| { |
| "best_metric": 0.8146001100540161, |
| "best_model_checkpoint": "experts/mistralic-expert-22/checkpoint-1800", |
| "epoch": 0.38496703719743997, |
| "eval_steps": 200, |
| "global_step": 2000, |
| "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": 0.9917, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8752, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8807, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7598, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6723, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.1704, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7503, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.624, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8103, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9605, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.905, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.759, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8095, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.1747, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8053, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6902, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.878, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.89, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.899, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0926, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9986, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8805, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6184, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9538, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0293, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4661, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8878, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8821, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.85, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9784, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7107, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7971, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8528, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8603, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7917, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6604, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8287, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1047, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6991, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9996, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8583, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7114, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7781, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9923, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8158, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7188, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8207, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7159, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8143, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9955, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7208, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9416, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7941, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7234, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0252, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7091, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7707, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.157, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5682, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.807, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6721, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8324, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8405, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7994, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7385, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4077, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8142, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6694, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8789, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.058, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1531, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9339, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7771, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0121, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7184, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9318, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.105, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6653, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.631, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9929, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.03, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7572, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8269, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6807, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9773, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.866, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9673, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8735, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8992, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8388, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8848, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6855, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7848, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1573, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8753, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.016, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8481, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8461, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8009, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8147, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.818, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7188, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8583, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8898, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5784, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6973, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8995, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8895, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7302, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.975, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7004, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6645, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7651, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8179, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8355, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8815, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8535, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7703, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8666, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9532, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7817, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0935, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7434, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8263, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9015, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.5507, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.923, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0015, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8959, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.996, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9214, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6441, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8035, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8738, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6934, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7723, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8101, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6368, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7078, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8225, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7866, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8628, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6387, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7479, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8173, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8456, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7523, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6802, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7845, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.622, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.842, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0277, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0264, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9195, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.797, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.87, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.6444, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8486, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9643, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6787, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7781, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6973, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7924, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9079, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7007, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8613, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.608, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9465, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9435, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7896, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6892, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8605, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7067, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6283, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9131, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0426, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7652, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0208, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7484, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8419, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9176, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.861, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.907, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.651, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6038, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7511, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6037, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.592, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5935, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9313, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8181, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.792, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7445, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6853, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9783, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9599, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9033, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.04, |
| "eval_loss": 0.8272572159767151, |
| "eval_runtime": 108.0162, |
| "eval_samples_per_second": 9.258, |
| "eval_steps_per_second": 4.629, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.04, |
| "mmlu_eval_accuracy": 0.599579385714013, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "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": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.33, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.1056992180898357, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8147, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2314, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9407, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7344, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9563, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9926, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8277, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7435, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9858, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8615, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7798, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8055, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8161, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8922, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7699, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9098, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8848, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7967, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7214, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6714, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8672, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7304, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7675, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5663, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9466, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0255, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9056, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8921, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7836, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8398, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8849, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8133, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9293, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6496, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5932, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9838, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8546, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9101, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.859, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9602, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9327, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5277, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.748, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7524, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9312, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2737, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0512, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7334, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7493, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5577, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0016, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9124, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8657, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9144, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9532, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7411, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7931, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6414, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7083, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9409, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.856, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9955, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5929, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7547, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6542, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8503, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8429, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.857, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8845, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8411, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9812, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8003, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7384, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7263, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.78, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9858, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7142, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7677, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6429, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8833, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8403, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5925, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7843, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8162, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7978, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7364, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8184, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8204, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9493, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8844, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7213, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7367, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8065, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7912, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9889, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6346, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8347, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.821, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0193, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7745, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5443, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0296, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8065, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7023, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7963, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6676, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