,
+ "step": 1901
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8245,
+ "step": 1902
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9334,
+ "step": 1903
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.5922,
+ "step": 1904
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7117,
+ "step": 1905
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7399,
+ "step": 1906
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9311,
+ "step": 1907
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9798,
+ "step": 1908
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.596,
+ "step": 1909
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9791,
+ "step": 1910
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7955,
+ "step": 1911
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7784,
+ "step": 1912
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.8823,
+ "step": 1913
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6896,
+ "step": 1914
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.9004,
+ "step": 1915
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.6769,
+ "step": 1916
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0002,
+ "loss": 0.7266,
+ "step": 1917
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8832,
+ "step": 1918
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.3216,
+ "step": 1919
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8418,
+ "step": 1920
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8421,
+ "step": 1921
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.9854,
+ "step": 1922
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6857,
+ "step": 1923
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.5681,
+ "step": 1924
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7502,
+ "step": 1925
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8959,
+ "step": 1926
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8317,
+ "step": 1927
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.761,
+ "step": 1928
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7882,
+ "step": 1929
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8495,
+ "step": 1930
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7368,
+ "step": 1931
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.657,
+ "step": 1932
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8864,
+ "step": 1933
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7591,
+ "step": 1934
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.5665,
+ "step": 1935
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6528,
+ "step": 1936
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.4354,
+ "step": 1937
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7752,
+ "step": 1938
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.569,
+ "step": 1939
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.5951,
+ "step": 1940
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.4796,
+ "step": 1941
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7293,
+ "step": 1942
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8563,
+ "step": 1943
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 1.0438,
+ "step": 1944
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7843,
+ "step": 1945
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6391,
+ "step": 1946
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6883,
+ "step": 1947
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6799,
+ "step": 1948
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 1.2362,
+ "step": 1949
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.9167,
+ "step": 1950
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7315,
+ "step": 1951
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6957,
+ "step": 1952
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6972,
+ "step": 1953
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 1.0928,
+ "step": 1954
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8077,
+ "step": 1955
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.9741,
+ "step": 1956
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6326,
+ "step": 1957
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6168,
+ "step": 1958
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8294,
+ "step": 1959
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8041,
+ "step": 1960
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7237,
+ "step": 1961
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7619,
+ "step": 1962
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6153,
+ "step": 1963
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.642,
+ "step": 1964
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8367,
+ "step": 1965
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.9133,
+ "step": 1966
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8438,
+ "step": 1967
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.6413,
+ "step": 1968
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8013,
+ "step": 1969
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.761,
+ "step": 1970
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7426,
+ "step": 1971
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8,
+ "step": 1972
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.7121,
+ "step": 1973
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8288,
+ "step": 1974
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.746,
+ "step": 1975
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8884,
+ "step": 1976
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8298,
+ "step": 1977
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.0002,
+ "loss": 0.8072,
+ "step": 1978
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.4938,
+ "step": 1979
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6872,
+ "step": 1980
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.9711,
+ "step": 1981
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.974,
+ "step": 1982
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8756,
+ "step": 1983
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8436,
+ "step": 1984
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6283,
+ "step": 1985
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7213,
+ "step": 1986
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7792,
+ "step": 1987
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7662,
+ "step": 1988
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.9639,
+ "step": 1989
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7354,
+ "step": 1990
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7979,
+ "step": 1991
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.9131,
+ "step": 1992
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.4003,
+ "step": 1993
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.914,
+ "step": 1994
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.4672,
+ "step": 1995
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7574,
+ "step": 1996
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7151,
+ "step": 1997
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7959,
+ "step": 1998
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6512,
+ "step": 1999
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7852,
+ "step": 2000
+ },
+ {
+ "epoch": 0.33,
+ "eval_loss": 0.752848744392395,
+ "eval_runtime": 131.602,
+ "eval_samples_per_second": 7.599,
+ "eval_steps_per_second": 3.799,
+ "step": 2000
+ },
+ {
+ "epoch": 0.33,
+ "mmlu_eval_accuracy": 0.5691849760946832,
+ "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.5,
+ "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
+ "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": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.4375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
+ "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.65625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "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": 0.6153846153846154,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
+ "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
+ "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
+ "mmlu_eval_accuracy_international_law": 1.0,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.92,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
+ "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
+ "mmlu_eval_accuracy_moral_scenarios": 0.43,
+ "mmlu_eval_accuracy_nutrition": 0.7878787878787878,
+ "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
+ "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
+ "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "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": 0.7727272727272727,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
+ "mmlu_loss": 1.1752762202727889,
+ "step": 2000
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7267,
+ "step": 2001
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 1.2648,
+ "step": 2002
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8298,
+ "step": 2003
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8376,
+ "step": 2004
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.702,
+ "step": 2005
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8086,
+ "step": 2006
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8528,
+ "step": 2007
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6656,
+ "step": 2008
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8634,
+ "step": 2009
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8844,
+ "step": 2010
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.5432,
+ "step": 2011
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.9794,
+ "step": 2012
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.839,
+ "step": 2013
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7317,
+ "step": 2014
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7027,
+ "step": 2015
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7342,
+ "step": 2016
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6352,
+ "step": 2017
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.5922,
+ "step": 2018
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.4453,
+ "step": 2019
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7501,
+ "step": 2020
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8115,
+ "step": 2021
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.893,
+ "step": 2022
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.789,
+ "step": 2023
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7879,
+ "step": 2024
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6548,
+ "step": 2025
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 1.006,
+ "step": 2026
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7132,
+ "step": 2027
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.7709,
+ "step": 2028
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6992,
+ "step": 2029
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.5689,
+ "step": 2030
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6236,
+ "step": 2031
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8304,
+ "step": 2032
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.9352,
+ "step": 2033
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8443,
+ "step": 2034
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6599,
+ "step": 2035
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.8409,
+ "step": 2036
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.6997,
+ "step": 2037
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.0002,
+ "loss": 0.786,
+ "step": 2038
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.9681,
+ "step": 2039
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8544,
+ "step": 2040
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.9446,
+ "step": 2041
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7879,
+ "step": 2042
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 1.0545,
+ "step": 2043
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.6719,
+ "step": 2044
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7428,
+ "step": 2045
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8388,
+ "step": 2046
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7171,
+ "step": 2047
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.646,
+ "step": 2048
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8184,
+ "step": 2049
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5731,
+ "step": 2050
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8029,
+ "step": 2051
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.98,
+ "step": 2052
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5321,
+ "step": 2053
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7811,
+ "step": 2054
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 1.0171,
+ "step": 2055
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.753,
+ "step": 2056
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.9288,
+ "step": 2057
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7526,
+ "step": 2058
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8926,
+ "step": 2059
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8702,
+ "step": 2060
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8068,
+ "step": 2061
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8542,
+ "step": 2062
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7184,
+ "step": 2063
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8021,
+ "step": 2064
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7745,
+ "step": 2065
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5975,
+ "step": 2066
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5209,
+ "step": 2067
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.9128,
+ "step": 2068
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5747,
+ "step": 2069
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5494,
+ "step": 2070
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8932,
+ "step": 2071
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5441,
+ "step": 2072
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5504,
+ "step": 2073
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5092,
+ "step": 2074
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 1.1967,
+ "step": 2075
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.694,
+ "step": 2076
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8577,
+ "step": 2077
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.6906,
+ "step": 2078
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.6218,
+ "step": 2079
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8694,
+ "step": 2080
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.6817,
+ "step": 2081
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5004,
+ "step": 2082
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7916,
+ "step": 2083
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8507,
+ "step": 2084
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5931,
+ "step": 2085
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.6575,
+ "step": 2086
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 1.0013,
+ "step": 2087
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7782,
+ "step": 2088
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7849,
+ "step": 2089
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.6948,
+ "step": 2090
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.488,
+ "step": 2091
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7317,
+ "step": 2092
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8818,
+ "step": 2093
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7518,
+ "step": 2094
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.889,
+ "step": 2095
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.5147,
+ "step": 2096
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.6197,
+ "step": 2097
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.7885,
+ "step": 2098
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.0002,
+ "loss": 0.8667,
+ "step": 2099
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7704,
+ "step": 2100
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.6682,
+ "step": 2101
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.4656,
+ "step": 2102
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7399,
+ "step": 2103
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7891,
+ "step": 2104
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.6717,
+ "step": 2105
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7552,
+ "step": 2106
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 1.0138,
+ "step": 2107
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7381,
+ "step": 2108
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 1.0103,
+ "step": 2109
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.851,
+ "step": 2110
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7569,
+ "step": 2111
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.9706,
+ "step": 2112
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.5157,
+ "step": 2113
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7643,
+ "step": 2114
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7947,
+ "step": 2115
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8266,
+ "step": 2116
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.5799,
+ "step": 2117
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8642,
+ "step": 2118
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8729,
+ "step": 2119
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.6575,
+ "step": 2120
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7454,
+ "step": 2121
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8058,
+ "step": 2122
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.5945,
+ "step": 2123
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.524,
+ "step": 2124
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7295,
+ "step": 2125
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.659,
+ "step": 2126
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7518,
+ "step": 2127
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.623,
+ "step": 2128
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8044,
+ "step": 2129
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.6818,
+ "step": 2130
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8399,
+ "step": 2131
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7292,
+ "step": 2132
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8489,
+ "step": 2133
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8093,
+ "step": 2134
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7857,
+ "step": 2135
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.583,
+ "step": 2136
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.837,
+ "step": 2137
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.64,
+ "step": 2138
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.5885,
+ "step": 2139
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8039,
+ "step": 2140
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8894,
+ "step": 2141
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.9435,
+ "step": 2142
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8138,
+ "step": 2143
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.9148,
+ "step": 2144
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8558,
+ "step": 2145
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.4714,
+ "step": 2146
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8708,
+ "step": 2147
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.6831,
+ "step": 2148
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7856,
+ "step": 2149
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.6415,
+ "step": 2150
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.5021,
+ "step": 2151
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7662,
+ "step": 2152
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.8857,
+ "step": 2153
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.7005,
+ "step": 2154
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.9997,
+ "step": 2155
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.5424,
+ "step": 2156
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 1.0604,
+ "step": 2157
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.9917,
+ "step": 2158
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 1.1453,
+ "step": 2159
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.0002,
+ "loss": 0.786,
+ "step": 2160
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7012,
+ "step": 2161
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7995,
+ "step": 2162
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 1.2655,
+ "step": 2163
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7892,
+ "step": 2164
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6177,
+ "step": 2165
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7411,
+ "step": 2166
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.4969,
+ "step": 2167
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.5629,
+ "step": 2168
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7161,
+ "step": 2169
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8582,
+ "step": 2170
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.9504,
+ "step": 2171
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6852,
+ "step": 2172
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7765,
+ "step": 2173
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6606,
+ "step": 2174
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.9861,
+ "step": 2175
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6297,
+ "step": 2176
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 1.0347,
+ "step": 2177
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.778,
+ "step": 2178
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7029,
+ "step": 2179
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6162,
+ "step": 2180
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8231,
+ "step": 2181
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.763,
+ "step": 2182
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6419,
+ "step": 2183
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.607,
+ "step": 2184
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6823,
+ "step": 2185
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7794,
+ "step": 2186
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8182,
+ "step": 2187
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.9136,
+ "step": 2188
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.4843,
+ "step": 2189
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8645,
+ "step": 2190
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 1.0224,
+ "step": 2191
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.9855,
+ "step": 2192
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.5692,
+ "step": 2193
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6979,
+ "step": 2194
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.5271,
+ "step": 2195
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.9303,
+ "step": 2196
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.929,
+ "step": 2197
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7371,
+ "step": 2198
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6342,
+ "step": 2199
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8963,
+ "step": 2200
+ },
+ {
+ "epoch": 0.36,
+ "eval_loss": 0.7507924437522888,
+ "eval_runtime": 131.5866,
+ "eval_samples_per_second": 7.6,
+ "eval_steps_per_second": 3.8,
+ "step": 2200
+ },
+ {
+ "epoch": 0.36,
+ "mmlu_eval_accuracy": 0.5536695878813044,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.5625,
+ "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
+ "mmlu_eval_accuracy_college_biology": 0.5625,
+ "mmlu_eval_accuracy_college_chemistry": 0.125,
+ "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.4375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.71875,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
+ "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+ "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
+ "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+ "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+ "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": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.6363636363636364,
+ "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.28,
+ "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.7714285714285715,
+ "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
+ "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
+ "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
+ "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+ "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": 0.7368421052631579,
+ "mmlu_loss": 1.1485521083323202,
+ "step": 2200
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.5828,
+ "step": 2201
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8653,
+ "step": 2202
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7077,
+ "step": 2203
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7007,
+ "step": 2204
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6926,
+ "step": 2205
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7959,
+ "step": 2206
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8934,
+ "step": 2207
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8543,
+ "step": 2208
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7842,
+ "step": 2209
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8277,
+ "step": 2210
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8518,
+ "step": 2211
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.9768,
+ "step": 2212
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.9592,
+ "step": 2213
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 1.1021,
+ "step": 2214
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 1.0696,
+ "step": 2215
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8984,
+ "step": 2216
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.8487,
+ "step": 2217
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.6364,
+ "step": 2218
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7703,
+ "step": 2219
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 1.0047,
+ "step": 2220
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0002,
+ "loss": 0.7262,
+ "step": 2221
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8957,
+ "step": 2222
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.641,
+ "step": 2223
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.5702,
+ "step": 2224
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.764,
+ "step": 2225
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7372,
+ "step": 2226
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.754,
+ "step": 2227
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8521,
+ "step": 2228
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 1.1058,
+ "step": 2229
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8239,
+ "step": 2230
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 1.0771,
+ "step": 2231
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6587,
+ "step": 2232
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7997,
+ "step": 2233
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 1.0849,
+ "step": 2234
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.5827,
+ "step": 2235
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7081,
+ "step": 2236
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6897,
+ "step": 2237
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7522,
+ "step": 2238
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7877,
+ "step": 2239
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.805,
+ "step": 2240
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.5968,
+ "step": 2241
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8151,
+ "step": 2242
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6776,
+ "step": 2243
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8338,
+ "step": 2244
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8798,
+ "step": 2245
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8186,
+ "step": 2246
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7515,
+ "step": 2247
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8388,
+ "step": 2248
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8149,
+ "step": 2249
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7716,
+ "step": 2250
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6499,
+ "step": 2251
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7321,
+ "step": 2252
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7596,
+ "step": 2253
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.9931,
+ "step": 2254
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.981,
+ "step": 2255
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.701,
+ "step": 2256
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6564,
+ "step": 2257
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6767,
+ "step": 2258
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.4725,
+ "step": 2259
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7185,
+ "step": 2260
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8727,
+ "step": 2261
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.5604,
+ "step": 2262
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6934,
+ "step": 2263
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7527,
+ "step": 2264
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.779,
+ "step": 2265
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.943,
+ "step": 2266
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7045,
+ "step": 2267
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7053,
+ "step": 2268
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.9895,
+ "step": 2269
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7772,
+ "step": 2270
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7792,
+ "step": 2271
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8361,
+ "step": 2272
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8189,
+ "step": 2273
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.646,
+ "step": 2274
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 1.0571,
+ "step": 2275
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.648,
+ "step": 2276
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.7741,
+ "step": 2277
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.863,
+ "step": 2278
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.6895,
+ "step": 2279
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.8982,
+ "step": 2280
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.9069,
+ "step": 2281
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.0002,
+ "loss": 0.9023,
+ "step": 2282
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.559,
+ "step": 2283
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7736,
+ "step": 2284
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6028,
+ "step": 2285
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.3765,
+ "step": 2286
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8312,
+ "step": 2287
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9462,
+ "step": 2288
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7134,
+ "step": 2289
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.539,
+ "step": 2290
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9708,
+ "step": 2291
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6094,
+ "step": 2292
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7205,
+ "step": 2293
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8019,
+ "step": 2294
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 1.0196,
+ "step": 2295
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.936,
+ "step": 2296
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8383,
+ "step": 2297
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 1.0003,
+ "step": 2298
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9363,
+ "step": 2299
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9608,
+ "step": 2300
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6128,
+ "step": 2301
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6391,
+ "step": 2302
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7403,
+ "step": 2303
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6169,
+ "step": 2304
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.5755,
+ "step": 2305
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.5686,
+ "step": 2306
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.4683,
+ "step": 2307
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6812,
+ "step": 2308
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.753,
+ "step": 2309
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7064,
+ "step": 2310
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8319,
+ "step": 2311
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.5496,
+ "step": 2312
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8146,
+ "step": 2313
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 1.0419,
+ "step": 2314
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9729,
+ "step": 2315
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9189,
+ "step": 2316
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.4441,
+ "step": 2317
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6014,
+ "step": 2318
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7556,
+ "step": 2319
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8206,
+ "step": 2320
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8836,
+ "step": 2321
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7174,
+ "step": 2322
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8926,
+ "step": 2323
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.88,
+ "step": 2324
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7144,
+ "step": 2325
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6857,
+ "step": 2326
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8967,
+ "step": 2327
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7278,
+ "step": 2328
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8208,
+ "step": 2329
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.759,
+ "step": 2330
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7028,
+ "step": 2331
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8076,
+ "step": 2332
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7903,
+ "step": 2333
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.888,
+ "step": 2334
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6905,
+ "step": 2335
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9466,
+ "step": 2336
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.965,
+ "step": 2337
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.9968,
+ "step": 2338
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 1.0619,
+ "step": 2339
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8506,
+ "step": 2340
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.6321,
+ "step": 2341
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.7377,
+ "step": 2342
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.0002,
+ "loss": 0.8507,
+ "step": 2343
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7841,
+ "step": 2344
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7395,
+ "step": 2345
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.844,
+ "step": 2346
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5406,
+ "step": 2347
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7586,
+ "step": 2348
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.8497,
+ "step": 2349
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7953,
+ "step": 2350
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.9833,
+ "step": 2351
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.6546,
+ "step": 2352
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.9985,
+ "step": 2353
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.6249,
+ "step": 2354
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.933,
+ "step": 2355
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.8876,
+ "step": 2356
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.6898,
+ "step": 2357
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 1.0062,
+ "step": 2358
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7294,
+ "step": 2359
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.9061,
+ "step": 2360
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.866,
+ "step": 2361
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7369,
+ "step": 2362
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.8759,
+ "step": 2363
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7783,
+ "step": 2364
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.714,
+ "step": 2365
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 1.1328,
+ "step": 2366
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7433,
+ "step": 2367
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.6027,
+ "step": 2368
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7573,
+ "step": 2369
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 1.094,
+ "step": 2370
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.9156,
+ "step": 2371
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.9604,
+ "step": 2372
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.784,
+ "step": 2373
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 1.1049,
+ "step": 2374
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7491,
+ "step": 2375
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5781,
+ "step": 2376
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.6156,
+ "step": 2377
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7122,
+ "step": 2378
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.736,
+ "step": 2379
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.8568,
+ "step": 2380
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5992,
+ "step": 2381
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.511,
+ "step": 2382
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5195,
+ "step": 2383
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.8785,
+ "step": 2384
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5884,
+ "step": 2385
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.772,
+ "step": 2386
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 1.1139,
+ "step": 2387
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.8165,
+ "step": 2388
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5812,
+ "step": 2389
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5957,
+ "step": 2390
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.9219,
+ "step": 2391
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7419,
+ "step": 2392
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5522,
+ "step": 2393
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7682,
+ "step": 2394
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7594,
+ "step": 2395
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.732,
+ "step": 2396
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7478,
+ "step": 2397
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.6377,
+ "step": 2398
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.9237,
+ "step": 2399
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 1.0092,
+ "step": 2400
+ },
+ {
+ "epoch": 0.39,
+ "eval_loss": 0.754124641418457,
+ "eval_runtime": 131.5696,
+ "eval_samples_per_second": 7.601,
+ "eval_steps_per_second": 3.8,
+ "step": 2400
+ },
+ {
+ "epoch": 0.39,
+ "mmlu_eval_accuracy": 0.5488713452213336,
+ "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.5625,
+ "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
+ "mmlu_eval_accuracy_college_biology": 0.5625,
+ "mmlu_eval_accuracy_college_chemistry": 0.0,
+ "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
+ "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "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": 0.375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
+ "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.65625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
+ "mmlu_eval_accuracy_high_school_physics": 0.4117647058823529,
+ "mmlu_eval_accuracy_high_school_psychology": 0.75,
+ "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.46153846153846156,
+ "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+ "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+ "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": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+ "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
+ "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": 0.7428571428571429,
+ "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
+ "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
+ "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": 0.4444444444444444,
+ "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
+ "mmlu_loss": 1.223464566017878,
+ "step": 2400
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.7608,
+ "step": 2401
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.6459,
+ "step": 2402
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 1.0475,
+ "step": 2403
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0002,
+ "loss": 0.5478,
+ "step": 2404
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8448,
+ "step": 2405
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6834,
+ "step": 2406
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8275,
+ "step": 2407
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8986,
+ "step": 2408
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8555,
+ "step": 2409
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7626,
+ "step": 2410
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7321,
+ "step": 2411
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 1.0853,
+ "step": 2412
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8623,
+ "step": 2413
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7921,
+ "step": 2414
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.5459,
+ "step": 2415
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6145,
+ "step": 2416
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7401,
+ "step": 2417
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8739,
+ "step": 2418
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.4755,
+ "step": 2419
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7535,
+ "step": 2420
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 1.1141,
+ "step": 2421
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8943,
+ "step": 2422
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.9819,
+ "step": 2423
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7231,
+ "step": 2424
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.4587,
+ "step": 2425
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7576,
+ "step": 2426
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8637,
+ "step": 2427
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 1.0658,
+ "step": 2428
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.5253,
+ "step": 2429
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 1.01,
+ "step": 2430
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7207,
+ "step": 2431
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7255,
+ "step": 2432
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6631,
+ "step": 2433
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8594,
+ "step": 2434
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8288,
+ "step": 2435
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8355,
+ "step": 2436
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8267,
+ "step": 2437
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8579,
+ "step": 2438
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8763,
+ "step": 2439
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7434,
+ "step": 2440
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.833,
+ "step": 2441
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.9564,
+ "step": 2442
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.9249,
+ "step": 2443
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8565,
+ "step": 2444
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.5787,
+ "step": 2445
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.5709,
+ "step": 2446
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8362,
+ "step": 2447
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6556,
+ "step": 2448
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8708,
+ "step": 2449
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7568,
+ "step": 2450
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7025,
+ "step": 2451
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6675,
+ "step": 2452
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6966,
+ "step": 2453
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.9525,
+ "step": 2454
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8237,
+ "step": 2455
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6726,
+ "step": 2456
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6591,
+ "step": 2457
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7169,
+ "step": 2458
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.6239,
+ "step": 2459
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 1.0067,
+ "step": 2460
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.5825,
+ "step": 2461
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.8051,
+ "step": 2462
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.9009,
+ "step": 2463
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0002,
+ "loss": 0.7616,
+ "step": 2464
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.868,
+ "step": 2465
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6286,
+ "step": 2466
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.772,
+ "step": 2467
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7254,
+ "step": 2468
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8407,
+ "step": 2469
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7503,
+ "step": 2470
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6329,
+ "step": 2471
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.712,
