| { |
| "best_metric": 0.7605424523353577, |
| "best_model_checkpoint": "experts/mistralic-expert-30/checkpoint-1800", |
| "epoch": 1.1129660545353366, |
| "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": 0.917, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8831, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9565, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0423, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8049, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.69, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.2058, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8205, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7703, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8955, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7268, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7995, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2621, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9095, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7625, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9512, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6878, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8436, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0459, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8211, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8762, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7982, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.848, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8634, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8242, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8436, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6039, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8928, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8944, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9128, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9446, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1871, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9247, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8389, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6557, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1418, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7204, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6996, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.849, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8153, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0103, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6409, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7049, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6787, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6913, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7891, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6736, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6203, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8579, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7625, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8473, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6805, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7965, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0887, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7472, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8054, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8281, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9568, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6355, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7689, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6185, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9125, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7082, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.64, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7568, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6326, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0269, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6974, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8957, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7612, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8794, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9091, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9245, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7243, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9059, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9854, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7721, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9271, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7339, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8063, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8364, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9442, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7905, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5438, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9456, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9273, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8956, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7529, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8644, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9585, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7955, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4851, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8139, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.876, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6603, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9285, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.637, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.667, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7723, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0664, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8124, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6158, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8176, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.829, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9279, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6661, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6813, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8523, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6664, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9014, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4691, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7214, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7615, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7261, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.768, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6997, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8658, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6748, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8024, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7302, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.885, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9024, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8236, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5508, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9144, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7062, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.676, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7982, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6412, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.565, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7098, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5431, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7696, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6103, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6949, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.852, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7568, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8673, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8329, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8403, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9229, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0478, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0403, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7259, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7757, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8211, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8058, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8081, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8451, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8486, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7606, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8513, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.645, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7491, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6105, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.818, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.026, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6129, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8144, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7729, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6458, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0529, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8144, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8033, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7475, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9598, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.918, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8383, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9462, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7227, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8182, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7564, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.0113, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9714, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6105, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7716, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.899, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7271, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8987, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7743, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7397, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7813, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8339, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8814, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6205, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8181, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9911, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7913, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.0493, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7468, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7198, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8347, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6803, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6587, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8048, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7418, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7057, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7048, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7441, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6067, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.11, |
| "eval_loss": 0.7916712760925293, |
| "eval_runtime": 82.9454, |
| "eval_samples_per_second": 12.056, |
| "eval_steps_per_second": 6.028, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.11, |
| "mmlu_eval_accuracy": 0.5872195320404554, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "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": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.07142857142857142, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "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": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.39, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_law": 0.4, |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "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": 0.8421052631578947, |
| "mmlu_loss": 1.2516618557371608, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7795, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5807, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8287, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9606, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9183, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8648, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5634, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6642, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9435, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8524, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9557, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.651, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5265, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8406, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.813, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.1195, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9042, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9216, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.1693, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8188, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9122, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6351, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5267, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7818, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6986, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8823, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8112, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4976, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8545, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6447, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8584, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.557, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8453, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7156, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.766, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7877, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6461, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6278, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0594, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8383, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8323, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6091, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6537, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9819, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6718, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4985, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.609, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7838, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7154, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6985, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5889, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9411, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5611, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7642, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8727, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5774, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8326, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5841, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7623, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.543, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.1922, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6953, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8819, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7631, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6368, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6696, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6758, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6634, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8266, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.71, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7797, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6015, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.892, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8356, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8191, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8106, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7587, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.773, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.821, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8614, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.981, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5964, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.874, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7826, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8872, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8336, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7758, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7344, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0131, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7578, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0172, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9115, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.787, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7284, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7582, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0598, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8959, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7597, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.008, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.642, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8525, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8235, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6835, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0166, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.728, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9295, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7358, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0691, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6312, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8472, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7948, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8899, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8352, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7379, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9018, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7815, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8065, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8694, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8247, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6853, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9481, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6872, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7663, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8693, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.0304, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7774, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7432, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7034, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9093, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.618, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.0275, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7891, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7131, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8858, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6563, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6493, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.581, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.569, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6081, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0199, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9034, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9528, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8019, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6197, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0557, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7312, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7088, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7954, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6534, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8635, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9934, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6114, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8228, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5536, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6938, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9188, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8113, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6773, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7933, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7914, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7919, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6259, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7255, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8117, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5816, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8125, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7813, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9051, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7769, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7823, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0435, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9003, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.054, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7153, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.544, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9158, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8449, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.2083, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9851, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9702, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7137, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5867, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9597, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7685, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7543, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5758, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4995, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5777, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.0203, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7143, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8262, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7772, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8867, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.2762, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6897, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8571, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7049, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.22, |
| "eval_loss": 0.7784976959228516, |
| "eval_runtime": 82.935, |
| "eval_samples_per_second": 12.058, |
| "eval_steps_per_second": 6.029, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.22, |
| "mmlu_eval_accuracy": 0.597359193082703, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.07142857142857142, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.38, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.6285714285714286, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.40588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.7037037037037037, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.3869162216665851, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7011, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6796, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.2008, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6586, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9062, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7695, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8481, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.004, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7552, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.0011, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.0865, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6273, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.758, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9304, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6851, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8144, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.755, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7335, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8084, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6346, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8916, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.0456, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7943, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8768, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.1603, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.584, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6938, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.622, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7227, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.863, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6609, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8388, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.924, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9388, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6788, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9342, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6548, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8123, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8848, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6967, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0012, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7062, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7387, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7116, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0237, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.1968, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8405, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6748, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7525, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8419, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8057, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5776, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9549, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9802, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8299, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7293, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8854, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7156, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8028, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7525, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9252, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0772, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8538, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0446, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9455, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0433, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0316, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7749, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7248, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7955, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.705, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6042, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8003, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7404, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8748, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8478, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8937, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8792, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.1139, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8187, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8801, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7372, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6295, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7979, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7443, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5995, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8917, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6565, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7472, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6585, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6609, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8794, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7859, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7741, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9115, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8919, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.0071, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5511, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9811, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6222, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7608, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8506, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9435, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7317, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8775, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5091, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8361, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9554, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7774, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7158, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7544, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8753, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7032, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7305, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7603, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7764, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6194, |
| "step": 517 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.1323, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.937, |
| "step": 519 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7652, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9415, |
| "step": 521 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7793, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8108, |
| "step": 523 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7914, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8662, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9023, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6402, |
| "step": 527 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.145, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6027, |
| "step": 529 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7041, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.618, |
| "step": 531 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6028, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9375, |
| "step": 533 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8439, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.807, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7757, |
| "step": 537 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7661, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8159, |
| "step": 539 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0684, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.664, |
| "step": 541 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.0304, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9596, |
| "step": 543 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7808, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6666, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5879, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.779, |
| "step": 547 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6015, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7296, |
| "step": 549 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7238, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8034, |
| "step": 551 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.1363, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9616, |
| "step": 553 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8757, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7032, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6976, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6586, |
| "step": 557 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6957, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6512, |
| "step": 559 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.0339, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.737, |
| "step": 561 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8371, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5524, |
| "step": 563 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6859, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6328, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.1986, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7582, |
| "step": 567 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6379, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.0128, |
| "step": 569 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6914, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9703, |
| "step": 571 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.749, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8763, |
| "step": 573 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5792, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.1809, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8333, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8755, |
| "step": 577 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6851, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.0471, |
| "step": 579 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.946, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7806, |
| "step": 581 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8829, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9606, |
| "step": 583 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9547, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6903, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.873, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7348, |
| "step": 587 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8593, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6386, |