7119, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8342, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8657, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9189, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9195, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.737, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6512, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9037, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.942, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9569, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0143, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0598, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8547, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6792, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9757, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7605, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7029, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9174, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.908, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0953, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6913, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7406, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6988, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4839, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9061, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7339, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7077, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0788, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7634, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.652, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.802, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7895, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8487, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7397, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.857, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9782, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.688, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5103, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.762, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9286, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7606, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8981, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7572, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7552, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6557, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7197, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6654, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8249, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9703, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9126, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6395, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8437, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8885, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9231, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5622, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9886, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6807, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7713, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8205, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6163, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6523, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7218, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7817, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1076, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9519, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9615, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8734, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7385, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6975, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6935, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6487, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8057, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6614, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6929, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9957, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7739, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.953, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5405, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8276, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6967, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7781, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8823, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5715, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9113, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8301, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9687, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.76, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7402, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7616, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_loss": 0.8319438099861145, |
| "eval_runtime": 107.9019, |
| "eval_samples_per_second": 9.268, |
| "eval_steps_per_second": 4.634, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.08, |
| "mmlu_eval_accuracy": 0.5644849203853226, |
| "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.36363636363636365, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464, |
| "mmlu_eval_accuracy_econometrics": 0.3333333333333333, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.5, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, |
| "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, |
| "mmlu_eval_accuracy_moral_scenarios": 0.33, |
| "mmlu_eval_accuracy_nutrition": 0.7878787878787878, |
| "mmlu_eval_accuracy_philosophy": 0.6176470588235294, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.37058823529411766, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7368421052631579, |
| "mmlu_loss": 1.3124866759886007, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8433, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9205, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7859, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8488, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6307, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5087, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8378, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8898, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8309, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8289, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8561, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8791, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6545, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9316, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7037, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6658, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9517, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8115, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8659, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6427, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6879, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0384, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5888, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9127, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7456, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8999, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9658, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9099, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8155, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5489, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7532, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7318, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8388, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7363, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8658, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.87, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8169, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9461, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.777, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8922, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6376, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7908, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.845, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.1059, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9725, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.001, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9184, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7949, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6397, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8519, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7851, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7957, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.2149, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9572, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9525, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0605, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7367, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8119, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6568, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7564, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7836, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8655, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9538, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7256, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6022, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5525, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8683, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5313, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7129, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9313, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9505, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8025, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8977, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8055, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9525, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0739, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0571, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8615, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8694, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8516, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0282, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9444, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6297, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9982, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.814, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8239, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8397, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7726, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8604, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9932, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9045, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9513, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9425, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6765, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8696, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8834, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8216, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8824, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7342, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8486, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.1737, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7487, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5858, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8642, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8384, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9313, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8897, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7311, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7789, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8578, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.0066, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6273, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6804, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7417, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8504, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.856, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8277, |
| "step": 517 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7622, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6808, |
| "step": 519 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9331, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7157, |
| "step": 521 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6574, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9497, |
| "step": 523 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7128, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7914, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7403, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8889, |
| "step": 527 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8651, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6389, |
| "step": 529 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9403, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7894, |
| "step": 531 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.882, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8738, |
| "step": 533 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.1446, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7493, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8229, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8648, |
| "step": 537 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9932, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8452, |
| "step": 539 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7869, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.1482, |
| "step": 541 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7878, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.788, |
| "step": 543 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6392, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9142, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5735, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7491, |
| "step": 547 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8111, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.1825, |
| "step": 549 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7141, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7931, |
| "step": 551 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7072, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6284, |
| "step": 553 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7792, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8095, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7571, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7008, |
| "step": 557 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.0028, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.84, |
| "step": 559 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8219, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6631, |
| "step": 561 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4663, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8435, |
| "step": 563 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9562, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6672, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6709, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8419, |
| "step": 567 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7651, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.699, |
| "step": 569 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9775, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8303, |
| "step": 571 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7964, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6591, |
| "step": 573 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8658, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9896, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4332, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8514, |
| "step": 577 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8869, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8641, |
| "step": 579 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8047, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6639, |
| "step": 581 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.0241, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8246, |
| "step": 583 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8675, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8013, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7683, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6556, |
| "step": 587 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.77, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7938, |
| "step": 589 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6935, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9622, |
| "step": 591 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.87, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7516, |
| "step": 593 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8894, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7505, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9013, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7466, |
| "step": 597 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0704, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7459, |
| "step": 599 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6249, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.12, |
