| { |
| "best_metric": 0.7547169327735901, |
| "best_model_checkpoint": "experts/mistralic-expert-27/checkpoint-2000", |
| "epoch": 0.8403949856432523, |
| "eval_steps": 200, |
| "global_step": 3000, |
| "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": 0.8515, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0016, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9244, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8938, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9381, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9047, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9177, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.776, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.2272, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.621, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7208, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8402, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9664, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8599, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 1.0764, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.7397, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.9149, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6371, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8259, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.885, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.627, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.4987, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7001, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7281, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8388, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8198, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.2694, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7079, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7991, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7802, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9949, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8857, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7527, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7332, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6222, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0137, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.744, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7011, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8345, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6677, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.0183, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9737, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7512, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6606, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8773, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5466, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9622, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5968, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 1.3467, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6072, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7429, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.8461, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7983, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5969, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5919, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5897, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0288, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9288, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8752, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8349, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9769, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9669, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8334, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7836, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0791, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8543, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0215, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0372, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9485, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8693, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8803, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6534, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7149, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8411, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7099, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7047, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8679, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8017, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6613, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7222, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8748, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7316, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6947, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0667, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.9676, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.833, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8182, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6474, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 1.0239, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7249, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6326, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8058, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9325, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.956, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9449, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9865, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6998, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.1191, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7545, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8301, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7196, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8251, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7913, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7251, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.3406, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.683, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9042, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.765, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8744, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.806, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5341, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8243, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5607, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 1.3209, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6228, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7746, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7126, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7528, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7139, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6889, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6784, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9756, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7036, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.9542, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7897, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.582, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.776, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.1599, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9125, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.943, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6111, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.928, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6738, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7624, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9895, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7896, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7711, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6223, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8233, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6787, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7175, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6508, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8722, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5994, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8695, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7595, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7585, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6953, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8081, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8345, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0605, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 1.0322, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.977, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9086, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6311, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9419, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.8587, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7774, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.9576, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4736, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8697, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.88, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6901, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9841, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.631, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7931, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7471, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8889, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.728, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8334, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7263, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7756, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7522, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0049, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7769, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5807, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7693, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5979, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7248, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7929, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.1816, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8134, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8823, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8537, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7626, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8402, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6839, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9483, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5432, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.9971, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5966, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 1.0149, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6789, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6106, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7499, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7392, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7065, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7361, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8511, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6838, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.06, |
| "eval_loss": 0.7837238311767578, |
| "eval_runtime": 91.3154, |
| "eval_samples_per_second": 10.951, |
| "eval_steps_per_second": 5.476, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.06, |
| "mmlu_eval_accuracy": 0.5974198354378633, |
| "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": 0.5625, |
| "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": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "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": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "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": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.37, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.3310329847952094, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7127, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9879, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6891, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6018, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6216, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7711, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8027, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5749, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8574, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6781, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8819, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8168, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6749, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.4493, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7036, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.9759, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.689, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8269, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.538, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5219, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8282, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5979, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6459, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7986, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7707, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 1.0793, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8393, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6534, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8249, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7749, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.8302, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4986, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 1.0351, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7491, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9273, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9643, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7227, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7973, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9233, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9661, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6018, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8716, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8563, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7334, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.693, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7452, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7814, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.835, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6131, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6444, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8895, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9792, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7697, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9047, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7994, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8345, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9951, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6706, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9174, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5901, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8755, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6768, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.769, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.8326, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.9974, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7752, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5528, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7047, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6334, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8251, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7164, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6664, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7146, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7151, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8774, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.1954, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7017, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7481, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.865, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8642, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8344, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6054, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8368, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7402, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9823, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5446, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6242, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5789, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8399, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6731, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9359, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8102, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8048, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5286, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7049, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8625, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.869, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5592, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9752, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7281, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.816, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.9537, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7122, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6704, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7399, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8786, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 1.0965, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5692, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.821, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8087, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9395, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7753, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6052, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8376, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7273, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7126, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7054, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8974, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9198, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9435, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5791, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6828, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7419, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.624, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4686, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7024, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6697, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7239, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7786, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7014, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8402, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7476, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7146, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9206, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4866, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6831, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.9031, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7632, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5149, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7999, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6832, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6583, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6552, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6879, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6099, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8569, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5892, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.866, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5922, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7183, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7622, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5705, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4467, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.0294, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8717, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7025, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6844, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.9568, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7038, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8106, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7659, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 1.0471, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7604, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7535, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8791, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8012, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.704, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5535, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.8617, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7809, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.937, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7343, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5308, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8121, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6756, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7405, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7266, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.0368, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7609, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5561, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.717, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8601, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6434, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8277, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 1.0131, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8164, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6569, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8263, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6882, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6537, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7471, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7786, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9951, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7691, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5795, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.724, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7112, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6747, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5958, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.11, |
| "eval_loss": 0.7711280584335327, |
| "eval_runtime": 91.3662, |
| "eval_samples_per_second": 10.945, |
| "eval_steps_per_second": 5.472, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.11, |
| "mmlu_eval_accuracy": 0.595131757652337, |
| "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": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "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": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "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": 0.6153846153846154, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.6363636363636364, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.32, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.149632434192899, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6746, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7625, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6698, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.649, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8052, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8485, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.738, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8199, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.8153, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6654, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.979, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6864, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6224, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7189, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6949, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7218, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9522, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7875, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7953, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5779, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9045, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9686, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.1377, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7117, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0006, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7475, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7934, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7729, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8744, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8148, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9379, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9524, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.0701, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7522, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7548, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9857, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8919, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 1.3602, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9267, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.579, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8845, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7242, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6653, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.7769, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9578, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8898, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6872, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.562, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.0301, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 1.3176, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.713, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7054, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6514, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7632, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7948, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.728, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.802, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7215, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.893, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6396, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6465, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.606, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6327, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8119, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7538, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.9011, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6115, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7021, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7106, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6753, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7095, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6058, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7475, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5297, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7267, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7617, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.8619, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6803, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6286, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6644, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7286, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9773, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8785, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6305, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7557, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9292, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5896, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6759, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 1.0595, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5028, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.711, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8302, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.808, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5937, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8843, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7989, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8322, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5414, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9976, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7051, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7493, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7871, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3547, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8038, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.966, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9355, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8424, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6497, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9517, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5614, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5163, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5838, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8741, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7621, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.9078, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5451, |