| "step": 589 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7687, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 1.0703, |
| "step": 591 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8591, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8562, |
| "step": 593 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6137, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.722, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6615, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.831, |
| "step": 597 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8047, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.683, |
| "step": 599 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.849, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.33, |
| "eval_loss": 0.7726691365242004, |
| "eval_runtime": 83.0392, |
| "eval_samples_per_second": 12.043, |
| "eval_steps_per_second": 6.021, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.33, |
| "mmlu_eval_accuracy": 0.5898730754800985, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.5625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.3, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.6686991858575737, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8094, |
| "step": 601 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8538, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7692, |
| "step": 603 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8488, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6124, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9341, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8824, |
| "step": 607 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8314, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.799, |
| "step": 609 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6911, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7006, |
| "step": 611 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.0695, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.547, |
| "step": 613 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.9192, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6776, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6046, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.704, |
| "step": 617 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8071, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.687, |
| "step": 619 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7949, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.931, |
| "step": 621 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9305, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6317, |
| "step": 623 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7995, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.655, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6989, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7784, |
| "step": 627 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7286, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7863, |
| "step": 629 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6692, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7514, |
| "step": 631 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9918, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7463, |
| "step": 633 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7426, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7359, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7011, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.872, |
| "step": 637 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6166, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6944, |
| "step": 639 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8465, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6953, |
| "step": 641 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6678, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6941, |
| "step": 643 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.755, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8597, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8673, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7136, |
| "step": 647 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 1.0527, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6628, |
| "step": 649 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7831, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7488, |
| "step": 651 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7073, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.932, |
| "step": 653 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8989, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5849, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9153, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8341, |
| "step": 657 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.601, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9564, |
| "step": 659 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8988, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7874, |
| "step": 661 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8275, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7766, |
| "step": 663 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7234, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8408, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0936, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7411, |
| "step": 667 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8804, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7184, |
| "step": 669 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8789, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9489, |
| "step": 671 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.1602, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7048, |
| "step": 673 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9023, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7527, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9468, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6498, |
| "step": 677 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.1241, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7673, |
| "step": 679 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.1057, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5039, |
| "step": 681 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7335, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6466, |
| "step": 683 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7219, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8325, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6342, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9708, |
| "step": 687 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7624, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.0402, |
| "step": 689 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.891, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7093, |
| "step": 691 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6578, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6603, |
| "step": 693 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6762, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6362, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6938, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7137, |
| "step": 697 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7714, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6363, |
| "step": 699 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.9081, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6552, |
| "step": 701 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8471, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6339, |
| "step": 703 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6266, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7092, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.617, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8699, |
| "step": 707 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7717, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 1.0034, |
| "step": 709 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8068, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8766, |
| "step": 711 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8226, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.5197, |
| "step": 713 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 1.006, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 1.0099, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7304, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 717 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7918, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7122, |
| "step": 719 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.5629, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.577, |
| "step": 721 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8255, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 1.1059, |
| "step": 723 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.9009, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8437, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8454, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.6206, |
| "step": 727 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7477, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7166, |
| "step": 729 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6266, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8614, |
| "step": 731 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8098, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8656, |
| "step": 733 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6536, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7785, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8529, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.5697, |
| "step": 737 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7945, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8824, |
| "step": 739 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8023, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6716, |
| "step": 741 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6102, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 1.1388, |
| "step": 743 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6859, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.5529, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8628, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.9995, |
| "step": 747 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 1.0449, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7464, |
| "step": 749 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7865, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7028, |
| "step": 751 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.9694, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6978, |
| "step": 753 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.9477, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.9811, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6546, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6885, |
| "step": 757 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6217, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8715, |
| "step": 759 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8769, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7903, |
| "step": 761 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7649, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8519, |
| "step": 763 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.9744, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7152, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7487, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7429, |
| "step": 767 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8211, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.5692, |
| "step": 769 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6741, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8843, |
| "step": 771 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.725, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7012, |
| "step": 773 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6027, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6644, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 1.0466, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8491, |
| "step": 777 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8558, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.793, |
| "step": 779 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6651, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7176, |
| "step": 781 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7185, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7267, |
| "step": 783 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 1.0453, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 1.0622, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 1.1521, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.807, |
| "step": 787 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7072, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9032, |
| "step": 789 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9485, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7886, |
| "step": 791 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9187, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7337, |
| "step": 793 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7969, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.5746, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.8626, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7413, |
| "step": 797 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9744, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7961, |
| "step": 799 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.5264, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.45, |
| "eval_loss": 0.7736682891845703, |
| "eval_runtime": 82.8894, |
| "eval_samples_per_second": 12.064, |
| "eval_steps_per_second": 6.032, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.45, |
| "mmlu_eval_accuracy": 0.5959330926028517, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, |
| "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": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.36363636363636365, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.6666666666666666, |
| "mmlu_eval_accuracy_electrical_engineering": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.53125, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.3, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, |
| "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.664157256056061, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7129, |
| "step": 801 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.888, |
| "step": 802 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6792, |
| "step": 803 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7045, |
| "step": 804 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7198, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8233, |
| "step": 806 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 1.0325, |
| "step": 807 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6939, |
| "step": 808 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8991, |
| "step": 809 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.9371, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 1.0169, |
| "step": 811 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6112, |
| "step": 812 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.5585, |
| "step": 813 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 1.2605, |
| "step": 814 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8294, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6615, |
| "step": 816 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6685, |
| "step": 817 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7242, |
| "step": 818 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8733, |
| "step": 819 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6434, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8741, |
| "step": 821 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7012, |
| "step": 822 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6215, |
| "step": 823 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8369, |
| "step": 824 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7389, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.5556, |
| "step": 826 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8524, |
| "step": 827 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6824, |
| "step": 828 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.708, |
| "step": 829 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7716, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9216, |
| "step": 831 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.5337, |
| "step": 832 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7778, |
| "step": 833 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7028, |
| "step": 834 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9445, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.788, |
| "step": 836 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.9585, |
| "step": 837 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7352, |
| "step": 838 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7828, |
| "step": 839 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6872, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7718, |
| "step": 841 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 1.0173, |
