| { |
| "best_metric": 0.731837809085846, |
| "best_model_checkpoint": "experts/mistralic-expert-21/checkpoint-1800", |
| "epoch": 0.3592814371257485, |
| "eval_steps": 200, |
| "global_step": 1800, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0613, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9108, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7569, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.787, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7733, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.3886, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4824, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8347, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.82, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8169, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5864, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8223, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7887, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7621, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7538, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6615, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9518, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5117, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0139, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7997, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7555, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6843, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7045, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0227, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6964, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8441, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7352, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5966, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8846, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7978, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1034, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8306, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0758, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7684, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8209, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.62, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.572, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6433, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7005, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.627, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0076, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6516, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7212, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7396, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.741, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9025, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8793, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6438, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7211, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7618, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8042, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7492, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.949, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8977, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8895, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8984, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7287, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5764, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1263, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8061, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4888, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7801, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9469, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6781, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0843, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6992, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8471, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7336, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1684, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5612, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9468, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.895, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9233, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.073, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.972, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6457, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6522, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6546, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0024, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0696, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6181, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0234, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8514, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7032, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0316, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8664, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.469, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8426, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5955, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7823, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.778, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5699, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.949, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7923, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9486, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7278, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8223, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5869, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.466, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7447, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6255, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6209, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0633, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9947, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7474, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8626, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8522, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6075, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.912, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.012, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7406, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6897, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.543, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4828, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9805, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7141, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9519, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8587, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7717, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6015, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7259, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0637, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7039, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1593, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9464, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5241, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8832, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7735, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9548, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6403, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9322, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7969, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.917, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6374, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9057, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5936, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5973, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6922, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.885, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6669, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7749, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8642, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7119, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4539, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6143, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9801, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8143, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9165, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8678, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9648, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7353, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7707, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0535, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.74, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6083, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6862, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6716, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6058, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7035, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7014, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6037, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8274, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6332, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7386, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8272, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6769, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8564, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.794, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6748, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6602, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8841, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.748, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0154, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.721, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9784, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5253, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8122, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6862, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8924, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0512, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8278, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6857, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6204, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7821, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9541, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8025, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0036, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9583, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5459, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8891, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6371, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7171, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8905, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5407, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8385, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8422, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8084, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6382, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.04, |
| "eval_loss": 0.7527853846549988, |
| "eval_runtime": 193.5321, |
| "eval_samples_per_second": 5.167, |
| "eval_steps_per_second": 2.584, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.04, |
| "mmlu_eval_accuracy": 0.6069425201408959, |
| "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "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": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.75, |
| "mmlu_eval_accuracy_electrical_engineering": 0.6875, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "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": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.96, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.36, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_law": 0.4117647058823529, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "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": 0.8947368421052632, |
| "mmlu_loss": 1.424748288122232, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.514, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9316, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5804, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4997, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1044, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8998, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9126, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7343, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8555, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8367, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9339, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5325, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.97, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7172, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7829, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9416, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9772, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9562, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9923, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.773, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8065, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4085, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7975, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7796, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9226, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9331, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8537, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5119, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9528, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8085, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6984, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7458, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5282, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7395, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7395, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7806, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0737, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8094, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8435, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6897, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9283, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.77, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7093, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6659, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6623, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2126, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8251, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6103, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5445, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6698, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6505, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6185, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8441, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7706, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7232, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.566, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7725, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.772, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8432, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7987, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5326, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