+ "step": 2472
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9667,
+ "step": 2473
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7194,
+ "step": 2474
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6196,
+ "step": 2475
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8703,
+ "step": 2476
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6371,
+ "step": 2477
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7197,
+ "step": 2478
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8817,
+ "step": 2479
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.601,
+ "step": 2480
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.5872,
+ "step": 2481
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6102,
+ "step": 2482
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9108,
+ "step": 2483
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7453,
+ "step": 2484
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.5531,
+ "step": 2485
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7493,
+ "step": 2486
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9283,
+ "step": 2487
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 1.0349,
+ "step": 2488
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9457,
+ "step": 2489
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6669,
+ "step": 2490
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8719,
+ "step": 2491
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.4049,
+ "step": 2492
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6762,
+ "step": 2493
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7887,
+ "step": 2494
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7616,
+ "step": 2495
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6861,
+ "step": 2496
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6862,
+ "step": 2497
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7137,
+ "step": 2498
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7956,
+ "step": 2499
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7892,
+ "step": 2500
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9866,
+ "step": 2501
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6448,
+ "step": 2502
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8001,
+ "step": 2503
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8254,
+ "step": 2504
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9118,
+ "step": 2505
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.5775,
+ "step": 2506
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.5753,
+ "step": 2507
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 1.0153,
+ "step": 2508
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9019,
+ "step": 2509
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.5547,
+ "step": 2510
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8908,
+ "step": 2511
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6691,
+ "step": 2512
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.5401,
+ "step": 2513
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8547,
+ "step": 2514
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9645,
+ "step": 2515
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 1.102,
+ "step": 2516
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.8085,
+ "step": 2517
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7589,
+ "step": 2518
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.5556,
+ "step": 2519
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.6201,
+ "step": 2520
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7761,
+ "step": 2521
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.628,
+ "step": 2522
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9584,
+ "step": 2523
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.9155,
+ "step": 2524
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0002,
+ "loss": 0.7465,
+ "step": 2525
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7352,
+ "step": 2526
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7346,
+ "step": 2527
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.6878,
+ "step": 2528
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7621,
+ "step": 2529
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.5621,
+ "step": 2530
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8292,
+ "step": 2531
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7636,
+ "step": 2532
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7687,
+ "step": 2533
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.9402,
+ "step": 2534
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8057,
+ "step": 2535
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8209,
+ "step": 2536
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.9906,
+ "step": 2537
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8246,
+ "step": 2538
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.4754,
+ "step": 2539
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7922,
+ "step": 2540
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.818,
+ "step": 2541
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7198,
+ "step": 2542
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8984,
+ "step": 2543
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.5432,
+ "step": 2544
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7439,
+ "step": 2545
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.878,
+ "step": 2546
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.5922,
+ "step": 2547
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7512,
+ "step": 2548
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.901,
+ "step": 2549
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.6229,
+ "step": 2550
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.5738,
+ "step": 2551
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8911,
+ "step": 2552
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.2983,
+ "step": 2553
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.846,
+ "step": 2554
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7868,
+ "step": 2555
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7506,
+ "step": 2556
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7047,
+ "step": 2557
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7249,
+ "step": 2558
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.6154,
+ "step": 2559
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8087,
+ "step": 2560
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.9299,
+ "step": 2561
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8161,
+ "step": 2562
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 1.0844,
+ "step": 2563
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.863,
+ "step": 2564
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.4191,
+ "step": 2565
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7249,
+ "step": 2566
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7854,
+ "step": 2567
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.9462,
+ "step": 2568
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8241,
+ "step": 2569
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.5093,
+ "step": 2570
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8147,
+ "step": 2571
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7835,
+ "step": 2572
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 1.0354,
+ "step": 2573
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.666,
+ "step": 2574
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8681,
+ "step": 2575
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7828,
+ "step": 2576
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8671,
+ "step": 2577
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8738,
+ "step": 2578
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8246,
+ "step": 2579
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.6113,
+ "step": 2580
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.6819,
+ "step": 2581
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7734,
+ "step": 2582
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.7119,
+ "step": 2583
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.832,
+ "step": 2584
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8859,
+ "step": 2585
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.0002,
+ "loss": 0.8273,
+ "step": 2586
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.784,
+ "step": 2587
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.5618,
+ "step": 2588
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6341,
+ "step": 2589
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8338,
+ "step": 2590
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8473,
+ "step": 2591
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 1.0108,
+ "step": 2592
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7801,
+ "step": 2593
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6627,
+ "step": 2594
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.5425,
+ "step": 2595
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7461,
+ "step": 2596
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.4163,
+ "step": 2597
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8292,
+ "step": 2598
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.9198,
+ "step": 2599
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 1.0873,
+ "step": 2600
+ },
+ {
+ "epoch": 0.43,
+ "eval_loss": 0.7570953369140625,
+ "eval_runtime": 131.5868,
+ "eval_samples_per_second": 7.6,
+ "eval_steps_per_second": 3.8,
+ "step": 2600
+ },
+ {
+ "epoch": 0.43,
+ "mmlu_eval_accuracy": 0.5581231690028787,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.6875,
+ "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
+ "mmlu_eval_accuracy_college_biology": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.25,
+ "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
+ "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
+ "mmlu_eval_accuracy_electrical_engineering": 0.3125,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.59375,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
+ "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.6744186046511628,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
+ "mmlu_eval_accuracy_high_school_physics": 0.47058823529411764,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
+ "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
+ "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
+ "mmlu_eval_accuracy_management": 0.9090909090909091,
+ "mmlu_eval_accuracy_marketing": 0.8,
+ "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+ "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
+ "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
+ "mmlu_eval_accuracy_moral_scenarios": 0.36,
+ "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
+ "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
+ "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
+ "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
+ "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
+ "mmlu_eval_accuracy_public_relations": 0.6666666666666666,
+ "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
+ "mmlu_eval_accuracy_sociology": 0.9545454545454546,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
+ "mmlu_loss": 1.0118672493310261,
+ "step": 2600
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8037,
+ "step": 2601
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7605,
+ "step": 2602
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8871,
+ "step": 2603
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.3938,
+ "step": 2604
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6924,
+ "step": 2605
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7479,
+ "step": 2606
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8584,
+ "step": 2607
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.5362,
+ "step": 2608
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8184,
+ "step": 2609
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.9594,
+ "step": 2610
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6906,
+ "step": 2611
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6427,
+ "step": 2612
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 1.1499,
+ "step": 2613
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7943,
+ "step": 2614
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8735,
+ "step": 2615
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8639,
+ "step": 2616
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6335,
+ "step": 2617
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.9219,
+ "step": 2618
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8471,
+ "step": 2619
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7907,
+ "step": 2620
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6762,
+ "step": 2621
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.9186,
+ "step": 2622
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6569,
+ "step": 2623
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7081,
+ "step": 2624
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7623,
+ "step": 2625
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7855,
+ "step": 2626
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8738,
+ "step": 2627
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7203,
+ "step": 2628
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.4544,
+ "step": 2629
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7572,
+ "step": 2630
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.4496,
+ "step": 2631
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.5314,
+ "step": 2632
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8078,
+ "step": 2633
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8064,
+ "step": 2634
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.4411,
+ "step": 2635
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.9895,
+ "step": 2636
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8596,
+ "step": 2637
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.5229,
+ "step": 2638
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7742,
+ "step": 2639
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.9026,
+ "step": 2640
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.4865,
+ "step": 2641
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.4471,
+ "step": 2642
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.8359,
+ "step": 2643
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.7448,
+ "step": 2644
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.9158,
+ "step": 2645
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6196,
+ "step": 2646
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.0002,
+ "loss": 0.6178,
+ "step": 2647
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7435,
+ "step": 2648
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7007,
+ "step": 2649
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8002,
+ "step": 2650
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.426,
+ "step": 2651
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7232,
+ "step": 2652
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.722,
+ "step": 2653
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6358,
+ "step": 2654
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7446,
+ "step": 2655
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6941,
+ "step": 2656
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7354,
+ "step": 2657
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6183,
+ "step": 2658
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8515,
+ "step": 2659
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8874,
+ "step": 2660
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6188,
+ "step": 2661
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.9337,
+ "step": 2662
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7617,
+ "step": 2663
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.5362,
+ "step": 2664
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7865,
+ "step": 2665
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7054,
+ "step": 2666
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7383,
+ "step": 2667
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7314,
+ "step": 2668
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8373,
+ "step": 2669
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8596,
+ "step": 2670
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.5253,
+ "step": 2671
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8387,
+ "step": 2672
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8043,
+ "step": 2673
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6454,
+ "step": 2674
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8217,
+ "step": 2675
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7386,
+ "step": 2676
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7064,
+ "step": 2677
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 1.0354,
+ "step": 2678
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7412,
+ "step": 2679
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6881,
+ "step": 2680
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6349,
+ "step": 2681
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.9402,
+ "step": 2682
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 1.0755,
+ "step": 2683
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7402,
+ "step": 2684
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.663,
+ "step": 2685
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 1.0114,
+ "step": 2686
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7499,
+ "step": 2687
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7386,
+ "step": 2688
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8142,
+ "step": 2689
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.5404,
+ "step": 2690
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.914,
+ "step": 2691
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.6213,
+ "step": 2692
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 1.0168,
+ "step": 2693
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.727,
+ "step": 2694
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8568,
+ "step": 2695
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7015,
+ "step": 2696
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8244,
+ "step": 2697
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8432,
+ "step": 2698
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7202,
+ "step": 2699
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8496,
+ "step": 2700
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.5937,
+ "step": 2701
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 1.0566,
+ "step": 2702
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7125,
+ "step": 2703
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7536,
+ "step": 2704
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7104,
+ "step": 2705
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 1.11,
+ "step": 2706
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.8753,
+ "step": 2707
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0002,
+ "loss": 0.7139,
+ "step": 2708
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8618,
+ "step": 2709
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7046,
+ "step": 2710
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6036,
+ "step": 2711
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.9455,
+ "step": 2712
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8699,
+ "step": 2713
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6548,
+ "step": 2714
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.817,
+ "step": 2715
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.799,
+ "step": 2716
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8412,
+ "step": 2717
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8949,
+ "step": 2718
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.2124,
+ "step": 2719
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.1396,
+ "step": 2720
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7156,
+ "step": 2721
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8667,
+ "step": 2722
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7494,
+ "step": 2723
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8531,
+ "step": 2724
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7537,
+ "step": 2725
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7779,
+ "step": 2726
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.5883,
+ "step": 2727
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.0681,
+ "step": 2728
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.5107,
+ "step": 2729
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.0429,
+ "step": 2730
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7311,
+ "step": 2731
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.0758,
+ "step": 2732
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.5067,
+ "step": 2733
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.796,
+ "step": 2734
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7282,
+ "step": 2735
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.664,
+ "step": 2736
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.9432,
+ "step": 2737
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6337,
+ "step": 2738
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6284,
+ "step": 2739
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6309,
+ "step": 2740
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.792,
+ "step": 2741
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.9338,
+ "step": 2742
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.0635,
+ "step": 2743
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8012,
+ "step": 2744
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.0958,
+ "step": 2745
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8876,
+ "step": 2746
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.9664,
+ "step": 2747
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7179,
+ "step": 2748
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.609,
+ "step": 2749
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6172,
+ "step": 2750
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.1386,
+ "step": 2751
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7308,
+ "step": 2752
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.5839,
+ "step": 2753
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8759,
+ "step": 2754
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8531,
+ "step": 2755
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7749,
+ "step": 2756
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8547,
+ "step": 2757
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8355,
+ "step": 2758
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8094,
+ "step": 2759
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7815,
+ "step": 2760
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.858,
+ "step": 2761
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6334,
+ "step": 2762
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.8427,
+ "step": 2763
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.6815,
+ "step": 2764
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.9356,
+ "step": 2765
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.7899,
+ "step": 2766
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 1.0199,
+ "step": 2767
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.73,
+ "step": 2768
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.0002,
+ "loss": 0.9788,
+ "step": 2769
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8299,
+ "step": 2770
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7428,
+ "step": 2771
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9578,
+ "step": 2772
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6592,
+ "step": 2773
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8675,
+ "step": 2774
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7752,
+ "step": 2775
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6614,
+ "step": 2776
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.4909,
+ "step": 2777
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6649,
+ "step": 2778
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8044,
+ "step": 2779
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7484,
+ "step": 2780
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9253,
+ "step": 2781
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.673,
+ "step": 2782
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.574,
+ "step": 2783
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9379,
+ "step": 2784
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8074,
+ "step": 2785
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6783,
+ "step": 2786
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7403,
+ "step": 2787
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8888,
+ "step": 2788
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6428,
+ "step": 2789
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.5931,
+ "step": 2790
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8189,
+ "step": 2791
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.5879,
+ "step": 2792
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8604,
+ "step": 2793
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.5792,
+ "step": 2794
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8081,
+ "step": 2795
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7078,
+ "step": 2796
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7481,
+ "step": 2797
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7828,
+ "step": 2798
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7156,
+ "step": 2799
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9685,
+ "step": 2800
+ },
+ {
+ "epoch": 0.46,
+ "eval_loss": 0.7530816197395325,
+ "eval_runtime": 131.6881,
+ "eval_samples_per_second": 7.594,
+ "eval_steps_per_second": 3.797,
+ "step": 2800
+ },
+ {
+ "epoch": 0.46,
+ "mmlu_eval_accuracy": 0.5530778539637836,
+ "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
+ "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
+ "mmlu_eval_accuracy_astronomy": 0.5625,
+ "mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931,
+ "mmlu_eval_accuracy_college_biology": 0.6875,
+ "mmlu_eval_accuracy_college_chemistry": 0.125,
+ "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.7272727272727273,
+ "mmlu_eval_accuracy_college_physics": 0.8181818181818182,
+ "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
+ "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.4375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
+ "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.78125,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
+ "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
+ "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8,
+ "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
+ "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
+ "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666,
+ "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": 0.2727272727272727,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.8,
+ "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+ "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
+ "mmlu_eval_accuracy_moral_disputes": 0.5,
+ "mmlu_eval_accuracy_moral_scenarios": 0.39,
+ "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
+ "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
+ "mmlu_eval_accuracy_prehistory": 0.6,
+ "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
+ "mmlu_eval_accuracy_professional_law": 0.4294117647058823,
+ "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
+ "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
+ "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": 0.4444444444444444,
+ "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+ "mmlu_loss": 1.0421120278712352,
+ "step": 2800
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8098,
+ "step": 2801
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9656,
+ "step": 2802
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8368,
+ "step": 2803
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6248,
+ "step": 2804
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6017,
+ "step": 2805
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8374,
+ "step": 2806
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 1.1513,
+ "step": 2807
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8444,
+ "step": 2808
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8169,
+ "step": 2809
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8228,
+ "step": 2810
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 1.0234,
+ "step": 2811
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8949,
+ "step": 2812
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.773,
+ "step": 2813
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9613,
+ "step": 2814
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.86,
+ "step": 2815
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6061,
+ "step": 2816
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8257,
+ "step": 2817
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8071,
+ "step": 2818
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 1.0167,
+ "step": 2819
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7324,
+ "step": 2820
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9094,
+ "step": 2821
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6907,
+ "step": 2822
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.6652,
+ "step": 2823
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8172,
+ "step": 2824
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.5711,
+ "step": 2825
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.5757,
+ "step": 2826
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7104,
+ "step": 2827
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.9677,
+ "step": 2828
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.8585,
+ "step": 2829
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.0002,
+ "loss": 0.7479,
+ "step": 2830
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.9201,
+ "step": 2831
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.9481,
+ "step": 2832
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.3661,
+ "step": 2833
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8192,
+ "step": 2834
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 1.4574,
+ "step": 2835
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8958,
+ "step": 2836
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7935,
+ "step": 2837
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8905,
+ "step": 2838
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6342,
+ "step": 2839
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.978,
+ "step": 2840
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7466,
+ "step": 2841
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8611,
+ "step": 2842
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7029,
+ "step": 2843
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6195,
+ "step": 2844
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7525,
+ "step": 2845
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.5484,
+ "step": 2846
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.9817,
+ "step": 2847
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.9278,
+ "step": 2848
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.671,
+ "step": 2849
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6675,
+ "step": 2850
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6611,
+ "step": 2851
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7286,
+ "step": 2852
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.9665,
+ "step": 2853
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.942,
+ "step": 2854
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6561,
+ "step": 2855
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.9651,
+ "step": 2856
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.5337,
+ "step": 2857
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.792,
+ "step": 2858
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 1.1503,
+ "step": 2859
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8918,
+ "step": 2860
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6808,
+ "step": 2861
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.602,
+ "step": 2862
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6973,
+ "step": 2863
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8459,
+ "step": 2864
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8665,
+ "step": 2865
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8927,
+ "step": 2866
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6571,
+ "step": 2867
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7067,
+ "step": 2868
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.5655,
+ "step": 2869
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.9379,
+ "step": 2870
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6435,
+ "step": 2871
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7375,
+ "step": 2872
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 1.0998,
+ "step": 2873
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8005,
+ "step": 2874
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6864,
+ "step": 2875
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6895,
+ "step": 2876
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.767,
+ "step": 2877
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6207,
+ "step": 2878
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.495,
+ "step": 2879
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7,
+ "step": 2880
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8074,
+ "step": 2881
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7798,
+ "step": 2882
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6458,
+ "step": 2883
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.8946,
+ "step": 2884
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.6924,
+ "step": 2885
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7649,
+ "step": 2886
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.7728,
+ "step": 2887
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.5118,
+ "step": 2888
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 0.826,
+ "step": 2889
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.0002,
+ "loss": 1.148,
+ "step": 2890
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6991,
+ "step": 2891
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.609,
+ "step": 2892
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7536,
+ "step": 2893
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7738,
+ "step": 2894
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6682,
+ "step": 2895
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5314,
+ "step": 2896
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.4833,
+ "step": 2897
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.59,
+ "step": 2898
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6338,
+ "step": 2899
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.4124,
+ "step": 2900
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7652,
+ "step": 2901
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7304,
+ "step": 2902
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.9406,
+ "step": 2903
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5314,
+ "step": 2904
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.8301,
+ "step": 2905
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.8097,
+ "step": 2906
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6065,
+ "step": 2907
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.405,
+ "step": 2908
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7592,
+ "step": 2909
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6914,
+ "step": 2910
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 1.0258,
+ "step": 2911
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.929,
+ "step": 2912
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.733,
+ "step": 2913
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6774,
+ "step": 2914
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.803,
+ "step": 2915
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5683,
+ "step": 2916
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.639,
+ "step": 2917
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.8319,
+ "step": 2918
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7761,
+ "step": 2919
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.9458,
+ "step": 2920
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7242,
+ "step": 2921
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7603,
+ "step": 2922
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7693,
+ "step": 2923
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.8695,
+ "step": 2924
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 1.1133,
+ "step": 2925
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5037,
+ "step": 2926
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7865,
+ "step": 2927
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6465,
+ "step": 2928
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.9632,
+ "step": 2929
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6949,
+ "step": 2930
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6652,
+ "step": 2931
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.4068,
+ "step": 2932
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7214,
+ "step": 2933
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 1.0307,
+ "step": 2934
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6681,
+ "step": 2935
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5881,
+ "step": 2936
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7022,
+ "step": 2937
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.8272,
+ "step": 2938
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5194,
+ "step": 2939
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5115,
+ "step": 2940
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.8319,
+ "step": 2941
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.7172,
+ "step": 2942
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.8612,
+ "step": 2943
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.9976,
+ "step": 2944
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.6673,
+ "step": 2945
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.979,
+ "step": 2946
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.9404,
+ "step": 2947
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.885,
+ "step": 2948
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.9053,
+ "step": 2949
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 0.5979,
+ "step": 2950
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.0002,
+ "loss": 1.0637,
+ "step": 2951
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8182,
+ "step": 2952
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7503,
+ "step": 2953
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7866,
+ "step": 2954
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8625,
+ "step": 2955
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8117,
+ "step": 2956
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.66,
+ "step": 2957
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.0334,
+ "step": 2958
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.0704,
+ "step": 2959
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8582,
+ "step": 2960
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7488,
+ "step": 2961
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.1001,
+ "step": 2962
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7449,
+ "step": 2963
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.0588,
+ "step": 2964
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7596,
+ "step": 2965
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7089,
+ "step": 2966
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7837,
+ "step": 2967
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.056,
+ "step": 2968
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.996,
+ "step": 2969
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8761,
+ "step": 2970
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8035,
+ "step": 2971
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7216,
+ "step": 2972
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7333,
+ "step": 2973
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6356,
+ "step": 2974
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7532,
+ "step": 2975
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6252,
+ "step": 2976
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7604,
+ "step": 2977
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.734,
+ "step": 2978
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.938,
+ "step": 2979
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6099,
+ "step": 2980
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.9421,
+ "step": 2981
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7869,
+ "step": 2982
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.9485,
+ "step": 2983
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8903,
+ "step": 2984
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.1069,
+ "step": 2985
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.5866,
+ "step": 2986
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6901,
+ "step": 2987
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.715,
+ "step": 2988
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.0569,
+ "step": 2989
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.5055,
+ "step": 2990
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.9878,
+ "step": 2991
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8048,
+ "step": 2992
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.9757,
+ "step": 2993
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.4064,
+ "step": 2994
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8918,
+ "step": 2995
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.703,
+ "step": 2996
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.0976,
+ "step": 2997
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.9383,
+ "step": 2998
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.5379,
+ "step": 2999
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.5319,
+ "step": 3000
+ },
+ {
+ "epoch": 0.49,
+ "eval_loss": 0.7562741041183472,
+ "eval_runtime": 131.6412,
+ "eval_samples_per_second": 7.596,
+ "eval_steps_per_second": 3.798,
+ "step": 3000
+ },
+ {
+ "epoch": 0.49,
+ "mmlu_eval_accuracy": 0.5657216197300882,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.75,
+ "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
+ "mmlu_eval_accuracy_college_biology": 0.6875,
+ "mmlu_eval_accuracy_college_chemistry": 0.125,
+ "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": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
+ "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5625,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.6,
+ "mmlu_eval_accuracy_high_school_biology": 0.6875,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "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": 0.13793103448275862,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
+ "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
+ "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.8076923076923077,
+ "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
+ "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": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.9090909090909091,
+ "mmlu_eval_accuracy_marketing": 0.76,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
+ "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
+ "mmlu_eval_accuracy_moral_scenarios": 0.39,
+ "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
+ "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
+ "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
+ "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
+ "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
+ "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
+ "mmlu_eval_accuracy_sociology": 0.9090909090909091,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
+ "mmlu_loss": 1.091038808109866,
+ "step": 3000
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.8781,
+ "step": 3001
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6487,
+ "step": 3002
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.0339,
+ "step": 3003
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.9526,
+ "step": 3004
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.5935,
+ "step": 3005
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6707,
+ "step": 3006
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6641,
+ "step": 3007
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6717,
+ "step": 3008
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.7678,
+ "step": 3009
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 1.0386,
+ "step": 3010
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.772,
+ "step": 3011
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.0002,
+ "loss": 0.6929,
+ "step": 3012
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8485,
+ "step": 3013
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.4891,
+ "step": 3014
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7512,
+ "step": 3015
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7122,
+ "step": 3016
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.838,
+ "step": 3017
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.9496,
+ "step": 3018
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 1.0716,
+ "step": 3019
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7698,
+ "step": 3020
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8338,
+ "step": 3021
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7043,
+ "step": 3022
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.6221,
+ "step": 3023
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8421,