| "step": 842 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 1.0552, |
| "step": 843 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5785, |
| "step": 844 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7474, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6906, |
| "step": 846 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6615, |
| "step": 847 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6512, |
| "step": 848 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.8627, |
| "step": 849 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.865, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.9555, |
| "step": 851 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6198, |
| "step": 852 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7384, |
| "step": 853 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6158, |
| "step": 854 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.78, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6948, |
| "step": 856 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.671, |
| "step": 857 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.812, |
| "step": 858 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8024, |
| "step": 859 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8047, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 1.076, |
| "step": 861 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8193, |
| "step": 862 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8441, |
| "step": 863 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7713, |
| "step": 864 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8932, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8177, |
| "step": 866 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.682, |
| "step": 867 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7099, |
| "step": 868 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8821, |
| "step": 869 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6317, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.708, |
| "step": 871 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7151, |
| "step": 872 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8716, |
| "step": 873 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8376, |
| "step": 874 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.5943, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 1.0619, |
| "step": 876 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7217, |
| "step": 877 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.833, |
| "step": 878 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7117, |
| "step": 879 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6553, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6977, |
| "step": 881 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8597, |
| "step": 882 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7501, |
| "step": 883 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7776, |
| "step": 884 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6864, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8728, |
| "step": 886 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.9045, |
| "step": 887 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6664, |
| "step": 888 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6195, |
| "step": 889 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8296, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.6577, |
| "step": 891 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7348, |
| "step": 892 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 893 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.977, |
| "step": 894 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.649, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5617, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7239, |
| "step": 897 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7526, |
| "step": 898 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8213, |
| "step": 899 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7565, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7491, |
| "step": 901 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7702, |
| "step": 902 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.6165, |
| "step": 903 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8112, |
| "step": 904 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7853, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.1472, |
| "step": 906 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.736, |
| "step": 907 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8443, |
| "step": 908 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8938, |
| "step": 909 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7457, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.9263, |
| "step": 911 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.9717, |
| "step": 912 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.856, |
| "step": 913 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8276, |
| "step": 914 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7799, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8549, |
| "step": 916 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7338, |
| "step": 917 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7271, |
| "step": 918 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8964, |
| "step": 919 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.9394, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7309, |
| "step": 921 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8603, |
| "step": 922 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.776, |
| "step": 923 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.5931, |
| "step": 924 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7017, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.9707, |
| "step": 926 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7626, |
| "step": 927 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8182, |
| "step": 928 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.5435, |
| "step": 929 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6299, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7159, |
| "step": 931 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.9819, |
| "step": 932 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8233, |
| "step": 933 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7573, |
| "step": 934 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8456, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7595, |
| "step": 936 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7223, |
| "step": 937 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7176, |
| "step": 938 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8873, |
| "step": 939 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7002, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6593, |
| "step": 941 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6588, |
| "step": 942 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7715, |
| "step": 943 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6065, |
| "step": 944 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6518, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7139, |
| "step": 946 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8269, |
| "step": 947 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6744, |
| "step": 948 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.9748, |
| "step": 949 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8436, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.587, |
| "step": 951 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6966, |
| "step": 952 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5487, |
| "step": 953 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7167, |
| "step": 954 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.9755, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7621, |
| "step": 956 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6022, |
| "step": 957 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.9283, |
| "step": 958 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5495, |
| "step": 959 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7548, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7022, |
| "step": 961 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7441, |
| "step": 962 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6605, |
| "step": 963 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.8305, |
| "step": 964 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 1.118, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.9325, |
| "step": 966 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7019, |
| "step": 967 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.787, |
| "step": 968 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7185, |
| "step": 969 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 1.2544, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 1.0763, |
| "step": 971 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.828, |
| "step": 972 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5774, |
| "step": 973 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5799, |
| "step": 974 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5827, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5796, |
| "step": 976 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.8542, |
| "step": 977 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 1.1648, |
| "step": 978 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4692, |
| "step": 979 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6877, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6807, |
| "step": 981 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8971, |
| "step": 982 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7504, |
| "step": 983 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.9374, |
| "step": 984 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.634, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7102, |
| "step": 986 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8151, |
| "step": 987 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 1.0117, |
| "step": 988 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8022, |
| "step": 989 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7131, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8999, |
| "step": 991 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8989, |
| "step": 992 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.618, |
| "step": 993 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6371, |
| "step": 994 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6816, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7382, |
| "step": 996 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8496, |
| "step": 997 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6922, |
| "step": 998 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5522, |
| "step": 999 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8282, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_loss": 0.7664738297462463, |
| "eval_runtime": 82.8968, |
| "eval_samples_per_second": 12.063, |
| "eval_steps_per_second": 6.032, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.56, |
| "mmlu_eval_accuracy": 0.5680395940467005, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, |
| "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": 0.09090909090909091, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.36363636363636365, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.07142857142857142, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_law": 0.38235294117647056, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.6406227096565087, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8444, |
| "step": 1001 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3956, |
| "step": 1002 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 1.1147, |
| "step": 1003 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6255, |
| "step": 1004 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8598, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6429, |
| "step": 1006 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6911, |
| "step": 1007 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7242, |
| "step": 1008 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6417, |
| "step": 1009 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.9359, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6635, |
| "step": 1011 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7047, |
| "step": 1012 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5438, |
| "step": 1013 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8008, |
| "step": 1014 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8163, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6284, |
| "step": 1016 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7703, |
| "step": 1017 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9294, |
| "step": 1018 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9036, |
| "step": 1019 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6277, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8175, |
| "step": 1021 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8527, |
| "step": 1022 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7374, |
| "step": 1023 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6697, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9748, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8438, |
| "step": 1026 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 1.0555, |
| "step": 1027 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.5869, |
| "step": 1028 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6569, |
| "step": 1029 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8185, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8111, |
| "step": 1031 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8657, |
| "step": 1032 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9645, |
| "step": 1033 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6885, |
| "step": 1034 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.9432, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.5565, |
| "step": 1036 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.5873, |
| "step": 1037 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8399, |
| "step": 1038 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7405, |
| "step": 1039 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7144, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8081, |
| "step": 1041 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6644, |
| "step": 1042 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.769, |
| "step": 1043 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6424, |
| "step": 1044 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 1.1516, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4286, |
| "step": 1046 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 1.0178, |
| "step": 1047 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.794, |
| "step": 1048 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8201, |
| "step": 1049 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6548, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7257, |
| "step": 1051 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5961, |
| "step": 1052 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8494, |
| "step": 1053 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7698, |
| "step": 1054 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5891, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6478, |
| "step": 1056 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.9793, |
| "step": 1057 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.4977, |
| "step": 1058 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6033, |
| "step": 1059 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7309, |
| "step": 1061 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 1.1113, |
| "step": 1062 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8072, |
| "step": 1063 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8315, |
| "step": 1064 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.9742, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.738, |
| "step": 1066 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6252, |
| "step": 1067 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5874, |
| "step": 1068 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7784, |
| "step": 1069 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 1.0135, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.9033, |
| "step": 1071 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6198, |
| "step": 1072 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8888, |
| "step": 1073 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8216, |
| "step": 1074 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7485, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7536, |
| "step": 1076 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.9004, |
| "step": 1077 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8388, |
| "step": 1078 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7983, |
| "step": 1079 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7184, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7404, |
| "step": 1081 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7025, |
| "step": 1082 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7985, |
| "step": 1083 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.854, |