6139, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8207, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5063, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5334, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9008, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6599, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8471, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6595, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5456, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0143, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.67, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6222, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7805, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6144, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7708, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7156, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7694, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6985, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.874, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5317, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6899, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6082, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8266, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7347, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.751, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0287, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8968, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6152, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6274, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8135, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9844, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.524, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9002, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8671, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.571, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6869, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7714, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8017, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7137, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0807, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0078, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1694, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7987, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6085, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5241, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5681, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5722, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6979, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6865, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9059, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8185, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6021, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7516, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.911, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7015, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.597, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0015, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7167, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9415, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8917, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7481, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5737, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9759, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9918, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9649, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7749, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9503, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7141, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6378, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7751, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.645, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9118, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7924, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6403, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8784, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8649, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8398, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5552, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5564, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9732, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1613, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7281, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6125, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8193, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8056, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7315, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8247, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8393, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7128, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9818, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8143, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6425, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7803, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.742, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4946, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8821, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8141, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9041, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1578, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5657, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7776, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7091, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6348, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6316, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6251, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0801, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7038, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5184, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7923, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8136, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9586, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9144, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.599, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7147, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7103, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6476, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8461, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0367, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7077, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8001, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.2422, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8415, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.736, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8057, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6561, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8917, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6604, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6119, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7716, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7671, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6078, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9074, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7811, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9108, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7685, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0709, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_loss": 0.744175374507904, |
| "eval_runtime": 193.6137, |
| "eval_samples_per_second": 5.165, |
| "eval_steps_per_second": 2.582, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.08, |
| "mmlu_eval_accuracy": 0.6043681736715669, |
| "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "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": 0.5833333333333334, |
| "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": 0.46875, |
| "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": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 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": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.32, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "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": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.4941282213979228, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8267, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7679, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8803, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6909, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7483, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0523, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9839, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5975, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0406, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9047, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7649, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7999, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7923, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7002, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7059, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6865, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7177, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8824, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.787, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8307, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7682, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8036, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9209, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9188, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7904, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9804, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6341, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8869, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8143, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8974, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6402, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5342, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7838, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8436, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.024, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8074, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6338, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9353, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6655, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.661, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7146, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8318, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5375, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8082, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6488, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8286, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5802, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7761, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7238, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9706, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5635, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7103, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4757, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5925, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6109, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.622, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7854, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5802, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7563, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7902, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8256, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.948, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.745, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6632, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6803, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8442, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6172, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5671, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7855, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6973, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5741, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7252, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7218, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5713, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8696, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6925, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.0487, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7919, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9096, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8272, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8518, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6736, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.802, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8978, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.657, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5339, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7309, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6462, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5552, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.1669, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7967, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7578, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8775, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7491, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9756, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8121, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7392, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9004, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7345, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7513, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7502, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8346, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9917, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8684, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9052, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6764, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7646, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4939, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.0915, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4308, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7663, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6548, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6723, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.0556, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7296, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9113, |
| "step": 517 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9695, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7099, |
| "step": 519 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5564, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8164, |
| "step": 521 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5783, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7821, |
| "step": 523 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6573, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8268, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6128, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6761, |
| "step": 527 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7277, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5248, |