+ "step": 3024
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7658,
+ "step": 3025
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.9514,
+ "step": 3026
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7065,
+ "step": 3027
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.5111,
+ "step": 3028
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8038,
+ "step": 3029
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.808,
+ "step": 3030
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8133,
+ "step": 3031
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7117,
+ "step": 3032
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.5798,
+ "step": 3033
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8345,
+ "step": 3034
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8441,
+ "step": 3035
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.6213,
+ "step": 3036
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.618,
+ "step": 3037
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7699,
+ "step": 3038
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7391,
+ "step": 3039
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7765,
+ "step": 3040
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7144,
+ "step": 3041
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.5507,
+ "step": 3042
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 1.0901,
+ "step": 3043
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7746,
+ "step": 3044
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.9164,
+ "step": 3045
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.614,
+ "step": 3046
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.6068,
+ "step": 3047
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.9984,
+ "step": 3048
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.9879,
+ "step": 3049
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.9075,
+ "step": 3050
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.6982,
+ "step": 3051
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.927,
+ "step": 3052
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7391,
+ "step": 3053
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.5627,
+ "step": 3054
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7256,
+ "step": 3055
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7678,
+ "step": 3056
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.5627,
+ "step": 3057
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7685,
+ "step": 3058
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8924,
+ "step": 3059
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.882,
+ "step": 3060
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.6875,
+ "step": 3061
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8683,
+ "step": 3062
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.768,
+ "step": 3063
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7801,
+ "step": 3064
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.9501,
+ "step": 3065
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.6802,
+ "step": 3066
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.6508,
+ "step": 3067
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.611,
+ "step": 3068
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7241,
+ "step": 3069
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.5179,
+ "step": 3070
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8495,
+ "step": 3071
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.7408,
+ "step": 3072
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.0002,
+ "loss": 0.8953,
+ "step": 3073
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9477,
+ "step": 3074
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9135,
+ "step": 3075
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6908,
+ "step": 3076
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8108,
+ "step": 3077
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8012,
+ "step": 3078
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8357,
+ "step": 3079
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7435,
+ "step": 3080
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.5473,
+ "step": 3081
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8405,
+ "step": 3082
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7903,
+ "step": 3083
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6618,
+ "step": 3084
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9276,
+ "step": 3085
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7161,
+ "step": 3086
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.654,
+ "step": 3087
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8015,
+ "step": 3088
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 1.0513,
+ "step": 3089
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8357,
+ "step": 3090
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6062,
+ "step": 3091
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.695,
+ "step": 3092
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7898,
+ "step": 3093
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7437,
+ "step": 3094
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7369,
+ "step": 3095
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.729,
+ "step": 3096
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7076,
+ "step": 3097
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.857,
+ "step": 3098
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6087,
+ "step": 3099
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8445,
+ "step": 3100
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.5858,
+ "step": 3101
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.738,
+ "step": 3102
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6476,
+ "step": 3103
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.697,
+ "step": 3104
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8828,
+ "step": 3105
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9003,
+ "step": 3106
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8344,
+ "step": 3107
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7103,
+ "step": 3108
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.637,
+ "step": 3109
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 1.1437,
+ "step": 3110
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7657,
+ "step": 3111
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6144,
+ "step": 3112
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 1.0407,
+ "step": 3113
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.596,
+ "step": 3114
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7938,
+ "step": 3115
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 1.3087,
+ "step": 3116
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9146,
+ "step": 3117
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8138,
+ "step": 3118
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7663,
+ "step": 3119
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8417,
+ "step": 3120
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.5309,
+ "step": 3121
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6276,
+ "step": 3122
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 1.0024,
+ "step": 3123
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9719,
+ "step": 3124
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.5478,
+ "step": 3125
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8333,
+ "step": 3126
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.5024,
+ "step": 3127
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.88,
+ "step": 3128
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7488,
+ "step": 3129
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9953,
+ "step": 3130
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.8652,
+ "step": 3131
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.6715,
+ "step": 3132
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.9498,
+ "step": 3133
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.0002,
+ "loss": 0.7319,
+ "step": 3134
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7297,
+ "step": 3135
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8408,
+ "step": 3136
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.955,
+ "step": 3137
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7728,
+ "step": 3138
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6917,
+ "step": 3139
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.929,
+ "step": 3140
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8458,
+ "step": 3141
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.915,
+ "step": 3142
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6284,
+ "step": 3143
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7155,
+ "step": 3144
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6397,
+ "step": 3145
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8762,
+ "step": 3146
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7542,
+ "step": 3147
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7003,
+ "step": 3148
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.9587,
+ "step": 3149
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.9843,
+ "step": 3150
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6883,
+ "step": 3151
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6476,
+ "step": 3152
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 1.017,
+ "step": 3153
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8428,
+ "step": 3154
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7956,
+ "step": 3155
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7353,
+ "step": 3156
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6851,
+ "step": 3157
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.5245,
+ "step": 3158
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7067,
+ "step": 3159
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8815,
+ "step": 3160
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7119,
+ "step": 3161
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.9293,
+ "step": 3162
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.9643,
+ "step": 3163
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6312,
+ "step": 3164
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7596,
+ "step": 3165
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6358,
+ "step": 3166
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8905,
+ "step": 3167
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6303,
+ "step": 3168
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7482,
+ "step": 3169
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.5716,
+ "step": 3170
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8766,
+ "step": 3171
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.9254,
+ "step": 3172
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7601,
+ "step": 3173
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7328,
+ "step": 3174
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 1.0365,
+ "step": 3175
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 1.131,
+ "step": 3176
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6634,
+ "step": 3177
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 1.039,
+ "step": 3178
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7729,
+ "step": 3179
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6849,
+ "step": 3180
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6879,
+ "step": 3181
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7109,
+ "step": 3182
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7998,
+ "step": 3183
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6229,
+ "step": 3184
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.711,
+ "step": 3185
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.7155,
+ "step": 3186
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8632,
+ "step": 3187
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 1.0824,
+ "step": 3188
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.9087,
+ "step": 3189
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8377,
+ "step": 3190
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.9201,
+ "step": 3191
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6367,
+ "step": 3192
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.8713,
+ "step": 3193
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.6782,
+ "step": 3194
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0002,
+ "loss": 0.657,
+ "step": 3195
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9913,
+ "step": 3196
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7744,
+ "step": 3197
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7068,
+ "step": 3198
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.5735,
+ "step": 3199
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7163,
+ "step": 3200
+ },
+ {
+ "epoch": 0.53,
+ "eval_loss": 0.7516635656356812,
+ "eval_runtime": 131.726,
+ "eval_samples_per_second": 7.592,
+ "eval_steps_per_second": 3.796,
+ "step": 3200
+ },
+ {
+ "epoch": 0.53,
+ "mmlu_eval_accuracy": 0.5737180213767722,
+ "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.6875,
+ "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": 0.25,
+ "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": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "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": 0.4146341463414634,
+ "mmlu_eval_accuracy_formal_logic": 0.42857142857142855,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.71875,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
+ "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
+ "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+ "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.8,
+ "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": 0.34,
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
+ "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
+ "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
+ "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
+ "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
+ "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
+ "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+ "mmlu_eval_accuracy_sociology": 0.8636363636363636,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.4444444444444444,
+ "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
+ "mmlu_loss": 0.8832502972636148,
+ "step": 3200
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8177,
+ "step": 3201
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.5147,
+ "step": 3202
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9274,
+ "step": 3203
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.6697,
+ "step": 3204
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.727,
+ "step": 3205
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.6729,
+ "step": 3206
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7744,
+ "step": 3207
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9278,
+ "step": 3208
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.676,
+ "step": 3209
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8493,
+ "step": 3210
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9212,
+ "step": 3211
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7283,
+ "step": 3212
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9432,
+ "step": 3213
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.6467,
+ "step": 3214
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7879,
+ "step": 3215
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8767,
+ "step": 3216
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.668,
+ "step": 3217
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8151,
+ "step": 3218
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.5956,
+ "step": 3219
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8103,
+ "step": 3220
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7316,
+ "step": 3221
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.665,
+ "step": 3222
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.6783,
+ "step": 3223
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9532,
+ "step": 3224
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.5412,
+ "step": 3225
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7871,
+ "step": 3226
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7102,
+ "step": 3227
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.882,
+ "step": 3228
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8141,
+ "step": 3229
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7232,
+ "step": 3230
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9752,
+ "step": 3231
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9711,
+ "step": 3232
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7897,
+ "step": 3233
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.752,
+ "step": 3234
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7319,
+ "step": 3235
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.731,
+ "step": 3236
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.6455,
+ "step": 3237
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.5552,
+ "step": 3238
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7528,
+ "step": 3239
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.5483,
+ "step": 3240
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7164,
+ "step": 3241
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.3298,
+ "step": 3242
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7539,
+ "step": 3243
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9876,
+ "step": 3244
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.4657,
+ "step": 3245
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7068,
+ "step": 3246
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.6767,
+ "step": 3247
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9049,
+ "step": 3248
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8022,
+ "step": 3249
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.7796,
+ "step": 3250
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.9956,
+ "step": 3251
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.6867,
+ "step": 3252
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.4339,
+ "step": 3253
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.775,
+ "step": 3254
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.802,
+ "step": 3255
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.0002,
+ "loss": 0.8255,
+ "step": 3256
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7897,
+ "step": 3257
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7778,
+ "step": 3258
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.6077,
+ "step": 3259
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.765,
+ "step": 3260
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.6567,
+ "step": 3261
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.6769,
+ "step": 3262
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.712,
+ "step": 3263
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.836,
+ "step": 3264
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 1.0661,
+ "step": 3265
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.6398,
+ "step": 3266
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8791,
+ "step": 3267
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.5367,
+ "step": 3268
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8647,
+ "step": 3269
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.901,
+ "step": 3270
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.553,
+ "step": 3271
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.59,
+ "step": 3272
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7861,
+ "step": 3273
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7848,
+ "step": 3274
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.6838,
+ "step": 3275
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.9176,
+ "step": 3276
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7397,
+ "step": 3277
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8605,
+ "step": 3278
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.5378,
+ "step": 3279
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7457,
+ "step": 3280
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8959,
+ "step": 3281
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.9157,
+ "step": 3282
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7365,
+ "step": 3283
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7628,
+ "step": 3284
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7365,
+ "step": 3285
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8435,
+ "step": 3286
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.5414,
+ "step": 3287
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.528,
+ "step": 3288
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8281,
+ "step": 3289
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.4378,
+ "step": 3290
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8019,
+ "step": 3291
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 1.0369,
+ "step": 3292
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.87,
+ "step": 3293
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7197,
+ "step": 3294
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7688,
+ "step": 3295
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7638,
+ "step": 3296
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.9255,
+ "step": 3297
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 1.0927,
+ "step": 3298
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7159,
+ "step": 3299
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8979,
+ "step": 3300
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8632,
+ "step": 3301
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8038,
+ "step": 3302
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7759,
+ "step": 3303
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7496,
+ "step": 3304
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.5661,
+ "step": 3305
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8136,
+ "step": 3306
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8669,
+ "step": 3307
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7742,
+ "step": 3308
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7886,
+ "step": 3309
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.7035,
+ "step": 3310
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.788,
+ "step": 3311
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.9369,
+ "step": 3312
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.9092,
+ "step": 3313
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.775,
+ "step": 3314
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.4413,
+ "step": 3315
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.501,
+ "step": 3316
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0002,
+ "loss": 0.8051,
+ "step": 3317
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.675,
+ "step": 3318
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5488,
+ "step": 3319
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7659,
+ "step": 3320
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7587,
+ "step": 3321
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8153,
+ "step": 3322
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8325,
+ "step": 3323
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8006,
+ "step": 3324
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5756,
+ "step": 3325
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8116,
+ "step": 3326
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8767,
+ "step": 3327
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5521,
+ "step": 3328
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7448,
+ "step": 3329
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5055,
+ "step": 3330
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8363,
+ "step": 3331
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.0262,
+ "step": 3332
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6862,
+ "step": 3333
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6646,
+ "step": 3334
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.9674,
+ "step": 3335
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6988,
+ "step": 3336
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7075,
+ "step": 3337
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6892,
+ "step": 3338
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.914,
+ "step": 3339
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7156,
+ "step": 3340
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.0676,
+ "step": 3341
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7941,
+ "step": 3342
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6558,
+ "step": 3343
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.002,
+ "step": 3344
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.1118,
+ "step": 3345
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7156,
+ "step": 3346
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6462,
+ "step": 3347
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5919,
+ "step": 3348
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.1466,
+ "step": 3349
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7977,
+ "step": 3350
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5158,
+ "step": 3351
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.904,
+ "step": 3352
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8702,
+ "step": 3353
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.9045,
+ "step": 3354
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.0569,
+ "step": 3355
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.63,
+ "step": 3356
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8772,
+ "step": 3357
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5319,
+ "step": 3358
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.0922,
+ "step": 3359
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5511,
+ "step": 3360
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8385,
+ "step": 3361
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.0059,
+ "step": 3362
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7977,
+ "step": 3363
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7771,
+ "step": 3364
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.9893,
+ "step": 3365
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7589,
+ "step": 3366
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6704,
+ "step": 3367
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.6237,
+ "step": 3368
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 1.049,
+ "step": 3369
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8292,
+ "step": 3370
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.5175,
+ "step": 3371
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.632,
+ "step": 3372
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7679,
+ "step": 3373
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8408,
+ "step": 3374
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8201,
+ "step": 3375
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.7912,
+ "step": 3376
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0002,
+ "loss": 0.8123,
+ "step": 3377
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5863,
+ "step": 3378
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7737,
+ "step": 3379
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.549,
+ "step": 3380
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7419,
+ "step": 3381
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.923,
+ "step": 3382
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7479,
+ "step": 3383
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5268,
+ "step": 3384
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7313,
+ "step": 3385
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.9283,
+ "step": 3386
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7676,
+ "step": 3387
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.9106,
+ "step": 3388
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.9585,
+ "step": 3389
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5391,
+ "step": 3390
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.745,
+ "step": 3391
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 1.3179,
+ "step": 3392
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7739,
+ "step": 3393
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5233,
+ "step": 3394
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6186,
+ "step": 3395
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8163,
+ "step": 3396
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8342,
+ "step": 3397
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7069,
+ "step": 3398
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.587,
+ "step": 3399
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.595,
+ "step": 3400
+ },
+ {
+ "epoch": 0.56,
+ "eval_loss": 0.749654233455658,
+ "eval_runtime": 131.6498,
+ "eval_samples_per_second": 7.596,
+ "eval_steps_per_second": 3.798,
+ "step": 3400
+ },
+ {
+ "epoch": 0.56,
+ "mmlu_eval_accuracy": 0.574503484588032,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.6875,
+ "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": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
+ "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
+ "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.625,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.6875,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
+ "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
+ "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
+ "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.88,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
+ "mmlu_eval_accuracy_moral_disputes": 0.5,
+ "mmlu_eval_accuracy_moral_scenarios": 0.26,
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
+ "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
+ "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
+ "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
+ "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": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.3888888888888889,
+ "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
+ "mmlu_loss": 1.0488827172326356,
+ "step": 3400
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7714,
+ "step": 3401
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6509,
+ "step": 3402
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8402,
+ "step": 3403
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7598,
+ "step": 3404
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.9918,
+ "step": 3405
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7602,
+ "step": 3406
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6071,
+ "step": 3407
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7109,
+ "step": 3408
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8946,
+ "step": 3409
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6201,
+ "step": 3410
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7355,
+ "step": 3411
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6829,
+ "step": 3412
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.9137,
+ "step": 3413
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5523,
+ "step": 3414
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6845,
+ "step": 3415
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 1.1485,
+ "step": 3416
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8524,
+ "step": 3417
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.934,
+ "step": 3418
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5627,
+ "step": 3419
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.9024,
+ "step": 3420
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7054,
+ "step": 3421
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7798,
+ "step": 3422
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8881,
+ "step": 3423
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6062,
+ "step": 3424
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6502,
+ "step": 3425
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.9398,
+ "step": 3426
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8441,
+ "step": 3427
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5296,
+ "step": 3428
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6628,
+ "step": 3429
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5828,
+ "step": 3430
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.813,
+ "step": 3431
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8269,
+ "step": 3432
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5469,
+ "step": 3433
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7982,
+ "step": 3434
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.5248,
+ "step": 3435
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.8588,
+ "step": 3436
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.7306,
+ "step": 3437
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0002,
+ "loss": 0.6578,
+ "step": 3438
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9038,
+ "step": 3439
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7253,
+ "step": 3440
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8225,
+ "step": 3441
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8381,
+ "step": 3442
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.6786,
+ "step": 3443
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9323,
+ "step": 3444
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.4972,
+ "step": 3445
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9031,
+ "step": 3446
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 1.0749,
+ "step": 3447
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7974,
+ "step": 3448
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9209,
+ "step": 3449
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7745,
+ "step": 3450
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8584,
+ "step": 3451
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.4844,
+ "step": 3452
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7297,
+ "step": 3453
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7858,
+ "step": 3454
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8104,
+ "step": 3455
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9209,
+ "step": 3456
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8718,
+ "step": 3457
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.533,
+ "step": 3458
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8577,
+ "step": 3459
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9081,
+ "step": 3460
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.6779,
+ "step": 3461
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 1.0281,
+ "step": 3462
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7136,
+ "step": 3463
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7337,
+ "step": 3464
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 1.1414,
+ "step": 3465
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.615,
+ "step": 3466
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8462,
+ "step": 3467
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.6567,
+ "step": 3468
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9807,
+ "step": 3469
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.79,
+ "step": 3470
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 1.1021,
+ "step": 3471
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7497,
+ "step": 3472
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9982,
+ "step": 3473
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9731,
+ "step": 3474
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 1.0387,
+ "step": 3475
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.5922,
+ "step": 3476
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8949,
+ "step": 3477
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7119,
+ "step": 3478
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.6747,
+ "step": 3479
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.5461,
+ "step": 3480
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.899,
+ "step": 3481
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 1.0509,
+ "step": 3482
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 1.0077,
+ "step": 3483
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.5613,
+ "step": 3484
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8587,
+ "step": 3485
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9096,
+ "step": 3486
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9492,
+ "step": 3487
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.5355,
+ "step": 3488
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8666,
+ "step": 3489
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8679,
+ "step": 3490
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.4945,
+ "step": 3491
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8946,
+ "step": 3492
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7361,
+ "step": 3493
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7939,
+ "step": 3494
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.6511,
+ "step": 3495
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7011,
+ "step": 3496
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.9052,
+ "step": 3497
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.7592,
+ "step": 3498
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.0002,
+ "loss": 0.8194,
+ "step": 3499
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6785,
+ "step": 3500
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7941,
+ "step": 3501
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8527,
+ "step": 3502
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9948,
+ "step": 3503
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.5943,
+ "step": 3504
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6741,
+ "step": 3505
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 1.2698,
+ "step": 3506
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6925,
+ "step": 3507
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6545,
+ "step": 3508
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 1.0184,
+ "step": 3509
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9142,
+ "step": 3510
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8805,
+ "step": 3511
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8892,
+ "step": 3512
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.5916,
+ "step": 3513
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6512,
+ "step": 3514
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6256,
+ "step": 3515
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.4009,
+ "step": 3516
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 1.0093,
+ "step": 3517
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7941,
+ "step": 3518
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7197,
+ "step": 3519
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8771,
+ "step": 3520
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8182,
+ "step": 3521
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9429,
+ "step": 3522
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8493,
+ "step": 3523
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9307,
+ "step": 3524
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9908,
+ "step": 3525
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7679,
+ "step": 3526
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6472,
+ "step": 3527
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.4572,
+ "step": 3528
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.4444,
+ "step": 3529
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8649,
+ "step": 3530
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.795,
+ "step": 3531
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6956,
+ "step": 3532
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9454,
+ "step": 3533
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9834,
+ "step": 3534
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8243,
+ "step": 3535
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9817,
+ "step": 3536
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.5175,
+ "step": 3537
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.5396,
+ "step": 3538
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9076,
+ "step": 3539
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7235,
+ "step": 3540
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7607,
+ "step": 3541
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7125,
+ "step": 3542
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9536,
+ "step": 3543
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7611,
+ "step": 3544
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9378,
+ "step": 3545
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8266,
+ "step": 3546
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8743,
+ "step": 3547
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 1.0378,
+ "step": 3548
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.969,
+ "step": 3549
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6474,
+ "step": 3550
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.5435,
+ "step": 3551
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.9508,
+ "step": 3552
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.5541,
+ "step": 3553
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7572,
+ "step": 3554
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.6647,
+ "step": 3555
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.902,
+ "step": 3556
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.8061,
+ "step": 3557
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.954,
+ "step": 3558
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7288,
+ "step": 3559
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.0002,
+ "loss": 0.7382,
+ "step": 3560
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6404,
+ "step": 3561
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.9829,
+ "step": 3562
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.827,
+ "step": 3563
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6734,
+ "step": 3564
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7555,
+ "step": 3565
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8572,
+ "step": 3566
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.9734,
+ "step": 3567
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.9008,
+ "step": 3568
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 1.0214,
+ "step": 3569
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.614,
+ "step": 3570
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6459,
+ "step": 3571
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7835,
+ "step": 3572
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8504,
+ "step": 3573
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7896,
+ "step": 3574
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 1.0251,
+ "step": 3575
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.874,
+ "step": 3576
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7773,
+ "step": 3577
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.5916,
+ "step": 3578
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7578,
+ "step": 3579
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8065,
+ "step": 3580
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8056,
+ "step": 3581
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7554,
+ "step": 3582
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7562,
+ "step": 3583
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6385,
+ "step": 3584
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7085,
+ "step": 3585
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.895,
+ "step": 3586
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.9523,
+ "step": 3587
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7818,
+ "step": 3588
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6468,
+ "step": 3589
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.9299,
+ "step": 3590
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.4926,
+ "step": 3591
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8229,
+ "step": 3592
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6897,
+ "step": 3593
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8863,
+ "step": 3594
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8045,
+ "step": 3595
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.742,
+ "step": 3596
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7566,
+ "step": 3597
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6584,
+ "step": 3598
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.621,
+ "step": 3599
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8159,
+ "step": 3600
+ },
+ {
+ "epoch": 0.59,
+ "eval_loss": 0.7506269812583923,
+ "eval_runtime": 131.59,
+ "eval_samples_per_second": 7.599,
+ "eval_steps_per_second": 3.8,
+ "step": 3600
+ },
+ {
+ "epoch": 0.59,
+ "mmlu_eval_accuracy": 0.5826522403805086,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.8125,
+ "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
+ "mmlu_eval_accuracy_college_biology": 0.5625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
+ "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5,
+ "mmlu_eval_accuracy_econometrics": 0.5,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
+ "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.71875,
+ "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": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
+ "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
+ "mmlu_eval_accuracy_high_school_psychology": 0.85,
+ "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
+ "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
+ "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": 0.5454545454545454,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.88,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
+ "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
+ "mmlu_eval_accuracy_moral_scenarios": 0.35,