| "eval_loss": 0.8239523768424988, |
| "eval_runtime": 108.0006, |
| "eval_samples_per_second": 9.259, |
| "eval_steps_per_second": 4.63, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.12, |
| "mmlu_eval_accuracy": 0.5767479059350066, |
| "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": 0.5517241379310345, |
| "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": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.5, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "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": 0.8, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.34, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7647058823529411, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_law": 0.37058823529411766, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478, |
| "mmlu_eval_accuracy_public_relations": 0.3333333333333333, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 1.0, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.0626621378230674, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8937, |
| "step": 601 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6484, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.1631, |
| "step": 603 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9261, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8708, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7974, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.593, |
| "step": 607 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.853, |
| "step": 609 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8493, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.767, |
| "step": 611 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7043, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6778, |
| "step": 613 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8997, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6661, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6566, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.1303, |
| "step": 617 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6522, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7512, |
| "step": 619 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8616, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6807, |
| "step": 621 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7494, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9582, |
| "step": 623 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9157, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.712, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7072, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9214, |
| "step": 627 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8604, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7721, |
| "step": 629 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6244, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0174, |
| "step": 631 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.933, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8022, |
| "step": 633 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6461, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8137, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.644, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.97, |
| "step": 637 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0134, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7135, |
| "step": 639 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8643, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9489, |
| "step": 641 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9385, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6434, |
| "step": 643 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8027, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5102, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7558, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8466, |
| "step": 647 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9113, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8914, |
| "step": 649 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8724, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7429, |
| "step": 651 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.797, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9116, |
| "step": 653 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0321, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6855, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8333, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8465, |
| "step": 657 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.841, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7407, |
| "step": 659 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.671, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8655, |
| "step": 661 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7531, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.797, |
| "step": 663 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7406, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8582, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7556, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0, |
| "step": 667 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9225, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7745, |
| "step": 669 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.672, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8807, |
| "step": 671 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9416, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9291, |
| "step": 673 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.07, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.865, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5955, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9277, |
| "step": 677 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7636, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6985, |
| "step": 679 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7898, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8206, |
| "step": 681 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8341, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6745, |
| "step": 683 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5721, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8839, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8846, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8544, |
| "step": 687 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9011, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.039, |
| "step": 689 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9518, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7659, |
| "step": 691 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7467, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7914, |
| "step": 693 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8516, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9462, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7714, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8217, |
| "step": 697 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0775, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6857, |
| "step": 699 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8516, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9827, |
| "step": 701 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7188, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8473, |
| "step": 703 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7903, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 1.1412, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9854, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9383, |
| "step": 707 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9324, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.772, |
| "step": 709 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6148, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8209, |
| "step": 711 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5247, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7913, |
| "step": 713 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8309, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.868, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8826, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8212, |
| "step": 717 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.991, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7799, |
| "step": 719 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5986, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 721 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8981, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.672, |
| "step": 723 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7597, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7075, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7639, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5174, |
| "step": 727 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9761, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8007, |
| "step": 729 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9283, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8481, |
| "step": 731 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.876, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9838, |
| "step": 733 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.873, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6769, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8994, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6173, |
| "step": 737 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9697, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9337, |
| "step": 739 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6562, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9336, |
| "step": 741 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9479, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7138, |
| "step": 743 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9331, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8263, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7274, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8048, |
| "step": 747 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8069, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7556, |
| "step": 749 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9893, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7542, |
| "step": 751 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8254, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6808, |
| "step": 753 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8977, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.0146, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8072, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7988, |
| "step": 757 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9259, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7222, |
| "step": 759 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6838, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7277, |
| "step": 761 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8344, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7687, |
| "step": 763 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7731, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6727, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8028, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7871, |
| "step": 767 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7153, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7676, |
| "step": 769 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9727, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6709, |
| "step": 771 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8173, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7286, |
| "step": 773 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.789, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5131, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8337, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8912, |
| "step": 777 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7301, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.0566, |
| "step": 779 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8174, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7458, |
| "step": 781 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7841, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8716, |
| "step": 783 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7436, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6822, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8617, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8575, |
| "step": 787 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7825, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9225, |
| "step": 789 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9179, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6989, |
| "step": 791 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9076, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8997, |
| "step": 793 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9768, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7592, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9849, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7044, |
| "step": 797 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7252, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.785, |
| "step": 799 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7535, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.15, |
| "eval_loss": 0.8314322233200073, |
| "eval_runtime": 107.7767, |
| "eval_samples_per_second": 9.278, |
| "eval_steps_per_second": 4.639, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.15, |
| "mmlu_eval_accuracy": 0.5852864021594637, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.625, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.125, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.6363636363636364, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.5, |
| "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": 0.7619047619047619, |
| "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": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.26, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7647058823529411, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_law": 0.37058823529411766, |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5185185185185185, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.6111111111111112, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.068377559941369, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9542, |
| "step": 801 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9369, |
| "step": 802 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.722, |
| "step": 803 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5851, |
| "step": 804 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.0236, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5474, |
| "step": 806 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9945, |
| "step": 807 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0455, |
| "step": 808 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7946, |
| "step": 809 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7688, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9149, |
| "step": 811 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7135, |
| "step": 812 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7656, |
| "step": 813 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7422, |
| "step": 814 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8751, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8876, |
| "step": 816 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0541, |
| "step": 817 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.993, |
| "step": 818 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.77, |
| "step": 819 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8067, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9629, |
| "step": 821 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9683, |
| "step": 822 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.722, |
| "step": 823 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7577, |
| "step": 824 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6923, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8585, |
| "step": 826 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8164, |
| "step": 827 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7618, |
| "step": 828 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8823, |
| "step": 829 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9447, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9193, |
| "step": 831 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7581, |
| "step": 832 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7875, |
| "step": 833 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7074, |
| "step": 834 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7871, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9631, |
| "step": 836 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9075, |
| "step": 837 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9951, |
| "step": 838 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7714, |
| "step": 839 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7699, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7919, |
| "step": 841 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9562, |
| "step": 842 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9486, |
| "step": 843 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7622, |
| "step": 844 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7765, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6023, |
| "step": 846 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7426, |