| "step": 517 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6455, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8835, |
| "step": 519 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.0077, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7221, |
| "step": 521 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.2393, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8231, |
| "step": 523 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7431, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7848, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8377, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8671, |
| "step": 527 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6551, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7995, |
| "step": 529 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6129, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8026, |
| "step": 531 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6679, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7536, |
| "step": 533 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6447, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7832, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7263, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5212, |
| "step": 537 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9244, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6809, |
| "step": 539 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 1.3082, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7221, |
| "step": 541 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7937, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5129, |
| "step": 543 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7195, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6749, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9323, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4464, |
| "step": 547 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.8001, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7704, |
| "step": 549 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6777, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7822, |
| "step": 551 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.782, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.9071, |
| "step": 553 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.789, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7624, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6902, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6423, |
| "step": 557 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9266, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8236, |
| "step": 559 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7496, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7672, |
| "step": 561 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8415, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.846, |
| "step": 563 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7179, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8169, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6179, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7729, |
| "step": 567 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8952, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6799, |
| "step": 569 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8388, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5752, |
| "step": 571 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7472, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7052, |
| "step": 573 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9083, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6484, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9202, |
| "step": 577 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9371, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.627, |
| "step": 579 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.8092, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5173, |
| "step": 581 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7339, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.9927, |
| "step": 583 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7343, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7429, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6479, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 1.0072, |
| "step": 587 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.7925, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.536, |
| "step": 589 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.783, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7394, |
| "step": 591 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7646, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7734, |
| "step": 593 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9106, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6778, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9966, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9279, |
| "step": 597 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7395, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7229, |
| "step": 599 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7157, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.17, |
| "eval_loss": 0.7653947472572327, |
| "eval_runtime": 91.3738, |
| "eval_samples_per_second": 10.944, |
| "eval_steps_per_second": 5.472, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.17, |
| "mmlu_eval_accuracy": 0.5902782641364419, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "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": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.07142857142857142, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.6875, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163, |
| "mmlu_eval_accuracy_moral_disputes": 0.5, |
| "mmlu_eval_accuracy_moral_scenarios": 0.29, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6470588235294118, |
| "mmlu_eval_accuracy_prehistory": 0.4857142857142857, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.4117647058823529, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "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": 0.3888888888888889, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.267489625758664, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7074, |
| "step": 601 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6404, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6072, |
| "step": 603 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.776, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.611, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9403, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7496, |
| "step": 607 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8005, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9798, |
| "step": 609 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.1622, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8528, |
| "step": 611 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6685, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.9202, |
| "step": 613 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6941, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7712, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8914, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0155, |
| "step": 617 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6328, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.4421, |
| "step": 619 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7168, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8016, |
| "step": 621 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 1.0827, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8356, |
| "step": 623 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.659, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9116, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6402, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.2751, |
| "step": 627 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8762, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6956, |
| "step": 629 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9037, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5796, |
| "step": 631 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7262, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8831, |
| "step": 633 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8247, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7031, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6222, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 637 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7805, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8202, |
| "step": 639 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7964, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.0204, |
| "step": 641 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7417, |
| "step": 643 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8101, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5506, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6402, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 1.1561, |
| "step": 647 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9295, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7977, |
| "step": 649 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6146, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7837, |
| "step": 651 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.9755, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7649, |
| "step": 653 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.8047, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6296, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7258, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7367, |
| "step": 657 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6122, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6361, |
| "step": 659 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5481, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7254, |
| "step": 661 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.2157, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6693, |
| "step": 663 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7579, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7433, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7958, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6886, |
| "step": 667 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5909, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.735, |
| "step": 669 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0948, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8579, |
| "step": 671 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7997, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8616, |
| "step": 673 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5389, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6547, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7415, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8623, |
| "step": 677 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5213, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7119, |
| "step": 679 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8255, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7747, |
| "step": 681 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1123, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8869, |
| "step": 683 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1725, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6241, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7738, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1263, |
| "step": 687 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8678, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.8083, |
| "step": 689 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7931, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.0405, |
| "step": 691 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.9127, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7048, |
| "step": 693 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6928, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7478, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 1.1453, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7099, |
| "step": 697 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7424, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6368, |
| "step": 699 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9105, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7722, |
| "step": 701 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7649, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8586, |
| "step": 703 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7569, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8333, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8529, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6152, |
| "step": 707 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6218, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 1.1182, |
| "step": 709 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7355, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6721, |
| "step": 711 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6152, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8021, |
| "step": 713 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6551, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.744, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8329, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8458, |
| "step": 717 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9324, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9079, |
| "step": 719 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5804, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9841, |
| "step": 721 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9636, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9126, |
| "step": 723 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7151, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6924, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7408, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7634, |
| "step": 727 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7254, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9239, |
| "step": 729 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.9962, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5296, |
| "step": 731 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6538, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6049, |
| "step": 733 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.2081, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8019, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.5458, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.809, |
| "step": 737 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7407, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7946, |
| "step": 739 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9359, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.9991, |
| "step": 741 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6081, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7142, |
| "step": 743 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7502, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6188, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6792, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8758, |
| "step": 747 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6326, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8321, |
| "step": 749 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.2388, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8242, |
| "step": 751 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.005, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 753 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8128, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.831, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5827, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7678, |
| "step": 757 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7541, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8041, |
| "step": 759 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7432, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.8951, |
| "step": 761 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6636, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.843, |
| "step": 763 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7363, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 1.0638, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6643, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6466, |
| "step": 767 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8272, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7093, |
| "step": 769 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7332, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5535, |
| "step": 771 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9422, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7921, |
| "step": 773 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.716, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5059, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8241, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6965, |
| "step": 777 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4428, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9968, |
| "step": 779 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9867, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 1.2924, |
| "step": 781 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7345, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6361, |
| "step": 783 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.654, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6404, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9022, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7192, |
| "step": 787 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6271, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.69, |
| "step": 789 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5543, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6825, |
| "step": 791 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6361, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8491, |
| "step": 793 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6575, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7651, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.9229, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6503, |
| "step": 797 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8475, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.7125, |
| "step": 799 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6716, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.22, |
| "eval_loss": 0.7683338522911072, |
| "eval_runtime": 91.4409, |
| "eval_samples_per_second": 10.936, |
| "eval_steps_per_second": 5.468, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.22, |
| "mmlu_eval_accuracy": 0.5996096231169502, |
| "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": 0.7241379310344828, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "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": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "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": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "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": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.6956521739130435, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.32, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.4117647058823529, |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.5552010817098243, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 801 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5989, |
| "step": 802 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.8127, |
| "step": 803 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7845, |
| "step": 804 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8664, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.032, |
| "step": 806 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.664, |
| "step": 807 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6123, |
| "step": 808 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8501, |
| "step": 809 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6433, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6522, |
| "step": 811 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6728, |
| "step": 812 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6853, |
| "step": 813 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9993, |
| "step": 814 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8953, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7662, |
| "step": 816 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7194, |
| "step": 817 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8953, |
| "step": 818 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7143, |
| "step": 819 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8962, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6724, |
| "step": 821 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8984, |
| "step": 822 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5593, |
| "step": 823 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7077, |
| "step": 824 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7892, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8517, |
| "step": 826 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8124, |
| "step": 827 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 1.0824, |
| "step": 828 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6711, |
| "step": 829 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6421, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5863, |
| "step": 831 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.5743, |
| "step": 832 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.6218, |
| "step": 833 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9091, |
| "step": 834 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9506, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.7968, |
| "step": 836 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.8053, |
| "step": 837 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.9938, |
| "step": 838 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8676, |
| "step": 839 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8604, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5693, |
| "step": 841 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9743, |
| "step": 842 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9974, |
| "step": 843 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.531, |
| "step": 844 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6083, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.594, |
| "step": 846 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.1782, |
| "step": 847 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9615, |
| "step": 848 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9456, |
| "step": 849 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7375, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7654, |
| "step": 851 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9911, |
| "step": 852 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8615, |
| "step": 853 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7104, |
| "step": 854 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7116, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8094, |
| "step": 856 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9496, |
| "step": 857 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 1.0314, |
| "step": 858 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.605, |