+ "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
+ "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
+ "mmlu_eval_accuracy_prehistory": 0.6,
+ "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
+ "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
+ "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+ "mmlu_eval_accuracy_sociology": 0.8181818181818182,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.4444444444444444,
+ "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+ "mmlu_loss": 1.1334416739078477,
+ "step": 3600
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6198,
+ "step": 3601
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.5997,
+ "step": 3602
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.798,
+ "step": 3603
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8733,
+ "step": 3604
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.661,
+ "step": 3605
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.4949,
+ "step": 3606
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7743,
+ "step": 3607
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.9543,
+ "step": 3608
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 1.0559,
+ "step": 3609
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.7182,
+ "step": 3610
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.9093,
+ "step": 3611
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8173,
+ "step": 3612
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.8253,
+ "step": 3613
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.5855,
+ "step": 3614
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.5945,
+ "step": 3615
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6092,
+ "step": 3616
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6732,
+ "step": 3617
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6782,
+ "step": 3618
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6654,
+ "step": 3619
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.6727,
+ "step": 3620
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.0002,
+ "loss": 0.4883,
+ "step": 3621
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7042,
+ "step": 3622
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6871,
+ "step": 3623
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8523,
+ "step": 3624
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7614,
+ "step": 3625
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.9922,
+ "step": 3626
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8352,
+ "step": 3627
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.796,
+ "step": 3628
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7674,
+ "step": 3629
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8229,
+ "step": 3630
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7351,
+ "step": 3631
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8084,
+ "step": 3632
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6287,
+ "step": 3633
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6576,
+ "step": 3634
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.433,
+ "step": 3635
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.9451,
+ "step": 3636
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8166,
+ "step": 3637
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8891,
+ "step": 3638
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6921,
+ "step": 3639
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.4516,
+ "step": 3640
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6829,
+ "step": 3641
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.9014,
+ "step": 3642
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.5958,
+ "step": 3643
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.3924,
+ "step": 3644
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7241,
+ "step": 3645
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7725,
+ "step": 3646
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8577,
+ "step": 3647
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7656,
+ "step": 3648
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 1.0481,
+ "step": 3649
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.9214,
+ "step": 3650
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6203,
+ "step": 3651
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8047,
+ "step": 3652
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8814,
+ "step": 3653
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.462,
+ "step": 3654
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7891,
+ "step": 3655
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6932,
+ "step": 3656
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6169,
+ "step": 3657
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6512,
+ "step": 3658
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8812,
+ "step": 3659
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8939,
+ "step": 3660
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.989,
+ "step": 3661
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7318,
+ "step": 3662
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8295,
+ "step": 3663
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.447,
+ "step": 3664
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.758,
+ "step": 3665
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7289,
+ "step": 3666
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6583,
+ "step": 3667
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.5963,
+ "step": 3668
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7499,
+ "step": 3669
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7756,
+ "step": 3670
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7339,
+ "step": 3671
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6942,
+ "step": 3672
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.8506,
+ "step": 3673
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6451,
+ "step": 3674
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7236,
+ "step": 3675
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.9986,
+ "step": 3676
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.5004,
+ "step": 3677
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.7134,
+ "step": 3678
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 1.0034,
+ "step": 3679
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.746,
+ "step": 3680
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.5623,
+ "step": 3681
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.0002,
+ "loss": 0.6818,
+ "step": 3682
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7176,
+ "step": 3683
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7562,
+ "step": 3684
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9916,
+ "step": 3685
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7662,
+ "step": 3686
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6481,
+ "step": 3687
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6961,
+ "step": 3688
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6693,
+ "step": 3689
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7448,
+ "step": 3690
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.605,
+ "step": 3691
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5274,
+ "step": 3692
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7783,
+ "step": 3693
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9291,
+ "step": 3694
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.8049,
+ "step": 3695
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5145,
+ "step": 3696
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 1.003,
+ "step": 3697
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9122,
+ "step": 3698
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 1.0403,
+ "step": 3699
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6969,
+ "step": 3700
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9441,
+ "step": 3701
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 1.0414,
+ "step": 3702
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7389,
+ "step": 3703
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.8022,
+ "step": 3704
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7118,
+ "step": 3705
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7156,
+ "step": 3706
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7644,
+ "step": 3707
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5756,
+ "step": 3708
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.8302,
+ "step": 3709
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.73,
+ "step": 3710
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9252,
+ "step": 3711
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.8429,
+ "step": 3712
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6095,
+ "step": 3713
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9014,
+ "step": 3714
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.563,
+ "step": 3715
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7037,
+ "step": 3716
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9826,
+ "step": 3717
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6815,
+ "step": 3718
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6904,
+ "step": 3719
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 1.0041,
+ "step": 3720
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5924,
+ "step": 3721
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.6953,
+ "step": 3722
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7888,
+ "step": 3723
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.8817,
+ "step": 3724
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9264,
+ "step": 3725
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7591,
+ "step": 3726
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7993,
+ "step": 3727
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7386,
+ "step": 3728
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7005,
+ "step": 3729
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5969,
+ "step": 3730
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.85,
+ "step": 3731
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5772,
+ "step": 3732
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7647,
+ "step": 3733
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.8237,
+ "step": 3734
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7258,
+ "step": 3735
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9647,
+ "step": 3736
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5279,
+ "step": 3737
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.5559,
+ "step": 3738
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7492,
+ "step": 3739
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9171,
+ "step": 3740
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.9443,
+ "step": 3741
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.7924,
+ "step": 3742
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.0002,
+ "loss": 0.8119,
+ "step": 3743
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.5553,
+ "step": 3744
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 1.0719,
+ "step": 3745
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6299,
+ "step": 3746
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7441,
+ "step": 3747
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.631,
+ "step": 3748
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9058,
+ "step": 3749
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7353,
+ "step": 3750
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7452,
+ "step": 3751
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9035,
+ "step": 3752
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6045,
+ "step": 3753
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6287,
+ "step": 3754
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.648,
+ "step": 3755
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7814,
+ "step": 3756
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 1.024,
+ "step": 3757
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6673,
+ "step": 3758
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 1.0115,
+ "step": 3759
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7928,
+ "step": 3760
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7938,
+ "step": 3761
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9024,
+ "step": 3762
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7677,
+ "step": 3763
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.5106,
+ "step": 3764
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6804,
+ "step": 3765
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7812,
+ "step": 3766
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7305,
+ "step": 3767
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6532,
+ "step": 3768
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9125,
+ "step": 3769
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7275,
+ "step": 3770
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7715,
+ "step": 3771
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 1.0193,
+ "step": 3772
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.5958,
+ "step": 3773
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.8214,
+ "step": 3774
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.5232,
+ "step": 3775
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.5943,
+ "step": 3776
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.97,
+ "step": 3777
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.8649,
+ "step": 3778
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7044,
+ "step": 3779
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6559,
+ "step": 3780
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 1.0746,
+ "step": 3781
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.836,
+ "step": 3782
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6087,
+ "step": 3783
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7576,
+ "step": 3784
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9034,
+ "step": 3785
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6595,
+ "step": 3786
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9853,
+ "step": 3787
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9735,
+ "step": 3788
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.8421,
+ "step": 3789
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6717,
+ "step": 3790
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7765,
+ "step": 3791
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9774,
+ "step": 3792
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.5509,
+ "step": 3793
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9114,
+ "step": 3794
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.4861,
+ "step": 3795
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.6866,
+ "step": 3796
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.4111,
+ "step": 3797
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9114,
+ "step": 3798
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7345,
+ "step": 3799
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9816,
+ "step": 3800
+ },
+ {
+ "epoch": 0.62,
+ "eval_loss": 0.7537699341773987,
+ "eval_runtime": 131.6884,
+ "eval_samples_per_second": 7.594,
+ "eval_steps_per_second": 3.797,
+ "step": 3800
+ },
+ {
+ "epoch": 0.62,
+ "mmlu_eval_accuracy": 0.5747248918142907,
+ "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
+ "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
+ "mmlu_eval_accuracy_astronomy": 0.6875,
+ "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
+ "mmlu_eval_accuracy_college_biology": 0.5625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.4375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
+ "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.6875,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.5454545454545454,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
+ "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": 0.5217391304347826,
+ "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
+ "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
+ "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": 0.5454545454545454,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.88,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
+ "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
+ "mmlu_eval_accuracy_moral_scenarios": 0.32,
+ "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
+ "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
+ "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
+ "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
+ "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
+ "mmlu_eval_accuracy_sociology": 0.7727272727272727,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.4444444444444444,
+ "mmlu_eval_accuracy_world_religions": 0.631578947368421,
+ "mmlu_loss": 1.0576872202730676,
+ "step": 3800
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.5402,
+ "step": 3801
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.9108,
+ "step": 3802
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.0002,
+ "loss": 0.7512,
+ "step": 3803
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7245,
+ "step": 3804
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6794,
+ "step": 3805
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7492,
+ "step": 3806
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9332,
+ "step": 3807
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 1.2136,
+ "step": 3808
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8751,
+ "step": 3809
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8575,
+ "step": 3810
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9675,
+ "step": 3811
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7018,
+ "step": 3812
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7989,
+ "step": 3813
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7021,
+ "step": 3814
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6984,
+ "step": 3815
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 1.1223,
+ "step": 3816
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6015,
+ "step": 3817
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7926,
+ "step": 3818
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8791,
+ "step": 3819
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9313,
+ "step": 3820
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.864,
+ "step": 3821
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.4597,
+ "step": 3822
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8548,
+ "step": 3823
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.742,
+ "step": 3824
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.5929,
+ "step": 3825
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6599,
+ "step": 3826
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9248,
+ "step": 3827
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7489,
+ "step": 3828
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7957,
+ "step": 3829
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8486,
+ "step": 3830
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8626,
+ "step": 3831
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9905,
+ "step": 3832
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7992,
+ "step": 3833
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8073,
+ "step": 3834
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.5318,
+ "step": 3835
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7159,
+ "step": 3836
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8805,
+ "step": 3837
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9001,
+ "step": 3838
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7076,
+ "step": 3839
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6394,
+ "step": 3840
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7299,
+ "step": 3841
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9415,
+ "step": 3842
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7803,
+ "step": 3843
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9901,
+ "step": 3844
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6428,
+ "step": 3845
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7608,
+ "step": 3846
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6747,
+ "step": 3847
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9323,
+ "step": 3848
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6931,
+ "step": 3849
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6605,
+ "step": 3850
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9666,
+ "step": 3851
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 1.0202,
+ "step": 3852
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7591,
+ "step": 3853
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6166,
+ "step": 3854
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 1.08,
+ "step": 3855
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 1.164,
+ "step": 3856
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9347,
+ "step": 3857
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.5875,
+ "step": 3858
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.8427,
+ "step": 3859
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.6039,
+ "step": 3860
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.3595,
+ "step": 3861
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.9754,
+ "step": 3862
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.7815,
+ "step": 3863
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0002,
+ "loss": 0.4558,
+ "step": 3864
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.9595,
+ "step": 3865
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6937,
+ "step": 3866
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.581,
+ "step": 3867
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7417,
+ "step": 3868
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7924,
+ "step": 3869
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.5623,
+ "step": 3870
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.488,
+ "step": 3871
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.864,
+ "step": 3872
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.701,
+ "step": 3873
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7831,
+ "step": 3874
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7006,
+ "step": 3875
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8439,
+ "step": 3876
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6861,
+ "step": 3877
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.752,
+ "step": 3878
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7008,
+ "step": 3879
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 1.1345,
+ "step": 3880
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.9909,
+ "step": 3881
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6671,
+ "step": 3882
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6767,
+ "step": 3883
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8261,
+ "step": 3884
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7584,
+ "step": 3885
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6682,
+ "step": 3886
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8094,
+ "step": 3887
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 1.03,
+ "step": 3888
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.878,
+ "step": 3889
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8439,
+ "step": 3890
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.4543,
+ "step": 3891
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.658,
+ "step": 3892
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6426,
+ "step": 3893
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8137,
+ "step": 3894
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8807,
+ "step": 3895
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.9436,
+ "step": 3896
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.892,
+ "step": 3897
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8417,
+ "step": 3898
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6452,
+ "step": 3899
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.828,
+ "step": 3900
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.9157,
+ "step": 3901
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6588,
+ "step": 3902
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6912,
+ "step": 3903
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7504,
+ "step": 3904
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8242,
+ "step": 3905
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.9903,
+ "step": 3906
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6505,
+ "step": 3907
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7512,
+ "step": 3908
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7766,
+ "step": 3909
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7185,
+ "step": 3910
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6193,
+ "step": 3911
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7262,
+ "step": 3912
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6989,
+ "step": 3913
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8782,
+ "step": 3914
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6939,
+ "step": 3915
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 1.1232,
+ "step": 3916
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.8039,
+ "step": 3917
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6064,
+ "step": 3918
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.7359,
+ "step": 3919
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.796,
+ "step": 3920
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6751,
+ "step": 3921
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.6136,
+ "step": 3922
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.5512,
+ "step": 3923
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.9652,
+ "step": 3924
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0002,
+ "loss": 0.5687,
+ "step": 3925
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.5255,
+ "step": 3926
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.5293,
+ "step": 3927
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.883,
+ "step": 3928
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6263,
+ "step": 3929
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8519,
+ "step": 3930
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6,
+ "step": 3931
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6254,
+ "step": 3932
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8341,
+ "step": 3933
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7435,
+ "step": 3934
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 1.1017,
+ "step": 3935
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 1.0322,
+ "step": 3936
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6318,
+ "step": 3937
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7687,
+ "step": 3938
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8693,
+ "step": 3939
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9442,
+ "step": 3940
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6218,
+ "step": 3941
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7335,
+ "step": 3942
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9228,
+ "step": 3943
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.4507,
+ "step": 3944
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6448,
+ "step": 3945
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9261,
+ "step": 3946
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8647,
+ "step": 3947
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7792,
+ "step": 3948
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6935,
+ "step": 3949
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 1.1047,
+ "step": 3950
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.5875,
+ "step": 3951
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9353,
+ "step": 3952
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7937,
+ "step": 3953
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9495,
+ "step": 3954
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8424,
+ "step": 3955
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6798,
+ "step": 3956
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 1.117,
+ "step": 3957
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.935,
+ "step": 3958
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7301,
+ "step": 3959
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6926,
+ "step": 3960
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9344,
+ "step": 3961
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.4965,
+ "step": 3962
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6938,
+ "step": 3963
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8308,
+ "step": 3964
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7292,
+ "step": 3965
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.701,
+ "step": 3966
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7405,
+ "step": 3967
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8253,
+ "step": 3968
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8218,
+ "step": 3969
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6495,
+ "step": 3970
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7938,
+ "step": 3971
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8966,
+ "step": 3972
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7902,
+ "step": 3973
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6365,
+ "step": 3974
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8687,
+ "step": 3975
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8928,
+ "step": 3976
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.6106,
+ "step": 3977
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.5016,
+ "step": 3978
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.906,
+ "step": 3979
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8617,
+ "step": 3980
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9383,
+ "step": 3981
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.7615,
+ "step": 3982
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 1.191,
+ "step": 3983
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.8554,
+ "step": 3984
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.4396,
+ "step": 3985
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0002,
+ "loss": 0.9333,
+ "step": 3986
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8592,
+ "step": 3987
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6116,
+ "step": 3988
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7387,
+ "step": 3989
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8976,
+ "step": 3990
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6268,
+ "step": 3991
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8299,
+ "step": 3992
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6538,
+ "step": 3993
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8059,
+ "step": 3994
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7929,
+ "step": 3995
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.9452,
+ "step": 3996
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 1.008,
+ "step": 3997
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8184,
+ "step": 3998
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8047,
+ "step": 3999
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.9196,
+ "step": 4000
+ },
+ {
+ "epoch": 0.66,
+ "eval_loss": 0.7537485957145691,
+ "eval_runtime": 131.7422,
+ "eval_samples_per_second": 7.591,
+ "eval_steps_per_second": 3.795,
+ "step": 4000
+ },
+ {
+ "epoch": 0.66,
+ "mmlu_eval_accuracy": 0.5449337961098338,
+ "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.6875,
+ "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
+ "mmlu_eval_accuracy_college_biology": 0.5,
+ "mmlu_eval_accuracy_college_chemistry": 0.125,
+ "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
+ "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5625,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "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": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
+ "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
+ "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
+ "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
+ "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+ "mmlu_eval_accuracy_international_law": 1.0,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.28,
+ "mmlu_eval_accuracy_nutrition": 0.696969696969697,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
+ "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
+ "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
+ "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+ "mmlu_eval_accuracy_sociology": 0.6818181818181818,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
+ "mmlu_loss": 0.9393934273299598,
+ "step": 4000
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8735,
+ "step": 4001
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8306,
+ "step": 4002
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6419,
+ "step": 4003
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.4954,
+ "step": 4004
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.606,
+ "step": 4005
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6402,
+ "step": 4006
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 1.1135,
+ "step": 4007
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.4582,
+ "step": 4008
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.5686,
+ "step": 4009
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 1.1929,
+ "step": 4010
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6899,
+ "step": 4011
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7652,
+ "step": 4012
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.4438,
+ "step": 4013
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6595,
+ "step": 4014
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8755,
+ "step": 4015
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.997,
+ "step": 4016
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.651,
+ "step": 4017
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7122,
+ "step": 4018
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6794,
+ "step": 4019
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8957,
+ "step": 4020
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7544,
+ "step": 4021
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.9787,
+ "step": 4022
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.4722,
+ "step": 4023
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8886,
+ "step": 4024
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6733,
+ "step": 4025
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6,
+ "step": 4026
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8649,
+ "step": 4027
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7694,
+ "step": 4028
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7215,
+ "step": 4029
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.9129,
+ "step": 4030
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7671,
+ "step": 4031
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.804,
+ "step": 4032
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.786,
+ "step": 4033
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 1.0315,
+ "step": 4034
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 1.0094,
+ "step": 4035
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.481,
+ "step": 4036
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.7125,
+ "step": 4037
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.5605,
+ "step": 4038
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6978,
+ "step": 4039
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6818,
+ "step": 4040
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.6538,
+ "step": 4041
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.8544,
+ "step": 4042
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.544,
+ "step": 4043
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.5517,
+ "step": 4044
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.9996,
+ "step": 4045
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.907,
+ "step": 4046
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0002,
+ "loss": 0.892,
+ "step": 4047
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7069,
+ "step": 4048
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.045,
+ "step": 4049
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.0,
+ "step": 4050
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.803,
+ "step": 4051
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8542,
+ "step": 4052
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.951,
+ "step": 4053
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6267,
+ "step": 4054
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.3254,
+ "step": 4055
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8324,
+ "step": 4056
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6778,
+ "step": 4057
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.1308,
+ "step": 4058
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6028,
+ "step": 4059
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7107,
+ "step": 4060
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6946,
+ "step": 4061
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7788,
+ "step": 4062
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.5525,
+ "step": 4063
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7175,
+ "step": 4064
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7034,
+ "step": 4065
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.9934,
+ "step": 4066
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.0798,
+ "step": 4067
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6854,
+ "step": 4068
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.3544,
+ "step": 4069
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.35,
+ "step": 4070
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6968,
+ "step": 4071
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.5759,
+ "step": 4072
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.9113,
+ "step": 4073
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.0737,
+ "step": 4074
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8199,
+ "step": 4075
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6423,
+ "step": 4076
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8007,
+ "step": 4077
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.5192,
+ "step": 4078
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.5552,
+ "step": 4079
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7254,
+ "step": 4080
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.706,
+ "step": 4081
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7497,
+ "step": 4082
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7493,
+ "step": 4083
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7074,
+ "step": 4084
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.0309,
+ "step": 4085
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8056,
+ "step": 4086
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8165,
+ "step": 4087
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.872,
+ "step": 4088
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6163,
+ "step": 4089
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8407,
+ "step": 4090
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8058,
+ "step": 4091
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.7988,
+ "step": 4092
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6226,
+ "step": 4093
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 1.0535,
+ "step": 4094
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8436,
+ "step": 4095
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.5542,
+ "step": 4096
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8667,
+ "step": 4097
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.9178,
+ "step": 4098
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8791,
+ "step": 4099
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.5041,
+ "step": 4100
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.6645,
+ "step": 4101
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.968,
+ "step": 4102
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.4261,
+ "step": 4103
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8508,
+ "step": 4104
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8927,
+ "step": 4105
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.4507,
+ "step": 4106
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.982,
+ "step": 4107
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0002,
+ "loss": 0.8964,
+ "step": 4108
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9159,
+ "step": 4109
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7705,
+ "step": 4110
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7375,
+ "step": 4111
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9107,
+ "step": 4112
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.699,
+ "step": 4113
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8554,
+ "step": 4114
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9986,
+ "step": 4115
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6024,
+ "step": 4116
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.5505,
+ "step": 4117
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8374,
+ "step": 4118
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.3902,
+ "step": 4119
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6852,
+ "step": 4120
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8564,
+ "step": 4121
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6535,
+ "step": 4122
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8605,
+ "step": 4123
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8209,
+ "step": 4124
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6501,
+ "step": 4125
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.5352,
+ "step": 4126
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8045,
+ "step": 4127
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6482,
+ "step": 4128
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.5674,
+ "step": 4129
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6092,
+ "step": 4130
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9681,
+ "step": 4131
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8404,
+ "step": 4132
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9622,
+ "step": 4133
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.698,
+ "step": 4134
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6511,
+ "step": 4135
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9056,
+ "step": 4136
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8652,
+ "step": 4137
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.5614,
+ "step": 4138
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.834,
+ "step": 4139
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7341,
+ "step": 4140
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7551,
+ "step": 4141
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8801,
+ "step": 4142
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7538,
+ "step": 4143
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6405,
+ "step": 4144
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6983,
+ "step": 4145
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 1.2172,
+ "step": 4146
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6614,
+ "step": 4147
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8717,
+ "step": 4148
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 1.003,
+ "step": 4149
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9082,
+ "step": 4150
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9477,
+ "step": 4151
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.9291,
+ "step": 4152
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 1.0087,
+ "step": 4153
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 1.063,
+ "step": 4154
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.849,
+ "step": 4155
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 1.0245,
+ "step": 4156
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.4341,
+ "step": 4157
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7575,
+ "step": 4158
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8221,
+ "step": 4159
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.6938,
+ "step": 4160
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.4986,
+ "step": 4161
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 1.0502,
+ "step": 4162
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.921,
+ "step": 4163
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7972,
+ "step": 4164
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7228,
+ "step": 4165
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7494,
+ "step": 4166
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.7726,
+ "step": 4167
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.8441,
+ "step": 4168
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0002,
+ "loss": 0.4793,
+ "step": 4169
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8861,
+ "step": 4170
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8571,
+ "step": 4171
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 1.1305,
+ "step": 4172
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.5088,
+ "step": 4173
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.5141,
+ "step": 4174
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.918,
+ "step": 4175
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.9921,
+ "step": 4176
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7473,
+ "step": 4177
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7213,
+ "step": 4178
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7158,
+ "step": 4179
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8161,
+ "step": 4180
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8173,
+ "step": 4181
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.768,
+ "step": 4182