| "step": 847 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6728, |
| "step": 848 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.825, |
| "step": 849 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9433, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9373, |
| "step": 851 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9356, |
| "step": 852 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9599, |
| "step": 853 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.825, |
| "step": 854 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9899, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9104, |
| "step": 856 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7832, |
| "step": 857 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7825, |
| "step": 858 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.661, |
| "step": 859 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6978, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7528, |
| "step": 861 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.553, |
| "step": 862 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7198, |
| "step": 863 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7452, |
| "step": 864 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4933, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5857, |
| "step": 866 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8252, |
| "step": 867 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8617, |
| "step": 868 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.64, |
| "step": 869 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7116, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9711, |
| "step": 871 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9874, |
| "step": 872 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8595, |
| "step": 873 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7947, |
| "step": 874 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6689, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9601, |
| "step": 876 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.834, |
| "step": 877 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5666, |
| "step": 878 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7491, |
| "step": 879 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7362, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6809, |
| "step": 881 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8278, |
| "step": 882 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8141, |
| "step": 883 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6499, |
| "step": 884 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7019, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9741, |
| "step": 886 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9711, |
| "step": 887 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7295, |
| "step": 888 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8351, |
| "step": 889 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8951, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6802, |
| "step": 891 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8653, |
| "step": 892 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9284, |
| "step": 893 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8818, |
| "step": 894 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7905, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8017, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7307, |
| "step": 897 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9555, |
| "step": 898 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8246, |
| "step": 899 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8623, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7665, |
| "step": 901 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7087, |
| "step": 902 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7632, |
| "step": 903 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8463, |
| "step": 904 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.67, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9897, |
| "step": 906 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.839, |
| "step": 907 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6107, |
| "step": 908 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0102, |
| "step": 909 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5473, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9059, |
| "step": 911 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9688, |
| "step": 912 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6527, |
| "step": 913 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7323, |
| "step": 914 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7994, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.0027, |
| "step": 916 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9274, |
| "step": 917 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5995, |
| "step": 918 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8599, |
| "step": 919 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7599, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.601, |
| "step": 921 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8202, |
| "step": 922 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8857, |
| "step": 923 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5073, |
| "step": 924 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.0193, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8295, |
| "step": 926 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9962, |
| "step": 927 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.1134, |
| "step": 928 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.833, |
| "step": 929 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9124, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.0302, |
| "step": 931 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9182, |
| "step": 932 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.532, |
| "step": 933 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7662, |
| "step": 934 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8333, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6508, |
| "step": 936 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.603, |
| "step": 937 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8424, |
| "step": 938 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8288, |
| "step": 939 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8063, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7095, |
| "step": 941 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9294, |
| "step": 942 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9205, |
| "step": 943 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8253, |
| "step": 944 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8448, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7412, |
| "step": 946 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9216, |
| "step": 947 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.822, |
| "step": 948 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8435, |
| "step": 949 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8546, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.575, |
| "step": 951 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.0692, |
| "step": 952 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8348, |
| "step": 953 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9301, |
| "step": 954 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7265, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8753, |
| "step": 956 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8021, |
| "step": 957 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7412, |
| "step": 958 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7696, |
| "step": 959 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9858, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8307, |
| "step": 961 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7608, |
| "step": 962 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8428, |
| "step": 963 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9388, |
| "step": 964 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5026, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6896, |
| "step": 966 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7954, |
| "step": 967 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9893, |
| "step": 968 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6204, |
| "step": 969 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8292, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6268, |
| "step": 971 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9288, |
| "step": 972 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8863, |
| "step": 973 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8266, |
| "step": 974 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8311, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7688, |
| "step": 976 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9308, |
| "step": 977 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0505, |
| "step": 978 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8594, |
| "step": 979 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.927, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6642, |
| "step": 981 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8662, |
| "step": 982 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8863, |
| "step": 983 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7942, |
| "step": 984 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9385, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1678, |
| "step": 986 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0085, |
| "step": 987 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0007, |
| "step": 988 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1553, |
| "step": 989 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8498, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1541, |
| "step": 991 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8701, |
| "step": 992 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.691, |
| "step": 993 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7448, |
| "step": 994 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9344, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1617, |
| "step": 996 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6735, |
| "step": 997 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9245, |
| "step": 998 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6245, |
| "step": 999 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7727, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.19, |
| "eval_loss": 0.8221133947372437, |
| "eval_runtime": 107.7614, |
| "eval_samples_per_second": 9.28, |
| "eval_steps_per_second": 4.64, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.19, |
| "mmlu_eval_accuracy": 0.5768050908161603, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.6428571428571429, |
| "mmlu_eval_accuracy_astronomy": 0.5625, |
| "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": 0.18181818181818182, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_physics": 0.36363636363636365, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.6956521739130435, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.7272727272727273, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.5882352941176471, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.3764705882352941, |
| "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.6666666666666666, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.0894444307443678, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7983, |
| "step": 1001 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7486, |
| "step": 1002 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8172, |
| "step": 1003 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6772, |
| "step": 1004 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7724, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8207, |
| "step": 1006 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7904, |
| "step": 1007 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8351, |
| "step": 1008 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7827, |
| "step": 1009 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9647, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9669, |
| "step": 1011 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7974, |
| "step": 1012 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7287, |
| "step": 1013 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9687, |
| "step": 1014 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9239, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9859, |
| "step": 1016 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6932, |
| "step": 1017 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.762, |
| "step": 1018 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8419, |
| "step": 1019 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8942, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6363, |
| "step": 1021 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.0143, |
| "step": 1022 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7529, |
| "step": 1023 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9264, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6283, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7133, |
| "step": 1026 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7825, |
| "step": 1027 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7986, |
| "step": 1028 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6617, |
| "step": 1029 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8952, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9269, |
| "step": 1031 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8428, |
| "step": 1032 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9976, |
| "step": 1033 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9051, |
| "step": 1034 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.704, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6956, |
| "step": 1036 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7071, |
| "step": 1037 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8563, |
| "step": 1038 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5753, |
| "step": 1039 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7108, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7634, |
| "step": 1041 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9367, |
| "step": 1042 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9209, |
| "step": 1043 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7426, |
| "step": 1044 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7212, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.172, |
| "step": 1046 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6434, |
| "step": 1047 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8692, |
| "step": 1048 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7776, |
| "step": 1049 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7578, |
| "step": 1051 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7549, |
| "step": 1052 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8046, |
| "step": 1053 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8274, |
| "step": 1054 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.0264, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8592, |
| "step": 1056 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5615, |
| "step": 1057 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7084, |
| "step": 1058 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8377, |
| "step": 1059 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9046, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9777, |
| "step": 1061 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.598, |
| "step": 1062 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7353, |
| "step": 1063 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7785, |
| "step": 1064 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7571, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8154, |
| "step": 1066 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6559, |
| "step": 1067 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8339, |
| "step": 1068 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9225, |
| "step": 1069 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9206, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8587, |
| "step": 1071 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6064, |
| "step": 1072 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.839, |
| "step": 1073 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7312, |
| "step": 1074 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8338, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6873, |
| "step": 1076 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9526, |
| "step": 1077 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7698, |
| "step": 1078 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8345, |
| "step": 1079 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.836, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6012, |
| "step": 1081 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8149, |
| "step": 1082 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.621, |
| "step": 1083 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9225, |
| "step": 1084 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8493, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6558, |
| "step": 1086 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7498, |
| "step": 1087 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.985, |
| "step": 1088 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.891, |