| "step": 529 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9179, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 531 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6512, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7375, |
| "step": 533 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9032, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7618, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7835, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6359, |
| "step": 537 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.052, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5695, |
| "step": 539 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8816, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6503, |
| "step": 541 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6225, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5026, |
| "step": 543 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8693, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7503, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7084, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9936, |
| "step": 547 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8152, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9392, |
| "step": 549 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.0249, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9395, |
| "step": 551 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4423, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9268, |
| "step": 553 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9497, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.468, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8133, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6607, |
| "step": 557 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8391, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8714, |
| "step": 559 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6308, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6394, |
| "step": 561 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.913, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.688, |
| "step": 563 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8688, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7826, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5856, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.0415, |
| "step": 567 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9789, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9218, |
| "step": 569 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7543, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.53, |
| "step": 571 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.1324, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7376, |
| "step": 573 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9111, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7983, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7959, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8756, |
| "step": 577 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0287, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4868, |
| "step": 579 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6186, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6415, |
| "step": 581 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8003, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8524, |
| "step": 583 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8741, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6316, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6371, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7191, |
| "step": 587 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7531, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6027, |
| "step": 589 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7606, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.938, |
| "step": 591 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7439, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9444, |
| "step": 593 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6378, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9355, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7833, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7256, |
| "step": 597 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6942, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7742, |
| "step": 599 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.684, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.12, |
| "eval_loss": 0.7413733601570129, |
| "eval_runtime": 193.4196, |
| "eval_samples_per_second": 5.17, |
| "eval_steps_per_second": 2.585, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.12, |
| "mmlu_eval_accuracy": 0.6066353154142871, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "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": 0.5769230769230769, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.33, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "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": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.546568229210906, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4465, |
| "step": 601 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8645, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6272, |
| "step": 603 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9447, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.769, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7628, |
| "step": 607 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6576, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6482, |
| "step": 609 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7616, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 611 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.585, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5834, |
| "step": 613 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7417, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8483, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5884, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8695, |
| "step": 617 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6002, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.758, |
| "step": 619 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0847, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7203, |
| "step": 621 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0074, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7413, |
| "step": 623 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7296, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5597, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.685, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6455, |
| "step": 627 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6522, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6444, |
| "step": 629 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4978, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8012, |
| "step": 631 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4906, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4867, |
| "step": 633 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.726, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3853, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6966, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7108, |
| "step": 637 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.667, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7569, |
| "step": 639 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9096, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0672, |
| "step": 641 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7803, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7002, |
| "step": 643 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8482, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8336, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8791, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9492, |
| "step": 647 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8084, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6602, |
| "step": 649 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6844, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.1414, |
| "step": 651 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6237, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9086, |
| "step": 653 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9113, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4281, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8587, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0444, |
| "step": 657 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7863, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8533, |
| "step": 659 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5408, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7821, |
| "step": 661 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7603, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.662, |
| "step": 663 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8689, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7313, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8914, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9887, |
| "step": 667 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7603, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7592, |
| "step": 669 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0364, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.735, |
| "step": 671 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0776, |
| "step": 673 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.775, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8048, |
| "step": 677 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.645, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8758, |
| "step": 679 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9819, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9876, |
| "step": 681 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.994, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6316, |
| "step": 683 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6357, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5663, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5252, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8512, |
| "step": 687 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6856, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6659, |
| "step": 689 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9902, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 1.0727, |
| "step": 691 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6238, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5903, |
| "step": 693 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6347, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7957, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 1.018, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4683, |
| "step": 697 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8655, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 1.0705, |
| "step": 699 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7668, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6937, |
| "step": 701 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6379, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9275, |
| "step": 703 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.772, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6673, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8489, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4449, |
| "step": 707 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7635, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6486, |
| "step": 709 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6111, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8594, |
| "step": 711 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5079, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6897, |
| "step": 713 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.869, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9091, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7522, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8457, |
| "step": 717 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6871, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7114, |
| "step": 719 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6767, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9387, |
| "step": 721 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6753, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7687, |
| "step": 723 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6395, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7196, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 1.0891, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6333, |
| "step": 727 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7339, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6865, |
| "step": 729 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7704, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6839, |
| "step": 731 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8703, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7393, |
| "step": 733 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9276, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5298, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8824, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9587, |
| "step": 737 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5957, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.788, |
| "step": 739 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6561, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6559, |
| "step": 741 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6629, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.438, |
| "step": 743 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7781, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.77, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6313, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7117, |
| "step": 747 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7473, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6311, |
| "step": 749 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7513, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.674, |
| "step": 751 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5662, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.543, |
| "step": 753 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8509, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7214, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8063, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8514, |
| "step": 757 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5505, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7168, |
| "step": 759 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7901, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6215, |
| "step": 761 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9305, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7226, |
| "step": 763 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7223, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9405, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.0112, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7048, |
| "step": 767 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6901, |
| "step": 769 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7391, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6669, |
| "step": 771 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5642, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.1581, |