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 1.0386,
+ "step": 4183
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7681,
+ "step": 4184
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7326,
+ "step": 4185
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.9131,
+ "step": 4186
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.6131,
+ "step": 4187
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 1.1967,
+ "step": 4188
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7489,
+ "step": 4189
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7104,
+ "step": 4190
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8456,
+ "step": 4191
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.878,
+ "step": 4192
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8159,
+ "step": 4193
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 1.0059,
+ "step": 4194
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 1.019,
+ "step": 4195
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.6476,
+ "step": 4196
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.9292,
+ "step": 4197
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.6789,
+ "step": 4198
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7217,
+ "step": 4199
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.5865,
+ "step": 4200
+ },
+ {
+ "epoch": 0.69,
+ "eval_loss": 0.7533943057060242,
+ "eval_runtime": 131.6714,
+ "eval_samples_per_second": 7.595,
+ "eval_steps_per_second": 3.797,
+ "step": 4200
+ },
+ {
+ "epoch": 0.69,
+ "mmlu_eval_accuracy": 0.5644995419094495,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.75,
+ "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
+ "mmlu_eval_accuracy_college_biology": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.25,
+ "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.5,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
+ "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5625,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
+ "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
+ "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+ "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
+ "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636,
+ "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
+ "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
+ "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.8,
+ "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+ "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.34,
+ "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.6,
+ "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
+ "mmlu_eval_accuracy_professional_law": 0.32941176470588235,
+ "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
+ "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
+ "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
+ "mmlu_eval_accuracy_sociology": 0.7727272727272727,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
+ "mmlu_loss": 1.0887044936262908,
+ "step": 4200
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7088,
+ "step": 4201
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.826,
+ "step": 4202
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.9575,
+ "step": 4203
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.5376,
+ "step": 4204
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.6873,
+ "step": 4205
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8586,
+ "step": 4206
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 1.183,
+ "step": 4207
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.9128,
+ "step": 4208
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.879,
+ "step": 4209
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 1.0587,
+ "step": 4210
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.6885,
+ "step": 4211
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8904,
+ "step": 4212
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8077,
+ "step": 4213
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7364,
+ "step": 4214
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8644,
+ "step": 4215
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7677,
+ "step": 4216
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.9061,
+ "step": 4217
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.5827,
+ "step": 4218
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7322,
+ "step": 4219
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.9292,
+ "step": 4220
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.6685,
+ "step": 4221
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.765,
+ "step": 4222
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7416,
+ "step": 4223
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.8206,
+ "step": 4224
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.947,
+ "step": 4225
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7442,
+ "step": 4226
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.6038,
+ "step": 4227
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.7065,
+ "step": 4228
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.0002,
+ "loss": 0.718,
+ "step": 4229
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.9411,
+ "step": 4230
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6908,
+ "step": 4231
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 1.0305,
+ "step": 4232
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6893,
+ "step": 4233
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.728,
+ "step": 4234
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.9497,
+ "step": 4235
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7119,
+ "step": 4236
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.5091,
+ "step": 4237
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7729,
+ "step": 4238
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7998,
+ "step": 4239
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.9126,
+ "step": 4240
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8319,
+ "step": 4241
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.788,
+ "step": 4242
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7115,
+ "step": 4243
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8995,
+ "step": 4244
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6007,
+ "step": 4245
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6406,
+ "step": 4246
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7189,
+ "step": 4247
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6078,
+ "step": 4248
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7395,
+ "step": 4249
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7255,
+ "step": 4250
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6547,
+ "step": 4251
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8872,
+ "step": 4252
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7467,
+ "step": 4253
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6842,
+ "step": 4254
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7927,
+ "step": 4255
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.858,
+ "step": 4256
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8082,
+ "step": 4257
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7945,
+ "step": 4258
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.5583,
+ "step": 4259
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7802,
+ "step": 4260
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7948,
+ "step": 4261
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7878,
+ "step": 4262
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.5858,
+ "step": 4263
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8051,
+ "step": 4264
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8607,
+ "step": 4265
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8733,
+ "step": 4266
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8997,
+ "step": 4267
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.9369,
+ "step": 4268
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8915,
+ "step": 4269
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7083,
+ "step": 4270
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 1.1058,
+ "step": 4271
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6546,
+ "step": 4272
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7363,
+ "step": 4273
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8596,
+ "step": 4274
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.9282,
+ "step": 4275
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6584,
+ "step": 4276
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.699,
+ "step": 4277
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7562,
+ "step": 4278
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8363,
+ "step": 4279
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7332,
+ "step": 4280
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7752,
+ "step": 4281
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7878,
+ "step": 4282
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.9379,
+ "step": 4283
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.5127,
+ "step": 4284
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6401,
+ "step": 4285
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.7128,
+ "step": 4286
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.8791,
+ "step": 4287
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.9435,
+ "step": 4288
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.6818,
+ "step": 4289
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.0002,
+ "loss": 0.731,
+ "step": 4290
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.6199,
+ "step": 4291
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8362,
+ "step": 4292
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7557,
+ "step": 4293
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8104,
+ "step": 4294
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.6304,
+ "step": 4295
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.6191,
+ "step": 4296
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7229,
+ "step": 4297
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8468,
+ "step": 4298
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9561,
+ "step": 4299
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7094,
+ "step": 4300
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7046,
+ "step": 4301
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7452,
+ "step": 4302
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7965,
+ "step": 4303
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9377,
+ "step": 4304
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8339,
+ "step": 4305
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.5201,
+ "step": 4306
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.693,
+ "step": 4307
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.4841,
+ "step": 4308
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8746,
+ "step": 4309
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7905,
+ "step": 4310
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8549,
+ "step": 4311
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7169,
+ "step": 4312
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7492,
+ "step": 4313
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8333,
+ "step": 4314
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.5296,
+ "step": 4315
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.5097,
+ "step": 4316
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.5333,
+ "step": 4317
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8874,
+ "step": 4318
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7506,
+ "step": 4319
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.899,
+ "step": 4320
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8063,
+ "step": 4321
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7164,
+ "step": 4322
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8668,
+ "step": 4323
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9056,
+ "step": 4324
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8459,
+ "step": 4325
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.6318,
+ "step": 4326
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9774,
+ "step": 4327
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7431,
+ "step": 4328
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7021,
+ "step": 4329
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8374,
+ "step": 4330
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9133,
+ "step": 4331
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9535,
+ "step": 4332
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9591,
+ "step": 4333
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.5714,
+ "step": 4334
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8052,
+ "step": 4335
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 1.0769,
+ "step": 4336
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8726,
+ "step": 4337
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8605,
+ "step": 4338
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.862,
+ "step": 4339
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8431,
+ "step": 4340
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7999,
+ "step": 4341
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8344,
+ "step": 4342
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8283,
+ "step": 4343
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.9465,
+ "step": 4344
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.6453,
+ "step": 4345
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.8787,
+ "step": 4346
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.6611,
+ "step": 4347
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7626,
+ "step": 4348
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.901,
+ "step": 4349
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7381,
+ "step": 4350
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.0002,
+ "loss": 0.7061,
+ "step": 4351
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 1.0379,
+ "step": 4352
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6639,
+ "step": 4353
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.865,
+ "step": 4354
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8033,
+ "step": 4355
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7056,
+ "step": 4356
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8245,
+ "step": 4357
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6756,
+ "step": 4358
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6938,
+ "step": 4359
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8485,
+ "step": 4360
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6693,
+ "step": 4361
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8517,
+ "step": 4362
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7159,
+ "step": 4363
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8121,
+ "step": 4364
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6454,
+ "step": 4365
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.5119,
+ "step": 4366
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7997,
+ "step": 4367
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7839,
+ "step": 4368
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.9646,
+ "step": 4369
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7841,
+ "step": 4370
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 1.0305,
+ "step": 4371
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6358,
+ "step": 4372
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8418,
+ "step": 4373
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.781,
+ "step": 4374
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.9208,
+ "step": 4375
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 1.0694,
+ "step": 4376
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6828,
+ "step": 4377
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7383,
+ "step": 4378
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8482,
+ "step": 4379
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8074,
+ "step": 4380
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.74,
+ "step": 4381
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.5378,
+ "step": 4382
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6981,
+ "step": 4383
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7824,
+ "step": 4384
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6468,
+ "step": 4385
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 1.186,
+ "step": 4386
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7951,
+ "step": 4387
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.9237,
+ "step": 4388
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.9582,
+ "step": 4389
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7297,
+ "step": 4390
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.3732,
+ "step": 4391
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6837,
+ "step": 4392
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8324,
+ "step": 4393
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8395,
+ "step": 4394
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.632,
+ "step": 4395
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8551,
+ "step": 4396
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8718,
+ "step": 4397
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7141,
+ "step": 4398
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6758,
+ "step": 4399
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7194,
+ "step": 4400
+ },
+ {
+ "epoch": 0.72,
+ "eval_loss": 0.7532736659049988,
+ "eval_runtime": 131.6439,
+ "eval_samples_per_second": 7.596,
+ "eval_steps_per_second": 3.798,
+ "step": 4400
+ },
+ {
+ "epoch": 0.72,
+ "mmlu_eval_accuracy": 0.5489116992002624,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.75,
+ "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
+ "mmlu_eval_accuracy_college_biology": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
+ "mmlu_eval_accuracy_econometrics": 0.5,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5625,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.5625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
+ "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8,
+ "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": 0.6956521739130435,
+ "mmlu_eval_accuracy_human_sexuality": 0.5,
+ "mmlu_eval_accuracy_international_law": 0.9230769230769231,
+ "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.29,
+ "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
+ "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
+ "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
+ "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
+ "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
+ "mmlu_eval_accuracy_sociology": 0.6363636363636364,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
+ "mmlu_eval_accuracy_virology": 0.4444444444444444,
+ "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
+ "mmlu_loss": 1.1256680759764215,
+ "step": 4400
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7289,
+ "step": 4401
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7805,
+ "step": 4402
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.9821,
+ "step": 4403
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.9217,
+ "step": 4404
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.4598,
+ "step": 4405
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.8441,
+ "step": 4406
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7677,
+ "step": 4407
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 1.0347,
+ "step": 4408
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7505,
+ "step": 4409
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6557,
+ "step": 4410
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.6402,
+ "step": 4411
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.0002,
+ "loss": 0.7251,
+ "step": 4412
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7225,
+ "step": 4413
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6499,
+ "step": 4414
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8984,
+ "step": 4415
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6837,
+ "step": 4416
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7705,
+ "step": 4417
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8128,
+ "step": 4418
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.5027,
+ "step": 4419
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.9676,
+ "step": 4420
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.9938,
+ "step": 4421
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 1.0139,
+ "step": 4422
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8246,
+ "step": 4423
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.4546,
+ "step": 4424
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.767,
+ "step": 4425
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.9061,
+ "step": 4426
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6387,
+ "step": 4427
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.5553,
+ "step": 4428
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8795,
+ "step": 4429
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6621,
+ "step": 4430
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7296,
+ "step": 4431
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8448,
+ "step": 4432
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.918,
+ "step": 4433
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8295,
+ "step": 4434
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.794,
+ "step": 4435
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7037,
+ "step": 4436
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6892,
+ "step": 4437
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.5753,
+ "step": 4438
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7756,
+ "step": 4439
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8451,
+ "step": 4440
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.5788,
+ "step": 4441
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.9984,
+ "step": 4442
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7915,
+ "step": 4443
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8622,
+ "step": 4444
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8023,
+ "step": 4445
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7619,
+ "step": 4446
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.5191,
+ "step": 4447
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7042,
+ "step": 4448
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7875,
+ "step": 4449
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8848,
+ "step": 4450
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6226,
+ "step": 4451
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7548,
+ "step": 4452
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.821,
+ "step": 4453
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.642,
+ "step": 4454
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7164,
+ "step": 4455
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.61,
+ "step": 4456
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7795,
+ "step": 4457
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.9688,
+ "step": 4458
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8398,
+ "step": 4459
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.9748,
+ "step": 4460
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6699,
+ "step": 4461
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7378,
+ "step": 4462
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.8977,
+ "step": 4463
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7809,
+ "step": 4464
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6606,
+ "step": 4465
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6417,
+ "step": 4466
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 1.0545,
+ "step": 4467
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.9741,
+ "step": 4468
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7331,
+ "step": 4469
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7184,
+ "step": 4470
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.811,
+ "step": 4471
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.7116,
+ "step": 4472
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0002,
+ "loss": 0.6309,
+ "step": 4473
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.5395,
+ "step": 4474
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.5559,
+ "step": 4475
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8749,
+ "step": 4476
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8352,
+ "step": 4477
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.724,
+ "step": 4478
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.692,
+ "step": 4479
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.9015,
+ "step": 4480
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.5229,
+ "step": 4481
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7055,
+ "step": 4482
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7741,
+ "step": 4483
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8178,
+ "step": 4484
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.788,
+ "step": 4485
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7283,
+ "step": 4486
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7197,
+ "step": 4487
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.9261,
+ "step": 4488
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7099,
+ "step": 4489
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8473,
+ "step": 4490
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.639,
+ "step": 4491
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.636,
+ "step": 4492
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.9413,
+ "step": 4493
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.9017,
+ "step": 4494
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8861,
+ "step": 4495
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8836,
+ "step": 4496
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7972,
+ "step": 4497
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8568,
+ "step": 4498
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7918,
+ "step": 4499
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.631,
+ "step": 4500
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7749,
+ "step": 4501
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 1.0638,
+ "step": 4502
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8119,
+ "step": 4503
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8776,
+ "step": 4504
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7677,
+ "step": 4505
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7324,
+ "step": 4506
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7398,
+ "step": 4507
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.732,
+ "step": 4508
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7114,
+ "step": 4509
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.772,
+ "step": 4510
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.886,
+ "step": 4511
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.6252,
+ "step": 4512
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 1.0659,
+ "step": 4513
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7199,
+ "step": 4514
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.6815,
+ "step": 4515
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.9449,
+ "step": 4516
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7396,
+ "step": 4517
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 1.0366,
+ "step": 4518
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.6383,
+ "step": 4519
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.765,
+ "step": 4520
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.916,
+ "step": 4521
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.651,
+ "step": 4522
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.4668,
+ "step": 4523
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7961,
+ "step": 4524
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.6233,
+ "step": 4525
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.5247,
+ "step": 4526
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.9137,
+ "step": 4527
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7782,
+ "step": 4528
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 1.1018,
+ "step": 4529
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.5824,
+ "step": 4530
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.7501,
+ "step": 4531
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.8408,
+ "step": 4532
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.9354,
+ "step": 4533
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0002,
+ "loss": 0.6268,
+ "step": 4534
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.864,
+ "step": 4535
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.576,
+ "step": 4536
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7936,
+ "step": 4537
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 1.0962,
+ "step": 4538
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7042,
+ "step": 4539
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7947,
+ "step": 4540
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6836,
+ "step": 4541
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6667,
+ "step": 4542
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9211,
+ "step": 4543
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7376,
+ "step": 4544
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6675,
+ "step": 4545
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9592,
+ "step": 4546
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.906,
+ "step": 4547
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6151,
+ "step": 4548
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 1.1113,
+ "step": 4549
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.723,
+ "step": 4550
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8942,
+ "step": 4551
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8589,
+ "step": 4552
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6644,
+ "step": 4553
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6037,
+ "step": 4554
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6093,
+ "step": 4555
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.758,
+ "step": 4556
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9568,
+ "step": 4557
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9693,
+ "step": 4558
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 1.0103,
+ "step": 4559
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8603,
+ "step": 4560
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7311,
+ "step": 4561
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8597,
+ "step": 4562
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.684,
+ "step": 4563
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.825,
+ "step": 4564
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.5722,
+ "step": 4565
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6409,
+ "step": 4566
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8368,
+ "step": 4567
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7797,
+ "step": 4568
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.4905,
+ "step": 4569
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8342,
+ "step": 4570
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9027,
+ "step": 4571
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.754,
+ "step": 4572
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8468,
+ "step": 4573
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7125,
+ "step": 4574
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6785,
+ "step": 4575
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7762,
+ "step": 4576
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6681,
+ "step": 4577
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9039,
+ "step": 4578
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.5755,
+ "step": 4579
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9517,
+ "step": 4580
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9827,
+ "step": 4581
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 1.1242,
+ "step": 4582
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7676,
+ "step": 4583
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6868,
+ "step": 4584
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8196,
+ "step": 4585
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8112,
+ "step": 4586
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.5174,
+ "step": 4587
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.9508,
+ "step": 4588
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6084,
+ "step": 4589
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8421,
+ "step": 4590
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6949,
+ "step": 4591
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.7422,
+ "step": 4592
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.8267,
+ "step": 4593
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6379,
+ "step": 4594
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.0002,
+ "loss": 0.6868,
+ "step": 4595
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.5569,
+ "step": 4596
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7502,
+ "step": 4597
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7659,
+ "step": 4598
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8215,
+ "step": 4599
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8463,
+ "step": 4600
+ },
+ {
+ "epoch": 0.76,
+ "eval_loss": 0.7545199990272522,
+ "eval_runtime": 131.6197,
+ "eval_samples_per_second": 7.598,
+ "eval_steps_per_second": 3.799,
+ "step": 4600
+ },
+ {
+ "epoch": 0.76,
+ "mmlu_eval_accuracy": 0.5579345039363026,
+ "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": 0.5625,
+ "mmlu_eval_accuracy_college_chemistry": 0.375,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
+ "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
+ "mmlu_eval_accuracy_electrical_engineering": 0.375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.59375,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
+ "mmlu_eval_accuracy_high_school_physics": 0.47058823529411764,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
+ "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
+ "mmlu_eval_accuracy_international_law": 1.0,
+ "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": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
+ "mmlu_eval_accuracy_moral_disputes": 0.5,
+ "mmlu_eval_accuracy_moral_scenarios": 0.29,
+ "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
+ "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
+ "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
+ "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
+ "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
+ "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
+ "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+ "mmlu_eval_accuracy_sociology": 0.7272727272727273,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
+ "mmlu_loss": 1.1306793517372316,
+ "step": 4600
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7911,
+ "step": 4601
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6099,
+ "step": 4602
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8163,
+ "step": 4603
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7926,
+ "step": 4604
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.5599,
+ "step": 4605
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8866,
+ "step": 4606
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7925,
+ "step": 4607
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.4836,
+ "step": 4608
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.629,
+ "step": 4609
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8442,
+ "step": 4610
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8111,
+ "step": 4611
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.723,
+ "step": 4612
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7955,
+ "step": 4613
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8198,
+ "step": 4614
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6983,
+ "step": 4615
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.5537,
+ "step": 4616
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7414,
+ "step": 4617
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.9449,
+ "step": 4618
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.959,
+ "step": 4619
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8925,
+ "step": 4620
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.5446,
+ "step": 4621
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6682,
+ "step": 4622
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.9294,
+ "step": 4623
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 1.0393,
+ "step": 4624
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8364,
+ "step": 4625
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8019,
+ "step": 4626
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8926,
+ "step": 4627
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7648,
+ "step": 4628
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.9909,
+ "step": 4629
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.4993,
+ "step": 4630
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8648,
+ "step": 4631
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7263,
+ "step": 4632
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7969,
+ "step": 4633
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6861,
+ "step": 4634
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7987,
+ "step": 4635
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8769,
+ "step": 4636
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6689,
+ "step": 4637
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6925,
+ "step": 4638
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.5434,
+ "step": 4639
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.9533,
+ "step": 4640
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.9459,
+ "step": 4641
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6551,
+ "step": 4642
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8431,
+ "step": 4643
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8653,
+ "step": 4644
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8077,
+ "step": 4645
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.8904,
+ "step": 4646
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.6417,
+ "step": 4647
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.9134,
+ "step": 4648
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7827,
+ "step": 4649
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.901,
+ "step": 4650
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.4452,
+ "step": 4651
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 1.049,
+ "step": 4652
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 1.0046,
+ "step": 4653
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7731,
+ "step": 4654
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.0002,
+ "loss": 0.7016,
+ "step": 4655
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7885,
+ "step": 4656
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8188,
+ "step": 4657
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7878,
+ "step": 4658
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5348,
+ "step": 4659
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5377,
+ "step": 4660
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9717,
+ "step": 4661
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9968,
+ "step": 4662
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6693,
+ "step": 4663
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9584,
+ "step": 4664
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8928,
+ "step": 4665
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8323,
+ "step": 4666
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5385,
+ "step": 4667
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8671,
+ "step": 4668
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 1.2198,
+ "step": 4669
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8682,
+ "step": 4670
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.693,
+ "step": 4671
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8047,
+ "step": 4672
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5764,
+ "step": 4673
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6759,
+ "step": 4674
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7589,
+ "step": 4675
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6284,
+ "step": 4676
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8198,
+ "step": 4677
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9619,
+ "step": 4678
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9287,
+ "step": 4679
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5996,
+ "step": 4680
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8517,
+ "step": 4681
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8839,
+ "step": 4682
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.574,
+ "step": 4683
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6976,
+ "step": 4684
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9371,
+ "step": 4685
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6739,
+ "step": 4686
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9652,
+ "step": 4687
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.736,
+ "step": 4688
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5818,
+ "step": 4689
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5674,
+ "step": 4690
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.806,
+ "step": 4691
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7965,
+ "step": 4692
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6628,
+ "step": 4693
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9155,
+ "step": 4694
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8472,
+ "step": 4695
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7586,
+ "step": 4696
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5,
+ "step": 4697
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6968,
+ "step": 4698
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.5077,
+ "step": 4699
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.4783,
+ "step": 4700
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8218,
+ "step": 4701
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7666,
+ "step": 4702
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 1.222,
+ "step": 4703
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 1.1747,
+ "step": 4704
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9013,
+ "step": 4705
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8178,
+ "step": 4706
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7968,
+ "step": 4707
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6571,
+ "step": 4708
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7033,
+ "step": 4709
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.9057,
+ "step": 4710
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7577,
+ "step": 4711
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.6935,
+ "step": 4712
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 1.1719,
+ "step": 4713
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7239,
+ "step": 4714
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.8716,
+ "step": 4715
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.0002,
+ "loss": 0.7261,
+ "step": 4716
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6858,
+ "step": 4717
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8988,
+ "step": 4718
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7271,
+ "step": 4719
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.5656,
+ "step": 4720
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.5795,
+ "step": 4721
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8601,
+ "step": 4722
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9158,
+ "step": 4723
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.717,
+ "step": 4724
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.908,
+ "step": 4725
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8905,
+ "step": 4726
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.831,
+ "step": 4727
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8061,
+ "step": 4728
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8823,
+ "step": 4729
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6111,
+ "step": 4730
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8214,
+ "step": 4731
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.855,
+ "step": 4732
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.489,
+ "step": 4733
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7378,
+ "step": 4734
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7705,
+ "step": 4735
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.649,
+ "step": 4736
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6189,
+ "step": 4737
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7611,
+ "step": 4738
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7817,
+ "step": 4739
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6566,
+ "step": 4740
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9041,
+ "step": 4741
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 1.4254,
+ "step": 4742
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 1.0887,
+ "step": 4743
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.4048,
+ "step": 4744
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7499,
+ "step": 4745
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9056,
+ "step": 4746
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8189,
+ "step": 4747
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9799,
+ "step": 4748
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7086,
+ "step": 4749
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.5226,
+ "step": 4750
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7229,
+ "step": 4751
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6861,
+ "step": 4752
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 1.123,
+ "step": 4753
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9267,
+ "step": 4754
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7782,
+ "step": 4755
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 1.0838,
+ "step": 4756
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9511,
+ "step": 4757
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6474,
+ "step": 4758
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8848,
+ "step": 4759
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.5661,
+ "step": 4760
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.634,
+ "step": 4761
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9702,
+ "step": 4762
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7348,
+ "step": 4763
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7856,
+ "step": 4764
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 1.0825,
+ "step": 4765
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6216,
+ "step": 4766
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.5991,
+ "step": 4767
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7493,
+ "step": 4768
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.7389,
+ "step": 4769
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.5324,
+ "step": 4770
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.908,
+ "step": 4771
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 1.0117,
+ "step": 4772
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8214,
+ "step": 4773
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.6918,
+ "step": 4774
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.8191,
+ "step": 4775
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.9084,
+ "step": 4776
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.0002,
+ "loss": 0.711,
+ "step": 4777
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7869,
+ "step": 4778
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.5487,
+ "step": 4779
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.5357,
+ "step": 4780
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6629,
+ "step": 4781
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6267,
+ "step": 4782
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7728,
+ "step": 4783
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.8346,
+ "step": 4784
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.696,
+ "step": 4785
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7651,
+ "step": 4786
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.9569,
+ "step": 4787
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6884,
+ "step": 4788
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6987,
+ "step": 4789
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.758,
+ "step": 4790
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.8554,
+ "step": 4791
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.8516,
+ "step": 4792
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6886,
+ "step": 4793
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.5816,
+ "step": 4794
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.828,
+ "step": 4795
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.822,
+ "step": 4796
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.5578,
+ "step": 4797
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7181,
+ "step": 4798
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.9475,
+ "step": 4799
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.8014,
+ "step": 4800
+ },
+ {
+ "epoch": 0.79,
+ "eval_loss": 0.7526820302009583,
+ "eval_runtime": 131.6004,
+ "eval_samples_per_second": 7.599,
+ "eval_steps_per_second": 3.799,
+ "step": 4800
+ },
+ {
+ "epoch": 0.79,
+ "mmlu_eval_accuracy": 0.5392650576015245,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.625,
+ "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
+ "mmlu_eval_accuracy_college_biology": 0.5,
+ "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": 0.5,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
+ "mmlu_eval_accuracy_econometrics": 0.25,
+ "mmlu_eval_accuracy_electrical_engineering": 0.625,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
+ "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.5625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
+ "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": 0.8166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
+ "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
+ "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+ "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+ "mmlu_eval_accuracy_international_law": 0.6923076923076923,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
+ "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
+ "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
+ "mmlu_eval_accuracy_moral_scenarios": 0.26,
+ "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
+ "mmlu_eval_accuracy_professional_law": 0.3235294117647059,
+ "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+ "mmlu_eval_accuracy_sociology": 0.6363636363636364,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
+ "mmlu_loss": 1.2467338922129598,
+ "step": 4800
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.4966,
+ "step": 4801
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6529,
+ "step": 4802
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.8714,
+ "step": 4803
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.5789,
+ "step": 4804
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.903,
+ "step": 4805
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.786,
+ "step": 4806
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6975,
+ "step": 4807
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.9127,
+ "step": 4808
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7049,
+ "step": 4809
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 1.0193,
+ "step": 4810
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7216,
+ "step": 4811
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.428,
+ "step": 4812
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.8916,
+ "step": 4813
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.5671,
+ "step": 4814
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7441,
+ "step": 4815
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.595,
+ "step": 4816
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7915,
+ "step": 4817
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7546,
+ "step": 4818
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6821,
+ "step": 4819
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.9311,
+ "step": 4820
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.3551,
+ "step": 4821
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7464,
+ "step": 4822
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.9629,
+ "step": 4823
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 1.0849,
+ "step": 4824
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.4646,
+ "step": 4825
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7067,
+ "step": 4826
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6032,
+ "step": 4827
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7114,
+ "step": 4828
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.725,
+ "step": 4829
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.9246,
+ "step": 4830
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6965,
+ "step": 4831
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6399,
+ "step": 4832
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6809,
+ "step": 4833
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.8358,
+ "step": 4834
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7491,
+ "step": 4835
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.6326,
+ "step": 4836
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 1.1249,
+ "step": 4837
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.0002,
+ "loss": 0.7432,
+ "step": 4838
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 1.0147,
+ "step": 4839
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7459,
+ "step": 4840
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7453,
+ "step": 4841
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6676,
+ "step": 4842
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9504,
+ "step": 4843
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8959,
+ "step": 4844
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.461,
+ "step": 4845
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6699,
+ "step": 4846
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.4154,
+ "step": 4847
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9485,
+ "step": 4848
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7389,
+ "step": 4849
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8109,
+ "step": 4850
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8148,
+ "step": 4851
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.4654,
+ "step": 4852
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8444,
+ "step": 4853
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.4985,
+ "step": 4854
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6692,
+ "step": 4855
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7235,
+ "step": 4856
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7402,
+ "step": 4857
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7794,
+ "step": 4858
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9156,
+ "step": 4859
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8724,
+ "step": 4860
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6646,
+ "step": 4861
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6552,
+ "step": 4862
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8261,
+ "step": 4863
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7549,
+ "step": 4864
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.4587,
+ "step": 4865
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9461,
+ "step": 4866
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9087,
+ "step": 4867
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7156,
+ "step": 4868
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8056,
+ "step": 4869
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9365,
+ "step": 4870
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6373,
+ "step": 4871
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9882,
+ "step": 4872
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7162,
+ "step": 4873
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7489,
+ "step": 4874
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 1.1003,
+ "step": 4875
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6933,
+ "step": 4876
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7414,
+ "step": 4877
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.4158,
+ "step": 4878
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.648,
+ "step": 4879
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7928,
+ "step": 4880
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8692,
+ "step": 4881
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7973,
+ "step": 4882
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.4256,
+ "step": 4883
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9049,
+ "step": 4884
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.907,
+ "step": 4885
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6737,
+ "step": 4886
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 1.1773,
+ "step": 4887
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7772,
+ "step": 4888
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7401,
+ "step": 4889
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7418,
+ "step": 4890
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8023,
+ "step": 4891
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6014,
+ "step": 4892
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.9101,
+ "step": 4893
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7898,
+ "step": 4894
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6695,
+ "step": 4895
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.7071,
+ "step": 4896
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8681,
+ "step": 4897
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.6333,
+ "step": 4898
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0002,
+ "loss": 0.8757,
+ "step": 4899
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7741,
+ "step": 4900
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 1.0184,
+ "step": 4901
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9164,
+ "step": 4902
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8519,
+ "step": 4903
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6396,
+ "step": 4904
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 1.2245,
+ "step": 4905
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8713,
+ "step": 4906
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.996,
+ "step": 4907
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8079,
+ "step": 4908
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9805,
+ "step": 4909
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.5796,
+ "step": 4910
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9672,
+ "step": 4911
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6402,
+ "step": 4912
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7811,
+ "step": 4913
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 1.0604,
+ "step": 4914
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7772,
+ "step": 4915
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9942,
+ "step": 4916
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.948,
+ "step": 4917
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7708,
+ "step": 4918
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.644,
+ "step": 4919
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8183,
+ "step": 4920
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9998,
+ "step": 4921
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6516,
+ "step": 4922
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8257,
+ "step": 4923
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8655,
+ "step": 4924
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8193,
+ "step": 4925
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6359,
+ "step": 4926
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6756,
+ "step": 4927
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.865,
+ "step": 4928
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6023,
+ "step": 4929
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8107,
+ "step": 4930
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6788,
+ "step": 4931
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.5452,
+ "step": 4932
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8593,
+ "step": 4933
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7688,
+ "step": 4934
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.574,
+ "step": 4935
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7644,
+ "step": 4936
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9559,
+ "step": 4937
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.5227,
+ "step": 4938
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8031,
+ "step": 4939
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7025,
+ "step": 4940
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8732,
+ "step": 4941
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8613,
+ "step": 4942
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7975,
+ "step": 4943
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 1.0406,
+ "step": 4944
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.5253,
+ "step": 4945
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7925,
+ "step": 4946
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.8209,
+ "step": 4947
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6693,
+ "step": 4948
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.892,
+ "step": 4949
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.5657,
+ "step": 4950
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6676,
+ "step": 4951
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.6019,
+ "step": 4952
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.4724,
+ "step": 4953
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9311,
+ "step": 4954
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7948,
+ "step": 4955
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.7516,
+ "step": 4956
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.4601,
+ "step": 4957
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.9001,
+ "step": 4958
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.5278,
+ "step": 4959
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.0002,
+ "loss": 0.55,
+ "step": 4960
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8358,
+ "step": 4961
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.904,
+ "step": 4962
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8669,
+ "step": 4963
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6948,
+ "step": 4964
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.9742,
+ "step": 4965
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8455,
+ "step": 4966
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.9986,
+ "step": 4967
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 1.1176,
+ "step": 4968
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7747,
+ "step": 4969
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7824,
+ "step": 4970
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 1.1169,
+ "step": 4971
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.44,
+ "step": 4972
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.4928,
+ "step": 4973
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6143,
+ "step": 4974
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8974,
+ "step": 4975
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8774,
+ "step": 4976
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8368,
+ "step": 4977
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.447,
+ "step": 4978
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.795,
+ "step": 4979
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7965,
+ "step": 4980
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7807,
+ "step": 4981
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8919,
+ "step": 4982
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.9949,
+ "step": 4983
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8772,
+ "step": 4984
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6358,
+ "step": 4985
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7467,
+ "step": 4986
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7547,
+ "step": 4987
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7838,
+ "step": 4988
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6139,
+ "step": 4989
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8463,
+ "step": 4990
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6306,
+ "step": 4991
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7618,
+ "step": 4992
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.5543,
+ "step": 4993
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 1.1244,
+ "step": 4994
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.9058,
+ "step": 4995
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6997,
+ "step": 4996
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.9399,
+ "step": 4997
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7832,
+ "step": 4998
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6344,
+ "step": 4999
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.691,
+ "step": 5000
+ },
+ {
+ "epoch": 0.82,
+ "eval_loss": 0.7546924948692322,
+ "eval_runtime": 131.6828,
+ "eval_samples_per_second": 7.594,
+ "eval_steps_per_second": 3.797,
+ "step": 5000
+ },
+ {
+ "epoch": 0.82,
+ "mmlu_eval_accuracy": 0.5418852977811036,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
+ "mmlu_eval_accuracy_astronomy": 0.625,
+ "mmlu_eval_accuracy_business_ethics": 0.7272727272727273,
+ "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": 0.18181818181818182,
+ "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
+ "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.5,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.6363636363636364,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
+ "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
+ "mmlu_eval_accuracy_high_school_psychology": 0.85,
+ "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
+ "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
+ "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
+ "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": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.72,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
+ "mmlu_eval_accuracy_moral_disputes": 0.5,
+ "mmlu_eval_accuracy_moral_scenarios": 0.36,
+ "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
+ "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
+ "mmlu_eval_accuracy_prehistory": 0.6,
+ "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
+ "mmlu_eval_accuracy_professional_law": 0.37058823529411766,
+ "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
+ "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": 0.6818181818181818,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.631578947368421,
+ "mmlu_loss": 1.0400860130047673,
+ "step": 5000
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8872,
+ "step": 5001
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7547,
+ "step": 5002
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7562,
+ "step": 5003
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.9166,
+ "step": 5004
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8079,
+ "step": 5005
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.9695,
+ "step": 5006
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8322,
+ "step": 5007
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.919,
+ "step": 5008
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6851,
+ "step": 5009
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6177,
+ "step": 5010
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.869,
+ "step": 5011
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.612,
+ "step": 5012
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.7186,
+ "step": 5013
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8265,
+ "step": 5014
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8337,
+ "step": 5015
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6366,
+ "step": 5016
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8563,
+ "step": 5017
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6517,
+ "step": 5018
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8611,
+ "step": 5019
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.6829,
+ "step": 5020
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.0002,
+ "loss": 0.8569,
+ "step": 5021
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6617,
+ "step": 5022
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6274,
+ "step": 5023
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.534,
+ "step": 5024
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 1.1059,
+ "step": 5025
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8075,
+ "step": 5026
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9163,
+ "step": 5027
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9912,
+ "step": 5028
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6242,
+ "step": 5029
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9923,
+ "step": 5030
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6908,
+ "step": 5031
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8626,
+ "step": 5032
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.3738,
+ "step": 5033
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9007,
+ "step": 5034
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9533,
+ "step": 5035
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7739,
+ "step": 5036
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9124,
+ "step": 5037
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.5033,
+ "step": 5038
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6412,
+ "step": 5039
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8114,
+ "step": 5040
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 1.0448,
+ "step": 5041
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6149,
+ "step": 5042
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7747,
+ "step": 5043
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7328,
+ "step": 5044
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7047,
+ "step": 5045
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9814,
+ "step": 5046
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 1.0747,
+ "step": 5047
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7094,
+ "step": 5048
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9872,
+ "step": 5049
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9597,
+ "step": 5050
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7743,
+ "step": 5051
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6882,
+ "step": 5052
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.5816,
+ "step": 5053
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8778,
+ "step": 5054
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8066,
+ "step": 5055
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8358,
+ "step": 5056
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8457,
+ "step": 5057
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9675,
+ "step": 5058
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7536,
+ "step": 5059
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9672,
+ "step": 5060
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8778,
+ "step": 5061
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.531,
+ "step": 5062
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7825,
+ "step": 5063
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8806,
+ "step": 5064
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9126,
+ "step": 5065
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8396,
+ "step": 5066
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7679,
+ "step": 5067
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.7333,
+ "step": 5068
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.4986,
+ "step": 5069
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.5319,
+ "step": 5070
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8008,
+ "step": 5071
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.9164,
+ "step": 5072
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8599,
+ "step": 5073
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.5823,
+ "step": 5074
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.5219,
+ "step": 5075
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.5349,
+ "step": 5076
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.6725,
+ "step": 5077
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8132,
+ "step": 5078
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8611,
+ "step": 5079
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8765,
+ "step": 5080
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.8415,
+ "step": 5081
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.0002,
+ "loss": 0.899,
+ "step": 5082
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.71,
+ "step": 5083
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.908,
+ "step": 5084
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8432,
+ "step": 5085
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6448,
+ "step": 5086
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7767,
+ "step": 5087
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7631,
+ "step": 5088
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8108,
+ "step": 5089
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6807,
+ "step": 5090
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7838,
+ "step": 5091
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7977,
+ "step": 5092
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8451,
+ "step": 5093
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.9068,
+ "step": 5094
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.9989,
+ "step": 5095
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8002,
+ "step": 5096
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7866,
+ "step": 5097
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7501,
+ "step": 5098
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.4009,
+ "step": 5099
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8492,
+ "step": 5100
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.0456,
+ "step": 5101
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.0063,
+ "step": 5102
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7048,
+ "step": 5103
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.93,
+ "step": 5104
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.1099,
+ "step": 5105
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.0477,
+ "step": 5106
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.9419,
+ "step": 5107
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7155,
+ "step": 5108
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.782,
+ "step": 5109
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.0007,
+ "step": 5110
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8387,
+ "step": 5111
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8626,
+ "step": 5112
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8566,
+ "step": 5113
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.821,
+ "step": 5114
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6982,
+ "step": 5115
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8734,
+ "step": 5116
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6952,
+ "step": 5117
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6359,
+ "step": 5118
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8674,
+ "step": 5119
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.9601,
+ "step": 5120
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8296,
+ "step": 5121
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.798,
+ "step": 5122
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6152,
+ "step": 5123
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.9122,
+ "step": 5124
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.3275,
+ "step": 5125
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.1467,
+ "step": 5126
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8109,
+ "step": 5127
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8532,
+ "step": 5128
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8827,
+ "step": 5129
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6189,
+ "step": 5130
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8095,
+ "step": 5131
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.5149,
+ "step": 5132
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7072,
+ "step": 5133
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8098,
+ "step": 5134
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6731,
+ "step": 5135
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.8144,
+ "step": 5136
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.0108,
+ "step": 5137
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.6936,
+ "step": 5138
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.7974,
+ "step": 5139
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.596,
+ "step": 5140
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 0.9056,
+ "step": 5141
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0002,
+ "loss": 1.1622,
+ "step": 5142
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.4363,
+ "step": 5143
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8767,
+ "step": 5144
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7319,
+ "step": 5145
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.591,
+ "step": 5146
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6934,
+ "step": 5147
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8133,
+ "step": 5148
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.869,
+ "step": 5149
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.5649,
+ "step": 5150
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6343,
+ "step": 5151
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.5792,
+ "step": 5152
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.765,
+ "step": 5153
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7543,
+ "step": 5154
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7557,
+ "step": 5155
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6972,
+ "step": 5156
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.5681,
+ "step": 5157
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.787,
+ "step": 5158
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8267,
+ "step": 5159
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.978,
+ "step": 5160
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.5671,
+ "step": 5161
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6536,
+ "step": 5162
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.736,
+ "step": 5163
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7226,
+ "step": 5164
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.4204,
+ "step": 5165
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6504,
+ "step": 5166
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8797,
+ "step": 5167
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6669,
+ "step": 5168
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8974,
+ "step": 5169
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6961,
+ "step": 5170
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7178,
+ "step": 5171
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.9813,
+ "step": 5172
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8421,
+ "step": 5173
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7525,
+ "step": 5174
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7244,
+ "step": 5175
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 1.0433,
+ "step": 5176
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6234,
+ "step": 5177
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7653,
+ "step": 5178
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6155,
+ "step": 5179
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6304,
+ "step": 5180
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6232,
+ "step": 5181
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8279,
+ "step": 5182
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6165,
+ "step": 5183
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.613,
+ "step": 5184
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7917,
+ "step": 5185
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7079,
+ "step": 5186
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 1.2307,
+ "step": 5187
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8599,
+ "step": 5188
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.65,
+ "step": 5189
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7754,
+ "step": 5190
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.6245,
+ "step": 5191
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.72,
+ "step": 5192
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.5067,
+ "step": 5193
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8264,
+ "step": 5194
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8361,
+ "step": 5195
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8509,
+ "step": 5196
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7017,
+ "step": 5197
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7595,
+ "step": 5198
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8285,
+ "step": 5199
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 1.0161,
+ "step": 5200
+ },
+ {
+ "epoch": 0.85,
+ "eval_loss": 0.7570607662200928,
+ "eval_runtime": 131.5936,
+ "eval_samples_per_second": 7.599,
+ "eval_steps_per_second": 3.8,
+ "step": 5200
+ },
+ {
+ "epoch": 0.85,
+ "mmlu_eval_accuracy": 0.5391922125688261,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.75,
+ "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
+ "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": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_medicine": 0.5,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
+ "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
+ "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
+ "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": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.5,
+ "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": 0.6363636363636364,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
+ "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
+ "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
+ "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+ "mmlu_eval_accuracy_international_law": 1.0,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
+ "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
+ "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
+ "mmlu_eval_accuracy_moral_scenarios": 0.35,
+ "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
+ "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
+ "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
+ "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
+ "mmlu_eval_accuracy_professional_law": 0.31176470588235294,
+ "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+ "mmlu_eval_accuracy_sociology": 0.6363636363636364,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+ "mmlu_loss": 1.2944577410622615,
+ "step": 5200
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.7148,
+ "step": 5201
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.8881,
+ "step": 5202
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0002,
+ "loss": 0.731,
+ "step": 5203
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6718,
+ "step": 5204
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5627,
+ "step": 5205
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7415,
+ "step": 5206
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8756,
+ "step": 5207
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7867,
+ "step": 5208
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7876,
+ "step": 5209
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.9018,
+ "step": 5210
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.662,
+ "step": 5211
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8326,
+ "step": 5212
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5151,
+ "step": 5213
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8341,
+ "step": 5214
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6266,
+ "step": 5215
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8856,
+ "step": 5216
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5794,
+ "step": 5217
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7445,
+ "step": 5218
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.3348,
+ "step": 5219
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6973,
+ "step": 5220
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7916,
+ "step": 5221
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7283,
+ "step": 5222
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 1.1066,
+ "step": 5223
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5296,
+ "step": 5224
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.9062,
+ "step": 5225
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5675,
+ "step": 5226
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5619,
+ "step": 5227
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5829,
+ "step": 5228
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8905,
+ "step": 5229
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7783,
+ "step": 5230
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8705,
+ "step": 5231
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7131,
+ "step": 5232
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8611,
+ "step": 5233
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.9494,
+ "step": 5234
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6996,
+ "step": 5235
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.91,
+ "step": 5236
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7032,
+ "step": 5237
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8203,
+ "step": 5238
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7117,
+ "step": 5239
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8512,
+ "step": 5240
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7162,
+ "step": 5241
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7058,
+ "step": 5242
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7749,
+ "step": 5243
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7576,
+ "step": 5244
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8583,
+ "step": 5245
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7959,
+ "step": 5246
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7251,
+ "step": 5247
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6967,
+ "step": 5248
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.8427,
+ "step": 5249
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5726,
+ "step": 5250
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6307,
+ "step": 5251
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7522,
+ "step": 5252
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5172,
+ "step": 5253
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.46,
+ "step": 5254
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6092,
+ "step": 5255
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6503,
+ "step": 5256
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7304,
+ "step": 5257
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6034,
+ "step": 5258
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.6499,
+ "step": 5259
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7869,
+ "step": 5260
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.5615,
+ "step": 5261
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.722,
+ "step": 5262
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7543,
+ "step": 5263
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.0002,
+ "loss": 0.7606,
+ "step": 5264
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7863,
+ "step": 5265
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 1.0483,
+ "step": 5266
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 1.0626,
+ "step": 5267
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7134,
+ "step": 5268
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 1.2534,
+ "step": 5269
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7122,
+ "step": 5270
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7735,
+ "step": 5271
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7985,
+ "step": 5272
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6574,
+ "step": 5273
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.5534,
+ "step": 5274
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.9228,
+ "step": 5275
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.845,
+ "step": 5276
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6409,
+ "step": 5277
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7251,
+ "step": 5278
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6542,
+ "step": 5279
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8726,
+ "step": 5280
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6983,
+ "step": 5281
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.5973,
+ "step": 5282
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.9301,
+ "step": 5283
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.4312,
+ "step": 5284
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6934,
+ "step": 5285
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8837,
+ "step": 5286
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6669,
+ "step": 5287
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8208,
+ "step": 5288
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8299,
+ "step": 5289
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6955,
+ "step": 5290
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6232,
+ "step": 5291
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6172,
+ "step": 5292
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7974,
+ "step": 5293
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7701,
+ "step": 5294
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7202,
+ "step": 5295
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8934,
+ "step": 5296
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 1.0895,
+ "step": 5297
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.9629,
+ "step": 5298
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8588,
+ "step": 5299
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.94,
+ "step": 5300
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8273,
+ "step": 5301
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.9395,
+ "step": 5302
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.5857,
+ "step": 5303
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6201,
+ "step": 5304
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8031,
+ "step": 5305
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6865,
+ "step": 5306
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.693,
+ "step": 5307
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6235,
+ "step": 5308
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 1.1614,
+ "step": 5309
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8339,
+ "step": 5310
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6874,
+ "step": 5311
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7138,
+ "step": 5312
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7386,
+ "step": 5313
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8596,
+ "step": 5314