| "step": 1089 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8744, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8277, |
| "step": 1091 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6847, |
| "step": 1092 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6816, |
| "step": 1093 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6313, |
| "step": 1094 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7649, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9559, |
| "step": 1096 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0069, |
| "step": 1097 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8068, |
| "step": 1098 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.826, |
| "step": 1099 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9864, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8207, |
| "step": 1101 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6934, |
| "step": 1102 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9991, |
| "step": 1103 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7827, |
| "step": 1104 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0234, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9579, |
| "step": 1106 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6391, |
| "step": 1107 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8214, |
| "step": 1108 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0001, |
| "step": 1109 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9233, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5531, |
| "step": 1111 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7158, |
| "step": 1112 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8638, |
| "step": 1113 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8938, |
| "step": 1114 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8129, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.89, |
| "step": 1116 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5949, |
| "step": 1117 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6555, |
| "step": 1118 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8881, |
| "step": 1119 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9674, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8433, |
| "step": 1121 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8145, |
| "step": 1122 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7719, |
| "step": 1123 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.0407, |
| "step": 1124 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7751, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9276, |
| "step": 1126 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.885, |
| "step": 1127 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9224, |
| "step": 1128 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.964, |
| "step": 1129 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.822, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9673, |
| "step": 1131 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8695, |
| "step": 1132 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9036, |
| "step": 1133 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.0815, |
| "step": 1134 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8784, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8444, |
| "step": 1136 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9152, |
| "step": 1137 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7576, |
| "step": 1138 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8372, |
| "step": 1139 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9207, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9243, |
| "step": 1141 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6946, |
| "step": 1142 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5998, |
| "step": 1143 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6083, |
| "step": 1144 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.0207, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8205, |
| "step": 1146 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7422, |
| "step": 1147 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.763, |
| "step": 1148 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8019, |
| "step": 1149 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9221, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8362, |
| "step": 1151 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8024, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8264, |
| "step": 1153 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6174, |
| "step": 1154 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6716, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.798, |
| "step": 1156 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8662, |
| "step": 1157 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7987, |
| "step": 1158 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9795, |
| "step": 1159 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8001, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7329, |
| "step": 1161 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8504, |
| "step": 1162 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9516, |
| "step": 1163 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9072, |
| "step": 1164 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7216, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8843, |
| "step": 1166 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7729, |
| "step": 1167 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8456, |
| "step": 1168 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7582, |
| "step": 1169 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7492, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9831, |
| "step": 1171 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7332, |
| "step": 1172 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7229, |
| "step": 1173 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7622, |
| "step": 1174 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5888, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8685, |
| "step": 1176 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.778, |
| "step": 1177 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9537, |
| "step": 1178 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6841, |
| "step": 1179 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8068, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.0086, |
| "step": 1181 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8699, |
| "step": 1182 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7676, |
| "step": 1183 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8789, |
| "step": 1184 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.375, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.759, |
| "step": 1186 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8446, |
| "step": 1187 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8148, |
| "step": 1188 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6919, |
| "step": 1189 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.608, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.749, |
| "step": 1191 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7304, |
| "step": 1192 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.1062, |
| "step": 1193 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7372, |
| "step": 1194 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9267, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9035, |
| "step": 1196 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8798, |
| "step": 1197 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9881, |
| "step": 1198 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.872, |
| "step": 1199 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.93, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.23, |
| "eval_loss": 0.8195345401763916, |
| "eval_runtime": 107.5294, |
| "eval_samples_per_second": 9.3, |
| "eval_steps_per_second": 4.65, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.23, |
| "mmlu_eval_accuracy": 0.5435130031626643, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.5625, |
| "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": 0.125, |
| "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091, |
| "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.36363636363636365, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.3125, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.7272727272727273, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.36, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.6176470588235294, |
| "mmlu_eval_accuracy_prehistory": 0.6571428571428571, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.37058823529411766, |
| "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7368421052631579, |
| "mmlu_loss": 1.2997911337461547, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8528, |
| "step": 1201 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9547, |
| "step": 1202 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8032, |
| "step": 1203 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.679, |
| "step": 1204 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.824, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.964, |
| "step": 1206 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6154, |
| "step": 1207 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5239, |
| "step": 1208 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7988, |
| "step": 1209 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5223, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7433, |
| "step": 1211 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7831, |
| "step": 1212 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7253, |
| "step": 1213 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8159, |
| "step": 1214 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6605, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5967, |
| "step": 1216 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8218, |
| "step": 1217 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9243, |
| "step": 1218 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8622, |
| "step": 1219 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8015, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6271, |
| "step": 1221 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8378, |
| "step": 1222 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8605, |
| "step": 1223 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7579, |
| "step": 1224 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8241, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7355, |
| "step": 1226 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7556, |
| "step": 1227 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.054, |
| "step": 1228 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8148, |
| "step": 1229 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.878, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.939, |
| "step": 1231 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.624, |
| "step": 1232 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0049, |
| "step": 1233 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8586, |
| "step": 1234 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8201, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8713, |
| "step": 1236 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7515, |
| "step": 1237 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5959, |
| "step": 1238 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8113, |
| "step": 1239 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5581, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7962, |
| "step": 1241 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8917, |
| "step": 1242 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0076, |
| "step": 1243 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8805, |
| "step": 1244 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8305, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8017, |
| "step": 1246 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.1182, |
| "step": 1247 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8506, |
| "step": 1248 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8716, |
| "step": 1249 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7659, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6689, |
| "step": 1251 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0241, |
| "step": 1252 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9119, |
| "step": 1253 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8612, |
| "step": 1254 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6326, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7158, |
| "step": 1256 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9439, |
| "step": 1257 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7628, |
| "step": 1258 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7597, |
| "step": 1259 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7789, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7511, |
| "step": 1261 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6601, |
| "step": 1262 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7342, |
| "step": 1263 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7831, |
| "step": 1264 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 1266 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5815, |
| "step": 1267 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7188, |
| "step": 1268 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7392, |
| "step": 1269 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6834, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8545, |
| "step": 1271 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0233, |
| "step": 1272 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7578, |
| "step": 1273 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.2378, |
| "step": 1274 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9399, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.668, |
| "step": 1276 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7894, |
| "step": 1277 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6307, |
| "step": 1278 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7293, |
| "step": 1279 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.629, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8629, |
| "step": 1281 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9917, |
| "step": 1282 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7331, |
| "step": 1283 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8318, |
| "step": 1284 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.725, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7665, |
| "step": 1286 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8015, |
| "step": 1287 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7864, |
| "step": 1288 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8412, |
| "step": 1289 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9087, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7792, |
| "step": 1291 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0072, |
| "step": 1292 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7536, |
| "step": 1293 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0158, |
| "step": 1294 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9052, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8222, |
| "step": 1296 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8344, |
| "step": 1297 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.825, |
| "step": 1298 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6271, |
| "step": 1299 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9525, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6363, |
| "step": 1301 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7225, |
| "step": 1302 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9168, |
| "step": 1303 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8018, |
| "step": 1304 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6198, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8333, |
| "step": 1306 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8377, |
| "step": 1307 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7898, |
| "step": 1308 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9539, |
| "step": 1309 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0578, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0193, |
| "step": 1311 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8825, |
| "step": 1312 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 1313 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.846, |
| "step": 1314 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5414, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.2466, |
| "step": 1316 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0793, |
| "step": 1317 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7391, |
| "step": 1318 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6585, |
| "step": 1319 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4362, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6499, |
| "step": 1321 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.1133, |
| "step": 1322 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8416, |
| "step": 1323 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5723, |
| "step": 1324 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9606, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6929, |
| "step": 1326 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.846, |
| "step": 1327 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4776, |
| "step": 1328 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.641, |