| "step": 773 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.737, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8622, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8514, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9713, |
| "step": 777 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9228, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.526, |
| "step": 779 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6225, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5743, |
| "step": 781 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5652, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4765, |
| "step": 783 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9622, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6645, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0493, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6555, |
| "step": 787 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9545, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6747, |
| "step": 789 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7934, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7103, |
| "step": 791 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5921, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7873, |
| "step": 793 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9653, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7854, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7418, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 797 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.836, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8236, |
| "step": 799 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6746, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_loss": 0.7383494973182678, |
| "eval_runtime": 193.4476, |
| "eval_samples_per_second": 5.169, |
| "eval_steps_per_second": 2.585, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.16, |
| "mmlu_eval_accuracy": 0.5824527930989486, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.625, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.6875, |
| "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": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, |
| "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.6875, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.6956521739130435, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.7272727272727273, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.8, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "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": 0.6176470588235294, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.38235294117647056, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7368421052631579, |
| "mmlu_loss": 1.5123703732269551, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.1325, |
| "step": 801 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0307, |
| "step": 802 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7898, |
| "step": 803 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7005, |
| "step": 804 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7593, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6625, |
| "step": 806 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7347, |
| "step": 807 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8784, |
| "step": 808 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4695, |
| "step": 809 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8117, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0179, |
| "step": 811 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3943, |
| "step": 812 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8658, |
| "step": 813 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.568, |
| "step": 814 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5006, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9943, |
| "step": 816 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5892, |
| "step": 817 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9133, |
| "step": 818 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8451, |
| "step": 819 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9722, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7591, |
| "step": 821 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6626, |
| "step": 822 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8656, |
| "step": 823 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7894, |
| "step": 824 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7845, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4706, |
| "step": 826 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.677, |
| "step": 827 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8349, |
| "step": 828 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9249, |
| "step": 829 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8127, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7091, |
| "step": 831 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.945, |
| "step": 832 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6807, |
| "step": 833 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8597, |
| "step": 834 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5821, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5261, |
| "step": 836 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7211, |
| "step": 837 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.531, |
| "step": 838 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7866, |
| "step": 839 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5671, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8422, |
| "step": 841 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6337, |
| "step": 842 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5958, |
| "step": 843 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 844 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8512, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6628, |
| "step": 846 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5743, |
| "step": 847 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.785, |
| "step": 848 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6344, |
| "step": 849 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5008, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8397, |
| "step": 851 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6754, |
| "step": 852 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7784, |
| "step": 853 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0332, |
| "step": 854 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7469, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8307, |
| "step": 856 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6985, |
| "step": 857 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7129, |
| "step": 858 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8013, |
| "step": 859 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5928, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7336, |
| "step": 861 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6602, |
| "step": 862 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7297, |
| "step": 863 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7492, |
| "step": 864 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7373, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8825, |
| "step": 866 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5926, |
| "step": 867 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8167, |
| "step": 868 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0685, |
| "step": 869 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6117, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.759, |
| "step": 871 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.602, |
| "step": 872 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8258, |
| "step": 873 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7127, |
| "step": 874 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5475, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8619, |
| "step": 876 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9897, |
| "step": 877 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8684, |
| "step": 878 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7956, |
| "step": 879 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9684, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7395, |
| "step": 881 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7849, |
| "step": 882 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8232, |
| "step": 883 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8302, |
| "step": 884 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9753, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5334, |
| "step": 886 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6797, |
| "step": 887 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7426, |
| "step": 888 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6695, |
| "step": 889 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6054, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7343, |
| "step": 891 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7936, |
| "step": 892 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8432, |
| "step": 893 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5937, |
| "step": 894 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7726, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7908, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7519, |
| "step": 897 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9247, |
| "step": 898 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6882, |
| "step": 899 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.837, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.712, |
| "step": 901 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5879, |
| "step": 902 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.896, |
| "step": 903 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6178, |
| "step": 904 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7161, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4227, |
| "step": 906 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7676, |
| "step": 907 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.69, |
| "step": 908 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6964, |
| "step": 909 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7591, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8413, |
| "step": 911 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8537, |
| "step": 912 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9511, |
| "step": 913 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6141, |
| "step": 914 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5237, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7318, |
| "step": 916 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6696, |
| "step": 917 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8418, |
| "step": 918 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.1654, |
| "step": 919 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7256, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.628, |
| "step": 921 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6198, |
| "step": 922 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8068, |
| "step": 923 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7606, |
| "step": 924 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7729, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.568, |
| "step": 926 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7844, |
| "step": 927 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8251, |
| "step": 928 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7107, |
| "step": 929 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7679, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0475, |
| "step": 931 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9008, |
| "step": 932 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9582, |
| "step": 933 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7771, |
| "step": 934 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9076, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7497, |
| "step": 936 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8194, |
| "step": 937 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9736, |
| "step": 938 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9531, |
| "step": 939 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5868, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7418, |
| "step": 941 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6473, |
| "step": 942 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5416, |
| "step": 943 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.835, |
| "step": 944 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9608, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6923, |
| "step": 946 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9096, |
| "step": 947 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5797, |
| "step": 948 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5413, |
| "step": 949 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9418, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8327, |
| "step": 951 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9431, |
| "step": 952 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5363, |
| "step": 953 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.501, |
| "step": 954 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9756, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7025, |
| "step": 956 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7918, |
| "step": 957 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7463, |
| "step": 958 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8343, |
| "step": 959 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8494, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8324, |
| "step": 961 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7082, |
| "step": 962 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7782, |
| "step": 963 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7745, |
| "step": 964 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9588, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6768, |
| "step": 966 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6051, |
| "step": 967 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6357, |
| "step": 968 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7516, |
| "step": 969 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5479, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6951, |
| "step": 971 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5802, |
| "step": 972 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8874, |
| "step": 973 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.888, |
| "step": 974 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.83, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6681, |
| "step": 976 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.573, |
| "step": 977 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6659, |
| "step": 978 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.1737, |
| "step": 979 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7718, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8225, |
| "step": 981 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7975, |
| "step": 982 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9496, |
| "step": 983 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9455, |
| "step": 984 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6933, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5385, |
| "step": 986 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7877, |
| "step": 987 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7844, |
| "step": 988 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4767, |