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7985,
+ "step": 5315
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6018,
+ "step": 5316
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.9183,
+ "step": 5317
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.5919,
+ "step": 5318
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.8873,
+ "step": 5319
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6674,
+ "step": 5320
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.9526,
+ "step": 5321
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7959,
+ "step": 5322
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.7442,
+ "step": 5323
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.9682,
+ "step": 5324
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.0002,
+ "loss": 0.6986,
+ "step": 5325
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 1.0101,
+ "step": 5326
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8869,
+ "step": 5327
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7313,
+ "step": 5328
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7464,
+ "step": 5329
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6972,
+ "step": 5330
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8185,
+ "step": 5331
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9128,
+ "step": 5332
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8034,
+ "step": 5333
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9641,
+ "step": 5334
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7911,
+ "step": 5335
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8467,
+ "step": 5336
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7767,
+ "step": 5337
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7157,
+ "step": 5338
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8093,
+ "step": 5339
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.858,
+ "step": 5340
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.878,
+ "step": 5341
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.5611,
+ "step": 5342
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9083,
+ "step": 5343
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.743,
+ "step": 5344
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6457,
+ "step": 5345
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7839,
+ "step": 5346
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6778,
+ "step": 5347
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9833,
+ "step": 5348
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.4485,
+ "step": 5349
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8594,
+ "step": 5350
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6182,
+ "step": 5351
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.801,
+ "step": 5352
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.5387,
+ "step": 5353
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8203,
+ "step": 5354
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.778,
+ "step": 5355
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8081,
+ "step": 5356
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6091,
+ "step": 5357
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8204,
+ "step": 5358
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7605,
+ "step": 5359
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7611,
+ "step": 5360
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9418,
+ "step": 5361
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.5927,
+ "step": 5362
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.761,
+ "step": 5363
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9497,
+ "step": 5364
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8418,
+ "step": 5365
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6609,
+ "step": 5366
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8712,
+ "step": 5367
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.5725,
+ "step": 5368
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6984,
+ "step": 5369
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6018,
+ "step": 5370
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6502,
+ "step": 5371
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 1.0874,
+ "step": 5372
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.5237,
+ "step": 5373
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6227,
+ "step": 5374
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.4831,
+ "step": 5375
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9457,
+ "step": 5376
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.704,
+ "step": 5377
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.712,
+ "step": 5378
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8411,
+ "step": 5379
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9111,
+ "step": 5380
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.7319,
+ "step": 5381
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6228,
+ "step": 5382
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.9846,
+ "step": 5383
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8681,
+ "step": 5384
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.6221,
+ "step": 5385
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.0002,
+ "loss": 0.8166,
+ "step": 5386
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 1.0206,
+ "step": 5387
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8194,
+ "step": 5388
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9925,
+ "step": 5389
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9388,
+ "step": 5390
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6106,
+ "step": 5391
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.4923,
+ "step": 5392
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6819,
+ "step": 5393
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8685,
+ "step": 5394
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.798,
+ "step": 5395
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.894,
+ "step": 5396
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7512,
+ "step": 5397
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.5001,
+ "step": 5398
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.752,
+ "step": 5399
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8313,
+ "step": 5400
+ },
+ {
+ "epoch": 0.89,
+ "eval_loss": 0.7573027610778809,
+ "eval_runtime": 131.61,
+ "eval_samples_per_second": 7.598,
+ "eval_steps_per_second": 3.799,
+ "step": 5400
+ },
+ {
+ "epoch": 0.89,
+ "mmlu_eval_accuracy": 0.5535363827788407,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.5,
+ "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": 0.125,
+ "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
+ "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
+ "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
+ "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": 0.4146341463414634,
+ "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": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
+ "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
+ "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": 0.8,
+ "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
+ "mmlu_eval_accuracy_human_aging": 0.782608695652174,
+ "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": 0.6111111111111112,
+ "mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
+ "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.34,
+ "mmlu_eval_accuracy_nutrition": 0.696969696969697,
+ "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
+ "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
+ "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
+ "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
+ "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
+ "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.7037037037037037,
+ "mmlu_eval_accuracy_sociology": 0.8636363636363636,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.631578947368421,
+ "mmlu_loss": 1.0044040840226116,
+ "step": 5400
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 1.0322,
+ "step": 5401
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7576,
+ "step": 5402
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9805,
+ "step": 5403
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8773,
+ "step": 5404
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9102,
+ "step": 5405
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9518,
+ "step": 5406
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8061,
+ "step": 5407
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.722,
+ "step": 5408
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 1.1114,
+ "step": 5409
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7466,
+ "step": 5410
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8599,
+ "step": 5411
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9889,
+ "step": 5412
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7419,
+ "step": 5413
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7471,
+ "step": 5414
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6773,
+ "step": 5415
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6627,
+ "step": 5416
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7799,
+ "step": 5417
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6111,
+ "step": 5418
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7181,
+ "step": 5419
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 1.0414,
+ "step": 5420
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7894,
+ "step": 5421
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7075,
+ "step": 5422
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 1.0978,
+ "step": 5423
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8059,
+ "step": 5424
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6369,
+ "step": 5425
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.752,
+ "step": 5426
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9878,
+ "step": 5427
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7314,
+ "step": 5428
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.977,
+ "step": 5429
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8139,
+ "step": 5430
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.4639,
+ "step": 5431
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8779,
+ "step": 5432
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7548,
+ "step": 5433
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.919,
+ "step": 5434
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7097,
+ "step": 5435
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7139,
+ "step": 5436
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7733,
+ "step": 5437
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7562,
+ "step": 5438
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6196,
+ "step": 5439
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7418,
+ "step": 5440
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.6483,
+ "step": 5441
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 1.0091,
+ "step": 5442
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.5953,
+ "step": 5443
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.9923,
+ "step": 5444
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8842,
+ "step": 5445
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.7261,
+ "step": 5446
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.0002,
+ "loss": 0.8973,
+ "step": 5447
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8355,
+ "step": 5448
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9119,
+ "step": 5449
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6546,
+ "step": 5450
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6234,
+ "step": 5451
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7652,
+ "step": 5452
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9246,
+ "step": 5453
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8217,
+ "step": 5454
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6573,
+ "step": 5455
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8254,
+ "step": 5456
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9003,
+ "step": 5457
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8375,
+ "step": 5458
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7139,
+ "step": 5459
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6549,
+ "step": 5460
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.603,
+ "step": 5461
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8928,
+ "step": 5462
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6254,
+ "step": 5463
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.709,
+ "step": 5464
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.909,
+ "step": 5465
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6673,
+ "step": 5466
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6139,
+ "step": 5467
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6717,
+ "step": 5468
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.5793,
+ "step": 5469
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9662,
+ "step": 5470
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6386,
+ "step": 5471
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7888,
+ "step": 5472
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8934,
+ "step": 5473
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8559,
+ "step": 5474
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9941,
+ "step": 5475
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9985,
+ "step": 5476
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9379,
+ "step": 5477
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9548,
+ "step": 5478
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9312,
+ "step": 5479
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8416,
+ "step": 5480
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.823,
+ "step": 5481
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6526,
+ "step": 5482
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7048,
+ "step": 5483
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8494,
+ "step": 5484
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6002,
+ "step": 5485
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.808,
+ "step": 5486
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6146,
+ "step": 5487
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8975,
+ "step": 5488
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8299,
+ "step": 5489
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.508,
+ "step": 5490
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8859,
+ "step": 5491
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.4986,
+ "step": 5492
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7364,
+ "step": 5493
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.5677,
+ "step": 5494
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9249,
+ "step": 5495
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.9316,
+ "step": 5496
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7841,
+ "step": 5497
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7884,
+ "step": 5498
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.3761,
+ "step": 5499
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8021,
+ "step": 5500
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6339,
+ "step": 5501
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8181,
+ "step": 5502
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8774,
+ "step": 5503
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.7942,
+ "step": 5504
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.8235,
+ "step": 5505
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.6374,
+ "step": 5506
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.4308,
+ "step": 5507
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.0002,
+ "loss": 0.4129,
+ "step": 5508
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9604,
+ "step": 5509
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8313,
+ "step": 5510
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7495,
+ "step": 5511
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9446,
+ "step": 5512
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6674,
+ "step": 5513
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9904,
+ "step": 5514
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.989,
+ "step": 5515
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6253,
+ "step": 5516
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.5869,
+ "step": 5517
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8086,
+ "step": 5518
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7037,
+ "step": 5519
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 1.1378,
+ "step": 5520
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.5666,
+ "step": 5521
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8733,
+ "step": 5522
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8578,
+ "step": 5523
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6893,
+ "step": 5524
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.738,
+ "step": 5525
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7215,
+ "step": 5526
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6664,
+ "step": 5527
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 1.0088,
+ "step": 5528
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8572,
+ "step": 5529
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9753,
+ "step": 5530
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9261,
+ "step": 5531
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9845,
+ "step": 5532
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7809,
+ "step": 5533
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.5954,
+ "step": 5534
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7221,
+ "step": 5535
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8189,
+ "step": 5536
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8694,
+ "step": 5537
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9574,
+ "step": 5538
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6938,
+ "step": 5539
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9122,
+ "step": 5540
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8863,
+ "step": 5541
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.4845,
+ "step": 5542
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.5207,
+ "step": 5543
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7122,
+ "step": 5544
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.925,
+ "step": 5545
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7193,
+ "step": 5546
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9306,
+ "step": 5547
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6802,
+ "step": 5548
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8898,
+ "step": 5549
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7773,
+ "step": 5550
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6858,
+ "step": 5551
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8281,
+ "step": 5552
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9012,
+ "step": 5553
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.575,
+ "step": 5554
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8427,
+ "step": 5555
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9052,
+ "step": 5556
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7926,
+ "step": 5557
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.6887,
+ "step": 5558
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7351,
+ "step": 5559
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9455,
+ "step": 5560
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7989,
+ "step": 5561
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.733,
+ "step": 5562
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8826,
+ "step": 5563
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7459,
+ "step": 5564
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7546,
+ "step": 5565
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.9101,
+ "step": 5566
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.7619,
+ "step": 5567
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0002,
+ "loss": 0.8338,
+ "step": 5568
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7217,
+ "step": 5569
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 1.1281,
+ "step": 5570
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.5675,
+ "step": 5571
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 1.0303,
+ "step": 5572
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6281,
+ "step": 5573
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7133,
+ "step": 5574
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.9057,
+ "step": 5575
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.9976,
+ "step": 5576
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 1.0461,
+ "step": 5577
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7773,
+ "step": 5578
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.942,
+ "step": 5579
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.8754,
+ "step": 5580
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.4891,
+ "step": 5581
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.8193,
+ "step": 5582
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6067,
+ "step": 5583
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.8549,
+ "step": 5584
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7848,
+ "step": 5585
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6281,
+ "step": 5586
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7282,
+ "step": 5587
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.9881,
+ "step": 5588
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 1.0898,
+ "step": 5589
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.8647,
+ "step": 5590
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6521,
+ "step": 5591
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.5586,
+ "step": 5592
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.589,
+ "step": 5593
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.8831,
+ "step": 5594
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6811,
+ "step": 5595
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.5972,
+ "step": 5596
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.5166,
+ "step": 5597
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6318,
+ "step": 5598
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7892,
+ "step": 5599
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6298,
+ "step": 5600
+ },
+ {
+ "epoch": 0.92,
+ "eval_loss": 0.7568509578704834,
+ "eval_runtime": 131.6466,
+ "eval_samples_per_second": 7.596,
+ "eval_steps_per_second": 3.798,
+ "step": 5600
+ },
+ {
+ "epoch": 0.92,
+ "mmlu_eval_accuracy": 0.5302614353168894,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5,
+ "mmlu_eval_accuracy_astronomy": 0.5,
+ "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
+ "mmlu_eval_accuracy_college_biology": 0.5625,
+ "mmlu_eval_accuracy_college_chemistry": 0.125,
+ "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
+ "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
+ "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
+ "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+ "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": 0.25,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.59375,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
+ "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
+ "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
+ "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
+ "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": 0.6956521739130435,
+ "mmlu_eval_accuracy_human_sexuality": 0.5,
+ "mmlu_eval_accuracy_international_law": 1.0,
+ "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
+ "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.88,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
+ "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
+ "mmlu_eval_accuracy_moral_scenarios": 0.29,
+ "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
+ "mmlu_eval_accuracy_philosophy": 0.5,
+ "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
+ "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
+ "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
+ "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
+ "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+ "mmlu_eval_accuracy_sociology": 0.7727272727272727,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.631578947368421,
+ "mmlu_loss": 1.2112150888102173,
+ "step": 5600
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.8411,
+ "step": 5601
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7866,
+ "step": 5602
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6554,
+ "step": 5603
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7234,
+ "step": 5604
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.4037,
+ "step": 5605
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.9236,
+ "step": 5606
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7781,
+ "step": 5607
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.917,
+ "step": 5608
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7848,
+ "step": 5609
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.5141,
+ "step": 5610
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 1.0792,
+ "step": 5611
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.9425,
+ "step": 5612
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6924,
+ "step": 5613
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.824,
+ "step": 5614
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.4964,
+ "step": 5615
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.767,
+ "step": 5616
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.5439,
+ "step": 5617
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6681,
+ "step": 5618
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.987,
+ "step": 5619
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.9177,
+ "step": 5620
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7167,
+ "step": 5621
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.9006,
+ "step": 5622
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.8574,
+ "step": 5623
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7821,
+ "step": 5624
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 1.3664,
+ "step": 5625
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6688,
+ "step": 5626
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7017,
+ "step": 5627
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.7292,
+ "step": 5628
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0002,
+ "loss": 0.6384,
+ "step": 5629
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.915,
+ "step": 5630
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8656,
+ "step": 5631
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7386,
+ "step": 5632
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8869,
+ "step": 5633
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.756,
+ "step": 5634
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.532,
+ "step": 5635
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7359,
+ "step": 5636
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.691,
+ "step": 5637
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8208,
+ "step": 5638
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8302,
+ "step": 5639
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6807,
+ "step": 5640
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7715,
+ "step": 5641
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 1.1177,
+ "step": 5642
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8522,
+ "step": 5643
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7695,
+ "step": 5644
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8631,
+ "step": 5645
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7542,
+ "step": 5646
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8105,
+ "step": 5647
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6049,
+ "step": 5648
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6164,
+ "step": 5649
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6373,
+ "step": 5650
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6218,
+ "step": 5651
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 1.0081,
+ "step": 5652
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8226,
+ "step": 5653
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6212,
+ "step": 5654
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.5167,
+ "step": 5655
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7182,
+ "step": 5656
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7793,
+ "step": 5657
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6875,
+ "step": 5658
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8738,
+ "step": 5659
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.5131,
+ "step": 5660
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.4931,
+ "step": 5661
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7163,
+ "step": 5662
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7372,
+ "step": 5663
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6434,
+ "step": 5664
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8497,
+ "step": 5665
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7592,
+ "step": 5666
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.5747,
+ "step": 5667
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6075,
+ "step": 5668
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7951,
+ "step": 5669
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7659,
+ "step": 5670
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6826,
+ "step": 5671
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8356,
+ "step": 5672
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 1.0523,
+ "step": 5673
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.76,
+ "step": 5674
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.973,
+ "step": 5675
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.6099,
+ "step": 5676
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.9101,
+ "step": 5677
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7137,
+ "step": 5678
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8016,
+ "step": 5679
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8851,
+ "step": 5680
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.77,
+ "step": 5681
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7092,
+ "step": 5682
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 1.1899,
+ "step": 5683
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.5347,
+ "step": 5684
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7378,
+ "step": 5685
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.8622,
+ "step": 5686
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7049,
+ "step": 5687
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.5822,
+ "step": 5688
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.9036,
+ "step": 5689
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.0002,
+ "loss": 0.7228,
+ "step": 5690
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9574,
+ "step": 5691
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.973,
+ "step": 5692
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.754,
+ "step": 5693
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7229,
+ "step": 5694
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.6008,
+ "step": 5695
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8347,
+ "step": 5696
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8039,
+ "step": 5697
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7682,
+ "step": 5698
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9045,
+ "step": 5699
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9489,
+ "step": 5700
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7213,
+ "step": 5701
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7936,
+ "step": 5702
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.6357,
+ "step": 5703
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.4616,
+ "step": 5704
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8291,
+ "step": 5705
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.5767,
+ "step": 5706
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8355,
+ "step": 5707
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7081,
+ "step": 5708
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8414,
+ "step": 5709
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9,
+ "step": 5710
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8502,
+ "step": 5711
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8572,
+ "step": 5712
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.73,
+ "step": 5713
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7497,
+ "step": 5714
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9662,
+ "step": 5715
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.713,
+ "step": 5716
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7176,
+ "step": 5717
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.5951,
+ "step": 5718
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8369,
+ "step": 5719
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9989,
+ "step": 5720
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.6209,
+ "step": 5721
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.5043,
+ "step": 5722
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.871,
+ "step": 5723
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8965,
+ "step": 5724
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.5495,
+ "step": 5725
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.6294,
+ "step": 5726
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.6597,
+ "step": 5727
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7899,
+ "step": 5728
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.5511,
+ "step": 5729
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8029,
+ "step": 5730
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9552,
+ "step": 5731
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.689,
+ "step": 5732
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8029,
+ "step": 5733
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9937,
+ "step": 5734
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7684,
+ "step": 5735
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7458,
+ "step": 5736
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8201,
+ "step": 5737
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 1.2752,
+ "step": 5738
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.884,
+ "step": 5739
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7943,
+ "step": 5740
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 1.0416,
+ "step": 5741
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7158,
+ "step": 5742
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8879,
+ "step": 5743
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8423,
+ "step": 5744
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.7611,
+ "step": 5745
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.5072,
+ "step": 5746
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 1.1542,
+ "step": 5747
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.6498,
+ "step": 5748
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.8678,
+ "step": 5749
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.6684,
+ "step": 5750
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.0002,
+ "loss": 0.9313,
+ "step": 5751
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7082,
+ "step": 5752
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7943,
+ "step": 5753
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.705,
+ "step": 5754
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.853,
+ "step": 5755
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8636,
+ "step": 5756
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7533,
+ "step": 5757
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.5873,
+ "step": 5758
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7943,
+ "step": 5759
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.666,
+ "step": 5760
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6332,
+ "step": 5761
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 1.4062,
+ "step": 5762
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8999,
+ "step": 5763
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.3884,
+ "step": 5764
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7841,
+ "step": 5765
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8468,
+ "step": 5766
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7001,
+ "step": 5767
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6359,
+ "step": 5768
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.42,
+ "step": 5769
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6682,
+ "step": 5770
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8387,
+ "step": 5771
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6663,
+ "step": 5772
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.9856,
+ "step": 5773
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.9537,
+ "step": 5774
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.794,
+ "step": 5775
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8019,
+ "step": 5776
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6929,
+ "step": 5777
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7452,
+ "step": 5778
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6483,
+ "step": 5779
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7669,
+ "step": 5780
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.5762,
+ "step": 5781
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.9958,
+ "step": 5782
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 1.0292,
+ "step": 5783
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.747,
+ "step": 5784
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6999,
+ "step": 5785
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7663,
+ "step": 5786
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.696,
+ "step": 5787
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7454,
+ "step": 5788
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7263,
+ "step": 5789
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8639,
+ "step": 5790
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7553,
+ "step": 5791
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.5242,
+ "step": 5792
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8186,
+ "step": 5793
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 1.2811,
+ "step": 5794
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8176,
+ "step": 5795
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7739,
+ "step": 5796
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8576,
+ "step": 5797
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7849,
+ "step": 5798
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7429,
+ "step": 5799
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6609,
+ "step": 5800
+ },
+ {
+ "epoch": 0.95,
+ "eval_loss": 0.7572384476661682,
+ "eval_runtime": 131.6327,
+ "eval_samples_per_second": 7.597,
+ "eval_steps_per_second": 3.798,
+ "step": 5800
+ },
+ {
+ "epoch": 0.95,
+ "mmlu_eval_accuracy": 0.5230331495492276,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
+ "mmlu_eval_accuracy_astronomy": 0.5625,
+ "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
+ "mmlu_eval_accuracy_college_biology": 0.625,
+ "mmlu_eval_accuracy_college_chemistry": 0.25,
+ "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
+ "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
+ "mmlu_eval_accuracy_college_medicine": 0.5,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
+ "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
+ "mmlu_eval_accuracy_electrical_engineering": 0.4375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
+ "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
+ "mmlu_eval_accuracy_global_facts": 0.1,
+ "mmlu_eval_accuracy_high_school_biology": 0.53125,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
+ "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
+ "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
+ "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
+ "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
+ "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
+ "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": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 1.0,
+ "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
+ "mmlu_eval_accuracy_moral_disputes": 0.5,
+ "mmlu_eval_accuracy_moral_scenarios": 0.35,
+ "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
+ "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
+ "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
+ "mmlu_eval_accuracy_professional_law": 0.32941176470588235,
+ "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
+ "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+ "mmlu_eval_accuracy_sociology": 0.8181818181818182,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
+ "mmlu_eval_accuracy_virology": 0.3888888888888889,
+ "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+ "mmlu_loss": 1.1871224955385722,
+ "step": 5800
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.9059,
+ "step": 5801
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.9874,
+ "step": 5802
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7752,
+ "step": 5803
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.9475,
+ "step": 5804
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8847,
+ "step": 5805
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7602,
+ "step": 5806
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.9615,
+ "step": 5807
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8496,
+ "step": 5808
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7267,
+ "step": 5809
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.7287,
+ "step": 5810
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.6387,
+ "step": 5811
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.0002,
+ "loss": 0.8825,
+ "step": 5812
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.5843,
+ "step": 5813
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7989,
+ "step": 5814
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8118,
+ "step": 5815
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6368,
+ "step": 5816
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.0179,
+ "step": 5817
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7691,
+ "step": 5818
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.2382,
+ "step": 5819
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7772,
+ "step": 5820
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.5415,
+ "step": 5821
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8365,
+ "step": 5822
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6377,
+ "step": 5823
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8377,
+ "step": 5824
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.0803,
+ "step": 5825
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7633,
+ "step": 5826
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6959,
+ "step": 5827
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.829,
+ "step": 5828
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.002,
+ "step": 5829
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.011,
+ "step": 5830
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6827,
+ "step": 5831
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.0655,
+ "step": 5832
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8833,
+ "step": 5833
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.9107,
+ "step": 5834
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6728,
+ "step": 5835
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.63,
+ "step": 5836
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8589,
+ "step": 5837
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7529,
+ "step": 5838
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6623,
+ "step": 5839
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.9483,
+ "step": 5840
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8995,
+ "step": 5841
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.9031,
+ "step": 5842
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.825,
+ "step": 5843
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6552,
+ "step": 5844
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8065,
+ "step": 5845
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8009,
+ "step": 5846
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7398,
+ "step": 5847
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.9893,
+ "step": 5848
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7149,
+ "step": 5849
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7349,
+ "step": 5850
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.4984,
+ "step": 5851
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6512,
+ "step": 5852
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.799,
+ "step": 5853
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7188,
+ "step": 5854
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8869,
+ "step": 5855
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.2405,
+ "step": 5856
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.5321,
+ "step": 5857
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.0541,
+ "step": 5858
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7558,
+ "step": 5859
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7913,
+ "step": 5860
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8548,
+ "step": 5861
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.9363,
+ "step": 5862
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7766,
+ "step": 5863
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.9109,
+ "step": 5864
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7699,
+ "step": 5865
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.2606,
+ "step": 5866
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.6255,
+ "step": 5867