| "step": 1084 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8476, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7515, |
| "step": 1086 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8254, |
| "step": 1087 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.9011, |
| "step": 1088 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6717, |
| "step": 1089 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.9705, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.9081, |
| "step": 1091 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 1.1384, |
| "step": 1092 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7289, |
| "step": 1093 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8871, |
| "step": 1094 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8542, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 1.0924, |
| "step": 1096 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7012, |
| "step": 1097 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8034, |
| "step": 1098 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7886, |
| "step": 1099 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8149, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8693, |
| "step": 1101 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 1.1006, |
| "step": 1102 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8262, |
| "step": 1103 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8043, |
| "step": 1104 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7822, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.5716, |
| "step": 1106 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.9173, |
| "step": 1107 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6354, |
| "step": 1108 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8665, |
| "step": 1109 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6939, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.9026, |
| "step": 1111 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7689, |
| "step": 1112 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6813, |
| "step": 1113 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.9875, |
| "step": 1114 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.831, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.61, |
| "step": 1116 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 1117 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.5594, |
| "step": 1118 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8151, |
| "step": 1119 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8597, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7591, |
| "step": 1121 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.777, |
| "step": 1122 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.9713, |
| "step": 1123 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7798, |
| "step": 1124 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 1.0202, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.716, |
| "step": 1126 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7293, |
| "step": 1127 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8491, |
| "step": 1128 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8255, |
| "step": 1129 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6905, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.9929, |
| "step": 1131 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7138, |
| "step": 1132 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7505, |
| "step": 1133 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7044, |
| "step": 1134 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8077, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7681, |
| "step": 1136 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.69, |
| "step": 1137 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.5508, |
| "step": 1138 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7423, |
| "step": 1139 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.556, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7653, |
| "step": 1141 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8008, |
| "step": 1142 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.5151, |
| "step": 1143 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7936, |
| "step": 1144 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.5705, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6211, |
| "step": 1146 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6109, |
| "step": 1147 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8355, |
| "step": 1148 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.9114, |
| "step": 1149 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8541, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.79, |
| "step": 1151 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7761, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.974, |
| "step": 1153 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7023, |
| "step": 1154 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.945, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4371, |
| "step": 1156 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7636, |
| "step": 1157 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8441, |
| "step": 1158 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6714, |
| "step": 1159 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.5947, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7405, |
| "step": 1161 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.8397, |
| "step": 1162 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6299, |
| "step": 1163 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7503, |
| "step": 1164 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.9472, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.5864, |
| "step": 1166 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.5782, |
| "step": 1167 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 1.1221, |
| "step": 1168 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.4995, |
| "step": 1169 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.57, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.8475, |
| "step": 1171 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6108, |
| "step": 1172 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.8776, |
| "step": 1173 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.9581, |
| "step": 1174 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.5297, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6639, |
| "step": 1176 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6564, |
| "step": 1177 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.4873, |
| "step": 1178 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6766, |
| "step": 1179 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8921, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8402, |
| "step": 1181 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.9247, |
| "step": 1182 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.9097, |
| "step": 1183 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8343, |
| "step": 1184 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.9645, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7769, |
| "step": 1186 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 1.0081, |
| "step": 1187 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7612, |
| "step": 1188 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6588, |
| "step": 1189 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6934, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7615, |
| "step": 1191 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 1192 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.67, |
| "step": 1193 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7805, |
| "step": 1194 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7156, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.666, |
| "step": 1196 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8581, |
| "step": 1197 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.9999, |
| "step": 1198 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.71, |
| "step": 1199 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.5967, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.67, |
| "eval_loss": 0.7641408443450928, |
| "eval_runtime": 82.914, |
| "eval_samples_per_second": 12.061, |
| "eval_steps_per_second": 6.03, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.67, |
| "mmlu_eval_accuracy": 0.5842121029509861, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.36363636363636365, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.6666666666666666, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.53125, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.33, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.6470588235294118, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.40588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5185185185185185, |
| "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": 1.5320068213009026, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7302, |
| "step": 1201 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7529, |
| "step": 1202 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7727, |
| "step": 1203 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.86, |
| "step": 1204 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7636, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7389, |
| "step": 1206 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.6758, |
| "step": 1207 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.664, |
| "step": 1208 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.9282, |
| "step": 1209 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7689, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7969, |
| "step": 1211 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7796, |
| "step": 1212 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.9702, |
| "step": 1213 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5672, |
| "step": 1214 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7377, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7183, |
| "step": 1216 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 1.0991, |
| "step": 1217 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8716, |
| "step": 1218 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8704, |
| "step": 1219 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.9228, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 1.0344, |
| "step": 1221 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5409, |
| "step": 1222 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 1.0712, |
| "step": 1223 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.546, |
| "step": 1224 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6392, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5817, |
| "step": 1226 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7669, |
| "step": 1227 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6865, |
| "step": 1228 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.501, |
| "step": 1229 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 1.0462, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8102, |
| "step": 1231 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.9663, |
| "step": 1232 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8491, |
| "step": 1233 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8199, |
| "step": 1234 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6845, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8145, |
| "step": 1236 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.9194, |
| "step": 1237 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6605, |
| "step": 1238 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8206, |
| "step": 1239 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7695, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6766, |
| "step": 1241 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7863, |
| "step": 1242 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8506, |
| "step": 1243 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8079, |
| "step": 1244 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7807, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.9115, |
| "step": 1246 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 1.0511, |
| "step": 1247 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8422, |
| "step": 1248 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.8199, |
| "step": 1249 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.8658, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6759, |
| "step": 1251 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.5732, |
| "step": 1252 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7558, |
| "step": 1253 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7876, |
| "step": 1254 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7982, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.5568, |
| "step": 1256 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6868, |
| "step": 1257 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7314, |
| "step": 1258 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 1.0657, |
| "step": 1259 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7123, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.78, |
| "step": 1261 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7292, |
| "step": 1262 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7889, |
| "step": 1263 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7212, |
| "step": 1264 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6678, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7044, |
| "step": 1266 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7401, |
| "step": 1267 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 1268 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.5394, |
| "step": 1269 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.9506, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.9098, |
| "step": 1271 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 1.0294, |
| "step": 1272 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.8086, |
| "step": 1273 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.9548, |
| "step": 1274 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6132, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7565, |
| "step": 1276 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.5131, |
| "step": 1277 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.8407, |
| "step": 1278 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6161, |
| "step": 1279 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7799, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6705, |
| "step": 1281 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7694, |
| "step": 1282 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.5894, |
| "step": 1283 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6241, |
| "step": 1284 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6625, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.9509, |
| "step": 1286 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6339, |
| "step": 1287 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.9656, |
| "step": 1288 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.4276, |
| "step": 1289 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6758, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8048, |
| "step": 1291 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7595, |
| "step": 1292 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6673, |
| "step": 1293 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.9202, |
| "step": 1294 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7749, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7844, |
| "step": 1296 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.668, |
| "step": 1297 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7629, |
| "step": 1298 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.9523, |
| "step": 1299 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6812, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.9166, |
| "step": 1301 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7941, |
| "step": 1302 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5727, |
| "step": 1303 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7959, |
| "step": 1304 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7619, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5163, |
| "step": 1306 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8782, |
| "step": 1307 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.717, |
| "step": 1308 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.6571, |
| "step": 1309 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.648, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.9448, |
| "step": 1311 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.6491, |
| "step": 1312 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8065, |
| "step": 1313 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.9544, |
| "step": 1314 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8259, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7838, |
| "step": 1316 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8349, |
| "step": 1317 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5241, |
| "step": 1318 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7542, |
| "step": 1319 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7634, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.5673, |
| "step": 1321 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8024, |