| "step": 989 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.649, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6118, |
| "step": 991 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7695, |
| "step": 992 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8787, |
| "step": 993 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7278, |
| "step": 994 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6649, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8817, |
| "step": 996 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9241, |
| "step": 997 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.0226, |
| "step": 998 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6312, |
| "step": 999 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.0413, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.2, |
| "eval_loss": 0.7385837435722351, |
| "eval_runtime": 193.3745, |
| "eval_samples_per_second": 5.171, |
| "eval_steps_per_second": 2.586, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.2, |
| "mmlu_eval_accuracy": 0.5902575064856496, |
| "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": 0.6896551724137931, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "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": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.5365853658536586, |
| "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, |
| "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": 0.6666666666666666, |
| "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": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.9, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.7272727272727273, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.6363636363636364, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "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": 0.31, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.6857142857142857, |
| "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, |
| "mmlu_eval_accuracy_professional_law": 0.38235294117647056, |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "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": 0.7894736842105263, |
| "mmlu_loss": 1.3963911063366086, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8493, |
| "step": 1001 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5994, |
| "step": 1002 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7318, |
| "step": 1003 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5181, |
| "step": 1004 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9893, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8404, |
| "step": 1006 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6006, |
| "step": 1007 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6347, |
| "step": 1008 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6482, |
| "step": 1009 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4894, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6805, |
| "step": 1011 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.743, |
| "step": 1012 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6163, |
| "step": 1013 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6122, |
| "step": 1014 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8754, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7474, |
| "step": 1016 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.068, |
| "step": 1017 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7107, |
| "step": 1018 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7884, |
| "step": 1019 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7634, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6653, |
| "step": 1021 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8652, |
| "step": 1022 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7117, |
| "step": 1023 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8491, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5426, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7367, |
| "step": 1026 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5975, |
| "step": 1027 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6938, |
| "step": 1028 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8798, |
| "step": 1029 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6943, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8041, |
| "step": 1031 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9096, |
| "step": 1032 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6555, |
| "step": 1033 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5689, |
| "step": 1034 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5898, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5444, |
| "step": 1036 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6294, |
| "step": 1037 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.814, |
| "step": 1038 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8768, |
| "step": 1039 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7185, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9066, |
| "step": 1041 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7584, |
| "step": 1042 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7111, |
| "step": 1043 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8036, |
| "step": 1044 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7198, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5807, |
| "step": 1046 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6591, |
| "step": 1047 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0319, |
| "step": 1048 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6876, |
| "step": 1049 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6837, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5331, |
| "step": 1051 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9072, |
| "step": 1052 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4844, |
| "step": 1053 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6604, |
| "step": 1054 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6808, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7692, |
| "step": 1056 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7271, |
| "step": 1057 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6018, |
| "step": 1058 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6975, |
| "step": 1059 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.737, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6803, |
| "step": 1061 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9012, |
| "step": 1062 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8243, |
| "step": 1063 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6736, |
| "step": 1064 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9098, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0275, |
| "step": 1066 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7305, |
| "step": 1067 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.927, |
| "step": 1068 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3883, |
| "step": 1069 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7116, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0463, |
| "step": 1071 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9751, |
| "step": 1072 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6838, |
| "step": 1073 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9358, |
| "step": 1074 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7106, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.1675, |
| "step": 1076 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.2714, |
| "step": 1077 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5542, |
| "step": 1078 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7961, |
| "step": 1079 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5015, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.691, |
| "step": 1081 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8486, |
| "step": 1082 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.967, |
| "step": 1083 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7693, |
| "step": 1084 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6872, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7146, |
| "step": 1086 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.741, |
| "step": 1087 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8334, |
| "step": 1088 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8693, |
| "step": 1089 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8379, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.707, |
| "step": 1091 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6885, |
| "step": 1092 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7412, |
| "step": 1093 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.0916, |
| "step": 1094 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7675, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6496, |
| "step": 1096 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8712, |
| "step": 1097 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7878, |
| "step": 1098 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9333, |
| "step": 1099 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7686, |
| "step": 1101 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9195, |
| "step": 1102 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8343, |
| "step": 1103 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7939, |
| "step": 1104 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7631, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9208, |
| "step": 1106 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5177, |
| "step": 1107 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8252, |
| "step": 1108 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7257, |
| "step": 1109 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5251, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7283, |
| "step": 1111 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7158, |
| "step": 1112 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5401, |
| "step": 1113 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6453, |
| "step": 1114 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.0929, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9226, |
| "step": 1116 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9246, |
| "step": 1117 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7314, |
| "step": 1118 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9544, |
| "step": 1119 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7326, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9662, |
| "step": 1121 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4423, |
| "step": 1122 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5184, |
| "step": 1123 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9587, |
| "step": 1124 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9054, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6653, |
| "step": 1126 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9978, |
| "step": 1127 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5819, |
| "step": 1128 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7805, |
| "step": 1129 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6538, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8143, |
| "step": 1131 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6377, |
| "step": 1132 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8567, |
| "step": 1133 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6649, |
| "step": 1134 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6762, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6681, |
| "step": 1136 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6237, |
| "step": 1137 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7035, |
| "step": 1138 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5982, |
| "step": 1139 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9579, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.602, |
| "step": 1141 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8965, |
| "step": 1142 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8469, |
| "step": 1143 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9311, |
| "step": 1144 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8779, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7089, |
| "step": 1146 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8793, |
| "step": 1147 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6426, |
| "step": 1148 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7687, |
| "step": 1149 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9376, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8068, |
| "step": 1151 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7344, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5577, |
| "step": 1153 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7571, |
| "step": 1154 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6942, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6628, |
| "step": 1156 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6766, |
| "step": 1157 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7471, |
| "step": 1158 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6496, |
| "step": 1159 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9536, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6883, |
| "step": 1161 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8751, |
| "step": 1162 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.994, |
| "step": 1163 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8015, |
| "step": 1164 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.1034, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9908, |
| "step": 1166 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6977, |
| "step": 1167 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.62, |
| "step": 1168 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 1169 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8775, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8037, |
| "step": 1171 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7462, |
| "step": 1172 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8134, |
| "step": 1173 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9558, |
| "step": 1174 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8952, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9713, |
| "step": 1176 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5715, |
| "step": 1177 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.815, |
| "step": 1178 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.825, |
| "step": 1179 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6262, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8169, |
| "step": 1181 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7321, |
| "step": 1182 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6139, |
| "step": 1183 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8301, |
| "step": 1184 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8124, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5715, |
| "step": 1186 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3773, |
| "step": 1187 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.74, |
| "step": 1188 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4801, |
| "step": 1189 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.66, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7036, |
| "step": 1191 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9719, |
| "step": 1192 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6145, |
| "step": 1193 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8648, |
| "step": 1194 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5101, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8194, |