| "step": 859 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5856, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8477, |
| "step": 861 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8679, |
| "step": 862 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8049, |
| "step": 863 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7652, |
| "step": 864 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5932, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6057, |
| "step": 866 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.803, |
| "step": 867 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7922, |
| "step": 868 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.8943, |
| "step": 869 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.6859, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.7846, |
| "step": 871 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9672, |
| "step": 872 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.9152, |
| "step": 873 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.848, |
| "step": 874 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9842, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8678, |
| "step": 876 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6134, |
| "step": 877 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6918, |
| "step": 878 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5717, |
| "step": 879 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7718, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7134, |
| "step": 881 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.1728, |
| "step": 882 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.913, |
| "step": 883 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.779, |
| "step": 884 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0638, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6272, |
| "step": 886 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5417, |
| "step": 887 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7229, |
| "step": 888 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5419, |
| "step": 889 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7964, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.686, |
| "step": 891 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7073, |
| "step": 892 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.8423, |
| "step": 893 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7349, |
| "step": 894 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7648, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6475, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.1131, |
| "step": 897 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 1.0033, |
| "step": 898 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7899, |
| "step": 899 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.868, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6742, |
| "step": 901 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9739, |
| "step": 902 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.5877, |
| "step": 903 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9477, |
| "step": 904 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.9342, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.6766, |
| "step": 906 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4548, |
| "step": 907 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.803, |
| "step": 908 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7811, |
| "step": 909 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.7205, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7532, |
| "step": 911 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.1386, |
| "step": 912 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7277, |
| "step": 913 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.643, |
| "step": 914 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6477, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8168, |
| "step": 916 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6483, |
| "step": 917 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9247, |
| "step": 918 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9149, |
| "step": 919 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8446, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8447, |
| "step": 921 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7336, |
| "step": 922 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8024, |
| "step": 923 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4926, |
| "step": 924 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.9098, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 1.0701, |
| "step": 926 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.704, |
| "step": 927 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8024, |
| "step": 928 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.5903, |
| "step": 929 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7855, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.947, |
| "step": 931 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6364, |
| "step": 932 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.778, |
| "step": 933 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4958, |
| "step": 934 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6898, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.695, |
| "step": 936 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.6967, |
| "step": 937 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7083, |
| "step": 938 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.842, |
| "step": 939 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8032, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8426, |
| "step": 941 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8608, |
| "step": 942 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.8572, |
| "step": 943 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.7389, |
| "step": 944 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.777, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8948, |
| "step": 946 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7159, |
| "step": 947 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8653, |
| "step": 948 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 949 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5906, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7516, |
| "step": 951 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6704, |
| "step": 952 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.3426, |
| "step": 953 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7007, |
| "step": 954 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7238, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.0506, |
| "step": 956 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8088, |
| "step": 957 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5377, |
| "step": 958 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7382, |
| "step": 959 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6837, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6682, |
| "step": 961 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6718, |
| "step": 962 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.405, |
| "step": 963 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6204, |
| "step": 964 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.671, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4728, |
| "step": 966 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6529, |
| "step": 967 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7026, |
| "step": 968 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.836, |
| "step": 969 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7985, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5906, |
| "step": 971 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 1.1345, |
| "step": 972 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.9188, |
| "step": 973 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8656, |
| "step": 974 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7944, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7229, |
| "step": 976 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6269, |
| "step": 977 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8268, |
| "step": 978 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.8836, |
| "step": 979 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.7128, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6174, |
| "step": 981 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5685, |
| "step": 982 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.647, |
| "step": 983 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7036, |
| "step": 984 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7851, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9488, |
| "step": 986 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8378, |
| "step": 987 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8402, |
| "step": 988 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9389, |
| "step": 989 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6176, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8237, |
| "step": 991 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6816, |
| "step": 992 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8002, |
| "step": 993 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6272, |
| "step": 994 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9922, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8507, |
| "step": 996 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8946, |
| "step": 997 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7328, |
| "step": 998 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7224, |
| "step": 999 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8249, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.28, |
| "eval_loss": 0.7592520713806152, |
| "eval_runtime": 91.4469, |
| "eval_samples_per_second": 10.935, |
| "eval_steps_per_second": 5.468, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.28, |
| "mmlu_eval_accuracy": 0.5855624113568011, |
| "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": 0.6896551724137931, |
| "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": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "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": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.3790606545483786, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.552, |
| "step": 1001 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6725, |
| "step": 1002 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7686, |
| "step": 1003 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8139, |
| "step": 1004 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.9294, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6221, |
| "step": 1006 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7523, |
| "step": 1007 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8402, |
| "step": 1008 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6475, |
| "step": 1009 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8218, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.8183, |
| "step": 1011 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7966, |
| "step": 1012 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 1.044, |
| "step": 1013 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5967, |
| "step": 1014 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.766, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5538, |
| "step": 1016 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.7015, |
| "step": 1017 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4749, |
| "step": 1018 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.821, |
| "step": 1019 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6149, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7798, |
| "step": 1021 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7494, |
| "step": 1022 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5728, |
| "step": 1023 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7222, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.141, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8858, |
| "step": 1026 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.0272, |
| "step": 1027 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9547, |
| "step": 1028 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.628, |
| "step": 1029 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7507, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7809, |
| "step": 1031 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.664, |
| "step": 1032 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8757, |
| "step": 1033 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5843, |
| "step": 1034 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7344, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9026, |
| "step": 1036 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.929, |
| "step": 1037 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7992, |
| "step": 1038 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6975, |
| "step": 1039 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4934, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8906, |
| "step": 1041 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8235, |
| "step": 1042 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7273, |
| "step": 1043 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.6494, |
| "step": 1044 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.1105, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 1.1272, |
| "step": 1046 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7007, |
| "step": 1047 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7703, |
| "step": 1048 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.732, |
| "step": 1049 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7608, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.8709, |
| "step": 1051 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.7752, |
| "step": 1052 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.9189, |
| "step": 1053 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6994, |
| "step": 1054 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8277, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6485, |
| "step": 1056 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8787, |
| "step": 1057 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5937, |
| "step": 1058 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8835, |
| "step": 1059 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7316, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7857, |
| "step": 1061 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8896, |
| "step": 1062 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7906, |
| "step": 1063 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9671, |
| "step": 1064 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6801, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5024, |
| "step": 1066 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8988, |
| "step": 1067 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7868, |
| "step": 1068 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7636, |
| "step": 1069 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9208, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7155, |
| "step": 1071 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7987, |
| "step": 1072 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6663, |
| "step": 1073 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.8094, |
| "step": 1074 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7287, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6826, |
| "step": 1076 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6902, |
| "step": 1077 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7612, |
| "step": 1078 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7896, |
| "step": 1079 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7998, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7112, |
| "step": 1081 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.505, |
| "step": 1082 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7186, |
| "step": 1083 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5452, |
| "step": 1084 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.9019, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 1.017, |
| "step": 1086 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.6172, |
| "step": 1087 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.7657, |
| "step": 1088 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.801, |
| "step": 1089 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5943, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.153, |
| "step": 1091 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5904, |
| "step": 1092 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5904, |
| "step": 1093 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7064, |
| "step": 1094 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6761, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6715, |
| "step": 1096 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7218, |
| "step": 1097 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8345, |
| "step": 1098 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7341, |
| "step": 1099 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8476, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6466, |
| "step": 1101 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6955, |
| "step": 1102 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9571, |
| "step": 1103 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.652, |
| "step": 1104 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9219, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7963, |
| "step": 1106 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.941, |
| "step": 1107 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8073, |
| "step": 1108 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9341, |
| "step": 1109 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8725, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5304, |
| "step": 1111 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7059, |
| "step": 1112 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6912, |
| "step": 1113 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8192, |
| "step": 1114 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6067, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.6474, |
| "step": 1116 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9105, |
| "step": 1117 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8472, |
| "step": 1118 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.0291, |
| "step": 1119 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.9686, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5913, |
| "step": 1121 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7594, |
| "step": 1122 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.7547, |
| "step": 1123 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 1.061, |
| "step": 1124 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7124, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6468, |
| "step": 1126 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.694, |
| "step": 1127 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9216, |
| "step": 1128 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9477, |
| "step": 1129 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7448, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.781, |
| "step": 1131 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7198, |
| "step": 1132 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8355, |
| "step": 1133 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7171, |
| "step": 1134 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6847, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8222, |
| "step": 1136 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6507, |
| "step": 1137 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7279, |
| "step": 1138 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8866, |
| "step": 1139 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9019, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9002, |
| "step": 1141 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7771, |
| "step": 1142 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9804, |
| "step": 1143 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.2389, |
| "step": 1144 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8247, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6745, |
| "step": 1146 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.555, |
| "step": 1147 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7471, |
| "step": 1148 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8144, |
| "step": 1149 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9233, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6868, |
| "step": 1151 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6514, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.9485, |
| "step": 1153 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7383, |
| "step": 1154 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.748, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6314, |
| "step": 1156 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7215, |
| "step": 1157 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 1.1403, |
| "step": 1158 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.7485, |
| "step": 1159 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.8363, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9214, |
| "step": 1161 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6929, |
| "step": 1162 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6725, |
| "step": 1163 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7494, |
| "step": 1164 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7087, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8887, |
| "step": 1166 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6416, |
| "step": 1167 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7489, |
| "step": 1168 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.77, |
| "step": 1169 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.796, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6206, |
| "step": 1171 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8002, |
| "step": 1172 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6849, |
| "step": 1173 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8694, |
| "step": 1174 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6111, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5592, |
| "step": 1176 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7278, |
| "step": 1177 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7031, |
| "step": 1178 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8665, |
| "step": 1179 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 1.0335, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8981, |
| "step": 1181 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6253, |
| "step": 1182 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 1.1058, |
| "step": 1183 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7876, |
| "step": 1184 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7832, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8095, |
| "step": 1186 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7478, |
| "step": 1187 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6722, |
| "step": 1188 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.9019, |
| "step": 1189 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.8746, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.5723, |
| "step": 1191 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.461, |
| "step": 1192 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6881, |
| "step": 1193 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.6982, |
| "step": 1194 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.7312, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8116, |
| "step": 1196 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7491, |
| "step": 1197 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5088, |
| "step": 1198 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7724, |
| "step": 1199 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8484, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.34, |
| "eval_loss": 0.7587723135948181, |
| "eval_runtime": 91.2566, |
| "eval_samples_per_second": 10.958, |
| "eval_steps_per_second": 5.479, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.34, |
| "mmlu_eval_accuracy": 0.595376849216796, |
| "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": 0.6896551724137931, |
| "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": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, |
| "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": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "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": 0.45454545454545453, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.29, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "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": 0.5833333333333334, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.358286567479761, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7993, |
| "step": 1201 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.2978, |
| "step": 1202 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.1393, |
| "step": 1203 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8681, |
| "step": 1204 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6696, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7722, |
| "step": 1206 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7295, |
| "step": 1207 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8314, |
| "step": 1208 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.769, |
| "step": 1209 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7002, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6315, |
| "step": 1211 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.6555, |
| "step": 1212 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7785, |
| "step": 1213 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8533, |
| "step": 1214 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7678, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7287, |
| "step": 1216 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5761, |
| "step": 1217 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5782, |
| "step": 1218 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7158, |
| "step": 1219 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7155, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7562, |
| "step": 1221 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.1387, |
| "step": 1222 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4348, |
| "step": 1223 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7921, |
| "step": 1224 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.65, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.5564, |
| "step": 1226 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 1.0329, |
| "step": 1227 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.772, |