| "step": 1322 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8666, |
| "step": 1323 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7932, |
| "step": 1324 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8828, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6335, |
| "step": 1326 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.9178, |
| "step": 1327 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8979, |
| "step": 1328 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7502, |
| "step": 1329 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.729, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6377, |
| "step": 1331 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8053, |
| "step": 1332 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.925, |
| "step": 1333 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8753, |
| "step": 1334 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7441, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7546, |
| "step": 1336 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.9319, |
| "step": 1337 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.9168, |
| "step": 1338 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.9055, |
| "step": 1339 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 1.0107, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4885, |
| "step": 1341 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6251, |
| "step": 1342 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7646, |
| "step": 1343 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7667, |
| "step": 1344 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5622, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6551, |
| "step": 1346 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7888, |
| "step": 1347 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8018, |
| "step": 1348 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7234, |
| "step": 1349 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7192, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6805, |
| "step": 1351 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5704, |
| "step": 1352 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6582, |
| "step": 1353 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7541, |
| "step": 1354 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5848, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7052, |
| "step": 1356 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.9419, |
| "step": 1357 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6591, |
| "step": 1358 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6736, |
| "step": 1359 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8165, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8926, |
| "step": 1361 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4963, |
| "step": 1362 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7309, |
| "step": 1363 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6716, |
| "step": 1364 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8792, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8633, |
| "step": 1366 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.9542, |
| "step": 1367 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5599, |
| "step": 1368 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6201, |
| "step": 1369 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7743, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7274, |
| "step": 1371 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6956, |
| "step": 1372 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.753, |
| "step": 1373 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5674, |
| "step": 1374 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8603, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7857, |
| "step": 1376 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5662, |
| "step": 1377 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 1.1161, |
| "step": 1378 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 1.0328, |
| "step": 1379 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6009, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5383, |
| "step": 1381 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6027, |
| "step": 1382 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.758, |
| "step": 1383 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6886, |
| "step": 1384 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6627, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7875, |
| "step": 1386 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5165, |
| "step": 1387 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6973, |
| "step": 1388 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7099, |
| "step": 1389 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7366, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5924, |
| "step": 1391 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 1.0348, |
| "step": 1392 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.8325, |
| "step": 1393 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.9108, |
| "step": 1394 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.8542, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7018, |
| "step": 1396 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7217, |
| "step": 1397 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7276, |
| "step": 1398 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7208, |
| "step": 1399 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7058, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.78, |
| "eval_loss": 0.7619088292121887, |
| "eval_runtime": 82.9201, |
| "eval_samples_per_second": 12.06, |
| "eval_steps_per_second": 6.03, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.78, |
| "mmlu_eval_accuracy": 0.5875448109730831, |
| "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.6666666666666666, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.32, |
| "mmlu_eval_accuracy_nutrition": 0.6363636363636364, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.4, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.6773303024295727, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.9387, |
| "step": 1401 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5772, |
| "step": 1402 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6517, |
| "step": 1403 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7033, |
| "step": 1404 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.8149, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 1.0402, |
| "step": 1406 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5618, |
| "step": 1407 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6943, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6082, |
| "step": 1409 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7612, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.6271, |
| "step": 1411 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8189, |
| "step": 1412 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8002, |
| "step": 1413 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.851, |
| "step": 1414 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.9041, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.581, |
| "step": 1416 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8611, |
| "step": 1417 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7943, |
| "step": 1418 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 1.0152, |
| "step": 1419 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.677, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7031, |
| "step": 1421 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8084, |
| "step": 1422 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.5578, |
| "step": 1423 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.687, |
| "step": 1424 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.5722, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8661, |
| "step": 1426 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.6619, |
| "step": 1427 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.6821, |
| "step": 1428 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.5552, |
| "step": 1429 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7146, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8584, |
| "step": 1431 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6001, |
| "step": 1432 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.652, |
| "step": 1433 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6072, |
| "step": 1434 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8237, |
| "step": 1435 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7881, |
| "step": 1436 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6464, |
| "step": 1437 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7378, |
| "step": 1438 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.9659, |
| "step": 1439 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.853, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7535, |
| "step": 1441 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7429, |
| "step": 1442 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.4976, |
| "step": 1443 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7948, |
| "step": 1444 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8077, |
| "step": 1445 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8397, |
| "step": 1446 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.9652, |
| "step": 1447 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6096, |
| "step": 1448 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6318, |
| "step": 1449 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7394, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8281, |
| "step": 1451 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.664, |
| "step": 1452 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7196, |
| "step": 1453 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6666, |
| "step": 1454 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7363, |
| "step": 1455 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6106, |
| "step": 1456 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.702, |
| "step": 1457 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6459, |
| "step": 1458 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8532, |
| "step": 1459 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 1.1809, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 1.0057, |
| "step": 1461 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7079, |
| "step": 1462 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 1.0675, |
| "step": 1463 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7603, |
| "step": 1464 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 1.0187, |
| "step": 1465 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.4513, |
| "step": 1466 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7967, |
| "step": 1467 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8326, |
| "step": 1468 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6886, |
| "step": 1469 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8602, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6915, |
| "step": 1471 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5106, |
| "step": 1472 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.9294, |
| "step": 1473 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8184, |
| "step": 1474 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6636, |
| "step": 1475 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.904, |
| "step": 1476 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7625, |
| "step": 1477 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7641, |
| "step": 1478 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6556, |
| "step": 1479 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7957, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.9936, |
| "step": 1481 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6222, |
| "step": 1482 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.697, |
| "step": 1483 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8885, |
| "step": 1484 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7991, |
| "step": 1485 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6679, |
| "step": 1486 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8357, |
| "step": 1487 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8287, |
| "step": 1488 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.9932, |
| "step": 1489 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6997, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7282, |
| "step": 1491 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8577, |
| "step": 1492 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.694, |
| "step": 1493 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.9417, |
| "step": 1494 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7577, |
| "step": 1495 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7849, |
| "step": 1496 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7439, |
| "step": 1497 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7389, |
| "step": 1498 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.9858, |
| "step": 1499 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.5772, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 1.0648, |
| "step": 1501 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7284, |
| "step": 1502 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8727, |
| "step": 1503 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8943, |
| "step": 1504 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8771, |
| "step": 1505 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7253, |
| "step": 1506 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7251, |
| "step": 1507 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7746, |
| "step": 1508 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8174, |
| "step": 1509 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7988, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.588, |
| "step": 1511 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8134, |
| "step": 1512 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.9415, |
| "step": 1513 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7124, |
| "step": 1514 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.6484, |
| "step": 1515 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.741, |
| "step": 1516 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8875, |
| "step": 1517 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.9591, |
| "step": 1518 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.7403, |
| "step": 1519 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.974, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.6597, |
| "step": 1521 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.6797, |
| "step": 1522 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.766, |
| "step": 1523 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.7836, |
| "step": 1524 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.9622, |
| "step": 1525 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.952, |
| "step": 1526 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.6942, |
| "step": 1527 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.6609, |
| "step": 1528 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.7129, |
| "step": 1529 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.6405, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.7159, |
| "step": 1531 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 1.078, |
| "step": 1532 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.8455, |
| "step": 1533 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.6938, |
| "step": 1534 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.7245, |
| "step": 1535 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.968, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.998, |
| "step": 1537 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 1538 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.874, |
| "step": 1539 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.8737, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 1.2467, |
| "step": 1541 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.7766, |
| "step": 1542 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.7575, |
| "step": 1543 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.5839, |
| "step": 1544 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.6405, |
| "step": 1545 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.8635, |
| "step": 1546 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.7865, |
| "step": 1547 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.82, |
| "step": 1548 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.974, |
| "step": 1549 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.8379, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.7702, |
| "step": 1551 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.8027, |
| "step": 1552 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.9111, |
| "step": 1553 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.8377, |
| "step": 1554 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 1.0793, |
| "step": 1555 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.7552, |
| "step": 1556 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.8357, |
| "step": 1557 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.7364, |
| "step": 1558 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.9625, |