| "step": 1196 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7549, |
| "step": 1197 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0408, |
| "step": 1198 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0106, |
| "step": 1199 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8807, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_loss": 0.7385936379432678, |
| "eval_runtime": 193.3658, |
| "eval_samples_per_second": 5.172, |
| "eval_steps_per_second": 2.586, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.24, |
| "mmlu_eval_accuracy": 0.5850567848032601, |
| "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.625, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, |
| "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": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.5365853658536586, |
| "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": 0.45454545454545453, |
| "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": 0.7619047619047619, |
| "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": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.7272727272727273, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, |
| "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": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.31, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.6470588235294118, |
| "mmlu_eval_accuracy_prehistory": 0.6285714285714286, |
| "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903, |
| "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": 0.5555555555555556, |
| "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": 0.7894736842105263, |
| "mmlu_loss": 1.2981049245315806, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7728, |
| "step": 1201 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6242, |
| "step": 1202 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7127, |
| "step": 1203 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.866, |
| "step": 1204 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.663, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7179, |
| "step": 1206 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7653, |
| "step": 1207 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9871, |
| "step": 1208 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9166, |
| "step": 1209 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.1658, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4651, |
| "step": 1211 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6344, |
| "step": 1212 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8869, |
| "step": 1213 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6786, |
| "step": 1214 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0322, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0817, |
| "step": 1216 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6136, |
| "step": 1217 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9182, |
| "step": 1218 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.609, |
| "step": 1219 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7963, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7568, |
| "step": 1221 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5355, |
| "step": 1222 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.607, |
| "step": 1223 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7562, |
| "step": 1224 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0653, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0218, |
| "step": 1226 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.2943, |
| "step": 1227 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7194, |
| "step": 1228 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8448, |
| "step": 1229 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6793, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5851, |
| "step": 1231 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5922, |
| "step": 1232 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7201, |
| "step": 1233 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5675, |
| "step": 1234 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7308, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6672, |
| "step": 1236 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6274, |
| "step": 1237 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0234, |
| "step": 1238 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8175, |
| "step": 1239 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7139, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6531, |
| "step": 1241 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4537, |
| "step": 1242 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6386, |
| "step": 1243 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6976, |
| "step": 1244 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8005, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.64, |
| "step": 1246 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.964, |
| "step": 1247 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7017, |
| "step": 1248 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7415, |
| "step": 1249 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7627, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9413, |
| "step": 1251 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.602, |
| "step": 1252 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.63, |
| "step": 1253 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6862, |
| "step": 1254 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7735, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8266, |
| "step": 1256 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.1041, |
| "step": 1257 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9596, |
| "step": 1258 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0606, |
| "step": 1259 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9781, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7358, |
| "step": 1261 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7159, |
| "step": 1262 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8248, |
| "step": 1263 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8255, |
| "step": 1264 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.777, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7657, |
| "step": 1266 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8742, |
| "step": 1267 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7048, |
| "step": 1268 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0851, |
| "step": 1269 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8084, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7662, |
| "step": 1271 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7076, |
| "step": 1272 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8632, |
| "step": 1273 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6383, |
| "step": 1274 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8594, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7838, |
| "step": 1276 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8323, |
| "step": 1277 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8696, |
| "step": 1278 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0206, |
| "step": 1279 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6374, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.482, |
| "step": 1281 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.666, |
| "step": 1282 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9236, |
| "step": 1283 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7483, |
| "step": 1284 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7885, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6145, |
| "step": 1286 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9731, |
| "step": 1287 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.5695, |
| "step": 1288 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8955, |
| "step": 1289 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9962, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8836, |
| "step": 1291 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7248, |
| "step": 1292 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.2259, |
| "step": 1293 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.497, |
| "step": 1294 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8431, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6612, |
| "step": 1296 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0634, |
| "step": 1297 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6303, |
| "step": 1298 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.5267, |
| "step": 1299 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7871, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7684, |
| "step": 1301 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8465, |
| "step": 1302 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9231, |
| "step": 1303 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8286, |
| "step": 1304 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9982, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6057, |
| "step": 1306 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9049, |
| "step": 1307 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8037, |
| "step": 1308 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0362, |
| "step": 1309 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8726, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0052, |
| "step": 1311 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6261, |
| "step": 1312 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.782, |
| "step": 1313 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7096, |
| "step": 1314 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8602, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.603, |
| "step": 1316 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8997, |
| "step": 1317 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4747, |
| "step": 1318 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4972, |
| "step": 1319 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4048, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.5627, |
| "step": 1321 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8923, |
| "step": 1322 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7151, |
| "step": 1323 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6126, |
| "step": 1324 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6479, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.5613, |
| "step": 1326 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9719, |
| "step": 1327 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6337, |
| "step": 1328 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7027, |
| "step": 1329 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.9991, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8137, |
| "step": 1331 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6427, |
| "step": 1332 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.791, |
| "step": 1333 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6689, |
| "step": 1334 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7175, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8023, |
| "step": 1336 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4474, |
| "step": 1337 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.781, |
| "step": 1338 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8098, |
| "step": 1339 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.755, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5796, |
| "step": 1341 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5143, |
| "step": 1342 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.117, |
| "step": 1343 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8023, |
| "step": 1344 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.0466, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7681, |
| "step": 1346 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6037, |
| "step": 1347 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7295, |
| "step": 1348 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6693, |
| "step": 1349 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6673, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6244, |
| "step": 1351 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4588, |
| "step": 1352 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.9678, |
| "step": 1353 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7069, |
| "step": 1354 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.699, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.728, |
| "step": 1356 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6508, |
| "step": 1357 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6429, |
| "step": 1358 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8893, |
| "step": 1359 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7653, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.0214, |
| "step": 1361 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5662, |
| "step": 1362 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7259, |
| "step": 1363 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6182, |
| "step": 1364 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5431, |
| "step": 1366 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.778, |
| "step": 1367 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6397, |
| "step": 1368 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.1543, |
| "step": 1369 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7573, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7691, |
| "step": 1371 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6315, |
| "step": 1372 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.0004, |
| "step": 1373 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6203, |
| "step": 1374 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8698, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6049, |
| "step": 1376 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8734, |
| "step": 1377 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.2827, |
| "step": 1378 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8979, |
| "step": 1379 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6003, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6169, |
| "step": 1381 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.1044, |
| "step": 1382 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5581, |
| "step": 1383 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8054, |
| "step": 1384 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6763, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6498, |
| "step": 1386 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5826, |
| "step": 1387 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0307, |
| "step": 1388 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6372, |
| "step": 1389 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6526, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0349, |
| "step": 1391 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5989, |
| "step": 1392 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6707, |
| "step": 1393 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5712, |
| "step": 1394 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8944, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6856, |
| "step": 1396 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.731, |
| "step": 1397 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6569, |
| "step": 1398 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7735, |
| "step": 1399 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.641, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.28, |
| "eval_loss": 0.737493634223938, |
