| { |
| "best_metric": 1.090264916419983, |
| "best_model_checkpoint": "experts/mistralic-expert-7/checkpoint-400", |
| "epoch": 0.09054382887216343, |
| "eval_steps": 200, |
| "global_step": 400, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.1069, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.2094, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9918, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9152, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.6613, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9858, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.08, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0209, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9899, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0958, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.901, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8789, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.1249, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.474, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.2876, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.1123, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.193, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.2487, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.2474, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5941, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9055, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.2227, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9262, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.3076, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1696, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.3313, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2686, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.3981, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1538, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.163, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0271, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2048, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.104, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9422, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1554, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0124, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.3659, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.198, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2263, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.4718, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2161, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.805, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1312, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1027, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.4121, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2392, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8876, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2035, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8813, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2105, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9209, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.5502, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8156, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.3101, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.063, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8669, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2532, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1871, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9992, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.3905, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1335, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9591, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.1383, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0769, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0011, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9631, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0007, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7713, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0919, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.3098, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0577, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2181, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.066, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1163, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9696, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.996, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1752, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7712, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1081, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1301, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0869, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2455, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2583, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.3496, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0519, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9926, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.252, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0356, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0417, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8122, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0815, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0187, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2092, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2072, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2845, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0897, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.087, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9407, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.1855, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2297, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2538, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2918, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.2886, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.3194, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.387, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.3296, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.023, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.3532, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.164, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.22, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1294, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.3594, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0209, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1782, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.4478, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9884, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.2223, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.2952, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0233, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9052, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0663, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9975, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1312, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0912, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1431, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0593, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9973, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.654, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9922, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.2856, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.2685, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9818, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9132, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1126, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0989, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0457, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9864, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8723, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1542, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.126, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8874, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9306, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1046, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.652, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1687, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9122, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.5094, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.3892, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.3784, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.2615, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.4813, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9087, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.2279, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.0277, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0921, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.329, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.3525, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9512, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.111, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0456, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9593, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1118, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9878, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7738, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9905, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2998, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9747, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.086, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9589, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.002, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.062, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1623, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.7756, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0481, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2943, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1221, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0195, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1282, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9557, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1124, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0878, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.3296, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.3528, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1276, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0291, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.6419, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2399, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2008, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0825, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9417, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2083, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1882, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9195, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1144, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.5104, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.3833, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.2207, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1321, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.3945, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.3293, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.05, |
| "eval_loss": 1.1023212671279907, |
| "eval_runtime": 184.3668, |
| "eval_samples_per_second": 5.424, |
| "eval_steps_per_second": 2.712, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.05, |
| "mmlu_eval_accuracy": 0.5754564885825776, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "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": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.28, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.6571428571428571, |
| "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, |
| "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7368421052631579, |
| "mmlu_loss": 1.4548778765039718, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0673, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.4929, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0702, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6805, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1698, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.3649, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7315, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0896, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2642, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2406, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0695, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2109, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0194, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.881, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1048, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0131, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1672, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.3431, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1419, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1728, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0923, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8012, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8913, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2599, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2398, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0219, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2604, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0559, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8025, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.5543, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8803, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0187, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9477, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.3309, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0872, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1225, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1212, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.2685, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0323, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1132, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9758, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.3243, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2639, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2757, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8589, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2434, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9988, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.926, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0675, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0737, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0925, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1135, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0306, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9488, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2267, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0911, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1731, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.245, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9823, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1964, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9566, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1863, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.194, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2929, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1267, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.056, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.3178, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.314, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2308, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1292, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9399, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2056, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0501, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1905, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9338, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.108, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.2321, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9377, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.025, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0819, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8704, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6804, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0458, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0824, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1325, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.1303, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0245, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0304, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9662, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9651, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8933, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.3649, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9489, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.4182, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1817, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1286, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9664, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0353, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8003, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1159, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.2123, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9077, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9209, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1074, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1486, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0098, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.209, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1762, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9637, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0725, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.5555, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8834, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.169, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9521, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.061, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0147, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9353, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.117, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.2968, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9233, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.2067, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.065, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.3042, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.1112, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0318, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.2159, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0696, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.3873, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0289, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9252, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.3741, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.029, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1885, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1669, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0341, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1607, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8112, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7137, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.2613, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.2306, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8955, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0075, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1575, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0755, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.203, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0073, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.154, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0882, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.3667, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9183, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0256, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1558, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7771, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9609, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0822, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0246, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9567, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.948, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9191, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.2471, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0428, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7941, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1061, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0724, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0029, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1191, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8988, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.905, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9681, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.995, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9311, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9495, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8281, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0885, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.3558, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0965, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0165, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.038, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0586, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0414, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.3588, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.2422, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.1097, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0915, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.1821, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.2693, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.2324, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.215, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.1658, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9563, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.1071, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8785, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0177, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.1641, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.1911, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0184, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.083, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.3112, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.083, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.09, |
| "eval_loss": 1.090264916419983, |
| "eval_runtime": 184.2447, |
| "eval_samples_per_second": 5.428, |
| "eval_steps_per_second": 2.714, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.09, |
| "mmlu_eval_accuracy": 0.5865762223797235, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.125, |
| "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6744186046511628, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.7692307692307693, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, |
| "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": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.31, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.6470588235294118, |
| "mmlu_eval_accuracy_prehistory": 0.6285714285714286, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.40588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "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": 0.6666666666666666, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.2847806720291666, |
| "step": 400 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 13251, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 1.0961197737128755e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