| "step": 1329 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8026, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8442, |
| "step": 1331 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.5105, |
| "step": 1332 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.778, |
| "step": 1333 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8846, |
| "step": 1334 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8107, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.844, |
| "step": 1336 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8696, |
| "step": 1337 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7486, |
| "step": 1338 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8199, |
| "step": 1339 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9825, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8303, |
| "step": 1341 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7669, |
| "step": 1342 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 1343 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8445, |
| "step": 1344 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7926, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.735, |
| "step": 1346 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8486, |
| "step": 1347 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6096, |
| "step": 1348 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6806, |
| "step": 1349 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8527, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.718, |
| "step": 1351 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8803, |
| "step": 1352 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9253, |
| "step": 1353 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6331, |
| "step": 1354 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7672, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8612, |
| "step": 1356 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7002, |
| "step": 1357 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7032, |
| "step": 1358 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9663, |
| "step": 1359 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8368, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6596, |
| "step": 1361 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0682, |
| "step": 1362 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6945, |
| "step": 1363 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7514, |
| "step": 1364 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.727, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9341, |
| "step": 1366 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8012, |
| "step": 1367 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7649, |
| "step": 1368 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8664, |
| "step": 1369 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8466, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9392, |
| "step": 1371 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8166, |
| "step": 1372 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9135, |
| "step": 1373 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8317, |
| "step": 1374 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9675, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8231, |
| "step": 1376 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5382, |
| "step": 1377 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8736, |
| "step": 1378 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8208, |
| "step": 1379 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8121, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8718, |
| "step": 1381 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6804, |
| "step": 1382 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7236, |
| "step": 1383 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.119, |
| "step": 1384 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8956, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.9316, |
| "step": 1386 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8747, |
| "step": 1387 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8069, |
| "step": 1388 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8237, |
| "step": 1389 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7534, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7788, |
| "step": 1391 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.0281, |
| "step": 1392 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8001, |
| "step": 1393 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8231, |
| "step": 1394 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7892, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.941, |
| "step": 1396 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5953, |
| "step": 1397 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.9596, |
| "step": 1398 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 1399 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6043, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.27, |
| "eval_loss": 0.825057864189148, |
| "eval_runtime": 107.5613, |
| "eval_samples_per_second": 9.297, |
| "eval_steps_per_second": 4.649, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.27, |
| "mmlu_eval_accuracy": 0.5261002332209418, |
| "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091, |
| "mmlu_eval_accuracy_anatomy": 0.35714285714285715, |
| "mmlu_eval_accuracy_astronomy": 0.375, |
| "mmlu_eval_accuracy_business_ethics": 0.36363636363636365, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.5, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.7272727272727273, |
| "mmlu_eval_accuracy_computer_security": 0.36363636363636365, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.3333333333333333, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293, |
| "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.6875, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_geography": 0.6363636363636364, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.7692307692307693, |
| "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, |
| "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, |
| "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.4782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.7692307692307693, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.68, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.6511627906976745, |
| "mmlu_eval_accuracy_moral_disputes": 0.42105263157894735, |
| "mmlu_eval_accuracy_moral_scenarios": 0.32, |
| "mmlu_eval_accuracy_nutrition": 0.5757575757575758, |
| "mmlu_eval_accuracy_philosophy": 0.5882352941176471, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.22580645161290322, |
| "mmlu_eval_accuracy_professional_law": 0.36470588235294116, |
| "mmlu_eval_accuracy_professional_medicine": 0.3225806451612903, |
| "mmlu_eval_accuracy_professional_psychology": 0.5072463768115942, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5185185185185185, |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364, |
| "mmlu_eval_accuracy_virology": 0.3333333333333333, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.0671936689484525, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8498, |
| "step": 1401 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5981, |
| "step": 1402 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7947, |
| "step": 1403 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7712, |
| "step": 1404 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.854, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8629, |
| "step": 1406 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6836, |
| "step": 1407 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6803, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.558, |
| "step": 1409 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8982, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.914, |
| "step": 1411 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5412, |
| "step": 1412 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7332, |
| "step": 1413 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.9627, |
| "step": 1414 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8164, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5844, |
| "step": 1416 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8248, |
| "step": 1417 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.2164, |
| "step": 1418 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8451, |
| "step": 1419 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7261, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6247, |
| "step": 1421 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.837, |
| "step": 1422 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.0168, |
| "step": 1423 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8536, |
| "step": 1424 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7846, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7055, |
| "step": 1426 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6929, |
| "step": 1427 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.717, |
| "step": 1428 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7653, |
| "step": 1429 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8804, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7957, |
| "step": 1431 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9933, |
| "step": 1432 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5636, |
| "step": 1433 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7725, |
| "step": 1434 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5726, |
| "step": 1435 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8531, |
| "step": 1436 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8233, |
| "step": 1437 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0856, |
| "step": 1438 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.686, |
| "step": 1439 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.673, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6611, |
| "step": 1441 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6216, |
| "step": 1442 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9692, |
| "step": 1443 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7918, |
| "step": 1444 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7039, |
| "step": 1445 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7026, |
| "step": 1446 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7067, |
| "step": 1447 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.848, |
| "step": 1448 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9224, |
| "step": 1449 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0266, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8069, |
| "step": 1451 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8536, |
| "step": 1452 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7476, |
| "step": 1453 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8078, |
| "step": 1454 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9525, |
| "step": 1455 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.649, |
| "step": 1456 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8569, |
| "step": 1457 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9695, |
| "step": 1458 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0852, |
| "step": 1459 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8398, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7028, |
| "step": 1461 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7751, |
| "step": 1462 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7162, |
| "step": 1463 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0478, |
| "step": 1464 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9565, |
| "step": 1465 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9746, |
| "step": 1466 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6842, |
| "step": 1467 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0284, |
| "step": 1468 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7857, |
| "step": 1469 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6987, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8753, |
| "step": 1471 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7316, |
| "step": 1472 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7754, |
| "step": 1473 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8902, |
| "step": 1474 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7006, |
| "step": 1475 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8815, |
| "step": 1476 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8732, |
| "step": 1477 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8545, |
| "step": 1478 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7443, |
| "step": 1479 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7169, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.898, |
| "step": 1481 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8762, |
| "step": 1482 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7574, |
| "step": 1483 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9288, |
| "step": 1484 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8225, |
| "step": 1485 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.589, |
| "step": 1486 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6509, |
| "step": 1487 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8877, |
| "step": 1488 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8212, |
| "step": 1489 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.938, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8443, |
| "step": 1491 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8144, |
| "step": 1492 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5863, |
| "step": 1493 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8429, |
| "step": 1494 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8094, |
| "step": 1495 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.889, |
| "step": 1496 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7291, |
| "step": 1497 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9432, |
| "step": 1498 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7819, |
| "step": 1499 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7555, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7687, |
| "step": 1501 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8002, |
| "step": 1502 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8006, |
| "step": 1503 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8311, |
| "step": 1504 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.687, |
| "step": 1505 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8844, |
| "step": 1506 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.964, |
| "step": 1507 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6618, |
| "step": 1508 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8479, |
| "step": 1509 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9227, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8525, |
| "step": 1511 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7547, |
| "step": 1512 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7651, |
| "step": 1513 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8823, |
| "step": 1514 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8152, |
| "step": 1515 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8002, |
| "step": 1516 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8979, |
| "step": 1517 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7671, |
| "step": 1518 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9666, |
| "step": 1519 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9958, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7264, |
| "step": 1521 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8294, |
| "step": 1522 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8473, |
| "step": 1523 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4256, |
| "step": 1524 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.0817, |
| "step": 1525 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8366, |
| "step": 1526 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9516, |
| "step": 1527 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5898, |
| "step": 1528 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7336, |
| "step": 1529 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8132, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7671, |
| "step": 1531 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7801, |
| "step": 1532 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0315, |
| "step": 1533 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5838, |
| "step": 1534 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9304, |
| "step": 1535 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.966, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7656, |
| "step": 1537 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6905, |
| "step": 1538 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.754, |
| "step": 1539 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9846, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8398, |
| "step": 1541 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9174, |
| "step": 1542 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6372, |
| "step": 1543 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7296, |
| "step": 1544 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9294, |
| "step": 1545 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7385, |
| "step": 1546 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7534, |
| "step": 1547 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7553, |
| "step": 1548 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0317, |
| "step": 1549 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8087, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6293, |