| "eval_runtime": 193.3981, |
| "eval_samples_per_second": 5.171, |
| "eval_steps_per_second": 2.585, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.28, |
| "mmlu_eval_accuracy": 0.5963366012888651, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, |
| "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": 0.45454545454545453, |
| "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": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.7272727272727273, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.37, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "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": 0.7894736842105263, |
| "mmlu_loss": 1.3717845350348297, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8039, |
| "step": 1401 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8371, |
| "step": 1402 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6907, |
| "step": 1403 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8172, |
| "step": 1404 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.612, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8182, |
| "step": 1406 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8467, |
| "step": 1407 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9056, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8473, |
| "step": 1409 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5926, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9594, |
| "step": 1411 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7882, |
| "step": 1412 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6995, |
| "step": 1413 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7341, |
| "step": 1414 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8583, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6016, |
| "step": 1416 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7078, |
| "step": 1417 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9061, |
| "step": 1418 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6072, |
| "step": 1419 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.804, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4595, |
| "step": 1421 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8988, |
| "step": 1422 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6457, |
| "step": 1423 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.327, |
| "step": 1424 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6058, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.1151, |
| "step": 1426 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5914, |
| "step": 1427 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7973, |
| "step": 1428 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8986, |
| "step": 1429 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9922, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5859, |
| "step": 1431 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.0249, |
| "step": 1432 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7148, |
| "step": 1433 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8038, |
| "step": 1434 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9982, |
| "step": 1435 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5926, |
| "step": 1436 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.63, |
| "step": 1437 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8477, |
| "step": 1438 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5802, |
| "step": 1439 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8729, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8008, |
| "step": 1441 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.3002, |
| "step": 1442 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6955, |
| "step": 1443 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7948, |
| "step": 1444 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6707, |
| "step": 1445 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7164, |
| "step": 1446 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6419, |
| "step": 1447 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7299, |
| "step": 1448 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9397, |
| "step": 1449 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4473, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6968, |
| "step": 1451 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5945, |
| "step": 1452 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.823, |
| "step": 1453 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5784, |
| "step": 1454 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8504, |
| "step": 1455 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9299, |
| "step": 1456 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.641, |
| "step": 1457 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7642, |
| "step": 1458 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6284, |
| "step": 1459 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.0075, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8246, |
| "step": 1461 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7743, |
| "step": 1462 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5919, |
| "step": 1463 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.645, |
| "step": 1464 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7434, |
| "step": 1465 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8994, |
| "step": 1466 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9009, |
| "step": 1467 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.1806, |
| "step": 1468 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7441, |
| "step": 1469 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7162, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.0153, |
| "step": 1471 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8883, |
| "step": 1472 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8876, |
| "step": 1473 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7395, |
| "step": 1474 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7949, |
| "step": 1475 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7775, |
| "step": 1476 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7387, |
| "step": 1477 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5592, |
| "step": 1478 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7048, |
| "step": 1479 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7409, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.876, |
| "step": 1481 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7626, |
| "step": 1482 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5746, |
| "step": 1483 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0608, |
| "step": 1484 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6692, |
| "step": 1485 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7718, |
| "step": 1486 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8303, |
| "step": 1487 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8256, |
| "step": 1488 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8302, |
| "step": 1489 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6504, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7018, |
| "step": 1491 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8037, |
| "step": 1492 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8941, |
| "step": 1493 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.749, |
| "step": 1494 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9105, |
| "step": 1495 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7904, |
| "step": 1496 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7485, |
| "step": 1497 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7993, |
| "step": 1498 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6333, |
| "step": 1499 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7026, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6228, |
| "step": 1501 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8741, |
| "step": 1502 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7702, |
| "step": 1503 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.769, |
| "step": 1504 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8118, |
| "step": 1505 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9701, |
| "step": 1506 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8376, |
| "step": 1507 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9531, |
| "step": 1508 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7551, |
| "step": 1509 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9609, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.611, |
| "step": 1511 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 1512 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8539, |
| "step": 1513 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7061, |
| "step": 1514 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7443, |
| "step": 1515 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8781, |
| "step": 1516 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6154, |
| "step": 1517 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8681, |
| "step": 1518 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6472, |
| "step": 1519 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5393, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.51, |
| "step": 1521 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5832, |
| "step": 1522 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0378, |
| "step": 1523 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.894, |
| "step": 1524 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8839, |
| "step": 1525 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6072, |
| "step": 1526 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6752, |
| "step": 1527 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6889, |
| "step": 1528 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7994, |
| "step": 1529 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7532, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7144, |
| "step": 1531 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6113, |
| "step": 1532 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8841, |
| "step": 1533 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7422, |
| "step": 1534 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5074, |
| "step": 1535 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5108, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9971, |
| "step": 1537 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9342, |
| "step": 1538 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7075, |
| "step": 1539 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6289, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6066, |
| "step": 1541 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.726, |
| "step": 1542 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5277, |
| "step": 1543 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.0274, |
| "step": 1544 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9127, |
| "step": 1545 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6878, |
| "step": 1546 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8866, |
| "step": 1547 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7599, |
| "step": 1548 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7409, |
| "step": 1549 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9534, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9316, |
| "step": 1551 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5377, |
| "step": 1552 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7862, |
| "step": 1553 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8891, |
| "step": 1554 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7137, |
| "step": 1555 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9805, |
| "step": 1556 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5044, |
| "step": 1557 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6009, |
| "step": 1558 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7629, |
| "step": 1559 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7634, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7604, |
| "step": 1561 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5707, |
| "step": 1562 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8544, |
| "step": 1563 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8021, |
| "step": 1564 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5862, |
| "step": 1565 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.722, |
| "step": 1566 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7889, |
| "step": 1567 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7455, |
| "step": 1568 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9378, |
| "step": 1569 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5532, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9331, |
| "step": 1571 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.666, |
| "step": 1572 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7396, |
| "step": 1573 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8025, |
| "step": 1574 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6942, |
| "step": 1575 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.67, |
| "step": 1576 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6749, |
| "step": 1577 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7419, |
| "step": 1578 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9045, |
| "step": 1579 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8431, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6031, |
| "step": 1581 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7918, |
| "step": 1582 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.59, |
| "step": 1583 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8004, |
| "step": 1584 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8104, |
| "step": 1585 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8527, |
| "step": 1586 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5993, |
| "step": 1587 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5955, |
| "step": 1588 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8226, |
| "step": 1589 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7617, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7278, |
| "step": 1591 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.1376, |
| "step": 1592 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6112, |
| "step": 1593 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.607, |
| "step": 1594 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9778, |
| "step": 1595 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6059, |
| "step": 1596 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7678, |
| "step": 1597 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8135, |
| "step": 1598 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6198, |
| "step": 1599 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6704, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_loss": 0.7339836359024048, |
| "eval_runtime": 193.3683, |
| "eval_samples_per_second": 5.171, |
| "eval_steps_per_second": 2.586, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.32, |
| "mmlu_eval_accuracy": 0.5962788695449543, |
| "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "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": 0.6818181818181818, |
| "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": 0.75, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "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": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, |
| "mmlu_eval_accuracy_high_school_psychology": 0.9, |
| "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": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.7272727272727273, |
| "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": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.36, |
| "mmlu_eval_accuracy_nutrition": 0.6060606060606061, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.6285714285714286, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.40588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.3888888888888889, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.411965688705133, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7315, |
| "step": 1601 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8499, |
| "step": 1602 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4946, |
| "step": 1603 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8385, |
| "step": 1604 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8326, |
| "step": 1605 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7248, |
| "step": 1606 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9317, |
| "step": 1607 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6783, |
| "step": 1608 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8575, |
| "step": 1609 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.614, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.705, |
| "step": 1611 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6059, |
| "step": 1612 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5645, |
| "step": 1613 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.757, |
| "step": 1614 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.018, |
| "step": 1615 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6826, |
| "step": 1616 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7791, |
| "step": 1617 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8688, |
| "step": 1618 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5269, |
| "step": 1619 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6839, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6163, |
| "step": 1621 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5816, |
| "step": 1622 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6612, |
| "step": 1623 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5393, |
| "step": 1624 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5663, |
| "step": 1625 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4988, |
| "step": 1626 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.589, |
| "step": 1627 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.728, |
| "step": 1628 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7513, |
| "step": 1629 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6099, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6395, |
| "step": 1631 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7323, |
| "step": 1632 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7442, |
| "step": 1633 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9926, |
| "step": 1634 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7851, |
| "step": 1635 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7614, |
| "step": 1636 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 1637 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.717, |
| "step": 1638 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6278, |
| "step": 1639 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6587, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7604, |
| "step": 1641 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7759, |
| "step": 1642 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.779, |
| "step": 1643 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9031, |
| "step": 1644 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7952, |
| "step": 1645 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6909, |
| "step": 1646 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6161, |
| "step": 1647 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7425, |
| "step": 1648 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.762, |
| "step": 1649 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.688, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6995, |
| "step": 1651 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6851, |
| "step": 1652 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6166, |
| "step": 1653 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.887, |
| "step": 1654 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.775, |
| "step": 1655 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6838, |
| "step": 1656 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8018, |
| "step": 1657 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6719, |
| "step": 1658 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5903, |
| "step": 1659 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7068, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 1.1788, |
| "step": 1661 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8269, |
| "step": 1662 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7854, |
| "step": 1663 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 1664 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8708, |
| "step": 1665 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7566, |
| "step": 1666 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6498, |
| "step": 1667 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6413, |
| "step": 1668 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9154, |
| "step": 1669 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5866, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4664, |
| "step": 1671 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7339, |
| "step": 1672 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5437, |
| "step": 1673 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8272, |
| "step": 1674 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6237, |
| "step": 1675 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8042, |
| "step": 1676 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6711, |
| "step": 1677 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5174, |
| "step": 1678 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6051, |
| "step": 1679 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.0672, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6812, |
| "step": 1681 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6988, |
| "step": 1682 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6626, |
| "step": 1683 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7337, |
| "step": 1684 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6562, |
| "step": 1685 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8744, |
| "step": 1686 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 1687 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8222, |
| "step": 1688 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7897, |
| "step": 1689 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8273, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9049, |
| "step": 1691 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5091, |
| "step": 1692 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9177, |
| "step": 1693 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.538, |
| "step": 1694 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6983, |
| "step": 1695 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9841, |
| "step": 1696 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.865, |
| "step": 1697 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8893, |
| "step": 1698 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6181, |
| "step": 1699 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6567, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7502, |
| "step": 1701 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6971, |
| "step": 1702 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7403, |
| "step": 1703 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.483, |
| "step": 1704 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5365, |
| "step": 1705 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6331, |
| "step": 1706 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9463, |
| "step": 1707 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7957, |
| "step": 1708 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5585, |
| "step": 1709 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9387, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8924, |
| "step": 1711 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7219, |
| "step": 1712 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.0721, |
| "step": 1713 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6027, |
| "step": 1714 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6024, |
| "step": 1715 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.0104, |
| "step": 1716 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7259, |
| "step": 1717 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7208, |
| "step": 1718 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8806, |
| "step": 1719 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7487, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8921, |
| "step": 1721 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7401, |
| "step": 1722 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.769, |
| "step": 1723 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7566, |
| "step": 1724 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 1725 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7079, |
| "step": 1726 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8469, |
| "step": 1727 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8595, |
| "step": 1728 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.632, |
| "step": 1729 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6818, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9119, |
| "step": 1731 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.575, |
| "step": 1732 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.649, |
| "step": 1733 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5105, |
| "step": 1734 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7621, |
| "step": 1735 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6576, |
| "step": 1736 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8492, |
| "step": 1737 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7724, |
| "step": 1738 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5132, |
| "step": 1739 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7034, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8683, |
| "step": 1741 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9222, |
| "step": 1742 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8111, |
| "step": 1743 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0202, |
| "step": 1744 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8754, |
| "step": 1745 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7585, |
| "step": 1746 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7622, |
| "step": 1747 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5034, |
| "step": 1748 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6989, |
| "step": 1749 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6681, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4827, |
| "step": 1751 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0283, |
| "step": 1752 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7714, |
| "step": 1753 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5526, |
| "step": 1754 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7419, |
| "step": 1755 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5953, |
| "step": 1756 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.892, |
| "step": 1757 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7527, |
| "step": 1758 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7673, |
| "step": 1759 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9816, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7577, |
| "step": 1761 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8456, |
| "step": 1762 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0854, |
| "step": 1763 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7411, |
| "step": 1764 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4924, |
| "step": 1765 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6813, |
| "step": 1766 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4921, |
| "step": 1767 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7531, |
| "step": 1768 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8107, |
| "step": 1769 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8341, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7441, |
| "step": 1771 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7778, |
| "step": 1772 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6772, |
| "step": 1773 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7325, |
| "step": 1774 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5785, |
| "step": 1775 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4894, |
| "step": 1776 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.64, |
| "step": 1777 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0618, |
| "step": 1778 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5968, |
| "step": 1779 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8235, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6847, |
| "step": 1781 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8263, |
| "step": 1782 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8101, |
| "step": 1783 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7963, |
| "step": 1784 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8573, |
| "step": 1785 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5593, |
| "step": 1786 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6083, |
| "step": 1787 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6859, |
| "step": 1788 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7312, |
| "step": 1789 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7605, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5251, |
| "step": 1791 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7297, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7584, |
| "step": 1793 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7508, |
| "step": 1794 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9912, |
| "step": 1795 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9799, |
| "step": 1796 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6311, |
| "step": 1797 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9597, |
| "step": 1798 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7767, |
| "step": 1799 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7836, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.36, |
| "eval_loss": 0.731837809085846, |
| "eval_runtime": 193.5482, |
| "eval_samples_per_second": 5.167, |
| "eval_steps_per_second": 2.583, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.36, |
| "mmlu_eval_accuracy": 0.5880298500679698, |
| "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": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "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": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_physics": 0.6363636363636364, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.7272727272727273, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.39, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.6857142857142857, |
| "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.5208221201252377, |
| "step": 1800 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 15030, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 5.259847139198239e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