| "step": 1559 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.7047, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.717, |
| "step": 1561 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.8069, |
| "step": 1562 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.7397, |
| "step": 1563 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.8465, |
| "step": 1564 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 1.07, |
| "step": 1565 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.6538, |
| "step": 1566 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.6398, |
| "step": 1567 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.8773, |
| "step": 1568 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.71, |
| "step": 1569 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.7761, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 1.04, |
| "step": 1571 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.6757, |
| "step": 1572 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 1573 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5938, |
| "step": 1574 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5952, |
| "step": 1575 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 1.0112, |
| "step": 1576 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7106, |
| "step": 1577 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7636, |
| "step": 1578 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7009, |
| "step": 1579 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.6757, |
| "step": 1581 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7912, |
| "step": 1582 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.6387, |
| "step": 1583 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.6465, |
| "step": 1584 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7663, |
| "step": 1585 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7259, |
| "step": 1586 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.888, |
| "step": 1587 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5473, |
| "step": 1588 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7334, |
| "step": 1589 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.7637, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.8107, |
| "step": 1591 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.7586, |
| "step": 1592 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.7676, |
| "step": 1593 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.7604, |
| "step": 1594 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.8433, |
| "step": 1595 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.571, |
| "step": 1596 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.9946, |
| "step": 1597 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.9161, |
| "step": 1598 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.7137, |
| "step": 1599 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.6692, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.89, |
| "eval_loss": 0.7644941806793213, |
| "eval_runtime": 82.9094, |
| "eval_samples_per_second": 12.061, |
| "eval_steps_per_second": 6.031, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.89, |
| "mmlu_eval_accuracy": 0.5911541740969178, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.53125, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.5172413793103449, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.8260869565217391, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.7272727272727273, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5, |
| "mmlu_eval_accuracy_moral_scenarios": 0.3, |
| "mmlu_eval_accuracy_nutrition": 0.6363636363636364, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.3764705882352941, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.9473684210526315, |
| "mmlu_loss": 1.3567504320614643, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.9034, |
| "step": 1601 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.6831, |
| "step": 1602 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.8018, |
| "step": 1603 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.8494, |
| "step": 1604 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.5818, |
| "step": 1605 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.7758, |
| "step": 1606 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.9035, |
| "step": 1607 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.5992, |
| "step": 1608 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.8484, |
| "step": 1609 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.4902, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.9208, |
| "step": 1611 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.7765, |
| "step": 1612 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.8134, |
| "step": 1613 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.9028, |
| "step": 1614 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.7175, |
| "step": 1615 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.7742, |
| "step": 1616 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.8701, |
| "step": 1617 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.6987, |
| "step": 1618 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.7247, |
| "step": 1619 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.7984, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.8313, |
| "step": 1621 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.6596, |
| "step": 1622 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.7419, |
| "step": 1623 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.7895, |
| "step": 1624 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.8078, |
| "step": 1625 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.8647, |
| "step": 1626 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.8197, |
| "step": 1627 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.8227, |
| "step": 1628 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.7802, |
| "step": 1629 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.7387, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.7734, |
| "step": 1631 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.8649, |
| "step": 1632 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.5799, |
| "step": 1633 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 1.17, |
| "step": 1634 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4727, |
| "step": 1635 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.8212, |
| "step": 1636 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.7042, |
| "step": 1637 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.7503, |
| "step": 1638 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.6897, |
| "step": 1639 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.8217, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.867, |
| "step": 1641 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.9215, |
| "step": 1642 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.8906, |
| "step": 1643 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.7576, |
| "step": 1644 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.843, |
| "step": 1645 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.6203, |
| "step": 1646 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.913, |
| "step": 1647 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.6104, |
| "step": 1648 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 1.0285, |
| "step": 1649 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.7595, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.7528, |
| "step": 1651 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.8649, |
| "step": 1652 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.6223, |
| "step": 1653 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.9104, |
| "step": 1654 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.823, |
| "step": 1655 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.5288, |
| "step": 1656 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.8207, |
| "step": 1657 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.6964, |
| "step": 1658 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.6802, |
| "step": 1659 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.6804, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.5973, |
| "step": 1661 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.7287, |
| "step": 1662 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.7482, |
| "step": 1663 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.8343, |
| "step": 1664 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.9013, |
| "step": 1665 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.8209, |
| "step": 1666 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 1.0045, |
| "step": 1667 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.6911, |
| "step": 1668 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.9624, |
| "step": 1669 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 1.1277, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.9194, |
| "step": 1671 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.6413, |
| "step": 1672 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.8579, |
| "step": 1673 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.8584, |
| "step": 1674 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.869, |
| "step": 1675 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.8223, |
| "step": 1676 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.8025, |
| "step": 1677 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.6106, |
| "step": 1678 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.8249, |
| "step": 1679 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.6829, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.9618, |
| "step": 1681 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.8106, |
| "step": 1682 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.9443, |
| "step": 1683 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.6772, |
| "step": 1684 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.7641, |
| "step": 1685 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.7033, |
| "step": 1686 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.6336, |
| "step": 1687 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.8461, |
| "step": 1688 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.8544, |
| "step": 1689 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.8562, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.5742, |
| "step": 1691 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.6741, |
| "step": 1692 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.923, |
| "step": 1693 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.8644, |
| "step": 1694 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.9874, |
| "step": 1695 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.6392, |
| "step": 1696 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.8831, |
| "step": 1697 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.7903, |
| "step": 1698 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.6539, |
| "step": 1699 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.6821, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.9394, |
| "step": 1701 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.9345, |
| "step": 1702 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.7406, |
| "step": 1703 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.6362, |
| "step": 1704 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.7822, |
| "step": 1705 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.7829, |
| "step": 1706 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.8096, |
| "step": 1707 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.7158, |
| "step": 1708 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.807, |
| "step": 1709 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.707, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.8495, |
| "step": 1711 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.7741, |
| "step": 1712 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.6473, |
| "step": 1713 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.6496, |
| "step": 1714 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.667, |
| "step": 1715 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.7813, |
| "step": 1716 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.9751, |
| "step": 1717 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.8566, |
| "step": 1718 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.6343, |
| "step": 1719 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.7653, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.5696, |
| "step": 1721 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.8283, |
| "step": 1722 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.6727, |
| "step": 1723 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.7969, |
| "step": 1724 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.5256, |
| "step": 1725 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.8542, |
| "step": 1726 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.6109, |
| "step": 1727 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 1.0262, |
| "step": 1728 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.7867, |
| "step": 1729 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.8284, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.9407, |
| "step": 1731 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.6253, |
| "step": 1732 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.7532, |
| "step": 1733 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.7209, |
| "step": 1734 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.9331, |
| "step": 1735 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.7362, |
| "step": 1736 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.8912, |
| "step": 1737 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.6255, |
| "step": 1738 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.9594, |
| "step": 1739 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.7707, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.7462, |
| "step": 1741 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.7791, |
| "step": 1742 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.7915, |
| "step": 1743 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.7206, |
| "step": 1744 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.8485, |
| "step": 1745 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.6939, |
| "step": 1746 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.9452, |
| "step": 1747 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.5399, |
| "step": 1748 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 1.1072, |
| "step": 1749 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.9985, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 1.0336, |
| "step": 1751 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.6382, |
| "step": 1752 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.9021, |
| "step": 1753 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.9304, |
| "step": 1754 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.8629, |
| "step": 1755 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.9639, |
| "step": 1756 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 1.0808, |
| "step": 1757 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.6663, |
| "step": 1758 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.5794, |
| "step": 1759 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.8016, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.6496, |
| "step": 1761 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.7395, |
| "step": 1762 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 1.0013, |
| "step": 1763 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.8077, |
| "step": 1764 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.8097, |
| "step": 1765 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.5588, |
| "step": 1766 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.6772, |
| "step": 1767 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.7536, |
| "step": 1768 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.904, |
| "step": 1769 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 1.001, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.5343, |
| "step": 1771 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.6421, |
| "step": 1772 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.7673, |
| "step": 1773 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.6657, |
| "step": 1774 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.8877, |
| "step": 1775 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.811, |
| "step": 1776 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 1777 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.7269, |
| "step": 1778 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.628, |
| "step": 1779 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.7409, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.9004, |
| "step": 1781 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.8736, |
| "step": 1782 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.9484, |
| "step": 1783 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.7814, |
| "step": 1784 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.6734, |
| "step": 1785 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.6495, |
| "step": 1786 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.6911, |
| "step": 1787 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.9102, |
| "step": 1788 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5416, |
| "step": 1789 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7291, |