| "step": 1551 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7301, |
| "step": 1552 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9018, |
| "step": 1553 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.787, |
| "step": 1554 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6241, |
| "step": 1555 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5388, |
| "step": 1556 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6282, |
| "step": 1557 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8723, |
| "step": 1558 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7209, |
| "step": 1559 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7809, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5963, |
| "step": 1561 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8324, |
| "step": 1562 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6484, |
| "step": 1563 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8241, |
| "step": 1564 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5794, |
| "step": 1565 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9628, |
| "step": 1566 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7016, |
| "step": 1567 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0955, |
| "step": 1568 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8643, |
| "step": 1569 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6446, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9405, |
| "step": 1571 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8703, |
| "step": 1572 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6421, |
| "step": 1573 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8491, |
| "step": 1574 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7685, |
| "step": 1575 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0141, |
| "step": 1576 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8018, |
| "step": 1577 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7598, |
| "step": 1578 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9007, |
| "step": 1579 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8525, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8368, |
| "step": 1581 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7355, |
| "step": 1582 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6625, |
| "step": 1583 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.841, |
| "step": 1584 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5787, |
| "step": 1585 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8488, |
| "step": 1586 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.3169, |
| "step": 1587 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9284, |
| "step": 1588 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8948, |
| "step": 1589 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7806, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.0476, |
| "step": 1591 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5354, |
| "step": 1592 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8623, |
| "step": 1593 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9032, |
| "step": 1594 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7472, |
| "step": 1595 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.1098, |
| "step": 1596 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.412, |
| "step": 1597 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.869, |
| "step": 1598 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7577, |
| "step": 1599 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8903, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.31, |
| "eval_loss": 0.8221507668495178, |
| "eval_runtime": 107.8499, |
| "eval_samples_per_second": 9.272, |
| "eval_steps_per_second": 4.636, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.31, |
| "mmlu_eval_accuracy": 0.5689444031497279, |
| "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.5625, |
| "mmlu_eval_accuracy_business_ethics": 0.2727272727272727, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.6875, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "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": 0.8571428571428571, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.8461538461538461, |
| "mmlu_eval_accuracy_human_aging": 0.6956521739130435, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.76, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.38, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.4, |
| "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.9545454545454546, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 0.8437367426936683, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6816, |
| "step": 1601 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6536, |
| "step": 1602 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7733, |
| "step": 1603 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.01, |
| "step": 1604 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.0122, |
| "step": 1605 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7618, |
| "step": 1606 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 1607 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.818, |
| "step": 1608 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8361, |
| "step": 1609 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5973, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.057, |
| "step": 1611 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.888, |
| "step": 1612 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8184, |
| "step": 1613 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7498, |
| "step": 1614 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9438, |
| "step": 1615 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8675, |
| "step": 1616 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7394, |
| "step": 1617 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9155, |
| "step": 1618 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.878, |
| "step": 1619 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9278, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.0698, |
| "step": 1621 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6272, |
| "step": 1622 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7526, |
| "step": 1623 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9892, |
| "step": 1624 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8515, |
| "step": 1625 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7535, |
| "step": 1626 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9034, |
| "step": 1627 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8185, |
| "step": 1628 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8484, |
| "step": 1629 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8227, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9481, |
| "step": 1631 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8076, |
| "step": 1632 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7329, |
| "step": 1633 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6009, |
| "step": 1634 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9391, |
| "step": 1635 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8281, |
| "step": 1636 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9875, |
| "step": 1637 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.705, |
| "step": 1638 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 1639 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9632, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9511, |
| "step": 1641 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9213, |
| "step": 1642 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9555, |
| "step": 1643 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7763, |
| "step": 1644 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8044, |
| "step": 1645 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8144, |
| "step": 1646 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7764, |
| "step": 1647 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7505, |
| "step": 1648 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9509, |
| "step": 1649 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5935, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9426, |
| "step": 1651 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9142, |
| "step": 1652 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.006, |
| "step": 1653 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8634, |
| "step": 1654 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.0491, |
| "step": 1655 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8147, |
| "step": 1656 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6093, |
| "step": 1657 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5288, |
| "step": 1658 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7538, |
| "step": 1659 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8122, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8889, |
| "step": 1661 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7966, |
| "step": 1662 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8071, |
| "step": 1663 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8796, |
| "step": 1664 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6675, |
| "step": 1665 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 1666 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.0125, |
| "step": 1667 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7006, |
| "step": 1668 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9563, |
| "step": 1669 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7269, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9715, |
| "step": 1671 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.0151, |
| "step": 1672 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6572, |
| "step": 1673 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5888, |
| "step": 1674 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9198, |
| "step": 1675 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9, |
| "step": 1676 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7485, |
| "step": 1677 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5904, |
| "step": 1678 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6453, |
| "step": 1679 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6435, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7521, |
| "step": 1681 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8277, |
| "step": 1682 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7807, |
| "step": 1683 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7882, |
| "step": 1684 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7859, |
| "step": 1685 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.0744, |
| "step": 1686 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7908, |
| "step": 1687 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8735, |
| "step": 1688 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5536, |
| "step": 1689 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7903, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 1.0194, |
| "step": 1691 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7377, |
| "step": 1692 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8358, |
| "step": 1693 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.524, |
| "step": 1694 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7802, |
| "step": 1695 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6332, |
| "step": 1696 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9569, |
| "step": 1697 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9413, |
| "step": 1698 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 1.0077, |
| "step": 1699 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7344, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6809, |
| "step": 1701 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7994, |
| "step": 1702 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7187, |
| "step": 1703 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 1704 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9343, |
| "step": 1705 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9431, |
| "step": 1706 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7519, |
| "step": 1707 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7143, |
| "step": 1708 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8279, |
| "step": 1709 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5734, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9228, |
| "step": 1711 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9653, |
| "step": 1712 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7466, |
| "step": 1713 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9613, |
| "step": 1714 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7572, |
| "step": 1715 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7698, |
| "step": 1716 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7907, |
| "step": 1717 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7198, |
| "step": 1718 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.866, |
| "step": 1719 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7706, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6453, |
| "step": 1721 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9814, |
| "step": 1722 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5207, |
| "step": 1723 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7474, |
| "step": 1724 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.585, |
| "step": 1725 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9092, |
| "step": 1726 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8555, |
| "step": 1727 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8701, |
| "step": 1728 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9159, |
| "step": 1729 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9023, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7989, |
| "step": 1731 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7434, |
| "step": 1732 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7934, |
| "step": 1733 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8429, |
| "step": 1734 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7717, |
| "step": 1735 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7028, |
| "step": 1736 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7803, |
| "step": 1737 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7464, |
| "step": 1738 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 1.0388, |
| "step": 1739 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8894, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.812, |
| "step": 1741 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8362, |
| "step": 1742 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8536, |
| "step": 1743 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9904, |
| "step": 1744 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7413, |
| "step": 1745 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9807, |
| "step": 1746 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8341, |
| "step": 1747 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8223, |
| "step": 1748 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7822, |
| "step": 1749 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7082, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4812, |
| "step": 1751 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8285, |
| "step": 1752 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8178, |
| "step": 1753 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6253, |
| "step": 1754 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7521, |
| "step": 1755 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8063, |
| "step": 1756 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.0525, |
| "step": 1757 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6469, |
| "step": 1758 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5751, |
| "step": 1759 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9385, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6278, |
| "step": 1761 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8664, |
| "step": 1762 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7919, |
| "step": 1763 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6463, |
| "step": 1764 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9496, |
| "step": 1765 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5857, |
| "step": 1766 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8012, |
| "step": 1767 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7307, |
| "step": 1768 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.0323, |
| "step": 1769 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9045, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4746, |
| "step": 1771 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7484, |
| "step": 1772 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7377, |
| "step": 1773 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9423, |
| "step": 1774 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8146, |
| "step": 1775 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8995, |
| "step": 1776 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5627, |
| "step": 1777 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6484, |
| "step": 1778 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9091, |
| "step": 1779 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7198, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6813, |
| "step": 1781 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9187, |
| "step": 1782 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9621, |
| "step": 1783 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6419, |