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.768,
+ "step": 5868
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7604,
+ "step": 5869
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7688,
+ "step": 5870
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.8899,
+ "step": 5871
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 1.1141,
+ "step": 5872
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0002,
+ "loss": 0.7358,
+ "step": 5873
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.5494,
+ "step": 5874
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6543,
+ "step": 5875
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9928,
+ "step": 5876
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9858,
+ "step": 5877
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6624,
+ "step": 5878
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.683,
+ "step": 5879
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8244,
+ "step": 5880
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6677,
+ "step": 5881
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8965,
+ "step": 5882
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.873,
+ "step": 5883
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 1.2479,
+ "step": 5884
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.5186,
+ "step": 5885
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8986,
+ "step": 5886
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6506,
+ "step": 5887
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6354,
+ "step": 5888
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6855,
+ "step": 5889
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6733,
+ "step": 5890
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.703,
+ "step": 5891
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.5373,
+ "step": 5892
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6051,
+ "step": 5893
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6242,
+ "step": 5894
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7562,
+ "step": 5895
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9275,
+ "step": 5896
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8822,
+ "step": 5897
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 1.0804,
+ "step": 5898
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8364,
+ "step": 5899
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6658,
+ "step": 5900
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8575,
+ "step": 5901
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7826,
+ "step": 5902
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6238,
+ "step": 5903
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 1.1837,
+ "step": 5904
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6654,
+ "step": 5905
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.5047,
+ "step": 5906
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9894,
+ "step": 5907
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.746,
+ "step": 5908
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7391,
+ "step": 5909
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6154,
+ "step": 5910
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9496,
+ "step": 5911
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6619,
+ "step": 5912
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 1.0598,
+ "step": 5913
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9148,
+ "step": 5914
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8715,
+ "step": 5915
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8047,
+ "step": 5916
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9877,
+ "step": 5917
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.828,
+ "step": 5918
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6213,
+ "step": 5919
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.5439,
+ "step": 5920
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.808,
+ "step": 5921
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 1.0274,
+ "step": 5922
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7757,
+ "step": 5923
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7065,
+ "step": 5924
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9578,
+ "step": 5925
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7515,
+ "step": 5926
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.6899,
+ "step": 5927
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7149,
+ "step": 5928
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.5282,
+ "step": 5929
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.9435,
+ "step": 5930
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8707,
+ "step": 5931
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.8243,
+ "step": 5932
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 1.0413,
+ "step": 5933
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0002,
+ "loss": 0.7702,
+ "step": 5934
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.4833,
+ "step": 5935
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8733,
+ "step": 5936
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6509,
+ "step": 5937
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7631,
+ "step": 5938
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.989,
+ "step": 5939
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7686,
+ "step": 5940
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6717,
+ "step": 5941
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7277,
+ "step": 5942
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6606,
+ "step": 5943
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7376,
+ "step": 5944
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6214,
+ "step": 5945
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7048,
+ "step": 5946
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6142,
+ "step": 5947
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9274,
+ "step": 5948
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9764,
+ "step": 5949
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9148,
+ "step": 5950
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.5132,
+ "step": 5951
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9114,
+ "step": 5952
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8297,
+ "step": 5953
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8272,
+ "step": 5954
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7253,
+ "step": 5955
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9494,
+ "step": 5956
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9263,
+ "step": 5957
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9341,
+ "step": 5958
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 1.0222,
+ "step": 5959
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6946,
+ "step": 5960
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7541,
+ "step": 5961
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6096,
+ "step": 5962
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7266,
+ "step": 5963
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7313,
+ "step": 5964
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8359,
+ "step": 5965
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7218,
+ "step": 5966
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8403,
+ "step": 5967
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6983,
+ "step": 5968
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9046,
+ "step": 5969
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.5978,
+ "step": 5970
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9524,
+ "step": 5971
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7185,
+ "step": 5972
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8263,
+ "step": 5973
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 1.0198,
+ "step": 5974
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7039,
+ "step": 5975
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9132,
+ "step": 5976
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6496,
+ "step": 5977
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7322,
+ "step": 5978
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8371,
+ "step": 5979
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.9671,
+ "step": 5980
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.6564,
+ "step": 5981
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8938,
+ "step": 5982
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.661,
+ "step": 5983
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8196,
+ "step": 5984
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 1.0196,
+ "step": 5985
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 1.1033,
+ "step": 5986
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.7012,
+ "step": 5987
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8286,
+ "step": 5988
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.586,
+ "step": 5989
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 1.1123,
+ "step": 5990
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8704,
+ "step": 5991
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8366,
+ "step": 5992
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8541,
+ "step": 5993
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0002,
+ "loss": 0.8266,
+ "step": 5994
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9463,
+ "step": 5995
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.7052,
+ "step": 5996
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 1.1841,
+ "step": 5997
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6531,
+ "step": 5998
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.806,
+ "step": 5999
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.7816,
+ "step": 6000
+ },
+ {
+ "epoch": 0.99,
+ "eval_loss": 0.7560673952102661,
+ "eval_runtime": 131.7253,
+ "eval_samples_per_second": 7.592,
+ "eval_steps_per_second": 3.796,
+ "step": 6000
+ },
+ {
+ "epoch": 0.99,
+ "mmlu_eval_accuracy": 0.5541866617056794,
+ "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.6875,
+ "mmlu_eval_accuracy_business_ethics": 0.7272727272727273,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
+ "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": 0.5,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
+ "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
+ "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
+ "mmlu_eval_accuracy_electrical_engineering": 0.4375,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
+ "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+ "mmlu_eval_accuracy_global_facts": 0.4,
+ "mmlu_eval_accuracy_high_school_biology": 0.46875,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
+ "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": 0.7619047619047619,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.4827586206896552,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
+ "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8,
+ "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
+ "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
+ "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": 0.7272727272727273,
+ "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
+ "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
+ "mmlu_eval_accuracy_management": 0.7272727272727273,
+ "mmlu_eval_accuracy_marketing": 0.8,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
+ "mmlu_eval_accuracy_moral_disputes": 0.5,
+ "mmlu_eval_accuracy_moral_scenarios": 0.28,
+ "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
+ "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
+ "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
+ "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
+ "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
+ "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
+ "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
+ "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
+ "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+ "mmlu_eval_accuracy_sociology": 0.8181818181818182,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
+ "mmlu_eval_accuracy_virology": 0.5,
+ "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+ "mmlu_loss": 0.9176231137921854,
+ "step": 6000
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.92,
+ "step": 6001
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6709,
+ "step": 6002
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6509,
+ "step": 6003
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.7731,
+ "step": 6004
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5218,
+ "step": 6005
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5127,
+ "step": 6006
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5466,
+ "step": 6007
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6999,
+ "step": 6008
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8014,
+ "step": 6009
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5503,
+ "step": 6010
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8462,
+ "step": 6011
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.91,
+ "step": 6012
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9856,
+ "step": 6013
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8973,
+ "step": 6014
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6229,
+ "step": 6015
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 1.0128,
+ "step": 6016
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8,
+ "step": 6017
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8851,
+ "step": 6018
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.7596,
+ "step": 6019
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5101,
+ "step": 6020
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.647,
+ "step": 6021
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6608,
+ "step": 6022
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.7162,
+ "step": 6023
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8862,
+ "step": 6024
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5658,
+ "step": 6025
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6894,
+ "step": 6026
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.631,
+ "step": 6027
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9762,
+ "step": 6028
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8543,
+ "step": 6029
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8214,
+ "step": 6030
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8155,
+ "step": 6031
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6961,
+ "step": 6032
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6051,
+ "step": 6033
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8938,
+ "step": 6034
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.964,
+ "step": 6035
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.7863,
+ "step": 6036
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9647,
+ "step": 6037
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6533,
+ "step": 6038
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6301,
+ "step": 6039
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5623,
+ "step": 6040
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9503,
+ "step": 6041
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9377,
+ "step": 6042
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8947,
+ "step": 6043
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.5348,
+ "step": 6044
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9553,
+ "step": 6045
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 1.0936,
+ "step": 6046
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8397,
+ "step": 6047
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6345,
+ "step": 6048
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.4864,
+ "step": 6049
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.9717,
+ "step": 6050
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.8439,
+ "step": 6051
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 1.1819,
+ "step": 6052
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.6844,
+ "step": 6053
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.7877,
+ "step": 6054
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0002,
+ "loss": 0.4692,
+ "step": 6055
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8353,
+ "step": 6056
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6864,
+ "step": 6057
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7034,
+ "step": 6058
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7006,
+ "step": 6059
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7265,
+ "step": 6060
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 1.0102,
+ "step": 6061
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6727,
+ "step": 6062
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.686,
+ "step": 6063
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.452,
+ "step": 6064
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 1.0397,
+ "step": 6065
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7481,
+ "step": 6066
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8179,
+ "step": 6067
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 1.1399,
+ "step": 6068
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7763,
+ "step": 6069
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8474,
+ "step": 6070
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7565,
+ "step": 6071
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7641,
+ "step": 6072
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7109,
+ "step": 6073
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7553,
+ "step": 6074
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.714,
+ "step": 6075
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6999,
+ "step": 6076
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6577,
+ "step": 6077
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 1.0083,
+ "step": 6078
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7476,
+ "step": 6079
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7481,
+ "step": 6080
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7459,
+ "step": 6081
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8133,
+ "step": 6082
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8167,
+ "step": 6083
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.868,
+ "step": 6084
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8323,
+ "step": 6085
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6757,
+ "step": 6086
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.717,
+ "step": 6087
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5235,
+ "step": 6088
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4337,
+ "step": 6089
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5811,
+ "step": 6090
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6134,
+ "step": 6091
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5729,
+ "step": 6092
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5863,
+ "step": 6093
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4666,
+ "step": 6094
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.621,
+ "step": 6095
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4437,
+ "step": 6096
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4508,
+ "step": 6097
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4723,
+ "step": 6098
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5512,
+ "step": 6099
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.3863,
+ "step": 6100
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6035,
+ "step": 6101
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4541,
+ "step": 6102
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4495,
+ "step": 6103
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7307,
+ "step": 6104
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7699,
+ "step": 6105
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5502,
+ "step": 6106
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5688,
+ "step": 6107
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.3051,
+ "step": 6108
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.8163,
+ "step": 6109
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7434,
+ "step": 6110
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.6556,
+ "step": 6111
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5602,
+ "step": 6112
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.4549,
+ "step": 6113
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.5315,
+ "step": 6114
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.7938,
+ "step": 6115
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.0002,
+ "loss": 0.9281,
+ "step": 6116
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.2253,
+ "step": 6117
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4907,
+ "step": 6118
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.605,
+ "step": 6119
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7445,
+ "step": 6120
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7316,
+ "step": 6121
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4795,
+ "step": 6122
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7582,
+ "step": 6123
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6188,
+ "step": 6124
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5787,
+ "step": 6125
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.3778,
+ "step": 6126
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.456,
+ "step": 6127
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5953,
+ "step": 6128
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.9266,
+ "step": 6129
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6317,
+ "step": 6130
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5758,
+ "step": 6131
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6978,
+ "step": 6132
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4863,
+ "step": 6133
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.538,
+ "step": 6134
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6262,
+ "step": 6135
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5846,
+ "step": 6136
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6534,
+ "step": 6137
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.3126,
+ "step": 6138
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.62,
+ "step": 6139
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5577,
+ "step": 6140
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6544,
+ "step": 6141
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.3914,
+ "step": 6142
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5244,
+ "step": 6143
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5589,
+ "step": 6144
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6551,
+ "step": 6145
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8224,
+ "step": 6146
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6967,
+ "step": 6147
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6565,
+ "step": 6148
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4591,
+ "step": 6149
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.454,
+ "step": 6150
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5841,
+ "step": 6151
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6853,
+ "step": 6152
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5635,
+ "step": 6153
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4419,
+ "step": 6154
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.3326,
+ "step": 6155
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5264,
+ "step": 6156
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6229,
+ "step": 6157
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6868,
+ "step": 6158
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.8161,
+ "step": 6159
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5327,
+ "step": 6160
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.7164,
+ "step": 6161
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.663,
+ "step": 6162
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6278,
+ "step": 6163
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5661,
+ "step": 6164
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5846,
+ "step": 6165
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6532,
+ "step": 6166
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6723,
+ "step": 6167
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6245,
+ "step": 6168
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.563,
+ "step": 6169
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4213,
+ "step": 6170
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5742,
+ "step": 6171
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.6358,
+ "step": 6172
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4959,
+ "step": 6173
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.4134,
+ "step": 6174
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.374,
+ "step": 6175
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5806,
+ "step": 6176
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.0002,
+ "loss": 0.5818,
+ "step": 6177
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8435,
+ "step": 6178
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7231,
+ "step": 6179
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7051,
+ "step": 6180
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6058,
+ "step": 6181
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8654,
+ "step": 6182
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5256,
+ "step": 6183
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5972,
+ "step": 6184
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7521,
+ "step": 6185
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5981,
+ "step": 6186
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.4882,
+ "step": 6187
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5822,
+ "step": 6188
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5135,
+ "step": 6189
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5801,
+ "step": 6190
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6471,
+ "step": 6191
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5207,
+ "step": 6192
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5081,
+ "step": 6193
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.4595,
+ "step": 6194
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.8701,
+ "step": 6195
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5111,
+ "step": 6196
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5928,
+ "step": 6197
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5746,
+ "step": 6198
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7244,
+ "step": 6199
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6446,
+ "step": 6200
+ },
+ {
+ "epoch": 1.02,
+ "eval_loss": 0.7689220905303955,
+ "eval_runtime": 131.671,
+ "eval_samples_per_second": 7.595,
+ "eval_steps_per_second": 3.797,
+ "step": 6200
+ },
+ {
+ "epoch": 1.02,
+ "mmlu_eval_accuracy": 0.5715176271997892,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+ "mmlu_eval_accuracy_astronomy": 0.625,
+ "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
+ "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
+ "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": 0.5909090909090909,
+ "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
+ "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
+ "mmlu_eval_accuracy_electrical_engineering": 0.5625,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
+ "mmlu_eval_accuracy_formal_logic": 0.42857142857142855,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.59375,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.7692307692307693,
+ "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
+ "mmlu_eval_accuracy_high_school_psychology": 0.85,
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
+ "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
+ "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
+ "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
+ "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": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.84,
+ "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
+ "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
+ "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
+ "mmlu_eval_accuracy_moral_scenarios": 0.29,
+ "mmlu_eval_accuracy_nutrition": 0.696969696969697,
+ "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
+ "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
+ "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
+ "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
+ "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
+ "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+ "mmlu_eval_accuracy_sociology": 0.9090909090909091,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
+ "mmlu_loss": 0.9558499650567692,
+ "step": 6200
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6062,
+ "step": 6201
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6032,
+ "step": 6202
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.561,
+ "step": 6203
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.4079,
+ "step": 6204
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6275,
+ "step": 6205
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.3475,
+ "step": 6206
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6517,
+ "step": 6207
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.645,
+ "step": 6208
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5111,
+ "step": 6209
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6819,
+ "step": 6210
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.3636,
+ "step": 6211
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6966,
+ "step": 6212
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5142,
+ "step": 6213
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6658,
+ "step": 6214
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.687,
+ "step": 6215
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7124,
+ "step": 6216
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7188,
+ "step": 6217
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7206,
+ "step": 6218
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5702,
+ "step": 6219
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6039,
+ "step": 6220
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6068,
+ "step": 6221
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7902,
+ "step": 6222
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7177,
+ "step": 6223
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7458,
+ "step": 6224
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.3964,
+ "step": 6225
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6472,
+ "step": 6226
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.4947,
+ "step": 6227
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6436,
+ "step": 6228
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5875,
+ "step": 6229
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5151,
+ "step": 6230
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5099,
+ "step": 6231
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6439,
+ "step": 6232
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7228,
+ "step": 6233
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6005,
+ "step": 6234
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6718,
+ "step": 6235
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7025,
+ "step": 6236
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.6174,
+ "step": 6237
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0002,
+ "loss": 0.5144,
+ "step": 6238
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6794,
+ "step": 6239
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6652,
+ "step": 6240
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5218,
+ "step": 6241
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6885,
+ "step": 6242
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4799,
+ "step": 6243
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5289,
+ "step": 6244
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7574,
+ "step": 6245
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.3729,
+ "step": 6246
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5993,
+ "step": 6247
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8746,
+ "step": 6248
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5014,
+ "step": 6249
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6427,
+ "step": 6250
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5425,
+ "step": 6251
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.549,
+ "step": 6252
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5774,
+ "step": 6253
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.503,
+ "step": 6254
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6014,
+ "step": 6255
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4089,
+ "step": 6256
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5722,
+ "step": 6257
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5259,
+ "step": 6258
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.48,
+ "step": 6259
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4591,
+ "step": 6260
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5597,
+ "step": 6261
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7679,
+ "step": 6262
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4424,
+ "step": 6263
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6155,
+ "step": 6264
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6558,
+ "step": 6265
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5359,
+ "step": 6266
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.416,
+ "step": 6267
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6122,
+ "step": 6268
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6334,
+ "step": 6269
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.466,
+ "step": 6270
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8003,
+ "step": 6271
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6888,
+ "step": 6272
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6223,
+ "step": 6273
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6497,
+ "step": 6274
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.489,
+ "step": 6275
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6443,
+ "step": 6276
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5104,
+ "step": 6277
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5377,
+ "step": 6278
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6392,
+ "step": 6279
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4018,
+ "step": 6280
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6095,
+ "step": 6281
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.3712,
+ "step": 6282
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4678,
+ "step": 6283
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5635,
+ "step": 6284
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.419,
+ "step": 6285
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5868,
+ "step": 6286
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6329,
+ "step": 6287
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5484,
+ "step": 6288
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6816,
+ "step": 6289
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.4478,
+ "step": 6290
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7833,
+ "step": 6291
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.8275,
+ "step": 6292
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7119,
+ "step": 6293
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6181,
+ "step": 6294
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5911,
+ "step": 6295
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.695,
+ "step": 6296
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.614,
+ "step": 6297
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.5145,
+ "step": 6298
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0002,
+ "loss": 0.6438,
+ "step": 6299
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.58,
+ "step": 6300
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5168,
+ "step": 6301
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5215,
+ "step": 6302
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5577,
+ "step": 6303
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.485,
+ "step": 6304
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6327,
+ "step": 6305
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7714,
+ "step": 6306
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9221,
+ "step": 6307
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4674,
+ "step": 6308
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5249,
+ "step": 6309
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8127,
+ "step": 6310
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7444,
+ "step": 6311
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5232,
+ "step": 6312
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6849,
+ "step": 6313
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4721,
+ "step": 6314
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5491,
+ "step": 6315
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7283,
+ "step": 6316
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.534,
+ "step": 6317
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8665,
+ "step": 6318
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5043,
+ "step": 6319
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4038,
+ "step": 6320
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5021,
+ "step": 6321
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7291,
+ "step": 6322
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6413,
+ "step": 6323
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6706,
+ "step": 6324
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 1.0441,
+ "step": 6325
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6939,
+ "step": 6326
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7659,
+ "step": 6327
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.9259,
+ "step": 6328
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6901,
+ "step": 6329
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5091,
+ "step": 6330
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7004,
+ "step": 6331
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7819,
+ "step": 6332
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.415,
+ "step": 6333
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7947,
+ "step": 6334
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5127,
+ "step": 6335
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5655,
+ "step": 6336
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5329,
+ "step": 6337
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6419,
+ "step": 6338
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.8009,
+ "step": 6339
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.365,
+ "step": 6340
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4507,
+ "step": 6341
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6468,
+ "step": 6342
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6173,
+ "step": 6343
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5255,
+ "step": 6344
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6876,
+ "step": 6345
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.7042,
+ "step": 6346
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6395,
+ "step": 6347
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.3292,
+ "step": 6348
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4478,
+ "step": 6349
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.585,
+ "step": 6350
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5233,
+ "step": 6351
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.2907,
+ "step": 6352
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.3312,
+ "step": 6353
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.6319,
+ "step": 6354
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4441,
+ "step": 6355
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4045,
+ "step": 6356
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 1.0694,
+ "step": 6357
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5241,
+ "step": 6358
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.4754,
+ "step": 6359
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.0002,
+ "loss": 0.5606,
+ "step": 6360
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7378,
+ "step": 6361
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7608,
+ "step": 6362
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5168,
+ "step": 6363
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5711,
+ "step": 6364
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6788,
+ "step": 6365
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.4122,
+ "step": 6366
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.635,
+ "step": 6367
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6023,
+ "step": 6368
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.4666,
+ "step": 6369
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7066,
+ "step": 6370
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5153,
+ "step": 6371
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5298,
+ "step": 6372
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6439,
+ "step": 6373
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.462,
+ "step": 6374
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5451,
+ "step": 6375
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5628,
+ "step": 6376
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5817,
+ "step": 6377
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7119,
+ "step": 6378
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6162,
+ "step": 6379
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5984,
+ "step": 6380
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.54,
+ "step": 6381
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7415,
+ "step": 6382
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.2781,
+ "step": 6383
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7376,
+ "step": 6384
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.7709,
+ "step": 6385
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.925,
+ "step": 6386
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5392,
+ "step": 6387
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9883,
+ "step": 6388
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6426,
+ "step": 6389
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.545,
+ "step": 6390
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.4511,
+ "step": 6391
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5265,
+ "step": 6392
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.68,
+ "step": 6393
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.4912,
+ "step": 6394
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6679,
+ "step": 6395
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.4239,
+ "step": 6396
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5228,
+ "step": 6397
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.5934,
+ "step": 6398
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.9043,
+ "step": 6399
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6576,
+ "step": 6400
+ },
+ {
+ "epoch": 1.05,
+ "eval_loss": 0.7675976753234863,
+ "eval_runtime": 131.7379,
+ "eval_samples_per_second": 7.591,
+ "eval_steps_per_second": 3.795,
+ "step": 6400
+ },
+ {
+ "epoch": 1.05,
+ "mmlu_eval_accuracy": 0.531372965510407,
+ "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+ "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
+ "mmlu_eval_accuracy_astronomy": 0.4375,
+ "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
+ "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": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
+ "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
+ "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
+ "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
+ "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
+ "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
+ "mmlu_eval_accuracy_electrical_engineering": 0.3125,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
+ "mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
+ "mmlu_eval_accuracy_global_facts": 0.3,
+ "mmlu_eval_accuracy_high_school_biology": 0.65625,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
+ "mmlu_eval_accuracy_high_school_physics": 0.4117647058823529,
+ "mmlu_eval_accuracy_high_school_psychology": 0.8,
+ "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
+ "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
+ "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
+ "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
+ "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": 0.7222222222222222,
+ "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
+ "mmlu_eval_accuracy_management": 0.8181818181818182,
+ "mmlu_eval_accuracy_marketing": 0.88,
+ "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+ "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
+ "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
+ "mmlu_eval_accuracy_moral_scenarios": 0.14,
+ "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
+ "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
+ "mmlu_eval_accuracy_prehistory": 0.7142857142857143,
+ "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
+ "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
+ "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
+ "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
+ "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
+ "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+ "mmlu_eval_accuracy_sociology": 0.8181818181818182,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
+ "mmlu_eval_accuracy_virology": 0.5555555555555556,
+ "mmlu_eval_accuracy_world_religions": 0.631578947368421,
+ "mmlu_loss": 0.9138225094262054,
+ "step": 6400
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 18258,
+ "num_train_epochs": 3,
+ "save_steps": 200,
+ "total_flos": 1.3171875913686712e+18,
+ "trial_name": null,
+ "trial_params": null
+}