| "step": 1790 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6832, |
| "step": 1791 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.786, |
| "step": 1792 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6105, |
| "step": 1793 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7175, |
| "step": 1794 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6779, |
| "step": 1795 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0176, |
| "step": 1796 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9064, |
| "step": 1797 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6435, |
| "step": 1798 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6721, |
| "step": 1799 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5863, |
| "step": 1800 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_loss": 0.7605424523353577, |
| "eval_runtime": 82.874, |
| "eval_samples_per_second": 12.067, |
| "eval_steps_per_second": 6.033, |
| "step": 1800 |
| }, |
| { |
| "epoch": 1.0, |
| "mmlu_eval_accuracy": 0.5733867254484578, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.625, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "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": 0.5833333333333334, |
| "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": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.27, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.4857142857142857, |
| "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_law": 0.3588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.3740841173238605, |
| "step": 1800 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6415, |
| "step": 1801 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6116, |
| "step": 1802 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3728, |
| "step": 1803 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7551, |
| "step": 1804 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.795, |
| "step": 1805 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5354, |
| "step": 1806 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8167, |
| "step": 1807 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8366, |
| "step": 1808 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6188, |
| "step": 1809 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6037, |
| "step": 1810 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7611, |
| "step": 1811 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.56, |
| "step": 1812 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.487, |
| "step": 1813 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7775, |
| "step": 1814 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6353, |
| "step": 1815 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5401, |
| "step": 1816 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6311, |
| "step": 1817 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6317, |
| "step": 1818 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.826, |
| "step": 1819 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7145, |
| "step": 1820 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6528, |
| "step": 1821 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5848, |
| "step": 1822 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9186, |
| "step": 1823 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7049, |
| "step": 1824 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5847, |
| "step": 1825 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6974, |
| "step": 1826 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6891, |
| "step": 1827 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7963, |
| "step": 1828 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6417, |
| "step": 1829 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5824, |
| "step": 1830 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6214, |
| "step": 1831 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5916, |
| "step": 1832 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6352, |
| "step": 1833 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6905, |
| "step": 1834 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7473, |
| "step": 1835 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4169, |
| "step": 1836 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.686, |
| "step": 1837 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6263, |
| "step": 1838 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4849, |
| "step": 1839 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4826, |
| "step": 1840 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4875, |
| "step": 1841 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6793, |
| "step": 1842 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6085, |
| "step": 1843 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7038, |
| "step": 1844 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6792, |
| "step": 1845 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7697, |
| "step": 1846 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5491, |
| "step": 1847 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5715, |
| "step": 1848 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7716, |
| "step": 1849 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5649, |
| "step": 1850 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5606, |
| "step": 1851 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7341, |
| "step": 1852 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 1.059, |
| "step": 1853 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5753, |
| "step": 1854 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.444, |
| "step": 1855 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5915, |
| "step": 1856 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5192, |
| "step": 1857 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7481, |
| "step": 1858 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4275, |
| "step": 1859 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7553, |
| "step": 1860 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7987, |
| "step": 1861 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7056, |
| "step": 1862 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5369, |
| "step": 1863 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6994, |
| "step": 1864 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7582, |
| "step": 1865 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6785, |
| "step": 1866 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5466, |
| "step": 1867 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6638, |
| "step": 1868 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8879, |
| "step": 1869 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6135, |
| "step": 1870 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7023, |
| "step": 1871 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7896, |
| "step": 1872 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5315, |
| "step": 1873 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4776, |
| "step": 1874 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6828, |
| "step": 1875 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8458, |
| "step": 1876 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8765, |
| "step": 1877 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5436, |
| "step": 1878 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.656, |
| "step": 1879 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4227, |
| "step": 1880 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3611, |
| "step": 1881 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7414, |
| "step": 1882 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5935, |
| "step": 1883 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5872, |
| "step": 1884 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7494, |
| "step": 1885 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.747, |
| "step": 1886 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6998, |
| "step": 1887 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4567, |
| "step": 1888 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9365, |
| "step": 1889 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6859, |
| "step": 1890 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5804, |
| "step": 1891 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6457, |
| "step": 1892 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 1893 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5267, |
| "step": 1894 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5794, |
| "step": 1895 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.393, |
| "step": 1896 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7009, |
| "step": 1897 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8995, |
| "step": 1898 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7277, |
| "step": 1899 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6192, |
| "step": 1900 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8317, |
| "step": 1901 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.767, |
| "step": 1902 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5209, |
| "step": 1903 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6561, |
| "step": 1904 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6175, |
| "step": 1905 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5987, |
| "step": 1906 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.702, |
| "step": 1907 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5278, |
| "step": 1908 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6624, |
| "step": 1909 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 1910 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5979, |
| "step": 1911 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7562, |
| "step": 1912 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5166, |
| "step": 1913 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4779, |
| "step": 1914 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9734, |
| "step": 1915 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6579, |
| "step": 1916 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5571, |
| "step": 1917 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8005, |
| "step": 1918 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4726, |
| "step": 1919 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7866, |
| "step": 1920 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.532, |
| "step": 1921 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5056, |
| "step": 1922 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7492, |
| "step": 1923 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5716, |
| "step": 1924 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8462, |
| "step": 1925 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7794, |
| "step": 1926 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5183, |
| "step": 1927 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.824, |
| "step": 1928 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8351, |
| "step": 1929 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6027, |
| "step": 1930 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6184, |
| "step": 1931 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6904, |
| "step": 1932 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7388, |
| "step": 1933 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5173, |
| "step": 1934 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5439, |
| "step": 1935 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7503, |
| "step": 1936 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5972, |
| "step": 1937 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.674, |
| "step": 1938 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6399, |
| "step": 1939 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.623, |
| "step": 1940 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8815, |
| "step": 1941 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5395, |
| "step": 1942 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6533, |
| "step": 1943 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.603, |
| "step": 1944 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5429, |
| "step": 1945 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6215, |
| "step": 1946 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7849, |
| "step": 1947 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5865, |
| "step": 1948 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7335, |
| "step": 1949 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5907, |
| "step": 1950 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7171, |
| "step": 1951 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5122, |
| "step": 1952 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6113, |
| "step": 1953 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4836, |
| "step": 1954 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7171, |
| "step": 1955 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7028, |
| "step": 1956 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6924, |
| "step": 1957 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7085, |
| "step": 1958 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5797, |
| "step": 1959 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4352, |
| "step": 1960 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7144, |
| "step": 1961 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5979, |
| "step": 1962 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 1963 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5317, |
| "step": 1964 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7218, |
| "step": 1965 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7897, |
| "step": 1966 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8909, |
| "step": 1967 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4301, |
| "step": 1968 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.558, |
| "step": 1969 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5524, |
| "step": 1970 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6589, |
| "step": 1971 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5222, |
| "step": 1972 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5859, |
| "step": 1973 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 1974 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7688, |
| "step": 1975 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.645, |
| "step": 1976 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5645, |
| "step": 1977 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7942, |
| "step": 1978 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 1.1408, |
| "step": 1979 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8611, |
| "step": 1980 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6411, |
| "step": 1981 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7876, |
| "step": 1982 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5883, |
| "step": 1983 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7339, |
| "step": 1984 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9912, |
| "step": 1985 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6813, |
| "step": 1986 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7725, |
| "step": 1987 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8112, |
| "step": 1988 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5289, |
| "step": 1989 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6401, |
| "step": 1990 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7115, |
| "step": 1991 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5656, |
| "step": 1992 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5108, |
| "step": 1993 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6239, |
| "step": 1994 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7939, |
| "step": 1995 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6274, |
| "step": 1996 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5827, |
| "step": 1997 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7131, |
| "step": 1998 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6048, |
| "step": 1999 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7982, |
| "step": 2000 |
| }, |
| { |
| "epoch": 1.11, |
| "eval_loss": 0.7635790109634399, |
| "eval_runtime": 82.9368, |
| "eval_samples_per_second": 12.057, |
| "eval_steps_per_second": 6.029, |
| "step": 2000 |
| }, |
| { |
| "epoch": 1.11, |
| "mmlu_eval_accuracy": 0.5815857169920476, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "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": 0.18181818181818182, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, |
| "mmlu_eval_accuracy_formal_logic": 0.07142857142857142, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "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": 0.7674418604651163, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.34, |
| "mmlu_eval_accuracy_nutrition": 0.6060606060606061, |
| "mmlu_eval_accuracy_philosophy": 0.7647058823529411, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_law": 0.3588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, |
| "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5185185185185185, |
| "mmlu_eval_accuracy_sociology": 0.9545454545454546, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.4468448488503773, |
| "step": 2000 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 5391, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 1.9215084862026547e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