| "step": 1228 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.8211, |
| "step": 1229 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7699, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.7914, |
| "step": 1231 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7374, |
| "step": 1232 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9654, |
| "step": 1233 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0583, |
| "step": 1234 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7629, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8505, |
| "step": 1236 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6189, |
| "step": 1237 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8853, |
| "step": 1238 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.764, |
| "step": 1239 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8702, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8117, |
| "step": 1241 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.932, |
| "step": 1242 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8377, |
| "step": 1243 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8209, |
| "step": 1244 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9794, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6946, |
| "step": 1246 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6477, |
| "step": 1247 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0626, |
| "step": 1248 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6545, |
| "step": 1249 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7822, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 1.0147, |
| "step": 1251 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.985, |
| "step": 1252 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8784, |
| "step": 1253 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4954, |
| "step": 1254 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8208, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7407, |
| "step": 1256 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8609, |
| "step": 1257 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7015, |
| "step": 1258 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5386, |
| "step": 1259 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.621, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.7619, |
| "step": 1261 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6456, |
| "step": 1262 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.709, |
| "step": 1263 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.8826, |
| "step": 1264 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.9888, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.986, |
| "step": 1266 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6564, |
| "step": 1267 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5294, |
| "step": 1268 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7207, |
| "step": 1269 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8746, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7005, |
| "step": 1271 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5602, |
| "step": 1272 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7792, |
| "step": 1273 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6572, |
| "step": 1274 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6797, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9142, |
| "step": 1276 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6843, |
| "step": 1277 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7543, |
| "step": 1278 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.77, |
| "step": 1279 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5221, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6851, |
| "step": 1281 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9185, |
| "step": 1282 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8634, |
| "step": 1283 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9415, |
| "step": 1284 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8496, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7637, |
| "step": 1286 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7228, |
| "step": 1287 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8413, |
| "step": 1288 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7082, |
| "step": 1289 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8749, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6908, |
| "step": 1291 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.819, |
| "step": 1292 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9855, |
| "step": 1293 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6183, |
| "step": 1294 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.9045, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8328, |
| "step": 1296 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7234, |
| "step": 1297 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6389, |
| "step": 1298 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.8613, |
| "step": 1299 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6095, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6333, |
| "step": 1301 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.7148, |
| "step": 1302 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7561, |
| "step": 1303 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8257, |
| "step": 1304 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6396, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.746, |
| "step": 1306 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6255, |
| "step": 1307 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4698, |
| "step": 1308 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6089, |
| "step": 1309 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7873, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8572, |
| "step": 1311 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8359, |
| "step": 1312 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4204, |
| "step": 1313 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7831, |
| "step": 1314 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8603, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6363, |
| "step": 1316 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7935, |
| "step": 1317 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8668, |
| "step": 1318 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9006, |
| "step": 1319 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.9704, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8005, |
| "step": 1321 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7223, |
| "step": 1322 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8968, |
| "step": 1323 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8935, |
| "step": 1324 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6483, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.6436, |
| "step": 1326 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8119, |
| "step": 1327 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8601, |
| "step": 1328 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.886, |
| "step": 1329 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7607, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.0093, |
| "step": 1331 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5804, |
| "step": 1332 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5635, |
| "step": 1333 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8009, |
| "step": 1334 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.7008, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 1.4542, |
| "step": 1336 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5373, |
| "step": 1337 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.8805, |
| "step": 1338 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.591, |
| "step": 1339 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8181, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8198, |
| "step": 1341 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5881, |
| "step": 1342 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8749, |
| "step": 1343 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.936, |
| "step": 1344 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7399, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6565, |
| "step": 1346 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.875, |
| "step": 1347 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7875, |
| "step": 1348 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7044, |
| "step": 1349 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.4295, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.745, |
| "step": 1351 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6656, |
| "step": 1352 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6348, |
| "step": 1353 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8962, |
| "step": 1354 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7303, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8027, |
| "step": 1356 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6525, |
| "step": 1357 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.705, |
| "step": 1358 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5459, |
| "step": 1359 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5559, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8749, |
| "step": 1361 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.2005, |
| "step": 1362 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7538, |
| "step": 1363 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.7038, |
| "step": 1364 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.8233, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5937, |
| "step": 1366 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6018, |
| "step": 1367 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6573, |
| "step": 1368 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.9257, |
| "step": 1369 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6851, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6514, |
| "step": 1371 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5193, |
| "step": 1372 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 1.4595, |
| "step": 1373 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6818, |
| "step": 1374 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6104, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7715, |
| "step": 1376 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.9542, |
| "step": 1377 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6765, |
| "step": 1378 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8361, |
| "step": 1379 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7025, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8366, |
| "step": 1381 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7079, |
| "step": 1382 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6756, |
| "step": 1383 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.5987, |
| "step": 1384 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7069, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6555, |
| "step": 1386 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7086, |
| "step": 1387 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.509, |
| "step": 1388 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6997, |
| "step": 1389 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7765, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8115, |
| "step": 1391 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7393, |
| "step": 1392 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.753, |
| "step": 1393 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6523, |
| "step": 1394 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8665, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7235, |
| "step": 1396 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6925, |
| "step": 1397 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7754, |
| "step": 1398 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.5876, |
| "step": 1399 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6701, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.39, |
| "eval_loss": 0.7606543898582458, |
| "eval_runtime": 91.2993, |
| "eval_samples_per_second": 10.953, |
| "eval_steps_per_second": 5.476, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.39, |
| "mmlu_eval_accuracy": 0.5855874777082944, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.125, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "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": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.686046511627907, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.36, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.6857142857142857, |
| "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, |
| "mmlu_eval_accuracy_professional_law": 0.4117647058823529, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.9545454545454546, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.2406748528191065, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7061, |
| "step": 1401 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6537, |
| "step": 1402 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8519, |
| "step": 1403 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6782, |
| "step": 1404 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.6131, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7058, |
| "step": 1406 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 1.0982, |
| "step": 1407 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.8705, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 1.0306, |
| "step": 1409 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.7703, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7901, |
| "step": 1411 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7444, |
| "step": 1412 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.6906, |
| "step": 1413 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7987, |
| "step": 1414 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.652, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.986, |
| "step": 1416 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7587, |
| "step": 1417 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.777, |
| "step": 1418 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.6487, |
| "step": 1419 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8019, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7025, |
| "step": 1421 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8354, |
| "step": 1422 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.6907, |
| "step": 1423 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8688, |
| "step": 1424 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.5869, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.9731, |
| "step": 1426 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7119, |
| "step": 1427 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7062, |
| "step": 1428 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.714, |
| "step": 1429 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 1.2432, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.6416, |
| "step": 1431 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7495, |
| "step": 1432 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.9611, |
| "step": 1433 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7343, |
| "step": 1434 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8011, |
| "step": 1435 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.5626, |
| "step": 1436 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4191, |
| "step": 1437 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8198, |
| "step": 1438 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8544, |
| "step": 1439 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7357, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.8431, |
| "step": 1441 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.6006, |
| "step": 1442 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.559, |
| "step": 1443 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7757, |
| "step": 1444 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.7111, |
| "step": 1445 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6789, |
| "step": 1446 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7668, |
| "step": 1447 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.5543, |
| "step": 1448 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8095, |
| "step": 1449 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7293, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7056, |
| "step": 1451 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7089, |
| "step": 1452 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8306, |
| "step": 1453 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8928, |
| "step": 1454 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8476, |
| "step": 1455 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7404, |
| "step": 1456 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.9528, |
| "step": 1457 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.85, |
| "step": 1458 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7252, |
| "step": 1459 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6299, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8698, |
| "step": 1461 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8021, |
| "step": 1462 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7791, |
| "step": 1463 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7604, |
| "step": 1464 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6063, |
| "step": 1465 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.9336, |
| "step": 1466 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6635, |
| "step": 1467 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 1.0151, |
| "step": 1468 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.911, |
| "step": 1469 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8036, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6156, |
| "step": 1471 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7366, |
| "step": 1472 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.9541, |
| "step": 1473 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 1.0236, |
| "step": 1474 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8172, |
| "step": 1475 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7454, |
| "step": 1476 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.5642, |
| "step": 1477 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.6124, |
| "step": 1478 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.9375, |
| "step": 1479 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.7211, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.8605, |
| "step": 1481 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8524, |
| "step": 1482 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7599, |
| "step": 1483 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7678, |
| "step": 1484 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8738, |
| "step": 1485 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.9267, |
| "step": 1486 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7784, |
| "step": 1487 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4685, |
| "step": 1488 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 1.1555, |
| "step": 1489 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8181, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6622, |
| "step": 1491 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.5818, |
| "step": 1492 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7313, |
| "step": 1493 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7137, |
| "step": 1494 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8606, |
| "step": 1495 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8784, |
| "step": 1496 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4853, |
| "step": 1497 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8424, |
| "step": 1498 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8878, |
| "step": 1499 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6286, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8793, |
| "step": 1501 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 1.0865, |
| "step": 1502 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7601, |
| "step": 1503 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.5933, |
| "step": 1504 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7641, |
| "step": 1505 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8542, |
| "step": 1506 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6614, |
| "step": 1507 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.7376, |
| "step": 1508 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6074, |
| "step": 1509 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.8445, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.5532, |
| "step": 1511 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.6887, |
| "step": 1512 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.875, |
| "step": 1513 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.869, |
| "step": 1514 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 1.0066, |
| "step": 1515 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.621, |
| "step": 1516 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 1.2734, |
| "step": 1517 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6531, |
| "step": 1518 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8748, |
| "step": 1519 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6397, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.926, |
| "step": 1521 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.748, |
| "step": 1522 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6582, |
| "step": 1523 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7718, |
| "step": 1524 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6313, |
| "step": 1525 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7337, |
| "step": 1526 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7701, |
| "step": 1527 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.548, |
| "step": 1528 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8146, |
| "step": 1529 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.654, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6167, |
| "step": 1531 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7398, |
| "step": 1532 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6911, |
| "step": 1533 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8596, |
| "step": 1534 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6333, |
| "step": 1535 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.9712, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8867, |
| "step": 1537 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8923, |
| "step": 1538 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6645, |
| "step": 1539 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.9824, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.5603, |
| "step": 1541 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6966, |
| "step": 1542 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7245, |
| "step": 1543 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.5627, |
| "step": 1544 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.8853, |
| "step": 1545 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.828, |
| "step": 1546 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.672, |
| "step": 1547 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.7949, |
| "step": 1548 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6874, |
| "step": 1549 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6691, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6731, |
| "step": 1551 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6298, |
| "step": 1552 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.6408, |
| "step": 1553 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7285, |
| "step": 1554 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9176, |
| "step": 1555 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.8929, |
| "step": 1556 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.6174, |
| "step": 1557 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.6638, |
| "step": 1558 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.793, |
| "step": 1559 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.8845, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.8588, |
| "step": 1561 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.5743, |
| "step": 1562 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.5905, |
| "step": 1563 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7798, |
| "step": 1564 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.796, |
| "step": 1565 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7868, |
| "step": 1566 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7659, |
| "step": 1567 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.745, |
| "step": 1568 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.6678, |
| "step": 1569 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9254, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9394, |
| "step": 1571 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9058, |
| "step": 1572 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7593, |
| "step": 1573 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.6573, |
| "step": 1574 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7545, |
| "step": 1575 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.621, |
| "step": 1576 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.9703, |
| "step": 1577 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7551, |
| "step": 1578 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7428, |
| "step": 1579 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7572, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.8903, |
| "step": 1581 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.8698, |
| "step": 1582 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.6777, |
| "step": 1583 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.8617, |
| "step": 1584 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.5994, |
| "step": 1585 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 1.1672, |
| "step": 1586 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7175, |
| "step": 1587 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.7591, |
| "step": 1588 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7576, |
| "step": 1589 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8101, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7161, |
| "step": 1591 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8309, |
| "step": 1592 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.9805, |
| "step": 1593 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.5597, |
| "step": 1594 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7209, |
| "step": 1595 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8612, |
| "step": 1596 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7736, |
| "step": 1597 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.707, |
| "step": 1598 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.596, |
| "step": 1599 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.763, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.45, |
| "eval_loss": 0.7634783983230591, |
| "eval_runtime": 91.3512, |
| "eval_samples_per_second": 10.947, |
| "eval_steps_per_second": 5.473, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.45, |
| "mmlu_eval_accuracy": 0.5861257346216938, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.125, |
| "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": 0.6363636363636364, |
| "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": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.27, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, |
| "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.3571751972869543, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8022, |
| "step": 1601 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7344, |
| "step": 1602 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.5221, |
| "step": 1603 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.9072, |
| "step": 1604 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8042, |
| "step": 1605 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.9127, |
| "step": 1606 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.5423, |
| "step": 1607 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6129, |
| "step": 1608 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7148, |
| "step": 1609 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6549, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6772, |
| "step": 1611 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7288, |
| "step": 1612 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.9472, |
| "step": 1613 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7271, |
| "step": 1614 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7048, |
| "step": 1615 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6166, |
| "step": 1616 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6216, |
| "step": 1617 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.6141, |
| "step": 1618 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8986, |
| "step": 1619 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.9313, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7519, |
| "step": 1621 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.7667, |
| "step": 1622 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 1.1411, |
| "step": 1623 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.8257, |
| "step": 1624 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.774, |
| "step": 1625 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6474, |
| "step": 1626 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8877, |
| "step": 1627 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9457, |
| "step": 1628 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8766, |
| "step": 1629 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7989, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6697, |
| "step": 1631 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6824, |
| "step": 1632 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7456, |
| "step": 1633 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8785, |
| "step": 1634 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8014, |
| "step": 1635 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6971, |
| "step": 1636 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6482, |
| "step": 1637 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.589, |
| "step": 1638 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7229, |
| "step": 1639 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7393, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8257, |
| "step": 1641 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7251, |
| "step": 1642 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7801, |
| "step": 1643 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8885, |
| "step": 1644 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7804, |
| "step": 1645 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7825, |
| "step": 1646 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8901, |
| "step": 1647 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9069, |
| "step": 1648 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.842, |
| "step": 1649 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.5946, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7756, |
| "step": 1651 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8786, |
| "step": 1652 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.8477, |
| "step": 1653 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9718, |
| "step": 1654 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9511, |
| "step": 1655 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7307, |
| "step": 1656 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9716, |
| "step": 1657 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.7274, |
| "step": 1658 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.9166, |
| "step": 1659 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 1.0036, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7457, |
| "step": 1661 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6718, |
| "step": 1662 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.8366, |
| "step": 1663 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5318, |
| "step": 1664 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7972, |
| "step": 1665 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5817, |
| "step": 1666 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7911, |
| "step": 1667 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7277, |
| "step": 1668 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6088, |
| "step": 1669 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7114, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.8194, |
| "step": 1671 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6716, |
| "step": 1672 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.8965, |
| "step": 1673 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7846, |
| "step": 1674 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7752, |
| "step": 1675 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 1.0487, |
| "step": 1676 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.8279, |
| "step": 1677 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7433, |
| "step": 1678 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5304, |
| "step": 1679 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.824, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5568, |
| "step": 1681 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.9079, |
| "step": 1682 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6286, |
| "step": 1683 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6298, |
| "step": 1684 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6712, |
| "step": 1685 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7066, |
| "step": 1686 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6609, |
| "step": 1687 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5824, |
| "step": 1688 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.667, |
| "step": 1689 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.8292, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6808, |
| "step": 1691 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5077, |
| "step": 1692 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 1.0128, |
| "step": 1693 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.6344, |
| "step": 1694 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.7885, |
| "step": 1695 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7602, |
| "step": 1696 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7265, |
| "step": 1697 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8088, |
| "step": 1698 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6756, |
| "step": 1699 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7957, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7733, |
| "step": 1701 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7661, |
| "step": 1702 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8455, |
| "step": 1703 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8565, |
| "step": 1704 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8444, |
| "step": 1705 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7478, |
| "step": 1706 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 1.2686, |
| "step": 1707 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7944, |
| "step": 1708 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4002, |
| "step": 1709 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.9041, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 1.0472, |
| "step": 1711 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8089, |
| "step": 1712 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6568, |
| "step": 1713 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8956, |
| "step": 1714 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6515, |
| "step": 1715 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7044, |
| "step": 1716 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8808, |
| "step": 1717 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.9463, |
| "step": 1718 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6471, |
| "step": 1719 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8389, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6148, |
| "step": 1721 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.6534, |
| "step": 1722 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.5906, |
| "step": 1723 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7967, |
| "step": 1724 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7427, |
| "step": 1725 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.627, |
| "step": 1726 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.7904, |
| "step": 1727 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.5737, |
| "step": 1728 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8265, |
| "step": 1729 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.8531, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4584, |
| "step": 1731 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.9181, |
| "step": 1732 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.685, |
| "step": 1733 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.718, |
| "step": 1734 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6735, |
| "step": 1735 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7619, |
| "step": 1736 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.5544, |
| "step": 1737 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 1738 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.543, |
| "step": 1739 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.5472, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6914, |
| "step": 1741 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8005, |
| "step": 1742 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6824, |
| "step": 1743 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.9142, |
| "step": 1744 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6842, |
| "step": 1745 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.9398, |
| "step": 1746 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6312, |
| "step": 1747 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7458, |
| "step": 1748 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.569, |
| "step": 1749 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.5507, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7954, |
| "step": 1751 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6042, |
| "step": 1752 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8462, |
| "step": 1753 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6703, |
| "step": 1754 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.9411, |
| "step": 1755 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8136, |
| "step": 1756 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8994, |
| "step": 1757 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.6908, |
| "step": 1758 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8346, |
| "step": 1759 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7257, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.8495, |
| "step": 1761 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7143, |
| "step": 1762 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.9639, |
| "step": 1763 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7449, |
| "step": 1764 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7003, |
| "step": 1765 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.705, |
| "step": 1766 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.7133, |
| "step": 1767 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.0357, |
| "step": 1768 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.0074, |
| "step": 1769 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8353, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.739, |
| "step": 1771 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.6385, |
| "step": 1772 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.2912, |
| "step": 1773 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.6711, |
| "step": 1774 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.0869, |
| "step": 1775 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7992, |
| "step": 1776 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7731, |
| "step": 1777 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7131, |
| "step": 1778 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.0448, |
| "step": 1779 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8659, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8926, |
| "step": 1781 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.6559, |
| "step": 1782 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.6973, |
| "step": 1783 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.118, |
| "step": 1784 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.729, |
| "step": 1785 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.637, |
| "step": 1786 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.719, |
| "step": 1787 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7396, |
| "step": 1788 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.9281, |
| "step": 1789 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8008, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7892, |
| "step": 1791 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.965, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.8576, |
| "step": 1793 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5958, |
| "step": 1794 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5759, |
| "step": 1795 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5354, |
| "step": 1796 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7849, |
| "step": 1797 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5876, |
| "step": 1798 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5072, |
| "step": 1799 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.7589, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.5, |
| "eval_loss": 0.760337769985199, |
| "eval_runtime": 91.1897, |
| "eval_samples_per_second": 10.966, |
| "eval_steps_per_second": 5.483, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.5, |
| "mmlu_eval_accuracy": 0.5748622064631259, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "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": 0.43902439024390244, |
| "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": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "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": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.29, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.38235294117647056, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5185185185185185, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.7368421052631579, |
| "mmlu_loss": 1.153787750060191, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5642, |
| "step": 1801 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 1.0193, |
| "step": 1802 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 1.0383, |
| "step": 1803 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8988, |
| "step": 1804 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7725, |
| "step": 1805 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.847, |
| "step": 1806 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6895, |
| "step": 1807 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.5211, |
| "step": 1808 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6876, |
| "step": 1809 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.9133, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8305, |
| "step": 1811 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7984, |
| "step": 1812 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6817, |
| "step": 1813 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8758, |
| "step": 1814 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7266, |
| "step": 1815 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6254, |
| "step": 1816 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.5157, |
| "step": 1817 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7535, |
| "step": 1818 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8322, |
| "step": 1819 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.5011, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6264, |
| "step": 1821 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7039, |
| "step": 1822 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8268, |
| "step": 1823 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7185, |
| "step": 1824 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6019, |
| "step": 1825 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7725, |
| "step": 1826 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.694, |
| "step": 1827 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7112, |
| "step": 1828 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.849, |
| "step": 1829 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8015, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7866, |
| "step": 1831 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8689, |
| "step": 1832 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6554, |
| "step": 1833 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7834, |
| "step": 1834 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.6349, |
| "step": 1835 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.8475, |
| "step": 1836 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7655, |
| "step": 1837 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.7218, |
| "step": 1838 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7619, |
| "step": 1839 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7978, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 1.3072, |
| "step": 1841 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8567, |
| "step": 1842 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7514, |
| "step": 1843 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7967, |
| "step": 1844 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6751, |
| "step": 1845 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8748, |
| "step": 1846 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.587, |
| "step": 1847 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.4418, |
| "step": 1848 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.9242, |
| "step": 1849 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.952, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.9603, |
| "step": 1851 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7161, |
| "step": 1852 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.5984, |
| "step": 1853 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.9094, |
| "step": 1854 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7623, |
| "step": 1855 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6426, |
| "step": 1856 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8381, |
| "step": 1857 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8639, |
| "step": 1858 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 1.0359, |
| "step": 1859 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.5515, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.5715, |
| "step": 1861 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 1.0494, |
| "step": 1862 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7088, |
| "step": 1863 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6379, |
| "step": 1864 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.5019, |
| "step": 1865 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.9137, |
| "step": 1866 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.7975, |
| "step": 1867 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.8245, |
| "step": 1868 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 1.3826, |
| "step": 1869 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 1.0986, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6268, |
| "step": 1871 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6342, |
| "step": 1872 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.6949, |
| "step": 1873 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.5793, |
| "step": 1874 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5814, |
| "step": 1875 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5231, |
| "step": 1876 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7788, |
| "step": 1877 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8654, |
| "step": 1878 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8793, |
| "step": 1879 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8175, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.732, |
| "step": 1881 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.9457, |
| "step": 1882 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7382, |
| "step": 1883 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8549, |
| "step": 1884 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8341, |
| "step": 1885 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7796, |
| "step": 1886 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6704, |
| "step": 1887 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.703, |
| "step": 1888 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6864, |
| "step": 1889 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7125, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6919, |
| "step": 1891 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7964, |
| "step": 1892 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5107, |
| "step": 1893 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6551, |
| "step": 1894 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8275, |
| "step": 1895 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7288, |
| "step": 1896 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7272, |
| "step": 1897 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.9544, |
| "step": 1898 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7776, |
| "step": 1899 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5247, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8944, |
| "step": 1901 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6986, |
| "step": 1902 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.8797, |
| "step": 1903 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5964, |
| "step": 1904 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.6702, |
| "step": 1905 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.9455, |
| "step": 1906 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7909, |
| "step": 1907 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.7214, |
| "step": 1908 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.9393, |
| "step": 1909 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.9573, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5434, |
| "step": 1911 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.763, |
| "step": 1912 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7691, |
| "step": 1913 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5564, |
| "step": 1914 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6202, |
| "step": 1915 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6602, |
| "step": 1916 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.8309, |
| "step": 1917 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6808, |
| "step": 1918 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.732, |
| "step": 1919 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.686, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7557, |
| "step": 1921 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7008, |
| "step": 1922 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7032, |
| "step": 1923 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.772, |
| "step": 1924 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5926, |
| "step": 1925 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6201, |
| "step": 1926 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 1927 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6901, |
| "step": 1928 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5761, |
| "step": 1929 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5735, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.8851, |
| "step": 1931 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.662, |
| "step": 1932 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7474, |
| "step": 1933 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7239, |
| "step": 1934 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7692, |
| "step": 1935 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7017, |
| "step": 1936 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.8345, |
| "step": 1937 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6148, |
| "step": 1938 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.6718, |
| "step": 1939 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7164, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.75, |
| "step": 1941 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.688, |
| "step": 1942 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4795, |
| "step": 1943 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7119, |
| "step": 1944 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.7677, |
| "step": 1945 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6147, |
| "step": 1946 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8072, |
| "step": 1947 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7937, |
| "step": 1948 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.9106, |
| "step": 1949 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8437, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.811, |
| "step": 1951 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8723, |
| "step": 1952 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.584, |
| "step": 1953 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7877, |
| "step": 1954 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8544, |
| "step": 1955 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.9139, |
| "step": 1956 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 1.1662, |
| "step": 1957 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.9457, |
| "step": 1958 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7551, |
| "step": 1959 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7289, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8665, |
| "step": 1961 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.5887, |
| "step": 1962 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6488, |
| "step": 1963 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8402, |
| "step": 1964 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 1.0034, |
| "step": 1965 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6688, |
| "step": 1966 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.712, |
| "step": 1967 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8344, |
| "step": 1968 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7148, |
| "step": 1969 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8158, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.767, |
| "step": 1971 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 1.2604, |
| "step": 1972 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6381, |
| "step": 1973 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6698, |
| "step": 1974 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.9755, |
| "step": 1975 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6752, |
| "step": 1976 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8847, |
| "step": 1977 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7669, |
| "step": 1978 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.8488, |
| "step": 1979 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.7186, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.9155, |
| "step": 1981 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7154, |
| "step": 1982 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4807, |
| "step": 1983 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5578, |
| "step": 1984 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.854, |
| "step": 1985 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 1.0284, |
| "step": 1986 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7866, |
| "step": 1987 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5041, |
| "step": 1988 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8031, |
| "step": 1989 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5446, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8468, |
| "step": 1991 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8612, |
| "step": 1992 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6295, |
| "step": 1993 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.754, |
| "step": 1994 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7321, |
| "step": 1995 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6553, |
| "step": 1996 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6448, |
| "step": 1997 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8073, |
| "step": 1998 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7388, |
| "step": 1999 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7321, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_loss": 0.7547169327735901, |
| "eval_runtime": 91.1045, |
| "eval_samples_per_second": 10.976, |
| "eval_steps_per_second": 5.488, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.56, |
| "mmlu_eval_accuracy": 0.5820074579956969, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, |
| "mmlu_eval_accuracy_college_biology": 0.6875, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.53125, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.6363636363636364, |
| "mmlu_eval_accuracy_management": 0.7272727272727273, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6470588235294118, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, |
| "mmlu_eval_accuracy_professional_law": 0.36470588235294116, |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.1355582054694702, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8017, |
| "step": 2001 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 1.0315, |
| "step": 2002 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7598, |
| "step": 2003 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8989, |
| "step": 2004 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7436, |
| "step": 2005 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7785, |
| "step": 2006 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8185, |
| "step": 2007 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7216, |
| "step": 2008 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7584, |
| "step": 2009 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6834, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7293, |
| "step": 2011 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6816, |
| "step": 2012 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.6946, |
| "step": 2013 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.7526, |
| "step": 2014 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.8041, |
| "step": 2015 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5529, |
| "step": 2016 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 1.0911, |
| "step": 2017 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 1.0868, |
| "step": 2018 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8094, |
| "step": 2019 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7376, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6409, |
| "step": 2021 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6549, |
| "step": 2022 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6947, |
| "step": 2023 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6202, |
| "step": 2024 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6624, |
| "step": 2025 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7151, |
| "step": 2026 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7334, |
| "step": 2027 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7312, |
| "step": 2028 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9528, |
| "step": 2029 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6649, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 2031 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.897, |
| "step": 2032 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9273, |
| "step": 2033 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 1.1978, |
| "step": 2034 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.534, |
| "step": 2035 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7802, |
| "step": 2036 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8946, |
| "step": 2037 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7839, |
| "step": 2038 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6388, |
| "step": 2039 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.5584, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8451, |
| "step": 2041 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6883, |
| "step": 2042 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7261, |
| "step": 2043 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9211, |
| "step": 2044 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6025, |
| "step": 2045 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.8946, |
| "step": 2046 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.806, |
| "step": 2047 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.6378, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7442, |
| "step": 2049 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.9729, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7615, |
| "step": 2051 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.7335, |
| "step": 2052 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8438, |
| "step": 2053 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8473, |
| "step": 2054 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8458, |
| "step": 2055 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8026, |
| "step": 2056 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7046, |
| "step": 2057 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6552, |
| "step": 2058 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.5021, |
| "step": 2059 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8582, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.9896, |
| "step": 2061 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8744, |
| "step": 2062 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.9558, |
| "step": 2063 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7503, |
| "step": 2064 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.5833, |
| "step": 2065 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6487, |
| "step": 2066 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7085, |
| "step": 2067 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6144, |
| "step": 2068 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6663, |
| "step": 2069 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7278, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 1.2791, |
| "step": 2071 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8181, |
| "step": 2072 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8522, |
| "step": 2073 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6319, |
| "step": 2074 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6021, |
| "step": 2075 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7327, |
| "step": 2076 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.917, |
| "step": 2077 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6251, |
| "step": 2078 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7305, |
| "step": 2079 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6256, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.574, |
| "step": 2081 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8796, |
| "step": 2082 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4998, |
| "step": 2083 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7552, |
| "step": 2084 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7845, |
| "step": 2085 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.7009, |
| "step": 2086 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.6677, |
| "step": 2087 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.8338, |
| "step": 2088 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.986, |
| "step": 2089 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7587, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8826, |
| "step": 2091 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6928, |
| "step": 2092 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6379, |
| "step": 2093 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.637, |
| "step": 2094 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6464, |
| "step": 2095 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5664, |
| "step": 2096 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5572, |
| "step": 2097 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5947, |
| "step": 2098 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8091, |
| "step": 2099 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.9163, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7663, |
| "step": 2101 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8748, |
| "step": 2102 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5589, |
| "step": 2103 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 1.1307, |
| "step": 2104 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6657, |
| "step": 2105 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.9973, |
| "step": 2106 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8885, |
| "step": 2107 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5856, |
| "step": 2108 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5579, |
| "step": 2109 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8722, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6948, |
| "step": 2111 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8412, |
| "step": 2112 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8311, |
| "step": 2113 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.9206, |
| "step": 2114 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7067, |
| "step": 2115 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5427, |
| "step": 2116 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7181, |
| "step": 2117 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6474, |
| "step": 2118 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6684, |
| "step": 2119 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.851, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.6535, |
| "step": 2121 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.8385, |
| "step": 2122 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3981, |
| "step": 2123 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.7207, |
| "step": 2124 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.837, |
| "step": 2125 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.9904, |
| "step": 2126 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6015, |
| "step": 2127 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 2128 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7122, |
| "step": 2129 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7918, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7153, |
| "step": 2131 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.5327, |
| "step": 2132 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4898, |
| "step": 2133 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.5885, |
| "step": 2134 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.5763, |
| "step": 2135 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 1.1243, |
| "step": 2136 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 1.0493, |
| "step": 2137 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8323, |
| "step": 2138 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8127, |
| "step": 2139 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6722, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8463, |
| "step": 2141 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 1.0171, |
| "step": 2142 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 1.0606, |
| "step": 2143 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6246, |
| "step": 2144 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7852, |
| "step": 2145 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7038, |
| "step": 2146 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.9737, |
| "step": 2147 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.5653, |
| "step": 2148 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6615, |
| "step": 2149 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4929, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7497, |
| "step": 2151 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6732, |
| "step": 2152 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6695, |
| "step": 2153 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.8614, |
| "step": 2154 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7395, |
| "step": 2155 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6907, |
| "step": 2156 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.6765, |
| "step": 2157 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7367, |
| "step": 2158 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.7189, |
| "step": 2159 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.5675, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6887, |
| "step": 2161 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6188, |
| "step": 2162 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 1.2986, |
| "step": 2163 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8073, |
| "step": 2164 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.496, |
| "step": 2165 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.5633, |
| "step": 2166 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6999, |
| "step": 2167 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7846, |
| "step": 2168 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7506, |
| "step": 2169 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6829, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 1.0657, |
| "step": 2171 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.9266, |
| "step": 2172 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7319, |
| "step": 2173 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8066, |
| "step": 2174 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8881, |
| "step": 2175 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7548, |
| "step": 2176 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6617, |
| "step": 2177 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.9168, |
| "step": 2178 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.608, |
| "step": 2179 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7542, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7797, |
| "step": 2181 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.7577, |
| "step": 2182 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6621, |
| "step": 2183 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6316, |
| "step": 2184 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 1.0896, |
| "step": 2185 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.704, |
| "step": 2186 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 1.0948, |
| "step": 2187 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8686, |
| "step": 2188 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6999, |
| "step": 2189 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.662, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.5723, |
| "step": 2191 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.704, |
| "step": 2192 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8778, |
| "step": 2193 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.9525, |
| "step": 2194 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.8278, |
| "step": 2195 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7953, |
| "step": 2196 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8234, |
| "step": 2197 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8384, |
| "step": 2198 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6044, |
| "step": 2199 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6389, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.62, |
| "eval_loss": 0.7563657164573669, |
| "eval_runtime": 91.4045, |
| "eval_samples_per_second": 10.94, |
| "eval_steps_per_second": 5.47, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.62, |
| "mmlu_eval_accuracy": 0.5826020549082809, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.5625, |
| "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": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "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": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.28, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6470588235294118, |
| "mmlu_eval_accuracy_prehistory": 0.6285714285714286, |
| "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": 0.6231884057971014, |
| "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": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7368421052631579, |
| "mmlu_loss": 1.2215042237980247, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.5522, |
| "step": 2201 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7791, |
| "step": 2202 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7212, |
| "step": 2203 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.707, |
| "step": 2204 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 1.0016, |
| "step": 2205 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.9329, |
| "step": 2206 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8271, |
| "step": 2207 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7952, |
| "step": 2208 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7006, |
| "step": 2209 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6713, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8508, |
| "step": 2211 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.5233, |
| "step": 2212 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7588, |
| "step": 2213 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8064, |
| "step": 2214 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.8623, |
| "step": 2215 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7301, |
| "step": 2216 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7093, |
| "step": 2217 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 1.3294, |
| "step": 2218 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6944, |
| "step": 2219 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6887, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.844, |
| "step": 2221 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.773, |
| "step": 2222 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6886, |
| "step": 2223 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4979, |
| "step": 2224 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.6316, |
| "step": 2225 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.9359, |
| "step": 2226 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.5983, |
| "step": 2227 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7013, |
| "step": 2228 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7404, |
| "step": 2229 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.5411, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.7451, |
| "step": 2231 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8263, |
| "step": 2232 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6482, |
| "step": 2233 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7222, |
| "step": 2234 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7032, |
| "step": 2235 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8442, |
| "step": 2236 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6567, |
| "step": 2237 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6364, |
| "step": 2238 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7432, |
| "step": 2239 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6094, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7313, |
| "step": 2241 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6054, |
| "step": 2242 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8277, |
| "step": 2243 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.9067, |
| "step": 2244 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7871, |
| "step": 2245 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8178, |
| "step": 2246 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8389, |
| "step": 2247 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7553, |
| "step": 2248 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8553, |
| "step": 2249 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7311, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6631, |
| "step": 2251 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.9109, |
| "step": 2252 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7242, |
| "step": 2253 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.9112, |
| "step": 2254 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.6063, |
| "step": 2255 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7715, |
| "step": 2256 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.5175, |
| "step": 2257 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7818, |
| "step": 2258 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8671, |
| "step": 2259 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7768, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3654, |
| "step": 2261 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8709, |
| "step": 2262 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.8799, |
| "step": 2263 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 1.2128, |
| "step": 2264 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.7254, |
| "step": 2265 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.5971, |
| "step": 2266 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6582, |
| "step": 2267 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8429, |
| "step": 2268 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.703, |
| "step": 2269 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 1.1426, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7958, |
| "step": 2271 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6682, |
| "step": 2272 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6697, |
| "step": 2273 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8747, |
| "step": 2274 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6009, |
| "step": 2275 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.9564, |
| "step": 2276 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 1.0858, |
| "step": 2277 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.711, |
| "step": 2278 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.9743, |
| "step": 2279 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7124, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.9975, |
| "step": 2281 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6864, |
| "step": 2282 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 1.0248, |
| "step": 2283 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.9004, |
| "step": 2284 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.5991, |
| "step": 2285 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 1.1939, |
| "step": 2286 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8208, |
| "step": 2287 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8265, |
| "step": 2288 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6785, |
| "step": 2289 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6545, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8645, |
| "step": 2291 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7824, |
| "step": 2292 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7636, |
| "step": 2293 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8188, |
| "step": 2294 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.5969, |
| "step": 2295 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.8113, |
| "step": 2296 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7089, |
| "step": 2297 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 1.1589, |
| "step": 2298 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 1.1851, |
| "step": 2299 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7016, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.7024, |
| "step": 2301 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6191, |
| "step": 2302 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6785, |
| "step": 2303 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6069, |
| "step": 2304 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.5976, |
| "step": 2305 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.8576, |
| "step": 2306 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7258, |
| "step": 2307 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7038, |
| "step": 2308 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.9179, |
| "step": 2309 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.464, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7847, |
| "step": 2311 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6322, |
| "step": 2312 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7427, |
| "step": 2313 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7923, |
| "step": 2314 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7644, |
| "step": 2315 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6974, |
| "step": 2316 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.8527, |
| "step": 2317 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6665, |
| "step": 2318 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 1.0779, |
| "step": 2319 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7974, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 1.2059, |
| "step": 2321 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7548, |
| "step": 2322 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.9205, |
| "step": 2323 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.5943, |
| "step": 2324 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.4771, |
| "step": 2325 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6195, |
| "step": 2326 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6307, |
| "step": 2327 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6425, |
| "step": 2328 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.7574, |
| "step": 2329 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6876, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.9513, |
| "step": 2331 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6448, |
| "step": 2332 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.742, |
| "step": 2333 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.9247, |
| "step": 2334 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.5039, |
| "step": 2335 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6449, |
| "step": 2336 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.847, |
| "step": 2337 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.6937, |
| "step": 2338 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6779, |
| "step": 2339 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6546, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.4663, |
| "step": 2341 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8397, |
| "step": 2342 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 1.0239, |
| "step": 2343 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8712, |
| "step": 2344 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8219, |
| "step": 2345 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8108, |
| "step": 2346 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7664, |
| "step": 2347 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7626, |
| "step": 2348 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.592, |
| "step": 2349 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.831, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7737, |
| "step": 2351 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8337, |
| "step": 2352 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7545, |
| "step": 2353 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.5448, |
| "step": 2354 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7259, |
| "step": 2355 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7829, |
| "step": 2356 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7474, |
| "step": 2357 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8029, |
| "step": 2358 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7423, |
| "step": 2359 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.5942, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.8439, |
| "step": 2361 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6984, |
| "step": 2362 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6368, |
| "step": 2363 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.5377, |
| "step": 2364 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.752, |
| "step": 2365 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.864, |
| "step": 2366 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6312, |
| "step": 2367 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.802, |
| "step": 2368 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7091, |
| "step": 2369 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7658, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.708, |
| "step": 2371 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7372, |
| "step": 2372 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.7961, |
| "step": 2373 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8454, |
| "step": 2374 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.5268, |
| "step": 2375 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.9048, |
| "step": 2376 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7116, |
| "step": 2377 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7304, |
| "step": 2378 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.6836, |
| "step": 2379 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7538, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7306, |
| "step": 2381 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.9091, |
| "step": 2382 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7369, |
| "step": 2383 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.5452, |
| "step": 2384 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7903, |
| "step": 2385 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7795, |
| "step": 2386 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.843, |
| "step": 2387 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8279, |
| "step": 2388 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7761, |
| "step": 2389 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7786, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7173, |
| "step": 2391 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8767, |
| "step": 2392 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8623, |
| "step": 2393 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.829, |
| "step": 2394 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.683, |
| "step": 2395 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8336, |
| "step": 2396 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8172, |
| "step": 2397 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7403, |
| "step": 2398 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.6563, |
| "step": 2399 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 1.8522, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.67, |
| "eval_loss": 0.7571815252304077, |
| "eval_runtime": 91.3363, |
| "eval_samples_per_second": 10.949, |
| "eval_steps_per_second": 5.474, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.67, |
| "mmlu_eval_accuracy": 0.5681354289903606, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.125, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "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": 0.34615384615384615, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.3125, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.5, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.6956521739130435, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.28, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.3941176470588235, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.9545454545454546, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7368421052631579, |
| "mmlu_loss": 1.126686368264044, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.805, |
| "step": 2401 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.68, |
| "step": 2402 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 1.1484, |
| "step": 2403 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.899, |
| "step": 2404 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.9028, |
| "step": 2405 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.633, |
| "step": 2406 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.8361, |
| "step": 2407 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.7135, |
| "step": 2408 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.6666, |
| "step": 2409 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8293, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7178, |
| "step": 2411 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7261, |
| "step": 2412 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7539, |
| "step": 2413 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5289, |
| "step": 2414 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6782, |
| "step": 2415 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7252, |
| "step": 2416 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8091, |
| "step": 2417 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8005, |
| "step": 2418 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8669, |
| "step": 2419 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7677, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6995, |
| "step": 2421 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.9417, |
| "step": 2422 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7283, |
| "step": 2423 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7084, |
| "step": 2424 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.9037, |
| "step": 2425 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7413, |
| "step": 2426 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.9289, |
| "step": 2427 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7222, |
| "step": 2428 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7824, |
| "step": 2429 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6167, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 2431 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6235, |
| "step": 2432 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.605, |
| "step": 2433 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8177, |
| "step": 2434 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7302, |
| "step": 2435 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7105, |
| "step": 2436 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 1.0482, |
| "step": 2437 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 1.1263, |
| "step": 2438 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5046, |
| "step": 2439 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6771, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.6466, |
| "step": 2441 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.8839, |
| "step": 2442 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.736, |
| "step": 2443 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.7751, |
| "step": 2444 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 1.0562, |
| "step": 2445 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6919, |
| "step": 2446 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 1.0881, |
| "step": 2447 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 1.0781, |
| "step": 2448 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7517, |
| "step": 2449 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.9344, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.5772, |
| "step": 2451 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7704, |
| "step": 2452 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8949, |
| "step": 2453 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.796, |
| "step": 2454 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6315, |
| "step": 2455 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8702, |
| "step": 2456 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8721, |
| "step": 2457 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.9889, |
| "step": 2458 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.687, |
| "step": 2459 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.613, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7001, |
| "step": 2461 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7077, |
| "step": 2462 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6304, |
| "step": 2463 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6094, |
| "step": 2464 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.781, |
| "step": 2465 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7933, |
| "step": 2466 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.588, |
| "step": 2467 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7817, |
| "step": 2468 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.5774, |
| "step": 2469 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8305, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8325, |
| "step": 2471 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7351, |
| "step": 2472 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7808, |
| "step": 2473 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6158, |
| "step": 2474 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.7638, |
| "step": 2475 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8675, |
| "step": 2476 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.5848, |
| "step": 2477 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.8513, |
| "step": 2478 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.6933, |
| "step": 2479 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.719, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.5601, |
| "step": 2481 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.8418, |
| "step": 2482 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6469, |
| "step": 2483 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.9286, |
| "step": 2484 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.8226, |
| "step": 2485 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7401, |
| "step": 2486 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.525, |
| "step": 2487 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6933, |
| "step": 2488 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6796, |
| "step": 2489 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7064, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7188, |
| "step": 2491 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.747, |
| "step": 2492 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7419, |
| "step": 2493 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.8267, |
| "step": 2494 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.5714, |
| "step": 2495 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.5506, |
| "step": 2496 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6133, |
| "step": 2497 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.9461, |
| "step": 2498 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.8866, |
| "step": 2499 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6088, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6175, |
| "step": 2501 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7424, |
| "step": 2502 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.9053, |
| "step": 2503 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7706, |
| "step": 2504 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7287, |
| "step": 2505 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6419, |
| "step": 2506 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 1.0895, |
| "step": 2507 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.9446, |
| "step": 2508 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7091, |
| "step": 2509 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.6207, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.8011, |
| "step": 2511 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7304, |
| "step": 2512 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7177, |
| "step": 2513 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 2514 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7138, |
| "step": 2515 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.7177, |
| "step": 2516 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7634, |
| "step": 2517 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 1.1124, |
| "step": 2518 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.9685, |
| "step": 2519 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.683, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6716, |
| "step": 2521 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 1.2181, |
| "step": 2522 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.8464, |
| "step": 2523 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.9175, |
| "step": 2524 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.836, |
| "step": 2525 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.648, |
| "step": 2526 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7535, |
| "step": 2527 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7726, |
| "step": 2528 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6119, |
| "step": 2529 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.5404, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6599, |
| "step": 2531 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.8547, |
| "step": 2532 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6001, |
| "step": 2533 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.8913, |
| "step": 2534 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7961, |
| "step": 2535 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.5459, |
| "step": 2536 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.935, |
| "step": 2537 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.9872, |
| "step": 2538 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7351, |
| "step": 2539 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7928, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 1.471, |
| "step": 2541 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7258, |
| "step": 2542 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.8, |
| "step": 2543 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.8448, |
| "step": 2544 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7711, |
| "step": 2545 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.9582, |
| "step": 2546 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6896, |
| "step": 2547 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6222, |
| "step": 2548 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 1.016, |
| "step": 2549 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.7552, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.6162, |
| "step": 2551 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.5994, |
| "step": 2552 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7636, |
| "step": 2553 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8764, |
| "step": 2554 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8378, |
| "step": 2555 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 1.0519, |
| "step": 2556 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6931, |
| "step": 2557 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6649, |
| "step": 2558 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6055, |
| "step": 2559 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6679, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7294, |
| "step": 2561 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8151, |
| "step": 2562 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5768, |
| "step": 2563 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6757, |
| "step": 2564 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6183, |
| "step": 2565 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.94, |
| "step": 2566 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7752, |
| "step": 2567 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8955, |
| "step": 2568 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 1.0498, |
| "step": 2569 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.9329, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7199, |
| "step": 2571 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.596, |
| "step": 2572 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7846, |
| "step": 2573 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8354, |
| "step": 2574 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.9311, |
| "step": 2575 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8795, |
| "step": 2576 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.8177, |
| "step": 2577 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7992, |
| "step": 2578 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.724, |
| "step": 2579 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 1.0655, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6926, |
| "step": 2581 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6707, |
| "step": 2582 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5537, |
| "step": 2583 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7054, |
| "step": 2584 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6989, |
| "step": 2585 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.7376, |
| "step": 2586 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.6225, |
| "step": 2587 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.738, |
| "step": 2588 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8593, |
| "step": 2589 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8773, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.827, |
| "step": 2591 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7705, |
| "step": 2592 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8033, |
| "step": 2593 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.617, |
| "step": 2594 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5893, |
| "step": 2595 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.946, |
| "step": 2596 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8673, |
| "step": 2597 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8062, |
| "step": 2598 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 1.0186, |
| "step": 2599 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.99, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.73, |
| "eval_loss": 0.7575185894966125, |
| "eval_runtime": 91.2378, |
| "eval_samples_per_second": 10.96, |
| "eval_steps_per_second": 5.48, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.73, |
| "mmlu_eval_accuracy": 0.5780307375109485, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, |
| "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": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.3125, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "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": 0.45454545454545453, |
| "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": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "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": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.6521739130434783, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.96, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.25, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.6285714285714286, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.3411764705882353, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "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": 0.7894736842105263, |
| "mmlu_loss": 1.1076063126053262, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.6863, |
| "step": 2601 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.9147, |
| "step": 2602 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 1.1818, |
| "step": 2603 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.6997, |
| "step": 2604 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7721, |
| "step": 2605 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8134, |
| "step": 2606 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.93, |
| "step": 2607 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8197, |
| "step": 2608 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.6309, |
| "step": 2609 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7308, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7151, |
| "step": 2611 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.6607, |
| "step": 2612 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7264, |
| "step": 2613 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.9095, |
| "step": 2614 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.6357, |
| "step": 2615 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8214, |
| "step": 2616 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.7806, |
| "step": 2617 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.689, |
| "step": 2618 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.774, |
| "step": 2619 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8243, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.8089, |
| "step": 2621 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5845, |
| "step": 2622 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.631, |
| "step": 2623 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.676, |
| "step": 2624 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7918, |
| "step": 2625 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7722, |
| "step": 2626 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7631, |
| "step": 2627 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 1.0392, |
| "step": 2628 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.9192, |
| "step": 2629 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 1.1118, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8029, |
| "step": 2631 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6555, |
| "step": 2632 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.78, |
| "step": 2633 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 1.0018, |
| "step": 2634 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7078, |
| "step": 2635 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6517, |
| "step": 2636 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8116, |
| "step": 2637 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6271, |
| "step": 2638 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7116, |
| "step": 2639 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7444, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6389, |
| "step": 2641 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.5785, |
| "step": 2642 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6188, |
| "step": 2643 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7751, |
| "step": 2644 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6626, |
| "step": 2645 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8956, |
| "step": 2646 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8264, |
| "step": 2647 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 1.1154, |
| "step": 2648 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.739, |
| "step": 2649 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.5685, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6656, |
| "step": 2651 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6447, |
| "step": 2652 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.811, |
| "step": 2653 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6222, |
| "step": 2654 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.815, |
| "step": 2655 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.8423, |
| "step": 2656 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.647, |
| "step": 2657 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 1.0308, |
| "step": 2658 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.7811, |
| "step": 2659 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6532, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7068, |
| "step": 2661 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6119, |
| "step": 2662 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8816, |
| "step": 2663 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4199, |
| "step": 2664 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8249, |
| "step": 2665 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7811, |
| "step": 2666 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6205, |
| "step": 2667 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5271, |
| "step": 2668 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.9576, |
| "step": 2669 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.9489, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6032, |
| "step": 2671 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8252, |
| "step": 2672 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5858, |
| "step": 2673 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7827, |
| "step": 2674 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5508, |
| "step": 2675 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6509, |
| "step": 2676 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 1.3898, |
| "step": 2677 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8785, |
| "step": 2678 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8603, |
| "step": 2679 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.6423, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8038, |
| "step": 2681 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8288, |
| "step": 2682 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.853, |
| "step": 2683 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.9176, |
| "step": 2684 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7609, |
| "step": 2685 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.683, |
| "step": 2686 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7221, |
| "step": 2687 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 1.0903, |
| "step": 2688 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7023, |
| "step": 2689 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7299, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.7339, |
| "step": 2691 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5666, |
| "step": 2692 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.9223, |
| "step": 2693 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.8732, |
| "step": 2694 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.652, |
| "step": 2695 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7492, |
| "step": 2696 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8646, |
| "step": 2697 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6614, |
| "step": 2698 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.9713, |
| "step": 2699 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7555, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8226, |
| "step": 2701 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6592, |
| "step": 2702 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6671, |
| "step": 2703 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8307, |
| "step": 2704 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7025, |
| "step": 2705 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7026, |
| "step": 2706 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 1.0298, |
| "step": 2707 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.9802, |
| "step": 2708 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 1.0015, |
| "step": 2709 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7341, |
| "step": 2710 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7344, |
| "step": 2711 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6071, |
| "step": 2712 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5255, |
| "step": 2713 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 1.016, |
| "step": 2714 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6802, |
| "step": 2715 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 1.0353, |
| "step": 2716 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7197, |
| "step": 2717 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5929, |
| "step": 2718 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6785, |
| "step": 2719 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7598, |
| "step": 2720 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6869, |
| "step": 2721 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8086, |
| "step": 2722 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7444, |
| "step": 2723 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7846, |
| "step": 2724 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7242, |
| "step": 2725 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7431, |
| "step": 2726 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.6745, |
| "step": 2727 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.8139, |
| "step": 2728 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.7185, |
| "step": 2729 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5248, |
| "step": 2730 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6873, |
| "step": 2731 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7981, |
| "step": 2732 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6804, |
| "step": 2733 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5838, |
| "step": 2734 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8216, |
| "step": 2735 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5961, |
| "step": 2736 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7644, |
| "step": 2737 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7209, |
| "step": 2738 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8624, |
| "step": 2739 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7849, |
| "step": 2740 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7566, |
| "step": 2741 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8479, |
| "step": 2742 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6297, |
| "step": 2743 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7997, |
| "step": 2744 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8213, |
| "step": 2745 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6947, |
| "step": 2746 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7439, |
| "step": 2747 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7862, |
| "step": 2748 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5428, |
| "step": 2749 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 1.418, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8284, |
| "step": 2751 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7701, |
| "step": 2752 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6617, |
| "step": 2753 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.67, |
| "step": 2754 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6748, |
| "step": 2755 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7765, |
| "step": 2756 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6851, |
| "step": 2757 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7057, |
| "step": 2758 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 1.0829, |
| "step": 2759 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.9808, |
| "step": 2760 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8318, |
| "step": 2761 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5886, |
| "step": 2762 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.7003, |
| "step": 2763 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5915, |
| "step": 2764 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.8089, |
| "step": 2765 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.6638, |
| "step": 2766 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.845, |
| "step": 2767 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6341, |
| "step": 2768 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.9023, |
| "step": 2769 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7687, |
| "step": 2770 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.9973, |
| "step": 2771 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5888, |
| "step": 2772 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.619, |
| "step": 2773 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.8734, |
| "step": 2774 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.8741, |
| "step": 2775 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6867, |
| "step": 2776 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 2777 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.678, |
| "step": 2778 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.996, |
| "step": 2779 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7752, |
| "step": 2780 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.8677, |
| "step": 2781 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6267, |
| "step": 2782 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6091, |
| "step": 2783 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6553, |
| "step": 2784 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.8714, |
| "step": 2785 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.9175, |
| "step": 2786 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7752, |
| "step": 2787 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7343, |
| "step": 2788 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7856, |
| "step": 2789 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7041, |
| "step": 2790 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.608, |
| "step": 2791 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5186, |
| "step": 2792 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7167, |
| "step": 2793 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6674, |
| "step": 2794 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6823, |
| "step": 2795 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.797, |
| "step": 2796 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.9534, |
| "step": 2797 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6489, |
| "step": 2798 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.843, |
| "step": 2799 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7413, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.78, |
| "eval_loss": 0.7597759366035461, |
| "eval_runtime": 91.3258, |
| "eval_samples_per_second": 10.95, |
| "eval_steps_per_second": 5.475, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.78, |
| "mmlu_eval_accuracy": 0.5651813012550893, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.125, |
| "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, |
| "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": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.3125, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "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": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.7272727272727273, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.25, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5925925925925926, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.6842105263157895, |
| "mmlu_loss": 1.1731486964007893, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.6538, |
| "step": 2801 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 2802 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 1.1318, |
| "step": 2803 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.6151, |
| "step": 2804 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7682, |
| "step": 2805 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.426, |
| "step": 2806 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7408, |
| "step": 2807 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.9768, |
| "step": 2808 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4385, |
| "step": 2809 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.811, |
| "step": 2810 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.9669, |
| "step": 2811 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 1.0524, |
| "step": 2812 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7177, |
| "step": 2813 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7233, |
| "step": 2814 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8689, |
| "step": 2815 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.9211, |
| "step": 2816 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.5655, |
| "step": 2817 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7111, |
| "step": 2818 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8235, |
| "step": 2819 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.483, |
| "step": 2820 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8211, |
| "step": 2821 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7125, |
| "step": 2822 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.6244, |
| "step": 2823 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7261, |
| "step": 2824 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7496, |
| "step": 2825 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.6995, |
| "step": 2826 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 1.3807, |
| "step": 2827 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 1.2601, |
| "step": 2828 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.608, |
| "step": 2829 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7543, |
| "step": 2830 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7647, |
| "step": 2831 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.7637, |
| "step": 2832 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 1.0041, |
| "step": 2833 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.6474, |
| "step": 2834 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.8472, |
| "step": 2835 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.541, |
| "step": 2836 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.9504, |
| "step": 2837 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.963, |
| "step": 2838 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8447, |
| "step": 2839 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6907, |
| "step": 2840 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8845, |
| "step": 2841 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6403, |
| "step": 2842 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7858, |
| "step": 2843 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8926, |
| "step": 2844 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6094, |
| "step": 2845 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7943, |
| "step": 2846 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7317, |
| "step": 2847 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7912, |
| "step": 2848 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7113, |
| "step": 2849 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8171, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6974, |
| "step": 2851 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6092, |
| "step": 2852 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7501, |
| "step": 2853 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7203, |
| "step": 2854 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.775, |
| "step": 2855 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6861, |
| "step": 2856 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.749, |
| "step": 2857 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 1.1374, |
| "step": 2858 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.9169, |
| "step": 2859 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8474, |
| "step": 2860 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7508, |
| "step": 2861 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7594, |
| "step": 2862 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7499, |
| "step": 2863 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 1.0716, |
| "step": 2864 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6006, |
| "step": 2865 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.8419, |
| "step": 2866 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7708, |
| "step": 2867 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6879, |
| "step": 2868 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7209, |
| "step": 2869 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.7885, |
| "step": 2870 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6373, |
| "step": 2871 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.501, |
| "step": 2872 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 1.0381, |
| "step": 2873 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 1.0052, |
| "step": 2874 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 1.1929, |
| "step": 2875 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8557, |
| "step": 2876 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7761, |
| "step": 2877 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.9506, |
| "step": 2878 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8873, |
| "step": 2879 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.943, |
| "step": 2880 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6318, |
| "step": 2881 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6701, |
| "step": 2882 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8235, |
| "step": 2883 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7881, |
| "step": 2884 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.645, |
| "step": 2885 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8086, |
| "step": 2886 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7192, |
| "step": 2887 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.9269, |
| "step": 2888 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7067, |
| "step": 2889 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7244, |
| "step": 2890 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8922, |
| "step": 2891 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.9388, |
| "step": 2892 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8357, |
| "step": 2893 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7249, |
| "step": 2894 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6074, |
| "step": 2895 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8311, |
| "step": 2896 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6632, |
| "step": 2897 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.5425, |
| "step": 2898 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8008, |
| "step": 2899 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6882, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.6462, |
| "step": 2901 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4784, |
| "step": 2902 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 1.0214, |
| "step": 2903 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8039, |
| "step": 2904 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8517, |
| "step": 2905 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7433, |
| "step": 2906 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.9723, |
| "step": 2907 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.7788, |
| "step": 2908 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.8046, |
| "step": 2909 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.9176, |
| "step": 2910 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8242, |
| "step": 2911 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.9189, |
| "step": 2912 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8093, |
| "step": 2913 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8615, |
| "step": 2914 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 1.0575, |
| "step": 2915 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 1.1014, |
| "step": 2916 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7195, |
| "step": 2917 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5717, |
| "step": 2918 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7283, |
| "step": 2919 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6309, |
| "step": 2920 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.4775, |
| "step": 2921 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5092, |
| "step": 2922 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.9345, |
| "step": 2923 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7489, |
| "step": 2924 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8967, |
| "step": 2925 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7071, |
| "step": 2926 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6745, |
| "step": 2927 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.9299, |
| "step": 2928 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8703, |
| "step": 2929 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8701, |
| "step": 2930 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7004, |
| "step": 2931 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 1.073, |
| "step": 2932 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5934, |
| "step": 2933 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7682, |
| "step": 2934 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7607, |
| "step": 2935 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.53, |
| "step": 2936 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.4823, |
| "step": 2937 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.8349, |
| "step": 2938 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6513, |
| "step": 2939 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.6813, |
| "step": 2940 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5454, |
| "step": 2941 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.9061, |
| "step": 2942 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.976, |
| "step": 2943 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.7665, |
| "step": 2944 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5163, |
| "step": 2945 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6312, |
| "step": 2946 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.9431, |
| "step": 2947 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7861, |
| "step": 2948 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.873, |
| "step": 2949 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8246, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 1.0154, |
| "step": 2951 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.612, |
| "step": 2952 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8523, |
| "step": 2953 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8405, |
| "step": 2954 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6062, |
| "step": 2955 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8244, |
| "step": 2956 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7715, |
| "step": 2957 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6319, |
| "step": 2958 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7833, |
| "step": 2959 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8021, |
| "step": 2960 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.9222, |
| "step": 2961 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8747, |
| "step": 2962 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8136, |
| "step": 2963 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.9563, |
| "step": 2964 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.5417, |
| "step": 2965 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7695, |
| "step": 2966 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6317, |
| "step": 2967 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6787, |
| "step": 2968 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7081, |
| "step": 2969 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.9994, |
| "step": 2970 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.693, |
| "step": 2971 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.8714, |
| "step": 2972 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6057, |
| "step": 2973 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7717, |
| "step": 2974 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7381, |
| "step": 2975 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.7956, |
| "step": 2976 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.481, |
| "step": 2977 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 1.0797, |
| "step": 2978 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.6421, |
| "step": 2979 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.4999, |
| "step": 2980 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8485, |
| "step": 2981 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7468, |
| "step": 2982 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7418, |
| "step": 2983 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7889, |
| "step": 2984 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8369, |
| "step": 2985 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8752, |
| "step": 2986 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.6974, |
| "step": 2987 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.561, |
| "step": 2988 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8189, |
| "step": 2989 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 1.337, |
| "step": 2990 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8616, |
| "step": 2991 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7246, |
| "step": 2992 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.6486, |
| "step": 2993 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.7902, |
| "step": 2994 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.777, |
| "step": 2995 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.5363, |
| "step": 2996 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.6483, |
| "step": 2997 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.6878, |
| "step": 2998 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.6585, |
| "step": 2999 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.8975, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.84, |
| "eval_loss": 0.7594294548034668, |
| "eval_runtime": 91.2658, |
| "eval_samples_per_second": 10.957, |
| "eval_steps_per_second": 5.479, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.84, |
| "mmlu_eval_accuracy": 0.5633183786636575, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.125, |
| "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.1875, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "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": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.24, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6470588235294118, |
| "mmlu_eval_accuracy_prehistory": 0.6857142857142857, |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_law": 0.37058823529411766, |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "mmlu_eval_accuracy_sociology": 0.9545454545454546, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.631578947368421, |
| "mmlu_loss": 1.0499589160162537, |
| "step": 3000 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 10707, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 3.2200985745516134e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