| "step": 1784 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.2731, |
| "step": 1785 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8213, |
| "step": 1786 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8048, |
| "step": 1787 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5645, |
| "step": 1788 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8758, |
| "step": 1789 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9589, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7625, |
| "step": 1791 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7643, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8189, |
| "step": 1793 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8299, |
| "step": 1794 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.789, |
| "step": 1795 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7415, |
| "step": 1796 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6801, |
| "step": 1797 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0205, |
| "step": 1798 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8781, |
| "step": 1799 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0804, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.35, |
| "eval_loss": 0.8146001100540161, |
| "eval_runtime": 108.0941, |
| "eval_samples_per_second": 9.251, |
| "eval_steps_per_second": 4.626, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.35, |
| "mmlu_eval_accuracy": 0.5759611012722413, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.625, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.5625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667, |
| "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": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.8, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.28, |
| "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": 0.31176470588235294, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.9545454545454546, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.0327665438831786, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8016, |
| "step": 1801 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 1802 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6574, |
| "step": 1803 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8124, |
| "step": 1804 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5276, |
| "step": 1805 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7502, |
| "step": 1806 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0469, |
| "step": 1807 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9149, |
| "step": 1808 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.889, |
| "step": 1809 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6277, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7256, |
| "step": 1811 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8466, |
| "step": 1812 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8437, |
| "step": 1813 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8968, |
| "step": 1814 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8476, |
| "step": 1815 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5397, |
| "step": 1816 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7074, |
| "step": 1817 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9776, |
| "step": 1818 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 1819 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9726, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7608, |
| "step": 1821 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8355, |
| "step": 1822 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7307, |
| "step": 1823 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7717, |
| "step": 1824 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.908, |
| "step": 1825 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6057, |
| "step": 1826 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6752, |
| "step": 1827 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9067, |
| "step": 1828 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8395, |
| "step": 1829 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6725, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7249, |
| "step": 1831 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7055, |
| "step": 1832 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9172, |
| "step": 1833 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9367, |
| "step": 1834 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8796, |
| "step": 1835 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8621, |
| "step": 1836 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8216, |
| "step": 1837 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0159, |
| "step": 1838 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9952, |
| "step": 1839 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8941, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9377, |
| "step": 1841 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7245, |
| "step": 1842 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7017, |
| "step": 1843 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0832, |
| "step": 1844 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6605, |
| "step": 1845 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8471, |
| "step": 1846 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6213, |
| "step": 1847 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8034, |
| "step": 1848 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7434, |
| "step": 1849 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7257, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6828, |
| "step": 1851 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 1.0467, |
| "step": 1852 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 1.0019, |
| "step": 1853 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8737, |
| "step": 1854 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6707, |
| "step": 1855 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6337, |
| "step": 1856 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8495, |
| "step": 1857 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7764, |
| "step": 1858 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7458, |
| "step": 1859 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7908, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9008, |
| "step": 1861 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.843, |
| "step": 1862 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8604, |
| "step": 1863 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7122, |
| "step": 1864 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.856, |
| "step": 1865 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9627, |
| "step": 1866 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8974, |
| "step": 1867 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.986, |
| "step": 1868 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8162, |
| "step": 1869 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7787, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8206, |
| "step": 1871 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6129, |
| "step": 1872 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9801, |
| "step": 1873 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7778, |
| "step": 1874 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6498, |
| "step": 1875 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8788, |
| "step": 1876 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7876, |
| "step": 1877 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7567, |
| "step": 1878 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8699, |
| "step": 1879 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5723, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8008, |
| "step": 1881 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9557, |
| "step": 1882 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 1.1339, |
| "step": 1883 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7568, |
| "step": 1884 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9133, |
| "step": 1885 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7743, |
| "step": 1886 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7552, |
| "step": 1887 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6108, |
| "step": 1888 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8523, |
| "step": 1889 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7165, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7073, |
| "step": 1891 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8273, |
| "step": 1892 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9308, |
| "step": 1893 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5477, |
| "step": 1894 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 1.0091, |
| "step": 1895 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6657, |
| "step": 1896 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7054, |
| "step": 1897 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.763, |
| "step": 1898 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8852, |
| "step": 1899 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7191, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.767, |
| "step": 1901 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6845, |
| "step": 1902 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8597, |
| "step": 1903 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0146, |
| "step": 1904 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8562, |
| "step": 1905 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9003, |
| "step": 1906 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7635, |
| "step": 1907 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9138, |
| "step": 1908 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6774, |
| "step": 1909 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.668, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9251, |
| "step": 1911 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9255, |
| "step": 1912 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7345, |
| "step": 1913 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8944, |
| "step": 1914 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0152, |
| "step": 1915 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8073, |
| "step": 1916 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8764, |
| "step": 1917 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6797, |
| "step": 1918 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.639, |
| "step": 1919 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6815, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0481, |
| "step": 1921 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5657, |
| "step": 1922 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7028, |
| "step": 1923 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8974, |
| "step": 1924 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.875, |
| "step": 1925 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7509, |
| "step": 1926 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.756, |
| "step": 1927 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0423, |
| "step": 1928 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8452, |
| "step": 1929 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7343, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8304, |
| "step": 1931 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8745, |
| "step": 1932 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7954, |
| "step": 1933 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7008, |
| "step": 1934 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6662, |
| "step": 1935 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5971, |
| "step": 1936 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.789, |
| "step": 1937 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6848, |
| "step": 1938 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8327, |
| "step": 1939 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6904, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7409, |
| "step": 1941 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9888, |
| "step": 1942 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0448, |
| "step": 1943 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0101, |
| "step": 1944 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7703, |
| "step": 1945 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8223, |
| "step": 1946 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0914, |
| "step": 1947 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0928, |
| "step": 1948 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7851, |
| "step": 1949 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6261, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7146, |
| "step": 1951 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8096, |
| "step": 1952 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7789, |
| "step": 1953 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9109, |
| "step": 1954 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8681, |
| "step": 1955 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6915, |
| "step": 1956 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7957, |
| "step": 1957 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.833, |
| "step": 1958 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6663, |
| "step": 1959 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7429, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 1961 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9757, |
| "step": 1962 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6649, |
| "step": 1963 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6528, |
| "step": 1964 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9402, |
| "step": 1965 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6765, |
| "step": 1966 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6915, |
| "step": 1967 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9004, |
| "step": 1968 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9036, |
| "step": 1969 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.128, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8759, |
| "step": 1971 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6407, |
| "step": 1972 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6665, |
| "step": 1973 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9832, |
| "step": 1974 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.951, |
| "step": 1975 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9842, |
| "step": 1976 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7394, |
| "step": 1977 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9248, |
| "step": 1978 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6551, |
| "step": 1979 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6816, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8979, |
| "step": 1981 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8824, |
| "step": 1982 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9177, |
| "step": 1983 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8617, |
| "step": 1984 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6392, |
| "step": 1985 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7309, |
| "step": 1986 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9224, |
| "step": 1987 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8798, |
| "step": 1988 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7272, |
| "step": 1989 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.0722, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7289, |
| "step": 1991 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9002, |
| "step": 1992 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8929, |
| "step": 1993 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.0633, |
| "step": 1994 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7856, |
| "step": 1995 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9059, |
| "step": 1996 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8342, |
| "step": 1997 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8652, |
| "step": 1998 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8609, |
| "step": 1999 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9752, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.38, |
| "eval_loss": 0.8149585127830505, |
| "eval_runtime": 108.0696, |
| "eval_samples_per_second": 9.253, |
| "eval_steps_per_second": 4.627, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.38, |
| "mmlu_eval_accuracy": 0.5738805591482586, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, |
| "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": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "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": 0.65625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, |
| "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": 0.5348837209302325, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818, |
| "mmlu_eval_accuracy_high_school_world_history": 0.8076923076923077, |
| "mmlu_eval_accuracy_human_aging": 0.6521739130434783, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512, |
| "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316, |
| "mmlu_eval_accuracy_moral_scenarios": 0.34, |
| "mmlu_eval_accuracy_nutrition": 0.6363636363636364, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, |
| "mmlu_eval_accuracy_professional_law": 0.37058823529411766, |
| "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.105458383538393, |
| "step": 2000 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 15585, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 2.991499226278134e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
