| { |
| "best_metric": 0.7704223394393921, |
| "best_model_checkpoint": "experts/mistralic-expert-13/checkpoint-400", |
| "epoch": 0.04203557260331556, |
| "eval_steps": 200, |
| "global_step": 400, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8057, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.868, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7422, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9211, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7284, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.128, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.691, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8822, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0556, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8944, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.3661, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.897, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4765, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8214, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6931, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8858, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7504, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9926, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8128, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0525, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7398, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7148, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.939, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8571, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8606, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6652, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.636, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8285, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7824, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7264, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9278, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0789, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7172, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8376, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8398, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8799, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7641, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6588, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.788, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7721, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6668, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8948, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7881, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6409, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.918, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.678, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8724, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5262, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6769, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.13, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7797, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9556, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7761, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6964, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7111, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2248, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7785, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7058, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0892, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5716, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6917, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9176, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5918, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9956, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1706, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8256, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7374, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0579, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7109, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0454, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9837, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7759, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6114, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.823, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6405, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9208, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9584, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8973, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0944, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6444, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6373, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8495, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8911, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7041, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9224, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0205, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1457, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.666, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7664, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4892, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6736, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8725, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8566, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6853, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5988, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6631, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8843, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.935, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9667, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6212, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9027, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7214, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7017, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7641, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6351, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.611, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7059, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8062, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7346, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.849, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6919, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0735, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0983, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7274, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7297, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4824, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0998, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9032, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8483, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5231, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.769, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6166, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.846, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8177, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9799, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.603, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7934, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6732, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.702, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6871, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8845, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.5723, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7545, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0616, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5688, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.873, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6499, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8858, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.627, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7416, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9393, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.767, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.879, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.641, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8841, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8503, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6859, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8448, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7325, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0813, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.722, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6801, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8911, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7792, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.946, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9398, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9148, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1395, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8569, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6701, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9436, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6362, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6857, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6296, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7952, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7484, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7465, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.4395, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7578, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.756, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9464, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8769, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7327, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0913, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8841, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5796, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9103, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6798, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0787, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8528, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9995, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5743, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.634, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.988, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8765, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9337, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4822, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7905, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1404, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6976, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8448, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8432, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5687, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6528, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2231, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6245, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9841, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8432, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7284, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5404, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.02, |
| "eval_loss": 0.7798383831977844, |
| "eval_runtime": 95.1866, |
| "eval_samples_per_second": 10.506, |
| "eval_steps_per_second": 5.253, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.02, |
| "mmlu_eval_accuracy": 0.5850749757313948, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "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": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.34, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_law": 0.40588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.5018870820420218, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7717, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.075, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8029, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6166, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.825, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5856, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8154, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5654, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5801, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6079, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6345, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6687, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9763, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9686, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.784, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8225, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1642, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8505, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6015, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9394, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6616, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8977, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5748, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0238, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8827, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8886, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.025, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7792, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7394, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8316, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6724, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7914, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6508, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7213, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9859, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7088, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8148, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5307, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8188, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6935, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5691, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6292, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7924, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8754, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7818, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6114, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7362, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9104, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8642, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.2486, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9316, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6619, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.725, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7513, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8932, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8265, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7006, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6295, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9926, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.737, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8356, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0866, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8264, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.757, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7694, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5351, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6854, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0508, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7364, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8047, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0033, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9712, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6332, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8437, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6464, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9993, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1226, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7422, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.824, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5478, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8713, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8185, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6583, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5111, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7658, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7131, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8029, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7282, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6571, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.3087, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8032, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7082, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7862, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7951, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8892, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6942, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8256, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6694, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.115, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7531, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8585, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7966, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8774, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7866, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6955, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0246, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7458, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7726, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7765, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8112, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7909, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7809, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6419, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8863, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8054, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7239, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7755, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5018, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9658, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5482, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7833, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7887, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7808, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7392, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.745, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0452, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6139, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6727, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6865, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0485, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7331, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2392, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6594, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.33, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6953, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8768, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5251, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.972, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8908, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5679, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5835, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7334, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.681, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7237, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7765, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7141, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7206, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4814, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7291, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3692, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7458, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0571, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.445, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9983, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9078, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5956, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8008, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9969, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.538, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8287, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.834, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6707, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1536, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9761, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6926, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8788, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8978, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6775, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9067, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8387, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7436, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6387, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7309, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.778, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7269, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7226, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8201, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6562, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6942, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8172, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8761, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5423, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6533, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7778, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.7521, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8417, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8692, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8148, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.6309, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6056, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2633, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6098, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5883, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5178, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.645, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.837, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6274, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.04, |
| "eval_loss": 0.7704223394393921, |
| "eval_runtime": 95.1262, |
| "eval_samples_per_second": 10.512, |
| "eval_steps_per_second": 5.256, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.04, |
| "mmlu_eval_accuracy": 0.5772222074989346, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.36363636363636365, |
| "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": 0.35714285714285715, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.6521739130434783, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.3, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "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": 0.7894736842105263, |
| "mmlu_loss": 1.4183595089081373, |
| "step": 400 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 28545, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 4.789781317248614e+16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
