| { | |
| "best_metric": 0.8657823204994202, | |
| "best_model_checkpoint": "experts/mistralic-expert-25/checkpoint-200", | |
| "epoch": 0.05823475887170155, | |
| "eval_steps": 200, | |
| "global_step": 400, | |
| "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": 0.6586, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7382, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9771, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.239, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1367, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9121, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8723, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7729, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9516, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7457, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9671, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8585, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8002, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0357, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6908, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9481, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0791, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9394, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8856, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7797, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1097, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8909, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0102, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4585, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8925, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9697, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0254, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0075, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8735, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0522, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.973, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.559, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4682, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0415, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6446, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8066, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5727, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0717, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9121, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8141, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6281, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0932, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.971, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9128, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8409, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0842, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8916, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7288, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.889, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9508, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.854, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6206, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8621, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4238, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2063, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7552, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8874, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8812, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7006, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8047, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1526, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5679, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.794, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9858, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9189, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0108, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1895, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5368, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7866, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.868, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9624, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.656, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7712, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8823, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6565, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0115, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.2936, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8947, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6353, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5608, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0677, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8707, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0677, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7594, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6097, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1323, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6166, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9671, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9475, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5055, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9368, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0028, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7944, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9518, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6992, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7948, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.5601, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8346, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9327, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0734, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7085, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8345, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8653, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0132, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6714, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1224, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9094, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7111, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8389, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.994, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7135, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6824, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8392, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6321, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5622, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0738, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9969, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1434, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7824, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6799, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6665, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2051, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7486, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8522, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.07, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9415, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.622, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1949, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9893, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6507, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6626, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6973, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0162, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8232, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0537, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9964, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0375, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9369, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7409, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7771, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7179, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0834, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6802, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0296, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9016, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6938, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.739, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7792, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3007, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3336, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8723, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1111, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1674, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8059, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7778, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7531, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6723, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8483, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0657, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1322, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9167, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9756, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7663, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.135, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7567, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0317, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1242, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.977, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7498, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8805, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7917, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5351, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.869, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2608, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8666, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7938, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8882, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0029, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7313, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8234, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.882, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8096, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8343, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6169, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0987, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8363, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0218, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.838, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9084, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5031, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8404, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8532, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9322, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0271, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9383, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8817, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8641, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6634, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8597, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.765, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_loss": 0.8657823204994202, | |
| "eval_runtime": 150.7329, | |
| "eval_samples_per_second": 6.634, | |
| "eval_steps_per_second": 3.317, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "mmlu_eval_accuracy": 0.5909590703251556, | |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, | |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, | |
| "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": 0.625, | |
| "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": 0.6363636363636364, | |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, | |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, | |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, | |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, | |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, | |
| "mmlu_eval_accuracy_global_facts": 0.3, | |
| "mmlu_eval_accuracy_high_school_biology": 0.5625, | |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, | |
| "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": 0.7142857142857143, | |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, | |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, | |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, | |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, | |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, | |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, | |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, | |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, | |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, | |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, | |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, | |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, | |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, | |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, | |
| "mmlu_eval_accuracy_management": 0.9090909090909091, | |
| "mmlu_eval_accuracy_marketing": 0.88, | |
| "mmlu_eval_accuracy_medical_genetics": 1.0, | |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, | |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, | |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, | |
| "mmlu_eval_accuracy_nutrition": 0.6363636363636364, | |
| "mmlu_eval_accuracy_philosophy": 0.7647058823529411, | |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, | |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, | |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, | |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, | |
| "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594, | |
| "mmlu_eval_accuracy_public_relations": 0.5, | |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, | |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, | |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, | |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, | |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, | |
| "mmlu_loss": 1.5003863334188885, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8749, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8911, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8477, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.59, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.68, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1289, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6094, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7295, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9877, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4836, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.902, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0939, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8077, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1104, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0361, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7082, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8439, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.76, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9476, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9157, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.116, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8784, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0402, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0433, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7033, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5797, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8211, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9549, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1082, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7167, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.72, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3027, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9015, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8538, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1021, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8727, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.14, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0598, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9005, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7569, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7344, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5442, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5014, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0667, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6199, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8805, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.943, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7799, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6903, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.078, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8822, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8223, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7609, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.66, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8807, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9438, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7693, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8977, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0147, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7311, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1428, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9863, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1905, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0178, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0601, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8216, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9855, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9618, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9469, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1452, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6944, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0098, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7943, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6379, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0303, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0032, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0306, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8565, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.264, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8809, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9519, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6438, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.734, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7444, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6762, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6035, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9632, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.78, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0164, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1476, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8958, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1306, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1924, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5887, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7278, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1009, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9431, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0868, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.962, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.944, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1359, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1752, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.2978, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1673, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6636, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6604, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9764, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9834, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6069, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8305, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8291, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4372, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6138, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4996, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9616, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1558, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5072, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8821, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9128, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.134, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.842, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0481, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0161, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0036, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6041, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1767, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6586, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6043, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7591, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7701, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8548, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9349, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7302, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.295, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7391, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5129, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.837, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0422, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9438, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7376, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5119, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6394, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1076, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0006, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7598, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8456, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4541, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7358, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8682, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9197, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7392, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.96, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8633, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7089, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7632, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8844, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1651, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8405, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9419, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9078, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5853, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3986, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8585, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6812, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9836, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8018, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7313, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.261, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9301, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7423, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4561, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8535, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7937, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0679, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6891, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9629, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5769, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2939, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4939, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4871, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9562, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0184, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6472, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.91, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8464, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4652, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9297, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1218, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7709, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7838, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0776, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9132, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 1.028, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1235, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8227, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0908, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5499, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9925, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.9207, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8515, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 0.876522958278656, | |
| "eval_runtime": 150.8131, | |
| "eval_samples_per_second": 6.631, | |
| "eval_steps_per_second": 3.315, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "mmlu_eval_accuracy": 0.5887418611751553, | |
| "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": 0.6206896551724138, | |
| "mmlu_eval_accuracy_college_biology": 0.625, | |
| "mmlu_eval_accuracy_college_chemistry": 0.25, | |
| "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, | |
| "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": 0.7272727272727273, | |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, | |
| "mmlu_eval_accuracy_econometrics": 0.5, | |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, | |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, | |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, | |
| "mmlu_eval_accuracy_global_facts": 0.3, | |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, | |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, | |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, | |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, | |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, | |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, | |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6744186046511628, | |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, | |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, | |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, | |
| "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": 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": 0.9090909090909091, | |
| "mmlu_eval_accuracy_marketing": 0.92, | |
| "mmlu_eval_accuracy_medical_genetics": 1.0, | |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, | |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, | |
| "mmlu_eval_accuracy_moral_scenarios": 0.29, | |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, | |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, | |
| "mmlu_eval_accuracy_prehistory": 0.4857142857142857, | |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, | |
| "mmlu_eval_accuracy_professional_law": 0.4, | |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, | |
| "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, | |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, | |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, | |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, | |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, | |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, | |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, | |
| "mmlu_loss": 1.1819022774851975, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20604, | |
| "num_train_epochs": 3, | |
| "save_steps": 200, | |
| "total_flos": 9.398696198956646e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |