{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 35.0, "eval_steps": 500, "global_step": 5600, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00625, "grad_norm": 4.96875, "learning_rate": 0.0, "loss": 1.8427, "step": 1 }, { "epoch": 0.0125, "grad_norm": 5.09375, "learning_rate": 3.1250000000000005e-08, "loss": 1.9223, "step": 2 }, { "epoch": 0.01875, "grad_norm": 4.875, "learning_rate": 6.250000000000001e-08, "loss": 1.8604, "step": 3 }, { "epoch": 0.025, "grad_norm": 5.0625, "learning_rate": 9.375e-08, "loss": 1.9811, "step": 4 }, { "epoch": 0.03125, "grad_norm": 5.0, "learning_rate": 1.2500000000000002e-07, "loss": 1.9454, "step": 5 }, { "epoch": 0.0375, "grad_norm": 4.875, "learning_rate": 1.5625e-07, "loss": 1.8717, "step": 6 }, { "epoch": 0.04375, "grad_norm": 4.9375, "learning_rate": 1.875e-07, "loss": 1.8809, "step": 7 }, { "epoch": 0.05, "grad_norm": 4.9375, "learning_rate": 2.1875000000000002e-07, "loss": 1.9136, "step": 8 }, { "epoch": 0.05625, "grad_norm": 4.75, "learning_rate": 2.5000000000000004e-07, "loss": 1.816, "step": 9 }, { "epoch": 0.0625, "grad_norm": 4.875, "learning_rate": 2.8125e-07, "loss": 1.9044, "step": 10 }, { "epoch": 0.06875, "grad_norm": 4.96875, "learning_rate": 3.125e-07, "loss": 1.9006, "step": 11 }, { "epoch": 0.075, "grad_norm": 4.875, "learning_rate": 3.4375000000000004e-07, "loss": 1.8707, "step": 12 }, { "epoch": 0.08125, "grad_norm": 4.84375, "learning_rate": 3.75e-07, "loss": 1.8739, "step": 13 }, { "epoch": 0.0875, "grad_norm": 4.9375, "learning_rate": 4.0625000000000003e-07, "loss": 1.9165, "step": 14 }, { "epoch": 0.09375, "grad_norm": 5.0, "learning_rate": 4.3750000000000005e-07, "loss": 1.9139, "step": 15 }, { "epoch": 0.1, "grad_norm": 5.125, "learning_rate": 4.6875e-07, "loss": 1.9617, "step": 16 }, { "epoch": 0.10625, "grad_norm": 4.90625, "learning_rate": 5.000000000000001e-07, "loss": 1.9119, "step": 17 }, { "epoch": 0.1125, "grad_norm": 4.84375, "learning_rate": 5.312500000000001e-07, "loss": 1.8866, "step": 18 }, { "epoch": 0.11875, "grad_norm": 4.875, "learning_rate": 5.625e-07, "loss": 1.884, "step": 19 }, { "epoch": 0.125, "grad_norm": 4.71875, "learning_rate": 5.9375e-07, "loss": 1.8448, "step": 20 }, { "epoch": 0.13125, "grad_norm": 4.90625, "learning_rate": 6.25e-07, "loss": 1.9064, "step": 21 }, { "epoch": 0.1375, "grad_norm": 4.90625, "learning_rate": 6.5625e-07, "loss": 1.8927, "step": 22 }, { "epoch": 0.14375, "grad_norm": 5.09375, "learning_rate": 6.875000000000001e-07, "loss": 1.9613, "step": 23 }, { "epoch": 0.15, "grad_norm": 4.90625, "learning_rate": 7.187500000000001e-07, "loss": 1.9237, "step": 24 }, { "epoch": 0.15625, "grad_norm": 4.96875, "learning_rate": 7.5e-07, "loss": 1.9201, "step": 25 }, { "epoch": 0.1625, "grad_norm": 5.0, "learning_rate": 7.8125e-07, "loss": 1.9074, "step": 26 }, { "epoch": 0.16875, "grad_norm": 5.0, "learning_rate": 8.125000000000001e-07, "loss": 1.8848, "step": 27 }, { "epoch": 0.175, "grad_norm": 4.90625, "learning_rate": 8.437500000000001e-07, "loss": 1.8885, "step": 28 }, { "epoch": 0.18125, "grad_norm": 4.9375, "learning_rate": 8.750000000000001e-07, "loss": 1.8314, "step": 29 }, { "epoch": 0.1875, "grad_norm": 4.9375, "learning_rate": 9.0625e-07, "loss": 1.9331, "step": 30 }, { "epoch": 0.19375, "grad_norm": 4.90625, "learning_rate": 9.375e-07, "loss": 1.8882, "step": 31 }, { "epoch": 0.2, "grad_norm": 4.96875, "learning_rate": 9.6875e-07, "loss": 1.8964, "step": 32 }, { "epoch": 0.20625, "grad_norm": 5.0625, "learning_rate": 1.0000000000000002e-06, "loss": 1.93, "step": 33 }, { "epoch": 0.2125, "grad_norm": 5.03125, "learning_rate": 1.03125e-06, "loss": 1.9356, "step": 34 }, { "epoch": 0.21875, "grad_norm": 4.96875, "learning_rate": 1.0625000000000002e-06, "loss": 1.8844, "step": 35 }, { "epoch": 0.225, "grad_norm": 4.875, "learning_rate": 1.09375e-06, "loss": 1.8917, "step": 36 }, { "epoch": 0.23125, "grad_norm": 5.09375, "learning_rate": 1.125e-06, "loss": 1.94, "step": 37 }, { "epoch": 0.2375, "grad_norm": 4.96875, "learning_rate": 1.15625e-06, "loss": 1.882, "step": 38 }, { "epoch": 0.24375, "grad_norm": 4.96875, "learning_rate": 1.1875e-06, "loss": 1.8865, "step": 39 }, { "epoch": 0.25, "grad_norm": 5.0, "learning_rate": 1.2187500000000002e-06, "loss": 1.8719, "step": 40 }, { "epoch": 0.25625, "grad_norm": 4.96875, "learning_rate": 1.25e-06, "loss": 1.8712, "step": 41 }, { "epoch": 0.2625, "grad_norm": 5.03125, "learning_rate": 1.28125e-06, "loss": 1.9236, "step": 42 }, { "epoch": 0.26875, "grad_norm": 4.8125, "learning_rate": 1.3125e-06, "loss": 1.8739, "step": 43 }, { "epoch": 0.275, "grad_norm": 4.875, "learning_rate": 1.34375e-06, "loss": 1.8913, "step": 44 }, { "epoch": 0.28125, "grad_norm": 4.9375, "learning_rate": 1.3750000000000002e-06, "loss": 1.9317, "step": 45 }, { "epoch": 0.2875, "grad_norm": 4.90625, "learning_rate": 1.40625e-06, "loss": 1.9537, "step": 46 }, { "epoch": 0.29375, "grad_norm": 4.78125, "learning_rate": 1.4375000000000002e-06, "loss": 1.8575, "step": 47 }, { "epoch": 0.3, "grad_norm": 4.90625, "learning_rate": 1.46875e-06, "loss": 1.8925, "step": 48 }, { "epoch": 0.30625, "grad_norm": 5.09375, "learning_rate": 1.5e-06, "loss": 1.9315, "step": 49 }, { "epoch": 0.3125, "grad_norm": 4.8125, "learning_rate": 1.53125e-06, "loss": 1.8744, "step": 50 }, { "epoch": 0.31875, "grad_norm": 4.875, "learning_rate": 1.5625e-06, "loss": 1.909, "step": 51 }, { "epoch": 0.325, "grad_norm": 5.03125, "learning_rate": 1.5937500000000002e-06, "loss": 1.9183, "step": 52 }, { "epoch": 0.33125, "grad_norm": 4.90625, "learning_rate": 1.6250000000000001e-06, "loss": 1.8723, "step": 53 }, { "epoch": 0.3375, "grad_norm": 4.9375, "learning_rate": 1.6562500000000002e-06, "loss": 1.8911, "step": 54 }, { "epoch": 0.34375, "grad_norm": 4.875, "learning_rate": 1.6875000000000001e-06, "loss": 1.8587, "step": 55 }, { "epoch": 0.35, "grad_norm": 4.875, "learning_rate": 1.7187500000000003e-06, "loss": 1.8857, "step": 56 }, { "epoch": 0.35625, "grad_norm": 4.84375, "learning_rate": 1.7500000000000002e-06, "loss": 1.8325, "step": 57 }, { "epoch": 0.3625, "grad_norm": 4.78125, "learning_rate": 1.7812499999999999e-06, "loss": 1.8202, "step": 58 }, { "epoch": 0.36875, "grad_norm": 4.78125, "learning_rate": 1.8125e-06, "loss": 1.8516, "step": 59 }, { "epoch": 0.375, "grad_norm": 4.9375, "learning_rate": 1.84375e-06, "loss": 1.9478, "step": 60 }, { "epoch": 0.38125, "grad_norm": 4.8125, "learning_rate": 1.875e-06, "loss": 1.8716, "step": 61 }, { "epoch": 0.3875, "grad_norm": 4.875, "learning_rate": 1.90625e-06, "loss": 1.8927, "step": 62 }, { "epoch": 0.39375, "grad_norm": 4.84375, "learning_rate": 1.9375e-06, "loss": 1.8314, "step": 63 }, { "epoch": 0.4, "grad_norm": 4.6875, "learning_rate": 1.96875e-06, "loss": 1.8523, "step": 64 }, { "epoch": 0.40625, "grad_norm": 5.09375, "learning_rate": 2.0000000000000003e-06, "loss": 1.8914, "step": 65 }, { "epoch": 0.4125, "grad_norm": 4.90625, "learning_rate": 2.0312500000000002e-06, "loss": 1.9207, "step": 66 }, { "epoch": 0.41875, "grad_norm": 4.8125, "learning_rate": 2.0625e-06, "loss": 1.8506, "step": 67 }, { "epoch": 0.425, "grad_norm": 4.8125, "learning_rate": 2.09375e-06, "loss": 1.8728, "step": 68 }, { "epoch": 0.43125, "grad_norm": 4.84375, "learning_rate": 2.1250000000000004e-06, "loss": 1.865, "step": 69 }, { "epoch": 0.4375, "grad_norm": 4.75, "learning_rate": 2.15625e-06, "loss": 1.8667, "step": 70 }, { "epoch": 0.44375, "grad_norm": 4.75, "learning_rate": 2.1875e-06, "loss": 1.8068, "step": 71 }, { "epoch": 0.45, "grad_norm": 4.875, "learning_rate": 2.21875e-06, "loss": 1.8489, "step": 72 }, { "epoch": 0.45625, "grad_norm": 4.75, "learning_rate": 2.25e-06, "loss": 1.8257, "step": 73 }, { "epoch": 0.4625, "grad_norm": 4.875, "learning_rate": 2.28125e-06, "loss": 1.8666, "step": 74 }, { "epoch": 0.46875, "grad_norm": 4.78125, "learning_rate": 2.3125e-06, "loss": 1.8228, "step": 75 }, { "epoch": 0.475, "grad_norm": 4.8125, "learning_rate": 2.3437500000000002e-06, "loss": 1.8321, "step": 76 }, { "epoch": 0.48125, "grad_norm": 4.9375, "learning_rate": 2.375e-06, "loss": 1.8716, "step": 77 }, { "epoch": 0.4875, "grad_norm": 5.09375, "learning_rate": 2.40625e-06, "loss": 1.8906, "step": 78 }, { "epoch": 0.49375, "grad_norm": 4.6875, "learning_rate": 2.4375000000000004e-06, "loss": 1.8207, "step": 79 }, { "epoch": 0.5, "grad_norm": 4.625, "learning_rate": 2.4687500000000003e-06, "loss": 1.8412, "step": 80 }, { "epoch": 0.50625, "grad_norm": 4.6875, "learning_rate": 2.5e-06, "loss": 1.8232, "step": 81 }, { "epoch": 0.5125, "grad_norm": 4.75, "learning_rate": 2.53125e-06, "loss": 1.8184, "step": 82 }, { "epoch": 0.51875, "grad_norm": 4.6875, "learning_rate": 2.5625e-06, "loss": 1.8337, "step": 83 }, { "epoch": 0.525, "grad_norm": 4.53125, "learning_rate": 2.59375e-06, "loss": 1.8243, "step": 84 }, { "epoch": 0.53125, "grad_norm": 4.6875, "learning_rate": 2.625e-06, "loss": 1.8344, "step": 85 }, { "epoch": 0.5375, "grad_norm": 4.65625, "learning_rate": 2.65625e-06, "loss": 1.8204, "step": 86 }, { "epoch": 0.54375, "grad_norm": 4.875, "learning_rate": 2.6875e-06, "loss": 1.8775, "step": 87 }, { "epoch": 0.55, "grad_norm": 4.8125, "learning_rate": 2.71875e-06, "loss": 1.7902, "step": 88 }, { "epoch": 0.55625, "grad_norm": 4.75, "learning_rate": 2.7500000000000004e-06, "loss": 1.8077, "step": 89 }, { "epoch": 0.5625, "grad_norm": 4.625, "learning_rate": 2.7812500000000003e-06, "loss": 1.8635, "step": 90 }, { "epoch": 0.56875, "grad_norm": 4.59375, "learning_rate": 2.8125e-06, "loss": 1.8004, "step": 91 }, { "epoch": 0.575, "grad_norm": 4.59375, "learning_rate": 2.84375e-06, "loss": 1.7621, "step": 92 }, { "epoch": 0.58125, "grad_norm": 4.59375, "learning_rate": 2.8750000000000004e-06, "loss": 1.8295, "step": 93 }, { "epoch": 0.5875, "grad_norm": 4.65625, "learning_rate": 2.9062500000000003e-06, "loss": 1.8124, "step": 94 }, { "epoch": 0.59375, "grad_norm": 4.625, "learning_rate": 2.9375e-06, "loss": 1.7933, "step": 95 }, { "epoch": 0.6, "grad_norm": 4.6875, "learning_rate": 2.96875e-06, "loss": 1.8479, "step": 96 }, { "epoch": 0.60625, "grad_norm": 4.46875, "learning_rate": 3e-06, "loss": 1.8326, "step": 97 }, { "epoch": 0.6125, "grad_norm": 4.6875, "learning_rate": 3.03125e-06, "loss": 1.8293, "step": 98 }, { "epoch": 0.61875, "grad_norm": 4.5, "learning_rate": 3.0625e-06, "loss": 1.7663, "step": 99 }, { "epoch": 0.625, "grad_norm": 4.59375, "learning_rate": 3.0937500000000002e-06, "loss": 1.8328, "step": 100 }, { "epoch": 0.63125, "grad_norm": 4.5625, "learning_rate": 3.125e-06, "loss": 1.8282, "step": 101 }, { "epoch": 0.6375, "grad_norm": 4.5, "learning_rate": 3.15625e-06, "loss": 1.6976, "step": 102 }, { "epoch": 0.64375, "grad_norm": 4.5, "learning_rate": 3.1875000000000004e-06, "loss": 1.8449, "step": 103 }, { "epoch": 0.65, "grad_norm": 4.34375, "learning_rate": 3.2187500000000003e-06, "loss": 1.7599, "step": 104 }, { "epoch": 0.65625, "grad_norm": 4.5, "learning_rate": 3.2500000000000002e-06, "loss": 1.7856, "step": 105 }, { "epoch": 0.6625, "grad_norm": 4.53125, "learning_rate": 3.28125e-06, "loss": 1.7982, "step": 106 }, { "epoch": 0.66875, "grad_norm": 4.65625, "learning_rate": 3.3125000000000005e-06, "loss": 1.8063, "step": 107 }, { "epoch": 0.675, "grad_norm": 4.4375, "learning_rate": 3.3437500000000004e-06, "loss": 1.7812, "step": 108 }, { "epoch": 0.68125, "grad_norm": 4.375, "learning_rate": 3.3750000000000003e-06, "loss": 1.7383, "step": 109 }, { "epoch": 0.6875, "grad_norm": 4.65625, "learning_rate": 3.40625e-06, "loss": 1.84, "step": 110 }, { "epoch": 0.69375, "grad_norm": 4.375, "learning_rate": 3.4375000000000005e-06, "loss": 1.7481, "step": 111 }, { "epoch": 0.7, "grad_norm": 4.5, "learning_rate": 3.4687500000000005e-06, "loss": 1.7983, "step": 112 }, { "epoch": 0.70625, "grad_norm": 4.25, "learning_rate": 3.5000000000000004e-06, "loss": 1.737, "step": 113 }, { "epoch": 0.7125, "grad_norm": 4.375, "learning_rate": 3.53125e-06, "loss": 1.7453, "step": 114 }, { "epoch": 0.71875, "grad_norm": 4.25, "learning_rate": 3.5624999999999998e-06, "loss": 1.7661, "step": 115 }, { "epoch": 0.725, "grad_norm": 4.3125, "learning_rate": 3.5937499999999997e-06, "loss": 1.7422, "step": 116 }, { "epoch": 0.73125, "grad_norm": 4.25, "learning_rate": 3.625e-06, "loss": 1.7513, "step": 117 }, { "epoch": 0.7375, "grad_norm": 4.40625, "learning_rate": 3.65625e-06, "loss": 1.7519, "step": 118 }, { "epoch": 0.74375, "grad_norm": 4.28125, "learning_rate": 3.6875e-06, "loss": 1.7247, "step": 119 }, { "epoch": 0.75, "grad_norm": 4.34375, "learning_rate": 3.71875e-06, "loss": 1.7371, "step": 120 }, { "epoch": 0.75625, "grad_norm": 4.3125, "learning_rate": 3.75e-06, "loss": 1.7107, "step": 121 }, { "epoch": 0.7625, "grad_norm": 4.3125, "learning_rate": 3.78125e-06, "loss": 1.7496, "step": 122 }, { "epoch": 0.76875, "grad_norm": 4.28125, "learning_rate": 3.8125e-06, "loss": 1.7402, "step": 123 }, { "epoch": 0.775, "grad_norm": 4.25, "learning_rate": 3.84375e-06, "loss": 1.7213, "step": 124 }, { "epoch": 0.78125, "grad_norm": 4.3125, "learning_rate": 3.875e-06, "loss": 1.7067, "step": 125 }, { "epoch": 0.7875, "grad_norm": 4.28125, "learning_rate": 3.90625e-06, "loss": 1.7573, "step": 126 }, { "epoch": 0.79375, "grad_norm": 4.21875, "learning_rate": 3.9375e-06, "loss": 1.7493, "step": 127 }, { "epoch": 0.8, "grad_norm": 4.21875, "learning_rate": 3.96875e-06, "loss": 1.7292, "step": 128 }, { "epoch": 0.80625, "grad_norm": 4.28125, "learning_rate": 4.000000000000001e-06, "loss": 1.7453, "step": 129 }, { "epoch": 0.8125, "grad_norm": 4.15625, "learning_rate": 4.031250000000001e-06, "loss": 1.6893, "step": 130 }, { "epoch": 0.81875, "grad_norm": 4.21875, "learning_rate": 4.0625000000000005e-06, "loss": 1.6744, "step": 131 }, { "epoch": 0.825, "grad_norm": 4.21875, "learning_rate": 4.09375e-06, "loss": 1.6944, "step": 132 }, { "epoch": 0.83125, "grad_norm": 4.0625, "learning_rate": 4.125e-06, "loss": 1.6998, "step": 133 }, { "epoch": 0.8375, "grad_norm": 4.09375, "learning_rate": 4.15625e-06, "loss": 1.6616, "step": 134 }, { "epoch": 0.84375, "grad_norm": 3.953125, "learning_rate": 4.1875e-06, "loss": 1.7039, "step": 135 }, { "epoch": 0.85, "grad_norm": 4.0, "learning_rate": 4.218750000000001e-06, "loss": 1.6936, "step": 136 }, { "epoch": 0.85625, "grad_norm": 4.125, "learning_rate": 4.250000000000001e-06, "loss": 1.6918, "step": 137 }, { "epoch": 0.8625, "grad_norm": 4.125, "learning_rate": 4.281250000000001e-06, "loss": 1.6677, "step": 138 }, { "epoch": 0.86875, "grad_norm": 4.03125, "learning_rate": 4.3125e-06, "loss": 1.6967, "step": 139 }, { "epoch": 0.875, "grad_norm": 3.96875, "learning_rate": 4.34375e-06, "loss": 1.6188, "step": 140 }, { "epoch": 0.88125, "grad_norm": 4.09375, "learning_rate": 4.375e-06, "loss": 1.6653, "step": 141 }, { "epoch": 0.8875, "grad_norm": 3.953125, "learning_rate": 4.40625e-06, "loss": 1.6563, "step": 142 }, { "epoch": 0.89375, "grad_norm": 3.984375, "learning_rate": 4.4375e-06, "loss": 1.676, "step": 143 }, { "epoch": 0.9, "grad_norm": 3.90625, "learning_rate": 4.46875e-06, "loss": 1.6581, "step": 144 }, { "epoch": 0.90625, "grad_norm": 3.890625, "learning_rate": 4.5e-06, "loss": 1.6089, "step": 145 }, { "epoch": 0.9125, "grad_norm": 3.890625, "learning_rate": 4.53125e-06, "loss": 1.6114, "step": 146 }, { "epoch": 0.91875, "grad_norm": 3.921875, "learning_rate": 4.5625e-06, "loss": 1.6139, "step": 147 }, { "epoch": 0.925, "grad_norm": 3.875, "learning_rate": 4.59375e-06, "loss": 1.6088, "step": 148 }, { "epoch": 0.93125, "grad_norm": 3.703125, "learning_rate": 4.625e-06, "loss": 1.5632, "step": 149 }, { "epoch": 0.9375, "grad_norm": 3.78125, "learning_rate": 4.6562500000000005e-06, "loss": 1.5829, "step": 150 }, { "epoch": 0.94375, "grad_norm": 3.765625, "learning_rate": 4.6875000000000004e-06, "loss": 1.5841, "step": 151 }, { "epoch": 0.95, "grad_norm": 3.5, "learning_rate": 4.71875e-06, "loss": 1.5643, "step": 152 }, { "epoch": 0.95625, "grad_norm": 3.640625, "learning_rate": 4.75e-06, "loss": 1.5443, "step": 153 }, { "epoch": 0.9625, "grad_norm": 3.609375, "learning_rate": 4.78125e-06, "loss": 1.594, "step": 154 }, { "epoch": 0.96875, "grad_norm": 3.6875, "learning_rate": 4.8125e-06, "loss": 1.5289, "step": 155 }, { "epoch": 0.975, "grad_norm": 3.609375, "learning_rate": 4.84375e-06, "loss": 1.5102, "step": 156 }, { "epoch": 0.98125, "grad_norm": 3.25, "learning_rate": 4.875000000000001e-06, "loss": 1.507, "step": 157 }, { "epoch": 0.9875, "grad_norm": 3.546875, "learning_rate": 4.906250000000001e-06, "loss": 1.5605, "step": 158 }, { "epoch": 0.99375, "grad_norm": 3.578125, "learning_rate": 4.937500000000001e-06, "loss": 1.5454, "step": 159 }, { "epoch": 1.0, "grad_norm": 3.40625, "learning_rate": 4.9687500000000005e-06, "loss": 1.5216, "step": 160 }, { "epoch": 1.00625, "grad_norm": 3.265625, "learning_rate": 5e-06, "loss": 1.5596, "step": 161 }, { "epoch": 1.0125, "grad_norm": 3.4375, "learning_rate": 5.03125e-06, "loss": 1.5526, "step": 162 }, { "epoch": 1.01875, "grad_norm": 3.40625, "learning_rate": 5.0625e-06, "loss": 1.544, "step": 163 }, { "epoch": 1.025, "grad_norm": 3.28125, "learning_rate": 5.09375e-06, "loss": 1.4833, "step": 164 }, { "epoch": 1.03125, "grad_norm": 3.40625, "learning_rate": 5.125e-06, "loss": 1.5141, "step": 165 }, { "epoch": 1.0375, "grad_norm": 3.390625, "learning_rate": 5.15625e-06, "loss": 1.4638, "step": 166 }, { "epoch": 1.04375, "grad_norm": 3.296875, "learning_rate": 5.1875e-06, "loss": 1.4466, "step": 167 }, { "epoch": 1.05, "grad_norm": 3.265625, "learning_rate": 5.21875e-06, "loss": 1.4904, "step": 168 }, { "epoch": 1.05625, "grad_norm": 3.1875, "learning_rate": 5.25e-06, "loss": 1.4676, "step": 169 }, { "epoch": 1.0625, "grad_norm": 3.0625, "learning_rate": 5.2812500000000005e-06, "loss": 1.4465, "step": 170 }, { "epoch": 1.06875, "grad_norm": 3.28125, "learning_rate": 5.3125e-06, "loss": 1.4346, "step": 171 }, { "epoch": 1.075, "grad_norm": 3.234375, "learning_rate": 5.34375e-06, "loss": 1.4594, "step": 172 }, { "epoch": 1.08125, "grad_norm": 3.109375, "learning_rate": 5.375e-06, "loss": 1.4692, "step": 173 }, { "epoch": 1.0875, "grad_norm": 3.09375, "learning_rate": 5.40625e-06, "loss": 1.4737, "step": 174 }, { "epoch": 1.09375, "grad_norm": 2.984375, "learning_rate": 5.4375e-06, "loss": 1.4325, "step": 175 }, { "epoch": 1.1, "grad_norm": 2.90625, "learning_rate": 5.46875e-06, "loss": 1.4427, "step": 176 }, { "epoch": 1.10625, "grad_norm": 3.03125, "learning_rate": 5.500000000000001e-06, "loss": 1.3965, "step": 177 }, { "epoch": 1.1125, "grad_norm": 2.9375, "learning_rate": 5.531250000000001e-06, "loss": 1.4321, "step": 178 }, { "epoch": 1.11875, "grad_norm": 3.03125, "learning_rate": 5.5625000000000005e-06, "loss": 1.4709, "step": 179 }, { "epoch": 1.125, "grad_norm": 2.96875, "learning_rate": 5.5937500000000004e-06, "loss": 1.3946, "step": 180 }, { "epoch": 1.13125, "grad_norm": 2.96875, "learning_rate": 5.625e-06, "loss": 1.3841, "step": 181 }, { "epoch": 1.1375, "grad_norm": 2.8125, "learning_rate": 5.65625e-06, "loss": 1.4395, "step": 182 }, { "epoch": 1.14375, "grad_norm": 2.828125, "learning_rate": 5.6875e-06, "loss": 1.4324, "step": 183 }, { "epoch": 1.15, "grad_norm": 2.734375, "learning_rate": 5.71875e-06, "loss": 1.363, "step": 184 }, { "epoch": 1.15625, "grad_norm": 2.921875, "learning_rate": 5.750000000000001e-06, "loss": 1.41, "step": 185 }, { "epoch": 1.1625, "grad_norm": 2.609375, "learning_rate": 5.781250000000001e-06, "loss": 1.3933, "step": 186 }, { "epoch": 1.16875, "grad_norm": 2.640625, "learning_rate": 5.812500000000001e-06, "loss": 1.3588, "step": 187 }, { "epoch": 1.175, "grad_norm": 2.640625, "learning_rate": 5.843750000000001e-06, "loss": 1.3631, "step": 188 }, { "epoch": 1.18125, "grad_norm": 2.609375, "learning_rate": 5.875e-06, "loss": 1.3436, "step": 189 }, { "epoch": 1.1875, "grad_norm": 2.796875, "learning_rate": 5.9062499999999996e-06, "loss": 1.383, "step": 190 }, { "epoch": 1.19375, "grad_norm": 2.59375, "learning_rate": 5.9375e-06, "loss": 1.3676, "step": 191 }, { "epoch": 1.2, "grad_norm": 2.578125, "learning_rate": 5.96875e-06, "loss": 1.3213, "step": 192 }, { "epoch": 1.20625, "grad_norm": 2.46875, "learning_rate": 6e-06, "loss": 1.3731, "step": 193 }, { "epoch": 1.2125, "grad_norm": 2.46875, "learning_rate": 6.03125e-06, "loss": 1.2929, "step": 194 }, { "epoch": 1.21875, "grad_norm": 2.40625, "learning_rate": 6.0625e-06, "loss": 1.3348, "step": 195 }, { "epoch": 1.225, "grad_norm": 2.40625, "learning_rate": 6.09375e-06, "loss": 1.3471, "step": 196 }, { "epoch": 1.23125, "grad_norm": 2.421875, "learning_rate": 6.125e-06, "loss": 1.3193, "step": 197 }, { "epoch": 1.2375, "grad_norm": 2.28125, "learning_rate": 6.1562500000000006e-06, "loss": 1.3293, "step": 198 }, { "epoch": 1.24375, "grad_norm": 2.3125, "learning_rate": 6.1875000000000005e-06, "loss": 1.2997, "step": 199 }, { "epoch": 1.25, "grad_norm": 2.328125, "learning_rate": 6.21875e-06, "loss": 1.367, "step": 200 }, { "epoch": 1.25625, "grad_norm": 2.25, "learning_rate": 6.25e-06, "loss": 1.3215, "step": 201 }, { "epoch": 1.2625, "grad_norm": 2.3125, "learning_rate": 6.281249999999999e-06, "loss": 1.2458, "step": 202 }, { "epoch": 1.26875, "grad_norm": 2.15625, "learning_rate": 6.3125e-06, "loss": 1.2492, "step": 203 }, { "epoch": 1.275, "grad_norm": 2.171875, "learning_rate": 6.34375e-06, "loss": 1.3051, "step": 204 }, { "epoch": 1.28125, "grad_norm": 2.109375, "learning_rate": 6.375000000000001e-06, "loss": 1.2732, "step": 205 }, { "epoch": 1.2875, "grad_norm": 2.203125, "learning_rate": 6.40625e-06, "loss": 1.2759, "step": 206 }, { "epoch": 1.29375, "grad_norm": 2.28125, "learning_rate": 6.437500000000001e-06, "loss": 1.299, "step": 207 }, { "epoch": 1.3, "grad_norm": 1.96875, "learning_rate": 6.46875e-06, "loss": 1.2847, "step": 208 }, { "epoch": 1.30625, "grad_norm": 1.875, "learning_rate": 6.5000000000000004e-06, "loss": 1.2747, "step": 209 }, { "epoch": 1.3125, "grad_norm": 1.96875, "learning_rate": 6.5312499999999995e-06, "loss": 1.2739, "step": 210 }, { "epoch": 1.31875, "grad_norm": 2.015625, "learning_rate": 6.5625e-06, "loss": 1.2496, "step": 211 }, { "epoch": 1.325, "grad_norm": 1.9453125, "learning_rate": 6.59375e-06, "loss": 1.2695, "step": 212 }, { "epoch": 1.33125, "grad_norm": 1.9375, "learning_rate": 6.625000000000001e-06, "loss": 1.2953, "step": 213 }, { "epoch": 1.3375, "grad_norm": 1.9453125, "learning_rate": 6.65625e-06, "loss": 1.2987, "step": 214 }, { "epoch": 1.34375, "grad_norm": 1.8515625, "learning_rate": 6.687500000000001e-06, "loss": 1.2487, "step": 215 }, { "epoch": 1.35, "grad_norm": 1.84375, "learning_rate": 6.71875e-06, "loss": 1.2802, "step": 216 }, { "epoch": 1.35625, "grad_norm": 1.75, "learning_rate": 6.750000000000001e-06, "loss": 1.1927, "step": 217 }, { "epoch": 1.3625, "grad_norm": 1.6640625, "learning_rate": 6.7812500000000005e-06, "loss": 1.229, "step": 218 }, { "epoch": 1.36875, "grad_norm": 1.6796875, "learning_rate": 6.8125e-06, "loss": 1.2589, "step": 219 }, { "epoch": 1.375, "grad_norm": 1.5703125, "learning_rate": 6.84375e-06, "loss": 1.2319, "step": 220 }, { "epoch": 1.38125, "grad_norm": 1.6796875, "learning_rate": 6.875000000000001e-06, "loss": 1.2203, "step": 221 }, { "epoch": 1.3875, "grad_norm": 1.609375, "learning_rate": 6.90625e-06, "loss": 1.217, "step": 222 }, { "epoch": 1.39375, "grad_norm": 1.6015625, "learning_rate": 6.937500000000001e-06, "loss": 1.1892, "step": 223 }, { "epoch": 1.4, "grad_norm": 1.4921875, "learning_rate": 6.96875e-06, "loss": 1.2394, "step": 224 }, { "epoch": 1.40625, "grad_norm": 1.59375, "learning_rate": 7.000000000000001e-06, "loss": 1.1997, "step": 225 }, { "epoch": 1.4125, "grad_norm": 1.453125, "learning_rate": 7.031250000000001e-06, "loss": 1.2016, "step": 226 }, { "epoch": 1.41875, "grad_norm": 1.5234375, "learning_rate": 7.0625e-06, "loss": 1.2152, "step": 227 }, { "epoch": 1.425, "grad_norm": 1.515625, "learning_rate": 7.0937500000000005e-06, "loss": 1.1882, "step": 228 }, { "epoch": 1.43125, "grad_norm": 1.46875, "learning_rate": 7.1249999999999995e-06, "loss": 1.1733, "step": 229 }, { "epoch": 1.4375, "grad_norm": 1.3125, "learning_rate": 7.15625e-06, "loss": 1.2043, "step": 230 }, { "epoch": 1.44375, "grad_norm": 1.40625, "learning_rate": 7.187499999999999e-06, "loss": 1.2042, "step": 231 }, { "epoch": 1.45, "grad_norm": 1.25, "learning_rate": 7.21875e-06, "loss": 1.1837, "step": 232 }, { "epoch": 1.45625, "grad_norm": 1.3828125, "learning_rate": 7.25e-06, "loss": 1.1887, "step": 233 }, { "epoch": 1.4625, "grad_norm": 1.234375, "learning_rate": 7.281250000000001e-06, "loss": 1.1682, "step": 234 }, { "epoch": 1.46875, "grad_norm": 1.203125, "learning_rate": 7.3125e-06, "loss": 1.1619, "step": 235 }, { "epoch": 1.475, "grad_norm": 1.203125, "learning_rate": 7.343750000000001e-06, "loss": 1.1843, "step": 236 }, { "epoch": 1.48125, "grad_norm": 1.359375, "learning_rate": 7.375e-06, "loss": 1.2554, "step": 237 }, { "epoch": 1.4875, "grad_norm": 1.1875, "learning_rate": 7.4062500000000005e-06, "loss": 1.1611, "step": 238 }, { "epoch": 1.49375, "grad_norm": 1.15625, "learning_rate": 7.4375e-06, "loss": 1.1707, "step": 239 }, { "epoch": 1.5, "grad_norm": 1.2734375, "learning_rate": 7.468750000000001e-06, "loss": 1.2262, "step": 240 }, { "epoch": 1.50625, "grad_norm": 1.1640625, "learning_rate": 7.5e-06, "loss": 1.2072, "step": 241 }, { "epoch": 1.5125, "grad_norm": 1.2421875, "learning_rate": 7.531250000000001e-06, "loss": 1.2141, "step": 242 }, { "epoch": 1.51875, "grad_norm": 1.2421875, "learning_rate": 7.5625e-06, "loss": 1.181, "step": 243 }, { "epoch": 1.525, "grad_norm": 1.1015625, "learning_rate": 7.593750000000001e-06, "loss": 1.1834, "step": 244 }, { "epoch": 1.53125, "grad_norm": 1.1171875, "learning_rate": 7.625e-06, "loss": 1.2303, "step": 245 }, { "epoch": 1.5375, "grad_norm": 1.21875, "learning_rate": 7.656250000000001e-06, "loss": 1.1583, "step": 246 }, { "epoch": 1.54375, "grad_norm": 1.0390625, "learning_rate": 7.6875e-06, "loss": 1.195, "step": 247 }, { "epoch": 1.55, "grad_norm": 1.203125, "learning_rate": 7.718750000000001e-06, "loss": 1.11, "step": 248 }, { "epoch": 1.55625, "grad_norm": 0.9453125, "learning_rate": 7.75e-06, "loss": 1.2071, "step": 249 }, { "epoch": 1.5625, "grad_norm": 1.078125, "learning_rate": 7.781250000000001e-06, "loss": 1.1732, "step": 250 }, { "epoch": 1.56875, "grad_norm": 1.0859375, "learning_rate": 7.8125e-06, "loss": 1.2093, "step": 251 }, { "epoch": 1.575, "grad_norm": 1.078125, "learning_rate": 7.84375e-06, "loss": 1.189, "step": 252 }, { "epoch": 1.58125, "grad_norm": 1.0703125, "learning_rate": 7.875e-06, "loss": 1.1825, "step": 253 }, { "epoch": 1.5875, "grad_norm": 0.92578125, "learning_rate": 7.906249999999999e-06, "loss": 1.2116, "step": 254 }, { "epoch": 1.59375, "grad_norm": 1.0859375, "learning_rate": 7.9375e-06, "loss": 1.1599, "step": 255 }, { "epoch": 1.6, "grad_norm": 0.96875, "learning_rate": 7.96875e-06, "loss": 1.1715, "step": 256 }, { "epoch": 1.60625, "grad_norm": 1.234375, "learning_rate": 8.000000000000001e-06, "loss": 1.1007, "step": 257 }, { "epoch": 1.6125, "grad_norm": 1.0859375, "learning_rate": 8.03125e-06, "loss": 1.1408, "step": 258 }, { "epoch": 1.61875, "grad_norm": 0.96875, "learning_rate": 8.062500000000001e-06, "loss": 1.1887, "step": 259 }, { "epoch": 1.625, "grad_norm": 0.9296875, "learning_rate": 8.09375e-06, "loss": 1.1657, "step": 260 }, { "epoch": 1.63125, "grad_norm": 1.1015625, "learning_rate": 8.125000000000001e-06, "loss": 1.1536, "step": 261 }, { "epoch": 1.6375, "grad_norm": 1.015625, "learning_rate": 8.15625e-06, "loss": 1.1987, "step": 262 }, { "epoch": 1.64375, "grad_norm": 1.03125, "learning_rate": 8.1875e-06, "loss": 1.1287, "step": 263 }, { "epoch": 1.65, "grad_norm": 0.96875, "learning_rate": 8.21875e-06, "loss": 1.1397, "step": 264 }, { "epoch": 1.65625, "grad_norm": 1.0234375, "learning_rate": 8.25e-06, "loss": 1.1774, "step": 265 }, { "epoch": 1.6625, "grad_norm": 0.9375, "learning_rate": 8.28125e-06, "loss": 1.176, "step": 266 }, { "epoch": 1.66875, "grad_norm": 0.94140625, "learning_rate": 8.3125e-06, "loss": 1.158, "step": 267 }, { "epoch": 1.675, "grad_norm": 0.92578125, "learning_rate": 8.34375e-06, "loss": 1.2086, "step": 268 }, { "epoch": 1.68125, "grad_norm": 0.8359375, "learning_rate": 8.375e-06, "loss": 1.1961, "step": 269 }, { "epoch": 1.6875, "grad_norm": 1.046875, "learning_rate": 8.406250000000001e-06, "loss": 1.1934, "step": 270 }, { "epoch": 1.69375, "grad_norm": 1.0703125, "learning_rate": 8.437500000000002e-06, "loss": 1.1741, "step": 271 }, { "epoch": 1.7, "grad_norm": 1.0, "learning_rate": 8.468750000000001e-06, "loss": 1.1198, "step": 272 }, { "epoch": 1.70625, "grad_norm": 0.91796875, "learning_rate": 8.500000000000002e-06, "loss": 1.1475, "step": 273 }, { "epoch": 1.7125, "grad_norm": 0.94140625, "learning_rate": 8.53125e-06, "loss": 1.1693, "step": 274 }, { "epoch": 1.71875, "grad_norm": 0.91796875, "learning_rate": 8.562500000000001e-06, "loss": 1.1507, "step": 275 }, { "epoch": 1.725, "grad_norm": 0.93359375, "learning_rate": 8.59375e-06, "loss": 1.157, "step": 276 }, { "epoch": 1.73125, "grad_norm": 0.93359375, "learning_rate": 8.625e-06, "loss": 1.1569, "step": 277 }, { "epoch": 1.7375, "grad_norm": 0.84375, "learning_rate": 8.65625e-06, "loss": 1.1777, "step": 278 }, { "epoch": 1.74375, "grad_norm": 0.91796875, "learning_rate": 8.6875e-06, "loss": 1.1658, "step": 279 }, { "epoch": 1.75, "grad_norm": 0.984375, "learning_rate": 8.71875e-06, "loss": 1.1974, "step": 280 }, { "epoch": 1.75625, "grad_norm": 0.95703125, "learning_rate": 8.75e-06, "loss": 1.1852, "step": 281 }, { "epoch": 1.7625, "grad_norm": 0.953125, "learning_rate": 8.78125e-06, "loss": 1.1177, "step": 282 }, { "epoch": 1.76875, "grad_norm": 0.953125, "learning_rate": 8.8125e-06, "loss": 1.1444, "step": 283 }, { "epoch": 1.775, "grad_norm": 0.84765625, "learning_rate": 8.84375e-06, "loss": 1.1678, "step": 284 }, { "epoch": 1.78125, "grad_norm": 0.87890625, "learning_rate": 8.875e-06, "loss": 1.1871, "step": 285 }, { "epoch": 1.7875, "grad_norm": 0.88671875, "learning_rate": 8.906250000000001e-06, "loss": 1.1478, "step": 286 }, { "epoch": 1.79375, "grad_norm": 0.859375, "learning_rate": 8.9375e-06, "loss": 1.168, "step": 287 }, { "epoch": 1.8, "grad_norm": 1.0078125, "learning_rate": 8.968750000000001e-06, "loss": 1.1408, "step": 288 }, { "epoch": 1.80625, "grad_norm": 1.0078125, "learning_rate": 9e-06, "loss": 1.0994, "step": 289 }, { "epoch": 1.8125, "grad_norm": 0.953125, "learning_rate": 9.031250000000001e-06, "loss": 1.1681, "step": 290 }, { "epoch": 1.81875, "grad_norm": 0.9375, "learning_rate": 9.0625e-06, "loss": 1.1889, "step": 291 }, { "epoch": 1.825, "grad_norm": 0.84375, "learning_rate": 9.09375e-06, "loss": 1.1701, "step": 292 }, { "epoch": 1.83125, "grad_norm": 0.9453125, "learning_rate": 9.125e-06, "loss": 1.1729, "step": 293 }, { "epoch": 1.8375, "grad_norm": 1.0234375, "learning_rate": 9.15625e-06, "loss": 1.2352, "step": 294 }, { "epoch": 1.84375, "grad_norm": 0.96484375, "learning_rate": 9.1875e-06, "loss": 1.188, "step": 295 }, { "epoch": 1.85, "grad_norm": 0.8984375, "learning_rate": 9.21875e-06, "loss": 1.1573, "step": 296 }, { "epoch": 1.85625, "grad_norm": 0.94140625, "learning_rate": 9.25e-06, "loss": 1.1526, "step": 297 }, { "epoch": 1.8625, "grad_norm": 0.97265625, "learning_rate": 9.28125e-06, "loss": 1.1434, "step": 298 }, { "epoch": 1.86875, "grad_norm": 1.0390625, "learning_rate": 9.312500000000001e-06, "loss": 1.2036, "step": 299 }, { "epoch": 1.875, "grad_norm": 0.93359375, "learning_rate": 9.343750000000002e-06, "loss": 1.1946, "step": 300 }, { "epoch": 1.88125, "grad_norm": 1.3203125, "learning_rate": 9.375000000000001e-06, "loss": 1.2162, "step": 301 }, { "epoch": 1.8875, "grad_norm": 0.84765625, "learning_rate": 9.40625e-06, "loss": 1.1688, "step": 302 }, { "epoch": 1.89375, "grad_norm": 0.86328125, "learning_rate": 9.4375e-06, "loss": 1.1499, "step": 303 }, { "epoch": 1.9, "grad_norm": 0.90625, "learning_rate": 9.46875e-06, "loss": 1.1509, "step": 304 }, { "epoch": 1.90625, "grad_norm": 0.9921875, "learning_rate": 9.5e-06, "loss": 1.1258, "step": 305 }, { "epoch": 1.9125, "grad_norm": 0.96484375, "learning_rate": 9.53125e-06, "loss": 1.2264, "step": 306 }, { "epoch": 1.91875, "grad_norm": 1.0078125, "learning_rate": 9.5625e-06, "loss": 1.1526, "step": 307 }, { "epoch": 1.925, "grad_norm": 0.92578125, "learning_rate": 9.59375e-06, "loss": 1.1491, "step": 308 }, { "epoch": 1.93125, "grad_norm": 0.8125, "learning_rate": 9.625e-06, "loss": 1.1597, "step": 309 }, { "epoch": 1.9375, "grad_norm": 0.859375, "learning_rate": 9.65625e-06, "loss": 1.1277, "step": 310 }, { "epoch": 1.94375, "grad_norm": 0.97265625, "learning_rate": 9.6875e-06, "loss": 1.1486, "step": 311 }, { "epoch": 1.95, "grad_norm": 0.890625, "learning_rate": 9.71875e-06, "loss": 1.1496, "step": 312 }, { "epoch": 1.95625, "grad_norm": 1.1171875, "learning_rate": 9.750000000000002e-06, "loss": 1.2064, "step": 313 }, { "epoch": 1.9625, "grad_norm": 0.93359375, "learning_rate": 9.78125e-06, "loss": 1.1211, "step": 314 }, { "epoch": 1.96875, "grad_norm": 0.89453125, "learning_rate": 9.812500000000001e-06, "loss": 1.1511, "step": 315 }, { "epoch": 1.975, "grad_norm": 0.890625, "learning_rate": 9.84375e-06, "loss": 1.1523, "step": 316 }, { "epoch": 1.98125, "grad_norm": 0.89453125, "learning_rate": 9.875000000000001e-06, "loss": 1.173, "step": 317 }, { "epoch": 1.9875, "grad_norm": 0.99609375, "learning_rate": 9.90625e-06, "loss": 1.1413, "step": 318 }, { "epoch": 1.99375, "grad_norm": 1.0390625, "learning_rate": 9.937500000000001e-06, "loss": 1.1384, "step": 319 }, { "epoch": 2.0, "grad_norm": 0.9375, "learning_rate": 9.96875e-06, "loss": 1.1185, "step": 320 }, { "epoch": 2.00625, "grad_norm": 0.94140625, "learning_rate": 1e-05, "loss": 1.1336, "step": 321 }, { "epoch": 2.0125, "grad_norm": 0.87890625, "learning_rate": 1.003125e-05, "loss": 1.1752, "step": 322 }, { "epoch": 2.01875, "grad_norm": 0.94921875, "learning_rate": 1.00625e-05, "loss": 1.1661, "step": 323 }, { "epoch": 2.025, "grad_norm": 0.9453125, "learning_rate": 1.009375e-05, "loss": 1.1708, "step": 324 }, { "epoch": 2.03125, "grad_norm": 0.84375, "learning_rate": 1.0125e-05, "loss": 1.1189, "step": 325 }, { "epoch": 2.0375, "grad_norm": 0.9453125, "learning_rate": 1.0156250000000001e-05, "loss": 1.1663, "step": 326 }, { "epoch": 2.04375, "grad_norm": 0.9921875, "learning_rate": 1.01875e-05, "loss": 1.0994, "step": 327 }, { "epoch": 2.05, "grad_norm": 1.0078125, "learning_rate": 1.0218750000000001e-05, "loss": 1.1827, "step": 328 }, { "epoch": 2.05625, "grad_norm": 0.8671875, "learning_rate": 1.025e-05, "loss": 1.1574, "step": 329 }, { "epoch": 2.0625, "grad_norm": 0.94140625, "learning_rate": 1.0281250000000001e-05, "loss": 1.1171, "step": 330 }, { "epoch": 2.06875, "grad_norm": 0.95703125, "learning_rate": 1.03125e-05, "loss": 1.182, "step": 331 }, { "epoch": 2.075, "grad_norm": 0.8671875, "learning_rate": 1.034375e-05, "loss": 1.1374, "step": 332 }, { "epoch": 2.08125, "grad_norm": 0.95703125, "learning_rate": 1.0375e-05, "loss": 1.1471, "step": 333 }, { "epoch": 2.0875, "grad_norm": 0.98046875, "learning_rate": 1.040625e-05, "loss": 1.1999, "step": 334 }, { "epoch": 2.09375, "grad_norm": 0.8515625, "learning_rate": 1.04375e-05, "loss": 1.1388, "step": 335 }, { "epoch": 2.1, "grad_norm": 0.89453125, "learning_rate": 1.046875e-05, "loss": 1.1605, "step": 336 }, { "epoch": 2.10625, "grad_norm": 0.88671875, "learning_rate": 1.05e-05, "loss": 1.1554, "step": 337 }, { "epoch": 2.1125, "grad_norm": 0.88671875, "learning_rate": 1.053125e-05, "loss": 1.1484, "step": 338 }, { "epoch": 2.11875, "grad_norm": 0.86328125, "learning_rate": 1.0562500000000001e-05, "loss": 1.1509, "step": 339 }, { "epoch": 2.125, "grad_norm": 0.96875, "learning_rate": 1.0593750000000002e-05, "loss": 1.1582, "step": 340 }, { "epoch": 2.13125, "grad_norm": 0.98046875, "learning_rate": 1.0625e-05, "loss": 1.1925, "step": 341 }, { "epoch": 2.1375, "grad_norm": 0.890625, "learning_rate": 1.0656250000000002e-05, "loss": 1.1816, "step": 342 }, { "epoch": 2.14375, "grad_norm": 0.83984375, "learning_rate": 1.06875e-05, "loss": 1.1675, "step": 343 }, { "epoch": 2.15, "grad_norm": 0.9765625, "learning_rate": 1.0718750000000001e-05, "loss": 1.1468, "step": 344 }, { "epoch": 2.15625, "grad_norm": 0.8984375, "learning_rate": 1.075e-05, "loss": 1.1914, "step": 345 }, { "epoch": 2.1625, "grad_norm": 0.88671875, "learning_rate": 1.0781250000000001e-05, "loss": 1.2043, "step": 346 }, { "epoch": 2.16875, "grad_norm": 0.7734375, "learning_rate": 1.08125e-05, "loss": 1.1657, "step": 347 }, { "epoch": 2.175, "grad_norm": 0.94140625, "learning_rate": 1.0843750000000001e-05, "loss": 1.1722, "step": 348 }, { "epoch": 2.18125, "grad_norm": 0.8671875, "learning_rate": 1.0875e-05, "loss": 1.1656, "step": 349 }, { "epoch": 2.1875, "grad_norm": 0.921875, "learning_rate": 1.090625e-05, "loss": 1.1443, "step": 350 }, { "epoch": 2.19375, "grad_norm": 0.8515625, "learning_rate": 1.09375e-05, "loss": 1.1976, "step": 351 }, { "epoch": 2.2, "grad_norm": 0.90234375, "learning_rate": 1.096875e-05, "loss": 1.1793, "step": 352 }, { "epoch": 2.20625, "grad_norm": 0.9609375, "learning_rate": 1.1000000000000001e-05, "loss": 1.1708, "step": 353 }, { "epoch": 2.2125, "grad_norm": 0.9609375, "learning_rate": 1.103125e-05, "loss": 1.214, "step": 354 }, { "epoch": 2.21875, "grad_norm": 0.890625, "learning_rate": 1.1062500000000001e-05, "loss": 1.1032, "step": 355 }, { "epoch": 2.225, "grad_norm": 0.91796875, "learning_rate": 1.109375e-05, "loss": 1.1576, "step": 356 }, { "epoch": 2.23125, "grad_norm": 0.8359375, "learning_rate": 1.1125000000000001e-05, "loss": 1.1535, "step": 357 }, { "epoch": 2.2375, "grad_norm": 0.9375, "learning_rate": 1.115625e-05, "loss": 1.1457, "step": 358 }, { "epoch": 2.24375, "grad_norm": 0.92578125, "learning_rate": 1.1187500000000001e-05, "loss": 1.1443, "step": 359 }, { "epoch": 2.25, "grad_norm": 0.953125, "learning_rate": 1.121875e-05, "loss": 1.1605, "step": 360 }, { "epoch": 2.25625, "grad_norm": 0.859375, "learning_rate": 1.125e-05, "loss": 1.1453, "step": 361 }, { "epoch": 2.2625, "grad_norm": 0.8828125, "learning_rate": 1.128125e-05, "loss": 1.1606, "step": 362 }, { "epoch": 2.26875, "grad_norm": 0.9921875, "learning_rate": 1.13125e-05, "loss": 1.16, "step": 363 }, { "epoch": 2.275, "grad_norm": 0.83203125, "learning_rate": 1.134375e-05, "loss": 1.1349, "step": 364 }, { "epoch": 2.28125, "grad_norm": 0.796875, "learning_rate": 1.1375e-05, "loss": 1.1302, "step": 365 }, { "epoch": 2.2875, "grad_norm": 0.92578125, "learning_rate": 1.140625e-05, "loss": 1.1228, "step": 366 }, { "epoch": 2.29375, "grad_norm": 1.046875, "learning_rate": 1.14375e-05, "loss": 1.1277, "step": 367 }, { "epoch": 2.3, "grad_norm": 0.93359375, "learning_rate": 1.1468750000000001e-05, "loss": 1.1457, "step": 368 }, { "epoch": 2.30625, "grad_norm": 0.875, "learning_rate": 1.1500000000000002e-05, "loss": 1.1354, "step": 369 }, { "epoch": 2.3125, "grad_norm": 1.0234375, "learning_rate": 1.153125e-05, "loss": 1.1992, "step": 370 }, { "epoch": 2.31875, "grad_norm": 0.91796875, "learning_rate": 1.1562500000000002e-05, "loss": 1.151, "step": 371 }, { "epoch": 2.325, "grad_norm": 0.9140625, "learning_rate": 1.159375e-05, "loss": 1.1397, "step": 372 }, { "epoch": 2.33125, "grad_norm": 0.9296875, "learning_rate": 1.1625000000000001e-05, "loss": 1.1336, "step": 373 }, { "epoch": 2.3375, "grad_norm": 0.90625, "learning_rate": 1.165625e-05, "loss": 1.0895, "step": 374 }, { "epoch": 2.34375, "grad_norm": 0.90234375, "learning_rate": 1.1687500000000001e-05, "loss": 1.1349, "step": 375 }, { "epoch": 2.35, "grad_norm": 0.9375, "learning_rate": 1.171875e-05, "loss": 1.1562, "step": 376 }, { "epoch": 2.35625, "grad_norm": 1.0078125, "learning_rate": 1.175e-05, "loss": 1.2002, "step": 377 }, { "epoch": 2.3625, "grad_norm": 1.0078125, "learning_rate": 1.178125e-05, "loss": 1.1794, "step": 378 }, { "epoch": 2.36875, "grad_norm": 0.89453125, "learning_rate": 1.1812499999999999e-05, "loss": 1.1519, "step": 379 }, { "epoch": 2.375, "grad_norm": 1.078125, "learning_rate": 1.184375e-05, "loss": 1.1942, "step": 380 }, { "epoch": 2.38125, "grad_norm": 0.98046875, "learning_rate": 1.1875e-05, "loss": 1.0964, "step": 381 }, { "epoch": 2.3875, "grad_norm": 0.95703125, "learning_rate": 1.1906250000000001e-05, "loss": 1.1435, "step": 382 }, { "epoch": 2.39375, "grad_norm": 0.8515625, "learning_rate": 1.19375e-05, "loss": 1.1388, "step": 383 }, { "epoch": 2.4, "grad_norm": 0.9453125, "learning_rate": 1.1968750000000001e-05, "loss": 1.0979, "step": 384 }, { "epoch": 2.40625, "grad_norm": 0.8828125, "learning_rate": 1.2e-05, "loss": 1.1804, "step": 385 }, { "epoch": 2.4125, "grad_norm": 0.81640625, "learning_rate": 1.2031250000000001e-05, "loss": 1.1114, "step": 386 }, { "epoch": 2.41875, "grad_norm": 0.94140625, "learning_rate": 1.20625e-05, "loss": 1.1572, "step": 387 }, { "epoch": 2.425, "grad_norm": 0.89453125, "learning_rate": 1.2093750000000001e-05, "loss": 1.157, "step": 388 }, { "epoch": 2.43125, "grad_norm": 0.92578125, "learning_rate": 1.2125e-05, "loss": 1.1518, "step": 389 }, { "epoch": 2.4375, "grad_norm": 0.94921875, "learning_rate": 1.215625e-05, "loss": 1.1267, "step": 390 }, { "epoch": 2.44375, "grad_norm": 0.8671875, "learning_rate": 1.21875e-05, "loss": 1.1452, "step": 391 }, { "epoch": 2.45, "grad_norm": 0.99609375, "learning_rate": 1.221875e-05, "loss": 1.1398, "step": 392 }, { "epoch": 2.45625, "grad_norm": 0.9296875, "learning_rate": 1.225e-05, "loss": 1.1817, "step": 393 }, { "epoch": 2.4625, "grad_norm": 0.875, "learning_rate": 1.228125e-05, "loss": 1.1591, "step": 394 }, { "epoch": 2.46875, "grad_norm": 0.87890625, "learning_rate": 1.2312500000000001e-05, "loss": 1.1603, "step": 395 }, { "epoch": 2.475, "grad_norm": 0.92578125, "learning_rate": 1.2343750000000002e-05, "loss": 1.104, "step": 396 }, { "epoch": 2.48125, "grad_norm": 0.921875, "learning_rate": 1.2375000000000001e-05, "loss": 1.1594, "step": 397 }, { "epoch": 2.4875, "grad_norm": 0.8671875, "learning_rate": 1.2406250000000002e-05, "loss": 1.1449, "step": 398 }, { "epoch": 2.49375, "grad_norm": 0.9609375, "learning_rate": 1.24375e-05, "loss": 1.1682, "step": 399 }, { "epoch": 2.5, "grad_norm": 0.9296875, "learning_rate": 1.2468750000000002e-05, "loss": 1.0976, "step": 400 }, { "epoch": 2.50625, "grad_norm": 0.91015625, "learning_rate": 1.25e-05, "loss": 1.1072, "step": 401 }, { "epoch": 2.5125, "grad_norm": 0.9140625, "learning_rate": 1.253125e-05, "loss": 1.1735, "step": 402 }, { "epoch": 2.51875, "grad_norm": 0.80078125, "learning_rate": 1.2562499999999999e-05, "loss": 1.1593, "step": 403 }, { "epoch": 2.525, "grad_norm": 1.0390625, "learning_rate": 1.2593750000000001e-05, "loss": 1.1821, "step": 404 }, { "epoch": 2.53125, "grad_norm": 1.0078125, "learning_rate": 1.2625e-05, "loss": 1.2196, "step": 405 }, { "epoch": 2.5375, "grad_norm": 0.89453125, "learning_rate": 1.265625e-05, "loss": 1.122, "step": 406 }, { "epoch": 2.54375, "grad_norm": 0.9296875, "learning_rate": 1.26875e-05, "loss": 1.0801, "step": 407 }, { "epoch": 2.55, "grad_norm": 1.046875, "learning_rate": 1.271875e-05, "loss": 1.17, "step": 408 }, { "epoch": 2.55625, "grad_norm": 1.1484375, "learning_rate": 1.2750000000000002e-05, "loss": 1.2018, "step": 409 }, { "epoch": 2.5625, "grad_norm": 0.84765625, "learning_rate": 1.278125e-05, "loss": 1.1582, "step": 410 }, { "epoch": 2.56875, "grad_norm": 0.9609375, "learning_rate": 1.28125e-05, "loss": 1.1847, "step": 411 }, { "epoch": 2.575, "grad_norm": 0.9765625, "learning_rate": 1.2843750000000002e-05, "loss": 1.1235, "step": 412 }, { "epoch": 2.58125, "grad_norm": 0.91015625, "learning_rate": 1.2875000000000001e-05, "loss": 1.1319, "step": 413 }, { "epoch": 2.5875, "grad_norm": 0.9375, "learning_rate": 1.290625e-05, "loss": 1.1123, "step": 414 }, { "epoch": 2.59375, "grad_norm": 0.9296875, "learning_rate": 1.29375e-05, "loss": 1.1408, "step": 415 }, { "epoch": 2.6, "grad_norm": 0.98828125, "learning_rate": 1.2968750000000002e-05, "loss": 1.1439, "step": 416 }, { "epoch": 2.60625, "grad_norm": 1.09375, "learning_rate": 1.3000000000000001e-05, "loss": 1.1195, "step": 417 }, { "epoch": 2.6125, "grad_norm": 0.89453125, "learning_rate": 1.303125e-05, "loss": 1.1861, "step": 418 }, { "epoch": 2.61875, "grad_norm": 1.125, "learning_rate": 1.3062499999999999e-05, "loss": 1.1266, "step": 419 }, { "epoch": 2.625, "grad_norm": 1.03125, "learning_rate": 1.3093750000000001e-05, "loss": 1.1427, "step": 420 }, { "epoch": 2.63125, "grad_norm": 0.921875, "learning_rate": 1.3125e-05, "loss": 1.173, "step": 421 }, { "epoch": 2.6375, "grad_norm": 0.83203125, "learning_rate": 1.3156250000000001e-05, "loss": 1.1517, "step": 422 }, { "epoch": 2.64375, "grad_norm": 0.9765625, "learning_rate": 1.31875e-05, "loss": 1.1134, "step": 423 }, { "epoch": 2.65, "grad_norm": 0.91015625, "learning_rate": 1.3218750000000001e-05, "loss": 1.1505, "step": 424 }, { "epoch": 2.65625, "grad_norm": 0.9765625, "learning_rate": 1.3250000000000002e-05, "loss": 1.1212, "step": 425 }, { "epoch": 2.6625, "grad_norm": 0.9375, "learning_rate": 1.3281250000000001e-05, "loss": 1.1769, "step": 426 }, { "epoch": 2.66875, "grad_norm": 0.91796875, "learning_rate": 1.33125e-05, "loss": 1.1481, "step": 427 }, { "epoch": 2.675, "grad_norm": 0.87109375, "learning_rate": 1.3343749999999999e-05, "loss": 1.0953, "step": 428 }, { "epoch": 2.68125, "grad_norm": 1.0625, "learning_rate": 1.3375000000000002e-05, "loss": 1.1446, "step": 429 }, { "epoch": 2.6875, "grad_norm": 0.94140625, "learning_rate": 1.340625e-05, "loss": 1.1332, "step": 430 }, { "epoch": 2.69375, "grad_norm": 1.109375, "learning_rate": 1.34375e-05, "loss": 1.1279, "step": 431 }, { "epoch": 2.7, "grad_norm": 0.88671875, "learning_rate": 1.3468749999999999e-05, "loss": 1.1747, "step": 432 }, { "epoch": 2.70625, "grad_norm": 0.921875, "learning_rate": 1.3500000000000001e-05, "loss": 1.1088, "step": 433 }, { "epoch": 2.7125, "grad_norm": 0.953125, "learning_rate": 1.353125e-05, "loss": 1.1392, "step": 434 }, { "epoch": 2.71875, "grad_norm": 1.03125, "learning_rate": 1.3562500000000001e-05, "loss": 1.1238, "step": 435 }, { "epoch": 2.725, "grad_norm": 0.84765625, "learning_rate": 1.359375e-05, "loss": 1.1266, "step": 436 }, { "epoch": 2.73125, "grad_norm": 0.92578125, "learning_rate": 1.3625e-05, "loss": 1.1572, "step": 437 }, { "epoch": 2.7375, "grad_norm": 0.9296875, "learning_rate": 1.3656250000000002e-05, "loss": 1.0914, "step": 438 }, { "epoch": 2.74375, "grad_norm": 0.9296875, "learning_rate": 1.36875e-05, "loss": 1.1505, "step": 439 }, { "epoch": 2.75, "grad_norm": 1.0234375, "learning_rate": 1.371875e-05, "loss": 1.1849, "step": 440 }, { "epoch": 2.75625, "grad_norm": 1.0859375, "learning_rate": 1.3750000000000002e-05, "loss": 1.1741, "step": 441 }, { "epoch": 2.7625, "grad_norm": 0.828125, "learning_rate": 1.3781250000000001e-05, "loss": 1.1224, "step": 442 }, { "epoch": 2.76875, "grad_norm": 1.0625, "learning_rate": 1.38125e-05, "loss": 1.1233, "step": 443 }, { "epoch": 2.775, "grad_norm": 0.921875, "learning_rate": 1.384375e-05, "loss": 1.1193, "step": 444 }, { "epoch": 2.78125, "grad_norm": 1.125, "learning_rate": 1.3875000000000002e-05, "loss": 1.1806, "step": 445 }, { "epoch": 2.7875, "grad_norm": 0.98046875, "learning_rate": 1.3906250000000001e-05, "loss": 1.1794, "step": 446 }, { "epoch": 2.79375, "grad_norm": 0.9609375, "learning_rate": 1.39375e-05, "loss": 1.0975, "step": 447 }, { "epoch": 2.8, "grad_norm": 0.97265625, "learning_rate": 1.396875e-05, "loss": 1.1675, "step": 448 }, { "epoch": 2.80625, "grad_norm": 1.0390625, "learning_rate": 1.4000000000000001e-05, "loss": 1.1582, "step": 449 }, { "epoch": 2.8125, "grad_norm": 0.8984375, "learning_rate": 1.403125e-05, "loss": 1.1332, "step": 450 }, { "epoch": 2.81875, "grad_norm": 0.94921875, "learning_rate": 1.4062500000000001e-05, "loss": 1.1292, "step": 451 }, { "epoch": 2.825, "grad_norm": 0.93359375, "learning_rate": 1.409375e-05, "loss": 1.1789, "step": 452 }, { "epoch": 2.83125, "grad_norm": 0.98828125, "learning_rate": 1.4125e-05, "loss": 1.1372, "step": 453 }, { "epoch": 2.8375, "grad_norm": 0.90625, "learning_rate": 1.4156250000000002e-05, "loss": 1.1395, "step": 454 }, { "epoch": 2.84375, "grad_norm": 1.078125, "learning_rate": 1.4187500000000001e-05, "loss": 1.1487, "step": 455 }, { "epoch": 2.85, "grad_norm": 0.88671875, "learning_rate": 1.421875e-05, "loss": 1.1412, "step": 456 }, { "epoch": 2.85625, "grad_norm": 0.8515625, "learning_rate": 1.4249999999999999e-05, "loss": 1.1204, "step": 457 }, { "epoch": 2.8625, "grad_norm": 0.9921875, "learning_rate": 1.4281250000000002e-05, "loss": 1.1437, "step": 458 }, { "epoch": 2.86875, "grad_norm": 0.96875, "learning_rate": 1.43125e-05, "loss": 1.1402, "step": 459 }, { "epoch": 2.875, "grad_norm": 0.85546875, "learning_rate": 1.434375e-05, "loss": 1.1104, "step": 460 }, { "epoch": 2.88125, "grad_norm": 0.91015625, "learning_rate": 1.4374999999999999e-05, "loss": 1.1232, "step": 461 }, { "epoch": 2.8875, "grad_norm": 0.984375, "learning_rate": 1.4406250000000001e-05, "loss": 1.1804, "step": 462 }, { "epoch": 2.89375, "grad_norm": 0.9375, "learning_rate": 1.44375e-05, "loss": 1.1507, "step": 463 }, { "epoch": 2.9, "grad_norm": 0.8828125, "learning_rate": 1.4468750000000001e-05, "loss": 1.1386, "step": 464 }, { "epoch": 2.90625, "grad_norm": 1.0, "learning_rate": 1.45e-05, "loss": 1.1175, "step": 465 }, { "epoch": 2.9125, "grad_norm": 0.9453125, "learning_rate": 1.4531250000000003e-05, "loss": 1.1501, "step": 466 }, { "epoch": 2.91875, "grad_norm": 0.96875, "learning_rate": 1.4562500000000002e-05, "loss": 1.1429, "step": 467 }, { "epoch": 2.925, "grad_norm": 0.8359375, "learning_rate": 1.459375e-05, "loss": 1.1506, "step": 468 }, { "epoch": 2.93125, "grad_norm": 0.97265625, "learning_rate": 1.4625e-05, "loss": 1.1357, "step": 469 }, { "epoch": 2.9375, "grad_norm": 0.93359375, "learning_rate": 1.4656250000000002e-05, "loss": 1.1132, "step": 470 }, { "epoch": 2.94375, "grad_norm": 0.89453125, "learning_rate": 1.4687500000000001e-05, "loss": 1.0797, "step": 471 }, { "epoch": 2.95, "grad_norm": 0.94921875, "learning_rate": 1.471875e-05, "loss": 1.1049, "step": 472 }, { "epoch": 2.95625, "grad_norm": 0.921875, "learning_rate": 1.475e-05, "loss": 1.1294, "step": 473 }, { "epoch": 2.9625, "grad_norm": 0.98828125, "learning_rate": 1.4781250000000002e-05, "loss": 1.132, "step": 474 }, { "epoch": 2.96875, "grad_norm": 0.984375, "learning_rate": 1.4812500000000001e-05, "loss": 1.1294, "step": 475 }, { "epoch": 2.975, "grad_norm": 1.015625, "learning_rate": 1.484375e-05, "loss": 1.145, "step": 476 }, { "epoch": 2.98125, "grad_norm": 0.94140625, "learning_rate": 1.4875e-05, "loss": 1.1368, "step": 477 }, { "epoch": 2.9875, "grad_norm": 0.83984375, "learning_rate": 1.490625e-05, "loss": 1.1303, "step": 478 }, { "epoch": 2.99375, "grad_norm": 0.953125, "learning_rate": 1.4937500000000002e-05, "loss": 1.1207, "step": 479 }, { "epoch": 3.0, "grad_norm": 0.92578125, "learning_rate": 1.4968750000000001e-05, "loss": 1.1193, "step": 480 }, { "epoch": 3.00625, "grad_norm": 0.921875, "learning_rate": 1.5e-05, "loss": 1.1542, "step": 481 }, { "epoch": 3.0125, "grad_norm": 0.9140625, "learning_rate": 1.503125e-05, "loss": 1.14, "step": 482 }, { "epoch": 3.01875, "grad_norm": 0.9140625, "learning_rate": 1.5062500000000002e-05, "loss": 1.1343, "step": 483 }, { "epoch": 3.025, "grad_norm": 0.83203125, "learning_rate": 1.5093750000000001e-05, "loss": 1.1526, "step": 484 }, { "epoch": 3.03125, "grad_norm": 1.0078125, "learning_rate": 1.5125e-05, "loss": 1.1192, "step": 485 }, { "epoch": 3.0375, "grad_norm": 0.8984375, "learning_rate": 1.5156249999999999e-05, "loss": 1.1641, "step": 486 }, { "epoch": 3.04375, "grad_norm": 0.94140625, "learning_rate": 1.5187500000000002e-05, "loss": 1.1517, "step": 487 }, { "epoch": 3.05, "grad_norm": 1.0, "learning_rate": 1.521875e-05, "loss": 1.1392, "step": 488 }, { "epoch": 3.05625, "grad_norm": 0.953125, "learning_rate": 1.525e-05, "loss": 1.1067, "step": 489 }, { "epoch": 3.0625, "grad_norm": 1.09375, "learning_rate": 1.528125e-05, "loss": 1.1694, "step": 490 }, { "epoch": 3.06875, "grad_norm": 1.0390625, "learning_rate": 1.5312500000000003e-05, "loss": 1.1294, "step": 491 }, { "epoch": 3.075, "grad_norm": 0.8984375, "learning_rate": 1.534375e-05, "loss": 1.1019, "step": 492 }, { "epoch": 3.08125, "grad_norm": 1.0625, "learning_rate": 1.5375e-05, "loss": 1.1747, "step": 493 }, { "epoch": 3.0875, "grad_norm": 0.953125, "learning_rate": 1.540625e-05, "loss": 1.0915, "step": 494 }, { "epoch": 3.09375, "grad_norm": 0.96484375, "learning_rate": 1.5437500000000003e-05, "loss": 1.1518, "step": 495 }, { "epoch": 3.1, "grad_norm": 0.83984375, "learning_rate": 1.546875e-05, "loss": 1.1, "step": 496 }, { "epoch": 3.10625, "grad_norm": 0.94921875, "learning_rate": 1.55e-05, "loss": 1.1096, "step": 497 }, { "epoch": 3.1125, "grad_norm": 0.99609375, "learning_rate": 1.553125e-05, "loss": 1.1682, "step": 498 }, { "epoch": 3.11875, "grad_norm": 0.92578125, "learning_rate": 1.5562500000000002e-05, "loss": 1.1221, "step": 499 }, { "epoch": 3.125, "grad_norm": 1.09375, "learning_rate": 1.5593750000000003e-05, "loss": 1.1116, "step": 500 }, { "epoch": 3.13125, "grad_norm": 0.9296875, "learning_rate": 1.5625e-05, "loss": 1.0917, "step": 501 }, { "epoch": 3.1375, "grad_norm": 0.99609375, "learning_rate": 1.565625e-05, "loss": 1.1279, "step": 502 }, { "epoch": 3.14375, "grad_norm": 0.9375, "learning_rate": 1.56875e-05, "loss": 1.1355, "step": 503 }, { "epoch": 3.15, "grad_norm": 0.94140625, "learning_rate": 1.5718750000000003e-05, "loss": 1.1306, "step": 504 }, { "epoch": 3.15625, "grad_norm": 0.859375, "learning_rate": 1.575e-05, "loss": 1.0887, "step": 505 }, { "epoch": 3.1625, "grad_norm": 0.9453125, "learning_rate": 1.578125e-05, "loss": 1.0981, "step": 506 }, { "epoch": 3.16875, "grad_norm": 0.87890625, "learning_rate": 1.5812499999999998e-05, "loss": 1.1525, "step": 507 }, { "epoch": 3.175, "grad_norm": 0.93359375, "learning_rate": 1.5843750000000002e-05, "loss": 1.1312, "step": 508 }, { "epoch": 3.18125, "grad_norm": 0.94921875, "learning_rate": 1.5875e-05, "loss": 1.1623, "step": 509 }, { "epoch": 3.1875, "grad_norm": 0.98828125, "learning_rate": 1.590625e-05, "loss": 1.1437, "step": 510 }, { "epoch": 3.19375, "grad_norm": 0.859375, "learning_rate": 1.59375e-05, "loss": 1.1138, "step": 511 }, { "epoch": 3.2, "grad_norm": 0.86328125, "learning_rate": 1.5968750000000002e-05, "loss": 1.1091, "step": 512 }, { "epoch": 3.20625, "grad_norm": 0.9453125, "learning_rate": 1.6000000000000003e-05, "loss": 1.0669, "step": 513 }, { "epoch": 3.2125, "grad_norm": 1.2265625, "learning_rate": 1.603125e-05, "loss": 1.1542, "step": 514 }, { "epoch": 3.21875, "grad_norm": 1.0625, "learning_rate": 1.60625e-05, "loss": 1.1466, "step": 515 }, { "epoch": 3.225, "grad_norm": 0.9609375, "learning_rate": 1.609375e-05, "loss": 1.1079, "step": 516 }, { "epoch": 3.23125, "grad_norm": 0.98828125, "learning_rate": 1.6125000000000002e-05, "loss": 1.1604, "step": 517 }, { "epoch": 3.2375, "grad_norm": 0.9140625, "learning_rate": 1.615625e-05, "loss": 1.0778, "step": 518 }, { "epoch": 3.24375, "grad_norm": 0.890625, "learning_rate": 1.61875e-05, "loss": 1.1041, "step": 519 }, { "epoch": 3.25, "grad_norm": 0.921875, "learning_rate": 1.621875e-05, "loss": 1.1423, "step": 520 }, { "epoch": 3.25625, "grad_norm": 0.91796875, "learning_rate": 1.6250000000000002e-05, "loss": 1.1241, "step": 521 }, { "epoch": 3.2625, "grad_norm": 0.94921875, "learning_rate": 1.628125e-05, "loss": 1.1161, "step": 522 }, { "epoch": 3.26875, "grad_norm": 0.93359375, "learning_rate": 1.63125e-05, "loss": 1.1, "step": 523 }, { "epoch": 3.275, "grad_norm": 1.015625, "learning_rate": 1.634375e-05, "loss": 1.105, "step": 524 }, { "epoch": 3.28125, "grad_norm": 0.94921875, "learning_rate": 1.6375e-05, "loss": 1.1319, "step": 525 }, { "epoch": 3.2875, "grad_norm": 0.9609375, "learning_rate": 1.6406250000000002e-05, "loss": 1.0946, "step": 526 }, { "epoch": 3.29375, "grad_norm": 0.91015625, "learning_rate": 1.64375e-05, "loss": 1.1234, "step": 527 }, { "epoch": 3.3, "grad_norm": 1.0078125, "learning_rate": 1.646875e-05, "loss": 1.1658, "step": 528 }, { "epoch": 3.30625, "grad_norm": 0.98046875, "learning_rate": 1.65e-05, "loss": 1.1312, "step": 529 }, { "epoch": 3.3125, "grad_norm": 1.15625, "learning_rate": 1.6531250000000002e-05, "loss": 1.0903, "step": 530 }, { "epoch": 3.31875, "grad_norm": 0.9453125, "learning_rate": 1.65625e-05, "loss": 1.12, "step": 531 }, { "epoch": 3.325, "grad_norm": 0.96875, "learning_rate": 1.659375e-05, "loss": 1.0855, "step": 532 }, { "epoch": 3.33125, "grad_norm": 0.984375, "learning_rate": 1.6625e-05, "loss": 1.0916, "step": 533 }, { "epoch": 3.3375, "grad_norm": 0.953125, "learning_rate": 1.665625e-05, "loss": 1.1209, "step": 534 }, { "epoch": 3.34375, "grad_norm": 0.98828125, "learning_rate": 1.66875e-05, "loss": 1.1342, "step": 535 }, { "epoch": 3.35, "grad_norm": 0.94921875, "learning_rate": 1.671875e-05, "loss": 1.103, "step": 536 }, { "epoch": 3.35625, "grad_norm": 1.03125, "learning_rate": 1.675e-05, "loss": 1.1597, "step": 537 }, { "epoch": 3.3625, "grad_norm": 0.8984375, "learning_rate": 1.678125e-05, "loss": 1.0973, "step": 538 }, { "epoch": 3.36875, "grad_norm": 0.9765625, "learning_rate": 1.6812500000000002e-05, "loss": 1.1176, "step": 539 }, { "epoch": 3.375, "grad_norm": 1.0625, "learning_rate": 1.684375e-05, "loss": 1.1218, "step": 540 }, { "epoch": 3.38125, "grad_norm": 0.85546875, "learning_rate": 1.6875000000000004e-05, "loss": 1.0994, "step": 541 }, { "epoch": 3.3875, "grad_norm": 1.3828125, "learning_rate": 1.690625e-05, "loss": 1.1539, "step": 542 }, { "epoch": 3.39375, "grad_norm": 0.90625, "learning_rate": 1.6937500000000002e-05, "loss": 1.105, "step": 543 }, { "epoch": 3.4, "grad_norm": 1.1796875, "learning_rate": 1.696875e-05, "loss": 1.1344, "step": 544 }, { "epoch": 3.40625, "grad_norm": 1.109375, "learning_rate": 1.7000000000000003e-05, "loss": 1.1831, "step": 545 }, { "epoch": 3.4125, "grad_norm": 0.953125, "learning_rate": 1.703125e-05, "loss": 1.0922, "step": 546 }, { "epoch": 3.41875, "grad_norm": 0.97265625, "learning_rate": 1.70625e-05, "loss": 1.1088, "step": 547 }, { "epoch": 3.425, "grad_norm": 0.9609375, "learning_rate": 1.709375e-05, "loss": 1.1062, "step": 548 }, { "epoch": 3.43125, "grad_norm": 1.015625, "learning_rate": 1.7125000000000003e-05, "loss": 1.1798, "step": 549 }, { "epoch": 3.4375, "grad_norm": 1.3203125, "learning_rate": 1.715625e-05, "loss": 1.1296, "step": 550 }, { "epoch": 3.44375, "grad_norm": 1.2421875, "learning_rate": 1.71875e-05, "loss": 1.1041, "step": 551 }, { "epoch": 3.45, "grad_norm": 0.94921875, "learning_rate": 1.7218750000000002e-05, "loss": 1.1052, "step": 552 }, { "epoch": 3.45625, "grad_norm": 1.0390625, "learning_rate": 1.725e-05, "loss": 1.096, "step": 553 }, { "epoch": 3.4625, "grad_norm": 1.015625, "learning_rate": 1.728125e-05, "loss": 1.1041, "step": 554 }, { "epoch": 3.46875, "grad_norm": 1.0703125, "learning_rate": 1.73125e-05, "loss": 1.0984, "step": 555 }, { "epoch": 3.475, "grad_norm": 0.9609375, "learning_rate": 1.734375e-05, "loss": 1.1224, "step": 556 }, { "epoch": 3.48125, "grad_norm": 1.1640625, "learning_rate": 1.7375e-05, "loss": 1.1653, "step": 557 }, { "epoch": 3.4875, "grad_norm": 0.953125, "learning_rate": 1.7406250000000003e-05, "loss": 1.113, "step": 558 }, { "epoch": 3.49375, "grad_norm": 1.0234375, "learning_rate": 1.74375e-05, "loss": 1.1055, "step": 559 }, { "epoch": 3.5, "grad_norm": 0.953125, "learning_rate": 1.746875e-05, "loss": 1.0901, "step": 560 }, { "epoch": 3.50625, "grad_norm": 0.9609375, "learning_rate": 1.75e-05, "loss": 1.0991, "step": 561 }, { "epoch": 3.5125, "grad_norm": 0.953125, "learning_rate": 1.7531250000000003e-05, "loss": 1.114, "step": 562 }, { "epoch": 3.51875, "grad_norm": 0.8515625, "learning_rate": 1.75625e-05, "loss": 1.0677, "step": 563 }, { "epoch": 3.525, "grad_norm": 0.94921875, "learning_rate": 1.759375e-05, "loss": 1.0707, "step": 564 }, { "epoch": 3.53125, "grad_norm": 0.9765625, "learning_rate": 1.7625e-05, "loss": 1.076, "step": 565 }, { "epoch": 3.5375, "grad_norm": 0.953125, "learning_rate": 1.7656250000000002e-05, "loss": 1.0695, "step": 566 }, { "epoch": 3.54375, "grad_norm": 0.98046875, "learning_rate": 1.76875e-05, "loss": 1.0577, "step": 567 }, { "epoch": 3.55, "grad_norm": 0.92578125, "learning_rate": 1.771875e-05, "loss": 1.0768, "step": 568 }, { "epoch": 3.55625, "grad_norm": 1.125, "learning_rate": 1.775e-05, "loss": 1.101, "step": 569 }, { "epoch": 3.5625, "grad_norm": 1.171875, "learning_rate": 1.7781250000000002e-05, "loss": 1.1245, "step": 570 }, { "epoch": 3.56875, "grad_norm": 0.96484375, "learning_rate": 1.7812500000000003e-05, "loss": 1.1026, "step": 571 }, { "epoch": 3.575, "grad_norm": 0.99609375, "learning_rate": 1.784375e-05, "loss": 1.1194, "step": 572 }, { "epoch": 3.58125, "grad_norm": 1.03125, "learning_rate": 1.7875e-05, "loss": 1.1123, "step": 573 }, { "epoch": 3.5875, "grad_norm": 1.140625, "learning_rate": 1.790625e-05, "loss": 1.0802, "step": 574 }, { "epoch": 3.59375, "grad_norm": 1.296875, "learning_rate": 1.7937500000000002e-05, "loss": 1.0806, "step": 575 }, { "epoch": 3.6, "grad_norm": 1.2265625, "learning_rate": 1.796875e-05, "loss": 1.0665, "step": 576 }, { "epoch": 3.60625, "grad_norm": 1.1953125, "learning_rate": 1.8e-05, "loss": 1.1339, "step": 577 }, { "epoch": 3.6125, "grad_norm": 1.0234375, "learning_rate": 1.803125e-05, "loss": 1.0659, "step": 578 }, { "epoch": 3.61875, "grad_norm": 1.015625, "learning_rate": 1.8062500000000002e-05, "loss": 1.1172, "step": 579 }, { "epoch": 3.625, "grad_norm": 0.97265625, "learning_rate": 1.809375e-05, "loss": 1.0466, "step": 580 }, { "epoch": 3.63125, "grad_norm": 1.015625, "learning_rate": 1.8125e-05, "loss": 1.1008, "step": 581 }, { "epoch": 3.6375, "grad_norm": 1.2109375, "learning_rate": 1.815625e-05, "loss": 1.1104, "step": 582 }, { "epoch": 3.64375, "grad_norm": 1.015625, "learning_rate": 1.81875e-05, "loss": 1.0767, "step": 583 }, { "epoch": 3.65, "grad_norm": 0.89453125, "learning_rate": 1.8218750000000002e-05, "loss": 1.0921, "step": 584 }, { "epoch": 3.65625, "grad_norm": 0.90625, "learning_rate": 1.825e-05, "loss": 1.0772, "step": 585 }, { "epoch": 3.6625, "grad_norm": 0.9453125, "learning_rate": 1.828125e-05, "loss": 1.0476, "step": 586 }, { "epoch": 3.66875, "grad_norm": 0.91015625, "learning_rate": 1.83125e-05, "loss": 1.054, "step": 587 }, { "epoch": 3.675, "grad_norm": 1.03125, "learning_rate": 1.8343750000000002e-05, "loss": 1.0794, "step": 588 }, { "epoch": 3.68125, "grad_norm": 0.9765625, "learning_rate": 1.8375e-05, "loss": 1.0771, "step": 589 }, { "epoch": 3.6875, "grad_norm": 0.99609375, "learning_rate": 1.840625e-05, "loss": 1.0883, "step": 590 }, { "epoch": 3.69375, "grad_norm": 0.9375, "learning_rate": 1.84375e-05, "loss": 1.0787, "step": 591 }, { "epoch": 3.7, "grad_norm": 1.046875, "learning_rate": 1.846875e-05, "loss": 1.0911, "step": 592 }, { "epoch": 3.70625, "grad_norm": 1.0546875, "learning_rate": 1.85e-05, "loss": 1.0671, "step": 593 }, { "epoch": 3.7125, "grad_norm": 1.0234375, "learning_rate": 1.853125e-05, "loss": 1.0751, "step": 594 }, { "epoch": 3.71875, "grad_norm": 1.0234375, "learning_rate": 1.85625e-05, "loss": 1.0805, "step": 595 }, { "epoch": 3.725, "grad_norm": 1.0234375, "learning_rate": 1.859375e-05, "loss": 1.0591, "step": 596 }, { "epoch": 3.73125, "grad_norm": 0.96484375, "learning_rate": 1.8625000000000002e-05, "loss": 1.0874, "step": 597 }, { "epoch": 3.7375, "grad_norm": 1.015625, "learning_rate": 1.865625e-05, "loss": 1.0838, "step": 598 }, { "epoch": 3.74375, "grad_norm": 0.9140625, "learning_rate": 1.8687500000000004e-05, "loss": 1.0824, "step": 599 }, { "epoch": 3.75, "grad_norm": 1.0625, "learning_rate": 1.871875e-05, "loss": 1.0845, "step": 600 }, { "epoch": 3.75625, "grad_norm": 1.0, "learning_rate": 1.8750000000000002e-05, "loss": 1.0735, "step": 601 }, { "epoch": 3.7625, "grad_norm": 1.0390625, "learning_rate": 1.878125e-05, "loss": 1.0585, "step": 602 }, { "epoch": 3.76875, "grad_norm": 0.98046875, "learning_rate": 1.88125e-05, "loss": 1.0836, "step": 603 }, { "epoch": 3.775, "grad_norm": 1.203125, "learning_rate": 1.884375e-05, "loss": 1.1024, "step": 604 }, { "epoch": 3.78125, "grad_norm": 1.0625, "learning_rate": 1.8875e-05, "loss": 1.0536, "step": 605 }, { "epoch": 3.7875, "grad_norm": 0.953125, "learning_rate": 1.890625e-05, "loss": 1.0512, "step": 606 }, { "epoch": 3.79375, "grad_norm": 0.87890625, "learning_rate": 1.89375e-05, "loss": 1.0713, "step": 607 }, { "epoch": 3.8, "grad_norm": 0.9765625, "learning_rate": 1.896875e-05, "loss": 1.0691, "step": 608 }, { "epoch": 3.80625, "grad_norm": 1.0703125, "learning_rate": 1.9e-05, "loss": 1.0654, "step": 609 }, { "epoch": 3.8125, "grad_norm": 1.0703125, "learning_rate": 1.9031250000000002e-05, "loss": 1.0899, "step": 610 }, { "epoch": 3.81875, "grad_norm": 0.97265625, "learning_rate": 1.90625e-05, "loss": 1.067, "step": 611 }, { "epoch": 3.825, "grad_norm": 1.1171875, "learning_rate": 1.9093750000000003e-05, "loss": 1.0837, "step": 612 }, { "epoch": 3.83125, "grad_norm": 1.1796875, "learning_rate": 1.9125e-05, "loss": 1.0864, "step": 613 }, { "epoch": 3.8375, "grad_norm": 1.0859375, "learning_rate": 1.915625e-05, "loss": 1.0809, "step": 614 }, { "epoch": 3.84375, "grad_norm": 1.0390625, "learning_rate": 1.91875e-05, "loss": 1.0379, "step": 615 }, { "epoch": 3.85, "grad_norm": 1.2578125, "learning_rate": 1.9218750000000003e-05, "loss": 1.0899, "step": 616 }, { "epoch": 3.85625, "grad_norm": 1.0390625, "learning_rate": 1.925e-05, "loss": 1.0833, "step": 617 }, { "epoch": 3.8625, "grad_norm": 1.0078125, "learning_rate": 1.928125e-05, "loss": 1.1044, "step": 618 }, { "epoch": 3.86875, "grad_norm": 0.98046875, "learning_rate": 1.93125e-05, "loss": 1.0381, "step": 619 }, { "epoch": 3.875, "grad_norm": 0.9609375, "learning_rate": 1.9343750000000003e-05, "loss": 1.0729, "step": 620 }, { "epoch": 3.88125, "grad_norm": 1.0546875, "learning_rate": 1.9375e-05, "loss": 1.1034, "step": 621 }, { "epoch": 3.8875, "grad_norm": 1.0078125, "learning_rate": 1.940625e-05, "loss": 1.0906, "step": 622 }, { "epoch": 3.89375, "grad_norm": 1.0390625, "learning_rate": 1.94375e-05, "loss": 1.0275, "step": 623 }, { "epoch": 3.9, "grad_norm": 1.1640625, "learning_rate": 1.9468750000000002e-05, "loss": 1.061, "step": 624 }, { "epoch": 3.90625, "grad_norm": 0.9453125, "learning_rate": 1.9500000000000003e-05, "loss": 1.0792, "step": 625 }, { "epoch": 3.9125, "grad_norm": 1.03125, "learning_rate": 1.953125e-05, "loss": 1.0583, "step": 626 }, { "epoch": 3.91875, "grad_norm": 1.1796875, "learning_rate": 1.95625e-05, "loss": 1.0345, "step": 627 }, { "epoch": 3.925, "grad_norm": 0.98046875, "learning_rate": 1.959375e-05, "loss": 1.0675, "step": 628 }, { "epoch": 3.93125, "grad_norm": 1.03125, "learning_rate": 1.9625000000000003e-05, "loss": 1.0408, "step": 629 }, { "epoch": 3.9375, "grad_norm": 1.015625, "learning_rate": 1.965625e-05, "loss": 1.0726, "step": 630 }, { "epoch": 3.94375, "grad_norm": 1.03125, "learning_rate": 1.96875e-05, "loss": 1.0434, "step": 631 }, { "epoch": 3.95, "grad_norm": 0.95703125, "learning_rate": 1.9718749999999998e-05, "loss": 1.0728, "step": 632 }, { "epoch": 3.95625, "grad_norm": 0.8671875, "learning_rate": 1.9750000000000002e-05, "loss": 1.0607, "step": 633 }, { "epoch": 3.9625, "grad_norm": 1.0, "learning_rate": 1.978125e-05, "loss": 1.0407, "step": 634 }, { "epoch": 3.96875, "grad_norm": 1.109375, "learning_rate": 1.98125e-05, "loss": 1.0266, "step": 635 }, { "epoch": 3.975, "grad_norm": 0.99609375, "learning_rate": 1.984375e-05, "loss": 1.0547, "step": 636 }, { "epoch": 3.98125, "grad_norm": 0.96875, "learning_rate": 1.9875000000000002e-05, "loss": 1.0994, "step": 637 }, { "epoch": 3.9875, "grad_norm": 1.3125, "learning_rate": 1.9906250000000003e-05, "loss": 1.0991, "step": 638 }, { "epoch": 3.99375, "grad_norm": 1.0546875, "learning_rate": 1.99375e-05, "loss": 1.0911, "step": 639 }, { "epoch": 4.0, "grad_norm": 0.91015625, "learning_rate": 1.996875e-05, "loss": 1.0251, "step": 640 }, { "epoch": 4.00625, "grad_norm": 1.046875, "learning_rate": 2e-05, "loss": 1.0847, "step": 641 }, { "epoch": 4.0125, "grad_norm": 1.0390625, "learning_rate": 2.0031250000000002e-05, "loss": 1.0567, "step": 642 }, { "epoch": 4.01875, "grad_norm": 0.96484375, "learning_rate": 2.00625e-05, "loss": 1.0487, "step": 643 }, { "epoch": 4.025, "grad_norm": 0.91796875, "learning_rate": 2.009375e-05, "loss": 1.0374, "step": 644 }, { "epoch": 4.03125, "grad_norm": 0.9296875, "learning_rate": 2.0125e-05, "loss": 1.0681, "step": 645 }, { "epoch": 4.0375, "grad_norm": 1.125, "learning_rate": 2.0156250000000002e-05, "loss": 1.046, "step": 646 }, { "epoch": 4.04375, "grad_norm": 0.9453125, "learning_rate": 2.01875e-05, "loss": 1.0203, "step": 647 }, { "epoch": 4.05, "grad_norm": 1.1171875, "learning_rate": 2.021875e-05, "loss": 1.0918, "step": 648 }, { "epoch": 4.05625, "grad_norm": 0.9609375, "learning_rate": 2.025e-05, "loss": 1.0853, "step": 649 }, { "epoch": 4.0625, "grad_norm": 1.015625, "learning_rate": 2.0281250000000002e-05, "loss": 1.0791, "step": 650 }, { "epoch": 4.06875, "grad_norm": 1.109375, "learning_rate": 2.0312500000000002e-05, "loss": 1.0606, "step": 651 }, { "epoch": 4.075, "grad_norm": 1.0390625, "learning_rate": 2.034375e-05, "loss": 1.0616, "step": 652 }, { "epoch": 4.08125, "grad_norm": 1.09375, "learning_rate": 2.0375e-05, "loss": 1.0889, "step": 653 }, { "epoch": 4.0875, "grad_norm": 1.0, "learning_rate": 2.040625e-05, "loss": 1.0363, "step": 654 }, { "epoch": 4.09375, "grad_norm": 1.078125, "learning_rate": 2.0437500000000002e-05, "loss": 1.0532, "step": 655 }, { "epoch": 4.1, "grad_norm": 1.0546875, "learning_rate": 2.046875e-05, "loss": 1.058, "step": 656 }, { "epoch": 4.10625, "grad_norm": 0.97265625, "learning_rate": 2.05e-05, "loss": 1.0708, "step": 657 }, { "epoch": 4.1125, "grad_norm": 0.9453125, "learning_rate": 2.053125e-05, "loss": 1.0243, "step": 658 }, { "epoch": 4.11875, "grad_norm": 1.015625, "learning_rate": 2.0562500000000002e-05, "loss": 1.0842, "step": 659 }, { "epoch": 4.125, "grad_norm": 0.94140625, "learning_rate": 2.059375e-05, "loss": 1.0552, "step": 660 }, { "epoch": 4.13125, "grad_norm": 1.0859375, "learning_rate": 2.0625e-05, "loss": 1.0352, "step": 661 }, { "epoch": 4.1375, "grad_norm": 1.2109375, "learning_rate": 2.065625e-05, "loss": 1.0719, "step": 662 }, { "epoch": 4.14375, "grad_norm": 1.25, "learning_rate": 2.06875e-05, "loss": 1.05, "step": 663 }, { "epoch": 4.15, "grad_norm": 1.0546875, "learning_rate": 2.0718750000000002e-05, "loss": 1.0607, "step": 664 }, { "epoch": 4.15625, "grad_norm": 1.1484375, "learning_rate": 2.075e-05, "loss": 1.103, "step": 665 }, { "epoch": 4.1625, "grad_norm": 1.203125, "learning_rate": 2.0781250000000004e-05, "loss": 1.0534, "step": 666 }, { "epoch": 4.16875, "grad_norm": 1.046875, "learning_rate": 2.08125e-05, "loss": 1.0425, "step": 667 }, { "epoch": 4.175, "grad_norm": 0.96484375, "learning_rate": 2.0843750000000002e-05, "loss": 1.0569, "step": 668 }, { "epoch": 4.18125, "grad_norm": 1.046875, "learning_rate": 2.0875e-05, "loss": 1.0213, "step": 669 }, { "epoch": 4.1875, "grad_norm": 0.9609375, "learning_rate": 2.0906250000000003e-05, "loss": 1.0598, "step": 670 }, { "epoch": 4.19375, "grad_norm": 1.0625, "learning_rate": 2.09375e-05, "loss": 1.052, "step": 671 }, { "epoch": 4.2, "grad_norm": 0.98828125, "learning_rate": 2.096875e-05, "loss": 1.0412, "step": 672 }, { "epoch": 4.20625, "grad_norm": 1.40625, "learning_rate": 2.1e-05, "loss": 1.0557, "step": 673 }, { "epoch": 4.2125, "grad_norm": 0.99609375, "learning_rate": 2.1031250000000003e-05, "loss": 1.0427, "step": 674 }, { "epoch": 4.21875, "grad_norm": 1.1015625, "learning_rate": 2.10625e-05, "loss": 1.0637, "step": 675 }, { "epoch": 4.225, "grad_norm": 1.265625, "learning_rate": 2.109375e-05, "loss": 1.0752, "step": 676 }, { "epoch": 4.23125, "grad_norm": 1.1015625, "learning_rate": 2.1125000000000002e-05, "loss": 1.0448, "step": 677 }, { "epoch": 4.2375, "grad_norm": 1.0625, "learning_rate": 2.115625e-05, "loss": 1.0874, "step": 678 }, { "epoch": 4.24375, "grad_norm": 1.0625, "learning_rate": 2.1187500000000003e-05, "loss": 1.0862, "step": 679 }, { "epoch": 4.25, "grad_norm": 1.390625, "learning_rate": 2.121875e-05, "loss": 1.0943, "step": 680 }, { "epoch": 4.25625, "grad_norm": 0.953125, "learning_rate": 2.125e-05, "loss": 1.056, "step": 681 }, { "epoch": 4.2625, "grad_norm": 1.15625, "learning_rate": 2.128125e-05, "loss": 1.0381, "step": 682 }, { "epoch": 4.26875, "grad_norm": 0.9609375, "learning_rate": 2.1312500000000003e-05, "loss": 1.0688, "step": 683 }, { "epoch": 4.275, "grad_norm": 1.3828125, "learning_rate": 2.134375e-05, "loss": 1.0438, "step": 684 }, { "epoch": 4.28125, "grad_norm": 1.1796875, "learning_rate": 2.1375e-05, "loss": 1.0462, "step": 685 }, { "epoch": 4.2875, "grad_norm": 1.140625, "learning_rate": 2.140625e-05, "loss": 1.0592, "step": 686 }, { "epoch": 4.29375, "grad_norm": 1.0234375, "learning_rate": 2.1437500000000003e-05, "loss": 1.0211, "step": 687 }, { "epoch": 4.3, "grad_norm": 1.0078125, "learning_rate": 2.146875e-05, "loss": 1.0521, "step": 688 }, { "epoch": 4.30625, "grad_norm": 1.03125, "learning_rate": 2.15e-05, "loss": 1.068, "step": 689 }, { "epoch": 4.3125, "grad_norm": 1.09375, "learning_rate": 2.153125e-05, "loss": 1.0426, "step": 690 }, { "epoch": 4.31875, "grad_norm": 1.1875, "learning_rate": 2.1562500000000002e-05, "loss": 1.0797, "step": 691 }, { "epoch": 4.325, "grad_norm": 1.015625, "learning_rate": 2.1593750000000003e-05, "loss": 1.0664, "step": 692 }, { "epoch": 4.33125, "grad_norm": 0.97265625, "learning_rate": 2.1625e-05, "loss": 1.0284, "step": 693 }, { "epoch": 4.3375, "grad_norm": 1.0390625, "learning_rate": 2.165625e-05, "loss": 1.0127, "step": 694 }, { "epoch": 4.34375, "grad_norm": 0.96875, "learning_rate": 2.1687500000000002e-05, "loss": 1.0303, "step": 695 }, { "epoch": 4.35, "grad_norm": 1.1484375, "learning_rate": 2.1718750000000003e-05, "loss": 1.1018, "step": 696 }, { "epoch": 4.35625, "grad_norm": 1.171875, "learning_rate": 2.175e-05, "loss": 1.0598, "step": 697 }, { "epoch": 4.3625, "grad_norm": 1.078125, "learning_rate": 2.178125e-05, "loss": 1.0785, "step": 698 }, { "epoch": 4.36875, "grad_norm": 1.09375, "learning_rate": 2.18125e-05, "loss": 1.0852, "step": 699 }, { "epoch": 4.375, "grad_norm": 1.0, "learning_rate": 2.1843750000000002e-05, "loss": 1.0393, "step": 700 }, { "epoch": 4.38125, "grad_norm": 1.109375, "learning_rate": 2.1875e-05, "loss": 1.0616, "step": 701 }, { "epoch": 4.3875, "grad_norm": 0.9765625, "learning_rate": 2.190625e-05, "loss": 1.0304, "step": 702 }, { "epoch": 4.39375, "grad_norm": 1.9140625, "learning_rate": 2.19375e-05, "loss": 1.0724, "step": 703 }, { "epoch": 4.4, "grad_norm": 1.6171875, "learning_rate": 2.1968750000000002e-05, "loss": 1.0331, "step": 704 }, { "epoch": 4.40625, "grad_norm": 1.0234375, "learning_rate": 2.2000000000000003e-05, "loss": 1.0287, "step": 705 }, { "epoch": 4.4125, "grad_norm": 1.234375, "learning_rate": 2.203125e-05, "loss": 1.0732, "step": 706 }, { "epoch": 4.41875, "grad_norm": 1.25, "learning_rate": 2.20625e-05, "loss": 1.0561, "step": 707 }, { "epoch": 4.425, "grad_norm": 1.28125, "learning_rate": 2.2093750000000002e-05, "loss": 1.028, "step": 708 }, { "epoch": 4.43125, "grad_norm": 0.9375, "learning_rate": 2.2125000000000002e-05, "loss": 1.0398, "step": 709 }, { "epoch": 4.4375, "grad_norm": 1.0546875, "learning_rate": 2.215625e-05, "loss": 1.0704, "step": 710 }, { "epoch": 4.44375, "grad_norm": 1.0703125, "learning_rate": 2.21875e-05, "loss": 1.0471, "step": 711 }, { "epoch": 4.45, "grad_norm": 1.0859375, "learning_rate": 2.221875e-05, "loss": 1.0107, "step": 712 }, { "epoch": 4.45625, "grad_norm": 1.0625, "learning_rate": 2.2250000000000002e-05, "loss": 1.0744, "step": 713 }, { "epoch": 4.4625, "grad_norm": 0.98828125, "learning_rate": 2.228125e-05, "loss": 1.0465, "step": 714 }, { "epoch": 4.46875, "grad_norm": 1.09375, "learning_rate": 2.23125e-05, "loss": 1.0488, "step": 715 }, { "epoch": 4.475, "grad_norm": 0.9765625, "learning_rate": 2.234375e-05, "loss": 1.046, "step": 716 }, { "epoch": 4.48125, "grad_norm": 0.96875, "learning_rate": 2.2375000000000002e-05, "loss": 1.0517, "step": 717 }, { "epoch": 4.4875, "grad_norm": 1.2265625, "learning_rate": 2.2406250000000003e-05, "loss": 1.0748, "step": 718 }, { "epoch": 4.49375, "grad_norm": 1.0546875, "learning_rate": 2.24375e-05, "loss": 1.0251, "step": 719 }, { "epoch": 4.5, "grad_norm": 1.1328125, "learning_rate": 2.246875e-05, "loss": 1.0621, "step": 720 }, { "epoch": 4.50625, "grad_norm": 1.125, "learning_rate": 2.25e-05, "loss": 1.0898, "step": 721 }, { "epoch": 4.5125, "grad_norm": 1.21875, "learning_rate": 2.2531250000000002e-05, "loss": 1.0384, "step": 722 }, { "epoch": 4.51875, "grad_norm": 1.078125, "learning_rate": 2.25625e-05, "loss": 1.025, "step": 723 }, { "epoch": 4.525, "grad_norm": 1.0703125, "learning_rate": 2.2593750000000004e-05, "loss": 1.0729, "step": 724 }, { "epoch": 4.53125, "grad_norm": 1.0234375, "learning_rate": 2.2625e-05, "loss": 1.0566, "step": 725 }, { "epoch": 4.5375, "grad_norm": 1.015625, "learning_rate": 2.2656250000000002e-05, "loss": 1.0773, "step": 726 }, { "epoch": 4.54375, "grad_norm": 1.109375, "learning_rate": 2.26875e-05, "loss": 1.0454, "step": 727 }, { "epoch": 4.55, "grad_norm": 1.328125, "learning_rate": 2.271875e-05, "loss": 1.0072, "step": 728 }, { "epoch": 4.55625, "grad_norm": 1.0234375, "learning_rate": 2.275e-05, "loss": 1.0428, "step": 729 }, { "epoch": 4.5625, "grad_norm": 1.21875, "learning_rate": 2.278125e-05, "loss": 1.0293, "step": 730 }, { "epoch": 4.56875, "grad_norm": 1.3671875, "learning_rate": 2.28125e-05, "loss": 1.0673, "step": 731 }, { "epoch": 4.575, "grad_norm": 0.9375, "learning_rate": 2.284375e-05, "loss": 1.0548, "step": 732 }, { "epoch": 4.58125, "grad_norm": 1.046875, "learning_rate": 2.2875e-05, "loss": 1.0311, "step": 733 }, { "epoch": 4.5875, "grad_norm": 1.0, "learning_rate": 2.290625e-05, "loss": 1.0277, "step": 734 }, { "epoch": 4.59375, "grad_norm": 1.0, "learning_rate": 2.2937500000000002e-05, "loss": 1.0444, "step": 735 }, { "epoch": 4.6, "grad_norm": 1.1328125, "learning_rate": 2.296875e-05, "loss": 1.0635, "step": 736 }, { "epoch": 4.60625, "grad_norm": 0.90234375, "learning_rate": 2.3000000000000003e-05, "loss": 1.0333, "step": 737 }, { "epoch": 4.6125, "grad_norm": 0.94140625, "learning_rate": 2.303125e-05, "loss": 1.0342, "step": 738 }, { "epoch": 4.61875, "grad_norm": 0.94140625, "learning_rate": 2.30625e-05, "loss": 0.998, "step": 739 }, { "epoch": 4.625, "grad_norm": 1.21875, "learning_rate": 2.309375e-05, "loss": 1.0136, "step": 740 }, { "epoch": 4.63125, "grad_norm": 1.0625, "learning_rate": 2.3125000000000003e-05, "loss": 1.098, "step": 741 }, { "epoch": 4.6375, "grad_norm": 1.1640625, "learning_rate": 2.315625e-05, "loss": 1.0017, "step": 742 }, { "epoch": 4.64375, "grad_norm": 1.03125, "learning_rate": 2.31875e-05, "loss": 1.0479, "step": 743 }, { "epoch": 4.65, "grad_norm": 1.09375, "learning_rate": 2.321875e-05, "loss": 1.0782, "step": 744 }, { "epoch": 4.65625, "grad_norm": 1.3203125, "learning_rate": 2.3250000000000003e-05, "loss": 1.0267, "step": 745 }, { "epoch": 4.6625, "grad_norm": 1.1875, "learning_rate": 2.328125e-05, "loss": 1.0221, "step": 746 }, { "epoch": 4.66875, "grad_norm": 1.2421875, "learning_rate": 2.33125e-05, "loss": 1.021, "step": 747 }, { "epoch": 4.675, "grad_norm": 1.4375, "learning_rate": 2.334375e-05, "loss": 1.0417, "step": 748 }, { "epoch": 4.68125, "grad_norm": 1.25, "learning_rate": 2.3375000000000002e-05, "loss": 1.0483, "step": 749 }, { "epoch": 4.6875, "grad_norm": 1.15625, "learning_rate": 2.3406250000000003e-05, "loss": 1.009, "step": 750 }, { "epoch": 4.69375, "grad_norm": 1.234375, "learning_rate": 2.34375e-05, "loss": 1.0332, "step": 751 }, { "epoch": 4.7, "grad_norm": 1.2421875, "learning_rate": 2.346875e-05, "loss": 1.0522, "step": 752 }, { "epoch": 4.70625, "grad_norm": 1.234375, "learning_rate": 2.35e-05, "loss": 1.0363, "step": 753 }, { "epoch": 4.7125, "grad_norm": 1.03125, "learning_rate": 2.3531250000000003e-05, "loss": 0.9953, "step": 754 }, { "epoch": 4.71875, "grad_norm": 1.1796875, "learning_rate": 2.35625e-05, "loss": 1.0273, "step": 755 }, { "epoch": 4.725, "grad_norm": 1.046875, "learning_rate": 2.359375e-05, "loss": 1.0335, "step": 756 }, { "epoch": 4.73125, "grad_norm": 1.09375, "learning_rate": 2.3624999999999998e-05, "loss": 1.0209, "step": 757 }, { "epoch": 4.7375, "grad_norm": 1.0234375, "learning_rate": 2.3656250000000002e-05, "loss": 1.0608, "step": 758 }, { "epoch": 4.74375, "grad_norm": 0.984375, "learning_rate": 2.36875e-05, "loss": 1.0451, "step": 759 }, { "epoch": 4.75, "grad_norm": 1.2890625, "learning_rate": 2.371875e-05, "loss": 1.0102, "step": 760 }, { "epoch": 4.75625, "grad_norm": 1.03125, "learning_rate": 2.375e-05, "loss": 0.9997, "step": 761 }, { "epoch": 4.7625, "grad_norm": 1.2109375, "learning_rate": 2.3781250000000002e-05, "loss": 1.0168, "step": 762 }, { "epoch": 4.76875, "grad_norm": 1.046875, "learning_rate": 2.3812500000000003e-05, "loss": 1.082, "step": 763 }, { "epoch": 4.775, "grad_norm": 1.125, "learning_rate": 2.384375e-05, "loss": 1.0346, "step": 764 }, { "epoch": 4.78125, "grad_norm": 2.015625, "learning_rate": 2.3875e-05, "loss": 1.0744, "step": 765 }, { "epoch": 4.7875, "grad_norm": 1.109375, "learning_rate": 2.3906250000000002e-05, "loss": 1.0619, "step": 766 }, { "epoch": 4.79375, "grad_norm": 1.28125, "learning_rate": 2.3937500000000002e-05, "loss": 1.0246, "step": 767 }, { "epoch": 4.8, "grad_norm": 1.171875, "learning_rate": 2.396875e-05, "loss": 1.0575, "step": 768 }, { "epoch": 4.80625, "grad_norm": 1.3359375, "learning_rate": 2.4e-05, "loss": 1.0557, "step": 769 }, { "epoch": 4.8125, "grad_norm": 1.109375, "learning_rate": 2.403125e-05, "loss": 1.0296, "step": 770 }, { "epoch": 4.81875, "grad_norm": 1.5234375, "learning_rate": 2.4062500000000002e-05, "loss": 1.0591, "step": 771 }, { "epoch": 4.825, "grad_norm": 1.03125, "learning_rate": 2.409375e-05, "loss": 1.0176, "step": 772 }, { "epoch": 4.83125, "grad_norm": 1.0390625, "learning_rate": 2.4125e-05, "loss": 1.0404, "step": 773 }, { "epoch": 4.8375, "grad_norm": 1.1328125, "learning_rate": 2.415625e-05, "loss": 1.0757, "step": 774 }, { "epoch": 4.84375, "grad_norm": 1.09375, "learning_rate": 2.4187500000000002e-05, "loss": 0.9851, "step": 775 }, { "epoch": 4.85, "grad_norm": 1.15625, "learning_rate": 2.4218750000000003e-05, "loss": 1.0158, "step": 776 }, { "epoch": 4.85625, "grad_norm": 1.21875, "learning_rate": 2.425e-05, "loss": 0.9907, "step": 777 }, { "epoch": 4.8625, "grad_norm": 1.1484375, "learning_rate": 2.428125e-05, "loss": 1.0343, "step": 778 }, { "epoch": 4.86875, "grad_norm": 1.0859375, "learning_rate": 2.43125e-05, "loss": 1.0063, "step": 779 }, { "epoch": 4.875, "grad_norm": 1.1171875, "learning_rate": 2.4343750000000002e-05, "loss": 1.0369, "step": 780 }, { "epoch": 4.88125, "grad_norm": 1.1171875, "learning_rate": 2.4375e-05, "loss": 1.0196, "step": 781 }, { "epoch": 4.8875, "grad_norm": 1.0390625, "learning_rate": 2.440625e-05, "loss": 1.0297, "step": 782 }, { "epoch": 4.89375, "grad_norm": 1.1953125, "learning_rate": 2.44375e-05, "loss": 1.0058, "step": 783 }, { "epoch": 4.9, "grad_norm": 1.140625, "learning_rate": 2.4468750000000002e-05, "loss": 1.0428, "step": 784 }, { "epoch": 4.90625, "grad_norm": 1.0, "learning_rate": 2.45e-05, "loss": 1.0213, "step": 785 }, { "epoch": 4.9125, "grad_norm": 1.234375, "learning_rate": 2.453125e-05, "loss": 1.0142, "step": 786 }, { "epoch": 4.91875, "grad_norm": 1.25, "learning_rate": 2.45625e-05, "loss": 1.0586, "step": 787 }, { "epoch": 4.925, "grad_norm": 1.203125, "learning_rate": 2.459375e-05, "loss": 1.0411, "step": 788 }, { "epoch": 4.93125, "grad_norm": 1.203125, "learning_rate": 2.4625000000000002e-05, "loss": 1.0134, "step": 789 }, { "epoch": 4.9375, "grad_norm": 0.96875, "learning_rate": 2.465625e-05, "loss": 1.0092, "step": 790 }, { "epoch": 4.94375, "grad_norm": 1.2109375, "learning_rate": 2.4687500000000004e-05, "loss": 1.0106, "step": 791 }, { "epoch": 4.95, "grad_norm": 1.1015625, "learning_rate": 2.471875e-05, "loss": 1.0244, "step": 792 }, { "epoch": 4.95625, "grad_norm": 1.21875, "learning_rate": 2.4750000000000002e-05, "loss": 0.988, "step": 793 }, { "epoch": 4.9625, "grad_norm": 1.109375, "learning_rate": 2.478125e-05, "loss": 1.0407, "step": 794 }, { "epoch": 4.96875, "grad_norm": 1.0703125, "learning_rate": 2.4812500000000003e-05, "loss": 1.0295, "step": 795 }, { "epoch": 4.975, "grad_norm": 1.109375, "learning_rate": 2.484375e-05, "loss": 1.037, "step": 796 }, { "epoch": 4.98125, "grad_norm": 1.1171875, "learning_rate": 2.4875e-05, "loss": 0.9739, "step": 797 }, { "epoch": 4.9875, "grad_norm": 1.171875, "learning_rate": 2.490625e-05, "loss": 1.0654, "step": 798 }, { "epoch": 4.99375, "grad_norm": 1.1953125, "learning_rate": 2.4937500000000003e-05, "loss": 1.0233, "step": 799 }, { "epoch": 5.0, "grad_norm": 1.234375, "learning_rate": 2.496875e-05, "loss": 1.0284, "step": 800 }, { "epoch": 5.00625, "grad_norm": 1.015625, "learning_rate": 2.5e-05, "loss": 0.9831, "step": 801 }, { "epoch": 5.0125, "grad_norm": 1.3125, "learning_rate": 2.5031250000000002e-05, "loss": 1.0463, "step": 802 }, { "epoch": 5.01875, "grad_norm": 1.140625, "learning_rate": 2.50625e-05, "loss": 0.9972, "step": 803 }, { "epoch": 5.025, "grad_norm": 1.1015625, "learning_rate": 2.509375e-05, "loss": 1.0424, "step": 804 }, { "epoch": 5.03125, "grad_norm": 1.25, "learning_rate": 2.5124999999999997e-05, "loss": 1.0318, "step": 805 }, { "epoch": 5.0375, "grad_norm": 1.15625, "learning_rate": 2.5156250000000005e-05, "loss": 1.035, "step": 806 }, { "epoch": 5.04375, "grad_norm": 1.1875, "learning_rate": 2.5187500000000002e-05, "loss": 1.0017, "step": 807 }, { "epoch": 5.05, "grad_norm": 1.1015625, "learning_rate": 2.5218750000000003e-05, "loss": 1.0479, "step": 808 }, { "epoch": 5.05625, "grad_norm": 1.0703125, "learning_rate": 2.525e-05, "loss": 1.0267, "step": 809 }, { "epoch": 5.0625, "grad_norm": 1.03125, "learning_rate": 2.528125e-05, "loss": 0.9754, "step": 810 }, { "epoch": 5.06875, "grad_norm": 1.09375, "learning_rate": 2.53125e-05, "loss": 1.0386, "step": 811 }, { "epoch": 5.075, "grad_norm": 1.015625, "learning_rate": 2.534375e-05, "loss": 0.9997, "step": 812 }, { "epoch": 5.08125, "grad_norm": 1.2265625, "learning_rate": 2.5375e-05, "loss": 1.0365, "step": 813 }, { "epoch": 5.0875, "grad_norm": 1.2578125, "learning_rate": 2.5406250000000004e-05, "loss": 1.0438, "step": 814 }, { "epoch": 5.09375, "grad_norm": 1.140625, "learning_rate": 2.54375e-05, "loss": 1.0026, "step": 815 }, { "epoch": 5.1, "grad_norm": 1.140625, "learning_rate": 2.5468750000000002e-05, "loss": 1.0218, "step": 816 }, { "epoch": 5.10625, "grad_norm": 1.046875, "learning_rate": 2.5500000000000003e-05, "loss": 1.0231, "step": 817 }, { "epoch": 5.1125, "grad_norm": 1.109375, "learning_rate": 2.553125e-05, "loss": 1.0284, "step": 818 }, { "epoch": 5.11875, "grad_norm": 1.3359375, "learning_rate": 2.55625e-05, "loss": 0.9996, "step": 819 }, { "epoch": 5.125, "grad_norm": 1.1015625, "learning_rate": 2.559375e-05, "loss": 1.0375, "step": 820 }, { "epoch": 5.13125, "grad_norm": 1.0625, "learning_rate": 2.5625e-05, "loss": 1.0011, "step": 821 }, { "epoch": 5.1375, "grad_norm": 1.1015625, "learning_rate": 2.5656250000000004e-05, "loss": 1.0089, "step": 822 }, { "epoch": 5.14375, "grad_norm": 1.171875, "learning_rate": 2.5687500000000004e-05, "loss": 1.0185, "step": 823 }, { "epoch": 5.15, "grad_norm": 1.1875, "learning_rate": 2.5718750000000002e-05, "loss": 0.9827, "step": 824 }, { "epoch": 5.15625, "grad_norm": 1.203125, "learning_rate": 2.5750000000000002e-05, "loss": 1.0282, "step": 825 }, { "epoch": 5.1625, "grad_norm": 1.4765625, "learning_rate": 2.578125e-05, "loss": 1.0266, "step": 826 }, { "epoch": 5.16875, "grad_norm": 1.09375, "learning_rate": 2.58125e-05, "loss": 0.989, "step": 827 }, { "epoch": 5.175, "grad_norm": 1.265625, "learning_rate": 2.584375e-05, "loss": 1.0004, "step": 828 }, { "epoch": 5.18125, "grad_norm": 1.40625, "learning_rate": 2.5875e-05, "loss": 1.0617, "step": 829 }, { "epoch": 5.1875, "grad_norm": 1.1328125, "learning_rate": 2.590625e-05, "loss": 0.9883, "step": 830 }, { "epoch": 5.19375, "grad_norm": 1.25, "learning_rate": 2.5937500000000004e-05, "loss": 0.9932, "step": 831 }, { "epoch": 5.2, "grad_norm": 1.1328125, "learning_rate": 2.5968750000000004e-05, "loss": 0.9889, "step": 832 }, { "epoch": 5.20625, "grad_norm": 1.1015625, "learning_rate": 2.6000000000000002e-05, "loss": 1.0209, "step": 833 }, { "epoch": 5.2125, "grad_norm": 1.203125, "learning_rate": 2.6031250000000003e-05, "loss": 1.0146, "step": 834 }, { "epoch": 5.21875, "grad_norm": 1.40625, "learning_rate": 2.60625e-05, "loss": 1.0178, "step": 835 }, { "epoch": 5.225, "grad_norm": 1.1015625, "learning_rate": 2.609375e-05, "loss": 0.9993, "step": 836 }, { "epoch": 5.23125, "grad_norm": 1.0625, "learning_rate": 2.6124999999999998e-05, "loss": 0.9916, "step": 837 }, { "epoch": 5.2375, "grad_norm": 1.328125, "learning_rate": 2.615625e-05, "loss": 1.0199, "step": 838 }, { "epoch": 5.24375, "grad_norm": 1.1640625, "learning_rate": 2.6187500000000003e-05, "loss": 1.0159, "step": 839 }, { "epoch": 5.25, "grad_norm": 1.21875, "learning_rate": 2.6218750000000004e-05, "loss": 1.0023, "step": 840 }, { "epoch": 5.25625, "grad_norm": 1.15625, "learning_rate": 2.625e-05, "loss": 1.0195, "step": 841 }, { "epoch": 5.2625, "grad_norm": 1.1640625, "learning_rate": 2.6281250000000002e-05, "loss": 1.0041, "step": 842 }, { "epoch": 5.26875, "grad_norm": 1.4921875, "learning_rate": 2.6312500000000003e-05, "loss": 1.0109, "step": 843 }, { "epoch": 5.275, "grad_norm": 1.1484375, "learning_rate": 2.634375e-05, "loss": 0.9911, "step": 844 }, { "epoch": 5.28125, "grad_norm": 1.453125, "learning_rate": 2.6375e-05, "loss": 0.9797, "step": 845 }, { "epoch": 5.2875, "grad_norm": 1.15625, "learning_rate": 2.6406249999999998e-05, "loss": 0.9968, "step": 846 }, { "epoch": 5.29375, "grad_norm": 1.109375, "learning_rate": 2.6437500000000002e-05, "loss": 0.9914, "step": 847 }, { "epoch": 5.3, "grad_norm": 1.21875, "learning_rate": 2.6468750000000003e-05, "loss": 1.0096, "step": 848 }, { "epoch": 5.30625, "grad_norm": 1.21875, "learning_rate": 2.6500000000000004e-05, "loss": 0.9805, "step": 849 }, { "epoch": 5.3125, "grad_norm": 1.046875, "learning_rate": 2.653125e-05, "loss": 0.9889, "step": 850 }, { "epoch": 5.31875, "grad_norm": 1.28125, "learning_rate": 2.6562500000000002e-05, "loss": 0.9816, "step": 851 }, { "epoch": 5.325, "grad_norm": 1.15625, "learning_rate": 2.659375e-05, "loss": 0.9624, "step": 852 }, { "epoch": 5.33125, "grad_norm": 1.65625, "learning_rate": 2.6625e-05, "loss": 1.0024, "step": 853 }, { "epoch": 5.3375, "grad_norm": 1.2890625, "learning_rate": 2.665625e-05, "loss": 1.0113, "step": 854 }, { "epoch": 5.34375, "grad_norm": 1.1640625, "learning_rate": 2.6687499999999998e-05, "loss": 0.988, "step": 855 }, { "epoch": 5.35, "grad_norm": 1.296875, "learning_rate": 2.6718750000000002e-05, "loss": 0.9882, "step": 856 }, { "epoch": 5.35625, "grad_norm": 1.578125, "learning_rate": 2.6750000000000003e-05, "loss": 0.9853, "step": 857 }, { "epoch": 5.3625, "grad_norm": 1.25, "learning_rate": 2.6781250000000004e-05, "loss": 1.0343, "step": 858 }, { "epoch": 5.36875, "grad_norm": 1.1640625, "learning_rate": 2.68125e-05, "loss": 0.9843, "step": 859 }, { "epoch": 5.375, "grad_norm": 1.15625, "learning_rate": 2.6843750000000002e-05, "loss": 1.0051, "step": 860 }, { "epoch": 5.38125, "grad_norm": 1.234375, "learning_rate": 2.6875e-05, "loss": 1.0446, "step": 861 }, { "epoch": 5.3875, "grad_norm": 1.2265625, "learning_rate": 2.690625e-05, "loss": 0.9987, "step": 862 }, { "epoch": 5.39375, "grad_norm": 1.109375, "learning_rate": 2.6937499999999997e-05, "loss": 0.9622, "step": 863 }, { "epoch": 5.4, "grad_norm": 1.21875, "learning_rate": 2.6968750000000005e-05, "loss": 1.0137, "step": 864 }, { "epoch": 5.40625, "grad_norm": 1.4609375, "learning_rate": 2.7000000000000002e-05, "loss": 1.0085, "step": 865 }, { "epoch": 5.4125, "grad_norm": 1.1953125, "learning_rate": 2.7031250000000003e-05, "loss": 1.004, "step": 866 }, { "epoch": 5.41875, "grad_norm": 1.15625, "learning_rate": 2.70625e-05, "loss": 1.0201, "step": 867 }, { "epoch": 5.425, "grad_norm": 1.09375, "learning_rate": 2.709375e-05, "loss": 1.0074, "step": 868 }, { "epoch": 5.43125, "grad_norm": 1.265625, "learning_rate": 2.7125000000000002e-05, "loss": 1.0043, "step": 869 }, { "epoch": 5.4375, "grad_norm": 1.3125, "learning_rate": 2.715625e-05, "loss": 0.983, "step": 870 }, { "epoch": 5.44375, "grad_norm": 1.28125, "learning_rate": 2.71875e-05, "loss": 0.9744, "step": 871 }, { "epoch": 5.45, "grad_norm": 1.71875, "learning_rate": 2.7218750000000004e-05, "loss": 1.0236, "step": 872 }, { "epoch": 5.45625, "grad_norm": 1.234375, "learning_rate": 2.725e-05, "loss": 0.9897, "step": 873 }, { "epoch": 5.4625, "grad_norm": 1.546875, "learning_rate": 2.7281250000000002e-05, "loss": 0.9626, "step": 874 }, { "epoch": 5.46875, "grad_norm": 1.15625, "learning_rate": 2.7312500000000003e-05, "loss": 0.977, "step": 875 }, { "epoch": 5.475, "grad_norm": 1.171875, "learning_rate": 2.734375e-05, "loss": 1.0037, "step": 876 }, { "epoch": 5.48125, "grad_norm": 1.375, "learning_rate": 2.7375e-05, "loss": 0.991, "step": 877 }, { "epoch": 5.4875, "grad_norm": 1.203125, "learning_rate": 2.740625e-05, "loss": 0.9677, "step": 878 }, { "epoch": 5.49375, "grad_norm": 1.4921875, "learning_rate": 2.74375e-05, "loss": 0.99, "step": 879 }, { "epoch": 5.5, "grad_norm": 1.140625, "learning_rate": 2.746875e-05, "loss": 0.9592, "step": 880 }, { "epoch": 5.50625, "grad_norm": 1.125, "learning_rate": 2.7500000000000004e-05, "loss": 0.9717, "step": 881 }, { "epoch": 5.5125, "grad_norm": 1.1484375, "learning_rate": 2.7531250000000002e-05, "loss": 0.9801, "step": 882 }, { "epoch": 5.51875, "grad_norm": 1.25, "learning_rate": 2.7562500000000002e-05, "loss": 0.9707, "step": 883 }, { "epoch": 5.525, "grad_norm": 1.15625, "learning_rate": 2.759375e-05, "loss": 0.9658, "step": 884 }, { "epoch": 5.53125, "grad_norm": 1.1875, "learning_rate": 2.7625e-05, "loss": 0.9959, "step": 885 }, { "epoch": 5.5375, "grad_norm": 1.2109375, "learning_rate": 2.765625e-05, "loss": 0.9592, "step": 886 }, { "epoch": 5.54375, "grad_norm": 1.1953125, "learning_rate": 2.76875e-05, "loss": 0.9917, "step": 887 }, { "epoch": 5.55, "grad_norm": 1.3203125, "learning_rate": 2.771875e-05, "loss": 0.9704, "step": 888 }, { "epoch": 5.55625, "grad_norm": 1.171875, "learning_rate": 2.7750000000000004e-05, "loss": 0.9359, "step": 889 }, { "epoch": 5.5625, "grad_norm": 1.2890625, "learning_rate": 2.7781250000000004e-05, "loss": 0.9611, "step": 890 }, { "epoch": 5.56875, "grad_norm": 1.328125, "learning_rate": 2.7812500000000002e-05, "loss": 0.9483, "step": 891 }, { "epoch": 5.575, "grad_norm": 1.359375, "learning_rate": 2.7843750000000003e-05, "loss": 0.9911, "step": 892 }, { "epoch": 5.58125, "grad_norm": 1.2109375, "learning_rate": 2.7875e-05, "loss": 0.9666, "step": 893 }, { "epoch": 5.5875, "grad_norm": 1.25, "learning_rate": 2.790625e-05, "loss": 0.9596, "step": 894 }, { "epoch": 5.59375, "grad_norm": 1.5234375, "learning_rate": 2.79375e-05, "loss": 0.9763, "step": 895 }, { "epoch": 5.6, "grad_norm": 1.328125, "learning_rate": 2.796875e-05, "loss": 0.985, "step": 896 }, { "epoch": 5.60625, "grad_norm": 1.1953125, "learning_rate": 2.8000000000000003e-05, "loss": 0.9703, "step": 897 }, { "epoch": 5.6125, "grad_norm": 1.3515625, "learning_rate": 2.8031250000000004e-05, "loss": 0.9867, "step": 898 }, { "epoch": 5.61875, "grad_norm": 1.328125, "learning_rate": 2.80625e-05, "loss": 0.9766, "step": 899 }, { "epoch": 5.625, "grad_norm": 1.2578125, "learning_rate": 2.8093750000000002e-05, "loss": 0.9488, "step": 900 }, { "epoch": 5.63125, "grad_norm": 1.3515625, "learning_rate": 2.8125000000000003e-05, "loss": 0.9649, "step": 901 }, { "epoch": 5.6375, "grad_norm": 1.390625, "learning_rate": 2.815625e-05, "loss": 0.973, "step": 902 }, { "epoch": 5.64375, "grad_norm": 1.28125, "learning_rate": 2.81875e-05, "loss": 0.9828, "step": 903 }, { "epoch": 5.65, "grad_norm": 1.4453125, "learning_rate": 2.8218749999999998e-05, "loss": 0.9738, "step": 904 }, { "epoch": 5.65625, "grad_norm": 1.5234375, "learning_rate": 2.825e-05, "loss": 0.9349, "step": 905 }, { "epoch": 5.6625, "grad_norm": 1.3046875, "learning_rate": 2.8281250000000003e-05, "loss": 0.9249, "step": 906 }, { "epoch": 5.66875, "grad_norm": 1.4375, "learning_rate": 2.8312500000000004e-05, "loss": 0.9337, "step": 907 }, { "epoch": 5.675, "grad_norm": 1.3671875, "learning_rate": 2.834375e-05, "loss": 0.9554, "step": 908 }, { "epoch": 5.68125, "grad_norm": 1.421875, "learning_rate": 2.8375000000000002e-05, "loss": 0.9098, "step": 909 }, { "epoch": 5.6875, "grad_norm": 1.296875, "learning_rate": 2.840625e-05, "loss": 0.9448, "step": 910 }, { "epoch": 5.69375, "grad_norm": 1.375, "learning_rate": 2.84375e-05, "loss": 0.9847, "step": 911 }, { "epoch": 5.7, "grad_norm": 1.4296875, "learning_rate": 2.846875e-05, "loss": 0.9724, "step": 912 }, { "epoch": 5.70625, "grad_norm": 1.7421875, "learning_rate": 2.8499999999999998e-05, "loss": 0.9491, "step": 913 }, { "epoch": 5.7125, "grad_norm": 1.34375, "learning_rate": 2.8531250000000002e-05, "loss": 0.9522, "step": 914 }, { "epoch": 5.71875, "grad_norm": 1.359375, "learning_rate": 2.8562500000000003e-05, "loss": 0.9727, "step": 915 }, { "epoch": 5.725, "grad_norm": 1.3125, "learning_rate": 2.8593750000000004e-05, "loss": 0.9486, "step": 916 }, { "epoch": 5.73125, "grad_norm": 1.359375, "learning_rate": 2.8625e-05, "loss": 0.9512, "step": 917 }, { "epoch": 5.7375, "grad_norm": 1.3203125, "learning_rate": 2.8656250000000002e-05, "loss": 0.9193, "step": 918 }, { "epoch": 5.74375, "grad_norm": 1.421875, "learning_rate": 2.86875e-05, "loss": 0.9505, "step": 919 }, { "epoch": 5.75, "grad_norm": 1.4765625, "learning_rate": 2.871875e-05, "loss": 0.9607, "step": 920 }, { "epoch": 5.75625, "grad_norm": 1.359375, "learning_rate": 2.8749999999999997e-05, "loss": 0.9301, "step": 921 }, { "epoch": 5.7625, "grad_norm": 1.3203125, "learning_rate": 2.8781250000000005e-05, "loss": 0.9836, "step": 922 }, { "epoch": 5.76875, "grad_norm": 1.8046875, "learning_rate": 2.8812500000000002e-05, "loss": 0.9546, "step": 923 }, { "epoch": 5.775, "grad_norm": 2.203125, "learning_rate": 2.8843750000000003e-05, "loss": 0.9384, "step": 924 }, { "epoch": 5.78125, "grad_norm": 1.4296875, "learning_rate": 2.8875e-05, "loss": 0.9535, "step": 925 }, { "epoch": 5.7875, "grad_norm": 2.0, "learning_rate": 2.890625e-05, "loss": 0.9659, "step": 926 }, { "epoch": 5.79375, "grad_norm": 1.625, "learning_rate": 2.8937500000000002e-05, "loss": 0.9054, "step": 927 }, { "epoch": 5.8, "grad_norm": 1.328125, "learning_rate": 2.896875e-05, "loss": 0.953, "step": 928 }, { "epoch": 5.80625, "grad_norm": 1.2578125, "learning_rate": 2.9e-05, "loss": 0.9348, "step": 929 }, { "epoch": 5.8125, "grad_norm": 1.96875, "learning_rate": 2.9031249999999998e-05, "loss": 0.9432, "step": 930 }, { "epoch": 5.81875, "grad_norm": 3.171875, "learning_rate": 2.9062500000000005e-05, "loss": 0.9421, "step": 931 }, { "epoch": 5.825, "grad_norm": 1.9375, "learning_rate": 2.9093750000000002e-05, "loss": 0.944, "step": 932 }, { "epoch": 5.83125, "grad_norm": 1.5078125, "learning_rate": 2.9125000000000003e-05, "loss": 0.9156, "step": 933 }, { "epoch": 5.8375, "grad_norm": 1.953125, "learning_rate": 2.915625e-05, "loss": 0.9479, "step": 934 }, { "epoch": 5.84375, "grad_norm": 2.0625, "learning_rate": 2.91875e-05, "loss": 0.9612, "step": 935 }, { "epoch": 5.85, "grad_norm": 1.765625, "learning_rate": 2.921875e-05, "loss": 0.9489, "step": 936 }, { "epoch": 5.85625, "grad_norm": 1.5390625, "learning_rate": 2.925e-05, "loss": 0.9276, "step": 937 }, { "epoch": 5.8625, "grad_norm": 1.453125, "learning_rate": 2.928125e-05, "loss": 0.9268, "step": 938 }, { "epoch": 5.86875, "grad_norm": 1.3359375, "learning_rate": 2.9312500000000004e-05, "loss": 0.9177, "step": 939 }, { "epoch": 5.875, "grad_norm": 2.453125, "learning_rate": 2.9343750000000002e-05, "loss": 0.9974, "step": 940 }, { "epoch": 5.88125, "grad_norm": 2.53125, "learning_rate": 2.9375000000000003e-05, "loss": 0.9513, "step": 941 }, { "epoch": 5.8875, "grad_norm": 2.421875, "learning_rate": 2.9406250000000003e-05, "loss": 0.9648, "step": 942 }, { "epoch": 5.89375, "grad_norm": 1.453125, "learning_rate": 2.94375e-05, "loss": 0.8853, "step": 943 }, { "epoch": 5.9, "grad_norm": 1.6640625, "learning_rate": 2.946875e-05, "loss": 0.901, "step": 944 }, { "epoch": 5.90625, "grad_norm": 2.203125, "learning_rate": 2.95e-05, "loss": 0.9136, "step": 945 }, { "epoch": 5.9125, "grad_norm": 1.2578125, "learning_rate": 2.953125e-05, "loss": 0.9284, "step": 946 }, { "epoch": 5.91875, "grad_norm": 1.34375, "learning_rate": 2.9562500000000004e-05, "loss": 0.9201, "step": 947 }, { "epoch": 5.925, "grad_norm": 1.3125, "learning_rate": 2.9593750000000004e-05, "loss": 0.9716, "step": 948 }, { "epoch": 5.93125, "grad_norm": 1.4765625, "learning_rate": 2.9625000000000002e-05, "loss": 0.9433, "step": 949 }, { "epoch": 5.9375, "grad_norm": 1.5390625, "learning_rate": 2.9656250000000003e-05, "loss": 0.8923, "step": 950 }, { "epoch": 5.94375, "grad_norm": 1.828125, "learning_rate": 2.96875e-05, "loss": 0.9648, "step": 951 }, { "epoch": 5.95, "grad_norm": 1.265625, "learning_rate": 2.971875e-05, "loss": 0.9085, "step": 952 }, { "epoch": 5.95625, "grad_norm": 1.484375, "learning_rate": 2.975e-05, "loss": 0.9158, "step": 953 }, { "epoch": 5.9625, "grad_norm": 1.453125, "learning_rate": 2.978125e-05, "loss": 0.9243, "step": 954 }, { "epoch": 5.96875, "grad_norm": 1.390625, "learning_rate": 2.98125e-05, "loss": 0.9305, "step": 955 }, { "epoch": 5.975, "grad_norm": 1.5859375, "learning_rate": 2.9843750000000004e-05, "loss": 0.9329, "step": 956 }, { "epoch": 5.98125, "grad_norm": 1.6484375, "learning_rate": 2.9875000000000004e-05, "loss": 0.9267, "step": 957 }, { "epoch": 5.9875, "grad_norm": 1.640625, "learning_rate": 2.9906250000000002e-05, "loss": 0.9107, "step": 958 }, { "epoch": 5.99375, "grad_norm": 3.265625, "learning_rate": 2.9937500000000003e-05, "loss": 0.9458, "step": 959 }, { "epoch": 6.0, "grad_norm": 2.03125, "learning_rate": 2.996875e-05, "loss": 0.9427, "step": 960 }, { "epoch": 6.00625, "grad_norm": 1.65625, "learning_rate": 3e-05, "loss": 0.9286, "step": 961 }, { "epoch": 6.0125, "grad_norm": 1.6953125, "learning_rate": 3.0031249999999998e-05, "loss": 0.9272, "step": 962 }, { "epoch": 6.01875, "grad_norm": 1.8125, "learning_rate": 3.00625e-05, "loss": 0.9258, "step": 963 }, { "epoch": 6.025, "grad_norm": 1.453125, "learning_rate": 3.0093750000000003e-05, "loss": 0.8617, "step": 964 }, { "epoch": 6.03125, "grad_norm": 2.09375, "learning_rate": 3.0125000000000004e-05, "loss": 0.9184, "step": 965 }, { "epoch": 6.0375, "grad_norm": 2.0, "learning_rate": 3.015625e-05, "loss": 0.8988, "step": 966 }, { "epoch": 6.04375, "grad_norm": 2.328125, "learning_rate": 3.0187500000000002e-05, "loss": 0.9235, "step": 967 }, { "epoch": 6.05, "grad_norm": 1.6328125, "learning_rate": 3.0218750000000003e-05, "loss": 0.9038, "step": 968 }, { "epoch": 6.05625, "grad_norm": 1.4375, "learning_rate": 3.025e-05, "loss": 0.9191, "step": 969 }, { "epoch": 6.0625, "grad_norm": 1.609375, "learning_rate": 3.028125e-05, "loss": 0.92, "step": 970 }, { "epoch": 6.06875, "grad_norm": 1.515625, "learning_rate": 3.0312499999999998e-05, "loss": 0.8909, "step": 971 }, { "epoch": 6.075, "grad_norm": 1.5546875, "learning_rate": 3.0343750000000006e-05, "loss": 0.9438, "step": 972 }, { "epoch": 6.08125, "grad_norm": 1.7109375, "learning_rate": 3.0375000000000003e-05, "loss": 0.8423, "step": 973 }, { "epoch": 6.0875, "grad_norm": 1.4140625, "learning_rate": 3.0406250000000004e-05, "loss": 0.8358, "step": 974 }, { "epoch": 6.09375, "grad_norm": 2.515625, "learning_rate": 3.04375e-05, "loss": 0.9094, "step": 975 }, { "epoch": 6.1, "grad_norm": 1.5703125, "learning_rate": 3.0468750000000002e-05, "loss": 0.919, "step": 976 }, { "epoch": 6.10625, "grad_norm": 1.640625, "learning_rate": 3.05e-05, "loss": 0.8612, "step": 977 }, { "epoch": 6.1125, "grad_norm": 1.71875, "learning_rate": 3.053125e-05, "loss": 0.8827, "step": 978 }, { "epoch": 6.11875, "grad_norm": 2.59375, "learning_rate": 3.05625e-05, "loss": 0.9205, "step": 979 }, { "epoch": 6.125, "grad_norm": 1.859375, "learning_rate": 3.059375e-05, "loss": 0.9103, "step": 980 }, { "epoch": 6.13125, "grad_norm": 1.71875, "learning_rate": 3.0625000000000006e-05, "loss": 0.8593, "step": 981 }, { "epoch": 6.1375, "grad_norm": 1.5078125, "learning_rate": 3.065625e-05, "loss": 0.8527, "step": 982 }, { "epoch": 6.14375, "grad_norm": 2.234375, "learning_rate": 3.06875e-05, "loss": 0.9023, "step": 983 }, { "epoch": 6.15, "grad_norm": 2.328125, "learning_rate": 3.0718750000000005e-05, "loss": 0.8387, "step": 984 }, { "epoch": 6.15625, "grad_norm": 1.6328125, "learning_rate": 3.075e-05, "loss": 0.9013, "step": 985 }, { "epoch": 6.1625, "grad_norm": 1.8359375, "learning_rate": 3.078125e-05, "loss": 0.9204, "step": 986 }, { "epoch": 6.16875, "grad_norm": 1.4453125, "learning_rate": 3.08125e-05, "loss": 0.8779, "step": 987 }, { "epoch": 6.175, "grad_norm": 1.9140625, "learning_rate": 3.084375e-05, "loss": 0.8228, "step": 988 }, { "epoch": 6.18125, "grad_norm": 1.7734375, "learning_rate": 3.0875000000000005e-05, "loss": 0.9012, "step": 989 }, { "epoch": 6.1875, "grad_norm": 2.0625, "learning_rate": 3.090625e-05, "loss": 0.8921, "step": 990 }, { "epoch": 6.19375, "grad_norm": 1.90625, "learning_rate": 3.09375e-05, "loss": 0.859, "step": 991 }, { "epoch": 6.2, "grad_norm": 1.765625, "learning_rate": 3.0968750000000004e-05, "loss": 0.9114, "step": 992 }, { "epoch": 6.20625, "grad_norm": 1.921875, "learning_rate": 3.1e-05, "loss": 0.8662, "step": 993 }, { "epoch": 6.2125, "grad_norm": 2.15625, "learning_rate": 3.103125e-05, "loss": 0.8255, "step": 994 }, { "epoch": 6.21875, "grad_norm": 1.71875, "learning_rate": 3.10625e-05, "loss": 0.838, "step": 995 }, { "epoch": 6.225, "grad_norm": 1.9140625, "learning_rate": 3.109375e-05, "loss": 0.8933, "step": 996 }, { "epoch": 6.23125, "grad_norm": 1.8515625, "learning_rate": 3.1125000000000004e-05, "loss": 0.9246, "step": 997 }, { "epoch": 6.2375, "grad_norm": 1.6015625, "learning_rate": 3.115625e-05, "loss": 0.8585, "step": 998 }, { "epoch": 6.24375, "grad_norm": 1.7578125, "learning_rate": 3.1187500000000006e-05, "loss": 0.8146, "step": 999 }, { "epoch": 6.25, "grad_norm": 1.71875, "learning_rate": 3.121875e-05, "loss": 0.8661, "step": 1000 }, { "epoch": 6.25625, "grad_norm": 1.7109375, "learning_rate": 3.125e-05, "loss": 0.8933, "step": 1001 }, { "epoch": 6.2625, "grad_norm": 1.578125, "learning_rate": 3.128125e-05, "loss": 0.8662, "step": 1002 }, { "epoch": 6.26875, "grad_norm": 1.734375, "learning_rate": 3.13125e-05, "loss": 0.8784, "step": 1003 }, { "epoch": 6.275, "grad_norm": 1.53125, "learning_rate": 3.134375e-05, "loss": 0.8549, "step": 1004 }, { "epoch": 6.28125, "grad_norm": 1.671875, "learning_rate": 3.1375e-05, "loss": 0.8947, "step": 1005 }, { "epoch": 6.2875, "grad_norm": 1.8125, "learning_rate": 3.140625e-05, "loss": 0.8935, "step": 1006 }, { "epoch": 6.29375, "grad_norm": 1.6875, "learning_rate": 3.1437500000000005e-05, "loss": 0.8642, "step": 1007 }, { "epoch": 6.3, "grad_norm": 1.8203125, "learning_rate": 3.146875e-05, "loss": 0.7961, "step": 1008 }, { "epoch": 6.30625, "grad_norm": 1.625, "learning_rate": 3.15e-05, "loss": 0.9, "step": 1009 }, { "epoch": 6.3125, "grad_norm": 1.5625, "learning_rate": 3.1531250000000004e-05, "loss": 0.8309, "step": 1010 }, { "epoch": 6.31875, "grad_norm": 1.5546875, "learning_rate": 3.15625e-05, "loss": 0.8031, "step": 1011 }, { "epoch": 6.325, "grad_norm": 1.6328125, "learning_rate": 3.159375e-05, "loss": 0.7979, "step": 1012 }, { "epoch": 6.33125, "grad_norm": 1.671875, "learning_rate": 3.1624999999999996e-05, "loss": 0.833, "step": 1013 }, { "epoch": 6.3375, "grad_norm": 1.640625, "learning_rate": 3.165625000000001e-05, "loss": 0.8728, "step": 1014 }, { "epoch": 6.34375, "grad_norm": 1.9765625, "learning_rate": 3.1687500000000005e-05, "loss": 0.8134, "step": 1015 }, { "epoch": 6.35, "grad_norm": 1.7421875, "learning_rate": 3.171875e-05, "loss": 0.8495, "step": 1016 }, { "epoch": 6.35625, "grad_norm": 1.875, "learning_rate": 3.175e-05, "loss": 0.8327, "step": 1017 }, { "epoch": 6.3625, "grad_norm": 1.4453125, "learning_rate": 3.1781250000000003e-05, "loss": 0.8162, "step": 1018 }, { "epoch": 6.36875, "grad_norm": 1.5859375, "learning_rate": 3.18125e-05, "loss": 0.8804, "step": 1019 }, { "epoch": 6.375, "grad_norm": 1.6796875, "learning_rate": 3.184375e-05, "loss": 0.7766, "step": 1020 }, { "epoch": 6.38125, "grad_norm": 2.40625, "learning_rate": 3.1875e-05, "loss": 0.8566, "step": 1021 }, { "epoch": 6.3875, "grad_norm": 2.640625, "learning_rate": 3.1906250000000006e-05, "loss": 0.8141, "step": 1022 }, { "epoch": 6.39375, "grad_norm": 2.140625, "learning_rate": 3.1937500000000004e-05, "loss": 0.8583, "step": 1023 }, { "epoch": 6.4, "grad_norm": 1.875, "learning_rate": 3.196875e-05, "loss": 0.8481, "step": 1024 }, { "epoch": 6.40625, "grad_norm": 2.484375, "learning_rate": 3.2000000000000005e-05, "loss": 0.8366, "step": 1025 }, { "epoch": 6.4125, "grad_norm": 2.921875, "learning_rate": 3.203125e-05, "loss": 0.8617, "step": 1026 }, { "epoch": 6.41875, "grad_norm": 2.140625, "learning_rate": 3.20625e-05, "loss": 0.8301, "step": 1027 }, { "epoch": 6.425, "grad_norm": 2.484375, "learning_rate": 3.209375e-05, "loss": 0.8341, "step": 1028 }, { "epoch": 6.43125, "grad_norm": 2.796875, "learning_rate": 3.2125e-05, "loss": 0.7899, "step": 1029 }, { "epoch": 6.4375, "grad_norm": 2.5, "learning_rate": 3.215625e-05, "loss": 0.8215, "step": 1030 }, { "epoch": 6.44375, "grad_norm": 1.6875, "learning_rate": 3.21875e-05, "loss": 0.8036, "step": 1031 }, { "epoch": 6.45, "grad_norm": 2.375, "learning_rate": 3.221875e-05, "loss": 0.8456, "step": 1032 }, { "epoch": 6.45625, "grad_norm": 2.0625, "learning_rate": 3.2250000000000005e-05, "loss": 0.8512, "step": 1033 }, { "epoch": 6.4625, "grad_norm": 2.125, "learning_rate": 3.228125e-05, "loss": 0.8248, "step": 1034 }, { "epoch": 6.46875, "grad_norm": 1.9375, "learning_rate": 3.23125e-05, "loss": 0.774, "step": 1035 }, { "epoch": 6.475, "grad_norm": 1.9296875, "learning_rate": 3.2343750000000004e-05, "loss": 0.7998, "step": 1036 }, { "epoch": 6.48125, "grad_norm": 2.09375, "learning_rate": 3.2375e-05, "loss": 0.8748, "step": 1037 }, { "epoch": 6.4875, "grad_norm": 1.7265625, "learning_rate": 3.240625e-05, "loss": 0.8221, "step": 1038 }, { "epoch": 6.49375, "grad_norm": 1.765625, "learning_rate": 3.24375e-05, "loss": 0.7615, "step": 1039 }, { "epoch": 6.5, "grad_norm": 1.7421875, "learning_rate": 3.2468750000000007e-05, "loss": 0.8375, "step": 1040 }, { "epoch": 6.50625, "grad_norm": 2.03125, "learning_rate": 3.2500000000000004e-05, "loss": 0.8175, "step": 1041 }, { "epoch": 6.5125, "grad_norm": 1.578125, "learning_rate": 3.253125e-05, "loss": 0.7404, "step": 1042 }, { "epoch": 6.51875, "grad_norm": 1.890625, "learning_rate": 3.25625e-05, "loss": 0.8222, "step": 1043 }, { "epoch": 6.525, "grad_norm": 1.703125, "learning_rate": 3.259375e-05, "loss": 0.8609, "step": 1044 }, { "epoch": 6.53125, "grad_norm": 1.75, "learning_rate": 3.2625e-05, "loss": 0.8007, "step": 1045 }, { "epoch": 6.5375, "grad_norm": 1.921875, "learning_rate": 3.265625e-05, "loss": 0.7505, "step": 1046 }, { "epoch": 6.54375, "grad_norm": 1.90625, "learning_rate": 3.26875e-05, "loss": 0.8026, "step": 1047 }, { "epoch": 6.55, "grad_norm": 2.265625, "learning_rate": 3.2718750000000006e-05, "loss": 0.8398, "step": 1048 }, { "epoch": 6.55625, "grad_norm": 2.15625, "learning_rate": 3.275e-05, "loss": 0.8505, "step": 1049 }, { "epoch": 6.5625, "grad_norm": 2.140625, "learning_rate": 3.278125e-05, "loss": 0.7861, "step": 1050 }, { "epoch": 6.56875, "grad_norm": 1.6484375, "learning_rate": 3.2812500000000005e-05, "loss": 0.767, "step": 1051 }, { "epoch": 6.575, "grad_norm": 1.84375, "learning_rate": 3.284375e-05, "loss": 0.7824, "step": 1052 }, { "epoch": 6.58125, "grad_norm": 1.65625, "learning_rate": 3.2875e-05, "loss": 0.8082, "step": 1053 }, { "epoch": 6.5875, "grad_norm": 1.9609375, "learning_rate": 3.290625e-05, "loss": 0.8284, "step": 1054 }, { "epoch": 6.59375, "grad_norm": 1.9140625, "learning_rate": 3.29375e-05, "loss": 0.8023, "step": 1055 }, { "epoch": 6.6, "grad_norm": 2.140625, "learning_rate": 3.2968750000000005e-05, "loss": 0.7846, "step": 1056 }, { "epoch": 6.60625, "grad_norm": 1.7890625, "learning_rate": 3.3e-05, "loss": 0.7637, "step": 1057 }, { "epoch": 6.6125, "grad_norm": 2.578125, "learning_rate": 3.303125e-05, "loss": 0.7952, "step": 1058 }, { "epoch": 6.61875, "grad_norm": 2.015625, "learning_rate": 3.3062500000000004e-05, "loss": 0.8303, "step": 1059 }, { "epoch": 6.625, "grad_norm": 1.78125, "learning_rate": 3.309375e-05, "loss": 0.7887, "step": 1060 }, { "epoch": 6.63125, "grad_norm": 1.828125, "learning_rate": 3.3125e-05, "loss": 0.7381, "step": 1061 }, { "epoch": 6.6375, "grad_norm": 2.3125, "learning_rate": 3.315625e-05, "loss": 0.8321, "step": 1062 }, { "epoch": 6.64375, "grad_norm": 2.171875, "learning_rate": 3.31875e-05, "loss": 0.7972, "step": 1063 }, { "epoch": 6.65, "grad_norm": 2.375, "learning_rate": 3.3218750000000004e-05, "loss": 0.7728, "step": 1064 }, { "epoch": 6.65625, "grad_norm": 2.078125, "learning_rate": 3.325e-05, "loss": 0.7749, "step": 1065 }, { "epoch": 6.6625, "grad_norm": 1.8515625, "learning_rate": 3.3281250000000006e-05, "loss": 0.7718, "step": 1066 }, { "epoch": 6.66875, "grad_norm": 2.09375, "learning_rate": 3.33125e-05, "loss": 0.7543, "step": 1067 }, { "epoch": 6.675, "grad_norm": 2.203125, "learning_rate": 3.334375e-05, "loss": 0.71, "step": 1068 }, { "epoch": 6.68125, "grad_norm": 1.9609375, "learning_rate": 3.3375e-05, "loss": 0.7529, "step": 1069 }, { "epoch": 6.6875, "grad_norm": 2.34375, "learning_rate": 3.340625e-05, "loss": 0.7382, "step": 1070 }, { "epoch": 6.69375, "grad_norm": 2.203125, "learning_rate": 3.34375e-05, "loss": 0.8304, "step": 1071 }, { "epoch": 6.7, "grad_norm": 1.796875, "learning_rate": 3.3468750000000004e-05, "loss": 0.7856, "step": 1072 }, { "epoch": 6.70625, "grad_norm": 2.140625, "learning_rate": 3.35e-05, "loss": 0.8299, "step": 1073 }, { "epoch": 6.7125, "grad_norm": 2.0, "learning_rate": 3.3531250000000005e-05, "loss": 0.8109, "step": 1074 }, { "epoch": 6.71875, "grad_norm": 1.96875, "learning_rate": 3.35625e-05, "loss": 0.7775, "step": 1075 }, { "epoch": 6.725, "grad_norm": 1.859375, "learning_rate": 3.359375e-05, "loss": 0.7738, "step": 1076 }, { "epoch": 6.73125, "grad_norm": 1.984375, "learning_rate": 3.3625000000000004e-05, "loss": 0.7211, "step": 1077 }, { "epoch": 6.7375, "grad_norm": 1.8515625, "learning_rate": 3.365625e-05, "loss": 0.8208, "step": 1078 }, { "epoch": 6.74375, "grad_norm": 2.21875, "learning_rate": 3.36875e-05, "loss": 0.7566, "step": 1079 }, { "epoch": 6.75, "grad_norm": 1.921875, "learning_rate": 3.3718749999999996e-05, "loss": 0.751, "step": 1080 }, { "epoch": 6.75625, "grad_norm": 2.328125, "learning_rate": 3.375000000000001e-05, "loss": 0.7704, "step": 1081 }, { "epoch": 6.7625, "grad_norm": 2.359375, "learning_rate": 3.3781250000000005e-05, "loss": 0.7839, "step": 1082 }, { "epoch": 6.76875, "grad_norm": 1.7890625, "learning_rate": 3.38125e-05, "loss": 0.7664, "step": 1083 }, { "epoch": 6.775, "grad_norm": 1.9921875, "learning_rate": 3.384375e-05, "loss": 0.7487, "step": 1084 }, { "epoch": 6.78125, "grad_norm": 1.859375, "learning_rate": 3.3875000000000003e-05, "loss": 0.6803, "step": 1085 }, { "epoch": 6.7875, "grad_norm": 1.7578125, "learning_rate": 3.390625e-05, "loss": 0.7709, "step": 1086 }, { "epoch": 6.79375, "grad_norm": 1.7890625, "learning_rate": 3.39375e-05, "loss": 0.7465, "step": 1087 }, { "epoch": 6.8, "grad_norm": 1.8984375, "learning_rate": 3.396875e-05, "loss": 0.7931, "step": 1088 }, { "epoch": 6.80625, "grad_norm": 1.875, "learning_rate": 3.4000000000000007e-05, "loss": 0.808, "step": 1089 }, { "epoch": 6.8125, "grad_norm": 2.109375, "learning_rate": 3.4031250000000004e-05, "loss": 0.7742, "step": 1090 }, { "epoch": 6.81875, "grad_norm": 2.09375, "learning_rate": 3.40625e-05, "loss": 0.7984, "step": 1091 }, { "epoch": 6.825, "grad_norm": 2.296875, "learning_rate": 3.4093750000000005e-05, "loss": 0.7635, "step": 1092 }, { "epoch": 6.83125, "grad_norm": 2.15625, "learning_rate": 3.4125e-05, "loss": 0.7097, "step": 1093 }, { "epoch": 6.8375, "grad_norm": 1.859375, "learning_rate": 3.415625e-05, "loss": 0.8105, "step": 1094 }, { "epoch": 6.84375, "grad_norm": 2.59375, "learning_rate": 3.41875e-05, "loss": 0.7358, "step": 1095 }, { "epoch": 6.85, "grad_norm": 2.171875, "learning_rate": 3.421875e-05, "loss": 0.7984, "step": 1096 }, { "epoch": 6.85625, "grad_norm": 1.9921875, "learning_rate": 3.4250000000000006e-05, "loss": 0.7303, "step": 1097 }, { "epoch": 6.8625, "grad_norm": 3.171875, "learning_rate": 3.428125e-05, "loss": 0.6962, "step": 1098 }, { "epoch": 6.86875, "grad_norm": 2.765625, "learning_rate": 3.43125e-05, "loss": 0.7662, "step": 1099 }, { "epoch": 6.875, "grad_norm": 2.234375, "learning_rate": 3.4343750000000005e-05, "loss": 0.6479, "step": 1100 }, { "epoch": 6.88125, "grad_norm": 2.203125, "learning_rate": 3.4375e-05, "loss": 0.7778, "step": 1101 }, { "epoch": 6.8875, "grad_norm": 2.953125, "learning_rate": 3.440625e-05, "loss": 0.7067, "step": 1102 }, { "epoch": 6.89375, "grad_norm": 2.984375, "learning_rate": 3.4437500000000004e-05, "loss": 0.7078, "step": 1103 }, { "epoch": 6.9, "grad_norm": 2.515625, "learning_rate": 3.446875e-05, "loss": 0.7499, "step": 1104 }, { "epoch": 6.90625, "grad_norm": 2.3125, "learning_rate": 3.45e-05, "loss": 0.6625, "step": 1105 }, { "epoch": 6.9125, "grad_norm": 3.0, "learning_rate": 3.453125e-05, "loss": 0.7265, "step": 1106 }, { "epoch": 6.91875, "grad_norm": 2.421875, "learning_rate": 3.45625e-05, "loss": 0.7516, "step": 1107 }, { "epoch": 6.925, "grad_norm": 2.078125, "learning_rate": 3.4593750000000004e-05, "loss": 0.7689, "step": 1108 }, { "epoch": 6.93125, "grad_norm": 2.078125, "learning_rate": 3.4625e-05, "loss": 0.7221, "step": 1109 }, { "epoch": 6.9375, "grad_norm": 2.078125, "learning_rate": 3.465625e-05, "loss": 0.7202, "step": 1110 }, { "epoch": 6.94375, "grad_norm": 2.25, "learning_rate": 3.46875e-05, "loss": 0.7404, "step": 1111 }, { "epoch": 6.95, "grad_norm": 2.5, "learning_rate": 3.471875e-05, "loss": 0.6933, "step": 1112 }, { "epoch": 6.95625, "grad_norm": 2.390625, "learning_rate": 3.475e-05, "loss": 0.7615, "step": 1113 }, { "epoch": 6.9625, "grad_norm": 2.015625, "learning_rate": 3.478125e-05, "loss": 0.7154, "step": 1114 }, { "epoch": 6.96875, "grad_norm": 1.9140625, "learning_rate": 3.4812500000000006e-05, "loss": 0.7407, "step": 1115 }, { "epoch": 6.975, "grad_norm": 2.28125, "learning_rate": 3.484375e-05, "loss": 0.6595, "step": 1116 }, { "epoch": 6.98125, "grad_norm": 2.171875, "learning_rate": 3.4875e-05, "loss": 0.7005, "step": 1117 }, { "epoch": 6.9875, "grad_norm": 1.9765625, "learning_rate": 3.4906250000000005e-05, "loss": 0.7144, "step": 1118 }, { "epoch": 6.99375, "grad_norm": 2.359375, "learning_rate": 3.49375e-05, "loss": 0.7697, "step": 1119 }, { "epoch": 7.0, "grad_norm": 2.421875, "learning_rate": 3.496875e-05, "loss": 0.6867, "step": 1120 }, { "epoch": 7.00625, "grad_norm": 1.875, "learning_rate": 3.5e-05, "loss": 0.7465, "step": 1121 }, { "epoch": 7.0125, "grad_norm": 2.15625, "learning_rate": 3.503125e-05, "loss": 0.7125, "step": 1122 }, { "epoch": 7.01875, "grad_norm": 3.078125, "learning_rate": 3.5062500000000005e-05, "loss": 0.6792, "step": 1123 }, { "epoch": 7.025, "grad_norm": 2.4375, "learning_rate": 3.509375e-05, "loss": 0.7739, "step": 1124 }, { "epoch": 7.03125, "grad_norm": 2.1875, "learning_rate": 3.5125e-05, "loss": 0.7285, "step": 1125 }, { "epoch": 7.0375, "grad_norm": 2.03125, "learning_rate": 3.5156250000000004e-05, "loss": 0.7305, "step": 1126 }, { "epoch": 7.04375, "grad_norm": 1.734375, "learning_rate": 3.51875e-05, "loss": 0.7175, "step": 1127 }, { "epoch": 7.05, "grad_norm": 1.796875, "learning_rate": 3.521875e-05, "loss": 0.7172, "step": 1128 }, { "epoch": 7.05625, "grad_norm": 1.859375, "learning_rate": 3.525e-05, "loss": 0.668, "step": 1129 }, { "epoch": 7.0625, "grad_norm": 1.8984375, "learning_rate": 3.528125e-05, "loss": 0.685, "step": 1130 }, { "epoch": 7.06875, "grad_norm": 1.7890625, "learning_rate": 3.5312500000000005e-05, "loss": 0.6368, "step": 1131 }, { "epoch": 7.075, "grad_norm": 2.109375, "learning_rate": 3.534375e-05, "loss": 0.7192, "step": 1132 }, { "epoch": 7.08125, "grad_norm": 2.765625, "learning_rate": 3.5375e-05, "loss": 0.7476, "step": 1133 }, { "epoch": 7.0875, "grad_norm": 2.578125, "learning_rate": 3.5406250000000003e-05, "loss": 0.7184, "step": 1134 }, { "epoch": 7.09375, "grad_norm": 2.34375, "learning_rate": 3.54375e-05, "loss": 0.722, "step": 1135 }, { "epoch": 7.1, "grad_norm": 2.203125, "learning_rate": 3.546875e-05, "loss": 0.7571, "step": 1136 }, { "epoch": 7.10625, "grad_norm": 2.203125, "learning_rate": 3.55e-05, "loss": 0.7244, "step": 1137 }, { "epoch": 7.1125, "grad_norm": 1.890625, "learning_rate": 3.553125e-05, "loss": 0.5938, "step": 1138 }, { "epoch": 7.11875, "grad_norm": 2.515625, "learning_rate": 3.5562500000000004e-05, "loss": 0.681, "step": 1139 }, { "epoch": 7.125, "grad_norm": 2.453125, "learning_rate": 3.559375e-05, "loss": 0.7179, "step": 1140 }, { "epoch": 7.13125, "grad_norm": 2.328125, "learning_rate": 3.5625000000000005e-05, "loss": 0.7456, "step": 1141 }, { "epoch": 7.1375, "grad_norm": 1.9140625, "learning_rate": 3.565625e-05, "loss": 0.6968, "step": 1142 }, { "epoch": 7.14375, "grad_norm": 2.125, "learning_rate": 3.56875e-05, "loss": 0.7032, "step": 1143 }, { "epoch": 7.15, "grad_norm": 2.359375, "learning_rate": 3.571875e-05, "loss": 0.669, "step": 1144 }, { "epoch": 7.15625, "grad_norm": 2.078125, "learning_rate": 3.575e-05, "loss": 0.7402, "step": 1145 }, { "epoch": 7.1625, "grad_norm": 2.34375, "learning_rate": 3.578125e-05, "loss": 0.6694, "step": 1146 }, { "epoch": 7.16875, "grad_norm": 2.09375, "learning_rate": 3.58125e-05, "loss": 0.6671, "step": 1147 }, { "epoch": 7.175, "grad_norm": 2.078125, "learning_rate": 3.584375e-05, "loss": 0.697, "step": 1148 }, { "epoch": 7.18125, "grad_norm": 2.21875, "learning_rate": 3.5875000000000005e-05, "loss": 0.6915, "step": 1149 }, { "epoch": 7.1875, "grad_norm": 2.234375, "learning_rate": 3.590625e-05, "loss": 0.7141, "step": 1150 }, { "epoch": 7.19375, "grad_norm": 2.34375, "learning_rate": 3.59375e-05, "loss": 0.6979, "step": 1151 }, { "epoch": 7.2, "grad_norm": 3.015625, "learning_rate": 3.5968750000000004e-05, "loss": 0.7125, "step": 1152 }, { "epoch": 7.20625, "grad_norm": 2.421875, "learning_rate": 3.6e-05, "loss": 0.6971, "step": 1153 }, { "epoch": 7.2125, "grad_norm": 2.046875, "learning_rate": 3.603125e-05, "loss": 0.7352, "step": 1154 }, { "epoch": 7.21875, "grad_norm": 3.546875, "learning_rate": 3.60625e-05, "loss": 0.7155, "step": 1155 }, { "epoch": 7.225, "grad_norm": 3.453125, "learning_rate": 3.6093750000000007e-05, "loss": 0.6646, "step": 1156 }, { "epoch": 7.23125, "grad_norm": 2.28125, "learning_rate": 3.6125000000000004e-05, "loss": 0.6302, "step": 1157 }, { "epoch": 7.2375, "grad_norm": 2.6875, "learning_rate": 3.615625e-05, "loss": 0.6816, "step": 1158 }, { "epoch": 7.24375, "grad_norm": 2.03125, "learning_rate": 3.61875e-05, "loss": 0.6697, "step": 1159 }, { "epoch": 7.25, "grad_norm": 2.15625, "learning_rate": 3.621875e-05, "loss": 0.6035, "step": 1160 }, { "epoch": 7.25625, "grad_norm": 2.0625, "learning_rate": 3.625e-05, "loss": 0.6678, "step": 1161 }, { "epoch": 7.2625, "grad_norm": 2.625, "learning_rate": 3.628125e-05, "loss": 0.7109, "step": 1162 }, { "epoch": 7.26875, "grad_norm": 2.296875, "learning_rate": 3.63125e-05, "loss": 0.7911, "step": 1163 }, { "epoch": 7.275, "grad_norm": 2.59375, "learning_rate": 3.6343750000000006e-05, "loss": 0.6791, "step": 1164 }, { "epoch": 7.28125, "grad_norm": 2.234375, "learning_rate": 3.6375e-05, "loss": 0.6766, "step": 1165 }, { "epoch": 7.2875, "grad_norm": 2.375, "learning_rate": 3.640625e-05, "loss": 0.7194, "step": 1166 }, { "epoch": 7.29375, "grad_norm": 2.328125, "learning_rate": 3.6437500000000005e-05, "loss": 0.672, "step": 1167 }, { "epoch": 7.3, "grad_norm": 1.921875, "learning_rate": 3.646875e-05, "loss": 0.668, "step": 1168 }, { "epoch": 7.30625, "grad_norm": 2.265625, "learning_rate": 3.65e-05, "loss": 0.6772, "step": 1169 }, { "epoch": 7.3125, "grad_norm": 2.109375, "learning_rate": 3.653125e-05, "loss": 0.6428, "step": 1170 }, { "epoch": 7.31875, "grad_norm": 2.3125, "learning_rate": 3.65625e-05, "loss": 0.6926, "step": 1171 }, { "epoch": 7.325, "grad_norm": 2.109375, "learning_rate": 3.6593750000000005e-05, "loss": 0.6609, "step": 1172 }, { "epoch": 7.33125, "grad_norm": 1.8671875, "learning_rate": 3.6625e-05, "loss": 0.7217, "step": 1173 }, { "epoch": 7.3375, "grad_norm": 2.390625, "learning_rate": 3.665625e-05, "loss": 0.7069, "step": 1174 }, { "epoch": 7.34375, "grad_norm": 2.546875, "learning_rate": 3.6687500000000004e-05, "loss": 0.6294, "step": 1175 }, { "epoch": 7.35, "grad_norm": 2.0625, "learning_rate": 3.671875e-05, "loss": 0.6609, "step": 1176 }, { "epoch": 7.35625, "grad_norm": 2.078125, "learning_rate": 3.675e-05, "loss": 0.6953, "step": 1177 }, { "epoch": 7.3625, "grad_norm": 3.0, "learning_rate": 3.678125e-05, "loss": 0.6211, "step": 1178 }, { "epoch": 7.36875, "grad_norm": 2.21875, "learning_rate": 3.68125e-05, "loss": 0.7062, "step": 1179 }, { "epoch": 7.375, "grad_norm": 2.28125, "learning_rate": 3.684375e-05, "loss": 0.661, "step": 1180 }, { "epoch": 7.38125, "grad_norm": 2.28125, "learning_rate": 3.6875e-05, "loss": 0.6487, "step": 1181 }, { "epoch": 7.3875, "grad_norm": 1.875, "learning_rate": 3.6906250000000006e-05, "loss": 0.6468, "step": 1182 }, { "epoch": 7.39375, "grad_norm": 2.359375, "learning_rate": 3.69375e-05, "loss": 0.6909, "step": 1183 }, { "epoch": 7.4, "grad_norm": 2.25, "learning_rate": 3.696875e-05, "loss": 0.6164, "step": 1184 }, { "epoch": 7.40625, "grad_norm": 2.421875, "learning_rate": 3.7e-05, "loss": 0.6546, "step": 1185 }, { "epoch": 7.4125, "grad_norm": 2.703125, "learning_rate": 3.703125e-05, "loss": 0.6625, "step": 1186 }, { "epoch": 7.41875, "grad_norm": 1.90625, "learning_rate": 3.70625e-05, "loss": 0.6359, "step": 1187 }, { "epoch": 7.425, "grad_norm": 2.015625, "learning_rate": 3.709375e-05, "loss": 0.6452, "step": 1188 }, { "epoch": 7.43125, "grad_norm": 1.953125, "learning_rate": 3.7125e-05, "loss": 0.7249, "step": 1189 }, { "epoch": 7.4375, "grad_norm": 2.546875, "learning_rate": 3.7156250000000005e-05, "loss": 0.6033, "step": 1190 }, { "epoch": 7.44375, "grad_norm": 2.546875, "learning_rate": 3.71875e-05, "loss": 0.697, "step": 1191 }, { "epoch": 7.45, "grad_norm": 2.015625, "learning_rate": 3.721875e-05, "loss": 0.6175, "step": 1192 }, { "epoch": 7.45625, "grad_norm": 2.421875, "learning_rate": 3.7250000000000004e-05, "loss": 0.6419, "step": 1193 }, { "epoch": 7.4625, "grad_norm": 2.609375, "learning_rate": 3.728125e-05, "loss": 0.6085, "step": 1194 }, { "epoch": 7.46875, "grad_norm": 2.9375, "learning_rate": 3.73125e-05, "loss": 0.6988, "step": 1195 }, { "epoch": 7.475, "grad_norm": 2.359375, "learning_rate": 3.7343749999999996e-05, "loss": 0.6682, "step": 1196 }, { "epoch": 7.48125, "grad_norm": 2.71875, "learning_rate": 3.737500000000001e-05, "loss": 0.6234, "step": 1197 }, { "epoch": 7.4875, "grad_norm": 2.40625, "learning_rate": 3.7406250000000005e-05, "loss": 0.6781, "step": 1198 }, { "epoch": 7.49375, "grad_norm": 2.078125, "learning_rate": 3.74375e-05, "loss": 0.6616, "step": 1199 }, { "epoch": 7.5, "grad_norm": 2.328125, "learning_rate": 3.746875e-05, "loss": 0.5977, "step": 1200 }, { "epoch": 7.50625, "grad_norm": 3.359375, "learning_rate": 3.7500000000000003e-05, "loss": 0.6061, "step": 1201 }, { "epoch": 7.5125, "grad_norm": 2.3125, "learning_rate": 3.753125e-05, "loss": 0.6944, "step": 1202 }, { "epoch": 7.51875, "grad_norm": 2.046875, "learning_rate": 3.75625e-05, "loss": 0.6492, "step": 1203 }, { "epoch": 7.525, "grad_norm": 3.234375, "learning_rate": 3.759375e-05, "loss": 0.7069, "step": 1204 }, { "epoch": 7.53125, "grad_norm": 2.046875, "learning_rate": 3.7625e-05, "loss": 0.6152, "step": 1205 }, { "epoch": 7.5375, "grad_norm": 1.90625, "learning_rate": 3.7656250000000004e-05, "loss": 0.6749, "step": 1206 }, { "epoch": 7.54375, "grad_norm": 2.796875, "learning_rate": 3.76875e-05, "loss": 0.7188, "step": 1207 }, { "epoch": 7.55, "grad_norm": 2.328125, "learning_rate": 3.7718750000000005e-05, "loss": 0.6371, "step": 1208 }, { "epoch": 7.55625, "grad_norm": 2.078125, "learning_rate": 3.775e-05, "loss": 0.7537, "step": 1209 }, { "epoch": 7.5625, "grad_norm": 1.953125, "learning_rate": 3.778125e-05, "loss": 0.5443, "step": 1210 }, { "epoch": 7.56875, "grad_norm": 2.546875, "learning_rate": 3.78125e-05, "loss": 0.6645, "step": 1211 }, { "epoch": 7.575, "grad_norm": 2.609375, "learning_rate": 3.784375e-05, "loss": 0.6577, "step": 1212 }, { "epoch": 7.58125, "grad_norm": 2.78125, "learning_rate": 3.7875e-05, "loss": 0.7054, "step": 1213 }, { "epoch": 7.5875, "grad_norm": 2.609375, "learning_rate": 3.790625e-05, "loss": 0.621, "step": 1214 }, { "epoch": 7.59375, "grad_norm": 2.109375, "learning_rate": 3.79375e-05, "loss": 0.6706, "step": 1215 }, { "epoch": 7.6, "grad_norm": 2.015625, "learning_rate": 3.7968750000000005e-05, "loss": 0.6165, "step": 1216 }, { "epoch": 7.60625, "grad_norm": 2.34375, "learning_rate": 3.8e-05, "loss": 0.6575, "step": 1217 }, { "epoch": 7.6125, "grad_norm": 2.203125, "learning_rate": 3.803125e-05, "loss": 0.6665, "step": 1218 }, { "epoch": 7.61875, "grad_norm": 2.265625, "learning_rate": 3.8062500000000004e-05, "loss": 0.6709, "step": 1219 }, { "epoch": 7.625, "grad_norm": 2.203125, "learning_rate": 3.809375e-05, "loss": 0.6657, "step": 1220 }, { "epoch": 7.63125, "grad_norm": 2.203125, "learning_rate": 3.8125e-05, "loss": 0.6807, "step": 1221 }, { "epoch": 7.6375, "grad_norm": 2.734375, "learning_rate": 3.815625e-05, "loss": 0.6481, "step": 1222 }, { "epoch": 7.64375, "grad_norm": 1.9765625, "learning_rate": 3.818750000000001e-05, "loss": 0.5155, "step": 1223 }, { "epoch": 7.65, "grad_norm": 2.109375, "learning_rate": 3.8218750000000004e-05, "loss": 0.6234, "step": 1224 }, { "epoch": 7.65625, "grad_norm": 2.109375, "learning_rate": 3.825e-05, "loss": 0.6504, "step": 1225 }, { "epoch": 7.6625, "grad_norm": 1.984375, "learning_rate": 3.828125e-05, "loss": 0.6483, "step": 1226 }, { "epoch": 7.66875, "grad_norm": 2.171875, "learning_rate": 3.83125e-05, "loss": 0.6239, "step": 1227 }, { "epoch": 7.675, "grad_norm": 2.390625, "learning_rate": 3.834375e-05, "loss": 0.5889, "step": 1228 }, { "epoch": 7.68125, "grad_norm": 2.34375, "learning_rate": 3.8375e-05, "loss": 0.5898, "step": 1229 }, { "epoch": 7.6875, "grad_norm": 2.046875, "learning_rate": 3.840625e-05, "loss": 0.5571, "step": 1230 }, { "epoch": 7.69375, "grad_norm": 2.3125, "learning_rate": 3.8437500000000006e-05, "loss": 0.6016, "step": 1231 }, { "epoch": 7.7, "grad_norm": 2.28125, "learning_rate": 3.846875e-05, "loss": 0.6169, "step": 1232 }, { "epoch": 7.70625, "grad_norm": 2.3125, "learning_rate": 3.85e-05, "loss": 0.6008, "step": 1233 }, { "epoch": 7.7125, "grad_norm": 2.046875, "learning_rate": 3.8531250000000005e-05, "loss": 0.5177, "step": 1234 }, { "epoch": 7.71875, "grad_norm": 1.8359375, "learning_rate": 3.85625e-05, "loss": 0.5818, "step": 1235 }, { "epoch": 7.725, "grad_norm": 2.15625, "learning_rate": 3.859375e-05, "loss": 0.6227, "step": 1236 }, { "epoch": 7.73125, "grad_norm": 2.265625, "learning_rate": 3.8625e-05, "loss": 0.5789, "step": 1237 }, { "epoch": 7.7375, "grad_norm": 2.203125, "learning_rate": 3.865625e-05, "loss": 0.5656, "step": 1238 }, { "epoch": 7.74375, "grad_norm": 2.078125, "learning_rate": 3.8687500000000005e-05, "loss": 0.6424, "step": 1239 }, { "epoch": 7.75, "grad_norm": 2.5, "learning_rate": 3.871875e-05, "loss": 0.6463, "step": 1240 }, { "epoch": 7.75625, "grad_norm": 2.375, "learning_rate": 3.875e-05, "loss": 0.6343, "step": 1241 }, { "epoch": 7.7625, "grad_norm": 2.703125, "learning_rate": 3.8781250000000004e-05, "loss": 0.5986, "step": 1242 }, { "epoch": 7.76875, "grad_norm": 2.140625, "learning_rate": 3.88125e-05, "loss": 0.5723, "step": 1243 }, { "epoch": 7.775, "grad_norm": 2.078125, "learning_rate": 3.884375e-05, "loss": 0.5967, "step": 1244 }, { "epoch": 7.78125, "grad_norm": 2.5625, "learning_rate": 3.8875e-05, "loss": 0.6215, "step": 1245 }, { "epoch": 7.7875, "grad_norm": 2.03125, "learning_rate": 3.890625e-05, "loss": 0.5455, "step": 1246 }, { "epoch": 7.79375, "grad_norm": 2.515625, "learning_rate": 3.8937500000000005e-05, "loss": 0.6531, "step": 1247 }, { "epoch": 7.8, "grad_norm": 2.015625, "learning_rate": 3.896875e-05, "loss": 0.523, "step": 1248 }, { "epoch": 7.80625, "grad_norm": 2.40625, "learning_rate": 3.9000000000000006e-05, "loss": 0.5977, "step": 1249 }, { "epoch": 7.8125, "grad_norm": 2.15625, "learning_rate": 3.9031250000000003e-05, "loss": 0.5346, "step": 1250 }, { "epoch": 7.81875, "grad_norm": 2.0, "learning_rate": 3.90625e-05, "loss": 0.5739, "step": 1251 }, { "epoch": 7.825, "grad_norm": 2.625, "learning_rate": 3.909375e-05, "loss": 0.5394, "step": 1252 }, { "epoch": 7.83125, "grad_norm": 2.28125, "learning_rate": 3.9125e-05, "loss": 0.6348, "step": 1253 }, { "epoch": 7.8375, "grad_norm": 2.0, "learning_rate": 3.915625e-05, "loss": 0.5, "step": 1254 }, { "epoch": 7.84375, "grad_norm": 1.90625, "learning_rate": 3.91875e-05, "loss": 0.6191, "step": 1255 }, { "epoch": 7.85, "grad_norm": 2.046875, "learning_rate": 3.921875e-05, "loss": 0.5083, "step": 1256 }, { "epoch": 7.85625, "grad_norm": 2.328125, "learning_rate": 3.9250000000000005e-05, "loss": 0.6199, "step": 1257 }, { "epoch": 7.8625, "grad_norm": 2.078125, "learning_rate": 3.928125e-05, "loss": 0.5164, "step": 1258 }, { "epoch": 7.86875, "grad_norm": 2.3125, "learning_rate": 3.93125e-05, "loss": 0.5496, "step": 1259 }, { "epoch": 7.875, "grad_norm": 2.015625, "learning_rate": 3.9343750000000004e-05, "loss": 0.5392, "step": 1260 }, { "epoch": 7.88125, "grad_norm": 2.40625, "learning_rate": 3.9375e-05, "loss": 0.5622, "step": 1261 }, { "epoch": 7.8875, "grad_norm": 2.125, "learning_rate": 3.940625e-05, "loss": 0.5113, "step": 1262 }, { "epoch": 7.89375, "grad_norm": 2.359375, "learning_rate": 3.9437499999999996e-05, "loss": 0.561, "step": 1263 }, { "epoch": 7.9, "grad_norm": 2.25, "learning_rate": 3.946875000000001e-05, "loss": 0.5804, "step": 1264 }, { "epoch": 7.90625, "grad_norm": 1.9921875, "learning_rate": 3.9500000000000005e-05, "loss": 0.53, "step": 1265 }, { "epoch": 7.9125, "grad_norm": 2.078125, "learning_rate": 3.953125e-05, "loss": 0.5131, "step": 1266 }, { "epoch": 7.91875, "grad_norm": 2.078125, "learning_rate": 3.95625e-05, "loss": 0.5828, "step": 1267 }, { "epoch": 7.925, "grad_norm": 1.9375, "learning_rate": 3.9593750000000004e-05, "loss": 0.5433, "step": 1268 }, { "epoch": 7.93125, "grad_norm": 1.921875, "learning_rate": 3.9625e-05, "loss": 0.5076, "step": 1269 }, { "epoch": 7.9375, "grad_norm": 1.96875, "learning_rate": 3.965625e-05, "loss": 0.5519, "step": 1270 }, { "epoch": 7.94375, "grad_norm": 2.6875, "learning_rate": 3.96875e-05, "loss": 0.6137, "step": 1271 }, { "epoch": 7.95, "grad_norm": 2.5, "learning_rate": 3.9718750000000007e-05, "loss": 0.6014, "step": 1272 }, { "epoch": 7.95625, "grad_norm": 2.484375, "learning_rate": 3.9750000000000004e-05, "loss": 0.5747, "step": 1273 }, { "epoch": 7.9625, "grad_norm": 2.0, "learning_rate": 3.978125e-05, "loss": 0.4472, "step": 1274 }, { "epoch": 7.96875, "grad_norm": 2.0625, "learning_rate": 3.9812500000000005e-05, "loss": 0.5008, "step": 1275 }, { "epoch": 7.975, "grad_norm": 2.4375, "learning_rate": 3.984375e-05, "loss": 0.6263, "step": 1276 }, { "epoch": 7.98125, "grad_norm": 2.21875, "learning_rate": 3.9875e-05, "loss": 0.5872, "step": 1277 }, { "epoch": 7.9875, "grad_norm": 2.609375, "learning_rate": 3.990625e-05, "loss": 0.6466, "step": 1278 }, { "epoch": 7.99375, "grad_norm": 2.4375, "learning_rate": 3.99375e-05, "loss": 0.5444, "step": 1279 }, { "epoch": 8.0, "grad_norm": 2.328125, "learning_rate": 3.996875e-05, "loss": 0.5368, "step": 1280 }, { "epoch": 8.00625, "grad_norm": 2.4375, "learning_rate": 4e-05, "loss": 0.4946, "step": 1281 }, { "epoch": 8.0125, "grad_norm": 2.203125, "learning_rate": 4.003125e-05, "loss": 0.5762, "step": 1282 }, { "epoch": 8.01875, "grad_norm": 2.453125, "learning_rate": 4.0062500000000005e-05, "loss": 0.5243, "step": 1283 }, { "epoch": 8.025, "grad_norm": 2.59375, "learning_rate": 4.009375e-05, "loss": 0.6309, "step": 1284 }, { "epoch": 8.03125, "grad_norm": 2.5625, "learning_rate": 4.0125e-05, "loss": 0.5481, "step": 1285 }, { "epoch": 8.0375, "grad_norm": 2.296875, "learning_rate": 4.0156250000000004e-05, "loss": 0.5764, "step": 1286 }, { "epoch": 8.04375, "grad_norm": 2.546875, "learning_rate": 4.01875e-05, "loss": 0.6363, "step": 1287 }, { "epoch": 8.05, "grad_norm": 2.65625, "learning_rate": 4.021875e-05, "loss": 0.5737, "step": 1288 }, { "epoch": 8.05625, "grad_norm": 2.40625, "learning_rate": 4.025e-05, "loss": 0.5698, "step": 1289 }, { "epoch": 8.0625, "grad_norm": 2.390625, "learning_rate": 4.028125000000001e-05, "loss": 0.4942, "step": 1290 }, { "epoch": 8.06875, "grad_norm": 2.640625, "learning_rate": 4.0312500000000004e-05, "loss": 0.6181, "step": 1291 }, { "epoch": 8.075, "grad_norm": 2.28125, "learning_rate": 4.034375e-05, "loss": 0.5741, "step": 1292 }, { "epoch": 8.08125, "grad_norm": 2.03125, "learning_rate": 4.0375e-05, "loss": 0.5555, "step": 1293 }, { "epoch": 8.0875, "grad_norm": 2.03125, "learning_rate": 4.040625e-05, "loss": 0.5021, "step": 1294 }, { "epoch": 8.09375, "grad_norm": 2.359375, "learning_rate": 4.04375e-05, "loss": 0.494, "step": 1295 }, { "epoch": 8.1, "grad_norm": 2.234375, "learning_rate": 4.046875e-05, "loss": 0.5168, "step": 1296 }, { "epoch": 8.10625, "grad_norm": 2.71875, "learning_rate": 4.05e-05, "loss": 0.6183, "step": 1297 }, { "epoch": 8.1125, "grad_norm": 2.234375, "learning_rate": 4.0531250000000006e-05, "loss": 0.5713, "step": 1298 }, { "epoch": 8.11875, "grad_norm": 2.125, "learning_rate": 4.0562500000000003e-05, "loss": 0.5545, "step": 1299 }, { "epoch": 8.125, "grad_norm": 2.15625, "learning_rate": 4.059375e-05, "loss": 0.5139, "step": 1300 }, { "epoch": 8.13125, "grad_norm": 1.875, "learning_rate": 4.0625000000000005e-05, "loss": 0.4471, "step": 1301 }, { "epoch": 8.1375, "grad_norm": 2.15625, "learning_rate": 4.065625e-05, "loss": 0.5553, "step": 1302 }, { "epoch": 8.14375, "grad_norm": 2.328125, "learning_rate": 4.06875e-05, "loss": 0.5113, "step": 1303 }, { "epoch": 8.15, "grad_norm": 2.4375, "learning_rate": 4.071875e-05, "loss": 0.5796, "step": 1304 }, { "epoch": 8.15625, "grad_norm": 2.65625, "learning_rate": 4.075e-05, "loss": 0.6, "step": 1305 }, { "epoch": 8.1625, "grad_norm": 2.5625, "learning_rate": 4.0781250000000005e-05, "loss": 0.5839, "step": 1306 }, { "epoch": 8.16875, "grad_norm": 2.328125, "learning_rate": 4.08125e-05, "loss": 0.4561, "step": 1307 }, { "epoch": 8.175, "grad_norm": 2.390625, "learning_rate": 4.084375e-05, "loss": 0.5586, "step": 1308 }, { "epoch": 8.18125, "grad_norm": 2.25, "learning_rate": 4.0875000000000004e-05, "loss": 0.4504, "step": 1309 }, { "epoch": 8.1875, "grad_norm": 2.359375, "learning_rate": 4.090625e-05, "loss": 0.4282, "step": 1310 }, { "epoch": 8.19375, "grad_norm": 2.890625, "learning_rate": 4.09375e-05, "loss": 0.546, "step": 1311 }, { "epoch": 8.2, "grad_norm": 2.046875, "learning_rate": 4.096875e-05, "loss": 0.3971, "step": 1312 }, { "epoch": 8.20625, "grad_norm": 2.203125, "learning_rate": 4.1e-05, "loss": 0.4953, "step": 1313 }, { "epoch": 8.2125, "grad_norm": 2.625, "learning_rate": 4.1031250000000005e-05, "loss": 0.5305, "step": 1314 }, { "epoch": 8.21875, "grad_norm": 2.8125, "learning_rate": 4.10625e-05, "loss": 0.5717, "step": 1315 }, { "epoch": 8.225, "grad_norm": 2.84375, "learning_rate": 4.1093750000000006e-05, "loss": 0.5779, "step": 1316 }, { "epoch": 8.23125, "grad_norm": 3.203125, "learning_rate": 4.1125000000000004e-05, "loss": 0.5827, "step": 1317 }, { "epoch": 8.2375, "grad_norm": 2.3125, "learning_rate": 4.115625e-05, "loss": 0.5096, "step": 1318 }, { "epoch": 8.24375, "grad_norm": 2.890625, "learning_rate": 4.11875e-05, "loss": 0.6171, "step": 1319 }, { "epoch": 8.25, "grad_norm": 2.625, "learning_rate": 4.121875e-05, "loss": 0.5281, "step": 1320 }, { "epoch": 8.25625, "grad_norm": 2.5625, "learning_rate": 4.125e-05, "loss": 0.5541, "step": 1321 }, { "epoch": 8.2625, "grad_norm": 2.640625, "learning_rate": 4.1281250000000004e-05, "loss": 0.503, "step": 1322 }, { "epoch": 8.26875, "grad_norm": 2.5625, "learning_rate": 4.13125e-05, "loss": 0.5916, "step": 1323 }, { "epoch": 8.275, "grad_norm": 2.640625, "learning_rate": 4.1343750000000005e-05, "loss": 0.5294, "step": 1324 }, { "epoch": 8.28125, "grad_norm": 2.59375, "learning_rate": 4.1375e-05, "loss": 0.4591, "step": 1325 }, { "epoch": 8.2875, "grad_norm": 2.9375, "learning_rate": 4.140625e-05, "loss": 0.5542, "step": 1326 }, { "epoch": 8.29375, "grad_norm": 2.171875, "learning_rate": 4.1437500000000004e-05, "loss": 0.4448, "step": 1327 }, { "epoch": 8.3, "grad_norm": 2.5625, "learning_rate": 4.146875e-05, "loss": 0.4935, "step": 1328 }, { "epoch": 8.30625, "grad_norm": 2.46875, "learning_rate": 4.15e-05, "loss": 0.4836, "step": 1329 }, { "epoch": 8.3125, "grad_norm": 2.6875, "learning_rate": 4.1531249999999996e-05, "loss": 0.5814, "step": 1330 }, { "epoch": 8.31875, "grad_norm": 2.765625, "learning_rate": 4.156250000000001e-05, "loss": 0.4501, "step": 1331 }, { "epoch": 8.325, "grad_norm": 2.5625, "learning_rate": 4.1593750000000005e-05, "loss": 0.5456, "step": 1332 }, { "epoch": 8.33125, "grad_norm": 3.21875, "learning_rate": 4.1625e-05, "loss": 0.5966, "step": 1333 }, { "epoch": 8.3375, "grad_norm": 2.9375, "learning_rate": 4.165625e-05, "loss": 0.5566, "step": 1334 }, { "epoch": 8.34375, "grad_norm": 2.640625, "learning_rate": 4.1687500000000004e-05, "loss": 0.5405, "step": 1335 }, { "epoch": 8.35, "grad_norm": 2.578125, "learning_rate": 4.171875e-05, "loss": 0.4815, "step": 1336 }, { "epoch": 8.35625, "grad_norm": 3.59375, "learning_rate": 4.175e-05, "loss": 0.5046, "step": 1337 }, { "epoch": 8.3625, "grad_norm": 2.578125, "learning_rate": 4.178125e-05, "loss": 0.5486, "step": 1338 }, { "epoch": 8.36875, "grad_norm": 2.25, "learning_rate": 4.181250000000001e-05, "loss": 0.5573, "step": 1339 }, { "epoch": 8.375, "grad_norm": 2.984375, "learning_rate": 4.1843750000000004e-05, "loss": 0.6026, "step": 1340 }, { "epoch": 8.38125, "grad_norm": 2.703125, "learning_rate": 4.1875e-05, "loss": 0.5751, "step": 1341 }, { "epoch": 8.3875, "grad_norm": 2.578125, "learning_rate": 4.1906250000000006e-05, "loss": 0.5147, "step": 1342 }, { "epoch": 8.39375, "grad_norm": 2.71875, "learning_rate": 4.19375e-05, "loss": 0.4909, "step": 1343 }, { "epoch": 8.4, "grad_norm": 2.578125, "learning_rate": 4.196875e-05, "loss": 0.5327, "step": 1344 }, { "epoch": 8.40625, "grad_norm": 2.265625, "learning_rate": 4.2e-05, "loss": 0.4524, "step": 1345 }, { "epoch": 8.4125, "grad_norm": 2.296875, "learning_rate": 4.203125e-05, "loss": 0.5296, "step": 1346 }, { "epoch": 8.41875, "grad_norm": 2.578125, "learning_rate": 4.2062500000000006e-05, "loss": 0.447, "step": 1347 }, { "epoch": 8.425, "grad_norm": 2.8125, "learning_rate": 4.209375e-05, "loss": 0.4791, "step": 1348 }, { "epoch": 8.43125, "grad_norm": 2.375, "learning_rate": 4.2125e-05, "loss": 0.4386, "step": 1349 }, { "epoch": 8.4375, "grad_norm": 2.875, "learning_rate": 4.2156250000000005e-05, "loss": 0.484, "step": 1350 }, { "epoch": 8.44375, "grad_norm": 3.09375, "learning_rate": 4.21875e-05, "loss": 0.4726, "step": 1351 }, { "epoch": 8.45, "grad_norm": 2.375, "learning_rate": 4.221875e-05, "loss": 0.5093, "step": 1352 }, { "epoch": 8.45625, "grad_norm": 3.1875, "learning_rate": 4.2250000000000004e-05, "loss": 0.586, "step": 1353 }, { "epoch": 8.4625, "grad_norm": 3.921875, "learning_rate": 4.228125e-05, "loss": 0.4744, "step": 1354 }, { "epoch": 8.46875, "grad_norm": 2.390625, "learning_rate": 4.23125e-05, "loss": 0.4316, "step": 1355 }, { "epoch": 8.475, "grad_norm": 2.453125, "learning_rate": 4.234375e-05, "loss": 0.4964, "step": 1356 }, { "epoch": 8.48125, "grad_norm": 3.625, "learning_rate": 4.237500000000001e-05, "loss": 0.5258, "step": 1357 }, { "epoch": 8.4875, "grad_norm": 3.609375, "learning_rate": 4.2406250000000004e-05, "loss": 0.5359, "step": 1358 }, { "epoch": 8.49375, "grad_norm": 2.84375, "learning_rate": 4.24375e-05, "loss": 0.5366, "step": 1359 }, { "epoch": 8.5, "grad_norm": 2.265625, "learning_rate": 4.246875e-05, "loss": 0.5509, "step": 1360 }, { "epoch": 8.50625, "grad_norm": 2.296875, "learning_rate": 4.25e-05, "loss": 0.4217, "step": 1361 }, { "epoch": 8.5125, "grad_norm": 2.59375, "learning_rate": 4.253125e-05, "loss": 0.5292, "step": 1362 }, { "epoch": 8.51875, "grad_norm": 2.90625, "learning_rate": 4.25625e-05, "loss": 0.4396, "step": 1363 }, { "epoch": 8.525, "grad_norm": 2.1875, "learning_rate": 4.259375e-05, "loss": 0.4863, "step": 1364 }, { "epoch": 8.53125, "grad_norm": 2.484375, "learning_rate": 4.2625000000000006e-05, "loss": 0.524, "step": 1365 }, { "epoch": 8.5375, "grad_norm": 2.421875, "learning_rate": 4.2656250000000003e-05, "loss": 0.3698, "step": 1366 }, { "epoch": 8.54375, "grad_norm": 2.078125, "learning_rate": 4.26875e-05, "loss": 0.4161, "step": 1367 }, { "epoch": 8.55, "grad_norm": 2.25, "learning_rate": 4.2718750000000005e-05, "loss": 0.5093, "step": 1368 }, { "epoch": 8.55625, "grad_norm": 2.75, "learning_rate": 4.275e-05, "loss": 0.4496, "step": 1369 }, { "epoch": 8.5625, "grad_norm": 2.328125, "learning_rate": 4.278125e-05, "loss": 0.4893, "step": 1370 }, { "epoch": 8.56875, "grad_norm": 2.09375, "learning_rate": 4.28125e-05, "loss": 0.409, "step": 1371 }, { "epoch": 8.575, "grad_norm": 2.5, "learning_rate": 4.284375000000001e-05, "loss": 0.4881, "step": 1372 }, { "epoch": 8.58125, "grad_norm": 2.9375, "learning_rate": 4.2875000000000005e-05, "loss": 0.5219, "step": 1373 }, { "epoch": 8.5875, "grad_norm": 2.421875, "learning_rate": 4.290625e-05, "loss": 0.4812, "step": 1374 }, { "epoch": 8.59375, "grad_norm": 2.125, "learning_rate": 4.29375e-05, "loss": 0.4545, "step": 1375 }, { "epoch": 8.6, "grad_norm": 3.0, "learning_rate": 4.2968750000000004e-05, "loss": 0.4783, "step": 1376 }, { "epoch": 8.60625, "grad_norm": 2.265625, "learning_rate": 4.3e-05, "loss": 0.4183, "step": 1377 }, { "epoch": 8.6125, "grad_norm": 1.90625, "learning_rate": 4.303125e-05, "loss": 0.3629, "step": 1378 }, { "epoch": 8.61875, "grad_norm": 2.59375, "learning_rate": 4.30625e-05, "loss": 0.4658, "step": 1379 }, { "epoch": 8.625, "grad_norm": 2.359375, "learning_rate": 4.309375e-05, "loss": 0.5708, "step": 1380 }, { "epoch": 8.63125, "grad_norm": 2.578125, "learning_rate": 4.3125000000000005e-05, "loss": 0.4508, "step": 1381 }, { "epoch": 8.6375, "grad_norm": 2.28125, "learning_rate": 4.315625e-05, "loss": 0.4049, "step": 1382 }, { "epoch": 8.64375, "grad_norm": 1.9921875, "learning_rate": 4.3187500000000006e-05, "loss": 0.3826, "step": 1383 }, { "epoch": 8.65, "grad_norm": 2.296875, "learning_rate": 4.3218750000000004e-05, "loss": 0.5061, "step": 1384 }, { "epoch": 8.65625, "grad_norm": 2.40625, "learning_rate": 4.325e-05, "loss": 0.4641, "step": 1385 }, { "epoch": 8.6625, "grad_norm": 2.453125, "learning_rate": 4.328125e-05, "loss": 0.4779, "step": 1386 }, { "epoch": 8.66875, "grad_norm": 2.1875, "learning_rate": 4.33125e-05, "loss": 0.3181, "step": 1387 }, { "epoch": 8.675, "grad_norm": 2.46875, "learning_rate": 4.334375e-05, "loss": 0.5415, "step": 1388 }, { "epoch": 8.68125, "grad_norm": 2.234375, "learning_rate": 4.3375000000000004e-05, "loss": 0.4818, "step": 1389 }, { "epoch": 8.6875, "grad_norm": 2.453125, "learning_rate": 4.340625e-05, "loss": 0.4779, "step": 1390 }, { "epoch": 8.69375, "grad_norm": 2.265625, "learning_rate": 4.3437500000000006e-05, "loss": 0.4027, "step": 1391 }, { "epoch": 8.7, "grad_norm": 2.375, "learning_rate": 4.346875e-05, "loss": 0.5562, "step": 1392 }, { "epoch": 8.70625, "grad_norm": 2.0625, "learning_rate": 4.35e-05, "loss": 0.4186, "step": 1393 }, { "epoch": 8.7125, "grad_norm": 2.8125, "learning_rate": 4.3531250000000004e-05, "loss": 0.4443, "step": 1394 }, { "epoch": 8.71875, "grad_norm": 2.625, "learning_rate": 4.35625e-05, "loss": 0.4706, "step": 1395 }, { "epoch": 8.725, "grad_norm": 2.328125, "learning_rate": 4.359375e-05, "loss": 0.4523, "step": 1396 }, { "epoch": 8.73125, "grad_norm": 2.6875, "learning_rate": 4.3625e-05, "loss": 0.4394, "step": 1397 }, { "epoch": 8.7375, "grad_norm": 2.171875, "learning_rate": 4.365625000000001e-05, "loss": 0.4184, "step": 1398 }, { "epoch": 8.74375, "grad_norm": 2.25, "learning_rate": 4.3687500000000005e-05, "loss": 0.4228, "step": 1399 }, { "epoch": 8.75, "grad_norm": 2.09375, "learning_rate": 4.371875e-05, "loss": 0.4158, "step": 1400 }, { "epoch": 8.75625, "grad_norm": 2.203125, "learning_rate": 4.375e-05, "loss": 0.4121, "step": 1401 }, { "epoch": 8.7625, "grad_norm": 2.140625, "learning_rate": 4.3781250000000004e-05, "loss": 0.3911, "step": 1402 }, { "epoch": 8.76875, "grad_norm": 2.6875, "learning_rate": 4.38125e-05, "loss": 0.584, "step": 1403 }, { "epoch": 8.775, "grad_norm": 2.484375, "learning_rate": 4.384375e-05, "loss": 0.4894, "step": 1404 }, { "epoch": 8.78125, "grad_norm": 2.4375, "learning_rate": 4.3875e-05, "loss": 0.4699, "step": 1405 }, { "epoch": 8.7875, "grad_norm": 2.453125, "learning_rate": 4.390625000000001e-05, "loss": 0.4503, "step": 1406 }, { "epoch": 8.79375, "grad_norm": 2.390625, "learning_rate": 4.3937500000000004e-05, "loss": 0.4533, "step": 1407 }, { "epoch": 8.8, "grad_norm": 2.359375, "learning_rate": 4.396875e-05, "loss": 0.4379, "step": 1408 }, { "epoch": 8.80625, "grad_norm": 2.421875, "learning_rate": 4.4000000000000006e-05, "loss": 0.399, "step": 1409 }, { "epoch": 8.8125, "grad_norm": 2.734375, "learning_rate": 4.403125e-05, "loss": 0.4939, "step": 1410 }, { "epoch": 8.81875, "grad_norm": 2.453125, "learning_rate": 4.40625e-05, "loss": 0.4669, "step": 1411 }, { "epoch": 8.825, "grad_norm": 2.40625, "learning_rate": 4.409375e-05, "loss": 0.4053, "step": 1412 }, { "epoch": 8.83125, "grad_norm": 2.40625, "learning_rate": 4.4125e-05, "loss": 0.5083, "step": 1413 }, { "epoch": 8.8375, "grad_norm": 2.265625, "learning_rate": 4.4156250000000006e-05, "loss": 0.3737, "step": 1414 }, { "epoch": 8.84375, "grad_norm": 2.453125, "learning_rate": 4.4187500000000003e-05, "loss": 0.4332, "step": 1415 }, { "epoch": 8.85, "grad_norm": 2.25, "learning_rate": 4.421875e-05, "loss": 0.3996, "step": 1416 }, { "epoch": 8.85625, "grad_norm": 2.515625, "learning_rate": 4.4250000000000005e-05, "loss": 0.4419, "step": 1417 }, { "epoch": 8.8625, "grad_norm": 2.703125, "learning_rate": 4.428125e-05, "loss": 0.4639, "step": 1418 }, { "epoch": 8.86875, "grad_norm": 2.453125, "learning_rate": 4.43125e-05, "loss": 0.4532, "step": 1419 }, { "epoch": 8.875, "grad_norm": 2.3125, "learning_rate": 4.4343750000000004e-05, "loss": 0.4232, "step": 1420 }, { "epoch": 8.88125, "grad_norm": 2.296875, "learning_rate": 4.4375e-05, "loss": 0.5864, "step": 1421 }, { "epoch": 8.8875, "grad_norm": 2.484375, "learning_rate": 4.4406250000000005e-05, "loss": 0.3833, "step": 1422 }, { "epoch": 8.89375, "grad_norm": 2.625, "learning_rate": 4.44375e-05, "loss": 0.4154, "step": 1423 }, { "epoch": 8.9, "grad_norm": 2.078125, "learning_rate": 4.446875e-05, "loss": 0.4023, "step": 1424 }, { "epoch": 8.90625, "grad_norm": 2.53125, "learning_rate": 4.4500000000000004e-05, "loss": 0.4026, "step": 1425 }, { "epoch": 8.9125, "grad_norm": 2.015625, "learning_rate": 4.453125e-05, "loss": 0.3266, "step": 1426 }, { "epoch": 8.91875, "grad_norm": 3.0, "learning_rate": 4.45625e-05, "loss": 0.4957, "step": 1427 }, { "epoch": 8.925, "grad_norm": 2.140625, "learning_rate": 4.459375e-05, "loss": 0.4667, "step": 1428 }, { "epoch": 8.93125, "grad_norm": 2.6875, "learning_rate": 4.4625e-05, "loss": 0.4704, "step": 1429 }, { "epoch": 8.9375, "grad_norm": 2.34375, "learning_rate": 4.465625e-05, "loss": 0.4579, "step": 1430 }, { "epoch": 8.94375, "grad_norm": 2.6875, "learning_rate": 4.46875e-05, "loss": 0.4347, "step": 1431 }, { "epoch": 8.95, "grad_norm": 2.3125, "learning_rate": 4.4718750000000006e-05, "loss": 0.4848, "step": 1432 }, { "epoch": 8.95625, "grad_norm": 2.234375, "learning_rate": 4.4750000000000004e-05, "loss": 0.3708, "step": 1433 }, { "epoch": 8.9625, "grad_norm": 2.421875, "learning_rate": 4.478125e-05, "loss": 0.3833, "step": 1434 }, { "epoch": 8.96875, "grad_norm": 2.359375, "learning_rate": 4.4812500000000005e-05, "loss": 0.3754, "step": 1435 }, { "epoch": 8.975, "grad_norm": 2.6875, "learning_rate": 4.484375e-05, "loss": 0.3409, "step": 1436 }, { "epoch": 8.98125, "grad_norm": 2.390625, "learning_rate": 4.4875e-05, "loss": 0.3946, "step": 1437 }, { "epoch": 8.9875, "grad_norm": 2.5625, "learning_rate": 4.490625e-05, "loss": 0.4274, "step": 1438 }, { "epoch": 8.99375, "grad_norm": 2.1875, "learning_rate": 4.49375e-05, "loss": 0.2988, "step": 1439 }, { "epoch": 9.0, "grad_norm": 2.453125, "learning_rate": 4.4968750000000005e-05, "loss": 0.4604, "step": 1440 }, { "epoch": 9.00625, "grad_norm": 2.421875, "learning_rate": 4.5e-05, "loss": 0.4719, "step": 1441 }, { "epoch": 9.0125, "grad_norm": 2.40625, "learning_rate": 4.503125e-05, "loss": 0.3659, "step": 1442 }, { "epoch": 9.01875, "grad_norm": 2.1875, "learning_rate": 4.5062500000000004e-05, "loss": 0.3402, "step": 1443 }, { "epoch": 9.025, "grad_norm": 2.375, "learning_rate": 4.509375e-05, "loss": 0.4241, "step": 1444 }, { "epoch": 9.03125, "grad_norm": 2.65625, "learning_rate": 4.5125e-05, "loss": 0.3853, "step": 1445 }, { "epoch": 9.0375, "grad_norm": 2.578125, "learning_rate": 4.515625e-05, "loss": 0.4268, "step": 1446 }, { "epoch": 9.04375, "grad_norm": 2.03125, "learning_rate": 4.518750000000001e-05, "loss": 0.3868, "step": 1447 }, { "epoch": 9.05, "grad_norm": 2.0, "learning_rate": 4.5218750000000005e-05, "loss": 0.29, "step": 1448 }, { "epoch": 9.05625, "grad_norm": 2.28125, "learning_rate": 4.525e-05, "loss": 0.4788, "step": 1449 }, { "epoch": 9.0625, "grad_norm": 2.328125, "learning_rate": 4.528125e-05, "loss": 0.333, "step": 1450 }, { "epoch": 9.06875, "grad_norm": 2.25, "learning_rate": 4.5312500000000004e-05, "loss": 0.3879, "step": 1451 }, { "epoch": 9.075, "grad_norm": 2.28125, "learning_rate": 4.534375e-05, "loss": 0.3563, "step": 1452 }, { "epoch": 9.08125, "grad_norm": 2.78125, "learning_rate": 4.5375e-05, "loss": 0.4246, "step": 1453 }, { "epoch": 9.0875, "grad_norm": 2.296875, "learning_rate": 4.540625e-05, "loss": 0.3578, "step": 1454 }, { "epoch": 9.09375, "grad_norm": 2.140625, "learning_rate": 4.54375e-05, "loss": 0.3866, "step": 1455 }, { "epoch": 9.1, "grad_norm": 3.046875, "learning_rate": 4.5468750000000004e-05, "loss": 0.4983, "step": 1456 }, { "epoch": 9.10625, "grad_norm": 2.84375, "learning_rate": 4.55e-05, "loss": 0.4424, "step": 1457 }, { "epoch": 9.1125, "grad_norm": 1.765625, "learning_rate": 4.5531250000000006e-05, "loss": 0.34, "step": 1458 }, { "epoch": 9.11875, "grad_norm": 2.078125, "learning_rate": 4.55625e-05, "loss": 0.4805, "step": 1459 }, { "epoch": 9.125, "grad_norm": 1.9375, "learning_rate": 4.559375e-05, "loss": 0.3195, "step": 1460 }, { "epoch": 9.13125, "grad_norm": 2.09375, "learning_rate": 4.5625e-05, "loss": 0.2955, "step": 1461 }, { "epoch": 9.1375, "grad_norm": 2.21875, "learning_rate": 4.565625e-05, "loss": 0.3699, "step": 1462 }, { "epoch": 9.14375, "grad_norm": 2.296875, "learning_rate": 4.56875e-05, "loss": 0.3282, "step": 1463 }, { "epoch": 9.15, "grad_norm": 1.9375, "learning_rate": 4.571875e-05, "loss": 0.2728, "step": 1464 }, { "epoch": 9.15625, "grad_norm": 2.5625, "learning_rate": 4.575e-05, "loss": 0.4365, "step": 1465 }, { "epoch": 9.1625, "grad_norm": 2.15625, "learning_rate": 4.5781250000000005e-05, "loss": 0.4595, "step": 1466 }, { "epoch": 9.16875, "grad_norm": 2.4375, "learning_rate": 4.58125e-05, "loss": 0.3395, "step": 1467 }, { "epoch": 9.175, "grad_norm": 2.796875, "learning_rate": 4.584375e-05, "loss": 0.3169, "step": 1468 }, { "epoch": 9.18125, "grad_norm": 1.84375, "learning_rate": 4.5875000000000004e-05, "loss": 0.2932, "step": 1469 }, { "epoch": 9.1875, "grad_norm": 2.4375, "learning_rate": 4.590625e-05, "loss": 0.3901, "step": 1470 }, { "epoch": 9.19375, "grad_norm": 2.5, "learning_rate": 4.59375e-05, "loss": 0.3703, "step": 1471 }, { "epoch": 9.2, "grad_norm": 2.671875, "learning_rate": 4.596875e-05, "loss": 0.374, "step": 1472 }, { "epoch": 9.20625, "grad_norm": 1.921875, "learning_rate": 4.600000000000001e-05, "loss": 0.3183, "step": 1473 }, { "epoch": 9.2125, "grad_norm": 2.125, "learning_rate": 4.6031250000000004e-05, "loss": 0.3315, "step": 1474 }, { "epoch": 9.21875, "grad_norm": 2.40625, "learning_rate": 4.60625e-05, "loss": 0.377, "step": 1475 }, { "epoch": 9.225, "grad_norm": 2.34375, "learning_rate": 4.609375e-05, "loss": 0.387, "step": 1476 }, { "epoch": 9.23125, "grad_norm": 2.109375, "learning_rate": 4.6125e-05, "loss": 0.3536, "step": 1477 }, { "epoch": 9.2375, "grad_norm": 2.25, "learning_rate": 4.615625e-05, "loss": 0.3399, "step": 1478 }, { "epoch": 9.24375, "grad_norm": 2.09375, "learning_rate": 4.61875e-05, "loss": 0.2887, "step": 1479 }, { "epoch": 9.25, "grad_norm": 2.21875, "learning_rate": 4.621875e-05, "loss": 0.2958, "step": 1480 }, { "epoch": 9.25625, "grad_norm": 2.546875, "learning_rate": 4.6250000000000006e-05, "loss": 0.31, "step": 1481 }, { "epoch": 9.2625, "grad_norm": 1.984375, "learning_rate": 4.6281250000000003e-05, "loss": 0.2997, "step": 1482 }, { "epoch": 9.26875, "grad_norm": 2.359375, "learning_rate": 4.63125e-05, "loss": 0.3785, "step": 1483 }, { "epoch": 9.275, "grad_norm": 2.484375, "learning_rate": 4.6343750000000005e-05, "loss": 0.3771, "step": 1484 }, { "epoch": 9.28125, "grad_norm": 2.140625, "learning_rate": 4.6375e-05, "loss": 0.3931, "step": 1485 }, { "epoch": 9.2875, "grad_norm": 2.78125, "learning_rate": 4.640625e-05, "loss": 0.3554, "step": 1486 }, { "epoch": 9.29375, "grad_norm": 2.203125, "learning_rate": 4.64375e-05, "loss": 0.3193, "step": 1487 }, { "epoch": 9.3, "grad_norm": 1.8359375, "learning_rate": 4.646875e-05, "loss": 0.3015, "step": 1488 }, { "epoch": 9.30625, "grad_norm": 2.109375, "learning_rate": 4.6500000000000005e-05, "loss": 0.3838, "step": 1489 }, { "epoch": 9.3125, "grad_norm": 2.453125, "learning_rate": 4.653125e-05, "loss": 0.4058, "step": 1490 }, { "epoch": 9.31875, "grad_norm": 2.515625, "learning_rate": 4.65625e-05, "loss": 0.4143, "step": 1491 }, { "epoch": 9.325, "grad_norm": 1.8828125, "learning_rate": 4.6593750000000004e-05, "loss": 0.2571, "step": 1492 }, { "epoch": 9.33125, "grad_norm": 1.9609375, "learning_rate": 4.6625e-05, "loss": 0.3497, "step": 1493 }, { "epoch": 9.3375, "grad_norm": 2.3125, "learning_rate": 4.665625e-05, "loss": 0.3578, "step": 1494 }, { "epoch": 9.34375, "grad_norm": 2.65625, "learning_rate": 4.66875e-05, "loss": 0.4893, "step": 1495 }, { "epoch": 9.35, "grad_norm": 2.59375, "learning_rate": 4.671875e-05, "loss": 0.4598, "step": 1496 }, { "epoch": 9.35625, "grad_norm": 2.5, "learning_rate": 4.6750000000000005e-05, "loss": 0.3069, "step": 1497 }, { "epoch": 9.3625, "grad_norm": 2.46875, "learning_rate": 4.678125e-05, "loss": 0.3972, "step": 1498 }, { "epoch": 9.36875, "grad_norm": 2.359375, "learning_rate": 4.6812500000000006e-05, "loss": 0.3491, "step": 1499 }, { "epoch": 9.375, "grad_norm": 2.3125, "learning_rate": 4.6843750000000004e-05, "loss": 0.3046, "step": 1500 }, { "epoch": 9.38125, "grad_norm": 2.609375, "learning_rate": 4.6875e-05, "loss": 0.3374, "step": 1501 }, { "epoch": 9.3875, "grad_norm": 2.203125, "learning_rate": 4.690625e-05, "loss": 0.3403, "step": 1502 }, { "epoch": 9.39375, "grad_norm": 1.8515625, "learning_rate": 4.69375e-05, "loss": 0.238, "step": 1503 }, { "epoch": 9.4, "grad_norm": 2.203125, "learning_rate": 4.696875e-05, "loss": 0.2915, "step": 1504 }, { "epoch": 9.40625, "grad_norm": 2.078125, "learning_rate": 4.7e-05, "loss": 0.2899, "step": 1505 }, { "epoch": 9.4125, "grad_norm": 2.203125, "learning_rate": 4.703125e-05, "loss": 0.335, "step": 1506 }, { "epoch": 9.41875, "grad_norm": 2.375, "learning_rate": 4.7062500000000006e-05, "loss": 0.3431, "step": 1507 }, { "epoch": 9.425, "grad_norm": 2.15625, "learning_rate": 4.709375e-05, "loss": 0.2838, "step": 1508 }, { "epoch": 9.43125, "grad_norm": 2.875, "learning_rate": 4.7125e-05, "loss": 0.3554, "step": 1509 }, { "epoch": 9.4375, "grad_norm": 1.71875, "learning_rate": 4.7156250000000004e-05, "loss": 0.2356, "step": 1510 }, { "epoch": 9.44375, "grad_norm": 2.5625, "learning_rate": 4.71875e-05, "loss": 0.4455, "step": 1511 }, { "epoch": 9.45, "grad_norm": 2.671875, "learning_rate": 4.721875e-05, "loss": 0.3241, "step": 1512 }, { "epoch": 9.45625, "grad_norm": 2.640625, "learning_rate": 4.7249999999999997e-05, "loss": 0.3184, "step": 1513 }, { "epoch": 9.4625, "grad_norm": 2.4375, "learning_rate": 4.728125000000001e-05, "loss": 0.3457, "step": 1514 }, { "epoch": 9.46875, "grad_norm": 2.421875, "learning_rate": 4.7312500000000005e-05, "loss": 0.3055, "step": 1515 }, { "epoch": 9.475, "grad_norm": 2.359375, "learning_rate": 4.734375e-05, "loss": 0.3366, "step": 1516 }, { "epoch": 9.48125, "grad_norm": 2.640625, "learning_rate": 4.7375e-05, "loss": 0.3747, "step": 1517 }, { "epoch": 9.4875, "grad_norm": 2.28125, "learning_rate": 4.7406250000000004e-05, "loss": 0.3584, "step": 1518 }, { "epoch": 9.49375, "grad_norm": 2.359375, "learning_rate": 4.74375e-05, "loss": 0.3436, "step": 1519 }, { "epoch": 9.5, "grad_norm": 2.140625, "learning_rate": 4.746875e-05, "loss": 0.3372, "step": 1520 }, { "epoch": 9.50625, "grad_norm": 2.234375, "learning_rate": 4.75e-05, "loss": 0.3083, "step": 1521 }, { "epoch": 9.5125, "grad_norm": 2.578125, "learning_rate": 4.753125000000001e-05, "loss": 0.4023, "step": 1522 }, { "epoch": 9.51875, "grad_norm": 2.71875, "learning_rate": 4.7562500000000004e-05, "loss": 0.3412, "step": 1523 }, { "epoch": 9.525, "grad_norm": 2.421875, "learning_rate": 4.759375e-05, "loss": 0.3973, "step": 1524 }, { "epoch": 9.53125, "grad_norm": 2.234375, "learning_rate": 4.7625000000000006e-05, "loss": 0.3353, "step": 1525 }, { "epoch": 9.5375, "grad_norm": 2.40625, "learning_rate": 4.765625e-05, "loss": 0.398, "step": 1526 }, { "epoch": 9.54375, "grad_norm": 2.625, "learning_rate": 4.76875e-05, "loss": 0.4245, "step": 1527 }, { "epoch": 9.55, "grad_norm": 2.515625, "learning_rate": 4.771875e-05, "loss": 0.2904, "step": 1528 }, { "epoch": 9.55625, "grad_norm": 2.46875, "learning_rate": 4.775e-05, "loss": 0.3767, "step": 1529 }, { "epoch": 9.5625, "grad_norm": 1.8203125, "learning_rate": 4.778125e-05, "loss": 0.3113, "step": 1530 }, { "epoch": 9.56875, "grad_norm": 2.375, "learning_rate": 4.7812500000000003e-05, "loss": 0.3005, "step": 1531 }, { "epoch": 9.575, "grad_norm": 1.9453125, "learning_rate": 4.784375e-05, "loss": 0.2505, "step": 1532 }, { "epoch": 9.58125, "grad_norm": 2.171875, "learning_rate": 4.7875000000000005e-05, "loss": 0.2517, "step": 1533 }, { "epoch": 9.5875, "grad_norm": 2.5625, "learning_rate": 4.790625e-05, "loss": 0.3609, "step": 1534 }, { "epoch": 9.59375, "grad_norm": 2.421875, "learning_rate": 4.79375e-05, "loss": 0.2908, "step": 1535 }, { "epoch": 9.6, "grad_norm": 2.34375, "learning_rate": 4.7968750000000004e-05, "loss": 0.3013, "step": 1536 }, { "epoch": 9.60625, "grad_norm": 2.15625, "learning_rate": 4.8e-05, "loss": 0.3348, "step": 1537 }, { "epoch": 9.6125, "grad_norm": 2.625, "learning_rate": 4.803125e-05, "loss": 0.374, "step": 1538 }, { "epoch": 9.61875, "grad_norm": 2.25, "learning_rate": 4.80625e-05, "loss": 0.317, "step": 1539 }, { "epoch": 9.625, "grad_norm": 2.71875, "learning_rate": 4.809375000000001e-05, "loss": 0.3942, "step": 1540 }, { "epoch": 9.63125, "grad_norm": 1.8828125, "learning_rate": 4.8125000000000004e-05, "loss": 0.2741, "step": 1541 }, { "epoch": 9.6375, "grad_norm": 2.171875, "learning_rate": 4.815625e-05, "loss": 0.3771, "step": 1542 }, { "epoch": 9.64375, "grad_norm": 2.640625, "learning_rate": 4.81875e-05, "loss": 0.3912, "step": 1543 }, { "epoch": 9.65, "grad_norm": 2.0625, "learning_rate": 4.821875e-05, "loss": 0.3014, "step": 1544 }, { "epoch": 9.65625, "grad_norm": 2.328125, "learning_rate": 4.825e-05, "loss": 0.4917, "step": 1545 }, { "epoch": 9.6625, "grad_norm": 2.09375, "learning_rate": 4.828125e-05, "loss": 0.2998, "step": 1546 }, { "epoch": 9.66875, "grad_norm": 2.109375, "learning_rate": 4.83125e-05, "loss": 0.3505, "step": 1547 }, { "epoch": 9.675, "grad_norm": 2.09375, "learning_rate": 4.8343750000000006e-05, "loss": 0.2923, "step": 1548 }, { "epoch": 9.68125, "grad_norm": 2.421875, "learning_rate": 4.8375000000000004e-05, "loss": 0.3144, "step": 1549 }, { "epoch": 9.6875, "grad_norm": 1.9375, "learning_rate": 4.840625e-05, "loss": 0.3306, "step": 1550 }, { "epoch": 9.69375, "grad_norm": 2.21875, "learning_rate": 4.8437500000000005e-05, "loss": 0.3251, "step": 1551 }, { "epoch": 9.7, "grad_norm": 2.625, "learning_rate": 4.846875e-05, "loss": 0.3309, "step": 1552 }, { "epoch": 9.70625, "grad_norm": 2.140625, "learning_rate": 4.85e-05, "loss": 0.282, "step": 1553 }, { "epoch": 9.7125, "grad_norm": 2.328125, "learning_rate": 4.853125e-05, "loss": 0.3133, "step": 1554 }, { "epoch": 9.71875, "grad_norm": 2.078125, "learning_rate": 4.85625e-05, "loss": 0.2335, "step": 1555 }, { "epoch": 9.725, "grad_norm": 2.609375, "learning_rate": 4.8593750000000005e-05, "loss": 0.3078, "step": 1556 }, { "epoch": 9.73125, "grad_norm": 2.515625, "learning_rate": 4.8625e-05, "loss": 0.371, "step": 1557 }, { "epoch": 9.7375, "grad_norm": 2.0, "learning_rate": 4.865625e-05, "loss": 0.2585, "step": 1558 }, { "epoch": 9.74375, "grad_norm": 2.046875, "learning_rate": 4.8687500000000004e-05, "loss": 0.2765, "step": 1559 }, { "epoch": 9.75, "grad_norm": 1.9453125, "learning_rate": 4.871875e-05, "loss": 0.2215, "step": 1560 }, { "epoch": 9.75625, "grad_norm": 2.25, "learning_rate": 4.875e-05, "loss": 0.2759, "step": 1561 }, { "epoch": 9.7625, "grad_norm": 2.1875, "learning_rate": 4.878125e-05, "loss": 0.2437, "step": 1562 }, { "epoch": 9.76875, "grad_norm": 2.3125, "learning_rate": 4.88125e-05, "loss": 0.3385, "step": 1563 }, { "epoch": 9.775, "grad_norm": 2.5, "learning_rate": 4.8843750000000005e-05, "loss": 0.369, "step": 1564 }, { "epoch": 9.78125, "grad_norm": 2.328125, "learning_rate": 4.8875e-05, "loss": 0.264, "step": 1565 }, { "epoch": 9.7875, "grad_norm": 2.5, "learning_rate": 4.8906250000000006e-05, "loss": 0.3437, "step": 1566 }, { "epoch": 9.79375, "grad_norm": 2.171875, "learning_rate": 4.8937500000000004e-05, "loss": 0.3311, "step": 1567 }, { "epoch": 9.8, "grad_norm": 1.7890625, "learning_rate": 4.896875e-05, "loss": 0.2327, "step": 1568 }, { "epoch": 9.80625, "grad_norm": 2.078125, "learning_rate": 4.9e-05, "loss": 0.2969, "step": 1569 }, { "epoch": 9.8125, "grad_norm": 2.515625, "learning_rate": 4.903125e-05, "loss": 0.3371, "step": 1570 }, { "epoch": 9.81875, "grad_norm": 2.140625, "learning_rate": 4.90625e-05, "loss": 0.2691, "step": 1571 }, { "epoch": 9.825, "grad_norm": 2.015625, "learning_rate": 4.9093750000000004e-05, "loss": 0.3201, "step": 1572 }, { "epoch": 9.83125, "grad_norm": 2.65625, "learning_rate": 4.9125e-05, "loss": 0.3303, "step": 1573 }, { "epoch": 9.8375, "grad_norm": 2.84375, "learning_rate": 4.9156250000000006e-05, "loss": 0.3043, "step": 1574 }, { "epoch": 9.84375, "grad_norm": 2.609375, "learning_rate": 4.91875e-05, "loss": 0.3103, "step": 1575 }, { "epoch": 9.85, "grad_norm": 3.21875, "learning_rate": 4.921875e-05, "loss": 0.3649, "step": 1576 }, { "epoch": 9.85625, "grad_norm": 2.15625, "learning_rate": 4.9250000000000004e-05, "loss": 0.2822, "step": 1577 }, { "epoch": 9.8625, "grad_norm": 2.296875, "learning_rate": 4.928125e-05, "loss": 0.2991, "step": 1578 }, { "epoch": 9.86875, "grad_norm": 2.625, "learning_rate": 4.93125e-05, "loss": 0.41, "step": 1579 }, { "epoch": 9.875, "grad_norm": 1.7890625, "learning_rate": 4.9343749999999997e-05, "loss": 0.282, "step": 1580 }, { "epoch": 9.88125, "grad_norm": 2.453125, "learning_rate": 4.937500000000001e-05, "loss": 0.2843, "step": 1581 }, { "epoch": 9.8875, "grad_norm": 1.9921875, "learning_rate": 4.9406250000000005e-05, "loss": 0.2686, "step": 1582 }, { "epoch": 9.89375, "grad_norm": 2.25, "learning_rate": 4.94375e-05, "loss": 0.271, "step": 1583 }, { "epoch": 9.9, "grad_norm": 2.28125, "learning_rate": 4.946875e-05, "loss": 0.3213, "step": 1584 }, { "epoch": 9.90625, "grad_norm": 2.3125, "learning_rate": 4.9500000000000004e-05, "loss": 0.3174, "step": 1585 }, { "epoch": 9.9125, "grad_norm": 2.546875, "learning_rate": 4.953125e-05, "loss": 0.2525, "step": 1586 }, { "epoch": 9.91875, "grad_norm": 2.328125, "learning_rate": 4.95625e-05, "loss": 0.4271, "step": 1587 }, { "epoch": 9.925, "grad_norm": 2.8125, "learning_rate": 4.959375e-05, "loss": 0.3903, "step": 1588 }, { "epoch": 9.93125, "grad_norm": 2.90625, "learning_rate": 4.962500000000001e-05, "loss": 0.3249, "step": 1589 }, { "epoch": 9.9375, "grad_norm": 2.40625, "learning_rate": 4.9656250000000004e-05, "loss": 0.3009, "step": 1590 }, { "epoch": 9.94375, "grad_norm": 2.78125, "learning_rate": 4.96875e-05, "loss": 0.3287, "step": 1591 }, { "epoch": 9.95, "grad_norm": 2.453125, "learning_rate": 4.9718750000000006e-05, "loss": 0.3587, "step": 1592 }, { "epoch": 9.95625, "grad_norm": 2.0, "learning_rate": 4.975e-05, "loss": 0.2054, "step": 1593 }, { "epoch": 9.9625, "grad_norm": 2.28125, "learning_rate": 4.978125e-05, "loss": 0.3201, "step": 1594 }, { "epoch": 9.96875, "grad_norm": 1.9375, "learning_rate": 4.98125e-05, "loss": 0.2318, "step": 1595 }, { "epoch": 9.975, "grad_norm": 1.921875, "learning_rate": 4.984375e-05, "loss": 0.2741, "step": 1596 }, { "epoch": 9.98125, "grad_norm": 1.96875, "learning_rate": 4.9875000000000006e-05, "loss": 0.2196, "step": 1597 }, { "epoch": 9.9875, "grad_norm": 1.9609375, "learning_rate": 4.9906250000000004e-05, "loss": 0.259, "step": 1598 }, { "epoch": 9.99375, "grad_norm": 1.9375, "learning_rate": 4.99375e-05, "loss": 0.3047, "step": 1599 }, { "epoch": 10.0, "grad_norm": 2.671875, "learning_rate": 4.9968750000000005e-05, "loss": 0.308, "step": 1600 }, { "epoch": 10.00625, "grad_norm": 2.71875, "learning_rate": 5e-05, "loss": 0.3439, "step": 1601 }, { "epoch": 10.0125, "grad_norm": 2.203125, "learning_rate": 4.999999986650574e-05, "loss": 0.2921, "step": 1602 }, { "epoch": 10.01875, "grad_norm": 2.4375, "learning_rate": 4.9999999466022965e-05, "loss": 0.2662, "step": 1603 }, { "epoch": 10.025, "grad_norm": 2.53125, "learning_rate": 4.999999879855167e-05, "loss": 0.3262, "step": 1604 }, { "epoch": 10.03125, "grad_norm": 1.96875, "learning_rate": 4.999999786409187e-05, "loss": 0.2357, "step": 1605 }, { "epoch": 10.0375, "grad_norm": 2.40625, "learning_rate": 4.999999666264359e-05, "loss": 0.3567, "step": 1606 }, { "epoch": 10.04375, "grad_norm": 1.859375, "learning_rate": 4.999999519420681e-05, "loss": 0.2602, "step": 1607 }, { "epoch": 10.05, "grad_norm": 1.8046875, "learning_rate": 4.999999345878156e-05, "loss": 0.32, "step": 1608 }, { "epoch": 10.05625, "grad_norm": 2.59375, "learning_rate": 4.9999991456367865e-05, "loss": 0.4517, "step": 1609 }, { "epoch": 10.0625, "grad_norm": 2.40625, "learning_rate": 4.9999989186965744e-05, "loss": 0.3519, "step": 1610 }, { "epoch": 10.06875, "grad_norm": 2.265625, "learning_rate": 4.9999986650575226e-05, "loss": 0.3438, "step": 1611 }, { "epoch": 10.075, "grad_norm": 2.109375, "learning_rate": 4.999998384719632e-05, "loss": 0.3143, "step": 1612 }, { "epoch": 10.08125, "grad_norm": 2.578125, "learning_rate": 4.9999980776829075e-05, "loss": 0.3826, "step": 1613 }, { "epoch": 10.0875, "grad_norm": 2.515625, "learning_rate": 4.999997743947351e-05, "loss": 0.2956, "step": 1614 }, { "epoch": 10.09375, "grad_norm": 2.1875, "learning_rate": 4.999997383512967e-05, "loss": 0.2646, "step": 1615 }, { "epoch": 10.1, "grad_norm": 2.171875, "learning_rate": 4.999996996379759e-05, "loss": 0.3362, "step": 1616 }, { "epoch": 10.10625, "grad_norm": 2.21875, "learning_rate": 4.9999965825477314e-05, "loss": 0.3476, "step": 1617 }, { "epoch": 10.1125, "grad_norm": 2.046875, "learning_rate": 4.9999961420168887e-05, "loss": 0.2777, "step": 1618 }, { "epoch": 10.11875, "grad_norm": 1.859375, "learning_rate": 4.999995674787235e-05, "loss": 0.2026, "step": 1619 }, { "epoch": 10.125, "grad_norm": 1.796875, "learning_rate": 4.999995180858774e-05, "loss": 0.2165, "step": 1620 }, { "epoch": 10.13125, "grad_norm": 1.875, "learning_rate": 4.999994660231514e-05, "loss": 0.2813, "step": 1621 }, { "epoch": 10.1375, "grad_norm": 1.953125, "learning_rate": 4.99999411290546e-05, "loss": 0.2196, "step": 1622 }, { "epoch": 10.14375, "grad_norm": 2.625, "learning_rate": 4.999993538880615e-05, "loss": 0.3521, "step": 1623 }, { "epoch": 10.15, "grad_norm": 1.703125, "learning_rate": 4.9999929381569886e-05, "loss": 0.283, "step": 1624 }, { "epoch": 10.15625, "grad_norm": 2.046875, "learning_rate": 4.9999923107345846e-05, "loss": 0.3281, "step": 1625 }, { "epoch": 10.1625, "grad_norm": 2.5625, "learning_rate": 4.999991656613412e-05, "loss": 0.367, "step": 1626 }, { "epoch": 10.16875, "grad_norm": 1.9765625, "learning_rate": 4.999990975793476e-05, "loss": 0.2767, "step": 1627 }, { "epoch": 10.175, "grad_norm": 2.0625, "learning_rate": 4.9999902682747845e-05, "loss": 0.2424, "step": 1628 }, { "epoch": 10.18125, "grad_norm": 1.7578125, "learning_rate": 4.9999895340573444e-05, "loss": 0.2889, "step": 1629 }, { "epoch": 10.1875, "grad_norm": 2.578125, "learning_rate": 4.9999887731411656e-05, "loss": 0.4246, "step": 1630 }, { "epoch": 10.19375, "grad_norm": 1.65625, "learning_rate": 4.9999879855262534e-05, "loss": 0.2067, "step": 1631 }, { "epoch": 10.2, "grad_norm": 2.078125, "learning_rate": 4.9999871712126175e-05, "loss": 0.3396, "step": 1632 }, { "epoch": 10.20625, "grad_norm": 1.890625, "learning_rate": 4.999986330200268e-05, "loss": 0.309, "step": 1633 }, { "epoch": 10.2125, "grad_norm": 2.640625, "learning_rate": 4.9999854624892116e-05, "loss": 0.4116, "step": 1634 }, { "epoch": 10.21875, "grad_norm": 1.9921875, "learning_rate": 4.999984568079459e-05, "loss": 0.2566, "step": 1635 }, { "epoch": 10.225, "grad_norm": 2.265625, "learning_rate": 4.9999836469710195e-05, "loss": 0.2391, "step": 1636 }, { "epoch": 10.23125, "grad_norm": 2.203125, "learning_rate": 4.999982699163902e-05, "loss": 0.3025, "step": 1637 }, { "epoch": 10.2375, "grad_norm": 2.125, "learning_rate": 4.999981724658118e-05, "loss": 0.2722, "step": 1638 }, { "epoch": 10.24375, "grad_norm": 2.40625, "learning_rate": 4.999980723453676e-05, "loss": 0.2209, "step": 1639 }, { "epoch": 10.25, "grad_norm": 2.25, "learning_rate": 4.9999796955505895e-05, "loss": 0.3059, "step": 1640 }, { "epoch": 10.25625, "grad_norm": 2.1875, "learning_rate": 4.9999786409488677e-05, "loss": 0.2935, "step": 1641 }, { "epoch": 10.2625, "grad_norm": 2.390625, "learning_rate": 4.9999775596485216e-05, "loss": 0.2354, "step": 1642 }, { "epoch": 10.26875, "grad_norm": 2.34375, "learning_rate": 4.9999764516495636e-05, "loss": 0.2754, "step": 1643 }, { "epoch": 10.275, "grad_norm": 2.546875, "learning_rate": 4.9999753169520044e-05, "loss": 0.2809, "step": 1644 }, { "epoch": 10.28125, "grad_norm": 2.140625, "learning_rate": 4.999974155555858e-05, "loss": 0.257, "step": 1645 }, { "epoch": 10.2875, "grad_norm": 2.140625, "learning_rate": 4.999972967461135e-05, "loss": 0.3137, "step": 1646 }, { "epoch": 10.29375, "grad_norm": 2.1875, "learning_rate": 4.999971752667849e-05, "loss": 0.2843, "step": 1647 }, { "epoch": 10.3, "grad_norm": 2.03125, "learning_rate": 4.999970511176013e-05, "loss": 0.2412, "step": 1648 }, { "epoch": 10.30625, "grad_norm": 1.78125, "learning_rate": 4.999969242985639e-05, "loss": 0.2133, "step": 1649 }, { "epoch": 10.3125, "grad_norm": 1.875, "learning_rate": 4.999967948096743e-05, "loss": 0.2449, "step": 1650 }, { "epoch": 10.31875, "grad_norm": 2.265625, "learning_rate": 4.999966626509336e-05, "loss": 0.2378, "step": 1651 }, { "epoch": 10.325, "grad_norm": 1.984375, "learning_rate": 4.999965278223435e-05, "loss": 0.3116, "step": 1652 }, { "epoch": 10.33125, "grad_norm": 1.84375, "learning_rate": 4.999963903239052e-05, "loss": 0.2242, "step": 1653 }, { "epoch": 10.3375, "grad_norm": 2.640625, "learning_rate": 4.9999625015562025e-05, "loss": 0.3642, "step": 1654 }, { "epoch": 10.34375, "grad_norm": 2.21875, "learning_rate": 4.9999610731749014e-05, "loss": 0.2714, "step": 1655 }, { "epoch": 10.35, "grad_norm": 2.390625, "learning_rate": 4.9999596180951646e-05, "loss": 0.3579, "step": 1656 }, { "epoch": 10.35625, "grad_norm": 2.75, "learning_rate": 4.9999581363170075e-05, "loss": 0.4247, "step": 1657 }, { "epoch": 10.3625, "grad_norm": 2.515625, "learning_rate": 4.9999566278404444e-05, "loss": 0.3632, "step": 1658 }, { "epoch": 10.36875, "grad_norm": 1.9453125, "learning_rate": 4.999955092665494e-05, "loss": 0.2486, "step": 1659 }, { "epoch": 10.375, "grad_norm": 1.984375, "learning_rate": 4.99995353079217e-05, "loss": 0.2313, "step": 1660 }, { "epoch": 10.38125, "grad_norm": 2.0625, "learning_rate": 4.999951942220491e-05, "loss": 0.2417, "step": 1661 }, { "epoch": 10.3875, "grad_norm": 2.171875, "learning_rate": 4.999950326950473e-05, "loss": 0.2379, "step": 1662 }, { "epoch": 10.39375, "grad_norm": 3.0, "learning_rate": 4.999948684982134e-05, "loss": 0.3088, "step": 1663 }, { "epoch": 10.4, "grad_norm": 1.9921875, "learning_rate": 4.999947016315491e-05, "loss": 0.2394, "step": 1664 }, { "epoch": 10.40625, "grad_norm": 2.09375, "learning_rate": 4.999945320950562e-05, "loss": 0.2963, "step": 1665 }, { "epoch": 10.4125, "grad_norm": 1.84375, "learning_rate": 4.9999435988873646e-05, "loss": 0.2629, "step": 1666 }, { "epoch": 10.41875, "grad_norm": 2.328125, "learning_rate": 4.9999418501259174e-05, "loss": 0.3324, "step": 1667 }, { "epoch": 10.425, "grad_norm": 1.9921875, "learning_rate": 4.9999400746662404e-05, "loss": 0.2472, "step": 1668 }, { "epoch": 10.43125, "grad_norm": 2.109375, "learning_rate": 4.999938272508351e-05, "loss": 0.2159, "step": 1669 }, { "epoch": 10.4375, "grad_norm": 2.375, "learning_rate": 4.999936443652269e-05, "loss": 0.3118, "step": 1670 }, { "epoch": 10.44375, "grad_norm": 2.296875, "learning_rate": 4.999934588098013e-05, "loss": 0.312, "step": 1671 }, { "epoch": 10.45, "grad_norm": 3.15625, "learning_rate": 4.999932705845605e-05, "loss": 0.3698, "step": 1672 }, { "epoch": 10.45625, "grad_norm": 1.9375, "learning_rate": 4.999930796895063e-05, "loss": 0.187, "step": 1673 }, { "epoch": 10.4625, "grad_norm": 1.90625, "learning_rate": 4.999928861246408e-05, "loss": 0.2209, "step": 1674 }, { "epoch": 10.46875, "grad_norm": 2.140625, "learning_rate": 4.9999268988996604e-05, "loss": 0.3351, "step": 1675 }, { "epoch": 10.475, "grad_norm": 1.9453125, "learning_rate": 4.999924909854843e-05, "loss": 0.2578, "step": 1676 }, { "epoch": 10.48125, "grad_norm": 2.328125, "learning_rate": 4.9999228941119745e-05, "loss": 0.2804, "step": 1677 }, { "epoch": 10.4875, "grad_norm": 2.078125, "learning_rate": 4.999920851671078e-05, "loss": 0.2131, "step": 1678 }, { "epoch": 10.49375, "grad_norm": 2.046875, "learning_rate": 4.999918782532175e-05, "loss": 0.1878, "step": 1679 }, { "epoch": 10.5, "grad_norm": 1.890625, "learning_rate": 4.9999166866952866e-05, "loss": 0.2673, "step": 1680 }, { "epoch": 10.50625, "grad_norm": 1.9375, "learning_rate": 4.999914564160437e-05, "loss": 0.247, "step": 1681 }, { "epoch": 10.5125, "grad_norm": 2.03125, "learning_rate": 4.999912414927648e-05, "loss": 0.2727, "step": 1682 }, { "epoch": 10.51875, "grad_norm": 2.015625, "learning_rate": 4.999910238996941e-05, "loss": 0.2579, "step": 1683 }, { "epoch": 10.525, "grad_norm": 2.3125, "learning_rate": 4.999908036368342e-05, "loss": 0.3004, "step": 1684 }, { "epoch": 10.53125, "grad_norm": 1.375, "learning_rate": 4.999905807041873e-05, "loss": 0.1723, "step": 1685 }, { "epoch": 10.5375, "grad_norm": 2.03125, "learning_rate": 4.999903551017558e-05, "loss": 0.3076, "step": 1686 }, { "epoch": 10.54375, "grad_norm": 2.1875, "learning_rate": 4.999901268295421e-05, "loss": 0.3387, "step": 1687 }, { "epoch": 10.55, "grad_norm": 2.25, "learning_rate": 4.9998989588754865e-05, "loss": 0.3555, "step": 1688 }, { "epoch": 10.55625, "grad_norm": 2.046875, "learning_rate": 4.999896622757779e-05, "loss": 0.3678, "step": 1689 }, { "epoch": 10.5625, "grad_norm": 2.234375, "learning_rate": 4.9998942599423236e-05, "loss": 0.309, "step": 1690 }, { "epoch": 10.56875, "grad_norm": 2.015625, "learning_rate": 4.9998918704291466e-05, "loss": 0.2155, "step": 1691 }, { "epoch": 10.575, "grad_norm": 2.671875, "learning_rate": 4.999889454218271e-05, "loss": 0.2918, "step": 1692 }, { "epoch": 10.58125, "grad_norm": 2.640625, "learning_rate": 4.999887011309725e-05, "loss": 0.2777, "step": 1693 }, { "epoch": 10.5875, "grad_norm": 2.125, "learning_rate": 4.999884541703533e-05, "loss": 0.3053, "step": 1694 }, { "epoch": 10.59375, "grad_norm": 2.421875, "learning_rate": 4.999882045399723e-05, "loss": 0.2317, "step": 1695 }, { "epoch": 10.6, "grad_norm": 2.65625, "learning_rate": 4.9998795223983205e-05, "loss": 0.4413, "step": 1696 }, { "epoch": 10.60625, "grad_norm": 1.625, "learning_rate": 4.999876972699352e-05, "loss": 0.2563, "step": 1697 }, { "epoch": 10.6125, "grad_norm": 1.75, "learning_rate": 4.999874396302847e-05, "loss": 0.1943, "step": 1698 }, { "epoch": 10.61875, "grad_norm": 2.296875, "learning_rate": 4.99987179320883e-05, "loss": 0.3602, "step": 1699 }, { "epoch": 10.625, "grad_norm": 2.15625, "learning_rate": 4.999869163417331e-05, "loss": 0.223, "step": 1700 }, { "epoch": 10.63125, "grad_norm": 2.421875, "learning_rate": 4.999866506928377e-05, "loss": 0.3031, "step": 1701 }, { "epoch": 10.6375, "grad_norm": 1.84375, "learning_rate": 4.999863823741997e-05, "loss": 0.294, "step": 1702 }, { "epoch": 10.64375, "grad_norm": 1.8359375, "learning_rate": 4.9998611138582196e-05, "loss": 0.271, "step": 1703 }, { "epoch": 10.65, "grad_norm": 2.328125, "learning_rate": 4.999858377277073e-05, "loss": 0.2096, "step": 1704 }, { "epoch": 10.65625, "grad_norm": 2.40625, "learning_rate": 4.9998556139985875e-05, "loss": 0.3149, "step": 1705 }, { "epoch": 10.6625, "grad_norm": 1.875, "learning_rate": 4.999852824022792e-05, "loss": 0.2916, "step": 1706 }, { "epoch": 10.66875, "grad_norm": 2.046875, "learning_rate": 4.9998500073497156e-05, "loss": 0.2142, "step": 1707 }, { "epoch": 10.675, "grad_norm": 2.015625, "learning_rate": 4.9998471639793895e-05, "loss": 0.2293, "step": 1708 }, { "epoch": 10.68125, "grad_norm": 1.78125, "learning_rate": 4.999844293911844e-05, "loss": 0.2036, "step": 1709 }, { "epoch": 10.6875, "grad_norm": 1.890625, "learning_rate": 4.9998413971471094e-05, "loss": 0.3033, "step": 1710 }, { "epoch": 10.69375, "grad_norm": 2.21875, "learning_rate": 4.9998384736852164e-05, "loss": 0.231, "step": 1711 }, { "epoch": 10.7, "grad_norm": 2.15625, "learning_rate": 4.9998355235261974e-05, "loss": 0.273, "step": 1712 }, { "epoch": 10.70625, "grad_norm": 2.453125, "learning_rate": 4.999832546670082e-05, "loss": 0.3586, "step": 1713 }, { "epoch": 10.7125, "grad_norm": 1.953125, "learning_rate": 4.999829543116904e-05, "loss": 0.2654, "step": 1714 }, { "epoch": 10.71875, "grad_norm": 1.8515625, "learning_rate": 4.999826512866693e-05, "loss": 0.2888, "step": 1715 }, { "epoch": 10.725, "grad_norm": 2.734375, "learning_rate": 4.9998234559194846e-05, "loss": 0.2338, "step": 1716 }, { "epoch": 10.73125, "grad_norm": 2.484375, "learning_rate": 4.999820372275308e-05, "loss": 0.2858, "step": 1717 }, { "epoch": 10.7375, "grad_norm": 2.015625, "learning_rate": 4.9998172619341996e-05, "loss": 0.2602, "step": 1718 }, { "epoch": 10.74375, "grad_norm": 2.046875, "learning_rate": 4.999814124896189e-05, "loss": 0.3197, "step": 1719 }, { "epoch": 10.75, "grad_norm": 1.9296875, "learning_rate": 4.999810961161313e-05, "loss": 0.309, "step": 1720 }, { "epoch": 10.75625, "grad_norm": 1.4140625, "learning_rate": 4.9998077707296033e-05, "loss": 0.2139, "step": 1721 }, { "epoch": 10.7625, "grad_norm": 1.7734375, "learning_rate": 4.9998045536010954e-05, "loss": 0.1967, "step": 1722 }, { "epoch": 10.76875, "grad_norm": 2.109375, "learning_rate": 4.9998013097758214e-05, "loss": 0.3606, "step": 1723 }, { "epoch": 10.775, "grad_norm": 1.3203125, "learning_rate": 4.9997980392538185e-05, "loss": 0.1706, "step": 1724 }, { "epoch": 10.78125, "grad_norm": 2.265625, "learning_rate": 4.99979474203512e-05, "loss": 0.3222, "step": 1725 }, { "epoch": 10.7875, "grad_norm": 1.7109375, "learning_rate": 4.999791418119762e-05, "loss": 0.2328, "step": 1726 }, { "epoch": 10.79375, "grad_norm": 1.9921875, "learning_rate": 4.99978806750778e-05, "loss": 0.2613, "step": 1727 }, { "epoch": 10.8, "grad_norm": 1.9296875, "learning_rate": 4.999784690199209e-05, "loss": 0.2067, "step": 1728 }, { "epoch": 10.80625, "grad_norm": 1.515625, "learning_rate": 4.999781286194085e-05, "loss": 0.2199, "step": 1729 }, { "epoch": 10.8125, "grad_norm": 1.8828125, "learning_rate": 4.999777855492446e-05, "loss": 0.2238, "step": 1730 }, { "epoch": 10.81875, "grad_norm": 2.3125, "learning_rate": 4.999774398094327e-05, "loss": 0.4341, "step": 1731 }, { "epoch": 10.825, "grad_norm": 1.8359375, "learning_rate": 4.999770913999765e-05, "loss": 0.2421, "step": 1732 }, { "epoch": 10.83125, "grad_norm": 2.4375, "learning_rate": 4.999767403208798e-05, "loss": 0.318, "step": 1733 }, { "epoch": 10.8375, "grad_norm": 2.0, "learning_rate": 4.999763865721463e-05, "loss": 0.2719, "step": 1734 }, { "epoch": 10.84375, "grad_norm": 2.109375, "learning_rate": 4.999760301537798e-05, "loss": 0.2113, "step": 1735 }, { "epoch": 10.85, "grad_norm": 1.6796875, "learning_rate": 4.99975671065784e-05, "loss": 0.176, "step": 1736 }, { "epoch": 10.85625, "grad_norm": 1.5234375, "learning_rate": 4.9997530930816295e-05, "loss": 0.1736, "step": 1737 }, { "epoch": 10.8625, "grad_norm": 2.171875, "learning_rate": 4.999749448809204e-05, "loss": 0.3865, "step": 1738 }, { "epoch": 10.86875, "grad_norm": 1.9921875, "learning_rate": 4.999745777840601e-05, "loss": 0.2747, "step": 1739 }, { "epoch": 10.875, "grad_norm": 1.875, "learning_rate": 4.999742080175863e-05, "loss": 0.2908, "step": 1740 }, { "epoch": 10.88125, "grad_norm": 2.0, "learning_rate": 4.999738355815026e-05, "loss": 0.2513, "step": 1741 }, { "epoch": 10.8875, "grad_norm": 1.8515625, "learning_rate": 4.999734604758132e-05, "loss": 0.2652, "step": 1742 }, { "epoch": 10.89375, "grad_norm": 2.28125, "learning_rate": 4.99973082700522e-05, "loss": 0.3041, "step": 1743 }, { "epoch": 10.9, "grad_norm": 2.0625, "learning_rate": 4.9997270225563316e-05, "loss": 0.2271, "step": 1744 }, { "epoch": 10.90625, "grad_norm": 2.234375, "learning_rate": 4.9997231914115064e-05, "loss": 0.2736, "step": 1745 }, { "epoch": 10.9125, "grad_norm": 1.859375, "learning_rate": 4.9997193335707845e-05, "loss": 0.209, "step": 1746 }, { "epoch": 10.91875, "grad_norm": 2.359375, "learning_rate": 4.9997154490342085e-05, "loss": 0.2961, "step": 1747 }, { "epoch": 10.925, "grad_norm": 1.59375, "learning_rate": 4.99971153780182e-05, "loss": 0.1517, "step": 1748 }, { "epoch": 10.93125, "grad_norm": 1.578125, "learning_rate": 4.9997075998736597e-05, "loss": 0.185, "step": 1749 }, { "epoch": 10.9375, "grad_norm": 2.3125, "learning_rate": 4.999703635249771e-05, "loss": 0.2183, "step": 1750 }, { "epoch": 10.94375, "grad_norm": 2.28125, "learning_rate": 4.999699643930194e-05, "loss": 0.3015, "step": 1751 }, { "epoch": 10.95, "grad_norm": 1.671875, "learning_rate": 4.999695625914973e-05, "loss": 0.199, "step": 1752 }, { "epoch": 10.95625, "grad_norm": 2.21875, "learning_rate": 4.999691581204152e-05, "loss": 0.2072, "step": 1753 }, { "epoch": 10.9625, "grad_norm": 2.359375, "learning_rate": 4.999687509797772e-05, "loss": 0.2728, "step": 1754 }, { "epoch": 10.96875, "grad_norm": 2.0625, "learning_rate": 4.9996834116958765e-05, "loss": 0.298, "step": 1755 }, { "epoch": 10.975, "grad_norm": 2.03125, "learning_rate": 4.999679286898512e-05, "loss": 0.2369, "step": 1756 }, { "epoch": 10.98125, "grad_norm": 2.078125, "learning_rate": 4.999675135405719e-05, "loss": 0.218, "step": 1757 }, { "epoch": 10.9875, "grad_norm": 1.8671875, "learning_rate": 4.999670957217545e-05, "loss": 0.2499, "step": 1758 }, { "epoch": 10.99375, "grad_norm": 1.8515625, "learning_rate": 4.999666752334032e-05, "loss": 0.3285, "step": 1759 }, { "epoch": 11.0, "grad_norm": 2.171875, "learning_rate": 4.999662520755226e-05, "loss": 0.2794, "step": 1760 }, { "epoch": 11.00625, "grad_norm": 1.90625, "learning_rate": 4.9996582624811725e-05, "loss": 0.2533, "step": 1761 }, { "epoch": 11.0125, "grad_norm": 2.0625, "learning_rate": 4.999653977511917e-05, "loss": 0.2151, "step": 1762 }, { "epoch": 11.01875, "grad_norm": 2.15625, "learning_rate": 4.999649665847504e-05, "loss": 0.2234, "step": 1763 }, { "epoch": 11.025, "grad_norm": 1.859375, "learning_rate": 4.9996453274879816e-05, "loss": 0.2741, "step": 1764 }, { "epoch": 11.03125, "grad_norm": 2.109375, "learning_rate": 4.9996409624333944e-05, "loss": 0.2493, "step": 1765 }, { "epoch": 11.0375, "grad_norm": 2.03125, "learning_rate": 4.99963657068379e-05, "loss": 0.2025, "step": 1766 }, { "epoch": 11.04375, "grad_norm": 2.28125, "learning_rate": 4.9996321522392146e-05, "loss": 0.2921, "step": 1767 }, { "epoch": 11.05, "grad_norm": 1.90625, "learning_rate": 4.999627707099716e-05, "loss": 0.2486, "step": 1768 }, { "epoch": 11.05625, "grad_norm": 1.984375, "learning_rate": 4.9996232352653414e-05, "loss": 0.2501, "step": 1769 }, { "epoch": 11.0625, "grad_norm": 1.8359375, "learning_rate": 4.9996187367361386e-05, "loss": 0.2631, "step": 1770 }, { "epoch": 11.06875, "grad_norm": 1.953125, "learning_rate": 4.999614211512156e-05, "loss": 0.2322, "step": 1771 }, { "epoch": 11.075, "grad_norm": 1.875, "learning_rate": 4.9996096595934404e-05, "loss": 0.2363, "step": 1772 }, { "epoch": 11.08125, "grad_norm": 2.046875, "learning_rate": 4.999605080980042e-05, "loss": 0.2655, "step": 1773 }, { "epoch": 11.0875, "grad_norm": 1.875, "learning_rate": 4.99960047567201e-05, "loss": 0.256, "step": 1774 }, { "epoch": 11.09375, "grad_norm": 2.234375, "learning_rate": 4.999595843669393e-05, "loss": 0.2884, "step": 1775 }, { "epoch": 11.1, "grad_norm": 1.5390625, "learning_rate": 4.999591184972239e-05, "loss": 0.1516, "step": 1776 }, { "epoch": 11.10625, "grad_norm": 1.921875, "learning_rate": 4.999586499580599e-05, "loss": 0.2041, "step": 1777 }, { "epoch": 11.1125, "grad_norm": 1.9453125, "learning_rate": 4.999581787494524e-05, "loss": 0.1901, "step": 1778 }, { "epoch": 11.11875, "grad_norm": 1.421875, "learning_rate": 4.9995770487140626e-05, "loss": 0.1865, "step": 1779 }, { "epoch": 11.125, "grad_norm": 1.7734375, "learning_rate": 4.999572283239267e-05, "loss": 0.1817, "step": 1780 }, { "epoch": 11.13125, "grad_norm": 2.109375, "learning_rate": 4.9995674910701875e-05, "loss": 0.2279, "step": 1781 }, { "epoch": 11.1375, "grad_norm": 2.125, "learning_rate": 4.9995626722068744e-05, "loss": 0.2721, "step": 1782 }, { "epoch": 11.14375, "grad_norm": 1.71875, "learning_rate": 4.9995578266493797e-05, "loss": 0.2857, "step": 1783 }, { "epoch": 11.15, "grad_norm": 2.640625, "learning_rate": 4.999552954397756e-05, "loss": 0.3467, "step": 1784 }, { "epoch": 11.15625, "grad_norm": 1.53125, "learning_rate": 4.999548055452054e-05, "loss": 0.1615, "step": 1785 }, { "epoch": 11.1625, "grad_norm": 2.09375, "learning_rate": 4.999543129812327e-05, "loss": 0.3463, "step": 1786 }, { "epoch": 11.16875, "grad_norm": 2.109375, "learning_rate": 4.999538177478627e-05, "loss": 0.3489, "step": 1787 }, { "epoch": 11.175, "grad_norm": 2.390625, "learning_rate": 4.999533198451007e-05, "loss": 0.2226, "step": 1788 }, { "epoch": 11.18125, "grad_norm": 1.9765625, "learning_rate": 4.9995281927295215e-05, "loss": 0.2518, "step": 1789 }, { "epoch": 11.1875, "grad_norm": 2.453125, "learning_rate": 4.9995231603142214e-05, "loss": 0.331, "step": 1790 }, { "epoch": 11.19375, "grad_norm": 1.734375, "learning_rate": 4.9995181012051625e-05, "loss": 0.1834, "step": 1791 }, { "epoch": 11.2, "grad_norm": 2.078125, "learning_rate": 4.999513015402398e-05, "loss": 0.3296, "step": 1792 }, { "epoch": 11.20625, "grad_norm": 1.9296875, "learning_rate": 4.9995079029059824e-05, "loss": 0.2396, "step": 1793 }, { "epoch": 11.2125, "grad_norm": 1.6796875, "learning_rate": 4.9995027637159696e-05, "loss": 0.206, "step": 1794 }, { "epoch": 11.21875, "grad_norm": 2.296875, "learning_rate": 4.999497597832417e-05, "loss": 0.372, "step": 1795 }, { "epoch": 11.225, "grad_norm": 1.96875, "learning_rate": 4.999492405255376e-05, "loss": 0.2077, "step": 1796 }, { "epoch": 11.23125, "grad_norm": 1.484375, "learning_rate": 4.999487185984905e-05, "loss": 0.169, "step": 1797 }, { "epoch": 11.2375, "grad_norm": 1.75, "learning_rate": 4.999481940021058e-05, "loss": 0.2144, "step": 1798 }, { "epoch": 11.24375, "grad_norm": 2.046875, "learning_rate": 4.999476667363892e-05, "loss": 0.2721, "step": 1799 }, { "epoch": 11.25, "grad_norm": 2.78125, "learning_rate": 4.999471368013463e-05, "loss": 0.2629, "step": 1800 }, { "epoch": 11.25625, "grad_norm": 2.59375, "learning_rate": 4.999466041969828e-05, "loss": 0.2283, "step": 1801 }, { "epoch": 11.2625, "grad_norm": 1.7265625, "learning_rate": 4.999460689233043e-05, "loss": 0.2783, "step": 1802 }, { "epoch": 11.26875, "grad_norm": 2.0625, "learning_rate": 4.999455309803166e-05, "loss": 0.1855, "step": 1803 }, { "epoch": 11.275, "grad_norm": 2.25, "learning_rate": 4.9994499036802546e-05, "loss": 0.2941, "step": 1804 }, { "epoch": 11.28125, "grad_norm": 1.8125, "learning_rate": 4.999444470864365e-05, "loss": 0.2209, "step": 1805 }, { "epoch": 11.2875, "grad_norm": 2.046875, "learning_rate": 4.9994390113555564e-05, "loss": 0.2095, "step": 1806 }, { "epoch": 11.29375, "grad_norm": 2.234375, "learning_rate": 4.999433525153888e-05, "loss": 0.2235, "step": 1807 }, { "epoch": 11.3, "grad_norm": 2.421875, "learning_rate": 4.9994280122594164e-05, "loss": 0.2564, "step": 1808 }, { "epoch": 11.30625, "grad_norm": 2.078125, "learning_rate": 4.999422472672202e-05, "loss": 0.2216, "step": 1809 }, { "epoch": 11.3125, "grad_norm": 1.8828125, "learning_rate": 4.9994169063923034e-05, "loss": 0.2511, "step": 1810 }, { "epoch": 11.31875, "grad_norm": 1.6875, "learning_rate": 4.99941131341978e-05, "loss": 0.2, "step": 1811 }, { "epoch": 11.325, "grad_norm": 1.3671875, "learning_rate": 4.999405693754692e-05, "loss": 0.1984, "step": 1812 }, { "epoch": 11.33125, "grad_norm": 1.6875, "learning_rate": 4.999400047397098e-05, "loss": 0.1814, "step": 1813 }, { "epoch": 11.3375, "grad_norm": 1.4296875, "learning_rate": 4.99939437434706e-05, "loss": 0.1668, "step": 1814 }, { "epoch": 11.34375, "grad_norm": 1.34375, "learning_rate": 4.9993886746046384e-05, "loss": 0.1697, "step": 1815 }, { "epoch": 11.35, "grad_norm": 1.1953125, "learning_rate": 4.9993829481698925e-05, "loss": 0.1544, "step": 1816 }, { "epoch": 11.35625, "grad_norm": 2.25, "learning_rate": 4.9993771950428846e-05, "loss": 0.3361, "step": 1817 }, { "epoch": 11.3625, "grad_norm": 1.6015625, "learning_rate": 4.9993714152236764e-05, "loss": 0.1908, "step": 1818 }, { "epoch": 11.36875, "grad_norm": 2.515625, "learning_rate": 4.99936560871233e-05, "loss": 0.25, "step": 1819 }, { "epoch": 11.375, "grad_norm": 1.96875, "learning_rate": 4.999359775508905e-05, "loss": 0.2066, "step": 1820 }, { "epoch": 11.38125, "grad_norm": 1.453125, "learning_rate": 4.999353915613467e-05, "loss": 0.1683, "step": 1821 }, { "epoch": 11.3875, "grad_norm": 1.6640625, "learning_rate": 4.999348029026076e-05, "loss": 0.2373, "step": 1822 }, { "epoch": 11.39375, "grad_norm": 1.890625, "learning_rate": 4.9993421157467964e-05, "loss": 0.2039, "step": 1823 }, { "epoch": 11.4, "grad_norm": 1.921875, "learning_rate": 4.999336175775691e-05, "loss": 0.2851, "step": 1824 }, { "epoch": 11.40625, "grad_norm": 1.8515625, "learning_rate": 4.999330209112822e-05, "loss": 0.2614, "step": 1825 }, { "epoch": 11.4125, "grad_norm": 1.9765625, "learning_rate": 4.999324215758255e-05, "loss": 0.2478, "step": 1826 }, { "epoch": 11.41875, "grad_norm": 1.859375, "learning_rate": 4.999318195712053e-05, "loss": 0.3243, "step": 1827 }, { "epoch": 11.425, "grad_norm": 1.7890625, "learning_rate": 4.99931214897428e-05, "loss": 0.2483, "step": 1828 }, { "epoch": 11.43125, "grad_norm": 1.5078125, "learning_rate": 4.9993060755450015e-05, "loss": 0.1777, "step": 1829 }, { "epoch": 11.4375, "grad_norm": 2.0625, "learning_rate": 4.999299975424282e-05, "loss": 0.2555, "step": 1830 }, { "epoch": 11.44375, "grad_norm": 1.8203125, "learning_rate": 4.999293848612186e-05, "loss": 0.2549, "step": 1831 }, { "epoch": 11.45, "grad_norm": 1.7109375, "learning_rate": 4.9992876951087795e-05, "loss": 0.2341, "step": 1832 }, { "epoch": 11.45625, "grad_norm": 2.015625, "learning_rate": 4.9992815149141294e-05, "loss": 0.3167, "step": 1833 }, { "epoch": 11.4625, "grad_norm": 1.875, "learning_rate": 4.9992753080282994e-05, "loss": 0.2445, "step": 1834 }, { "epoch": 11.46875, "grad_norm": 2.015625, "learning_rate": 4.999269074451357e-05, "loss": 0.2441, "step": 1835 }, { "epoch": 11.475, "grad_norm": 1.8125, "learning_rate": 4.9992628141833685e-05, "loss": 0.2746, "step": 1836 }, { "epoch": 11.48125, "grad_norm": 1.8125, "learning_rate": 4.999256527224401e-05, "loss": 0.2388, "step": 1837 }, { "epoch": 11.4875, "grad_norm": 2.0, "learning_rate": 4.9992502135745214e-05, "loss": 0.2353, "step": 1838 }, { "epoch": 11.49375, "grad_norm": 2.484375, "learning_rate": 4.9992438732337974e-05, "loss": 0.2851, "step": 1839 }, { "epoch": 11.5, "grad_norm": 1.765625, "learning_rate": 4.999237506202297e-05, "loss": 0.2841, "step": 1840 }, { "epoch": 11.50625, "grad_norm": 1.4765625, "learning_rate": 4.9992311124800875e-05, "loss": 0.2294, "step": 1841 }, { "epoch": 11.5125, "grad_norm": 2.609375, "learning_rate": 4.999224692067237e-05, "loss": 0.2545, "step": 1842 }, { "epoch": 11.51875, "grad_norm": 1.78125, "learning_rate": 4.9992182449638146e-05, "loss": 0.2255, "step": 1843 }, { "epoch": 11.525, "grad_norm": 1.796875, "learning_rate": 4.9992117711698894e-05, "loss": 0.1981, "step": 1844 }, { "epoch": 11.53125, "grad_norm": 1.78125, "learning_rate": 4.9992052706855304e-05, "loss": 0.2207, "step": 1845 }, { "epoch": 11.5375, "grad_norm": 1.78125, "learning_rate": 4.999198743510807e-05, "loss": 0.2145, "step": 1846 }, { "epoch": 11.54375, "grad_norm": 1.5859375, "learning_rate": 4.999192189645788e-05, "loss": 0.2531, "step": 1847 }, { "epoch": 11.55, "grad_norm": 1.765625, "learning_rate": 4.9991856090905444e-05, "loss": 0.2083, "step": 1848 }, { "epoch": 11.55625, "grad_norm": 1.9609375, "learning_rate": 4.9991790018451465e-05, "loss": 0.1956, "step": 1849 }, { "epoch": 11.5625, "grad_norm": 1.9921875, "learning_rate": 4.999172367909664e-05, "loss": 0.1881, "step": 1850 }, { "epoch": 11.56875, "grad_norm": 1.8203125, "learning_rate": 4.999165707284169e-05, "loss": 0.2364, "step": 1851 }, { "epoch": 11.575, "grad_norm": 2.328125, "learning_rate": 4.999159019968731e-05, "loss": 0.2942, "step": 1852 }, { "epoch": 11.58125, "grad_norm": 2.0625, "learning_rate": 4.999152305963423e-05, "loss": 0.1927, "step": 1853 }, { "epoch": 11.5875, "grad_norm": 2.21875, "learning_rate": 4.999145565268316e-05, "loss": 0.2541, "step": 1854 }, { "epoch": 11.59375, "grad_norm": 2.046875, "learning_rate": 4.9991387978834826e-05, "loss": 0.2209, "step": 1855 }, { "epoch": 11.6, "grad_norm": 2.0625, "learning_rate": 4.999132003808994e-05, "loss": 0.2294, "step": 1856 }, { "epoch": 11.60625, "grad_norm": 2.1875, "learning_rate": 4.999125183044924e-05, "loss": 0.308, "step": 1857 }, { "epoch": 11.6125, "grad_norm": 2.078125, "learning_rate": 4.9991183355913436e-05, "loss": 0.3321, "step": 1858 }, { "epoch": 11.61875, "grad_norm": 1.5625, "learning_rate": 4.999111461448328e-05, "loss": 0.1611, "step": 1859 }, { "epoch": 11.625, "grad_norm": 1.609375, "learning_rate": 4.9991045606159484e-05, "loss": 0.1747, "step": 1860 }, { "epoch": 11.63125, "grad_norm": 1.78125, "learning_rate": 4.999097633094281e-05, "loss": 0.2185, "step": 1861 }, { "epoch": 11.6375, "grad_norm": 2.09375, "learning_rate": 4.999090678883398e-05, "loss": 0.1592, "step": 1862 }, { "epoch": 11.64375, "grad_norm": 1.8359375, "learning_rate": 4.999083697983374e-05, "loss": 0.2614, "step": 1863 }, { "epoch": 11.65, "grad_norm": 2.296875, "learning_rate": 4.9990766903942844e-05, "loss": 0.2039, "step": 1864 }, { "epoch": 11.65625, "grad_norm": 1.640625, "learning_rate": 4.999069656116203e-05, "loss": 0.149, "step": 1865 }, { "epoch": 11.6625, "grad_norm": 2.25, "learning_rate": 4.9990625951492056e-05, "loss": 0.2712, "step": 1866 }, { "epoch": 11.66875, "grad_norm": 1.25, "learning_rate": 4.999055507493368e-05, "loss": 0.1649, "step": 1867 }, { "epoch": 11.675, "grad_norm": 1.7265625, "learning_rate": 4.999048393148764e-05, "loss": 0.1929, "step": 1868 }, { "epoch": 11.68125, "grad_norm": 2.3125, "learning_rate": 4.9990412521154715e-05, "loss": 0.2253, "step": 1869 }, { "epoch": 11.6875, "grad_norm": 1.9609375, "learning_rate": 4.999034084393566e-05, "loss": 0.2265, "step": 1870 }, { "epoch": 11.69375, "grad_norm": 2.09375, "learning_rate": 4.9990268899831235e-05, "loss": 0.2916, "step": 1871 }, { "epoch": 11.7, "grad_norm": 1.7890625, "learning_rate": 4.9990196688842216e-05, "loss": 0.2006, "step": 1872 }, { "epoch": 11.70625, "grad_norm": 1.625, "learning_rate": 4.999012421096938e-05, "loss": 0.2664, "step": 1873 }, { "epoch": 11.7125, "grad_norm": 1.7890625, "learning_rate": 4.9990051466213494e-05, "loss": 0.1924, "step": 1874 }, { "epoch": 11.71875, "grad_norm": 1.828125, "learning_rate": 4.998997845457532e-05, "loss": 0.2231, "step": 1875 }, { "epoch": 11.725, "grad_norm": 2.234375, "learning_rate": 4.998990517605567e-05, "loss": 0.2394, "step": 1876 }, { "epoch": 11.73125, "grad_norm": 2.0625, "learning_rate": 4.99898316306553e-05, "loss": 0.2531, "step": 1877 }, { "epoch": 11.7375, "grad_norm": 1.59375, "learning_rate": 4.998975781837501e-05, "loss": 0.1768, "step": 1878 }, { "epoch": 11.74375, "grad_norm": 1.84375, "learning_rate": 4.9989683739215574e-05, "loss": 0.2193, "step": 1879 }, { "epoch": 11.75, "grad_norm": 1.5546875, "learning_rate": 4.9989609393177805e-05, "loss": 0.1614, "step": 1880 }, { "epoch": 11.75625, "grad_norm": 1.4609375, "learning_rate": 4.998953478026247e-05, "loss": 0.1403, "step": 1881 }, { "epoch": 11.7625, "grad_norm": 2.515625, "learning_rate": 4.998945990047039e-05, "loss": 0.3066, "step": 1882 }, { "epoch": 11.76875, "grad_norm": 2.203125, "learning_rate": 4.998938475380235e-05, "loss": 0.2067, "step": 1883 }, { "epoch": 11.775, "grad_norm": 2.578125, "learning_rate": 4.9989309340259156e-05, "loss": 0.3646, "step": 1884 }, { "epoch": 11.78125, "grad_norm": 2.5, "learning_rate": 4.9989233659841616e-05, "loss": 0.3218, "step": 1885 }, { "epoch": 11.7875, "grad_norm": 1.8125, "learning_rate": 4.998915771255053e-05, "loss": 0.2891, "step": 1886 }, { "epoch": 11.79375, "grad_norm": 1.8828125, "learning_rate": 4.9989081498386726e-05, "loss": 0.2408, "step": 1887 }, { "epoch": 11.8, "grad_norm": 2.109375, "learning_rate": 4.9989005017351e-05, "loss": 0.2182, "step": 1888 }, { "epoch": 11.80625, "grad_norm": 2.25, "learning_rate": 4.998892826944418e-05, "loss": 0.3004, "step": 1889 }, { "epoch": 11.8125, "grad_norm": 1.703125, "learning_rate": 4.998885125466708e-05, "loss": 0.1973, "step": 1890 }, { "epoch": 11.81875, "grad_norm": 2.234375, "learning_rate": 4.9988773973020524e-05, "loss": 0.2907, "step": 1891 }, { "epoch": 11.825, "grad_norm": 1.5, "learning_rate": 4.998869642450534e-05, "loss": 0.1667, "step": 1892 }, { "epoch": 11.83125, "grad_norm": 2.046875, "learning_rate": 4.998861860912235e-05, "loss": 0.2642, "step": 1893 }, { "epoch": 11.8375, "grad_norm": 1.5546875, "learning_rate": 4.9988540526872396e-05, "loss": 0.1716, "step": 1894 }, { "epoch": 11.84375, "grad_norm": 1.65625, "learning_rate": 4.99884621777563e-05, "loss": 0.1564, "step": 1895 }, { "epoch": 11.85, "grad_norm": 1.6875, "learning_rate": 4.99883835617749e-05, "loss": 0.2565, "step": 1896 }, { "epoch": 11.85625, "grad_norm": 1.8515625, "learning_rate": 4.998830467892904e-05, "loss": 0.2589, "step": 1897 }, { "epoch": 11.8625, "grad_norm": 2.0, "learning_rate": 4.998822552921957e-05, "loss": 0.2179, "step": 1898 }, { "epoch": 11.86875, "grad_norm": 1.890625, "learning_rate": 4.998814611264733e-05, "loss": 0.2389, "step": 1899 }, { "epoch": 11.875, "grad_norm": 1.890625, "learning_rate": 4.9988066429213153e-05, "loss": 0.23, "step": 1900 }, { "epoch": 11.88125, "grad_norm": 1.828125, "learning_rate": 4.99879864789179e-05, "loss": 0.2393, "step": 1901 }, { "epoch": 11.8875, "grad_norm": 2.1875, "learning_rate": 4.998790626176243e-05, "loss": 0.2157, "step": 1902 }, { "epoch": 11.89375, "grad_norm": 2.40625, "learning_rate": 4.99878257777476e-05, "loss": 0.2266, "step": 1903 }, { "epoch": 11.9, "grad_norm": 1.421875, "learning_rate": 4.998774502687427e-05, "loss": 0.1741, "step": 1904 }, { "epoch": 11.90625, "grad_norm": 1.3125, "learning_rate": 4.998766400914329e-05, "loss": 0.1597, "step": 1905 }, { "epoch": 11.9125, "grad_norm": 1.96875, "learning_rate": 4.998758272455554e-05, "loss": 0.2909, "step": 1906 }, { "epoch": 11.91875, "grad_norm": 1.5078125, "learning_rate": 4.9987501173111876e-05, "loss": 0.1967, "step": 1907 }, { "epoch": 11.925, "grad_norm": 1.6953125, "learning_rate": 4.998741935481318e-05, "loss": 0.1833, "step": 1908 }, { "epoch": 11.93125, "grad_norm": 1.78125, "learning_rate": 4.9987337269660314e-05, "loss": 0.2543, "step": 1909 }, { "epoch": 11.9375, "grad_norm": 2.46875, "learning_rate": 4.998725491765417e-05, "loss": 0.3472, "step": 1910 }, { "epoch": 11.94375, "grad_norm": 1.46875, "learning_rate": 4.998717229879561e-05, "loss": 0.1754, "step": 1911 }, { "epoch": 11.95, "grad_norm": 2.375, "learning_rate": 4.998708941308553e-05, "loss": 0.3145, "step": 1912 }, { "epoch": 11.95625, "grad_norm": 1.4765625, "learning_rate": 4.998700626052481e-05, "loss": 0.1412, "step": 1913 }, { "epoch": 11.9625, "grad_norm": 2.140625, "learning_rate": 4.998692284111433e-05, "loss": 0.1967, "step": 1914 }, { "epoch": 11.96875, "grad_norm": 1.828125, "learning_rate": 4.998683915485499e-05, "loss": 0.1835, "step": 1915 }, { "epoch": 11.975, "grad_norm": 1.3359375, "learning_rate": 4.9986755201747694e-05, "loss": 0.1567, "step": 1916 }, { "epoch": 11.98125, "grad_norm": 1.375, "learning_rate": 4.998667098179332e-05, "loss": 0.1518, "step": 1917 }, { "epoch": 11.9875, "grad_norm": 1.7265625, "learning_rate": 4.9986586494992775e-05, "loss": 0.2406, "step": 1918 }, { "epoch": 11.99375, "grad_norm": 1.8125, "learning_rate": 4.998650174134696e-05, "loss": 0.1948, "step": 1919 }, { "epoch": 12.0, "grad_norm": 1.484375, "learning_rate": 4.998641672085679e-05, "loss": 0.1698, "step": 1920 }, { "epoch": 12.00625, "grad_norm": 1.703125, "learning_rate": 4.9986331433523156e-05, "loss": 0.2894, "step": 1921 }, { "epoch": 12.0125, "grad_norm": 1.9453125, "learning_rate": 4.998624587934697e-05, "loss": 0.2482, "step": 1922 }, { "epoch": 12.01875, "grad_norm": 1.6875, "learning_rate": 4.9986160058329165e-05, "loss": 0.2117, "step": 1923 }, { "epoch": 12.025, "grad_norm": 1.9453125, "learning_rate": 4.998607397047064e-05, "loss": 0.2721, "step": 1924 }, { "epoch": 12.03125, "grad_norm": 1.84375, "learning_rate": 4.998598761577232e-05, "loss": 0.2359, "step": 1925 }, { "epoch": 12.0375, "grad_norm": 1.4375, "learning_rate": 4.998590099423512e-05, "loss": 0.1639, "step": 1926 }, { "epoch": 12.04375, "grad_norm": 1.59375, "learning_rate": 4.9985814105859986e-05, "loss": 0.1725, "step": 1927 }, { "epoch": 12.05, "grad_norm": 1.7109375, "learning_rate": 4.998572695064783e-05, "loss": 0.2076, "step": 1928 }, { "epoch": 12.05625, "grad_norm": 1.5234375, "learning_rate": 4.998563952859957e-05, "loss": 0.1562, "step": 1929 }, { "epoch": 12.0625, "grad_norm": 2.015625, "learning_rate": 4.9985551839716165e-05, "loss": 0.2807, "step": 1930 }, { "epoch": 12.06875, "grad_norm": 2.375, "learning_rate": 4.998546388399854e-05, "loss": 0.2778, "step": 1931 }, { "epoch": 12.075, "grad_norm": 2.09375, "learning_rate": 4.998537566144763e-05, "loss": 0.2179, "step": 1932 }, { "epoch": 12.08125, "grad_norm": 1.2734375, "learning_rate": 4.998528717206439e-05, "loss": 0.1183, "step": 1933 }, { "epoch": 12.0875, "grad_norm": 1.9140625, "learning_rate": 4.998519841584975e-05, "loss": 0.1955, "step": 1934 }, { "epoch": 12.09375, "grad_norm": 1.7109375, "learning_rate": 4.9985109392804666e-05, "loss": 0.2168, "step": 1935 }, { "epoch": 12.1, "grad_norm": 2.046875, "learning_rate": 4.998502010293009e-05, "loss": 0.2369, "step": 1936 }, { "epoch": 12.10625, "grad_norm": 1.78125, "learning_rate": 4.9984930546226975e-05, "loss": 0.2013, "step": 1937 }, { "epoch": 12.1125, "grad_norm": 1.375, "learning_rate": 4.9984840722696275e-05, "loss": 0.1432, "step": 1938 }, { "epoch": 12.11875, "grad_norm": 1.7578125, "learning_rate": 4.998475063233895e-05, "loss": 0.1752, "step": 1939 }, { "epoch": 12.125, "grad_norm": 1.6796875, "learning_rate": 4.9984660275155956e-05, "loss": 0.154, "step": 1940 }, { "epoch": 12.13125, "grad_norm": 2.03125, "learning_rate": 4.9984569651148275e-05, "loss": 0.2758, "step": 1941 }, { "epoch": 12.1375, "grad_norm": 2.03125, "learning_rate": 4.9984478760316854e-05, "loss": 0.2406, "step": 1942 }, { "epoch": 12.14375, "grad_norm": 2.390625, "learning_rate": 4.9984387602662675e-05, "loss": 0.2881, "step": 1943 }, { "epoch": 12.15, "grad_norm": 1.59375, "learning_rate": 4.998429617818672e-05, "loss": 0.2057, "step": 1944 }, { "epoch": 12.15625, "grad_norm": 1.8203125, "learning_rate": 4.998420448688994e-05, "loss": 0.2293, "step": 1945 }, { "epoch": 12.1625, "grad_norm": 2.125, "learning_rate": 4.998411252877334e-05, "loss": 0.2689, "step": 1946 }, { "epoch": 12.16875, "grad_norm": 1.703125, "learning_rate": 4.998402030383789e-05, "loss": 0.1799, "step": 1947 }, { "epoch": 12.175, "grad_norm": 1.9765625, "learning_rate": 4.998392781208457e-05, "loss": 0.2483, "step": 1948 }, { "epoch": 12.18125, "grad_norm": 2.453125, "learning_rate": 4.9983835053514386e-05, "loss": 0.2816, "step": 1949 }, { "epoch": 12.1875, "grad_norm": 2.203125, "learning_rate": 4.998374202812831e-05, "loss": 0.1863, "step": 1950 }, { "epoch": 12.19375, "grad_norm": 2.28125, "learning_rate": 4.998364873592735e-05, "loss": 0.2842, "step": 1951 }, { "epoch": 12.2, "grad_norm": 2.25, "learning_rate": 4.998355517691249e-05, "loss": 0.2203, "step": 1952 }, { "epoch": 12.20625, "grad_norm": 2.046875, "learning_rate": 4.9983461351084735e-05, "loss": 0.2201, "step": 1953 }, { "epoch": 12.2125, "grad_norm": 1.9375, "learning_rate": 4.9983367258445076e-05, "loss": 0.2255, "step": 1954 }, { "epoch": 12.21875, "grad_norm": 1.5078125, "learning_rate": 4.9983272898994546e-05, "loss": 0.1573, "step": 1955 }, { "epoch": 12.225, "grad_norm": 1.4453125, "learning_rate": 4.998317827273412e-05, "loss": 0.1532, "step": 1956 }, { "epoch": 12.23125, "grad_norm": 2.3125, "learning_rate": 4.998308337966483e-05, "loss": 0.2615, "step": 1957 }, { "epoch": 12.2375, "grad_norm": 1.734375, "learning_rate": 4.998298821978769e-05, "loss": 0.2614, "step": 1958 }, { "epoch": 12.24375, "grad_norm": 1.765625, "learning_rate": 4.998289279310369e-05, "loss": 0.1994, "step": 1959 }, { "epoch": 12.25, "grad_norm": 1.7109375, "learning_rate": 4.998279709961388e-05, "loss": 0.2846, "step": 1960 }, { "epoch": 12.25625, "grad_norm": 1.6875, "learning_rate": 4.998270113931927e-05, "loss": 0.1657, "step": 1961 }, { "epoch": 12.2625, "grad_norm": 1.9296875, "learning_rate": 4.9982604912220876e-05, "loss": 0.2153, "step": 1962 }, { "epoch": 12.26875, "grad_norm": 1.9921875, "learning_rate": 4.998250841831974e-05, "loss": 0.2528, "step": 1963 }, { "epoch": 12.275, "grad_norm": 1.9296875, "learning_rate": 4.998241165761689e-05, "loss": 0.2248, "step": 1964 }, { "epoch": 12.28125, "grad_norm": 1.6015625, "learning_rate": 4.9982314630113355e-05, "loss": 0.1608, "step": 1965 }, { "epoch": 12.2875, "grad_norm": 1.8828125, "learning_rate": 4.998221733581018e-05, "loss": 0.2563, "step": 1966 }, { "epoch": 12.29375, "grad_norm": 1.53125, "learning_rate": 4.998211977470838e-05, "loss": 0.1486, "step": 1967 }, { "epoch": 12.3, "grad_norm": 1.96875, "learning_rate": 4.9982021946809025e-05, "loss": 0.344, "step": 1968 }, { "epoch": 12.30625, "grad_norm": 1.8046875, "learning_rate": 4.9981923852113145e-05, "loss": 0.2359, "step": 1969 }, { "epoch": 12.3125, "grad_norm": 2.171875, "learning_rate": 4.998182549062179e-05, "loss": 0.2006, "step": 1970 }, { "epoch": 12.31875, "grad_norm": 1.7734375, "learning_rate": 4.9981726862336014e-05, "loss": 0.288, "step": 1971 }, { "epoch": 12.325, "grad_norm": 1.8828125, "learning_rate": 4.998162796725686e-05, "loss": 0.2308, "step": 1972 }, { "epoch": 12.33125, "grad_norm": 1.8515625, "learning_rate": 4.99815288053854e-05, "loss": 0.1513, "step": 1973 }, { "epoch": 12.3375, "grad_norm": 1.5546875, "learning_rate": 4.998142937672269e-05, "loss": 0.1782, "step": 1974 }, { "epoch": 12.34375, "grad_norm": 1.875, "learning_rate": 4.998132968126978e-05, "loss": 0.1957, "step": 1975 }, { "epoch": 12.35, "grad_norm": 2.015625, "learning_rate": 4.998122971902774e-05, "loss": 0.3081, "step": 1976 }, { "epoch": 12.35625, "grad_norm": 2.140625, "learning_rate": 4.998112948999764e-05, "loss": 0.2783, "step": 1977 }, { "epoch": 12.3625, "grad_norm": 1.625, "learning_rate": 4.9981028994180546e-05, "loss": 0.1902, "step": 1978 }, { "epoch": 12.36875, "grad_norm": 1.3515625, "learning_rate": 4.998092823157754e-05, "loss": 0.1288, "step": 1979 }, { "epoch": 12.375, "grad_norm": 1.828125, "learning_rate": 4.9980827202189687e-05, "loss": 0.2056, "step": 1980 }, { "epoch": 12.38125, "grad_norm": 1.625, "learning_rate": 4.9980725906018074e-05, "loss": 0.1485, "step": 1981 }, { "epoch": 12.3875, "grad_norm": 2.0, "learning_rate": 4.9980624343063786e-05, "loss": 0.2202, "step": 1982 }, { "epoch": 12.39375, "grad_norm": 1.53125, "learning_rate": 4.998052251332789e-05, "loss": 0.1735, "step": 1983 }, { "epoch": 12.4, "grad_norm": 2.171875, "learning_rate": 4.99804204168115e-05, "loss": 0.2392, "step": 1984 }, { "epoch": 12.40625, "grad_norm": 1.984375, "learning_rate": 4.9980318053515686e-05, "loss": 0.2335, "step": 1985 }, { "epoch": 12.4125, "grad_norm": 1.921875, "learning_rate": 4.998021542344154e-05, "loss": 0.2367, "step": 1986 }, { "epoch": 12.41875, "grad_norm": 2.015625, "learning_rate": 4.998011252659017e-05, "loss": 0.2489, "step": 1987 }, { "epoch": 12.425, "grad_norm": 1.2421875, "learning_rate": 4.998000936296267e-05, "loss": 0.1277, "step": 1988 }, { "epoch": 12.43125, "grad_norm": 2.21875, "learning_rate": 4.997990593256015e-05, "loss": 0.2542, "step": 1989 }, { "epoch": 12.4375, "grad_norm": 1.8515625, "learning_rate": 4.997980223538369e-05, "loss": 0.1757, "step": 1990 }, { "epoch": 12.44375, "grad_norm": 1.4921875, "learning_rate": 4.9979698271434425e-05, "loss": 0.1504, "step": 1991 }, { "epoch": 12.45, "grad_norm": 2.15625, "learning_rate": 4.9979594040713454e-05, "loss": 0.2805, "step": 1992 }, { "epoch": 12.45625, "grad_norm": 1.6640625, "learning_rate": 4.997948954322189e-05, "loss": 0.2167, "step": 1993 }, { "epoch": 12.4625, "grad_norm": 1.3515625, "learning_rate": 4.997938477896084e-05, "loss": 0.1907, "step": 1994 }, { "epoch": 12.46875, "grad_norm": 1.7734375, "learning_rate": 4.997927974793144e-05, "loss": 0.2496, "step": 1995 }, { "epoch": 12.475, "grad_norm": 1.625, "learning_rate": 4.99791744501348e-05, "loss": 0.1947, "step": 1996 }, { "epoch": 12.48125, "grad_norm": 1.7890625, "learning_rate": 4.9979068885572055e-05, "loss": 0.2643, "step": 1997 }, { "epoch": 12.4875, "grad_norm": 1.765625, "learning_rate": 4.997896305424432e-05, "loss": 0.2159, "step": 1998 }, { "epoch": 12.49375, "grad_norm": 1.3828125, "learning_rate": 4.997885695615273e-05, "loss": 0.16, "step": 1999 }, { "epoch": 12.5, "grad_norm": 1.984375, "learning_rate": 4.997875059129843e-05, "loss": 0.251, "step": 2000 }, { "epoch": 12.50625, "grad_norm": 1.5234375, "learning_rate": 4.997864395968252e-05, "loss": 0.1556, "step": 2001 }, { "epoch": 12.5125, "grad_norm": 1.359375, "learning_rate": 4.997853706130619e-05, "loss": 0.1453, "step": 2002 }, { "epoch": 12.51875, "grad_norm": 1.671875, "learning_rate": 4.997842989617054e-05, "loss": 0.1715, "step": 2003 }, { "epoch": 12.525, "grad_norm": 1.8203125, "learning_rate": 4.997832246427674e-05, "loss": 0.2558, "step": 2004 }, { "epoch": 12.53125, "grad_norm": 1.5859375, "learning_rate": 4.997821476562592e-05, "loss": 0.1891, "step": 2005 }, { "epoch": 12.5375, "grad_norm": 1.7421875, "learning_rate": 4.997810680021924e-05, "loss": 0.1692, "step": 2006 }, { "epoch": 12.54375, "grad_norm": 1.296875, "learning_rate": 4.997799856805785e-05, "loss": 0.1567, "step": 2007 }, { "epoch": 12.55, "grad_norm": 1.578125, "learning_rate": 4.9977890069142894e-05, "loss": 0.2929, "step": 2008 }, { "epoch": 12.55625, "grad_norm": 1.734375, "learning_rate": 4.9977781303475556e-05, "loss": 0.2045, "step": 2009 }, { "epoch": 12.5625, "grad_norm": 1.3046875, "learning_rate": 4.9977672271056985e-05, "loss": 0.1617, "step": 2010 }, { "epoch": 12.56875, "grad_norm": 1.6484375, "learning_rate": 4.9977562971888345e-05, "loss": 0.2296, "step": 2011 }, { "epoch": 12.575, "grad_norm": 1.46875, "learning_rate": 4.9977453405970795e-05, "loss": 0.1693, "step": 2012 }, { "epoch": 12.58125, "grad_norm": 1.6640625, "learning_rate": 4.997734357330553e-05, "loss": 0.2177, "step": 2013 }, { "epoch": 12.5875, "grad_norm": 1.7578125, "learning_rate": 4.997723347389368e-05, "loss": 0.2973, "step": 2014 }, { "epoch": 12.59375, "grad_norm": 1.7265625, "learning_rate": 4.997712310773647e-05, "loss": 0.1985, "step": 2015 }, { "epoch": 12.6, "grad_norm": 1.234375, "learning_rate": 4.997701247483504e-05, "loss": 0.1319, "step": 2016 }, { "epoch": 12.60625, "grad_norm": 1.0234375, "learning_rate": 4.997690157519059e-05, "loss": 0.1201, "step": 2017 }, { "epoch": 12.6125, "grad_norm": 1.4375, "learning_rate": 4.997679040880431e-05, "loss": 0.1593, "step": 2018 }, { "epoch": 12.61875, "grad_norm": 1.34375, "learning_rate": 4.9976678975677376e-05, "loss": 0.1785, "step": 2019 }, { "epoch": 12.625, "grad_norm": 1.3828125, "learning_rate": 4.997656727581098e-05, "loss": 0.1592, "step": 2020 }, { "epoch": 12.63125, "grad_norm": 1.578125, "learning_rate": 4.9976455309206316e-05, "loss": 0.1825, "step": 2021 }, { "epoch": 12.6375, "grad_norm": 2.03125, "learning_rate": 4.9976343075864576e-05, "loss": 0.2777, "step": 2022 }, { "epoch": 12.64375, "grad_norm": 1.578125, "learning_rate": 4.997623057578696e-05, "loss": 0.2138, "step": 2023 }, { "epoch": 12.65, "grad_norm": 1.640625, "learning_rate": 4.997611780897468e-05, "loss": 0.1971, "step": 2024 }, { "epoch": 12.65625, "grad_norm": 1.828125, "learning_rate": 4.9976004775428924e-05, "loss": 0.2559, "step": 2025 }, { "epoch": 12.6625, "grad_norm": 2.140625, "learning_rate": 4.997589147515091e-05, "loss": 0.2739, "step": 2026 }, { "epoch": 12.66875, "grad_norm": 1.6328125, "learning_rate": 4.997577790814185e-05, "loss": 0.1868, "step": 2027 }, { "epoch": 12.675, "grad_norm": 1.859375, "learning_rate": 4.997566407440294e-05, "loss": 0.1927, "step": 2028 }, { "epoch": 12.68125, "grad_norm": 2.265625, "learning_rate": 4.9975549973935413e-05, "loss": 0.2512, "step": 2029 }, { "epoch": 12.6875, "grad_norm": 1.5, "learning_rate": 4.9975435606740486e-05, "loss": 0.1866, "step": 2030 }, { "epoch": 12.69375, "grad_norm": 1.375, "learning_rate": 4.997532097281937e-05, "loss": 0.1689, "step": 2031 }, { "epoch": 12.7, "grad_norm": 1.40625, "learning_rate": 4.9975206072173304e-05, "loss": 0.1582, "step": 2032 }, { "epoch": 12.70625, "grad_norm": 1.453125, "learning_rate": 4.99750909048035e-05, "loss": 0.1729, "step": 2033 }, { "epoch": 12.7125, "grad_norm": 1.5859375, "learning_rate": 4.997497547071119e-05, "loss": 0.1626, "step": 2034 }, { "epoch": 12.71875, "grad_norm": 1.625, "learning_rate": 4.997485976989762e-05, "loss": 0.2007, "step": 2035 }, { "epoch": 12.725, "grad_norm": 1.9921875, "learning_rate": 4.997474380236402e-05, "loss": 0.2706, "step": 2036 }, { "epoch": 12.73125, "grad_norm": 1.3828125, "learning_rate": 4.997462756811162e-05, "loss": 0.1875, "step": 2037 }, { "epoch": 12.7375, "grad_norm": 1.9140625, "learning_rate": 4.997451106714166e-05, "loss": 0.2291, "step": 2038 }, { "epoch": 12.74375, "grad_norm": 2.234375, "learning_rate": 4.9974394299455404e-05, "loss": 0.2263, "step": 2039 }, { "epoch": 12.75, "grad_norm": 1.84375, "learning_rate": 4.997427726505408e-05, "loss": 0.1642, "step": 2040 }, { "epoch": 12.75625, "grad_norm": 1.7265625, "learning_rate": 4.997415996393894e-05, "loss": 0.1959, "step": 2041 }, { "epoch": 12.7625, "grad_norm": 1.734375, "learning_rate": 4.9974042396111246e-05, "loss": 0.2413, "step": 2042 }, { "epoch": 12.76875, "grad_norm": 1.78125, "learning_rate": 4.997392456157225e-05, "loss": 0.2068, "step": 2043 }, { "epoch": 12.775, "grad_norm": 1.8984375, "learning_rate": 4.99738064603232e-05, "loss": 0.2358, "step": 2044 }, { "epoch": 12.78125, "grad_norm": 1.8671875, "learning_rate": 4.997368809236537e-05, "loss": 0.215, "step": 2045 }, { "epoch": 12.7875, "grad_norm": 1.3984375, "learning_rate": 4.997356945770002e-05, "loss": 0.1719, "step": 2046 }, { "epoch": 12.79375, "grad_norm": 1.578125, "learning_rate": 4.997345055632842e-05, "loss": 0.1846, "step": 2047 }, { "epoch": 12.8, "grad_norm": 1.640625, "learning_rate": 4.997333138825183e-05, "loss": 0.2067, "step": 2048 }, { "epoch": 12.80625, "grad_norm": 1.5078125, "learning_rate": 4.997321195347154e-05, "loss": 0.1537, "step": 2049 }, { "epoch": 12.8125, "grad_norm": 1.890625, "learning_rate": 4.997309225198881e-05, "loss": 0.1923, "step": 2050 }, { "epoch": 12.81875, "grad_norm": 1.1640625, "learning_rate": 4.997297228380492e-05, "loss": 0.1188, "step": 2051 }, { "epoch": 12.825, "grad_norm": 1.9140625, "learning_rate": 4.997285204892116e-05, "loss": 0.1887, "step": 2052 }, { "epoch": 12.83125, "grad_norm": 1.6796875, "learning_rate": 4.99727315473388e-05, "loss": 0.2215, "step": 2053 }, { "epoch": 12.8375, "grad_norm": 1.640625, "learning_rate": 4.997261077905915e-05, "loss": 0.251, "step": 2054 }, { "epoch": 12.84375, "grad_norm": 1.765625, "learning_rate": 4.997248974408347e-05, "loss": 0.1763, "step": 2055 }, { "epoch": 12.85, "grad_norm": 1.453125, "learning_rate": 4.9972368442413076e-05, "loss": 0.2313, "step": 2056 }, { "epoch": 12.85625, "grad_norm": 2.15625, "learning_rate": 4.9972246874049254e-05, "loss": 0.2941, "step": 2057 }, { "epoch": 12.8625, "grad_norm": 1.828125, "learning_rate": 4.99721250389933e-05, "loss": 0.2295, "step": 2058 }, { "epoch": 12.86875, "grad_norm": 1.4375, "learning_rate": 4.997200293724652e-05, "loss": 0.1365, "step": 2059 }, { "epoch": 12.875, "grad_norm": 1.7265625, "learning_rate": 4.997188056881022e-05, "loss": 0.2211, "step": 2060 }, { "epoch": 12.88125, "grad_norm": 1.71875, "learning_rate": 4.99717579336857e-05, "loss": 0.1836, "step": 2061 }, { "epoch": 12.8875, "grad_norm": 2.25, "learning_rate": 4.9971635031874273e-05, "loss": 0.2819, "step": 2062 }, { "epoch": 12.89375, "grad_norm": 2.921875, "learning_rate": 4.997151186337726e-05, "loss": 0.1728, "step": 2063 }, { "epoch": 12.9, "grad_norm": 2.921875, "learning_rate": 4.997138842819596e-05, "loss": 0.31, "step": 2064 }, { "epoch": 12.90625, "grad_norm": 1.625, "learning_rate": 4.99712647263317e-05, "loss": 0.1981, "step": 2065 }, { "epoch": 12.9125, "grad_norm": 1.5546875, "learning_rate": 4.99711407577858e-05, "loss": 0.2046, "step": 2066 }, { "epoch": 12.91875, "grad_norm": 2.421875, "learning_rate": 4.9971016522559586e-05, "loss": 0.2368, "step": 2067 }, { "epoch": 12.925, "grad_norm": 2.59375, "learning_rate": 4.997089202065438e-05, "loss": 0.3217, "step": 2068 }, { "epoch": 12.93125, "grad_norm": 1.8828125, "learning_rate": 4.997076725207152e-05, "loss": 0.1776, "step": 2069 }, { "epoch": 12.9375, "grad_norm": 1.5, "learning_rate": 4.9970642216812334e-05, "loss": 0.165, "step": 2070 }, { "epoch": 12.94375, "grad_norm": 1.5625, "learning_rate": 4.9970516914878154e-05, "loss": 0.1764, "step": 2071 }, { "epoch": 12.95, "grad_norm": 1.90625, "learning_rate": 4.9970391346270315e-05, "loss": 0.2285, "step": 2072 }, { "epoch": 12.95625, "grad_norm": 2.125, "learning_rate": 4.997026551099017e-05, "loss": 0.2622, "step": 2073 }, { "epoch": 12.9625, "grad_norm": 1.6171875, "learning_rate": 4.997013940903906e-05, "loss": 0.2179, "step": 2074 }, { "epoch": 12.96875, "grad_norm": 2.25, "learning_rate": 4.997001304041832e-05, "loss": 0.2751, "step": 2075 }, { "epoch": 12.975, "grad_norm": 1.71875, "learning_rate": 4.996988640512931e-05, "loss": 0.2035, "step": 2076 }, { "epoch": 12.98125, "grad_norm": 2.015625, "learning_rate": 4.996975950317339e-05, "loss": 0.1752, "step": 2077 }, { "epoch": 12.9875, "grad_norm": 1.6015625, "learning_rate": 4.996963233455189e-05, "loss": 0.201, "step": 2078 }, { "epoch": 12.99375, "grad_norm": 1.3515625, "learning_rate": 4.99695048992662e-05, "loss": 0.1386, "step": 2079 }, { "epoch": 13.0, "grad_norm": 1.8359375, "learning_rate": 4.9969377197317647e-05, "loss": 0.2039, "step": 2080 }, { "epoch": 13.00625, "grad_norm": 2.109375, "learning_rate": 4.996924922870762e-05, "loss": 0.2199, "step": 2081 }, { "epoch": 13.0125, "grad_norm": 1.90625, "learning_rate": 4.996912099343748e-05, "loss": 0.2017, "step": 2082 }, { "epoch": 13.01875, "grad_norm": 0.984375, "learning_rate": 4.9968992491508595e-05, "loss": 0.1103, "step": 2083 }, { "epoch": 13.025, "grad_norm": 1.59375, "learning_rate": 4.996886372292233e-05, "loss": 0.2421, "step": 2084 }, { "epoch": 13.03125, "grad_norm": 1.6875, "learning_rate": 4.996873468768008e-05, "loss": 0.1496, "step": 2085 }, { "epoch": 13.0375, "grad_norm": 1.5703125, "learning_rate": 4.99686053857832e-05, "loss": 0.1566, "step": 2086 }, { "epoch": 13.04375, "grad_norm": 1.609375, "learning_rate": 4.996847581723309e-05, "loss": 0.2316, "step": 2087 }, { "epoch": 13.05, "grad_norm": 1.7109375, "learning_rate": 4.996834598203111e-05, "loss": 0.1982, "step": 2088 }, { "epoch": 13.05625, "grad_norm": 1.5625, "learning_rate": 4.996821588017867e-05, "loss": 0.1687, "step": 2089 }, { "epoch": 13.0625, "grad_norm": 1.890625, "learning_rate": 4.996808551167715e-05, "loss": 0.1972, "step": 2090 }, { "epoch": 13.06875, "grad_norm": 1.859375, "learning_rate": 4.996795487652794e-05, "loss": 0.2166, "step": 2091 }, { "epoch": 13.075, "grad_norm": 2.015625, "learning_rate": 4.996782397473244e-05, "loss": 0.2818, "step": 2092 }, { "epoch": 13.08125, "grad_norm": 1.1875, "learning_rate": 4.9967692806292056e-05, "loss": 0.1198, "step": 2093 }, { "epoch": 13.0875, "grad_norm": 1.8125, "learning_rate": 4.996756137120817e-05, "loss": 0.2125, "step": 2094 }, { "epoch": 13.09375, "grad_norm": 1.125, "learning_rate": 4.996742966948219e-05, "loss": 0.1359, "step": 2095 }, { "epoch": 13.1, "grad_norm": 1.59375, "learning_rate": 4.9967297701115534e-05, "loss": 0.1609, "step": 2096 }, { "epoch": 13.10625, "grad_norm": 2.15625, "learning_rate": 4.99671654661096e-05, "loss": 0.2207, "step": 2097 }, { "epoch": 13.1125, "grad_norm": 1.421875, "learning_rate": 4.996703296446581e-05, "loss": 0.209, "step": 2098 }, { "epoch": 13.11875, "grad_norm": 2.1875, "learning_rate": 4.9966900196185574e-05, "loss": 0.2144, "step": 2099 }, { "epoch": 13.125, "grad_norm": 1.390625, "learning_rate": 4.99667671612703e-05, "loss": 0.1517, "step": 2100 }, { "epoch": 13.13125, "grad_norm": 1.359375, "learning_rate": 4.996663385972142e-05, "loss": 0.1309, "step": 2101 }, { "epoch": 13.1375, "grad_norm": 1.9375, "learning_rate": 4.996650029154036e-05, "loss": 0.249, "step": 2102 }, { "epoch": 13.14375, "grad_norm": 1.671875, "learning_rate": 4.996636645672855e-05, "loss": 0.2243, "step": 2103 }, { "epoch": 13.15, "grad_norm": 1.65625, "learning_rate": 4.9966232355287404e-05, "loss": 0.2153, "step": 2104 }, { "epoch": 13.15625, "grad_norm": 1.7265625, "learning_rate": 4.9966097987218365e-05, "loss": 0.1638, "step": 2105 }, { "epoch": 13.1625, "grad_norm": 1.8828125, "learning_rate": 4.996596335252286e-05, "loss": 0.1984, "step": 2106 }, { "epoch": 13.16875, "grad_norm": 1.75, "learning_rate": 4.9965828451202334e-05, "loss": 0.2154, "step": 2107 }, { "epoch": 13.175, "grad_norm": 1.640625, "learning_rate": 4.996569328325823e-05, "loss": 0.2373, "step": 2108 }, { "epoch": 13.18125, "grad_norm": 2.21875, "learning_rate": 4.996555784869198e-05, "loss": 0.3068, "step": 2109 }, { "epoch": 13.1875, "grad_norm": 1.78125, "learning_rate": 4.9965422147505036e-05, "loss": 0.1612, "step": 2110 }, { "epoch": 13.19375, "grad_norm": 1.6328125, "learning_rate": 4.9965286179698855e-05, "loss": 0.1486, "step": 2111 }, { "epoch": 13.2, "grad_norm": 2.125, "learning_rate": 4.996514994527488e-05, "loss": 0.3026, "step": 2112 }, { "epoch": 13.20625, "grad_norm": 1.46875, "learning_rate": 4.996501344423456e-05, "loss": 0.1705, "step": 2113 }, { "epoch": 13.2125, "grad_norm": 1.9296875, "learning_rate": 4.996487667657937e-05, "loss": 0.2389, "step": 2114 }, { "epoch": 13.21875, "grad_norm": 2.046875, "learning_rate": 4.996473964231076e-05, "loss": 0.2833, "step": 2115 }, { "epoch": 13.225, "grad_norm": 2.078125, "learning_rate": 4.99646023414302e-05, "loss": 0.2992, "step": 2116 }, { "epoch": 13.23125, "grad_norm": 1.390625, "learning_rate": 4.9964464773939155e-05, "loss": 0.2026, "step": 2117 }, { "epoch": 13.2375, "grad_norm": 2.109375, "learning_rate": 4.996432693983908e-05, "loss": 0.2179, "step": 2118 }, { "epoch": 13.24375, "grad_norm": 2.125, "learning_rate": 4.9964188839131464e-05, "loss": 0.2747, "step": 2119 }, { "epoch": 13.25, "grad_norm": 2.53125, "learning_rate": 4.996405047181777e-05, "loss": 0.2114, "step": 2120 }, { "epoch": 13.25625, "grad_norm": 1.453125, "learning_rate": 4.9963911837899494e-05, "loss": 0.1391, "step": 2121 }, { "epoch": 13.2625, "grad_norm": 1.2265625, "learning_rate": 4.9963772937378086e-05, "loss": 0.1924, "step": 2122 }, { "epoch": 13.26875, "grad_norm": 2.484375, "learning_rate": 4.9963633770255056e-05, "loss": 0.2845, "step": 2123 }, { "epoch": 13.275, "grad_norm": 2.296875, "learning_rate": 4.996349433653188e-05, "loss": 0.2588, "step": 2124 }, { "epoch": 13.28125, "grad_norm": 1.5546875, "learning_rate": 4.996335463621005e-05, "loss": 0.1351, "step": 2125 }, { "epoch": 13.2875, "grad_norm": 1.3671875, "learning_rate": 4.996321466929106e-05, "loss": 0.2079, "step": 2126 }, { "epoch": 13.29375, "grad_norm": 1.921875, "learning_rate": 4.99630744357764e-05, "loss": 0.2456, "step": 2127 }, { "epoch": 13.3, "grad_norm": 2.625, "learning_rate": 4.9962933935667565e-05, "loss": 0.3224, "step": 2128 }, { "epoch": 13.30625, "grad_norm": 1.8359375, "learning_rate": 4.996279316896606e-05, "loss": 0.2165, "step": 2129 }, { "epoch": 13.3125, "grad_norm": 1.90625, "learning_rate": 4.996265213567339e-05, "loss": 0.1873, "step": 2130 }, { "epoch": 13.31875, "grad_norm": 2.015625, "learning_rate": 4.996251083579106e-05, "loss": 0.2442, "step": 2131 }, { "epoch": 13.325, "grad_norm": 1.9765625, "learning_rate": 4.996236926932057e-05, "loss": 0.2417, "step": 2132 }, { "epoch": 13.33125, "grad_norm": 1.921875, "learning_rate": 4.9962227436263453e-05, "loss": 0.2605, "step": 2133 }, { "epoch": 13.3375, "grad_norm": 1.4921875, "learning_rate": 4.99620853366212e-05, "loss": 0.2375, "step": 2134 }, { "epoch": 13.34375, "grad_norm": 1.46875, "learning_rate": 4.996194297039535e-05, "loss": 0.134, "step": 2135 }, { "epoch": 13.35, "grad_norm": 2.140625, "learning_rate": 4.99618003375874e-05, "loss": 0.2277, "step": 2136 }, { "epoch": 13.35625, "grad_norm": 1.2578125, "learning_rate": 4.996165743819889e-05, "loss": 0.1241, "step": 2137 }, { "epoch": 13.3625, "grad_norm": 1.3984375, "learning_rate": 4.9961514272231345e-05, "loss": 0.1789, "step": 2138 }, { "epoch": 13.36875, "grad_norm": 1.5625, "learning_rate": 4.996137083968628e-05, "loss": 0.1642, "step": 2139 }, { "epoch": 13.375, "grad_norm": 1.546875, "learning_rate": 4.996122714056525e-05, "loss": 0.2038, "step": 2140 }, { "epoch": 13.38125, "grad_norm": 1.8359375, "learning_rate": 4.9961083174869774e-05, "loss": 0.1975, "step": 2141 }, { "epoch": 13.3875, "grad_norm": 1.953125, "learning_rate": 4.996093894260139e-05, "loss": 0.2495, "step": 2142 }, { "epoch": 13.39375, "grad_norm": 1.9609375, "learning_rate": 4.996079444376165e-05, "loss": 0.2395, "step": 2143 }, { "epoch": 13.4, "grad_norm": 1.4921875, "learning_rate": 4.996064967835208e-05, "loss": 0.1368, "step": 2144 }, { "epoch": 13.40625, "grad_norm": 1.8203125, "learning_rate": 4.996050464637423e-05, "loss": 0.1951, "step": 2145 }, { "epoch": 13.4125, "grad_norm": 1.6171875, "learning_rate": 4.9960359347829667e-05, "loss": 0.1633, "step": 2146 }, { "epoch": 13.41875, "grad_norm": 1.84375, "learning_rate": 4.9960213782719916e-05, "loss": 0.1559, "step": 2147 }, { "epoch": 13.425, "grad_norm": 1.828125, "learning_rate": 4.996006795104655e-05, "loss": 0.1993, "step": 2148 }, { "epoch": 13.43125, "grad_norm": 1.2578125, "learning_rate": 4.995992185281112e-05, "loss": 0.1532, "step": 2149 }, { "epoch": 13.4375, "grad_norm": 1.6171875, "learning_rate": 4.995977548801519e-05, "loss": 0.1264, "step": 2150 }, { "epoch": 13.44375, "grad_norm": 1.6015625, "learning_rate": 4.9959628856660315e-05, "loss": 0.146, "step": 2151 }, { "epoch": 13.45, "grad_norm": 1.6015625, "learning_rate": 4.9959481958748065e-05, "loss": 0.2564, "step": 2152 }, { "epoch": 13.45625, "grad_norm": 1.2578125, "learning_rate": 4.9959334794280014e-05, "loss": 0.1444, "step": 2153 }, { "epoch": 13.4625, "grad_norm": 1.453125, "learning_rate": 4.995918736325773e-05, "loss": 0.1332, "step": 2154 }, { "epoch": 13.46875, "grad_norm": 1.6171875, "learning_rate": 4.995903966568278e-05, "loss": 0.2444, "step": 2155 }, { "epoch": 13.475, "grad_norm": 1.9453125, "learning_rate": 4.995889170155676e-05, "loss": 0.2857, "step": 2156 }, { "epoch": 13.48125, "grad_norm": 1.96875, "learning_rate": 4.995874347088123e-05, "loss": 0.25, "step": 2157 }, { "epoch": 13.4875, "grad_norm": 1.5078125, "learning_rate": 4.995859497365779e-05, "loss": 0.215, "step": 2158 }, { "epoch": 13.49375, "grad_norm": 1.7734375, "learning_rate": 4.995844620988801e-05, "loss": 0.1437, "step": 2159 }, { "epoch": 13.5, "grad_norm": 1.6875, "learning_rate": 4.995829717957349e-05, "loss": 0.2011, "step": 2160 }, { "epoch": 13.50625, "grad_norm": 1.765625, "learning_rate": 4.995814788271582e-05, "loss": 0.1766, "step": 2161 }, { "epoch": 13.5125, "grad_norm": 1.515625, "learning_rate": 4.995799831931659e-05, "loss": 0.1498, "step": 2162 }, { "epoch": 13.51875, "grad_norm": 1.8359375, "learning_rate": 4.9957848489377394e-05, "loss": 0.2568, "step": 2163 }, { "epoch": 13.525, "grad_norm": 1.90625, "learning_rate": 4.995769839289984e-05, "loss": 0.2043, "step": 2164 }, { "epoch": 13.53125, "grad_norm": 1.3046875, "learning_rate": 4.995754802988554e-05, "loss": 0.1483, "step": 2165 }, { "epoch": 13.5375, "grad_norm": 1.203125, "learning_rate": 4.995739740033607e-05, "loss": 0.1248, "step": 2166 }, { "epoch": 13.54375, "grad_norm": 1.9453125, "learning_rate": 4.995724650425307e-05, "loss": 0.184, "step": 2167 }, { "epoch": 13.55, "grad_norm": 1.9921875, "learning_rate": 4.995709534163814e-05, "loss": 0.2004, "step": 2168 }, { "epoch": 13.55625, "grad_norm": 1.421875, "learning_rate": 4.995694391249288e-05, "loss": 0.1491, "step": 2169 }, { "epoch": 13.5625, "grad_norm": 1.3203125, "learning_rate": 4.9956792216818934e-05, "loss": 0.1323, "step": 2170 }, { "epoch": 13.56875, "grad_norm": 1.6015625, "learning_rate": 4.9956640254617906e-05, "loss": 0.1883, "step": 2171 }, { "epoch": 13.575, "grad_norm": 1.9609375, "learning_rate": 4.9956488025891414e-05, "loss": 0.1558, "step": 2172 }, { "epoch": 13.58125, "grad_norm": 1.109375, "learning_rate": 4.9956335530641093e-05, "loss": 0.1071, "step": 2173 }, { "epoch": 13.5875, "grad_norm": 1.234375, "learning_rate": 4.995618276886857e-05, "loss": 0.1253, "step": 2174 }, { "epoch": 13.59375, "grad_norm": 1.5703125, "learning_rate": 4.9956029740575475e-05, "loss": 0.144, "step": 2175 }, { "epoch": 13.6, "grad_norm": 2.140625, "learning_rate": 4.995587644576344e-05, "loss": 0.2117, "step": 2176 }, { "epoch": 13.60625, "grad_norm": 1.46875, "learning_rate": 4.9955722884434114e-05, "loss": 0.1634, "step": 2177 }, { "epoch": 13.6125, "grad_norm": 1.546875, "learning_rate": 4.995556905658912e-05, "loss": 0.1544, "step": 2178 }, { "epoch": 13.61875, "grad_norm": 1.3984375, "learning_rate": 4.995541496223012e-05, "loss": 0.1579, "step": 2179 }, { "epoch": 13.625, "grad_norm": 1.671875, "learning_rate": 4.995526060135874e-05, "loss": 0.1349, "step": 2180 }, { "epoch": 13.63125, "grad_norm": 1.6484375, "learning_rate": 4.9955105973976636e-05, "loss": 0.2492, "step": 2181 }, { "epoch": 13.6375, "grad_norm": 1.578125, "learning_rate": 4.9954951080085465e-05, "loss": 0.1637, "step": 2182 }, { "epoch": 13.64375, "grad_norm": 1.375, "learning_rate": 4.995479591968688e-05, "loss": 0.1321, "step": 2183 }, { "epoch": 13.65, "grad_norm": 1.953125, "learning_rate": 4.9954640492782526e-05, "loss": 0.2465, "step": 2184 }, { "epoch": 13.65625, "grad_norm": 1.5859375, "learning_rate": 4.9954484799374076e-05, "loss": 0.1783, "step": 2185 }, { "epoch": 13.6625, "grad_norm": 1.765625, "learning_rate": 4.9954328839463184e-05, "loss": 0.1866, "step": 2186 }, { "epoch": 13.66875, "grad_norm": 2.09375, "learning_rate": 4.995417261305152e-05, "loss": 0.1881, "step": 2187 }, { "epoch": 13.675, "grad_norm": 1.3671875, "learning_rate": 4.9954016120140755e-05, "loss": 0.1392, "step": 2188 }, { "epoch": 13.68125, "grad_norm": 1.6015625, "learning_rate": 4.9953859360732566e-05, "loss": 0.2186, "step": 2189 }, { "epoch": 13.6875, "grad_norm": 1.9765625, "learning_rate": 4.995370233482861e-05, "loss": 0.2323, "step": 2190 }, { "epoch": 13.69375, "grad_norm": 2.4375, "learning_rate": 4.995354504243057e-05, "loss": 0.2271, "step": 2191 }, { "epoch": 13.7, "grad_norm": 1.515625, "learning_rate": 4.9953387483540135e-05, "loss": 0.1646, "step": 2192 }, { "epoch": 13.70625, "grad_norm": 1.234375, "learning_rate": 4.995322965815898e-05, "loss": 0.1146, "step": 2193 }, { "epoch": 13.7125, "grad_norm": 1.3359375, "learning_rate": 4.995307156628879e-05, "loss": 0.1428, "step": 2194 }, { "epoch": 13.71875, "grad_norm": 1.6875, "learning_rate": 4.995291320793125e-05, "loss": 0.1516, "step": 2195 }, { "epoch": 13.725, "grad_norm": 1.734375, "learning_rate": 4.995275458308806e-05, "loss": 0.244, "step": 2196 }, { "epoch": 13.73125, "grad_norm": 1.7890625, "learning_rate": 4.995259569176092e-05, "loss": 0.2535, "step": 2197 }, { "epoch": 13.7375, "grad_norm": 1.921875, "learning_rate": 4.995243653395151e-05, "loss": 0.2539, "step": 2198 }, { "epoch": 13.74375, "grad_norm": 1.234375, "learning_rate": 4.995227710966153e-05, "loss": 0.135, "step": 2199 }, { "epoch": 13.75, "grad_norm": 1.46875, "learning_rate": 4.9952117418892696e-05, "loss": 0.1489, "step": 2200 }, { "epoch": 13.75625, "grad_norm": 1.984375, "learning_rate": 4.9951957461646706e-05, "loss": 0.2759, "step": 2201 }, { "epoch": 13.7625, "grad_norm": 1.5078125, "learning_rate": 4.995179723792527e-05, "loss": 0.1566, "step": 2202 }, { "epoch": 13.76875, "grad_norm": 1.5078125, "learning_rate": 4.9951636747730094e-05, "loss": 0.1816, "step": 2203 }, { "epoch": 13.775, "grad_norm": 1.8046875, "learning_rate": 4.99514759910629e-05, "loss": 0.1827, "step": 2204 }, { "epoch": 13.78125, "grad_norm": 1.5, "learning_rate": 4.99513149679254e-05, "loss": 0.1628, "step": 2205 }, { "epoch": 13.7875, "grad_norm": 1.6328125, "learning_rate": 4.995115367831932e-05, "loss": 0.1567, "step": 2206 }, { "epoch": 13.79375, "grad_norm": 1.421875, "learning_rate": 4.995099212224637e-05, "loss": 0.1887, "step": 2207 }, { "epoch": 13.8, "grad_norm": 1.8359375, "learning_rate": 4.995083029970828e-05, "loss": 0.3454, "step": 2208 }, { "epoch": 13.80625, "grad_norm": 1.4609375, "learning_rate": 4.995066821070679e-05, "loss": 0.1793, "step": 2209 }, { "epoch": 13.8125, "grad_norm": 1.4453125, "learning_rate": 4.995050585524362e-05, "loss": 0.1708, "step": 2210 }, { "epoch": 13.81875, "grad_norm": 1.4609375, "learning_rate": 4.9950343233320503e-05, "loss": 0.2194, "step": 2211 }, { "epoch": 13.825, "grad_norm": 1.53125, "learning_rate": 4.995018034493918e-05, "loss": 0.2182, "step": 2212 }, { "epoch": 13.83125, "grad_norm": 1.40625, "learning_rate": 4.995001719010138e-05, "loss": 0.1776, "step": 2213 }, { "epoch": 13.8375, "grad_norm": 1.1640625, "learning_rate": 4.994985376880887e-05, "loss": 0.1268, "step": 2214 }, { "epoch": 13.84375, "grad_norm": 1.2421875, "learning_rate": 4.994969008106337e-05, "loss": 0.1539, "step": 2215 }, { "epoch": 13.85, "grad_norm": 1.3984375, "learning_rate": 4.994952612686664e-05, "loss": 0.1455, "step": 2216 }, { "epoch": 13.85625, "grad_norm": 1.609375, "learning_rate": 4.9949361906220425e-05, "loss": 0.2085, "step": 2217 }, { "epoch": 13.8625, "grad_norm": 1.5546875, "learning_rate": 4.994919741912649e-05, "loss": 0.2885, "step": 2218 }, { "epoch": 13.86875, "grad_norm": 1.53125, "learning_rate": 4.994903266558658e-05, "loss": 0.1658, "step": 2219 }, { "epoch": 13.875, "grad_norm": 1.6953125, "learning_rate": 4.9948867645602456e-05, "loss": 0.1885, "step": 2220 }, { "epoch": 13.88125, "grad_norm": 1.359375, "learning_rate": 4.994870235917588e-05, "loss": 0.1577, "step": 2221 }, { "epoch": 13.8875, "grad_norm": 1.5, "learning_rate": 4.994853680630863e-05, "loss": 0.2046, "step": 2222 }, { "epoch": 13.89375, "grad_norm": 2.421875, "learning_rate": 4.994837098700246e-05, "loss": 0.2643, "step": 2223 }, { "epoch": 13.9, "grad_norm": 1.5625, "learning_rate": 4.9948204901259144e-05, "loss": 0.1983, "step": 2224 }, { "epoch": 13.90625, "grad_norm": 1.7109375, "learning_rate": 4.9948038549080456e-05, "loss": 0.1989, "step": 2225 }, { "epoch": 13.9125, "grad_norm": 1.8203125, "learning_rate": 4.994787193046818e-05, "loss": 0.21, "step": 2226 }, { "epoch": 13.91875, "grad_norm": 1.9296875, "learning_rate": 4.9947705045424076e-05, "loss": 0.227, "step": 2227 }, { "epoch": 13.925, "grad_norm": 1.46875, "learning_rate": 4.994753789394995e-05, "loss": 0.1844, "step": 2228 }, { "epoch": 13.93125, "grad_norm": 1.3203125, "learning_rate": 4.994737047604757e-05, "loss": 0.17, "step": 2229 }, { "epoch": 13.9375, "grad_norm": 1.203125, "learning_rate": 4.9947202791718735e-05, "loss": 0.1339, "step": 2230 }, { "epoch": 13.94375, "grad_norm": 2.140625, "learning_rate": 4.9947034840965224e-05, "loss": 0.2609, "step": 2231 }, { "epoch": 13.95, "grad_norm": 1.9140625, "learning_rate": 4.9946866623788844e-05, "loss": 0.1954, "step": 2232 }, { "epoch": 13.95625, "grad_norm": 1.359375, "learning_rate": 4.994669814019138e-05, "loss": 0.1283, "step": 2233 }, { "epoch": 13.9625, "grad_norm": 1.328125, "learning_rate": 4.994652939017465e-05, "loss": 0.1531, "step": 2234 }, { "epoch": 13.96875, "grad_norm": 1.453125, "learning_rate": 4.9946360373740424e-05, "loss": 0.1175, "step": 2235 }, { "epoch": 13.975, "grad_norm": 1.5859375, "learning_rate": 4.994619109089054e-05, "loss": 0.1878, "step": 2236 }, { "epoch": 13.98125, "grad_norm": 1.25, "learning_rate": 4.994602154162678e-05, "loss": 0.1188, "step": 2237 }, { "epoch": 13.9875, "grad_norm": 2.1875, "learning_rate": 4.994585172595097e-05, "loss": 0.2682, "step": 2238 }, { "epoch": 13.99375, "grad_norm": 2.109375, "learning_rate": 4.994568164386492e-05, "loss": 0.2731, "step": 2239 }, { "epoch": 14.0, "grad_norm": 1.4609375, "learning_rate": 4.9945511295370446e-05, "loss": 0.1195, "step": 2240 }, { "epoch": 14.00625, "grad_norm": 1.7734375, "learning_rate": 4.994534068046937e-05, "loss": 0.226, "step": 2241 }, { "epoch": 14.0125, "grad_norm": 1.6640625, "learning_rate": 4.99451697991635e-05, "loss": 0.2187, "step": 2242 }, { "epoch": 14.01875, "grad_norm": 1.75, "learning_rate": 4.994499865145469e-05, "loss": 0.1993, "step": 2243 }, { "epoch": 14.025, "grad_norm": 1.578125, "learning_rate": 4.994482723734474e-05, "loss": 0.1399, "step": 2244 }, { "epoch": 14.03125, "grad_norm": 1.75, "learning_rate": 4.994465555683549e-05, "loss": 0.2241, "step": 2245 }, { "epoch": 14.0375, "grad_norm": 1.8515625, "learning_rate": 4.994448360992877e-05, "loss": 0.2216, "step": 2246 }, { "epoch": 14.04375, "grad_norm": 1.734375, "learning_rate": 4.994431139662643e-05, "loss": 0.1776, "step": 2247 }, { "epoch": 14.05, "grad_norm": 1.640625, "learning_rate": 4.9944138916930295e-05, "loss": 0.1896, "step": 2248 }, { "epoch": 14.05625, "grad_norm": 1.7109375, "learning_rate": 4.994396617084222e-05, "loss": 0.2608, "step": 2249 }, { "epoch": 14.0625, "grad_norm": 1.5078125, "learning_rate": 4.9943793158364026e-05, "loss": 0.1677, "step": 2250 }, { "epoch": 14.06875, "grad_norm": 1.4140625, "learning_rate": 4.9943619879497585e-05, "loss": 0.1133, "step": 2251 }, { "epoch": 14.075, "grad_norm": 2.28125, "learning_rate": 4.9943446334244737e-05, "loss": 0.2989, "step": 2252 }, { "epoch": 14.08125, "grad_norm": 2.125, "learning_rate": 4.9943272522607345e-05, "loss": 0.3468, "step": 2253 }, { "epoch": 14.0875, "grad_norm": 1.734375, "learning_rate": 4.9943098444587246e-05, "loss": 0.1895, "step": 2254 }, { "epoch": 14.09375, "grad_norm": 1.9140625, "learning_rate": 4.9942924100186317e-05, "loss": 0.1954, "step": 2255 }, { "epoch": 14.1, "grad_norm": 2.1875, "learning_rate": 4.994274948940641e-05, "loss": 0.2746, "step": 2256 }, { "epoch": 14.10625, "grad_norm": 0.96875, "learning_rate": 4.9942574612249394e-05, "loss": 0.1198, "step": 2257 }, { "epoch": 14.1125, "grad_norm": 1.796875, "learning_rate": 4.9942399468717135e-05, "loss": 0.2113, "step": 2258 }, { "epoch": 14.11875, "grad_norm": 1.75, "learning_rate": 4.994222405881151e-05, "loss": 0.2192, "step": 2259 }, { "epoch": 14.125, "grad_norm": 1.125, "learning_rate": 4.994204838253437e-05, "loss": 0.1638, "step": 2260 }, { "epoch": 14.13125, "grad_norm": 1.546875, "learning_rate": 4.994187243988763e-05, "loss": 0.1499, "step": 2261 }, { "epoch": 14.1375, "grad_norm": 1.484375, "learning_rate": 4.994169623087313e-05, "loss": 0.1873, "step": 2262 }, { "epoch": 14.14375, "grad_norm": 1.6953125, "learning_rate": 4.994151975549277e-05, "loss": 0.2349, "step": 2263 }, { "epoch": 14.15, "grad_norm": 1.453125, "learning_rate": 4.994134301374844e-05, "loss": 0.1963, "step": 2264 }, { "epoch": 14.15625, "grad_norm": 1.3359375, "learning_rate": 4.9941166005642014e-05, "loss": 0.1229, "step": 2265 }, { "epoch": 14.1625, "grad_norm": 1.3984375, "learning_rate": 4.99409887311754e-05, "loss": 0.1425, "step": 2266 }, { "epoch": 14.16875, "grad_norm": 1.4609375, "learning_rate": 4.9940811190350466e-05, "loss": 0.2068, "step": 2267 }, { "epoch": 14.175, "grad_norm": 2.09375, "learning_rate": 4.9940633383169134e-05, "loss": 0.2859, "step": 2268 }, { "epoch": 14.18125, "grad_norm": 2.359375, "learning_rate": 4.9940455309633284e-05, "loss": 0.2061, "step": 2269 }, { "epoch": 14.1875, "grad_norm": 1.59375, "learning_rate": 4.994027696974483e-05, "loss": 0.2225, "step": 2270 }, { "epoch": 14.19375, "grad_norm": 1.609375, "learning_rate": 4.994009836350566e-05, "loss": 0.2153, "step": 2271 }, { "epoch": 14.2, "grad_norm": 2.109375, "learning_rate": 4.99399194909177e-05, "loss": 0.3176, "step": 2272 }, { "epoch": 14.20625, "grad_norm": 1.234375, "learning_rate": 4.9939740351982856e-05, "loss": 0.133, "step": 2273 }, { "epoch": 14.2125, "grad_norm": 1.46875, "learning_rate": 4.993956094670303e-05, "loss": 0.179, "step": 2274 }, { "epoch": 14.21875, "grad_norm": 1.4453125, "learning_rate": 4.993938127508016e-05, "loss": 0.1612, "step": 2275 }, { "epoch": 14.225, "grad_norm": 1.1953125, "learning_rate": 4.9939201337116146e-05, "loss": 0.121, "step": 2276 }, { "epoch": 14.23125, "grad_norm": 1.484375, "learning_rate": 4.993902113281291e-05, "loss": 0.1431, "step": 2277 }, { "epoch": 14.2375, "grad_norm": 1.5, "learning_rate": 4.9938840662172384e-05, "loss": 0.1865, "step": 2278 }, { "epoch": 14.24375, "grad_norm": 1.59375, "learning_rate": 4.993865992519649e-05, "loss": 0.2338, "step": 2279 }, { "epoch": 14.25, "grad_norm": 1.828125, "learning_rate": 4.993847892188717e-05, "loss": 0.1823, "step": 2280 }, { "epoch": 14.25625, "grad_norm": 2.8125, "learning_rate": 4.993829765224634e-05, "loss": 0.2657, "step": 2281 }, { "epoch": 14.2625, "grad_norm": 1.328125, "learning_rate": 4.993811611627593e-05, "loss": 0.1874, "step": 2282 }, { "epoch": 14.26875, "grad_norm": 1.6953125, "learning_rate": 4.993793431397792e-05, "loss": 0.1569, "step": 2283 }, { "epoch": 14.275, "grad_norm": 1.4375, "learning_rate": 4.993775224535421e-05, "loss": 0.1945, "step": 2284 }, { "epoch": 14.28125, "grad_norm": 2.21875, "learning_rate": 4.9937569910406756e-05, "loss": 0.2776, "step": 2285 }, { "epoch": 14.2875, "grad_norm": 1.46875, "learning_rate": 4.993738730913751e-05, "loss": 0.1315, "step": 2286 }, { "epoch": 14.29375, "grad_norm": 1.578125, "learning_rate": 4.993720444154842e-05, "loss": 0.1983, "step": 2287 }, { "epoch": 14.3, "grad_norm": 1.8359375, "learning_rate": 4.993702130764144e-05, "loss": 0.1865, "step": 2288 }, { "epoch": 14.30625, "grad_norm": 1.4296875, "learning_rate": 4.993683790741852e-05, "loss": 0.1464, "step": 2289 }, { "epoch": 14.3125, "grad_norm": 1.5, "learning_rate": 4.993665424088163e-05, "loss": 0.24, "step": 2290 }, { "epoch": 14.31875, "grad_norm": 1.375, "learning_rate": 4.9936470308032716e-05, "loss": 0.1671, "step": 2291 }, { "epoch": 14.325, "grad_norm": 1.78125, "learning_rate": 4.993628610887376e-05, "loss": 0.1857, "step": 2292 }, { "epoch": 14.33125, "grad_norm": 1.8046875, "learning_rate": 4.9936101643406715e-05, "loss": 0.1967, "step": 2293 }, { "epoch": 14.3375, "grad_norm": 1.984375, "learning_rate": 4.993591691163355e-05, "loss": 0.253, "step": 2294 }, { "epoch": 14.34375, "grad_norm": 1.640625, "learning_rate": 4.9935731913556254e-05, "loss": 0.1898, "step": 2295 }, { "epoch": 14.35, "grad_norm": 1.859375, "learning_rate": 4.993554664917679e-05, "loss": 0.2008, "step": 2296 }, { "epoch": 14.35625, "grad_norm": 2.453125, "learning_rate": 4.9935361118497134e-05, "loss": 0.2729, "step": 2297 }, { "epoch": 14.3625, "grad_norm": 1.7578125, "learning_rate": 4.993517532151928e-05, "loss": 0.1688, "step": 2298 }, { "epoch": 14.36875, "grad_norm": 1.46875, "learning_rate": 4.99349892582452e-05, "loss": 0.1377, "step": 2299 }, { "epoch": 14.375, "grad_norm": 1.3125, "learning_rate": 4.993480292867689e-05, "loss": 0.1572, "step": 2300 }, { "epoch": 14.38125, "grad_norm": 1.28125, "learning_rate": 4.993461633281633e-05, "loss": 0.1428, "step": 2301 }, { "epoch": 14.3875, "grad_norm": 1.859375, "learning_rate": 4.993442947066552e-05, "loss": 0.1983, "step": 2302 }, { "epoch": 14.39375, "grad_norm": 1.6171875, "learning_rate": 4.993424234222646e-05, "loss": 0.2149, "step": 2303 }, { "epoch": 14.4, "grad_norm": 1.5859375, "learning_rate": 4.993405494750113e-05, "loss": 0.1909, "step": 2304 }, { "epoch": 14.40625, "grad_norm": 1.6640625, "learning_rate": 4.993386728649156e-05, "loss": 0.2437, "step": 2305 }, { "epoch": 14.4125, "grad_norm": 1.125, "learning_rate": 4.9933679359199726e-05, "loss": 0.1146, "step": 2306 }, { "epoch": 14.41875, "grad_norm": 1.171875, "learning_rate": 4.993349116562766e-05, "loss": 0.1391, "step": 2307 }, { "epoch": 14.425, "grad_norm": 1.140625, "learning_rate": 4.993330270577734e-05, "loss": 0.1133, "step": 2308 }, { "epoch": 14.43125, "grad_norm": 1.2421875, "learning_rate": 4.99331139796508e-05, "loss": 0.1725, "step": 2309 }, { "epoch": 14.4375, "grad_norm": 1.5234375, "learning_rate": 4.993292498725006e-05, "loss": 0.1884, "step": 2310 }, { "epoch": 14.44375, "grad_norm": 1.4375, "learning_rate": 4.993273572857714e-05, "loss": 0.1442, "step": 2311 }, { "epoch": 14.45, "grad_norm": 1.4140625, "learning_rate": 4.993254620363404e-05, "loss": 0.1392, "step": 2312 }, { "epoch": 14.45625, "grad_norm": 1.6796875, "learning_rate": 4.99323564124228e-05, "loss": 0.1941, "step": 2313 }, { "epoch": 14.4625, "grad_norm": 1.15625, "learning_rate": 4.993216635494544e-05, "loss": 0.1335, "step": 2314 }, { "epoch": 14.46875, "grad_norm": 2.390625, "learning_rate": 4.9931976031204e-05, "loss": 0.3507, "step": 2315 }, { "epoch": 14.475, "grad_norm": 2.1875, "learning_rate": 4.9931785441200496e-05, "loss": 0.2489, "step": 2316 }, { "epoch": 14.48125, "grad_norm": 2.03125, "learning_rate": 4.993159458493698e-05, "loss": 0.1758, "step": 2317 }, { "epoch": 14.4875, "grad_norm": 1.6875, "learning_rate": 4.993140346241548e-05, "loss": 0.1999, "step": 2318 }, { "epoch": 14.49375, "grad_norm": 2.03125, "learning_rate": 4.993121207363804e-05, "loss": 0.1922, "step": 2319 }, { "epoch": 14.5, "grad_norm": 2.109375, "learning_rate": 4.993102041860671e-05, "loss": 0.2117, "step": 2320 }, { "epoch": 14.50625, "grad_norm": 1.3984375, "learning_rate": 4.9930828497323526e-05, "loss": 0.165, "step": 2321 }, { "epoch": 14.5125, "grad_norm": 1.6875, "learning_rate": 4.993063630979055e-05, "loss": 0.1744, "step": 2322 }, { "epoch": 14.51875, "grad_norm": 2.46875, "learning_rate": 4.9930443856009826e-05, "loss": 0.2439, "step": 2323 }, { "epoch": 14.525, "grad_norm": 1.6015625, "learning_rate": 4.99302511359834e-05, "loss": 0.1421, "step": 2324 }, { "epoch": 14.53125, "grad_norm": 1.6171875, "learning_rate": 4.993005814971335e-05, "loss": 0.167, "step": 2325 }, { "epoch": 14.5375, "grad_norm": 2.421875, "learning_rate": 4.9929864897201725e-05, "loss": 0.2531, "step": 2326 }, { "epoch": 14.54375, "grad_norm": 2.53125, "learning_rate": 4.992967137845059e-05, "loss": 0.2347, "step": 2327 }, { "epoch": 14.55, "grad_norm": 2.09375, "learning_rate": 4.9929477593462014e-05, "loss": 0.2536, "step": 2328 }, { "epoch": 14.55625, "grad_norm": 1.6015625, "learning_rate": 4.992928354223807e-05, "loss": 0.1639, "step": 2329 }, { "epoch": 14.5625, "grad_norm": 2.203125, "learning_rate": 4.9929089224780826e-05, "loss": 0.1953, "step": 2330 }, { "epoch": 14.56875, "grad_norm": 1.859375, "learning_rate": 4.992889464109235e-05, "loss": 0.1788, "step": 2331 }, { "epoch": 14.575, "grad_norm": 1.2578125, "learning_rate": 4.992869979117473e-05, "loss": 0.1469, "step": 2332 }, { "epoch": 14.58125, "grad_norm": 1.515625, "learning_rate": 4.9928504675030045e-05, "loss": 0.1716, "step": 2333 }, { "epoch": 14.5875, "grad_norm": 1.171875, "learning_rate": 4.992830929266038e-05, "loss": 0.1026, "step": 2334 }, { "epoch": 14.59375, "grad_norm": 1.7578125, "learning_rate": 4.9928113644067817e-05, "loss": 0.2514, "step": 2335 }, { "epoch": 14.6, "grad_norm": 1.5078125, "learning_rate": 4.992791772925445e-05, "loss": 0.1384, "step": 2336 }, { "epoch": 14.60625, "grad_norm": 1.453125, "learning_rate": 4.9927721548222374e-05, "loss": 0.1984, "step": 2337 }, { "epoch": 14.6125, "grad_norm": 1.4921875, "learning_rate": 4.9927525100973674e-05, "loss": 0.1807, "step": 2338 }, { "epoch": 14.61875, "grad_norm": 2.0, "learning_rate": 4.9927328387510454e-05, "loss": 0.2114, "step": 2339 }, { "epoch": 14.625, "grad_norm": 1.8046875, "learning_rate": 4.9927131407834814e-05, "loss": 0.1616, "step": 2340 }, { "epoch": 14.63125, "grad_norm": 1.4140625, "learning_rate": 4.9926934161948855e-05, "loss": 0.1584, "step": 2341 }, { "epoch": 14.6375, "grad_norm": 1.3515625, "learning_rate": 4.99267366498547e-05, "loss": 0.208, "step": 2342 }, { "epoch": 14.64375, "grad_norm": 1.6328125, "learning_rate": 4.992653887155443e-05, "loss": 0.2473, "step": 2343 }, { "epoch": 14.65, "grad_norm": 1.3671875, "learning_rate": 4.992634082705018e-05, "loss": 0.123, "step": 2344 }, { "epoch": 14.65625, "grad_norm": 1.5, "learning_rate": 4.992614251634405e-05, "loss": 0.1869, "step": 2345 }, { "epoch": 14.6625, "grad_norm": 1.0546875, "learning_rate": 4.992594393943817e-05, "loss": 0.1136, "step": 2346 }, { "epoch": 14.66875, "grad_norm": 1.4609375, "learning_rate": 4.992574509633466e-05, "loss": 0.229, "step": 2347 }, { "epoch": 14.675, "grad_norm": 1.140625, "learning_rate": 4.992554598703563e-05, "loss": 0.1301, "step": 2348 }, { "epoch": 14.68125, "grad_norm": 1.25, "learning_rate": 4.992534661154322e-05, "loss": 0.1536, "step": 2349 }, { "epoch": 14.6875, "grad_norm": 1.6953125, "learning_rate": 4.992514696985956e-05, "loss": 0.2238, "step": 2350 }, { "epoch": 14.69375, "grad_norm": 1.2578125, "learning_rate": 4.992494706198677e-05, "loss": 0.1371, "step": 2351 }, { "epoch": 14.7, "grad_norm": 1.453125, "learning_rate": 4.9924746887926986e-05, "loss": 0.1792, "step": 2352 }, { "epoch": 14.70625, "grad_norm": 1.328125, "learning_rate": 4.992454644768236e-05, "loss": 0.1639, "step": 2353 }, { "epoch": 14.7125, "grad_norm": 1.8515625, "learning_rate": 4.992434574125503e-05, "loss": 0.1631, "step": 2354 }, { "epoch": 14.71875, "grad_norm": 1.8515625, "learning_rate": 4.992414476864712e-05, "loss": 0.2593, "step": 2355 }, { "epoch": 14.725, "grad_norm": 1.5625, "learning_rate": 4.99239435298608e-05, "loss": 0.1281, "step": 2356 }, { "epoch": 14.73125, "grad_norm": 1.3671875, "learning_rate": 4.992374202489821e-05, "loss": 0.1583, "step": 2357 }, { "epoch": 14.7375, "grad_norm": 1.8125, "learning_rate": 4.992354025376149e-05, "loss": 0.2012, "step": 2358 }, { "epoch": 14.74375, "grad_norm": 1.6875, "learning_rate": 4.992333821645281e-05, "loss": 0.1618, "step": 2359 }, { "epoch": 14.75, "grad_norm": 2.0625, "learning_rate": 4.9923135912974326e-05, "loss": 0.1805, "step": 2360 }, { "epoch": 14.75625, "grad_norm": 1.453125, "learning_rate": 4.99229333433282e-05, "loss": 0.1829, "step": 2361 }, { "epoch": 14.7625, "grad_norm": 1.078125, "learning_rate": 4.992273050751659e-05, "loss": 0.1095, "step": 2362 }, { "epoch": 14.76875, "grad_norm": 2.5625, "learning_rate": 4.9922527405541664e-05, "loss": 0.4237, "step": 2363 }, { "epoch": 14.775, "grad_norm": 1.6953125, "learning_rate": 4.9922324037405585e-05, "loss": 0.2027, "step": 2364 }, { "epoch": 14.78125, "grad_norm": 1.359375, "learning_rate": 4.992212040311053e-05, "loss": 0.1262, "step": 2365 }, { "epoch": 14.7875, "grad_norm": 2.0625, "learning_rate": 4.992191650265868e-05, "loss": 0.2342, "step": 2366 }, { "epoch": 14.79375, "grad_norm": 1.4765625, "learning_rate": 4.9921712336052203e-05, "loss": 0.2344, "step": 2367 }, { "epoch": 14.8, "grad_norm": 1.4140625, "learning_rate": 4.992150790329329e-05, "loss": 0.1338, "step": 2368 }, { "epoch": 14.80625, "grad_norm": 1.7109375, "learning_rate": 4.9921303204384104e-05, "loss": 0.1669, "step": 2369 }, { "epoch": 14.8125, "grad_norm": 1.421875, "learning_rate": 4.9921098239326855e-05, "loss": 0.1414, "step": 2370 }, { "epoch": 14.81875, "grad_norm": 1.34375, "learning_rate": 4.992089300812372e-05, "loss": 0.1395, "step": 2371 }, { "epoch": 14.825, "grad_norm": 1.515625, "learning_rate": 4.9920687510776895e-05, "loss": 0.1844, "step": 2372 }, { "epoch": 14.83125, "grad_norm": 2.0, "learning_rate": 4.992048174728857e-05, "loss": 0.2028, "step": 2373 }, { "epoch": 14.8375, "grad_norm": 1.5546875, "learning_rate": 4.992027571766094e-05, "loss": 0.1685, "step": 2374 }, { "epoch": 14.84375, "grad_norm": 1.296875, "learning_rate": 4.9920069421896207e-05, "loss": 0.1214, "step": 2375 }, { "epoch": 14.85, "grad_norm": 1.5859375, "learning_rate": 4.991986285999659e-05, "loss": 0.2053, "step": 2376 }, { "epoch": 14.85625, "grad_norm": 1.3515625, "learning_rate": 4.991965603196427e-05, "loss": 0.1803, "step": 2377 }, { "epoch": 14.8625, "grad_norm": 1.4765625, "learning_rate": 4.991944893780147e-05, "loss": 0.2006, "step": 2378 }, { "epoch": 14.86875, "grad_norm": 1.59375, "learning_rate": 4.99192415775104e-05, "loss": 0.208, "step": 2379 }, { "epoch": 14.875, "grad_norm": 1.9375, "learning_rate": 4.9919033951093274e-05, "loss": 0.1993, "step": 2380 }, { "epoch": 14.88125, "grad_norm": 1.6953125, "learning_rate": 4.9918826058552316e-05, "loss": 0.2118, "step": 2381 }, { "epoch": 14.8875, "grad_norm": 1.5078125, "learning_rate": 4.991861789988973e-05, "loss": 0.1836, "step": 2382 }, { "epoch": 14.89375, "grad_norm": 1.6640625, "learning_rate": 4.991840947510775e-05, "loss": 0.1758, "step": 2383 }, { "epoch": 14.9, "grad_norm": 2.0, "learning_rate": 4.99182007842086e-05, "loss": 0.1839, "step": 2384 }, { "epoch": 14.90625, "grad_norm": 1.5703125, "learning_rate": 4.991799182719451e-05, "loss": 0.2476, "step": 2385 }, { "epoch": 14.9125, "grad_norm": 1.65625, "learning_rate": 4.991778260406771e-05, "loss": 0.345, "step": 2386 }, { "epoch": 14.91875, "grad_norm": 1.546875, "learning_rate": 4.991757311483044e-05, "loss": 0.1564, "step": 2387 }, { "epoch": 14.925, "grad_norm": 1.6015625, "learning_rate": 4.9917363359484926e-05, "loss": 0.1661, "step": 2388 }, { "epoch": 14.93125, "grad_norm": 1.15625, "learning_rate": 4.9917153338033416e-05, "loss": 0.1593, "step": 2389 }, { "epoch": 14.9375, "grad_norm": 1.5234375, "learning_rate": 4.991694305047815e-05, "loss": 0.2307, "step": 2390 }, { "epoch": 14.94375, "grad_norm": 1.4453125, "learning_rate": 4.991673249682138e-05, "loss": 0.1515, "step": 2391 }, { "epoch": 14.95, "grad_norm": 1.5078125, "learning_rate": 4.9916521677065344e-05, "loss": 0.1531, "step": 2392 }, { "epoch": 14.95625, "grad_norm": 1.5078125, "learning_rate": 4.99163105912123e-05, "loss": 0.1527, "step": 2393 }, { "epoch": 14.9625, "grad_norm": 1.484375, "learning_rate": 4.9916099239264506e-05, "loss": 0.1728, "step": 2394 }, { "epoch": 14.96875, "grad_norm": 1.65625, "learning_rate": 4.991588762122421e-05, "loss": 0.1948, "step": 2395 }, { "epoch": 14.975, "grad_norm": 1.6015625, "learning_rate": 4.9915675737093673e-05, "loss": 0.2247, "step": 2396 }, { "epoch": 14.98125, "grad_norm": 1.8828125, "learning_rate": 4.9915463586875165e-05, "loss": 0.2017, "step": 2397 }, { "epoch": 14.9875, "grad_norm": 1.546875, "learning_rate": 4.9915251170570954e-05, "loss": 0.1961, "step": 2398 }, { "epoch": 14.99375, "grad_norm": 1.3203125, "learning_rate": 4.9915038488183295e-05, "loss": 0.1333, "step": 2399 }, { "epoch": 15.0, "grad_norm": 1.3046875, "learning_rate": 4.991482553971447e-05, "loss": 0.1222, "step": 2400 }, { "epoch": 15.00625, "grad_norm": 1.328125, "learning_rate": 4.991461232516675e-05, "loss": 0.1141, "step": 2401 }, { "epoch": 15.0125, "grad_norm": 1.671875, "learning_rate": 4.991439884454241e-05, "loss": 0.221, "step": 2402 }, { "epoch": 15.01875, "grad_norm": 1.578125, "learning_rate": 4.991418509784373e-05, "loss": 0.2072, "step": 2403 }, { "epoch": 15.025, "grad_norm": 1.4921875, "learning_rate": 4.9913971085073e-05, "loss": 0.1426, "step": 2404 }, { "epoch": 15.03125, "grad_norm": 1.7578125, "learning_rate": 4.99137568062325e-05, "loss": 0.2175, "step": 2405 }, { "epoch": 15.0375, "grad_norm": 1.46875, "learning_rate": 4.9913542261324515e-05, "loss": 0.1711, "step": 2406 }, { "epoch": 15.04375, "grad_norm": 1.2890625, "learning_rate": 4.991332745035134e-05, "loss": 0.1697, "step": 2407 }, { "epoch": 15.05, "grad_norm": 1.5546875, "learning_rate": 4.991311237331527e-05, "loss": 0.2333, "step": 2408 }, { "epoch": 15.05625, "grad_norm": 1.78125, "learning_rate": 4.9912897030218605e-05, "loss": 0.2144, "step": 2409 }, { "epoch": 15.0625, "grad_norm": 1.671875, "learning_rate": 4.991268142106363e-05, "loss": 0.2372, "step": 2410 }, { "epoch": 15.06875, "grad_norm": 1.3828125, "learning_rate": 4.9912465545852674e-05, "loss": 0.1801, "step": 2411 }, { "epoch": 15.075, "grad_norm": 1.2890625, "learning_rate": 4.991224940458801e-05, "loss": 0.1867, "step": 2412 }, { "epoch": 15.08125, "grad_norm": 1.359375, "learning_rate": 4.9912032997271974e-05, "loss": 0.2262, "step": 2413 }, { "epoch": 15.0875, "grad_norm": 1.2578125, "learning_rate": 4.9911816323906865e-05, "loss": 0.1502, "step": 2414 }, { "epoch": 15.09375, "grad_norm": 1.6875, "learning_rate": 4.9911599384494996e-05, "loss": 0.22, "step": 2415 }, { "epoch": 15.1, "grad_norm": 1.0859375, "learning_rate": 4.991138217903868e-05, "loss": 0.1439, "step": 2416 }, { "epoch": 15.10625, "grad_norm": 1.3125, "learning_rate": 4.991116470754025e-05, "loss": 0.1662, "step": 2417 }, { "epoch": 15.1125, "grad_norm": 1.640625, "learning_rate": 4.991094697000202e-05, "loss": 0.2399, "step": 2418 }, { "epoch": 15.11875, "grad_norm": 1.4609375, "learning_rate": 4.991072896642631e-05, "loss": 0.1347, "step": 2419 }, { "epoch": 15.125, "grad_norm": 1.7109375, "learning_rate": 4.9910510696815466e-05, "loss": 0.173, "step": 2420 }, { "epoch": 15.13125, "grad_norm": 1.9609375, "learning_rate": 4.99102921611718e-05, "loss": 0.2299, "step": 2421 }, { "epoch": 15.1375, "grad_norm": 1.15625, "learning_rate": 4.991007335949766e-05, "loss": 0.1268, "step": 2422 }, { "epoch": 15.14375, "grad_norm": 1.328125, "learning_rate": 4.9909854291795365e-05, "loss": 0.1432, "step": 2423 }, { "epoch": 15.15, "grad_norm": 1.375, "learning_rate": 4.990963495806728e-05, "loss": 0.1485, "step": 2424 }, { "epoch": 15.15625, "grad_norm": 1.6484375, "learning_rate": 4.990941535831572e-05, "loss": 0.2271, "step": 2425 }, { "epoch": 15.1625, "grad_norm": 1.4921875, "learning_rate": 4.990919549254305e-05, "loss": 0.1533, "step": 2426 }, { "epoch": 15.16875, "grad_norm": 1.0859375, "learning_rate": 4.990897536075161e-05, "loss": 0.1643, "step": 2427 }, { "epoch": 15.175, "grad_norm": 1.6015625, "learning_rate": 4.990875496294375e-05, "loss": 0.1477, "step": 2428 }, { "epoch": 15.18125, "grad_norm": 1.5078125, "learning_rate": 4.990853429912183e-05, "loss": 0.1298, "step": 2429 }, { "epoch": 15.1875, "grad_norm": 1.2109375, "learning_rate": 4.99083133692882e-05, "loss": 0.1223, "step": 2430 }, { "epoch": 15.19375, "grad_norm": 1.3359375, "learning_rate": 4.990809217344522e-05, "loss": 0.1969, "step": 2431 }, { "epoch": 15.2, "grad_norm": 1.4609375, "learning_rate": 4.990787071159526e-05, "loss": 0.1327, "step": 2432 }, { "epoch": 15.20625, "grad_norm": 1.6484375, "learning_rate": 4.990764898374067e-05, "loss": 0.2243, "step": 2433 }, { "epoch": 15.2125, "grad_norm": 1.3359375, "learning_rate": 4.9907426989883834e-05, "loss": 0.1672, "step": 2434 }, { "epoch": 15.21875, "grad_norm": 1.2109375, "learning_rate": 4.9907204730027116e-05, "loss": 0.1296, "step": 2435 }, { "epoch": 15.225, "grad_norm": 1.2890625, "learning_rate": 4.990698220417289e-05, "loss": 0.1425, "step": 2436 }, { "epoch": 15.23125, "grad_norm": 1.7109375, "learning_rate": 4.990675941232353e-05, "loss": 0.1896, "step": 2437 }, { "epoch": 15.2375, "grad_norm": 1.6640625, "learning_rate": 4.9906536354481424e-05, "loss": 0.1631, "step": 2438 }, { "epoch": 15.24375, "grad_norm": 1.515625, "learning_rate": 4.990631303064894e-05, "loss": 0.132, "step": 2439 }, { "epoch": 15.25, "grad_norm": 2.1875, "learning_rate": 4.9906089440828475e-05, "loss": 0.272, "step": 2440 }, { "epoch": 15.25625, "grad_norm": 1.4140625, "learning_rate": 4.990586558502242e-05, "loss": 0.1655, "step": 2441 }, { "epoch": 15.2625, "grad_norm": 1.8828125, "learning_rate": 4.990564146323314e-05, "loss": 0.2161, "step": 2442 }, { "epoch": 15.26875, "grad_norm": 1.421875, "learning_rate": 4.990541707546306e-05, "loss": 0.1361, "step": 2443 }, { "epoch": 15.275, "grad_norm": 1.9296875, "learning_rate": 4.990519242171456e-05, "loss": 0.2479, "step": 2444 }, { "epoch": 15.28125, "grad_norm": 1.6171875, "learning_rate": 4.990496750199004e-05, "loss": 0.2437, "step": 2445 }, { "epoch": 15.2875, "grad_norm": 1.7890625, "learning_rate": 4.9904742316291906e-05, "loss": 0.2431, "step": 2446 }, { "epoch": 15.29375, "grad_norm": 1.09375, "learning_rate": 4.990451686462256e-05, "loss": 0.0978, "step": 2447 }, { "epoch": 15.3, "grad_norm": 1.671875, "learning_rate": 4.990429114698441e-05, "loss": 0.1792, "step": 2448 }, { "epoch": 15.30625, "grad_norm": 1.1796875, "learning_rate": 4.990406516337987e-05, "loss": 0.1087, "step": 2449 }, { "epoch": 15.3125, "grad_norm": 1.5859375, "learning_rate": 4.990383891381135e-05, "loss": 0.1555, "step": 2450 }, { "epoch": 15.31875, "grad_norm": 1.3984375, "learning_rate": 4.990361239828127e-05, "loss": 0.1447, "step": 2451 }, { "epoch": 15.325, "grad_norm": 1.4921875, "learning_rate": 4.9903385616792045e-05, "loss": 0.1703, "step": 2452 }, { "epoch": 15.33125, "grad_norm": 1.953125, "learning_rate": 4.9903158569346096e-05, "loss": 0.2291, "step": 2453 }, { "epoch": 15.3375, "grad_norm": 1.8359375, "learning_rate": 4.990293125594585e-05, "loss": 0.1797, "step": 2454 }, { "epoch": 15.34375, "grad_norm": 1.75, "learning_rate": 4.990270367659373e-05, "loss": 0.1689, "step": 2455 }, { "epoch": 15.35, "grad_norm": 1.546875, "learning_rate": 4.9902475831292185e-05, "loss": 0.2288, "step": 2456 }, { "epoch": 15.35625, "grad_norm": 1.2734375, "learning_rate": 4.9902247720043626e-05, "loss": 0.1702, "step": 2457 }, { "epoch": 15.3625, "grad_norm": 2.390625, "learning_rate": 4.990201934285049e-05, "loss": 0.3197, "step": 2458 }, { "epoch": 15.36875, "grad_norm": 1.671875, "learning_rate": 4.990179069971523e-05, "loss": 0.1471, "step": 2459 }, { "epoch": 15.375, "grad_norm": 1.515625, "learning_rate": 4.990156179064028e-05, "loss": 0.153, "step": 2460 }, { "epoch": 15.38125, "grad_norm": 1.5, "learning_rate": 4.99013326156281e-05, "loss": 0.2018, "step": 2461 }, { "epoch": 15.3875, "grad_norm": 1.7890625, "learning_rate": 4.99011031746811e-05, "loss": 0.1767, "step": 2462 }, { "epoch": 15.39375, "grad_norm": 1.375, "learning_rate": 4.990087346780177e-05, "loss": 0.1385, "step": 2463 }, { "epoch": 15.4, "grad_norm": 1.09375, "learning_rate": 4.990064349499254e-05, "loss": 0.1104, "step": 2464 }, { "epoch": 15.40625, "grad_norm": 1.4140625, "learning_rate": 4.9900413256255876e-05, "loss": 0.217, "step": 2465 }, { "epoch": 15.4125, "grad_norm": 1.1171875, "learning_rate": 4.9900182751594226e-05, "loss": 0.1086, "step": 2466 }, { "epoch": 15.41875, "grad_norm": 1.359375, "learning_rate": 4.9899951981010065e-05, "loss": 0.1429, "step": 2467 }, { "epoch": 15.425, "grad_norm": 1.109375, "learning_rate": 4.9899720944505854e-05, "loss": 0.1634, "step": 2468 }, { "epoch": 15.43125, "grad_norm": 1.3046875, "learning_rate": 4.989948964208405e-05, "loss": 0.1402, "step": 2469 }, { "epoch": 15.4375, "grad_norm": 1.9375, "learning_rate": 4.989925807374714e-05, "loss": 0.1702, "step": 2470 }, { "epoch": 15.44375, "grad_norm": 1.3828125, "learning_rate": 4.989902623949758e-05, "loss": 0.1261, "step": 2471 }, { "epoch": 15.45, "grad_norm": 1.796875, "learning_rate": 4.989879413933787e-05, "loss": 0.3144, "step": 2472 }, { "epoch": 15.45625, "grad_norm": 1.5703125, "learning_rate": 4.9898561773270456e-05, "loss": 0.2095, "step": 2473 }, { "epoch": 15.4625, "grad_norm": 1.453125, "learning_rate": 4.9898329141297844e-05, "loss": 0.1579, "step": 2474 }, { "epoch": 15.46875, "grad_norm": 1.34375, "learning_rate": 4.989809624342251e-05, "loss": 0.1316, "step": 2475 }, { "epoch": 15.475, "grad_norm": 1.8125, "learning_rate": 4.989786307964694e-05, "loss": 0.2058, "step": 2476 }, { "epoch": 15.48125, "grad_norm": 1.8203125, "learning_rate": 4.989762964997363e-05, "loss": 0.163, "step": 2477 }, { "epoch": 15.4875, "grad_norm": 1.9296875, "learning_rate": 4.989739595440507e-05, "loss": 0.2017, "step": 2478 }, { "epoch": 15.49375, "grad_norm": 1.578125, "learning_rate": 4.989716199294375e-05, "loss": 0.1865, "step": 2479 }, { "epoch": 15.5, "grad_norm": 1.265625, "learning_rate": 4.989692776559218e-05, "loss": 0.1482, "step": 2480 }, { "epoch": 15.50625, "grad_norm": 1.7265625, "learning_rate": 4.9896693272352846e-05, "loss": 0.1959, "step": 2481 }, { "epoch": 15.5125, "grad_norm": 1.625, "learning_rate": 4.989645851322827e-05, "loss": 0.1445, "step": 2482 }, { "epoch": 15.51875, "grad_norm": 1.6953125, "learning_rate": 4.989622348822095e-05, "loss": 0.2568, "step": 2483 }, { "epoch": 15.525, "grad_norm": 1.8203125, "learning_rate": 4.9895988197333386e-05, "loss": 0.2033, "step": 2484 }, { "epoch": 15.53125, "grad_norm": 1.7734375, "learning_rate": 4.989575264056812e-05, "loss": 0.1504, "step": 2485 }, { "epoch": 15.5375, "grad_norm": 1.921875, "learning_rate": 4.9895516817927636e-05, "loss": 0.2081, "step": 2486 }, { "epoch": 15.54375, "grad_norm": 1.65625, "learning_rate": 4.989528072941446e-05, "loss": 0.2234, "step": 2487 }, { "epoch": 15.55, "grad_norm": 1.109375, "learning_rate": 4.989504437503113e-05, "loss": 0.1191, "step": 2488 }, { "epoch": 15.55625, "grad_norm": 1.5625, "learning_rate": 4.989480775478015e-05, "loss": 0.1776, "step": 2489 }, { "epoch": 15.5625, "grad_norm": 1.6796875, "learning_rate": 4.989457086866406e-05, "loss": 0.1808, "step": 2490 }, { "epoch": 15.56875, "grad_norm": 2.109375, "learning_rate": 4.989433371668539e-05, "loss": 0.1499, "step": 2491 }, { "epoch": 15.575, "grad_norm": 1.6796875, "learning_rate": 4.989409629884666e-05, "loss": 0.2025, "step": 2492 }, { "epoch": 15.58125, "grad_norm": 1.203125, "learning_rate": 4.989385861515042e-05, "loss": 0.1212, "step": 2493 }, { "epoch": 15.5875, "grad_norm": 1.640625, "learning_rate": 4.98936206655992e-05, "loss": 0.2437, "step": 2494 }, { "epoch": 15.59375, "grad_norm": 1.4765625, "learning_rate": 4.989338245019554e-05, "loss": 0.1615, "step": 2495 }, { "epoch": 15.6, "grad_norm": 1.1875, "learning_rate": 4.989314396894199e-05, "loss": 0.1404, "step": 2496 }, { "epoch": 15.60625, "grad_norm": 1.6015625, "learning_rate": 4.98929052218411e-05, "loss": 0.1992, "step": 2497 }, { "epoch": 15.6125, "grad_norm": 1.265625, "learning_rate": 4.9892666208895406e-05, "loss": 0.2011, "step": 2498 }, { "epoch": 15.61875, "grad_norm": 1.8984375, "learning_rate": 4.9892426930107474e-05, "loss": 0.1988, "step": 2499 }, { "epoch": 15.625, "grad_norm": 1.546875, "learning_rate": 4.989218738547985e-05, "loss": 0.1729, "step": 2500 }, { "epoch": 15.63125, "grad_norm": 1.7109375, "learning_rate": 4.98919475750151e-05, "loss": 0.2591, "step": 2501 }, { "epoch": 15.6375, "grad_norm": 2.046875, "learning_rate": 4.9891707498715775e-05, "loss": 0.2293, "step": 2502 }, { "epoch": 15.64375, "grad_norm": 0.921875, "learning_rate": 4.989146715658445e-05, "loss": 0.1021, "step": 2503 }, { "epoch": 15.65, "grad_norm": 1.2421875, "learning_rate": 4.989122654862368e-05, "loss": 0.1991, "step": 2504 }, { "epoch": 15.65625, "grad_norm": 1.4453125, "learning_rate": 4.989098567483605e-05, "loss": 0.1772, "step": 2505 }, { "epoch": 15.6625, "grad_norm": 1.421875, "learning_rate": 4.989074453522412e-05, "loss": 0.1473, "step": 2506 }, { "epoch": 15.66875, "grad_norm": 2.25, "learning_rate": 4.9890503129790467e-05, "loss": 0.2389, "step": 2507 }, { "epoch": 15.675, "grad_norm": 1.46875, "learning_rate": 4.989026145853768e-05, "loss": 0.1848, "step": 2508 }, { "epoch": 15.68125, "grad_norm": 1.5703125, "learning_rate": 4.989001952146832e-05, "loss": 0.1729, "step": 2509 }, { "epoch": 15.6875, "grad_norm": 2.140625, "learning_rate": 4.988977731858499e-05, "loss": 0.2423, "step": 2510 }, { "epoch": 15.69375, "grad_norm": 1.921875, "learning_rate": 4.9889534849890264e-05, "loss": 0.1973, "step": 2511 }, { "epoch": 15.7, "grad_norm": 1.328125, "learning_rate": 4.988929211538674e-05, "loss": 0.1573, "step": 2512 }, { "epoch": 15.70625, "grad_norm": 1.796875, "learning_rate": 4.9889049115077005e-05, "loss": 0.1917, "step": 2513 }, { "epoch": 15.7125, "grad_norm": 1.234375, "learning_rate": 4.988880584896365e-05, "loss": 0.1077, "step": 2514 }, { "epoch": 15.71875, "grad_norm": 1.2890625, "learning_rate": 4.988856231704928e-05, "loss": 0.1399, "step": 2515 }, { "epoch": 15.725, "grad_norm": 1.5859375, "learning_rate": 4.9888318519336504e-05, "loss": 0.1804, "step": 2516 }, { "epoch": 15.73125, "grad_norm": 1.1796875, "learning_rate": 4.9888074455827904e-05, "loss": 0.1012, "step": 2517 }, { "epoch": 15.7375, "grad_norm": 0.9921875, "learning_rate": 4.9887830126526105e-05, "loss": 0.1073, "step": 2518 }, { "epoch": 15.74375, "grad_norm": 1.140625, "learning_rate": 4.988758553143371e-05, "loss": 0.1156, "step": 2519 }, { "epoch": 15.75, "grad_norm": 1.1796875, "learning_rate": 4.988734067055333e-05, "loss": 0.1767, "step": 2520 }, { "epoch": 15.75625, "grad_norm": 1.7421875, "learning_rate": 4.988709554388757e-05, "loss": 0.2487, "step": 2521 }, { "epoch": 15.7625, "grad_norm": 1.7890625, "learning_rate": 4.988685015143907e-05, "loss": 0.2088, "step": 2522 }, { "epoch": 15.76875, "grad_norm": 1.40625, "learning_rate": 4.988660449321044e-05, "loss": 0.1694, "step": 2523 }, { "epoch": 15.775, "grad_norm": 1.203125, "learning_rate": 4.98863585692043e-05, "loss": 0.1203, "step": 2524 }, { "epoch": 15.78125, "grad_norm": 1.609375, "learning_rate": 4.9886112379423275e-05, "loss": 0.1923, "step": 2525 }, { "epoch": 15.7875, "grad_norm": 1.1484375, "learning_rate": 4.988586592387e-05, "loss": 0.0965, "step": 2526 }, { "epoch": 15.79375, "grad_norm": 1.34375, "learning_rate": 4.9885619202547114e-05, "loss": 0.1549, "step": 2527 }, { "epoch": 15.8, "grad_norm": 1.171875, "learning_rate": 4.9885372215457235e-05, "loss": 0.1138, "step": 2528 }, { "epoch": 15.80625, "grad_norm": 1.125, "learning_rate": 4.988512496260301e-05, "loss": 0.1148, "step": 2529 }, { "epoch": 15.8125, "grad_norm": 1.046875, "learning_rate": 4.988487744398708e-05, "loss": 0.1188, "step": 2530 }, { "epoch": 15.81875, "grad_norm": 1.90625, "learning_rate": 4.9884629659612094e-05, "loss": 0.2555, "step": 2531 }, { "epoch": 15.825, "grad_norm": 1.515625, "learning_rate": 4.988438160948068e-05, "loss": 0.1807, "step": 2532 }, { "epoch": 15.83125, "grad_norm": 2.125, "learning_rate": 4.98841332935955e-05, "loss": 0.2992, "step": 2533 }, { "epoch": 15.8375, "grad_norm": 1.4296875, "learning_rate": 4.988388471195921e-05, "loss": 0.1779, "step": 2534 }, { "epoch": 15.84375, "grad_norm": 1.7890625, "learning_rate": 4.988363586457445e-05, "loss": 0.2124, "step": 2535 }, { "epoch": 15.85, "grad_norm": 1.796875, "learning_rate": 4.9883386751443896e-05, "loss": 0.251, "step": 2536 }, { "epoch": 15.85625, "grad_norm": 1.96875, "learning_rate": 4.988313737257019e-05, "loss": 0.2659, "step": 2537 }, { "epoch": 15.8625, "grad_norm": 1.65625, "learning_rate": 4.988288772795601e-05, "loss": 0.1438, "step": 2538 }, { "epoch": 15.86875, "grad_norm": 2.0, "learning_rate": 4.988263781760402e-05, "loss": 0.2263, "step": 2539 }, { "epoch": 15.875, "grad_norm": 1.4453125, "learning_rate": 4.9882387641516875e-05, "loss": 0.158, "step": 2540 }, { "epoch": 15.88125, "grad_norm": 1.84375, "learning_rate": 4.988213719969726e-05, "loss": 0.1606, "step": 2541 }, { "epoch": 15.8875, "grad_norm": 1.4765625, "learning_rate": 4.988188649214785e-05, "loss": 0.1327, "step": 2542 }, { "epoch": 15.89375, "grad_norm": 1.25, "learning_rate": 4.988163551887132e-05, "loss": 0.1117, "step": 2543 }, { "epoch": 15.9, "grad_norm": 1.2109375, "learning_rate": 4.988138427987035e-05, "loss": 0.1822, "step": 2544 }, { "epoch": 15.90625, "grad_norm": 1.4453125, "learning_rate": 4.988113277514761e-05, "loss": 0.1568, "step": 2545 }, { "epoch": 15.9125, "grad_norm": 1.4921875, "learning_rate": 4.988088100470581e-05, "loss": 0.1528, "step": 2546 }, { "epoch": 15.91875, "grad_norm": 1.5703125, "learning_rate": 4.9880628968547624e-05, "loss": 0.1516, "step": 2547 }, { "epoch": 15.925, "grad_norm": 1.734375, "learning_rate": 4.9880376666675745e-05, "loss": 0.1427, "step": 2548 }, { "epoch": 15.93125, "grad_norm": 1.6015625, "learning_rate": 4.9880124099092876e-05, "loss": 0.2092, "step": 2549 }, { "epoch": 15.9375, "grad_norm": 1.65625, "learning_rate": 4.98798712658017e-05, "loss": 0.1872, "step": 2550 }, { "epoch": 15.94375, "grad_norm": 1.1171875, "learning_rate": 4.987961816680492e-05, "loss": 0.1139, "step": 2551 }, { "epoch": 15.95, "grad_norm": 1.6015625, "learning_rate": 4.987936480210525e-05, "loss": 0.1824, "step": 2552 }, { "epoch": 15.95625, "grad_norm": 1.7734375, "learning_rate": 4.9879111171705394e-05, "loss": 0.2843, "step": 2553 }, { "epoch": 15.9625, "grad_norm": 1.640625, "learning_rate": 4.987885727560805e-05, "loss": 0.2029, "step": 2554 }, { "epoch": 15.96875, "grad_norm": 1.6875, "learning_rate": 4.987860311381594e-05, "loss": 0.1534, "step": 2555 }, { "epoch": 15.975, "grad_norm": 1.796875, "learning_rate": 4.9878348686331774e-05, "loss": 0.2309, "step": 2556 }, { "epoch": 15.98125, "grad_norm": 1.1953125, "learning_rate": 4.987809399315827e-05, "loss": 0.1758, "step": 2557 }, { "epoch": 15.9875, "grad_norm": 1.4765625, "learning_rate": 4.987783903429814e-05, "loss": 0.1574, "step": 2558 }, { "epoch": 15.99375, "grad_norm": 1.5703125, "learning_rate": 4.987758380975411e-05, "loss": 0.3067, "step": 2559 }, { "epoch": 16.0, "grad_norm": 1.5703125, "learning_rate": 4.987732831952893e-05, "loss": 0.173, "step": 2560 }, { "epoch": 16.00625, "grad_norm": 1.578125, "learning_rate": 4.9877072563625285e-05, "loss": 0.2065, "step": 2561 }, { "epoch": 16.0125, "grad_norm": 1.6015625, "learning_rate": 4.987681654204595e-05, "loss": 0.1473, "step": 2562 }, { "epoch": 16.01875, "grad_norm": 1.65625, "learning_rate": 4.987656025479362e-05, "loss": 0.1894, "step": 2563 }, { "epoch": 16.025, "grad_norm": 1.4765625, "learning_rate": 4.987630370187106e-05, "loss": 0.2204, "step": 2564 }, { "epoch": 16.03125, "grad_norm": 1.828125, "learning_rate": 4.9876046883281005e-05, "loss": 0.205, "step": 2565 }, { "epoch": 16.0375, "grad_norm": 1.734375, "learning_rate": 4.987578979902619e-05, "loss": 0.1587, "step": 2566 }, { "epoch": 16.04375, "grad_norm": 1.203125, "learning_rate": 4.987553244910935e-05, "loss": 0.1645, "step": 2567 }, { "epoch": 16.05, "grad_norm": 1.671875, "learning_rate": 4.987527483353326e-05, "loss": 0.1903, "step": 2568 }, { "epoch": 16.05625, "grad_norm": 1.1953125, "learning_rate": 4.987501695230066e-05, "loss": 0.1518, "step": 2569 }, { "epoch": 16.0625, "grad_norm": 1.7578125, "learning_rate": 4.98747588054143e-05, "loss": 0.201, "step": 2570 }, { "epoch": 16.06875, "grad_norm": 1.2734375, "learning_rate": 4.987450039287694e-05, "loss": 0.1258, "step": 2571 }, { "epoch": 16.075, "grad_norm": 2.15625, "learning_rate": 4.987424171469134e-05, "loss": 0.2384, "step": 2572 }, { "epoch": 16.08125, "grad_norm": 1.5859375, "learning_rate": 4.987398277086026e-05, "loss": 0.1461, "step": 2573 }, { "epoch": 16.0875, "grad_norm": 1.546875, "learning_rate": 4.987372356138647e-05, "loss": 0.1332, "step": 2574 }, { "epoch": 16.09375, "grad_norm": 1.171875, "learning_rate": 4.987346408627272e-05, "loss": 0.107, "step": 2575 }, { "epoch": 16.1, "grad_norm": 1.5625, "learning_rate": 4.9873204345521815e-05, "loss": 0.2079, "step": 2576 }, { "epoch": 16.10625, "grad_norm": 1.609375, "learning_rate": 4.9872944339136503e-05, "loss": 0.3042, "step": 2577 }, { "epoch": 16.1125, "grad_norm": 2.0625, "learning_rate": 4.9872684067119566e-05, "loss": 0.2394, "step": 2578 }, { "epoch": 16.11875, "grad_norm": 2.328125, "learning_rate": 4.987242352947379e-05, "loss": 0.2141, "step": 2579 }, { "epoch": 16.125, "grad_norm": 1.5859375, "learning_rate": 4.987216272620195e-05, "loss": 0.1221, "step": 2580 }, { "epoch": 16.13125, "grad_norm": 1.2734375, "learning_rate": 4.987190165730684e-05, "loss": 0.1976, "step": 2581 }, { "epoch": 16.1375, "grad_norm": 1.65625, "learning_rate": 4.9871640322791234e-05, "loss": 0.1375, "step": 2582 }, { "epoch": 16.14375, "grad_norm": 2.1875, "learning_rate": 4.9871378722657934e-05, "loss": 0.1572, "step": 2583 }, { "epoch": 16.15, "grad_norm": 1.9140625, "learning_rate": 4.987111685690974e-05, "loss": 0.1585, "step": 2584 }, { "epoch": 16.15625, "grad_norm": 1.4609375, "learning_rate": 4.987085472554943e-05, "loss": 0.161, "step": 2585 }, { "epoch": 16.1625, "grad_norm": 1.1953125, "learning_rate": 4.987059232857981e-05, "loss": 0.1117, "step": 2586 }, { "epoch": 16.16875, "grad_norm": 2.0, "learning_rate": 4.987032966600369e-05, "loss": 0.1674, "step": 2587 }, { "epoch": 16.175, "grad_norm": 1.5390625, "learning_rate": 4.9870066737823874e-05, "loss": 0.1605, "step": 2588 }, { "epoch": 16.18125, "grad_norm": 1.3984375, "learning_rate": 4.9869803544043166e-05, "loss": 0.1875, "step": 2589 }, { "epoch": 16.1875, "grad_norm": 1.1328125, "learning_rate": 4.9869540084664376e-05, "loss": 0.1189, "step": 2590 }, { "epoch": 16.19375, "grad_norm": 1.3515625, "learning_rate": 4.986927635969031e-05, "loss": 0.1365, "step": 2591 }, { "epoch": 16.2, "grad_norm": 1.5546875, "learning_rate": 4.9869012369123805e-05, "loss": 0.1745, "step": 2592 }, { "epoch": 16.20625, "grad_norm": 1.6953125, "learning_rate": 4.986874811296767e-05, "loss": 0.199, "step": 2593 }, { "epoch": 16.2125, "grad_norm": 1.6484375, "learning_rate": 4.9868483591224713e-05, "loss": 0.2464, "step": 2594 }, { "epoch": 16.21875, "grad_norm": 2.1875, "learning_rate": 4.986821880389778e-05, "loss": 0.2776, "step": 2595 }, { "epoch": 16.225, "grad_norm": 1.484375, "learning_rate": 4.986795375098969e-05, "loss": 0.1685, "step": 2596 }, { "epoch": 16.23125, "grad_norm": 1.6875, "learning_rate": 4.986768843250328e-05, "loss": 0.1676, "step": 2597 }, { "epoch": 16.2375, "grad_norm": 1.59375, "learning_rate": 4.986742284844137e-05, "loss": 0.1373, "step": 2598 }, { "epoch": 16.24375, "grad_norm": 1.3671875, "learning_rate": 4.986715699880681e-05, "loss": 0.1376, "step": 2599 }, { "epoch": 16.25, "grad_norm": 1.2578125, "learning_rate": 4.9866890883602424e-05, "loss": 0.126, "step": 2600 }, { "epoch": 16.25625, "grad_norm": 1.4140625, "learning_rate": 4.986662450283107e-05, "loss": 0.1609, "step": 2601 }, { "epoch": 16.2625, "grad_norm": 1.7421875, "learning_rate": 4.986635785649558e-05, "loss": 0.187, "step": 2602 }, { "epoch": 16.26875, "grad_norm": 1.546875, "learning_rate": 4.986609094459881e-05, "loss": 0.1727, "step": 2603 }, { "epoch": 16.275, "grad_norm": 2.296875, "learning_rate": 4.9865823767143614e-05, "loss": 0.2682, "step": 2604 }, { "epoch": 16.28125, "grad_norm": 1.3671875, "learning_rate": 4.9865556324132836e-05, "loss": 0.1352, "step": 2605 }, { "epoch": 16.2875, "grad_norm": 1.984375, "learning_rate": 4.9865288615569335e-05, "loss": 0.1808, "step": 2606 }, { "epoch": 16.29375, "grad_norm": 1.3515625, "learning_rate": 4.986502064145597e-05, "loss": 0.1439, "step": 2607 }, { "epoch": 16.3, "grad_norm": 1.328125, "learning_rate": 4.98647524017956e-05, "loss": 0.1776, "step": 2608 }, { "epoch": 16.30625, "grad_norm": 1.6015625, "learning_rate": 4.9864483896591094e-05, "loss": 0.1995, "step": 2609 }, { "epoch": 16.3125, "grad_norm": 1.84375, "learning_rate": 4.986421512584533e-05, "loss": 0.174, "step": 2610 }, { "epoch": 16.31875, "grad_norm": 1.234375, "learning_rate": 4.986394608956115e-05, "loss": 0.1457, "step": 2611 }, { "epoch": 16.325, "grad_norm": 1.609375, "learning_rate": 4.986367678774146e-05, "loss": 0.1876, "step": 2612 }, { "epoch": 16.33125, "grad_norm": 1.5625, "learning_rate": 4.9863407220389116e-05, "loss": 0.168, "step": 2613 }, { "epoch": 16.3375, "grad_norm": 1.3203125, "learning_rate": 4.986313738750699e-05, "loss": 0.1342, "step": 2614 }, { "epoch": 16.34375, "grad_norm": 1.4609375, "learning_rate": 4.9862867289097986e-05, "loss": 0.1687, "step": 2615 }, { "epoch": 16.35, "grad_norm": 1.75, "learning_rate": 4.9862596925164976e-05, "loss": 0.1983, "step": 2616 }, { "epoch": 16.35625, "grad_norm": 1.7578125, "learning_rate": 4.986232629571085e-05, "loss": 0.1886, "step": 2617 }, { "epoch": 16.3625, "grad_norm": 1.5390625, "learning_rate": 4.9862055400738486e-05, "loss": 0.1897, "step": 2618 }, { "epoch": 16.36875, "grad_norm": 1.7265625, "learning_rate": 4.98617842402508e-05, "loss": 0.1746, "step": 2619 }, { "epoch": 16.375, "grad_norm": 1.015625, "learning_rate": 4.986151281425067e-05, "loss": 0.0991, "step": 2620 }, { "epoch": 16.38125, "grad_norm": 1.984375, "learning_rate": 4.9861241122741e-05, "loss": 0.2872, "step": 2621 }, { "epoch": 16.3875, "grad_norm": 1.921875, "learning_rate": 4.986096916572469e-05, "loss": 0.2989, "step": 2622 }, { "epoch": 16.39375, "grad_norm": 1.0859375, "learning_rate": 4.986069694320465e-05, "loss": 0.1024, "step": 2623 }, { "epoch": 16.4, "grad_norm": 1.609375, "learning_rate": 4.986042445518378e-05, "loss": 0.2415, "step": 2624 }, { "epoch": 16.40625, "grad_norm": 1.015625, "learning_rate": 4.9860151701665004e-05, "loss": 0.1097, "step": 2625 }, { "epoch": 16.4125, "grad_norm": 1.1875, "learning_rate": 4.9859878682651215e-05, "loss": 0.1448, "step": 2626 }, { "epoch": 16.41875, "grad_norm": 0.875, "learning_rate": 4.985960539814535e-05, "loss": 0.0839, "step": 2627 }, { "epoch": 16.425, "grad_norm": 1.2890625, "learning_rate": 4.98593318481503e-05, "loss": 0.1248, "step": 2628 }, { "epoch": 16.43125, "grad_norm": 1.3046875, "learning_rate": 4.9859058032669014e-05, "loss": 0.1612, "step": 2629 }, { "epoch": 16.4375, "grad_norm": 1.4609375, "learning_rate": 4.98587839517044e-05, "loss": 0.2506, "step": 2630 }, { "epoch": 16.44375, "grad_norm": 1.53125, "learning_rate": 4.985850960525938e-05, "loss": 0.2045, "step": 2631 }, { "epoch": 16.45, "grad_norm": 1.5078125, "learning_rate": 4.9858234993336906e-05, "loss": 0.1866, "step": 2632 }, { "epoch": 16.45625, "grad_norm": 1.703125, "learning_rate": 4.98579601159399e-05, "loss": 0.1628, "step": 2633 }, { "epoch": 16.4625, "grad_norm": 1.171875, "learning_rate": 4.9857684973071285e-05, "loss": 0.1156, "step": 2634 }, { "epoch": 16.46875, "grad_norm": 1.484375, "learning_rate": 4.9857409564734016e-05, "loss": 0.1664, "step": 2635 }, { "epoch": 16.475, "grad_norm": 2.296875, "learning_rate": 4.985713389093103e-05, "loss": 0.2625, "step": 2636 }, { "epoch": 16.48125, "grad_norm": 1.40625, "learning_rate": 4.985685795166526e-05, "loss": 0.149, "step": 2637 }, { "epoch": 16.4875, "grad_norm": 1.6171875, "learning_rate": 4.9856581746939675e-05, "loss": 0.2111, "step": 2638 }, { "epoch": 16.49375, "grad_norm": 1.4140625, "learning_rate": 4.985630527675721e-05, "loss": 0.1705, "step": 2639 }, { "epoch": 16.5, "grad_norm": 1.546875, "learning_rate": 4.9856028541120807e-05, "loss": 0.1537, "step": 2640 }, { "epoch": 16.50625, "grad_norm": 1.625, "learning_rate": 4.985575154003345e-05, "loss": 0.2027, "step": 2641 }, { "epoch": 16.5125, "grad_norm": 1.5703125, "learning_rate": 4.985547427349807e-05, "loss": 0.1619, "step": 2642 }, { "epoch": 16.51875, "grad_norm": 1.4375, "learning_rate": 4.985519674151764e-05, "loss": 0.1982, "step": 2643 }, { "epoch": 16.525, "grad_norm": 1.3359375, "learning_rate": 4.985491894409512e-05, "loss": 0.1175, "step": 2644 }, { "epoch": 16.53125, "grad_norm": 1.1328125, "learning_rate": 4.985464088123349e-05, "loss": 0.1008, "step": 2645 }, { "epoch": 16.5375, "grad_norm": 1.3671875, "learning_rate": 4.9854362552935706e-05, "loss": 0.1258, "step": 2646 }, { "epoch": 16.54375, "grad_norm": 1.453125, "learning_rate": 4.985408395920474e-05, "loss": 0.2166, "step": 2647 }, { "epoch": 16.55, "grad_norm": 1.7421875, "learning_rate": 4.9853805100043584e-05, "loss": 0.1301, "step": 2648 }, { "epoch": 16.55625, "grad_norm": 1.6484375, "learning_rate": 4.985352597545519e-05, "loss": 0.1539, "step": 2649 }, { "epoch": 16.5625, "grad_norm": 1.359375, "learning_rate": 4.985324658544256e-05, "loss": 0.1418, "step": 2650 }, { "epoch": 16.56875, "grad_norm": 1.296875, "learning_rate": 4.985296693000866e-05, "loss": 0.1453, "step": 2651 }, { "epoch": 16.575, "grad_norm": 1.3515625, "learning_rate": 4.9852687009156496e-05, "loss": 0.1292, "step": 2652 }, { "epoch": 16.58125, "grad_norm": 1.4609375, "learning_rate": 4.985240682288904e-05, "loss": 0.1716, "step": 2653 }, { "epoch": 16.5875, "grad_norm": 1.7578125, "learning_rate": 4.98521263712093e-05, "loss": 0.1715, "step": 2654 }, { "epoch": 16.59375, "grad_norm": 1.28125, "learning_rate": 4.985184565412026e-05, "loss": 0.1199, "step": 2655 }, { "epoch": 16.6, "grad_norm": 1.609375, "learning_rate": 4.9851564671624916e-05, "loss": 0.2277, "step": 2656 }, { "epoch": 16.60625, "grad_norm": 1.1796875, "learning_rate": 4.985128342372629e-05, "loss": 0.1505, "step": 2657 }, { "epoch": 16.6125, "grad_norm": 1.3125, "learning_rate": 4.985100191042735e-05, "loss": 0.1362, "step": 2658 }, { "epoch": 16.61875, "grad_norm": 1.5625, "learning_rate": 4.9850720131731135e-05, "loss": 0.1695, "step": 2659 }, { "epoch": 16.625, "grad_norm": 1.15625, "learning_rate": 4.985043808764064e-05, "loss": 0.1322, "step": 2660 }, { "epoch": 16.63125, "grad_norm": 1.390625, "learning_rate": 4.985015577815887e-05, "loss": 0.209, "step": 2661 }, { "epoch": 16.6375, "grad_norm": 2.015625, "learning_rate": 4.9849873203288854e-05, "loss": 0.1958, "step": 2662 }, { "epoch": 16.64375, "grad_norm": 1.8046875, "learning_rate": 4.984959036303361e-05, "loss": 0.1983, "step": 2663 }, { "epoch": 16.65, "grad_norm": 1.3359375, "learning_rate": 4.984930725739615e-05, "loss": 0.1354, "step": 2664 }, { "epoch": 16.65625, "grad_norm": 1.390625, "learning_rate": 4.98490238863795e-05, "loss": 0.1806, "step": 2665 }, { "epoch": 16.6625, "grad_norm": 1.7265625, "learning_rate": 4.984874024998668e-05, "loss": 0.1521, "step": 2666 }, { "epoch": 16.66875, "grad_norm": 1.921875, "learning_rate": 4.984845634822073e-05, "loss": 0.203, "step": 2667 }, { "epoch": 16.675, "grad_norm": 1.640625, "learning_rate": 4.984817218108467e-05, "loss": 0.1346, "step": 2668 }, { "epoch": 16.68125, "grad_norm": 2.0625, "learning_rate": 4.9847887748581554e-05, "loss": 0.2119, "step": 2669 }, { "epoch": 16.6875, "grad_norm": 1.515625, "learning_rate": 4.9847603050714405e-05, "loss": 0.19, "step": 2670 }, { "epoch": 16.69375, "grad_norm": 1.3671875, "learning_rate": 4.984731808748626e-05, "loss": 0.1166, "step": 2671 }, { "epoch": 16.7, "grad_norm": 1.1171875, "learning_rate": 4.984703285890018e-05, "loss": 0.1034, "step": 2672 }, { "epoch": 16.70625, "grad_norm": 2.28125, "learning_rate": 4.984674736495919e-05, "loss": 0.2642, "step": 2673 }, { "epoch": 16.7125, "grad_norm": 1.09375, "learning_rate": 4.984646160566635e-05, "loss": 0.117, "step": 2674 }, { "epoch": 16.71875, "grad_norm": 1.515625, "learning_rate": 4.984617558102471e-05, "loss": 0.1393, "step": 2675 }, { "epoch": 16.725, "grad_norm": 0.86328125, "learning_rate": 4.984588929103733e-05, "loss": 0.0929, "step": 2676 }, { "epoch": 16.73125, "grad_norm": 1.609375, "learning_rate": 4.9845602735707266e-05, "loss": 0.207, "step": 2677 }, { "epoch": 16.7375, "grad_norm": 1.453125, "learning_rate": 4.984531591503756e-05, "loss": 0.183, "step": 2678 }, { "epoch": 16.74375, "grad_norm": 1.515625, "learning_rate": 4.9845028829031306e-05, "loss": 0.2157, "step": 2679 }, { "epoch": 16.75, "grad_norm": 1.5703125, "learning_rate": 4.984474147769154e-05, "loss": 0.1676, "step": 2680 }, { "epoch": 16.75625, "grad_norm": 1.25, "learning_rate": 4.9844453861021356e-05, "loss": 0.1385, "step": 2681 }, { "epoch": 16.7625, "grad_norm": 1.125, "learning_rate": 4.9844165979023805e-05, "loss": 0.1207, "step": 2682 }, { "epoch": 16.76875, "grad_norm": 1.4921875, "learning_rate": 4.984387783170197e-05, "loss": 0.1357, "step": 2683 }, { "epoch": 16.775, "grad_norm": 1.6640625, "learning_rate": 4.984358941905894e-05, "loss": 0.1755, "step": 2684 }, { "epoch": 16.78125, "grad_norm": 1.8359375, "learning_rate": 4.984330074109777e-05, "loss": 0.1584, "step": 2685 }, { "epoch": 16.7875, "grad_norm": 1.484375, "learning_rate": 4.9843011797821574e-05, "loss": 0.1854, "step": 2686 }, { "epoch": 16.79375, "grad_norm": 1.609375, "learning_rate": 4.984272258923341e-05, "loss": 0.1564, "step": 2687 }, { "epoch": 16.8, "grad_norm": 1.5078125, "learning_rate": 4.9842433115336376e-05, "loss": 0.1762, "step": 2688 }, { "epoch": 16.80625, "grad_norm": 1.3671875, "learning_rate": 4.984214337613357e-05, "loss": 0.1388, "step": 2689 }, { "epoch": 16.8125, "grad_norm": 1.21875, "learning_rate": 4.9841853371628075e-05, "loss": 0.1222, "step": 2690 }, { "epoch": 16.81875, "grad_norm": 2.046875, "learning_rate": 4.9841563101823e-05, "loss": 0.215, "step": 2691 }, { "epoch": 16.825, "grad_norm": 1.203125, "learning_rate": 4.9841272566721444e-05, "loss": 0.1153, "step": 2692 }, { "epoch": 16.83125, "grad_norm": 1.28125, "learning_rate": 4.98409817663265e-05, "loss": 0.1162, "step": 2693 }, { "epoch": 16.8375, "grad_norm": 2.09375, "learning_rate": 4.984069070064127e-05, "loss": 0.1567, "step": 2694 }, { "epoch": 16.84375, "grad_norm": 2.0, "learning_rate": 4.984039936966888e-05, "loss": 0.1854, "step": 2695 }, { "epoch": 16.85, "grad_norm": 1.421875, "learning_rate": 4.984010777341243e-05, "loss": 0.1862, "step": 2696 }, { "epoch": 16.85625, "grad_norm": 1.4921875, "learning_rate": 4.983981591187504e-05, "loss": 0.192, "step": 2697 }, { "epoch": 16.8625, "grad_norm": 3.171875, "learning_rate": 4.9839523785059815e-05, "loss": 0.2372, "step": 2698 }, { "epoch": 16.86875, "grad_norm": 1.84375, "learning_rate": 4.9839231392969884e-05, "loss": 0.135, "step": 2699 }, { "epoch": 16.875, "grad_norm": 1.4609375, "learning_rate": 4.9838938735608384e-05, "loss": 0.1304, "step": 2700 }, { "epoch": 16.88125, "grad_norm": 1.1640625, "learning_rate": 4.983864581297841e-05, "loss": 0.1112, "step": 2701 }, { "epoch": 16.8875, "grad_norm": 1.3671875, "learning_rate": 4.983835262508311e-05, "loss": 0.1182, "step": 2702 }, { "epoch": 16.89375, "grad_norm": 1.40625, "learning_rate": 4.983805917192561e-05, "loss": 0.1871, "step": 2703 }, { "epoch": 16.9, "grad_norm": 1.578125, "learning_rate": 4.9837765453509046e-05, "loss": 0.1778, "step": 2704 }, { "epoch": 16.90625, "grad_norm": 1.3046875, "learning_rate": 4.983747146983656e-05, "loss": 0.1193, "step": 2705 }, { "epoch": 16.9125, "grad_norm": 1.4609375, "learning_rate": 4.983717722091127e-05, "loss": 0.1182, "step": 2706 }, { "epoch": 16.91875, "grad_norm": 1.3828125, "learning_rate": 4.983688270673634e-05, "loss": 0.1722, "step": 2707 }, { "epoch": 16.925, "grad_norm": 1.578125, "learning_rate": 4.9836587927314914e-05, "loss": 0.2191, "step": 2708 }, { "epoch": 16.93125, "grad_norm": 1.1875, "learning_rate": 4.983629288265013e-05, "loss": 0.1192, "step": 2709 }, { "epoch": 16.9375, "grad_norm": 1.375, "learning_rate": 4.983599757274514e-05, "loss": 0.1439, "step": 2710 }, { "epoch": 16.94375, "grad_norm": 1.609375, "learning_rate": 4.9835701997603114e-05, "loss": 0.1532, "step": 2711 }, { "epoch": 16.95, "grad_norm": 1.28125, "learning_rate": 4.983540615722718e-05, "loss": 0.1604, "step": 2712 }, { "epoch": 16.95625, "grad_norm": 1.3984375, "learning_rate": 4.9835110051620526e-05, "loss": 0.137, "step": 2713 }, { "epoch": 16.9625, "grad_norm": 1.6171875, "learning_rate": 4.9834813680786294e-05, "loss": 0.1366, "step": 2714 }, { "epoch": 16.96875, "grad_norm": 1.484375, "learning_rate": 4.983451704472766e-05, "loss": 0.1691, "step": 2715 }, { "epoch": 16.975, "grad_norm": 1.890625, "learning_rate": 4.9834220143447794e-05, "loss": 0.2296, "step": 2716 }, { "epoch": 16.98125, "grad_norm": 2.015625, "learning_rate": 4.9833922976949855e-05, "loss": 0.2704, "step": 2717 }, { "epoch": 16.9875, "grad_norm": 1.203125, "learning_rate": 4.983362554523703e-05, "loss": 0.1131, "step": 2718 }, { "epoch": 16.99375, "grad_norm": 1.84375, "learning_rate": 4.983332784831248e-05, "loss": 0.2381, "step": 2719 }, { "epoch": 17.0, "grad_norm": 1.453125, "learning_rate": 4.9833029886179396e-05, "loss": 0.1551, "step": 2720 }, { "epoch": 17.00625, "grad_norm": 0.76171875, "learning_rate": 4.9832731658840956e-05, "loss": 0.0747, "step": 2721 }, { "epoch": 17.0125, "grad_norm": 1.078125, "learning_rate": 4.9832433166300355e-05, "loss": 0.0963, "step": 2722 }, { "epoch": 17.01875, "grad_norm": 1.1953125, "learning_rate": 4.983213440856076e-05, "loss": 0.0982, "step": 2723 }, { "epoch": 17.025, "grad_norm": 1.1953125, "learning_rate": 4.983183538562538e-05, "loss": 0.1184, "step": 2724 }, { "epoch": 17.03125, "grad_norm": 1.8203125, "learning_rate": 4.983153609749739e-05, "loss": 0.1967, "step": 2725 }, { "epoch": 17.0375, "grad_norm": 1.3046875, "learning_rate": 4.9831236544180013e-05, "loss": 0.1109, "step": 2726 }, { "epoch": 17.04375, "grad_norm": 1.609375, "learning_rate": 4.9830936725676425e-05, "loss": 0.1938, "step": 2727 }, { "epoch": 17.05, "grad_norm": 1.5234375, "learning_rate": 4.983063664198984e-05, "loss": 0.1308, "step": 2728 }, { "epoch": 17.05625, "grad_norm": 1.28125, "learning_rate": 4.983033629312346e-05, "loss": 0.1304, "step": 2729 }, { "epoch": 17.0625, "grad_norm": 1.2734375, "learning_rate": 4.983003567908049e-05, "loss": 0.137, "step": 2730 }, { "epoch": 17.06875, "grad_norm": 1.71875, "learning_rate": 4.982973479986414e-05, "loss": 0.1337, "step": 2731 }, { "epoch": 17.075, "grad_norm": 1.375, "learning_rate": 4.9829433655477616e-05, "loss": 0.1194, "step": 2732 }, { "epoch": 17.08125, "grad_norm": 1.6796875, "learning_rate": 4.982913224592416e-05, "loss": 0.1894, "step": 2733 }, { "epoch": 17.0875, "grad_norm": 1.3359375, "learning_rate": 4.982883057120696e-05, "loss": 0.118, "step": 2734 }, { "epoch": 17.09375, "grad_norm": 1.5390625, "learning_rate": 4.982852863132926e-05, "loss": 0.1202, "step": 2735 }, { "epoch": 17.1, "grad_norm": 1.265625, "learning_rate": 4.982822642629428e-05, "loss": 0.167, "step": 2736 }, { "epoch": 17.10625, "grad_norm": 1.2421875, "learning_rate": 4.982792395610524e-05, "loss": 0.1204, "step": 2737 }, { "epoch": 17.1125, "grad_norm": 1.046875, "learning_rate": 4.9827621220765376e-05, "loss": 0.0956, "step": 2738 }, { "epoch": 17.11875, "grad_norm": 1.515625, "learning_rate": 4.982731822027791e-05, "loss": 0.1735, "step": 2739 }, { "epoch": 17.125, "grad_norm": 1.671875, "learning_rate": 4.982701495464609e-05, "loss": 0.218, "step": 2740 }, { "epoch": 17.13125, "grad_norm": 1.5, "learning_rate": 4.982671142387316e-05, "loss": 0.1757, "step": 2741 }, { "epoch": 17.1375, "grad_norm": 1.484375, "learning_rate": 4.982640762796234e-05, "loss": 0.1728, "step": 2742 }, { "epoch": 17.14375, "grad_norm": 1.6953125, "learning_rate": 4.9826103566916904e-05, "loss": 0.1324, "step": 2743 }, { "epoch": 17.15, "grad_norm": 1.4765625, "learning_rate": 4.9825799240740076e-05, "loss": 0.193, "step": 2744 }, { "epoch": 17.15625, "grad_norm": 0.87109375, "learning_rate": 4.982549464943511e-05, "loss": 0.107, "step": 2745 }, { "epoch": 17.1625, "grad_norm": 1.2578125, "learning_rate": 4.9825189793005264e-05, "loss": 0.141, "step": 2746 }, { "epoch": 17.16875, "grad_norm": 2.296875, "learning_rate": 4.982488467145379e-05, "loss": 0.2012, "step": 2747 }, { "epoch": 17.175, "grad_norm": 1.8359375, "learning_rate": 4.982457928478395e-05, "loss": 0.197, "step": 2748 }, { "epoch": 17.18125, "grad_norm": 1.34375, "learning_rate": 4.9824273632999e-05, "loss": 0.1592, "step": 2749 }, { "epoch": 17.1875, "grad_norm": 1.7734375, "learning_rate": 4.982396771610222e-05, "loss": 0.2719, "step": 2750 }, { "epoch": 17.19375, "grad_norm": 1.640625, "learning_rate": 4.982366153409685e-05, "loss": 0.2567, "step": 2751 }, { "epoch": 17.2, "grad_norm": 0.9609375, "learning_rate": 4.9823355086986176e-05, "loss": 0.0924, "step": 2752 }, { "epoch": 17.20625, "grad_norm": 1.5390625, "learning_rate": 4.982304837477348e-05, "loss": 0.2035, "step": 2753 }, { "epoch": 17.2125, "grad_norm": 1.3671875, "learning_rate": 4.982274139746202e-05, "loss": 0.1701, "step": 2754 }, { "epoch": 17.21875, "grad_norm": 1.375, "learning_rate": 4.982243415505508e-05, "loss": 0.1415, "step": 2755 }, { "epoch": 17.225, "grad_norm": 1.3984375, "learning_rate": 4.9822126647555946e-05, "loss": 0.1613, "step": 2756 }, { "epoch": 17.23125, "grad_norm": 1.6640625, "learning_rate": 4.98218188749679e-05, "loss": 0.19, "step": 2757 }, { "epoch": 17.2375, "grad_norm": 1.421875, "learning_rate": 4.982151083729422e-05, "loss": 0.1294, "step": 2758 }, { "epoch": 17.24375, "grad_norm": 1.3046875, "learning_rate": 4.9821202534538216e-05, "loss": 0.116, "step": 2759 }, { "epoch": 17.25, "grad_norm": 1.265625, "learning_rate": 4.9820893966703155e-05, "loss": 0.1256, "step": 2760 }, { "epoch": 17.25625, "grad_norm": 1.9453125, "learning_rate": 4.982058513379235e-05, "loss": 0.2122, "step": 2761 }, { "epoch": 17.2625, "grad_norm": 1.2578125, "learning_rate": 4.982027603580909e-05, "loss": 0.165, "step": 2762 }, { "epoch": 17.26875, "grad_norm": 1.40625, "learning_rate": 4.981996667275669e-05, "loss": 0.1254, "step": 2763 }, { "epoch": 17.275, "grad_norm": 1.3046875, "learning_rate": 4.9819657044638436e-05, "loss": 0.1205, "step": 2764 }, { "epoch": 17.28125, "grad_norm": 1.296875, "learning_rate": 4.9819347151457644e-05, "loss": 0.1515, "step": 2765 }, { "epoch": 17.2875, "grad_norm": 1.28125, "learning_rate": 4.9819036993217624e-05, "loss": 0.1148, "step": 2766 }, { "epoch": 17.29375, "grad_norm": 1.3046875, "learning_rate": 4.981872656992169e-05, "loss": 0.1228, "step": 2767 }, { "epoch": 17.3, "grad_norm": 1.5625, "learning_rate": 4.981841588157315e-05, "loss": 0.1512, "step": 2768 }, { "epoch": 17.30625, "grad_norm": 0.89453125, "learning_rate": 4.981810492817532e-05, "loss": 0.1016, "step": 2769 }, { "epoch": 17.3125, "grad_norm": 1.53125, "learning_rate": 4.981779370973154e-05, "loss": 0.1686, "step": 2770 }, { "epoch": 17.31875, "grad_norm": 2.3125, "learning_rate": 4.9817482226245115e-05, "loss": 0.2106, "step": 2771 }, { "epoch": 17.325, "grad_norm": 1.7265625, "learning_rate": 4.981717047771937e-05, "loss": 0.1448, "step": 2772 }, { "epoch": 17.33125, "grad_norm": 1.359375, "learning_rate": 4.981685846415764e-05, "loss": 0.1648, "step": 2773 }, { "epoch": 17.3375, "grad_norm": 1.3515625, "learning_rate": 4.9816546185563274e-05, "loss": 0.184, "step": 2774 }, { "epoch": 17.34375, "grad_norm": 1.1484375, "learning_rate": 4.981623364193958e-05, "loss": 0.1005, "step": 2775 }, { "epoch": 17.35, "grad_norm": 1.546875, "learning_rate": 4.9815920833289905e-05, "loss": 0.111, "step": 2776 }, { "epoch": 17.35625, "grad_norm": 2.25, "learning_rate": 4.9815607759617606e-05, "loss": 0.2071, "step": 2777 }, { "epoch": 17.3625, "grad_norm": 1.1875, "learning_rate": 4.9815294420926004e-05, "loss": 0.1623, "step": 2778 }, { "epoch": 17.36875, "grad_norm": 1.140625, "learning_rate": 4.9814980817218447e-05, "loss": 0.1029, "step": 2779 }, { "epoch": 17.375, "grad_norm": 1.6015625, "learning_rate": 4.98146669484983e-05, "loss": 0.1675, "step": 2780 }, { "epoch": 17.38125, "grad_norm": 1.5703125, "learning_rate": 4.9814352814768904e-05, "loss": 0.1736, "step": 2781 }, { "epoch": 17.3875, "grad_norm": 1.921875, "learning_rate": 4.9814038416033614e-05, "loss": 0.2095, "step": 2782 }, { "epoch": 17.39375, "grad_norm": 1.796875, "learning_rate": 4.9813723752295785e-05, "loss": 0.1896, "step": 2783 }, { "epoch": 17.4, "grad_norm": 1.4453125, "learning_rate": 4.981340882355878e-05, "loss": 0.1415, "step": 2784 }, { "epoch": 17.40625, "grad_norm": 1.3828125, "learning_rate": 4.981309362982598e-05, "loss": 0.1481, "step": 2785 }, { "epoch": 17.4125, "grad_norm": 1.546875, "learning_rate": 4.981277817110072e-05, "loss": 0.186, "step": 2786 }, { "epoch": 17.41875, "grad_norm": 1.3359375, "learning_rate": 4.9812462447386384e-05, "loss": 0.1879, "step": 2787 }, { "epoch": 17.425, "grad_norm": 1.34375, "learning_rate": 4.9812146458686355e-05, "loss": 0.1435, "step": 2788 }, { "epoch": 17.43125, "grad_norm": 1.3984375, "learning_rate": 4.981183020500399e-05, "loss": 0.1314, "step": 2789 }, { "epoch": 17.4375, "grad_norm": 2.0, "learning_rate": 4.981151368634267e-05, "loss": 0.3048, "step": 2790 }, { "epoch": 17.44375, "grad_norm": 1.125, "learning_rate": 4.9811196902705784e-05, "loss": 0.1245, "step": 2791 }, { "epoch": 17.45, "grad_norm": 1.53125, "learning_rate": 4.9810879854096704e-05, "loss": 0.1924, "step": 2792 }, { "epoch": 17.45625, "grad_norm": 1.4921875, "learning_rate": 4.9810562540518825e-05, "loss": 0.1364, "step": 2793 }, { "epoch": 17.4625, "grad_norm": 1.6640625, "learning_rate": 4.9810244961975535e-05, "loss": 0.1778, "step": 2794 }, { "epoch": 17.46875, "grad_norm": 1.3046875, "learning_rate": 4.9809927118470215e-05, "loss": 0.1349, "step": 2795 }, { "epoch": 17.475, "grad_norm": 0.77734375, "learning_rate": 4.980960901000627e-05, "loss": 0.0796, "step": 2796 }, { "epoch": 17.48125, "grad_norm": 1.296875, "learning_rate": 4.9809290636587093e-05, "loss": 0.1456, "step": 2797 }, { "epoch": 17.4875, "grad_norm": 1.640625, "learning_rate": 4.9808971998216095e-05, "loss": 0.1943, "step": 2798 }, { "epoch": 17.49375, "grad_norm": 1.703125, "learning_rate": 4.980865309489665e-05, "loss": 0.2078, "step": 2799 }, { "epoch": 17.5, "grad_norm": 1.5078125, "learning_rate": 4.9808333926632203e-05, "loss": 0.2015, "step": 2800 }, { "epoch": 17.50625, "grad_norm": 1.6171875, "learning_rate": 4.980801449342613e-05, "loss": 0.1813, "step": 2801 }, { "epoch": 17.5125, "grad_norm": 1.2890625, "learning_rate": 4.9807694795281854e-05, "loss": 0.1594, "step": 2802 }, { "epoch": 17.51875, "grad_norm": 1.3671875, "learning_rate": 4.9807374832202805e-05, "loss": 0.1343, "step": 2803 }, { "epoch": 17.525, "grad_norm": 1.2890625, "learning_rate": 4.980705460419237e-05, "loss": 0.1276, "step": 2804 }, { "epoch": 17.53125, "grad_norm": 1.296875, "learning_rate": 4.980673411125398e-05, "loss": 0.1603, "step": 2805 }, { "epoch": 17.5375, "grad_norm": 1.6328125, "learning_rate": 4.980641335339108e-05, "loss": 0.1768, "step": 2806 }, { "epoch": 17.54375, "grad_norm": 1.3203125, "learning_rate": 4.9806092330607065e-05, "loss": 0.1176, "step": 2807 }, { "epoch": 17.55, "grad_norm": 1.2734375, "learning_rate": 4.980577104290538e-05, "loss": 0.1083, "step": 2808 }, { "epoch": 17.55625, "grad_norm": 1.46875, "learning_rate": 4.980544949028945e-05, "loss": 0.1654, "step": 2809 }, { "epoch": 17.5625, "grad_norm": 1.3671875, "learning_rate": 4.980512767276271e-05, "loss": 0.1658, "step": 2810 }, { "epoch": 17.56875, "grad_norm": 1.7109375, "learning_rate": 4.980480559032861e-05, "loss": 0.1525, "step": 2811 }, { "epoch": 17.575, "grad_norm": 1.0546875, "learning_rate": 4.9804483242990565e-05, "loss": 0.092, "step": 2812 }, { "epoch": 17.58125, "grad_norm": 1.71875, "learning_rate": 4.9804160630752026e-05, "loss": 0.1813, "step": 2813 }, { "epoch": 17.5875, "grad_norm": 1.4453125, "learning_rate": 4.980383775361645e-05, "loss": 0.1454, "step": 2814 }, { "epoch": 17.59375, "grad_norm": 0.80078125, "learning_rate": 4.9803514611587275e-05, "loss": 0.0969, "step": 2815 }, { "epoch": 17.6, "grad_norm": 1.6953125, "learning_rate": 4.980319120466796e-05, "loss": 0.1763, "step": 2816 }, { "epoch": 17.60625, "grad_norm": 1.78125, "learning_rate": 4.980286753286195e-05, "loss": 0.1795, "step": 2817 }, { "epoch": 17.6125, "grad_norm": 1.28125, "learning_rate": 4.98025435961727e-05, "loss": 0.1336, "step": 2818 }, { "epoch": 17.61875, "grad_norm": 1.1484375, "learning_rate": 4.980221939460368e-05, "loss": 0.126, "step": 2819 }, { "epoch": 17.625, "grad_norm": 1.578125, "learning_rate": 4.9801894928158346e-05, "loss": 0.1629, "step": 2820 }, { "epoch": 17.63125, "grad_norm": 0.81640625, "learning_rate": 4.980157019684016e-05, "loss": 0.1052, "step": 2821 }, { "epoch": 17.6375, "grad_norm": 1.28125, "learning_rate": 4.98012452006526e-05, "loss": 0.1371, "step": 2822 }, { "epoch": 17.64375, "grad_norm": 1.59375, "learning_rate": 4.980091993959913e-05, "loss": 0.2144, "step": 2823 }, { "epoch": 17.65, "grad_norm": 1.5, "learning_rate": 4.980059441368323e-05, "loss": 0.2615, "step": 2824 }, { "epoch": 17.65625, "grad_norm": 1.5, "learning_rate": 4.9800268622908366e-05, "loss": 0.1555, "step": 2825 }, { "epoch": 17.6625, "grad_norm": 1.625, "learning_rate": 4.979994256727802e-05, "loss": 0.1622, "step": 2826 }, { "epoch": 17.66875, "grad_norm": 1.4296875, "learning_rate": 4.979961624679568e-05, "loss": 0.1916, "step": 2827 }, { "epoch": 17.675, "grad_norm": 0.89453125, "learning_rate": 4.979928966146482e-05, "loss": 0.0863, "step": 2828 }, { "epoch": 17.68125, "grad_norm": 1.2734375, "learning_rate": 4.979896281128894e-05, "loss": 0.1757, "step": 2829 }, { "epoch": 17.6875, "grad_norm": 1.53125, "learning_rate": 4.979863569627153e-05, "loss": 0.1982, "step": 2830 }, { "epoch": 17.69375, "grad_norm": 1.421875, "learning_rate": 4.979830831641608e-05, "loss": 0.1494, "step": 2831 }, { "epoch": 17.7, "grad_norm": 1.75, "learning_rate": 4.9797980671726074e-05, "loss": 0.2221, "step": 2832 }, { "epoch": 17.70625, "grad_norm": 1.03125, "learning_rate": 4.9797652762205025e-05, "loss": 0.1003, "step": 2833 }, { "epoch": 17.7125, "grad_norm": 1.25, "learning_rate": 4.979732458785644e-05, "loss": 0.1412, "step": 2834 }, { "epoch": 17.71875, "grad_norm": 1.5546875, "learning_rate": 4.979699614868381e-05, "loss": 0.1786, "step": 2835 }, { "epoch": 17.725, "grad_norm": 1.7109375, "learning_rate": 4.979666744469065e-05, "loss": 0.1433, "step": 2836 }, { "epoch": 17.73125, "grad_norm": 1.375, "learning_rate": 4.979633847588048e-05, "loss": 0.1359, "step": 2837 }, { "epoch": 17.7375, "grad_norm": 1.4609375, "learning_rate": 4.979600924225679e-05, "loss": 0.1976, "step": 2838 }, { "epoch": 17.74375, "grad_norm": 1.671875, "learning_rate": 4.979567974382311e-05, "loss": 0.1853, "step": 2839 }, { "epoch": 17.75, "grad_norm": 1.9921875, "learning_rate": 4.979534998058296e-05, "loss": 0.2081, "step": 2840 }, { "epoch": 17.75625, "grad_norm": 1.1484375, "learning_rate": 4.979501995253986e-05, "loss": 0.1224, "step": 2841 }, { "epoch": 17.7625, "grad_norm": 1.015625, "learning_rate": 4.979468965969732e-05, "loss": 0.1007, "step": 2842 }, { "epoch": 17.76875, "grad_norm": 1.4765625, "learning_rate": 4.9794359102058905e-05, "loss": 0.2069, "step": 2843 }, { "epoch": 17.775, "grad_norm": 1.46875, "learning_rate": 4.97940282796281e-05, "loss": 0.2187, "step": 2844 }, { "epoch": 17.78125, "grad_norm": 1.1953125, "learning_rate": 4.979369719240847e-05, "loss": 0.1531, "step": 2845 }, { "epoch": 17.7875, "grad_norm": 1.484375, "learning_rate": 4.9793365840403547e-05, "loss": 0.1359, "step": 2846 }, { "epoch": 17.79375, "grad_norm": 1.4375, "learning_rate": 4.979303422361685e-05, "loss": 0.1508, "step": 2847 }, { "epoch": 17.8, "grad_norm": 1.6640625, "learning_rate": 4.979270234205194e-05, "loss": 0.1503, "step": 2848 }, { "epoch": 17.80625, "grad_norm": 1.65625, "learning_rate": 4.979237019571235e-05, "loss": 0.1784, "step": 2849 }, { "epoch": 17.8125, "grad_norm": 1.0546875, "learning_rate": 4.979203778460163e-05, "loss": 0.0825, "step": 2850 }, { "epoch": 17.81875, "grad_norm": 1.171875, "learning_rate": 4.979170510872334e-05, "loss": 0.1154, "step": 2851 }, { "epoch": 17.825, "grad_norm": 1.890625, "learning_rate": 4.9791372168081026e-05, "loss": 0.1807, "step": 2852 }, { "epoch": 17.83125, "grad_norm": 1.6796875, "learning_rate": 4.9791038962678236e-05, "loss": 0.1451, "step": 2853 }, { "epoch": 17.8375, "grad_norm": 1.6484375, "learning_rate": 4.979070549251853e-05, "loss": 0.1968, "step": 2854 }, { "epoch": 17.84375, "grad_norm": 1.546875, "learning_rate": 4.979037175760548e-05, "loss": 0.1888, "step": 2855 }, { "epoch": 17.85, "grad_norm": 1.8359375, "learning_rate": 4.979003775794265e-05, "loss": 0.1884, "step": 2856 }, { "epoch": 17.85625, "grad_norm": 1.8359375, "learning_rate": 4.9789703493533596e-05, "loss": 0.1269, "step": 2857 }, { "epoch": 17.8625, "grad_norm": 1.9765625, "learning_rate": 4.978936896438189e-05, "loss": 0.1189, "step": 2858 }, { "epoch": 17.86875, "grad_norm": 1.3671875, "learning_rate": 4.978903417049111e-05, "loss": 0.1792, "step": 2859 }, { "epoch": 17.875, "grad_norm": 1.5546875, "learning_rate": 4.978869911186483e-05, "loss": 0.1664, "step": 2860 }, { "epoch": 17.88125, "grad_norm": 1.359375, "learning_rate": 4.9788363788506635e-05, "loss": 0.1315, "step": 2861 }, { "epoch": 17.8875, "grad_norm": 2.3125, "learning_rate": 4.9788028200420086e-05, "loss": 0.2028, "step": 2862 }, { "epoch": 17.89375, "grad_norm": 1.59375, "learning_rate": 4.9787692347608785e-05, "loss": 0.2107, "step": 2863 }, { "epoch": 17.9, "grad_norm": 1.4453125, "learning_rate": 4.978735623007631e-05, "loss": 0.1599, "step": 2864 }, { "epoch": 17.90625, "grad_norm": 1.2734375, "learning_rate": 4.978701984782625e-05, "loss": 0.1763, "step": 2865 }, { "epoch": 17.9125, "grad_norm": 1.203125, "learning_rate": 4.978668320086221e-05, "loss": 0.1275, "step": 2866 }, { "epoch": 17.91875, "grad_norm": 1.7890625, "learning_rate": 4.9786346289187776e-05, "loss": 0.2416, "step": 2867 }, { "epoch": 17.925, "grad_norm": 2.0625, "learning_rate": 4.978600911280654e-05, "loss": 0.2023, "step": 2868 }, { "epoch": 17.93125, "grad_norm": 1.5, "learning_rate": 4.978567167172211e-05, "loss": 0.2122, "step": 2869 }, { "epoch": 17.9375, "grad_norm": 1.4453125, "learning_rate": 4.9785333965938094e-05, "loss": 0.1752, "step": 2870 }, { "epoch": 17.94375, "grad_norm": 1.6171875, "learning_rate": 4.9784995995458095e-05, "loss": 0.1763, "step": 2871 }, { "epoch": 17.95, "grad_norm": 1.8515625, "learning_rate": 4.9784657760285715e-05, "loss": 0.206, "step": 2872 }, { "epoch": 17.95625, "grad_norm": 0.859375, "learning_rate": 4.978431926042457e-05, "loss": 0.082, "step": 2873 }, { "epoch": 17.9625, "grad_norm": 1.375, "learning_rate": 4.978398049587828e-05, "loss": 0.1257, "step": 2874 }, { "epoch": 17.96875, "grad_norm": 0.91796875, "learning_rate": 4.978364146665047e-05, "loss": 0.0813, "step": 2875 }, { "epoch": 17.975, "grad_norm": 1.6484375, "learning_rate": 4.978330217274474e-05, "loss": 0.2087, "step": 2876 }, { "epoch": 17.98125, "grad_norm": 0.89453125, "learning_rate": 4.978296261416473e-05, "loss": 0.0831, "step": 2877 }, { "epoch": 17.9875, "grad_norm": 1.3359375, "learning_rate": 4.978262279091406e-05, "loss": 0.1359, "step": 2878 }, { "epoch": 17.99375, "grad_norm": 1.6640625, "learning_rate": 4.978228270299635e-05, "loss": 0.1782, "step": 2879 }, { "epoch": 18.0, "grad_norm": 1.46875, "learning_rate": 4.978194235041525e-05, "loss": 0.1088, "step": 2880 }, { "epoch": 18.00625, "grad_norm": 1.71875, "learning_rate": 4.978160173317438e-05, "loss": 0.168, "step": 2881 }, { "epoch": 18.0125, "grad_norm": 0.953125, "learning_rate": 4.978126085127739e-05, "loss": 0.0977, "step": 2882 }, { "epoch": 18.01875, "grad_norm": 1.140625, "learning_rate": 4.978091970472791e-05, "loss": 0.1092, "step": 2883 }, { "epoch": 18.025, "grad_norm": 1.953125, "learning_rate": 4.97805782935296e-05, "loss": 0.2698, "step": 2884 }, { "epoch": 18.03125, "grad_norm": 1.7734375, "learning_rate": 4.9780236617686085e-05, "loss": 0.1474, "step": 2885 }, { "epoch": 18.0375, "grad_norm": 1.625, "learning_rate": 4.977989467720102e-05, "loss": 0.1752, "step": 2886 }, { "epoch": 18.04375, "grad_norm": 1.3359375, "learning_rate": 4.977955247207806e-05, "loss": 0.2124, "step": 2887 }, { "epoch": 18.05, "grad_norm": 1.3125, "learning_rate": 4.9779210002320864e-05, "loss": 0.1423, "step": 2888 }, { "epoch": 18.05625, "grad_norm": 1.6875, "learning_rate": 4.9778867267933084e-05, "loss": 0.1732, "step": 2889 }, { "epoch": 18.0625, "grad_norm": 1.671875, "learning_rate": 4.977852426891838e-05, "loss": 0.1715, "step": 2890 }, { "epoch": 18.06875, "grad_norm": 1.28125, "learning_rate": 4.977818100528041e-05, "loss": 0.17, "step": 2891 }, { "epoch": 18.075, "grad_norm": 1.265625, "learning_rate": 4.977783747702286e-05, "loss": 0.1132, "step": 2892 }, { "epoch": 18.08125, "grad_norm": 1.1875, "learning_rate": 4.9777493684149375e-05, "loss": 0.0967, "step": 2893 }, { "epoch": 18.0875, "grad_norm": 1.3515625, "learning_rate": 4.9777149626663633e-05, "loss": 0.1296, "step": 2894 }, { "epoch": 18.09375, "grad_norm": 1.6796875, "learning_rate": 4.977680530456932e-05, "loss": 0.1872, "step": 2895 }, { "epoch": 18.1, "grad_norm": 1.515625, "learning_rate": 4.97764607178701e-05, "loss": 0.1739, "step": 2896 }, { "epoch": 18.10625, "grad_norm": 1.671875, "learning_rate": 4.9776115866569654e-05, "loss": 0.1905, "step": 2897 }, { "epoch": 18.1125, "grad_norm": 1.453125, "learning_rate": 4.977577075067168e-05, "loss": 0.2214, "step": 2898 }, { "epoch": 18.11875, "grad_norm": 1.3125, "learning_rate": 4.977542537017984e-05, "loss": 0.1522, "step": 2899 }, { "epoch": 18.125, "grad_norm": 1.4609375, "learning_rate": 4.977507972509784e-05, "loss": 0.1427, "step": 2900 }, { "epoch": 18.13125, "grad_norm": 0.96484375, "learning_rate": 4.977473381542936e-05, "loss": 0.1007, "step": 2901 }, { "epoch": 18.1375, "grad_norm": 1.2734375, "learning_rate": 4.977438764117811e-05, "loss": 0.1646, "step": 2902 }, { "epoch": 18.14375, "grad_norm": 1.21875, "learning_rate": 4.9774041202347776e-05, "loss": 0.1094, "step": 2903 }, { "epoch": 18.15, "grad_norm": 1.859375, "learning_rate": 4.977369449894206e-05, "loss": 0.1673, "step": 2904 }, { "epoch": 18.15625, "grad_norm": 1.1875, "learning_rate": 4.977334753096466e-05, "loss": 0.1346, "step": 2905 }, { "epoch": 18.1625, "grad_norm": 1.203125, "learning_rate": 4.977300029841929e-05, "loss": 0.1193, "step": 2906 }, { "epoch": 18.16875, "grad_norm": 1.46875, "learning_rate": 4.977265280130964e-05, "loss": 0.1366, "step": 2907 }, { "epoch": 18.175, "grad_norm": 1.515625, "learning_rate": 4.977230503963945e-05, "loss": 0.1558, "step": 2908 }, { "epoch": 18.18125, "grad_norm": 1.5859375, "learning_rate": 4.9771957013412414e-05, "loss": 0.1646, "step": 2909 }, { "epoch": 18.1875, "grad_norm": 1.4375, "learning_rate": 4.9771608722632245e-05, "loss": 0.1819, "step": 2910 }, { "epoch": 18.19375, "grad_norm": 1.7578125, "learning_rate": 4.977126016730268e-05, "loss": 0.2325, "step": 2911 }, { "epoch": 18.2, "grad_norm": 1.6875, "learning_rate": 4.977091134742743e-05, "loss": 0.1313, "step": 2912 }, { "epoch": 18.20625, "grad_norm": 1.8828125, "learning_rate": 4.9770562263010225e-05, "loss": 0.2238, "step": 2913 }, { "epoch": 18.2125, "grad_norm": 1.3203125, "learning_rate": 4.977021291405479e-05, "loss": 0.1465, "step": 2914 }, { "epoch": 18.21875, "grad_norm": 1.828125, "learning_rate": 4.976986330056485e-05, "loss": 0.2021, "step": 2915 }, { "epoch": 18.225, "grad_norm": 1.4375, "learning_rate": 4.976951342254415e-05, "loss": 0.1641, "step": 2916 }, { "epoch": 18.23125, "grad_norm": 1.234375, "learning_rate": 4.976916327999642e-05, "loss": 0.1578, "step": 2917 }, { "epoch": 18.2375, "grad_norm": 1.0703125, "learning_rate": 4.9768812872925406e-05, "loss": 0.1477, "step": 2918 }, { "epoch": 18.24375, "grad_norm": 1.265625, "learning_rate": 4.9768462201334845e-05, "loss": 0.1141, "step": 2919 }, { "epoch": 18.25, "grad_norm": 1.234375, "learning_rate": 4.976811126522847e-05, "loss": 0.1239, "step": 2920 }, { "epoch": 18.25625, "grad_norm": 1.2109375, "learning_rate": 4.976776006461006e-05, "loss": 0.1215, "step": 2921 }, { "epoch": 18.2625, "grad_norm": 1.5703125, "learning_rate": 4.9767408599483336e-05, "loss": 0.1272, "step": 2922 }, { "epoch": 18.26875, "grad_norm": 1.53125, "learning_rate": 4.9767056869852055e-05, "loss": 0.1585, "step": 2923 }, { "epoch": 18.275, "grad_norm": 1.171875, "learning_rate": 4.976670487571999e-05, "loss": 0.117, "step": 2924 }, { "epoch": 18.28125, "grad_norm": 1.0859375, "learning_rate": 4.976635261709089e-05, "loss": 0.1127, "step": 2925 }, { "epoch": 18.2875, "grad_norm": 1.5625, "learning_rate": 4.976600009396852e-05, "loss": 0.229, "step": 2926 }, { "epoch": 18.29375, "grad_norm": 1.3515625, "learning_rate": 4.976564730635664e-05, "loss": 0.1684, "step": 2927 }, { "epoch": 18.3, "grad_norm": 1.3515625, "learning_rate": 4.976529425425902e-05, "loss": 0.1376, "step": 2928 }, { "epoch": 18.30625, "grad_norm": 1.828125, "learning_rate": 4.976494093767943e-05, "loss": 0.1941, "step": 2929 }, { "epoch": 18.3125, "grad_norm": 1.171875, "learning_rate": 4.976458735662164e-05, "loss": 0.1068, "step": 2930 }, { "epoch": 18.31875, "grad_norm": 1.84375, "learning_rate": 4.976423351108943e-05, "loss": 0.2194, "step": 2931 }, { "epoch": 18.325, "grad_norm": 1.328125, "learning_rate": 4.976387940108659e-05, "loss": 0.1373, "step": 2932 }, { "epoch": 18.33125, "grad_norm": 1.4765625, "learning_rate": 4.976352502661687e-05, "loss": 0.16, "step": 2933 }, { "epoch": 18.3375, "grad_norm": 1.5625, "learning_rate": 4.9763170387684096e-05, "loss": 0.1767, "step": 2934 }, { "epoch": 18.34375, "grad_norm": 1.65625, "learning_rate": 4.976281548429203e-05, "loss": 0.1807, "step": 2935 }, { "epoch": 18.35, "grad_norm": 1.140625, "learning_rate": 4.976246031644446e-05, "loss": 0.1156, "step": 2936 }, { "epoch": 18.35625, "grad_norm": 1.6328125, "learning_rate": 4.976210488414519e-05, "loss": 0.2527, "step": 2937 }, { "epoch": 18.3625, "grad_norm": 1.46875, "learning_rate": 4.976174918739801e-05, "loss": 0.1335, "step": 2938 }, { "epoch": 18.36875, "grad_norm": 1.8125, "learning_rate": 4.9761393226206723e-05, "loss": 0.215, "step": 2939 }, { "epoch": 18.375, "grad_norm": 1.3828125, "learning_rate": 4.9761037000575116e-05, "loss": 0.1252, "step": 2940 }, { "epoch": 18.38125, "grad_norm": 1.9765625, "learning_rate": 4.976068051050702e-05, "loss": 0.2581, "step": 2941 }, { "epoch": 18.3875, "grad_norm": 1.5703125, "learning_rate": 4.9760323756006214e-05, "loss": 0.1449, "step": 2942 }, { "epoch": 18.39375, "grad_norm": 0.99609375, "learning_rate": 4.975996673707653e-05, "loss": 0.111, "step": 2943 }, { "epoch": 18.4, "grad_norm": 1.203125, "learning_rate": 4.9759609453721765e-05, "loss": 0.1261, "step": 2944 }, { "epoch": 18.40625, "grad_norm": 1.46875, "learning_rate": 4.975925190594575e-05, "loss": 0.1379, "step": 2945 }, { "epoch": 18.4125, "grad_norm": 2.125, "learning_rate": 4.9758894093752296e-05, "loss": 0.2502, "step": 2946 }, { "epoch": 18.41875, "grad_norm": 1.1875, "learning_rate": 4.975853601714522e-05, "loss": 0.0927, "step": 2947 }, { "epoch": 18.425, "grad_norm": 1.796875, "learning_rate": 4.975817767612835e-05, "loss": 0.2649, "step": 2948 }, { "epoch": 18.43125, "grad_norm": 1.484375, "learning_rate": 4.9757819070705505e-05, "loss": 0.1946, "step": 2949 }, { "epoch": 18.4375, "grad_norm": 1.3984375, "learning_rate": 4.975746020088052e-05, "loss": 0.1242, "step": 2950 }, { "epoch": 18.44375, "grad_norm": 1.765625, "learning_rate": 4.975710106665724e-05, "loss": 0.228, "step": 2951 }, { "epoch": 18.45, "grad_norm": 1.828125, "learning_rate": 4.975674166803949e-05, "loss": 0.1976, "step": 2952 }, { "epoch": 18.45625, "grad_norm": 1.65625, "learning_rate": 4.9756382005031106e-05, "loss": 0.2363, "step": 2953 }, { "epoch": 18.4625, "grad_norm": 1.3515625, "learning_rate": 4.975602207763593e-05, "loss": 0.1555, "step": 2954 }, { "epoch": 18.46875, "grad_norm": 1.359375, "learning_rate": 4.975566188585781e-05, "loss": 0.1031, "step": 2955 }, { "epoch": 18.475, "grad_norm": 1.5390625, "learning_rate": 4.9755301429700584e-05, "loss": 0.1458, "step": 2956 }, { "epoch": 18.48125, "grad_norm": 1.609375, "learning_rate": 4.9754940709168106e-05, "loss": 0.1226, "step": 2957 }, { "epoch": 18.4875, "grad_norm": 1.3203125, "learning_rate": 4.975457972426424e-05, "loss": 0.1074, "step": 2958 }, { "epoch": 18.49375, "grad_norm": 1.3046875, "learning_rate": 4.975421847499282e-05, "loss": 0.1513, "step": 2959 }, { "epoch": 18.5, "grad_norm": 1.6484375, "learning_rate": 4.975385696135773e-05, "loss": 0.1814, "step": 2960 }, { "epoch": 18.50625, "grad_norm": 1.171875, "learning_rate": 4.9753495183362796e-05, "loss": 0.103, "step": 2961 }, { "epoch": 18.5125, "grad_norm": 1.3046875, "learning_rate": 4.975313314101191e-05, "loss": 0.1535, "step": 2962 }, { "epoch": 18.51875, "grad_norm": 1.1015625, "learning_rate": 4.975277083430893e-05, "loss": 0.0927, "step": 2963 }, { "epoch": 18.525, "grad_norm": 1.7734375, "learning_rate": 4.975240826325772e-05, "loss": 0.1684, "step": 2964 }, { "epoch": 18.53125, "grad_norm": 1.53125, "learning_rate": 4.975204542786216e-05, "loss": 0.158, "step": 2965 }, { "epoch": 18.5375, "grad_norm": 1.28125, "learning_rate": 4.975168232812613e-05, "loss": 0.1187, "step": 2966 }, { "epoch": 18.54375, "grad_norm": 1.2421875, "learning_rate": 4.975131896405349e-05, "loss": 0.1555, "step": 2967 }, { "epoch": 18.55, "grad_norm": 1.9296875, "learning_rate": 4.975095533564814e-05, "loss": 0.1891, "step": 2968 }, { "epoch": 18.55625, "grad_norm": 1.8046875, "learning_rate": 4.975059144291394e-05, "loss": 0.2039, "step": 2969 }, { "epoch": 18.5625, "grad_norm": 1.703125, "learning_rate": 4.97502272858548e-05, "loss": 0.189, "step": 2970 }, { "epoch": 18.56875, "grad_norm": 1.5390625, "learning_rate": 4.9749862864474595e-05, "loss": 0.1667, "step": 2971 }, { "epoch": 18.575, "grad_norm": 1.3515625, "learning_rate": 4.974949817877722e-05, "loss": 0.1289, "step": 2972 }, { "epoch": 18.58125, "grad_norm": 1.4296875, "learning_rate": 4.974913322876657e-05, "loss": 0.1406, "step": 2973 }, { "epoch": 18.5875, "grad_norm": 1.2421875, "learning_rate": 4.9748768014446546e-05, "loss": 0.1107, "step": 2974 }, { "epoch": 18.59375, "grad_norm": 2.0, "learning_rate": 4.974840253582104e-05, "loss": 0.1853, "step": 2975 }, { "epoch": 18.6, "grad_norm": 1.0546875, "learning_rate": 4.9748036792893966e-05, "loss": 0.0919, "step": 2976 }, { "epoch": 18.60625, "grad_norm": 1.3046875, "learning_rate": 4.974767078566922e-05, "loss": 0.1635, "step": 2977 }, { "epoch": 18.6125, "grad_norm": 1.359375, "learning_rate": 4.9747304514150715e-05, "loss": 0.1538, "step": 2978 }, { "epoch": 18.61875, "grad_norm": 1.46875, "learning_rate": 4.9746937978342364e-05, "loss": 0.1607, "step": 2979 }, { "epoch": 18.625, "grad_norm": 1.421875, "learning_rate": 4.974657117824808e-05, "loss": 0.137, "step": 2980 }, { "epoch": 18.63125, "grad_norm": 1.3671875, "learning_rate": 4.974620411387179e-05, "loss": 0.1259, "step": 2981 }, { "epoch": 18.6375, "grad_norm": 1.3515625, "learning_rate": 4.974583678521739e-05, "loss": 0.1568, "step": 2982 }, { "epoch": 18.64375, "grad_norm": 1.0859375, "learning_rate": 4.974546919228882e-05, "loss": 0.0901, "step": 2983 }, { "epoch": 18.65, "grad_norm": 1.5078125, "learning_rate": 4.9745101335090003e-05, "loss": 0.1328, "step": 2984 }, { "epoch": 18.65625, "grad_norm": 1.5, "learning_rate": 4.974473321362487e-05, "loss": 0.2522, "step": 2985 }, { "epoch": 18.6625, "grad_norm": 1.421875, "learning_rate": 4.974436482789735e-05, "loss": 0.14, "step": 2986 }, { "epoch": 18.66875, "grad_norm": 1.796875, "learning_rate": 4.9743996177911375e-05, "loss": 0.2085, "step": 2987 }, { "epoch": 18.675, "grad_norm": 1.9296875, "learning_rate": 4.974362726367089e-05, "loss": 0.2263, "step": 2988 }, { "epoch": 18.68125, "grad_norm": 1.7109375, "learning_rate": 4.974325808517982e-05, "loss": 0.1367, "step": 2989 }, { "epoch": 18.6875, "grad_norm": 1.4453125, "learning_rate": 4.974288864244213e-05, "loss": 0.1583, "step": 2990 }, { "epoch": 18.69375, "grad_norm": 1.78125, "learning_rate": 4.974251893546174e-05, "loss": 0.2481, "step": 2991 }, { "epoch": 18.7, "grad_norm": 1.03125, "learning_rate": 4.974214896424262e-05, "loss": 0.098, "step": 2992 }, { "epoch": 18.70625, "grad_norm": 1.2265625, "learning_rate": 4.9741778728788705e-05, "loss": 0.1677, "step": 2993 }, { "epoch": 18.7125, "grad_norm": 1.1640625, "learning_rate": 4.9741408229103953e-05, "loss": 0.1529, "step": 2994 }, { "epoch": 18.71875, "grad_norm": 2.015625, "learning_rate": 4.974103746519233e-05, "loss": 0.2727, "step": 2995 }, { "epoch": 18.725, "grad_norm": 1.3203125, "learning_rate": 4.974066643705778e-05, "loss": 0.1504, "step": 2996 }, { "epoch": 18.73125, "grad_norm": 1.734375, "learning_rate": 4.9740295144704285e-05, "loss": 0.1737, "step": 2997 }, { "epoch": 18.7375, "grad_norm": 1.8359375, "learning_rate": 4.973992358813579e-05, "loss": 0.1739, "step": 2998 }, { "epoch": 18.74375, "grad_norm": 1.859375, "learning_rate": 4.973955176735628e-05, "loss": 0.3139, "step": 2999 }, { "epoch": 18.75, "grad_norm": 1.34375, "learning_rate": 4.973917968236971e-05, "loss": 0.1259, "step": 3000 }, { "epoch": 18.75625, "grad_norm": 1.375, "learning_rate": 4.973880733318007e-05, "loss": 0.1499, "step": 3001 }, { "epoch": 18.7625, "grad_norm": 1.515625, "learning_rate": 4.973843471979133e-05, "loss": 0.1667, "step": 3002 }, { "epoch": 18.76875, "grad_norm": 1.078125, "learning_rate": 4.973806184220746e-05, "loss": 0.1027, "step": 3003 }, { "epoch": 18.775, "grad_norm": 1.609375, "learning_rate": 4.973768870043245e-05, "loss": 0.1756, "step": 3004 }, { "epoch": 18.78125, "grad_norm": 1.71875, "learning_rate": 4.9737315294470296e-05, "loss": 0.1572, "step": 3005 }, { "epoch": 18.7875, "grad_norm": 0.98046875, "learning_rate": 4.973694162432497e-05, "loss": 0.0861, "step": 3006 }, { "epoch": 18.79375, "grad_norm": 1.7421875, "learning_rate": 4.973656769000046e-05, "loss": 0.2654, "step": 3007 }, { "epoch": 18.8, "grad_norm": 1.609375, "learning_rate": 4.9736193491500775e-05, "loss": 0.1767, "step": 3008 }, { "epoch": 18.80625, "grad_norm": 1.078125, "learning_rate": 4.97358190288299e-05, "loss": 0.1107, "step": 3009 }, { "epoch": 18.8125, "grad_norm": 1.265625, "learning_rate": 4.973544430199184e-05, "loss": 0.1265, "step": 3010 }, { "epoch": 18.81875, "grad_norm": 1.3828125, "learning_rate": 4.973506931099059e-05, "loss": 0.1504, "step": 3011 }, { "epoch": 18.825, "grad_norm": 1.46875, "learning_rate": 4.9734694055830164e-05, "loss": 0.1849, "step": 3012 }, { "epoch": 18.83125, "grad_norm": 1.7265625, "learning_rate": 4.9734318536514564e-05, "loss": 0.1593, "step": 3013 }, { "epoch": 18.8375, "grad_norm": 1.6171875, "learning_rate": 4.97339427530478e-05, "loss": 0.1606, "step": 3014 }, { "epoch": 18.84375, "grad_norm": 1.734375, "learning_rate": 4.973356670543389e-05, "loss": 0.1866, "step": 3015 }, { "epoch": 18.85, "grad_norm": 1.640625, "learning_rate": 4.973319039367683e-05, "loss": 0.1806, "step": 3016 }, { "epoch": 18.85625, "grad_norm": 1.2578125, "learning_rate": 4.973281381778067e-05, "loss": 0.1523, "step": 3017 }, { "epoch": 18.8625, "grad_norm": 1.25, "learning_rate": 4.973243697774941e-05, "loss": 0.1067, "step": 3018 }, { "epoch": 18.86875, "grad_norm": 1.9140625, "learning_rate": 4.973205987358709e-05, "loss": 0.2506, "step": 3019 }, { "epoch": 18.875, "grad_norm": 1.359375, "learning_rate": 4.973168250529772e-05, "loss": 0.1253, "step": 3020 }, { "epoch": 18.88125, "grad_norm": 1.609375, "learning_rate": 4.973130487288534e-05, "loss": 0.1963, "step": 3021 }, { "epoch": 18.8875, "grad_norm": 1.4453125, "learning_rate": 4.973092697635399e-05, "loss": 0.1899, "step": 3022 }, { "epoch": 18.89375, "grad_norm": 1.1640625, "learning_rate": 4.973054881570768e-05, "loss": 0.0988, "step": 3023 }, { "epoch": 18.9, "grad_norm": 1.765625, "learning_rate": 4.973017039095048e-05, "loss": 0.1912, "step": 3024 }, { "epoch": 18.90625, "grad_norm": 1.7734375, "learning_rate": 4.9729791702086414e-05, "loss": 0.2408, "step": 3025 }, { "epoch": 18.9125, "grad_norm": 1.4375, "learning_rate": 4.972941274911953e-05, "loss": 0.1468, "step": 3026 }, { "epoch": 18.91875, "grad_norm": 0.734375, "learning_rate": 4.972903353205387e-05, "loss": 0.071, "step": 3027 }, { "epoch": 18.925, "grad_norm": 1.890625, "learning_rate": 4.972865405089349e-05, "loss": 0.2052, "step": 3028 }, { "epoch": 18.93125, "grad_norm": 1.1171875, "learning_rate": 4.972827430564244e-05, "loss": 0.0968, "step": 3029 }, { "epoch": 18.9375, "grad_norm": 1.0703125, "learning_rate": 4.9727894296304775e-05, "loss": 0.1203, "step": 3030 }, { "epoch": 18.94375, "grad_norm": 1.609375, "learning_rate": 4.9727514022884566e-05, "loss": 0.1916, "step": 3031 }, { "epoch": 18.95, "grad_norm": 1.515625, "learning_rate": 4.972713348538586e-05, "loss": 0.1569, "step": 3032 }, { "epoch": 18.95625, "grad_norm": 1.4296875, "learning_rate": 4.9726752683812715e-05, "loss": 0.1348, "step": 3033 }, { "epoch": 18.9625, "grad_norm": 1.453125, "learning_rate": 4.972637161816921e-05, "loss": 0.1399, "step": 3034 }, { "epoch": 18.96875, "grad_norm": 1.5, "learning_rate": 4.972599028845941e-05, "loss": 0.1773, "step": 3035 }, { "epoch": 18.975, "grad_norm": 0.875, "learning_rate": 4.972560869468739e-05, "loss": 0.0863, "step": 3036 }, { "epoch": 18.98125, "grad_norm": 1.3515625, "learning_rate": 4.972522683685723e-05, "loss": 0.1844, "step": 3037 }, { "epoch": 18.9875, "grad_norm": 1.640625, "learning_rate": 4.9724844714973e-05, "loss": 0.2799, "step": 3038 }, { "epoch": 18.99375, "grad_norm": 1.375, "learning_rate": 4.972446232903878e-05, "loss": 0.1361, "step": 3039 }, { "epoch": 19.0, "grad_norm": 1.7890625, "learning_rate": 4.972407967905865e-05, "loss": 0.1845, "step": 3040 }, { "epoch": 19.00625, "grad_norm": 1.4453125, "learning_rate": 4.972369676503672e-05, "loss": 0.1267, "step": 3041 }, { "epoch": 19.0125, "grad_norm": 1.5625, "learning_rate": 4.972331358697705e-05, "loss": 0.2637, "step": 3042 }, { "epoch": 19.01875, "grad_norm": 1.375, "learning_rate": 4.9722930144883746e-05, "loss": 0.1255, "step": 3043 }, { "epoch": 19.025, "grad_norm": 1.421875, "learning_rate": 4.97225464387609e-05, "loss": 0.1758, "step": 3044 }, { "epoch": 19.03125, "grad_norm": 1.671875, "learning_rate": 4.972216246861262e-05, "loss": 0.1493, "step": 3045 }, { "epoch": 19.0375, "grad_norm": 2.09375, "learning_rate": 4.972177823444299e-05, "loss": 0.2178, "step": 3046 }, { "epoch": 19.04375, "grad_norm": 2.296875, "learning_rate": 4.972139373625613e-05, "loss": 0.2559, "step": 3047 }, { "epoch": 19.05, "grad_norm": 1.796875, "learning_rate": 4.9721008974056125e-05, "loss": 0.1547, "step": 3048 }, { "epoch": 19.05625, "grad_norm": 1.5546875, "learning_rate": 4.97206239478471e-05, "loss": 0.1417, "step": 3049 }, { "epoch": 19.0625, "grad_norm": 1.875, "learning_rate": 4.972023865763318e-05, "loss": 0.2538, "step": 3050 }, { "epoch": 19.06875, "grad_norm": 1.0859375, "learning_rate": 4.971985310341844e-05, "loss": 0.1052, "step": 3051 }, { "epoch": 19.075, "grad_norm": 1.3828125, "learning_rate": 4.971946728520704e-05, "loss": 0.1338, "step": 3052 }, { "epoch": 19.08125, "grad_norm": 1.171875, "learning_rate": 4.971908120300307e-05, "loss": 0.1181, "step": 3053 }, { "epoch": 19.0875, "grad_norm": 1.3828125, "learning_rate": 4.9718694856810675e-05, "loss": 0.1645, "step": 3054 }, { "epoch": 19.09375, "grad_norm": 1.0546875, "learning_rate": 4.971830824663396e-05, "loss": 0.1035, "step": 3055 }, { "epoch": 19.1, "grad_norm": 1.4453125, "learning_rate": 4.971792137247707e-05, "loss": 0.1436, "step": 3056 }, { "epoch": 19.10625, "grad_norm": 1.1171875, "learning_rate": 4.971753423434413e-05, "loss": 0.0956, "step": 3057 }, { "epoch": 19.1125, "grad_norm": 1.03125, "learning_rate": 4.971714683223928e-05, "loss": 0.1517, "step": 3058 }, { "epoch": 19.11875, "grad_norm": 1.6953125, "learning_rate": 4.971675916616665e-05, "loss": 0.163, "step": 3059 }, { "epoch": 19.125, "grad_norm": 1.3984375, "learning_rate": 4.9716371236130374e-05, "loss": 0.1321, "step": 3060 }, { "epoch": 19.13125, "grad_norm": 1.1796875, "learning_rate": 4.971598304213462e-05, "loss": 0.1052, "step": 3061 }, { "epoch": 19.1375, "grad_norm": 1.4921875, "learning_rate": 4.9715594584183505e-05, "loss": 0.1522, "step": 3062 }, { "epoch": 19.14375, "grad_norm": 1.078125, "learning_rate": 4.9715205862281195e-05, "loss": 0.1122, "step": 3063 }, { "epoch": 19.15, "grad_norm": 1.0625, "learning_rate": 4.971481687643184e-05, "loss": 0.1217, "step": 3064 }, { "epoch": 19.15625, "grad_norm": 0.91015625, "learning_rate": 4.971442762663958e-05, "loss": 0.0936, "step": 3065 }, { "epoch": 19.1625, "grad_norm": 1.328125, "learning_rate": 4.9714038112908587e-05, "loss": 0.1412, "step": 3066 }, { "epoch": 19.16875, "grad_norm": 1.078125, "learning_rate": 4.9713648335243026e-05, "loss": 0.1082, "step": 3067 }, { "epoch": 19.175, "grad_norm": 1.5390625, "learning_rate": 4.971325829364703e-05, "loss": 0.1571, "step": 3068 }, { "epoch": 19.18125, "grad_norm": 1.890625, "learning_rate": 4.9712867988124804e-05, "loss": 0.2729, "step": 3069 }, { "epoch": 19.1875, "grad_norm": 1.2421875, "learning_rate": 4.971247741868049e-05, "loss": 0.1375, "step": 3070 }, { "epoch": 19.19375, "grad_norm": 1.078125, "learning_rate": 4.971208658531827e-05, "loss": 0.1503, "step": 3071 }, { "epoch": 19.2, "grad_norm": 1.7890625, "learning_rate": 4.971169548804231e-05, "loss": 0.2, "step": 3072 }, { "epoch": 19.20625, "grad_norm": 1.375, "learning_rate": 4.971130412685679e-05, "loss": 0.1847, "step": 3073 }, { "epoch": 19.2125, "grad_norm": 0.84765625, "learning_rate": 4.97109125017659e-05, "loss": 0.0913, "step": 3074 }, { "epoch": 19.21875, "grad_norm": 2.015625, "learning_rate": 4.9710520612773794e-05, "loss": 0.273, "step": 3075 }, { "epoch": 19.225, "grad_norm": 1.5546875, "learning_rate": 4.971012845988469e-05, "loss": 0.1914, "step": 3076 }, { "epoch": 19.23125, "grad_norm": 1.3125, "learning_rate": 4.970973604310276e-05, "loss": 0.1314, "step": 3077 }, { "epoch": 19.2375, "grad_norm": 1.203125, "learning_rate": 4.970934336243219e-05, "loss": 0.1654, "step": 3078 }, { "epoch": 19.24375, "grad_norm": 1.609375, "learning_rate": 4.970895041787719e-05, "loss": 0.1839, "step": 3079 }, { "epoch": 19.25, "grad_norm": 1.6484375, "learning_rate": 4.9708557209441937e-05, "loss": 0.2676, "step": 3080 }, { "epoch": 19.25625, "grad_norm": 1.5, "learning_rate": 4.9708163737130644e-05, "loss": 0.2024, "step": 3081 }, { "epoch": 19.2625, "grad_norm": 1.2890625, "learning_rate": 4.970777000094751e-05, "loss": 0.2007, "step": 3082 }, { "epoch": 19.26875, "grad_norm": 1.5, "learning_rate": 4.9707376000896736e-05, "loss": 0.1327, "step": 3083 }, { "epoch": 19.275, "grad_norm": 0.87109375, "learning_rate": 4.9706981736982535e-05, "loss": 0.1168, "step": 3084 }, { "epoch": 19.28125, "grad_norm": 1.9765625, "learning_rate": 4.9706587209209115e-05, "loss": 0.1938, "step": 3085 }, { "epoch": 19.2875, "grad_norm": 1.8984375, "learning_rate": 4.970619241758069e-05, "loss": 0.196, "step": 3086 }, { "epoch": 19.29375, "grad_norm": 1.3359375, "learning_rate": 4.970579736210148e-05, "loss": 0.1078, "step": 3087 }, { "epoch": 19.3, "grad_norm": 1.1171875, "learning_rate": 4.97054020427757e-05, "loss": 0.1098, "step": 3088 }, { "epoch": 19.30625, "grad_norm": 1.40625, "learning_rate": 4.970500645960756e-05, "loss": 0.1316, "step": 3089 }, { "epoch": 19.3125, "grad_norm": 1.578125, "learning_rate": 4.970461061260131e-05, "loss": 0.1502, "step": 3090 }, { "epoch": 19.31875, "grad_norm": 1.90625, "learning_rate": 4.970421450176116e-05, "loss": 0.1719, "step": 3091 }, { "epoch": 19.325, "grad_norm": 1.296875, "learning_rate": 4.970381812709134e-05, "loss": 0.1647, "step": 3092 }, { "epoch": 19.33125, "grad_norm": 1.828125, "learning_rate": 4.970342148859609e-05, "loss": 0.1624, "step": 3093 }, { "epoch": 19.3375, "grad_norm": 1.484375, "learning_rate": 4.970302458627964e-05, "loss": 0.21, "step": 3094 }, { "epoch": 19.34375, "grad_norm": 1.4765625, "learning_rate": 4.970262742014624e-05, "loss": 0.1697, "step": 3095 }, { "epoch": 19.35, "grad_norm": 1.875, "learning_rate": 4.970222999020012e-05, "loss": 0.1775, "step": 3096 }, { "epoch": 19.35625, "grad_norm": 1.4609375, "learning_rate": 4.9701832296445526e-05, "loss": 0.1754, "step": 3097 }, { "epoch": 19.3625, "grad_norm": 1.296875, "learning_rate": 4.970143433888671e-05, "loss": 0.13, "step": 3098 }, { "epoch": 19.36875, "grad_norm": 1.046875, "learning_rate": 4.970103611752791e-05, "loss": 0.0956, "step": 3099 }, { "epoch": 19.375, "grad_norm": 0.90234375, "learning_rate": 4.97006376323734e-05, "loss": 0.0764, "step": 3100 }, { "epoch": 19.38125, "grad_norm": 1.1171875, "learning_rate": 4.970023888342742e-05, "loss": 0.1172, "step": 3101 }, { "epoch": 19.3875, "grad_norm": 1.5859375, "learning_rate": 4.9699839870694225e-05, "loss": 0.2057, "step": 3102 }, { "epoch": 19.39375, "grad_norm": 1.4375, "learning_rate": 4.969944059417809e-05, "loss": 0.1299, "step": 3103 }, { "epoch": 19.4, "grad_norm": 1.3046875, "learning_rate": 4.9699041053883274e-05, "loss": 0.1789, "step": 3104 }, { "epoch": 19.40625, "grad_norm": 1.71875, "learning_rate": 4.9698641249814036e-05, "loss": 0.2264, "step": 3105 }, { "epoch": 19.4125, "grad_norm": 1.2578125, "learning_rate": 4.969824118197466e-05, "loss": 0.1516, "step": 3106 }, { "epoch": 19.41875, "grad_norm": 1.4375, "learning_rate": 4.96978408503694e-05, "loss": 0.129, "step": 3107 }, { "epoch": 19.425, "grad_norm": 1.6953125, "learning_rate": 4.969744025500255e-05, "loss": 0.1323, "step": 3108 }, { "epoch": 19.43125, "grad_norm": 1.171875, "learning_rate": 4.9697039395878373e-05, "loss": 0.1165, "step": 3109 }, { "epoch": 19.4375, "grad_norm": 1.171875, "learning_rate": 4.9696638273001164e-05, "loss": 0.1357, "step": 3110 }, { "epoch": 19.44375, "grad_norm": 1.1796875, "learning_rate": 4.9696236886375194e-05, "loss": 0.1644, "step": 3111 }, { "epoch": 19.45, "grad_norm": 2.03125, "learning_rate": 4.969583523600476e-05, "loss": 0.26, "step": 3112 }, { "epoch": 19.45625, "grad_norm": 1.4765625, "learning_rate": 4.9695433321894145e-05, "loss": 0.1238, "step": 3113 }, { "epoch": 19.4625, "grad_norm": 1.21875, "learning_rate": 4.9695031144047645e-05, "loss": 0.1065, "step": 3114 }, { "epoch": 19.46875, "grad_norm": 1.015625, "learning_rate": 4.969462870246955e-05, "loss": 0.0965, "step": 3115 }, { "epoch": 19.475, "grad_norm": 0.91015625, "learning_rate": 4.969422599716417e-05, "loss": 0.0842, "step": 3116 }, { "epoch": 19.48125, "grad_norm": 1.5078125, "learning_rate": 4.969382302813579e-05, "loss": 0.1865, "step": 3117 }, { "epoch": 19.4875, "grad_norm": 1.4765625, "learning_rate": 4.969341979538872e-05, "loss": 0.1552, "step": 3118 }, { "epoch": 19.49375, "grad_norm": 0.94921875, "learning_rate": 4.969301629892726e-05, "loss": 0.0933, "step": 3119 }, { "epoch": 19.5, "grad_norm": 1.640625, "learning_rate": 4.969261253875573e-05, "loss": 0.1992, "step": 3120 }, { "epoch": 19.50625, "grad_norm": 1.6015625, "learning_rate": 4.9692208514878444e-05, "loss": 0.2083, "step": 3121 }, { "epoch": 19.5125, "grad_norm": 1.484375, "learning_rate": 4.969180422729971e-05, "loss": 0.1288, "step": 3122 }, { "epoch": 19.51875, "grad_norm": 1.609375, "learning_rate": 4.969139967602384e-05, "loss": 0.2029, "step": 3123 }, { "epoch": 19.525, "grad_norm": 1.21875, "learning_rate": 4.9690994861055164e-05, "loss": 0.1201, "step": 3124 }, { "epoch": 19.53125, "grad_norm": 1.6171875, "learning_rate": 4.9690589782398e-05, "loss": 0.1407, "step": 3125 }, { "epoch": 19.5375, "grad_norm": 1.359375, "learning_rate": 4.9690184440056686e-05, "loss": 0.1277, "step": 3126 }, { "epoch": 19.54375, "grad_norm": 0.83203125, "learning_rate": 4.9689778834035524e-05, "loss": 0.0854, "step": 3127 }, { "epoch": 19.55, "grad_norm": 1.5078125, "learning_rate": 4.9689372964338875e-05, "loss": 0.1478, "step": 3128 }, { "epoch": 19.55625, "grad_norm": 1.2578125, "learning_rate": 4.968896683097106e-05, "loss": 0.1095, "step": 3129 }, { "epoch": 19.5625, "grad_norm": 1.6171875, "learning_rate": 4.968856043393642e-05, "loss": 0.159, "step": 3130 }, { "epoch": 19.56875, "grad_norm": 1.15625, "learning_rate": 4.9688153773239275e-05, "loss": 0.1318, "step": 3131 }, { "epoch": 19.575, "grad_norm": 1.1640625, "learning_rate": 4.968774684888401e-05, "loss": 0.0994, "step": 3132 }, { "epoch": 19.58125, "grad_norm": 1.7578125, "learning_rate": 4.968733966087493e-05, "loss": 0.1243, "step": 3133 }, { "epoch": 19.5875, "grad_norm": 1.578125, "learning_rate": 4.9686932209216394e-05, "loss": 0.1857, "step": 3134 }, { "epoch": 19.59375, "grad_norm": 1.28125, "learning_rate": 4.968652449391277e-05, "loss": 0.1247, "step": 3135 }, { "epoch": 19.6, "grad_norm": 1.3203125, "learning_rate": 4.968611651496841e-05, "loss": 0.1613, "step": 3136 }, { "epoch": 19.60625, "grad_norm": 1.7578125, "learning_rate": 4.968570827238764e-05, "loss": 0.1556, "step": 3137 }, { "epoch": 19.6125, "grad_norm": 1.140625, "learning_rate": 4.968529976617486e-05, "loss": 0.1036, "step": 3138 }, { "epoch": 19.61875, "grad_norm": 1.5, "learning_rate": 4.968489099633441e-05, "loss": 0.2005, "step": 3139 }, { "epoch": 19.625, "grad_norm": 1.3984375, "learning_rate": 4.9684481962870656e-05, "loss": 0.1361, "step": 3140 }, { "epoch": 19.63125, "grad_norm": 1.2734375, "learning_rate": 4.968407266578797e-05, "loss": 0.1354, "step": 3141 }, { "epoch": 19.6375, "grad_norm": 1.5078125, "learning_rate": 4.968366310509074e-05, "loss": 0.1641, "step": 3142 }, { "epoch": 19.64375, "grad_norm": 1.671875, "learning_rate": 4.96832532807833e-05, "loss": 0.1418, "step": 3143 }, { "epoch": 19.65, "grad_norm": 1.2109375, "learning_rate": 4.968284319287007e-05, "loss": 0.1603, "step": 3144 }, { "epoch": 19.65625, "grad_norm": 1.4765625, "learning_rate": 4.96824328413554e-05, "loss": 0.1523, "step": 3145 }, { "epoch": 19.6625, "grad_norm": 0.80078125, "learning_rate": 4.968202222624368e-05, "loss": 0.0851, "step": 3146 }, { "epoch": 19.66875, "grad_norm": 1.109375, "learning_rate": 4.9681611347539304e-05, "loss": 0.1579, "step": 3147 }, { "epoch": 19.675, "grad_norm": 1.1015625, "learning_rate": 4.9681200205246655e-05, "loss": 0.1139, "step": 3148 }, { "epoch": 19.68125, "grad_norm": 1.203125, "learning_rate": 4.968078879937011e-05, "loss": 0.1411, "step": 3149 }, { "epoch": 19.6875, "grad_norm": 1.1015625, "learning_rate": 4.968037712991409e-05, "loss": 0.121, "step": 3150 }, { "epoch": 19.69375, "grad_norm": 1.296875, "learning_rate": 4.967996519688297e-05, "loss": 0.1214, "step": 3151 }, { "epoch": 19.7, "grad_norm": 1.234375, "learning_rate": 4.9679553000281157e-05, "loss": 0.1661, "step": 3152 }, { "epoch": 19.70625, "grad_norm": 1.203125, "learning_rate": 4.967914054011305e-05, "loss": 0.1288, "step": 3153 }, { "epoch": 19.7125, "grad_norm": 1.21875, "learning_rate": 4.967872781638305e-05, "loss": 0.1243, "step": 3154 }, { "epoch": 19.71875, "grad_norm": 1.6796875, "learning_rate": 4.967831482909559e-05, "loss": 0.1585, "step": 3155 }, { "epoch": 19.725, "grad_norm": 1.453125, "learning_rate": 4.967790157825504e-05, "loss": 0.14, "step": 3156 }, { "epoch": 19.73125, "grad_norm": 1.9921875, "learning_rate": 4.967748806386584e-05, "loss": 0.1837, "step": 3157 }, { "epoch": 19.7375, "grad_norm": 1.3203125, "learning_rate": 4.96770742859324e-05, "loss": 0.1338, "step": 3158 }, { "epoch": 19.74375, "grad_norm": 1.2734375, "learning_rate": 4.967666024445914e-05, "loss": 0.1189, "step": 3159 }, { "epoch": 19.75, "grad_norm": 1.625, "learning_rate": 4.967624593945048e-05, "loss": 0.1568, "step": 3160 }, { "epoch": 19.75625, "grad_norm": 1.84375, "learning_rate": 4.967583137091085e-05, "loss": 0.2562, "step": 3161 }, { "epoch": 19.7625, "grad_norm": 1.0234375, "learning_rate": 4.967541653884467e-05, "loss": 0.1245, "step": 3162 }, { "epoch": 19.76875, "grad_norm": 1.25, "learning_rate": 4.967500144325638e-05, "loss": 0.0992, "step": 3163 }, { "epoch": 19.775, "grad_norm": 1.1640625, "learning_rate": 4.9674586084150394e-05, "loss": 0.1071, "step": 3164 }, { "epoch": 19.78125, "grad_norm": 1.2421875, "learning_rate": 4.967417046153117e-05, "loss": 0.1227, "step": 3165 }, { "epoch": 19.7875, "grad_norm": 1.578125, "learning_rate": 4.967375457540313e-05, "loss": 0.2195, "step": 3166 }, { "epoch": 19.79375, "grad_norm": 1.6015625, "learning_rate": 4.967333842577072e-05, "loss": 0.1642, "step": 3167 }, { "epoch": 19.8, "grad_norm": 0.8828125, "learning_rate": 4.967292201263839e-05, "loss": 0.0836, "step": 3168 }, { "epoch": 19.80625, "grad_norm": 1.125, "learning_rate": 4.967250533601059e-05, "loss": 0.1183, "step": 3169 }, { "epoch": 19.8125, "grad_norm": 1.1875, "learning_rate": 4.967208839589176e-05, "loss": 0.1688, "step": 3170 }, { "epoch": 19.81875, "grad_norm": 1.21875, "learning_rate": 4.9671671192286354e-05, "loss": 0.1247, "step": 3171 }, { "epoch": 19.825, "grad_norm": 1.484375, "learning_rate": 4.967125372519883e-05, "loss": 0.2155, "step": 3172 }, { "epoch": 19.83125, "grad_norm": 1.3359375, "learning_rate": 4.9670835994633644e-05, "loss": 0.1426, "step": 3173 }, { "epoch": 19.8375, "grad_norm": 1.1640625, "learning_rate": 4.967041800059526e-05, "loss": 0.122, "step": 3174 }, { "epoch": 19.84375, "grad_norm": 1.203125, "learning_rate": 4.966999974308815e-05, "loss": 0.1338, "step": 3175 }, { "epoch": 19.85, "grad_norm": 1.4921875, "learning_rate": 4.9669581222116756e-05, "loss": 0.1155, "step": 3176 }, { "epoch": 19.85625, "grad_norm": 1.671875, "learning_rate": 4.966916243768558e-05, "loss": 0.3056, "step": 3177 }, { "epoch": 19.8625, "grad_norm": 1.2578125, "learning_rate": 4.9668743389799065e-05, "loss": 0.1115, "step": 3178 }, { "epoch": 19.86875, "grad_norm": 1.328125, "learning_rate": 4.966832407846171e-05, "loss": 0.1374, "step": 3179 }, { "epoch": 19.875, "grad_norm": 1.328125, "learning_rate": 4.9667904503677975e-05, "loss": 0.1272, "step": 3180 }, { "epoch": 19.88125, "grad_norm": 1.1015625, "learning_rate": 4.9667484665452354e-05, "loss": 0.1086, "step": 3181 }, { "epoch": 19.8875, "grad_norm": 1.03125, "learning_rate": 4.966706456378932e-05, "loss": 0.1049, "step": 3182 }, { "epoch": 19.89375, "grad_norm": 0.84765625, "learning_rate": 4.966664419869337e-05, "loss": 0.0881, "step": 3183 }, { "epoch": 19.9, "grad_norm": 1.375, "learning_rate": 4.966622357016898e-05, "loss": 0.1451, "step": 3184 }, { "epoch": 19.90625, "grad_norm": 1.640625, "learning_rate": 4.966580267822065e-05, "loss": 0.2005, "step": 3185 }, { "epoch": 19.9125, "grad_norm": 1.25, "learning_rate": 4.966538152285288e-05, "loss": 0.1269, "step": 3186 }, { "epoch": 19.91875, "grad_norm": 1.3046875, "learning_rate": 4.966496010407017e-05, "loss": 0.1494, "step": 3187 }, { "epoch": 19.925, "grad_norm": 1.109375, "learning_rate": 4.966453842187701e-05, "loss": 0.1331, "step": 3188 }, { "epoch": 19.93125, "grad_norm": 1.125, "learning_rate": 4.9664116476277886e-05, "loss": 0.1001, "step": 3189 }, { "epoch": 19.9375, "grad_norm": 1.9375, "learning_rate": 4.966369426727735e-05, "loss": 0.2107, "step": 3190 }, { "epoch": 19.94375, "grad_norm": 1.5703125, "learning_rate": 4.966327179487987e-05, "loss": 0.2281, "step": 3191 }, { "epoch": 19.95, "grad_norm": 1.3828125, "learning_rate": 4.9662849059089976e-05, "loss": 0.0993, "step": 3192 }, { "epoch": 19.95625, "grad_norm": 2.25, "learning_rate": 4.9662426059912184e-05, "loss": 0.2211, "step": 3193 }, { "epoch": 19.9625, "grad_norm": 1.4921875, "learning_rate": 4.9662002797351014e-05, "loss": 0.1296, "step": 3194 }, { "epoch": 19.96875, "grad_norm": 1.3671875, "learning_rate": 4.966157927141097e-05, "loss": 0.1542, "step": 3195 }, { "epoch": 19.975, "grad_norm": 1.3828125, "learning_rate": 4.966115548209659e-05, "loss": 0.1214, "step": 3196 }, { "epoch": 19.98125, "grad_norm": 1.40625, "learning_rate": 4.966073142941239e-05, "loss": 0.1395, "step": 3197 }, { "epoch": 19.9875, "grad_norm": 1.3515625, "learning_rate": 4.96603071133629e-05, "loss": 0.1593, "step": 3198 }, { "epoch": 19.99375, "grad_norm": 1.828125, "learning_rate": 4.965988253395266e-05, "loss": 0.2245, "step": 3199 }, { "epoch": 20.0, "grad_norm": 1.7578125, "learning_rate": 4.965945769118621e-05, "loss": 0.1939, "step": 3200 }, { "epoch": 20.00625, "grad_norm": 0.88671875, "learning_rate": 4.965903258506806e-05, "loss": 0.0829, "step": 3201 }, { "epoch": 20.0125, "grad_norm": 0.84765625, "learning_rate": 4.965860721560277e-05, "loss": 0.0831, "step": 3202 }, { "epoch": 20.01875, "grad_norm": 1.078125, "learning_rate": 4.965818158279488e-05, "loss": 0.099, "step": 3203 }, { "epoch": 20.025, "grad_norm": 1.359375, "learning_rate": 4.965775568664894e-05, "loss": 0.1409, "step": 3204 }, { "epoch": 20.03125, "grad_norm": 1.515625, "learning_rate": 4.96573295271695e-05, "loss": 0.1335, "step": 3205 }, { "epoch": 20.0375, "grad_norm": 1.4375, "learning_rate": 4.965690310436108e-05, "loss": 0.2657, "step": 3206 }, { "epoch": 20.04375, "grad_norm": 1.2890625, "learning_rate": 4.965647641822827e-05, "loss": 0.1305, "step": 3207 }, { "epoch": 20.05, "grad_norm": 1.2109375, "learning_rate": 4.965604946877562e-05, "loss": 0.1028, "step": 3208 }, { "epoch": 20.05625, "grad_norm": 1.0, "learning_rate": 4.9655622256007685e-05, "loss": 0.1082, "step": 3209 }, { "epoch": 20.0625, "grad_norm": 1.15625, "learning_rate": 4.965519477992903e-05, "loss": 0.1087, "step": 3210 }, { "epoch": 20.06875, "grad_norm": 1.4140625, "learning_rate": 4.96547670405442e-05, "loss": 0.1355, "step": 3211 }, { "epoch": 20.075, "grad_norm": 1.2578125, "learning_rate": 4.9654339037857795e-05, "loss": 0.1249, "step": 3212 }, { "epoch": 20.08125, "grad_norm": 1.6953125, "learning_rate": 4.965391077187437e-05, "loss": 0.1688, "step": 3213 }, { "epoch": 20.0875, "grad_norm": 1.375, "learning_rate": 4.9653482242598495e-05, "loss": 0.1428, "step": 3214 }, { "epoch": 20.09375, "grad_norm": 1.2109375, "learning_rate": 4.965305345003475e-05, "loss": 0.0923, "step": 3215 }, { "epoch": 20.1, "grad_norm": 1.703125, "learning_rate": 4.965262439418772e-05, "loss": 0.1828, "step": 3216 }, { "epoch": 20.10625, "grad_norm": 1.0, "learning_rate": 4.965219507506198e-05, "loss": 0.1115, "step": 3217 }, { "epoch": 20.1125, "grad_norm": 1.28125, "learning_rate": 4.965176549266212e-05, "loss": 0.1719, "step": 3218 }, { "epoch": 20.11875, "grad_norm": 1.6015625, "learning_rate": 4.9651335646992726e-05, "loss": 0.1576, "step": 3219 }, { "epoch": 20.125, "grad_norm": 1.1484375, "learning_rate": 4.965090553805839e-05, "loss": 0.1132, "step": 3220 }, { "epoch": 20.13125, "grad_norm": 0.984375, "learning_rate": 4.9650475165863695e-05, "loss": 0.095, "step": 3221 }, { "epoch": 20.1375, "grad_norm": 1.3984375, "learning_rate": 4.965004453041325e-05, "loss": 0.1693, "step": 3222 }, { "epoch": 20.14375, "grad_norm": 1.1484375, "learning_rate": 4.964961363171166e-05, "loss": 0.1156, "step": 3223 }, { "epoch": 20.15, "grad_norm": 1.3828125, "learning_rate": 4.964918246976351e-05, "loss": 0.1736, "step": 3224 }, { "epoch": 20.15625, "grad_norm": 1.796875, "learning_rate": 4.9648751044573405e-05, "loss": 0.1751, "step": 3225 }, { "epoch": 20.1625, "grad_norm": 1.4140625, "learning_rate": 4.964831935614596e-05, "loss": 0.1524, "step": 3226 }, { "epoch": 20.16875, "grad_norm": 1.4765625, "learning_rate": 4.9647887404485785e-05, "loss": 0.1227, "step": 3227 }, { "epoch": 20.175, "grad_norm": 1.203125, "learning_rate": 4.964745518959749e-05, "loss": 0.1279, "step": 3228 }, { "epoch": 20.18125, "grad_norm": 1.109375, "learning_rate": 4.96470227114857e-05, "loss": 0.1429, "step": 3229 }, { "epoch": 20.1875, "grad_norm": 0.9609375, "learning_rate": 4.9646589970155024e-05, "loss": 0.0775, "step": 3230 }, { "epoch": 20.19375, "grad_norm": 1.4296875, "learning_rate": 4.964615696561008e-05, "loss": 0.1241, "step": 3231 }, { "epoch": 20.2, "grad_norm": 1.4375, "learning_rate": 4.96457236978555e-05, "loss": 0.1591, "step": 3232 }, { "epoch": 20.20625, "grad_norm": 1.3046875, "learning_rate": 4.9645290166895906e-05, "loss": 0.1962, "step": 3233 }, { "epoch": 20.2125, "grad_norm": 1.65625, "learning_rate": 4.9644856372735936e-05, "loss": 0.1215, "step": 3234 }, { "epoch": 20.21875, "grad_norm": 1.9609375, "learning_rate": 4.9644422315380225e-05, "loss": 0.1113, "step": 3235 }, { "epoch": 20.225, "grad_norm": 1.75, "learning_rate": 4.9643987994833385e-05, "loss": 0.1215, "step": 3236 }, { "epoch": 20.23125, "grad_norm": 1.3125, "learning_rate": 4.964355341110009e-05, "loss": 0.1231, "step": 3237 }, { "epoch": 20.2375, "grad_norm": 1.609375, "learning_rate": 4.9643118564184944e-05, "loss": 0.1445, "step": 3238 }, { "epoch": 20.24375, "grad_norm": 1.609375, "learning_rate": 4.964268345409261e-05, "loss": 0.1286, "step": 3239 }, { "epoch": 20.25, "grad_norm": 1.3203125, "learning_rate": 4.964224808082775e-05, "loss": 0.1184, "step": 3240 }, { "epoch": 20.25625, "grad_norm": 1.1484375, "learning_rate": 4.964181244439498e-05, "loss": 0.1105, "step": 3241 }, { "epoch": 20.2625, "grad_norm": 1.6484375, "learning_rate": 4.964137654479898e-05, "loss": 0.1565, "step": 3242 }, { "epoch": 20.26875, "grad_norm": 2.46875, "learning_rate": 4.964094038204439e-05, "loss": 0.2701, "step": 3243 }, { "epoch": 20.275, "grad_norm": 1.8828125, "learning_rate": 4.964050395613586e-05, "loss": 0.2068, "step": 3244 }, { "epoch": 20.28125, "grad_norm": 1.421875, "learning_rate": 4.964006726707808e-05, "loss": 0.1454, "step": 3245 }, { "epoch": 20.2875, "grad_norm": 1.203125, "learning_rate": 4.9639630314875696e-05, "loss": 0.1101, "step": 3246 }, { "epoch": 20.29375, "grad_norm": 1.4921875, "learning_rate": 4.9639193099533376e-05, "loss": 0.1502, "step": 3247 }, { "epoch": 20.3, "grad_norm": 1.34375, "learning_rate": 4.9638755621055785e-05, "loss": 0.101, "step": 3248 }, { "epoch": 20.30625, "grad_norm": 1.3359375, "learning_rate": 4.9638317879447606e-05, "loss": 0.1274, "step": 3249 }, { "epoch": 20.3125, "grad_norm": 0.9453125, "learning_rate": 4.963787987471349e-05, "loss": 0.0915, "step": 3250 }, { "epoch": 20.31875, "grad_norm": 1.28125, "learning_rate": 4.963744160685815e-05, "loss": 0.125, "step": 3251 }, { "epoch": 20.325, "grad_norm": 1.21875, "learning_rate": 4.963700307588623e-05, "loss": 0.0925, "step": 3252 }, { "epoch": 20.33125, "grad_norm": 1.625, "learning_rate": 4.963656428180245e-05, "loss": 0.1641, "step": 3253 }, { "epoch": 20.3375, "grad_norm": 0.88671875, "learning_rate": 4.963612522461147e-05, "loss": 0.0988, "step": 3254 }, { "epoch": 20.34375, "grad_norm": 0.98046875, "learning_rate": 4.9635685904317984e-05, "loss": 0.0994, "step": 3255 }, { "epoch": 20.35, "grad_norm": 1.640625, "learning_rate": 4.963524632092669e-05, "loss": 0.2307, "step": 3256 }, { "epoch": 20.35625, "grad_norm": 0.95703125, "learning_rate": 4.9634806474442274e-05, "loss": 0.0917, "step": 3257 }, { "epoch": 20.3625, "grad_norm": 1.28125, "learning_rate": 4.9634366364869444e-05, "loss": 0.1381, "step": 3258 }, { "epoch": 20.36875, "grad_norm": 1.0546875, "learning_rate": 4.963392599221289e-05, "loss": 0.109, "step": 3259 }, { "epoch": 20.375, "grad_norm": 1.703125, "learning_rate": 4.963348535647732e-05, "loss": 0.1471, "step": 3260 }, { "epoch": 20.38125, "grad_norm": 1.6640625, "learning_rate": 4.963304445766744e-05, "loss": 0.1811, "step": 3261 }, { "epoch": 20.3875, "grad_norm": 1.078125, "learning_rate": 4.9632603295787955e-05, "loss": 0.1079, "step": 3262 }, { "epoch": 20.39375, "grad_norm": 1.4140625, "learning_rate": 4.963216187084359e-05, "loss": 0.1533, "step": 3263 }, { "epoch": 20.4, "grad_norm": 1.7265625, "learning_rate": 4.963172018283903e-05, "loss": 0.2209, "step": 3264 }, { "epoch": 20.40625, "grad_norm": 1.25, "learning_rate": 4.963127823177902e-05, "loss": 0.1062, "step": 3265 }, { "epoch": 20.4125, "grad_norm": 1.6953125, "learning_rate": 4.963083601766827e-05, "loss": 0.2067, "step": 3266 }, { "epoch": 20.41875, "grad_norm": 1.3671875, "learning_rate": 4.96303935405115e-05, "loss": 0.1472, "step": 3267 }, { "epoch": 20.425, "grad_norm": 1.234375, "learning_rate": 4.962995080031344e-05, "loss": 0.1099, "step": 3268 }, { "epoch": 20.43125, "grad_norm": 2.65625, "learning_rate": 4.9629507797078816e-05, "loss": 0.213, "step": 3269 }, { "epoch": 20.4375, "grad_norm": 1.484375, "learning_rate": 4.9629064530812363e-05, "loss": 0.1065, "step": 3270 }, { "epoch": 20.44375, "grad_norm": 1.203125, "learning_rate": 4.96286210015188e-05, "loss": 0.119, "step": 3271 }, { "epoch": 20.45, "grad_norm": 1.3984375, "learning_rate": 4.9628177209202885e-05, "loss": 0.1259, "step": 3272 }, { "epoch": 20.45625, "grad_norm": 1.1953125, "learning_rate": 4.962773315386935e-05, "loss": 0.0976, "step": 3273 }, { "epoch": 20.4625, "grad_norm": 0.78125, "learning_rate": 4.9627288835522934e-05, "loss": 0.0847, "step": 3274 }, { "epoch": 20.46875, "grad_norm": 1.375, "learning_rate": 4.962684425416838e-05, "loss": 0.1785, "step": 3275 }, { "epoch": 20.475, "grad_norm": 1.2734375, "learning_rate": 4.962639940981044e-05, "loss": 0.1164, "step": 3276 }, { "epoch": 20.48125, "grad_norm": 1.28125, "learning_rate": 4.962595430245386e-05, "loss": 0.124, "step": 3277 }, { "epoch": 20.4875, "grad_norm": 1.65625, "learning_rate": 4.9625508932103397e-05, "loss": 0.1977, "step": 3278 }, { "epoch": 20.49375, "grad_norm": 1.453125, "learning_rate": 4.962506329876381e-05, "loss": 0.1812, "step": 3279 }, { "epoch": 20.5, "grad_norm": 1.796875, "learning_rate": 4.9624617402439853e-05, "loss": 0.1882, "step": 3280 }, { "epoch": 20.50625, "grad_norm": 1.4921875, "learning_rate": 4.96241712431363e-05, "loss": 0.1383, "step": 3281 }, { "epoch": 20.5125, "grad_norm": 1.2734375, "learning_rate": 4.96237248208579e-05, "loss": 0.1027, "step": 3282 }, { "epoch": 20.51875, "grad_norm": 1.6484375, "learning_rate": 4.962327813560942e-05, "loss": 0.1507, "step": 3283 }, { "epoch": 20.525, "grad_norm": 1.4609375, "learning_rate": 4.9622831187395645e-05, "loss": 0.1361, "step": 3284 }, { "epoch": 20.53125, "grad_norm": 1.4921875, "learning_rate": 4.962238397622134e-05, "loss": 0.2264, "step": 3285 }, { "epoch": 20.5375, "grad_norm": 0.91796875, "learning_rate": 4.962193650209128e-05, "loss": 0.0917, "step": 3286 }, { "epoch": 20.54375, "grad_norm": 1.203125, "learning_rate": 4.962148876501025e-05, "loss": 0.1334, "step": 3287 }, { "epoch": 20.55, "grad_norm": 0.9609375, "learning_rate": 4.962104076498302e-05, "loss": 0.0892, "step": 3288 }, { "epoch": 20.55625, "grad_norm": 1.203125, "learning_rate": 4.9620592502014393e-05, "loss": 0.1219, "step": 3289 }, { "epoch": 20.5625, "grad_norm": 1.390625, "learning_rate": 4.9620143976109134e-05, "loss": 0.1671, "step": 3290 }, { "epoch": 20.56875, "grad_norm": 1.296875, "learning_rate": 4.961969518727205e-05, "loss": 0.1639, "step": 3291 }, { "epoch": 20.575, "grad_norm": 1.15625, "learning_rate": 4.9619246135507924e-05, "loss": 0.1608, "step": 3292 }, { "epoch": 20.58125, "grad_norm": 1.546875, "learning_rate": 4.9618796820821556e-05, "loss": 0.1967, "step": 3293 }, { "epoch": 20.5875, "grad_norm": 1.2265625, "learning_rate": 4.9618347243217746e-05, "loss": 0.2024, "step": 3294 }, { "epoch": 20.59375, "grad_norm": 2.109375, "learning_rate": 4.9617897402701286e-05, "loss": 0.2097, "step": 3295 }, { "epoch": 20.6, "grad_norm": 1.5546875, "learning_rate": 4.9617447299276993e-05, "loss": 0.1381, "step": 3296 }, { "epoch": 20.60625, "grad_norm": 1.6484375, "learning_rate": 4.9616996932949666e-05, "loss": 0.1876, "step": 3297 }, { "epoch": 20.6125, "grad_norm": 1.96875, "learning_rate": 4.961654630372412e-05, "loss": 0.2493, "step": 3298 }, { "epoch": 20.61875, "grad_norm": 2.46875, "learning_rate": 4.961609541160516e-05, "loss": 0.2379, "step": 3299 }, { "epoch": 20.625, "grad_norm": 1.171875, "learning_rate": 4.9615644256597607e-05, "loss": 0.0965, "step": 3300 }, { "epoch": 20.63125, "grad_norm": 1.515625, "learning_rate": 4.961519283870628e-05, "loss": 0.1127, "step": 3301 }, { "epoch": 20.6375, "grad_norm": 1.3984375, "learning_rate": 4.9614741157936e-05, "loss": 0.111, "step": 3302 }, { "epoch": 20.64375, "grad_norm": 1.25, "learning_rate": 4.961428921429158e-05, "loss": 0.0791, "step": 3303 }, { "epoch": 20.65, "grad_norm": 1.3203125, "learning_rate": 4.961383700777786e-05, "loss": 0.1107, "step": 3304 }, { "epoch": 20.65625, "grad_norm": 1.375, "learning_rate": 4.9613384538399665e-05, "loss": 0.194, "step": 3305 }, { "epoch": 20.6625, "grad_norm": 1.6640625, "learning_rate": 4.9612931806161824e-05, "loss": 0.1951, "step": 3306 }, { "epoch": 20.66875, "grad_norm": 1.2734375, "learning_rate": 4.9612478811069175e-05, "loss": 0.1155, "step": 3307 }, { "epoch": 20.675, "grad_norm": 0.90625, "learning_rate": 4.9612025553126555e-05, "loss": 0.074, "step": 3308 }, { "epoch": 20.68125, "grad_norm": 1.328125, "learning_rate": 4.9611572032338804e-05, "loss": 0.1361, "step": 3309 }, { "epoch": 20.6875, "grad_norm": 0.87109375, "learning_rate": 4.961111824871077e-05, "loss": 0.0983, "step": 3310 }, { "epoch": 20.69375, "grad_norm": 1.640625, "learning_rate": 4.9610664202247294e-05, "loss": 0.184, "step": 3311 }, { "epoch": 20.7, "grad_norm": 1.703125, "learning_rate": 4.9610209892953225e-05, "loss": 0.1476, "step": 3312 }, { "epoch": 20.70625, "grad_norm": 1.2265625, "learning_rate": 4.960975532083342e-05, "loss": 0.1079, "step": 3313 }, { "epoch": 20.7125, "grad_norm": 2.09375, "learning_rate": 4.9609300485892725e-05, "loss": 0.14, "step": 3314 }, { "epoch": 20.71875, "grad_norm": 1.609375, "learning_rate": 4.9608845388136004e-05, "loss": 0.1426, "step": 3315 }, { "epoch": 20.725, "grad_norm": 1.2578125, "learning_rate": 4.960839002756812e-05, "loss": 0.1318, "step": 3316 }, { "epoch": 20.73125, "grad_norm": 1.234375, "learning_rate": 4.960793440419392e-05, "loss": 0.1111, "step": 3317 }, { "epoch": 20.7375, "grad_norm": 1.28125, "learning_rate": 4.960747851801829e-05, "loss": 0.1381, "step": 3318 }, { "epoch": 20.74375, "grad_norm": 1.15625, "learning_rate": 4.960702236904609e-05, "loss": 0.0789, "step": 3319 }, { "epoch": 20.75, "grad_norm": 1.7421875, "learning_rate": 4.960656595728219e-05, "loss": 0.1168, "step": 3320 }, { "epoch": 20.75625, "grad_norm": 1.015625, "learning_rate": 4.9606109282731463e-05, "loss": 0.1155, "step": 3321 }, { "epoch": 20.7625, "grad_norm": 1.78125, "learning_rate": 4.960565234539879e-05, "loss": 0.1763, "step": 3322 }, { "epoch": 20.76875, "grad_norm": 1.515625, "learning_rate": 4.960519514528906e-05, "loss": 0.1409, "step": 3323 }, { "epoch": 20.775, "grad_norm": 2.015625, "learning_rate": 4.960473768240713e-05, "loss": 0.3053, "step": 3324 }, { "epoch": 20.78125, "grad_norm": 1.390625, "learning_rate": 4.960427995675792e-05, "loss": 0.1673, "step": 3325 }, { "epoch": 20.7875, "grad_norm": 1.921875, "learning_rate": 4.960382196834628e-05, "loss": 0.1865, "step": 3326 }, { "epoch": 20.79375, "grad_norm": 1.15625, "learning_rate": 4.960336371717713e-05, "loss": 0.1117, "step": 3327 }, { "epoch": 20.8, "grad_norm": 1.046875, "learning_rate": 4.9602905203255354e-05, "loss": 0.0985, "step": 3328 }, { "epoch": 20.80625, "grad_norm": 1.1171875, "learning_rate": 4.960244642658585e-05, "loss": 0.1014, "step": 3329 }, { "epoch": 20.8125, "grad_norm": 1.4296875, "learning_rate": 4.960198738717351e-05, "loss": 0.1197, "step": 3330 }, { "epoch": 20.81875, "grad_norm": 1.1640625, "learning_rate": 4.960152808502324e-05, "loss": 0.1563, "step": 3331 }, { "epoch": 20.825, "grad_norm": 1.078125, "learning_rate": 4.960106852013996e-05, "loss": 0.0948, "step": 3332 }, { "epoch": 20.83125, "grad_norm": 1.5859375, "learning_rate": 4.9600608692528563e-05, "loss": 0.119, "step": 3333 }, { "epoch": 20.8375, "grad_norm": 1.46875, "learning_rate": 4.960014860219395e-05, "loss": 0.1063, "step": 3334 }, { "epoch": 20.84375, "grad_norm": 0.87890625, "learning_rate": 4.9599688249141066e-05, "loss": 0.0741, "step": 3335 }, { "epoch": 20.85, "grad_norm": 1.234375, "learning_rate": 4.9599227633374793e-05, "loss": 0.1733, "step": 3336 }, { "epoch": 20.85625, "grad_norm": 2.140625, "learning_rate": 4.959876675490007e-05, "loss": 0.2188, "step": 3337 }, { "epoch": 20.8625, "grad_norm": 1.6484375, "learning_rate": 4.959830561372182e-05, "loss": 0.1241, "step": 3338 }, { "epoch": 20.86875, "grad_norm": 1.4453125, "learning_rate": 4.959784420984496e-05, "loss": 0.1867, "step": 3339 }, { "epoch": 20.875, "grad_norm": 1.1015625, "learning_rate": 4.9597382543274415e-05, "loss": 0.1143, "step": 3340 }, { "epoch": 20.88125, "grad_norm": 1.5078125, "learning_rate": 4.959692061401513e-05, "loss": 0.1733, "step": 3341 }, { "epoch": 20.8875, "grad_norm": 1.765625, "learning_rate": 4.959645842207202e-05, "loss": 0.168, "step": 3342 }, { "epoch": 20.89375, "grad_norm": 1.2421875, "learning_rate": 4.9595995967450035e-05, "loss": 0.1155, "step": 3343 }, { "epoch": 20.9, "grad_norm": 1.1015625, "learning_rate": 4.959553325015411e-05, "loss": 0.1153, "step": 3344 }, { "epoch": 20.90625, "grad_norm": 1.59375, "learning_rate": 4.959507027018918e-05, "loss": 0.204, "step": 3345 }, { "epoch": 20.9125, "grad_norm": 1.296875, "learning_rate": 4.9594607027560196e-05, "loss": 0.0929, "step": 3346 }, { "epoch": 20.91875, "grad_norm": 1.609375, "learning_rate": 4.959414352227211e-05, "loss": 0.2112, "step": 3347 }, { "epoch": 20.925, "grad_norm": 1.3203125, "learning_rate": 4.9593679754329855e-05, "loss": 0.1603, "step": 3348 }, { "epoch": 20.93125, "grad_norm": 1.53125, "learning_rate": 4.9593215723738404e-05, "loss": 0.1502, "step": 3349 }, { "epoch": 20.9375, "grad_norm": 1.53125, "learning_rate": 4.9592751430502696e-05, "loss": 0.1195, "step": 3350 }, { "epoch": 20.94375, "grad_norm": 1.375, "learning_rate": 4.95922868746277e-05, "loss": 0.1181, "step": 3351 }, { "epoch": 20.95, "grad_norm": 1.125, "learning_rate": 4.959182205611837e-05, "loss": 0.119, "step": 3352 }, { "epoch": 20.95625, "grad_norm": 1.390625, "learning_rate": 4.959135697497967e-05, "loss": 0.2208, "step": 3353 }, { "epoch": 20.9625, "grad_norm": 2.546875, "learning_rate": 4.9590891631216584e-05, "loss": 0.1596, "step": 3354 }, { "epoch": 20.96875, "grad_norm": 1.7578125, "learning_rate": 4.959042602483406e-05, "loss": 0.1504, "step": 3355 }, { "epoch": 20.975, "grad_norm": 1.8828125, "learning_rate": 4.9589960155837076e-05, "loss": 0.1944, "step": 3356 }, { "epoch": 20.98125, "grad_norm": 1.4921875, "learning_rate": 4.958949402423062e-05, "loss": 0.1968, "step": 3357 }, { "epoch": 20.9875, "grad_norm": 1.421875, "learning_rate": 4.958902763001966e-05, "loss": 0.1441, "step": 3358 }, { "epoch": 20.99375, "grad_norm": 2.21875, "learning_rate": 4.958856097320917e-05, "loss": 0.204, "step": 3359 }, { "epoch": 21.0, "grad_norm": 1.0703125, "learning_rate": 4.958809405380414e-05, "loss": 0.0747, "step": 3360 }, { "epoch": 21.00625, "grad_norm": 1.515625, "learning_rate": 4.958762687180956e-05, "loss": 0.1205, "step": 3361 }, { "epoch": 21.0125, "grad_norm": 1.6328125, "learning_rate": 4.9587159427230415e-05, "loss": 0.1429, "step": 3362 }, { "epoch": 21.01875, "grad_norm": 2.0, "learning_rate": 4.958669172007171e-05, "loss": 0.1837, "step": 3363 }, { "epoch": 21.025, "grad_norm": 1.5234375, "learning_rate": 4.958622375033841e-05, "loss": 0.1637, "step": 3364 }, { "epoch": 21.03125, "grad_norm": 1.203125, "learning_rate": 4.9585755518035544e-05, "loss": 0.1076, "step": 3365 }, { "epoch": 21.0375, "grad_norm": 1.03125, "learning_rate": 4.958528702316809e-05, "loss": 0.1058, "step": 3366 }, { "epoch": 21.04375, "grad_norm": 1.0390625, "learning_rate": 4.958481826574107e-05, "loss": 0.1461, "step": 3367 }, { "epoch": 21.05, "grad_norm": 2.140625, "learning_rate": 4.9584349245759474e-05, "loss": 0.3048, "step": 3368 }, { "epoch": 21.05625, "grad_norm": 1.8359375, "learning_rate": 4.958387996322833e-05, "loss": 0.1219, "step": 3369 }, { "epoch": 21.0625, "grad_norm": 1.6484375, "learning_rate": 4.9583410418152625e-05, "loss": 0.2024, "step": 3370 }, { "epoch": 21.06875, "grad_norm": 1.296875, "learning_rate": 4.958294061053739e-05, "loss": 0.1393, "step": 3371 }, { "epoch": 21.075, "grad_norm": 1.3984375, "learning_rate": 4.958247054038765e-05, "loss": 0.1468, "step": 3372 }, { "epoch": 21.08125, "grad_norm": 1.3515625, "learning_rate": 4.95820002077084e-05, "loss": 0.1042, "step": 3373 }, { "epoch": 21.0875, "grad_norm": 1.3984375, "learning_rate": 4.9581529612504675e-05, "loss": 0.1358, "step": 3374 }, { "epoch": 21.09375, "grad_norm": 1.7890625, "learning_rate": 4.958105875478151e-05, "loss": 0.1381, "step": 3375 }, { "epoch": 21.1, "grad_norm": 1.1875, "learning_rate": 4.9580587634543926e-05, "loss": 0.1642, "step": 3376 }, { "epoch": 21.10625, "grad_norm": 0.9140625, "learning_rate": 4.958011625179695e-05, "loss": 0.0876, "step": 3377 }, { "epoch": 21.1125, "grad_norm": 1.4375, "learning_rate": 4.9579644606545625e-05, "loss": 0.1456, "step": 3378 }, { "epoch": 21.11875, "grad_norm": 1.1875, "learning_rate": 4.9579172698794976e-05, "loss": 0.1388, "step": 3379 }, { "epoch": 21.125, "grad_norm": 1.1171875, "learning_rate": 4.957870052855006e-05, "loss": 0.1191, "step": 3380 }, { "epoch": 21.13125, "grad_norm": 1.1640625, "learning_rate": 4.95782280958159e-05, "loss": 0.1155, "step": 3381 }, { "epoch": 21.1375, "grad_norm": 1.4296875, "learning_rate": 4.957775540059756e-05, "loss": 0.1286, "step": 3382 }, { "epoch": 21.14375, "grad_norm": 0.91015625, "learning_rate": 4.957728244290007e-05, "loss": 0.0849, "step": 3383 }, { "epoch": 21.15, "grad_norm": 1.078125, "learning_rate": 4.957680922272851e-05, "loss": 0.101, "step": 3384 }, { "epoch": 21.15625, "grad_norm": 1.4296875, "learning_rate": 4.9576335740087886e-05, "loss": 0.1431, "step": 3385 }, { "epoch": 21.1625, "grad_norm": 1.4765625, "learning_rate": 4.9575861994983305e-05, "loss": 0.1491, "step": 3386 }, { "epoch": 21.16875, "grad_norm": 1.15625, "learning_rate": 4.957538798741979e-05, "loss": 0.1412, "step": 3387 }, { "epoch": 21.175, "grad_norm": 0.79296875, "learning_rate": 4.957491371740243e-05, "loss": 0.0817, "step": 3388 }, { "epoch": 21.18125, "grad_norm": 1.3671875, "learning_rate": 4.9574439184936265e-05, "loss": 0.1963, "step": 3389 }, { "epoch": 21.1875, "grad_norm": 1.796875, "learning_rate": 4.957396439002638e-05, "loss": 0.2483, "step": 3390 }, { "epoch": 21.19375, "grad_norm": 1.0234375, "learning_rate": 4.957348933267784e-05, "loss": 0.1048, "step": 3391 }, { "epoch": 21.2, "grad_norm": 1.9921875, "learning_rate": 4.9573014012895724e-05, "loss": 0.1753, "step": 3392 }, { "epoch": 21.20625, "grad_norm": 1.25, "learning_rate": 4.9572538430685094e-05, "loss": 0.1218, "step": 3393 }, { "epoch": 21.2125, "grad_norm": 1.2578125, "learning_rate": 4.9572062586051046e-05, "loss": 0.1367, "step": 3394 }, { "epoch": 21.21875, "grad_norm": 1.03125, "learning_rate": 4.957158647899865e-05, "loss": 0.1031, "step": 3395 }, { "epoch": 21.225, "grad_norm": 1.15625, "learning_rate": 4.957111010953299e-05, "loss": 0.141, "step": 3396 }, { "epoch": 21.23125, "grad_norm": 0.69140625, "learning_rate": 4.9570633477659165e-05, "loss": 0.0784, "step": 3397 }, { "epoch": 21.2375, "grad_norm": 1.140625, "learning_rate": 4.957015658338226e-05, "loss": 0.1114, "step": 3398 }, { "epoch": 21.24375, "grad_norm": 1.765625, "learning_rate": 4.956967942670736e-05, "loss": 0.3232, "step": 3399 }, { "epoch": 21.25, "grad_norm": 1.3984375, "learning_rate": 4.956920200763957e-05, "loss": 0.1482, "step": 3400 }, { "epoch": 21.25625, "grad_norm": 1.046875, "learning_rate": 4.956872432618399e-05, "loss": 0.0845, "step": 3401 }, { "epoch": 21.2625, "grad_norm": 1.375, "learning_rate": 4.956824638234571e-05, "loss": 0.1384, "step": 3402 }, { "epoch": 21.26875, "grad_norm": 1.1328125, "learning_rate": 4.9567768176129845e-05, "loss": 0.0892, "step": 3403 }, { "epoch": 21.275, "grad_norm": 1.3359375, "learning_rate": 4.95672897075415e-05, "loss": 0.1201, "step": 3404 }, { "epoch": 21.28125, "grad_norm": 1.15625, "learning_rate": 4.956681097658578e-05, "loss": 0.0927, "step": 3405 }, { "epoch": 21.2875, "grad_norm": 1.15625, "learning_rate": 4.95663319832678e-05, "loss": 0.1329, "step": 3406 }, { "epoch": 21.29375, "grad_norm": 1.4453125, "learning_rate": 4.956585272759268e-05, "loss": 0.1288, "step": 3407 }, { "epoch": 21.3, "grad_norm": 1.5390625, "learning_rate": 4.9565373209565524e-05, "loss": 0.1699, "step": 3408 }, { "epoch": 21.30625, "grad_norm": 1.40625, "learning_rate": 4.956489342919147e-05, "loss": 0.0975, "step": 3409 }, { "epoch": 21.3125, "grad_norm": 1.5859375, "learning_rate": 4.956441338647564e-05, "loss": 0.1971, "step": 3410 }, { "epoch": 21.31875, "grad_norm": 1.390625, "learning_rate": 4.9563933081423155e-05, "loss": 0.1425, "step": 3411 }, { "epoch": 21.325, "grad_norm": 1.328125, "learning_rate": 4.956345251403913e-05, "loss": 0.1519, "step": 3412 }, { "epoch": 21.33125, "grad_norm": 1.3671875, "learning_rate": 4.9562971684328727e-05, "loss": 0.1179, "step": 3413 }, { "epoch": 21.3375, "grad_norm": 1.5, "learning_rate": 4.956249059229707e-05, "loss": 0.1254, "step": 3414 }, { "epoch": 21.34375, "grad_norm": 2.109375, "learning_rate": 4.9562009237949284e-05, "loss": 0.2613, "step": 3415 }, { "epoch": 21.35, "grad_norm": 1.4453125, "learning_rate": 4.9561527621290525e-05, "loss": 0.2128, "step": 3416 }, { "epoch": 21.35625, "grad_norm": 1.625, "learning_rate": 4.956104574232593e-05, "loss": 0.1441, "step": 3417 }, { "epoch": 21.3625, "grad_norm": 1.625, "learning_rate": 4.9560563601060644e-05, "loss": 0.1852, "step": 3418 }, { "epoch": 21.36875, "grad_norm": 1.3125, "learning_rate": 4.956008119749982e-05, "loss": 0.1476, "step": 3419 }, { "epoch": 21.375, "grad_norm": 1.7890625, "learning_rate": 4.9559598531648605e-05, "loss": 0.1722, "step": 3420 }, { "epoch": 21.38125, "grad_norm": 0.94921875, "learning_rate": 4.955911560351216e-05, "loss": 0.0784, "step": 3421 }, { "epoch": 21.3875, "grad_norm": 0.984375, "learning_rate": 4.955863241309563e-05, "loss": 0.0874, "step": 3422 }, { "epoch": 21.39375, "grad_norm": 1.4296875, "learning_rate": 4.9558148960404194e-05, "loss": 0.2052, "step": 3423 }, { "epoch": 21.4, "grad_norm": 1.5078125, "learning_rate": 4.9557665245442995e-05, "loss": 0.1578, "step": 3424 }, { "epoch": 21.40625, "grad_norm": 1.6484375, "learning_rate": 4.9557181268217227e-05, "loss": 0.1941, "step": 3425 }, { "epoch": 21.4125, "grad_norm": 1.1796875, "learning_rate": 4.9556697028732026e-05, "loss": 0.1285, "step": 3426 }, { "epoch": 21.41875, "grad_norm": 1.5078125, "learning_rate": 4.955621252699257e-05, "loss": 0.1161, "step": 3427 }, { "epoch": 21.425, "grad_norm": 1.203125, "learning_rate": 4.955572776300406e-05, "loss": 0.1469, "step": 3428 }, { "epoch": 21.43125, "grad_norm": 0.890625, "learning_rate": 4.9555242736771644e-05, "loss": 0.0894, "step": 3429 }, { "epoch": 21.4375, "grad_norm": 1.140625, "learning_rate": 4.955475744830051e-05, "loss": 0.1103, "step": 3430 }, { "epoch": 21.44375, "grad_norm": 1.7109375, "learning_rate": 4.955427189759585e-05, "loss": 0.1897, "step": 3431 }, { "epoch": 21.45, "grad_norm": 1.6484375, "learning_rate": 4.955378608466285e-05, "loss": 0.1979, "step": 3432 }, { "epoch": 21.45625, "grad_norm": 1.4140625, "learning_rate": 4.9553300009506676e-05, "loss": 0.1938, "step": 3433 }, { "epoch": 21.4625, "grad_norm": 1.140625, "learning_rate": 4.9552813672132536e-05, "loss": 0.093, "step": 3434 }, { "epoch": 21.46875, "grad_norm": 1.6953125, "learning_rate": 4.9552327072545625e-05, "loss": 0.2286, "step": 3435 }, { "epoch": 21.475, "grad_norm": 1.1953125, "learning_rate": 4.955184021075113e-05, "loss": 0.1239, "step": 3436 }, { "epoch": 21.48125, "grad_norm": 1.6015625, "learning_rate": 4.9551353086754266e-05, "loss": 0.1524, "step": 3437 }, { "epoch": 21.4875, "grad_norm": 1.2265625, "learning_rate": 4.955086570056022e-05, "loss": 0.1864, "step": 3438 }, { "epoch": 21.49375, "grad_norm": 1.75, "learning_rate": 4.9550378052174206e-05, "loss": 0.1732, "step": 3439 }, { "epoch": 21.5, "grad_norm": 1.5703125, "learning_rate": 4.954989014160143e-05, "loss": 0.136, "step": 3440 }, { "epoch": 21.50625, "grad_norm": 1.78125, "learning_rate": 4.95494019688471e-05, "loss": 0.1688, "step": 3441 }, { "epoch": 21.5125, "grad_norm": 1.3828125, "learning_rate": 4.954891353391642e-05, "loss": 0.1339, "step": 3442 }, { "epoch": 21.51875, "grad_norm": 1.65625, "learning_rate": 4.9548424836814634e-05, "loss": 0.1908, "step": 3443 }, { "epoch": 21.525, "grad_norm": 1.28125, "learning_rate": 4.954793587754694e-05, "loss": 0.1588, "step": 3444 }, { "epoch": 21.53125, "grad_norm": 1.4375, "learning_rate": 4.954744665611856e-05, "loss": 0.0947, "step": 3445 }, { "epoch": 21.5375, "grad_norm": 1.6484375, "learning_rate": 4.954695717253472e-05, "loss": 0.1544, "step": 3446 }, { "epoch": 21.54375, "grad_norm": 1.3046875, "learning_rate": 4.954646742680065e-05, "loss": 0.1353, "step": 3447 }, { "epoch": 21.55, "grad_norm": 1.1171875, "learning_rate": 4.9545977418921584e-05, "loss": 0.149, "step": 3448 }, { "epoch": 21.55625, "grad_norm": 1.6015625, "learning_rate": 4.954548714890276e-05, "loss": 0.1609, "step": 3449 }, { "epoch": 21.5625, "grad_norm": 1.1875, "learning_rate": 4.95449966167494e-05, "loss": 0.153, "step": 3450 }, { "epoch": 21.56875, "grad_norm": 1.3828125, "learning_rate": 4.954450582246674e-05, "loss": 0.1282, "step": 3451 }, { "epoch": 21.575, "grad_norm": 1.171875, "learning_rate": 4.954401476606003e-05, "loss": 0.1364, "step": 3452 }, { "epoch": 21.58125, "grad_norm": 1.1953125, "learning_rate": 4.954352344753452e-05, "loss": 0.1048, "step": 3453 }, { "epoch": 21.5875, "grad_norm": 1.34375, "learning_rate": 4.954303186689546e-05, "loss": 0.1552, "step": 3454 }, { "epoch": 21.59375, "grad_norm": 1.078125, "learning_rate": 4.9542540024148074e-05, "loss": 0.1085, "step": 3455 }, { "epoch": 21.6, "grad_norm": 1.46875, "learning_rate": 4.954204791929764e-05, "loss": 0.1369, "step": 3456 }, { "epoch": 21.60625, "grad_norm": 1.3671875, "learning_rate": 4.95415555523494e-05, "loss": 0.1506, "step": 3457 }, { "epoch": 21.6125, "grad_norm": 1.7109375, "learning_rate": 4.9541062923308625e-05, "loss": 0.1581, "step": 3458 }, { "epoch": 21.61875, "grad_norm": 1.96875, "learning_rate": 4.954057003218056e-05, "loss": 0.2586, "step": 3459 }, { "epoch": 21.625, "grad_norm": 1.125, "learning_rate": 4.9540076878970485e-05, "loss": 0.1282, "step": 3460 }, { "epoch": 21.63125, "grad_norm": 1.40625, "learning_rate": 4.9539583463683646e-05, "loss": 0.185, "step": 3461 }, { "epoch": 21.6375, "grad_norm": 1.4375, "learning_rate": 4.9539089786325335e-05, "loss": 0.1251, "step": 3462 }, { "epoch": 21.64375, "grad_norm": 1.921875, "learning_rate": 4.953859584690082e-05, "loss": 0.1956, "step": 3463 }, { "epoch": 21.65, "grad_norm": 1.40625, "learning_rate": 4.953810164541536e-05, "loss": 0.1644, "step": 3464 }, { "epoch": 21.65625, "grad_norm": 1.609375, "learning_rate": 4.953760718187425e-05, "loss": 0.2075, "step": 3465 }, { "epoch": 21.6625, "grad_norm": 1.1640625, "learning_rate": 4.9537112456282754e-05, "loss": 0.1014, "step": 3466 }, { "epoch": 21.66875, "grad_norm": 1.65625, "learning_rate": 4.953661746864617e-05, "loss": 0.1634, "step": 3467 }, { "epoch": 21.675, "grad_norm": 1.2265625, "learning_rate": 4.953612221896978e-05, "loss": 0.1127, "step": 3468 }, { "epoch": 21.68125, "grad_norm": 0.84375, "learning_rate": 4.9535626707258875e-05, "loss": 0.0807, "step": 3469 }, { "epoch": 21.6875, "grad_norm": 1.5546875, "learning_rate": 4.953513093351874e-05, "loss": 0.1562, "step": 3470 }, { "epoch": 21.69375, "grad_norm": 1.5625, "learning_rate": 4.9534634897754674e-05, "loss": 0.1261, "step": 3471 }, { "epoch": 21.7, "grad_norm": 1.6484375, "learning_rate": 4.953413859997198e-05, "loss": 0.1245, "step": 3472 }, { "epoch": 21.70625, "grad_norm": 1.5859375, "learning_rate": 4.953364204017595e-05, "loss": 0.1474, "step": 3473 }, { "epoch": 21.7125, "grad_norm": 1.09375, "learning_rate": 4.953314521837189e-05, "loss": 0.1119, "step": 3474 }, { "epoch": 21.71875, "grad_norm": 1.0234375, "learning_rate": 4.95326481345651e-05, "loss": 0.108, "step": 3475 }, { "epoch": 21.725, "grad_norm": 1.6953125, "learning_rate": 4.95321507887609e-05, "loss": 0.1482, "step": 3476 }, { "epoch": 21.73125, "grad_norm": 1.6796875, "learning_rate": 4.95316531809646e-05, "loss": 0.1669, "step": 3477 }, { "epoch": 21.7375, "grad_norm": 0.98828125, "learning_rate": 4.95311553111815e-05, "loss": 0.1086, "step": 3478 }, { "epoch": 21.74375, "grad_norm": 1.578125, "learning_rate": 4.953065717941693e-05, "loss": 0.1685, "step": 3479 }, { "epoch": 21.75, "grad_norm": 1.46875, "learning_rate": 4.953015878567621e-05, "loss": 0.1194, "step": 3480 }, { "epoch": 21.75625, "grad_norm": 2.0625, "learning_rate": 4.952966012996466e-05, "loss": 0.2255, "step": 3481 }, { "epoch": 21.7625, "grad_norm": 1.171875, "learning_rate": 4.95291612122876e-05, "loss": 0.1033, "step": 3482 }, { "epoch": 21.76875, "grad_norm": 1.5625, "learning_rate": 4.9528662032650364e-05, "loss": 0.1518, "step": 3483 }, { "epoch": 21.775, "grad_norm": 1.4921875, "learning_rate": 4.9528162591058283e-05, "loss": 0.2108, "step": 3484 }, { "epoch": 21.78125, "grad_norm": 1.125, "learning_rate": 4.952766288751669e-05, "loss": 0.1038, "step": 3485 }, { "epoch": 21.7875, "grad_norm": 0.80078125, "learning_rate": 4.952716292203092e-05, "loss": 0.0819, "step": 3486 }, { "epoch": 21.79375, "grad_norm": 1.40625, "learning_rate": 4.9526662694606316e-05, "loss": 0.1163, "step": 3487 }, { "epoch": 21.8, "grad_norm": 1.0625, "learning_rate": 4.9526162205248216e-05, "loss": 0.1148, "step": 3488 }, { "epoch": 21.80625, "grad_norm": 1.4375, "learning_rate": 4.952566145396197e-05, "loss": 0.2186, "step": 3489 }, { "epoch": 21.8125, "grad_norm": 1.6953125, "learning_rate": 4.952516044075292e-05, "loss": 0.1674, "step": 3490 }, { "epoch": 21.81875, "grad_norm": 1.4921875, "learning_rate": 4.952465916562642e-05, "loss": 0.1616, "step": 3491 }, { "epoch": 21.825, "grad_norm": 0.9375, "learning_rate": 4.9524157628587824e-05, "loss": 0.0802, "step": 3492 }, { "epoch": 21.83125, "grad_norm": 1.4609375, "learning_rate": 4.9523655829642485e-05, "loss": 0.1414, "step": 3493 }, { "epoch": 21.8375, "grad_norm": 0.97265625, "learning_rate": 4.9523153768795765e-05, "loss": 0.0787, "step": 3494 }, { "epoch": 21.84375, "grad_norm": 1.59375, "learning_rate": 4.9522651446053024e-05, "loss": 0.1378, "step": 3495 }, { "epoch": 21.85, "grad_norm": 1.828125, "learning_rate": 4.952214886141963e-05, "loss": 0.1953, "step": 3496 }, { "epoch": 21.85625, "grad_norm": 1.4296875, "learning_rate": 4.952164601490094e-05, "loss": 0.1517, "step": 3497 }, { "epoch": 21.8625, "grad_norm": 1.109375, "learning_rate": 4.9521142906502335e-05, "loss": 0.1198, "step": 3498 }, { "epoch": 21.86875, "grad_norm": 1.5625, "learning_rate": 4.952063953622919e-05, "loss": 0.1767, "step": 3499 }, { "epoch": 21.875, "grad_norm": 1.6171875, "learning_rate": 4.9520135904086874e-05, "loss": 0.1368, "step": 3500 }, { "epoch": 21.88125, "grad_norm": 2.0, "learning_rate": 4.951963201008076e-05, "loss": 0.2, "step": 3501 }, { "epoch": 21.8875, "grad_norm": 1.359375, "learning_rate": 4.9519127854216246e-05, "loss": 0.1367, "step": 3502 }, { "epoch": 21.89375, "grad_norm": 1.484375, "learning_rate": 4.9518623436498704e-05, "loss": 0.1981, "step": 3503 }, { "epoch": 21.9, "grad_norm": 0.953125, "learning_rate": 4.9518118756933516e-05, "loss": 0.0849, "step": 3504 }, { "epoch": 21.90625, "grad_norm": 1.171875, "learning_rate": 4.951761381552609e-05, "loss": 0.1024, "step": 3505 }, { "epoch": 21.9125, "grad_norm": 1.203125, "learning_rate": 4.951710861228181e-05, "loss": 0.1012, "step": 3506 }, { "epoch": 21.91875, "grad_norm": 1.6328125, "learning_rate": 4.9516603147206054e-05, "loss": 0.1385, "step": 3507 }, { "epoch": 21.925, "grad_norm": 1.4296875, "learning_rate": 4.9516097420304244e-05, "loss": 0.153, "step": 3508 }, { "epoch": 21.93125, "grad_norm": 1.703125, "learning_rate": 4.951559143158177e-05, "loss": 0.142, "step": 3509 }, { "epoch": 21.9375, "grad_norm": 1.5234375, "learning_rate": 4.951508518104404e-05, "loss": 0.1764, "step": 3510 }, { "epoch": 21.94375, "grad_norm": 1.328125, "learning_rate": 4.951457866869646e-05, "loss": 0.1179, "step": 3511 }, { "epoch": 21.95, "grad_norm": 1.3984375, "learning_rate": 4.9514071894544436e-05, "loss": 0.2044, "step": 3512 }, { "epoch": 21.95625, "grad_norm": 1.6015625, "learning_rate": 4.9513564858593376e-05, "loss": 0.1819, "step": 3513 }, { "epoch": 21.9625, "grad_norm": 1.390625, "learning_rate": 4.9513057560848705e-05, "loss": 0.1577, "step": 3514 }, { "epoch": 21.96875, "grad_norm": 1.9453125, "learning_rate": 4.9512550001315836e-05, "loss": 0.1916, "step": 3515 }, { "epoch": 21.975, "grad_norm": 1.0703125, "learning_rate": 4.95120421800002e-05, "loss": 0.0937, "step": 3516 }, { "epoch": 21.98125, "grad_norm": 1.40625, "learning_rate": 4.95115340969072e-05, "loss": 0.1348, "step": 3517 }, { "epoch": 21.9875, "grad_norm": 1.4140625, "learning_rate": 4.951102575204227e-05, "loss": 0.1144, "step": 3518 }, { "epoch": 21.99375, "grad_norm": 1.75, "learning_rate": 4.951051714541085e-05, "loss": 0.174, "step": 3519 }, { "epoch": 22.0, "grad_norm": 1.3046875, "learning_rate": 4.951000827701836e-05, "loss": 0.1362, "step": 3520 }, { "epoch": 22.00625, "grad_norm": 1.8203125, "learning_rate": 4.9509499146870236e-05, "loss": 0.1514, "step": 3521 }, { "epoch": 22.0125, "grad_norm": 0.99609375, "learning_rate": 4.950898975497192e-05, "loss": 0.0963, "step": 3522 }, { "epoch": 22.01875, "grad_norm": 1.2265625, "learning_rate": 4.9508480101328845e-05, "loss": 0.1318, "step": 3523 }, { "epoch": 22.025, "grad_norm": 0.93359375, "learning_rate": 4.950797018594646e-05, "loss": 0.0949, "step": 3524 }, { "epoch": 22.03125, "grad_norm": 1.875, "learning_rate": 4.950746000883021e-05, "loss": 0.1823, "step": 3525 }, { "epoch": 22.0375, "grad_norm": 1.078125, "learning_rate": 4.950694956998553e-05, "loss": 0.0975, "step": 3526 }, { "epoch": 22.04375, "grad_norm": 1.0390625, "learning_rate": 4.9506438869417895e-05, "loss": 0.0892, "step": 3527 }, { "epoch": 22.05, "grad_norm": 1.6796875, "learning_rate": 4.9505927907132744e-05, "loss": 0.2017, "step": 3528 }, { "epoch": 22.05625, "grad_norm": 1.140625, "learning_rate": 4.950541668313554e-05, "loss": 0.11, "step": 3529 }, { "epoch": 22.0625, "grad_norm": 1.1171875, "learning_rate": 4.950490519743174e-05, "loss": 0.1317, "step": 3530 }, { "epoch": 22.06875, "grad_norm": 1.4453125, "learning_rate": 4.9504393450026796e-05, "loss": 0.1807, "step": 3531 }, { "epoch": 22.075, "grad_norm": 2.015625, "learning_rate": 4.950388144092619e-05, "loss": 0.1839, "step": 3532 }, { "epoch": 22.08125, "grad_norm": 0.890625, "learning_rate": 4.950336917013538e-05, "loss": 0.0772, "step": 3533 }, { "epoch": 22.0875, "grad_norm": 1.421875, "learning_rate": 4.950285663765985e-05, "loss": 0.1845, "step": 3534 }, { "epoch": 22.09375, "grad_norm": 1.1640625, "learning_rate": 4.9502343843505055e-05, "loss": 0.1192, "step": 3535 }, { "epoch": 22.1, "grad_norm": 1.4140625, "learning_rate": 4.9501830787676486e-05, "loss": 0.1857, "step": 3536 }, { "epoch": 22.10625, "grad_norm": 1.4921875, "learning_rate": 4.9501317470179606e-05, "loss": 0.163, "step": 3537 }, { "epoch": 22.1125, "grad_norm": 1.90625, "learning_rate": 4.950080389101992e-05, "loss": 0.1493, "step": 3538 }, { "epoch": 22.11875, "grad_norm": 1.5859375, "learning_rate": 4.9500290050202894e-05, "loss": 0.1787, "step": 3539 }, { "epoch": 22.125, "grad_norm": 1.40625, "learning_rate": 4.949977594773402e-05, "loss": 0.1944, "step": 3540 }, { "epoch": 22.13125, "grad_norm": 0.9765625, "learning_rate": 4.9499261583618795e-05, "loss": 0.0914, "step": 3541 }, { "epoch": 22.1375, "grad_norm": 1.4609375, "learning_rate": 4.9498746957862696e-05, "loss": 0.1445, "step": 3542 }, { "epoch": 22.14375, "grad_norm": 1.4765625, "learning_rate": 4.949823207047124e-05, "loss": 0.1664, "step": 3543 }, { "epoch": 22.15, "grad_norm": 1.71875, "learning_rate": 4.949771692144992e-05, "loss": 0.1816, "step": 3544 }, { "epoch": 22.15625, "grad_norm": 2.40625, "learning_rate": 4.949720151080422e-05, "loss": 0.2147, "step": 3545 }, { "epoch": 22.1625, "grad_norm": 1.078125, "learning_rate": 4.949668583853967e-05, "loss": 0.0749, "step": 3546 }, { "epoch": 22.16875, "grad_norm": 0.859375, "learning_rate": 4.949616990466176e-05, "loss": 0.0765, "step": 3547 }, { "epoch": 22.175, "grad_norm": 1.234375, "learning_rate": 4.9495653709176e-05, "loss": 0.1464, "step": 3548 }, { "epoch": 22.18125, "grad_norm": 1.28125, "learning_rate": 4.949513725208792e-05, "loss": 0.1172, "step": 3549 }, { "epoch": 22.1875, "grad_norm": 1.484375, "learning_rate": 4.9494620533403013e-05, "loss": 0.2088, "step": 3550 }, { "epoch": 22.19375, "grad_norm": 1.578125, "learning_rate": 4.949410355312681e-05, "loss": 0.1711, "step": 3551 }, { "epoch": 22.2, "grad_norm": 1.515625, "learning_rate": 4.949358631126483e-05, "loss": 0.1855, "step": 3552 }, { "epoch": 22.20625, "grad_norm": 1.640625, "learning_rate": 4.94930688078226e-05, "loss": 0.1815, "step": 3553 }, { "epoch": 22.2125, "grad_norm": 1.7265625, "learning_rate": 4.949255104280564e-05, "loss": 0.1849, "step": 3554 }, { "epoch": 22.21875, "grad_norm": 1.0703125, "learning_rate": 4.9492033016219485e-05, "loss": 0.0874, "step": 3555 }, { "epoch": 22.225, "grad_norm": 1.40625, "learning_rate": 4.949151472806967e-05, "loss": 0.1346, "step": 3556 }, { "epoch": 22.23125, "grad_norm": 1.3359375, "learning_rate": 4.9490996178361706e-05, "loss": 0.1413, "step": 3557 }, { "epoch": 22.2375, "grad_norm": 1.3515625, "learning_rate": 4.949047736710116e-05, "loss": 0.1192, "step": 3558 }, { "epoch": 22.24375, "grad_norm": 1.3359375, "learning_rate": 4.948995829429357e-05, "loss": 0.1262, "step": 3559 }, { "epoch": 22.25, "grad_norm": 1.046875, "learning_rate": 4.948943895994447e-05, "loss": 0.0825, "step": 3560 }, { "epoch": 22.25625, "grad_norm": 1.390625, "learning_rate": 4.948891936405941e-05, "loss": 0.1264, "step": 3561 }, { "epoch": 22.2625, "grad_norm": 0.9375, "learning_rate": 4.9488399506643924e-05, "loss": 0.087, "step": 3562 }, { "epoch": 22.26875, "grad_norm": 1.3203125, "learning_rate": 4.948787938770358e-05, "loss": 0.1313, "step": 3563 }, { "epoch": 22.275, "grad_norm": 2.0, "learning_rate": 4.948735900724394e-05, "loss": 0.2067, "step": 3564 }, { "epoch": 22.28125, "grad_norm": 1.0078125, "learning_rate": 4.948683836527055e-05, "loss": 0.0915, "step": 3565 }, { "epoch": 22.2875, "grad_norm": 1.5, "learning_rate": 4.948631746178897e-05, "loss": 0.1722, "step": 3566 }, { "epoch": 22.29375, "grad_norm": 1.5859375, "learning_rate": 4.948579629680476e-05, "loss": 0.1494, "step": 3567 }, { "epoch": 22.3, "grad_norm": 1.6484375, "learning_rate": 4.9485274870323495e-05, "loss": 0.2248, "step": 3568 }, { "epoch": 22.30625, "grad_norm": 1.0625, "learning_rate": 4.948475318235073e-05, "loss": 0.0828, "step": 3569 }, { "epoch": 22.3125, "grad_norm": 1.171875, "learning_rate": 4.9484231232892056e-05, "loss": 0.1099, "step": 3570 }, { "epoch": 22.31875, "grad_norm": 1.53125, "learning_rate": 4.948370902195304e-05, "loss": 0.2123, "step": 3571 }, { "epoch": 22.325, "grad_norm": 1.609375, "learning_rate": 4.948318654953924e-05, "loss": 0.2316, "step": 3572 }, { "epoch": 22.33125, "grad_norm": 1.421875, "learning_rate": 4.9482663815656264e-05, "loss": 0.1569, "step": 3573 }, { "epoch": 22.3375, "grad_norm": 1.1015625, "learning_rate": 4.9482140820309676e-05, "loss": 0.1377, "step": 3574 }, { "epoch": 22.34375, "grad_norm": 1.015625, "learning_rate": 4.9481617563505064e-05, "loss": 0.0987, "step": 3575 }, { "epoch": 22.35, "grad_norm": 1.3203125, "learning_rate": 4.948109404524803e-05, "loss": 0.1288, "step": 3576 }, { "epoch": 22.35625, "grad_norm": 1.3125, "learning_rate": 4.9480570265544144e-05, "loss": 0.118, "step": 3577 }, { "epoch": 22.3625, "grad_norm": 1.6484375, "learning_rate": 4.948004622439902e-05, "loss": 0.1791, "step": 3578 }, { "epoch": 22.36875, "grad_norm": 0.9765625, "learning_rate": 4.9479521921818235e-05, "loss": 0.09, "step": 3579 }, { "epoch": 22.375, "grad_norm": 1.34375, "learning_rate": 4.94789973578074e-05, "loss": 0.2181, "step": 3580 }, { "epoch": 22.38125, "grad_norm": 0.8515625, "learning_rate": 4.9478472532372114e-05, "loss": 0.0805, "step": 3581 }, { "epoch": 22.3875, "grad_norm": 1.46875, "learning_rate": 4.947794744551799e-05, "loss": 0.1475, "step": 3582 }, { "epoch": 22.39375, "grad_norm": 1.21875, "learning_rate": 4.947742209725062e-05, "loss": 0.1448, "step": 3583 }, { "epoch": 22.4, "grad_norm": 1.5234375, "learning_rate": 4.947689648757563e-05, "loss": 0.1428, "step": 3584 }, { "epoch": 22.40625, "grad_norm": 1.2734375, "learning_rate": 4.9476370616498617e-05, "loss": 0.1232, "step": 3585 }, { "epoch": 22.4125, "grad_norm": 1.1171875, "learning_rate": 4.947584448402521e-05, "loss": 0.1016, "step": 3586 }, { "epoch": 22.41875, "grad_norm": 1.6640625, "learning_rate": 4.9475318090161026e-05, "loss": 0.1335, "step": 3587 }, { "epoch": 22.425, "grad_norm": 1.6796875, "learning_rate": 4.947479143491169e-05, "loss": 0.1592, "step": 3588 }, { "epoch": 22.43125, "grad_norm": 0.94140625, "learning_rate": 4.947426451828281e-05, "loss": 0.0976, "step": 3589 }, { "epoch": 22.4375, "grad_norm": 1.390625, "learning_rate": 4.947373734028004e-05, "loss": 0.1796, "step": 3590 }, { "epoch": 22.44375, "grad_norm": 2.234375, "learning_rate": 4.9473209900908976e-05, "loss": 0.2388, "step": 3591 }, { "epoch": 22.45, "grad_norm": 1.2265625, "learning_rate": 4.947268220017528e-05, "loss": 0.1152, "step": 3592 }, { "epoch": 22.45625, "grad_norm": 1.28125, "learning_rate": 4.9472154238084574e-05, "loss": 0.1705, "step": 3593 }, { "epoch": 22.4625, "grad_norm": 1.5390625, "learning_rate": 4.94716260146425e-05, "loss": 0.1338, "step": 3594 }, { "epoch": 22.46875, "grad_norm": 1.1796875, "learning_rate": 4.947109752985469e-05, "loss": 0.1022, "step": 3595 }, { "epoch": 22.475, "grad_norm": 2.640625, "learning_rate": 4.947056878372681e-05, "loss": 0.2282, "step": 3596 }, { "epoch": 22.48125, "grad_norm": 0.65234375, "learning_rate": 4.947003977626449e-05, "loss": 0.0711, "step": 3597 }, { "epoch": 22.4875, "grad_norm": 1.375, "learning_rate": 4.946951050747338e-05, "loss": 0.1222, "step": 3598 }, { "epoch": 22.49375, "grad_norm": 0.78125, "learning_rate": 4.946898097735914e-05, "loss": 0.0792, "step": 3599 }, { "epoch": 22.5, "grad_norm": 1.15625, "learning_rate": 4.946845118592741e-05, "loss": 0.1467, "step": 3600 }, { "epoch": 22.50625, "grad_norm": 0.9609375, "learning_rate": 4.946792113318386e-05, "loss": 0.0861, "step": 3601 }, { "epoch": 22.5125, "grad_norm": 0.96484375, "learning_rate": 4.9467390819134155e-05, "loss": 0.0874, "step": 3602 }, { "epoch": 22.51875, "grad_norm": 1.015625, "learning_rate": 4.946686024378395e-05, "loss": 0.1112, "step": 3603 }, { "epoch": 22.525, "grad_norm": 1.59375, "learning_rate": 4.946632940713891e-05, "loss": 0.2046, "step": 3604 }, { "epoch": 22.53125, "grad_norm": 1.25, "learning_rate": 4.946579830920471e-05, "loss": 0.1123, "step": 3605 }, { "epoch": 22.5375, "grad_norm": 1.3125, "learning_rate": 4.9465266949987026e-05, "loss": 0.1177, "step": 3606 }, { "epoch": 22.54375, "grad_norm": 1.921875, "learning_rate": 4.9464735329491524e-05, "loss": 0.1668, "step": 3607 }, { "epoch": 22.55, "grad_norm": 1.3359375, "learning_rate": 4.946420344772388e-05, "loss": 0.177, "step": 3608 }, { "epoch": 22.55625, "grad_norm": 1.2890625, "learning_rate": 4.946367130468978e-05, "loss": 0.1118, "step": 3609 }, { "epoch": 22.5625, "grad_norm": 1.7890625, "learning_rate": 4.94631389003949e-05, "loss": 0.2167, "step": 3610 }, { "epoch": 22.56875, "grad_norm": 1.109375, "learning_rate": 4.946260623484494e-05, "loss": 0.1113, "step": 3611 }, { "epoch": 22.575, "grad_norm": 1.078125, "learning_rate": 4.9462073308045576e-05, "loss": 0.091, "step": 3612 }, { "epoch": 22.58125, "grad_norm": 1.5234375, "learning_rate": 4.94615401200025e-05, "loss": 0.1491, "step": 3613 }, { "epoch": 22.5875, "grad_norm": 1.4921875, "learning_rate": 4.946100667072141e-05, "loss": 0.1578, "step": 3614 }, { "epoch": 22.59375, "grad_norm": 1.0078125, "learning_rate": 4.9460472960208e-05, "loss": 0.1168, "step": 3615 }, { "epoch": 22.6, "grad_norm": 1.578125, "learning_rate": 4.945993898846798e-05, "loss": 0.1565, "step": 3616 }, { "epoch": 22.60625, "grad_norm": 1.2734375, "learning_rate": 4.945940475550703e-05, "loss": 0.1371, "step": 3617 }, { "epoch": 22.6125, "grad_norm": 1.4453125, "learning_rate": 4.945887026133089e-05, "loss": 0.1398, "step": 3618 }, { "epoch": 22.61875, "grad_norm": 1.5, "learning_rate": 4.945833550594523e-05, "loss": 0.1672, "step": 3619 }, { "epoch": 22.625, "grad_norm": 1.3515625, "learning_rate": 4.945780048935579e-05, "loss": 0.1606, "step": 3620 }, { "epoch": 22.63125, "grad_norm": 1.484375, "learning_rate": 4.9457265211568274e-05, "loss": 0.1714, "step": 3621 }, { "epoch": 22.6375, "grad_norm": 1.90625, "learning_rate": 4.9456729672588385e-05, "loss": 0.2296, "step": 3622 }, { "epoch": 22.64375, "grad_norm": 1.84375, "learning_rate": 4.945619387242186e-05, "loss": 0.1809, "step": 3623 }, { "epoch": 22.65, "grad_norm": 1.25, "learning_rate": 4.945565781107442e-05, "loss": 0.114, "step": 3624 }, { "epoch": 22.65625, "grad_norm": 1.234375, "learning_rate": 4.945512148855178e-05, "loss": 0.1237, "step": 3625 }, { "epoch": 22.6625, "grad_norm": 1.3984375, "learning_rate": 4.945458490485968e-05, "loss": 0.137, "step": 3626 }, { "epoch": 22.66875, "grad_norm": 1.0390625, "learning_rate": 4.945404806000384e-05, "loss": 0.0927, "step": 3627 }, { "epoch": 22.675, "grad_norm": 1.5703125, "learning_rate": 4.945351095398999e-05, "loss": 0.1548, "step": 3628 }, { "epoch": 22.68125, "grad_norm": 1.1875, "learning_rate": 4.945297358682388e-05, "loss": 0.1078, "step": 3629 }, { "epoch": 22.6875, "grad_norm": 1.1328125, "learning_rate": 4.945243595851125e-05, "loss": 0.1058, "step": 3630 }, { "epoch": 22.69375, "grad_norm": 1.2421875, "learning_rate": 4.945189806905782e-05, "loss": 0.146, "step": 3631 }, { "epoch": 22.7, "grad_norm": 1.0546875, "learning_rate": 4.9451359918469355e-05, "loss": 0.0914, "step": 3632 }, { "epoch": 22.70625, "grad_norm": 0.89453125, "learning_rate": 4.945082150675159e-05, "loss": 0.0778, "step": 3633 }, { "epoch": 22.7125, "grad_norm": 1.359375, "learning_rate": 4.945028283391028e-05, "loss": 0.1059, "step": 3634 }, { "epoch": 22.71875, "grad_norm": 1.1015625, "learning_rate": 4.944974389995118e-05, "loss": 0.0958, "step": 3635 }, { "epoch": 22.725, "grad_norm": 1.59375, "learning_rate": 4.944920470488005e-05, "loss": 0.128, "step": 3636 }, { "epoch": 22.73125, "grad_norm": 1.453125, "learning_rate": 4.944866524870263e-05, "loss": 0.151, "step": 3637 }, { "epoch": 22.7375, "grad_norm": 1.109375, "learning_rate": 4.94481255314247e-05, "loss": 0.0896, "step": 3638 }, { "epoch": 22.74375, "grad_norm": 1.1796875, "learning_rate": 4.944758555305201e-05, "loss": 0.1033, "step": 3639 }, { "epoch": 22.75, "grad_norm": 1.4609375, "learning_rate": 4.944704531359034e-05, "loss": 0.1952, "step": 3640 }, { "epoch": 22.75625, "grad_norm": 1.4375, "learning_rate": 4.944650481304545e-05, "loss": 0.1648, "step": 3641 }, { "epoch": 22.7625, "grad_norm": 1.34375, "learning_rate": 4.944596405142312e-05, "loss": 0.1531, "step": 3642 }, { "epoch": 22.76875, "grad_norm": 1.9375, "learning_rate": 4.944542302872912e-05, "loss": 0.1772, "step": 3643 }, { "epoch": 22.775, "grad_norm": 1.2734375, "learning_rate": 4.944488174496922e-05, "loss": 0.1325, "step": 3644 }, { "epoch": 22.78125, "grad_norm": 1.34375, "learning_rate": 4.9444340200149216e-05, "loss": 0.1669, "step": 3645 }, { "epoch": 22.7875, "grad_norm": 1.015625, "learning_rate": 4.944379839427488e-05, "loss": 0.1325, "step": 3646 }, { "epoch": 22.79375, "grad_norm": 1.421875, "learning_rate": 4.9443256327352004e-05, "loss": 0.146, "step": 3647 }, { "epoch": 22.8, "grad_norm": 1.65625, "learning_rate": 4.944271399938638e-05, "loss": 0.185, "step": 3648 }, { "epoch": 22.80625, "grad_norm": 1.484375, "learning_rate": 4.944217141038379e-05, "loss": 0.1496, "step": 3649 }, { "epoch": 22.8125, "grad_norm": 1.2265625, "learning_rate": 4.944162856035004e-05, "loss": 0.1231, "step": 3650 }, { "epoch": 22.81875, "grad_norm": 1.7109375, "learning_rate": 4.9441085449290916e-05, "loss": 0.1459, "step": 3651 }, { "epoch": 22.825, "grad_norm": 1.484375, "learning_rate": 4.944054207721223e-05, "loss": 0.1522, "step": 3652 }, { "epoch": 22.83125, "grad_norm": 2.140625, "learning_rate": 4.943999844411977e-05, "loss": 0.1903, "step": 3653 }, { "epoch": 22.8375, "grad_norm": 1.03125, "learning_rate": 4.9439454550019354e-05, "loss": 0.0929, "step": 3654 }, { "epoch": 22.84375, "grad_norm": 1.2421875, "learning_rate": 4.9438910394916785e-05, "loss": 0.1319, "step": 3655 }, { "epoch": 22.85, "grad_norm": 1.078125, "learning_rate": 4.943836597881788e-05, "loss": 0.1267, "step": 3656 }, { "epoch": 22.85625, "grad_norm": 1.1640625, "learning_rate": 4.943782130172845e-05, "loss": 0.0928, "step": 3657 }, { "epoch": 22.8625, "grad_norm": 1.109375, "learning_rate": 4.943727636365431e-05, "loss": 0.0845, "step": 3658 }, { "epoch": 22.86875, "grad_norm": 1.5078125, "learning_rate": 4.943673116460128e-05, "loss": 0.1247, "step": 3659 }, { "epoch": 22.875, "grad_norm": 1.046875, "learning_rate": 4.943618570457518e-05, "loss": 0.1128, "step": 3660 }, { "epoch": 22.88125, "grad_norm": 1.7109375, "learning_rate": 4.9435639983581846e-05, "loss": 0.1603, "step": 3661 }, { "epoch": 22.8875, "grad_norm": 1.375, "learning_rate": 4.943509400162709e-05, "loss": 0.1211, "step": 3662 }, { "epoch": 22.89375, "grad_norm": 1.2265625, "learning_rate": 4.9434547758716756e-05, "loss": 0.1495, "step": 3663 }, { "epoch": 22.9, "grad_norm": 2.15625, "learning_rate": 4.943400125485668e-05, "loss": 0.1825, "step": 3664 }, { "epoch": 22.90625, "grad_norm": 1.125, "learning_rate": 4.9433454490052675e-05, "loss": 0.1095, "step": 3665 }, { "epoch": 22.9125, "grad_norm": 1.4609375, "learning_rate": 4.9432907464310615e-05, "loss": 0.1499, "step": 3666 }, { "epoch": 22.91875, "grad_norm": 1.390625, "learning_rate": 4.9432360177636316e-05, "loss": 0.1397, "step": 3667 }, { "epoch": 22.925, "grad_norm": 1.1484375, "learning_rate": 4.943181263003563e-05, "loss": 0.1288, "step": 3668 }, { "epoch": 22.93125, "grad_norm": 0.95703125, "learning_rate": 4.94312648215144e-05, "loss": 0.0909, "step": 3669 }, { "epoch": 22.9375, "grad_norm": 1.2890625, "learning_rate": 4.9430716752078485e-05, "loss": 0.1669, "step": 3670 }, { "epoch": 22.94375, "grad_norm": 1.328125, "learning_rate": 4.943016842173374e-05, "loss": 0.1219, "step": 3671 }, { "epoch": 22.95, "grad_norm": 1.375, "learning_rate": 4.9429619830486013e-05, "loss": 0.1236, "step": 3672 }, { "epoch": 22.95625, "grad_norm": 0.73046875, "learning_rate": 4.942907097834116e-05, "loss": 0.0735, "step": 3673 }, { "epoch": 22.9625, "grad_norm": 1.4453125, "learning_rate": 4.9428521865305054e-05, "loss": 0.123, "step": 3674 }, { "epoch": 22.96875, "grad_norm": 1.8515625, "learning_rate": 4.942797249138355e-05, "loss": 0.1533, "step": 3675 }, { "epoch": 22.975, "grad_norm": 1.484375, "learning_rate": 4.942742285658252e-05, "loss": 0.1664, "step": 3676 }, { "epoch": 22.98125, "grad_norm": 1.1875, "learning_rate": 4.9426872960907836e-05, "loss": 0.139, "step": 3677 }, { "epoch": 22.9875, "grad_norm": 1.0078125, "learning_rate": 4.9426322804365355e-05, "loss": 0.1005, "step": 3678 }, { "epoch": 22.99375, "grad_norm": 1.171875, "learning_rate": 4.9425772386960965e-05, "loss": 0.1138, "step": 3679 }, { "epoch": 23.0, "grad_norm": 1.15625, "learning_rate": 4.9425221708700554e-05, "loss": 0.1404, "step": 3680 }, { "epoch": 23.00625, "grad_norm": 1.4375, "learning_rate": 4.9424670769589984e-05, "loss": 0.1462, "step": 3681 }, { "epoch": 23.0125, "grad_norm": 1.8671875, "learning_rate": 4.942411956963515e-05, "loss": 0.1633, "step": 3682 }, { "epoch": 23.01875, "grad_norm": 1.4765625, "learning_rate": 4.9423568108841936e-05, "loss": 0.1438, "step": 3683 }, { "epoch": 23.025, "grad_norm": 1.6953125, "learning_rate": 4.942301638721623e-05, "loss": 0.2046, "step": 3684 }, { "epoch": 23.03125, "grad_norm": 1.4921875, "learning_rate": 4.942246440476392e-05, "loss": 0.1261, "step": 3685 }, { "epoch": 23.0375, "grad_norm": 1.484375, "learning_rate": 4.942191216149091e-05, "loss": 0.15, "step": 3686 }, { "epoch": 23.04375, "grad_norm": 1.1953125, "learning_rate": 4.9421359657403096e-05, "loss": 0.132, "step": 3687 }, { "epoch": 23.05, "grad_norm": 1.21875, "learning_rate": 4.9420806892506374e-05, "loss": 0.0977, "step": 3688 }, { "epoch": 23.05625, "grad_norm": 1.4609375, "learning_rate": 4.942025386680664e-05, "loss": 0.1562, "step": 3689 }, { "epoch": 23.0625, "grad_norm": 1.3046875, "learning_rate": 4.9419700580309825e-05, "loss": 0.1223, "step": 3690 }, { "epoch": 23.06875, "grad_norm": 1.453125, "learning_rate": 4.9419147033021814e-05, "loss": 0.1472, "step": 3691 }, { "epoch": 23.075, "grad_norm": 1.4921875, "learning_rate": 4.941859322494853e-05, "loss": 0.165, "step": 3692 }, { "epoch": 23.08125, "grad_norm": 1.5, "learning_rate": 4.941803915609588e-05, "loss": 0.1198, "step": 3693 }, { "epoch": 23.0875, "grad_norm": 1.5625, "learning_rate": 4.941748482646978e-05, "loss": 0.1407, "step": 3694 }, { "epoch": 23.09375, "grad_norm": 1.625, "learning_rate": 4.941693023607617e-05, "loss": 0.1573, "step": 3695 }, { "epoch": 23.1, "grad_norm": 1.28125, "learning_rate": 4.941637538492095e-05, "loss": 0.1399, "step": 3696 }, { "epoch": 23.10625, "grad_norm": 1.375, "learning_rate": 4.941582027301006e-05, "loss": 0.1193, "step": 3697 }, { "epoch": 23.1125, "grad_norm": 1.421875, "learning_rate": 4.941526490034941e-05, "loss": 0.1384, "step": 3698 }, { "epoch": 23.11875, "grad_norm": 1.2421875, "learning_rate": 4.941470926694496e-05, "loss": 0.0968, "step": 3699 }, { "epoch": 23.125, "grad_norm": 1.390625, "learning_rate": 4.9414153372802616e-05, "loss": 0.1705, "step": 3700 }, { "epoch": 23.13125, "grad_norm": 1.53125, "learning_rate": 4.941359721792832e-05, "loss": 0.1954, "step": 3701 }, { "epoch": 23.1375, "grad_norm": 1.03125, "learning_rate": 4.9413040802328024e-05, "loss": 0.0933, "step": 3702 }, { "epoch": 23.14375, "grad_norm": 1.375, "learning_rate": 4.941248412600767e-05, "loss": 0.107, "step": 3703 }, { "epoch": 23.15, "grad_norm": 1.28125, "learning_rate": 4.941192718897319e-05, "loss": 0.0909, "step": 3704 }, { "epoch": 23.15625, "grad_norm": 1.421875, "learning_rate": 4.9411369991230536e-05, "loss": 0.1571, "step": 3705 }, { "epoch": 23.1625, "grad_norm": 1.546875, "learning_rate": 4.9410812532785666e-05, "loss": 0.1551, "step": 3706 }, { "epoch": 23.16875, "grad_norm": 1.3515625, "learning_rate": 4.941025481364453e-05, "loss": 0.1689, "step": 3707 }, { "epoch": 23.175, "grad_norm": 1.296875, "learning_rate": 4.940969683381308e-05, "loss": 0.1228, "step": 3708 }, { "epoch": 23.18125, "grad_norm": 1.8359375, "learning_rate": 4.940913859329728e-05, "loss": 0.2439, "step": 3709 }, { "epoch": 23.1875, "grad_norm": 1.3046875, "learning_rate": 4.940858009210309e-05, "loss": 0.1484, "step": 3710 }, { "epoch": 23.19375, "grad_norm": 2.109375, "learning_rate": 4.9408021330236465e-05, "loss": 0.2658, "step": 3711 }, { "epoch": 23.2, "grad_norm": 1.1328125, "learning_rate": 4.940746230770339e-05, "loss": 0.1052, "step": 3712 }, { "epoch": 23.20625, "grad_norm": 1.8671875, "learning_rate": 4.940690302450982e-05, "loss": 0.252, "step": 3713 }, { "epoch": 23.2125, "grad_norm": 0.98828125, "learning_rate": 4.940634348066174e-05, "loss": 0.1027, "step": 3714 }, { "epoch": 23.21875, "grad_norm": 1.40625, "learning_rate": 4.940578367616511e-05, "loss": 0.1572, "step": 3715 }, { "epoch": 23.225, "grad_norm": 1.5546875, "learning_rate": 4.940522361102593e-05, "loss": 0.1838, "step": 3716 }, { "epoch": 23.23125, "grad_norm": 1.109375, "learning_rate": 4.9404663285250166e-05, "loss": 0.1009, "step": 3717 }, { "epoch": 23.2375, "grad_norm": 1.765625, "learning_rate": 4.9404102698843804e-05, "loss": 0.1738, "step": 3718 }, { "epoch": 23.24375, "grad_norm": 1.25, "learning_rate": 4.940354185181283e-05, "loss": 0.1107, "step": 3719 }, { "epoch": 23.25, "grad_norm": 1.5546875, "learning_rate": 4.940298074416324e-05, "loss": 0.1587, "step": 3720 }, { "epoch": 23.25625, "grad_norm": 1.765625, "learning_rate": 4.940241937590102e-05, "loss": 0.1931, "step": 3721 }, { "epoch": 23.2625, "grad_norm": 1.546875, "learning_rate": 4.940185774703218e-05, "loss": 0.1562, "step": 3722 }, { "epoch": 23.26875, "grad_norm": 1.25, "learning_rate": 4.940129585756269e-05, "loss": 0.1176, "step": 3723 }, { "epoch": 23.275, "grad_norm": 1.0859375, "learning_rate": 4.940073370749857e-05, "loss": 0.0828, "step": 3724 }, { "epoch": 23.28125, "grad_norm": 1.34375, "learning_rate": 4.940017129684582e-05, "loss": 0.1219, "step": 3725 }, { "epoch": 23.2875, "grad_norm": 1.71875, "learning_rate": 4.9399608625610446e-05, "loss": 0.1696, "step": 3726 }, { "epoch": 23.29375, "grad_norm": 0.95703125, "learning_rate": 4.939904569379846e-05, "loss": 0.0876, "step": 3727 }, { "epoch": 23.3, "grad_norm": 2.0625, "learning_rate": 4.939848250141587e-05, "loss": 0.2463, "step": 3728 }, { "epoch": 23.30625, "grad_norm": 1.453125, "learning_rate": 4.939791904846869e-05, "loss": 0.1201, "step": 3729 }, { "epoch": 23.3125, "grad_norm": 0.8828125, "learning_rate": 4.939735533496294e-05, "loss": 0.083, "step": 3730 }, { "epoch": 23.31875, "grad_norm": 1.25, "learning_rate": 4.9396791360904635e-05, "loss": 0.1772, "step": 3731 }, { "epoch": 23.325, "grad_norm": 0.62890625, "learning_rate": 4.93962271262998e-05, "loss": 0.0733, "step": 3732 }, { "epoch": 23.33125, "grad_norm": 1.875, "learning_rate": 4.939566263115447e-05, "loss": 0.1124, "step": 3733 }, { "epoch": 23.3375, "grad_norm": 1.6640625, "learning_rate": 4.939509787547467e-05, "loss": 0.1393, "step": 3734 }, { "epoch": 23.34375, "grad_norm": 1.578125, "learning_rate": 4.939453285926642e-05, "loss": 0.2, "step": 3735 }, { "epoch": 23.35, "grad_norm": 1.8125, "learning_rate": 4.939396758253577e-05, "loss": 0.247, "step": 3736 }, { "epoch": 23.35625, "grad_norm": 1.3046875, "learning_rate": 4.939340204528874e-05, "loss": 0.1354, "step": 3737 }, { "epoch": 23.3625, "grad_norm": 1.203125, "learning_rate": 4.9392836247531383e-05, "loss": 0.1528, "step": 3738 }, { "epoch": 23.36875, "grad_norm": 0.99609375, "learning_rate": 4.939227018926973e-05, "loss": 0.0798, "step": 3739 }, { "epoch": 23.375, "grad_norm": 1.734375, "learning_rate": 4.939170387050984e-05, "loss": 0.2047, "step": 3740 }, { "epoch": 23.38125, "grad_norm": 2.015625, "learning_rate": 4.939113729125776e-05, "loss": 0.2196, "step": 3741 }, { "epoch": 23.3875, "grad_norm": 1.3125, "learning_rate": 4.9390570451519524e-05, "loss": 0.1366, "step": 3742 }, { "epoch": 23.39375, "grad_norm": 1.5234375, "learning_rate": 4.9390003351301204e-05, "loss": 0.159, "step": 3743 }, { "epoch": 23.4, "grad_norm": 1.2578125, "learning_rate": 4.938943599060885e-05, "loss": 0.1574, "step": 3744 }, { "epoch": 23.40625, "grad_norm": 1.7734375, "learning_rate": 4.938886836944851e-05, "loss": 0.1658, "step": 3745 }, { "epoch": 23.4125, "grad_norm": 1.125, "learning_rate": 4.938830048782626e-05, "loss": 0.0952, "step": 3746 }, { "epoch": 23.41875, "grad_norm": 0.83203125, "learning_rate": 4.938773234574816e-05, "loss": 0.0767, "step": 3747 }, { "epoch": 23.425, "grad_norm": 1.328125, "learning_rate": 4.938716394322028e-05, "loss": 0.1419, "step": 3748 }, { "epoch": 23.43125, "grad_norm": 1.2578125, "learning_rate": 4.9386595280248696e-05, "loss": 0.1602, "step": 3749 }, { "epoch": 23.4375, "grad_norm": 1.3203125, "learning_rate": 4.9386026356839465e-05, "loss": 0.1539, "step": 3750 }, { "epoch": 23.44375, "grad_norm": 1.375, "learning_rate": 4.938545717299867e-05, "loss": 0.1172, "step": 3751 }, { "epoch": 23.45, "grad_norm": 0.9765625, "learning_rate": 4.9384887728732395e-05, "loss": 0.1046, "step": 3752 }, { "epoch": 23.45625, "grad_norm": 1.109375, "learning_rate": 4.938431802404672e-05, "loss": 0.1102, "step": 3753 }, { "epoch": 23.4625, "grad_norm": 1.390625, "learning_rate": 4.9383748058947724e-05, "loss": 0.1397, "step": 3754 }, { "epoch": 23.46875, "grad_norm": 1.328125, "learning_rate": 4.93831778334415e-05, "loss": 0.1077, "step": 3755 }, { "epoch": 23.475, "grad_norm": 1.5390625, "learning_rate": 4.9382607347534125e-05, "loss": 0.1142, "step": 3756 }, { "epoch": 23.48125, "grad_norm": 1.390625, "learning_rate": 4.938203660123171e-05, "loss": 0.1557, "step": 3757 }, { "epoch": 23.4875, "grad_norm": 1.3828125, "learning_rate": 4.9381465594540346e-05, "loss": 0.1257, "step": 3758 }, { "epoch": 23.49375, "grad_norm": 1.1484375, "learning_rate": 4.938089432746611e-05, "loss": 0.1076, "step": 3759 }, { "epoch": 23.5, "grad_norm": 0.78515625, "learning_rate": 4.938032280001513e-05, "loss": 0.0725, "step": 3760 }, { "epoch": 23.50625, "grad_norm": 1.09375, "learning_rate": 4.93797510121935e-05, "loss": 0.1205, "step": 3761 }, { "epoch": 23.5125, "grad_norm": 0.71484375, "learning_rate": 4.937917896400732e-05, "loss": 0.0695, "step": 3762 }, { "epoch": 23.51875, "grad_norm": 1.2734375, "learning_rate": 4.93786066554627e-05, "loss": 0.1331, "step": 3763 }, { "epoch": 23.525, "grad_norm": 0.97265625, "learning_rate": 4.9378034086565765e-05, "loss": 0.0788, "step": 3764 }, { "epoch": 23.53125, "grad_norm": 1.21875, "learning_rate": 4.937746125732262e-05, "loss": 0.1563, "step": 3765 }, { "epoch": 23.5375, "grad_norm": 1.1875, "learning_rate": 4.937688816773938e-05, "loss": 0.0962, "step": 3766 }, { "epoch": 23.54375, "grad_norm": 1.390625, "learning_rate": 4.937631481782218e-05, "loss": 0.1161, "step": 3767 }, { "epoch": 23.55, "grad_norm": 1.296875, "learning_rate": 4.937574120757712e-05, "loss": 0.1726, "step": 3768 }, { "epoch": 23.55625, "grad_norm": 1.3125, "learning_rate": 4.9375167337010345e-05, "loss": 0.1483, "step": 3769 }, { "epoch": 23.5625, "grad_norm": 0.828125, "learning_rate": 4.9374593206127976e-05, "loss": 0.0834, "step": 3770 }, { "epoch": 23.56875, "grad_norm": 1.0234375, "learning_rate": 4.937401881493613e-05, "loss": 0.0878, "step": 3771 }, { "epoch": 23.575, "grad_norm": 1.078125, "learning_rate": 4.9373444163440975e-05, "loss": 0.0856, "step": 3772 }, { "epoch": 23.58125, "grad_norm": 1.4375, "learning_rate": 4.937286925164862e-05, "loss": 0.1959, "step": 3773 }, { "epoch": 23.5875, "grad_norm": 1.1484375, "learning_rate": 4.9372294079565225e-05, "loss": 0.1222, "step": 3774 }, { "epoch": 23.59375, "grad_norm": 1.296875, "learning_rate": 4.937171864719691e-05, "loss": 0.1294, "step": 3775 }, { "epoch": 23.6, "grad_norm": 1.5546875, "learning_rate": 4.937114295454984e-05, "loss": 0.1259, "step": 3776 }, { "epoch": 23.60625, "grad_norm": 1.21875, "learning_rate": 4.937056700163015e-05, "loss": 0.1423, "step": 3777 }, { "epoch": 23.6125, "grad_norm": 0.8203125, "learning_rate": 4.9369990788444006e-05, "loss": 0.0733, "step": 3778 }, { "epoch": 23.61875, "grad_norm": 1.0703125, "learning_rate": 4.936941431499754e-05, "loss": 0.1452, "step": 3779 }, { "epoch": 23.625, "grad_norm": 0.94921875, "learning_rate": 4.936883758129693e-05, "loss": 0.0667, "step": 3780 }, { "epoch": 23.63125, "grad_norm": 1.5625, "learning_rate": 4.936826058734832e-05, "loss": 0.2024, "step": 3781 }, { "epoch": 23.6375, "grad_norm": 1.375, "learning_rate": 4.936768333315788e-05, "loss": 0.1482, "step": 3782 }, { "epoch": 23.64375, "grad_norm": 1.0234375, "learning_rate": 4.936710581873177e-05, "loss": 0.0855, "step": 3783 }, { "epoch": 23.65, "grad_norm": 1.3359375, "learning_rate": 4.936652804407616e-05, "loss": 0.1597, "step": 3784 }, { "epoch": 23.65625, "grad_norm": 1.25, "learning_rate": 4.9365950009197224e-05, "loss": 0.0944, "step": 3785 }, { "epoch": 23.6625, "grad_norm": 1.1328125, "learning_rate": 4.936537171410113e-05, "loss": 0.0869, "step": 3786 }, { "epoch": 23.66875, "grad_norm": 1.5390625, "learning_rate": 4.936479315879405e-05, "loss": 0.1594, "step": 3787 }, { "epoch": 23.675, "grad_norm": 1.15625, "learning_rate": 4.9364214343282175e-05, "loss": 0.1292, "step": 3788 }, { "epoch": 23.68125, "grad_norm": 1.4296875, "learning_rate": 4.9363635267571676e-05, "loss": 0.2379, "step": 3789 }, { "epoch": 23.6875, "grad_norm": 1.6015625, "learning_rate": 4.936305593166874e-05, "loss": 0.1776, "step": 3790 }, { "epoch": 23.69375, "grad_norm": 1.125, "learning_rate": 4.936247633557957e-05, "loss": 0.1104, "step": 3791 }, { "epoch": 23.7, "grad_norm": 1.8359375, "learning_rate": 4.936189647931032e-05, "loss": 0.2431, "step": 3792 }, { "epoch": 23.70625, "grad_norm": 1.3359375, "learning_rate": 4.9361316362867215e-05, "loss": 0.1341, "step": 3793 }, { "epoch": 23.7125, "grad_norm": 1.65625, "learning_rate": 4.936073598625643e-05, "loss": 0.1169, "step": 3794 }, { "epoch": 23.71875, "grad_norm": 1.1796875, "learning_rate": 4.936015534948418e-05, "loss": 0.0967, "step": 3795 }, { "epoch": 23.725, "grad_norm": 1.875, "learning_rate": 4.935957445255666e-05, "loss": 0.2534, "step": 3796 }, { "epoch": 23.73125, "grad_norm": 0.90625, "learning_rate": 4.935899329548007e-05, "loss": 0.0971, "step": 3797 }, { "epoch": 23.7375, "grad_norm": 1.125, "learning_rate": 4.9358411878260614e-05, "loss": 0.1497, "step": 3798 }, { "epoch": 23.74375, "grad_norm": 1.2421875, "learning_rate": 4.935783020090451e-05, "loss": 0.1998, "step": 3799 }, { "epoch": 23.75, "grad_norm": 1.15625, "learning_rate": 4.9357248263417963e-05, "loss": 0.1178, "step": 3800 }, { "epoch": 23.75625, "grad_norm": 1.703125, "learning_rate": 4.935666606580719e-05, "loss": 0.1619, "step": 3801 }, { "epoch": 23.7625, "grad_norm": 1.875, "learning_rate": 4.935608360807841e-05, "loss": 0.2034, "step": 3802 }, { "epoch": 23.76875, "grad_norm": 1.2421875, "learning_rate": 4.935550089023784e-05, "loss": 0.1462, "step": 3803 }, { "epoch": 23.775, "grad_norm": 1.4296875, "learning_rate": 4.935491791229171e-05, "loss": 0.105, "step": 3804 }, { "epoch": 23.78125, "grad_norm": 1.078125, "learning_rate": 4.935433467424624e-05, "loss": 0.0932, "step": 3805 }, { "epoch": 23.7875, "grad_norm": 1.546875, "learning_rate": 4.935375117610765e-05, "loss": 0.1562, "step": 3806 }, { "epoch": 23.79375, "grad_norm": 1.1171875, "learning_rate": 4.93531674178822e-05, "loss": 0.1399, "step": 3807 }, { "epoch": 23.8, "grad_norm": 1.2109375, "learning_rate": 4.935258339957609e-05, "loss": 0.115, "step": 3808 }, { "epoch": 23.80625, "grad_norm": 1.390625, "learning_rate": 4.935199912119558e-05, "loss": 0.1264, "step": 3809 }, { "epoch": 23.8125, "grad_norm": 1.0390625, "learning_rate": 4.93514145827469e-05, "loss": 0.0827, "step": 3810 }, { "epoch": 23.81875, "grad_norm": 1.2421875, "learning_rate": 4.93508297842363e-05, "loss": 0.1484, "step": 3811 }, { "epoch": 23.825, "grad_norm": 2.03125, "learning_rate": 4.935024472567001e-05, "loss": 0.2793, "step": 3812 }, { "epoch": 23.83125, "grad_norm": 1.21875, "learning_rate": 4.93496594070543e-05, "loss": 0.0908, "step": 3813 }, { "epoch": 23.8375, "grad_norm": 1.265625, "learning_rate": 4.93490738283954e-05, "loss": 0.1034, "step": 3814 }, { "epoch": 23.84375, "grad_norm": 1.6875, "learning_rate": 4.934848798969959e-05, "loss": 0.172, "step": 3815 }, { "epoch": 23.85, "grad_norm": 1.5859375, "learning_rate": 4.934790189097309e-05, "loss": 0.1567, "step": 3816 }, { "epoch": 23.85625, "grad_norm": 1.0703125, "learning_rate": 4.934731553222219e-05, "loss": 0.0949, "step": 3817 }, { "epoch": 23.8625, "grad_norm": 1.265625, "learning_rate": 4.934672891345315e-05, "loss": 0.1127, "step": 3818 }, { "epoch": 23.86875, "grad_norm": 1.7890625, "learning_rate": 4.93461420346722e-05, "loss": 0.1907, "step": 3819 }, { "epoch": 23.875, "grad_norm": 1.953125, "learning_rate": 4.934555489588566e-05, "loss": 0.2099, "step": 3820 }, { "epoch": 23.88125, "grad_norm": 1.1015625, "learning_rate": 4.9344967497099756e-05, "loss": 0.1027, "step": 3821 }, { "epoch": 23.8875, "grad_norm": 1.6875, "learning_rate": 4.934437983832079e-05, "loss": 0.1396, "step": 3822 }, { "epoch": 23.89375, "grad_norm": 1.2265625, "learning_rate": 4.934379191955503e-05, "loss": 0.1082, "step": 3823 }, { "epoch": 23.9, "grad_norm": 1.6015625, "learning_rate": 4.934320374080874e-05, "loss": 0.1534, "step": 3824 }, { "epoch": 23.90625, "grad_norm": 1.140625, "learning_rate": 4.934261530208823e-05, "loss": 0.1136, "step": 3825 }, { "epoch": 23.9125, "grad_norm": 0.88671875, "learning_rate": 4.934202660339975e-05, "loss": 0.0791, "step": 3826 }, { "epoch": 23.91875, "grad_norm": 1.59375, "learning_rate": 4.934143764474961e-05, "loss": 0.1549, "step": 3827 }, { "epoch": 23.925, "grad_norm": 1.71875, "learning_rate": 4.934084842614409e-05, "loss": 0.1772, "step": 3828 }, { "epoch": 23.93125, "grad_norm": 1.421875, "learning_rate": 4.9340258947589505e-05, "loss": 0.1642, "step": 3829 }, { "epoch": 23.9375, "grad_norm": 0.7734375, "learning_rate": 4.933966920909211e-05, "loss": 0.0749, "step": 3830 }, { "epoch": 23.94375, "grad_norm": 1.984375, "learning_rate": 4.933907921065823e-05, "loss": 0.1898, "step": 3831 }, { "epoch": 23.95, "grad_norm": 1.828125, "learning_rate": 4.9338488952294175e-05, "loss": 0.1419, "step": 3832 }, { "epoch": 23.95625, "grad_norm": 1.3984375, "learning_rate": 4.9337898434006226e-05, "loss": 0.1144, "step": 3833 }, { "epoch": 23.9625, "grad_norm": 0.94921875, "learning_rate": 4.93373076558007e-05, "loss": 0.0768, "step": 3834 }, { "epoch": 23.96875, "grad_norm": 1.4375, "learning_rate": 4.933671661768391e-05, "loss": 0.1965, "step": 3835 }, { "epoch": 23.975, "grad_norm": 1.3203125, "learning_rate": 4.933612531966215e-05, "loss": 0.1622, "step": 3836 }, { "epoch": 23.98125, "grad_norm": 1.40625, "learning_rate": 4.9335533761741757e-05, "loss": 0.1355, "step": 3837 }, { "epoch": 23.9875, "grad_norm": 1.640625, "learning_rate": 4.933494194392904e-05, "loss": 0.1811, "step": 3838 }, { "epoch": 23.99375, "grad_norm": 1.328125, "learning_rate": 4.933434986623031e-05, "loss": 0.144, "step": 3839 }, { "epoch": 24.0, "grad_norm": 1.4140625, "learning_rate": 4.933375752865191e-05, "loss": 0.1265, "step": 3840 }, { "epoch": 24.00625, "grad_norm": 1.7890625, "learning_rate": 4.933316493120015e-05, "loss": 0.2254, "step": 3841 }, { "epoch": 24.0125, "grad_norm": 1.640625, "learning_rate": 4.9332572073881364e-05, "loss": 0.1559, "step": 3842 }, { "epoch": 24.01875, "grad_norm": 1.03125, "learning_rate": 4.9331978956701875e-05, "loss": 0.1025, "step": 3843 }, { "epoch": 24.025, "grad_norm": 1.3515625, "learning_rate": 4.933138557966803e-05, "loss": 0.1289, "step": 3844 }, { "epoch": 24.03125, "grad_norm": 1.84375, "learning_rate": 4.933079194278617e-05, "loss": 0.1368, "step": 3845 }, { "epoch": 24.0375, "grad_norm": 1.4375, "learning_rate": 4.933019804606261e-05, "loss": 0.1339, "step": 3846 }, { "epoch": 24.04375, "grad_norm": 0.9296875, "learning_rate": 4.932960388950372e-05, "loss": 0.0923, "step": 3847 }, { "epoch": 24.05, "grad_norm": 1.0234375, "learning_rate": 4.932900947311583e-05, "loss": 0.0955, "step": 3848 }, { "epoch": 24.05625, "grad_norm": 1.421875, "learning_rate": 4.9328414796905285e-05, "loss": 0.1149, "step": 3849 }, { "epoch": 24.0625, "grad_norm": 1.28125, "learning_rate": 4.932781986087845e-05, "loss": 0.1031, "step": 3850 }, { "epoch": 24.06875, "grad_norm": 1.25, "learning_rate": 4.9327224665041674e-05, "loss": 0.1392, "step": 3851 }, { "epoch": 24.075, "grad_norm": 1.3203125, "learning_rate": 4.93266292094013e-05, "loss": 0.1103, "step": 3852 }, { "epoch": 24.08125, "grad_norm": 1.2265625, "learning_rate": 4.93260334939637e-05, "loss": 0.0991, "step": 3853 }, { "epoch": 24.0875, "grad_norm": 0.96484375, "learning_rate": 4.9325437518735244e-05, "loss": 0.1489, "step": 3854 }, { "epoch": 24.09375, "grad_norm": 1.8046875, "learning_rate": 4.9324841283722276e-05, "loss": 0.2033, "step": 3855 }, { "epoch": 24.1, "grad_norm": 1.4921875, "learning_rate": 4.9324244788931174e-05, "loss": 0.1631, "step": 3856 }, { "epoch": 24.10625, "grad_norm": 1.3125, "learning_rate": 4.9323648034368316e-05, "loss": 0.1713, "step": 3857 }, { "epoch": 24.1125, "grad_norm": 1.0625, "learning_rate": 4.9323051020040056e-05, "loss": 0.0954, "step": 3858 }, { "epoch": 24.11875, "grad_norm": 1.3671875, "learning_rate": 4.932245374595279e-05, "loss": 0.1766, "step": 3859 }, { "epoch": 24.125, "grad_norm": 1.703125, "learning_rate": 4.9321856212112885e-05, "loss": 0.1336, "step": 3860 }, { "epoch": 24.13125, "grad_norm": 1.34375, "learning_rate": 4.932125841852673e-05, "loss": 0.1455, "step": 3861 }, { "epoch": 24.1375, "grad_norm": 1.7734375, "learning_rate": 4.93206603652007e-05, "loss": 0.1964, "step": 3862 }, { "epoch": 24.14375, "grad_norm": 1.4765625, "learning_rate": 4.932006205214118e-05, "loss": 0.1268, "step": 3863 }, { "epoch": 24.15, "grad_norm": 1.15625, "learning_rate": 4.931946347935458e-05, "loss": 0.1046, "step": 3864 }, { "epoch": 24.15625, "grad_norm": 1.375, "learning_rate": 4.931886464684727e-05, "loss": 0.1717, "step": 3865 }, { "epoch": 24.1625, "grad_norm": 1.6640625, "learning_rate": 4.931826555462565e-05, "loss": 0.1092, "step": 3866 }, { "epoch": 24.16875, "grad_norm": 1.2734375, "learning_rate": 4.931766620269613e-05, "loss": 0.1104, "step": 3867 }, { "epoch": 24.175, "grad_norm": 1.03125, "learning_rate": 4.9317066591065096e-05, "loss": 0.124, "step": 3868 }, { "epoch": 24.18125, "grad_norm": 1.953125, "learning_rate": 4.931646671973896e-05, "loss": 0.2536, "step": 3869 }, { "epoch": 24.1875, "grad_norm": 1.328125, "learning_rate": 4.9315866588724124e-05, "loss": 0.1272, "step": 3870 }, { "epoch": 24.19375, "grad_norm": 1.2578125, "learning_rate": 4.9315266198027005e-05, "loss": 0.0944, "step": 3871 }, { "epoch": 24.2, "grad_norm": 1.171875, "learning_rate": 4.9314665547654005e-05, "loss": 0.0941, "step": 3872 }, { "epoch": 24.20625, "grad_norm": 1.3828125, "learning_rate": 4.931406463761154e-05, "loss": 0.1878, "step": 3873 }, { "epoch": 24.2125, "grad_norm": 1.40625, "learning_rate": 4.931346346790604e-05, "loss": 0.1178, "step": 3874 }, { "epoch": 24.21875, "grad_norm": 1.203125, "learning_rate": 4.931286203854392e-05, "loss": 0.0967, "step": 3875 }, { "epoch": 24.225, "grad_norm": 1.3515625, "learning_rate": 4.9312260349531584e-05, "loss": 0.109, "step": 3876 }, { "epoch": 24.23125, "grad_norm": 1.4921875, "learning_rate": 4.931165840087548e-05, "loss": 0.1831, "step": 3877 }, { "epoch": 24.2375, "grad_norm": 1.75, "learning_rate": 4.931105619258203e-05, "loss": 0.1801, "step": 3878 }, { "epoch": 24.24375, "grad_norm": 2.609375, "learning_rate": 4.931045372465766e-05, "loss": 0.1897, "step": 3879 }, { "epoch": 24.25, "grad_norm": 1.71875, "learning_rate": 4.930985099710881e-05, "loss": 0.0955, "step": 3880 }, { "epoch": 24.25625, "grad_norm": 1.125, "learning_rate": 4.9309248009941914e-05, "loss": 0.1189, "step": 3881 }, { "epoch": 24.2625, "grad_norm": 1.0546875, "learning_rate": 4.930864476316342e-05, "loss": 0.1453, "step": 3882 }, { "epoch": 24.26875, "grad_norm": 1.5390625, "learning_rate": 4.930804125677976e-05, "loss": 0.0906, "step": 3883 }, { "epoch": 24.275, "grad_norm": 1.609375, "learning_rate": 4.930743749079738e-05, "loss": 0.0966, "step": 3884 }, { "epoch": 24.28125, "grad_norm": 2.109375, "learning_rate": 4.9306833465222726e-05, "loss": 0.2018, "step": 3885 }, { "epoch": 24.2875, "grad_norm": 1.453125, "learning_rate": 4.9306229180062255e-05, "loss": 0.1605, "step": 3886 }, { "epoch": 24.29375, "grad_norm": 1.6953125, "learning_rate": 4.9305624635322426e-05, "loss": 0.1138, "step": 3887 }, { "epoch": 24.3, "grad_norm": 2.171875, "learning_rate": 4.930501983100969e-05, "loss": 0.1962, "step": 3888 }, { "epoch": 24.30625, "grad_norm": 2.09375, "learning_rate": 4.930441476713049e-05, "loss": 0.1629, "step": 3889 }, { "epoch": 24.3125, "grad_norm": 1.34375, "learning_rate": 4.930380944369132e-05, "loss": 0.1565, "step": 3890 }, { "epoch": 24.31875, "grad_norm": 1.6796875, "learning_rate": 4.930320386069861e-05, "loss": 0.1874, "step": 3891 }, { "epoch": 24.325, "grad_norm": 0.921875, "learning_rate": 4.9302598018158854e-05, "loss": 0.0819, "step": 3892 }, { "epoch": 24.33125, "grad_norm": 1.3828125, "learning_rate": 4.9301991916078505e-05, "loss": 0.1464, "step": 3893 }, { "epoch": 24.3375, "grad_norm": 1.2421875, "learning_rate": 4.930138555446404e-05, "loss": 0.1156, "step": 3894 }, { "epoch": 24.34375, "grad_norm": 1.8203125, "learning_rate": 4.930077893332194e-05, "loss": 0.183, "step": 3895 }, { "epoch": 24.35, "grad_norm": 1.03125, "learning_rate": 4.9300172052658686e-05, "loss": 0.1312, "step": 3896 }, { "epoch": 24.35625, "grad_norm": 0.703125, "learning_rate": 4.929956491248076e-05, "loss": 0.071, "step": 3897 }, { "epoch": 24.3625, "grad_norm": 1.4921875, "learning_rate": 4.929895751279463e-05, "loss": 0.146, "step": 3898 }, { "epoch": 24.36875, "grad_norm": 1.40625, "learning_rate": 4.929834985360679e-05, "loss": 0.1384, "step": 3899 }, { "epoch": 24.375, "grad_norm": 1.3359375, "learning_rate": 4.9297741934923735e-05, "loss": 0.1392, "step": 3900 }, { "epoch": 24.38125, "grad_norm": 1.375, "learning_rate": 4.929713375675196e-05, "loss": 0.1213, "step": 3901 }, { "epoch": 24.3875, "grad_norm": 1.4765625, "learning_rate": 4.9296525319097945e-05, "loss": 0.1635, "step": 3902 }, { "epoch": 24.39375, "grad_norm": 1.5078125, "learning_rate": 4.929591662196821e-05, "loss": 0.1568, "step": 3903 }, { "epoch": 24.4, "grad_norm": 1.328125, "learning_rate": 4.929530766536923e-05, "loss": 0.1112, "step": 3904 }, { "epoch": 24.40625, "grad_norm": 1.1015625, "learning_rate": 4.929469844930753e-05, "loss": 0.1269, "step": 3905 }, { "epoch": 24.4125, "grad_norm": 0.8203125, "learning_rate": 4.9294088973789605e-05, "loss": 0.0798, "step": 3906 }, { "epoch": 24.41875, "grad_norm": 1.3203125, "learning_rate": 4.9293479238821964e-05, "loss": 0.1239, "step": 3907 }, { "epoch": 24.425, "grad_norm": 1.046875, "learning_rate": 4.929286924441112e-05, "loss": 0.111, "step": 3908 }, { "epoch": 24.43125, "grad_norm": 0.68359375, "learning_rate": 4.92922589905636e-05, "loss": 0.0786, "step": 3909 }, { "epoch": 24.4375, "grad_norm": 0.9921875, "learning_rate": 4.92916484772859e-05, "loss": 0.0925, "step": 3910 }, { "epoch": 24.44375, "grad_norm": 0.97265625, "learning_rate": 4.929103770458455e-05, "loss": 0.0802, "step": 3911 }, { "epoch": 24.45, "grad_norm": 1.6796875, "learning_rate": 4.929042667246607e-05, "loss": 0.17, "step": 3912 }, { "epoch": 24.45625, "grad_norm": 1.875, "learning_rate": 4.9289815380937e-05, "loss": 0.1669, "step": 3913 }, { "epoch": 24.4625, "grad_norm": 0.8046875, "learning_rate": 4.928920383000385e-05, "loss": 0.0733, "step": 3914 }, { "epoch": 24.46875, "grad_norm": 1.7578125, "learning_rate": 4.928859201967315e-05, "loss": 0.2022, "step": 3915 }, { "epoch": 24.475, "grad_norm": 1.1953125, "learning_rate": 4.928797994995145e-05, "loss": 0.0853, "step": 3916 }, { "epoch": 24.48125, "grad_norm": 1.484375, "learning_rate": 4.928736762084528e-05, "loss": 0.1758, "step": 3917 }, { "epoch": 24.4875, "grad_norm": 0.984375, "learning_rate": 4.928675503236118e-05, "loss": 0.0838, "step": 3918 }, { "epoch": 24.49375, "grad_norm": 1.125, "learning_rate": 4.928614218450568e-05, "loss": 0.0811, "step": 3919 }, { "epoch": 24.5, "grad_norm": 1.4375, "learning_rate": 4.928552907728533e-05, "loss": 0.1546, "step": 3920 }, { "epoch": 24.50625, "grad_norm": 1.5703125, "learning_rate": 4.9284915710706695e-05, "loss": 0.1566, "step": 3921 }, { "epoch": 24.5125, "grad_norm": 1.65625, "learning_rate": 4.9284302084776306e-05, "loss": 0.1947, "step": 3922 }, { "epoch": 24.51875, "grad_norm": 1.578125, "learning_rate": 4.928368819950072e-05, "loss": 0.1452, "step": 3923 }, { "epoch": 24.525, "grad_norm": 2.140625, "learning_rate": 4.9283074054886506e-05, "loss": 0.1097, "step": 3924 }, { "epoch": 24.53125, "grad_norm": 1.6484375, "learning_rate": 4.928245965094021e-05, "loss": 0.1179, "step": 3925 }, { "epoch": 24.5375, "grad_norm": 1.6796875, "learning_rate": 4.928184498766839e-05, "loss": 0.1407, "step": 3926 }, { "epoch": 24.54375, "grad_norm": 1.6484375, "learning_rate": 4.928123006507762e-05, "loss": 0.1711, "step": 3927 }, { "epoch": 24.55, "grad_norm": 2.21875, "learning_rate": 4.928061488317447e-05, "loss": 0.2086, "step": 3928 }, { "epoch": 24.55625, "grad_norm": 2.09375, "learning_rate": 4.9279999441965495e-05, "loss": 0.1932, "step": 3929 }, { "epoch": 24.5625, "grad_norm": 1.7578125, "learning_rate": 4.927938374145727e-05, "loss": 0.1631, "step": 3930 }, { "epoch": 24.56875, "grad_norm": 1.5625, "learning_rate": 4.927876778165639e-05, "loss": 0.1858, "step": 3931 }, { "epoch": 24.575, "grad_norm": 1.4453125, "learning_rate": 4.927815156256942e-05, "loss": 0.134, "step": 3932 }, { "epoch": 24.58125, "grad_norm": 1.375, "learning_rate": 4.9277535084202935e-05, "loss": 0.2187, "step": 3933 }, { "epoch": 24.5875, "grad_norm": 1.4921875, "learning_rate": 4.927691834656353e-05, "loss": 0.1165, "step": 3934 }, { "epoch": 24.59375, "grad_norm": 1.5078125, "learning_rate": 4.927630134965778e-05, "loss": 0.1686, "step": 3935 }, { "epoch": 24.6, "grad_norm": 1.5234375, "learning_rate": 4.927568409349228e-05, "loss": 0.1846, "step": 3936 }, { "epoch": 24.60625, "grad_norm": 1.921875, "learning_rate": 4.9275066578073626e-05, "loss": 0.2161, "step": 3937 }, { "epoch": 24.6125, "grad_norm": 1.09375, "learning_rate": 4.92744488034084e-05, "loss": 0.0812, "step": 3938 }, { "epoch": 24.61875, "grad_norm": 1.1640625, "learning_rate": 4.927383076950321e-05, "loss": 0.1012, "step": 3939 }, { "epoch": 24.625, "grad_norm": 1.3671875, "learning_rate": 4.927321247636467e-05, "loss": 0.1302, "step": 3940 }, { "epoch": 24.63125, "grad_norm": 1.46875, "learning_rate": 4.9272593923999356e-05, "loss": 0.1536, "step": 3941 }, { "epoch": 24.6375, "grad_norm": 0.68359375, "learning_rate": 4.9271975112413886e-05, "loss": 0.067, "step": 3942 }, { "epoch": 24.64375, "grad_norm": 1.9375, "learning_rate": 4.927135604161487e-05, "loss": 0.1474, "step": 3943 }, { "epoch": 24.65, "grad_norm": 1.328125, "learning_rate": 4.927073671160892e-05, "loss": 0.1327, "step": 3944 }, { "epoch": 24.65625, "grad_norm": 1.1953125, "learning_rate": 4.9270117122402645e-05, "loss": 0.1183, "step": 3945 }, { "epoch": 24.6625, "grad_norm": 1.203125, "learning_rate": 4.926949727400266e-05, "loss": 0.1026, "step": 3946 }, { "epoch": 24.66875, "grad_norm": 1.5078125, "learning_rate": 4.92688771664156e-05, "loss": 0.2156, "step": 3947 }, { "epoch": 24.675, "grad_norm": 1.1171875, "learning_rate": 4.926825679964807e-05, "loss": 0.1031, "step": 3948 }, { "epoch": 24.68125, "grad_norm": 1.3046875, "learning_rate": 4.926763617370671e-05, "loss": 0.1673, "step": 3949 }, { "epoch": 24.6875, "grad_norm": 1.796875, "learning_rate": 4.926701528859813e-05, "loss": 0.129, "step": 3950 }, { "epoch": 24.69375, "grad_norm": 2.078125, "learning_rate": 4.9266394144328976e-05, "loss": 0.2319, "step": 3951 }, { "epoch": 24.7, "grad_norm": 1.4296875, "learning_rate": 4.926577274090588e-05, "loss": 0.128, "step": 3952 }, { "epoch": 24.70625, "grad_norm": 1.484375, "learning_rate": 4.926515107833547e-05, "loss": 0.1118, "step": 3953 }, { "epoch": 24.7125, "grad_norm": 1.6328125, "learning_rate": 4.926452915662439e-05, "loss": 0.2097, "step": 3954 }, { "epoch": 24.71875, "grad_norm": 1.1171875, "learning_rate": 4.926390697577928e-05, "loss": 0.1022, "step": 3955 }, { "epoch": 24.725, "grad_norm": 2.53125, "learning_rate": 4.9263284535806794e-05, "loss": 0.2428, "step": 3956 }, { "epoch": 24.73125, "grad_norm": 1.5546875, "learning_rate": 4.9262661836713564e-05, "loss": 0.1398, "step": 3957 }, { "epoch": 24.7375, "grad_norm": 1.3984375, "learning_rate": 4.926203887850625e-05, "loss": 0.1221, "step": 3958 }, { "epoch": 24.74375, "grad_norm": 1.8125, "learning_rate": 4.92614156611915e-05, "loss": 0.2759, "step": 3959 }, { "epoch": 24.75, "grad_norm": 1.6171875, "learning_rate": 4.926079218477598e-05, "loss": 0.1199, "step": 3960 }, { "epoch": 24.75625, "grad_norm": 1.125, "learning_rate": 4.9260168449266335e-05, "loss": 0.0926, "step": 3961 }, { "epoch": 24.7625, "grad_norm": 1.859375, "learning_rate": 4.925954445466924e-05, "loss": 0.2226, "step": 3962 }, { "epoch": 24.76875, "grad_norm": 1.2578125, "learning_rate": 4.925892020099134e-05, "loss": 0.0966, "step": 3963 }, { "epoch": 24.775, "grad_norm": 1.6484375, "learning_rate": 4.925829568823932e-05, "loss": 0.1289, "step": 3964 }, { "epoch": 24.78125, "grad_norm": 1.671875, "learning_rate": 4.925767091641984e-05, "loss": 0.1774, "step": 3965 }, { "epoch": 24.7875, "grad_norm": 1.59375, "learning_rate": 4.9257045885539576e-05, "loss": 0.1476, "step": 3966 }, { "epoch": 24.79375, "grad_norm": 1.3984375, "learning_rate": 4.9256420595605196e-05, "loss": 0.119, "step": 3967 }, { "epoch": 24.8, "grad_norm": 1.625, "learning_rate": 4.925579504662339e-05, "loss": 0.1557, "step": 3968 }, { "epoch": 24.80625, "grad_norm": 1.203125, "learning_rate": 4.925516923860083e-05, "loss": 0.0903, "step": 3969 }, { "epoch": 24.8125, "grad_norm": 1.5859375, "learning_rate": 4.92545431715442e-05, "loss": 0.1907, "step": 3970 }, { "epoch": 24.81875, "grad_norm": 1.453125, "learning_rate": 4.925391684546019e-05, "loss": 0.1425, "step": 3971 }, { "epoch": 24.825, "grad_norm": 1.3671875, "learning_rate": 4.925329026035548e-05, "loss": 0.1442, "step": 3972 }, { "epoch": 24.83125, "grad_norm": 1.15625, "learning_rate": 4.925266341623678e-05, "loss": 0.0888, "step": 3973 }, { "epoch": 24.8375, "grad_norm": 1.1796875, "learning_rate": 4.9252036313110764e-05, "loss": 0.094, "step": 3974 }, { "epoch": 24.84375, "grad_norm": 1.25, "learning_rate": 4.925140895098413e-05, "loss": 0.1374, "step": 3975 }, { "epoch": 24.85, "grad_norm": 1.078125, "learning_rate": 4.9250781329863606e-05, "loss": 0.1113, "step": 3976 }, { "epoch": 24.85625, "grad_norm": 1.21875, "learning_rate": 4.9250153449755856e-05, "loss": 0.1142, "step": 3977 }, { "epoch": 24.8625, "grad_norm": 1.0625, "learning_rate": 4.924952531066761e-05, "loss": 0.0792, "step": 3978 }, { "epoch": 24.86875, "grad_norm": 1.6171875, "learning_rate": 4.9248896912605574e-05, "loss": 0.1922, "step": 3979 }, { "epoch": 24.875, "grad_norm": 1.171875, "learning_rate": 4.924826825557645e-05, "loss": 0.1208, "step": 3980 }, { "epoch": 24.88125, "grad_norm": 1.4921875, "learning_rate": 4.924763933958696e-05, "loss": 0.1238, "step": 3981 }, { "epoch": 24.8875, "grad_norm": 0.7890625, "learning_rate": 4.924701016464382e-05, "loss": 0.08, "step": 3982 }, { "epoch": 24.89375, "grad_norm": 1.78125, "learning_rate": 4.924638073075374e-05, "loss": 0.1612, "step": 3983 }, { "epoch": 24.9, "grad_norm": 1.4140625, "learning_rate": 4.924575103792345e-05, "loss": 0.1386, "step": 3984 }, { "epoch": 24.90625, "grad_norm": 1.484375, "learning_rate": 4.9245121086159674e-05, "loss": 0.123, "step": 3985 }, { "epoch": 24.9125, "grad_norm": 1.4375, "learning_rate": 4.924449087546914e-05, "loss": 0.1859, "step": 3986 }, { "epoch": 24.91875, "grad_norm": 0.80859375, "learning_rate": 4.9243860405858576e-05, "loss": 0.0885, "step": 3987 }, { "epoch": 24.925, "grad_norm": 0.96484375, "learning_rate": 4.924322967733471e-05, "loss": 0.0806, "step": 3988 }, { "epoch": 24.93125, "grad_norm": 1.3203125, "learning_rate": 4.924259868990429e-05, "loss": 0.1192, "step": 3989 }, { "epoch": 24.9375, "grad_norm": 1.578125, "learning_rate": 4.9241967443574055e-05, "loss": 0.1467, "step": 3990 }, { "epoch": 24.94375, "grad_norm": 1.2265625, "learning_rate": 4.9241335938350744e-05, "loss": 0.0944, "step": 3991 }, { "epoch": 24.95, "grad_norm": 1.0390625, "learning_rate": 4.924070417424108e-05, "loss": 0.1065, "step": 3992 }, { "epoch": 24.95625, "grad_norm": 1.4609375, "learning_rate": 4.9240072151251834e-05, "loss": 0.1118, "step": 3993 }, { "epoch": 24.9625, "grad_norm": 2.09375, "learning_rate": 4.923943986938976e-05, "loss": 0.2074, "step": 3994 }, { "epoch": 24.96875, "grad_norm": 1.2578125, "learning_rate": 4.923880732866159e-05, "loss": 0.0839, "step": 3995 }, { "epoch": 24.975, "grad_norm": 1.640625, "learning_rate": 4.9238174529074093e-05, "loss": 0.1501, "step": 3996 }, { "epoch": 24.98125, "grad_norm": 1.28125, "learning_rate": 4.923754147063402e-05, "loss": 0.115, "step": 3997 }, { "epoch": 24.9875, "grad_norm": 1.5625, "learning_rate": 4.923690815334814e-05, "loss": 0.1231, "step": 3998 }, { "epoch": 24.99375, "grad_norm": 0.8671875, "learning_rate": 4.923627457722321e-05, "loss": 0.0794, "step": 3999 }, { "epoch": 25.0, "grad_norm": 1.609375, "learning_rate": 4.9235640742265986e-05, "loss": 0.1183, "step": 4000 }, { "epoch": 25.00625, "grad_norm": 1.3515625, "learning_rate": 4.923500664848326e-05, "loss": 0.1585, "step": 4001 }, { "epoch": 25.0125, "grad_norm": 1.4296875, "learning_rate": 4.923437229588179e-05, "loss": 0.1105, "step": 4002 }, { "epoch": 25.01875, "grad_norm": 1.359375, "learning_rate": 4.923373768446836e-05, "loss": 0.1065, "step": 4003 }, { "epoch": 25.025, "grad_norm": 1.21875, "learning_rate": 4.923310281424973e-05, "loss": 0.1296, "step": 4004 }, { "epoch": 25.03125, "grad_norm": 1.1328125, "learning_rate": 4.9232467685232696e-05, "loss": 0.107, "step": 4005 }, { "epoch": 25.0375, "grad_norm": 1.375, "learning_rate": 4.923183229742404e-05, "loss": 0.152, "step": 4006 }, { "epoch": 25.04375, "grad_norm": 1.5234375, "learning_rate": 4.923119665083054e-05, "loss": 0.1738, "step": 4007 }, { "epoch": 25.05, "grad_norm": 1.3359375, "learning_rate": 4.923056074545898e-05, "loss": 0.1391, "step": 4008 }, { "epoch": 25.05625, "grad_norm": 1.125, "learning_rate": 4.9229924581316164e-05, "loss": 0.0974, "step": 4009 }, { "epoch": 25.0625, "grad_norm": 1.1953125, "learning_rate": 4.922928815840888e-05, "loss": 0.1178, "step": 4010 }, { "epoch": 25.06875, "grad_norm": 1.6484375, "learning_rate": 4.922865147674393e-05, "loss": 0.1256, "step": 4011 }, { "epoch": 25.075, "grad_norm": 1.34375, "learning_rate": 4.9228014536328105e-05, "loss": 0.0963, "step": 4012 }, { "epoch": 25.08125, "grad_norm": 1.2109375, "learning_rate": 4.922737733716821e-05, "loss": 0.1207, "step": 4013 }, { "epoch": 25.0875, "grad_norm": 1.40625, "learning_rate": 4.9226739879271056e-05, "loss": 0.1645, "step": 4014 }, { "epoch": 25.09375, "grad_norm": 1.5078125, "learning_rate": 4.9226102162643436e-05, "loss": 0.1294, "step": 4015 }, { "epoch": 25.1, "grad_norm": 1.53125, "learning_rate": 4.922546418729218e-05, "loss": 0.1348, "step": 4016 }, { "epoch": 25.10625, "grad_norm": 1.234375, "learning_rate": 4.9224825953224084e-05, "loss": 0.1005, "step": 4017 }, { "epoch": 25.1125, "grad_norm": 1.21875, "learning_rate": 4.922418746044598e-05, "loss": 0.1089, "step": 4018 }, { "epoch": 25.11875, "grad_norm": 1.796875, "learning_rate": 4.922354870896467e-05, "loss": 0.105, "step": 4019 }, { "epoch": 25.125, "grad_norm": 2.1875, "learning_rate": 4.922290969878699e-05, "loss": 0.1524, "step": 4020 }, { "epoch": 25.13125, "grad_norm": 1.984375, "learning_rate": 4.922227042991976e-05, "loss": 0.183, "step": 4021 }, { "epoch": 25.1375, "grad_norm": 1.203125, "learning_rate": 4.922163090236981e-05, "loss": 0.1266, "step": 4022 }, { "epoch": 25.14375, "grad_norm": 0.91796875, "learning_rate": 4.9220991116143956e-05, "loss": 0.078, "step": 4023 }, { "epoch": 25.15, "grad_norm": 1.78125, "learning_rate": 4.922035107124904e-05, "loss": 0.2268, "step": 4024 }, { "epoch": 25.15625, "grad_norm": 1.2734375, "learning_rate": 4.92197107676919e-05, "loss": 0.1416, "step": 4025 }, { "epoch": 25.1625, "grad_norm": 1.0703125, "learning_rate": 4.9219070205479374e-05, "loss": 0.0864, "step": 4026 }, { "epoch": 25.16875, "grad_norm": 1.2890625, "learning_rate": 4.921842938461829e-05, "loss": 0.1341, "step": 4027 }, { "epoch": 25.175, "grad_norm": 1.34375, "learning_rate": 4.921778830511552e-05, "loss": 0.1256, "step": 4028 }, { "epoch": 25.18125, "grad_norm": 1.3671875, "learning_rate": 4.921714696697788e-05, "loss": 0.1046, "step": 4029 }, { "epoch": 25.1875, "grad_norm": 1.125, "learning_rate": 4.921650537021223e-05, "loss": 0.0889, "step": 4030 }, { "epoch": 25.19375, "grad_norm": 1.453125, "learning_rate": 4.9215863514825436e-05, "loss": 0.1211, "step": 4031 }, { "epoch": 25.2, "grad_norm": 1.5234375, "learning_rate": 4.921522140082433e-05, "loss": 0.1761, "step": 4032 }, { "epoch": 25.20625, "grad_norm": 1.4375, "learning_rate": 4.9214579028215776e-05, "loss": 0.1192, "step": 4033 }, { "epoch": 25.2125, "grad_norm": 0.9921875, "learning_rate": 4.921393639700665e-05, "loss": 0.089, "step": 4034 }, { "epoch": 25.21875, "grad_norm": 0.859375, "learning_rate": 4.9213293507203796e-05, "loss": 0.079, "step": 4035 }, { "epoch": 25.225, "grad_norm": 1.7578125, "learning_rate": 4.9212650358814094e-05, "loss": 0.2081, "step": 4036 }, { "epoch": 25.23125, "grad_norm": 1.4140625, "learning_rate": 4.9212006951844396e-05, "loss": 0.1557, "step": 4037 }, { "epoch": 25.2375, "grad_norm": 1.6328125, "learning_rate": 4.921136328630159e-05, "loss": 0.1529, "step": 4038 }, { "epoch": 25.24375, "grad_norm": 1.3359375, "learning_rate": 4.921071936219253e-05, "loss": 0.1051, "step": 4039 }, { "epoch": 25.25, "grad_norm": 1.15625, "learning_rate": 4.9210075179524124e-05, "loss": 0.1568, "step": 4040 }, { "epoch": 25.25625, "grad_norm": 1.125, "learning_rate": 4.920943073830322e-05, "loss": 0.1032, "step": 4041 }, { "epoch": 25.2625, "grad_norm": 1.5546875, "learning_rate": 4.920878603853672e-05, "loss": 0.216, "step": 4042 }, { "epoch": 25.26875, "grad_norm": 1.359375, "learning_rate": 4.920814108023151e-05, "loss": 0.1277, "step": 4043 }, { "epoch": 25.275, "grad_norm": 1.15625, "learning_rate": 4.920749586339446e-05, "loss": 0.1046, "step": 4044 }, { "epoch": 25.28125, "grad_norm": 1.53125, "learning_rate": 4.920685038803247e-05, "loss": 0.137, "step": 4045 }, { "epoch": 25.2875, "grad_norm": 0.9453125, "learning_rate": 4.920620465415243e-05, "loss": 0.0861, "step": 4046 }, { "epoch": 25.29375, "grad_norm": 1.359375, "learning_rate": 4.9205558661761255e-05, "loss": 0.1572, "step": 4047 }, { "epoch": 25.3, "grad_norm": 1.25, "learning_rate": 4.920491241086582e-05, "loss": 0.162, "step": 4048 }, { "epoch": 25.30625, "grad_norm": 1.375, "learning_rate": 4.920426590147304e-05, "loss": 0.1325, "step": 4049 }, { "epoch": 25.3125, "grad_norm": 1.28125, "learning_rate": 4.920361913358982e-05, "loss": 0.1112, "step": 4050 }, { "epoch": 25.31875, "grad_norm": 2.265625, "learning_rate": 4.9202972107223055e-05, "loss": 0.2473, "step": 4051 }, { "epoch": 25.325, "grad_norm": 0.69140625, "learning_rate": 4.920232482237966e-05, "loss": 0.0703, "step": 4052 }, { "epoch": 25.33125, "grad_norm": 0.90234375, "learning_rate": 4.920167727906656e-05, "loss": 0.0898, "step": 4053 }, { "epoch": 25.3375, "grad_norm": 1.0234375, "learning_rate": 4.9201029477290656e-05, "loss": 0.1046, "step": 4054 }, { "epoch": 25.34375, "grad_norm": 1.1484375, "learning_rate": 4.920038141705887e-05, "loss": 0.1297, "step": 4055 }, { "epoch": 25.35, "grad_norm": 1.2109375, "learning_rate": 4.9199733098378126e-05, "loss": 0.1268, "step": 4056 }, { "epoch": 25.35625, "grad_norm": 1.7421875, "learning_rate": 4.9199084521255345e-05, "loss": 0.2001, "step": 4057 }, { "epoch": 25.3625, "grad_norm": 1.1015625, "learning_rate": 4.919843568569745e-05, "loss": 0.0988, "step": 4058 }, { "epoch": 25.36875, "grad_norm": 1.125, "learning_rate": 4.9197786591711386e-05, "loss": 0.1399, "step": 4059 }, { "epoch": 25.375, "grad_norm": 1.125, "learning_rate": 4.919713723930407e-05, "loss": 0.1098, "step": 4060 }, { "epoch": 25.38125, "grad_norm": 1.6796875, "learning_rate": 4.9196487628482435e-05, "loss": 0.1393, "step": 4061 }, { "epoch": 25.3875, "grad_norm": 1.4296875, "learning_rate": 4.919583775925343e-05, "loss": 0.1433, "step": 4062 }, { "epoch": 25.39375, "grad_norm": 1.1640625, "learning_rate": 4.919518763162398e-05, "loss": 0.0889, "step": 4063 }, { "epoch": 25.4, "grad_norm": 1.6875, "learning_rate": 4.9194537245601044e-05, "loss": 0.2375, "step": 4064 }, { "epoch": 25.40625, "grad_norm": 1.28125, "learning_rate": 4.919388660119156e-05, "loss": 0.1532, "step": 4065 }, { "epoch": 25.4125, "grad_norm": 1.5859375, "learning_rate": 4.919323569840248e-05, "loss": 0.1415, "step": 4066 }, { "epoch": 25.41875, "grad_norm": 1.9609375, "learning_rate": 4.9192584537240754e-05, "loss": 0.2158, "step": 4067 }, { "epoch": 25.425, "grad_norm": 1.671875, "learning_rate": 4.919193311771334e-05, "loss": 0.1814, "step": 4068 }, { "epoch": 25.43125, "grad_norm": 1.34375, "learning_rate": 4.9191281439827176e-05, "loss": 0.1018, "step": 4069 }, { "epoch": 25.4375, "grad_norm": 1.6953125, "learning_rate": 4.9190629503589245e-05, "loss": 0.1253, "step": 4070 }, { "epoch": 25.44375, "grad_norm": 1.125, "learning_rate": 4.9189977309006495e-05, "loss": 0.0788, "step": 4071 }, { "epoch": 25.45, "grad_norm": 1.3359375, "learning_rate": 4.9189324856085906e-05, "loss": 0.1451, "step": 4072 }, { "epoch": 25.45625, "grad_norm": 1.609375, "learning_rate": 4.9188672144834423e-05, "loss": 0.1572, "step": 4073 }, { "epoch": 25.4625, "grad_norm": 1.71875, "learning_rate": 4.9188019175259034e-05, "loss": 0.1377, "step": 4074 }, { "epoch": 25.46875, "grad_norm": 1.7265625, "learning_rate": 4.918736594736672e-05, "loss": 0.1502, "step": 4075 }, { "epoch": 25.475, "grad_norm": 1.3515625, "learning_rate": 4.918671246116443e-05, "loss": 0.1457, "step": 4076 }, { "epoch": 25.48125, "grad_norm": 1.4609375, "learning_rate": 4.9186058716659166e-05, "loss": 0.2004, "step": 4077 }, { "epoch": 25.4875, "grad_norm": 1.265625, "learning_rate": 4.91854047138579e-05, "loss": 0.104, "step": 4078 }, { "epoch": 25.49375, "grad_norm": 1.2265625, "learning_rate": 4.918475045276761e-05, "loss": 0.1157, "step": 4079 }, { "epoch": 25.5, "grad_norm": 1.4375, "learning_rate": 4.918409593339531e-05, "loss": 0.1601, "step": 4080 }, { "epoch": 25.50625, "grad_norm": 1.8359375, "learning_rate": 4.918344115574796e-05, "loss": 0.2234, "step": 4081 }, { "epoch": 25.5125, "grad_norm": 1.3671875, "learning_rate": 4.918278611983257e-05, "loss": 0.1621, "step": 4082 }, { "epoch": 25.51875, "grad_norm": 1.4140625, "learning_rate": 4.9182130825656126e-05, "loss": 0.1324, "step": 4083 }, { "epoch": 25.525, "grad_norm": 1.265625, "learning_rate": 4.918147527322563e-05, "loss": 0.1134, "step": 4084 }, { "epoch": 25.53125, "grad_norm": 1.234375, "learning_rate": 4.918081946254808e-05, "loss": 0.1237, "step": 4085 }, { "epoch": 25.5375, "grad_norm": 1.5078125, "learning_rate": 4.918016339363049e-05, "loss": 0.1396, "step": 4086 }, { "epoch": 25.54375, "grad_norm": 1.1015625, "learning_rate": 4.917950706647986e-05, "loss": 0.1139, "step": 4087 }, { "epoch": 25.55, "grad_norm": 1.5859375, "learning_rate": 4.917885048110319e-05, "loss": 0.1391, "step": 4088 }, { "epoch": 25.55625, "grad_norm": 1.4375, "learning_rate": 4.91781936375075e-05, "loss": 0.1604, "step": 4089 }, { "epoch": 25.5625, "grad_norm": 1.21875, "learning_rate": 4.9177536535699814e-05, "loss": 0.1166, "step": 4090 }, { "epoch": 25.56875, "grad_norm": 1.421875, "learning_rate": 4.917687917568714e-05, "loss": 0.1099, "step": 4091 }, { "epoch": 25.575, "grad_norm": 0.8671875, "learning_rate": 4.917622155747649e-05, "loss": 0.0739, "step": 4092 }, { "epoch": 25.58125, "grad_norm": 1.1953125, "learning_rate": 4.917556368107491e-05, "loss": 0.1171, "step": 4093 }, { "epoch": 25.5875, "grad_norm": 1.7578125, "learning_rate": 4.9174905546489404e-05, "loss": 0.2076, "step": 4094 }, { "epoch": 25.59375, "grad_norm": 1.7265625, "learning_rate": 4.9174247153727e-05, "loss": 0.2018, "step": 4095 }, { "epoch": 25.6, "grad_norm": 1.21875, "learning_rate": 4.917358850279475e-05, "loss": 0.1156, "step": 4096 }, { "epoch": 25.60625, "grad_norm": 1.109375, "learning_rate": 4.917292959369968e-05, "loss": 0.1115, "step": 4097 }, { "epoch": 25.6125, "grad_norm": 1.28125, "learning_rate": 4.917227042644882e-05, "loss": 0.1371, "step": 4098 }, { "epoch": 25.61875, "grad_norm": 1.296875, "learning_rate": 4.917161100104921e-05, "loss": 0.1638, "step": 4099 }, { "epoch": 25.625, "grad_norm": 1.2890625, "learning_rate": 4.9170951317507895e-05, "loss": 0.094, "step": 4100 }, { "epoch": 25.63125, "grad_norm": 0.90625, "learning_rate": 4.9170291375831915e-05, "loss": 0.0913, "step": 4101 }, { "epoch": 25.6375, "grad_norm": 0.66796875, "learning_rate": 4.916963117602833e-05, "loss": 0.0698, "step": 4102 }, { "epoch": 25.64375, "grad_norm": 1.1953125, "learning_rate": 4.916897071810418e-05, "loss": 0.1061, "step": 4103 }, { "epoch": 25.65, "grad_norm": 1.1640625, "learning_rate": 4.9168310002066525e-05, "loss": 0.1267, "step": 4104 }, { "epoch": 25.65625, "grad_norm": 1.2109375, "learning_rate": 4.9167649027922415e-05, "loss": 0.114, "step": 4105 }, { "epoch": 25.6625, "grad_norm": 0.9765625, "learning_rate": 4.916698779567892e-05, "loss": 0.1201, "step": 4106 }, { "epoch": 25.66875, "grad_norm": 1.40625, "learning_rate": 4.916632630534308e-05, "loss": 0.1546, "step": 4107 }, { "epoch": 25.675, "grad_norm": 1.6171875, "learning_rate": 4.9165664556921985e-05, "loss": 0.205, "step": 4108 }, { "epoch": 25.68125, "grad_norm": 0.81640625, "learning_rate": 4.916500255042268e-05, "loss": 0.0612, "step": 4109 }, { "epoch": 25.6875, "grad_norm": 1.0546875, "learning_rate": 4.9164340285852256e-05, "loss": 0.0972, "step": 4110 }, { "epoch": 25.69375, "grad_norm": 1.625, "learning_rate": 4.916367776321777e-05, "loss": 0.1726, "step": 4111 }, { "epoch": 25.7, "grad_norm": 1.078125, "learning_rate": 4.91630149825263e-05, "loss": 0.0899, "step": 4112 }, { "epoch": 25.70625, "grad_norm": 1.125, "learning_rate": 4.9162351943784935e-05, "loss": 0.0995, "step": 4113 }, { "epoch": 25.7125, "grad_norm": 0.9921875, "learning_rate": 4.916168864700074e-05, "loss": 0.0873, "step": 4114 }, { "epoch": 25.71875, "grad_norm": 1.2421875, "learning_rate": 4.91610250921808e-05, "loss": 0.1626, "step": 4115 }, { "epoch": 25.725, "grad_norm": 1.40625, "learning_rate": 4.9160361279332226e-05, "loss": 0.1717, "step": 4116 }, { "epoch": 25.73125, "grad_norm": 1.265625, "learning_rate": 4.9159697208462074e-05, "loss": 0.1309, "step": 4117 }, { "epoch": 25.7375, "grad_norm": 1.296875, "learning_rate": 4.9159032879577446e-05, "loss": 0.1438, "step": 4118 }, { "epoch": 25.74375, "grad_norm": 0.90625, "learning_rate": 4.915836829268546e-05, "loss": 0.0889, "step": 4119 }, { "epoch": 25.75, "grad_norm": 1.1484375, "learning_rate": 4.915770344779318e-05, "loss": 0.1096, "step": 4120 }, { "epoch": 25.75625, "grad_norm": 1.359375, "learning_rate": 4.915703834490773e-05, "loss": 0.1417, "step": 4121 }, { "epoch": 25.7625, "grad_norm": 1.515625, "learning_rate": 4.91563729840362e-05, "loss": 0.1379, "step": 4122 }, { "epoch": 25.76875, "grad_norm": 1.2265625, "learning_rate": 4.91557073651857e-05, "loss": 0.1197, "step": 4123 }, { "epoch": 25.775, "grad_norm": 1.46875, "learning_rate": 4.915504148836334e-05, "loss": 0.1335, "step": 4124 }, { "epoch": 25.78125, "grad_norm": 1.1328125, "learning_rate": 4.9154375353576235e-05, "loss": 0.0993, "step": 4125 }, { "epoch": 25.7875, "grad_norm": 1.0625, "learning_rate": 4.915370896083148e-05, "loss": 0.1053, "step": 4126 }, { "epoch": 25.79375, "grad_norm": 0.58984375, "learning_rate": 4.915304231013622e-05, "loss": 0.0685, "step": 4127 }, { "epoch": 25.8, "grad_norm": 0.71484375, "learning_rate": 4.915237540149755e-05, "loss": 0.0686, "step": 4128 }, { "epoch": 25.80625, "grad_norm": 1.6171875, "learning_rate": 4.9151708234922603e-05, "loss": 0.1506, "step": 4129 }, { "epoch": 25.8125, "grad_norm": 1.0703125, "learning_rate": 4.9151040810418505e-05, "loss": 0.1004, "step": 4130 }, { "epoch": 25.81875, "grad_norm": 1.2734375, "learning_rate": 4.9150373127992386e-05, "loss": 0.1166, "step": 4131 }, { "epoch": 25.825, "grad_norm": 0.91015625, "learning_rate": 4.9149705187651376e-05, "loss": 0.0896, "step": 4132 }, { "epoch": 25.83125, "grad_norm": 1.3203125, "learning_rate": 4.9149036989402596e-05, "loss": 0.1183, "step": 4133 }, { "epoch": 25.8375, "grad_norm": 1.7109375, "learning_rate": 4.9148368533253195e-05, "loss": 0.1315, "step": 4134 }, { "epoch": 25.84375, "grad_norm": 1.2109375, "learning_rate": 4.91476998192103e-05, "loss": 0.1048, "step": 4135 }, { "epoch": 25.85, "grad_norm": 1.5390625, "learning_rate": 4.914703084728107e-05, "loss": 0.1819, "step": 4136 }, { "epoch": 25.85625, "grad_norm": 1.125, "learning_rate": 4.9146361617472633e-05, "loss": 0.1066, "step": 4137 }, { "epoch": 25.8625, "grad_norm": 0.90234375, "learning_rate": 4.914569212979215e-05, "loss": 0.0955, "step": 4138 }, { "epoch": 25.86875, "grad_norm": 1.6640625, "learning_rate": 4.9145022384246764e-05, "loss": 0.1255, "step": 4139 }, { "epoch": 25.875, "grad_norm": 1.03125, "learning_rate": 4.914435238084363e-05, "loss": 0.1434, "step": 4140 }, { "epoch": 25.88125, "grad_norm": 1.2578125, "learning_rate": 4.914368211958989e-05, "loss": 0.1288, "step": 4141 }, { "epoch": 25.8875, "grad_norm": 1.46875, "learning_rate": 4.9143011600492716e-05, "loss": 0.13, "step": 4142 }, { "epoch": 25.89375, "grad_norm": 0.94140625, "learning_rate": 4.914234082355927e-05, "loss": 0.0736, "step": 4143 }, { "epoch": 25.9, "grad_norm": 1.390625, "learning_rate": 4.914166978879671e-05, "loss": 0.1229, "step": 4144 }, { "epoch": 25.90625, "grad_norm": 1.3359375, "learning_rate": 4.91409984962122e-05, "loss": 0.1115, "step": 4145 }, { "epoch": 25.9125, "grad_norm": 1.4375, "learning_rate": 4.914032694581291e-05, "loss": 0.1417, "step": 4146 }, { "epoch": 25.91875, "grad_norm": 1.3046875, "learning_rate": 4.9139655137606015e-05, "loss": 0.1481, "step": 4147 }, { "epoch": 25.925, "grad_norm": 1.3671875, "learning_rate": 4.913898307159869e-05, "loss": 0.1474, "step": 4148 }, { "epoch": 25.93125, "grad_norm": 1.5234375, "learning_rate": 4.9138310747798115e-05, "loss": 0.1856, "step": 4149 }, { "epoch": 25.9375, "grad_norm": 1.4765625, "learning_rate": 4.9137638166211465e-05, "loss": 0.1259, "step": 4150 }, { "epoch": 25.94375, "grad_norm": 1.296875, "learning_rate": 4.913696532684592e-05, "loss": 0.1137, "step": 4151 }, { "epoch": 25.95, "grad_norm": 0.984375, "learning_rate": 4.913629222970868e-05, "loss": 0.1011, "step": 4152 }, { "epoch": 25.95625, "grad_norm": 1.1171875, "learning_rate": 4.9135618874806914e-05, "loss": 0.1012, "step": 4153 }, { "epoch": 25.9625, "grad_norm": 1.375, "learning_rate": 4.9134945262147826e-05, "loss": 0.1689, "step": 4154 }, { "epoch": 25.96875, "grad_norm": 1.1875, "learning_rate": 4.9134271391738605e-05, "loss": 0.1097, "step": 4155 }, { "epoch": 25.975, "grad_norm": 1.140625, "learning_rate": 4.913359726358645e-05, "loss": 0.0874, "step": 4156 }, { "epoch": 25.98125, "grad_norm": 0.98828125, "learning_rate": 4.913292287769856e-05, "loss": 0.0869, "step": 4157 }, { "epoch": 25.9875, "grad_norm": 1.9375, "learning_rate": 4.913224823408214e-05, "loss": 0.1558, "step": 4158 }, { "epoch": 25.99375, "grad_norm": 0.70703125, "learning_rate": 4.913157333274438e-05, "loss": 0.0782, "step": 4159 }, { "epoch": 26.0, "grad_norm": 0.7578125, "learning_rate": 4.913089817369251e-05, "loss": 0.08, "step": 4160 }, { "epoch": 26.00625, "grad_norm": 1.421875, "learning_rate": 4.913022275693372e-05, "loss": 0.1391, "step": 4161 }, { "epoch": 26.0125, "grad_norm": 0.578125, "learning_rate": 4.912954708247524e-05, "loss": 0.0613, "step": 4162 }, { "epoch": 26.01875, "grad_norm": 1.4453125, "learning_rate": 4.9128871150324276e-05, "loss": 0.1557, "step": 4163 }, { "epoch": 26.025, "grad_norm": 1.515625, "learning_rate": 4.912819496048804e-05, "loss": 0.166, "step": 4164 }, { "epoch": 26.03125, "grad_norm": 1.2578125, "learning_rate": 4.9127518512973774e-05, "loss": 0.1008, "step": 4165 }, { "epoch": 26.0375, "grad_norm": 1.0546875, "learning_rate": 4.912684180778869e-05, "loss": 0.1187, "step": 4166 }, { "epoch": 26.04375, "grad_norm": 0.86328125, "learning_rate": 4.912616484494001e-05, "loss": 0.0855, "step": 4167 }, { "epoch": 26.05, "grad_norm": 1.234375, "learning_rate": 4.912548762443497e-05, "loss": 0.0986, "step": 4168 }, { "epoch": 26.05625, "grad_norm": 1.5234375, "learning_rate": 4.91248101462808e-05, "loss": 0.1257, "step": 4169 }, { "epoch": 26.0625, "grad_norm": 0.9921875, "learning_rate": 4.912413241048474e-05, "loss": 0.0796, "step": 4170 }, { "epoch": 26.06875, "grad_norm": 2.3125, "learning_rate": 4.9123454417054025e-05, "loss": 0.246, "step": 4171 }, { "epoch": 26.075, "grad_norm": 1.4609375, "learning_rate": 4.91227761659959e-05, "loss": 0.1351, "step": 4172 }, { "epoch": 26.08125, "grad_norm": 1.4140625, "learning_rate": 4.91220976573176e-05, "loss": 0.1496, "step": 4173 }, { "epoch": 26.0875, "grad_norm": 1.3125, "learning_rate": 4.9121418891026364e-05, "loss": 0.1166, "step": 4174 }, { "epoch": 26.09375, "grad_norm": 1.171875, "learning_rate": 4.912073986712946e-05, "loss": 0.0954, "step": 4175 }, { "epoch": 26.1, "grad_norm": 1.890625, "learning_rate": 4.912006058563413e-05, "loss": 0.127, "step": 4176 }, { "epoch": 26.10625, "grad_norm": 1.328125, "learning_rate": 4.9119381046547636e-05, "loss": 0.1363, "step": 4177 }, { "epoch": 26.1125, "grad_norm": 0.83984375, "learning_rate": 4.9118701249877224e-05, "loss": 0.0682, "step": 4178 }, { "epoch": 26.11875, "grad_norm": 0.95703125, "learning_rate": 4.911802119563016e-05, "loss": 0.0905, "step": 4179 }, { "epoch": 26.125, "grad_norm": 1.5859375, "learning_rate": 4.91173408838137e-05, "loss": 0.1206, "step": 4180 }, { "epoch": 26.13125, "grad_norm": 1.421875, "learning_rate": 4.9116660314435124e-05, "loss": 0.1169, "step": 4181 }, { "epoch": 26.1375, "grad_norm": 0.77734375, "learning_rate": 4.9115979487501686e-05, "loss": 0.0745, "step": 4182 }, { "epoch": 26.14375, "grad_norm": 0.80859375, "learning_rate": 4.9115298403020657e-05, "loss": 0.0776, "step": 4183 }, { "epoch": 26.15, "grad_norm": 2.734375, "learning_rate": 4.911461706099933e-05, "loss": 0.1999, "step": 4184 }, { "epoch": 26.15625, "grad_norm": 2.0625, "learning_rate": 4.9113935461444955e-05, "loss": 0.1272, "step": 4185 }, { "epoch": 26.1625, "grad_norm": 1.296875, "learning_rate": 4.9113253604364826e-05, "loss": 0.1293, "step": 4186 }, { "epoch": 26.16875, "grad_norm": 1.890625, "learning_rate": 4.911257148976622e-05, "loss": 0.2086, "step": 4187 }, { "epoch": 26.175, "grad_norm": 1.0546875, "learning_rate": 4.911188911765643e-05, "loss": 0.1007, "step": 4188 }, { "epoch": 26.18125, "grad_norm": 1.0546875, "learning_rate": 4.9111206488042733e-05, "loss": 0.1475, "step": 4189 }, { "epoch": 26.1875, "grad_norm": 1.4375, "learning_rate": 4.9110523600932423e-05, "loss": 0.1104, "step": 4190 }, { "epoch": 26.19375, "grad_norm": 0.78515625, "learning_rate": 4.91098404563328e-05, "loss": 0.0782, "step": 4191 }, { "epoch": 26.2, "grad_norm": 1.2421875, "learning_rate": 4.910915705425114e-05, "loss": 0.1149, "step": 4192 }, { "epoch": 26.20625, "grad_norm": 1.5625, "learning_rate": 4.910847339469477e-05, "loss": 0.155, "step": 4193 }, { "epoch": 26.2125, "grad_norm": 1.234375, "learning_rate": 4.9107789477670964e-05, "loss": 0.13, "step": 4194 }, { "epoch": 26.21875, "grad_norm": 0.953125, "learning_rate": 4.910710530318704e-05, "loss": 0.0876, "step": 4195 }, { "epoch": 26.225, "grad_norm": 0.83203125, "learning_rate": 4.91064208712503e-05, "loss": 0.0835, "step": 4196 }, { "epoch": 26.23125, "grad_norm": 1.609375, "learning_rate": 4.910573618186807e-05, "loss": 0.1565, "step": 4197 }, { "epoch": 26.2375, "grad_norm": 1.453125, "learning_rate": 4.9105051235047636e-05, "loss": 0.1441, "step": 4198 }, { "epoch": 26.24375, "grad_norm": 1.046875, "learning_rate": 4.910436603079632e-05, "loss": 0.0936, "step": 4199 }, { "epoch": 26.25, "grad_norm": 1.265625, "learning_rate": 4.910368056912145e-05, "loss": 0.1174, "step": 4200 }, { "epoch": 26.25625, "grad_norm": 1.546875, "learning_rate": 4.9102994850030344e-05, "loss": 0.1255, "step": 4201 }, { "epoch": 26.2625, "grad_norm": 1.3203125, "learning_rate": 4.9102308873530314e-05, "loss": 0.1621, "step": 4202 }, { "epoch": 26.26875, "grad_norm": 1.0859375, "learning_rate": 4.91016226396287e-05, "loss": 0.1006, "step": 4203 }, { "epoch": 26.275, "grad_norm": 1.140625, "learning_rate": 4.9100936148332824e-05, "loss": 0.088, "step": 4204 }, { "epoch": 26.28125, "grad_norm": 1.3046875, "learning_rate": 4.910024939965002e-05, "loss": 0.1371, "step": 4205 }, { "epoch": 26.2875, "grad_norm": 0.953125, "learning_rate": 4.909956239358762e-05, "loss": 0.0971, "step": 4206 }, { "epoch": 26.29375, "grad_norm": 1.75, "learning_rate": 4.909887513015295e-05, "loss": 0.1314, "step": 4207 }, { "epoch": 26.3, "grad_norm": 1.4453125, "learning_rate": 4.9098187609353376e-05, "loss": 0.0987, "step": 4208 }, { "epoch": 26.30625, "grad_norm": 1.15625, "learning_rate": 4.9097499831196216e-05, "loss": 0.1176, "step": 4209 }, { "epoch": 26.3125, "grad_norm": 1.09375, "learning_rate": 4.909681179568882e-05, "loss": 0.0772, "step": 4210 }, { "epoch": 26.31875, "grad_norm": 1.2109375, "learning_rate": 4.9096123502838557e-05, "loss": 0.1065, "step": 4211 }, { "epoch": 26.325, "grad_norm": 1.0703125, "learning_rate": 4.909543495265275e-05, "loss": 0.0918, "step": 4212 }, { "epoch": 26.33125, "grad_norm": 1.234375, "learning_rate": 4.909474614513876e-05, "loss": 0.1307, "step": 4213 }, { "epoch": 26.3375, "grad_norm": 1.53125, "learning_rate": 4.909405708030395e-05, "loss": 0.1806, "step": 4214 }, { "epoch": 26.34375, "grad_norm": 1.5546875, "learning_rate": 4.9093367758155674e-05, "loss": 0.1193, "step": 4215 }, { "epoch": 26.35, "grad_norm": 0.90234375, "learning_rate": 4.9092678178701304e-05, "loss": 0.0817, "step": 4216 }, { "epoch": 26.35625, "grad_norm": 1.4296875, "learning_rate": 4.909198834194818e-05, "loss": 0.1863, "step": 4217 }, { "epoch": 26.3625, "grad_norm": 1.2421875, "learning_rate": 4.90912982479037e-05, "loss": 0.1127, "step": 4218 }, { "epoch": 26.36875, "grad_norm": 1.0390625, "learning_rate": 4.90906078965752e-05, "loss": 0.0982, "step": 4219 }, { "epoch": 26.375, "grad_norm": 1.09375, "learning_rate": 4.9089917287970086e-05, "loss": 0.0945, "step": 4220 }, { "epoch": 26.38125, "grad_norm": 1.140625, "learning_rate": 4.908922642209571e-05, "loss": 0.0976, "step": 4221 }, { "epoch": 26.3875, "grad_norm": 1.75, "learning_rate": 4.9088535298959464e-05, "loss": 0.1725, "step": 4222 }, { "epoch": 26.39375, "grad_norm": 1.25, "learning_rate": 4.908784391856872e-05, "loss": 0.1142, "step": 4223 }, { "epoch": 26.4, "grad_norm": 0.78125, "learning_rate": 4.908715228093087e-05, "loss": 0.0714, "step": 4224 }, { "epoch": 26.40625, "grad_norm": 1.7890625, "learning_rate": 4.908646038605329e-05, "loss": 0.1623, "step": 4225 }, { "epoch": 26.4125, "grad_norm": 1.109375, "learning_rate": 4.9085768233943374e-05, "loss": 0.0994, "step": 4226 }, { "epoch": 26.41875, "grad_norm": 1.328125, "learning_rate": 4.9085075824608516e-05, "loss": 0.1371, "step": 4227 }, { "epoch": 26.425, "grad_norm": 1.609375, "learning_rate": 4.908438315805611e-05, "loss": 0.1474, "step": 4228 }, { "epoch": 26.43125, "grad_norm": 1.703125, "learning_rate": 4.9083690234293553e-05, "loss": 0.1456, "step": 4229 }, { "epoch": 26.4375, "grad_norm": 1.28125, "learning_rate": 4.908299705332824e-05, "loss": 0.0885, "step": 4230 }, { "epoch": 26.44375, "grad_norm": 0.87109375, "learning_rate": 4.908230361516758e-05, "loss": 0.0738, "step": 4231 }, { "epoch": 26.45, "grad_norm": 1.5546875, "learning_rate": 4.908160991981898e-05, "loss": 0.15, "step": 4232 }, { "epoch": 26.45625, "grad_norm": 1.1484375, "learning_rate": 4.908091596728984e-05, "loss": 0.1391, "step": 4233 }, { "epoch": 26.4625, "grad_norm": 1.3125, "learning_rate": 4.9080221757587585e-05, "loss": 0.1005, "step": 4234 }, { "epoch": 26.46875, "grad_norm": 1.203125, "learning_rate": 4.9079527290719614e-05, "loss": 0.0992, "step": 4235 }, { "epoch": 26.475, "grad_norm": 0.96875, "learning_rate": 4.907883256669335e-05, "loss": 0.077, "step": 4236 }, { "epoch": 26.48125, "grad_norm": 1.203125, "learning_rate": 4.9078137585516215e-05, "loss": 0.112, "step": 4237 }, { "epoch": 26.4875, "grad_norm": 1.3828125, "learning_rate": 4.907744234719562e-05, "loss": 0.1133, "step": 4238 }, { "epoch": 26.49375, "grad_norm": 1.46875, "learning_rate": 4.907674685173901e-05, "loss": 0.0842, "step": 4239 }, { "epoch": 26.5, "grad_norm": 1.7890625, "learning_rate": 4.907605109915379e-05, "loss": 0.1985, "step": 4240 }, { "epoch": 26.50625, "grad_norm": 1.2734375, "learning_rate": 4.907535508944741e-05, "loss": 0.1154, "step": 4241 }, { "epoch": 26.5125, "grad_norm": 1.5546875, "learning_rate": 4.907465882262728e-05, "loss": 0.1549, "step": 4242 }, { "epoch": 26.51875, "grad_norm": 1.1953125, "learning_rate": 4.907396229870086e-05, "loss": 0.1042, "step": 4243 }, { "epoch": 26.525, "grad_norm": 0.703125, "learning_rate": 4.9073265517675584e-05, "loss": 0.0789, "step": 4244 }, { "epoch": 26.53125, "grad_norm": 1.6796875, "learning_rate": 4.907256847955888e-05, "loss": 0.2085, "step": 4245 }, { "epoch": 26.5375, "grad_norm": 1.4765625, "learning_rate": 4.90718711843582e-05, "loss": 0.1929, "step": 4246 }, { "epoch": 26.54375, "grad_norm": 1.390625, "learning_rate": 4.907117363208099e-05, "loss": 0.1487, "step": 4247 }, { "epoch": 26.55, "grad_norm": 0.65234375, "learning_rate": 4.90704758227347e-05, "loss": 0.0624, "step": 4248 }, { "epoch": 26.55625, "grad_norm": 2.03125, "learning_rate": 4.906977775632678e-05, "loss": 0.265, "step": 4249 }, { "epoch": 26.5625, "grad_norm": 1.5859375, "learning_rate": 4.906907943286469e-05, "loss": 0.1528, "step": 4250 }, { "epoch": 26.56875, "grad_norm": 1.2265625, "learning_rate": 4.906838085235588e-05, "loss": 0.1158, "step": 4251 }, { "epoch": 26.575, "grad_norm": 1.3984375, "learning_rate": 4.9067682014807825e-05, "loss": 0.1608, "step": 4252 }, { "epoch": 26.58125, "grad_norm": 1.40625, "learning_rate": 4.906698292022798e-05, "loss": 0.177, "step": 4253 }, { "epoch": 26.5875, "grad_norm": 1.1484375, "learning_rate": 4.90662835686238e-05, "loss": 0.0965, "step": 4254 }, { "epoch": 26.59375, "grad_norm": 1.21875, "learning_rate": 4.906558396000277e-05, "loss": 0.0954, "step": 4255 }, { "epoch": 26.6, "grad_norm": 1.25, "learning_rate": 4.9064884094372347e-05, "loss": 0.1154, "step": 4256 }, { "epoch": 26.60625, "grad_norm": 0.9921875, "learning_rate": 4.906418397174002e-05, "loss": 0.0911, "step": 4257 }, { "epoch": 26.6125, "grad_norm": 0.70703125, "learning_rate": 4.9063483592113254e-05, "loss": 0.0659, "step": 4258 }, { "epoch": 26.61875, "grad_norm": 1.2109375, "learning_rate": 4.906278295549954e-05, "loss": 0.1182, "step": 4259 }, { "epoch": 26.625, "grad_norm": 0.78125, "learning_rate": 4.9062082061906355e-05, "loss": 0.0799, "step": 4260 }, { "epoch": 26.63125, "grad_norm": 1.234375, "learning_rate": 4.906138091134118e-05, "loss": 0.1225, "step": 4261 }, { "epoch": 26.6375, "grad_norm": 0.94921875, "learning_rate": 4.9060679503811516e-05, "loss": 0.0964, "step": 4262 }, { "epoch": 26.64375, "grad_norm": 1.125, "learning_rate": 4.905997783932483e-05, "loss": 0.1105, "step": 4263 }, { "epoch": 26.65, "grad_norm": 1.9453125, "learning_rate": 4.905927591788864e-05, "loss": 0.2028, "step": 4264 }, { "epoch": 26.65625, "grad_norm": 1.78125, "learning_rate": 4.905857373951043e-05, "loss": 0.1823, "step": 4265 }, { "epoch": 26.6625, "grad_norm": 1.0625, "learning_rate": 4.90578713041977e-05, "loss": 0.1057, "step": 4266 }, { "epoch": 26.66875, "grad_norm": 1.4296875, "learning_rate": 4.9057168611957954e-05, "loss": 0.131, "step": 4267 }, { "epoch": 26.675, "grad_norm": 1.2265625, "learning_rate": 4.905646566279869e-05, "loss": 0.1317, "step": 4268 }, { "epoch": 26.68125, "grad_norm": 1.3125, "learning_rate": 4.905576245672743e-05, "loss": 0.1641, "step": 4269 }, { "epoch": 26.6875, "grad_norm": 1.7890625, "learning_rate": 4.905505899375167e-05, "loss": 0.1325, "step": 4270 }, { "epoch": 26.69375, "grad_norm": 1.0625, "learning_rate": 4.905435527387892e-05, "loss": 0.1061, "step": 4271 }, { "epoch": 26.7, "grad_norm": 0.96484375, "learning_rate": 4.905365129711671e-05, "loss": 0.0885, "step": 4272 }, { "epoch": 26.70625, "grad_norm": 1.2578125, "learning_rate": 4.905294706347255e-05, "loss": 0.1736, "step": 4273 }, { "epoch": 26.7125, "grad_norm": 0.84375, "learning_rate": 4.9052242572953956e-05, "loss": 0.0789, "step": 4274 }, { "epoch": 26.71875, "grad_norm": 1.71875, "learning_rate": 4.905153782556846e-05, "loss": 0.1587, "step": 4275 }, { "epoch": 26.725, "grad_norm": 1.5234375, "learning_rate": 4.905083282132359e-05, "loss": 0.1351, "step": 4276 }, { "epoch": 26.73125, "grad_norm": 1.0625, "learning_rate": 4.905012756022687e-05, "loss": 0.1182, "step": 4277 }, { "epoch": 26.7375, "grad_norm": 1.703125, "learning_rate": 4.904942204228582e-05, "loss": 0.1459, "step": 4278 }, { "epoch": 26.74375, "grad_norm": 1.1875, "learning_rate": 4.9048716267508e-05, "loss": 0.1117, "step": 4279 }, { "epoch": 26.75, "grad_norm": 1.734375, "learning_rate": 4.9048010235900935e-05, "loss": 0.1683, "step": 4280 }, { "epoch": 26.75625, "grad_norm": 1.609375, "learning_rate": 4.904730394747216e-05, "loss": 0.2264, "step": 4281 }, { "epoch": 26.7625, "grad_norm": 1.78125, "learning_rate": 4.904659740222921e-05, "loss": 0.1975, "step": 4282 }, { "epoch": 26.76875, "grad_norm": 2.328125, "learning_rate": 4.904589060017966e-05, "loss": 0.2172, "step": 4283 }, { "epoch": 26.775, "grad_norm": 1.2890625, "learning_rate": 4.904518354133103e-05, "loss": 0.1129, "step": 4284 }, { "epoch": 26.78125, "grad_norm": 1.921875, "learning_rate": 4.9044476225690894e-05, "loss": 0.2107, "step": 4285 }, { "epoch": 26.7875, "grad_norm": 1.3984375, "learning_rate": 4.904376865326679e-05, "loss": 0.1743, "step": 4286 }, { "epoch": 26.79375, "grad_norm": 1.5625, "learning_rate": 4.9043060824066274e-05, "loss": 0.1227, "step": 4287 }, { "epoch": 26.8, "grad_norm": 0.69921875, "learning_rate": 4.9042352738096906e-05, "loss": 0.0695, "step": 4288 }, { "epoch": 26.80625, "grad_norm": 1.8984375, "learning_rate": 4.904164439536626e-05, "loss": 0.2142, "step": 4289 }, { "epoch": 26.8125, "grad_norm": 2.03125, "learning_rate": 4.904093579588189e-05, "loss": 0.2136, "step": 4290 }, { "epoch": 26.81875, "grad_norm": 0.87109375, "learning_rate": 4.904022693965137e-05, "loss": 0.0822, "step": 4291 }, { "epoch": 26.825, "grad_norm": 1.3359375, "learning_rate": 4.9039517826682266e-05, "loss": 0.1031, "step": 4292 }, { "epoch": 26.83125, "grad_norm": 1.3671875, "learning_rate": 4.9038808456982146e-05, "loss": 0.1875, "step": 4293 }, { "epoch": 26.8375, "grad_norm": 1.6015625, "learning_rate": 4.903809883055859e-05, "loss": 0.1524, "step": 4294 }, { "epoch": 26.84375, "grad_norm": 1.0703125, "learning_rate": 4.9037388947419185e-05, "loss": 0.0936, "step": 4295 }, { "epoch": 26.85, "grad_norm": 0.94140625, "learning_rate": 4.90366788075715e-05, "loss": 0.0743, "step": 4296 }, { "epoch": 26.85625, "grad_norm": 0.83203125, "learning_rate": 4.9035968411023126e-05, "loss": 0.0756, "step": 4297 }, { "epoch": 26.8625, "grad_norm": 1.203125, "learning_rate": 4.903525775778165e-05, "loss": 0.0933, "step": 4298 }, { "epoch": 26.86875, "grad_norm": 1.046875, "learning_rate": 4.9034546847854656e-05, "loss": 0.1008, "step": 4299 }, { "epoch": 26.875, "grad_norm": 1.1484375, "learning_rate": 4.9033835681249735e-05, "loss": 0.1108, "step": 4300 }, { "epoch": 26.88125, "grad_norm": 1.5078125, "learning_rate": 4.903312425797449e-05, "loss": 0.1348, "step": 4301 }, { "epoch": 26.8875, "grad_norm": 1.1328125, "learning_rate": 4.9032412578036514e-05, "loss": 0.0835, "step": 4302 }, { "epoch": 26.89375, "grad_norm": 0.65234375, "learning_rate": 4.9031700641443414e-05, "loss": 0.0623, "step": 4303 }, { "epoch": 26.9, "grad_norm": 0.85546875, "learning_rate": 4.903098844820278e-05, "loss": 0.0725, "step": 4304 }, { "epoch": 26.90625, "grad_norm": 1.9140625, "learning_rate": 4.903027599832223e-05, "loss": 0.2193, "step": 4305 }, { "epoch": 26.9125, "grad_norm": 0.75390625, "learning_rate": 4.9029563291809363e-05, "loss": 0.0761, "step": 4306 }, { "epoch": 26.91875, "grad_norm": 1.0, "learning_rate": 4.902885032867179e-05, "loss": 0.0955, "step": 4307 }, { "epoch": 26.925, "grad_norm": 1.40625, "learning_rate": 4.902813710891715e-05, "loss": 0.174, "step": 4308 }, { "epoch": 26.93125, "grad_norm": 1.390625, "learning_rate": 4.9027423632553025e-05, "loss": 0.1167, "step": 4309 }, { "epoch": 26.9375, "grad_norm": 1.6875, "learning_rate": 4.902670989958705e-05, "loss": 0.2334, "step": 4310 }, { "epoch": 26.94375, "grad_norm": 1.15625, "learning_rate": 4.9025995910026856e-05, "loss": 0.1207, "step": 4311 }, { "epoch": 26.95, "grad_norm": 1.171875, "learning_rate": 4.902528166388005e-05, "loss": 0.0981, "step": 4312 }, { "epoch": 26.95625, "grad_norm": 1.3671875, "learning_rate": 4.902456716115428e-05, "loss": 0.1153, "step": 4313 }, { "epoch": 26.9625, "grad_norm": 1.0234375, "learning_rate": 4.902385240185715e-05, "loss": 0.1214, "step": 4314 }, { "epoch": 26.96875, "grad_norm": 1.25, "learning_rate": 4.902313738599632e-05, "loss": 0.1107, "step": 4315 }, { "epoch": 26.975, "grad_norm": 1.6484375, "learning_rate": 4.9022422113579414e-05, "loss": 0.1536, "step": 4316 }, { "epoch": 26.98125, "grad_norm": 1.734375, "learning_rate": 4.902170658461407e-05, "loss": 0.1335, "step": 4317 }, { "epoch": 26.9875, "grad_norm": 1.0, "learning_rate": 4.9020990799107935e-05, "loss": 0.0874, "step": 4318 }, { "epoch": 26.99375, "grad_norm": 0.95703125, "learning_rate": 4.902027475706864e-05, "loss": 0.0903, "step": 4319 }, { "epoch": 27.0, "grad_norm": 1.984375, "learning_rate": 4.901955845850385e-05, "loss": 0.2274, "step": 4320 }, { "epoch": 27.00625, "grad_norm": 2.140625, "learning_rate": 4.901884190342121e-05, "loss": 0.211, "step": 4321 }, { "epoch": 27.0125, "grad_norm": 1.0859375, "learning_rate": 4.901812509182836e-05, "loss": 0.1049, "step": 4322 }, { "epoch": 27.01875, "grad_norm": 1.046875, "learning_rate": 4.9017408023732967e-05, "loss": 0.0934, "step": 4323 }, { "epoch": 27.025, "grad_norm": 1.0234375, "learning_rate": 4.901669069914269e-05, "loss": 0.085, "step": 4324 }, { "epoch": 27.03125, "grad_norm": 1.34375, "learning_rate": 4.901597311806518e-05, "loss": 0.1749, "step": 4325 }, { "epoch": 27.0375, "grad_norm": 0.609375, "learning_rate": 4.90152552805081e-05, "loss": 0.055, "step": 4326 }, { "epoch": 27.04375, "grad_norm": 1.40625, "learning_rate": 4.901453718647914e-05, "loss": 0.1605, "step": 4327 }, { "epoch": 27.05, "grad_norm": 1.0546875, "learning_rate": 4.9013818835985934e-05, "loss": 0.096, "step": 4328 }, { "epoch": 27.05625, "grad_norm": 0.9140625, "learning_rate": 4.901310022903618e-05, "loss": 0.0856, "step": 4329 }, { "epoch": 27.0625, "grad_norm": 1.2578125, "learning_rate": 4.901238136563754e-05, "loss": 0.1739, "step": 4330 }, { "epoch": 27.06875, "grad_norm": 1.0, "learning_rate": 4.90116622457977e-05, "loss": 0.0918, "step": 4331 }, { "epoch": 27.075, "grad_norm": 1.09375, "learning_rate": 4.901094286952433e-05, "loss": 0.1314, "step": 4332 }, { "epoch": 27.08125, "grad_norm": 0.9375, "learning_rate": 4.9010223236825116e-05, "loss": 0.0864, "step": 4333 }, { "epoch": 27.0875, "grad_norm": 1.2890625, "learning_rate": 4.9009503347707744e-05, "loss": 0.121, "step": 4334 }, { "epoch": 27.09375, "grad_norm": 1.1484375, "learning_rate": 4.9008783202179904e-05, "loss": 0.1347, "step": 4335 }, { "epoch": 27.1, "grad_norm": 0.7734375, "learning_rate": 4.900806280024929e-05, "loss": 0.0761, "step": 4336 }, { "epoch": 27.10625, "grad_norm": 1.453125, "learning_rate": 4.900734214192358e-05, "loss": 0.1083, "step": 4337 }, { "epoch": 27.1125, "grad_norm": 1.0078125, "learning_rate": 4.9006621227210486e-05, "loss": 0.0815, "step": 4338 }, { "epoch": 27.11875, "grad_norm": 1.1328125, "learning_rate": 4.9005900056117706e-05, "loss": 0.0952, "step": 4339 }, { "epoch": 27.125, "grad_norm": 0.99609375, "learning_rate": 4.900517862865293e-05, "loss": 0.0967, "step": 4340 }, { "epoch": 27.13125, "grad_norm": 1.0390625, "learning_rate": 4.900445694482387e-05, "loss": 0.1417, "step": 4341 }, { "epoch": 27.1375, "grad_norm": 1.109375, "learning_rate": 4.900373500463824e-05, "loss": 0.0954, "step": 4342 }, { "epoch": 27.14375, "grad_norm": 1.4375, "learning_rate": 4.9003012808103744e-05, "loss": 0.1197, "step": 4343 }, { "epoch": 27.15, "grad_norm": 1.21875, "learning_rate": 4.9002290355228084e-05, "loss": 0.1298, "step": 4344 }, { "epoch": 27.15625, "grad_norm": 0.87890625, "learning_rate": 4.9001567646018996e-05, "loss": 0.0758, "step": 4345 }, { "epoch": 27.1625, "grad_norm": 1.46875, "learning_rate": 4.9000844680484187e-05, "loss": 0.1591, "step": 4346 }, { "epoch": 27.16875, "grad_norm": 1.1953125, "learning_rate": 4.900012145863138e-05, "loss": 0.0838, "step": 4347 }, { "epoch": 27.175, "grad_norm": 1.3984375, "learning_rate": 4.899939798046829e-05, "loss": 0.1486, "step": 4348 }, { "epoch": 27.18125, "grad_norm": 1.078125, "learning_rate": 4.899867424600266e-05, "loss": 0.0897, "step": 4349 }, { "epoch": 27.1875, "grad_norm": 1.09375, "learning_rate": 4.89979502552422e-05, "loss": 0.114, "step": 4350 }, { "epoch": 27.19375, "grad_norm": 1.671875, "learning_rate": 4.899722600819466e-05, "loss": 0.1542, "step": 4351 }, { "epoch": 27.2, "grad_norm": 1.6953125, "learning_rate": 4.899650150486776e-05, "loss": 0.1619, "step": 4352 }, { "epoch": 27.20625, "grad_norm": 1.53125, "learning_rate": 4.8995776745269254e-05, "loss": 0.1559, "step": 4353 }, { "epoch": 27.2125, "grad_norm": 1.1015625, "learning_rate": 4.8995051729406865e-05, "loss": 0.0925, "step": 4354 }, { "epoch": 27.21875, "grad_norm": 1.3515625, "learning_rate": 4.8994326457288345e-05, "loss": 0.1794, "step": 4355 }, { "epoch": 27.225, "grad_norm": 2.40625, "learning_rate": 4.8993600928921434e-05, "loss": 0.1773, "step": 4356 }, { "epoch": 27.23125, "grad_norm": 1.3203125, "learning_rate": 4.899287514431389e-05, "loss": 0.0894, "step": 4357 }, { "epoch": 27.2375, "grad_norm": 1.3671875, "learning_rate": 4.899214910347346e-05, "loss": 0.161, "step": 4358 }, { "epoch": 27.24375, "grad_norm": 2.046875, "learning_rate": 4.8991422806407885e-05, "loss": 0.2554, "step": 4359 }, { "epoch": 27.25, "grad_norm": 2.328125, "learning_rate": 4.8990696253124935e-05, "loss": 0.2312, "step": 4360 }, { "epoch": 27.25625, "grad_norm": 0.8984375, "learning_rate": 4.8989969443632366e-05, "loss": 0.0712, "step": 4361 }, { "epoch": 27.2625, "grad_norm": 1.640625, "learning_rate": 4.898924237793795e-05, "loss": 0.0879, "step": 4362 }, { "epoch": 27.26875, "grad_norm": 1.7109375, "learning_rate": 4.898851505604943e-05, "loss": 0.1505, "step": 4363 }, { "epoch": 27.275, "grad_norm": 1.0859375, "learning_rate": 4.898778747797459e-05, "loss": 0.1268, "step": 4364 }, { "epoch": 27.28125, "grad_norm": 0.80078125, "learning_rate": 4.89870596437212e-05, "loss": 0.072, "step": 4365 }, { "epoch": 27.2875, "grad_norm": 1.3828125, "learning_rate": 4.8986331553297026e-05, "loss": 0.0881, "step": 4366 }, { "epoch": 27.29375, "grad_norm": 1.4609375, "learning_rate": 4.898560320670985e-05, "loss": 0.1021, "step": 4367 }, { "epoch": 27.3, "grad_norm": 1.6796875, "learning_rate": 4.898487460396743e-05, "loss": 0.1142, "step": 4368 }, { "epoch": 27.30625, "grad_norm": 1.5625, "learning_rate": 4.8984145745077584e-05, "loss": 0.1497, "step": 4369 }, { "epoch": 27.3125, "grad_norm": 0.8984375, "learning_rate": 4.898341663004807e-05, "loss": 0.0647, "step": 4370 }, { "epoch": 27.31875, "grad_norm": 1.3046875, "learning_rate": 4.8982687258886675e-05, "loss": 0.1303, "step": 4371 }, { "epoch": 27.325, "grad_norm": 1.3671875, "learning_rate": 4.89819576316012e-05, "loss": 0.0889, "step": 4372 }, { "epoch": 27.33125, "grad_norm": 1.71875, "learning_rate": 4.898122774819943e-05, "loss": 0.1738, "step": 4373 }, { "epoch": 27.3375, "grad_norm": 1.28125, "learning_rate": 4.898049760868916e-05, "loss": 0.1573, "step": 4374 }, { "epoch": 27.34375, "grad_norm": 1.5234375, "learning_rate": 4.897976721307819e-05, "loss": 0.1109, "step": 4375 }, { "epoch": 27.35, "grad_norm": 1.578125, "learning_rate": 4.897903656137431e-05, "loss": 0.0796, "step": 4376 }, { "epoch": 27.35625, "grad_norm": 1.703125, "learning_rate": 4.897830565358534e-05, "loss": 0.1307, "step": 4377 }, { "epoch": 27.3625, "grad_norm": 1.2890625, "learning_rate": 4.897757448971907e-05, "loss": 0.1277, "step": 4378 }, { "epoch": 27.36875, "grad_norm": 1.1484375, "learning_rate": 4.8976843069783324e-05, "loss": 0.1166, "step": 4379 }, { "epoch": 27.375, "grad_norm": 1.8671875, "learning_rate": 4.89761113937859e-05, "loss": 0.1093, "step": 4380 }, { "epoch": 27.38125, "grad_norm": 2.234375, "learning_rate": 4.8975379461734615e-05, "loss": 0.1712, "step": 4381 }, { "epoch": 27.3875, "grad_norm": 1.7734375, "learning_rate": 4.89746472736373e-05, "loss": 0.1, "step": 4382 }, { "epoch": 27.39375, "grad_norm": 1.421875, "learning_rate": 4.897391482950175e-05, "loss": 0.1266, "step": 4383 }, { "epoch": 27.4, "grad_norm": 1.671875, "learning_rate": 4.89731821293358e-05, "loss": 0.1715, "step": 4384 }, { "epoch": 27.40625, "grad_norm": 1.421875, "learning_rate": 4.8972449173147276e-05, "loss": 0.1106, "step": 4385 }, { "epoch": 27.4125, "grad_norm": 1.2109375, "learning_rate": 4.897171596094401e-05, "loss": 0.0889, "step": 4386 }, { "epoch": 27.41875, "grad_norm": 1.4609375, "learning_rate": 4.897098249273382e-05, "loss": 0.1087, "step": 4387 }, { "epoch": 27.425, "grad_norm": 1.4921875, "learning_rate": 4.8970248768524545e-05, "loss": 0.1216, "step": 4388 }, { "epoch": 27.43125, "grad_norm": 1.359375, "learning_rate": 4.896951478832402e-05, "loss": 0.1824, "step": 4389 }, { "epoch": 27.4375, "grad_norm": 1.3203125, "learning_rate": 4.896878055214009e-05, "loss": 0.0979, "step": 4390 }, { "epoch": 27.44375, "grad_norm": 1.03125, "learning_rate": 4.896804605998059e-05, "loss": 0.0848, "step": 4391 }, { "epoch": 27.45, "grad_norm": 1.203125, "learning_rate": 4.896731131185336e-05, "loss": 0.1273, "step": 4392 }, { "epoch": 27.45625, "grad_norm": 0.984375, "learning_rate": 4.896657630776626e-05, "loss": 0.1165, "step": 4393 }, { "epoch": 27.4625, "grad_norm": 1.375, "learning_rate": 4.8965841047727125e-05, "loss": 0.1106, "step": 4394 }, { "epoch": 27.46875, "grad_norm": 1.34375, "learning_rate": 4.8965105531743815e-05, "loss": 0.1068, "step": 4395 }, { "epoch": 27.475, "grad_norm": 0.99609375, "learning_rate": 4.896436975982418e-05, "loss": 0.088, "step": 4396 }, { "epoch": 27.48125, "grad_norm": 0.94921875, "learning_rate": 4.896363373197609e-05, "loss": 0.1006, "step": 4397 }, { "epoch": 27.4875, "grad_norm": 1.171875, "learning_rate": 4.896289744820739e-05, "loss": 0.0943, "step": 4398 }, { "epoch": 27.49375, "grad_norm": 1.1015625, "learning_rate": 4.896216090852595e-05, "loss": 0.0829, "step": 4399 }, { "epoch": 27.5, "grad_norm": 0.7265625, "learning_rate": 4.8961424112939634e-05, "loss": 0.0758, "step": 4400 }, { "epoch": 27.50625, "grad_norm": 0.93359375, "learning_rate": 4.8960687061456324e-05, "loss": 0.0845, "step": 4401 }, { "epoch": 27.5125, "grad_norm": 1.390625, "learning_rate": 4.895994975408387e-05, "loss": 0.1388, "step": 4402 }, { "epoch": 27.51875, "grad_norm": 1.1171875, "learning_rate": 4.895921219083016e-05, "loss": 0.0957, "step": 4403 }, { "epoch": 27.525, "grad_norm": 1.859375, "learning_rate": 4.895847437170305e-05, "loss": 0.1698, "step": 4404 }, { "epoch": 27.53125, "grad_norm": 1.109375, "learning_rate": 4.895773629671046e-05, "loss": 0.0737, "step": 4405 }, { "epoch": 27.5375, "grad_norm": 1.671875, "learning_rate": 4.895699796586024e-05, "loss": 0.1266, "step": 4406 }, { "epoch": 27.54375, "grad_norm": 0.75390625, "learning_rate": 4.8956259379160284e-05, "loss": 0.0753, "step": 4407 }, { "epoch": 27.55, "grad_norm": 1.09375, "learning_rate": 4.8955520536618476e-05, "loss": 0.1091, "step": 4408 }, { "epoch": 27.55625, "grad_norm": 1.71875, "learning_rate": 4.895478143824271e-05, "loss": 0.1315, "step": 4409 }, { "epoch": 27.5625, "grad_norm": 1.6484375, "learning_rate": 4.895404208404089e-05, "loss": 0.1536, "step": 4410 }, { "epoch": 27.56875, "grad_norm": 1.4453125, "learning_rate": 4.895330247402089e-05, "loss": 0.1429, "step": 4411 }, { "epoch": 27.575, "grad_norm": 1.2109375, "learning_rate": 4.895256260819062e-05, "loss": 0.1147, "step": 4412 }, { "epoch": 27.58125, "grad_norm": 1.2734375, "learning_rate": 4.8951822486557986e-05, "loss": 0.1373, "step": 4413 }, { "epoch": 27.5875, "grad_norm": 1.53125, "learning_rate": 4.895108210913088e-05, "loss": 0.1172, "step": 4414 }, { "epoch": 27.59375, "grad_norm": 1.1015625, "learning_rate": 4.895034147591722e-05, "loss": 0.0885, "step": 4415 }, { "epoch": 27.6, "grad_norm": 1.0859375, "learning_rate": 4.8949600586924914e-05, "loss": 0.1014, "step": 4416 }, { "epoch": 27.60625, "grad_norm": 0.8359375, "learning_rate": 4.8948859442161874e-05, "loss": 0.0816, "step": 4417 }, { "epoch": 27.6125, "grad_norm": 1.3046875, "learning_rate": 4.894811804163601e-05, "loss": 0.1656, "step": 4418 }, { "epoch": 27.61875, "grad_norm": 1.03125, "learning_rate": 4.894737638535524e-05, "loss": 0.0784, "step": 4419 }, { "epoch": 27.625, "grad_norm": 1.4140625, "learning_rate": 4.894663447332749e-05, "loss": 0.148, "step": 4420 }, { "epoch": 27.63125, "grad_norm": 1.0078125, "learning_rate": 4.8945892305560685e-05, "loss": 0.0785, "step": 4421 }, { "epoch": 27.6375, "grad_norm": 0.82421875, "learning_rate": 4.894514988206275e-05, "loss": 0.0834, "step": 4422 }, { "epoch": 27.64375, "grad_norm": 0.94140625, "learning_rate": 4.89444072028416e-05, "loss": 0.0875, "step": 4423 }, { "epoch": 27.65, "grad_norm": 1.2890625, "learning_rate": 4.894366426790518e-05, "loss": 0.1331, "step": 4424 }, { "epoch": 27.65625, "grad_norm": 1.46875, "learning_rate": 4.8942921077261425e-05, "loss": 0.1212, "step": 4425 }, { "epoch": 27.6625, "grad_norm": 0.8125, "learning_rate": 4.894217763091827e-05, "loss": 0.0679, "step": 4426 }, { "epoch": 27.66875, "grad_norm": 1.40625, "learning_rate": 4.8941433928883654e-05, "loss": 0.1716, "step": 4427 }, { "epoch": 27.675, "grad_norm": 0.796875, "learning_rate": 4.894068997116551e-05, "loss": 0.071, "step": 4428 }, { "epoch": 27.68125, "grad_norm": 1.1171875, "learning_rate": 4.89399457577718e-05, "loss": 0.116, "step": 4429 }, { "epoch": 27.6875, "grad_norm": 1.21875, "learning_rate": 4.893920128871046e-05, "loss": 0.1131, "step": 4430 }, { "epoch": 27.69375, "grad_norm": 0.875, "learning_rate": 4.893845656398944e-05, "loss": 0.0927, "step": 4431 }, { "epoch": 27.7, "grad_norm": 1.0390625, "learning_rate": 4.89377115836167e-05, "loss": 0.1024, "step": 4432 }, { "epoch": 27.70625, "grad_norm": 1.0703125, "learning_rate": 4.893696634760019e-05, "loss": 0.0974, "step": 4433 }, { "epoch": 27.7125, "grad_norm": 0.9921875, "learning_rate": 4.8936220855947876e-05, "loss": 0.0909, "step": 4434 }, { "epoch": 27.71875, "grad_norm": 1.4140625, "learning_rate": 4.893547510866772e-05, "loss": 0.1993, "step": 4435 }, { "epoch": 27.725, "grad_norm": 1.671875, "learning_rate": 4.893472910576767e-05, "loss": 0.1611, "step": 4436 }, { "epoch": 27.73125, "grad_norm": 1.9609375, "learning_rate": 4.8933982847255724e-05, "loss": 0.2462, "step": 4437 }, { "epoch": 27.7375, "grad_norm": 1.46875, "learning_rate": 4.893323633313982e-05, "loss": 0.1, "step": 4438 }, { "epoch": 27.74375, "grad_norm": 1.1484375, "learning_rate": 4.893248956342794e-05, "loss": 0.1005, "step": 4439 }, { "epoch": 27.75, "grad_norm": 1.265625, "learning_rate": 4.893174253812807e-05, "loss": 0.0896, "step": 4440 }, { "epoch": 27.75625, "grad_norm": 2.03125, "learning_rate": 4.893099525724818e-05, "loss": 0.2353, "step": 4441 }, { "epoch": 27.7625, "grad_norm": 1.296875, "learning_rate": 4.893024772079624e-05, "loss": 0.1603, "step": 4442 }, { "epoch": 27.76875, "grad_norm": 1.828125, "learning_rate": 4.8929499928780257e-05, "loss": 0.1328, "step": 4443 }, { "epoch": 27.775, "grad_norm": 0.78125, "learning_rate": 4.892875188120819e-05, "loss": 0.0657, "step": 4444 }, { "epoch": 27.78125, "grad_norm": 1.0859375, "learning_rate": 4.892800357808806e-05, "loss": 0.0731, "step": 4445 }, { "epoch": 27.7875, "grad_norm": 1.046875, "learning_rate": 4.892725501942782e-05, "loss": 0.1087, "step": 4446 }, { "epoch": 27.79375, "grad_norm": 1.3515625, "learning_rate": 4.89265062052355e-05, "loss": 0.1379, "step": 4447 }, { "epoch": 27.8, "grad_norm": 1.390625, "learning_rate": 4.892575713551907e-05, "loss": 0.1543, "step": 4448 }, { "epoch": 27.80625, "grad_norm": 1.21875, "learning_rate": 4.8925007810286546e-05, "loss": 0.1324, "step": 4449 }, { "epoch": 27.8125, "grad_norm": 0.9375, "learning_rate": 4.8924258229545936e-05, "loss": 0.0854, "step": 4450 }, { "epoch": 27.81875, "grad_norm": 0.89453125, "learning_rate": 4.892350839330522e-05, "loss": 0.0847, "step": 4451 }, { "epoch": 27.825, "grad_norm": 1.6953125, "learning_rate": 4.892275830157244e-05, "loss": 0.1761, "step": 4452 }, { "epoch": 27.83125, "grad_norm": 1.4140625, "learning_rate": 4.892200795435556e-05, "loss": 0.1269, "step": 4453 }, { "epoch": 27.8375, "grad_norm": 1.21875, "learning_rate": 4.892125735166264e-05, "loss": 0.1316, "step": 4454 }, { "epoch": 27.84375, "grad_norm": 1.3515625, "learning_rate": 4.892050649350167e-05, "loss": 0.1062, "step": 4455 }, { "epoch": 27.85, "grad_norm": 1.5703125, "learning_rate": 4.8919755379880686e-05, "loss": 0.1708, "step": 4456 }, { "epoch": 27.85625, "grad_norm": 0.9296875, "learning_rate": 4.891900401080769e-05, "loss": 0.0769, "step": 4457 }, { "epoch": 27.8625, "grad_norm": 1.4765625, "learning_rate": 4.891825238629072e-05, "loss": 0.1735, "step": 4458 }, { "epoch": 27.86875, "grad_norm": 1.2734375, "learning_rate": 4.89175005063378e-05, "loss": 0.1022, "step": 4459 }, { "epoch": 27.875, "grad_norm": 1.875, "learning_rate": 4.891674837095695e-05, "loss": 0.2183, "step": 4460 }, { "epoch": 27.88125, "grad_norm": 1.046875, "learning_rate": 4.891599598015621e-05, "loss": 0.077, "step": 4461 }, { "epoch": 27.8875, "grad_norm": 1.3984375, "learning_rate": 4.891524333394363e-05, "loss": 0.1138, "step": 4462 }, { "epoch": 27.89375, "grad_norm": 1.28125, "learning_rate": 4.891449043232722e-05, "loss": 0.0995, "step": 4463 }, { "epoch": 27.9, "grad_norm": 1.25, "learning_rate": 4.891373727531504e-05, "loss": 0.1014, "step": 4464 }, { "epoch": 27.90625, "grad_norm": 1.921875, "learning_rate": 4.891298386291513e-05, "loss": 0.1831, "step": 4465 }, { "epoch": 27.9125, "grad_norm": 1.5234375, "learning_rate": 4.891223019513553e-05, "loss": 0.1461, "step": 4466 }, { "epoch": 27.91875, "grad_norm": 1.2890625, "learning_rate": 4.8911476271984294e-05, "loss": 0.0976, "step": 4467 }, { "epoch": 27.925, "grad_norm": 1.515625, "learning_rate": 4.8910722093469475e-05, "loss": 0.1515, "step": 4468 }, { "epoch": 27.93125, "grad_norm": 1.8203125, "learning_rate": 4.8909967659599126e-05, "loss": 0.2116, "step": 4469 }, { "epoch": 27.9375, "grad_norm": 1.109375, "learning_rate": 4.89092129703813e-05, "loss": 0.0745, "step": 4470 }, { "epoch": 27.94375, "grad_norm": 1.75, "learning_rate": 4.8908458025824055e-05, "loss": 0.1035, "step": 4471 }, { "epoch": 27.95, "grad_norm": 2.484375, "learning_rate": 4.890770282593546e-05, "loss": 0.1558, "step": 4472 }, { "epoch": 27.95625, "grad_norm": 1.140625, "learning_rate": 4.8906947370723586e-05, "loss": 0.1111, "step": 4473 }, { "epoch": 27.9625, "grad_norm": 0.875, "learning_rate": 4.8906191660196485e-05, "loss": 0.0711, "step": 4474 }, { "epoch": 27.96875, "grad_norm": 1.109375, "learning_rate": 4.890543569436224e-05, "loss": 0.111, "step": 4475 }, { "epoch": 27.975, "grad_norm": 0.7421875, "learning_rate": 4.890467947322892e-05, "loss": 0.064, "step": 4476 }, { "epoch": 27.98125, "grad_norm": 1.3359375, "learning_rate": 4.89039229968046e-05, "loss": 0.0928, "step": 4477 }, { "epoch": 27.9875, "grad_norm": 1.1328125, "learning_rate": 4.8903166265097355e-05, "loss": 0.0848, "step": 4478 }, { "epoch": 27.99375, "grad_norm": 1.21875, "learning_rate": 4.890240927811528e-05, "loss": 0.1376, "step": 4479 }, { "epoch": 28.0, "grad_norm": 0.90234375, "learning_rate": 4.890165203586644e-05, "loss": 0.085, "step": 4480 }, { "epoch": 28.00625, "grad_norm": 1.8828125, "learning_rate": 4.8900894538358944e-05, "loss": 0.129, "step": 4481 }, { "epoch": 28.0125, "grad_norm": 1.125, "learning_rate": 4.890013678560086e-05, "loss": 0.0979, "step": 4482 }, { "epoch": 28.01875, "grad_norm": 0.78125, "learning_rate": 4.88993787776003e-05, "loss": 0.0715, "step": 4483 }, { "epoch": 28.025, "grad_norm": 0.9921875, "learning_rate": 4.8898620514365346e-05, "loss": 0.0719, "step": 4484 }, { "epoch": 28.03125, "grad_norm": 0.74609375, "learning_rate": 4.88978619959041e-05, "loss": 0.07, "step": 4485 }, { "epoch": 28.0375, "grad_norm": 0.93359375, "learning_rate": 4.889710322222467e-05, "loss": 0.0825, "step": 4486 }, { "epoch": 28.04375, "grad_norm": 1.2734375, "learning_rate": 4.889634419333514e-05, "loss": 0.1862, "step": 4487 }, { "epoch": 28.05, "grad_norm": 1.1875, "learning_rate": 4.889558490924364e-05, "loss": 0.1419, "step": 4488 }, { "epoch": 28.05625, "grad_norm": 1.0234375, "learning_rate": 4.8894825369958255e-05, "loss": 0.0917, "step": 4489 }, { "epoch": 28.0625, "grad_norm": 1.2578125, "learning_rate": 4.8894065575487114e-05, "loss": 0.129, "step": 4490 }, { "epoch": 28.06875, "grad_norm": 1.171875, "learning_rate": 4.889330552583833e-05, "loss": 0.1695, "step": 4491 }, { "epoch": 28.075, "grad_norm": 1.484375, "learning_rate": 4.889254522102e-05, "loss": 0.1486, "step": 4492 }, { "epoch": 28.08125, "grad_norm": 1.1015625, "learning_rate": 4.8891784661040275e-05, "loss": 0.1434, "step": 4493 }, { "epoch": 28.0875, "grad_norm": 1.1171875, "learning_rate": 4.889102384590726e-05, "loss": 0.1066, "step": 4494 }, { "epoch": 28.09375, "grad_norm": 1.6484375, "learning_rate": 4.8890262775629084e-05, "loss": 0.2635, "step": 4495 }, { "epoch": 28.1, "grad_norm": 1.4140625, "learning_rate": 4.888950145021386e-05, "loss": 0.1634, "step": 4496 }, { "epoch": 28.10625, "grad_norm": 1.6796875, "learning_rate": 4.888873986966974e-05, "loss": 0.1874, "step": 4497 }, { "epoch": 28.1125, "grad_norm": 1.1953125, "learning_rate": 4.888797803400485e-05, "loss": 0.1194, "step": 4498 }, { "epoch": 28.11875, "grad_norm": 0.92578125, "learning_rate": 4.8887215943227324e-05, "loss": 0.0938, "step": 4499 }, { "epoch": 28.125, "grad_norm": 0.8828125, "learning_rate": 4.88864535973453e-05, "loss": 0.0841, "step": 4500 }, { "epoch": 28.13125, "grad_norm": 0.69140625, "learning_rate": 4.888569099636692e-05, "loss": 0.0666, "step": 4501 }, { "epoch": 28.1375, "grad_norm": 1.2890625, "learning_rate": 4.8884928140300335e-05, "loss": 0.1129, "step": 4502 }, { "epoch": 28.14375, "grad_norm": 1.5546875, "learning_rate": 4.888416502915368e-05, "loss": 0.1415, "step": 4503 }, { "epoch": 28.15, "grad_norm": 1.2265625, "learning_rate": 4.8883401662935116e-05, "loss": 0.1292, "step": 4504 }, { "epoch": 28.15625, "grad_norm": 0.65625, "learning_rate": 4.888263804165279e-05, "loss": 0.0624, "step": 4505 }, { "epoch": 28.1625, "grad_norm": 1.1015625, "learning_rate": 4.888187416531485e-05, "loss": 0.0965, "step": 4506 }, { "epoch": 28.16875, "grad_norm": 1.3125, "learning_rate": 4.8881110033929466e-05, "loss": 0.1103, "step": 4507 }, { "epoch": 28.175, "grad_norm": 1.421875, "learning_rate": 4.888034564750479e-05, "loss": 0.124, "step": 4508 }, { "epoch": 28.18125, "grad_norm": 0.859375, "learning_rate": 4.8879581006049e-05, "loss": 0.0845, "step": 4509 }, { "epoch": 28.1875, "grad_norm": 1.21875, "learning_rate": 4.8878816109570245e-05, "loss": 0.1904, "step": 4510 }, { "epoch": 28.19375, "grad_norm": 1.4453125, "learning_rate": 4.88780509580767e-05, "loss": 0.2069, "step": 4511 }, { "epoch": 28.2, "grad_norm": 1.3984375, "learning_rate": 4.887728555157654e-05, "loss": 0.1468, "step": 4512 }, { "epoch": 28.20625, "grad_norm": 1.2890625, "learning_rate": 4.8876519890077926e-05, "loss": 0.1096, "step": 4513 }, { "epoch": 28.2125, "grad_norm": 1.6015625, "learning_rate": 4.887575397358905e-05, "loss": 0.2063, "step": 4514 }, { "epoch": 28.21875, "grad_norm": 0.84375, "learning_rate": 4.887498780211809e-05, "loss": 0.0744, "step": 4515 }, { "epoch": 28.225, "grad_norm": 1.15625, "learning_rate": 4.887422137567322e-05, "loss": 0.1367, "step": 4516 }, { "epoch": 28.23125, "grad_norm": 1.8125, "learning_rate": 4.887345469426263e-05, "loss": 0.168, "step": 4517 }, { "epoch": 28.2375, "grad_norm": 1.59375, "learning_rate": 4.887268775789451e-05, "loss": 0.1688, "step": 4518 }, { "epoch": 28.24375, "grad_norm": 1.4453125, "learning_rate": 4.887192056657704e-05, "loss": 0.1604, "step": 4519 }, { "epoch": 28.25, "grad_norm": 1.203125, "learning_rate": 4.8871153120318434e-05, "loss": 0.1281, "step": 4520 }, { "epoch": 28.25625, "grad_norm": 1.765625, "learning_rate": 4.887038541912686e-05, "loss": 0.1704, "step": 4521 }, { "epoch": 28.2625, "grad_norm": 1.0078125, "learning_rate": 4.886961746301054e-05, "loss": 0.0863, "step": 4522 }, { "epoch": 28.26875, "grad_norm": 1.0546875, "learning_rate": 4.8868849251977675e-05, "loss": 0.0902, "step": 4523 }, { "epoch": 28.275, "grad_norm": 1.6484375, "learning_rate": 4.886808078603645e-05, "loss": 0.1362, "step": 4524 }, { "epoch": 28.28125, "grad_norm": 1.109375, "learning_rate": 4.8867312065195094e-05, "loss": 0.0889, "step": 4525 }, { "epoch": 28.2875, "grad_norm": 1.3828125, "learning_rate": 4.886654308946179e-05, "loss": 0.1649, "step": 4526 }, { "epoch": 28.29375, "grad_norm": 1.3125, "learning_rate": 4.8865773858844776e-05, "loss": 0.1229, "step": 4527 }, { "epoch": 28.3, "grad_norm": 1.3671875, "learning_rate": 4.886500437335226e-05, "loss": 0.13, "step": 4528 }, { "epoch": 28.30625, "grad_norm": 1.0859375, "learning_rate": 4.8864234632992457e-05, "loss": 0.1088, "step": 4529 }, { "epoch": 28.3125, "grad_norm": 1.28125, "learning_rate": 4.886346463777358e-05, "loss": 0.1169, "step": 4530 }, { "epoch": 28.31875, "grad_norm": 0.9921875, "learning_rate": 4.8862694387703866e-05, "loss": 0.0888, "step": 4531 }, { "epoch": 28.325, "grad_norm": 1.1875, "learning_rate": 4.8861923882791524e-05, "loss": 0.1051, "step": 4532 }, { "epoch": 28.33125, "grad_norm": 0.890625, "learning_rate": 4.886115312304481e-05, "loss": 0.0761, "step": 4533 }, { "epoch": 28.3375, "grad_norm": 1.1875, "learning_rate": 4.886038210847192e-05, "loss": 0.0825, "step": 4534 }, { "epoch": 28.34375, "grad_norm": 1.6171875, "learning_rate": 4.885961083908113e-05, "loss": 0.101, "step": 4535 }, { "epoch": 28.35, "grad_norm": 1.265625, "learning_rate": 4.8858839314880636e-05, "loss": 0.1154, "step": 4536 }, { "epoch": 28.35625, "grad_norm": 1.2890625, "learning_rate": 4.8858067535878696e-05, "loss": 0.1058, "step": 4537 }, { "epoch": 28.3625, "grad_norm": 1.0, "learning_rate": 4.8857295502083554e-05, "loss": 0.09, "step": 4538 }, { "epoch": 28.36875, "grad_norm": 1.5546875, "learning_rate": 4.8856523213503455e-05, "loss": 0.1371, "step": 4539 }, { "epoch": 28.375, "grad_norm": 1.3203125, "learning_rate": 4.885575067014664e-05, "loss": 0.1053, "step": 4540 }, { "epoch": 28.38125, "grad_norm": 1.4921875, "learning_rate": 4.8854977872021364e-05, "loss": 0.2346, "step": 4541 }, { "epoch": 28.3875, "grad_norm": 0.8046875, "learning_rate": 4.885420481913588e-05, "loss": 0.0669, "step": 4542 }, { "epoch": 28.39375, "grad_norm": 1.390625, "learning_rate": 4.8853431511498443e-05, "loss": 0.111, "step": 4543 }, { "epoch": 28.4, "grad_norm": 0.89453125, "learning_rate": 4.885265794911731e-05, "loss": 0.075, "step": 4544 }, { "epoch": 28.40625, "grad_norm": 1.9140625, "learning_rate": 4.885188413200075e-05, "loss": 0.165, "step": 4545 }, { "epoch": 28.4125, "grad_norm": 1.2734375, "learning_rate": 4.885111006015701e-05, "loss": 0.1227, "step": 4546 }, { "epoch": 28.41875, "grad_norm": 1.4453125, "learning_rate": 4.8850335733594376e-05, "loss": 0.1743, "step": 4547 }, { "epoch": 28.425, "grad_norm": 2.1875, "learning_rate": 4.884956115232111e-05, "loss": 0.2139, "step": 4548 }, { "epoch": 28.43125, "grad_norm": 1.3203125, "learning_rate": 4.884878631634549e-05, "loss": 0.1118, "step": 4549 }, { "epoch": 28.4375, "grad_norm": 1.328125, "learning_rate": 4.8848011225675775e-05, "loss": 0.1395, "step": 4550 }, { "epoch": 28.44375, "grad_norm": 1.703125, "learning_rate": 4.884723588032025e-05, "loss": 0.1798, "step": 4551 }, { "epoch": 28.45, "grad_norm": 1.2109375, "learning_rate": 4.88464602802872e-05, "loss": 0.0949, "step": 4552 }, { "epoch": 28.45625, "grad_norm": 0.9375, "learning_rate": 4.88456844255849e-05, "loss": 0.0814, "step": 4553 }, { "epoch": 28.4625, "grad_norm": 1.71875, "learning_rate": 4.884490831622165e-05, "loss": 0.1186, "step": 4554 }, { "epoch": 28.46875, "grad_norm": 1.3046875, "learning_rate": 4.884413195220573e-05, "loss": 0.1437, "step": 4555 }, { "epoch": 28.475, "grad_norm": 1.0546875, "learning_rate": 4.8843355333545424e-05, "loss": 0.0831, "step": 4556 }, { "epoch": 28.48125, "grad_norm": 1.8984375, "learning_rate": 4.884257846024903e-05, "loss": 0.1888, "step": 4557 }, { "epoch": 28.4875, "grad_norm": 0.9296875, "learning_rate": 4.884180133232486e-05, "loss": 0.0842, "step": 4558 }, { "epoch": 28.49375, "grad_norm": 1.5, "learning_rate": 4.884102394978119e-05, "loss": 0.1488, "step": 4559 }, { "epoch": 28.5, "grad_norm": 1.828125, "learning_rate": 4.884024631262634e-05, "loss": 0.1627, "step": 4560 }, { "epoch": 28.50625, "grad_norm": 1.5703125, "learning_rate": 4.8839468420868606e-05, "loss": 0.2027, "step": 4561 }, { "epoch": 28.5125, "grad_norm": 1.1953125, "learning_rate": 4.8838690274516293e-05, "loss": 0.1085, "step": 4562 }, { "epoch": 28.51875, "grad_norm": 1.03125, "learning_rate": 4.8837911873577725e-05, "loss": 0.0956, "step": 4563 }, { "epoch": 28.525, "grad_norm": 0.890625, "learning_rate": 4.8837133218061196e-05, "loss": 0.0786, "step": 4564 }, { "epoch": 28.53125, "grad_norm": 1.2109375, "learning_rate": 4.8836354307975026e-05, "loss": 0.1255, "step": 4565 }, { "epoch": 28.5375, "grad_norm": 1.171875, "learning_rate": 4.883557514332755e-05, "loss": 0.1169, "step": 4566 }, { "epoch": 28.54375, "grad_norm": 0.96484375, "learning_rate": 4.883479572412708e-05, "loss": 0.0884, "step": 4567 }, { "epoch": 28.55, "grad_norm": 1.4609375, "learning_rate": 4.883401605038193e-05, "loss": 0.118, "step": 4568 }, { "epoch": 28.55625, "grad_norm": 1.328125, "learning_rate": 4.883323612210043e-05, "loss": 0.1008, "step": 4569 }, { "epoch": 28.5625, "grad_norm": 1.03125, "learning_rate": 4.883245593929092e-05, "loss": 0.0854, "step": 4570 }, { "epoch": 28.56875, "grad_norm": 1.4453125, "learning_rate": 4.883167550196173e-05, "loss": 0.162, "step": 4571 }, { "epoch": 28.575, "grad_norm": 1.5546875, "learning_rate": 4.883089481012119e-05, "loss": 0.1721, "step": 4572 }, { "epoch": 28.58125, "grad_norm": 2.0, "learning_rate": 4.8830113863777624e-05, "loss": 0.2072, "step": 4573 }, { "epoch": 28.5875, "grad_norm": 1.6015625, "learning_rate": 4.88293326629394e-05, "loss": 0.151, "step": 4574 }, { "epoch": 28.59375, "grad_norm": 1.3359375, "learning_rate": 4.882855120761484e-05, "loss": 0.114, "step": 4575 }, { "epoch": 28.6, "grad_norm": 0.8359375, "learning_rate": 4.882776949781229e-05, "loss": 0.0723, "step": 4576 }, { "epoch": 28.60625, "grad_norm": 1.09375, "learning_rate": 4.8826987533540114e-05, "loss": 0.0906, "step": 4577 }, { "epoch": 28.6125, "grad_norm": 1.796875, "learning_rate": 4.882620531480664e-05, "loss": 0.1684, "step": 4578 }, { "epoch": 28.61875, "grad_norm": 1.1171875, "learning_rate": 4.882542284162025e-05, "loss": 0.0996, "step": 4579 }, { "epoch": 28.625, "grad_norm": 0.91015625, "learning_rate": 4.8824640113989286e-05, "loss": 0.0705, "step": 4580 }, { "epoch": 28.63125, "grad_norm": 1.1484375, "learning_rate": 4.8823857131922096e-05, "loss": 0.1026, "step": 4581 }, { "epoch": 28.6375, "grad_norm": 1.2890625, "learning_rate": 4.882307389542706e-05, "loss": 0.1149, "step": 4582 }, { "epoch": 28.64375, "grad_norm": 1.34375, "learning_rate": 4.8822290404512525e-05, "loss": 0.1206, "step": 4583 }, { "epoch": 28.65, "grad_norm": 0.85546875, "learning_rate": 4.882150665918688e-05, "loss": 0.0648, "step": 4584 }, { "epoch": 28.65625, "grad_norm": 1.421875, "learning_rate": 4.882072265945848e-05, "loss": 0.1254, "step": 4585 }, { "epoch": 28.6625, "grad_norm": 1.2421875, "learning_rate": 4.8819938405335705e-05, "loss": 0.1473, "step": 4586 }, { "epoch": 28.66875, "grad_norm": 1.359375, "learning_rate": 4.881915389682692e-05, "loss": 0.1032, "step": 4587 }, { "epoch": 28.675, "grad_norm": 1.234375, "learning_rate": 4.881836913394051e-05, "loss": 0.103, "step": 4588 }, { "epoch": 28.68125, "grad_norm": 1.453125, "learning_rate": 4.881758411668486e-05, "loss": 0.1368, "step": 4589 }, { "epoch": 28.6875, "grad_norm": 0.859375, "learning_rate": 4.8816798845068344e-05, "loss": 0.0636, "step": 4590 }, { "epoch": 28.69375, "grad_norm": 1.265625, "learning_rate": 4.8816013319099365e-05, "loss": 0.116, "step": 4591 }, { "epoch": 28.7, "grad_norm": 0.9765625, "learning_rate": 4.881522753878629e-05, "loss": 0.0749, "step": 4592 }, { "epoch": 28.70625, "grad_norm": 0.8515625, "learning_rate": 4.8814441504137534e-05, "loss": 0.0734, "step": 4593 }, { "epoch": 28.7125, "grad_norm": 0.98828125, "learning_rate": 4.8813655215161466e-05, "loss": 0.0806, "step": 4594 }, { "epoch": 28.71875, "grad_norm": 1.0390625, "learning_rate": 4.88128686718665e-05, "loss": 0.0938, "step": 4595 }, { "epoch": 28.725, "grad_norm": 1.328125, "learning_rate": 4.881208187426103e-05, "loss": 0.123, "step": 4596 }, { "epoch": 28.73125, "grad_norm": 1.2734375, "learning_rate": 4.881129482235347e-05, "loss": 0.1325, "step": 4597 }, { "epoch": 28.7375, "grad_norm": 1.5390625, "learning_rate": 4.881050751615221e-05, "loss": 0.1525, "step": 4598 }, { "epoch": 28.74375, "grad_norm": 1.7109375, "learning_rate": 4.880971995566567e-05, "loss": 0.1423, "step": 4599 }, { "epoch": 28.75, "grad_norm": 1.109375, "learning_rate": 4.8808932140902244e-05, "loss": 0.1064, "step": 4600 }, { "epoch": 28.75625, "grad_norm": 1.25, "learning_rate": 4.8808144071870364e-05, "loss": 0.1113, "step": 4601 }, { "epoch": 28.7625, "grad_norm": 1.1875, "learning_rate": 4.8807355748578445e-05, "loss": 0.1116, "step": 4602 }, { "epoch": 28.76875, "grad_norm": 1.296875, "learning_rate": 4.880656717103489e-05, "loss": 0.1805, "step": 4603 }, { "epoch": 28.775, "grad_norm": 1.25, "learning_rate": 4.880577833924813e-05, "loss": 0.1115, "step": 4604 }, { "epoch": 28.78125, "grad_norm": 1.4375, "learning_rate": 4.88049892532266e-05, "loss": 0.1755, "step": 4605 }, { "epoch": 28.7875, "grad_norm": 1.5859375, "learning_rate": 4.8804199912978704e-05, "loss": 0.2303, "step": 4606 }, { "epoch": 28.79375, "grad_norm": 0.99609375, "learning_rate": 4.880341031851289e-05, "loss": 0.0816, "step": 4607 }, { "epoch": 28.8, "grad_norm": 1.109375, "learning_rate": 4.880262046983759e-05, "loss": 0.1179, "step": 4608 }, { "epoch": 28.80625, "grad_norm": 1.375, "learning_rate": 4.880183036696123e-05, "loss": 0.1244, "step": 4609 }, { "epoch": 28.8125, "grad_norm": 1.5859375, "learning_rate": 4.880104000989225e-05, "loss": 0.1334, "step": 4610 }, { "epoch": 28.81875, "grad_norm": 1.5234375, "learning_rate": 4.88002493986391e-05, "loss": 0.1398, "step": 4611 }, { "epoch": 28.825, "grad_norm": 1.2890625, "learning_rate": 4.8799458533210206e-05, "loss": 0.1131, "step": 4612 }, { "epoch": 28.83125, "grad_norm": 1.1953125, "learning_rate": 4.879866741361403e-05, "loss": 0.1162, "step": 4613 }, { "epoch": 28.8375, "grad_norm": 1.515625, "learning_rate": 4.8797876039859016e-05, "loss": 0.1138, "step": 4614 }, { "epoch": 28.84375, "grad_norm": 1.5390625, "learning_rate": 4.879708441195362e-05, "loss": 0.1602, "step": 4615 }, { "epoch": 28.85, "grad_norm": 1.0859375, "learning_rate": 4.879629252990628e-05, "loss": 0.1247, "step": 4616 }, { "epoch": 28.85625, "grad_norm": 0.84765625, "learning_rate": 4.879550039372547e-05, "loss": 0.0787, "step": 4617 }, { "epoch": 28.8625, "grad_norm": 1.1953125, "learning_rate": 4.8794708003419646e-05, "loss": 0.0912, "step": 4618 }, { "epoch": 28.86875, "grad_norm": 1.5859375, "learning_rate": 4.8793915358997264e-05, "loss": 0.2167, "step": 4619 }, { "epoch": 28.875, "grad_norm": 1.015625, "learning_rate": 4.879312246046679e-05, "loss": 0.0736, "step": 4620 }, { "epoch": 28.88125, "grad_norm": 1.171875, "learning_rate": 4.8792329307836706e-05, "loss": 0.0891, "step": 4621 }, { "epoch": 28.8875, "grad_norm": 1.140625, "learning_rate": 4.8791535901115457e-05, "loss": 0.0882, "step": 4622 }, { "epoch": 28.89375, "grad_norm": 1.015625, "learning_rate": 4.8790742240311536e-05, "loss": 0.0912, "step": 4623 }, { "epoch": 28.9, "grad_norm": 1.03125, "learning_rate": 4.878994832543342e-05, "loss": 0.1171, "step": 4624 }, { "epoch": 28.90625, "grad_norm": 1.4765625, "learning_rate": 4.878915415648957e-05, "loss": 0.0961, "step": 4625 }, { "epoch": 28.9125, "grad_norm": 1.296875, "learning_rate": 4.878835973348849e-05, "loss": 0.1558, "step": 4626 }, { "epoch": 28.91875, "grad_norm": 1.2265625, "learning_rate": 4.878756505643864e-05, "loss": 0.1163, "step": 4627 }, { "epoch": 28.925, "grad_norm": 0.83203125, "learning_rate": 4.878677012534853e-05, "loss": 0.0709, "step": 4628 }, { "epoch": 28.93125, "grad_norm": 1.0, "learning_rate": 4.8785974940226635e-05, "loss": 0.0848, "step": 4629 }, { "epoch": 28.9375, "grad_norm": 0.99609375, "learning_rate": 4.878517950108145e-05, "loss": 0.0887, "step": 4630 }, { "epoch": 28.94375, "grad_norm": 1.1953125, "learning_rate": 4.878438380792147e-05, "loss": 0.0958, "step": 4631 }, { "epoch": 28.95, "grad_norm": 0.87109375, "learning_rate": 4.8783587860755195e-05, "loss": 0.0809, "step": 4632 }, { "epoch": 28.95625, "grad_norm": 1.25, "learning_rate": 4.8782791659591134e-05, "loss": 0.1348, "step": 4633 }, { "epoch": 28.9625, "grad_norm": 1.4140625, "learning_rate": 4.8781995204437766e-05, "loss": 0.1592, "step": 4634 }, { "epoch": 28.96875, "grad_norm": 1.2421875, "learning_rate": 4.8781198495303615e-05, "loss": 0.1097, "step": 4635 }, { "epoch": 28.975, "grad_norm": 1.0546875, "learning_rate": 4.8780401532197185e-05, "loss": 0.0895, "step": 4636 }, { "epoch": 28.98125, "grad_norm": 1.453125, "learning_rate": 4.877960431512699e-05, "loss": 0.1222, "step": 4637 }, { "epoch": 28.9875, "grad_norm": 1.25, "learning_rate": 4.877880684410153e-05, "loss": 0.1081, "step": 4638 }, { "epoch": 28.99375, "grad_norm": 1.7109375, "learning_rate": 4.877800911912935e-05, "loss": 0.2042, "step": 4639 }, { "epoch": 29.0, "grad_norm": 1.15625, "learning_rate": 4.877721114021894e-05, "loss": 0.1057, "step": 4640 }, { "epoch": 29.00625, "grad_norm": 1.25, "learning_rate": 4.877641290737884e-05, "loss": 0.1499, "step": 4641 }, { "epoch": 29.0125, "grad_norm": 0.8984375, "learning_rate": 4.8775614420617574e-05, "loss": 0.0643, "step": 4642 }, { "epoch": 29.01875, "grad_norm": 0.96484375, "learning_rate": 4.877481567994365e-05, "loss": 0.0765, "step": 4643 }, { "epoch": 29.025, "grad_norm": 1.015625, "learning_rate": 4.8774016685365625e-05, "loss": 0.1001, "step": 4644 }, { "epoch": 29.03125, "grad_norm": 1.8671875, "learning_rate": 4.8773217436892014e-05, "loss": 0.2176, "step": 4645 }, { "epoch": 29.0375, "grad_norm": 1.328125, "learning_rate": 4.877241793453137e-05, "loss": 0.1029, "step": 4646 }, { "epoch": 29.04375, "grad_norm": 1.5234375, "learning_rate": 4.87716181782922e-05, "loss": 0.1374, "step": 4647 }, { "epoch": 29.05, "grad_norm": 1.2578125, "learning_rate": 4.8770818168183074e-05, "loss": 0.1183, "step": 4648 }, { "epoch": 29.05625, "grad_norm": 1.3046875, "learning_rate": 4.8770017904212525e-05, "loss": 0.1018, "step": 4649 }, { "epoch": 29.0625, "grad_norm": 1.046875, "learning_rate": 4.876921738638911e-05, "loss": 0.1454, "step": 4650 }, { "epoch": 29.06875, "grad_norm": 1.2265625, "learning_rate": 4.876841661472136e-05, "loss": 0.1216, "step": 4651 }, { "epoch": 29.075, "grad_norm": 1.6484375, "learning_rate": 4.876761558921784e-05, "loss": 0.1413, "step": 4652 }, { "epoch": 29.08125, "grad_norm": 0.921875, "learning_rate": 4.87668143098871e-05, "loss": 0.0769, "step": 4653 }, { "epoch": 29.0875, "grad_norm": 1.3515625, "learning_rate": 4.876601277673769e-05, "loss": 0.1147, "step": 4654 }, { "epoch": 29.09375, "grad_norm": 1.4765625, "learning_rate": 4.8765210989778185e-05, "loss": 0.1599, "step": 4655 }, { "epoch": 29.1, "grad_norm": 1.078125, "learning_rate": 4.876440894901714e-05, "loss": 0.0892, "step": 4656 }, { "epoch": 29.10625, "grad_norm": 1.78125, "learning_rate": 4.876360665446312e-05, "loss": 0.2762, "step": 4657 }, { "epoch": 29.1125, "grad_norm": 1.140625, "learning_rate": 4.8762804106124695e-05, "loss": 0.0984, "step": 4658 }, { "epoch": 29.11875, "grad_norm": 0.76171875, "learning_rate": 4.876200130401043e-05, "loss": 0.064, "step": 4659 }, { "epoch": 29.125, "grad_norm": 1.6640625, "learning_rate": 4.8761198248128915e-05, "loss": 0.2169, "step": 4660 }, { "epoch": 29.13125, "grad_norm": 1.234375, "learning_rate": 4.8760394938488706e-05, "loss": 0.1804, "step": 4661 }, { "epoch": 29.1375, "grad_norm": 1.2109375, "learning_rate": 4.875959137509839e-05, "loss": 0.1441, "step": 4662 }, { "epoch": 29.14375, "grad_norm": 0.8515625, "learning_rate": 4.875878755796655e-05, "loss": 0.0817, "step": 4663 }, { "epoch": 29.15, "grad_norm": 1.1484375, "learning_rate": 4.875798348710177e-05, "loss": 0.0944, "step": 4664 }, { "epoch": 29.15625, "grad_norm": 1.5, "learning_rate": 4.875717916251264e-05, "loss": 0.1315, "step": 4665 }, { "epoch": 29.1625, "grad_norm": 1.1640625, "learning_rate": 4.875637458420774e-05, "loss": 0.1066, "step": 4666 }, { "epoch": 29.16875, "grad_norm": 1.84375, "learning_rate": 4.8755569752195676e-05, "loss": 0.1553, "step": 4667 }, { "epoch": 29.175, "grad_norm": 2.0625, "learning_rate": 4.875476466648504e-05, "loss": 0.2295, "step": 4668 }, { "epoch": 29.18125, "grad_norm": 1.3125, "learning_rate": 4.875395932708442e-05, "loss": 0.1375, "step": 4669 }, { "epoch": 29.1875, "grad_norm": 0.9296875, "learning_rate": 4.875315373400242e-05, "loss": 0.0646, "step": 4670 }, { "epoch": 29.19375, "grad_norm": 1.625, "learning_rate": 4.875234788724765e-05, "loss": 0.1476, "step": 4671 }, { "epoch": 29.2, "grad_norm": 1.0703125, "learning_rate": 4.875154178682871e-05, "loss": 0.1067, "step": 4672 }, { "epoch": 29.20625, "grad_norm": 1.34375, "learning_rate": 4.875073543275422e-05, "loss": 0.1234, "step": 4673 }, { "epoch": 29.2125, "grad_norm": 1.3125, "learning_rate": 4.874992882503277e-05, "loss": 0.1277, "step": 4674 }, { "epoch": 29.21875, "grad_norm": 1.25, "learning_rate": 4.8749121963673e-05, "loss": 0.0931, "step": 4675 }, { "epoch": 29.225, "grad_norm": 1.1875, "learning_rate": 4.874831484868351e-05, "loss": 0.1156, "step": 4676 }, { "epoch": 29.23125, "grad_norm": 1.625, "learning_rate": 4.874750748007292e-05, "loss": 0.1018, "step": 4677 }, { "epoch": 29.2375, "grad_norm": 1.28125, "learning_rate": 4.874669985784985e-05, "loss": 0.1366, "step": 4678 }, { "epoch": 29.24375, "grad_norm": 1.4453125, "learning_rate": 4.874589198202294e-05, "loss": 0.1022, "step": 4679 }, { "epoch": 29.25, "grad_norm": 1.125, "learning_rate": 4.874508385260081e-05, "loss": 0.1044, "step": 4680 }, { "epoch": 29.25625, "grad_norm": 1.125, "learning_rate": 4.874427546959208e-05, "loss": 0.0986, "step": 4681 }, { "epoch": 29.2625, "grad_norm": 1.2265625, "learning_rate": 4.87434668330054e-05, "loss": 0.1056, "step": 4682 }, { "epoch": 29.26875, "grad_norm": 2.0, "learning_rate": 4.87426579428494e-05, "loss": 0.2585, "step": 4683 }, { "epoch": 29.275, "grad_norm": 1.3515625, "learning_rate": 4.874184879913271e-05, "loss": 0.0965, "step": 4684 }, { "epoch": 29.28125, "grad_norm": 1.1328125, "learning_rate": 4.874103940186398e-05, "loss": 0.081, "step": 4685 }, { "epoch": 29.2875, "grad_norm": 1.28125, "learning_rate": 4.874022975105186e-05, "loss": 0.1386, "step": 4686 }, { "epoch": 29.29375, "grad_norm": 0.8515625, "learning_rate": 4.873941984670498e-05, "loss": 0.0906, "step": 4687 }, { "epoch": 29.3, "grad_norm": 1.484375, "learning_rate": 4.873860968883199e-05, "loss": 0.1939, "step": 4688 }, { "epoch": 29.30625, "grad_norm": 0.91796875, "learning_rate": 4.8737799277441566e-05, "loss": 0.0735, "step": 4689 }, { "epoch": 29.3125, "grad_norm": 1.0703125, "learning_rate": 4.8736988612542337e-05, "loss": 0.0781, "step": 4690 }, { "epoch": 29.31875, "grad_norm": 1.078125, "learning_rate": 4.873617769414297e-05, "loss": 0.1025, "step": 4691 }, { "epoch": 29.325, "grad_norm": 1.5546875, "learning_rate": 4.873536652225214e-05, "loss": 0.1588, "step": 4692 }, { "epoch": 29.33125, "grad_norm": 1.140625, "learning_rate": 4.8734555096878475e-05, "loss": 0.0987, "step": 4693 }, { "epoch": 29.3375, "grad_norm": 1.2265625, "learning_rate": 4.8733743418030677e-05, "loss": 0.1209, "step": 4694 }, { "epoch": 29.34375, "grad_norm": 1.1953125, "learning_rate": 4.873293148571739e-05, "loss": 0.1255, "step": 4695 }, { "epoch": 29.35, "grad_norm": 1.6796875, "learning_rate": 4.87321192999473e-05, "loss": 0.159, "step": 4696 }, { "epoch": 29.35625, "grad_norm": 1.5234375, "learning_rate": 4.873130686072907e-05, "loss": 0.1372, "step": 4697 }, { "epoch": 29.3625, "grad_norm": 1.0390625, "learning_rate": 4.873049416807138e-05, "loss": 0.0792, "step": 4698 }, { "epoch": 29.36875, "grad_norm": 1.359375, "learning_rate": 4.872968122198292e-05, "loss": 0.1538, "step": 4699 }, { "epoch": 29.375, "grad_norm": 1.875, "learning_rate": 4.872886802247235e-05, "loss": 0.1666, "step": 4700 }, { "epoch": 29.38125, "grad_norm": 1.7734375, "learning_rate": 4.872805456954837e-05, "loss": 0.1991, "step": 4701 }, { "epoch": 29.3875, "grad_norm": 1.2578125, "learning_rate": 4.8727240863219665e-05, "loss": 0.108, "step": 4702 }, { "epoch": 29.39375, "grad_norm": 1.53125, "learning_rate": 4.872642690349492e-05, "loss": 0.106, "step": 4703 }, { "epoch": 29.4, "grad_norm": 0.953125, "learning_rate": 4.8725612690382836e-05, "loss": 0.0731, "step": 4704 }, { "epoch": 29.40625, "grad_norm": 1.0859375, "learning_rate": 4.872479822389211e-05, "loss": 0.0952, "step": 4705 }, { "epoch": 29.4125, "grad_norm": 0.859375, "learning_rate": 4.8723983504031424e-05, "loss": 0.0764, "step": 4706 }, { "epoch": 29.41875, "grad_norm": 1.40625, "learning_rate": 4.8723168530809494e-05, "loss": 0.1109, "step": 4707 }, { "epoch": 29.425, "grad_norm": 1.078125, "learning_rate": 4.872235330423502e-05, "loss": 0.0821, "step": 4708 }, { "epoch": 29.43125, "grad_norm": 1.7578125, "learning_rate": 4.87215378243167e-05, "loss": 0.1416, "step": 4709 }, { "epoch": 29.4375, "grad_norm": 1.15625, "learning_rate": 4.872072209106326e-05, "loss": 0.1522, "step": 4710 }, { "epoch": 29.44375, "grad_norm": 0.81640625, "learning_rate": 4.8719906104483395e-05, "loss": 0.0855, "step": 4711 }, { "epoch": 29.45, "grad_norm": 1.3125, "learning_rate": 4.871908986458583e-05, "loss": 0.0984, "step": 4712 }, { "epoch": 29.45625, "grad_norm": 1.6171875, "learning_rate": 4.8718273371379275e-05, "loss": 0.1562, "step": 4713 }, { "epoch": 29.4625, "grad_norm": 1.15625, "learning_rate": 4.871745662487245e-05, "loss": 0.0871, "step": 4714 }, { "epoch": 29.46875, "grad_norm": 0.95703125, "learning_rate": 4.871663962507408e-05, "loss": 0.1133, "step": 4715 }, { "epoch": 29.475, "grad_norm": 1.6953125, "learning_rate": 4.8715822371992895e-05, "loss": 0.134, "step": 4716 }, { "epoch": 29.48125, "grad_norm": 1.78125, "learning_rate": 4.8715004865637614e-05, "loss": 0.1134, "step": 4717 }, { "epoch": 29.4875, "grad_norm": 1.3046875, "learning_rate": 4.871418710601698e-05, "loss": 0.1097, "step": 4718 }, { "epoch": 29.49375, "grad_norm": 1.0625, "learning_rate": 4.871336909313971e-05, "loss": 0.0851, "step": 4719 }, { "epoch": 29.5, "grad_norm": 1.796875, "learning_rate": 4.871255082701455e-05, "loss": 0.1831, "step": 4720 }, { "epoch": 29.50625, "grad_norm": 1.015625, "learning_rate": 4.871173230765024e-05, "loss": 0.0953, "step": 4721 }, { "epoch": 29.5125, "grad_norm": 1.0078125, "learning_rate": 4.871091353505551e-05, "loss": 0.0951, "step": 4722 }, { "epoch": 29.51875, "grad_norm": 1.203125, "learning_rate": 4.8710094509239125e-05, "loss": 0.1433, "step": 4723 }, { "epoch": 29.525, "grad_norm": 0.8359375, "learning_rate": 4.87092752302098e-05, "loss": 0.062, "step": 4724 }, { "epoch": 29.53125, "grad_norm": 1.28125, "learning_rate": 4.8708455697976324e-05, "loss": 0.0816, "step": 4725 }, { "epoch": 29.5375, "grad_norm": 0.76953125, "learning_rate": 4.8707635912547426e-05, "loss": 0.0718, "step": 4726 }, { "epoch": 29.54375, "grad_norm": 1.453125, "learning_rate": 4.8706815873931854e-05, "loss": 0.1442, "step": 4727 }, { "epoch": 29.55, "grad_norm": 1.2578125, "learning_rate": 4.870599558213838e-05, "loss": 0.1103, "step": 4728 }, { "epoch": 29.55625, "grad_norm": 0.953125, "learning_rate": 4.870517503717576e-05, "loss": 0.0928, "step": 4729 }, { "epoch": 29.5625, "grad_norm": 0.99609375, "learning_rate": 4.870435423905276e-05, "loss": 0.0854, "step": 4730 }, { "epoch": 29.56875, "grad_norm": 1.3671875, "learning_rate": 4.870353318777814e-05, "loss": 0.1949, "step": 4731 }, { "epoch": 29.575, "grad_norm": 1.0703125, "learning_rate": 4.8702711883360666e-05, "loss": 0.0835, "step": 4732 }, { "epoch": 29.58125, "grad_norm": 1.5234375, "learning_rate": 4.870189032580912e-05, "loss": 0.2801, "step": 4733 }, { "epoch": 29.5875, "grad_norm": 0.92578125, "learning_rate": 4.870106851513226e-05, "loss": 0.0896, "step": 4734 }, { "epoch": 29.59375, "grad_norm": 1.1953125, "learning_rate": 4.870024645133889e-05, "loss": 0.1058, "step": 4735 }, { "epoch": 29.6, "grad_norm": 0.8828125, "learning_rate": 4.8699424134437765e-05, "loss": 0.0831, "step": 4736 }, { "epoch": 29.60625, "grad_norm": 1.46875, "learning_rate": 4.8698601564437675e-05, "loss": 0.1189, "step": 4737 }, { "epoch": 29.6125, "grad_norm": 1.5078125, "learning_rate": 4.86977787413474e-05, "loss": 0.134, "step": 4738 }, { "epoch": 29.61875, "grad_norm": 1.078125, "learning_rate": 4.869695566517573e-05, "loss": 0.098, "step": 4739 }, { "epoch": 29.625, "grad_norm": 0.953125, "learning_rate": 4.8696132335931455e-05, "loss": 0.0848, "step": 4740 }, { "epoch": 29.63125, "grad_norm": 1.15625, "learning_rate": 4.869530875362337e-05, "loss": 0.1034, "step": 4741 }, { "epoch": 29.6375, "grad_norm": 0.8671875, "learning_rate": 4.8694484918260275e-05, "loss": 0.0784, "step": 4742 }, { "epoch": 29.64375, "grad_norm": 0.796875, "learning_rate": 4.869366082985095e-05, "loss": 0.0733, "step": 4743 }, { "epoch": 29.65, "grad_norm": 1.5078125, "learning_rate": 4.8692836488404226e-05, "loss": 0.1887, "step": 4744 }, { "epoch": 29.65625, "grad_norm": 1.0625, "learning_rate": 4.8692011893928876e-05, "loss": 0.1159, "step": 4745 }, { "epoch": 29.6625, "grad_norm": 1.375, "learning_rate": 4.869118704643372e-05, "loss": 0.1293, "step": 4746 }, { "epoch": 29.66875, "grad_norm": 1.1328125, "learning_rate": 4.869036194592757e-05, "loss": 0.0831, "step": 4747 }, { "epoch": 29.675, "grad_norm": 0.796875, "learning_rate": 4.8689536592419235e-05, "loss": 0.0639, "step": 4748 }, { "epoch": 29.68125, "grad_norm": 1.3046875, "learning_rate": 4.8688710985917525e-05, "loss": 0.1062, "step": 4749 }, { "epoch": 29.6875, "grad_norm": 1.2421875, "learning_rate": 4.868788512643126e-05, "loss": 0.1113, "step": 4750 }, { "epoch": 29.69375, "grad_norm": 1.140625, "learning_rate": 4.868705901396926e-05, "loss": 0.1095, "step": 4751 }, { "epoch": 29.7, "grad_norm": 1.4140625, "learning_rate": 4.868623264854035e-05, "loss": 0.0997, "step": 4752 }, { "epoch": 29.70625, "grad_norm": 1.5078125, "learning_rate": 4.868540603015335e-05, "loss": 0.1515, "step": 4753 }, { "epoch": 29.7125, "grad_norm": 1.125, "learning_rate": 4.868457915881709e-05, "loss": 0.0972, "step": 4754 }, { "epoch": 29.71875, "grad_norm": 0.65234375, "learning_rate": 4.868375203454041e-05, "loss": 0.0582, "step": 4755 }, { "epoch": 29.725, "grad_norm": 1.5390625, "learning_rate": 4.868292465733213e-05, "loss": 0.1259, "step": 4756 }, { "epoch": 29.73125, "grad_norm": 1.109375, "learning_rate": 4.868209702720109e-05, "loss": 0.0851, "step": 4757 }, { "epoch": 29.7375, "grad_norm": 0.84765625, "learning_rate": 4.868126914415613e-05, "loss": 0.0636, "step": 4758 }, { "epoch": 29.74375, "grad_norm": 1.4140625, "learning_rate": 4.868044100820609e-05, "loss": 0.171, "step": 4759 }, { "epoch": 29.75, "grad_norm": 1.6640625, "learning_rate": 4.8679612619359815e-05, "loss": 0.1497, "step": 4760 }, { "epoch": 29.75625, "grad_norm": 1.203125, "learning_rate": 4.8678783977626155e-05, "loss": 0.1238, "step": 4761 }, { "epoch": 29.7625, "grad_norm": 1.6640625, "learning_rate": 4.8677955083013946e-05, "loss": 0.1937, "step": 4762 }, { "epoch": 29.76875, "grad_norm": 1.171875, "learning_rate": 4.867712593553206e-05, "loss": 0.118, "step": 4763 }, { "epoch": 29.775, "grad_norm": 1.3984375, "learning_rate": 4.8676296535189334e-05, "loss": 0.1187, "step": 4764 }, { "epoch": 29.78125, "grad_norm": 1.046875, "learning_rate": 4.8675466881994644e-05, "loss": 0.0783, "step": 4765 }, { "epoch": 29.7875, "grad_norm": 1.2109375, "learning_rate": 4.867463697595683e-05, "loss": 0.1109, "step": 4766 }, { "epoch": 29.79375, "grad_norm": 0.89453125, "learning_rate": 4.8673806817084774e-05, "loss": 0.0949, "step": 4767 }, { "epoch": 29.8, "grad_norm": 1.5, "learning_rate": 4.867297640538734e-05, "loss": 0.147, "step": 4768 }, { "epoch": 29.80625, "grad_norm": 1.609375, "learning_rate": 4.8672145740873374e-05, "loss": 0.1286, "step": 4769 }, { "epoch": 29.8125, "grad_norm": 2.09375, "learning_rate": 4.867131482355177e-05, "loss": 0.2183, "step": 4770 }, { "epoch": 29.81875, "grad_norm": 1.2890625, "learning_rate": 4.867048365343139e-05, "loss": 0.0762, "step": 4771 }, { "epoch": 29.825, "grad_norm": 2.203125, "learning_rate": 4.866965223052112e-05, "loss": 0.2505, "step": 4772 }, { "epoch": 29.83125, "grad_norm": 1.15625, "learning_rate": 4.866882055482983e-05, "loss": 0.0913, "step": 4773 }, { "epoch": 29.8375, "grad_norm": 1.046875, "learning_rate": 4.8667988626366414e-05, "loss": 0.0968, "step": 4774 }, { "epoch": 29.84375, "grad_norm": 0.84375, "learning_rate": 4.866715644513974e-05, "loss": 0.0854, "step": 4775 }, { "epoch": 29.85, "grad_norm": 1.7265625, "learning_rate": 4.86663240111587e-05, "loss": 0.1435, "step": 4776 }, { "epoch": 29.85625, "grad_norm": 1.2734375, "learning_rate": 4.86654913244322e-05, "loss": 0.1994, "step": 4777 }, { "epoch": 29.8625, "grad_norm": 1.9609375, "learning_rate": 4.866465838496912e-05, "loss": 0.1644, "step": 4778 }, { "epoch": 29.86875, "grad_norm": 0.94140625, "learning_rate": 4.8663825192778344e-05, "loss": 0.0715, "step": 4779 }, { "epoch": 29.875, "grad_norm": 1.6640625, "learning_rate": 4.8662991747868795e-05, "loss": 0.1486, "step": 4780 }, { "epoch": 29.88125, "grad_norm": 1.3359375, "learning_rate": 4.866215805024935e-05, "loss": 0.1336, "step": 4781 }, { "epoch": 29.8875, "grad_norm": 1.0234375, "learning_rate": 4.8661324099928936e-05, "loss": 0.0857, "step": 4782 }, { "epoch": 29.89375, "grad_norm": 1.6484375, "learning_rate": 4.866048989691644e-05, "loss": 0.1783, "step": 4783 }, { "epoch": 29.9, "grad_norm": 1.3203125, "learning_rate": 4.865965544122078e-05, "loss": 0.0954, "step": 4784 }, { "epoch": 29.90625, "grad_norm": 1.7421875, "learning_rate": 4.865882073285086e-05, "loss": 0.2191, "step": 4785 }, { "epoch": 29.9125, "grad_norm": 0.88671875, "learning_rate": 4.86579857718156e-05, "loss": 0.1184, "step": 4786 }, { "epoch": 29.91875, "grad_norm": 1.2265625, "learning_rate": 4.865715055812392e-05, "loss": 0.1181, "step": 4787 }, { "epoch": 29.925, "grad_norm": 1.4375, "learning_rate": 4.865631509178473e-05, "loss": 0.1561, "step": 4788 }, { "epoch": 29.93125, "grad_norm": 1.09375, "learning_rate": 4.865547937280697e-05, "loss": 0.092, "step": 4789 }, { "epoch": 29.9375, "grad_norm": 1.8125, "learning_rate": 4.865464340119955e-05, "loss": 0.1509, "step": 4790 }, { "epoch": 29.94375, "grad_norm": 1.1328125, "learning_rate": 4.86538071769714e-05, "loss": 0.094, "step": 4791 }, { "epoch": 29.95, "grad_norm": 0.93359375, "learning_rate": 4.865297070013145e-05, "loss": 0.0835, "step": 4792 }, { "epoch": 29.95625, "grad_norm": 1.140625, "learning_rate": 4.8652133970688636e-05, "loss": 0.0964, "step": 4793 }, { "epoch": 29.9625, "grad_norm": 0.88671875, "learning_rate": 4.86512969886519e-05, "loss": 0.0917, "step": 4794 }, { "epoch": 29.96875, "grad_norm": 2.25, "learning_rate": 4.8650459754030165e-05, "loss": 0.2711, "step": 4795 }, { "epoch": 29.975, "grad_norm": 1.421875, "learning_rate": 4.8649622266832376e-05, "loss": 0.136, "step": 4796 }, { "epoch": 29.98125, "grad_norm": 0.8359375, "learning_rate": 4.8648784527067495e-05, "loss": 0.0825, "step": 4797 }, { "epoch": 29.9875, "grad_norm": 0.85546875, "learning_rate": 4.864794653474445e-05, "loss": 0.0711, "step": 4798 }, { "epoch": 29.99375, "grad_norm": 1.6171875, "learning_rate": 4.864710828987219e-05, "loss": 0.1507, "step": 4799 }, { "epoch": 30.0, "grad_norm": 1.3203125, "learning_rate": 4.864626979245968e-05, "loss": 0.115, "step": 4800 }, { "epoch": 30.00625, "grad_norm": 0.83203125, "learning_rate": 4.864543104251587e-05, "loss": 0.0808, "step": 4801 }, { "epoch": 30.0125, "grad_norm": 1.40625, "learning_rate": 4.8644592040049706e-05, "loss": 0.1413, "step": 4802 }, { "epoch": 30.01875, "grad_norm": 1.046875, "learning_rate": 4.8643752785070175e-05, "loss": 0.085, "step": 4803 }, { "epoch": 30.025, "grad_norm": 1.609375, "learning_rate": 4.8642913277586206e-05, "loss": 0.1235, "step": 4804 }, { "epoch": 30.03125, "grad_norm": 1.6015625, "learning_rate": 4.864207351760678e-05, "loss": 0.1366, "step": 4805 }, { "epoch": 30.0375, "grad_norm": 1.3984375, "learning_rate": 4.864123350514088e-05, "loss": 0.1339, "step": 4806 }, { "epoch": 30.04375, "grad_norm": 1.03125, "learning_rate": 4.864039324019745e-05, "loss": 0.1025, "step": 4807 }, { "epoch": 30.05, "grad_norm": 1.1640625, "learning_rate": 4.863955272278549e-05, "loss": 0.1238, "step": 4808 }, { "epoch": 30.05625, "grad_norm": 1.671875, "learning_rate": 4.863871195291395e-05, "loss": 0.1663, "step": 4809 }, { "epoch": 30.0625, "grad_norm": 0.890625, "learning_rate": 4.863787093059182e-05, "loss": 0.0669, "step": 4810 }, { "epoch": 30.06875, "grad_norm": 1.6171875, "learning_rate": 4.8637029655828095e-05, "loss": 0.1541, "step": 4811 }, { "epoch": 30.075, "grad_norm": 1.21875, "learning_rate": 4.8636188128631746e-05, "loss": 0.1083, "step": 4812 }, { "epoch": 30.08125, "grad_norm": 1.5625, "learning_rate": 4.863534634901176e-05, "loss": 0.1189, "step": 4813 }, { "epoch": 30.0875, "grad_norm": 1.296875, "learning_rate": 4.863450431697713e-05, "loss": 0.0864, "step": 4814 }, { "epoch": 30.09375, "grad_norm": 1.1171875, "learning_rate": 4.863366203253685e-05, "loss": 0.0767, "step": 4815 }, { "epoch": 30.1, "grad_norm": 0.5859375, "learning_rate": 4.863281949569991e-05, "loss": 0.058, "step": 4816 }, { "epoch": 30.10625, "grad_norm": 0.5546875, "learning_rate": 4.863197670647531e-05, "loss": 0.0558, "step": 4817 }, { "epoch": 30.1125, "grad_norm": 1.3359375, "learning_rate": 4.8631133664872054e-05, "loss": 0.0919, "step": 4818 }, { "epoch": 30.11875, "grad_norm": 1.234375, "learning_rate": 4.863029037089914e-05, "loss": 0.1579, "step": 4819 }, { "epoch": 30.125, "grad_norm": 1.609375, "learning_rate": 4.862944682456558e-05, "loss": 0.1447, "step": 4820 }, { "epoch": 30.13125, "grad_norm": 1.2421875, "learning_rate": 4.862860302588037e-05, "loss": 0.1112, "step": 4821 }, { "epoch": 30.1375, "grad_norm": 1.6796875, "learning_rate": 4.8627758974852544e-05, "loss": 0.1907, "step": 4822 }, { "epoch": 30.14375, "grad_norm": 0.96484375, "learning_rate": 4.86269146714911e-05, "loss": 0.0647, "step": 4823 }, { "epoch": 30.15, "grad_norm": 1.1015625, "learning_rate": 4.8626070115805053e-05, "loss": 0.0886, "step": 4824 }, { "epoch": 30.15625, "grad_norm": 1.1796875, "learning_rate": 4.862522530780342e-05, "loss": 0.1017, "step": 4825 }, { "epoch": 30.1625, "grad_norm": 0.9296875, "learning_rate": 4.862438024749524e-05, "loss": 0.0799, "step": 4826 }, { "epoch": 30.16875, "grad_norm": 1.7265625, "learning_rate": 4.862353493488953e-05, "loss": 0.1738, "step": 4827 }, { "epoch": 30.175, "grad_norm": 1.140625, "learning_rate": 4.862268936999531e-05, "loss": 0.0721, "step": 4828 }, { "epoch": 30.18125, "grad_norm": 1.46875, "learning_rate": 4.862184355282162e-05, "loss": 0.1061, "step": 4829 }, { "epoch": 30.1875, "grad_norm": 1.6640625, "learning_rate": 4.862099748337749e-05, "loss": 0.1307, "step": 4830 }, { "epoch": 30.19375, "grad_norm": 1.4609375, "learning_rate": 4.862015116167196e-05, "loss": 0.1169, "step": 4831 }, { "epoch": 30.2, "grad_norm": 1.2421875, "learning_rate": 4.861930458771405e-05, "loss": 0.1418, "step": 4832 }, { "epoch": 30.20625, "grad_norm": 1.375, "learning_rate": 4.861845776151281e-05, "loss": 0.1094, "step": 4833 }, { "epoch": 30.2125, "grad_norm": 1.2734375, "learning_rate": 4.86176106830773e-05, "loss": 0.188, "step": 4834 }, { "epoch": 30.21875, "grad_norm": 1.015625, "learning_rate": 4.861676335241655e-05, "loss": 0.1068, "step": 4835 }, { "epoch": 30.225, "grad_norm": 1.1796875, "learning_rate": 4.8615915769539615e-05, "loss": 0.1179, "step": 4836 }, { "epoch": 30.23125, "grad_norm": 1.5625, "learning_rate": 4.861506793445554e-05, "loss": 0.1287, "step": 4837 }, { "epoch": 30.2375, "grad_norm": 1.0390625, "learning_rate": 4.861421984717339e-05, "loss": 0.0828, "step": 4838 }, { "epoch": 30.24375, "grad_norm": 1.0390625, "learning_rate": 4.861337150770222e-05, "loss": 0.0931, "step": 4839 }, { "epoch": 30.25, "grad_norm": 0.68359375, "learning_rate": 4.861252291605108e-05, "loss": 0.0588, "step": 4840 }, { "epoch": 30.25625, "grad_norm": 1.5, "learning_rate": 4.8611674072229044e-05, "loss": 0.1329, "step": 4841 }, { "epoch": 30.2625, "grad_norm": 1.4453125, "learning_rate": 4.8610824976245165e-05, "loss": 0.1109, "step": 4842 }, { "epoch": 30.26875, "grad_norm": 1.609375, "learning_rate": 4.8609975628108516e-05, "loss": 0.1621, "step": 4843 }, { "epoch": 30.275, "grad_norm": 1.171875, "learning_rate": 4.8609126027828176e-05, "loss": 0.1213, "step": 4844 }, { "epoch": 30.28125, "grad_norm": 1.0, "learning_rate": 4.860827617541321e-05, "loss": 0.0994, "step": 4845 }, { "epoch": 30.2875, "grad_norm": 0.8671875, "learning_rate": 4.86074260708727e-05, "loss": 0.0809, "step": 4846 }, { "epoch": 30.29375, "grad_norm": 1.40625, "learning_rate": 4.8606575714215716e-05, "loss": 0.1041, "step": 4847 }, { "epoch": 30.3, "grad_norm": 0.71875, "learning_rate": 4.860572510545135e-05, "loss": 0.0697, "step": 4848 }, { "epoch": 30.30625, "grad_norm": 1.5546875, "learning_rate": 4.860487424458867e-05, "loss": 0.1371, "step": 4849 }, { "epoch": 30.3125, "grad_norm": 1.1328125, "learning_rate": 4.860402313163678e-05, "loss": 0.098, "step": 4850 }, { "epoch": 30.31875, "grad_norm": 1.84375, "learning_rate": 4.8603171766604767e-05, "loss": 0.1473, "step": 4851 }, { "epoch": 30.325, "grad_norm": 1.2734375, "learning_rate": 4.860232014950171e-05, "loss": 0.1057, "step": 4852 }, { "epoch": 30.33125, "grad_norm": 0.9921875, "learning_rate": 4.8601468280336726e-05, "loss": 0.0774, "step": 4853 }, { "epoch": 30.3375, "grad_norm": 0.9609375, "learning_rate": 4.860061615911889e-05, "loss": 0.0779, "step": 4854 }, { "epoch": 30.34375, "grad_norm": 1.2109375, "learning_rate": 4.859976378585731e-05, "loss": 0.1186, "step": 4855 }, { "epoch": 30.35, "grad_norm": 1.375, "learning_rate": 4.8598911160561095e-05, "loss": 0.1413, "step": 4856 }, { "epoch": 30.35625, "grad_norm": 1.0078125, "learning_rate": 4.8598058283239344e-05, "loss": 0.1019, "step": 4857 }, { "epoch": 30.3625, "grad_norm": 1.4453125, "learning_rate": 4.8597205153901173e-05, "loss": 0.1405, "step": 4858 }, { "epoch": 30.36875, "grad_norm": 1.0546875, "learning_rate": 4.859635177255568e-05, "loss": 0.1291, "step": 4859 }, { "epoch": 30.375, "grad_norm": 1.546875, "learning_rate": 4.859549813921199e-05, "loss": 0.155, "step": 4860 }, { "epoch": 30.38125, "grad_norm": 1.625, "learning_rate": 4.859464425387922e-05, "loss": 0.2259, "step": 4861 }, { "epoch": 30.3875, "grad_norm": 1.6796875, "learning_rate": 4.8593790116566473e-05, "loss": 0.165, "step": 4862 }, { "epoch": 30.39375, "grad_norm": 1.1171875, "learning_rate": 4.859293572728289e-05, "loss": 0.0738, "step": 4863 }, { "epoch": 30.4, "grad_norm": 1.234375, "learning_rate": 4.859208108603759e-05, "loss": 0.135, "step": 4864 }, { "epoch": 30.40625, "grad_norm": 2.03125, "learning_rate": 4.8591226192839696e-05, "loss": 0.2317, "step": 4865 }, { "epoch": 30.4125, "grad_norm": 0.765625, "learning_rate": 4.859037104769834e-05, "loss": 0.0645, "step": 4866 }, { "epoch": 30.41875, "grad_norm": 1.234375, "learning_rate": 4.858951565062265e-05, "loss": 0.1497, "step": 4867 }, { "epoch": 30.425, "grad_norm": 0.875, "learning_rate": 4.858866000162178e-05, "loss": 0.0709, "step": 4868 }, { "epoch": 30.43125, "grad_norm": 1.203125, "learning_rate": 4.8587804100704845e-05, "loss": 0.1316, "step": 4869 }, { "epoch": 30.4375, "grad_norm": 1.640625, "learning_rate": 4.858694794788099e-05, "loss": 0.1503, "step": 4870 }, { "epoch": 30.44375, "grad_norm": 0.90625, "learning_rate": 4.858609154315937e-05, "loss": 0.0632, "step": 4871 }, { "epoch": 30.45, "grad_norm": 1.40625, "learning_rate": 4.8585234886549125e-05, "loss": 0.165, "step": 4872 }, { "epoch": 30.45625, "grad_norm": 1.265625, "learning_rate": 4.8584377978059395e-05, "loss": 0.1372, "step": 4873 }, { "epoch": 30.4625, "grad_norm": 0.9140625, "learning_rate": 4.8583520817699344e-05, "loss": 0.0769, "step": 4874 }, { "epoch": 30.46875, "grad_norm": 1.0625, "learning_rate": 4.8582663405478125e-05, "loss": 0.0931, "step": 4875 }, { "epoch": 30.475, "grad_norm": 1.703125, "learning_rate": 4.8581805741404886e-05, "loss": 0.1657, "step": 4876 }, { "epoch": 30.48125, "grad_norm": 0.9453125, "learning_rate": 4.858094782548879e-05, "loss": 0.1227, "step": 4877 }, { "epoch": 30.4875, "grad_norm": 1.4296875, "learning_rate": 4.8580089657738996e-05, "loss": 0.1176, "step": 4878 }, { "epoch": 30.49375, "grad_norm": 1.515625, "learning_rate": 4.857923123816468e-05, "loss": 0.1557, "step": 4879 }, { "epoch": 30.5, "grad_norm": 0.859375, "learning_rate": 4.8578372566775006e-05, "loss": 0.0691, "step": 4880 }, { "epoch": 30.50625, "grad_norm": 1.640625, "learning_rate": 4.857751364357913e-05, "loss": 0.1296, "step": 4881 }, { "epoch": 30.5125, "grad_norm": 1.328125, "learning_rate": 4.857665446858625e-05, "loss": 0.1103, "step": 4882 }, { "epoch": 30.51875, "grad_norm": 1.2890625, "learning_rate": 4.857579504180552e-05, "loss": 0.1346, "step": 4883 }, { "epoch": 30.525, "grad_norm": 1.2265625, "learning_rate": 4.857493536324612e-05, "loss": 0.0992, "step": 4884 }, { "epoch": 30.53125, "grad_norm": 1.6328125, "learning_rate": 4.857407543291725e-05, "loss": 0.1539, "step": 4885 }, { "epoch": 30.5375, "grad_norm": 1.8203125, "learning_rate": 4.857321525082808e-05, "loss": 0.1768, "step": 4886 }, { "epoch": 30.54375, "grad_norm": 1.28125, "learning_rate": 4.857235481698778e-05, "loss": 0.1244, "step": 4887 }, { "epoch": 30.55, "grad_norm": 1.4609375, "learning_rate": 4.857149413140557e-05, "loss": 0.1383, "step": 4888 }, { "epoch": 30.55625, "grad_norm": 1.2578125, "learning_rate": 4.857063319409062e-05, "loss": 0.1163, "step": 4889 }, { "epoch": 30.5625, "grad_norm": 1.3515625, "learning_rate": 4.856977200505214e-05, "loss": 0.114, "step": 4890 }, { "epoch": 30.56875, "grad_norm": 0.85546875, "learning_rate": 4.856891056429931e-05, "loss": 0.0741, "step": 4891 }, { "epoch": 30.575, "grad_norm": 1.4140625, "learning_rate": 4.856804887184134e-05, "loss": 0.122, "step": 4892 }, { "epoch": 30.58125, "grad_norm": 1.640625, "learning_rate": 4.8567186927687445e-05, "loss": 0.1324, "step": 4893 }, { "epoch": 30.5875, "grad_norm": 1.2421875, "learning_rate": 4.85663247318468e-05, "loss": 0.1092, "step": 4894 }, { "epoch": 30.59375, "grad_norm": 1.6875, "learning_rate": 4.8565462284328636e-05, "loss": 0.1525, "step": 4895 }, { "epoch": 30.6, "grad_norm": 1.2890625, "learning_rate": 4.8564599585142154e-05, "loss": 0.0969, "step": 4896 }, { "epoch": 30.60625, "grad_norm": 1.3125, "learning_rate": 4.856373663429657e-05, "loss": 0.1294, "step": 4897 }, { "epoch": 30.6125, "grad_norm": 1.4375, "learning_rate": 4.8562873431801115e-05, "loss": 0.1136, "step": 4898 }, { "epoch": 30.61875, "grad_norm": 0.98828125, "learning_rate": 4.856200997766498e-05, "loss": 0.073, "step": 4899 }, { "epoch": 30.625, "grad_norm": 1.375, "learning_rate": 4.85611462718974e-05, "loss": 0.0948, "step": 4900 }, { "epoch": 30.63125, "grad_norm": 1.0, "learning_rate": 4.8560282314507596e-05, "loss": 0.1284, "step": 4901 }, { "epoch": 30.6375, "grad_norm": 1.3125, "learning_rate": 4.8559418105504795e-05, "loss": 0.1166, "step": 4902 }, { "epoch": 30.64375, "grad_norm": 1.0078125, "learning_rate": 4.855855364489823e-05, "loss": 0.0853, "step": 4903 }, { "epoch": 30.65, "grad_norm": 1.65625, "learning_rate": 4.8557688932697135e-05, "loss": 0.0898, "step": 4904 }, { "epoch": 30.65625, "grad_norm": 1.625, "learning_rate": 4.8556823968910734e-05, "loss": 0.1938, "step": 4905 }, { "epoch": 30.6625, "grad_norm": 1.3984375, "learning_rate": 4.855595875354828e-05, "loss": 0.1664, "step": 4906 }, { "epoch": 30.66875, "grad_norm": 1.515625, "learning_rate": 4.8555093286618995e-05, "loss": 0.1731, "step": 4907 }, { "epoch": 30.675, "grad_norm": 1.328125, "learning_rate": 4.855422756813214e-05, "loss": 0.1017, "step": 4908 }, { "epoch": 30.68125, "grad_norm": 0.9765625, "learning_rate": 4.8553361598096945e-05, "loss": 0.073, "step": 4909 }, { "epoch": 30.6875, "grad_norm": 1.3359375, "learning_rate": 4.8552495376522665e-05, "loss": 0.1266, "step": 4910 }, { "epoch": 30.69375, "grad_norm": 2.1875, "learning_rate": 4.8551628903418556e-05, "loss": 0.1904, "step": 4911 }, { "epoch": 30.7, "grad_norm": 1.375, "learning_rate": 4.855076217879386e-05, "loss": 0.1427, "step": 4912 }, { "epoch": 30.70625, "grad_norm": 1.390625, "learning_rate": 4.8549895202657844e-05, "loss": 0.1032, "step": 4913 }, { "epoch": 30.7125, "grad_norm": 1.265625, "learning_rate": 4.8549027975019754e-05, "loss": 0.1307, "step": 4914 }, { "epoch": 30.71875, "grad_norm": 1.109375, "learning_rate": 4.8548160495888874e-05, "loss": 0.0974, "step": 4915 }, { "epoch": 30.725, "grad_norm": 1.40625, "learning_rate": 4.854729276527444e-05, "loss": 0.1077, "step": 4916 }, { "epoch": 30.73125, "grad_norm": 1.328125, "learning_rate": 4.854642478318574e-05, "loss": 0.123, "step": 4917 }, { "epoch": 30.7375, "grad_norm": 1.8046875, "learning_rate": 4.8545556549632034e-05, "loss": 0.1888, "step": 4918 }, { "epoch": 30.74375, "grad_norm": 1.6875, "learning_rate": 4.85446880646226e-05, "loss": 0.178, "step": 4919 }, { "epoch": 30.75, "grad_norm": 1.0625, "learning_rate": 4.8543819328166706e-05, "loss": 0.0922, "step": 4920 }, { "epoch": 30.75625, "grad_norm": 1.328125, "learning_rate": 4.854295034027364e-05, "loss": 0.1136, "step": 4921 }, { "epoch": 30.7625, "grad_norm": 1.1015625, "learning_rate": 4.854208110095267e-05, "loss": 0.0794, "step": 4922 }, { "epoch": 30.76875, "grad_norm": 2.109375, "learning_rate": 4.854121161021309e-05, "loss": 0.1787, "step": 4923 }, { "epoch": 30.775, "grad_norm": 1.5078125, "learning_rate": 4.854034186806418e-05, "loss": 0.1338, "step": 4924 }, { "epoch": 30.78125, "grad_norm": 1.15625, "learning_rate": 4.853947187451522e-05, "loss": 0.0896, "step": 4925 }, { "epoch": 30.7875, "grad_norm": 0.859375, "learning_rate": 4.853860162957552e-05, "loss": 0.0774, "step": 4926 }, { "epoch": 30.79375, "grad_norm": 1.0625, "learning_rate": 4.853773113325436e-05, "loss": 0.0804, "step": 4927 }, { "epoch": 30.8, "grad_norm": 1.515625, "learning_rate": 4.8536860385561046e-05, "loss": 0.1321, "step": 4928 }, { "epoch": 30.80625, "grad_norm": 0.79296875, "learning_rate": 4.853598938650487e-05, "loss": 0.0712, "step": 4929 }, { "epoch": 30.8125, "grad_norm": 1.375, "learning_rate": 4.853511813609514e-05, "loss": 0.1194, "step": 4930 }, { "epoch": 30.81875, "grad_norm": 1.3046875, "learning_rate": 4.853424663434115e-05, "loss": 0.1243, "step": 4931 }, { "epoch": 30.825, "grad_norm": 1.1875, "learning_rate": 4.853337488125221e-05, "loss": 0.1339, "step": 4932 }, { "epoch": 30.83125, "grad_norm": 0.734375, "learning_rate": 4.853250287683765e-05, "loss": 0.0687, "step": 4933 }, { "epoch": 30.8375, "grad_norm": 0.80078125, "learning_rate": 4.8531630621106746e-05, "loss": 0.0777, "step": 4934 }, { "epoch": 30.84375, "grad_norm": 1.078125, "learning_rate": 4.853075811406885e-05, "loss": 0.1114, "step": 4935 }, { "epoch": 30.85, "grad_norm": 1.6640625, "learning_rate": 4.852988535573325e-05, "loss": 0.1666, "step": 4936 }, { "epoch": 30.85625, "grad_norm": 1.1953125, "learning_rate": 4.852901234610929e-05, "loss": 0.1172, "step": 4937 }, { "epoch": 30.8625, "grad_norm": 1.4140625, "learning_rate": 4.8528139085206284e-05, "loss": 0.17, "step": 4938 }, { "epoch": 30.86875, "grad_norm": 1.6015625, "learning_rate": 4.8527265573033553e-05, "loss": 0.1362, "step": 4939 }, { "epoch": 30.875, "grad_norm": 1.359375, "learning_rate": 4.852639180960043e-05, "loss": 0.1268, "step": 4940 }, { "epoch": 30.88125, "grad_norm": 0.96484375, "learning_rate": 4.852551779491625e-05, "loss": 0.0805, "step": 4941 }, { "epoch": 30.8875, "grad_norm": 0.953125, "learning_rate": 4.852464352899034e-05, "loss": 0.0776, "step": 4942 }, { "epoch": 30.89375, "grad_norm": 1.0703125, "learning_rate": 4.852376901183204e-05, "loss": 0.0835, "step": 4943 }, { "epoch": 30.9, "grad_norm": 1.5625, "learning_rate": 4.852289424345069e-05, "loss": 0.1407, "step": 4944 }, { "epoch": 30.90625, "grad_norm": 0.6171875, "learning_rate": 4.852201922385564e-05, "loss": 0.0577, "step": 4945 }, { "epoch": 30.9125, "grad_norm": 1.21875, "learning_rate": 4.852114395305622e-05, "loss": 0.1168, "step": 4946 }, { "epoch": 30.91875, "grad_norm": 1.6875, "learning_rate": 4.852026843106179e-05, "loss": 0.1704, "step": 4947 }, { "epoch": 30.925, "grad_norm": 1.140625, "learning_rate": 4.851939265788169e-05, "loss": 0.1127, "step": 4948 }, { "epoch": 30.93125, "grad_norm": 1.3984375, "learning_rate": 4.851851663352527e-05, "loss": 0.136, "step": 4949 }, { "epoch": 30.9375, "grad_norm": 0.96875, "learning_rate": 4.8517640358001904e-05, "loss": 0.1091, "step": 4950 }, { "epoch": 30.94375, "grad_norm": 1.5078125, "learning_rate": 4.8516763831320935e-05, "loss": 0.1823, "step": 4951 }, { "epoch": 30.95, "grad_norm": 1.234375, "learning_rate": 4.851588705349173e-05, "loss": 0.1013, "step": 4952 }, { "epoch": 30.95625, "grad_norm": 1.5078125, "learning_rate": 4.851501002452364e-05, "loss": 0.1291, "step": 4953 }, { "epoch": 30.9625, "grad_norm": 0.94921875, "learning_rate": 4.851413274442606e-05, "loss": 0.0728, "step": 4954 }, { "epoch": 30.96875, "grad_norm": 1.453125, "learning_rate": 4.851325521320833e-05, "loss": 0.1718, "step": 4955 }, { "epoch": 30.975, "grad_norm": 1.34375, "learning_rate": 4.851237743087983e-05, "loss": 0.133, "step": 4956 }, { "epoch": 30.98125, "grad_norm": 1.640625, "learning_rate": 4.8511499397449945e-05, "loss": 0.1506, "step": 4957 }, { "epoch": 30.9875, "grad_norm": 1.34375, "learning_rate": 4.851062111292803e-05, "loss": 0.1219, "step": 4958 }, { "epoch": 30.99375, "grad_norm": 1.265625, "learning_rate": 4.85097425773235e-05, "loss": 0.1299, "step": 4959 }, { "epoch": 31.0, "grad_norm": 1.4140625, "learning_rate": 4.85088637906457e-05, "loss": 0.1618, "step": 4960 }, { "epoch": 31.00625, "grad_norm": 1.734375, "learning_rate": 4.850798475290403e-05, "loss": 0.1687, "step": 4961 }, { "epoch": 31.0125, "grad_norm": 0.93359375, "learning_rate": 4.850710546410789e-05, "loss": 0.0754, "step": 4962 }, { "epoch": 31.01875, "grad_norm": 0.59765625, "learning_rate": 4.8506225924266646e-05, "loss": 0.0543, "step": 4963 }, { "epoch": 31.025, "grad_norm": 0.96484375, "learning_rate": 4.8505346133389716e-05, "loss": 0.091, "step": 4964 }, { "epoch": 31.03125, "grad_norm": 1.1484375, "learning_rate": 4.850446609148648e-05, "loss": 0.1014, "step": 4965 }, { "epoch": 31.0375, "grad_norm": 1.2734375, "learning_rate": 4.850358579856633e-05, "loss": 0.1253, "step": 4966 }, { "epoch": 31.04375, "grad_norm": 1.1953125, "learning_rate": 4.850270525463869e-05, "loss": 0.1189, "step": 4967 }, { "epoch": 31.05, "grad_norm": 1.1875, "learning_rate": 4.8501824459712945e-05, "loss": 0.0996, "step": 4968 }, { "epoch": 31.05625, "grad_norm": 2.203125, "learning_rate": 4.850094341379851e-05, "loss": 0.1718, "step": 4969 }, { "epoch": 31.0625, "grad_norm": 1.1953125, "learning_rate": 4.850006211690479e-05, "loss": 0.0859, "step": 4970 }, { "epoch": 31.06875, "grad_norm": 0.828125, "learning_rate": 4.84991805690412e-05, "loss": 0.0798, "step": 4971 }, { "epoch": 31.075, "grad_norm": 0.7890625, "learning_rate": 4.849829877021715e-05, "loss": 0.0717, "step": 4972 }, { "epoch": 31.08125, "grad_norm": 1.3203125, "learning_rate": 4.849741672044207e-05, "loss": 0.1243, "step": 4973 }, { "epoch": 31.0875, "grad_norm": 0.92578125, "learning_rate": 4.8496534419725356e-05, "loss": 0.0655, "step": 4974 }, { "epoch": 31.09375, "grad_norm": 1.890625, "learning_rate": 4.849565186807646e-05, "loss": 0.1414, "step": 4975 }, { "epoch": 31.1, "grad_norm": 1.5390625, "learning_rate": 4.849476906550479e-05, "loss": 0.1272, "step": 4976 }, { "epoch": 31.10625, "grad_norm": 0.98046875, "learning_rate": 4.8493886012019765e-05, "loss": 0.095, "step": 4977 }, { "epoch": 31.1125, "grad_norm": 0.91796875, "learning_rate": 4.849300270763083e-05, "loss": 0.0802, "step": 4978 }, { "epoch": 31.11875, "grad_norm": 0.9765625, "learning_rate": 4.849211915234742e-05, "loss": 0.0932, "step": 4979 }, { "epoch": 31.125, "grad_norm": 1.0859375, "learning_rate": 4.849123534617897e-05, "loss": 0.1066, "step": 4980 }, { "epoch": 31.13125, "grad_norm": 0.85546875, "learning_rate": 4.84903512891349e-05, "loss": 0.0846, "step": 4981 }, { "epoch": 31.1375, "grad_norm": 1.125, "learning_rate": 4.8489466981224675e-05, "loss": 0.1015, "step": 4982 }, { "epoch": 31.14375, "grad_norm": 0.90234375, "learning_rate": 4.848858242245773e-05, "loss": 0.1001, "step": 4983 }, { "epoch": 31.15, "grad_norm": 1.3203125, "learning_rate": 4.848769761284351e-05, "loss": 0.1803, "step": 4984 }, { "epoch": 31.15625, "grad_norm": 0.6875, "learning_rate": 4.848681255239146e-05, "loss": 0.0689, "step": 4985 }, { "epoch": 31.1625, "grad_norm": 0.859375, "learning_rate": 4.848592724111105e-05, "loss": 0.0849, "step": 4986 }, { "epoch": 31.16875, "grad_norm": 1.28125, "learning_rate": 4.848504167901171e-05, "loss": 0.1496, "step": 4987 }, { "epoch": 31.175, "grad_norm": 1.859375, "learning_rate": 4.848415586610292e-05, "loss": 0.2033, "step": 4988 }, { "epoch": 31.18125, "grad_norm": 0.734375, "learning_rate": 4.848326980239413e-05, "loss": 0.067, "step": 4989 }, { "epoch": 31.1875, "grad_norm": 0.99609375, "learning_rate": 4.84823834878948e-05, "loss": 0.0906, "step": 4990 }, { "epoch": 31.19375, "grad_norm": 1.4140625, "learning_rate": 4.84814969226144e-05, "loss": 0.1726, "step": 4991 }, { "epoch": 31.2, "grad_norm": 1.1796875, "learning_rate": 4.848061010656239e-05, "loss": 0.1206, "step": 4992 }, { "epoch": 31.20625, "grad_norm": 1.21875, "learning_rate": 4.8479723039748246e-05, "loss": 0.0976, "step": 4993 }, { "epoch": 31.2125, "grad_norm": 1.1875, "learning_rate": 4.847883572218145e-05, "loss": 0.1344, "step": 4994 }, { "epoch": 31.21875, "grad_norm": 0.92578125, "learning_rate": 4.847794815387147e-05, "loss": 0.0838, "step": 4995 }, { "epoch": 31.225, "grad_norm": 1.5703125, "learning_rate": 4.847706033482778e-05, "loss": 0.1547, "step": 4996 }, { "epoch": 31.23125, "grad_norm": 1.0859375, "learning_rate": 4.8476172265059884e-05, "loss": 0.1018, "step": 4997 }, { "epoch": 31.2375, "grad_norm": 1.5859375, "learning_rate": 4.8475283944577234e-05, "loss": 0.128, "step": 4998 }, { "epoch": 31.24375, "grad_norm": 1.4140625, "learning_rate": 4.847439537338934e-05, "loss": 0.1699, "step": 4999 }, { "epoch": 31.25, "grad_norm": 1.015625, "learning_rate": 4.847350655150567e-05, "loss": 0.1289, "step": 5000 }, { "epoch": 31.25625, "grad_norm": 1.46875, "learning_rate": 4.8472617478935746e-05, "loss": 0.1979, "step": 5001 }, { "epoch": 31.2625, "grad_norm": 1.640625, "learning_rate": 4.8471728155689035e-05, "loss": 0.1291, "step": 5002 }, { "epoch": 31.26875, "grad_norm": 1.0234375, "learning_rate": 4.8470838581775046e-05, "loss": 0.0729, "step": 5003 }, { "epoch": 31.275, "grad_norm": 1.296875, "learning_rate": 4.846994875720329e-05, "loss": 0.0831, "step": 5004 }, { "epoch": 31.28125, "grad_norm": 1.546875, "learning_rate": 4.8469058681983256e-05, "loss": 0.1186, "step": 5005 }, { "epoch": 31.2875, "grad_norm": 1.6796875, "learning_rate": 4.846816835612445e-05, "loss": 0.1581, "step": 5006 }, { "epoch": 31.29375, "grad_norm": 1.453125, "learning_rate": 4.8467277779636385e-05, "loss": 0.1005, "step": 5007 }, { "epoch": 31.3, "grad_norm": 0.8515625, "learning_rate": 4.8466386952528565e-05, "loss": 0.0774, "step": 5008 }, { "epoch": 31.30625, "grad_norm": 1.59375, "learning_rate": 4.846549587481052e-05, "loss": 0.1738, "step": 5009 }, { "epoch": 31.3125, "grad_norm": 1.171875, "learning_rate": 4.846460454649174e-05, "loss": 0.14, "step": 5010 }, { "epoch": 31.31875, "grad_norm": 1.1796875, "learning_rate": 4.8463712967581776e-05, "loss": 0.0945, "step": 5011 }, { "epoch": 31.325, "grad_norm": 0.6953125, "learning_rate": 4.8462821138090125e-05, "loss": 0.0723, "step": 5012 }, { "epoch": 31.33125, "grad_norm": 1.21875, "learning_rate": 4.8461929058026314e-05, "loss": 0.1154, "step": 5013 }, { "epoch": 31.3375, "grad_norm": 1.1796875, "learning_rate": 4.846103672739989e-05, "loss": 0.0823, "step": 5014 }, { "epoch": 31.34375, "grad_norm": 1.28125, "learning_rate": 4.846014414622036e-05, "loss": 0.1127, "step": 5015 }, { "epoch": 31.35, "grad_norm": 0.94140625, "learning_rate": 4.845925131449725e-05, "loss": 0.1043, "step": 5016 }, { "epoch": 31.35625, "grad_norm": 1.4375, "learning_rate": 4.845835823224013e-05, "loss": 0.2017, "step": 5017 }, { "epoch": 31.3625, "grad_norm": 1.8984375, "learning_rate": 4.84574648994585e-05, "loss": 0.1903, "step": 5018 }, { "epoch": 31.36875, "grad_norm": 1.046875, "learning_rate": 4.845657131616194e-05, "loss": 0.0976, "step": 5019 }, { "epoch": 31.375, "grad_norm": 1.3203125, "learning_rate": 4.845567748235995e-05, "loss": 0.138, "step": 5020 }, { "epoch": 31.38125, "grad_norm": 1.796875, "learning_rate": 4.8454783398062106e-05, "loss": 0.1885, "step": 5021 }, { "epoch": 31.3875, "grad_norm": 1.359375, "learning_rate": 4.8453889063277946e-05, "loss": 0.1083, "step": 5022 }, { "epoch": 31.39375, "grad_norm": 1.21875, "learning_rate": 4.8452994478017015e-05, "loss": 0.1485, "step": 5023 }, { "epoch": 31.4, "grad_norm": 0.640625, "learning_rate": 4.845209964228888e-05, "loss": 0.0579, "step": 5024 }, { "epoch": 31.40625, "grad_norm": 1.046875, "learning_rate": 4.845120455610309e-05, "loss": 0.0791, "step": 5025 }, { "epoch": 31.4125, "grad_norm": 1.2109375, "learning_rate": 4.84503092194692e-05, "loss": 0.1166, "step": 5026 }, { "epoch": 31.41875, "grad_norm": 1.0859375, "learning_rate": 4.8449413632396783e-05, "loss": 0.1359, "step": 5027 }, { "epoch": 31.425, "grad_norm": 1.1328125, "learning_rate": 4.84485177948954e-05, "loss": 0.0782, "step": 5028 }, { "epoch": 31.43125, "grad_norm": 1.390625, "learning_rate": 4.8447621706974604e-05, "loss": 0.1118, "step": 5029 }, { "epoch": 31.4375, "grad_norm": 1.609375, "learning_rate": 4.8446725368643985e-05, "loss": 0.1718, "step": 5030 }, { "epoch": 31.44375, "grad_norm": 0.9609375, "learning_rate": 4.844582877991311e-05, "loss": 0.0784, "step": 5031 }, { "epoch": 31.45, "grad_norm": 1.15625, "learning_rate": 4.844493194079154e-05, "loss": 0.1051, "step": 5032 }, { "epoch": 31.45625, "grad_norm": 1.2578125, "learning_rate": 4.8444034851288866e-05, "loss": 0.1093, "step": 5033 }, { "epoch": 31.4625, "grad_norm": 0.875, "learning_rate": 4.844313751141467e-05, "loss": 0.0781, "step": 5034 }, { "epoch": 31.46875, "grad_norm": 0.625, "learning_rate": 4.844223992117853e-05, "loss": 0.0632, "step": 5035 }, { "epoch": 31.475, "grad_norm": 1.0703125, "learning_rate": 4.8441342080590036e-05, "loss": 0.1156, "step": 5036 }, { "epoch": 31.48125, "grad_norm": 1.40625, "learning_rate": 4.844044398965877e-05, "loss": 0.1481, "step": 5037 }, { "epoch": 31.4875, "grad_norm": 1.640625, "learning_rate": 4.843954564839432e-05, "loss": 0.124, "step": 5038 }, { "epoch": 31.49375, "grad_norm": 1.1875, "learning_rate": 4.8438647056806294e-05, "loss": 0.1091, "step": 5039 }, { "epoch": 31.5, "grad_norm": 1.1015625, "learning_rate": 4.843774821490429e-05, "loss": 0.1016, "step": 5040 }, { "epoch": 31.50625, "grad_norm": 1.4609375, "learning_rate": 4.8436849122697883e-05, "loss": 0.1191, "step": 5041 }, { "epoch": 31.5125, "grad_norm": 1.1328125, "learning_rate": 4.84359497801967e-05, "loss": 0.1135, "step": 5042 }, { "epoch": 31.51875, "grad_norm": 1.09375, "learning_rate": 4.843505018741034e-05, "loss": 0.0758, "step": 5043 }, { "epoch": 31.525, "grad_norm": 1.2890625, "learning_rate": 4.843415034434839e-05, "loss": 0.1219, "step": 5044 }, { "epoch": 31.53125, "grad_norm": 1.4453125, "learning_rate": 4.843325025102049e-05, "loss": 0.1568, "step": 5045 }, { "epoch": 31.5375, "grad_norm": 0.9375, "learning_rate": 4.843234990743624e-05, "loss": 0.0887, "step": 5046 }, { "epoch": 31.54375, "grad_norm": 0.9609375, "learning_rate": 4.843144931360525e-05, "loss": 0.0844, "step": 5047 }, { "epoch": 31.55, "grad_norm": 1.40625, "learning_rate": 4.843054846953714e-05, "loss": 0.1322, "step": 5048 }, { "epoch": 31.55625, "grad_norm": 1.1640625, "learning_rate": 4.842964737524153e-05, "loss": 0.1033, "step": 5049 }, { "epoch": 31.5625, "grad_norm": 1.375, "learning_rate": 4.842874603072804e-05, "loss": 0.1552, "step": 5050 }, { "epoch": 31.56875, "grad_norm": 0.7421875, "learning_rate": 4.842784443600632e-05, "loss": 0.063, "step": 5051 }, { "epoch": 31.575, "grad_norm": 0.91015625, "learning_rate": 4.842694259108597e-05, "loss": 0.0748, "step": 5052 }, { "epoch": 31.58125, "grad_norm": 0.9296875, "learning_rate": 4.8426040495976635e-05, "loss": 0.0751, "step": 5053 }, { "epoch": 31.5875, "grad_norm": 1.140625, "learning_rate": 4.842513815068794e-05, "loss": 0.0982, "step": 5054 }, { "epoch": 31.59375, "grad_norm": 1.171875, "learning_rate": 4.842423555522953e-05, "loss": 0.1111, "step": 5055 }, { "epoch": 31.6, "grad_norm": 1.1328125, "learning_rate": 4.842333270961104e-05, "loss": 0.064, "step": 5056 }, { "epoch": 31.60625, "grad_norm": 0.78515625, "learning_rate": 4.842242961384211e-05, "loss": 0.0657, "step": 5057 }, { "epoch": 31.6125, "grad_norm": 1.25, "learning_rate": 4.8421526267932406e-05, "loss": 0.1317, "step": 5058 }, { "epoch": 31.61875, "grad_norm": 1.6484375, "learning_rate": 4.8420622671891533e-05, "loss": 0.1612, "step": 5059 }, { "epoch": 31.625, "grad_norm": 1.109375, "learning_rate": 4.841971882572919e-05, "loss": 0.1075, "step": 5060 }, { "epoch": 31.63125, "grad_norm": 0.84765625, "learning_rate": 4.8418814729454985e-05, "loss": 0.0604, "step": 5061 }, { "epoch": 31.6375, "grad_norm": 1.0859375, "learning_rate": 4.84179103830786e-05, "loss": 0.0938, "step": 5062 }, { "epoch": 31.64375, "grad_norm": 1.671875, "learning_rate": 4.841700578660969e-05, "loss": 0.1254, "step": 5063 }, { "epoch": 31.65, "grad_norm": 1.4296875, "learning_rate": 4.8416100940057906e-05, "loss": 0.1184, "step": 5064 }, { "epoch": 31.65625, "grad_norm": 2.4375, "learning_rate": 4.8415195843432925e-05, "loss": 0.2943, "step": 5065 }, { "epoch": 31.6625, "grad_norm": 1.265625, "learning_rate": 4.8414290496744404e-05, "loss": 0.1355, "step": 5066 }, { "epoch": 31.66875, "grad_norm": 0.8125, "learning_rate": 4.841338490000201e-05, "loss": 0.073, "step": 5067 }, { "epoch": 31.675, "grad_norm": 1.375, "learning_rate": 4.841247905321542e-05, "loss": 0.1317, "step": 5068 }, { "epoch": 31.68125, "grad_norm": 1.1328125, "learning_rate": 4.8411572956394305e-05, "loss": 0.0964, "step": 5069 }, { "epoch": 31.6875, "grad_norm": 1.1328125, "learning_rate": 4.8410666609548335e-05, "loss": 0.0998, "step": 5070 }, { "epoch": 31.69375, "grad_norm": 1.140625, "learning_rate": 4.840976001268721e-05, "loss": 0.1155, "step": 5071 }, { "epoch": 31.7, "grad_norm": 1.8125, "learning_rate": 4.840885316582059e-05, "loss": 0.2074, "step": 5072 }, { "epoch": 31.70625, "grad_norm": 1.609375, "learning_rate": 4.8407946068958165e-05, "loss": 0.184, "step": 5073 }, { "epoch": 31.7125, "grad_norm": 1.5390625, "learning_rate": 4.840703872210963e-05, "loss": 0.1609, "step": 5074 }, { "epoch": 31.71875, "grad_norm": 1.2578125, "learning_rate": 4.840613112528467e-05, "loss": 0.0649, "step": 5075 }, { "epoch": 31.725, "grad_norm": 1.2734375, "learning_rate": 4.8405223278492986e-05, "loss": 0.1003, "step": 5076 }, { "epoch": 31.73125, "grad_norm": 1.4375, "learning_rate": 4.8404315181744256e-05, "loss": 0.1076, "step": 5077 }, { "epoch": 31.7375, "grad_norm": 1.109375, "learning_rate": 4.840340683504819e-05, "loss": 0.0942, "step": 5078 }, { "epoch": 31.74375, "grad_norm": 1.2109375, "learning_rate": 4.840249823841449e-05, "loss": 0.1604, "step": 5079 }, { "epoch": 31.75, "grad_norm": 1.421875, "learning_rate": 4.8401589391852855e-05, "loss": 0.1064, "step": 5080 }, { "epoch": 31.75625, "grad_norm": 1.3203125, "learning_rate": 4.840068029537299e-05, "loss": 0.121, "step": 5081 }, { "epoch": 31.7625, "grad_norm": 1.1484375, "learning_rate": 4.839977094898461e-05, "loss": 0.0916, "step": 5082 }, { "epoch": 31.76875, "grad_norm": 0.94921875, "learning_rate": 4.839886135269742e-05, "loss": 0.0839, "step": 5083 }, { "epoch": 31.775, "grad_norm": 1.0625, "learning_rate": 4.8397951506521134e-05, "loss": 0.1246, "step": 5084 }, { "epoch": 31.78125, "grad_norm": 0.93359375, "learning_rate": 4.8397041410465474e-05, "loss": 0.0699, "step": 5085 }, { "epoch": 31.7875, "grad_norm": 0.99609375, "learning_rate": 4.839613106454016e-05, "loss": 0.105, "step": 5086 }, { "epoch": 31.79375, "grad_norm": 1.2421875, "learning_rate": 4.8395220468754906e-05, "loss": 0.1646, "step": 5087 }, { "epoch": 31.8, "grad_norm": 1.59375, "learning_rate": 4.839430962311944e-05, "loss": 0.1903, "step": 5088 }, { "epoch": 31.80625, "grad_norm": 1.6171875, "learning_rate": 4.83933985276435e-05, "loss": 0.1983, "step": 5089 }, { "epoch": 31.8125, "grad_norm": 1.40625, "learning_rate": 4.839248718233679e-05, "loss": 0.1014, "step": 5090 }, { "epoch": 31.81875, "grad_norm": 1.453125, "learning_rate": 4.839157558720907e-05, "loss": 0.1461, "step": 5091 }, { "epoch": 31.825, "grad_norm": 1.3203125, "learning_rate": 4.8390663742270065e-05, "loss": 0.1188, "step": 5092 }, { "epoch": 31.83125, "grad_norm": 1.3046875, "learning_rate": 4.838975164752951e-05, "loss": 0.0959, "step": 5093 }, { "epoch": 31.8375, "grad_norm": 1.5078125, "learning_rate": 4.838883930299715e-05, "loss": 0.1882, "step": 5094 }, { "epoch": 31.84375, "grad_norm": 1.75, "learning_rate": 4.838792670868273e-05, "loss": 0.1801, "step": 5095 }, { "epoch": 31.85, "grad_norm": 0.93359375, "learning_rate": 4.838701386459599e-05, "loss": 0.0871, "step": 5096 }, { "epoch": 31.85625, "grad_norm": 1.6328125, "learning_rate": 4.838610077074669e-05, "loss": 0.2152, "step": 5097 }, { "epoch": 31.8625, "grad_norm": 1.328125, "learning_rate": 4.838518742714456e-05, "loss": 0.1186, "step": 5098 }, { "epoch": 31.86875, "grad_norm": 1.0546875, "learning_rate": 4.8384273833799375e-05, "loss": 0.0892, "step": 5099 }, { "epoch": 31.875, "grad_norm": 1.65625, "learning_rate": 4.838335999072089e-05, "loss": 0.1689, "step": 5100 }, { "epoch": 31.88125, "grad_norm": 1.1015625, "learning_rate": 4.838244589791885e-05, "loss": 0.0952, "step": 5101 }, { "epoch": 31.8875, "grad_norm": 1.578125, "learning_rate": 4.838153155540302e-05, "loss": 0.1683, "step": 5102 }, { "epoch": 31.89375, "grad_norm": 2.0, "learning_rate": 4.8380616963183187e-05, "loss": 0.1624, "step": 5103 }, { "epoch": 31.9, "grad_norm": 1.515625, "learning_rate": 4.8379702121269096e-05, "loss": 0.1587, "step": 5104 }, { "epoch": 31.90625, "grad_norm": 1.2578125, "learning_rate": 4.837878702967052e-05, "loss": 0.0964, "step": 5105 }, { "epoch": 31.9125, "grad_norm": 1.59375, "learning_rate": 4.837787168839724e-05, "loss": 0.1669, "step": 5106 }, { "epoch": 31.91875, "grad_norm": 0.83203125, "learning_rate": 4.837695609745902e-05, "loss": 0.0735, "step": 5107 }, { "epoch": 31.925, "grad_norm": 1.28125, "learning_rate": 4.837604025686564e-05, "loss": 0.1131, "step": 5108 }, { "epoch": 31.93125, "grad_norm": 0.90234375, "learning_rate": 4.8375124166626894e-05, "loss": 0.068, "step": 5109 }, { "epoch": 31.9375, "grad_norm": 1.5625, "learning_rate": 4.837420782675256e-05, "loss": 0.1206, "step": 5110 }, { "epoch": 31.94375, "grad_norm": 1.2578125, "learning_rate": 4.837329123725242e-05, "loss": 0.1415, "step": 5111 }, { "epoch": 31.95, "grad_norm": 1.53125, "learning_rate": 4.8372374398136256e-05, "loss": 0.1038, "step": 5112 }, { "epoch": 31.95625, "grad_norm": 0.90234375, "learning_rate": 4.837145730941387e-05, "loss": 0.0714, "step": 5113 }, { "epoch": 31.9625, "grad_norm": 0.890625, "learning_rate": 4.837053997109505e-05, "loss": 0.1206, "step": 5114 }, { "epoch": 31.96875, "grad_norm": 1.125, "learning_rate": 4.836962238318959e-05, "loss": 0.1034, "step": 5115 }, { "epoch": 31.975, "grad_norm": 1.1015625, "learning_rate": 4.836870454570731e-05, "loss": 0.1298, "step": 5116 }, { "epoch": 31.98125, "grad_norm": 1.234375, "learning_rate": 4.836778645865799e-05, "loss": 0.1121, "step": 5117 }, { "epoch": 31.9875, "grad_norm": 0.9921875, "learning_rate": 4.836686812205144e-05, "loss": 0.0917, "step": 5118 }, { "epoch": 31.99375, "grad_norm": 1.5546875, "learning_rate": 4.836594953589748e-05, "loss": 0.1463, "step": 5119 }, { "epoch": 32.0, "grad_norm": 0.9296875, "learning_rate": 4.83650307002059e-05, "loss": 0.0692, "step": 5120 }, { "epoch": 32.00625, "grad_norm": 1.65625, "learning_rate": 4.8364111614986527e-05, "loss": 0.1623, "step": 5121 }, { "epoch": 32.0125, "grad_norm": 1.0859375, "learning_rate": 4.836319228024916e-05, "loss": 0.0992, "step": 5122 }, { "epoch": 32.01875, "grad_norm": 1.3984375, "learning_rate": 4.836227269600364e-05, "loss": 0.1443, "step": 5123 }, { "epoch": 32.025, "grad_norm": 0.90234375, "learning_rate": 4.836135286225978e-05, "loss": 0.0777, "step": 5124 }, { "epoch": 32.03125, "grad_norm": 1.0234375, "learning_rate": 4.836043277902739e-05, "loss": 0.0953, "step": 5125 }, { "epoch": 32.0375, "grad_norm": 1.1171875, "learning_rate": 4.8359512446316306e-05, "loss": 0.0907, "step": 5126 }, { "epoch": 32.04375, "grad_norm": 0.9140625, "learning_rate": 4.835859186413636e-05, "loss": 0.0883, "step": 5127 }, { "epoch": 32.05, "grad_norm": 1.1015625, "learning_rate": 4.835767103249738e-05, "loss": 0.1065, "step": 5128 }, { "epoch": 32.05625, "grad_norm": 0.97265625, "learning_rate": 4.8356749951409204e-05, "loss": 0.0918, "step": 5129 }, { "epoch": 32.0625, "grad_norm": 1.1484375, "learning_rate": 4.835582862088166e-05, "loss": 0.1198, "step": 5130 }, { "epoch": 32.06875, "grad_norm": 1.109375, "learning_rate": 4.835490704092459e-05, "loss": 0.1002, "step": 5131 }, { "epoch": 32.075, "grad_norm": 1.09375, "learning_rate": 4.8353985211547844e-05, "loss": 0.1261, "step": 5132 }, { "epoch": 32.08125, "grad_norm": 1.125, "learning_rate": 4.8353063132761256e-05, "loss": 0.1059, "step": 5133 }, { "epoch": 32.0875, "grad_norm": 1.6328125, "learning_rate": 4.835214080457468e-05, "loss": 0.1884, "step": 5134 }, { "epoch": 32.09375, "grad_norm": 0.95703125, "learning_rate": 4.835121822699796e-05, "loss": 0.0734, "step": 5135 }, { "epoch": 32.1, "grad_norm": 1.6953125, "learning_rate": 4.8350295400040966e-05, "loss": 0.1835, "step": 5136 }, { "epoch": 32.10625, "grad_norm": 1.1796875, "learning_rate": 4.834937232371353e-05, "loss": 0.1121, "step": 5137 }, { "epoch": 32.1125, "grad_norm": 0.72265625, "learning_rate": 4.834844899802552e-05, "loss": 0.0811, "step": 5138 }, { "epoch": 32.11875, "grad_norm": 1.640625, "learning_rate": 4.8347525422986806e-05, "loss": 0.1951, "step": 5139 }, { "epoch": 32.125, "grad_norm": 1.140625, "learning_rate": 4.8346601598607235e-05, "loss": 0.1007, "step": 5140 }, { "epoch": 32.13125, "grad_norm": 1.8984375, "learning_rate": 4.834567752489668e-05, "loss": 0.1975, "step": 5141 }, { "epoch": 32.1375, "grad_norm": 1.8203125, "learning_rate": 4.834475320186501e-05, "loss": 0.148, "step": 5142 }, { "epoch": 32.14375, "grad_norm": 1.015625, "learning_rate": 4.83438286295221e-05, "loss": 0.0803, "step": 5143 }, { "epoch": 32.15, "grad_norm": 1.4765625, "learning_rate": 4.8342903807877815e-05, "loss": 0.1381, "step": 5144 }, { "epoch": 32.15625, "grad_norm": 1.015625, "learning_rate": 4.834197873694205e-05, "loss": 0.089, "step": 5145 }, { "epoch": 32.1625, "grad_norm": 1.296875, "learning_rate": 4.834105341672466e-05, "loss": 0.0781, "step": 5146 }, { "epoch": 32.16875, "grad_norm": 1.25, "learning_rate": 4.834012784723554e-05, "loss": 0.1213, "step": 5147 }, { "epoch": 32.175, "grad_norm": 0.96875, "learning_rate": 4.833920202848458e-05, "loss": 0.0852, "step": 5148 }, { "epoch": 32.18125, "grad_norm": 1.15625, "learning_rate": 4.833827596048165e-05, "loss": 0.1293, "step": 5149 }, { "epoch": 32.1875, "grad_norm": 1.578125, "learning_rate": 4.8337349643236654e-05, "loss": 0.1369, "step": 5150 }, { "epoch": 32.19375, "grad_norm": 1.1015625, "learning_rate": 4.8336423076759484e-05, "loss": 0.0696, "step": 5151 }, { "epoch": 32.2, "grad_norm": 1.0859375, "learning_rate": 4.8335496261060034e-05, "loss": 0.1002, "step": 5152 }, { "epoch": 32.20625, "grad_norm": 0.921875, "learning_rate": 4.83345691961482e-05, "loss": 0.078, "step": 5153 }, { "epoch": 32.2125, "grad_norm": 1.0703125, "learning_rate": 4.833364188203387e-05, "loss": 0.0717, "step": 5154 }, { "epoch": 32.21875, "grad_norm": 1.25, "learning_rate": 4.833271431872698e-05, "loss": 0.1409, "step": 5155 }, { "epoch": 32.225, "grad_norm": 1.8125, "learning_rate": 4.833178650623741e-05, "loss": 0.141, "step": 5156 }, { "epoch": 32.23125, "grad_norm": 1.1015625, "learning_rate": 4.833085844457507e-05, "loss": 0.1373, "step": 5157 }, { "epoch": 32.2375, "grad_norm": 1.2421875, "learning_rate": 4.832993013374988e-05, "loss": 0.1016, "step": 5158 }, { "epoch": 32.24375, "grad_norm": 1.5, "learning_rate": 4.832900157377175e-05, "loss": 0.1107, "step": 5159 }, { "epoch": 32.25, "grad_norm": 1.2421875, "learning_rate": 4.832807276465059e-05, "loss": 0.1231, "step": 5160 }, { "epoch": 32.25625, "grad_norm": 1.375, "learning_rate": 4.832714370639633e-05, "loss": 0.1171, "step": 5161 }, { "epoch": 32.2625, "grad_norm": 0.95703125, "learning_rate": 4.832621439901889e-05, "loss": 0.1058, "step": 5162 }, { "epoch": 32.26875, "grad_norm": 1.8046875, "learning_rate": 4.8325284842528195e-05, "loss": 0.1086, "step": 5163 }, { "epoch": 32.275, "grad_norm": 1.3046875, "learning_rate": 4.8324355036934165e-05, "loss": 0.1238, "step": 5164 }, { "epoch": 32.28125, "grad_norm": 1.6171875, "learning_rate": 4.832342498224673e-05, "loss": 0.1379, "step": 5165 }, { "epoch": 32.2875, "grad_norm": 0.953125, "learning_rate": 4.832249467847584e-05, "loss": 0.0864, "step": 5166 }, { "epoch": 32.29375, "grad_norm": 1.46875, "learning_rate": 4.8321564125631415e-05, "loss": 0.0956, "step": 5167 }, { "epoch": 32.3, "grad_norm": 1.4140625, "learning_rate": 4.8320633323723385e-05, "loss": 0.1476, "step": 5168 }, { "epoch": 32.30625, "grad_norm": 0.84765625, "learning_rate": 4.831970227276171e-05, "loss": 0.0643, "step": 5169 }, { "epoch": 32.3125, "grad_norm": 1.5859375, "learning_rate": 4.831877097275632e-05, "loss": 0.1569, "step": 5170 }, { "epoch": 32.31875, "grad_norm": 1.0234375, "learning_rate": 4.831783942371716e-05, "loss": 0.0903, "step": 5171 }, { "epoch": 32.325, "grad_norm": 1.703125, "learning_rate": 4.8316907625654196e-05, "loss": 0.227, "step": 5172 }, { "epoch": 32.33125, "grad_norm": 1.8671875, "learning_rate": 4.8315975578577355e-05, "loss": 0.15, "step": 5173 }, { "epoch": 32.3375, "grad_norm": 2.109375, "learning_rate": 4.8315043282496606e-05, "loss": 0.1522, "step": 5174 }, { "epoch": 32.34375, "grad_norm": 1.5546875, "learning_rate": 4.83141107374219e-05, "loss": 0.2048, "step": 5175 }, { "epoch": 32.35, "grad_norm": 1.4375, "learning_rate": 4.8313177943363206e-05, "loss": 0.1544, "step": 5176 }, { "epoch": 32.35625, "grad_norm": 1.6015625, "learning_rate": 4.831224490033047e-05, "loss": 0.1288, "step": 5177 }, { "epoch": 32.3625, "grad_norm": 1.0078125, "learning_rate": 4.8311311608333665e-05, "loss": 0.0676, "step": 5178 }, { "epoch": 32.36875, "grad_norm": 1.078125, "learning_rate": 4.831037806738276e-05, "loss": 0.0757, "step": 5179 }, { "epoch": 32.375, "grad_norm": 0.96484375, "learning_rate": 4.8309444277487725e-05, "loss": 0.0683, "step": 5180 }, { "epoch": 32.38125, "grad_norm": 1.4609375, "learning_rate": 4.830851023865852e-05, "loss": 0.1409, "step": 5181 }, { "epoch": 32.3875, "grad_norm": 1.3046875, "learning_rate": 4.8307575950905136e-05, "loss": 0.0943, "step": 5182 }, { "epoch": 32.39375, "grad_norm": 1.4140625, "learning_rate": 4.830664141423755e-05, "loss": 0.1218, "step": 5183 }, { "epoch": 32.4, "grad_norm": 0.7734375, "learning_rate": 4.830570662866572e-05, "loss": 0.0676, "step": 5184 }, { "epoch": 32.40625, "grad_norm": 1.0703125, "learning_rate": 4.830477159419966e-05, "loss": 0.0849, "step": 5185 }, { "epoch": 32.4125, "grad_norm": 1.453125, "learning_rate": 4.830383631084935e-05, "loss": 0.1289, "step": 5186 }, { "epoch": 32.41875, "grad_norm": 1.0390625, "learning_rate": 4.830290077862475e-05, "loss": 0.0805, "step": 5187 }, { "epoch": 32.425, "grad_norm": 1.671875, "learning_rate": 4.830196499753589e-05, "loss": 0.1552, "step": 5188 }, { "epoch": 32.43125, "grad_norm": 1.1484375, "learning_rate": 4.830102896759273e-05, "loss": 0.1306, "step": 5189 }, { "epoch": 32.4375, "grad_norm": 1.4140625, "learning_rate": 4.830009268880529e-05, "loss": 0.1192, "step": 5190 }, { "epoch": 32.44375, "grad_norm": 1.703125, "learning_rate": 4.829915616118356e-05, "loss": 0.1398, "step": 5191 }, { "epoch": 32.45, "grad_norm": 1.109375, "learning_rate": 4.829821938473753e-05, "loss": 0.1127, "step": 5192 }, { "epoch": 32.45625, "grad_norm": 1.28125, "learning_rate": 4.8297282359477225e-05, "loss": 0.1098, "step": 5193 }, { "epoch": 32.4625, "grad_norm": 0.8125, "learning_rate": 4.8296345085412644e-05, "loss": 0.0767, "step": 5194 }, { "epoch": 32.46875, "grad_norm": 2.078125, "learning_rate": 4.82954075625538e-05, "loss": 0.1482, "step": 5195 }, { "epoch": 32.475, "grad_norm": 2.0, "learning_rate": 4.82944697909107e-05, "loss": 0.199, "step": 5196 }, { "epoch": 32.48125, "grad_norm": 1.2265625, "learning_rate": 4.829353177049335e-05, "loss": 0.0829, "step": 5197 }, { "epoch": 32.4875, "grad_norm": 1.9765625, "learning_rate": 4.829259350131179e-05, "loss": 0.1634, "step": 5198 }, { "epoch": 32.49375, "grad_norm": 1.609375, "learning_rate": 4.8291654983376024e-05, "loss": 0.1454, "step": 5199 }, { "epoch": 32.5, "grad_norm": 1.3984375, "learning_rate": 4.829071621669607e-05, "loss": 0.0935, "step": 5200 }, { "epoch": 32.50625, "grad_norm": 1.7265625, "learning_rate": 4.8289777201281974e-05, "loss": 0.1444, "step": 5201 }, { "epoch": 32.5125, "grad_norm": 1.1015625, "learning_rate": 4.8288837937143746e-05, "loss": 0.0947, "step": 5202 }, { "epoch": 32.51875, "grad_norm": 1.3046875, "learning_rate": 4.828789842429143e-05, "loss": 0.1269, "step": 5203 }, { "epoch": 32.525, "grad_norm": 1.234375, "learning_rate": 4.828695866273506e-05, "loss": 0.1333, "step": 5204 }, { "epoch": 32.53125, "grad_norm": 0.6328125, "learning_rate": 4.828601865248465e-05, "loss": 0.0668, "step": 5205 }, { "epoch": 32.5375, "grad_norm": 0.67578125, "learning_rate": 4.828507839355027e-05, "loss": 0.0744, "step": 5206 }, { "epoch": 32.54375, "grad_norm": 1.3125, "learning_rate": 4.8284137885941936e-05, "loss": 0.1248, "step": 5207 }, { "epoch": 32.55, "grad_norm": 1.53125, "learning_rate": 4.8283197129669696e-05, "loss": 0.2068, "step": 5208 }, { "epoch": 32.55625, "grad_norm": 1.015625, "learning_rate": 4.8282256124743616e-05, "loss": 0.0876, "step": 5209 }, { "epoch": 32.5625, "grad_norm": 0.98046875, "learning_rate": 4.828131487117373e-05, "loss": 0.0925, "step": 5210 }, { "epoch": 32.56875, "grad_norm": 1.109375, "learning_rate": 4.828037336897009e-05, "loss": 0.102, "step": 5211 }, { "epoch": 32.575, "grad_norm": 1.0546875, "learning_rate": 4.827943161814276e-05, "loss": 0.0907, "step": 5212 }, { "epoch": 32.58125, "grad_norm": 1.1171875, "learning_rate": 4.8278489618701784e-05, "loss": 0.0954, "step": 5213 }, { "epoch": 32.5875, "grad_norm": 0.73828125, "learning_rate": 4.827754737065723e-05, "loss": 0.0603, "step": 5214 }, { "epoch": 32.59375, "grad_norm": 1.609375, "learning_rate": 4.827660487401916e-05, "loss": 0.1231, "step": 5215 }, { "epoch": 32.6, "grad_norm": 1.15625, "learning_rate": 4.827566212879764e-05, "loss": 0.069, "step": 5216 }, { "epoch": 32.60625, "grad_norm": 1.046875, "learning_rate": 4.827471913500274e-05, "loss": 0.0788, "step": 5217 }, { "epoch": 32.6125, "grad_norm": 1.5859375, "learning_rate": 4.827377589264452e-05, "loss": 0.1282, "step": 5218 }, { "epoch": 32.61875, "grad_norm": 1.015625, "learning_rate": 4.827283240173307e-05, "loss": 0.0753, "step": 5219 }, { "epoch": 32.625, "grad_norm": 1.7890625, "learning_rate": 4.8271888662278454e-05, "loss": 0.1443, "step": 5220 }, { "epoch": 32.63125, "grad_norm": 1.0703125, "learning_rate": 4.8270944674290755e-05, "loss": 0.0739, "step": 5221 }, { "epoch": 32.6375, "grad_norm": 1.4296875, "learning_rate": 4.827000043778006e-05, "loss": 0.0966, "step": 5222 }, { "epoch": 32.64375, "grad_norm": 1.1875, "learning_rate": 4.8269055952756434e-05, "loss": 0.1411, "step": 5223 }, { "epoch": 32.65, "grad_norm": 1.578125, "learning_rate": 4.8268111219229985e-05, "loss": 0.1894, "step": 5224 }, { "epoch": 32.65625, "grad_norm": 1.5546875, "learning_rate": 4.8267166237210796e-05, "loss": 0.1459, "step": 5225 }, { "epoch": 32.6625, "grad_norm": 1.84375, "learning_rate": 4.826622100670895e-05, "loss": 0.1511, "step": 5226 }, { "epoch": 32.66875, "grad_norm": 1.1875, "learning_rate": 4.826527552773455e-05, "loss": 0.076, "step": 5227 }, { "epoch": 32.675, "grad_norm": 0.97265625, "learning_rate": 4.826432980029769e-05, "loss": 0.0952, "step": 5228 }, { "epoch": 32.68125, "grad_norm": 0.96875, "learning_rate": 4.826338382440848e-05, "loss": 0.0792, "step": 5229 }, { "epoch": 32.6875, "grad_norm": 2.09375, "learning_rate": 4.826243760007701e-05, "loss": 0.1761, "step": 5230 }, { "epoch": 32.69375, "grad_norm": 1.890625, "learning_rate": 4.826149112731338e-05, "loss": 0.1311, "step": 5231 }, { "epoch": 32.7, "grad_norm": 1.25, "learning_rate": 4.8260544406127724e-05, "loss": 0.1361, "step": 5232 }, { "epoch": 32.70625, "grad_norm": 0.8203125, "learning_rate": 4.8259597436530125e-05, "loss": 0.0746, "step": 5233 }, { "epoch": 32.7125, "grad_norm": 1.4765625, "learning_rate": 4.825865021853071e-05, "loss": 0.1257, "step": 5234 }, { "epoch": 32.71875, "grad_norm": 1.078125, "learning_rate": 4.825770275213959e-05, "loss": 0.0698, "step": 5235 }, { "epoch": 32.725, "grad_norm": 1.390625, "learning_rate": 4.8256755037366884e-05, "loss": 0.1377, "step": 5236 }, { "epoch": 32.73125, "grad_norm": 1.609375, "learning_rate": 4.825580707422272e-05, "loss": 0.1309, "step": 5237 }, { "epoch": 32.7375, "grad_norm": 0.9609375, "learning_rate": 4.825485886271721e-05, "loss": 0.0816, "step": 5238 }, { "epoch": 32.74375, "grad_norm": 1.234375, "learning_rate": 4.8253910402860494e-05, "loss": 0.1095, "step": 5239 }, { "epoch": 32.75, "grad_norm": 0.765625, "learning_rate": 4.825296169466269e-05, "loss": 0.068, "step": 5240 }, { "epoch": 32.75625, "grad_norm": 1.25, "learning_rate": 4.825201273813393e-05, "loss": 0.0959, "step": 5241 }, { "epoch": 32.7625, "grad_norm": 1.03125, "learning_rate": 4.8251063533284356e-05, "loss": 0.0899, "step": 5242 }, { "epoch": 32.76875, "grad_norm": 0.83984375, "learning_rate": 4.8250114080124105e-05, "loss": 0.0831, "step": 5243 }, { "epoch": 32.775, "grad_norm": 1.421875, "learning_rate": 4.824916437866331e-05, "loss": 0.1079, "step": 5244 }, { "epoch": 32.78125, "grad_norm": 1.3828125, "learning_rate": 4.8248214428912116e-05, "loss": 0.1297, "step": 5245 }, { "epoch": 32.7875, "grad_norm": 0.95703125, "learning_rate": 4.824726423088066e-05, "loss": 0.0879, "step": 5246 }, { "epoch": 32.79375, "grad_norm": 0.8515625, "learning_rate": 4.824631378457911e-05, "loss": 0.0835, "step": 5247 }, { "epoch": 32.8, "grad_norm": 1.1796875, "learning_rate": 4.82453630900176e-05, "loss": 0.0959, "step": 5248 }, { "epoch": 32.80625, "grad_norm": 1.625, "learning_rate": 4.8244412147206284e-05, "loss": 0.1295, "step": 5249 }, { "epoch": 32.8125, "grad_norm": 1.34375, "learning_rate": 4.824346095615533e-05, "loss": 0.1069, "step": 5250 }, { "epoch": 32.81875, "grad_norm": 0.984375, "learning_rate": 4.8242509516874875e-05, "loss": 0.0661, "step": 5251 }, { "epoch": 32.825, "grad_norm": 1.359375, "learning_rate": 4.824155782937511e-05, "loss": 0.1705, "step": 5252 }, { "epoch": 32.83125, "grad_norm": 1.0546875, "learning_rate": 4.8240605893666164e-05, "loss": 0.1292, "step": 5253 }, { "epoch": 32.8375, "grad_norm": 1.390625, "learning_rate": 4.8239653709758224e-05, "loss": 0.1132, "step": 5254 }, { "epoch": 32.84375, "grad_norm": 2.046875, "learning_rate": 4.823870127766146e-05, "loss": 0.2111, "step": 5255 }, { "epoch": 32.85, "grad_norm": 0.81640625, "learning_rate": 4.823774859738603e-05, "loss": 0.0735, "step": 5256 }, { "epoch": 32.85625, "grad_norm": 1.1640625, "learning_rate": 4.823679566894213e-05, "loss": 0.1096, "step": 5257 }, { "epoch": 32.8625, "grad_norm": 1.046875, "learning_rate": 4.8235842492339904e-05, "loss": 0.0801, "step": 5258 }, { "epoch": 32.86875, "grad_norm": 0.796875, "learning_rate": 4.823488906758956e-05, "loss": 0.0631, "step": 5259 }, { "epoch": 32.875, "grad_norm": 0.98828125, "learning_rate": 4.823393539470128e-05, "loss": 0.0829, "step": 5260 }, { "epoch": 32.88125, "grad_norm": 1.6015625, "learning_rate": 4.823298147368523e-05, "loss": 0.1419, "step": 5261 }, { "epoch": 32.8875, "grad_norm": 1.2734375, "learning_rate": 4.8232027304551606e-05, "loss": 0.1513, "step": 5262 }, { "epoch": 32.89375, "grad_norm": 1.296875, "learning_rate": 4.82310728873106e-05, "loss": 0.1141, "step": 5263 }, { "epoch": 32.9, "grad_norm": 1.0625, "learning_rate": 4.823011822197241e-05, "loss": 0.0893, "step": 5264 }, { "epoch": 32.90625, "grad_norm": 1.015625, "learning_rate": 4.822916330854722e-05, "loss": 0.0734, "step": 5265 }, { "epoch": 32.9125, "grad_norm": 1.046875, "learning_rate": 4.822820814704524e-05, "loss": 0.0804, "step": 5266 }, { "epoch": 32.91875, "grad_norm": 1.65625, "learning_rate": 4.8227252737476656e-05, "loss": 0.1485, "step": 5267 }, { "epoch": 32.925, "grad_norm": 0.84765625, "learning_rate": 4.822629707985169e-05, "loss": 0.0697, "step": 5268 }, { "epoch": 32.93125, "grad_norm": 1.0546875, "learning_rate": 4.822534117418053e-05, "loss": 0.1017, "step": 5269 }, { "epoch": 32.9375, "grad_norm": 1.359375, "learning_rate": 4.8224385020473394e-05, "loss": 0.1037, "step": 5270 }, { "epoch": 32.94375, "grad_norm": 0.953125, "learning_rate": 4.82234286187405e-05, "loss": 0.078, "step": 5271 }, { "epoch": 32.95, "grad_norm": 1.0, "learning_rate": 4.822247196899203e-05, "loss": 0.0828, "step": 5272 }, { "epoch": 32.95625, "grad_norm": 1.265625, "learning_rate": 4.822151507123824e-05, "loss": 0.1061, "step": 5273 }, { "epoch": 32.9625, "grad_norm": 0.8125, "learning_rate": 4.822055792548934e-05, "loss": 0.0785, "step": 5274 }, { "epoch": 32.96875, "grad_norm": 0.95703125, "learning_rate": 4.821960053175553e-05, "loss": 0.0823, "step": 5275 }, { "epoch": 32.975, "grad_norm": 1.2578125, "learning_rate": 4.8218642890047063e-05, "loss": 0.1092, "step": 5276 }, { "epoch": 32.98125, "grad_norm": 1.078125, "learning_rate": 4.821768500037415e-05, "loss": 0.0677, "step": 5277 }, { "epoch": 32.9875, "grad_norm": 0.94921875, "learning_rate": 4.8216726862747014e-05, "loss": 0.0747, "step": 5278 }, { "epoch": 32.99375, "grad_norm": 1.0625, "learning_rate": 4.82157684771759e-05, "loss": 0.1009, "step": 5279 }, { "epoch": 33.0, "grad_norm": 1.2734375, "learning_rate": 4.821480984367104e-05, "loss": 0.0901, "step": 5280 }, { "epoch": 33.00625, "grad_norm": 1.1640625, "learning_rate": 4.821385096224268e-05, "loss": 0.1015, "step": 5281 }, { "epoch": 33.0125, "grad_norm": 1.5390625, "learning_rate": 4.8212891832901045e-05, "loss": 0.1238, "step": 5282 }, { "epoch": 33.01875, "grad_norm": 1.03125, "learning_rate": 4.8211932455656386e-05, "loss": 0.0762, "step": 5283 }, { "epoch": 33.025, "grad_norm": 0.9609375, "learning_rate": 4.8210972830518956e-05, "loss": 0.0743, "step": 5284 }, { "epoch": 33.03125, "grad_norm": 1.0625, "learning_rate": 4.821001295749899e-05, "loss": 0.0932, "step": 5285 }, { "epoch": 33.0375, "grad_norm": 1.234375, "learning_rate": 4.820905283660674e-05, "loss": 0.1251, "step": 5286 }, { "epoch": 33.04375, "grad_norm": 1.4375, "learning_rate": 4.820809246785247e-05, "loss": 0.1289, "step": 5287 }, { "epoch": 33.05, "grad_norm": 1.390625, "learning_rate": 4.820713185124643e-05, "loss": 0.1148, "step": 5288 }, { "epoch": 33.05625, "grad_norm": 1.5, "learning_rate": 4.8206170986798876e-05, "loss": 0.179, "step": 5289 }, { "epoch": 33.0625, "grad_norm": 1.28125, "learning_rate": 4.8205209874520076e-05, "loss": 0.1132, "step": 5290 }, { "epoch": 33.06875, "grad_norm": 1.9296875, "learning_rate": 4.820424851442029e-05, "loss": 0.206, "step": 5291 }, { "epoch": 33.075, "grad_norm": 1.421875, "learning_rate": 4.820328690650978e-05, "loss": 0.1181, "step": 5292 }, { "epoch": 33.08125, "grad_norm": 1.2265625, "learning_rate": 4.8202325050798835e-05, "loss": 0.1089, "step": 5293 }, { "epoch": 33.0875, "grad_norm": 1.34375, "learning_rate": 4.82013629472977e-05, "loss": 0.1002, "step": 5294 }, { "epoch": 33.09375, "grad_norm": 1.3984375, "learning_rate": 4.820040059601668e-05, "loss": 0.1232, "step": 5295 }, { "epoch": 33.1, "grad_norm": 0.90625, "learning_rate": 4.8199437996966025e-05, "loss": 0.0858, "step": 5296 }, { "epoch": 33.10625, "grad_norm": 1.2734375, "learning_rate": 4.819847515015602e-05, "loss": 0.1055, "step": 5297 }, { "epoch": 33.1125, "grad_norm": 1.9140625, "learning_rate": 4.8197512055596966e-05, "loss": 0.187, "step": 5298 }, { "epoch": 33.11875, "grad_norm": 1.421875, "learning_rate": 4.8196548713299126e-05, "loss": 0.1722, "step": 5299 }, { "epoch": 33.125, "grad_norm": 0.91796875, "learning_rate": 4.819558512327281e-05, "loss": 0.0704, "step": 5300 }, { "epoch": 33.13125, "grad_norm": 1.578125, "learning_rate": 4.819462128552829e-05, "loss": 0.2511, "step": 5301 }, { "epoch": 33.1375, "grad_norm": 1.390625, "learning_rate": 4.819365720007587e-05, "loss": 0.1533, "step": 5302 }, { "epoch": 33.14375, "grad_norm": 1.1328125, "learning_rate": 4.8192692866925836e-05, "loss": 0.1109, "step": 5303 }, { "epoch": 33.15, "grad_norm": 1.640625, "learning_rate": 4.81917282860885e-05, "loss": 0.182, "step": 5304 }, { "epoch": 33.15625, "grad_norm": 0.8671875, "learning_rate": 4.819076345757415e-05, "loss": 0.0689, "step": 5305 }, { "epoch": 33.1625, "grad_norm": 1.96875, "learning_rate": 4.8189798381393104e-05, "loss": 0.2066, "step": 5306 }, { "epoch": 33.16875, "grad_norm": 1.34375, "learning_rate": 4.818883305755566e-05, "loss": 0.1793, "step": 5307 }, { "epoch": 33.175, "grad_norm": 0.86328125, "learning_rate": 4.818786748607212e-05, "loss": 0.0693, "step": 5308 }, { "epoch": 33.18125, "grad_norm": 1.15625, "learning_rate": 4.818690166695281e-05, "loss": 0.0871, "step": 5309 }, { "epoch": 33.1875, "grad_norm": 1.609375, "learning_rate": 4.8185935600208046e-05, "loss": 0.1757, "step": 5310 }, { "epoch": 33.19375, "grad_norm": 0.75, "learning_rate": 4.818496928584812e-05, "loss": 0.0581, "step": 5311 }, { "epoch": 33.2, "grad_norm": 1.4609375, "learning_rate": 4.818400272388338e-05, "loss": 0.1342, "step": 5312 }, { "epoch": 33.20625, "grad_norm": 1.1796875, "learning_rate": 4.8183035914324136e-05, "loss": 0.1204, "step": 5313 }, { "epoch": 33.2125, "grad_norm": 1.1953125, "learning_rate": 4.818206885718072e-05, "loss": 0.1074, "step": 5314 }, { "epoch": 33.21875, "grad_norm": 0.99609375, "learning_rate": 4.8181101552463445e-05, "loss": 0.0901, "step": 5315 }, { "epoch": 33.225, "grad_norm": 1.609375, "learning_rate": 4.818013400018266e-05, "loss": 0.1645, "step": 5316 }, { "epoch": 33.23125, "grad_norm": 0.9921875, "learning_rate": 4.817916620034868e-05, "loss": 0.0915, "step": 5317 }, { "epoch": 33.2375, "grad_norm": 1.8125, "learning_rate": 4.817819815297185e-05, "loss": 0.1888, "step": 5318 }, { "epoch": 33.24375, "grad_norm": 1.359375, "learning_rate": 4.8177229858062525e-05, "loss": 0.1168, "step": 5319 }, { "epoch": 33.25, "grad_norm": 1.421875, "learning_rate": 4.8176261315631014e-05, "loss": 0.1126, "step": 5320 }, { "epoch": 33.25625, "grad_norm": 1.6796875, "learning_rate": 4.817529252568768e-05, "loss": 0.1568, "step": 5321 }, { "epoch": 33.2625, "grad_norm": 1.4765625, "learning_rate": 4.8174323488242856e-05, "loss": 0.1762, "step": 5322 }, { "epoch": 33.26875, "grad_norm": 1.828125, "learning_rate": 4.817335420330691e-05, "loss": 0.1426, "step": 5323 }, { "epoch": 33.275, "grad_norm": 0.58984375, "learning_rate": 4.817238467089018e-05, "loss": 0.0542, "step": 5324 }, { "epoch": 33.28125, "grad_norm": 1.3828125, "learning_rate": 4.817141489100302e-05, "loss": 0.1056, "step": 5325 }, { "epoch": 33.2875, "grad_norm": 1.7265625, "learning_rate": 4.8170444863655795e-05, "loss": 0.2003, "step": 5326 }, { "epoch": 33.29375, "grad_norm": 1.0703125, "learning_rate": 4.816947458885886e-05, "loss": 0.104, "step": 5327 }, { "epoch": 33.3, "grad_norm": 1.359375, "learning_rate": 4.816850406662258e-05, "loss": 0.1571, "step": 5328 }, { "epoch": 33.30625, "grad_norm": 0.9296875, "learning_rate": 4.81675332969573e-05, "loss": 0.079, "step": 5329 }, { "epoch": 33.3125, "grad_norm": 1.578125, "learning_rate": 4.816656227987342e-05, "loss": 0.0996, "step": 5330 }, { "epoch": 33.31875, "grad_norm": 1.640625, "learning_rate": 4.8165591015381285e-05, "loss": 0.1166, "step": 5331 }, { "epoch": 33.325, "grad_norm": 0.9375, "learning_rate": 4.816461950349128e-05, "loss": 0.0888, "step": 5332 }, { "epoch": 33.33125, "grad_norm": 1.484375, "learning_rate": 4.816364774421378e-05, "loss": 0.1514, "step": 5333 }, { "epoch": 33.3375, "grad_norm": 1.5390625, "learning_rate": 4.8162675737559154e-05, "loss": 0.1288, "step": 5334 }, { "epoch": 33.34375, "grad_norm": 1.3203125, "learning_rate": 4.816170348353779e-05, "loss": 0.1389, "step": 5335 }, { "epoch": 33.35, "grad_norm": 1.5546875, "learning_rate": 4.8160730982160076e-05, "loss": 0.1038, "step": 5336 }, { "epoch": 33.35625, "grad_norm": 1.1796875, "learning_rate": 4.815975823343638e-05, "loss": 0.0783, "step": 5337 }, { "epoch": 33.3625, "grad_norm": 1.5234375, "learning_rate": 4.815878523737711e-05, "loss": 0.1168, "step": 5338 }, { "epoch": 33.36875, "grad_norm": 0.76171875, "learning_rate": 4.815781199399264e-05, "loss": 0.067, "step": 5339 }, { "epoch": 33.375, "grad_norm": 1.0546875, "learning_rate": 4.815683850329338e-05, "loss": 0.0888, "step": 5340 }, { "epoch": 33.38125, "grad_norm": 1.1796875, "learning_rate": 4.815586476528972e-05, "loss": 0.09, "step": 5341 }, { "epoch": 33.3875, "grad_norm": 1.484375, "learning_rate": 4.815489077999206e-05, "loss": 0.1377, "step": 5342 }, { "epoch": 33.39375, "grad_norm": 1.2890625, "learning_rate": 4.8153916547410794e-05, "loss": 0.1051, "step": 5343 }, { "epoch": 33.4, "grad_norm": 0.8359375, "learning_rate": 4.815294206755633e-05, "loss": 0.0685, "step": 5344 }, { "epoch": 33.40625, "grad_norm": 1.0625, "learning_rate": 4.815196734043909e-05, "loss": 0.0737, "step": 5345 }, { "epoch": 33.4125, "grad_norm": 1.078125, "learning_rate": 4.815099236606946e-05, "loss": 0.0841, "step": 5346 }, { "epoch": 33.41875, "grad_norm": 1.2890625, "learning_rate": 4.815001714445786e-05, "loss": 0.1049, "step": 5347 }, { "epoch": 33.425, "grad_norm": 0.54296875, "learning_rate": 4.814904167561472e-05, "loss": 0.0608, "step": 5348 }, { "epoch": 33.43125, "grad_norm": 1.75, "learning_rate": 4.8148065959550444e-05, "loss": 0.3116, "step": 5349 }, { "epoch": 33.4375, "grad_norm": 1.15625, "learning_rate": 4.814708999627545e-05, "loss": 0.0909, "step": 5350 }, { "epoch": 33.44375, "grad_norm": 1.0703125, "learning_rate": 4.814611378580016e-05, "loss": 0.1329, "step": 5351 }, { "epoch": 33.45, "grad_norm": 1.1875, "learning_rate": 4.8145137328135016e-05, "loss": 0.1392, "step": 5352 }, { "epoch": 33.45625, "grad_norm": 1.09375, "learning_rate": 4.8144160623290424e-05, "loss": 0.0771, "step": 5353 }, { "epoch": 33.4625, "grad_norm": 1.0390625, "learning_rate": 4.814318367127683e-05, "loss": 0.0764, "step": 5354 }, { "epoch": 33.46875, "grad_norm": 1.2578125, "learning_rate": 4.8142206472104666e-05, "loss": 0.1421, "step": 5355 }, { "epoch": 33.475, "grad_norm": 1.40625, "learning_rate": 4.814122902578436e-05, "loss": 0.0993, "step": 5356 }, { "epoch": 33.48125, "grad_norm": 1.21875, "learning_rate": 4.814025133232635e-05, "loss": 0.1115, "step": 5357 }, { "epoch": 33.4875, "grad_norm": 0.984375, "learning_rate": 4.813927339174109e-05, "loss": 0.0816, "step": 5358 }, { "epoch": 33.49375, "grad_norm": 1.7109375, "learning_rate": 4.813829520403902e-05, "loss": 0.179, "step": 5359 }, { "epoch": 33.5, "grad_norm": 1.1875, "learning_rate": 4.813731676923058e-05, "loss": 0.1036, "step": 5360 }, { "epoch": 33.50625, "grad_norm": 2.03125, "learning_rate": 4.8136338087326216e-05, "loss": 0.1588, "step": 5361 }, { "epoch": 33.5125, "grad_norm": 1.21875, "learning_rate": 4.813535915833639e-05, "loss": 0.1085, "step": 5362 }, { "epoch": 33.51875, "grad_norm": 0.671875, "learning_rate": 4.8134379982271556e-05, "loss": 0.0587, "step": 5363 }, { "epoch": 33.525, "grad_norm": 1.0390625, "learning_rate": 4.813340055914216e-05, "loss": 0.0776, "step": 5364 }, { "epoch": 33.53125, "grad_norm": 1.0234375, "learning_rate": 4.813242088895868e-05, "loss": 0.0812, "step": 5365 }, { "epoch": 33.5375, "grad_norm": 0.85546875, "learning_rate": 4.813144097173156e-05, "loss": 0.0697, "step": 5366 }, { "epoch": 33.54375, "grad_norm": 1.078125, "learning_rate": 4.813046080747128e-05, "loss": 0.0931, "step": 5367 }, { "epoch": 33.55, "grad_norm": 1.3671875, "learning_rate": 4.8129480396188296e-05, "loss": 0.1467, "step": 5368 }, { "epoch": 33.55625, "grad_norm": 0.87109375, "learning_rate": 4.8128499737893086e-05, "loss": 0.0868, "step": 5369 }, { "epoch": 33.5625, "grad_norm": 1.0703125, "learning_rate": 4.8127518832596116e-05, "loss": 0.084, "step": 5370 }, { "epoch": 33.56875, "grad_norm": 1.3203125, "learning_rate": 4.812653768030787e-05, "loss": 0.1169, "step": 5371 }, { "epoch": 33.575, "grad_norm": 1.71875, "learning_rate": 4.812555628103882e-05, "loss": 0.1327, "step": 5372 }, { "epoch": 33.58125, "grad_norm": 1.5390625, "learning_rate": 4.812457463479945e-05, "loss": 0.1524, "step": 5373 }, { "epoch": 33.5875, "grad_norm": 1.4140625, "learning_rate": 4.812359274160024e-05, "loss": 0.1452, "step": 5374 }, { "epoch": 33.59375, "grad_norm": 1.25, "learning_rate": 4.812261060145169e-05, "loss": 0.1036, "step": 5375 }, { "epoch": 33.6, "grad_norm": 1.3984375, "learning_rate": 4.8121628214364266e-05, "loss": 0.1859, "step": 5376 }, { "epoch": 33.60625, "grad_norm": 1.2578125, "learning_rate": 4.812064558034847e-05, "loss": 0.1367, "step": 5377 }, { "epoch": 33.6125, "grad_norm": 1.015625, "learning_rate": 4.8119662699414805e-05, "loss": 0.0818, "step": 5378 }, { "epoch": 33.61875, "grad_norm": 1.2578125, "learning_rate": 4.811867957157375e-05, "loss": 0.1036, "step": 5379 }, { "epoch": 33.625, "grad_norm": 0.84765625, "learning_rate": 4.8117696196835824e-05, "loss": 0.0713, "step": 5380 }, { "epoch": 33.63125, "grad_norm": 1.1328125, "learning_rate": 4.8116712575211515e-05, "loss": 0.1087, "step": 5381 }, { "epoch": 33.6375, "grad_norm": 1.5234375, "learning_rate": 4.8115728706711335e-05, "loss": 0.1432, "step": 5382 }, { "epoch": 33.64375, "grad_norm": 1.2890625, "learning_rate": 4.811474459134578e-05, "loss": 0.1129, "step": 5383 }, { "epoch": 33.65, "grad_norm": 1.28125, "learning_rate": 4.811376022912537e-05, "loss": 0.1046, "step": 5384 }, { "epoch": 33.65625, "grad_norm": 1.296875, "learning_rate": 4.8112775620060626e-05, "loss": 0.1488, "step": 5385 }, { "epoch": 33.6625, "grad_norm": 1.6796875, "learning_rate": 4.811179076416204e-05, "loss": 0.1373, "step": 5386 }, { "epoch": 33.66875, "grad_norm": 1.453125, "learning_rate": 4.811080566144015e-05, "loss": 0.1406, "step": 5387 }, { "epoch": 33.675, "grad_norm": 1.125, "learning_rate": 4.810982031190546e-05, "loss": 0.1175, "step": 5388 }, { "epoch": 33.68125, "grad_norm": 1.546875, "learning_rate": 4.810883471556851e-05, "loss": 0.1556, "step": 5389 }, { "epoch": 33.6875, "grad_norm": 1.546875, "learning_rate": 4.8107848872439814e-05, "loss": 0.165, "step": 5390 }, { "epoch": 33.69375, "grad_norm": 0.83203125, "learning_rate": 4.8106862782529906e-05, "loss": 0.0716, "step": 5391 }, { "epoch": 33.7, "grad_norm": 0.98046875, "learning_rate": 4.810587644584932e-05, "loss": 0.0918, "step": 5392 }, { "epoch": 33.70625, "grad_norm": 1.0625, "learning_rate": 4.810488986240858e-05, "loss": 0.0927, "step": 5393 }, { "epoch": 33.7125, "grad_norm": 1.296875, "learning_rate": 4.810390303221822e-05, "loss": 0.1727, "step": 5394 }, { "epoch": 33.71875, "grad_norm": 1.46875, "learning_rate": 4.810291595528879e-05, "loss": 0.1071, "step": 5395 }, { "epoch": 33.725, "grad_norm": 1.3203125, "learning_rate": 4.810192863163083e-05, "loss": 0.1394, "step": 5396 }, { "epoch": 33.73125, "grad_norm": 1.15625, "learning_rate": 4.810094106125488e-05, "loss": 0.1234, "step": 5397 }, { "epoch": 33.7375, "grad_norm": 1.4453125, "learning_rate": 4.809995324417149e-05, "loss": 0.1597, "step": 5398 }, { "epoch": 33.74375, "grad_norm": 1.65625, "learning_rate": 4.809896518039121e-05, "loss": 0.1294, "step": 5399 }, { "epoch": 33.75, "grad_norm": 1.078125, "learning_rate": 4.8097976869924575e-05, "loss": 0.0921, "step": 5400 }, { "epoch": 33.75625, "grad_norm": 0.953125, "learning_rate": 4.8096988312782174e-05, "loss": 0.0767, "step": 5401 }, { "epoch": 33.7625, "grad_norm": 1.4375, "learning_rate": 4.8095999508974535e-05, "loss": 0.1133, "step": 5402 }, { "epoch": 33.76875, "grad_norm": 0.63671875, "learning_rate": 4.809501045851222e-05, "loss": 0.0604, "step": 5403 }, { "epoch": 33.775, "grad_norm": 1.2421875, "learning_rate": 4.8094021161405816e-05, "loss": 0.1168, "step": 5404 }, { "epoch": 33.78125, "grad_norm": 1.546875, "learning_rate": 4.8093031617665864e-05, "loss": 0.1342, "step": 5405 }, { "epoch": 33.7875, "grad_norm": 1.5078125, "learning_rate": 4.809204182730293e-05, "loss": 0.1891, "step": 5406 }, { "epoch": 33.79375, "grad_norm": 0.9765625, "learning_rate": 4.809105179032761e-05, "loss": 0.1164, "step": 5407 }, { "epoch": 33.8, "grad_norm": 1.75, "learning_rate": 4.809006150675046e-05, "loss": 0.1334, "step": 5408 }, { "epoch": 33.80625, "grad_norm": 1.4765625, "learning_rate": 4.8089070976582054e-05, "loss": 0.1213, "step": 5409 }, { "epoch": 33.8125, "grad_norm": 1.0078125, "learning_rate": 4.8088080199832965e-05, "loss": 0.0799, "step": 5410 }, { "epoch": 33.81875, "grad_norm": 1.328125, "learning_rate": 4.808708917651379e-05, "loss": 0.1204, "step": 5411 }, { "epoch": 33.825, "grad_norm": 1.1875, "learning_rate": 4.8086097906635104e-05, "loss": 0.0844, "step": 5412 }, { "epoch": 33.83125, "grad_norm": 1.2109375, "learning_rate": 4.8085106390207494e-05, "loss": 0.1473, "step": 5413 }, { "epoch": 33.8375, "grad_norm": 1.1875, "learning_rate": 4.8084114627241546e-05, "loss": 0.0926, "step": 5414 }, { "epoch": 33.84375, "grad_norm": 1.0234375, "learning_rate": 4.808312261774786e-05, "loss": 0.1293, "step": 5415 }, { "epoch": 33.85, "grad_norm": 1.578125, "learning_rate": 4.808213036173702e-05, "loss": 0.1629, "step": 5416 }, { "epoch": 33.85625, "grad_norm": 1.5390625, "learning_rate": 4.808113785921963e-05, "loss": 0.1554, "step": 5417 }, { "epoch": 33.8625, "grad_norm": 1.1953125, "learning_rate": 4.808014511020629e-05, "loss": 0.1398, "step": 5418 }, { "epoch": 33.86875, "grad_norm": 0.63671875, "learning_rate": 4.8079152114707594e-05, "loss": 0.0593, "step": 5419 }, { "epoch": 33.875, "grad_norm": 0.96875, "learning_rate": 4.807815887273416e-05, "loss": 0.0713, "step": 5420 }, { "epoch": 33.88125, "grad_norm": 1.3515625, "learning_rate": 4.807716538429658e-05, "loss": 0.1253, "step": 5421 }, { "epoch": 33.8875, "grad_norm": 1.59375, "learning_rate": 4.8076171649405476e-05, "loss": 0.1519, "step": 5422 }, { "epoch": 33.89375, "grad_norm": 1.1484375, "learning_rate": 4.8075177668071454e-05, "loss": 0.0812, "step": 5423 }, { "epoch": 33.9, "grad_norm": 1.125, "learning_rate": 4.807418344030513e-05, "loss": 0.1053, "step": 5424 }, { "epoch": 33.90625, "grad_norm": 1.3515625, "learning_rate": 4.8073188966117126e-05, "loss": 0.142, "step": 5425 }, { "epoch": 33.9125, "grad_norm": 1.09375, "learning_rate": 4.8072194245518064e-05, "loss": 0.0851, "step": 5426 }, { "epoch": 33.91875, "grad_norm": 1.109375, "learning_rate": 4.8071199278518553e-05, "loss": 0.0906, "step": 5427 }, { "epoch": 33.925, "grad_norm": 1.0546875, "learning_rate": 4.8070204065129234e-05, "loss": 0.0824, "step": 5428 }, { "epoch": 33.93125, "grad_norm": 1.375, "learning_rate": 4.806920860536073e-05, "loss": 0.1138, "step": 5429 }, { "epoch": 33.9375, "grad_norm": 0.96484375, "learning_rate": 4.8068212899223676e-05, "loss": 0.0815, "step": 5430 }, { "epoch": 33.94375, "grad_norm": 1.1875, "learning_rate": 4.80672169467287e-05, "loss": 0.1185, "step": 5431 }, { "epoch": 33.95, "grad_norm": 1.359375, "learning_rate": 4.806622074788644e-05, "loss": 0.1143, "step": 5432 }, { "epoch": 33.95625, "grad_norm": 1.4609375, "learning_rate": 4.806522430270753e-05, "loss": 0.1075, "step": 5433 }, { "epoch": 33.9625, "grad_norm": 1.3671875, "learning_rate": 4.806422761120263e-05, "loss": 0.1431, "step": 5434 }, { "epoch": 33.96875, "grad_norm": 1.21875, "learning_rate": 4.8063230673382356e-05, "loss": 0.124, "step": 5435 }, { "epoch": 33.975, "grad_norm": 1.625, "learning_rate": 4.806223348925738e-05, "loss": 0.2171, "step": 5436 }, { "epoch": 33.98125, "grad_norm": 1.46875, "learning_rate": 4.8061236058838336e-05, "loss": 0.1327, "step": 5437 }, { "epoch": 33.9875, "grad_norm": 1.0078125, "learning_rate": 4.806023838213588e-05, "loss": 0.0784, "step": 5438 }, { "epoch": 33.99375, "grad_norm": 1.2578125, "learning_rate": 4.805924045916067e-05, "loss": 0.0928, "step": 5439 }, { "epoch": 34.0, "grad_norm": 1.3359375, "learning_rate": 4.805824228992337e-05, "loss": 0.107, "step": 5440 }, { "epoch": 34.00625, "grad_norm": 0.953125, "learning_rate": 4.805724387443462e-05, "loss": 0.082, "step": 5441 }, { "epoch": 34.0125, "grad_norm": 1.078125, "learning_rate": 4.8056245212705097e-05, "loss": 0.1252, "step": 5442 }, { "epoch": 34.01875, "grad_norm": 1.921875, "learning_rate": 4.8055246304745466e-05, "loss": 0.1925, "step": 5443 }, { "epoch": 34.025, "grad_norm": 1.390625, "learning_rate": 4.805424715056639e-05, "loss": 0.1158, "step": 5444 }, { "epoch": 34.03125, "grad_norm": 1.0234375, "learning_rate": 4.805324775017854e-05, "loss": 0.0847, "step": 5445 }, { "epoch": 34.0375, "grad_norm": 1.640625, "learning_rate": 4.80522481035926e-05, "loss": 0.189, "step": 5446 }, { "epoch": 34.04375, "grad_norm": 1.5390625, "learning_rate": 4.8051248210819226e-05, "loss": 0.1396, "step": 5447 }, { "epoch": 34.05, "grad_norm": 1.90625, "learning_rate": 4.805024807186911e-05, "loss": 0.1825, "step": 5448 }, { "epoch": 34.05625, "grad_norm": 0.99609375, "learning_rate": 4.804924768675294e-05, "loss": 0.0727, "step": 5449 }, { "epoch": 34.0625, "grad_norm": 1.5625, "learning_rate": 4.804824705548138e-05, "loss": 0.0904, "step": 5450 }, { "epoch": 34.06875, "grad_norm": 1.015625, "learning_rate": 4.8047246178065125e-05, "loss": 0.0857, "step": 5451 }, { "epoch": 34.075, "grad_norm": 1.21875, "learning_rate": 4.804624505451487e-05, "loss": 0.1261, "step": 5452 }, { "epoch": 34.08125, "grad_norm": 1.484375, "learning_rate": 4.804524368484129e-05, "loss": 0.1176, "step": 5453 }, { "epoch": 34.0875, "grad_norm": 1.1953125, "learning_rate": 4.8044242069055104e-05, "loss": 0.0882, "step": 5454 }, { "epoch": 34.09375, "grad_norm": 1.3828125, "learning_rate": 4.8043240207167e-05, "loss": 0.1009, "step": 5455 }, { "epoch": 34.1, "grad_norm": 0.859375, "learning_rate": 4.8042238099187656e-05, "loss": 0.0816, "step": 5456 }, { "epoch": 34.10625, "grad_norm": 0.93359375, "learning_rate": 4.80412357451278e-05, "loss": 0.0827, "step": 5457 }, { "epoch": 34.1125, "grad_norm": 1.2265625, "learning_rate": 4.804023314499813e-05, "loss": 0.0994, "step": 5458 }, { "epoch": 34.11875, "grad_norm": 1.890625, "learning_rate": 4.8039230298809345e-05, "loss": 0.1421, "step": 5459 }, { "epoch": 34.125, "grad_norm": 1.8984375, "learning_rate": 4.803822720657216e-05, "loss": 0.1619, "step": 5460 }, { "epoch": 34.13125, "grad_norm": 1.1640625, "learning_rate": 4.8037223868297294e-05, "loss": 0.0799, "step": 5461 }, { "epoch": 34.1375, "grad_norm": 1.3046875, "learning_rate": 4.803622028399545e-05, "loss": 0.1952, "step": 5462 }, { "epoch": 34.14375, "grad_norm": 1.4296875, "learning_rate": 4.803521645367736e-05, "loss": 0.1634, "step": 5463 }, { "epoch": 34.15, "grad_norm": 1.09375, "learning_rate": 4.8034212377353735e-05, "loss": 0.129, "step": 5464 }, { "epoch": 34.15625, "grad_norm": 1.0859375, "learning_rate": 4.8033208055035296e-05, "loss": 0.0905, "step": 5465 }, { "epoch": 34.1625, "grad_norm": 1.171875, "learning_rate": 4.803220348673278e-05, "loss": 0.0817, "step": 5466 }, { "epoch": 34.16875, "grad_norm": 1.2265625, "learning_rate": 4.80311986724569e-05, "loss": 0.102, "step": 5467 }, { "epoch": 34.175, "grad_norm": 1.390625, "learning_rate": 4.80301936122184e-05, "loss": 0.1409, "step": 5468 }, { "epoch": 34.18125, "grad_norm": 1.2265625, "learning_rate": 4.802918830602801e-05, "loss": 0.0855, "step": 5469 }, { "epoch": 34.1875, "grad_norm": 1.9296875, "learning_rate": 4.8028182753896456e-05, "loss": 0.2481, "step": 5470 }, { "epoch": 34.19375, "grad_norm": 1.3359375, "learning_rate": 4.8027176955834486e-05, "loss": 0.115, "step": 5471 }, { "epoch": 34.2, "grad_norm": 1.15625, "learning_rate": 4.802617091185285e-05, "loss": 0.0901, "step": 5472 }, { "epoch": 34.20625, "grad_norm": 0.8203125, "learning_rate": 4.8025164621962284e-05, "loss": 0.0705, "step": 5473 }, { "epoch": 34.2125, "grad_norm": 0.78125, "learning_rate": 4.802415808617352e-05, "loss": 0.0787, "step": 5474 }, { "epoch": 34.21875, "grad_norm": 2.09375, "learning_rate": 4.802315130449733e-05, "loss": 0.1649, "step": 5475 }, { "epoch": 34.225, "grad_norm": 0.83203125, "learning_rate": 4.802214427694446e-05, "loss": 0.0613, "step": 5476 }, { "epoch": 34.23125, "grad_norm": 0.97265625, "learning_rate": 4.8021137003525664e-05, "loss": 0.0757, "step": 5477 }, { "epoch": 34.2375, "grad_norm": 1.28125, "learning_rate": 4.802012948425169e-05, "loss": 0.1116, "step": 5478 }, { "epoch": 34.24375, "grad_norm": 0.82421875, "learning_rate": 4.80191217191333e-05, "loss": 0.0704, "step": 5479 }, { "epoch": 34.25, "grad_norm": 1.2578125, "learning_rate": 4.8018113708181265e-05, "loss": 0.1033, "step": 5480 }, { "epoch": 34.25625, "grad_norm": 1.4296875, "learning_rate": 4.801710545140635e-05, "loss": 0.0951, "step": 5481 }, { "epoch": 34.2625, "grad_norm": 1.28125, "learning_rate": 4.801609694881932e-05, "loss": 0.101, "step": 5482 }, { "epoch": 34.26875, "grad_norm": 0.9140625, "learning_rate": 4.801508820043095e-05, "loss": 0.0614, "step": 5483 }, { "epoch": 34.275, "grad_norm": 0.875, "learning_rate": 4.8014079206251995e-05, "loss": 0.1095, "step": 5484 }, { "epoch": 34.28125, "grad_norm": 1.0546875, "learning_rate": 4.8013069966293246e-05, "loss": 0.0893, "step": 5485 }, { "epoch": 34.2875, "grad_norm": 1.9296875, "learning_rate": 4.801206048056548e-05, "loss": 0.1388, "step": 5486 }, { "epoch": 34.29375, "grad_norm": 1.234375, "learning_rate": 4.801105074907948e-05, "loss": 0.0811, "step": 5487 }, { "epoch": 34.3, "grad_norm": 1.5625, "learning_rate": 4.801004077184602e-05, "loss": 0.1224, "step": 5488 }, { "epoch": 34.30625, "grad_norm": 1.28125, "learning_rate": 4.8009030548875896e-05, "loss": 0.105, "step": 5489 }, { "epoch": 34.3125, "grad_norm": 1.1796875, "learning_rate": 4.800802008017989e-05, "loss": 0.0767, "step": 5490 }, { "epoch": 34.31875, "grad_norm": 1.171875, "learning_rate": 4.80070093657688e-05, "loss": 0.0729, "step": 5491 }, { "epoch": 34.325, "grad_norm": 1.078125, "learning_rate": 4.8005998405653405e-05, "loss": 0.0928, "step": 5492 }, { "epoch": 34.33125, "grad_norm": 0.83984375, "learning_rate": 4.800498719984452e-05, "loss": 0.1144, "step": 5493 }, { "epoch": 34.3375, "grad_norm": 0.62109375, "learning_rate": 4.800397574835293e-05, "loss": 0.0534, "step": 5494 }, { "epoch": 34.34375, "grad_norm": 1.078125, "learning_rate": 4.800296405118945e-05, "loss": 0.1233, "step": 5495 }, { "epoch": 34.35, "grad_norm": 0.953125, "learning_rate": 4.8001952108364876e-05, "loss": 0.0933, "step": 5496 }, { "epoch": 34.35625, "grad_norm": 1.5078125, "learning_rate": 4.8000939919890015e-05, "loss": 0.1305, "step": 5497 }, { "epoch": 34.3625, "grad_norm": 1.328125, "learning_rate": 4.799992748577568e-05, "loss": 0.1092, "step": 5498 }, { "epoch": 34.36875, "grad_norm": 1.0078125, "learning_rate": 4.799891480603268e-05, "loss": 0.0696, "step": 5499 }, { "epoch": 34.375, "grad_norm": 1.484375, "learning_rate": 4.799790188067184e-05, "loss": 0.1237, "step": 5500 }, { "epoch": 34.38125, "grad_norm": 1.453125, "learning_rate": 4.799688870970396e-05, "loss": 0.1334, "step": 5501 }, { "epoch": 34.3875, "grad_norm": 1.65625, "learning_rate": 4.799587529313987e-05, "loss": 0.1498, "step": 5502 }, { "epoch": 34.39375, "grad_norm": 1.3671875, "learning_rate": 4.7994861630990394e-05, "loss": 0.1071, "step": 5503 }, { "epoch": 34.4, "grad_norm": 1.0390625, "learning_rate": 4.799384772326636e-05, "loss": 0.0807, "step": 5504 }, { "epoch": 34.40625, "grad_norm": 0.98828125, "learning_rate": 4.799283356997859e-05, "loss": 0.0824, "step": 5505 }, { "epoch": 34.4125, "grad_norm": 1.578125, "learning_rate": 4.7991819171137913e-05, "loss": 0.1388, "step": 5506 }, { "epoch": 34.41875, "grad_norm": 1.0078125, "learning_rate": 4.7990804526755166e-05, "loss": 0.0818, "step": 5507 }, { "epoch": 34.425, "grad_norm": 1.0390625, "learning_rate": 4.798978963684119e-05, "loss": 0.0948, "step": 5508 }, { "epoch": 34.43125, "grad_norm": 0.77734375, "learning_rate": 4.798877450140681e-05, "loss": 0.0568, "step": 5509 }, { "epoch": 34.4375, "grad_norm": 1.6328125, "learning_rate": 4.798775912046288e-05, "loss": 0.1592, "step": 5510 }, { "epoch": 34.44375, "grad_norm": 1.0078125, "learning_rate": 4.798674349402025e-05, "loss": 0.1042, "step": 5511 }, { "epoch": 34.45, "grad_norm": 1.4375, "learning_rate": 4.7985727622089746e-05, "loss": 0.1063, "step": 5512 }, { "epoch": 34.45625, "grad_norm": 1.5703125, "learning_rate": 4.7984711504682225e-05, "loss": 0.145, "step": 5513 }, { "epoch": 34.4625, "grad_norm": 0.91796875, "learning_rate": 4.798369514180854e-05, "loss": 0.0809, "step": 5514 }, { "epoch": 34.46875, "grad_norm": 1.6171875, "learning_rate": 4.7982678533479555e-05, "loss": 0.156, "step": 5515 }, { "epoch": 34.475, "grad_norm": 1.1484375, "learning_rate": 4.798166167970611e-05, "loss": 0.1044, "step": 5516 }, { "epoch": 34.48125, "grad_norm": 0.72265625, "learning_rate": 4.798064458049907e-05, "loss": 0.0585, "step": 5517 }, { "epoch": 34.4875, "grad_norm": 1.15625, "learning_rate": 4.7979627235869305e-05, "loss": 0.1116, "step": 5518 }, { "epoch": 34.49375, "grad_norm": 1.0234375, "learning_rate": 4.797860964582766e-05, "loss": 0.1226, "step": 5519 }, { "epoch": 34.5, "grad_norm": 1.2890625, "learning_rate": 4.797759181038504e-05, "loss": 0.1244, "step": 5520 }, { "epoch": 34.50625, "grad_norm": 1.2421875, "learning_rate": 4.797657372955228e-05, "loss": 0.1049, "step": 5521 }, { "epoch": 34.5125, "grad_norm": 1.1953125, "learning_rate": 4.797555540334027e-05, "loss": 0.1015, "step": 5522 }, { "epoch": 34.51875, "grad_norm": 1.375, "learning_rate": 4.797453683175987e-05, "loss": 0.0964, "step": 5523 }, { "epoch": 34.525, "grad_norm": 1.1640625, "learning_rate": 4.797351801482197e-05, "loss": 0.0991, "step": 5524 }, { "epoch": 34.53125, "grad_norm": 0.97265625, "learning_rate": 4.797249895253745e-05, "loss": 0.0791, "step": 5525 }, { "epoch": 34.5375, "grad_norm": 0.91796875, "learning_rate": 4.797147964491719e-05, "loss": 0.063, "step": 5526 }, { "epoch": 34.54375, "grad_norm": 1.40625, "learning_rate": 4.797046009197208e-05, "loss": 0.1361, "step": 5527 }, { "epoch": 34.55, "grad_norm": 1.484375, "learning_rate": 4.7969440293712996e-05, "loss": 0.1758, "step": 5528 }, { "epoch": 34.55625, "grad_norm": 1.4140625, "learning_rate": 4.7968420250150846e-05, "loss": 0.1623, "step": 5529 }, { "epoch": 34.5625, "grad_norm": 1.171875, "learning_rate": 4.796739996129651e-05, "loss": 0.1224, "step": 5530 }, { "epoch": 34.56875, "grad_norm": 1.015625, "learning_rate": 4.79663794271609e-05, "loss": 0.0779, "step": 5531 }, { "epoch": 34.575, "grad_norm": 1.296875, "learning_rate": 4.79653586477549e-05, "loss": 0.1185, "step": 5532 }, { "epoch": 34.58125, "grad_norm": 1.6328125, "learning_rate": 4.796433762308941e-05, "loss": 0.1125, "step": 5533 }, { "epoch": 34.5875, "grad_norm": 0.8671875, "learning_rate": 4.796331635317535e-05, "loss": 0.0657, "step": 5534 }, { "epoch": 34.59375, "grad_norm": 0.96875, "learning_rate": 4.796229483802361e-05, "loss": 0.1043, "step": 5535 }, { "epoch": 34.6, "grad_norm": 0.97265625, "learning_rate": 4.796127307764511e-05, "loss": 0.0704, "step": 5536 }, { "epoch": 34.60625, "grad_norm": 1.3125, "learning_rate": 4.796025107205075e-05, "loss": 0.1201, "step": 5537 }, { "epoch": 34.6125, "grad_norm": 0.7578125, "learning_rate": 4.795922882125146e-05, "loss": 0.0666, "step": 5538 }, { "epoch": 34.61875, "grad_norm": 1.3984375, "learning_rate": 4.7958206325258145e-05, "loss": 0.1628, "step": 5539 }, { "epoch": 34.625, "grad_norm": 1.140625, "learning_rate": 4.7957183584081736e-05, "loss": 0.1052, "step": 5540 }, { "epoch": 34.63125, "grad_norm": 0.87890625, "learning_rate": 4.795616059773315e-05, "loss": 0.0925, "step": 5541 }, { "epoch": 34.6375, "grad_norm": 0.875, "learning_rate": 4.7955137366223306e-05, "loss": 0.0644, "step": 5542 }, { "epoch": 34.64375, "grad_norm": 1.7734375, "learning_rate": 4.795411388956314e-05, "loss": 0.0922, "step": 5543 }, { "epoch": 34.65, "grad_norm": 1.5625, "learning_rate": 4.795309016776358e-05, "loss": 0.1548, "step": 5544 }, { "epoch": 34.65625, "grad_norm": 1.0546875, "learning_rate": 4.7952066200835555e-05, "loss": 0.0852, "step": 5545 }, { "epoch": 34.6625, "grad_norm": 0.86328125, "learning_rate": 4.7951041988790004e-05, "loss": 0.0691, "step": 5546 }, { "epoch": 34.66875, "grad_norm": 0.95703125, "learning_rate": 4.7950017531637864e-05, "loss": 0.0884, "step": 5547 }, { "epoch": 34.675, "grad_norm": 1.53125, "learning_rate": 4.7948992829390074e-05, "loss": 0.1212, "step": 5548 }, { "epoch": 34.68125, "grad_norm": 1.0703125, "learning_rate": 4.794796788205759e-05, "loss": 0.0929, "step": 5549 }, { "epoch": 34.6875, "grad_norm": 1.4296875, "learning_rate": 4.7946942689651334e-05, "loss": 0.1975, "step": 5550 }, { "epoch": 34.69375, "grad_norm": 1.1171875, "learning_rate": 4.794591725218227e-05, "loss": 0.1183, "step": 5551 }, { "epoch": 34.7, "grad_norm": 1.0546875, "learning_rate": 4.7944891569661354e-05, "loss": 0.0808, "step": 5552 }, { "epoch": 34.70625, "grad_norm": 1.359375, "learning_rate": 4.794386564209953e-05, "loss": 0.1266, "step": 5553 }, { "epoch": 34.7125, "grad_norm": 1.9375, "learning_rate": 4.794283946950776e-05, "loss": 0.2098, "step": 5554 }, { "epoch": 34.71875, "grad_norm": 1.8203125, "learning_rate": 4.7941813051897e-05, "loss": 0.1853, "step": 5555 }, { "epoch": 34.725, "grad_norm": 0.94921875, "learning_rate": 4.794078638927821e-05, "loss": 0.078, "step": 5556 }, { "epoch": 34.73125, "grad_norm": 1.2265625, "learning_rate": 4.793975948166236e-05, "loss": 0.1091, "step": 5557 }, { "epoch": 34.7375, "grad_norm": 1.421875, "learning_rate": 4.793873232906041e-05, "loss": 0.1138, "step": 5558 }, { "epoch": 34.74375, "grad_norm": 1.4765625, "learning_rate": 4.793770493148333e-05, "loss": 0.133, "step": 5559 }, { "epoch": 34.75, "grad_norm": 1.265625, "learning_rate": 4.79366772889421e-05, "loss": 0.1105, "step": 5560 }, { "epoch": 34.75625, "grad_norm": 1.4375, "learning_rate": 4.793564940144769e-05, "loss": 0.1127, "step": 5561 }, { "epoch": 34.7625, "grad_norm": 1.5546875, "learning_rate": 4.7934621269011075e-05, "loss": 0.147, "step": 5562 }, { "epoch": 34.76875, "grad_norm": 1.875, "learning_rate": 4.793359289164324e-05, "loss": 0.1839, "step": 5563 }, { "epoch": 34.775, "grad_norm": 1.0625, "learning_rate": 4.793256426935516e-05, "loss": 0.0808, "step": 5564 }, { "epoch": 34.78125, "grad_norm": 1.28125, "learning_rate": 4.7931535402157836e-05, "loss": 0.095, "step": 5565 }, { "epoch": 34.7875, "grad_norm": 1.9296875, "learning_rate": 4.793050629006223e-05, "loss": 0.122, "step": 5566 }, { "epoch": 34.79375, "grad_norm": 1.6171875, "learning_rate": 4.792947693307936e-05, "loss": 0.1023, "step": 5567 }, { "epoch": 34.8, "grad_norm": 1.8125, "learning_rate": 4.79284473312202e-05, "loss": 0.1184, "step": 5568 }, { "epoch": 34.80625, "grad_norm": 1.2265625, "learning_rate": 4.792741748449575e-05, "loss": 0.1042, "step": 5569 }, { "epoch": 34.8125, "grad_norm": 1.140625, "learning_rate": 4.7926387392917016e-05, "loss": 0.0838, "step": 5570 }, { "epoch": 34.81875, "grad_norm": 1.1328125, "learning_rate": 4.792535705649499e-05, "loss": 0.0727, "step": 5571 }, { "epoch": 34.825, "grad_norm": 0.734375, "learning_rate": 4.792432647524068e-05, "loss": 0.0532, "step": 5572 }, { "epoch": 34.83125, "grad_norm": 0.625, "learning_rate": 4.792329564916509e-05, "loss": 0.0605, "step": 5573 }, { "epoch": 34.8375, "grad_norm": 1.3515625, "learning_rate": 4.792226457827923e-05, "loss": 0.1046, "step": 5574 }, { "epoch": 34.84375, "grad_norm": 1.0703125, "learning_rate": 4.792123326259411e-05, "loss": 0.0771, "step": 5575 }, { "epoch": 34.85, "grad_norm": 1.171875, "learning_rate": 4.7920201702120747e-05, "loss": 0.1029, "step": 5576 }, { "epoch": 34.85625, "grad_norm": 1.2578125, "learning_rate": 4.791916989687015e-05, "loss": 0.1211, "step": 5577 }, { "epoch": 34.8625, "grad_norm": 0.625, "learning_rate": 4.7918137846853355e-05, "loss": 0.0611, "step": 5578 }, { "epoch": 34.86875, "grad_norm": 1.1796875, "learning_rate": 4.7917105552081367e-05, "loss": 0.1005, "step": 5579 }, { "epoch": 34.875, "grad_norm": 1.6328125, "learning_rate": 4.791607301256522e-05, "loss": 0.1525, "step": 5580 }, { "epoch": 34.88125, "grad_norm": 0.94921875, "learning_rate": 4.7915040228315936e-05, "loss": 0.0737, "step": 5581 }, { "epoch": 34.8875, "grad_norm": 1.1796875, "learning_rate": 4.791400719934454e-05, "loss": 0.1048, "step": 5582 }, { "epoch": 34.89375, "grad_norm": 1.25, "learning_rate": 4.791297392566207e-05, "loss": 0.1007, "step": 5583 }, { "epoch": 34.9, "grad_norm": 1.1640625, "learning_rate": 4.791194040727956e-05, "loss": 0.0867, "step": 5584 }, { "epoch": 34.90625, "grad_norm": 1.234375, "learning_rate": 4.7910906644208054e-05, "loss": 0.1185, "step": 5585 }, { "epoch": 34.9125, "grad_norm": 0.8203125, "learning_rate": 4.790987263645859e-05, "loss": 0.1114, "step": 5586 }, { "epoch": 34.91875, "grad_norm": 1.3046875, "learning_rate": 4.79088383840422e-05, "loss": 0.1125, "step": 5587 }, { "epoch": 34.925, "grad_norm": 1.1953125, "learning_rate": 4.790780388696994e-05, "loss": 0.1089, "step": 5588 }, { "epoch": 34.93125, "grad_norm": 1.0546875, "learning_rate": 4.7906769145252844e-05, "loss": 0.0833, "step": 5589 }, { "epoch": 34.9375, "grad_norm": 1.375, "learning_rate": 4.790573415890198e-05, "loss": 0.1299, "step": 5590 }, { "epoch": 34.94375, "grad_norm": 1.9296875, "learning_rate": 4.7904698927928406e-05, "loss": 0.2255, "step": 5591 }, { "epoch": 34.95, "grad_norm": 1.140625, "learning_rate": 4.790366345234315e-05, "loss": 0.0989, "step": 5592 }, { "epoch": 34.95625, "grad_norm": 1.765625, "learning_rate": 4.7902627732157294e-05, "loss": 0.1432, "step": 5593 }, { "epoch": 34.9625, "grad_norm": 1.2890625, "learning_rate": 4.790159176738189e-05, "loss": 0.1014, "step": 5594 }, { "epoch": 34.96875, "grad_norm": 0.8046875, "learning_rate": 4.7900555558028006e-05, "loss": 0.0817, "step": 5595 }, { "epoch": 34.975, "grad_norm": 1.2890625, "learning_rate": 4.78995191041067e-05, "loss": 0.1024, "step": 5596 }, { "epoch": 34.98125, "grad_norm": 1.6640625, "learning_rate": 4.7898482405629055e-05, "loss": 0.1859, "step": 5597 }, { "epoch": 34.9875, "grad_norm": 0.91015625, "learning_rate": 4.789744546260612e-05, "loss": 0.0859, "step": 5598 }, { "epoch": 34.99375, "grad_norm": 1.078125, "learning_rate": 4.7896408275049e-05, "loss": 0.1248, "step": 5599 }, { "epoch": 35.0, "grad_norm": 1.265625, "learning_rate": 4.789537084296874e-05, "loss": 0.1159, "step": 5600 } ], "logging_steps": 1, "max_steps": 32000, "num_input_tokens_seen": 0, "num_train_epochs": 200, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }